This page intentionally left blank

Perspectives on Animal Behavior

This page intentionally left blank
Perspectives on Animal Behavior
Third Edition

Judith Goodenough
Biology Department / University of Massachusetts, Amherst

Betty McGuire
Department of Ecology and Evolutionary Biology / Cornell University, Ithaca

Elizabeth Jakob
Psychology Department / University of Massachusetts, Amherst

John Wiley & Sons, Inc.

VICE PRESIDENT AND EXECUTIVE PUBLISHER Kaye Pace
SENIOR ACQUISITIONS EDITOR Kevin Witt
EDITORIAL PROGRAM ASSISTANT Alissa Etrheim
EDITORIAL ASSISTANT Jenna Paleski
PRODUCTION SERVICES MANAGER Dorothy Sinclair
SENIOR PRODUCTION EDITOR Janet Foxman
EXECUTIVE MARKETING MANAGER Clay Stone
CREATIVE DIRECTOR Harry Nolan
SENIOR DESIGNER Kevin Murphy
PHOTO EDITOR Sheena Goldstein
SENIOR ILLUSTRATION EDITOR Anna Melhorn
SENIOR MEDIA EDITOR Linda Muriello
PRODUCTION SERVICES Cécile Billioti de Gage/ Preparé
COVER DESIGN M77 Design
COVER PHOTO Richard Packwood/Getty Images, Inc.

This book was set in 10/12 Janson by Preparé and printed and bound
by Courier/Kendallville. The cover was printed by Courier/Kendallville.

This book is printed on acid-free paper.

8

Copyright © 2010, 2001, 1993 John Wiley & Sons, Inc. All rights reserved.
No part of this publication may be reproduced, stored in a retrieval system or transmitted
in any form or by any means, electronic, mechanical, photocopying, recording, scanning or
otherwise, except as permitted under Sections 107 or 108 of the 1976 United States
Copyright Act, without either the prior written permission of the Publisher, or authorization
through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc.,
222 Rosewood Drive, Danvers, MA 01923, website www.copyright.com. Requests to the
Publisher for permission should be addressed to the Permissions Department, John Wiley &
Sons, Inc., 111 River Street, Hoboken, NJ 07030-5774, (201) 748-6011, fax (201) 748-6008,
website www.wiley.com/go/permissions.

To order books or for customer service, please call 1-800-CALL WILEY (225-5945).

Library of Congress Cataloging-in-Publication Data

Goodenough, Judith.
Perspectives on animal behavior / Judith Goodenough, Betty
McGuire, Elizabeth Jakob.—3rd ed.
p. cm.
Includes bibliographical references (p. ).
ISBN 978–0–470–04517–6 (cloth)
1. Animal behavior. I. McGuire, Betty. II. Jakob, Elizabeth M.
III. Title.

QL751.G59 2010
591.5—dc22 2009013360

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1
 To my husband, Stephen Goodenough,
my daughters Aimee and Heather,
and my mother, Betty Levrat.
J. G.
In loving memory of James Patrick McGuire.
To Willy, Kate, and Owen Bemis,
and to Dora, Kevin, and Cathy McGuire.
B. M.
To Adam Porter and Margaret Dringoli, with love and gratitude.
E. J.
This page intentionally left blank
 Preface

Outside the window, male ruby-throated humming- underpinnings, outward to its environment, and back in
birds newly returned from wintering in Central time to its evolutionary origins. Thus, animal behavior
America are brightening this early May day in the draws together fields of study that might often seem,
northeastern United States. The remarkable annual especially to a student early in his or her academic
migration of these tiny birds is a behavioral story in its career, to have little to do with one another. Our hope
own right, complete with a long, foodless crossing of is that readers of this book will see that the facts learned
the Gulf of Mexico to return to breeding areas they left in different courses are connected after all.
here more than six months before. But today it is the The authors of this third edition share a deep
behavior of the males that is the most immediate–they interest in animal behavior, from the standpoints of
are claiming or in some cases reclaiming feeding terri- teaching, writing, and research. Yet each brings to this
tories through dramatic aerial displays and battles. In a project a very different perspective. Judith
week or so the females will arrive and the interactions Goodenough, from the Department of Biology at the
will further intensify, as the breeding season gets under University of Massachusetts Amherst, has led the
way. Watching and understanding these behaviors charge from the very first edition of this text. Judith has
offers ways to engage nature in new and deeper ways, studied biological rhythms in creatures from planaria
and it places our own lives in the context of the other to finches to deer mice (not to mention in the green
animals that share the planet with us. alga Chlamydomonas). Betty McGuire, from the
This brief foray into the easily observed lives of Department of Ecology and Evolutionary Biology at
hummingbirds illustrates a simple point: animal behav- Cornell University, who wrote eight chapters for the
ior is engaging to watch, as evidenced by the popularity first edition of the book but stepped down from the
of nature shows on both small and big screens. In fact, second edition because of other obligations, returns to
those of us fortunate enough to offer a course in animal this third edition. Betty’s research focuses on parental
behavior are often told by our colleagues how lucky we behavior, reproduction, and ecology of small mammals,
are to teach a subject that people of all backgrounds and she occasionally dabbles in work with larger
find fascinating. There is something very appealing in domestic animals such as dogs and horses. In addition
learning about the often bizarre lives going on around to Perspectives in Animal Behavior, Judith and Betty have
us, sometimes literally under our noses. coauthored Human Biology: Personal, Environmental, and
Besides satisfying our natural curiosity, the study of Social Concerns and Biology of Humans: Concepts,
animal behavior is truly interdisciplinary. To deeply Applications, and Issues. Elizabeth Jakob, from the
understand an animal’s behavior, we must look both Department of Psychology at the University of
inward to its genetic, morphological, and physiological Massachusetts Amherst, joins the team of authors. Beth

vii
viii Preface

studies the behavior of spiders, asking questions about

their learning, perception, and interactions with con- CHANGES AND UPDATES TO
specifics and with other species. She has carried out THE THIRD EDITION
field projects in California, Mexico, Massachusetts, and
Maine, and she also does laboratory experiments when We added new features to this edition to promote crit-
the questions or the long Massachusetts winters ical thinking, active learning, and the development of
demand it. She coauthored the fifth edition of Animal vocabulary. We retained certain features from previous
Behavior: Mechanisms, Ecology, Evolution and co-edited editions, such as the outlines at the start of each chapter
the laboratory manual Learning the Skills of Research: and the summary at the end, to help students organize
Laboratory and Field Exercises in Animal Behavior. material and identify key concepts.
Collectively, the three of us have taught courses rang-
ing from large lectures for nonmajors to upper division NEW STOP AND THINK QUESTIONS
courses in animal behavior, vertebrate biology, and Each chapter has at least one Stop and Think Question
evolution. designed to encourage students to pause from reading
Our goals and basic approach remain the same as in the text, think about the information they have just read,
the first two editions: to help students understand the and apply it to a new situation. Many of these questions
history, mechanisms, development, function, and evo- focus on experimental design, methodology, or ethics.
lution of behavior. To this end we introduce the major
approaches to the study of animal behavior and show NEW BOLD TERMS AND GLOSSARY
how these diverse approaches can be integrated to pro-
vide a more complete understanding of any particular Key terms in each chapter are now set in bold type
behavior. Because courses in animal behavior are offered when first formally introduced. These terms and others
in biology, psychology, and animal science departments, of lesser importance are included in the new glossary at
we include basic background material-for example, in the end of the book.
genetics, neurophysiology, and endocrinology–to ensure
that each chapter can stand alone and that each con- INCREASED EMPHASIS ON ALTERNATIVE
tains information accessible to students with different HYPOTHESES
backgrounds. Within each chapter, the examples we
To increase our coverage of the ways in which scientists
develop provide a balance between classic and recent
design and carry out research in animal behavior, we
studies. Throughout the text we use clear and engaging
added many new in-depth examples to illustrate the
writing to explain potentially complex topics such as
importance of developing and testing alternative
behavioral genomics and mathematical models of
hypotheses. These examples will help students under-
behavior.
stand how research is conducted and allow them to bet-
In organization, this third edition begins in much
ter design their own experiments and to evaluate the
the same way as previous editions, with an introduc-
primary literature of animal behavior.
tory chapter on hypothesis testing, followed by Part 1
on approaches to the study of animal behavior. These
early chapters present students with the questions
INCREASED COVERAGE OF SEVERAL KEY
asked and methods used by scientists working in sub- AREAS
fields of animal behavior such as behavioral genetics, In response to current events and reviewer comments,
behavioral endocrinology, and behavioral ecology. Part we increased our coverage of the growing interactions
2 focuses on how behavior relates to the survival of between the fields of animal behavior and conservation
individuals and includes chapters, for example, examin- biology. In particular, we include examples that demon-
ing how animals find places to live, feed, and avoid strate how studies of behavior yield information critical
being eaten. The chapters in Part 3 focus on interac- to conservation efforts. We also address ethical issues in
tions between individuals, such as those between behavioral research, and on the flip side, we examine
mates, parents and offspring, competitors, or collabo- how behavioral research provides information essential
rators. We also describe communication in detail to improving the lives of captive animals. This edition
before moving on to exploring how it is used in con- also includes more examples from the fields of human
flict and cooperation. behavior and applied animal behavior.
 Acknowledgments

Our editor, Kevin Witt, had confidence that three Ethan Cloftfelter, Amherst College
academics with very different schedules could work Fiona Cross, University of Canterbury
together to produce a current text with a single voice. Perri Eason, University of Louisville
We thank him for his encouragement and support. Sarah Huber, Randolph-Macon College
Many other dedicated people at John Wiley & Sons Brian Kelly, University of Massachusetts Amherst
helped get this book into your hands. Alissa Etrheim, Chris Maher, University of Southern Maine
our editorial program assistant, helped us put this book Sarah Partan, Hampshire College
together. She provided gentle reminders to keep us on Mark Petrie, University of Canterbury
schedule and always offered to help in any way possible. Jeff Podos, University of Massachusetts Amherst
We are particularly grateful for her help with permis- Denise Pope, Mount Holyoke College
sions. The text was improved by the skills of our copy Adam Porter, University of Massachusetts Amherst
editor, Betty Pessagno. Anna Melhorn, Senior Nancy Solomon, Miami University of Ohio
Illustration Editor, and Sheena Goldstein, our photo Theodore Stankowich, University of Massachusetts
editor, are largely responsible for the appearance of the Amherst
text. We also thank Will Sillin for his beautiful illustra- Zoran Tadic, University of Zagreb
tions that begin each chapter and each of the three Christina Veino, University of Massachusetts
parts of the book. Elaine Soares, our photo researcher, Amherst
worked diligently to find the perfect photographs for Paige Warren, University of Massachusetts Amherst
our needs. Cécile Billioti de Gage/Preparé headed pro- Gordon Wyse, University of Massachusetts
duction of our book. She helped set priorities and suc- Amherst
cessfully moved the project along even at the busiest
We are also grateful to the many authors who were
times of the academic year. We are now happy to pass
generous enough to check over our portrayal of their
the project to the capable hands of Lucy Parkinson, our
work. Any remaining errors are, of course, ours.
marketing director. Each person is a professional, a
pleasure to work with, and a team player.
FROM JUDITH GOODENOUGH
We thank the reviewers who provided essential feed-
back that helped shape this new edition: I would like to thank many of my friends and col-
Mitchell Baker, Queens College leagues at UMass for lending their expertise to various
Zane Barlow, University of Massachusetts Amherst aspects of this book. Zane Barlow’s comments on
Willy Bemis, Cornell University genetics greatly improved the chapter. As always,
Renae Brodie, Mount Holyoke College Gordon Wyse provided valuable suggestions on the
Bruce Byers, University of Massachusetts Amherst nervous system. Adam Porter helped smooth the writ-

ix
x Acknowledgments

ing of the most difficult sections. My e-mail friend refuge at the Jersey shore where I could escape the
from Croatia, Zoran Tadic, Faculty of Science, deadlines and e-mail (not to mention the weather in
University of Zagreb, provided useful comments and a Ithaca), and relax and reconnect for a few days. My
fresh perspective on many chapters. Although she did father, James McGuire, died during the writing of this
not help with writing, Margaret Ludlam helped me edition. He was an avid birder and a dedicated volun-
cope when things got tense, as they often did. She is teer at his local animal shelter, and I know that he
always cheerful and calm. Most of all, I would like to would have enjoyed our new text. Finally, I thank
thank my family who have always been my cheering Lowell Getz, close friend and colleague of almost 25
squad, and who have encouraged and supported me at years, for listening patiently and always being there. We
every stage of this revision. My husband Steve bore the will get those manuscripts done, Lowell, I promise!
brunt of my obsession. After months of my spending
many more hours with the computer than with him,
FROM BETH JAKOB
Steve merely quipped that he was getting rather fond
of the back of my head and continued to do what was I thank my husband, Adam Porter, who was incredibly
necessary to keep the household running smoothly. supportive throughout this process, was always ready
He’s the funniest person I know, and his wit kept me to read something and offer his frank opinion and
sane. After 38 years of marriage, he still keeps me on detailed edits, and saved our credit rating by taking
my toes by introducing me as his first wife. My daugh- over paying the bills when I was too distracted to
ters, Aimee and Heather, still the joy of my life, remind notice them. I began work on my section of the book
me that the people you love must always come first. during my sabbatical leave, and I thank the faculty,
They’ll find time to spend with me when they know I staff, and students of my host institution at the
need a break. Their courage and confidence in reach- University of Canterbury, Christchurch, New Zealand
ing for their dreams encourages me to keep stretching for their help and friendship, Robert Jackson for his
toward mine—the most immediate of which is the sponsorship, and the Fulbright Foundation for sup-
completion of this revision. My mother, Betty Levrat, port. I am grateful for the patience of my graduate stu-
instilled in me a love of learning and was always willing dents and collaborators whose manuscripts and paper-
to do whatever needed to be done to free time for my work languished during the more intense periods of
writing. book preparation. Penny Jaques, Betsy Dumont, Joe
Elkinton, and my other colleagues in the Organismic
and Evolutionary Biology Program kept me sane and
FROM BETTY MCGUIRE
at least somewhat on track when I foolishly agreed to
I thank my husband, Willy Bemis, for his continuous be Graduate Program Director when this book was
encouragement throughout this project and willingness underway. I am grateful to additional friends and fam-
to read (and edit) everything I passed his way. My ily for forgiving my incessant conversation, deep sighs,
daughter Kate let us use some of her wonderful photo- and continuous e-mails about the book and for
graphs in the third edition–at 16 she already has a keen reminding me that there are other things beyond my
eye and understanding of animal behavior. I thank my computer screen, particularly Doreen Jakob, Alexis
son, Owen, who always reminds me, by his near contin- Jakob, Carolyn Truini, Jack Dringoli, Margaret
uous recitation of animal facts, how fascinating animals Dringoli, Perri Eason, Marta Hersek, Chris Maher,
are to all who watch them. My mother, Dora McGuire, Maggie Hodge, and Nancy Reynolds. Now it’s finally
and siblings, Kevin and Cathy McGuire, provided time to get back outside!
 Contents

Chapter 1 • Introduction 1 Chapter 3 • Genetic Analysis of Behavior 27

Basics of Gene Action 28
Four Questions About Animal Behavior 2
Animal Behavior as an Interdisciplinary Study 2 Goals of Behavioral Genetics 31
The Interplay of Questions: A Case Study 3 Methods of Behavioral Genetics 31
Inbreeding 31
Hypothesis Testing 5 Artificial Selection 33
Inducing Mutations and Screening for Change in
Behavior 35
Finding Natural Variants and Looking for Genetic
Part 1 • Approaches to the Study Differences 36
of Animal Behavior Hybridization 37
Chapter 2 • History of the Study of Animal The Foraging Gene as an Example of Behavioral
Behavior 11 Genetics in Action 38
Candidate Genes 38
The Beginnings 11 Linking a Protein to a Trait 40
Intellectual Continuity in the Animal World 11 Locating all the Genes Associated with a Trait 41
Darwin's Evolutionary Framework 12 Microarray Analysis 41
Classical Ethology 13 Important Principles of Behavioral Genetics 42
The Approach: Evolutionary, Comparative, One Gene Usually Affects Several Traits 42
Descriptive, Field-Oriented 13 Genes Work in Interacting Networks 44
Classical Ethological Concepts 15
Behavioral Variation and Genes 45
Comparative Psychology 17
The Approach: Physiological, Developmental, Environmental Regulation of Gene Expression 47
Quantitative, Laboratory-Oriented 17 Dominance Relationships in Cichlid Fish 47
Early Concepts of Comparative Psychology 17 Song Learning in Male Songbirds 48
The Roots of Physiological Psychology 20 The Importance of Genetic Background to Behavioral
Genetics 50
Sociobiology and Behavioral Ecology 20 Networks of Genes are Responsive to the
More Recent Trends 22 Environment 51
Field Studies 22
Cellular and Molecular Basis of Behavior 22 Epigenetics and Behavioral Genetics 52
Behavioral Biology 23 Complex Relationships Among Genes 54
Applied Animal Behavior 23 A Broader Perspective 54

xi
xii Contents

Chapter 4 • Natural Selection Predators and Prey: The Neuroethology of Life-and-

and Behavior 57 Death Struggles 109

Natural Selection 58 Processing in the Central Nervous System 113

Common Misunderstandings About Natural Brain Changes Underlying Behavioral Change 113
Selection 60 Social Behavior Network 120
Genetic Variation 61 Responding—Motor Systems 122
Variation is Common 61 Neural Control in Motor Systems 122
The Raw Material of Genetic Variation 61 Locust Flight 122
Variation and the Response to Natural
Selection 62 Chapter 7 • Physiological Analysis of
Behavior—The Endocrine
The Maintenance of Variation 63
Gene Flow and Genetic Drift 63 System 127
Correlated Traits 65 The Endocrine System 129
Changing Environmental Conditions 65 Endocrine Glands and Hormones 129
Frequency-Dependent Selection 67
Hormonal Versus Neural Communication 129
Negative-Assortative Mating 68
Types of Hormones and Their Modes of Action 130
Evolutionarily Stable Strategies: Fitness and the
Behavior of Others 68 How Hormones Influence Behavior 132
Testing Hypotheses About Natural Selection Effects on Sensation and Perception 132
Effects on Development and Activity of the Central
and Adaptation 70
Nervous System 133
The Experimental Approach 72
The Comparative Approach 73 Effects on Muscles 133
Monitoring Selection in the Field 75 Methods of Studying Hormone–Behavior
Modeling the Costs and Benefits of Traits 75 Relationships 135
Chapter 5 • Learning and Cognition 77 Interventional Studies 135
Correlational Studies 137
Definition of Learning 78
Types of Learning 78 Organizational and Activational Effects of
Habituation 78 Hormones 139
Classical Conditioning 80 Defining the Dichotomy 139
Operant Conditioning 82 Sex Differences in the Behavior of Norway
Latent Learning 84 Rats 139
Social Learning 85 Individual Differences in the Behavior of Male Tree
Lizards 141
Species Differences in Learning: Questioning the Dichotomy 141
Comparative Studies 87
The Ability to Learn as a Heritable Trait 87 The Dynamic Relationship Between Hormones
Evolution and the Variation in Learning and Behavior 142
Across Species 88 A Reciprocal Relationship 142
Hormonal Suppression of Behavior 142
Other Evidence of Cognitive Abilities
in Animals 91 Interactions Between Hormones, Behavior, and
Tool Use 92 the Environment 142
Detours 94 Adjusting to the Harshness and Predictability of the
Understanding Numbers and Other Physical Environment 143
Abstract Concepts 95 Adjusting to Onlookers in the Social
Self-Recognition and Perspective Taking 97 Environment 145
Chapter 6 • Physiological Analysis—Nerve A Detailed Look at the Hormonal Basis of
Cells and Behavior 99 Selected Behaviors 146
Helping at the Nest 146
Concepts from Cellular Neurobiology 100
Scent-Marking 147
Types of Neurons and Their Jobs 100
Migrating 148
The Message of a Neuron 101
Ions, Membrane Permeability, and Behavior 104 Chapter 8 • The Development
Behavioral Change and Synaptic Transmission 104 of Behavior 151
The Structure of the Synapse 104
Integration 106 Influences on Behavioral Development 152
Development of the Nervous System 152
Specializations for Perception of Biologically Development of Nonneural Structures 155
Relevant Stimuli—Sensory Processing 107 Hormonal Milieu 155
Processing of Sensory Information for Sound Physical Characteristics of the Environment 157
Localization 108 Experience Through Play 158
 Contents xiii

The Concept of Sensitive Periods 159 Magnetic Cues 214

Changing Terminology—From Critical Periods to Cues from the Earth’s Magnetic Field 214
Sensitive Periods 159 Directional Information from the Earth’s
Timing of Sensitive Periods 160 Magnetic Field: A Magnetic Compass 217
Multiple Sensitive Periods 161 Positional Information from the Earth’s
Some Examples of Sensitive Periods in Behavioral Magnetic Field: A Magnetic Map? 218
Development 161 Magnetoreception 222
Pulling It All Together—The Development of Chemical Cues 224
Bird Song 171 Olfaction and Salmon Homing 224
Genetic, Hormonal, and Neural Control of Song 172 Olfaction and Pigeon Homing 226
Role of Learning in Song Development 174
Sensitive Periods in Song Learning 176 Electrical Cues and Electrolocation 229
Own-Species Bias in Song Learning 178 Chapter 11 • The Ecology and Evolution
Social Factors and Song Development 178
A Diversity of Song Learning Strategies 179 of Spatial Distribution 233
Developmental Homeostasis 180 Remaining at Home Versus Leaving 234
Rehabilitation of Chimpanzees After Long-Term Costs and Benefits of Natal Philopatry 234
Isolation 180 Costs and Benefits of Natal Dispersal 235
Sex Biases in Natal Dispersal 238
Natal Dispersal and Conservation Biology 240

Part 2 • Survival Habitat Selection 241

Indicators of Habitat Quality 241
Chapter 9 • Biological Clocks 185 Search Tactics 243
Effects of Natal Experience 243
Defining Properties of Clock-Controlled Habitat Selection and Conservation
Rhythms 186 Biology 244
Persistence in Constant Conditions 186
Entrainment by Environmental Cycles 187 Migration 245
Temperature Compensation 188 Costs of Migration 246
Benefits of Migration 248
Rhythmic Behavior 189 Migration and Conservation Biology 251
Daily Rhythms 189
Lunar Day Rhythms 189 Chapter 12 • Foraging Behavior 253
Semilunar Rhythms 189
Monthly Rhythms 190 Obtaining Food 253
Annual Rhythms 191 Suspension Feeding 254
Omnivory 254
The Clock Versus the Hands of the Clock 192 Herbivory 254
Advantages of Clock-Controlled Behavior 192 Carnivory 256
Anticipation of Environmental Change 192 Adaptations for Detecting Prey 259
Synchronization of a Behavior with an Event
That Cannot Be Sensed Directly 193 Optimal Foraging 264
Continuous Measurement of Time 194 Diet Selection: A Simple Model 265
Deciding When to Leave a Patch: The Marginal
Adaptiveness of Biological Clocks 194
Value Theorem 267
Organization of Circadian Systems 194 Adding Complexity and Realism 268
Multiple Clocks 195 The Utility of Models 274
Coordination of Circadian Timing 195
Human Implications of Circadian Rhythms 200 Chapter 13 • Antipredator Behavior 275
Jet Lag 200 Camouflage 277
Human Health 200 Coloration Matching the Visual Background 277
Disruptive Coloration 280
Chapter 10 • Mechanisms of Orientation Countershading 282
and Navigation 203 Transparency 283
Levels of Navigational Ability 204 Masquerade 284
Piloting 204 Other Functions of Color 284
Compass Orientation 204 Polymorphism 285
True Navigation 206
Warning Coloration 287
Multiplicity of Orientation Cues 206 Batesian Mimicry 289
Visual Cues 206 Diverting Coloration, Structures,
Landmarks 206 and Behavior 290
Sun Compass 207 False Heads 290
Star Compass 211 Autotomy 291
Polarized Light and Orientation 212 Feigning Injury or Death 292
xiv Contents

Intimidation and Fighting Back 294 Chemical Senses 361

Enhancement of Body Size and Display Touch 364
of Weaponry 294 Electrical Fields 365
Eyespots 294
Chemical Repellents 296 Multimodal Communication 366
Functions of Communication 367
Pronouncement of Vigilance 297 Species Recognition 368
Group Defense 298 Mate Attraction 369
Alarm Signals 298 Courtship and Mating 369
Improved Detection 299 Maintaining Social Bonds 372
Dilution Effect 299 Alarm 372
Selfish Herd 300 Aggregation 373
Confusion Effect 301 Agonistic Encounters 374
Mobbing 301 Communication about Resources:
A Case Study 374
Maintenance of Antipredator Behavior 302
Chapter 17 • The Evolution of
Communication 381
Part 3 • Interactions Between The Changing Views of Communication 381
Individuals Sharing Information 381
Manipulating Others 382
Chapter 14 • Reproductive Behavior 305
Signals and Honesty 382
Sexual Selection: Historical and Theoretical When are Honest Signals Likely? 382
Background 306 When are Dishonest Signals Likely? 386
Explanations for Sex Differences in Reproductive Can Honest and Dishonest Signals Coexist? 386
Behavior 306
Revisiting the Ideas of Bateman 308
The Evolutionary Origins of Signals 387
Ritualization 387
Intrasexual Selection—Competition for Mates 309 Receiver-Bias Mechanisms 390
Adaptations That Help a Male Secure Copulations 309
Adaptations That Favor the Use of a Male’s Sperm 313
Selective Forces That Shape Signals 391
Characteristics of the Sender 391
Sexual Interference: Decreasing the Reproductive
Characteristics of the Environment 392
Success of Rival Males 318
Characteristics of the Receiver 395
Intersexual Selection—Mate Choice 320
Criteria by Which Females Choose Mates 320
Language and Apes 396
What is Language? 396
Origin and Maintenance of Mate-Choice
Ape Language Studies 397
Preferences 327
Cryptic Female Choice 329 Communication and Animal Cognition 401
Sexual Conflict 330 Chapter 18 • Conflict 405
Chapter 15 • Parental Care and Mating Aggression and Conflict 406
Systems 333 Why Do Animals Fight? 406
Parental Care 334 An Evolutionary View of Conflict 406
Conflicts Among Family Members Over Parental The Evolution of Fighting Behavior 406
Investment 334 Using Game Theory to Understand
Some Factors That Influence the Allocation the Evolution of Conflict 406
of Parental Resources 336 Asymmetries in Contests 410
Overall Patterns of Parental Care 339
Dispensing with Parental Conflict Among Group Members 413
Care—Brood Parasitism 343 How Dominance is Determined 413
The Benefits of Being Dominant 413
Mating Systems 345 The Benefits of Being Subordinate 414
Classifying Mating Systems 345
Monogamy 346 Conflict Over Space 415
Polygyny 349 Home Ranges, Core Areas, and Territories 415
Polyandry 353 The Ideal Free Distribution and Space Use 415
The Economics of Holding a Territory 416
Chapter 16 • Communication: Channels and The Economics of Territory Size 416
Functions 355 Strategies for Reducing the Cost
of Territorial Defense 417
The Definition of Communication 356
Channels for Communication 356 A Proximate View of Conflict 419
Aggression and Testosterone 419
Vision 356
Stress, Aggression, and Dominance 420
Audition 358
Substrate Vibrations 361
 Contents xv

Chapter 19 • Group Living, Altruism, and Examples of Cooperation Among Animals 434
Cooperation 423 Alarm Calls 434
Cooperation in Acquiring a Mate 436
Living in Groups: From Aggregations to Cooperative Breeding and Helping 439
Structured Societies 423 Eusociality 445
Benefits of Group Living 424
Costs of Living in Groups 426 Glossary 451
Balancing Costs and Benefits 427
References 461
The Puzzle of Altruism 427
Individual Selection and “Altruism” 428
Photo Credits 513
Kin Selection 428 Permissions 517
Reciprocal Altruism 432
Manipulation 434 Index 523
This page intentionally left blank
 1
Introduction

Four Questions About Animal Behavior In one way or another, people have been studying ani-
Animal Behavior as an Interdisciplinary Study mal behavior for thousands of years. The most skillful
hunters and fishermen are usually those who can make
The Interplay of Questions: A Case Study
predictions about the behavior of their prey (Figure 1.1).
Hypothesis Testing It is important to know that when salmon are spawning,
they will not respond to a fisherman’s bait; that many
A New Caledonian crow picks up a twig, bends it into a rodents escape toward the dark, whereas most birds
hook with her foot, pokes it into a hole, and pulls out an escape toward the light; and that many kinds of animals
insect. A male wolf spider does a rhythmic courtship will fight, some ferociously, if they are trapped.
dance, waving his tufted legs as he approaches an atten- The study of animal behavior may have occupied the
tive female. An albatross spreads its massive wings, lifts fringes of human consciousness for centuries for just such
off into the ocean breezes, and does not touch land again a practical reason. Later, when animals were domesti-
for two years. Two male rattlesnakes entwine themselves cated and put to work, it was necessary to learn new
in a wrestling match, settling their dispute without things about them. Horses could be trained for riding or
resorting to a venomous bite. And an emperor penguin, for pulling wagons or tools. Dogs could be trained to
huddled on the ice in the endless blizzards of an track prey or to protect individual humans; cats could not.
Antarctic winter, forgoes food for months while incu- In time, the study of animal behavior took on new
bating the egg delicately balanced on his feet. dimensions. The goals, as well as the techniques,
The behavior of animals is featured on endless changed. Animals are no longer studied simply so that
nature shows and even the occasional big-screen block- we can exploit them more efficiently, although this may
buster for a reason—we find fascinating these glimpses still be one reason for our attention. Now, however, we
into the worlds of the other creatures on our planet. have become aware that increased knowledge of the
How does the world appear to them? Do they think and behavior of specific species in their natural habitats may
feel like we do? How did such strange behaviors come help us save some endangered groups from extinction.
to exist? In addition, information on their normal behavior may
At this point in human history, we have the luxury help us ensure their welfare, not just in the wild, but also
of studying animal behavior for curiosity’s sake. In the in laboratories or zoos (Blumstein et al. 2004; Sutherland
past, our interest in it was based on more practical needs. et al. 1998; Swaisgood 2007). We may be interested in

1
2 Chapter 1 / Introduction

focus on function (survival value) and evolution. It has

also been suggested that consideration of the mecha-
nisms of behavior should, at least in some cases, include
both cognitive and emotional mental processes (Emery
and Clayton 2005).
To better appreciate the types of questions we may
ask about animal behavior, consider a familiar phenom-
enon: the seasonal migration of songbirds between
northern and southern latitudes. As new birds appear
daily at backyard feeders in early spring, we may each
become curious about migration, but depending on our
personal interests, each of us may ask different questions.
How do they “know” it is time? How do they find their
way? Such questions focus on the mechanisms that
underlie the behavior. Must those making this journey
for the first time learn the route from experienced trav-
elers? Do they inherit a directional tendency from their
parents? Questions such as these concern development.
Why do they do it? What do they gain that outweighs
the risks and demands of such a journey? These are ques-
tions about the survival value, or adaptiveness of migra-
tion. Finally, how did it all begin? Were the advancing
glaciers responsible? Were the migratory paths modified
during the thousands of years each species has been
FIGURE 1.1 People have been studying animal behavior migrating? These questions center on the evolution of
for centuries, sometimes for very practical reasons. the behavior. So we see that when we ask why an animal
Knowledge of the behavior of game species may make it behaves in a certain way, some of us may be asking about
easier to put food on the table. immediate causes (the machinery underlying the
response) and others may be asking about the evolu-
tionary causes.
No one type of question is better than the others.
behavior as an example of a broader intellectual concern, Answers to all types are necessary as we weave the fab-
such as evolutionary theory. Or we may be interested in ric of our understanding. These are not competing
studying animal behavior because it may serve as a model avenues of investigation. Rather, they are complemen-
to help us understand human behavior. And, as we said tary. Each may feed back on the others, deepening our
before, sometimes we are fortunate enough to be able to understanding and broadening our avenues of investi-
study animal behavior simply because our curiosity gation (Armstrong 1991; Halpin 1991; Stamps 1991).
prompts us to ask questions about some of the organisms
with which we share the earth.

ANIMAL BEHAVIOR AS
AN INTERDISCIPLINARY STUDY
FOUR QUESTIONS
ABOUT ANIMAL BEHAVIOR Marion Stamp Dawkins (1989) has drawn an analogy
between Tinbergen’s (1963) four aspects of investigation
As casual observations of animal behavior crystallized of behavior and the four legs of an animal. An animal
into a field of scientific study, Niko Tinbergen (1963) that lacks one of its legs can only hobble along.
identified four types of questions that should be asked Similarly, progress in the study of animal behavior is
about behavior: What are the mechanisms that cause it? hampered by a lack of information in any one of these
How does it develop? What is its survival value? How areas of study. This is not to imply that each investiga-
did it evolve? Tinbergen believed that ethology—the tor must ask all types of questions. Often we find that
biological study of behavior—should “give equal atten- individuals are more excited by one type of question than
tion to each of them and to their integration.” by others. However, each investigator will be more suc-
Tinbergen’s four questions are sometimes condensed cessful in finding the answer to the question of personal
into two categories: “how” questions, which focus on interest if he or she is armed with information and tech-
causation and development, and “why” questions, which niques from all four areas of study.
 Animal Behavior as an Interdisciplinary Study 3

THE INTERPLAY OF QUESTIONS: the dominant pair raise their offspring. They baby-sit,
A CASE STUDY attack predators, drive away intruding mongoose groups,
and warn others of a predator’s approach. If a subordi-
To illustrate the richness of a multidimensional nate female does give birth, she will nurse the young of
approach, let’s consider the dwarf mongoose (Helogale the top-ranking female along with her own, even
parvula)—an appealing and active animal that has some though she has fewer young than the dominant female.
very unusual behavior. Dwarf mongooses are the small- The efforts of these helpers allow the breeders to raise
est African carnivores (meat eaters), weighing only more offspring than they could without help: in fact,
300–340 g (11–12 ounces) and only about 43 cm (16 in.) unaided breeding pairs are almost never able to raise
long, half of which is tail (Figure 1.2). their offspring to independence.
What makes dwarf mongooses so unusual is that In Chapter 4 we will explore in detail how natural
they live in social groups. Mongoose packs have around selection works, but even without more in-depth back-
9 adults and yearlings (Creel and Waser 1994; Rood ground you can easily imagine why the observation that
1990), but may contain up to 40 individuals. Because we mongooses give up some of their own chances to have
are such social creatures ourselves, sociality may not offspring in favor of helping others is intriguing to ani-
strike us as particularly unusual, but in fact it is quite mal behaviorists. This raises a host of “how” and “why”
rare. Most carnivores, in fact, are solitary and find each questions that involve both the mechanism and the evo-
other only when it is time to mate. lution of behavior. Let’s begin by examining some ques-
Dwarf mongooses don’t simply live as close neigh- tions about evolution.
bors, but they take social behavior to an extreme: most
of the breeding is done by a dominant female (reviewed
in Creel 2005). Over 70% of the pregnancies in one Evolutionary Questions
long-term study area were by dominant females. The About Dwarf Mongoose Behavior
top-ranking female also gets priority access to food and When we study the evolution of behavior, we can take sev-
initiates any movements that the pack undertakes. The eral different approaches. First, we can ask about the cur-
rest of the pack falls into a dominance hierarchy, each rent costs and benefits of a behavior. Evolution, as we will
with their own place in the chain of command. see later in the book, is all about reproductive success, so
In spite of the fact that subordinate animals low in an evolutionary approach to asking questions about mon-
the hierarchy rarely get to personally breed, they help goose behavior concerns how behavior affects the pro-
duction of offspring. We can also look deeper into history,
and study how a behavior first originated in mongoose
ancestors.
First let’s think about why dwarf mongooses might
currently benefit from living in groups. If you have been
lucky enough to watch dwarf mongooses in the wild or
in a zoo, one of the behaviors you will notice is their con-
stant vigilance—they stand on their tiptoes and peer
alertly around. Their seeming paranoia is justified: dwarf
mongooses are small and thus an appetizing prey for
many other carnivores. This leads us to a hypothesis for
why dwarf mongooses live in groups: to reduce the risk
of predation. And in fact, researchers have found that by
living in groups, dwarf mongooses benefit by each other’s
vigilance and by sheer safety in numbers (if a predator
is going to grab a single mongoose and run off, it’s bet-
ter to be in a group than all alone). More than half of
mongooses that venture off on their own are eaten.
FIGURE 1.2 The dwarf mongoose lives in groups in Are there costs to being in a group? Certainly. As we
which the members cooperate in raising the young. discuss in Chapter 19, being in a group means facing
The dominant male and female are often the only
competition for all sorts of resources. For dwarf mon-
group members that breed. Reproduction by other
gooses, however, sharing one particularly important
group members is usually suppressed. However, other
high-ranking individuals are sometimes allowed to resource is not a problem—food. Dwarf mongooses feed
breed. The variability of reproductive suppression primarily on arthropods—crickets, grasshoppers, ter-
raises many “why” questions. The answers illustrate mites, spiders, and scorpions—and there are plenty to go
the interaction of physiological, behavioral, and evolu- around. So, one of the biggest potential problems with
tionary mechanisms. being in a group isn’t too important for dwarf mongooses.
4 Chapter 1 / Introduction

So, there are plenty of advantages to being in a group behavior. For example, we can ask a simple question:
for dwarf mongooses. But what about those subordinates which characteristics make an animal likely to be dom-
that don’t get to reproduce much, if at all, in a group? Do inant? In one study, the dominant female was without
the costs of losing the chance to reproduce outweigh the exception the oldest in a pack. Within an age class,
benefits of being in a group? Let’s approach this question heavy mongooses were more likely to be dominant
in a way animal behaviorists often do: let’s carefully out- (Creel 2005).
line the choices available to a subordinate mongoose. One We can also ask more complex and interesting ques-
option is to leave the original pack with a few other sub- tions about the interactions among mongooses in a pack.
ordinates and together form a new pack, where it may Up until now, you may have the impression that subor-
rank higher in the hierarchy. Or it can go off and try to dinate mongooses quietly give up their reproductive
join an already existing pack—the chances of a new immi- capabilities, but in fact reproduction is a point of con-
grant getting an opportunity to breed are fairly good, tention. Dominant mongooses can reproductively sup-
especially if the incoming mongoose manages to drive off press subordinates (Creel et al. 1992, 1995).
the resident breeders (Rood 1990). However, the prob- In general, reproductive suppression can be accom-
lem with either of these choices is that half of dispersers plished through either chemical or behavioral means. In
die in the process, generally getting eaten by predators. the first way, chemicals released by dominant individu-
Alternatively, a subordinate can stay in the pack and als, perhaps in urine or feces, suppress reproductive
hope for an improvement in its social rank, either by development or function in subordinates. In the second
slowly gaining status over its lifetime or by benefiting way, dominant individuals behave aggressively toward
from a tragedy that befalls the dominant mongoose. In subordinates who attempt to breed. Sometimes both
addition, subordinates do have some chance at breeding chemicals and aggressive behavior play a role.
themselves. In one study, 12% of subordinate mongooses Reproductive suppression in male and female dwarf
became pregnant, and DNA fingerprinting revealed that mongooses involves different mechanisms (Creel et al.
15% of the offspring in a pack had subordinate mothers 1992). In males, reproductive suppression of subordi-
and 25% had subordinate fathers (Keane et al. 1994). nates is accomplished entirely through behavioral means.
The chances of subordinates getting to reproduce are The dominant male attacks other males to keep them
especially good when extra food is around, as researchers from mating with fertile females. Levels of androgens
have shown by providing food supplements. (Creel and (male hormones) of subordinate and dominant males are
Waser 1997). Finally, even dwarf mongooses that don’t indistinguishable. In females, suppression can occur
breed can help their genes pass on to the next genera- through hormonal as well as behavioral controls. The
tion by helping to raise their relatives. Thus, for any ability to breed depends on the female’s peak level of the
given mongoose, the current costs and benefits of its dif- hormone estrogen. Low-ranking subordinates have low
ferent options may vary, and indeed different individu- estrogen levels compared to the breeder. However, high
als make different choices. ranking, older subordinates have estrogen levels similar
Using a different set of techniques, we can also to that of the breeder, and have a better chance of breed-
investigate the evolutionary origins of dwarf mongoose ing if they disperse. As a result, the top-ranking female
social behavior. As we will see in many other examples must ease her behavioral suppression of the older
later in this book, to study evolutionary origins usually females to keep them as helpers.
involves the construction of a phylogeny, sort of a fam- Why might both kinds of reproductive suppres-
ily tree of a taxonomic group. Usually we are not fortu- sion evolve for females but not for males? The answer
nate enough to have a good fossil record, so we have to may lie in the certainty of parentage and the ultimate
use other techniques. Veron et al. (2004) used both gene means of suppressing reproduction by subordinates—
sequences and behavioral traits of the 37 species of mon- infanticide. The top female is likely to be able to iden-
gooses that are alive today in order to construct a phy- tify her own offspring by odor cues learned at the time
logeny. Their work suggests that the ancestor of of birth. However, the young of a single litter can have
modern-day mongooses was solitary, lived in a forest, different fathers, so a male cannot easily determine
and ate vertebrates. When the environment changed which of them are his. If resources become too scarce
during the Pliocene, some mongoose lineages switched to support the young of a subordinate, a dominant can
to eating insects. Insectivory made sociality more likely kill them. Because a female can recognize her own
because competition for food was reduced. young, she can selectively kill those of subordinates. A
dominant male practicing the same infanticidal policy
Proximate Questions would risk killing some of his own offspring. So, a top-
About Dwarf Mongoose Behavior ranking female will allow reproduction of subordinate
Besides questions about current costs and benefits and females who are most likely to leave and breed else-
the evolutionary history of mongoose social behavior, where because she can veto that decision later if need
we can also ask about the mechanisms that underlie the be. Thus, the variation in reproductive suppression in
 Hypothesis Testing 5

dwarf mongooses results from an interplay between

the mechanisms that control it and evolution.
The dwarf mongoose studies illustrate the extent to
which proximate and ultimate questions about behavior
are intertwined. In the pages that follow, we will consider
aspects of the behavior of other species in light of all four
of Tinbergen’s questions. First, we will introduce some
of the approaches to the study of animal behavior—
genetics, evolution, ecology, learning, neurobiology,
endocrinology, and development. We will discuss some
of the basic principles and techniques used in each
approach. Then, in the chapters that follow, we will con-
sider specific types of behavior, including orientation,
foraging, antipredator defense, mating, parental care,
and other social behavior, and see how the various per- FIGURE 1.3 Male burrowing owls scatter mammal
manure around the entrance of their nest. There are
spectives may act synergistically to broaden our under-
at least four alternative hypotheses for this behavior:
standing of animal behavior. Before we begin, however,
attracting mates, signaling that the nest burrow is occu-
we should take a moment to consider how scientists pose pied, camouflaging the scents of an active owl nest, and
questions about animal behavior and then go about attracting prey. A different set of predictions accompa-
answering them. nies each hypothesis, and each prediction can be tested.

The observation of manure scattered around the

HYPOTHESIS TESTING nest entrance prompts the question, “What is the func-
tion of manure around the nest opening?” Matthew
The study of animal behavior usually begins with an Smith and Courtney Conway (2007) developed and
observation that prompts a question, which is followed tested several hypotheses to answer this question.
by forming hypotheses (logical guesses) about a possi-
Mate attraction The first hypothesis was that males
ble answer. It is necessary to be able to test each hypo-
scatter manure to attract females. Perhaps females
thesis. Generally, the hypothesis leads to a prediction,
use the manure to assess a male’s quality, much like
which will support the hypothesis if it holds true when
female bowerbirds use the objects in a male’s bower
tested. Depending on the hypothesis, the test may
to assess his quality. Only males scatter manure,
involve further observations, comparisons of behavior
which is consistent with the hypothesis that the
among species, or experimental manipulation.
manure attracts females.
Different hypotheses can sometimes lead to identi-
cal predictions, and then both hypotheses are supported Sign that burrow is occupied Suitable nest sites
or refuted, depending on the outcome of the test. In this are limited, and resident males defend their site vig-
event, it is necessary to make other predictions that will orously. Perhaps the manure signals to other males
allow us to reject one of the hypotheses. that the burrow is occupied. Nonresident males
There may also be alternate hypotheses to explain could then avoid the costs of fighting by looking
a particular behavior. If so, each hypothesis could lead elsewhere for a burrow.
to different predictions, and each prediction would be Olfactory camouflage Common predators of bur-
tested by observations and experiments. rowing owls include coyotes, badgers, and skunks,
Studies of an unusual nesting practice of burrowing which are predators that use odors to locate prey.
owls provide an example of an observation leading to By masking the scents of an active owl nest, the
several hypotheses for the function of the behavior. manure may reduce the risk of predators detecting
Burrowing owls (Athene cunicularis) live and nest in desert the owls within.
grasslands in Canada, the western United States, and
Prey attraction Burrowing owls eat what is avail-
some parts of South America (Figure 1.3). The male pre-
able: mice, voles, toads, small birds, insects, spiders,
pares the underground nest burrow by digging and
and centipedes. Indeed, small arthropods make up
scraping out dirt or by taking over a burrow of another
a large part of their diet. Perhaps the manure attracts
small mammal, such as a prairie dog or kangaroo rat.
small arthropods to the owl nest, creating a home-
The unusual part of nest building is that the male lines
delivery service for owl nestlings.
the tunnel leading to the nest cup with 3 to 7 cm of
mammal manure and then scatters manure around the As you can see in Table 1.1, a different set of predic-
entrance to the nest. tions accompanies each of these hypotheses, and each
 TABLE 1.1 Hypotheses, Predictions, and Tests of the Function of Manure Scattering by Burrowing Owls
Hypothesis Hypothesis Hypothesis Hypothesis
1: Mate 2: Burrow 3: Olfactory 4: Prey
Prediction attraction occupied Camouflage attraction Test Result Conclusion

1. Stage of nesting Before pair Soon after Just before All stages Estimated date of pair Scattering began after Contradicts mate-
cycle when manure is formation arrival incubation bond formation and dates pair bond formation attraction, burrow-
collected should be: of manure scattering occupied, and olfactory
camouflage hypotheses
2. Presence of manure Yes/no Yes No No Scatter manure around 58% of nests with manure Pattern in direction of
at traditional nest before entrance of some burrows; became nests; 78% of burrow-occupied
owls return from remove manure from nests with removed hypothesis but results
migration discourages some burrows of previous manure became nests are not statistically
nesting by other owls years significant
3. Increased Yes/no Yes No No Experimental group = Resident male more likely Pattern in direction of
perception of presented with mount of to attack an owl mount prediction of burrow-
competition burrowing owl and tape than starling mount; occupied prediction but
increases manure of its primary call; control Increased scattering results are not statistically
scattering group = presented with after presentation of owl significant
mount of European mount than of European
6

starling of its primary call; starling

Measure amount of
manure spread
4. More surviving young No No No Yes Count number of Manure at entrance of nest Supports prey-attraction
at nests with manure surviving offspring in made no difference in the hypothesis
manure-supplemented and number of surviving young Contradicts olfactory-
manure-removed nests camouflage hypothesis
5. Fewer predatory attacks No No Yes No Select active nests and add Probability of predatory Contradicts olfactory-
in natural nests with or remove manure attacks was not altered camouflage hypothesis
manure by treatment
6. Fewer predatory attacks No No Yes No Probability of predatory Contradicts olfactory-
in artificial nests with attacks was not altered camouflage hypothesis
manure by treatment
7. Greater arthropod No No No Yes Compare biomass of Pitfall traps with manure Supports prey-attraction
biomass in nests with arthropods at nest sites collected more arthropod hypothesis
manure with and without manure biomass than did pitfall
and pitfall traps with and traps without manure
without manure
 Hypothesis Testing 7

prediction can be tested. For example, each hypothesis played for ten minutes or until the resident male attacked
makes a different prediction about the timing of manure the mount. Before the first presentation of calls, all scat-
scattering. If the manure attracts a mate, then one would tered material was removed from around the burrow
predict that manure should be present before pair bond- entrance. After the third presentation, all scattered mate-
ing occurs. However, if the manure signals that the bur- rial was again collected and weighed. As predicted by the
row is occupied, then one would predict that manure burrow-occupied hypothesis, resident males were more
should be scattered soon after the owls return from migra- likely to attack a mount of a burrowing owl, which would
tion, well before pair bonding. On the other hand, if be perceived as a threat, than they would a European
manure masks the odor of the owl nest to lower the risk starling mount. Furthermore, after three presentations
of predation, then one would predict that manure scat- of a burrowing owl mount, resident males scattered more
tering occurred later, just before incubation. If the func- manure than they did after three presentations of a
tion of manure is to attract arthropod prey, then one European starling mount. However, once again, these
would predict that scattering should increase after pair results are in the right direction but are not statistically
formation and be most common when nestlings are pre- significant. Recall that the third prediction, that manure
sent in the nest. scattering will begin soon after arrival at the nesting site,
The obvious test of these predictions is to find out was not met; manure scattering actually occurred several
when manure scattering takes place. The usual course of weeks after arrival.
events is that single males arrive at the nesting grounds The olfactory-camouflage hypothesis predicts that
first, followed by previously mated pairs, and finally sin- the amount of manure scattered around the burrow
gle females. Pair bonding occurs when the females entrance will influence the risk of predation. Smith and
arrive. The female incubates seven to ten eggs for 28 to Conway tested this prediction on active natural nests and
30 days until they hatch. She continues to sit on the on artificial nests. They randomly selected 26 active nest
brood for another week or two. During this entire time, sites and added or removed manure from the nest
the male brings food to the nest. For about the next six entrances every two to four days. The artificial nests
weeks, both adults feed the young until they can hunt for were created by baiting unoccupied burrows with
themselves. chicken eggs and assigning each to one of four treat-
Smith and Conway (2007) observed burrows every ments: horse manure and signs of owls, such as feathers
two to four days to determine the number of adults and and pellets; coyote scat plus signs of owls; no manure or
juveniles present, when manure scattering took place, scat but signs of owls; no manure or scat or signs of owls.
and whether any owls were killed by predators. In 87% If the manure at the nest entrance functions to hide the
of the 46 observed burrows, manure scattering occurred scent of an active owl nest, then one would predict less
after pair bond formation, which is inconsistent with the predation on natural or artificial burrows with manure
mate-attraction and burrow-occupied hypotheses and or scat around the entrance. This prediction was not sup-
best fits the prey-attraction hypothesis. ported; there was no difference in the probability of pre-
The burrow-occupied hypothesis predicts that a dation between nests with manure and those without
nonresident male would be less likely to enter a bur- manure.
row if manure was scattered at the entrance. To test this The first prediction of the prey-attraction hypo-
prediction, Smith and Conway added or removed thesis is that manure attracts small arthropod species that
manure from burrows that had been used in the previ- burrowing owls typically eat. To test this prediction,
ous two years. They visited the nests twice a week to Smith and Conway created 75 sampling areas, each con-
see which nests were used. Whereas 78% of the nests taining two treatment sites—one with manure and a con-
from which manure had been removed became nests, trol without manure. Each sampling area also contained
only 58% of the burrows with added manure became three pitfall traps to collect any small arthropods that
nests. This result is not statistically significant, but is were attracted to and approached the treatment site. As
in the direction that would be predicted by the hypoth- predicted, the average biomass (dry weight) of arthro-
esis that the manure functions as an indicator that the pods collected was higher at manure sites.
burrow is occupied, suggesting that it is biologically If more small arthropods are attracted to nests with
meaningful. more manure, it follows that a second prediction of the
A second prediction of the burrow-occupied prey-attraction hypothesis is that manure increases the
hypothesis is that the amount of manure scattered would survival rate of nestlings by attracting arthropods to sup-
increase with perceived competition. Smith and Conway plement the food brought to the nest by the parents. The
tested this prediction by presenting each resident male data do not support this prediction. There was no sig-
with a mount of a burrowing owl accompanied by a nificant difference in the number of young surviving in
recording of its primary call and, separately, a mount of nests with or without manure. Furthermore, there were
a European starling accompanied by a recording of its slightly more survivors in nests without manure, which
primary call. On three separate occasions, the calls were is in the opposite direction of the prediction.
8 Chapter 1 / Introduction

These studies point out the usefulness of testing their nesting behavior in general and about the function
alternate hypotheses and alternate predictions from a of manure scattering in particular may provide insight
single hypothesis. The data do not support the hypoth- into whether or how the disappearance of large grazing
esis that scattered manure functions to attract a mate. mammals from prairies has caused the owl population to
Other data, though not statistically significant, suggest decline (Smith and Conway 2007). A study performed in
that manure may function as a sign that a burrow is occu- Oklahoma revealed that the declining population of bur-
pied: burrows with manure were 36% less likely to rowing owls is related to the elimination of prairie dogs
become nests than burrows without manure, and the (Cynomys ludovicianus). In this region of the country, most
perception of competition increased manure scattering burrowing owls made nests in abandoned prairie dog
by resident males. Although the timing of manure scat- burrows. Unfortunately, prairie dogs are not welcomed
tering fits the olfactory-camouflage hypothesis, the fail- guests on many ranches because they eat some of the
ure of manure to increase survival of the young does not. available food and because their burrow holes may pose
The hypothesis that the function of manure scattering a risk to livestock. As prairie dogs disappear, so do their
is to attract prey to the nest burrows is supported by both burrows. Fewer prairie dog burrows mean fewer nesting
the timing of manure scattering and the increased burrows for burrowing owls, so their numbers are also
amount of prey biomass trapped near nests with manure declining. To increase the number of burrowing owls, it
compared to nests without manure. However, the num- would seem that the prairie dog population must also be
ber of young surviving in nests was not increased by the maintained (Butts and Lewis 1982).
presence of manure, which does not at first appear to be In the pages that follow, we will explore various per-
consistent with the prey-attraction hypothesis. But per- spectives on animal behavior, sometimes stopping to
haps attracting prey to the nest allows juveniles to learn consider how the information might be used to serve
to handle prey near the safety of the burrow, or perhaps animal welfare or conservation. We will also see many
attracting prey to the nest means that the parents can examples of hypothesis testing. Keep these general pro-
make fewer foraging trips. cedures in mind as you read them. Instead of passively
accepting the given explanation for a particular behav-
ior, be critical of the evidence. Try to think of alternate
STOP AND THINK hypotheses for the behaviors described. Make predic-
Some males use materials such as feathers, grass, or dried tions and design your own tests of those predictions.
moss to scatter at the nest entrance. Which hypothesis or
hypotheses would be consistent with this observation?

SUMMARY
STOP AND THINK
There are other possible hypotheses for the function of Animal behavior is studied for many reasons, both prac-
manure scattering at the nest entrance by male burrow- tical and intellectual. A full understanding of animal
ing owls. Mammal manure is abundant in areas where behavior requires answers to four types of questions,
burrowing owls nest, so perhaps manure is good nest- those about (1) immediate mechanisms, (2) development,
building material, and some gets dropped around the (3) survival value, and (4) evolution. Our progress in
entrance by sloppy males. Perhaps manure serves as insu- understanding the behavior of animals will be enhanced
lation or absorbs water to prevent flooding in the nest.
by considering all four types of questions.
How would you test these hypotheses? Can you think of
The study of animal behavior usually begins with an
others? How would you test them?
observation that prompts a question. The next step is to
think of tentative explanations, called hypotheses, to
Populations of burrowing owls are declining so pre- answer that question. Each hypothesis should produce
cipitously that they are listed as endangered in some testable predictions. The tests of those predictions sup-
states and as threatened in others. Learning more about port or refute the hypothesis.
 PART ONE

Approaches to the Study

of Animal Behavior

9
This page intentionally left blank
 2
History of the Study
of Animal Behavior
The Beginnings
Intellectual Continuity in the Animal World
THE BEGINNINGS
Darwin’s Evolutionary Framework INTELLECTUAL CONTINUITY
Classical Ethology IN THE ANIMAL WORLD
The Approach: Evolutionary, Comparative, Descriptive,
Field-Oriented It is difficult, perhaps impossible, to pinpoint the precise
Classical Ethological Concepts beginnings of the study of animal behavior. Rather than
attempting this feat, we will simply consider some of the
Comparative Psychology
highlights in the development of the discipline. One
The Approach: Physiological, Developmental, idea, that of intellectual continuity among animals, was
Quantitative, Laboratory-Oriented
important in shaping some of the earliest views of ani-
Early Concepts of Comparative Psychology
mal behavior. Although the idea of intellectual continu-
The Roots of Physiological Psychology ity was summarized in 1855 by Herbert Spencer in his
Sociobiology and Behavioral Ecology book Principles of Psychology (Spencer 1855), its roots can
More Recent Trends be traced back much further in history, to the ideas of
Field Studies the ancient Greek philosophers. The concept focused on
Cellular and Molecular Basis of Behavior continuity in mental states between “lower” and
“higher” animals and was based on a picture of evolu-
Behavioral Biology
tion similar to Aristotle’s scala naturae, the great chain of
Applied Animal Behavior
being. In the scala naturae, the evolution of species was
viewed as linear and continuous. This classification sys-
tem was hierarchical, with animals ranked according to
To understand a field of study today, we must know their degree of relationship, the highest point of evolu-
something about its past. In this chapter, we consider the tion being humans (Hodos and Campbell 1969). At the
history of the study of animal behavior. Our focus is on bottom of the scale were creatures such as sponges; then
the development of key concepts in the field. further up the scale were insects, fish, amphibians,

11
12 Chapter 2 / History of the Study of Animal Behavior

reptiles, birds, nonhuman mammals, and finally humans. based variation in their behavior, morphology, physiol-
Because evolution was seen as a linear process, with each ogy, and so on. Evolutionary change occurs as the her-
higher species evolving from a lower one until, finally, itable traits of successful individuals (i.e., those that
humans emerged, it was thought that the animal mind survive and reproduce) are spread throughout the pop-
and the human mind were simply points on a continuum. ulation, whereas those traits of less successful individu-
als are lost.
In two later books, The Descent of Man, and Selection
DARWIN’S EVOLUTIONARY in Relation to Sex (1871) and Expression of the Emotions in
FRAMEWORK Man and Animals (1873), Darwin applied his evolution-
ary theory to behavior. In these volumes he recorded his
A few years after the publication of Spencer’s book,
careful and thorough observations on animal behavior,
Charles Darwin (Figure 2.1) published his thoughts on
but his records were anecdotal and often anthropomor-
evolution by natural selection in On the Origin of Species
phic. This was not, however, sloppy science. In the tra-
(1859). Although Darwin’s focus in this book was not on
dition of his day, he believed that careful observations
animal behavior, his ideas provided a conceptual frame-
were useless unless they were connected to a general the-
work within which the field of animal behavior could
ory. Darwin’s general theory was evolution by natural
develop. Discussed in more detail in Chapter 4, these
selection. Because humans evolved from other animals,
ideas can be briefly summarized as follows:
he considered the minds of humans and animals to be
1. Within a species, there is usually variation among similar in kind and to differ only in complexity. As a
individuals. reflection of this belief, he described the behavior of ani-
2. Some of this variation is inherited, and is passed on mals by using terms that denote human emotions and
from mother to offspring. feelings: Ants despaired, and dogs expressed pleasure,
shame, and love. Darwin’s opinion was influential, and
3. Most of the offspring produced by animals do not it is not surprising that others interested in animal behav-
survive to reproduce. Some individuals survive ior followed his lead. Both ethologists and comparative
longer and produce more offspring than others, psychologists trace the beginnings of their respective
because of their particular inherited characteristics. fields to the ideas of Darwin.
Natural selection is the differential survival and For about a decade after the publication of Expression
reproduction of individuals that results from genetically of the Emotions in Man and Animals (1873), descriptions
of animal behavior usually took the form of stories about
the accomplishments of individual animals that were
believed to think and experience emotions as humans do.
For example, based on his subjective interpretation of
what he observed, George J. Romanes, a protégé of
Darwin, constructed a table of emotions that charted the
evolution of the mind and listed the emotions in order
of their historical or evolutionary appearance (Figure
2.2). In his books, Animal Intelligence (1882), Mental
Evolution in Animals (1884), and Mental Evolution in Man
(1889), Romanes examined the implications of
Darwinian thinking about the continuity of species for
the behavior of nonhuman and human animals.
In addition to Romanes, several other scientists
made notable contributions to the study of animal
behavior at the turn of the twentieth century. Jacques
Loeb (1918) believed that all patterns of behavior were
simply “forced movements” or tropisms, physiochemi-
cal reactions toward or away from stimuli. Herbert
Spencer Jennings, perhaps best known for his book
Behavior of the Lower Organisms (1906), disagreed with
Loeb’s ideas and instead emphasized the variability and
modifiability of behavior. Of course, there were many
FIGURE 2.1 Charles Darwin. His ideas on evolution by other pioneers in the study of animal behavior, but we
natural selection provided an evolutionary framework will move on to the turmoil that developed in the disci-
for the study of animal behavior. pline during the twentieth century.
 Classical Ethology 13

Animals in which where variables could be controlled. Finally, ethologists

Emotion emotion first appears were interested in differences between the species,
Shame, deceit Apes, dogs whereas comparative psychologists searched for general
“laws” of behavior.
Revenge, anger Monkeys, elephants Of course there were many exceptions to this char-
Grief, hate, cruelty Carnivores, rodents acterization. Some ethologists made remarkable discov-

Emotional development
eries indoors in their homes and laboratories and
Pride, resentment Birds explored the influence of experience on behavioral devel-
Sympathy Ants, bees, wasps opment (Barlow 1989), and some comparative psychol-
ogists studied a wide range of species and patterns of
Affection Crustaceans behavior and conducted their studies in the field
Jealousy, anger Fish (Dewsbury 1989). Indeed, within each field there were
individuals interested in all four questions of animal
Pugnacity, industry, curiosity Insects, spiders behavior. Although traditionally ethology and compar-
Sexual feelings Molluscs ative psychology have been portrayed as very different
approaches, some accounts of the history of animal
Surprise, fear Larvae of insects, behavior have tended to downplay the differences
segmented worms
between them (e.g., Dewsbury 1984).
FIGURE 2.2 The ideas of George J. Romanes on the
evolutionary appearance of emotions in animals.
(Modified from Romanes 1889.)

CLASSICAL ETHOLOGY
As interest in animal behavior grew, differences in THE APPROACH: EVOLUTIONARY,
opinion developed. These differences eventually led to COMPARATIVE, DESCRIPTIVE, FIELD-
the development of two major approaches to the study
ORIENTED
of animal behavior, ethology and comparative psychol-
ogy, centered in Europe and the United States, respec- “Why is that animal doing that?” is perhaps the funda-
tively. The split that developed between the two mental question of ethology, the approach to the study
approaches seemed at times quite severe. Indeed, one of behavior founded largely by Konrad Lorenz, Niko
must wonder why two groups of scientists, each striving Tinbergen, and Karl von Frisch, European zoologists
for a greater understanding of the marvels of animal who shared the Nobel Prize in medicine and psychol-
behavior, could have stood worlds apart. The European ogy in 1973 (Figure 2.3). Traditionally, ethology con-
ethologists and the American comparative psychologists centrated on the evolutionary basis of animal behavior.
were separated by more than the Atlantic Ocean. In fact, Because natural selection can act only on traits that are
the basic questions they asked about animal behavior genetically determined, it seems a logical outcome of the
were different. (Recall from Chapter 1 that the four ethologists’ basic interest in the evolution of behavior to
questions outlined by Niko Tinbergen in 1963 con- focus on those behavior patterns that are inherited. An
cerned the mechanism, function, development, and evo- emphasis on phylogeny (the evolutionary history of a
lution of behavior.) Whereas the ethologists focused species) is particularly characteristic of the work of
their attention on the evolution and function of behav- Konrad Lorenz.
ior, the comparative psychologists concentrated on the The studies of ethologists often involve comparisons
mechanism and development of behavior. Because they among closely related species. In the words of Lorenz
asked different questions, the types of behavior they (1958), “Every time a biologist seeks to know why an
studied and even their experimental organisms differed. organism looks and acts as it does, he must resort to the
Whereas ethologists, by and large, studied innate behav- comparative method.” Here Lorenz was referring to the
ior in birds, fish, and insects, comparative psychologists, method employed by comparative anatomists when they
particularly those of the behaviorist school (see discus- ask the same question about morphology. If comparative
sion later in chapter), emphasized learned behavior in anatomists were to ask why a whale’s flipper is structured
mammals such as the Norway rat. Furthermore, to as it is, they might compare the skeleton of the flipper
determine the normal function of a behavior, ethologists with that of the forelimb of other vertebrates. They
often attempted to observe the animal in its natural habi- could then see that the typical vertebrate forelimb has
tat or in environments designed to simulate that habi- been specialized for the aquatic life of this mammal.
tat. On the other hand, comparative psychologists Similarly, if one were to wonder why a male fly of the
believed that learning was best studied in the laboratory, species Hilara sartor spins an elaborate silken balloon to
14 Chapter 2 / History of the Study of Animal Behavior

present to a female before mating (Figure 2.4), the sig- place food inside cocoons; others enclose something
nificance of the behavior would become apparent after meaningless, such as a daisy petal. Finally, the males of
comparing it with the mating behavior of the other H. sartor present the females with an empty gauze case
species of flies in the family Empididae. Let us consider that turns off the predatory behavior of the female,
the gift-giving behavior of H. sartor in more detail to thereby allowing them to mate (Kessel 1955). Long-
illustrate the ethologist’s comparative method. tailed dance fly males (Rhamphonyia sulcata) usually offer
Among the empidid flies are species that display a female a genuine nuptial gift. Natasha LeBas and Leon
almost every imaginable evolutionary step in the pro-
gression toward the balloon display. By observing the
manner in which the male empidid fly approaches the
female for mating, one sees that at the heart of the prob-
lem is the fact that the male may be a meal, rather than
a mate, for the predacious female. In one species, Empis
trigramma, the male approaches the female while she is
eating. Because her mouth is already full, his well-timed
approach increases his chances of surviving the
encounter. In another species, E. poplitea, the male cap-
tures a prey, perhaps a fly, and gives it to the female, pro-
viding her with a meal before attempting to copulate.
Males of the species H. quadrivittata gift-wrap the meal
in a silky cocoon before offering it to the female. In
another species, H. thoracica, the cocoon, or case, is large
and elaborate, but the food inside is small and of little
value. In yet another species, H. maura, only some males

FIGURE 2.3 Konrad Lorenz (above), Niko Tinbergen

(top right), and Karl von Frisch (right), three ethologists
who shared the Nobel Prize in 1973.
 Classical Ethology 15

Female
Male
Male

Female

FIGURE 2.4 Male flies of the species Hilara sartor present females with an empty silken bal-
loon before mating. The evolution and function of this behavior can be understood by com-
paring the behavior of closely related species, that is, by using the comparative method
characteristic of the ethological approach. (Drawn from descriptions in Kessel 1955.)

Hockham (2005) altered the gift by filling it with a large these patterns of movement to be just as reliable as mor-
prey, small prey, or a worthless nonedible token. phological characters in defining a particular group.
Although females did copulate longer with males offer- The stereotyped patterns of behavior that intrigued
ing the largest prey, the females copulated for the same ethologists such as Whitman, Heinroth, and Craig were
amount of time with males bearing small prey and those named fixed action patterns by Lorenz. By definition, a
bearing token gifts. Thus, male cheaters can invade the fixed action pattern (FAP) is a motor response that is
population. Without a comparison of the behavior with initiated by some environmental stimulus but that can
that of other species, an observer would be hard pressed continue to completion without the influence of exter-
to explain why males offer silken balloons to females. nal stimuli. For example, Lorenz and Tinbergen (1938)
In addition to utilizing the comparative method, showed that a female greylag goose (Anser anser) will
ethologists often work in the field rather than in the lab- retrieve an egg that has rolled just outside her nest by
oratory. After all, they reason, it is in the natural setting reaching beyond it with her bill and rolling it toward her
that the normal context in which the behavior is dis- with the underside of the bill (Figure 2.5). If the egg is
played is apparent. From this setting, the function of the experimentally removed once the rolling behavior has
behavior may be deduced, and knowledge of the func- begun, the goose will continue the retrieval response
tion may allow us to understand why the behavior has until the now imaginary egg is safely returned to the
been shaped to its present form by natural selection. nest. We have emphasized the fact that once initiated,
Tinbergen and his students conducted much of their FAPs continue to completion. There is little consensus,
research in the field. Lorenz and his followers, on the however, on their defining attributes. Other character-
other hand, studied captive animals, but they often istics that have been used to describe them include the
attempted to simulate in captivity some characteristics of following: (1) the sequence of component acts of an FAP
the animal’s natural habitat (Barlow 1991). As we will see is unalterable, (2) an FAP is not learned, (3) it may be
in Chapter 16, von Frisch’s carefully designed field triggered under inappropriate circumstances, and (4) it
experiments reveal that scout honeybees communicate is performed by all appropriate members of a species
the distance and direction of a rich food source to (Dewsbury 1978).
recruits by “dancing” within the hive. The concept of a fixed action pattern was questioned
in the years following Lorenz’s first introduction of the
term. George Barlow (1968) suggested that, in reality,
CLASSICAL ETHOLOGICAL CONCEPTS most patterns of behavior are not as stereotyped as the
notion of the FAP suggests, and furthermore, most can-
The Fixed Action Pattern not easily be separated into fixed and orientation com-
At the turn of the twentieth century, Charles Otis ponents. He suggested the alternative term modal action
Whitman of the University of Chicago and Oskar pattern (MAP). In specific cases, however, the term fixed
Heinroth of the Berlin Aquarium were pioneering the action pattern may be appropriate. Finley and colleagues
field of ethology (Lorenz 1981). Both scientists were (1983) examined the courtship displays of mallard ducks
interested in the behavior of birds, and each indepen- (Anas platyrhyncos) and concluded that the patterns of
dently concluded that the displays of different species are behavior were indeed as highly stereotyped as suggested
often exceptionally constant. In fact, they considered by the notion of FAP. We will continue to use the
16 Chapter 2 / History of the Study of Animal Behavior

FIGURE 2.5 The egg retrieval response of the greylag goose. The chin-tucking move-
ments used by the female as she rolls the egg back to the nest are highly stereotyped
and are an example of a fixed action pattern. The side-to-side movements that correct
for any deviations in the path of the egg are called the taxis component of the
response. If the egg is removed, the female will continue to roll an imaginary egg back
to the nest. One defining characteristic of a fixed action pattern is that it will continue
to completion even in the absence of guiding stimuli. (Drawn from a photograph in
Lorenz and Tinbergen 1938.)

traditional term here, keeping in mind that there is some The final product is an intricate pattern which etholo-
debate over the appropriateness of its use. gists call a chain of reactions. Here, each component
FAP brings the animal into the situation that evokes the
next FAP.
Sign Stimuli and Releasers as Triggers One of the earliest analyses of a chain of reactions
A fixed action pattern is obviously produced in response was conducted by N. Tinbergen (1951) on the courtship
to something in the environment. Let’s consider the ritual of the three-spined stickleback. This complex
nature of the stimulus that might trigger the behavior. sequence of behaviors culminates in the synchronization
Ethologists called such a stimulus a sign stimulus. If the of gamete release, an event of obvious adaptive value in
sign stimulus is emitted by a member of the same species, species with external fertilization. Each female behavior
it is called a social releaser or, simply, a releaser. is triggered by the preceding male behavior, which in
Releasers are important in communication among ani- turn was triggered by the preceding female behavior
mals, as we will see in Chapter 18. Although releasers are (Figure 2.6).
technically a type of sign stimulus, the terms are often A male stickleback in reproductive condition may
used interchangeably. sometimes attack a female entering his territory. If the
Sign stimuli may be only a small part of any envi- female does not flee and instead begins to display the
ronmental situation. For example, a male European robin appropriate head-up posture in which she hangs
(Erithacus rubecula) will attack another male robin that obliquely in the water, exposing her egg-swollen
enters its territory. Experiments have shown, however, abdomen, the male will begin his courtship with a zigzag
that a tuft of red feathers is attacked as vigorously as an dance. He repeatedly alternates a quick movement
intruding male (Lack 1943). The attack is not stimulated toward her with a sideways turn away. This dance
by the sight of another bird but only by the sight of red releases the approach behavior of the female. Her move-
feathers. Of course, in the world of male robins, red ment induces the male to turn and swim rapidly toward
feathers usually appear on the breast of a competitor. the nest, an action that entices the female to follow. At
Sign stimuli, simple cues that may be indicative of the nest, he lies on his side and makes a series of rapid
very complex situations, get through to the animal’s ner- thrusts with his snout into the entrance while raising his
vous system, where they release patterns of behavior that dorsal spines toward his mate. This behavior is the
may consist, in large part, of fixed action patterns. For releaser for the female to enter the nest. The presence
example, the attack of the male European robin may be of the female in the nest is the releaser for the male to
composed of FAPs that involve pecking, clutching, and begin to rhythmically prod the base of her tail with his
wing fluttering. The end result is that when an intrud- snout, causing the female to release her eggs. She then
ing male robin appears, it is immediately identified and swims out of the nest, making room for the male to enter
effectively attacked. and fertilize the eggs. At the completion of this ritual,
the male chases the female away and continues to defend
his territory against other males until another female can
Chain of Reactions be enticed into the courtship routine. The male mates
So far we have considered only relatively simple behav- with three to five females and then cares for the devel-
iors, but a great deal of complexity can be added to the oping eggs by guarding them from predators and fan-
behavioral repertoire by building sequences of FAPs. ning water over them for aeration. We see, then, that this
 Comparative Psychology 17

vated than the other (Morris 1958). Such flexibility

begins to make sense when the function for which the
1 ritual evolved is considered. For the stickleback,
Female appears, courtship is important to time the release of the
gives head-up
display
gametes, and thus males and females seem to adjust their
activities so that they are physiologically ready for
gamete release at the same time. Despite some flexibil-
ity, however, the component behaviors do not occur ran-
2 domly. In the display, a particular behavior is more likely
Male swims
zigzag to
to be followed by certain actions than by others.
female 3
Female swims,
head up,
toward male COMPARATIVE PSYCHOLOGY
THE APPROACH: PHYSIOLOGICAL,
DEVELOPMENTAL, QUANTITATIVE,
4 LABORATORY-ORIENTED
Male swims
toward nest The comparative psychologists’ emphasis on labora-
tory studies of observable, quantifiable patterns of
behavior distinguished them from the European ethol-
5 ogists during the first half of the twentieth century.
Female follows Recall that, at this time, many ethologists preferred to
study animal behavior under natural conditions. This
meant that they went into the field and observed behav-
6
Male shows nest ior. The problem was that in the field, the unexpected is
expected; one cannot control all the variables. The com-
7
parative psychologists argued that good, experimental
Female enters nest science cannot be done under such uncontrolled condi-
tions. The ethologists were further criticized because,
although they described changes in behavior, they often
neglected to quantify their results and rarely analyzed the
8 data with statistical procedures. Given the psychologists’
Male tremble-thrusts
penchant for laboratory studies that produce quantifiable
results, it is not too surprising that much of their early
research focused on learning and the physiological basis
9
of behavior. Again, however, we wish to emphasize that
Female spawns
although learning and physiology were popular areas of
study among comparative psychologists, the evolution
10
Female leaves and function of behavior were also examined by some
of comparative psychology’s practitioners. We will now
consider some of the major conceptual developments in
11
Male enters the field of comparative psychology.
and fertilizes
EARLY CONCEPTS OF COMPARATIVE
FIGURE 2.6 Courtship behavior in the three-spined PSYCHOLOGY
stickleback. (From N. Tinbergen, 1989.)
Morgan’s Canon
Recall from our previous discussion of the ideas and
complex sequence is largely a chain of FAPs, each trig- writings of Darwin and Romanes that the early descrip-
gered by its own sign stimulus, or releaser. tions of animal behavior were often subjective, anthro-
The chain of reactions is not as rigid as the above pomorphic accounts. C. Lloyd Morgan helped stop the
description of courtship in the three-spined stickleback anecdotal tradition, thereby helping comparative psy-
implies. There are actually many deviations in the pre- chology to become the objective science it is today. He
cise order of the events in the ritual, and some actions argued that behavior must be explained in the simplest
must be repeated several times if one partner is less moti- way that is consistent with the evidence and without the
18 Chapter 2 / History of the Study of Animal Behavior

assumption that human emotions or mental abilities are a

involved. This idea was crystallized in Morgan’s Canon
(1894): “In no case may we interpret an action as the out-
come of the exercise of a higher psychical faculty if it can
be interpreted as the outcome of the exercise of one
which stands lower in the psychological scale.” In other
words, when two explanations for a behavior appear
equally valid, the simpler is preferred. People were urged
to offer explanations of an animal’s behavior without
referring to the animal’s presumed feelings or thought
processes.

Learning and Reinforcement

We have already mentioned that many of the early com- b
parative psychologists focused their research efforts on 400
learning. The early days of these studies were exciting
300

Time (sec)
times indeed, and some of the most important work was
done by scientists in America. E. L. Thorndike (1898), 200
for example, devised experimental techniques to study
learning in the laboratory. He was a pioneer in research 100
on what was called trial-and-error learning, now usually
called operant conditioning. In operant conditioning, the 0
10 20 30 40 50 60 10 15
animal is required to perform a behavior to receive a Trials
reward. In one series of experiments, Thorndike
invented boxes that presented different problems to ani- FIGURE 2.7 (a) A problem box. Thorndike invented
many “problem boxes” to measure the learning ability
mals. For instance, one problem box was a crate with a
of animals. An animal would be placed inside the box
trapdoor at the top through which an experimenter and would have to learn how to operate an escape
might drop a cat to the inside of the box. A hungry cat mechanism. (b) The time required for escape on succes-
was left in the box until it accidentally operated a mech- sive trials was a measure of how quickly the animals
anism, perhaps pulling a loop or pressing a lever that mastered the task. (From Thorndike 1911.)
opened an escape door on the side of the box, allowing
it access to food that had been placed nearby. The length
Pavlov rang a bell immediately before feeding a dog and
of time it took for each escape provides an objective,
found that, in time, the dog came to salivate at the sound
quantifiable measure of learning progress (Figure 2.7).
of the bell alone (Pavlov 1927). We will discuss classical
During repeated trials, the animal became more efficient
conditioning in more detail in Chapter 5.
and required less time to hit the escape latch.
At first, comparative psychologists used operant and
Thorndike’s studies led him to develop the Law of
classical conditioning techniques to study the learning
Effect, a cornerstone of operant conditioning. His basic
abilities of a wide variety of species. Thorndike, for
notion was that responses that are rewarded, that is, fol-
example, examined learning in fish, chickens, cats, dogs,
lowed by a “satisfying” state of affairs, will tend to be
and monkeys and noted striking similarities in the learn-
repeated (this idea was also described by C. Lloyd
ing processes of these animals. His results were there-
Morgan and other investigators of animal behavior).
fore consistent with the idea of intellectual continuity.
Thorndike began publishing studies on animal intellect
Thorndike concluded that although animals might dif-
and behavior in the late 1800s, and in 1911 he published
fer in what they learned or in how rapidly they learned
a collection of his writings in a book entitled Animal
it, the process of learning must be the same in all species.
Intelligence: Experimental Studies.
In his 1911 collection of papers, he summarized his idea
Just a few years after Thorndike introduced the idea
of intellectual continuity as follows (p. 294):
of trial-and-error learning, Ivan Pavlov (Figure 2.8), a
Russian physiologist, described the conditioned reflex. [Intellect’s] general law is that when in a certain situa-
Pavlov noticed that a dog begins to salivate at the sight tion an animal acts so that pleasure results, that act is
of food, and he reasoned that the sight of food must have selected from all those performed and associated with
come to signal the presence of food. (In science, the key that situation, so that, when that situation recurs, the
observations that trigger great ideas are often quite com- act will be more likely to follow than it was before. . . .
monplace, as in this case. It is not what you observe; it The intellectual evolution of the race consists of an
is what you make of it.) In his well-known experiment, increase in the number, delicacy, complexity, perma-
 Comparative Psychology 19

B. F. Skinner, one of the most famous behaviorists,

devised an apparatus that was similar to Thorndike’s
problem box but lacked the Houdini quality. Instead of
learning to operate a contrivance that provides a means
of escape, a hungry animal placed in a “Skinner box”
must manipulate a mechanism that provides a small food
reward (Figure 2.9). A rat may learn to press a lever, and
a pigeon may learn to peck at a key. Patterns of behav-
ior that are rewarded tend to be repeated, or to increase
in frequency, and so learning can be measured as the
number of responses over time. Skinner believed that the
control of behavior was a matter of reinforcement.
Behaviorists began to see basic principles underly-
ing learning that were common to all species. They
expected to find similarities in the learning process
because at that time, the minds of all species were con-
sidered similar in kind. Thus, according to the tradi-
tional view of learning held by the followers of
behaviorism, the minds of humans and animals were
similar in kind and differed only in complexity. In short,
there were general laws of learning that transcended all
species and problems. If this was true, then it was rea-
sonable to study the laws of acquisition, extinction, delay

FIGURE 2.8 Ivan Pavlov described a conditioned reflex

in the dog.

nence and speed of formation of such associations. In

man his increase reaches such a point that an appar-
ently new type of mind results, which conceals the real
continuity of the process. . . . Amongst the minds of
animals that of man leads, not as a demigod from
another planet, but as a king from the same race.

Behaviorism
Another important event that steered comparative psy-
chology toward objectivity and laboratory analysis was
the birth of behaviorism, a school of psychology that
restricts the study of behavior to events that can be
seen—a description of the stimulus and the response it
elicits. Behaviorists sought to eliminate subjectivity from
their studies by concentrating their research efforts on
identifying the stimuli that elicit responses and the
rewards and punishments that maintain them. This was,
indeed, a step toward better science. They designed
experiments that would yield quantifiable data, invented
equipment to measure and record responses, and devel-
oped statistical techniques that could be used to analyze
behavioral data. The assumptions that an animal’s men-
tal capacity could not be measured directly, but its abil-
ity to solve a problem could, again focused attention on
learning ability as a popular research subject. A learned
response could be described objectively, and experiments FIGURE 2.9 B. F. Skinner and his apparatus, the Skinner
could be conducted under the controlled conditions of box. Animals placed in the box learned to operate a
the laboratory. mechanism to obtain a food reward.
20 Chapter 2 / History of the Study of Animal Behavior

of reinforcement, or any other aspect of the learning basis of learning in the rat. He also examined the role of
process in a simple and convenient animal, such as the the brain in emotion and in vision.
domesticated form of the Norway rat (Rattus norvegicus), The comparative psychologist Frank Beach began
and the results could then be broadly applied to other his career by using brain surgery to determine the effects
species. of lesions on the maternal behavior of the rat, but he
later went on to study the effects of hormones on behav-
ior. He analyzed the roles of nerves, hormones, and
STOP AND THINK experience in the sexual behavior of fishes, amphibians,
Cancer is a disease that gives off odors. Assume that you reptiles, birds, and mammals. We will discuss some of his
are a researcher interested in knowing whether dogs can work in the field of behavioral endocrinology in more
detect bladder cancer. What experiment would you detail in Chapter 7.
design? Recognizing that animal behavior is concerned with
the activities of groups of animals, as well as of individ-
uals, some comparative psychologists studied social
THE ROOTS OF PHYSIOLOGICAL behavior. Robert Yerkes, for example, established a
research facility (later named the Yerkes Laboratory of
PSYCHOLOGY
Primate Biology) at Orange Park, Florida, to study a
Although learning was a dominant focus of research dur- wide range of behavior in primates. Some researchers
ing the middle of the twentieth century, it was not the also began to see that although it is often easier to make
only research interest of comparative psychologists. measurements in the laboratory, it is not impossible to
Another research topic was the physiological basis of get good measurements in the field. C. R. Carpenter
behavior. Part of the psychological foundation of behav- studied a variety of primate species, each in its natural
ior is, of course, the nervous system. The comparative setting: howler monkeys in Panama, spider monkeys in
psychologists’ interest in the neurological mechanisms of Central America, and gibbons in Thailand, to name a
behavior can be traced back to Pierre Flourens, a pro- few. T. C. Schneirla used both field observation and lab-
tégé of Baron Cuvier, a famous scientist of nineteenth- oratory experimentation to investigate the social behav-
century France who stressed the importance of ior in army ants. In doing so, he applied the rigorous
laboratory research (Jaynes 1969). Flourens earned his methodology of laboratory researchers to his field stud-
reputation for his studies of the relationship between ies. Such pioneering studies began to help weave the two
behavior and brain structure. For example, he did exper- independent sciences of ethology and comparative psy-
iments in which parts of the brain were removed, such chology together.
as the cerebral hemispheres from a pigeon, to look for
the effect on the animal’s behavior.
Karl Lashley was one comparative psychologist who SOCIOBIOLOGY AND
maintained an interest in physiology, as well as a com- BEHAVIORAL ECOLOGY
parative base of study, during the years when learning by
the laboratory rat dominated the field. His attempts to The field of animal behavior has grown enormously. In
localize learning in the cerebral cortex resulted in the the 1960s and early 1970s, for example, field researchers
rejection of some hypotheses that were widely accepted such as John Crook (1964; 1970) and John Eisenberg and
at the time. For example, based on Pavlov’s ideas, it was colleagues (1972) suggested that ecological context was
assumed that learning depended on the growth or sometimes a better correlate of social behavior than was
strengthening of neural connections between one part of phylogeny (remember that ethologists often focused on
the cerebral cortex and another. To test this idea, Lashley phylogenetic analyses of behavior). Another dramatic
(1950) trained rats on a variety of mazes and discrimi- development was the birth of a new discipline that focused
nation tests and then tried to disrupt the memory by on the application of evolutionary theory to social
making a cut into the cerebral cortex in a different place behavior. This new discipline was called sociobiology.
in each animal. After destroying varying amounts of Another discipline, behavioral ecology examines the
brain tissue, he would then retest the animals to see if ways in which animals interact with their environment to
their behavior changed. In general, he found that when learn how behavior contributes to the animal’s reproduc-
it came to complex problem solving, the entire cerebral tive success and survival. In either case, the key element
cortex was involved, and any particular area was just as of this approach is the idea that behavior should, on aver-
important as any other. He also experimentally addressed age, maximize fitness of individuals. Thus, behavioral
questions such as whether the learned response ecologists attempt to identify the payoffs and costs that
depended on a fixed pattern of muscle movements. play a role in the evolution of behavior (Owens 2006).
Contrary to expectations, he found that they do not. But There has been some question concerning the
Lashley was not solely concerned with the neurological uniqueness of the approach of sociobiology. Some sci-
 Sociobiology and Behavioral Ecology 21

entists, for example, question whether sociobiology is However, it was not until 1975, when E. O. Wilson pub-
really a new discipline or simply part of contemporary lished his landmark text, Sociobiology, that the true impact
ethology (e.g., Dawkins 1989). In contrast, others of sociobiological ideas was felt. The text, an engaging
believe that at least early on, ethology and sociobiology integration of ideas from fields such as ethology, ecol-
could be separated in several ways (Barlow 1989). For ogy, and population biology, gained almost instant noto-
example, whereas classical ethologists tended to derive riety from both within and outside the scientific
hypotheses from detailed observations (i.e., through community. Wilson defined sociobiology as the “sys-
induction), sociobiologists tended to be more deductive, tematic study of the biological basis of all social behav-
typically deriving hypotheses from larger theoretical ior” and proposed that a knowledge of demography (e.g.,
frameworks. Whereas classical ethologists were inter- information on population growth and age structure)
ested in species differences, sociobiologists began to and of the genetic structure of populations was essential
investigate individual differences, examining the costs in understanding the evolution of social behavior.
and benefits of a particular act. Having mentioned some Although sociobiological ideas had been developing
of the questions concerning the precise relationship for several years before the publication of Wilson’s book,
between sociobiology and other fields of animal behav- the text crystallized many of the relevant issues and soon
ior, let us consider the relatively recent “history” of became the focal point for proponents and critics alike.
sociobiology. Criticism arose from both the scientific and political are-
During the late 1960s and early 1970s, most scien- nas. First, in attempting to establish sociobiology,
tists were quite comfortable with the idea that natural Wilson attacked fields such as ethology and comparative
selection acted primarily on individuals. Despite the exis- psychology and made the bold prediction that in due
tence of widespread agreement, however, some nagging time sociobiology would engulf these disciplines. He
issues that seemed inconsistent with selection at the level specifically predicted that ethology and comparative psy-
of the individual remained (Hinde 1982). For example, chology would be “cannibalized by neurophysiology and
how could one explain the evolution of sterile castes in sensory physiology from one end and sociobiology and
species of ants, bees, and wasps? How could the evolu- behavioral ecology from the other” (Wilson 1975).
tion of nonreproducing individuals be consistent with Another area of great concern, this time from the polit-
Darwinian selection? Similarly, how was one to explain ical arena, was the extension of sociobiological thinking,
the evolution of certain patterns of behavior, called altru- in the absence of sound evidence, to human social behav-
istic behavior, that seemed to benefit others but were ior (Cooper 1985). Opponents of sociobiology claimed
costly (with respect to survival and reproduction) to the that Wilson advocated biological determinism, the idea
performer? Why, for example, do some animals give that the present conditions of human societies are sim-
alarm calls when they spot a predator, when calling may ply the result of the biology of the human species and
actually increase their own chances of being detected? therefore cannot be altered. Although only the final
The answer to these questions came in 1964 when W. D. chapter of his text was devoted specifically to humans,
Hamilton published his seminal papers, “The Genetical heated debate over the social and political implications
Evolution of Social Behaviour, I, II.” Hamilton showed of sociobiological theories ensued (e.g., see the collec-
that evolutionary success (the contribution of genes to tion of papers in Caplan 1978).
subsequent generations) should be measured not only by During the 1970s and early 1980s, research on
the number of surviving offspring produced by an indi- sociobiological topics in animal behavior flourished.
vidual but also by the effects of that individual’s actions George Barlow (1989) suggested, “The study of animal
on nondescendant kin (e.g., siblings, nieces, and behavior had indeed begun to stagnate by 1975, and the
nephews). He coined the term inclusive fitness to describe advent of sociobiology was just the kick in the pants the
an individual’s collective genetic success—that is, a com- field needed to get moving again.” The field of animal
bination of direct fitness (own reproduction) and indi- behavior became revitalized because sociobiology pro-
rect fitness (effects on reproduction by nondescendant vided a framework that could be used to test hypotheses
kin). When quantifying an individual’s inclusive fitness, about the adaptiveness or survival value of behavior. But
we count—to varying degrees, depending on how the “kick in the pants” was so strong that for a time
closely they are related—all the offspring, personal or of almost all research in animal behavior was done under
relatives, that are alive because of the actions of that indi- the banner of sociobiology (Bateson and Klopfer 1989).
vidual. This concept of inclusive fitness paved the way Of all possible questions about animal behavior, one—
toward an understanding of the evolution of sterile castes its function, or survival value—had come to dominate
and altruistic acts (these topics are discussed in more the field.
detail in Chapter 19). By the end of the 1980s, however, many researchers
In the years following Hamilton’s (1964) paper, began to notice the imbalance in the study of animal
many studies were conducted in which the idea of inclu- behavior. It became apparent that our understanding of
sive fitness was used to interpret social behavior. animal behavior would be fuller if both its immediate and
22 Chapter 2 / History of the Study of Animal Behavior

evolutionary causes are considered. As Marian Stamp depends on what other members of the population are
Dawkins (Dawkins 1989, p. 53) has said, doing. In such situations, it is often helpful to consider
whether an individual’s choice of action is an evolution-
Genes operate through making bodies do things. These
arily stable strategy (ESS). An ESS is a strategy that,
bodies have to develop and they need machinery (sense
when adopted by most members of the population,
organs, decision centers, and means of executing action)
cannot be invaded by the spread of any rare alternative
to be able to pass their genes on to the next generation.
strategy. The concept of ESS has been applied in stud-
To understand this process fully, we need a science that
ies of mating systems, communication, conflict, and
is not only aware of the evolutionary ebb and flow of
cooperation (Krebs and Davies 1997).
genotypes over evolutionary time, but can look at the
bridge between generations, at the bodies that grow and
move and court and find food and pass their genetic CELLULAR AND MOLECULAR BASIS
cargo on through time with the frailest and most mar- OF BEHAVIOR
velous of flesh-and-blood machinery.
We are also making great strides in understanding the
mechanisms of behavior, largely because of tools and
techniques that were not available even a few years ago.
MORE RECENT TRENDS Today, sign stimuli may be interpreted as filtering at the
level of sensory receptors or as feature detection neurons
The study of animal behavior has seen some changes in that respond to specific features of a stimulus. We can
the areas that are the focus of study. On the whole, how- identify neurons in circuits that underlie FAPs. For
ever, it seems to have returned to research that consid- example, new recording techniques have made it possi-
ers all of Tinbergen’s four questions. Michael Taborsky ble to map the nervous systems of several invertebrates.
(2006) points out that ethology has largely regained its In some animals, specific neurons have been linked to
balance and addresses questions of mechanism as well as specific behaviors. For instance, in the grasshopper
function. Let’s consider some of the recent changes of (Omocestrus viridulus) three different hind leg movements
focus in behavioral research. (FAPs) are involved in producing the courtship sound
signals. By using microinjection techniques and intra-
cellular recording, it has been shown that a specific type
FIELD STUDIES
of brain nerve cell is responsible for each of these three
Following the assumption that natural selection shaped FAPs. During courtship, these nerve cell types are acti-
behavior, we should expect that behavior observed in the vated in a specific sequence (Hedwig and Heinrich
field should increase the animal’s chance of survival. 1997). Fixed action patterns are now discussed in terms
Many of the pioneering field studies were purely of neural networks, command neurons, or central pat-
descriptive, and some also included an explanation of tern generators. In the chapters that follow, we will con-
what was described. However, today’s field studies usu- sider many other examples of how physiology,
ally begin with a clearly stated hypothesis to be tested by neurobiology, and molecular biology have enhanced the
data collection and analysis. The hypothesis generally study of behavior.
relates either to the short-term function of the behavior During the last decade, some avenues of research
or to the long-term fitness (relative number of surviving have focused on the cellular, or even the molecular,
offspring) consequences of a behavior. Thus, the data underpinnings of animal behavior. One of the most
collected are often the outcome or consequence of a exciting areas of research is behavioral genomics—study
behavior, such as the amount of food collected. of the role of an organism’s genetic material in behav-
Researchers might then determine the fitness conse- ior. An organism’s genome consists of all of its DNA.
quences of natural variation in the expression of that One goal of genomics is to learn the sequence of all the
behavior within a population (Altmann and Altmann genes in an organism’s genome. Indeed, scientists have
2003). As we will discuss in later chapters of this book, now determined the sequence of bases in the entire
particularly in Chapters 4 and 12, today’s field studies of genomes of more than 100 organisms. This information
animal behavior usually focus on the costs and benefits allows scientists to zero in on the sequence of a partic-
of a particular behavior, with the common currency ular gene. That sequence can then be compared to a
being reproductive success. Natural selection is often database of known gene sequences, which contains some
assumed to have shaped not just an efficient but also an genes whose function is known. If the location or
optimal form of behavior. For example, we would expect sequence matches that of a gene whose function is
a starling to select the type of prey that will maximize known, it may provide a clue as to the function of the
the amount of food (energy) that can be delivered to its gene of interest.
brood. Also, because an animal’s environment includes Genomics provides a way for researchers to consider
competitors, an individual’s best choice of action often the activity of networks of genes, instead of looking at one
 More Recent Trends 23

gene at a time. Gene sequences reveal which mRNAs to TABLE 2.1 Disciplines in Animal Behavior
look for if the gene is active. This is often accomplished
using DNA microarray analysis, which compares the Discipline Focus
activity of thousands of genes simultaneously. We will see Neuroethology The neurological study of behavior
subsets of genes become active under different condi-
Behavioral The study of the hormonal basis of
tions. Genomics also tells us which proteins to look for
endocrinology behavior
if the gene is active. The next goal is to determine the
functions of the proteins that are produced. The biggest Neuroecology The study of adaptive variation in
cognition and the brain
challenge will be to figure out how the environment and
genome work together to direct the structure and Cognitive ecology An approach that views cognition as
behavior of an individual. As Gene Robinson (2005; p. an adaptive trait shaped by natural
selection
257) declared, the time has come to “achieve a compre-
hensive understanding of social life in molecular terms: Evolutionary An approach to psychology that
how it evolved, how it is governed, and how it influences psychology attempts to explain human mental and
psychological traits as adaptations
all aspects of genome structure, gene expression and
shaped by natural selection
organismal development, physiology and behavior.”
This reductionist approach to animal behavior (try- Behavioral genetics The study of the influence of genetic
or genomics information on behavior
ing to understand the behavior by understanding its
components) has allowed many exciting discoveries, but Applied ethology The study of the behavior of domestic
by the early twenty-first century appeals were being animals or other animals kept in
captivity
made to “return to the whole organism” (e.g., Bateson
2003, 2005). According to these appeals, if you want to
know why an animals behaves a certain way, you must The integration of research focusing on Tinbergen’s
look at the whole organism, not just neurons, genes, and four questions has led to the development of new sub-
molecules (Hogan 2005). disciplines of animal behavior (Table 2.1). A fascinating
subdiscipline is animal cognition or cognitive ethology.
Animal behavior is no longer viewed only as the result
BEHAVIORAL BIOLOGY
of genetic programming or neural wiring or as the result
In recent years, the term behavioral biology has been of a simple stimulus–response reaction. Instead, an
coined to describe behavioral research that includes animal’s mental capabilities are seen as a product of nat-
more than one of Tinbergen’s four questions (Taborsky ural selection. The field of study began with Donald
2006). The themes of returning to studying behavior at Griffin’s controversial book, The Question of Animal
the level of the whole organism and integrated studies Awareness (1976). Griffin later (2001) named the field
of the four questions are threaded through the papers cognitive ethology. It is an interdisciplinary area of
celebrating the fortieth anniversary of Tinbergen’s clas- research that brings Tinbergen’s four questions to bear
sic paper on the four questions1. Michael Ryan (2005) on the study of animals’ mental experiences. Three areas
argues that an approach to animal behavior that integrates of research, in particular, are making rapid progress: ani-
Tinbergen’s four questions—cause, development, sur- mal communication, seed caching and recovery, and nav-
vival value, and evolution—is needed to provide a com- igation and orientation (Balda et al. 1998). You will read
plete and correct understanding of behavior. More is more about these subjects in later chapters of this book.
learned by integrating the aims and methods associated
with each question than by studying each question in iso-
lation. Ryan illustrates this idea by explaining how APPLIED ANIMAL BEHAVIOR
knowledge of the evolutionary history of calls of túngara Professional opportunities are growing in applied animal
frogs (Physalaemus pustulosus) helped researchers under- behavior, the study of animal behavior with practical
stand the mechanism of male calling and female implications rather than just for the sake of accumulating
response, as well as how it develops and increases fitness. knowledge. This subfield of applied animal behavior is
And David Sherry (2005) suggests that knowledge of the itself divided into other disciplines. Many applied animal
survival value or function of a behavior can assist behaviorists focus on captive animals. Some work with
research on the causes of the behavior. For example, companion animals, such as dogs and cats, training them
knowing that the ability to sense the earth’s magnetic and solving behavioral problems. Others study the posi-
field serves an orientation function suggests the proper- tive effects of the human–animal bond: pets, for example,
ties that a magnetoreceptor must have. improve the mental health of many elderly. Still other
researchers work with laboratory, zoo, and farm animals.
1
A collection of these papers can be found in Animal Biology Many applied animal behaviorists work to improve
2005(4): 55. the welfare of captive animals (Fraser and Weary 2005).
24 Chapter 2 / History of the Study of Animal Behavior

To understand the challenges of this discipline, consider Other applied animal behaviorists work with wild
the “Five Freedoms” for captive animals proposed by the animals. Among them are professionals who work in
Farm Animal Welfare Council in the United Kingdom: wildlife management (e.g., increasing the population of
(1) freedom from thirst, hunger, and malnutrition, (2) game species) and pest management. An increasingly
freedom from discomfort due to environment, (3) free- important field is conservation behavior, in which the
dom from pain, injury, and disease, (4) freedom to principles of animal behavior are used in efforts to con-
express the normal behavior of the species, and (5) free- serve biodiversity. As human populations spill over into
dom from fear and distress. Whether certain of these the habitats of animals, many populations of animals are
freedoms are provided is easy to determine. For instance, declining or disappearing. To halt or reverse these
we can see whether the animals have ready access to fresh losses, we need behavioral data about habitat prefer-
water and a healthy diet. It is generally possible to see ences, migratory routes, territory size, social organiza-
that an animal is injured or diseased and to provide rapid tion, food requirements, risk of predation, mating
diagnosis and treatment. It is more difficult to be sure habits, and more. These data are crucial for designing
that we are providing animals with the other freedoms. effective nature preserves. For example, studies of trop-
For instance, in order to ensure that animals have the ical birds revealed the paths of their migratory routes
freedom to express normal behavior we must know what up and down mountains, and conservationists subse-
normal behavior is. This may require new studies of cap- quently protected corridors of land that connected pre-
tive species (or their close relatives) in their natural habi- serves on the mountaintop and in the valleys.
tat. In order to ensure that animals are free from Conservation behaviorists also may breed animals in
discomfort, pain, fear, and distress, we must understand captivity for return to the wild. This requires knowledge
something about the mental state of other species. Table of the communication signals used in mating.
2.2 categorizes some of the ways that researchers assess Reintroduction of a captive-born animal to the natural
stress in animals. habitat requires training to recognize and avoid preda-
tors (see Chapter 5). Throughout the book, we will
mention the work of applied animal behaviorists.
STOP AND THINK
Dust bathing is a natural behavior that chickens perform Today there is a sense of rejuvenation in the study
to keep their feathers in good condition and rid themselves of animal behavior, largely because many disciplines are
of mites. If you were charged with designing commercial
now contributing to its study (van Staaden 1998). New
chicken cages with the welfare of chickens in mind, what
techniques and interactions among disciplines allow us
experiment would you perform to determine the “impor-
tance” of the opportunity to dust bathe to chickens? to ask and answer many questions about behavior that
could not be addressed previously.

TABLE 2.2 Three Conceptions of Animal Welfare and Typical Measures Used
to Provide Positive Evidence of Animal Welfare
Conception of animal welfare Typical measures

Biological function Increase in stress hormones (–)

Reduction in immune competence (–)
Incidence of disease and injury (–)
Survival rate (+)
Growth rate (+)
Reproductive success (+)
Affective states Behavioral signs of fear, pain, frustration, etc. (–)
Physiological changes thought to reflect fear, pain, etc. (–)
Behavioral signs of aversion or learned avoidance (–)
Behavioral indicators of comfort/contentment (+)
Performance of behavior (e.g., play) thought to be pleasurable (+)
Behavioral signs of approach/preference (+)
Natural living Performance of natural behavior (+)
Behavioral/physiological indicators of thwarted natural behavior (–)
Performance of abnormal behavior (–)

Source: D. Fraser and D. M. Weary. 2005. Applied animal behavior and animal welfare. In The Behavior of Animals:
Mechanisms, Function, and Evolution, edited by J. J. Bolhuis and L.-A. Giraldeau, Table 15.1, p. 364. Malden, MA:
Blackwell Publishing.
 Summary 25

Study of the history of animal behavior will show us methodology for classical conditioning. Behaviorism is
that whether our primary interest is the mechanism or a school of psychology that proposes limiting the study
the function of behavior, our efforts will be most fruit- of behavior to actions that can be observed. B. F. Skinner,
ful if we keep a clear focus on behavior as the driving a prominent behaviorist, found that patterns of behav-
interest of research. ior that are rewarded tend to be repeated or to increase
in frequency, and he concluded that the control of behav-
ior was largely a matter of reinforcement.
SUMMARY The physiological basis of behavior is another tra-
ditional subject investigated by comparative psycholo-
Perhaps the most important concept in the study of ani- gists. Despite their emphasis on learning and physiology
mal behavior is Darwin’s idea of evolution through nat- in the laboratory, some comparative psychologists stud-
ural selection, which provides the evolutionary ied the social behavior of animals in the field.
framework necessary for the development of animal In the 1960s, a new discipline emerged in the study
behavior. of animal behavior; this discipline, called sociobiology
In the early 1900s, the two dominant approaches to (or sometimes behavioral ecology), focused on the appli-
the study of animal behavior were ethology, centered in cation of evolutionary theory to social behavior. W. D.
Europe, and comparative psychology, headquartered in Hamilton articulated one of its central concepts, that of
the United States. Ethologists focused primarily on the inclusive fitness, in 1964. According to Hamilton, indi-
function and evolution of behavior. Because the context viduals behave in such a manner as to maximize their
in which a behavior is displayed is sometimes a clue to inclusive fitness (i.e., their own survival and reproduc-
its function, ethologists often studied behavior under tion plus that of their relatives) rather than acting sim-
field conditions. It followed from their interest in evo- ply to maximize their own fitness. Suddenly, certain
lution that ethologists used a comparative approach and issues that seemed inconsistent with selection at the level
studied primarily innate behaviors. of the individual, such as the evolution of sterile castes
Early ethologists were interested in stereotyped pat- in insects and altruistic behavior (behavior that benefits
terns of behavior, considering them to be just as reliable others at the expense of the performer), were explain-
as morphological characters in defining a particular able.
group. These stereotyped behaviors were called fixed Approximately ten years after Hamilton’s paper,
action patterns (FAPs). An FAP is triggered by a very E. O. Wilson crystallized sociobiological ideas in his
specific stimulus. That portion of the total stimulus that landmark text, Sociobiology. Sociobiology and an interest
releases the FAP is called the sign stimulus or releaser. in the survival value of behavior dominated the study of
Because most behaviors are not so stereotyped as animal behavior for approximately a decade, but it soon
implied by the notion of FAP, they have more recently became apparent that a complete understanding of
been described as modal action patterns (MAPs). behavior requires knowledge of both mechanism and
In contrast to the early ethologists, comparative psy- function. As new technologies became available,
chologists emphasized laboratory studies of observable, researchers began to explore the mechanisms of behav-
quantifiable patterns of behavior. In general, they asked ior on a molecular or cellular level. Today, the study of
questions that concerned the development or causation animal behavior has returned to a more balanced
of behavior. Learning and the physiological bases of approach that considers mechanism and function.
behavior were the focus of much of their research. Research is conducted in the laboratory, as well as in the
Many exciting advances were made in the study of field. Information gathered in this research is being
learning. Thorndike developed the techniques for study- applied to assist the welfare of captive animals and to
ing trial-and-error learning, and Pavlov provided the study conservation biology.
This page intentionally left blank
 3
Genetic Analysis of Behavior

Basics of Gene Action Networks of Genes Are Responsive

to the Environment
Goals of Behavioral Genetics
Epigenetics and Behavioral Genetics
Methods of Behavioral Genetics
Complex Relationships Among Genes
Inbreeding
Artificial Selection A Broader Perspective
Inducing Mutations and Screening for Change
in Behavior
Picture this: In front of you are two cages, each contain-
Finding Natural Variants and Looking
for Genetic Differences
ing a female rat and her litter of week-old pups. Your col-
league enters the room, and the door slams behind her.
Hybridization
The pups in one cage jump in response to the sound,
The Foraging Gene as an Example of Behavioral while those in the other cage continue grooming or sleep-
Genetics in Action
ing. Why do the pups behave so differently when they
Candidate Genes hear loud sound? In this case, it is fair to blame the
Linking a Protein to a Trait mother because, in addition to influencing the future
Locating All the Genes Associated with a Trait mothering style of the female pups, the quality of mater-
Microarray Analysis nal care affects the pups’ response to stress, both now and
Important Principles of Behavioral Genetics in adulthood. The mother of the calm pups spends a great
One Gene Usually Affects Several Traits deal of time grooming and nursing her pups (Figure 3.1);
Genes Work in Interacting Networks
the mother of the skittish pups is neglectful.
Although it might be tempting to explain these dif-
Behavioral Variation and Genes
ferences in behavior as being due to differences in either
Environmental Regulation of Gene Expression learning or genes, we will learn in this chapter that the
Dominance Relationships in Cichlid Fish relationships among genes, experience, and behavior are
Song Learning in Male Songbirds not that simple. For example, we will see that the expe-
The Importance of Genetic Background riences during the first week of life alter the activity of
to Behavioral Genetics two genes in a nearly permanent way and influence both

27
28 Chapter 3 / Genetic Analysis of Behavior

FIGURE 3.1 The quality of maternal care that

rat pups receive determines how timid the
pups will be and the quality of maternal care
that the females will give to their own pups
later in life.

how the adult rats will respond to stress and how attentive changes in gene expression alter the nervous system and
the females will be to their own pups. Furthermore, this physiology, and behavior is modified (Robinson 2008).
change in gene activity is transmitted to the next gener- Thus, we see that the effect of genes on behavior can be
ation. So, is the behavior determined by gene activity? dynamic. When we see animals that differ from one
Yes. Is it determined by early experience? Yes. How is another in behavior, it seems almost impossible that we
this possible? We will answer that question later in this will be able to say precisely how they differ. Yet this is
chapter, after discussing some of the ways that genes can exactly what the thriving field of behavioral genetics is
influence behavior. Surprisingly, the answer suggests that giving us the ability to do, at least for some behaviors.
your behavior today may be influenced by your great-
grandmother’s lifestyle.
The relationship between genes and behavior is BASICS OF GENE ACTION
often difficult to decipher, but we do know that there is
not a one-to-one correspondence between genes and What do genes really do, then? How do they work? As
behavior. Consider a very simple behavior: a fly extends you may already know, genes can direct the synthesis of
its proboscis (a tubular structure of mouthparts) when the proteins. Each protein is specified by a different gene,
sense organs on its feet detect sugar. Think of all the parts or if the protein consists of more than one chain of
of the fly that need to be in good working order for the amino acids, a gene specifies one of those chains. The
fly to successfully perform this behavior: the physical protein may be structural and be used as a building block
structures (the sense organs on the feet, for example) of the organism, or it may be regulatory. A regulatory
must be functional, as must the neural circuitry to carry protein may modify the activity of other genes. In other
the information from the sense organs to the fly’s brain, words, genes can code for specific proteins, and the pro-
which must assess the information and send signals via teins affect the composition and organization of the ani-
motor neurons to the muscles of the proboscis, which mal in ways that influence how it behaves. An animal’s
must be able to contract. Each of these pieces is essen- sensory receptors detect stimuli and send information to
tial to behavior, and their structure and function are influ- the nervous system, where it is further interpreted and
enced by many genes. In addition, most steps are analyzed. The nervous system may then initiate a
influenced by both the internal physiological state of the response by effectors (muscles and glands) that results in
animal (Is it hungry? Has it already learned anything behavior. Genes direct the development of the structure
about the environment?) and the external environment. and function of receptors, nerves, muscles, and glands.
There are two ways that genes can affect behavior. Alterations in genes may change the proteins they code
In the example of fruit fly foraging just described, genes for, with the result that anatomy or physiology may be
alter behavior through their effects on development of altered in a manner that changes behavior. We will look
the nervous system and physiology. We will see other at a few examples of the links between genes and behav-
examples of this relationship throughout the chapter. ior later in this chapter.
But we will also consider examples of behavior in which To understand the relationship between genes and
the environment, especially social interactions, trigger proteins, it is helpful to know a little biochemistry.
changes in the nervous system or physiology. These Genes are made up of DNA (deoxyribonucleic acid).
changes then alter the pattern of gene expression; that The structure of DNA is somewhat like a long ladder,
is, they turn some genes on and others off. In turn, twisted about itself like a spiral staircase. The DNA lad-
 Basics of Gene Action 29

der is composed of two long strings of smaller molecules the accurate production of new DNA molecules, but
called nucleotides. Each nucleotide chain makes up one also for conversion of the information in the gene into
side of the ladder and half of each rung. A nucleotide a protein.
consists of a phosphate, a nitrogenous base, and a sugar The instructions for each protein are written as the
called deoxyribose. There are only four different nitro- sequence of bases in the DNA molecule. The first step
gen-containing bases in DNA: adenine (A), thymine is to transcribe the information in DNA into RNA
(T), cytosine (C), and guanine (G). Although DNA has (ribonucleic acid), specifically messenger RNA (mRNA).
only four different nucleotides, a DNA molecule is very The DNA is unzipped for part of its length so that an
long and has thousands of nucleotides. When forming mRNA molecule can be formed. This first mRNA strand
a rung of the ladder, adenine must pair with thymine is modified or edited before it leaves the nucleus. Some
and cytosine must pair with guanine (Figure 3.2). The of the regions of the mRNA strand that do not code for
specificity of these base pairs is important, not only for a protein are then snipped out.

DNA molecule Transcription Translation

3'
A T 5'
T A
G C
C 5'

TA
A•••T
G C An enzyme G
A•••T A T RNA polymerase Glutamic
CG A glu
acid
G
C•••G A 3' T
A T U A
A•••T
O G G C A
T•••A C G
C
C thr Threonine
C G•••C G G C
O G G C U
Sugar A T
A
(deoxyribose)
C•••G T U A
Phosphate A
C•••G G G C
Nucleotide G ser Serine
C C G
A•••T
G G C C
T•••A
U

AT
A

G
A

AT
G

G•••C C G A U val Valine

T A
T•••A G
CG A
C•••G 5'
AT
A•••T m RNA
G C
5' 3' 3'

DNA Part of Part of

mRNA a protein
molecule

DNA RNA Protein

FIGURE 3.2 A diagrammatic representation of the biochemistry of gene expression. A gene is a

region of a DNA molecule that has the information needed to make a specific protein. A DNA
molecule is composed of nucleotides. A nucleotide, shown on the extreme left of the figures, is a
nitrogenous base, a sugar (deoxyribose), and a phosphate. In the ladderlike DNA molecule, the
two uprights are composed of alternating sugar and phosphate groups, and the rungs are paired
nitrogenous bases. The pairing of bases is specific: adenine with thymine and cytosine with gua-
nine. During transcription, the synthesis of mRNA, the sequence of bases on the DNA molecule
is converted to the complementary sequence of bases on the mRNA molecule. Each unit of
three bases on the mRNA molecule signifies a particular amino acid. The message of messenger
RNA is, therefore, its sequence of bases that determines the order and kinds of amino acids in
the protein product.
30 Chapter 3 / Genetic Analysis of Behavior

The mRNA has a structure similar to DNA except the membranes of nerve cells that allow potassium ions
for three differences: It is single-stranded, its sugar is to pass through. Potassium channels are critical to nor-
ribose, and the base uracil substitutes for thymine. mal nerve cell function, and so the mutation causes a
Because adenine must pair with uracil and cytosine must nerve cell defect and abnormal behavior (Kaplan and
pair with guanine, the sequence of bases on RNA is spec- Trout 1969).
ified by the sequence of bases on DNA. Because the A recurrent theme in behavioral genetics is that
DNA molecule is so long, the four bases can be ordered behavior depends on which genes are expressed in which
in many ways. Therefore, with only four bases, DNA can tissues and when. Although all the cells of an animal’s
encode the information needed for the synthesis of a body have the same genes, some of them are turned off
myriad of different proteins. during development. When genes are turned off, they do
The next step is to translate the order of bases on not produce a protein. If these genes are then turned on,
the mRNA molecule into a protein. Proteins are long however, their proteins are produced, and in some cases,
chains of amino acids. Each different protein has a they will modify a structure or function in a way that will
unique order of amino acids. A group of three bases on alter behavior. Thus, an organism’s behavior may change
the mRNA molecule is translated, three bases at a time, as specific genes are turned on or off in specific tissues.
into a protein. The order of bases on DNA specifies the This applies even in the short term, as the animal expe-
sequence of bases in mRNA, which can be translated into riences new stimuli.
only one array of amino acids. Thus, the information in The activity of specific genes is often influenced by
the gene is its sequence of bases that codes for a specific regulatory genes, of which there are many types. Certain
protein. The transfer of information can be summarized regulatory genes modify the activity of other genes
as follows: through the production of proteins called transcription
factors. Transcription factors and other products of reg-
Sequence of bases in DNA S Sequence of bases in
ulatory genes increase gene expression or decrease gene
mRNA S Sequence of amino acids in a protein.
expression, altering the amount of mRNA produced.
For most proteins, and ultimately the traits they Thus, gene regulation determines when and where a
influence, different forms exist because the underlying protein will be produced, as well as how much of that
DNA sequences are different. For example, coat color protein is produced. Mutations in regulatory genes often
in mice might be black or brown. In other words, this have a more widespread effect on the organism than do
gene for coat color can be expressed in two different structural genes because these genes produce proteins
ways; one resulting from intense black pigment granules, that regulate many other genes.
and the other from chocolate brown pigment granules. Consider the mechanisms by which genes regulate
These alternative forms of a gene are called alleles. courtship and mating in fruit flies. If you have ever left
Sometimes, as in the eye color of fruit flies, there are a ripe banana in your kitchen too long, you may know
many possible alleles. The wild type, or most common that mating is something fruit flies do well, but you may
eye color of fruit flies, is red, but other alleles of this gene not know that the choreography is complex. The dance
can result in white or vermilion (a brilliant red tinged begins with orientation, during which the male faces the
with orange) eyes. Furthermore, most of the animal female and taps her on the abdomen with his foreleg. If
species used in genetic studies are diploid, meaning that she wanders away, he follows her. Next, he begins to
an individual possesses two alleles of each gene, one from “sing” a courtship song by fluttering a single out-
each parent, and this influences how the gene is stretched wing. If the female does not show interest, he
expressed. If the two alleles of a gene are identical, the will repeat these actions. When the female seems recep-
individual is said to be homozygous for the trait. tive, he extends his proboscis and licks the female’s gen-
However, an individual may inherit different alleles for italia. Next, he will try to copulate with her. If the
a gene from its mother and father. Such an individual is attempt fails, he will wait a few moments before starting
heterozygous for the gene. The genetic diversity the ritual from the beginning (Hall 1994).
among unrelated individuals results from the particular The regulatory gene fruitless (fru) affects nearly
alleles that they possess and from whether or not they every aspect of male courtship in fruit flies and provides
are heterozygous. an example of gene regulation at several levels—sensory
Structural genes produce proteins that become part processing, choice of behavior, and carrying out the
of a structure or that have a specific function within an behavior (Dickson 2008). Fru is but one gene in a hier-
organism. Consider, for example, the Shaker gene in the archy of regulatory genes that influence the final behav-
fruit fly, Drosophila melanogaster. A particular mutation— ior pattern. In such a hierarchy, the protein product of
that is, a specific change in the sequence of bases in the one gene regulates the activity of another gene whose
DNA—in the Shaker gene results in flies that shake vio- protein product affects the activity of perhaps a third
lently under anesthesia. It turns out that the mutation gene and fourth gene, and so on. Recall that a newly
changes a protein used in the formation of channels in formed mRNA is edited or spliced before it is translated
 Methods of Behavioral Genetics 31

into protein. Fru mRNA transcripts are edited differ- als. Then, typically by comparing relatives, we can deter-
ently in males and females, under the direction of a gene mine how much of the observed variation is due to
higher in the hierarchy (transformer or tra) that responds genetic differences among the individuals and how much
to the number of X chromosomes present. When only is caused by differences in their environments. The
one X chromosome is present, the male-specific versions heritability of a particular trait in a specific population
of these fru transcription factors are produced. In turn, is the ratio of the variation caused by genetic differences
these fru transcription factors regulate genes in certain to the total amount of variability in the trait in that pop-
neurons to build the correct neural circuitry for male ulation. Therefore, heritability can vary from 0 to 1. A
courtship and sex determination in fruit flies. The value of 0.5 indicates that 50% of the variability in the
female-specific mRNA transcripts of fru do not direct population studied is due to genetic differences. The
the development of the same circuitry (Baker et al. 2001; heritability of a complex trait such as behavior is rarely
Dulac 2005). more than 50% (Plomin et al. 2003).
Male-specific fru proteins alter the way that the Genomics, the study of an organism’s entire
neural circuit underlying sexual behavior functions to genome—all of its DNA—has had a major impact on
produce male courtship behavior. This difference is most behavioral genetics. The field took off during the 1990s
likely due to an identifiable subset of fru-expressing with the beginning of gene-sequencing projects designed
interneurons that are present only in males. In the to discover the sequence of nucleotides in entire genomes.
female, these neurons are programmed to die. The pres- The goal of functional genomics is to understand the
ence of the male-specific fru protein prevents the death function of genes and noncoding regions of the genome.
of these interneurons, which alters the way in which the It often begins with the wealth of information created by
neural circuit for sexual behavior functions in a male gene-sequencing projects, but functional genomics is pri-
(Kimura et al. 2005). marily interested in patterns of gene activity under dif-
Fru is expressed in only about 500 neurons, roughly ferent conditions or at different developmental stages.
1.5% of the neurons in the central nervous system. Comparative genomics analyzes differences in the
However, fru is also expressed in nearly all of the sen- genomes of different species. Its goal is to understand how
sory neurons involved in courtship, especially olfactory traits have evolved, as well as how they work physiologi-
sensory neurons. Thus, as you might expect, disruptions cally. The next level of study is proteomics, which strives
of fru lead to changes in sexual behavior. Male fru to study the full set of proteins coded for by an organism’s
mutants are bisexual. A group of only male fru mutants genes and to understand how these proteins work
form courtship chains in which each male chases and together to produce and modify traits. As we will see, dif-
courts the male in front, forming revolving circles of ferent genes are expressed in different degrees and in dif-
courting flies. Expression of male-specific fru causes ferent tissues at different times during development and
females to display male courtship behavior instead of under different environmental conditions.
female behavior (Manoli et al. 2005). We see, then, that
this single regulatory gene has profound influences on
many aspects of mating behavior.
METHODS OF BEHAVIORAL
GENETICS
GOALS OF BEHAVIORAL
As we have seen, the observed variation in behavior
GENETICS among individuals results from differences in genes and
When we observe an animal in nature, its actions are in environments. Therefore, when we are interested in
generally a result of many genes interacting with one exploring the influence of genes on behavior, it is impor-
another and the environment. A goal of behavioral tant to rule out environmental effects by raising the
genetics, then, is to identify the gene, or more commonly animals in the same environment. If environmental con-
the genes, that underlie a behavior and to learn the func- ditions are identical for all animals, then observed behav-
tions of these genes. Another goal of behavior genetics ioral differences are due to genetic differences.
is to decipher the interactions among genes and their
products and between genes and the environment to
INBREEDING
understand why a particular behavior takes the form it
does. In order to understand these interactions, it is often Inbred lines, one of the tools used by behavioral geneti-
useful to quantify the heritability of a behavior. cists, are laboratory colonies of individuals that have no,
Heritability is a statistical measure that suggests how or virtually no, genetic diversity: they are homozygous for
strongly a behavior is influenced by genes. Because it is nearly all their genes. Inbred strains are usually created
a statistical measure, we must measure the differences in by mating close family members with one another for
the behavior of a sufficiently large number of individu- many generations and are useful in behavioral genetics
32 Chapter 3 / Genetic Analysis of Behavior

because they create a population with nearly identical a Model predator

2.0
genes. By using inbred strains, therefore, it is possible to with eyes
separate the effects of genes from those of the environ-
ment. To show the effects of genes, the behavior of
members of two different inbred strains is compared in 1.5
the same environment. In this case, any observed differ-

Frequency of fleeing
Responses of P strain
ence in behavior must be caused by a difference between
the genes of the strain. Moreover, the influence of the Responses of S strain
1.0
environment on behavior can also be shown by using
inbred strains. If members of the same inbred strain,
individuals who are almost genetically identical, behave
differently when they are raised under different condi- 0.5 Model predator
tions, then the variation must be caused by environ- without eyes
mental effects.

0.0
Comparisons of Inbred Strains P S P S
to Show the Role of Genes
The work of Ádám Miklósi and his colleagues (1997) on b 1.5
the antipredator behavior of paradise fish (Macropodus
opercularis) larvae is an example of studies that compare
Frequency of backing
the behavior of inbred strains. Paradise fish live in
densely vegetated, shallow marshes and rice fields in 1.0
Southeast Asia, along with several predator fish species.
Avoiding predation is crucial to survival, and so one
might suspect it has a genetic basis. Having raised strains
0.5
of paradise fish that had been inbred for over 30 gener-
ations, Miklósi could investigate this possibility. He
crafted model predators from plastic centrifuge tubes.
Since eyes are known to be an important cue in preda- 0.0
P S
tor recognition in many species, some models were cre-
ated with black eyespots. Miklósi individually placed the FIGURE 3.3 The antipredator responses of larvae of two
larvae of two inbred strains (S and P) of paradise fish into inbred strains of paradise fish when presented with
an experimental tank and observed their responses to model predators. Fleeing consists of darting by slapping
model predators for three minutes. There are two com- the caudal fin. Backing involves swimming backward
mon antipredator responses: fleeing, in which the larva with a curved body in a direction perpendicular to the
suddenly darts by slapping with its caudal fin, and back- body axes of the larva. Larvae of strain P showed a sig-
nificantly higher frequency of both fleeing and backing
ing, in which the larva swims backward with its body
than did larvae of strain S, indicating genetic differences
curved. Larvae of strain P were significantly more likely in antipredator behavior between the strains. (Data
to show antipredator responses, both fleeing and back- modified from Miklósi et al. 1997.)
ing, than were the larvae of strain S (Figure 3.3).

Comparisons of Inbred Strains almost every gene, the hybrid offspring of both of these
to Show the Role of Environment crosses have the same genotype, even though their
Because individuals of inbred strains are identical in 98% mothers are from different strains. Therefore, if the
of their genes, such strains, as we mentioned, can pro- behavior of the hybrid offspring of reciprocal crosses dif-
vide a way to hold the genetic input constant while vary- fers, it must be an effect of the parental environment. If
ing the environment. Thus, if differences in behavior are the young are raised solely by their mothers, as they
found, they must be due to the environment. often are in the laboratory, the difference must be an
Some of the environmental effects on behavior take effect of the maternal environment.
place very early in life. Eliminating the effects of early Theoretically, it is even possible to determine
learning is tricky, but it can be done. The simplest way whether the maternal environment had its greatest
is by a reciprocal cross, one in which males of inbred influence on the offspring before or after birth. Cross-
strain A are mated with females of inbred strain B and fostering, transferring the offspring shortly after birth
males of strain B are mated with females of strain A. to a mother of a different strain, is a technique for
Since the individuals of each strain are homozygous for detecting maternal influences that occur after birth. If
 Methods of Behavioral Genetics 33

offspring that were transferred to a foster mother imme- blance. For example, when siblings are adopted into
diately after birth behave more like individuals of the different families, we observe genetically related indi-
foster mother’s strain than like those of their own strain, viduals who do not share the same family environment.
postnatal maternal influences are implicated. The degree to which these siblings are similar is an
A cross-fostering experiment involving two differ- indication of the contribution of genes to the family
ent species of voles helped separate the influences of resemblance. On the other hand, when families with
genes and parental environment in the expression of a children of their own adopt additional children, we
particular behavior. Prairie voles, Microtus ochrogaster, observe children with different genetic backgrounds
show more parental care than do meadow voles, M. living in the same environment. The degree to which
pennsylvanicus. For example, prairie vole females spend the family’s own children resemble the adopted chil-
more time in the nest with their young, contacting their dren is an indication of environment in the develop-
offspring by huddling over them and nursing them. Male ment of that trait.
prairie voles also show more parental care than do male Twin studies are another favorite tool of behavioral
meadow voles. Prairie vole males share a nest with the geneticists interested in human behavior. In these stud-
female and frequently groom and huddle over the young, ies, twins are raised in similar family environments.
in contrast to meadow vole males, who nest separately Researchers then compare the similarities in a particu-
and rarely enter the female’s nest. lar trait between sets of identical twins (who have 100%
To determine whether the species difference in of their genes in common) and between sets of fraternal
parental care was due to genes or early experience, Betty twins (who, on average, have 50% of their genes in com-
McGuire (1988) fostered meadow vole pups with prairie mon). It is reasoned that similarities between identical
vole parents. As a control, she fostered meadow vole twins that are not found in fraternal twins must be due
pups to other meadow vole parents. When the foster to their genes. Perhaps the most powerful design is the
pups became adults and had their own families, she mea- adoption-twin studies. In this design, twins adopted into
sured the amount of parental care they gave to their sec- different families are compared to twins raised in the
ond litters. The meadow vole pups that were raised by same family.
prairie voles gave more care to their own offspring than
did those pups that had been fostered to other meadow
ARTIFICIAL SELECTION
vole parents. Cross-fostered females, for example, spent
more time huddling over and nursing their young. The Artificial selection is another means of demonstrating
early experience of male voles influenced their parental that a behavior has a genetic basis. It differs from nat-
behavior in much the same way as it affected the females’ ural selection in that an experimenter, not “nature,’’
behavior. Male meadow vole pups that were fostered to decides which individuals will breed and leave offspring.
parents of their own species behaved as meadow vole The rationale for artificial selection is that if the fre-
males usually do, in that they rarely entered the nest with quency of a trait in a population can be altered by choos-
the young. However, four of the eight meadow vole ing the appropriate breeders, it must have a genetic basis.
males raised by prairie vole parents nested with their Usually the first step is to test individuals of a genetically
mates and spent time in contact with the young. This variable population for a particular behavior trait. Those
cross-fostering experiment shows that the experience a individuals who show the desired attribute are mated
vole has with its own mother can influence the way it with one another, and those who lack the trait are pre-
treats its own offspring. However, not all behaviors were vented from breeding. If the trait has a genetic basis, the
modified by early experience. Nonsocial behaviors such alleles responsible for it will increase in frequency in the
as food caching and tunnel building and overall activity population because only those possessing them are pro-
level were unaffected by the species of the foster parents. ducing offspring. As a result, the behavior becomes more
Thus, using the technique of cross-fostering, it is possi- common or exaggerated with each successive generation.
ble to determine whether a particular behavior is influ- If the environment is held constant, traits that change
enced by the parental environment. under artificial selection must have some genetic basis.
Of course, it would be unethical to create inbred Humans often use artificial selection to create
strains in humans. However, we can take advantage of nat- breeds of animals with traits that they consider useful.
urally occurring differences in genetic relationships among For example, even if you are not a dog lover, you have
family members. Parents and offspring share 50% of their probably noticed that different breeds of dogs have dif-
genes, and they may also share a similar environment by ferent personalities. All dogs belong to the same species,
living together in the same household. So if children raised but various behavior traits and hunting skills have been
by strict, aggressive parents grow up to become play- selected for in different breeds (strains). Terriers, for
ground bullies, is it due to genes or environment? example, are fighters. Aggressiveness was selected for so
Adoption studies are one way to tease apart the role that they could be used to attack small game. The bea-
of genes and environment as causes for family resem- gle, on the other hand, was bred to be a scent hound.
34 Chapter 3 / Genetic Analysis of Behavior

Since beagles usually work in packs to sniff out game, their nests. Then, she selectively mated mice to create
they must be more tolerant with companions. Therefore, lines of high and low nest-building behavior. She mated
their aggressiveness was reduced by selective breeding. males and females that built large nests to create high
Shetland sheep dogs were bred for their ability to learn lines and males and females that built small nests to cre-
to herd sheep. Spaniels, used for hunting birds, are “peo- ate low lines. In addition, she created control lines by
ple dogs,’’ being more affectionate than aggressive. randomly mating males and females from each genera-
These behavioral differences among breeds of dogs were tion. After 15 generations of artificial selection, the mice
brought about by people who placed a premium on cer- of the high nest-building line used an average of 40
tain traits and then arranged matings between individ- grams of cotton for their nests. In contrast, the mice in
ual dogs that showed the desired behavior. In other the low nest-building line used an average of only 5
words, the frequency of particular behavior patterns pre- grams of cotton in nest construction. The mice of the
sent in all breeds has been modified through artificial control line built nests about the same size as those built
selection (Scott and Fuller 1965). by the mice in the initial population—15 grams of cot-
Artificial selection for nesting behavior in house ton (Figure 3.5). Selection for high and low nesting
mice (Mus domesticus) has not only demonstrated a behavior was continued for over 40 generations, at which
genetic basis for the behavior but has also shed some time mice from the high nest-building line collected
light on how natural selection might work on the trait more than 40 times the amount of cotton than did the
in wild populations. House mice usually live in fields, mice of the low line.
where they build nests of grasses and other soft plant The results of this experiment confirm that nest
material. In the laboratory, both male and female house building in house mice does indeed have a genetic basis.
mice will use cotton as nesting material, which makes it Thus, natural selection could have a similar influence on
easy to quantify the size of the nest constructed. Carol nest building if nest size influences fitness (the number
Lynch (1980) noticed that some mice built larger nests of offspring successfully raised). House mice do, in fact,
than others did (Figure 3.4). These differences could be build larger nests in the north than in the south (Lynch
due to genetic or environmental factors, or both. Lynch 1992). This suggests that large nests may be a factor that
suspected that there was some genetic basis in nesting helps mice in cold environments raise more offspring. In
behavior. To separate the genetic and environmental the laboratory, Lynch bred groups of mice from both
influences, she began to selectively mate mice, based on lines at 22ºC or at 4ºC and counted the number of off-
the size of the nest they built, and she raised all mice spring that survived to 40 days of age. Pup survival in
under the same environmental conditions. She began both lines was reduced at the lower temperature.
with a population of house mice that gathered between Nonetheless, mice from the lines that built larger nests
13 and 18 grams of cotton over a four-day period to build raised more pups that lived to be 40 days old at both
environmental temperatures. Thus, nest building is an
important component of fitness, and its genetic basis
allows it to be shaped by natural selection (Bult and
Lynch 1997).
It is interesting to note, however, that selection can
lead to the same observed behavior—constructing large
or small nests—by favoring different sets of genes.
Notice in Figure 3.5 that two high lines and two low
strains were created through artificial selection. Crosses
between mice of different high strains (or between mice
of different low strains) revealed that there were still
genetic differences between strains that expressed the
behavior in similar ways. This suggests that natural selec-
tion can follow different paths in different populations
to produce the same adaptive behavior (Bult and Lynch
1996).
In some cases, selection influences behavior by affect-
ing genes that code for proteins that affect the structure
FIGURE 3.4 Individual differences in the size of nests
or function of the nervous system. For example, it is pos-
built by house mice in the laboratory. The genetic basis
of these differences in nest-building behavior has been sible to selectively breed mice to be very active and to
demonstrated by artificial selection experiments. explore an open test arena or to be less active. Such selec-
Whereas the mouse on the left was selected to build a tion results in specific differences in brain structure
small nest, the one on the right was selected to build between the high- and low-activity lines: Specific regions
a large nest. of the hippocampus are more developed in the mice that
 Methods of Behavioral Genetics 35

55
FIGURE 3.5 Artificial selection for large or
50
small nest size in house mice. The nests of
45 mice in the original population consisted
High 1 High 2
Control 1 Control 2 of 13 to 18 grams of cotton. Individuals
40
Grams of cotton collected

Low 1 Low 2 who built the largest nests were bred with
35 others who built large nests to create a
high nest-building line. Individuals who
30 built the smallest nests were bred with
25
others who built small nests to create a
low nest-building line. A control line was
20 developed by randomly mating individuals
of each generation. After 15 generations,
15
the nests of the high line were an average
10 of 8 times larger than those of the low
line—40 grams and 5 grams, respectively.
5 The nests of mice in the control line were
0 roughly the same size as those built by
S0 S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 mice in the initial population (15 grams).
Generation of selection (Data from Lynch 1980.)

are selected for high activity (Hausheer-Zarmakupi et al. surprising because odors are important in the daily life
1996). As the field of behavioral genomics expands, we of fruit flies for locating both food and appropriate
will be able to discover the actual genes and proteins mates. William Quinn and his colleagues (1974) demon-
responsible for differences such as these. strated this avoidance conditioning by shocking a group
of flies for 15 seconds in the presence of one odor but
not in the presence of a second odor. Then they pre-
INDUCING MUTATIONS AND sented the odors, one at a time, to the flies without
SCREENING FOR CHANGE shocking them to see how many would avoid each odor.
Most normal flies avoid only the odor connected with
IN BEHAVIOR
the electric shock, an association that lasts for three to
A mutation, a change in a gene’s instructions for pro- six hours (Dudai et al. 1976).
ducing a protein, can be induced by agents that change Two mutants, dunce1 and dunce2, were isolated by
the DNA bases. When organisms are exposed to muta- exposing a population of fruit flies to a chemical known
genic agents, some of them become behaviorally aber- to cause mutations and then screening the flies based on
rant and can be separated from the population of normal their learning ability. The dunce mutations are alleles, or
individuals on this basis. Appropriate genetic crosses can alternate forms, of a single gene on the X chromosome.
then determine whether the behavioral change is caused In contrast to normal flies, dunces fail to learn to avoid
by an alteration in a single gene. Even a small change odors associated with shock when they are taught using
may result in a difference in a specific aspect of an Quinn’s experimental design (Dudai et al. 1976). Even
anatomical structure or a physiological process that larval dunces are deficient in olfactory learning (Aceves-
mediates a behavior. Identifying the anatomical or phys- Pina and Quinn 1979).
iological differences between mutant and normal indi- Why don’t dunce fruit flies learn as well as normal
viduals brings us closer to understanding how genes can flies? The first guess—that the sensory system was defec-
influence behavior. tive so that the dunce flies could not detect either the
Studies on learning in the fruit fly have helped fill odors or the shock—was incorrect. Experiments revealed
in a few of the missing links between genes and behav- that the mutants are able to detect both (Dudai et al.
ior. You have probably never met a fruit fly with remark- 1976). In spite of this ability, mutant flies are unable to
able intelligence, but some mutant strains are so poor at remember the association between the shock and an
olfactory learning that they have earned the epithet odor.
dunce. Before considering the deficiency caused by the Apparently, the dunce mutants have a problem with
dunce mutation, olfactory learning in normal fruit flies the early stages of memory formation. If they are
and the way of demonstrating it should be described. shocked in the presence of an odor without subsequent
Normal fruit flies can associate an odor with an unpleas- exposure to a second odor, they do in fact associate the
ant event, such as an electric shock, and learn to avoid odor with the aversive stimulus and avoid it if tested
that odor if it is encountered again. This should not be immediately after training, but the association fades
36 Chapter 3 / Genetic Analysis of Behavior

quickly. The association between the shock and an odor Rutabaga is deficient here Dunce is deficient here
is short-lived, and the experience with a control odor
ATP
during the interval between the training and the test of adenylyl-cyclase
learning seems to eliminate associations that may have
formed. For these two reasons, we conclude that an early
stage of memory formation is defective in the dunce cAMP
mutants (Dudai 1979).
What does the dunce gene do? It codes for a form of Binds to cAMP phosphodiesterase
5⬘ AMP
an enzyme called cyclic AMP (adenosine monophos- and activates
phate) phosphodiesterase. This enzyme is important
because it breaks down cyclic AMP (cAMP), which is a
PKA (protein kinase A)
mediator of many biochemical processes in different
types of cells. Thus, we see that the dunce mutation Activates
causes a reduced level of cAMP phosphodiesterase and,
CREB gene
therefore, an increased level of cAMP. In addition, the
mutation impairs an early stage of memory formation in
olfactory learning. This should lead you to suspect that
the enzyme or cAMP might play a role in the underly- CREB protein turns on genes that
cause changes in the structure and
ing physiological process involved in this type of mem- function of nerve cells that underlay memory
ory formation. This idea has been tested by inhibiting
cAMP phosphodiesterase in normal flies and testing the FIGURE 3.6 A summary of the molecular events that
olfactory learning abilities. When treated in this way, accompany memory formation in fruit flies. Single-
normal flies learn no better than dunces do (Byers et al. gene mutations that result in poor olfactory learning
1981). Thus, the behavior of dunce flies suggests that were helpful in uncovering many of the details of
memory formation. The dunce mutant has a defective
cAMP has a role in learning and memory.
form of cAMP phosphodiesterase, an enzyme that
Another single-gene mutation in fruit flies,
breaks down cAMP. Rutabaga has a defective form of
rutabaga, also causes poor learning and memory. This adenylyl cyclase, an enzyme that forms cAMP from
mutation has filled in some of the details of the con- ATP. Thus, both mutations affect the levels of cAMP
nection between cAMP and memory formation. The within nerve cells. The cAMP binds to another
level of cAMP within a cell actually depends on two enzyme, PKA (protein kinase A), and activates it.
enzymes. As we’ve seen, cAMP phosphodiesterase Active PKA then turns on the CREB gene, whose
breaks down cAMP, lowering its concentration. In con- protein regulates the activity of other genes so that
trast, another enzyme, called adenylyl cyclase, raises new connections can be made among nerve cells.
cAMP levels by causing the formation of cAMP from These connections are responsible, in part, for
ATP (adenosine triphosphate) (Figure 3.6). Rutabaga long-term memory.
mutants have a defective form of adenylyl cyclase (Levin
et al. 1992). As a result, rutabaga’s adenylyl cyclase is not
activated by the stimulus involved in learning in the way
it would be in a normal fly. Here, the learning stimulus that, in turn, activates other genes. It is not cAMP itselt
doesn’t cause cAMP levels to rise. Thus, dunce and that produces the behavior, but rather cAMP is part of
rutabaga have opposite effects on the level of cAMP the regulatory pathway that affects the behavior. These
within a cell. other genes affected by the regulatory hierarcy then
As is often the case in science, the more we learn, control the growth of connections between brain cells,
the more questions we have. In this case, the dunce and changes in the nervous system that are responsible for
rutabaga mutants prompt us to wonder how cAMP is memory (Davis et al. 1995).
involved in memory formation. Figure 3.6 summarizes
a pathway involving cAMP that is thought to be impor- FINDING NATURAL VARIANTS AND
tant in part of a regulatory pathway for memory for-
LOOKING FOR GENETIC DIFFERENCES
mation. Notice that cAMP binds to another enzyme,
protein kinase A (PKA), and activates it. All protein Most of us would not find foraging fruit fly larvae of
kinases work by adding a phosphate group to another great interest, unless the foraging was taking place in a
molecule, activating or inactivating the other molecule. fruit bowl on our kitchen table. But it is fascinating to
In this case, PKA activates the CREB gene, which codes many behavioral geneticists, because it helps us to
for the protein CREB (cAMP response binding protein) understand much more complicated behaviors. The
 Methods of Behavioral Genetics 37

interest began when Marla Sokolowski (1980) noticed pared these “hybrid” offspring to pure rovers and pure
two forms of feeding behavior in natural populations of sitters.
larval fruit flies (D. melanogaster). “Rover” larvae move Sokolowski began by crossing adult rovers with
around continually on their food and often leave their other rovers and adult sitters with other sitters, creating
food source to look for another. In contrast, “sitter” lar- parental strains of sitters and rovers. The responses of
vae travel only short distances and tend to remain on a male larvae resulting from these crosses are shown in
food source. In fact, when these larvae were brought Figure 3.7a. Notice that the frequency distribution of
into the laboratory and allowed to feed on yeast paste path lengths of rovers is easily distinguished from that
in a petri dish, the distances traveled by rovers were of sitters. Next, Sokolowski crossed adults of sitter lar-
nearly four times longer than those of sitters. When vae with adults of rover larvae. The path lengths of the
food is not present, all flies move rapidly, as rovers do. first generation (F1) offspring of these crosses are shown
Thus, the difference in foraging styles is not because sit- in Figure 3.7b. Nearly all the larvae were rovers. When
ters are energy-deficient or sick (Pereira and Sokolowski the F1 adults were crossed with one another, the result-
1993; Sokolowski et al. 1997). ing second-generation (F2) larvae consisted of both
rovers and sitters in a ratio of three rovers to one sitter
(Figure 3.7c). This ratio of offspring is what would be
HYBRIDIZATION
expected if the variation in foraging strategies in a nat-
Sokolowski immediately suspected that a behavior with ural population could be explained by variation in a
two distinct forms would be genetically controlled. To single gene, dubbed foraging (for), with two alleles (alter-
investigate the genetic basis of these rover and sitter for- native forms). The two alleles of for are forR (rover) and
aging strategies, Sokolowski performed a series of hy- forS (sitter), and forR is dominant to forS. Thus, rover flies
bridization experiments using adult rovers and adult are forRforR or forRforS and sitters are forSforS (de Belle and
sitters in which she mated rovers to sitters, and she com- Sokolowski 1987).

a Parental strain Parental strain FIGURE 3.7 The results of mating experi-
sitters rovers ments on fruit flies showing “rover” or
7.59 cm “sitter” foraging strategies as larvae. Rover
Sitter Rover larvae forage over significantly longer
Number of individuals

Number of individuals

10 7 distances than do sitter larvae. (a) The

8 6 parental strains were created by crossing
5 Hypothesis: rovers with rovers or sitters with sitters.
6 4 F–rover is dominant
(b) The frequency distributions show for-
4 3 f –sitter is recessive
2 aging path distances of male (F1) larvae of
2 crosses between adults of the parental
1
0 0 rover strain and adults of the parental sit-
0 0
Path length (cm) Path length (cm) ter strain. Almost all the resulting larvae
were rovers. (c) The frequency distribu-
Number of individuals

b tions show foraging path distances of male

8 (F2) larvae resulting from crosses between
6 F1 F F F1 flies. The results are consistent with
f Ff Ff the hypothesis that the trait is controlled
4
f Ff Ff by a single gene, called foraging (for), and
2
All Rovers that the rover allele is dominant to the
0
0 sitter allele. Punnett squares of the crosses
Path length (cm) are shown on the right. (Data modified
from de Belle and Sokolowski 1987.)
Number of individuals

c
16
F2 F f
12
F FF Ff
8
f Ff ff
4
3 Rovers: 1 Sitter
0
0 4 8 12 16 20 24 28 32
Path length (cm)
38 Chapter 3 / Genetic Analysis of Behavior

FIGURE 3.8 The candidate gene

Behavior of interest
approach for identifying genes underly-
ing a behavior of interest nominates
genes based on their location or their
role in other organisms.

Nomination of candidate gene based on:

• Literature search
• Genome data
• Gene known to have a similar role in other organisms
• Microarray analysis

Alter gene expression Clone and sequence gene

• Knock in genes
• Knock out genes
• RNA interference Compare with Look for differences
other organisms in sequence

Expression analysis

between species. Thus, once researchers know the

THE FORAGING GENE AS AN genetic basis of a trait in one species, they have a good
EXAMPLE OF BEHAVIORAL hint about the genetic basis of the same trait in other
GENETICS IN ACTION species. For example, after the links between the for
gene, PKG, and fruit fly foraging behavior were identi-
We will continue with the story of the quest to understand fied, researchers began to suspect that the for gene might
the role of genes and environment in fruit fly foraging. play a role in food-related behaviors in other organisms.
This story will illustrate some basic principles of behav- In other words, for became a candidate gene for forag-
ioral genetics and some of the techniques used to look at ing behavior. The nomination of a candidate gene may
the role of genes in behavior. be based on a search of the literature for genes known
to be involved in producing a similar behavior in another
CANDIDATE GENES organism or by comparing the sequence of the gene to
the sequences of genes in other organisms using
A series of breeding experiments showed that for is genome-sequencing data (Figure 3.8).
located on Drosophila chromosome 2 (de Belle et al. In this case, a search of databases of gene sequences
1989) close to that of a gene for an enzyme PKG (cyclic revealed genes with similar sequences and function as the
GMP1-dependent kinase), which was already known to fruit fly for gene in three other organisms: the honeybee
be important in signaling pathways within the cell (Apis mellifera), red harvester ants (Pogonomyrmex barbatus),
(reviewed in Barinaga 1994). Next, the gene was cloned and the nematode Caenorhabditis elegans (Figure 3.9). Slight
(many identical copies of it were created) and was shown differences in the sequence of the gene have arisen during
to be identical to a known Drosophila gene (dg2) that pro- evolution, but the versions of for in all these species affect
duces PKG. Indeed, the heads of adult rover flies had the level of PKG (Table 3.1). The genes in these organ-
significantly more of the enzyme PKG than did the isms are orthologues, meaning that they descended from
heads of adult sitter flies. The function of for and the dif- the same ancestral gene and have the same function. As we
ference in the level of PKG in rover and sitter flies are will see, however, for activity and for-PKG affects forag-
consistent with the idea that PKG causes the difference ing through different mechanisms in these four species and
in behavior, and is therefore part of the regulatory hier- may be regulated in different ways (Tan and Tang 2006).
archy affecting foraging behavior of these flies (Osborne
et al. 1997).
Over the course of evolution, the DNA sequences Honeybees
that influence a particular behavior tend to be conserved Fruit flies forage to satisfy their own hunger. In contrast,
honeybee workers forage in order to bring food back to
1
Guanosine monophosphate. their colony, and their own hunger is not lessened by
 The Foraging Gene as an Example of Behavioral Genetics in Action 39

emerges from her brood cell, her responsibilities are jan-

itorial—she cleans the cells. Next, she begins taking on
the duties of a “nurse,” which include caring for and
feeding the brood. When she is about 2 or 3 weeks old,
a Honeybee she ventures out of the hive and begins to collect pollen
or nectar in her new role as a forager.
Nursing Foraging The age-related switch from nurse to forager is asso-
ciated with an increase in the activity of the for gene in
the brains of honeybees. The level of for mRNA indi-
cates for gene activity, that is, how much the gene is
b Red harvester being expressed. The greater the for gene activity, the
ant greater the for mRNA level. Recall that for mRNA
encodes PKG. As you can see in Figure 3.10, levels of
for mRNA in the brain are higher in foragers than in
Foraging Nursing newly
nurses (Ben-Shahar et al. 2002).
emerged brood Is this conclusive evidence that for gene activity
underlies the switch from nurse to forager? No. Age is
another variable that differs between nurses and foragers.
c Nematode To rule out age as the cause of the change in colony
responsibilities, Yehuda Ben-Shahar and his colleagues
(2002) manipulated colonies so that all members were
Solitary Social initially one day old. The absence of foragers caused
some members of the colony to switch to foraging as
much as two weeks sooner than usual. The nurse bees
d Fruit fly and the forager bees were the same age, but the brains
of foragers were still higher in for mRNA. Thus, the
switch in behavior from nurse to forager is due to the for
Sitter Rover
gene activity and is not an age-related progression.
FIGURE 3.9 When the foraging gene is expressed, PKG
levels increase, changing food-related behavior in
honeybees, red harvester ants, nematodes (C. elegans), Harvester Ants
and fruit flies. (Modified from Tan and Tang 2006.)
Still another variation on exactly how for affects behav-
ior comes from red harvester ants. Like bees, red har-
foraging. Nonetheless, the same for gene regulates for- vester ants live in large colonies with 10,000 to 12,000
aging in both organisms, albeit in very different ways. workers. Some workers perform tasks inside the colony,
Unlike fly larvae, whose main job in life is to eat, the such as caring for newly emerged ants, while others for-
duties assigned to a honeybee worker depend largely on age outside the colony. The number of workers changes
her age. For the first few days after a honeybee worker with environmental conditions. As we saw in honeybees,

TABLE 3.1 Orthologues of the Foraging Gene

Gene Organism Behavior Inheritance

for Fruit fly Natural variation in an Allelic variation: rover (forR) is

(Drosophila melanogaster) individual fly’s foraging dominant to forS (Sokolowski
behavior; rovers have longer et al. 1997)
foraging paths than sitters
egl-4 A nematode worm Solitary to social feeding Mutation in egl-4 causes soli-
(Caenorhabditis elegans) tary feeding; expression of
egl-4 causes worms to
aggregate in social feeding
(Fugiwara et al. 2002)
amfor Honey bee (Apis mellifera) Change in caste from nurse Gene expression of amfor is
(within-hive duties) to forager lowest in nurses and highest
(out-of-hive duties) in foragers (Ben-Shahar et al.
2002)
40 Chapter 3 / Genetic Analysis of Behavior

a b Typical colonies

Colony 1 Colony 2 Colony 3 Nurses

Amfor 3 3 12 Foragers
10 Precocious foragers
2 2 8

mRNA abundance (arbitrary units)

6
1 1
4
2
EF1a

Single-cohort colonies

c Colony 1 Colony 2 Colony 3 Colony 4

5 2 2 3

4
2
3
1 1
2 1

Behavioral groups

FIGURE 3.10 The foraging gene becomes active as nurse honeybees leave the hive and become foragers. The
level of for mRNA, which encodes PKG, is higher in the brains of forager honeybees than in the brains of nurse
honeybees. (a) Northern blots of the brains of nurse and forager honeybees. The darkness of the band reflects
the level of for mRNA and, therefore, the degree of gene activity. (b) Histograms of arbitrary units of for mRNA
in the brains of nurses and foragers indicate greater for gene expression in foragers. (c) The activity of the for
gene in foragers is greater than that in nurse bees of the same age. (From Ben-Shahar et al. 2002.)

the harvester ant version of for varies in expression as didate gene underlies a behavior is to alter the gene’s
workers switch from within-colony duties to foraging. expression. This can be done by increasing gene activ-
However, in harvester ants the relationship between for ity by adding copies (knocking in) of the gene or by
expression and behavior is the opposite of the relation- decreasing gene activity by disabling (knocking out)
ship in honeybees: for expression is greater in the brains genes. Knocking in genes increases the amount of that
of workers within the colony than it is in the brains of gene’s protein product, causing a greater effect on
foragers (Ingram et al. 2005). behavior. Knocking out a gene eliminates the product
of the disabled gene. If the probability that the behav-
ior of interest is displayed is changed by altering the
Nematodes activity of the gene, it suggests that the gene is involved
Similar to the fruit fly, the nematode (roundworm) in producing that behavior (Fitzpatrick et al. 2005).
Caenorhabditis elegans has two forms of foraging: roam- The technique of knocking in was used to confirm
ing and dwelling. A roamer travels long distances with- the mechanisms by which the for gene alters the feeding
out stopping. In contrast, a dweller travels short strategy. By adding four extra copies of the for gene to the
distances and makes frequent stops. As in the fruit fly, nuclei of eggs from a sitter fruit fly, Sokolowski’s team
differences in the alleles of the nematode for gene result showed conclusively that the for gene affects the rover vs.
in roaming or dwelling during foraging (Fujiwara et al. sitter behavior through its effects on PKG activity. Recall
2002). However, unlike the fruit fly where roving that sitter flies normally have lower PKG activity than do
increases with for gene activity, in C. elegans roaming rover flies. When extra copies of for were added to sitter
decreases with for gene activity. fly egg nuclei, the resulting flies had PKG levels similar
to those of rover flies and their foraging behavior was
similar to that of rovers (Osborne et al. 1997). In nature,
LINKING A PROTEIN TO A TRAIT then, mutations in for reduce the amount of PKG pro-
So far, we have presented very strong circumstantial duced, causing the sitter foraging behavior.
evidence that foraging behavior is associated with the The technique of knocking out genes has con-
protein PKG. How can we confirm this? A particularly firmed that for may act differently depending on the
elegant technique to confirm whether a particular can- organism it is in. When the nematode for gene is
 The Foraging Gene as an Example of Behavioral Genetics in Action 41

Pln1 Pln2 Pln3 Pln4 FIGURE 3.11 Interactions between the

“pollen” QTLs (pln 1-4), the genes involved
in foraging and sucrose responsiveness in
honeybees. (Modified from Hunt et al. 2007.)
Sucrose responsiveness Foraging age Foraging behavior

knocked out, there is an increase in roaming behavior ciated with pollen-hoarding were not so easy to predict:
(Fujiwara et al. 2002). Note that this is the reverse of workers mature to foragers at a younger age, they are
what happens in the fruit fly, where roving increases more active after emerging, they are better at learning
with PKG activity. various tasks, and they are more responsive to sucrose
(Rüppell et al. 2006).
There are four genes known to underlie this collec-
LOCATING ALL THE GENES
tion of behaviors associated with foraging. These
ASSOCIATED WITH A TRAIT
“pollen” QTLs are named pln-1, pln-2, pln-3, and pln-4.
The expression of most behaviors does not fall into dis- Pln-1 and pln-2, which are associated with the size of the
crete categories, such as rover or sitter. Instead, the pollen loads collected by workers, are responsible for
expression varies continuously, similar to the way that 59% of the variation in the amount of pollen stored in
height varies throughout a population. The reason for honeybee colonies (Rüppell et al. 2004). Pln-2 and
continuous variation is that many genes play a role in pln-3 influence the bee’s ability to determine the con-
shaping behavior. We describe a trait (behavior in this centration of sugar in the nectar collected. Pln-4 is either
case) as a quantitative trait2 when many genes are the honeybee for or a gene located close to it.
involved. A region of DNA associated with a particular Interactions of these genes influence the age at the onset
quantitative trait is called a quantitative trait locus of foraging and sucrose responsiveness, which affect the
(QTL). Thus, there are many quantitative trait loci choice of food source (Figure 3.11). Thus, each of these
(QTLs) associated with any complex behavior. QTL genes has a very specific role to play in creating the
analysis tells us whether the trait is determined by a few behavior of pollen collection.
genes, each having a large effect, or many genes, each hav-
ing a small effect. Thus a goal of QTL research is to iden-
tify the many genes underlying a trait and the extent to MICROARRAY ANALYSIS
which each of them alters the trait (Plomin et al. 2003).
Furthermore, once a QTL has been identified, that region Another innovative technique that allows us to simulta-
of DNA can be sequenced and compared to a database of neously investigate the effect of many genes on a behav-
DNA containing genes whose functions are known. This ior is microarray analysis. As you now know, a number
information can help identify candidate genes, that is, of genes may be present in an animal, but not all of them
genes that might be involved in the trait in question. are expressed (active) during a particular behavior.
We have seen that a honeybee worker’s change from Microarray analysis enables us to create a gene expres-
within-nest duties, such as caring for the brood, to for- sion profile by monitoring the expression of hundreds or
aging is caused by an increase in the activity of the for- even thousands of genes at once. A DNA microarray
aging gene, which causes an increase in PKG, an consists of thousands of DNA sequences stamped onto
important cell-signaling molecule. Once a worker has a solid surface, such as a glass slide. Molecular tags are
become a forager, she has many decisions to make. The used to identify mRNA produced by each of the genes.
first decision concerns the type of food. Bees forage for The greater the mRNA production, the more active the
pollen, nectar, or both. Many factors influence this gene. In this way, it is possible to compare gene activity
choice, including genetic makeup, brood status, colony in different tissues or in the same tissue at different
resources, and availability of food resources (Rüppell et times. The genes that are active only during a particular
al. 2004). Strains of bees were bred and selected on the behavior may play a role in producing that behavior
basis of the amount of pollen the workers stored in the (Hofmann 2003).
colony. The workers in colonies that store large amounts Charles Whitfield and his colleagues (2003) looked
of pollen tend to specialize more in pollen (rather than at patterns of gene expression of about 5500 genes in
nectar) collection and carry heavier loads of pollen than nurse bees (5 to 9 days old) and forager bees (28 to 32
do bees from colonies that do not hoard pollen. Those days old) from a typical colony. Nurses and foragers
traits might be expected, but other behavioral traits asso- showed significant differences in the expression of 2200
(39%) of the 5500 genes tested (Figure 3.12). Although
the expression levels of many genes differ, the magni-
2
Quantitative traits are also called polygenic traits. tude of the difference is not great. Thus, modest
42 Chapter 3 / Genetic Analysis of Behavior

FIGURE 3.12 The gene expression profile in the brain of a honeybee nurse differs from that of
the forager brain (left), even if the nurse and forager are the same age (right). Each bar shows
the activity of a particular gene, named on the left. The intensity of activity is indicated by color.
(From Whitfield et al. 2003.)

changes in gene activity are correlated with large behav- ioral genetics, we will include both the for gene and
ioral differences. genes that influence other behaviors, including social
DNA microanalysis was also used to confirm the behaviors.
earlier study that looked only at for gene activity—the
pattern of gene activity is correlated with the behavioral
chores performed by nurses or foragers, not with their ONE GENE USUALLY AFFECTS
age. Recall that in colonies created so that they are ini- SEVERAL TRAITS
tially composed of only young bees, some workers will PKG is an important cell-signaling molecule in the reg-
mature into foragers as much as two weeks earlier than ulatory hierarchy of many traits, so it should not be sur-
usual and others will remain nurses for much longer prising that the for gene that codes for it influences
than usual. Thus, there are precocious foragers and behaviors other than locomotion during foraging. For
overage nurses. The gene expression patterns in the example, for influences how fruit flies respond to sugar.
brains of age-matched nurses and precocious foragers When stimulated with a drop of sucrose, a fruit fly
were compared, as were the gene expression patterns extends its proboscis to feed. Rovers do this more read-
of age-matched foragers and overage nurses. Those ily than do sitters. Furthermore, rover flies continue to
four groups were analyzed along with nurses and respond to repeated sucrose stimulation for a longer time
foragers from typical colonies. The gene expression than sitter flies do. In nature, adult rovers move further
patterns reflect the bee’s behavior, not its age. away from the food source after feeding than adult sit-
Furthermore, the team was able to predict with 92% ters. Perhaps rovers and sitters differ in their evaluation
to 95% accuracy whether an individual bee was a nurse of the food source. Rovers may continue searching for
or a forager by its pattern of gene expression alone food because they remain responsive to food; sitters may
(Whitfield et al. 2003). stay put because they lose responsiveness to food. As a
result, rover foraging behavior may have an advantage
in places where food sources are scattered (Scheiner et
IMPORTANT PRINCIPLES al. 2004).
OF BEHAVIORAL GENETICS Rovers and sitters also differ in the loss of respon-
siveness to repeated electrical stimulation of the brain.
The for gene has been useful to illustrate the techniques When a stimulus occurs repeatedly without conse-
used to discover the relationship between genes and quence, animals usually lose responsiveness to that stim-
behavior. Now, as we turn to broader issues of behav- ulus. This loss of responsiveness, called habituation, is
 Important Principles of Behavioral Genetics 43

discussed in Chapter 5. In fruit flies, electrical brain stim- vitellogenin naturally declines with age, removing the
ulation activates the giant fiber jump-and-flight escape inhibition of JH and allowing the transition to forager.
response. When fruit fly brains are electrically stimulated In addition, high levels of vitellogenin earlier in life
repeatedly, sitter fruit flies habituate more rapidly than prime foragers to specialize in collecting pollen.
do rovers. Thus, as we saw with responsiveness to sugar, Vitellogenin also slows aging by protecting against free
rovers remain responsive to electrical stimulation longer radicals.
than sitters do (Engel and Hoy 2000). Similar patterns Mindy Nelson and her co-workers (2007) used a
of responsiveness to stimuli affecting different sensory technique called gene knockdown to demonstrate that
modalities indicate that the for gene affects behavior at the honeybee vitellogenin gene regulates social organi-
the level of the brain, not at the sensory level (Scheiner zation through multiple (pleiotropic) effects on the
et al. 2004). onset of foraging and foraging specialization on nectar
In honeybees, the for gene is regulated by vitel- or pollen (Figure 3.13a). In this case, gene activity was
logenin, the protein product of the vitellogenin gene, knocked down using RNA interference (RNAi), which
which has multiple effects on social organization. is double-stranded RNA that is complementary to the
Vitellogenin is responsible for the onset of foraging mRNA from the gene of interest. When introduced
because it inhibits juvenile hormone (JH), a hormone into a cell, the RNAi binds to a protein to form a com-
that stimulates the transition to forager. The level of plex that destroys the particular mRNA of interest. In

a
Vitellogenin gene

Inhibit juvenile hormone High level of vitellogenin Slows aging by inhibiting

(JH stimulates the transition to forager) early in life primes worker to collect pollen free radicals

b 7 Theoretical trajectory
of a pollen forager
6
Theoretical trajectory
Vitellogenin titer (μg/μl)

of a nectar forager
5
Vitellogenin RNAi
Pollen threshold phenotype
4
Treatment with dsRNA
3
Foraging threshold

0
0 5 10 15 20 25 30
Age (days)

FIGURE 3.13 The vitellogenin gene has several effects. (a) The vitellogenin gene affects the timing of the
transition to foraging, foraging specialization, and aging. (b) A single gene, vitellogenin, coordinates the transi-
tion from nurse to forager and primes the forgers to collect either nectar or pollen. The activity of the vitel-
logenin gene was lowered using RNA interference, which resulted in very low vitellogenin production. If
vitellogenin levels are above the foraging threshold, the change from nurse to forager is suppressed. The level
of vitellogenin decreases with age, and when it drops below the foraging threshold, the likelihood of the onset
of forging increases. The level of vitellogenin before it drops below the foraging threshold influences whether
the forager will specialize in collecting pollen or nectar. If the vitellogenin level was high, the forager special-
ized in pollen collection. When vitellogenin production was knocked down using RNAi, the workers began
foraging at an earlier age than usual and collected pollen. (From Nelson et al. 2007.)
44 Chapter 3 / Genetic Analysis of Behavior

this experiment, RNAi lowered the mRNA produced

by the vitellogenin gene and lowered vitellogenin levels.
Worker honeybees whose vitellogenin levels were low- 1
ered by RNAi switched from nurse to forager at an ear- 2
lier age, preferred nectar as a food source, and lived
shorter lives (Figure 3.13b). Many genes, like vitel- 3
logenin, affect many traits. This has implications for the
6
evolution of behavior, as a variety of traits may share
an underlying genetic basis and may thus not evolve 4
independently from one another. We’ll return to this
point in the next chapter. 5

GENES WORK IN INTERACTING FIGURE 3.14 Hypothetical diagram of overlapping genetic

NETWORKS modules. Each circle represents a genetic module that
controls a specific behavior. Each genetic module consists
While genes such as for have strong influence on behav- of interacting networks of genes. The arrows represent
ior, they do not act alone. Microarray analysis has taught mRNA transcripts that can affect certain genetic mod-
us that genes work in functional networks and that those ules. In this scheme, gene module 1 affects gene modules
networks interact to form genetic modules, genes that 2 and 3 in addition to itself. Gene module 3 affects gene
work together closely, that are responsible for a certain modules 5 and 6 in addition to itself. As a result, gene
behavior. The genetic modules have regions that over- module 1 can affect gene module 5 through its effects on
lap in some of their genes, allowing the modules to work gene module 3. (From Amholt 2004.)
together and to regulate one another. Figure 3.14 shows
a hypothetical diagram of overlapping genetic modules.
Each circle represents a genetic module that controls a hormone. If it does, the hormone binds to the receptor,
specific behavior. Each genetic module consists of inter- and then the cell responds the way that type of cell is
acting networks of genes. In this scheme, gene module programmed to respond (see Chapter 7).
1 affects gene modules 2 and 3 in addition to itself. Gene For social bonding to occur, the estrogen and oxy-
module 3 affects gene modules 5 and 6 in addition to tocin receptors must be located on cells in specific
itself. As a result, gene module 1 can affect gene mod- regions of the brain. The estrogen receptors are
ule 5 through its effects on gene module 3. Thus, a important because estrogen plays a role in regulating
change in a single gene can cause a change that can have the levels of both oxytocin and oxytocin receptors.
a ripple effect that causes changes in interacting gene Estrogen increases the level of expression of the oxy-
modules and have effects on many related traits or tocin receptor genes in the hypothalamus. Oxytocin
behaviors (Amholt 2004). This property allows many and its receptors are needed to process olfactory infor-
behaviors and parts of an animal’s physiology to be co- mation in the amygdala of the brain, which is essen-
regulated in important ways. tial for social recognition.
We see the importance of each gene in a network in Elena Choleris and her co-workers (2004) demon-
the interactions among four genes that affect the for- strated that interactions of all four genes are needed for
mation of social bonds in rodents. Social bonding social bonding to occur. They created knockout mice
requires that an individual be motivated to approach and that lacked one of the four components. If even one of
engage another individual and that the two animals are the genes was disabled, the mutant mice couldn’t distin-
able to identify one another through remembered social guish a familiar mouse from a stranger. Based on this evi-
cues. Then, under appropriate conditions a social bond dence, Choleris and her colleagues have proposed that
can form (Lim and Young 2006). a small network involving these four genes underlies the
Social recognition between female mice depends on olfactory basis of social recognition (Figure 3.15). In this
interactions among the protein products of four genes: model, olfactory cues are detected by the usual sensory
estrogen receptor α, estrogen receptor β, oxytocin, and receptors, and the information is sent to the amygdala,
an oxytocin receptor. Estrogen and oxytocin are hor- a reward region of the brain. Estrogens from the ovaries
mones, which are chemical messengers that are released bind to estrogen receptors β in a region of the hypo-
into the bloodstream and carried to cells throughout the thalamus (known as the paraventricular nucleus).
body. (Hormones and behavior are discussed in more Estrogen is a steroid hormone that often acts by bind-
detail in Chapter 7.) Some hormones, including oxy- ing to a receptor, entering a cell’s nucleus, and turning
tocin, are produced by neurons. For a cell to respond on certain genes. In this case, estrogen turns on the oxy-
to a hormone, the cell must have a receptor for that tocin gene. The oxytocin reaches the amygdala through
 Behavioral Variation and Genes 45

FIGURE 3.15 A diagram of a four-gene

Individual-specific micronet underlying social recognition.
olfactory cues E = estrogen Estrogen produced by the ovaries binds
to estrogen receptor ␣ in the amygdala of
ER = estrogen receptor
the brain, regulating the expression of
Non-volatile Volatile OT = oxytocin the oxytocin receptor gene. Estrogens
also bind to estrogen receptor ␤ in the
OTR = oxytocin receptor
hypothalamus of the brain, which regu-
lates the expression of oxytocin. The
Main olfactory binding of oxytocin to its receptors
Vomeronasal bulb/system
organ underlies social recognition. (From
Hypothalamus Choleris et al. 2004.)
PVN and SON

OT
Accessory OT
olfactory ERβ
bulb/system OT

OTR E
Blood stream
ERα

Social
recognition
Ovaries:
estrogen (E)
production

axons from the hypothalamus. In the amygdala, oxytocin regions of DNA control gene activity and are, therefore,
binds to estrogen receptor α and turns on oxytocin responsible for the pattern of gene activity.
receptor genes. We see, then, that any disruption to the Gene regulation might vary over time, as the hon-
gene interactions in this module causes the entire system eybee for gene does. Recall that the transition from nurse
to fail. to forager occurs as the activity of for increases, thereby
raising the level of PKG (Ben-Shahar et al. 2002).
The behavior of other organisms may vary because
BEHAVIORAL VARIATION of differences in the tissues in which the gene is
AND GENES expressed. We see this effect in social bonding of voles.
Prairie voles are monogamous, meaning that they
As we have seen, genetically controlled behavior is not remain with a female after mating. In contrast, a male
necessarily fixed and stereotyped. Instead, it can be meadow vole is nonmonogamous; he loves them and
dynamic and responsive to the environment. What roles leaves them. Why?
might genes play in such variation in behavior? Male prairie voles respond to the hormone vaso-
Behavioral variability might be caused by differences pressin differently than male meadow voles because of
in the alleles present in an individual. Recall that the for- differences in the distribution of vasopressin receptors
aging style of fruit flies depends on which alleles of for (V1aR) in the brains of the two species (Figure 3.16).
are present. If forR is present, the larva or fly is a rover, As with all hormones, vasopressin must bind to a
but if both alleles are forS it would be a sitter. receptor to bring about its effects. Only cells with
Alternatively, behavioral variation might be caused vasopressin receptors can respond to vasopressin.
by differences in gene regulation. Microarray analysis Male prairie voles have many more vasopressin recep-
shows us the pattern of gene activity in a given tissue tors than do male meadow voles, and their receptors
under certain conditions. If the conditions change, so are concentrated in the “pleasure center,” a reward
does the pattern of gene activity. New genes will be system of the brain. In contrast, male meadow voles
turned on, and other genes will be turned off. A gene’s have few vasopressin receptors in the reward system of
level of expression can also be modified. Regulatory the brain (Young et al. 1999).
46 Chapter 3 / Genetic Analysis of Behavior

a c 80 e

60

Time spent huddling (min)

40

20

0
Prairie vole
b d 80 f
60

40

20

0
Meadow vole

FIGURE 3.16 Social bond formation and vasopressin receptor distribution. Whereas a (a) prairie vole
male forms a pair bond with his partner after mating, as shown by huddling, (b) meadow vole males
remain solitary. Histograms showing the amount of time a (c) male prairie vole and a (d) male
meadow vole spend huddling with a female after they have mated with her. Monogamous male
prairie voles spend time huddling with a female after he has mated with her. Promiscuous male
meadow voles spend little time with a female after mating. (e) Male prairie voles have many vaso-
pressin receptors (VP) in the reward centers of the brain (ventral pallidum), as compared to ( f ) the
number of vasopressin receptors in meadow voles. The difference in receptor distribution is thought
to be responsible for the difference in pair bonding in these species. Prairie voles are monogamous;
meadow voles are not. A promoter found in prairie voles but not meadow voles is responsible for the
differences in receptor distribution. (From Lim et al. 2004.)

After mating, equal amounts of vasopressin surge in Lim and her colleagues then demonstrated that this
the brains of males of both species. In the prairie vole, promoter is likely to be the cause for the difference in
the vasopressin binds to its receptors and stimulates the the number of and distribution of receptors, as well as
reward system. Afterward, the male prairie vole associ- the difference in partner preference in these species.
ates the odor of his female mate with the feeling of They injected the promoter attached to a harmless viral
reward caused by vasopressin, resulting in a preference carrier into a reward center (ventral pallium) of meadow
for the female mate over a strange female. Male meadow vole males and measured the length of time each male
voles have fewer vasopressin receptors in the brain’s spent huddling in close contact with the female after
reward system, so the vasopressin released during mat- mating. Compared with untreated meadow vole males,
ing does not stimulate the reward system, and no pair males who received the promoter spend significantly
bond forms. more time huddling with the female after mating.
Miranda Lim and her colleagues (Lim et al. 2004) Similarly, if that promoter sequence is inserted into male
hypothesized that genetic variation in the vasopressin mice, which are not closely related to voles and are never
receptor gene underlies the differences in the distribu- monogamous, the mice show an increased preference for
tion of vasopressin receptors in the brains of monoga- a partner with whom they have mated, similar to the
mous and polygamous voles. They tested the hypothesis effect on monogamous prairie vole.
by comparing the DNA sequences of the vasopressin
receptor gene in several vole species, and found that the
coding sequences are 99% identical. Thus the protein STOP AND THINK
produced by the gene is virtually the same in every vole Before identifying the vasopressin receptor distribution as
species tested. the cause of the difference in social bonding in monoga-
Monogamous and polygamous voles do differ in the mous prairie voles and in nonmonogamous montane
promoter region of the vasopressin receptor gene. The voles, researchers injected vasopressin into the brains of
promoter is a region of DNA needed to turn on the both prairie voles and montane voles (Young et al. 1999).
Why was this step necessary? What outcome led the
gene. Monogamous species, such as the prairie vole, have
researchers to conclude that the distribution of receptors
an expanded region of repetitive DNA in the promoter
was the critical difference between the responses of these
region. The polygamous meadow vole lacks this region species?
in the promoter.
 Environmental Regulation of Gene Expression 47

Based on the important roles for vasopressin, oxy-

Heredity Environment
tocin, their receptors, and estrogen receptors in social
bonding in rodents, researchers suspect that these hor-
mones play a role in human social behavior, including
autism (Lim et al. 2005). Autism spectrum disorders are Genome
developmental disorders characterized by poor commu-
nication and social skills, usually accompanied by repet-
itive, stereotyped behavior patterns. One in every 166
children in the United States is diagnosed with an autism Variations in environment
spectrum disorder. One way to explore this hypothesis
that alterations in the oxytocin or vasopressin signaling
pathways are related to autism is to look for differences
in levels of these hormones among autistic children or Development Learning and memory
age-matched nonautistic children. In fact, oxytocin and
vasopressin levels are lower in autistic children (Lim and
Young 2006). One study has shown that oxytocin
Changes in gene Changes in nerve
administration boosts some social skills in autistic
expression cell activity
patients (Bartz and Hollander 2006). In addition, some
studies, taken together, suggest that variation in the vaso-
pressin receptor promoter gene in humans is associated
with autism (Insel 2006). Behavior

FIGURE 3.17 Genes and the environment interact to

produce behavior.
ENVIRONMENTAL REGULATION
OF GENE EXPRESSION can turn on immediate early genes, altering gene
Earlier in this chapter, we presented the idea of heri- expression in the brain to produce behavior adapted to
tability, which is the part of the variation in a trait that the environment.
can be attributed to genetics. The remainder of the vari-
ability is attributed to environmental influences. The
DOMINANCE RELATIONSHIPS
implication is that the effects of genes and the effects of
IN CICHLID FISH
the environment can be neatly divided into distinct
parcels. There are two types of males in the cichlid fish
We now know that it is not nature or nurture, but Astatotilapia3 burtoni. Dominant males are brightly col-
instead nature and nurture. The observed behavior is ored, aggressively defend territories, and have greater
the product of genes and environment acting on the reproductive success than subordinate males. Sub-
genome. A change in the pattern of gene expression ordinates are nonterritorial, have camouflage coloration,
is often the first quantifiable sign of those interactions and are less likely to be chosen as mates. The physical
of genes and environment affecting the behavior environment of the cichlid’s natural habitat fluctuates,
(Figure 3.17; Robinson 2004). causing frequent changes in dominance relationships. As
As we will see in Chapter 6, a nerve cell can a male rises in social status, he becomes sexually mature
respond to appropriate stimuli within milliseconds by and his growth rate slows.
generating an action potential. Thus, the immediate Interacting genes simultaneously affect physiology,
response to a stimulus is neural. However, that neural behavior, and social status in these fish. Gonadotropin-
response often initiates a broader response by activat- releasing hormone (GnRH), a small protein encoded by
ing immediate early genes. These genes become the GnRH gene, plays an important role in vertebrate
active within minutes to hours after a nerve cell is stim- reproductive physiology. Dominant males have larger
ulated and code for proteins that regulate the activity GnRH-releasing neurons in the preoptic area of the
of other genes. These other genes then produce pro- hypothalamus of the brain than do subordinate males.
teins that are important in the behavioral response, The larger size of these neurons is a result of increased
including turning on a gene for a specific hormone or activity of the GnRH gene.
its receptor or affecting the growth of nerve cells and In A. burtoni, the GnRH gene is regulated by social
nerve cell activity. We will consider dominance rela- stimuli resulting from dominance or submission. As a
tionships in cichlid fish and song learning in male
songbirds to better understand how the environment 3
Astatotilapia was formerly known as Haplochromis.
48 Chapter 3 / Genetic Analysis of Behavior

nonterritorial male begins to win aggressive encounters, lights came on, the remaining male perceived an oppor-
his GnRH gene increases activity, as does the number of tunity to rise in social status and within minutes his col-
GnRH receptors in the pituitary gland. As GnRH, the oration and behavior changed (Figure 3.18). The
protein product of the GnRH gene, increases, he moves changes in coloration and behavior were accompanied by
up in social rank and acquires a territory. GnRH triggers an increase in the activity of the erg-1 gene in the brain
the release of hormones that, in turn, trigger the release region rich in GnRH-releasing neurons. Erg-1 codes for
of sex hormones (gonadotropins) from the pituitary proteins that regulate the activity of other genes.
gland, which leads to development of the testes and the Among these genes is GnRH. As we have seen, the pro-
production of sperm. At the same time, GnRH brings tein product of GnRH acts in the pituitary gland to
about changes in coloration. His gray body becomes blue increase the levels of sex hormones. Other genes acti-
or yellow and a black bar appears across his face. The vated by erg-1 produce proteins that are thought to be
colorful transformation signals to both male and female important in the growth and activity of nerve cells, which
fish that he is now dominant (Hofmann 2006). underlie behavioral changes.
Body size is an important factor in a male’s ability
to attain dominance, and so it should not be surprising
SONG LEARNING IN MALE SONGBIRDS
that these characteristics are also affected by social sta-
tus. Subordinate males typically grow faster than domi- We see an example of interactions among genes, and the
nant males. Once dominance is attained, growth rate internal and external environment when a male zebra
slows and the male devotes more of his energy to repro- finch (Taeniopygia guttata) or canary (Serinus canaria) is
duction. Social stimuli associated with dominance bring first exposed to the song of its own species. Young males
about the changes in growth rate by increasing the learn to sing their species’ song by imitating the songs
expression of a gene encoding the hormone somato- of adult males. It is important that they learn the cor-
statin. Somatostatin inhibits the release of growth hor- rect song because, like other songbirds, these birds sing
mone (GH). Thus, once a male has established to defend territories and to attract mates.
dominance and becomes reproductively active (because But what changes in the nervous system underlie a
of GnRH), his somatostatin level increases, causing a male songbird’s learning his species’ song? Claudio
decrease in GH. Lower levels of GH lead to slower Mello, David Clayton, and their colleagues have
growth. If he then loses challenges by subordinate males uncovered some answers to this question. They sus-
and falls in dominance, becoming a nonterritorial male, pected that changes, if they occurred, would be found
somatostatin levels decrease and GH levels increase, in the bird’s forebrain because this region is important
causing his growth rate to increase again. The change in in auditory processing. They further hypothesized that
growth rate may reflect energy trade-offs between if gene activity changed, it would most likely be the
growth and reproduction. A subordinate male invests in activity of one of the so-called immediate-early genes,
growth; a dominant male invests in reproduction which become active within minutes after a nerve cell
(Hofmann 2006). is stimulated and code for proteins that regulate the
Opportunity to rise in social status may occur activity of other genes. In this example, the other genes
rapidly, as when a dominant male is plucked from the then produce proteins that are thought to be important
water by a predator. The immediate early gene, erg-1, in the growth of nerve cells and to affect nerve cell
orchestrates a subordinate male’s response to this oppor- activity, which, in turn, are involved in forming long-term
tunity to rise in social status. Although changes in a memories. Zenk4 is one of these genes. If zenk were
male’s aggressiveness and fertility may take a week or so turned on by exposure to the song, the levels of zenk
to occur, within minutes of the opportunity to rise in mRNA would be expected to rise. So, Mello and his
social status he becomes brilliantly colored and begins colleagues (1992) played a 45-minute tape recording to
to make threatening displays and chases other males. young male zebra finches or canaries. The recording
Sabrina Burmeister and her colleagues (2005) observed was that of the song of its own species, one of another
a group of four females, a dominant male, and a subor- species, or simply bursts of sound. They then measured
dinate male. Adjacent to the observation tank were tanks the level of zenk mRNA in the birds’ forebrains. It
containing large communities of fish. The fish in the turns out that zenk activity is increased greatly when
observation tank could see the other fish but could not males hear the song of their own species and much less
interact with them. The researchers used infrared night- so in response to the song of another species. Exposure
vision goggles to remove the resident dominant male to simple bursts of sound did not increase zenk activity
from the observation tank 1 hour before the lights were above that found in birds without any auditory stimu-
turned on. Because cichlids respond primarily to visual
cues, they do very little in the dark. However, when the 4
Zenk is a homologue of erg-1.
 Environmental Regulation of Gene Expression 49

a Subordinate b Dominant

FPO FPO

c Ascending males (previous day) d Ascending males

12 12
Behaviors per min

Behaviors per min

Male Latency
10 10
2:30
8 8 13:50
6 6 10:20
3:25
4 4
2 2
0 0
0-5 5-10 10-15 15-20 20-25 25-30 30-35 0-5 5-10 10-15 15-20 20-25 25-30 30-35
Min after lights on Min after lights on

e Stable subordinates f Stable dominant

12 12
Behaviors per min

Behaviors per min

Male Latency Male Latency

10 10
0:00 6:10
8 0:00 8 6:00
6 0:00 6 3:25

4 4
2 2
0 0
0-5 5-10 10-15 15-20 20-25 25-30 30-35 0-5 5-10 10-15 15-20 20-25 25-30 30-35
Min after lights on Min after lights on

FIGURE 3.18 The behavioral response to social opportunity. (a) Subordinate male appearance.
(b) Dominant male appearance. (c) The day before the opportunity to rise in social status,
subordinate males do not display dominant behaviors, such as chasing, threats, courtship
solicitations, and spawning site entries. (d) Within minutes of the perception that the
dominant male was gone, a subordinate male displays behavior associated with dominance.
(e) These changes do not occur in a subordinate male if the dominat male remains in the
tank. ( f ) The behavioral changes in a male rising in rank are similar to the behaviors
displayed by a stable dominant male. (From Burmeister et al. 2005.)

lation (Figure 3.19). The results of this experiment are the auditory forebrain regions of male zebra finches fol-
consistent with the idea that zenk is activated when a lowing repeated exposure to the song of the same zebra
male songbird hears the song of its species. finch. As expected, zenk mRNA increased during the
One function of bird song is to defend territories, first 30 minutes. However, in spite of continued stimu-
and it would be a waste of energy to continuously defend lation by the same song, zenk mRNA fell to baseline lev-
borders that are not contested. It should not be surpris- els. Furthermore, when the same male’s song was played
ing, then, that male songbirds can discriminate among after a full day of silence, there was no increase in zenk
the songs of other singers. Such discrimination requires activity. Nonetheless, zenk activity did increase follow-
a male to learn the characteristics of songs of specific ing exposure to the species song of another, unfamiliar
males, its neighbors, for instance. Mello and his col- zebra finch individual (Figure 3.20).
leagues (1995) hypothesized that zenk activity underlies Similar changes in zenk activity have been shown
the formation of long-term memories. Therefore, they in freely ranging song sparrows (Melospiza melodia).
predicted that zenk activity would increase in response When the species’ song was played through a loud-
to the songs of unfamiliar males but would not increase speaker within a male’s territory, he approached the
in response to the songs of familiar males. To test this speaker, searched for the intruder, and began actively
hypothesis, they measured the levels of zenk mRNA in singing in defense of his territory. Zenk activity in
50 Chapter 3 / Genetic Analysis of Behavior

10 Canary 5
Normalized density of

8
4

Zenk fold-induction in forebrain

zenk mRNA

4 3

2
2
0
Unstim Tone Different Own
n=4 n=2 species species
n=3 n=4 1

10 Zebra finch
0
Normalized density of

8 0/0 0/S1 0/S2 S1/S1 S1/S2 S2/S2 S2/S1

zenk mRNA

6 FIGURE 3.20 Zenk activity in auditory regions of the

forebrain of a male zebra finch drops following repeated
4 exposure to the song of one male, but it can be reacti-
vated by a new song. The birds were exposed to
2 repeated song stimulus for 2.5 hours, which was
immediately followed by a second “test” stimulus last-
0 ing 30 minutes. The levels of zenk mRNA in the birds’
Unstim Tone Different Own forebrains were then measured. Each dot represents
n=3 n=1 species species the amount of zenk mRNA in a single bird. The bars
n=3 n=4
represent the mean zenk mRNA level for the experi-
Auditory stimulus mental group. These results are consistent with the
FIGURE 3.19 Zenk gene activity in the forebrains of male hypothesis that zenk activity underlies the formation of
songbirds following exposure to songs of their own long-term memories. We would not expect zenk to be
species, songs of another species, bursts of tones that activated after the memories have been formed. The 0
are not song, or no auditory stimulation. The density of represents silence. S1 and S2 indicate specific songs of
zenk mRNA, which reflects gene activity, increases fol- two different individuals from another aviary. (Data
lowing exposure to birdsong. The increase following from Mello et al. 1995.)
exposure to the song of the listener’s own species is
significantly greater than that following the song of
another species. Zenk activity is not a simple response
to noise because it is not greater following nonsong genetic background) are present. In other words, differ-
tones than when the bird is not stimulated by any ent genetic backgrounds may respond differently to the
sound. (Data from Mello et al. 1992.) same mutation. The fruit fly rover versus sitter behav-
iors again provide an example. Recall that the rover
several brain auditory structures was higher in the
genotype is either forRforR or forRforS. The sitter geno-
males whose territories had been challenged by a
type, on the other hand, is forSforS. We have seen that as
loudspeaker intruder than in unstimulated controls.
cGMP increases, so do PKG levels and rover foraging
Thus, natural behaviors in the field affect gene activ-
behavior. Rovers have higher levels of PKG than do sit-
ity in much the same way they do in the laboratory
ters. Although it seemed logical to predict that a muta-
(Jarvis et al. 1997).
tion in a gene other than for that decreased cGMP would
We see, then, that the relationship between genes
also decrease PKG and locomotion during foraging, it
and behavior need not be static. Instead, the environ-
didn’t happen that way. Instead, both rover and sitter lar-
ment can continuously alter gene activity and, therefore,
vae with mutations in this gene had increased PKG and
modify an organism’s behavior.
increased locomotor activity.
Craig Riedl and his colleagues (2005) used microar-
THE IMPORTANCE OF GENETIC ray analysis to compare gene activity in rover and in sit-
BACKGROUND TO BEHAVIORAL ter flies after this mutation decreasing cGMP was
introduced. Some genes responded in the same manner
GENETICS
in both rovers and sitters. For example, this mutation
The expression of a gene of interest may depend on increased the activity of 37 genes and reduced the gene
which alleles of all the other genes in the genome (the activity of 14 other genes in both rovers and sitters.
 Environmental Regulation of Gene Expression 51

However, in a subset of nine commonly affected genes and become sneakers; that is, they steal matings.
the changes in gene activity were in the opposite direc- Sneaker males are ten times smaller than migratory
tion and depended on whether the fly was a rover or a males and are able to reproduce without leaving fresh-
sitter. Thus, the mutation had a different effect on the water. The early sexual maturity of sneaker males is
pattern of gene expression depending on whether the associated with larger gonads, smaller body size,
fly’s genotype was rover or sitter. changes in feeding and hormone levels, increased
response to female pheromones, and a male body scent
(Aubin-Horth et al. 2005a). Other males migrate out to
NETWORKS OF GENES ARE sea, where they mature, and return years later as large
RESPONSIVE TO THE ENVIRONMENT fish to breed in freshwater (Aubin-Horth et al. 2005b).
Complex behaviors are produced by genes interacting Thus, within the same population, males of the same
with the environment. For those genes, some of the age will either be immature, nonreproductive males that
“environmental” factors are hormones within the same will migrate out to sea before breeding or reproductive
organism. We will consider the ways in which hormones sneaker males.
affect behavior in Chapter 7, so we will not cover that A different subset of genes is expressed in the
here. When you read Chapter 7, keep in mind that an brains of sneaker males compared with those of
important way that certain hormones affect behavior is immature males. Microarray analysis of 2917 genes
by altering gene expression patterns. revealed that 15% (432) of the genes differed in activ-
ity in the brains of sneaker males compared with
immature males (Figure 3.21). Most of these genes
Rearing Conditions Affect Gene Networks have only small differences in activity but caused large
Males in the same population of Atlantic salmon (Salmo differences in morphology and behavior. Whereas the
salar) have alternative reproductive life histories. Some genes that showed increased activity in sneaker males
males reach sexual maturity in the first three years of life were associated with reproduction, the genes showing

a Growth Growth hormone related FIGURE 3.21 Proportion of

Circadian rhythms Transcription regulation genes with increased activity
Protein folding and maturation in (a) sneaker male and
Development (b) immature male Atlantic
Protein synthesis
Blood
salmon. Genes showing higher
activity in sneaker males tend
Immune function
Neural plasticity to be associated with repro-
Neuroprotection duction. Immature males had
Protein degradation higher activity in genes associ-
ated with body growth. (From
Feeding, reproduction, Aubin-Horth et al. 2005b.)
Metabolism other

Reproduction

Neural signalling

b Neurodegeneration Growth
Transcription regulation
Development
Neural Protein folding and maturation
plasticity

Protein synthesis
Metabolism

Neural
signalling
Blood
Reproduction
Immune function
Feeding, reproduction,
other Protein degradation Neuroprotection
52 Chapter 3 / Genetic Analysis of Behavior

increased activity in immature males were associated packaged the DNA molecule is. DNA is packaged with
with growth (Aubin-Horth et al. 2005b). proteins, mostly histones to form chromatin, which
Reproductive strategy is not the only factor affecting forms chromosomes. DNA methylation (adding a
gene expression. Environmental cues, such as population methyl group to the cytosine bases in DNA) turns off the
density, food availability, or temperature and lighting also activity of a gene by bringing in proteins that act to com-
influence gene activity. Nadia Aubin-Horth and her col- pact DNA into a tighter form, reducing access by regu-
leagues (2005a) compared gene expression patterns in the latory proteins that promote transcription. On the other
brains of sneaker males to the gene expression patterns in hand, histone acetylation makes the DNA less tightly
the brains of immature males as in the previous experi- coiled and gene expression easier. At one time,
ment. This time, however, she also considered rearing researchers believed that DNA methylation and acety-
conditions. Gene expression patterns of sneaker and lation occurred during development and permanently
immature males from a wild population were compared altered gene expression in certain tissues. We now know,
with those from laboratory-raised salmon. The fish that however, that these processes can be affected by the envi-
were raised in the lab were originally obtained from the ronment and that the pattern of DNA methylation is
same population as the wild-caught fish. Reproductive dynamic and changes over time.
strategy and rearing environment interact in interesting One of the first demonstrations of epigenetic con-
ways. The activity of one subset of genes differs between trol of the activity of a gene being nearly permanently
wild-caught fish and laboratory-raised fish whose ances- turned off by early life experiences come from Norway
tors came from the same wild population, regardless of rats (Rattus norwegicus). As described in the chapter open-
their reproductive strategy. Presumably these genes are ing, some rat mothers nurture their young more than
related to survival in different environments. other mothers do. Nurturing mothers extensively lick,
There is also a subset of genes that is more active in groom, and nurse their young. Their pups tolerate stress
sneaker males than in immature males, regardless of the better than pups of less attentive mothers. When the
rearing environment. As we saw in the previous experi- female pups are old enough to reproduce, they too
ment, the genes that are more active in sneaker males are become nurturing mothers.
associated with reproduction, and the genes that are Through what physiological mechanism does the
more active in immature males are associated with quality of maternal care affect adult stress responses and
growth. Interestingly, more than half the genes that var- maternal care? The frequent contact with pups increases
ied in activity between sneakers and immature males var- the expression of a gene for a glucocorticoid receptor in
ied in only one environment. In other words, the activity a region of the brain called the hippocampus.
of a subset of the genes in males with the same repro- Glucocorticoids are hormones released by the adrenal
ductive strategy varied depending on which environment glands in response to stress. A feedback relationship
they experienced. So, in fish raised in different environ- between the binding of glucocorticoids to their recep-
ments, different regulatory networks may lead to the tors and the production of glucocorticoids maintains the
same reproductive tactic. hormones at the appropriate level to cope with stress.
The binding of glucocorticoids to their receptors in the
hippocampus increases the activity of genes that dampen
EPIGENETICS AND BEHAVIORAL production of glucocorticoids. Thus, the increase in the
GENETICS number of glucocorticoid receptors helps the well-cared-
for pups to better regulate their response to stress, and
Now that we are familiar with some of the ways that so they are less fearful. In contrast, pups that did not
genes influence behavior and some of the ways that receive appropriate maternal care during the first week
genes and environment interact, let’s return to the story of life do not produce as many glucocorticoid receptors
that opened this chapter: the maternal care in rats influ- in the hippocampus, so they are anxious and fearful in
encing stress responses and the maternal behavior of rats response to stress. Furthermore, when the female pups
in future generations. The effects of maternal care on become mothers, they do not adequately lick or groom
later behavior of the pups occurs through epigenetics, a their own pups.
means of gene regulation that is only recently becoming These differences in maternal care and response
understood as a mechanism that plays a role in shaping to stress are due to changes in gene regulation
behavior. Epigenetics involves a stable alteration in gene brought about by changes in chromosome structure.
expression without changes in DNA sequence. In other At birth the promoter region of the glucocorticoid
words, it regulates how genes are expressed without receptors of all pups is demethylated, and at the end
changing the proteins they encode. We will consider the of the first day of life methylation of the promoter
roles that two epigenetic processes—DNA methylation region had begun in all pups. However, by the end of
and histone acetylation—play in behavior. These the first week of life, the promoter of a glucocorticoid
processes alter gene expression by affecting how tightly receptor in the pups of neglectful mothers was more
 Epigenetics and Behavioral Genetics 53

methylated than the same gene in pups of nurturing DNA methylation patterns can be affected by envi-
mothers. Licking and grooming increase the level of ronmental factors, cause disease, be transmitted through
serotonin, a neurotransmitter used in communication generations, and, potentially, influence evolution.
between certain neurons. In turn, serotonin results in Consider, for example, the effects that exposure to the
the removal of methyl groups and the addition of fungicide vinclozolin has on rats. Vinclozolin, which is
acetyl groups to histones. Both changes make DNA still used today on some agricultural crops, is an envi-
less tightly coiled and more easily expressed. The ronmental endocrine disrupter. It interferes with hor-
resulting increased expression of the glucocorticoid mone signaling of androgens (male sex hormones),
receptor increases the number of glucocorticoid recep- progestins, glucocorticoids, and mineralcorticoids. The
tors in the hippocampus and lessens the pups’ DNA methylation pattern differs between rats that were
responses to stress (Weaver et al. 2006). exposed to vinclozolin and rats that were not. The dif-
Recall that the quality of maternal care also influ- ferences in DNA methylation patterns are believed to be
ences how nurturing a female pup will be toward her the reason that brief exposure of a rat embryo to vin-
own young when she becomes an adult. The effect on clozolin increases the likelihood of diseases that do not
future mothering styles is brought about by epigenetic begin until adulthood. If male embryos are exposed to
changes in another gene; a particular estrogen receptor vinclozolin during the time that testes are forming, the
(ER α) in the preoptic area of the hypothalamus adult male has an increased risk of abnormal testes and
becomes methylated in female offspring of mothers who low fertility. The increased risk of adult-onset diseases
show little maternal care. As we saw earlier in this chap- and reduced fertility are passed through four genera-
ter, this estrogen receptor is necessary for the regulation tions, even if the offspring have never been in contact
of oxytocin receptor binding in the same region of the with vinclozolin (Anway et al. 2006). Vinclozolin causes
brain. Oxytocin is a hormone crucial for maternal care. these effects by changing the methylation pattern of 15
Thus, poor maternal care results in methylation of genes genes, many of which are regulatory regions of DNA,
for receptors of glucocorticoid hormone, a hormone that reducing the expression of these genes (Chang et al.
mediates stress responses, and genes for ER α, which is 2006). The acquired patterns of DNA methylation and
needed for proper response to oxytocin. In this way, chromatin condensation are then passed from generation
mother–pup interactions cause stable genetic changes to generation.
that influence the behavior of adult offspring Perhaps the most interesting aspect of vinclozolin
(Champagne et al. 2006). exposure is that it influences mate choice for at least
Although DNA methylation patterns are considered three generations. As we will see in Chapter 4, evolution
to be stable, some studies suggest that methylation can occurs when individuals with one set of alleles enjoy
be reversed in adulthood. Injection of trichostatin A, a more reproductive success than individuals with other
chemical that causes demethylation, into the brains of alleles. Therefore, it is in an individual’s best interest to
pups who had received poor maternal care made the choose a mate with “good genes.” Because vinclozolin
pups less fearful and better mothers. The effects of good increases the risk of adult diseases and lowers male fer-
maternal care can also be reversed. Methionine, which tility, it seems reasonable that a female should avoid mat-
is a chemical found in the diet, can alter DNA methyla- ing with a male who had been exposed to vinclozolin.
tion by donating a methyl group. When methionine is Regardless of whether a female has been exposed to vin-
infused into the brains of the adult offspring who had clozolin herself, when given equal access to both a male
received either good or poor maternal care, the degree with a history of exposure and a male that has had no
of DNA methylation increased. The adult offspring who exposure, she shows a significant preference for the
had received good maternal care now became anxious in unexposed male. In fact, she prefers an unexposed male
stressful situations and the females showed little mater- to the great-grandson of an exposed male! This female
nal care (Weaver et al. 2006). choice could affect evolution (Crews et al. 2007).
Epigenetic changes in DNA may also underlie
STOP AND THINK learning. Courtney Miller and David Sweatt (2007) con-
Experiments were performed in which the pups of mothers ditioned rats to be fearful of a location by giving them
who offered poor maternal care were cross-fostered to an electric shock when they were in the training cham-
mothers who offered good maternal care. Pups from good ber. If the rats froze when they were placed in the cham-
mothers were cross-fostered to mothers who offered poor ber at a later time, the researchers concluded that the rats
maternal care. As adults, the pups raised by good mothers had formed fearful memories of the chamber. Miller and
handled stress well, and the females were good mothers. Sweatt demonstrated that the pattern of DNA methyla-
The pups raised by poor mothers were fearful as adults, and
tion in the hippocampus of the brain changes when
the females were poor mothers (Francis et al. 1999). Why
memories are formed by treating the rats with drugs that
were these experiments necessary? What can you conclude
from them? prevent methylation between their training and testing.
Specifically, as fearful memories form, rapid methylation
54 Chapter 3 / Genetic Analysis of Behavior

(silencing) of a memory-suppressing gene (protein phos- Behavioral ecologists are interested in adaptive
phase 1, PP1) and demethylation of a memory-promoting behavior, which results from natural selection acting on
gene reelin take place. variations in phenotypes (traits) that have a genetic basis.
One take-home message is that we should look more In nature, individuals of the same species may have slight
closely at the role of epigenetics and disease. Epigenetics differences, perhaps only a single base pair, in their
is believed to play a role in human behavioral disorders, genes. These differences (polymorphisms) may produce
such as autism spectrum disorders, Rett syndrome (a subtle differences in phenotype on which natural selec-
developmental disorder that affects the nervous system), tion can act. Sequencing genes identifies these poly-
and fragile-X syndrome (an inherited form of mental morphisms. Microarray analysis would be useful in
impairment; Schanen 2006). identifying genes whose activity is correlated with
A second take-home message is that DNA is sensi- important ecological conditions (Feder and Mitchell-
tive to the environment, so what we eat and the chemi- Olds 2003). In this chapter, we have considered several
cals we are exposed to may influence our health by examples of ecological genomics: social recognition in
affecting our gene expression patterns. We are beginning rodents (Choleris et al. 2004), dominance relations in
to realize that the lifestyle of a person’s ancestors, and cichlid fish (Hofmann 2006; Trainor and Hofmann
not just the individual’s behavior, can influence health. 2006), monogamy vs. polygamy in voles (Lim et al.
Or, put another way, your lifestyle can influence the 2004), and alternative reproductive strategies in salmon
health of your great-grandchildren. Maternal nutrition (Aubin-Horth et al. 2005b).
during pregnancy causes epigenetic changes in gene Our knowledge of genome sequences has
activity in the fetus that increases susceptibility to obe- expanded well beyond lab animals and now includes a
sity, type-2 diabetes, heart disease, and cancer (Martin- wide variety of taxa. Even if we have only the gene
Gronert and Ozanne 2006). sequences of lab animals, such as the house mouse or
Norway rat, the similarity in the sequences often
allows us to transfer what we know about their DNA
COMPLEX RELATIONSHIPS sequences to wild populations of rodents, such as the
AMONG GENES deer mouse. These genome sequences may help iden-
tify the specific gene changes that are responsible for
We have seen that complex interactions among genes adaptation to a specific environment and to speciation
can affect behavior. The environment and social inter- (Storz and Hoekstra 2007). Polymorphisms occur in a
actions can also alter the pattern of gene expression, natural population. These small changes in gene
bringing about changes in behavior. As you can see in structure may result in subtle changes in the way in
Figure 3.22, these relationships may occur rapidly by which genes interact, giving rise to differences in the
effects on the nervous system. The relationships can also trait (behavior). The genetic polymorphisms that
change behavior on a slower time scale by affecting brain result in changes in the trait are what natural selection
development or altering the genome. As we will see in “selects” during evolution. Techniques have been
Chapter 4, evolution can also modify the genome developed to use microarray analysis to look at gene
(Robinson et al. 2008). expression in native populations in the field to gain an
understanding of the genetic basis of an organism’s
response to environmental change (Travers et al.
A BROADER PERSPECTIVE 2007). For example, there are about a dozen species of
finches on the Galapagos Islands, each differing in
Technical revolutions have vastly changed the field of beak shape and size. Different beak shapes adapt dif-
behavioral genetics in just a few short years. As never ferent species to eating different types of food. DNA
before, we are able to follow the thread that ties microanalysis has been used to discover that natural
together genes, cells, the entire organism and its behav- selection worked on a calcium-signaling molecule,
iors, all the way to the adaptation of the organism in calmodulin, to change beak shapes in these finches
its natural environment. A new interdisciplinary field (Abzhanov et al. 2006).
called evolutionary and ecological functional genomics Evolutionary functional genetics focuses on the
seeks to understand the processes that are biologically interface of genomics and evolution. In this chapter, we
important to both adaptation to the environment and have seen that evolution can occur by selection of new
evolutionary fitness. It brings together people with structural alleles or by altering gene regulation. In the
divergent interests—molecular, cellular, organismal, examples of food-related behavior in fruit flies, honey-
and ecological—to investigate how the mechanisms bees, harvester ants, and nematodes, we have seen that
that underlie a behavior increase function in a natural a relationship between a signaling molecule (PKG) and
environment to increase evolutionary fitness a type of behavior (obtaining food) can be conserved in
(Fitzpatrick et al. 2005). evolution, but molecular pathways between the two may
 A Broader Perspective 55

Social
interactions

Individual
behavior Epigenetic
changes

Brain
Development function Evolution

Genome

FIGURE 3.22 Complex relationships among genes, the nervous system, and behavior occur over
different time scales. These relationships can vary over physiological time by affecting brain activity.
These relationships can also work over developmental time, that is, via effects on brain development
and modification of the pattern of gene activity. The genome and behavior can also be altered over
evolutionary time. (Modified from Robinson et al. 2008.)

differ. Recall that high PKG increased foraging activity actions among members of the same species. This grow-
in fruit flies and honeybees but decreased foraging activ- ing field, called sociogenomics, endeavors to understand
ity in harvester ants. The fact that increased levels of how genes influence social behavior. Sociogenomics
PKG have the opposite effect on behavior in the fruit fly goes beyond simply identifying genes. It seeks to under-
and honeybee as they do in the harvester ant and nema- stand the functions of those genes, and how they affect
tode could suggest that the association evolved twice the mechanisms that bring about behavior, primarily
(Fitzpatrick and Sokolowski 2004). those of the nervous and endocrine systems (Robinson
Some researchers apply the techniques of evolu- 1999). Indeed, most of the examples we discussed in this
tionary and ecological functional genomics specifically chapter could be considered examples of sociogenomics
to social behavior, that is, behavior that involves inter- at work.
56 Chapter 3 / Genetic Analysis of Behavior

two alleles forR and forS. ForR is dominant and results in

SUMMARY larvae and adult flies called rovers because they travel
Genes are made up of DNA (deoxyribonucleic acid). long distances while foraging. ForS is recessive and
Behavior depends on which genes are being expressed, results in flies described as sitters because they travel only
when, and in which tissues. short distances during foraging. PKG, an important cell-
A goal of behavior genetics is to decipher the inter- signaling molecule, is for’s protein product. For was a
actions among genes and between genes and the envi- candidate gene for foraging behaviors in other organ-
ronment to understand why a particular behavior takes isms. For underlies foraging behaviors in honeybees, har-
the form it does. vester ants, and the nematode.
The classical methods of experimentally demon- Among the important principles of behavioral
strating the genetic basis of behavior include inbreeding, genetics are that one gene usually affects several traits
artificial selection, and hybridization. Inbreeding is the and that genes work in interacting networks. Thus,
mating of close relatives. though we often use the shorthand that a gene is “for a
Many animal studies on behavioral genetics begin behavior,” what we mean is that such genes are embed-
with a mutation, which may be induced or occur natu- ded in regulatory networks that control the behavior.
rally. Then a researcher identifies the gene and discov- The activity of immediate-early genes is often the
ers how the gene influences a behavior. Artificial first quantifiable indication of genes interacting with the
selection is a different breeding regimen in which indi- environment. These genes become active within minutes
viduals showing a desired behavior are bred with one to hours after nerve cell activity. Their protein products
another. If the frequency of the trait in the population help orchestrate the behavioral response to stimuli.
increases when the appropriate breeders are mated, then Changes in the tightness of the coiling of DNA
the behavior must have a genetic basis. Hybridization is influence gene expression without altering DNA
another breeding system used to demonstrate that a sequence. We see this in the effect of the quality of
given behavior has a genetic basis. Individuals that dis- maternal care on the stress responses and future mater-
play the behavior in distinct but different ways are mated nal behavior of pups.
with one another, and the behavior of the hybrid off- Evolutionary and ecological evolution seeks to
spring is observed. understand how natural selection works to select the
The foraging (for) gene provides a good example of genes that are responsible for ecologically adaptive
the techniques used by behavioral geneticists. For was behavior. Sociogenomics is the evolutionary and eco-
first identified in fruit flies, in which it exists naturally in logical genomics of social behavior.
 4
Natural Selection and Behavior

Natural Selection too severe. The returning parents would adopt and feed
Common Misunderstandings the new chicks, as if they were following the rule that any
About Natural Selection chick in the nest should be treated as offspring. However,
playing the same trick on herring gulls would cause
Genetic Variation
greater commotion: these gulls recognize even very
Variation Is Common
young chicks as their own and refuse to care for the
The Raw Material of Genetic Variation chicks of others.
Variation and the Response to Natural Selection These observations make sense in light of the ecol-
The Maintenance of Variation ogy of the two species. Even without an experimenter’s
Gene Flow and Genetic Drift intervention, herring gulls are likely to encounter chicks
Correlated Traits from other nests: they nest on the ground in large
Changing Environmental Conditions
colonies, and neighboring chicks often wander (Figure
4.1). Kittiwake gulls, on the other hand, nest on steep
Frequency-Dependent Selection
cliffs, where the chances of the wrong chick ending up
Negative-Assortative Mating
in the nest are minimal (Figure 4.2). Herring gulls and
Evolutionarily Stable Strategies: Fitness and the kittiwakes differ in other traits as well: herring gulls have
Behavior of Others
wider and shallower nests, whereas the deeper nests of
Testing Hypotheses About Natural Selection kittiwakes are less likely to allow eggs to roll off cliffs.
and Adaptation Predators such as foxes can move freely through a her-
The Experimental Approach ring gull colony but can’t reach the tiny ledges where the
The Comparative Approach kittiwake nests are perched. Thus, herring gull parents
Monitoring Selection in the Field would do well to make the nest less obvious to preda-
Modeling the Costs and Benefits of Traits tors. In fact, this seems to be the case: herring gulls
remove the tell-tale signs of eggshells and droppings
from the nest area, thus reducing the chance that preda-
If a meddlesome biologist scooped up kittiwake gull tors will detect the nests. Kittiwakes, in contrast, are less
chicks from neighboring nests and swapped them when tidy, leaving egg shells and droppings in the vicinity of
their parents were away, the consequences would not be their nest (Cullen 1957a, b).

57
58 Chapter 4 / Natural Selection and Behavior

FIGURE 4.1 The nesting grounds of the

herring gull. Nests are close together
and accessible to predators.

The behaviors of these species are therefore well

suited to their respective environments. But by what
process does this match come about? How does it happen
that even closely related species such as these can differ so
much? The answer, of course, lies in evolution. As the sci-
entist Theodosius Dobzhansky famously put it, “Nothing
in biology makes sense except in the light of evolution.”
We define evolution as a change in the frequencies
of different alleles in a population of organisms over gen-
erations. An allele is an alternative form of a gene
(Chapter 3). A population is an interbreeding group of
organisms of the same species. Although other types of
evolutionary forces have an impact on animal behavior,
natural selection is arguably the most important and is
the reason why species so often appear to be well suited
for their environments. We’ll begin by examining nat-
ural selection in detail, and we’ll save our discussion of
other evolutionary forces for later in the chapter.

NATURAL SELECTION
Charles Darwin was already convinced that species were
not fixed, but could change over time, when he had a
flash of insight about the process by which this change
occurs. He spent several more decades immersed in
research, deep thought, and extensive writing before he
published his first paper on the subject of the causes of
evolutionary change. During those years, Darwin con-
templated many esoteric topics, including pigeons.
Pigeon breeding was a common pastime in the 1800s.
FIGURE 4.2 Nesting area of the kittiwake. The shift in Darwin was interested in the great variety of pigeon
nesting sites from the ground to cliffside has rendered breeds. Some had fancy feathers around their heads,
the kittiwake nests inaccessible to predators and is cor- others had flashy colors, and still others had strange tum-
related with many behavioral changes. bling behavior during flight. Although no one yet knew
 Natural Selection 59

about the existence of genes, pigeon fanciers understood Observation 2: Some of these variable traits are
that they could obtain pigeons with the traits they genetically based. They have been inherited from
wanted by mating males and females that had these parents and can be passed on to offspring. As we
traits. Often, the offspring had even more extreme traits have seen in Chapter 3, many behavioral traits are
than did either of their parents. Slowly, over generations, influenced by the genotype.
extraordinarily bizarre pigeons could be bred through
Observation 3: Among the inherited traits are a few
this process of artificial selection. Other breeders had
that improve an individual’s chances of leaving more
long made similar modifications in other species of
offspring than other individuals can. These traits
domesticated animals, selecting for chickens that reliably
may be important to an animal’s survival (e.g., how
laid eggs, cows that produced a great deal of milk, and
well can it avoid a predator or tolerate a cold win-
dogs that were good at hunting or herding.
ter?), its reproduction (e.g., how well can it attract
Darwin noticed a parallel between the process that
mates?), or both. Notice that Darwin’s quote above
was happening in pigeon lofts and what might be going
from The Origin of Species focuses on competition,
on in nature. Reluctant to publish his ideas before they
but in fact, many different traits may determine an
were fully fleshed out, he painstakingly assembled his
animal’s reproductive success.
evidence over the decades. Finally, pushed by the fact
that another naturalist, Alfred Russell Wallace, had con- Conclusion: Because offspring are likely to inherit
verged on the same ideas, Darwin published On the their parents’ beneficial traits, these traits become
Origin of Species by Natural Selection or The Preservation of more common in the population relative to the traits
Favored Races in the Struggle for Life. This quote sums up borne by less successful reproducers. This is evolu-
the essence of Darwin’s view (1859, Chapter 1): tion by natural selection.
As many more individuals of a species are born than
can possibly survive and, as there is a frequently recur- Whenever we find the simple prerequisites of her-
ring struggle for existence, it follows that any being, if itable variation along with differential survival and
it vary however slightly in any manner profitable to reproduction of some phenotypes, we face a logically
itself, under the complex and sometimes varying condi- inevitable conclusion: evolution must occur. This type
tions of life, will have a better chance of surviving, and of evolutionary change is called natural selection
thus be naturally selected. From the strong principle of because nature “selects’’ those traits that enhance
inheritance, any selected variety will tend to propagate reproductive success. For example, male bighorn
its new and modified form. sheep that are victorious in head-to-head butting con-
tests with other males generally leave more offspring
Since 1859, Darwin’s ideas have been expanded and than do the losers (Figure 4.3, Geist 1971). Winning
built upon in an explosion of both empirical and theoreti- males are likely to have a suite of characteristics: per-
cal research. However, the core concepts of natural selec- haps they are larger, or their horns are stronger, or
tion remain the same. Let’s put them into simpler language. they are better strategists during fights. Any of these
Observation 1: Individuals in a population vary: traits with a heritable component will be passed on to
they differ in appearance, behavior, physiology, or their offspring. Because the winning males leave more
some other part of their phenotype. offspring than losing males, the composition of the

FIGURE 4.3 Two bighorn rams clashing

heads. The winner will have priority in
mating and will be more likely to leave
offspring.
60 Chapter 4 / Natural Selection and Behavior

population will change in the next generation: more traits can be more important to evolutionary success.
individuals will have alleles that code for the “winning” For example, sometimes small sneaky males mate with
traits rather than the “losing” traits. These traits are more females than do large dominant males. We can
then said to be under selection. Traits that have identify two components of fitness. The most com-
evolved because they allow individuals to survive and monly measured is direct fitness, the number of sur-
reproduce better than their competitors are called viving offspring an individual produces. An individual
adaptations. (We will restrict the term adaptation to can also increase the number of its alleles that survive
those traits that have a genetic basis. This definition in the next generation by helping relatives who share
excludes learned behaviors, although the capacity to its alleles (for example, by helping a sister raise her off-
learn may be an adaptation. We’ll return to this point spring). Fitness gained by helping relatives is called
in Chapter 5.) The word “adaptation” also refers to the indirect fitness. Direct and indirect fitness are
process of change over evolutionary time that occurs together called inclusive fitness. We’ll discuss inclu-
through natural selection. Note that biologists say sive fitness in detail in Chapter 17, but for the present,
“adaptation” but not “adaption.” we’ll focus on direct fitness. We are generally most
interested in relative fitness, or the average fitness of
a gene or individual compared with the rest of the
population.
COMMON MISUNDERSTANDINGS A related phrase that should be avoided is “for the
ABOUT NATURAL SELECTION good of the species.” A common misconception is that
traits evolve in order to help a species survive. However,
Although the logic of natural selection is straightfor- natural selection cannot act with the future of the species
ward, nonscientists often share some common misun- in mind. If a trait increases an animal’s fitness relative to
derstandings about the concept, which are sometimes that of other animals, the trait will increase in the pop-
reinforced by poorly written articles in the popular ulation, even if it means trouble for the species in the
press. This problem is exacerbated by the terminology long term. For instance, if a genetic mutation arises that
of evolution, which uses words that have other, more increases the number of individuals’ offspring, natural
commonly used meanings. For example, by now you selection will likely cause the mutation to become much
should see that the word “selection” is not used in more frequent. The population may grow quickly and
quite the same way in the phrase “natural selection” as outstrip its available food resources, and then crash.
in the phrase “artificial selection.” In artificial selec- Thus, a trait that is favorable to the individual may
tion, the selective force is imposed by humans’ partic- increase in frequency, even though it is not favorable for
ular goals. In contrast, natural selection is not capable the population or species. We will explore this concept
of long-range or even short-range planning. Although in more detail in later chapters.
it might be beneficial for humans to have wheels built For many people, “evolution” conjures up images
into their feet, natural selection cannot see us through only of macroevolution, or large-scale changes over
generations of humans with rudimentary and useless geological time, such as birds evolving from reptilian
axles in our arches, or other necessary precursors of ancestors. This is indeed an example of evolutionary
those useful wheels. change, but it is not the whole story. Remember our def-
Another troublesome phrase is “survival of the inition: evolution is a change in allele frequencies in a
fittest.” First, “survival” is only one of many traits that population of organisms over generations. Thus, the
natural selection might act upon: an animal must not concept of evolution also encompasses small changes
only survive, but must compete for resources, find a that happen over only a few generations. For example,
home and a mate, and ultimately produce offspring the Colorado potato beetle damages crops. When farm-
that carry its genes. Any one of these abilities, plus ers spray pesticides on their fields, most beetles die, but
many more, might be improved through natural selec- a lucky few happen to have alleles that allow them to sur-
tion. The second part of the phrase, “the fittest,” is also vive the poison. Soon the population is primarily com-
misleading: it evokes the picture of the most physically prised of the offspring of these fortunate beetles, all of
fit, strongest, and most aggressive individuals domi- which have genes that confer resistance to the pesticide.
nating all others in order to pass their genes along. This is an example of microevolution, an evolutionary
However, in an evolutionary sense, fitness has a more change within species.
subtle meaning: it is the reproductive success of an Finally, remember that it is populations, not indi-
allele or an individual relative to other alleles or indi- viduals, that evolve. Animals may develop over the
viduals in the same population. Sometimes the biggest course of their lives, but evolutionary change only
and most muscular individuals do have the highest fit- happens in populations from one generation to the
ness in an evolutionary sense, but sometimes other next.
 Genetic Variation 61

basis, evolution does not act directly on the genotype

GENETIC VARIATION (the genetic makeup), but rather on the phenotype (the
observable traits). Selection therefore cannot act on
VARIATION IS COMMON genetic differences if they have no effect on the phe-
Variation across individuals is the rule rather than the notype. For example, recessive alleles in heterozygous
exception. All house flies may look much the same to individuals are not expressed in the phenotype, so nat-
you, but if you examined them under a microscope you ural selection cannot act on these traits when they
would see many differences: their wings are of slightly occur in heterozygous individuals: the number of off-
different lengths, the color is different, the hairs on their spring that a heterozygous individual has is not affected
abdomens vary. In most cases, not even the offspring of by the presence of that recessive allele. Natural selec-
the same parents are identical, as is easily demonstrated tion can only “see” recessive traits when they are in
by a glance at your own parents or your cat’s kittens homozygous individuals. Thus, to fully understand how
(Figure 4.4). Sometimes the differences between indi- selection acts on a trait, we need to understand the rela-
viduals are obvious, such as in size or color pattern; in tionship between the genotype and the phenotype. In
other cases, they are harder to detect, such as differences Chapter 3, we discussed some of the continually evolv-
in metabolic rate or in mating behaviors that are exhib- ing methods that behavioral geneticists use to link phe-
ited very briefly at particular times of the year. notype to genotype.
Sometimes variants fall into distinct classes. In other
cases, variation is continuous, changing gradually from
one extreme to the other. A common pattern is the famil- THE RAW MATERIAL OF
iar bell-shaped normal distribution, with most individ- GENETIC VARIATION
uals falling about midway between the extremes (Figure
4.5a). Regardless of the exact form it takes, it is clear that Natural selection does not create genetic variability, but
variation is a common feature of populations. only acts on the variability present in a population. So
Variation in the traits of organisms is called pheno- where does genetic variability come from? The two main
typic variation. As described in Chapter 3, phenotypic sources are mutation and recombination.
variation arises from two sources: the underlying genet-
ics and the environment. For evolution, the key com- Mutation
ponent of variation is genetic variation. Only traits that Mutation is a change in the DNA sequence of an organ-
are at least partially based on genes can evolve. Even if ism. Of concern to evolutionary biologists are mutations
there is differential survival and reproduction, if all indi- that occur in sperm and eggs, or the tissue that gives rise
viduals are genetically identical for a particular trait, that to them, and thus can be passed on to offspring. (In con-
trait cannot evolve by natural selection. trast, somatic mutations, such as most cancers, cannot be
It is also important to remember, however, that passed onto offspring.) Mutations come in many forms.
although traits cannot evolve unless they have a genetic Some mutations affect only a small part of the genotype:

FIGURE 4.4 Variation in offspring

of the same two parents. During
the meiotic cell division that
forms gametes, the alleles of each
parent are shuffled and recom-
bined. Thus, sexual reproduction
would scramble any “perfect”
combination of genes.
62 Chapter 4 / Natural Selection and Behavior

a Directional selection the affected gene’s job is to make proteins (a structural

gene) or to turn other genes on or off (a regulatory
gene). Other mutations are larger: genes may be dupli-
cated or deleted, and entire pieces of chromosomes can
move from one chromosome to another or be reversed
in their orientation.
Any type of mutation, if its owner survives, increases
genetic variation in the population. However, variation
produced by mutation is likely to be disadvantageous.
After all, a random change in any finely tuned machine,
whether it is a living body or a laptop computer, is
unlikely to be an improvement. Nonetheless, in spite of
the fact that only a tiny percentage of mutations are
beneficial, mutation is the ultimate source of variation.
b
Recombination
In organisms that undergo sexual reproduction, another
source of genetic diversity is the recombination of alleles
Frequency

that occurs during meiosis, the type of cell division that

results in the formation of gametes (eggs or sperm).
During meiosis, a diploid parent (with a pair of homol-
ogous chromosomes) produces haploid gametes. A
gamete thus contains only half of the genetic material of
the parent that produced it, so a parent that produces an
offspring by sexual reproduction shares only half its al-
leles. Chromosomes containing different copies of an
allele can show up in different gametes in any combina-
c tion. Thus, two sisters share half of each parent’s genes
but not necessarily the same half. Furthermore, at the
start of meiosis, when homologous chromosomes are
lined up, pieces of chromosomes containing alleles for
the same gene are swapped in a process called crossing-
over. Individual chromosomes in the gametes are there-
fore unlikely to be identical to those in either parent,
further increasing the differences among siblings. Thus,
even if natural selection were somehow able to produce
an individual with a perfect combination of alleles, the
offspring of a sexually-reproducing individual would not
Nest-building behavior
have that same combination.

FIGURE 4.5 Directional selection. (a) The familiar bell

curve, illustrating a fictional data set of nest-building
VARIATION AND THE RESPONSE
behavior in mice. The dotted line indicates the popula- TO NATURAL SELECTION
tion average. Most individuals build medium-sized
nests, with some building very large nests and others The genetic variation provided by mutation and recom-
very small nests. (b) If the weather grows colder, only bination provides raw material on which natural selec-
the mice with the largest and warmest nests will suc- tion can work. Remember that evolution happens at the
cessfully reproduce. These lucky mice are at the right- population level, not the individual level, so let’s think
hand side of the bell curve. (c) In the next generation, explicitly about how populations change when they
the population is shifted to the right. Note that the undergo natural selection.
average is larger than in (a). Return to the normal distribution in Figure 4.5a.
What might we expect to happen to this distribution
perhaps one base pair of nucleotides is substituted for over subsequent generations? Under stable environ-
another, or base pairs are inserted or deleted. However, mental conditions, the animals with traits at the center
in many cases even these small changes have a dramatic of the distribution would be expected to do best: they
effect on the function of a gene, regardless of whether might be most successful in the current environment,
 The Maintenance of Variation 63

whereas those at the extreme ends of the distribution are GENE FLOW AND GENETIC DRIFT
less well suited to current conditions. Under conditions
Imagine a group of populations, all physically near
of stabilizing selection such as this one, where the optimum
each other but in slightly different ecological condi-
phenotype is near the population’s average, the mean (or
tions and thus under different selection pressures. If
average) phenotype in the population is unlikely to shift
these populations were completely separated from one
much, if at all, from one generation to the next, although
another, then one would expect them to diverge over
the distribution may become narrower. In other cases,
time, with local adaptation to the conditions of each
the environment may change and the optimum pheno-
region. But what would happen if the populations are
type may shift over time. In this case, those at one
not completely isolated and individuals moved between
extreme would come to be favored (Figure 4.5b), and the
them? Genes from the populations would mix in a
curve that represents the population’s phenotype would
process called gene flow. Gene flow makes popula-
shift in that direction (Figure 4.5c). This phenomenon,
tions more similar to one another. Depending on its
called directional selection, can be weak—resulting in a
strength, gene flow can slow or even halt the effect of
curve that shifts just a little from generation to genera-
local adaptation.
tion—or strong—resulting in a big change from one
An excellent example of the effect of gene flow on
generation to the next. Recall the example of microevo-
behavior is provided by Susan Riechert and her col-
lution in a field of insect pests sprayed with pesticide.
leagues. They documented genetically based differences
Pesticide application is an extremely strong directional
in territorial behavior among populations of the funnel-
selection pressure, and generally only a handful of indi-
web-building spider (Agelenopsis aperta) living in differ-
viduals are resistant. Their offspring quickly make up the
ent habitats. These spiders compete for sites in which to
bulk of the population, and thus the change in the
build their webs, and they defend these sites from con-
genetic makeup of the population can be enormous in
specifics (Figure 4.6; Riechert 1986). This species occu-
just a few generations.
pies a wide variety of habitats in its range from northern
We have already seen another way to visualize the
Wyoming to southern Mexico. Some spiders live in rel-
effects of selection. Refer to Figure 3.5, which illustrates
atively lush riparian vegetation along the rivers and lakes
the change over time in nest-building behavior in house
of Arizona where insect prey are abundant but predators
mice. The top lines are groups that underwent direc-
are also common. Other spiders live in desert grassland,
tional selection for large nests, and you can see that nest-
a much harsher environment. In this environment, insect
building behavior increased over time. The low lines
prey are few, and the scorching sun makes it impossible
show the response of lines undergoing directional selection
to forage during much of the day. There are fewer good
for small nests. The middle lines were from a control
places in which to build a web in the grassland than in
group. You can see that in the low lines there was an ini-
the riparian habitat.
tial decrease in nest size, but then the response to selec-
tion flattened out around the tenth generation. We
commonly see this pattern: as directional selection
proceeds, eventually we “run out” of genetic variation
upon which natural selection can act. So, when directional
selection is applied consistently, we expect to see the pop-
ulation respond by becoming more adapted, eventually
leading to a decrease in genetic variation. Mutation can
restore this variation, but on a longer time scale.

THE MAINTENANCE
OF VARIATION
Since natural selection favors those individuals with cer-
tain traits, why hasn’t it eliminated from the population
those individuals that bear other traits? Shouldn’t nat-
ural selection weed out the less optimal animals and keep
only the best? Given that mutations are rare, why do we
still see variation in most populations? Now we return FIGURE 4.6 A funnel-web spider at the entrance to the
to the topic we set aside earlier: natural selection is not funnel that extends from its web. There are differences
the only evolutionary force that changes the allele fre- in the expression of territorial behavior in populations
quencies in populations. of this spider that live in different ecological conditions.
64 Chapter 4 / Natural Selection and Behavior

Spiders living in these two habitats are strikingly dif-

ferent in behavior. Grassland spiders are much more
aggressive and will not allow other spiders near their
webs (Figure 4.7). Furthermore, the intensity of territo-
rial disputes between grassland spiders is greater than
those between riparian spiders. Threat displays of grass-
land spiders are more likely to escalate into battles, and
the fighting more often results in physical injury or death
(Riechert 1979, 1981, 1982). Here, in this unforgiving
environment, web sites are extremely valuable, so it is
worth taking the risk of engaging in a dangerous fight.
Grassland spiders are also very aggressive toward prey
and attack prey that land in their web much more quickly
than do riparian spiders (Hedrick and Riechert 1989). In
contrast, in a riparian habitat, web sites are easier to find
and therefore not so valuable. Risking injury is not as
likely to be worthwhile. Prey are abundant, so missing
the occasional insect by being a bit slow to attack is not
a big deal. Riparian spiders have other problems, how-
ever: birds and other predators also prefer the riparian
habitat. When researchers mimic a predation attempt by
disturbing their webs, riparian spiders are very cautious,
hiding for a long time compared to desert spiders before
venturing from the safety of their funnel (Riechert and
Hedrick 1990).
These behavioral differences are not simply a
response to environmental conditions, such as food avail-
ability, nor are they learned from territorial disputes or
other experiences. Rather, they are traceable to genetic
differences between the desert and riparian populations
(Maynard Smith and Riechert 1984; Riechert and
Maynard Smith 1989). Spiders were collected from an
arid grassland environment in New Mexico and from a
riparian environment in Arizona. Purebred lines were
established by allowing individuals from a particular
habitat to mate with one another. After the spiderlings
emerged from the eggs, each was raised separately and
given all the prey they could eat. When they were
mature, females from each population line were placed
in experimental enclosures where they could build webs.
FIGURE 4.7 Variation in ecological conditions in differ-
Just as in field populations, the average distance between ent regions of the range of the funnel-web spider. The
laboratory-raised females from riparian populations was desert grassland of New Mexico (top) has few suitable
less than that between females from grassland popula- web sites and low prey abundance. In contrast, most of
tions. Even under lush conditions, grassland spiders still the woodlands near rivers in Arizona (bottom) have
maintain a large web—an indication that the behaviors many suitable web sites and prey is plentiful.
responsible for territory size are genetically, rather than
environmentally, determined. In another experiment,
spiders were collected from either grassland or riparian low behavior typical of populations in riparian areas, this
habitats and transplanted to the other (Riechert and Hall population showed significantly more variability in
2000). Phenotypes that were inappropriate to the new behavioral traits, including the highly aggressive terri-
habitats were selected against. That is, they did not sur- torial behavior typical of desert populations. Why was
vive as well as phenotypes native to the habitat. this? It turns out that this riparian population was not
These populations of spiders thus seem well adapted isolated from desert populations but was constantly faced
to their respective environments. However, spiders of with an influx of immigrants. When researchers installed
one particular riparian population had characteristics a drift fence to prevent individuals from moving from
that do not fit this pattern. Instead of showing the mel- one population to another, they cut off gene flow. The
 The Maintenance of Variation 65

population then evolved over several generations to gene, may affect several traits (a phenomenon known as
become less aggressive and more cautious, and thus more pleiotropy). Similarly, genes are sometimes tightly linked
adapted to its local environment (Riechert 1993a). Thus, when they are physically close together on the same
gene flow, by making populations genetically more sim- chromosome. One gene may be dragged along for a time
ilar to one another, can keep populations from being as when there is selection on its linked partner, at least until
well adapted to their local environment as they might be. recombination and selection allow the correlation to be
Genetic drift is another evolutionary process that broken. Finally, two traits may share an underlying mor-
may influence behavior. Genetic drift is the change in phological and physiological basis that may make it dif-
allele frequencies in a population due purely to chance ficult to uncouple them. Whatever the reason, when
events. For example, imagine that the only individuals traits are tightly correlated, even negative traits might be
in a population that carry a rare allele happen to die maintained in the population if the net effect on the
before they breed, not because the allele itself is unfa- genotype is a positive one. For example, an individual’s
vorable, but just because those individuals were unlucky. behavior is often consistent across different environ-
The rare allele would then be gone from the population. ments (Sih 2004): an animal that is bold and thus very
Allele frequencies in populations drift up or down over successful in seeking out mates may also be inappropri-
generations. A particular allele might even drift to fixa- ately bold when investigating that strange noise in the
tion, so that it is carried in every member of the popu- underbrush. Boldness in both situations may be influ-
lation. As you may have deduced, genetic drift is more enced by the same underlying genotype, levels of hor-
important as population size gets smaller. mones, and the like, and thus behaving optimally in
Populations that have gone through a bottleneck— every situation might be impossible.
a sharp reduction in population size analogous to the Let’s look at an example of correlated traits. One of
narrowing of a bottle at its neck—may show long-last- the most famous studies is on the evolution of beak size
ing evidence of genetic drift. Populations sometimes and shape in Darwin’s finches in the Galapagos Islands in
experience bottlenecks because of natural events, but Ecuador. Charles Darwin was first to posit why many finch
bottlenecks are especially likely to occur in rarer animals species—all quite similar in appearance—nonetheless
of conservation concern. For example, a bottleneck in differed in striking ways, particularly in their beaks. He
cheetah population size may account for the increase in suggested that the species shared a common ancestor,
the frequency of deformed sperm in male cheetahs but that over time they diverged and specialized on dif-
(O’Brien 1994). Other populations for which drift is ferent food resources. Some species, for example, have
important include any stock of animals built up from a crushing beaks useful for seeds, and others have beaks
small number of individuals, such as laboratory animals, useful for poking into flowers. So far, this is simply a
domesticated livestock, fish in hatcheries, or insects wonderful illustration of the effect of natural selection
reared for biological control. In practice, it can be very on the traits of beak shape and size.
hard to be certain when genetic drift has acted: demon- However, finches use their beaks for more than just
strating with confidence that allele frequencies have feeding: male birds also sing, and how they sing is influ-
changed by chance means that all other explanations enced by their beak shape. Females base their choice of
need to be examined and eliminated. For example, island mates on song (Podos and Nowicki 2004). Thus, selec-
populations of bumblebees vary in how they search for tive forces may not act on just the single behavioral pat-
and handle flowers. Chittka et al. (2004) examined a tern of feeding, but there may be a ripple effect on
number of possible adaptive explanations for this varia- singing behavior and mating behavior. Correlated traits
tion, such as differences in the array of flower species on such as these do not evolve independently from one
the different islands, before concluding that drift was the another.
most likely explanation for the pattern. We can become
more confident of the importance of drift in a particu- CHANGING ENVIRONMENTAL
lar population if we have historical information about
CONDITIONS
population size, but unless the underlying genetic vari-
ation does not affect the phenotype in any way, it is chal- When we examine a population of animals today, we
lenging to be confident that the changes could have been must remember that we are looking only at a single snap-
caused by some selective factor that was not studied. shot in time. A population may appear to be poorly
adapted to current conditions because there is an evolu-
tionary time lag between selection and its effects.
CORRELATED TRAITS
Fluctuation of selection pressures from generation to
It would be a mistake to think of any organism as an generation means that natural selection must play catch-
assortment of traits that evolve independently of one up, so today’s traits may reflect past evolutionary pres-
another. Traits may be correlated with one another for sures, not current ones. Thus, paradoxically, natural
a number of reasons. One gene, such as a regulatory selection may actually be the reason that a trait we see
66 Chapter 4 / Natural Selection and Behavior

today does not appear to be adaptive. Experiments are factors, such as other animals. For example, until
needed to sort this out. recently the island nation of New Zealand had only two
Humans are the cause of a great deal of rapid envi- species of bats but no other mammals. Birds there have
ronmental change. Not all of the change is detrimental not faced mammalian predators in their evolutionary
to animals. In some cases, animals actually benefit from past, so they do not have the antipredator skills needed
close proximity to humans. For example, in northern to cope with them. Many New Zealand birds will alight
Massachusetts, opossums (Didelphis virginiana) are at the fearlessly near dangerous animals. Some, such as a par-
northernmost edge of their range; their naked tails and rot called the kakapo, have even lost their ability to fly
ears make them vulnerable to the cold. Opossums that (Figure 4.9). Now that cats, rats, stoats, weasels, and
survive a harsh winter are those that live near humans and other predators have become established, many bird
can take advantage of shelter and food, such as leftover populations are in dramatic decline. Some researchers
corn in the fields inadvertently provided by humans have tried to instill fear of predators into native birds by
(Kanda 2005; Kanda et al. 2006; Figure 4.8). One might using aversive conditioning techniques. For example,
predict that over time there will be selection for opossums McLean and his colleagues (1999) presented dead
that are less fearful of humans. stuffed cats and ferrets to New Zealand robin chicks
Global warming is also causing new selection pres- (Petroica australis) while playing robin alarm calls and
sures on species (Walther et al. 2002). For example, newts distress calls. Robins learned to associate cats and ferrets
and frogs both breed in ponds in the spring. In Britain, with danger and reduced their tendency to approach
newts (Triturus spp.) have responded to warming tem- them. Techniques such as these are obviously extremely
peratures by entering ponds earlier than they used to, but time-intensive but may be useful as a last-ditch measure
frogs (Rana temporaria) still reproduce at the same time. to save severely threatened populations. Most preserva-
This means that frog eggs and tadpoles are now exposed tion efforts revolve instead around intensive trapping of
to more newt predators than before (Beebe 1995). the predators, but even that approach is only feasible on
Humans cause changing selection pressures in other small islands or fenced-in reserves. Antipredator behav-
ways. To take just a few examples, habitat is lost to devel- ior may sometimes evolve in response to new predators,
opment, pollutants and fertilizers change water chem- but some species are likely to be lost forever.
istry, traditional migratory stopovers disappear, and light Adaptations that evolve in one species may change the
pollution (caused by use of artificial lights throughout the selection pressures on other species, which in turn may
night) can interfere with animal navigation. These envi- change the selection pressures on the first (Van Valen
ronmental changes are occurring so rapidly that they are 1973). For example, insectivorous bats use sonar to locate
just a blink of an eye in evolutionary time, and many pop- flying moths. In response, some moth species have
ulations cannot evolve fast enough to keep up. evolved the ability to detect the ultrasonic signals emitted
The environment of a species includes not only abi- by the bat and to undertake evasive action with a fast erratic
otic factors, such as climate and weather, but also biotic flight. Bats are then under even greater pressure to detect

FIGURE 4.8 North American

oppossums can benefit from
proximity to human
habitation.
 The Maintenance of Variation 67

Evidence from many species shows that predators do

indeed choose the most common type of prey, especially
if the prey density is low (reviewed in Allen 1988). Alan
Bond and Alan Kamil (1998) found experimental support
for the hypothesis that frequency-dependent predation
can maintain genetic diversity in a population. In these
ingenious experiments, blue jays (Cyanocitta cristata)
“preyed on” virtual moths presented on touchpad com-
puter screens. Photographs of the dark form of a moth
(Catocala relicta) commonly eaten by blue jays were
scanned to create digital images that were then modified
to create four different forms of moths. A fifth form was
generated in a similar manner but with a different moth
species (C. retecta). The virtual moths were presented on
FIGURE 4.9 A flightless New Zealand parrot called the a computer screen against one of five different back-
kakapo. This species is vulnerable to introduced preda-
grounds that altered the difficulty of detecting the moths.
tors, such as stoats (a kind of weasel).
The blue jays preyed on the moths by pecking at the
screen. The “dead” moth was removed from the display,
and follow moths. This sort of coevolution is known as an and the jay received a food reward. The relative numbers
evolutionary arms race, analogous to the mutual back-and- of moth forms that escaped detection determined the
forth escalation of weaponry between the United States subsequent abundance of each prey type. The blue jays
and Russia in the decades after World War II. Species preyed on the most common form of moth and switched
engaged in arms races are also said to be behaving like the to alternative forms when that form became less com-
Red Queen in Lewis Carroll’s Alice in Wonderland, who had mon. This was true as long as the virtual moths were not
to run as fast as possible just to stay in place. too cryptic. Thus, frequency-dependent predation can
maintain variation in the appearance of prey. In nature,
FREQUENCY-DEPENDENT SELECTION the maintenance of this prey polymorphism (literally,
“many forms”) would also maintain the genetic variation
Sometimes variation is maintained in a population underlying it. As we discussed earlier, selection acts on the
because different genotypes are favored at different times. phenotype, but evolution happens through changes in
One such mechanism is frequency-dependent selection, allele frequencies. We’ll see more examples of frequency-
in which an allele has a greater selective advantage when dependent predation in Chapter 13.
it is rare than when it is common in the population. As a
result, the frequency of any given allele fluctuates: it
increases until it is common and then decreases once the Frequency-Dependent Reproduction
alternative allele is favored. There are many examples of
In this type of mating, sometimes called the rare-male
frequency-dependent selection (Ayala and Campbell
effect, a male with a rare phenotype enjoys more than
1974). We will consider two types—frequency-dependent
his expected share of matings. The alleles of the rare
predation and frequency-dependent reproduction.
phenotype increase in the population until they become
common and are no longer favored. The allele frequen-
Frequency-Dependent Predation cies of different phenotypes can thus seesaw back and
It is easy to see how frequency-dependent predation forth over time. The rare-male effect has been demon-
could maintain variation in a population of prey. strated in guppies (Poecilia reticulata), a fish species in
Although predators usually have a varied diet, they often which male coloration is extremely variable even within
attack one prey type more often than is expected by a single population. Females were allowed to examine
chance. For example, when the members of a prey species males through a glass partition. They were then given a
differ in some characteristic, such as color, a predator choice between mating with a male of a familiar pheno-
might concentrate on the most common form and ignore type or a male of a novel phenotype. Females chose
the less common forms. The more common individuals males with novel color patterns—rare males—over
are preferentially attacked until their numbers, and thus males with a color pattern with which they were famil-
their alleles, decline in frequency. Meanwhile, the rarer iar (Hughes et al. 1999). Over time, frequency-depen-
form survives and reproduces, and its relative frequency dent mating can maintain a variety of male phenotypes
increases. Then the predator switches to the new most in the population, in the same way that frequency-
common form of prey, which eventually then decreases dependent predation maintained variation in the types
in number, and the cycle begins again. of moths in the example above.
68 Chapter 4 / Natural Selection and Behavior

An example that combines these types of frequency EVOLUTIONARILY STABLE

dependence comes from Texas field crickets, Gryllus STRATEGIES: FITNESS AND
texensis. Male crickets chirp by rubbing together a spe- THE BEHAVIOR OF OTHERS
cial “file” and “scraper” on their wings. This calling
behavior attracts females. Males vary tremendously in Sometimes the success of a strategy depends on what
the time they devote to calling every night—some other individuals are doing. Recall the rules of the child-
rarely or never call, and others call for more than ten hood game of rock–paper–scissors: rock breaks scissors;
hours in a night. Why would there be this much vari- scissors cut paper; paper covers rock. Any one of the
ation? Unfortunately for the males, calling also attracts moves could win or lose depending on the actions of the
parasitoid flies, which lay their eggs on the crickets. other players. Play scissors and you win if your opponent
The fly larvae burrow into the males, killing them plays paper. But if you play scissors too often, your oppo-
within about a week. Flies are especially active early in nent will catch on and defeat you by playing rock. The
the evening and are most common in the fall. Thus, optimal behavior is dependent on frequency: whether
when flies are common, the calling males are soon par- you win depends on the frequency of the strategies
asitized, and the males calling less end up with more played by others. If you don’t have any knowledge of
mates over their (longer) lives. Of course, when the flies their plans, the best way to have a chance of breaking
are rare, the calling males have the obvious advantage even is to play all three strategies—rock, paper, and scis-
(Bertram 2002). sors—in random order with equal frequency (Maynard
Smith 1976).
Just as rock, paper, and scissors may be considered
alternate strategies in a game, an animal’s behavior may
NEGATIVE-ASSORTATIVE MATING
be described as a strategy (Maynard Smith 1976, 1982).
Negative-assortative mating also preserves genetic Our use of the word “strategy” does not imply that the
variation in a population. This, in essence, means that animal consciously plans the best course of action for
opposites attract. More generally, the term describes the maximizing reproductive success. “Strategies” are sim-
situation in which individuals tend to choose mates with ply the set of behaviors available to an animal, and “win-
a different phenotype than their own. It differs from ning” means that the individual’s fitness increases more
rare-male advantage, where females of all phenotypes than its competitor’s does (i.e., it leaves more offspring).
prefer the unusual males. Here, females of different The optimal strategy for an individual to follow
phenotypes have different preferences. Obviously, if the when the rewards (called payoffs) depend on what oth-
phenotypic difference has a genetic basis, genetic vari- ers are doing is called an evolutionarily stable strategy
ability will be enhanced. Admittedly, such assortment is (ESS). By definition, an ESS is a strategy that, when
not common in nature. adopted by nearly all members of a population, cannot
Negative-assortative mating maintains the tan- be beaten by a different strategy: no other strategy con-
striped and white-striped morphs of white-throated fers more fitness benefits. For example, remember that
sparrows (Zonotrichia albicollis) in approximately equal a herring gull will not take care of a neighboring chick
numbers in a population. Both female morphs prefer that wanders into its nest. This is an ESS because there
tan-striped males, which are better parents because is no alternative behavior that will yield greater repro-
they spend more time feeding the chicks. However, ductive success: the alternative strategy of caring for
the white-striped females outcompete the tan-striped other birds’ chicks would mean that herring gull parents
females for access to the tan-striped males. Tan-striped would waste time and energy caring for offspring that are
females then pair with the leftover white-striped not their own. As a result, an ESS is unbeatable and
males. As a result, 93 to 98% of the population mates uncheatable in the long run.
with an individual of the opposite morph (Houtman An ESS may be “pure” and consist of a single strat-
and Falls 1994). egy, as in the example above, or it may be “mixed,” con-
In some species, negative-assortative mating is a sisting of several strategies in a stable equilibrium.
mechanism that prevents inbreeding. For example, in Consider, for instance, a hypothetical population of fish-
some strains of mice, individuals can determine whether catching birds. There are two strategies for getting din-
others share certain of their alleles by the smell of urine. ner—catch your own fish or steal one from another bird.
They then choose mates that have different odors from Natural selection is assumed to favor the strategy that
themselves (reviewed in Penn and Potts 1999). Patrick maximizes benefit. Since the thief minimizes its costs and
Bateson (1983) has suggested that some species are able gets full benefits from the efforts of others, thievery is
to recognize kin (individuals sharing many alleles) and favored initially. However, as the proportion of bandits
then choose mates that differ from kin. Bateson’s ideas in the population increases, so does the likelihood of
are discussed further in Chapter 8. encountering either another robber or a bird that has
 The Maintenance of Variation 69

already had its fish stolen. Then, honesty becomes the Decisions Outcomes
best policy. When hard-working birds become common,
thievery once again becomes profitable (Dawkins 1980). Abandons
As the relative frequencies of alternative strategies fluc-
Dig
tuate, they reach some ratio at which both strategies Is alone
result in equal reproductive success. That particular mix Establish
of strategies will be an ESS. a new nest
Mixed ESSs can arise in two ways. First, different Is joined
Enter
genotypes could be responsible for producing each strat-
egy. In this case, each individual of the population would Joins
always adopt the same strategy, but individuals would
differ on which strategy they adopted. At equilibrium, FIGURE 4.11 A female digger wasp’s alternative nesting
there would be a particular frequency of each type of strategies and their outcomes. The two available
individual. Second, each individual could vary its strat- strategies are to “dig” and to “enter” a burrow dug by
egy and play each one with a certain frequency. Using someone else. There are three possible outcomes of a
the rock–paper–scissors example again to illustrate this decision to dig: the female may remain alone and retain
difference, we see that stability would result if one-third exclusive use of the burrow; she may be joined by
another female; or she may have to abandon the nest
of the population played rock, one-third played scissors,
because of some catastrophe. If a female decides to enter
and one-third played paper. Alternatively, each member
an existing burrow, she may be alone or she may be
of the population could play rock, paper, and scissors joining another female. (From Brockmann et al. 1979.)
with equal frequency.
Now let’s examine a few biological examples of evo-
lutionarily stable strategies. a burrow. Furthermore, the investment is not risk-free.
There is no guarantee that she will not be joined by
The Nesting Strategies of Digger Wasps The nest- another female, and if she is, she may lose her invest-
ing behavior of female digger wasps (Sphex ichneumoneus) ment. In addition, temporary catastrophes can take
is a clear illustration of a mixed ESS in which two strate- place—for instance, an invasion by ants or a centipede—
gies coexist (Brockmann et al. 1979). A female lays her that may force a female to abandon her burrow. Once
eggs in underground nests that consist of a burrow with the intruders have gone, the abandoned nest is quite suit-
one or more side tunnels, each ending in a brood cham- able for nesting again. A female that finds an abandoned
ber (Figure 4.10). She lays a single egg in a brood cham- nest reaps the benefits without incurring the costs. So,
ber after provisioning it with one to six katydids, a it might seem that entering an existing burrow would be
process that can take as long as ten days. the favored strategy. Indeed, it is—if the burrow is
Here is the choice female wasps face: to dig or not actually abandoned. Unfortunately, there is no way to
to dig (Figure 4.11). On some occasions a female will dig determine whether the nest is abandoned or whether the
her own nest, while on others she will enter an existing resident is just out hunting. A female who is provisioning
burrow. Digging has an obvious cost in time and energy a nest is gone most of the time; it may be hours or days
since it takes a female an average of 100 minutes to dig before she learns that another female is occupying the
nest. Eventually the two females will meet, and when
they do, they fight, sometimes to the death, and winner
takes all. If the intruder wins, she continues to provision
the nest and lays her eggs on the jointly provided sup-
ply of katydids.
Whether it is best to dig or to enter depends on
what the other members of the population are doing.
The strategy of entering an existing burrow is most suc-
cessful when it is rare. As entering becomes more com-
mon, there are fewer diggers. As a result, the chances of
entering an occupied nest increase, along with costly
fights, and eventually digging becomes a better strategy.
For one digger wasp population studied in New
Hampshire, 41% of the decisions made by wasps were
to “enter” and 59% were to “dig” (Brockmann et al.
FIGURE 4.10 A female digger wasp at the entrance of 1979; Figure 4.12). Is that an evolutionary stable mix of
the burrow. strategies? If so, the reproductive success of females who
70 Chapter 4 / Natural Selection and Behavior

ries. Instead, these “sneaker” males mate covertly with

females on the territories of orange-throated males.
Dig They get away with this because orange-throated males
cannot defend all their females. However, a population
Relative fitness

of yellow-throated males is also not evolutionary stable

because this strategy can be invaded by males with blue
throats. Blue-throated males defend territories that are
only big enough to hold a single female, so they can suc-
Enter cessfully defend her against sneaky yellow-throated males.
However, when the yellow-throated sneaker males are
rare, it once again pays to defend large territories with
41% several females, and the reproductive success of orange-
Frequency of “enter” decisions
throated males exceeds that of blue-throated males.
FIGURE 4.12 The digger wasp’s strategies to “dig” and Thus, orange-throated males can successfully invade a
to “enter” are a mixed ESS: neither can win out over population of blue-throated males, completing the
the other under all conditions. The success of either dynamic cycle. In this mixed ESS, then, yellow beats
strategy depends on that adopted by other members of orange, blue beats yellow, and orange beats blue. Notice
the population. For a New Hampshire population of the similarity between changes in morph frequency and
digger wasps, the strategies have equal fitness when the rock–paper–scissors game. The predominant color
they exist in the ratio of 41% “enter” and 59% “dig.” morph in a natural population was observed to fluctuate
in the manner predicted by these frequency-dependent
enter must equal that of females who dig. If the payoffs changes in fitness: blue was predominant in 1991, orange
of the strategies are not equal, females choosing the in 1992, yellow in 1993-1994, and blue again in 1995
more successful strategy will become more prevalent in (Sinervo and Lively 1996).
the population. To determine the reproductive success
of these strategies, Jane Brockmann and her colleagues
(1979) spent over 1500 hours observing 410 burrows. TESTING HYPOTHESES ABOUT
The reproductive success of dig versus enter decisions NATURAL SELECTION AND
was calculated. In the study population, the researchers ADAPTATION
found no significant difference in the reproductive suc-
cess of individuals who adopted the digging versus enter- Questions about the evolution and adaptive significance
ing strategies. Whereas the reproductive success of of behavior have been central to ethology since its begin-
individuals who decided to dig a nest was an average of ning. Recall that two of Tinbergen’s (1963) four ques-
0.96 egg laid per 100 hours, the reproductive success of tions were: What is a trait’s function (survival value)?
those who decided to enter an existing nest was 0.84 egg How did the trait evolve? These are questions about,
per 100 hours. Although the average for nondiggers was respectively, the current adaptive value of a trait and its
slightly less than that for diggers, the range of variation evolutionary history. These questions still drive the
in reproductive success between the two groups over- research of many ethologists, behavioral ecologists, and
lapped broadly. Therefore, the nesting strategies of sociobiologists.
female digger wasps comprise a mixed ESS. We began the chapter by observing that many ani-
mal traits appear to be well suited to their environments,
Reproductive Strategies of Male Lizards The and we attributed this match to natural selection.
changes in fitness among alternative male reproductive Natural selection is a powerful force, but we must keep
strategies of the side-blotched lizard (Uta stansburiana) in mind that the notion that any particular trait is an
provide an example in which the strategies of a mixed adaptation is a hypothesis that must be tested. Stephen
ESS cycle from one generation to the next. This small J. Gould and Richard Lewontin (1979) called behavior-
iguanid lizard lives in the inner coast range of California. ists (and others) to task for too often accepting the adap-
Male lizards come in three throat colors: orange, yellow, tive hypothesis without proof. Indeed, they ridiculed the
or blue. Each of these three genetically determined color adaptationist approach, which they claim breaks an indi-
morphs also displays a different reproductive strategy. vidual into separate traits and assumes that each of those
Males with orange throats are very aggressive and defend traits is adaptive. Gould and Lewontin named this prac-
large territories, within which live several females. tice a Panglossian paradigm, after Dr. Pangloss in
However, a population of only orange-throated males is Voltaire’s satire Candide who made the obviously absurd
not evolutionarily stable because yellow-throated males assertion that “Things cannot be other than they are. . . .
can steal their mates. Yellow males don’t defend territo- Everything is made for the best purpose. Our noses were
 Testing Hypotheses about Natural Selection and Adaptation 71

made to carry spectacles, so we have spectacles. Legs behave in a certain way survive and reproduce better
were clearly intended for breeches, and we wear them.” than those who behave in some other way. In our consid-
Although Gould and Lewontin suggested that a eration of territoriality in funnel-web spiders, the alter-
Panglossian philosophy is inherent in the thinking of all native forms of behavior were easy to identify. The
adaptationists, Ernst Mayr (1983) argued that it is not. fitness of spiders that defend territories in areas that are
Mayr asserted that adaptationists claim only that natural hot and have few prey could be compared to the fitness
selection produces the best genotypes possible given the of others that build webs in areas with more favorable
many constraints placed on a population, not that all thermal conditions and features that attract prey. But the
traits are perfect. As we have seen, a trait may not be opti- alternatives are not always this easily identifiable because
mal for several reasons. First, natural selection must act the losers of the competition may be long gone.
on the total phenotype of the individual, which is usually Nonetheless, if we are to demonstrate adaptiveness, we
a mixed bag of traits—some good, some bad—and many must always identify the alternatives from which natural
traits (whether good or bad) are influenced by the same selection had to choose (Dawkins 1986).
sets of genes so that perfection is elusive. Second, natural Testing whether particular traits are adaptive has
selection can act only on the available alternatives. Those stimulated interesting research that might have been
alternatives will depend on the constraints imposed by neglected if one readily assumed the nonadaptiveness of
the population’s evolutionary history and each individual’s traits. Consider, for instance, Niko Tinbergen’s obser-
present conditions—ecological, anatomical, and physio- vation of a seemingly nonadaptive behavior of black-
logical. The “perfect” mutation or allele combination headed gulls (Larus ridibundus). Tinbergen observed that
may not have yet arisen, if it can at all. Finally, natural the gull parent does not immediately remove the broken
selection works in a given environment, and as we have eggshells from the nest. Tinbergen and his colleagues
seen, conditions may vary from place to place or change had already demonstrated that the presence of eggshells
over time. In other words, Mayr argued that Gould and in the nest attracted predators such as herring gulls and
Lewontin were setting up a straw man (an argument that carrion crows. But here was the black-headed gull, sit-
is particularly weak and thus easily countered). Mayr sug- ting for hours among the conspicuous shell fragments.
gested that scientists who study adaptation are (generally) Tinbergen first thought that this delay must be danger-
well aware of other forces that can cause traits to be sub- ous and only explainable as a pleiotropic and nonadap-
optimal, and that adaptationists do not actually hold the tive effect associated with the removal behavior. He
views that Gould and Lewontin ascribed to them. warned, however, that leaping to such conclusions is, in
Nevertheless, the cautionary tale has had an important essence, a refusal to investigate.
influence on evolutionary biologists, including behavior- Tinbergen then observed the black-headed gull
ists, who now pay much more attention to testing adap- colonies more carefully, looking for evidence of the
tive hypotheses instead of simply assuming that a trait has adaptiveness of the delay. He noted that chicks were
a current adaptive function. commonly eaten by neighboring gulls and that canni-
So if we cannot just assume that a given trait is an balistic neighbors took many more chicks that were
adaptation, how do we decide whether or not it is? Let’s newly hatched and still wet than chicks that had dried
look in more detail at funnel-web spiders. Earlier, we saw and become fluffy. Whereas a gull could swallow a wet
that the expression of territoriality in these spiders is well chick within a few seconds, it took about ten minutes to
suited to the existing ecological conditions. We hypoth- down a dry chick. One might imagine the difference as
esized that natural selection was responsible for the similar to trying to swallow a peeled grape as opposed
appropriateness of the behavior and, therefore, that ter- to a cotton ball. Nest-robbing gulls were observed to
ritoriality is an adaptation. For any characteristic to be snatch the wet young within a fraction of a second if the
an adaptation, individuals bearing the trait must leave parent was distracted by a predator. In fact, one chick was
more offspring than those lacking it. Data support the gulped down while its parent was carrying off some
hypothesis that territory quality influences reproductive eggshells. So, Tinbergen deduced that although remov-
success. In a lava bed in central New Mexico, spiders that ing the shells reduced predation by other species of birds,
defend high-quality web sites, those with the best ambi- delaying removal until the chicks were dry decreased the
ent temperatures and prey abundance, have 13 times the likelihood of the chicks’ being cannibalized by neigh-
reproductive potential of their neighbors in poor- boring gulls while their parents were away on their
quality areas (Riechert and Tracy 1975). These data sup- chores (Tinbergen et al. 1962).
port our hypothesis that territoriality is adaptive. In this case, Tinbergen tested his hypothesis by sim-
When we ask questions about the adaptiveness of ple observation, but other research on the adaptiveness
behavior, we are necessarily asking about its value for of behavior incorporates several different approaches—
enhancing reproductive success. The aim in answering experiments, comparative studies, monitoring natural
such a question is to understand why those animals that selection in the field, and mathematical modeling. Each
72 Chapter 4 / Natural Selection and Behavior

of these approaches to testing hypotheses about adapta-

tion involves determining the reproductive success of
different forms of the same trait. The observational
approach compares the observed forms to other (some-
times hypothetical) forms. The comparative approach
compares the behavior of the same or related species in
different environments. The experimental approach
compares different forms of the behavior, and usually
conditions are manipulated and the behavior is observed.
Monitoring selection in the field involves documenting
changes in the frequency of behaviors over time.
Mathematical models are used to understand the logic
of how complex suites of variables influence behaviors
and to compare the potential success of different
strategies under a range of conditions that is difficult to
replicate in the field or laboratory.

THE EXPERIMENTAL APPROACH

We opened the chapter by describing some differences
between cliff-nesting kittiwakes and ground-nesting
gulls that are correlated with the degree of the risk of
FIGURE 4.13 A bittern removing pieces of eggshell from
predation. Cullen (1957a, b) found that kittiwakes, which
its nest. This behavior is typical of many bird species that
have low predation rates, leave eggshell pieces in the rely heavily on nest concealment to reduce predation of
nest, but ground-nesting gulls, which have high preda- the young. The white inner surface of the eggshell may
tion rates, generally remove broken eggshells. This cor- make the nest more noticeable to a predator.
relation prompted Tinbergen to hypothesize that the
survival value of eggshell removal was to reduce preda-
tion on the young (Figure 4.13). However, several other at various distances from some of the nests. The broken
hypotheses are possible: the sharp edges on shells might eggshell bits did attract predators. Furthermore, the risk
injure chicks or interfere with brooding; an empty shell of predation decreased as eggshell pieces were placed at
might slip over an unhatched egg, encasing the chick in increasing distances from the nest. Note, however, that
an impenetrable double layer of shell; or the egg remains this is a test of only one of the possible hypotheses listed
might serve as a breeding ground for parasites, bacteria, above, and other hypotheses may also stand up to testing:
or fungi. these hypotheses are not mutually exclusive.
Tinbergen and his colleagues (1962) tested the Another example of testing multiple competing
hypothesis that eggshell removal reduced predation on hypotheses is the study of the peculiar behavior of snake
chicks. The study began with observation. It was noted scent application (Clucas et al. 2008). California ground
that the eggs, chicks, and nest were camouflaged and squirrels (Spermophilus beecheyi) and rock squirrels
might be difficult for a predator to spot. However, the (S. variegatus) chew the shed skins of rattlesnakes, one of
bright white inner surface of a piece of eggshell might their major predators, and then lick their fur. Clucas and
catch a predator’s eye and reveal the nest site. So the her colleagues considered three hypotheses for this
researchers began by painting some black-headed gull behavior: it may serve as a defense against ectoparasites
eggs white to test the idea that white eggs might be more (e.g., fleas, ticks, or mites); it may distract conspecifics
vulnerable to predators than the naturally camouflaged during aggressive interaction; or it may deter predators.
eggs. Of 68 naturally colored gull eggs, only 13 were Each hypothesis leads to different predictions about which
taken by predators. However, 43 of the 69 white eggs squirrels should engage in this behavior more frequently.
were taken. The difference in predation rates lent cre- Because juveniles have more fleas than adults, the antipar-
dence to the idea that the white inner surface of eggshell asite hypothesis predicts that juveniles should apply snake
pieces might endanger nearby eggs or chicks. Because all scent more frequently. Males engage in aggressive inter-
black-headed gulls remove eggshells, Tinbergen could actions more often than do juveniles and adult females, so
not compare survival rates in natural nests with and with- if distracting conspecifics is the primary function, males
out eggshell pieces. Instead, he created artificial variation should be the most likely to apply snake scent. Finally,
to observe the effects of natural selection. He made his juveniles and adult females are most vulnerable to preda-
own gull nests with eggs and placed white pieces of shell tors, so the antipredator hypothesis predicts that they will
 Testing Hypotheses about Natural Selection and Adaptation 73

be the ones most likely to apply scent. Juveniles and adult in which behavioral and morphological traits evolved.
females were indeed most likely to apply scent, so an For example, males of some species in the swordtail fish
antipredator function seems most likely and can now be genus Xiphophorus have long tailfin extensions called
further tested using controlled experiments. swords, but males in other species do not. Females pre-
fer males with swords. Strangely, even females in species
in which the males have no swords prefer males that have
THE COMPARATIVE APPROACH plastic swords artificially attached. A phylogenetic recon-
The comparative approach to the investigation of struction of the swordtail genus suggested that the
adaptation involves taking into account the evolutionary female’s preference for swords evolved before the sword
relationships among a set of study species. For example, itself: females were predisposed to be attracted to swords
one might study individuals of the same or related even before males evolved them (Basolo 1990, 1995a, b).
species that inhabit different kinds of environments. This sort of study requires a robust phylogeny, which is
These individuals will have inherited some common often difficult to obtain. Phylogenies are hypotheses, too,
genes because they have a common ancestor. But if they and can change depending on the data on which they are
have come to live in different ecological situations, they based. For example, a swordtail phylogeny deduced from
now experience different selection pressures and thus different data had a different shape and suggested a dif-
may have diverged in their traits. Similarly, the converse ferent evolutionary order for these behavioral traits
might also be true—unrelated species that have come to (Meyer 1997), but even further evidence supports the
inhabit the same environment, and thus experience the initial interpretation (Meyer et al. 2006). As more and
same selection pressures, may display similar traits. more data are used to reconstruct a phylogeny, its shape
We have already seen the comparative approach in becomes more stable, along with the behavioral deduc-
action in the herring and kittiwake gull example: these are tions we can make from it.
species that descended from a recent, common ground- Besides determining the relative order in which
nesting ancestor but now live in very different ecologi- phylogenetic traits evolve, we can also use phyloge-
cal situations and have behavioral differences to match netic information to examine the relationship between
their ecological circumstances. However, a comparison of behavior and various ecological variables (such as the
only kittiwake and herring gulls is quite limited: these two risk of predation in the gull example). As we have men-
species essentially provide a single observation of differ- tioned, data from a single pair of species limits the con-
ent environmental effects on the behavior of close rela- fidence we can place in the conclusions that we draw.
tives. The presence of a correlation between the behavior Ideally, it would be best to have multiple species found
and the environmental conditions, however logical, in different environments: for example, the Galapagos
might still be just a coincidence. Larger sample sizes are swallow-tailed gull, which chooses nest sites with char-
needed in order to rule out random chance as an expla- acteristics intermediate between those of kittiwakes
nation. (If you flip a coin once and get “heads,” does that and herring gulls, also shows behavioral patterns that
mean the coin always lands on “heads”?) are intermediate (Hailman 1967). Better still is to
How can the comparative method be used to its best include not just multiple species, but those with mul-
effect? If an animal behaviorist is lucky enough to work tiple evolutionary origins. If a behavior has evolved a
on a taxonomic group that has been well studied by sys- number of times and is significantly correlated with a
tematists, she may be fortunate enough to have a phy- particular ecological context, we can be more confident
logeny available. A phylogeny is an “evolutionary tree” that a common evolutionary explanation exists.
that shows the historical relationships among a group of Examine the phylogenetic trees in Figure 4.14 to bet-
organisms, in particular the order in which different sub- ter understand the role of phylogeny in drawing strong
groups branched off from one another. The reconstruc- conclusions.
tion of a phylogeny is generally done by comparing the
living members of a group (and fossils, where possible)
and grouping together those that share relatively newer Limitations of the Comparative Approach
(derived) traits that differ from the traits of their imme- Although the comparative method can be helpful in the
diate ancestors. Historically, morphological traits were study of adaptation, it must be applied carefully
used to construct phylogenies. Now, however, DNA (Clutton-Brock and Harvey 1979, 1984; Gittleman
sequences are often used. (The methods of phylogenetic 1989). Just as in other evolutionary interpretations, we
reconstruction are worthy of several books on their own, should consider, test, and rule out alternative hypothe-
so we will leave aside the details.) ses. Sometimes this can be done by listing the compet-
Once a phylogeny is established, behavioral traits ing hypotheses and developing predictions for each. The
can be mapped onto it. If the phylogeny is definitive and confirmation of the predictions should lend more weight
detailed enough, it may allow us to determine the order to some hypotheses than to others.
74 Chapter 4 / Natural Selection and Behavior

a W W W WW W W W WC C C C C C CCCC C
FIGURE 4.14 A hypothetical example illustrating the
effect of wise choice of species on our ability to test
hypotheses about the influence of the environment
on traits. Imagine that we are interested in testing
the hypothesis that sleeping in large groups (rather
GENUS A GENUS B
than alone) is associated with cold temperatures.
Pictured are two hypothetical phylogenies that
illustrate the possible relationships among 20
species. The modern-day species are at the “tips”
of the phylogeny at the top of each tree, and their
ancestors are nearer the bottom. Branch points
Sleep in groups
indicate places where lineages diverged. In both
Sleep alone phylogenies, we have 20 species. In each, 10 species
sleep in large groups (indicated by a colored line),
and all of these are found in cold regions (indicated
by C). The remaining 10 species sleep alone (indi-
cated by a black line) and are found in warm
b W C W CC W C W WC W W C W C WC C C C regions (W). From other evidence, we know that
the common ancestor (at the bottom) slept alone
and lived in a cold region. If the phylogeny is as
illustrated in (a), we essentially have a sample size
of one for our test: sleeping in groups evolved only
GENUS A GENUS B once. All the species that sleep in groups are
related, and all the species that sleep alone are
closely related. Thus, it is hard to draw any conclu-
sions about the evolutionary origin of sleeping in
groups. In (b), sleeping in groups evolved eight
times, so our sample size for our test of the corre-
lation between sleeping in groups and temperature
is eight. Here we can be more confident that tem-
perature and sleeping in groups are related.
(Modified from Rosenheim 1993.)

We should also keep in mind that correlations lation does not answer the question. As with any corre-
between traits and ecological variables, such as those lational study, it is also important to remember that a
shown in Figure 4.14, identified by the comparative third, unmeasured variable may be the underlying
approach, do not prove there is a common cause. connection between the variables under study.
Consider the difficulties in determining whether the diet In some cases, the lack of a pattern between traits
was a cause or an effect of sociality in John Crook’s and ecological variables helps to eliminate a hypothesis.
(1964) comparative study of over 90 species of weaver For example, thermoregulation is one hypothesis
birds (Ploceinae). He noticed that species living in the for- explaining why birds often roost in groups at night. By
est generally eat insects and forage alone. In contrast, huddling together as they sleep, they may conserve body
species inhabiting the savannah eat seeds and feed in warmth. This hypothesis predicts that species that spend
flocks (Figure 4.15). Crook identified correlations more time in cold areas and species that have lower body
between the degree of sociality and two factors—diet and masses would be especially likely to need this ther-
predation. But which is cause and which is effect? Seeds moregulatory boost. A phylogenetically based study that
often have a patchy distribution, and groups of foragers examined species from many distantly related groups of
are more likely to find a patch that can feed them all. birds did not find this pattern, so the thermoregulatory
One might infer, therefore, that diet is the cause of flock- hypothesis is not supported (Beauchamp 1999). As with
ing in weaver birds. However, living in groups is also a any study where we accept the null hypothesis, consid-
good defense against predators, and after animals began erations of statistical power are key: in this case, the sam-
living in groups, seeds may have been the only food ple size is large enough to provide confidence that we
source that could supply enough food for an entire flock. would have been able to detect a biologically meaningful
So, was sociality a cause or an effect of diet? The corre- pattern if one had existed.
 Testing Hypotheses about Natural Selection and Adaptation 75

FIGURE 4.15 Weaver birds. John Crook

compared the diet and degree of sociality
of weaver birds that inhabit different
environments. He observed that solitary
species, such as Ploceus nelicouri (shown on
the left; nest not in proportion), are usually
insectivores that defend large territories in
the forest. In contrast, social species, such
as P. phillipinus (shown on the right), eat
seeds and live in the open savannah.

MONITORING SELECTION they identify factors that should be considered in the

IN THE FIELD decision. But the decision may still be difficult because it
may require integration of concerns along different
It’s relatively straightforward to measure evolution over dimensions. For instance, if you were deciding whether to
generations in the laboratory (although even that can be get your own apartment or live with your parents, you
extremely time consuming), but it is much, much more might have to weigh factors that represent two dimen-
difficult to do so in the field. We’ve already mentioned sions—money (the difference in rent) against freedom
the variation in beak size and shape across Darwin’s finch (gained by being on your own). These different factors
species on the Galapagos Islands. Beaks of these closely might seem to be apples and oranges, and so they might
related species range from robust to slender, according be difficult to compare.
to the kind of food they eat. In most such cases of adap- A similar example for animals is the decision of
tive radiation (the evolution of an ancestral species into whether to stay in a safe place where there is not much
many different ecological niches), we must rely on com- to eat or whether to go out to forage in a place where
parative analyses in order to trace the pattern that evo- there is abundant food but where dangerous predators
lution has taken. In finches, we can see natural selection might lurk. How do we decide which of these strategies
in action. In an intensive 30-year-long field study, is optimal under current conditions? As the number of
Rosemary and Peter Grant and their colleagues docu- influences on a behavior increases, it can be increasingly
mented changes in beak size in medium ground finches difficult to identify the best solution.
(Geospiza fortis) in response to the environment (Boag Animal behaviorists often use optimality modeling to
and Grant 1981; Price et al. 1984). During periods of weigh the pros and cons or, to use the proper jargon, the
drought, medium ground finches that had deeper beaks costs and benefits, of each available strategy. A model is a
were better able to eat the hard seeds available for food. mathematical expression of the costs and benefits of each
These birds produced more offspring. Because beak size strategy. First, all costs and benefits are translated into com-
has a genetic component, the offspring also had deeper mon units that represent a measure of fitness. This com-
beaks and the population mean shifted. During rainy mon measure of fitness is called “currency” (a holdover
years, many more small seeds were available; birds with from economic theory, where modeling approaches now
smaller bills had the advantage, shifting the population used in animal behavior were first developed), and it allows
mean back. The finch work is beautifully summarized in different strategies to be compared. The optimal strategy
Jonathan Weiner’s (1995) book The Beak of the Finch: A is the behavioral alternative that maximizes the difference
Story of Evolution in Our Time, winner of a Pulitzer Prize. between the costs and benefits. In economic terms, the
alternative that maximizes the difference between costs and
benefits is the one that yields the greatest profit. In evolu-
MODELING THE COSTS
tionary terms, this is the choice that maximizes fitness. If
AND BENEFITS OF TRAITS
a behavior is at least in part genetically based, the success-
Most actions have costs and benefits. When considering ful alternative is the one that would contribute most to the
the best course of action, people often create lists of the next generation. We’ll explore some detailed examples of
pros and cons of each strategy. Lists are helpful because mathematical models in later chapters.
76 Chapter 4 / Natural Selection and Behavior

Measuring costs and benefits in terms of fitness is rules may generate the same result—but it gives us a
extremely difficult to do empirically. It is generally logically coherent hypothesis to test against data.
impossible, especially for long-lived species, to follow
individuals in the field over the course of their lives and
to measure their reproductive success. Instead, SUMMARY
researchers often measure a behavior that they assume
correlates with fitness. For example, the rate at which ani- Animals frequently appear to match their environments
mals acquire calories has been used as a way to compare very well, and natural selection is the process that cre-
different foraging strategies. The assumption is that the ates this match. Natural selection occurs when there is
faster the rate of energy intake, the higher the fitness. Of phenotypic variation in a population, the variation has
course, this assumption is very much an oversimplifica- at least some genetic basis, and some of these inherited
tion: surely an animal that is intent on eating as fast as traits improve their owners’ chance of leaving more off-
possible may miss an approaching predator. Therefore, spring. When these conditions are present, the allele fre-
in constructing a model, it is important to think carefully quencies in a population change over generations, and
about what an animal faces during the time period under evolution by natural selection occurs.
consideration—is it really free to focus on foraging, or In spite of the power of natural selection to shape
should it take into account predation risk as well? If the animals and their behavior, we cannot assume that organ-
latter is true, then predation risk should be included as a isms are perfectly adapted to their environment. Animals
variable in the model. Once all the variables are identi- may migrate into a population from nearby areas, bring-
fied and included, the model can be used to identify the ing their genes with them. In small populations, genetic
optimal behaviors under different conditions, such as drift (changes in allele frequencies by chance alone)
high prey abundance and low predation risk. These pre- becomes increasingly important. Phenotypic traits are
dictions can then be tested using experiments in which often correlated with one another, so selection on one
conditions are manipulated. trait might drag along another trait. Selection pressures
This may sound like complex, higher-level thinking also change over time, and evolution lags behind: the
that is above and beyond the capabilities of most animals. genetic makeup of today’s populations results from selec-
Surely a bumblebee does not calculate the rate of nec- tion on previous generations.
tar production by different flower species, the density of Selection pressure on a particular genotype may
the flowers, and the energy it uses in flying from plant depend on its frequency in the population: for example,
to plant before choosing where to forage. In using lan- predators may preferentially feed on the most common
guage like “decision rules,” behaviorists are not imply- prey, driving down its numbers, and then switch to an alter-
ing that animals make conscious decisions or “think native prey whose numbers are on the rise. The success of
things through” to find the optimal course of action. a particular genotype may also depend on the behavior of
Instead, we assume that natural selection has shaped other members of the population. An evolutionarily stable
behavior over generations so that the animal responds strategy, or ESS, is one that cannot be beaten by another.
appropriately under a particular set of circumstances. An ESS might be a single strategy or a combination of
What may appear to be very complex behavior may strategies, each played with a particular frequency.
come down to following a fairly simple strategy. How are hypotheses about the adaptive nature of
Models can give us insight into how well simple behavior tested? Several approaches are available to us.
behavioral rules can generate complex-looking behavior. We can conduct experiments to measure the present-day
For example, in one model, researchers defined a habi- costs and benefits of particular behaviors. This is prob-
tat grid and added simulated animals. The animals were ably the most common approach taken by modern-day
allowed to move around the grid as if they were playing behavioral ecologists. If we know something about the
a video game. Given just a few rules to follow, such as phylogenetic relationships among a group of species, we
“Fight other animals that you encounter” and “Reduce can use the comparative method to tease apart the evo-
the probability of returning to places where you have had lutionary history of a particular behavioral trait. In rare
a fight,” the animals settled into a pattern that looks cases, we can monitor populations in the field and actu-
remarkably like territoriality: each stayed in its own area ally see natural selection at work. Finally, we can use
and defended it against other animals (Stamps and mathematical techniques to model the costs and bene-
Krishnan 2001). This sort of result does not prove, of fits of behavior and compare the value of different
course, that these are the rules that animals follow—other strategies.
 5
Learning and Cognition

Definition of Learning leaves off, carry it to a termite mound, and insert the
Types of Learning stick into the hole on top. It then draws out the stick,
now covered with swarming termites, and licks the ter-
Habituation
mites off. Does it seem to you that the chimp planned
Classical Conditioning
its actions? When it first broke off the twig, did it under-
Operant Conditioning
stand that it would result in a snack? Now imagine
Latent Learning watching an ant lion, a little larval insect, dig a pit in the
Social Learning sand. The ant lion waits in the bottom of the pit. Finally,
Species Differences in Learning: an ant slips over the edge and begins to slide toward the
Comparative Studies ant lion’s waiting jaws. The ant tries to scramble back
The Ability to Learn as a Heritable Trait out, and the ant lion hurls sand upward. The ant is
Evolution and the Variation in Learning Across Species knocked back into the pit, and the ant lion feeds. What
Other Evidence of Cognitive Abilities in Animals do you think the ant lion understands about its own
behavior? Did it anticipate catching ants when it dug its
Tool Use
pit? Would your opinion change if you knew that ant
Detours
lions can learn to associate a human-made vibratory cue
Understanding Numbers and Other Abstract Concepts
with the presentation of food (Guillette et al. 2009)?
Self-Recognition and Perspective Taking Most of us would attribute more cognitive abilities to the
chimp than the insect, but are the underlying mecha-
Most people have presuppositions about the mental lives nisms really different? How might we critically examine
of animals. We tend to discount the abilities of some these questions? In this chapter, we will discuss how we
species, especially those very different from us, but we know what animals know: how they learn, why species
anthropomorphize our pets and other primates and differ, how we can test their ability to solve challenging
assume that they think like we do. For example, imag- cognitive problems, and even how they view themselves
ine watching a chimpanzee break off a twig, strip its and others.

77
78 Chapter 5 / Learning and Cognition

Researchers classify patterns of learning into differ-

DEFINITION OF LEARNING ent categories, but keep in mind that the relationships
among them may be more complex than they first
Learning has proved so tricky to define that one textbook
appear—they may overlap, and the distinctions between
on the topic begins by defining it as “a term devised to
them may not be clear-cut. In fact, researchers do not
embarrass learning psychologists, who tie themselves in
all agree on the nature of the categories or how many
knots trying to define it” (Lieberman 1993). For exam-
there should be. We’ll postpone discussion of several
ple, we could say that learning is “a process through
types of learning until later chapters. For example,
which experience changes an individual’s behavior.” This
imprinting and song learning, which generally occur
certainly encompasses what we usually think of as learn-
early in life, are discussed in the chapter on the devel-
ing, but it also includes phenomena that we would not
opment of behavior (Chapter 8).
call learning. For instance, an athlete might run the last
mile of a marathon at a slower speed than she ran the
first mile. This is the result of the experience of running,
but it isn’t learning. Similarly, a man entering a dark TYPES OF LEARNING
movie theater from the sunlit street may inadvertently
HABITUATION
step on a discarded candy box on the floor because the
photoreceptors in his eyes have not yet adapted to the We usually think of learning as resulting in the expansion
dim light. In a few minutes, his eyes adjust and he steps of an individual’s behavioral repertoire—perhaps learning
over the trash. Thus, sensory adaptation can also lead to a new skill or a new association. However, in habituation,
a change in behavior, but we wouldn’t want to call this the animal learns not to respond to a particular stimulus
learning either. because the stimulus has proven to be harmless. A bird
Let’s expand our definition: learning is a change in must learn not to fly away every time the wind rustles the
our capacity for behavior as a result of experience, exclud- leaves. Habituation can be defined more precisely as the
ing the effects of fatigue, sensory adaptation, or matura- waning of a response after repeated presentation of a stim-
tion of the nervous system (Hinde 1970). “Experience” ulus. Once habituation occurs, its effects are long lasting.
includes exposure to particular combinations of environ- Habituation is everywhere, from unicellular protozoans to
mental stimuli, as well as practicing a behavior. humans (Wyers et al. 1973). It is generally considered to
Complicating matters further is that behavioral be the simplest form of learning.
changes that result from learning are not always A classic example that illustrates the essential char-
expressed immediately. For example, a student may not acteristics of habituation is the clamworm, Nereis pelagica.
demonstrate that she has learned course material until This marine polychaete lives in underwater tube-shaped
the day of the exam. In addition, the learned behavior burrows it constructs out of mud. It filters tiny bits of
may not be consistently expressed every time the food from the water. When it feeds, it partially emerges
opportunity presents itself—sometimes the student from its tube. However, it withdraws quickly back into
may be able to articulate what she has learned, and the safety of the tube when it senses sudden stimuli such
sometimes she may not. Thus, the change in behavior as a shadow that could signal the approach of a predator.
that results from learning is perhaps more accurately Clark (1960) kept clamworms in shallow pans of
described as a change in the probability that a certain water in the laboratory. When he passed a shadow over
behavior will occur. them, they withdrew into their tubes. The second time he
This discussion should make it clear that we can’t presented the shadow, slightly fewer worms responded.
necessarily know whether an animal has learned some- The third presentation elicited even fewer withdrawals. As
thing just by seeing a change in its behavior. To be shown in Figure 5.1, subsequent presentations resulted in
absolutely sure that learning has occurred, we must a continued decline in escape responses. The clamworms
manipulate the experiences of different groups of ani- had habituated, and the effects of habituation lasted for
mals and then compare their performance on the same several hours. The clamworms’ decline in responsiveness
test (Shettleworth 1998). Because of this need for care- was not because the sense organs became adapted to the
fully controlled experiments, the vast majority of work stimulus, because sensory adaptation occurs much more
on learning has been done in laboratory settings. quickly than this. Nor was the decline due to muscle
However, as we will see, we can still address evolution- fatigue, because habituated worms still withdrew in
arily important questions, and we can even do some field response to prodding. The clamworms had learned to
studies. The modern study of learning incorporates stop responding to the shadow.
both proximate and ultimate questions, all the way from A characteristic of habituation is that it is specific to
the cellular level to the phylogenetic level, where we a particular stimulus. For instance, young turkeys,
take into account evolutionary relationships among chickens, and pheasants innately show antipredator
species. behaviors, such as crouching and giving alarm calls, at
 Types of Learning 79

Shadow Mechanical shock Shadow sun as it undulates with the waves. In this case, respond-
100 (Group A) ing to the shadow every time it appears would mean that
80 the clamworm loses opportunities to feed. Unnecessary
60 responses also waste energy. Habituation, like other
40 kinds of learning, thus focuses attention and energy on
Worms reacting, percent

20 the important aspects of the environment (Leibrecht and

0
Askew 1980).
0 10 20 30 40 50 60 Habituation has been documented in nearly every
Trials species that has been tested, but it may vary in its details
Shadow Rest Shadow in ways that make sense given the habitat of the species.
100 (Group B) For instance, consider two closely related species of
80 crabs. On one hand, Chasmagnathus is a semiterrestrial
60 crab that lives on the mudflats amid patches of cord grass
40
along the coast of South America. On the other hand,
Interval of 40 minutes Pachygrapsus inhabits the rocky intertidal zone. Crabs of
20
both species begin to run when a shadow passes over-
0 head, and both habituate to repeated presentation of
10 10
Trials Trials shadows. However, habituation lasts much longer in
Chasmagnathus. This makes sense because the wind-
FIGURE 5.1 Habituation of the withdrawal response to a blown grass of the habitat of Chasmagnathus casts many
shadow by the clamworm, Nereis. In habituation, the harmless moving shadows, but shadows on the bare
simplest form of learning, the animal learns not to rocks of the habitat of Pachygrapsus are more likely to sig-
respond to frequently encountered stimuli that are not nal the arrival of a predator (Tomsic et al. 1993).
associated with reward or punishment. Both groups of
Habituation may also occur in the context of inter-
worms habituated to the shadow during trials 1–10. A
mechanical shock was then administered to group A at
actions within species. For instance, animals that defend
1-minute intervals. The worms initially responded to territories encounter their next-door neighbors time and
the new stimulus, showing that the loss of the response again. Over time, many species reduce their aggressive
to the shadow was not a result of fatigue. Group B responses directed toward these familiar neighbors:
rested for 40 minutes while group A received the there is little point to fighting day after day over a bound-
mechanical shocks. Both groups were equally responsive ary that has already been settled. However, unfamiliar
to the shadow during the last 10 stimulus presentations. intruders will still provoke a territory holder to attack.
Thus, habituation was specific to a stimulus. (Modified This phenomenon can be nicely demonstrated in species
from R. B. Clark 1960.) with acoustic communication by playing back recorded
calls. A number of bird species (e.g., Falls 1982) as well
as bullfrogs (Davis 1987) respond aggressively to a play-
the sight of objects moving overhead. The chicks initially back of a stranger’s call but not to that of a familiar call.
respond to a great variety of objects, only a few of which One hypothesis is that habituation is the mechanism that
are dangerous. By the time they are adults, the birds mediates this process: perhaps frogs stop responding
respond only to the image of a predator such as a hawk aggressively to a call when they hear it repeatedly (Peeke
flying overhead. Schleidt (1961a, b) tested whether 1984). Bee and Gerhardt (2001) created a “new neighbor”
habituation could underlie the development of the speci- bullfrog by synthesizing a new call and playing it back
ficity of these responses. He showed that models of var- from a previously unoccupied territory. Initially, this mys-
ious shapes—a gooselike silhouette, and even a circle and terious new voice elicited quite a response: male bullfrogs
a square—all effectively elicited alarm calls from young called back and charged at the speaker. After repeated
turkey chicks during the initial two days of testing. presentations, their aggressiveness declined. Because this
When any one of these models was presented frequently, decline carried over between nights and was specific to
it elicited fewer and fewer alarm calls. Chicks still called particular characteristics of the calls, it meets the criteria
in response to stimuli they encountered only occasionally. of “relatively permanent” and “specific to a stimulus.”

The Adaptive Value of Habituation

Obviously, it is important for a clamworm to withdraw STOP AND THINK
to the safety of its burrow when a shadow is that of a Can you think of other hypotheses, besides habituation,
predator. However, a recurring shadow that is not followed explaining why animals might stop responding to
by an attack is more likely caused by something harmless, neighbors’ calls?
perhaps a patch of algae that is repeatedly blocking the
80 Chapter 5 / Learning and Cognition

Habituation as an Experimental Tool UR. The new connection between the US and CS is
Habituation is a very useful tool for the study of cogni- called a conditioned reflex.
tive processes in animals, including humans. In a typical Over thousands of controlled studies, researchers
protocol, a subject is habituated to a stimulus, and then have found some remarkably consistent characteristics of
a new stimulus is presented. If the subject’s response classical conditioning. These general features hold
changes, the experimenter knows it can detect the dif- across a wide range of stimuli, as well as across species.
ference between the two stimuli. This is especially use- Next, we’ll describe three of the most important char-
ful in the study of cognition in infants, where it’s not acteristics of classical conditioning.
always easy to determine what the infant perceives. First, the order of the presentation of the US and
CS is important. Conditioning is most effective when
the CS (such as a tone) precedes the US (such as food).
CLASSICAL CONDITIONING The CS serves as a signal that the US will appear; a cue
is of little value if it occurs after the fact. Also, the two
In associative learning, some sort of a mental connec- stimuli must occur fairly close together if an association
tion is formed between representations of two stimuli between them is to be made. Thus, if you want your new
(Shettleworth 1998). The first type of associative learn- love to associate you with something nice, stand on the
ing we will examine is classical conditioning. For many doorstep with flowers hidden behind your back. Right
people, the phrases “classical conditioning” and “Pavlov’s after the object of your affection sees your smiling face,
dogs” are intertwined. Pavlov first laid out the principles present the flowers.
of classical conditioning in 1927. Pavlov was a Russian A second characteristic arises from the fact that use-
physiologist whose main interest was not learning but ful signals are reliable: they predict that a particular event
digestion. He wondered why a dog salivates at the antic- or stimulus will follow. A signal is useless if it merely
ipation of food rather than just its presence. He hypothe- indicates that any one of a dozen events may follow.
sized that the animal had made a connection between the Therefore, it should not be surprising that for classical
sight or smell of food and the food itself, and he became conditioning to be most effective, the CS must precede
interested in exactly how dogs made these associations. the US more often than it does other stimuli (Rescorla
To measure saliva, Pavlov made a small opening, or 1988a, b).
fistula, in the dog’s cheek so that the saliva would drain Finally, after an association between a CS and US is
into a funnel outside the dog’s body. The hungry dog was formed, it can be lost again. If the CS is no longer reli-
harnessed into position on a stand and then presented able because it is presented frequently without being fol-
with various stimuli. As expected, the dog salivated when lowed by the US (for example, if a tone is given time and
powdered food was blown into its mouth. In contrast, it again with no food), the subject stops responding to the
did not salivate when it heard the sound of a bell. Then tone. The loss of the conditioned response is called
Pavlov began pairing these two stimuli: immediately extinction. Thus, it is important to remember to con-
before food powder was presented, the bell sounded. tinue to bring flowers to your love! Of course, in a
Pavlov presented these paired stimuli repeatedly at inter- changing environment, it is fortunate that learned
vals over several days. After 30 presentations, the dog responses can be extinguished.
salivated in response to the bell alone. As trials contin-
ued, the dog salivated more profusely and responded
more quickly to the bell (Pavlov 1927). The Adaptive Value of Classical Conditioning
Let’s phrase these results in more general terms. To Pavlov (1927) suggested, and other researchers agree,
begin, an animal has a particular inborn response to a that learning through classical conditioning is likely to
certain stimulus. This is called the unconditioned provide fitness advantages to wild animals. However,
stimulus (US) because the animal did not have to learn most studies of classical conditioning over the decades
the response to it. In Pavlov’s study, the US is food. The have focused on determining the rules under which it
response to the US is called the unconditioned functions. Most effort has been channeled toward char-
response (UR). In Pavlov’s study, the UR is salivation. acterizing the process of conditioning, such as the most
A second stimulus is paired repeatedly with the US until effective interval between the conditioned and uncondi-
eventually it, too, is able to elicit the response. At this tioned stimuli and the time course of extinction.
point, the new stimulus is called the conditioned Relatively few studies have addressed the potential value
stimulus (CS) because the animal’s response has become of classical conditioning in the everyday life of an ani-
conditional upon its presentation; here, the CS is the mal. We’ll discuss three of them here. Our first example
bell. The response to the conditioned stimulus is called comes from Karen Hollis (1984, 1999) on territorial and
the conditioned response (CR). The conditioned reproductive behaviors in blue gouramis (Trichogaster
response may differ slightly from the unconditioned trichopterus), fish that inhabit shallow pools and streams
response. In Pavlov’s study, the CR is salivation, like the in Africa and Southeast Asia.
 Types of Learning 81

A male blue gourami defends its territory with an der with their fins already erect. During the ensuing
aggressive display: it swims rapidly toward the intruding fights, they delivered significantly more tailbeats and
fish, with all fins erect. If the intruder does not respond bites than did their competitors (Hollis 1984). This
with a submissive posture or retreat, the contest escalates response may have been mediated through hormones: in
into a heated battle that can result in serious injury. The conditioned males, the presentation of the light led to
males bite each other and flip their tails to beat water an increase in androgens, male sex hormones known to
against the opponent’s sensitive lateral line organ. The heighten aggressiveness in many species of vertebrates
lateral line is a row of receptors running in a line down (Hollis 1990; see Chapter 7).
the side of fish that detect movement and vibration in the The conditioned male gains a long-term benefit in
water. Dangerous fights such as these gourami fights are addition to his immediate competitive edge: the experi-
most likely to evolve when the value of the resource is ence of winning increases the probability of winning
great. Success is crucial for male gouramis because again in battles with new opponents. Thus, conditioned
females rarely mate with a male without a territory. males not only win the first battle but are likely to keep
If a male were to learn the signals that indicate the winning. In contrast, fish that lose their first battle are
approach of a rival—perhaps visual, chemical, or likely to lose later battles as well. In one experiment, all
mechanical cues—he might be better prepared for bat- fish that lost the first battle also lost the second one
tle and gain a competitive edge. Hollis selected pairs of (Hollis et al. 1995). Many species, both vertebrates and
male fish with similar body size and aggression levels, invertebrates, show a similar “winner effect.”
and placed them on opposite sides of a divider in an Male blue gouramis that successfully defend a ter-
aquarium. For one member of each pair, a 10-second ritory are more likely to attract females, but excessive
light (the CS) preceded a 15-second viewing of a rival aggressiveness could actually harm mating success. A ter-
(the US). As a control, the CS and US were also shown ritorial male is likely to attack all visitors to his territory,
to the other member of the pair, but their presentations even females. If a male is conditioned with a light signal
were not paired: they occurred randomly with respect to to expect the arrival of a female, he is less likely to attack
one another. During the test trials, the light signal was her when she appears (Hollis et al. 1997). As can be seen
given, and then the barrier that separated the fish was in Figure 5.2, conditioned males bit females fewer times
removed, allowing them to interact. The males that had than did unconditioned males. These conditioned males
been classically conditioned to associate the light with also spent more time building a nest. The shift in
the imminent appearance of a rival were superior in behavior from aggressive to reproductive activities paid
territorial defense. They approached the territorial bor- off in reproductive success. Conditioned males spawned

a Biting b Nest building

35 300
Mean total number of responses

Mean total number of bouts

30 250
25
200
20
150
15
100
10

5 50

0 0
Conditioned Unconditioned Conditioned Unconditioned

c Spawning d Reproduction
150 FIGURE 5.2 Male blue gourami fish
Mean latency to spawn (in hours)

1500 learned, through classical conditioning,

125
Mean number of young

that a light signaled the imminent appear-

1200 ance of a receptive female. Conditioned
100
1000
males had greater reproductive success
75 than unconditioned males. Conditioned
750 males (a) bit approaching females less
50 frequently, (b) spent more time building
500
nests, (c) were quicker to spawn, and
25 250 (d) produced more young than did
0 0 unconditioned males. (Data from Hollis
Conditioned Unconditioned Conditioned Unconditioned et al. 1997.)
82 Chapter 5 / Learning and Cognition

more quickly and produced more fry than did uncondi- Treatment 1
tioned males.
In nature, of course, flashing lights are unlikely cues.
However, many natural signals are likely to be available
to be learned (Hollis 1999). For example, territorial Male present No male present
invaders might be seen, heard, or scented by a territor-
ial holder. The shape of a gravid belly might reliably
indicate a willing female’s approach.
Let’s look at another taxon in which classical condi-
tioning functions in an evolutionarily relevant context.
Subject Subject
Male field crickets (Gryllus bimaculatus) mate by trans-
ferring a sperm in a packet called a spermatophore.
Because spermatophores are costly to produce, males
would do best not to transfer larger spermatophores than
necessary. However, when females mate more than once, Treatment 2
the sperm from different mates compete inside her body
to fertilize her eggs. Thus, if a male faces another male
in a competition for a female’s attentions, he will increase
Male present No male present
his chances of fathering more offspring by transferring
a larger spermatophore and thus more sperm to the
female (Mallard and Barnard 2003). Interestingly, males
can learn to associate environmental cues with the pres-
ence of male competitors (Lyons and Barnard 2006).
The experimental design is illustrated in Figure 5.3.
Subject Subject
Males were placed in one side of a terrarium along with
topographical cues: either two or four Lego bricks. The
other side of the terrarium, visible through a clear wall,
either held a potential competitor or was empty. Each
subject male had four opportunities to mate during train- FIGURE 5.3 Field crickets learned about spatial cues that
ing. For any particular subject, a particular quantity of signaled the presence of a competitor. Males were
placed on one side of a terrarium and either two or four
bricks (two or four) was always associated with the pres-
Lego bricks. For males in Treatment 1, four bricks were
ence of a competitor, while the other quantity of bricks
always associated with the presence of a male competi-
never was. Thus, males had a chance to learn that a par- tor on the other side of the partition, and two bricks
ticular number of bricks signaled the presence of a com- were never associated with a competitor. For males in
petitor. After training, the males were allowed to mate Treatment 2, the situation was reversed. After training,
next to each arrangement of bricks, with no competitors subjects were allowed to mate next to each arrangement
present. Males produced larger spermatophores in the of bricks, with no competitors present. Males produced
environment that, for them, had been associated with a larger sperm packets in the environment associated with
competitor. Odor cues (peppermint- and vanilla-scented a competitor. (From Lyons and Barnard 2006.)
oils from a cosmetics store) could also serve as signals.
When tested with scents that had been associated with
the presence of a competitor, males again produced are held in tubes (Takeda 1961). When the antennae of
larger spermatophores. a bee are touched with a sucrose solution (the US), the
Our final example of classical conditioning in an eco- bee extends its proboscis to lick it (the UR). When an
logical context concerns feeding behaviors. A wonderful odor is presented just before the sucrose solution is pre-
model system is the honeybee. A foraging bee has a lot sented, the bee rapidly forms an association between the
to learn. Flowers bloom and fade, so the best places to odor (the CS) and the sucrose, and begins to extend its
forage are constantly changing. Individual flowers can proboscis to the odor alone (the CR). This easy-to-use
vary in the amount and quality of nectar and pollen they protocol has enabled researchers to test numerous
produce, as well as whether they’ve recently been visited hypotheses about learning, including how bees general-
by another forager. Flower species also vary in their ize from one stimulus to others (Menzel 2007).
shape, and bees must access each flower shape differently.
Location, color, shape, pattern, texture, and odor of flow-
OPERANT CONDITIONING
ers are all characteristics that bees learn about. For exam-
ple, bees can be rapidly conditioned to respond to odor. Another form of associative learning is operant (or
Carefully controlled odor cues are presented to bees that instrumental) conditioning. The formal study of this
 Types of Learning 83

topic was begun by Thorndike, who invented a “puzzle A stimulus, such as a bit of food, that changes the
box” with a door that could be opened with a latch on probability that an animal will repeat its behavior is
the inside. We already described this box in Chapter 2, called a reinforcer. In the experiments described so far,
but will review it briefly here. Thorndike would place a positive reinforcers were used. A positive reinforcer is
hungry cat in the box with a tempting bowl of food out- one that increases the probability that a behavior will be
side the box. The cat would leap around in an attempt repeated, such as food offered to a hungry rat or a drink
to get to the food. Eventually, the cat would accidentally to a thirsty one. The definition of a negative reinforcer
hit the lever in the correct way, the door would open, and may seem counterintuitive: it increases the probability of
the cat would get to eat. Thorndike would then scoop a response once it is removed. If an unpleasant or painful
up the cat and pop it back in the box again. Over suc- stimulus stops when an animal performs a certain act, it
cessive trials, a typical cat would get faster and faster at is likely to repeat that action. For example, a rat will learn
performing the correct behavior to release the latch. to push a bar to turn off a bright electric light for 60 sec-
This type of learning is called operant conditioning to onds (Keller 1941). Negative reinforcement is thus dif-
emphasize that the animal operates on the environment ferent from punishment, which is an aversive stimulus
to produce consequences. It is also called trial-and-error that results in a decrease in a response.
learning. In operant conditioning, as in classical conditioning,
B. F. Skinner later invented the Skinner box, an the timing of events is critical. When the animal spon-
apparatus that was even easier to use than Thorndike’s taneously performs a behavior, reinforcement must fol-
puzzle box, and is still used today. A hungry animal is low closely. In a sense, a cause-and-effect relationship
placed in the Skinner box, where it must learn to manip- develops between the performance of the act and the
ulate a mechanism (such as pressing a lever or pecking delivery of the reinforcer. When reinforcement is with-
a key) in order to get a food reward (Figure 5.4). Data held, the response rate will gradually decline and
collection (number of bar or key presses) is generally become extinguished, just as the strength of the condi-
automated and very rapid. tioned reflex decreases when the CS is presented many
times without the US.

Shaping
Operant conditioning can be used to teach animals to
perform novel and sometimes complex acts. Hollywood
animal trainers rely on a method called shaping, which
has parallels to the gradual way in which a sculptor molds
a lump of clay (Skinner 1953). At first, the trainer rein-
forces any gross approximation of the desired act but
then requires better and better performances to get a
reward. For example, to train a dolphin to jump from the
water through a hoop, first reward it for approaching the
hoop. When it learns to approach, reward it only when
it swims through the hoop. Then raise the hoop on suc-
cessive trials until it is clear of the water, and offer your
dolphin a fish only when it makes the leap.
Shaping works on people as well as other animals.
When a writer, Amy Sutherland, was researching animal
training techniques for a book, it struck her that many
of the techniques might be useful in her marriage. “After
all,” she writes, “you don’t get a sea lion to balance a ball
on the end of its nose by nagging.” She quietly began to
use shaping techniques to train her husband not to throw
laundry on the floor and to change some of his other
annoying habits. Eventually, she couldn’t resist explain-
ing what she was doing—it worked very well!— and he
FIGURE 5.4 A rat in a Skinner box. The hungry animal
explores the box and eventually presses the bar. This began using the same techniques on her. Sutherland
automatically results in the delivery of a small food pel- described her experiences in a humorous piece in the
let that the rat quickly consumes. The food reward New York Times (“What Shamu Taught Me About a
increases the probability that the rat will press the Happy Marriage”). Apparently the piece resonated with
bar again. readers: it was the paper’s most emailed article in 2006.
84 Chapter 5 / Learning and Cognition

Reinforcement Schedules for a few days. The other mouse had no experience in
In real life, reward seldom follows every performance of the room. On 13 of the 17 trials, the owl caught one of
an act. Instead, the reward is usually intermittent. For the mice. Only two of the captured mice were from the
example, a honeybee will find nectar rewards in a flower group that was familiar with the room, suggesting that
only if the flower hasn’t been recently visited by another their knowledge of the environment helped them evade
bee. The frequency with which rewards are offered is the predator.
called the reinforcement schedule. Partial reinforcement Even ants seem to be able to gather information for
schedules vary either the ratio of nonreinforced to rein- later use. Temnothorax albipennis ants build nests in flat
forced response or the time period between successive rock crevices. When their nests are damaged, they have
reinforcements. Alternatively, rewards may be doled out to move to a new area. Like apartment hunters, ants eval-
in no particular pattern (Ferster and Skinner 1957). uate prospective nest sites based on a range of criteria:
Each reinforcement schedule has predictable effects floor area, headroom, entrance size, darkness, hygiene,
on the rate of response and on how long the animal will and the proximity of hostile neighbors (reviewed in
continue responding when it is no longer rewarded. We Franks et al. 2007). All this evaluation takes time, and if
will highlight just a few examples from an extensive body their nest is destroyed, the ants must find a new home
of work. A continuous reinforcement schedule, in very quickly. Researchers tested whether ants keep track
which each occurrence of the behavior is rewarded, is of the local housing options even before they need to
best during the initial training to establish and shape a move. Ants turn out to be quite content with nest sites
response. A fixed ratio schedule, one in which the ani- made of cardboard sandwiched between glass slides, so
mal must respond a set number of times before rein- it is easy to design laboratory experiments, as shown in
forcement is given, usually results in very high response Figure 5.5. Researchers placed a new nest site,
rates because the individual has control over how quickly Alternative #1, near the ants’ current nest for a week so
it will be rewarded. The faster it responds, the sooner it that the ants could become familiar with it. Next, the
completes the number of responses required to receive researchers introduced a second nest of exactly the same
the reward. A fixed ratio reinforcement schedule is sim- quality, Alternative #2. They then immediately destroyed
ilar to piecework in factories, in which the employee gets the ants’ current nest and forced them to move. Any dif-
paid when a certain number of items are completed. ference in whether ants favored Alternative #1 or #2
Employers like the system because of the very high pro- would suggest that they had learned something about
duction rate it generates. In a variable ratio schedule, Alternative #1 during the week of reconnaissance.
the number of responses required for reinforcement In the first experiment, the ants’ initial nest was of
varies randomly. This also generates very high response high quality, but both Alternatives #1 and #2 were of low
rates because the individual is rewarded for fast quality. Of 30 colonies tested, only two chose Alternative
responses. The variability means that there aren’t #1, 23 chose Alternative #2, and 5 were split in their
detectable patterns of reinforcement, so the subject is
unable to discern immediately when reinforcement has
Alternative Alternative
stopped. Thus, the response tends to persist even if the nest #1 nest #2
reward is withheld for a while. This is exactly the behav-
ior that casino owners want to encourage in their cus- 10 cm
tomers, so slot machines are programmed with a
variable ratio schedule. 10 cm 10 cm

Old nest
LATENT LEARNING (destroyed)

Sometimes animals seem to learn without any obvious

immediate reward. For instance, an animal can learn FIGURE 5.5 The experimental design for a study of
important characteristics of its environment during latent learning in ants. Ants were living in the nest site
unrewarded explorations and then use this information pictured at the bottom of the figure. A new nest
later. Even though the knowledge is not put to immedi- (Alternative nest #1) was added to their cage and left
there for a week. Then, a second nest (Alternative nest
ate use (i.e., it is latent), it may later prove to be lifesaving.
#2) was added, and the original nest was immediately
The value of latent learning seems intuitively obvi- destroyed. When the two alternative nests were of
ous. Several studies have shown that familiarity with the identical low quality, ants preferred to settle in nest #2.
terrain improves survival (Metzgar 1967). Pairs of When the two alternative nests were of identical high
white-footed mice (Peromyscus leucopus) were released quality, ants had no preference. The locations of the
into a room with a screech owl (Otus asio). One of the two alternative nests were randomized. (Modified from
pair previously had the opportunity to explore the room Franks et al. 2007.)
 Types of Learning 85

choice, colonizing both nests. Thus, ants discarded a Observational conditioning is a type of classical
familiar, but low-quality, alternative in favor of an iden- conditioning that occurs in social situations. For
tical nest. This decision might seem illogical—why instance, some animals can learn to avoid dangerous sit-
choose one poor nest over an identical one?—but the uations by watching conspecifics. For example, rhesus
authors argue that it makes sense for ants to ignore a monkeys learn to fear and avoid snakes by watching
familiar but unattractive option in favor of exploring for other monkeys show their fear (Mineka and Cook 1988).
a better one, even if they ultimately settle for something Interestingly, monkeys did not show a fear of flowers
that is also unattractive. The experiment was repeated, after watching other monkeys respond fearfully to them
but this time with both alternatives of the same high (Cook and Mineka 1990), suggesting that fear of snakes
quality as the old nests. Here, the ants had no preference: results from a combination of experience and a predis-
23 of 24 colonies were split between the choices, sug- position to learn this particular association. Similarly, fat-
gesting that they indeed are assessing site quality when head minnows do not innately show fear of one of their
they explore, and they can use this information later at natural predators, the northern pike. However, minnows
an appropriate time. learn to show fear responses to pike odor when they are
paired with minnows that have had experience with pike,
but not when paired with inexperienced minnows. After
SOCIAL LEARNING
learning to recognize the pike as a predator, the minnows
Learning from others is a fundamental part of human have a better chance of surviving future encounters, and
learning (Bandura 1962; Meltzoff 1988), but it is not part they are able to transmit the information to naive fathead
of every animal’s behavioral repertoire. Clearly, social minnows (Mathis et al. 1996). Note that in observational
species have much greater opportunity for social learn- learning, we do not need to assume that observers under-
ing than do solitary species. stand anything about the mental state of the animals they
The term social learning encompasses a broad are learning from.
range of phenomena, some of which suggest a higher In goal-directed emulation, an observer seems to
level of cognitive skill on the part of the animal than do learn from observation what goal is to be achieved but
others. In some cases, animals inadvertently provide does not copy precisely what the demonstrator does. For
information to other animals. In other cases, individu- example, chimpanzees and children both watched an
als actively share information through specific signals. adult human demonstrator retrieve artificial fruit from
Researchers who study social learning distinguish several a clear plastic box by untwisting bolts. Chimps showed
categories. goal-directed emulation: they directed their attention at
In stimulus enhancement, an animal may be the correct part of the box but did not imitate the action
attracted to a particular object because a conspecific is of the demonstrator. Children, in contrast, imitated the
near it or is interacting with it. Similarly, in local actions of the observer exactly, even extraneous motions
enhancement, an animal may be attracted to a partic- (Whiten et al. 1999).
ular location because a conspecific is there. Thus, in both That brings us to the evidence for imitation, where
types of enhancement, information is not being actively an observer copies exactly what a demonstrator does.
communicated by one animal to another (Galef 1988; Imitation is quite rare and difficult to document, espe-
Giraldeau 1997; Marler 1996). As an example of stim- cially without knowing the detailed history of what the
ulus enhancement, rats can learn dietary preferences animal has previously learned or seen. One method for
from other rats by smelling their breath. In one exper- studying imitation is the two-action test. The subject is
iment, a “demonstrator” rat ate food flavored with cocoa presented with a task that has two equally easy solutions.
or cinnamon. The demonstrator was then anesthetized If subjects are more likely to choose the solution that
and placed 2 inches away from the wire cage of an awake they have just seen demonstrated, it is taken as evidence
“observer” rat. Although the demonstrator slept of imitation. For example, budgerigars (pet-store para-
through the demonstration, the observer later showed keets, Melopsittacus undulatus) were trained as demon-
a preference for the food the demonstrator had eaten strators. Each learned one of a series of techniques for
(Galef 1990a). removing the cover from a food dish: using their feet,
In nature, local and stimulus enhancement occur pecking with their bills, or pulling with their bills.
frequently in the context of foraging. For instance, when Observer budgies watched a demonstrator open a dish.
bumblebees first visit a new flower species, they are more When presented with a similar dish, the observer used
likely to land on flowers that are already occupied by the same technique it had just witnessed (Dawson and
other bees (Worden and Papaj 2005). Once they learn Foss 1965). Subsequent efforts by others to replicate this
about nectar availability and quantity, they decide for test either failed to do so or produced only transient
themselves rather than following conspecifics (reviewed effects. One reason for these conflicting results might
in Leadbeater and Chittka 2007). Many other animals have been variation in the performance of the demon-
also use conspecifics as cues to good foraging patches. strators. Mottley and Heyes (2003) controlled for this
86 Chapter 5 / Learning and Cognition

variation by letting budgies watch videos of demonstra- doves, but territorial doves learn more readily from
tors rather than live ones. Budgies were more likely to Carib grackles, the species they most often feed with in
remove a stopper from a food box using the method they mixed flocks (Carlier and Lefebvre 1997).
had seen demonstrated in the videos than the alternative
method.
Traditions
Many socially learned behaviors are transient and dis-
The Adaptive Value of Social Learning appear quickly. Others, called traditions, spread through
The potential adaptive value of social learning is clear. a group and are stable over time. For instance, a larce-
It saves some of the time and energy that might be nous tradition began in England around 1921 when a
wasted as an individual learned the business of survival bird species called the blue tit (Parus caeruleus) learned
by trial and error. Although each member of a popula- to break into milk bottles to steal the cream, which, in
tion may have the capacity to learn appropriate the days before homogenization, floated to the top. This
responses for themselves, it is often more efficient and technique spread throughout Great Britain as other birds
less dangerous to learn about the world from others acquired the habit (Fisher and Hinde 1949) (Figure 5.6).
(Galef 1976). Primate groups show a great deal of behavioral vari-
Let’s examine some potential benefits of social learn- ation, suggesting the importance of traditional behaviors.
ing about food. Rats eat a wide range of food types. For example, chimpanzee groups vary markedly in their
However, not all potential foods are safe or nutritious. behaviors: 39 behavior patterns, including tool use,
Rats not only learn about food by smelling each other’s grooming, and courtship, occur frequently in some com-
breath, as described above, but by observation. When a munities but are absent in others (Whiten et al. 1999).
rat observes another eating a novel food, the observer is An interesting tradition is the food-washing habit that
more likely to try it than if it observed another rat eat- spread within a group of snow monkeys. As the story
ing a familiar food. This is a safe way to add breadth to goes, a young female snow monkey of Japan, named Imo,
the diet (Galef 1993). As a result, groups of rats will learn developed new techniques for the treatment of sweet
to select a nutritionally balanced diet more quickly than potatoes and wheat, food provided by the researchers
do rats that are housed alone (Galef and Wright 1995). who study the social behavior of the snow monkeys.
Other species may learn routes to food from con- First, Imo discovered that washing the sweet potato in
specifics. Guppies (Poecilia reticulata), for example, the sea not only cleaned it but also enhanced the flavor
quickly learn a safe route by shoaling, or swimming in by lightly salting it (Figure 5.7). One of Imo’s playmates
large groups (Laland and Williams 1997). observed her and followed suit. Then Imo’s mother
Animals may also learn from other species. We see caught on. And so the tradition spread, usually from
this in different populations of Zenaida doves in youngsters to mothers and siblings. When the young-
Barbados, which live only a few hundred meters apart. sters became mothers, their offspring imitated the
Group-foraging doves learn more quickly from other behavior as if food had always been cleaned in this way.

FIGURE 5.6 The tradition among birds of

opening milk bottles to steal sips of
cream spread rapidly from one area in
England. This trick may have been
spread by social learning.
 Species Differences in Learning: Comparative Studies 87

not groundless: many essential characteristics of learning,

such as the most effective order of presentation of condi-
tioned versus unconditioned stimuli, are indeed similar,
whether studied in rats or humans. However, in recent
years, researchers have been intrigued not just by similar-
ities across species in how and what they learn, but also by
their differences. We’ve already mentioned several studies
that document differences across species that seem to cor-
relate with the ecological conditions they face. In this sec-
tion, we will more explicitly consider the evidence that
differences across species are rooted in natural selection.

THE ABILITY TO LEARN

AS A HERITABLE TRAIT
As we have seen, in order for natural selection to act, the
trait in question must be at least partly heritable. What,
FIGURE 5.7 The tradition of washing sweet potatoes in exactly, about a learned behavior is inherited? A jump-
the sea was begun by a young Japanese snow monkey, ing spider (Phidippus princeps) can learn that red and black
and it spread rapidly to other members of the troop.
milkweed bugs (Oncopeltus fasciatus) are not good to eat:
the first time a spider sees one, it leaps on it, but by the
eighth trial, the spider ignores a bug that crawls right
Several years later Imo started a new custom. The past it (Skow and Jakob 2006). However, the knowledge
researchers spread wheat on the sand, from which the that milkweed bugs taste nasty is not passed onto to the
snow monkeys had to painstakingly pick each grain. One spider’s offspring. Its offspring must learn this for them-
day Imo tossed a handful of sand and wheat into the sea. selves. Learned knowledge is not genetically heritable,
The sand sank but the wheat floated so that it could be although, in some species, offspring can learn from
scooped up from the surface. This ploy was also picked watching their parents. What is heritable, and thus sub-
up by most monkeys in the troop during the next few ject to natural selection, is the capacity to learn.
years (Kawai 1965; Kawamura 1959; Lefebvre 1995). The heritability of the ability to learn has been
However, we must be careful. The division between experimentally demonstrated in several species, includ-
individual learning (learning through one’s own experi- ing that standby of behavioral genetics, Drosophila
ence) and social learning is not always clear-cut: both melanogaster. Mery and Kawecki (2002) carried out an
may occur simultaneously and can be difficult to distin- artificial selection experiment like those experiments
guish. For example, sweet potato washing may occur described in Chapter 3. They gave the fruit flies a choice
through stimulus enhancement. A monkey may pick up of two places to oviposit: media flavored with pineapple
a dropped potato that has been washed, like the taste, versus that flavored with orange juice. One of these
and then be primed to learn to wash potatoes on its own media also contained a quinine solution. Quinine tastes
(de Waal 2001). In addition, differential reinforcement bitter to humans, and it also deters flies. After experience
may maintain the behavior. The monkeys’ only source with this arrangement, flies were then offered a choice
of sweet potatoes is the caretaker. Since the food wash- between orange and pineapple media that had no qui-
ing interests researchers and amuses tourists, the care- nine. Flies that had learned the association between qui-
takers give more sweet potatoes to those members of the nine and a particular flavor avoided that flavor and laid
troop that were known to wash them than to those that their eggs on the neutral flavor. These eggs were col-
did not (Galef 1990b). Although the habit clearly spread lected and reared up to adulthood on an unflavored
throughout the population, we cannot be sure of the cornmeal mix. Thus, only fruit flies that learned to avoid
mechanism. the flavor associated with quinine contributed their al-
leles to the next generation. Each generation, flies were
tested on the same learning task. After 15 generations,
SPECIES DIFFERENCES flies from these selected lines were able to learn the task
IN LEARNING: faster and remember it longer than were flies from con-
COMPARATIVE STUDIES trol lines. These abilities were not confined to the orig-
inal task: they could also learn about novel flavors (apple
For decades, the dominant view in the study of learning and tomato). Thus, the experimenters were able to select
was that it is a general process that occurs in essentially the for an increased ability to learn to identify odors of fruits
same way across mammal species. This view is certainly suitable for egg laying.
88 Chapter 5 / Learning and Cognition

EVOLUTION AND THE VARIATION a

IN LEARNING ACROSS SPECIES
Learning allows an animal to adjust its behavior to new
situations, even those to which its ancestors were never
exposed. Our anthropocentric view is that the ability to
learn is undeniably a positive trait, and it may seem coun-
terintuitive that the ability to learn may not always be
advantageous (Shettleworth 1998). Learning has its
costs. First, it takes time to learn: a spider that is born
with an innate aversion to eating red and black bugs is
saved the time and trouble of repeated mistakes. Second,
the ability to learn requires the dedication of neurons to
the task. Because neurons cannot be infinitely reduced
in size, there is just so much space available for differ-
ent functions in a brain of a given size. If, like a spider,
your brain were smaller than the size of a pinhead, per-
haps it would be better to devote your neuronal space to b
something else, such as large olfactory centers that might
allow you to detect and interpret the chemical scents left
by prey. Finally, learning seems to have an “operating
cost”—it takes energy to collect, process, and store infor-
mation. Drosophila lines forced to use their ability to
learn had fewer offspring than ones that were not (Mery
and Kawecki 2004), implying that natural selection
should act against flies that learn “too well.”
Because the ability to learn can be heritable and has
costs and benefits, we predict that we should see differ-
ences in learning ability across species (Kamil and
Mauldin 1988; Kamil and Yoerg 1982). The environ-
ment and evolutionary history of a species should influ-
ence the degree to which a particular type of learning
will increase their fitness. We have already seen that ani-
mals cannot learn all tasks with the same ease: there
appear to be biological constraints on learning, and
members of a particular species may be prepared to learn
certain things and not others (Chapter 2). Now we will
focus on the evidence for differences among species in
their ability to learn different tasks.
The most complete example of species-specific dif-
ferences in learning ability comes to us from three c
related species of birds: Clark’s nutcrackers (Nucifraga
columbiana), pinyon jays (Gymnorhinus cyanocephalus), and
scrub jays (Aphelocoma coerulescens) (Figure 5.8). These
birds are among the species that cache (store) seeds: they
collect pine seeds in autumn and dig small holes in which
to hide them so that the seeds will not be stolen by other
animals. The birds recover and eat the seeds through-
out the winter and spring when food is scarce.

FIGURE 5.8 Seed-caching birds. (a) Clark’s nutcrackers,

(b) pinyon jays, and (c) scrub jays are birds that hide
seeds in holes in the ground during the autumn and
return to find the seeds during the winter and spring,
when food is less plentiful.
 Species Differences in Learning: Comparative Studies 89

These three species differ in their ecology and in the 1.0

extent to which they rely on seeds (see review in Gibson

Mean proportion correct

and Kamil 2005). Clark’s nutcrackers are the champion 0.8
seed-storers (Balda 1980; Balda and Kamil 1998). This
species lives at high elevations in coniferous forests in 0.6
western North America, where winters are harsh and
0.4
long. During a three-week period in the fall, they may
cache as many as 33,000 pine seeds in several thousand
0.2
separate locations as far as 22 km from the harvesting
site. Throughout the winter, nutcrackers survive almost 0
entirely on these stored seeds. The second species, the Scrub jays Pinyon jays Nutcrackers
pinyon jay, lives in pine woodlands at lower elevations.
This species relies less on caching than do nutcrackers FIGURE 5.9 Histograms that show the accuracy with
which scrub jays, pinyon jays, and Clark’s nutcrackers
but still caches an impressive 20,000 seeds as far as 11
find their caches. Each bird first hid seeds in sand-filled
km away from the collecting site (Balda 1980; Balda and
holes in an indoor aviary. The aviary floor had 90 holes,
Kamil 1998). About 70 to 90% of the pinyon jays’ win- each of which could be filled with sand or plugged. In
ter diet consists of cached seeds. Scrub jays, the final one experimental condition, indicated with the solid
species in our trio, store “only” about 6000 seeds a year, bars, 15 holes were available for caching. In the other,
and these comprise less than 60% of the winter diet indicated by striped bars, all 90 holes were available.
(Balda 1980; Vander Wall and Balda 1977). After caching, the birds were removed from the room
These birds routinely find individual beakfuls of for a week. When they were returned, all 90 holes had
seeds, months after they’ve hidden them, in a landscape been filled with sand. To recover the seeds, the birds
that may be transformed by snow. How do they do this? would probe the sand with their beaks. Accuracy was
It’s not that they simply smell the seeds or sense them measured as the proportion of holes probed that con-
tained seeds. Clark’s nutcrackers and pinyon jays, the
in some other way: they actually remember the exact
species most dependent on cached seeds for winter sur-
locations (Balda 1980; Vander Wall 1982). This is one
vival, were significantly better than scrub jays at recov-
of the most impressive examples of spatial memory on ering their caches. (From Balda and Kamil 1989.)
record, far surpassing human abilities.
The ecological differences among these three species
led investigators to a prediction: species that rely more These data make sense, then, in light of the eco-
heavily on caching to survive the winter have a better spa- logical differences between the species. But is this test
tial memory. To test this prediction, Balda and Kamil enough? Perhaps there is something about these partic-
(1989) devised an elegant experiment that mimics the ular experimental conditions that make the test difficult
process of caching but under controlled conditions. Birds for scrub jays. Perhaps, for example, they are not as
were permitted to store seeds in sand-filled holes in the motivated as the other species to cache and recover seeds
floor of an indoor aviary. The floor had 90 holes, any of in the aviary. This problem gets at the heart of one of
which could be open and filled with sand suitable for the difficulties of interpreting comparative data from dif-
burying seeds, or blocked with a wooden plug. This ferent species on learning: are differences the result of
arrangement allowed the experimenters to vary the posi- the way the species respond to the test conditions, or do
tion and number of the holes available for caching. Each they reflect true cognitive differences?
bird’s ability to recover caches was tested in two condi- One way to resolve this difficulty is to test species
tions. In one, only 15 holes were open; in the other, all in multiple experimental situations that present the same
90 holes were available. After a bird had placed eight type of task but in different ways (Kamil and Mauldin
caches, it was removed from the room. One week later, 1988). In this case, we need other tests of spatial mem-
when it was returned to the aviary, all 90 holes were open, ory that do not require birds to cache seeds. In a second
and the bird’s task was to probe in the subset of the holes test, birds were given the chance to collect seeds from
where it had cached seeds. The accuracy of recovery was holes in an aviary, much like the first test (Kamil et al.
measured as the proportion of holes probed that had con- 1994). However, this task differed because they were not
tained their seeds. All three species performed better than allowed to cache seeds themselves but were required to
expected by chance alone. However, nutcrackers and learn the locations of seeds cached by the experimenter.
pinyon jays, the species that depend most heavily on find- Birds were trained in a room with four open holes, each
ing their stored seeds to survive the winter, did signifi- with seeds. After training, they were given access to the
cantly better than the scrub jays in both experimental room with those same four holes open but with no seeds
conditions (Figure 5.9). Species differences were small in them. In addition, an extra four holes were also open,
when only 15 holes were available for caching but much and these did have seeds. Thus, the task was to learn to
larger when all 90 holes were available. remember, and then bypass, the holes where they had
90 Chapter 5 / Learning and Cognition

cached seeds earlier. Again, Clark’s nutcrackers and nutcrackers, pinyon jays, scrub jays, along with a fourth
pinyon jays were better at this task than were scrub jays. species, Mexican jays. Mexican jays also cache food, and
In contrast to the previous test, when the task got harder they live at a higher elevation than scrub jays. Birds were
such that the birds had to remember the locations for first trained to peck at a monitor at an illuminated cir-
a longer time, the species differences decreased. cle in a particular location. During the testing session,
Differences between the tests reiterate the value of mul- two circles were illuminated: one in the location the bird
tiple kinds of tests in unraveling species differences in had been trained with and one in a new location. Birds
learning ability: even small differences in experimental were rewarded for pecking at the key in the new loca-
design can have meaningful effects on the outcome. This, tion. By increasing the delay between the training ses-
of course, complicates our interpretation of the results. sions and the testing sessions, the researchers could
A third test of spatial memory was quite different determine how long the birds could remember the loca-
from the first two. Olson et al. (1995) used an experi- tion of the key that was rewarded during training.
mental design with the descriptive name of a “delayed Nutcrackers performed better than other species
operant nonmatching-to-sample procedure” to test (Figure 5.10a).

a 50 Spatial task – location of circle

45
Scrub jay
40
Mexican jay
35
Retention interval (sec)

Pinyon jay

30 Nutcracker

25

20

15

10

5
FIGURE 5.10 Comparison of learning
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 abilities among food-storing and non-
Block (100 trials) storing corvids on spatial and nonspatial
tasks. Shown here is the length of time
each type of bird could remember a par-
b 50
Nonspatial task – color of circle ticular task. (a) Clark’s nutcracker, the
species most highly dependent on stored
Scrub jay food, was an obvious champion on a spa-
40 tial task, remembering the location of a
Mexican jay
circle displayed on a screen. (b) However,
Retention interval (sec)

Pinyon jay species differences in memory on a non-

30 spatial task, remembering the color of a
Nutcracker circle, were not related to the species’
dependence on stored food. Thus, the
20 outstanding performance of Clark’s nut-
cracker on spatial memory tasks is not
simply because these birds adapt to life
10 in captivity better than other species or
because nutcrackers are generally more
“intelligent.” These observations add
strength to the hypothesis that learning
0
5 10 15 20 25 30 abilities are shaped by natural selection.
Block (100 trials) (Data from Olson et al. 1995.)
 Other Evidence of Cognitive Abilities in Animals 91

These experiments show that differences in spa- 1995), although the evidence is not as consistent as that
tial memory among Clark’s nutcrackers, pinyon jays, in corvids (reviewed in Shettleworth 1995 and
and scrub jays are not limited to the recovery of cached Pravosudov 2007).
seeds. These results increase our confidence in our Far fewer comparative studies of spatial learning
conclusion that spatial memory is an adaptive special- have been done in caching mammals, but a similar
ization. However, we still can’t dismiss the alternative relationship between ecology and spatial skills has been
hypotheses that the species differences arise because documented in several species. For instance, the Great
the food-caching species happen to be better suited to Basin kangaroo rat (Dipodomys microps) eats leaves of
life in the laboratory environment, or possibly are the saltbush. This is a common and abundant food, and
more motivated to complete the tasks. The likelihood presents no navigational challenges. In contrast, the
of these alternatives can be examined by testing species Merriam’s kangaroo rat (D. merriami) prefers seeds and
on a nonspatial task. If the differences between stores them in scattered locations, much as nutcrack-
species persist even in a nonspatial task, then we ers do. As predicted from their ecological differences,
should suspect that they are driven by something the Merriam’s kangaroo rat performed better on a
besides an adaptive specialization for spatial learning. spatial test than the Great Basin rat (Barkley and
Olson et al. (1995) gave birds a nonspatial task, a non- Jacobs 2007).
matching-to-sample procedure based on color, rather Earlier in this section we suggested that learning
than location. In contrast to the spatial test, there were ability depends on the number of neurons devoted to the
no differences among species (Figure 5.10b). This task. In birds and mammals, spatial learning is known to
again supports the idea that spatial learning ability, be at least partly based in the region of the brain called
rather than learning in general, is predicted by the the hippocampus. Does the size of the hippocampus vary
ecology of the species. with the amount of food hoarding? Results from earlier
Yet another hypothesis could explain species differ- studies were mixed, with some finding a relationship
ences that we see in various spatial tests. Perhaps these (e.g., Basil et al. 1996; Krebs et al. 1989; Sherry et al.
species differ not in their spatial learning ability, but in 1989) and others not (e.g., Brodin and Lundbord 2003).
their ability to perceive appropriate details of the envi- A recent large comparative study supports a relationship
ronment (Macphail and Bolhuis 2001). Gibson and between hippocampal size and spatial learning. This
Kamil (2005) tested the perceptual ability of these study included 55 bird species and controlled carefully
species by requiring them to discriminate the distance for phylogenetic relationships. Closely related species
between two landmarks on a computer screen. All three may have more similar cognitive abilities than more dis-
species were able to make fine discriminations, but nut- tantly related species, which can bias the results of a
crackers were no better—indeed, they were slightly comparative study unless phylogeny is accounted for.
worse—than the other species. Thus, this experiment Both the relative volume of the hippocampus and the
provides no evidence that differences among these total volume of the brain were positively correlated with
species in the learning tasks were due to how they per- the amount of food hoarding (Garamszegi and Eens
ceive spatial information. However, proving that per- 2004). However, Provosudov and de Kort (2006) found
ceptual differences are not driving differences among that scrub jay brains are much larger than previously
species—in essence, proving a null hypothesis—is reported, and argued that methodological issues cloud
incredibly difficult, if not impossible. the interpretation of the existing data on hippocampal
This correlation between spatial memory and food size. Clearly, the issue is not yet settled, and research
caching is consistent with our hypothesis that evolution continues.
may shape learning ability, but it is also possible that the
observed differences in spatial memory are simply
chance differences among this admittedly small group of OTHER EVIDENCE
species that have been studied. The case would be OF COGNITIVE ABILITIES
stronger if the same patterns were found among other IN ANIMALS
groups of related species. Food-storing behavior has also
evolved among some species of the Paridae (titmice and The study of the mental processes of animals goes
chickadees), a family of birds that is phylogenetically dis- beyond the study of learning. Some scientists have won-
tinct from the Corvidae (nutcrackers and jays). Parids dered whether animals have mental experiences—
store seeds and insects in hundreds of widely scattered thoughts and feelings, for instance (Bateson and Klopfer
sites for shorter time periods than do the corvids—hours 1991; Griffin 1981, 1982, 1984, 1991; Hoage and
to weeks as opposed to months. In several tests, food- Goldman 1986; Mellgren 1983; Ristau 1991). But how
storing species perform better than non–food-storing could we ever know whether other animals think or
species (Giraldeau 1997; Krebs et al. 1996; Shettleworth whether they are self-aware? What constitutes good
92 Chapter 5 / Learning and Cognition

evidence for complex cognitive abilities? In this section,

we’ll discuss experiments that address the idea of animal
understanding.

TOOL USE
A sea otter, floating on its back, uses a rock to break a
clam shell on its belly. A vulture drops a rock on an
ostrich egg, which cracks open. A chimpanzee strips the
leaves off a stick and uses it to pull juicy termites out of
a mound. All of these are examples of tool use—the use
of an object in order to obtain a goal. Once considered
to be a hallmark trait that separated humans from other
animals, tool use is now known in many species.
Tool use seems to demonstrate a high level of cogni-
tion. In order to use a tool, must not an animal under-
stand how it works? Animals using a tool to solve a
problem can often appear to be thinking it through. Let’s
begin by considering the famous example of Köhler’s
(1927) chimpanzees, particularly one named Sultan.
Sultan first learned to use a stick as a tool to extend his
reach and rake in a banana on the ground outside his FIGURE 5.11 Sultan playing with sticks. After getting
cage. Having mastered this, he was given two sticks that experience with the sticks, Sultan fit them together
when put together end to end were just long enough to end to end to reach bananas.
reach the fruit. Sultan tried unsuccessfully to reach the
reward with each of the sticks. He even managed to prod
one stick with the tip of the other until it touched the targets and climbing on an object to reach another
banana, but since the sticks were not joined, he could not object. Pigeons can chain together similar learned tasks.
retrieve the fruit. For over an hour, Sultan persistently Pigeons that were trained both to push a box in a par-
tried, and failed, to get the banana. Finally, he seemed to ticular direction and to climb on the box to get a reward
give up and began to play with the sticks (Figure 5.11). were able to put both behaviors together. Epstein et al.
As he was playing, he happened to hold one in each hand (1984) concluded that seemingly insightful behavior
so the ends were pointed toward one another. At this might be built from specific stimulus–response relation-
point he fitted one end of the stick into the other, thus ships learned through operant conditioning.
lengthening the tool. Immediately, he ran to the bars of These studies emphasize the need to control for an
his cage and began to rake in the banana. As he was draw- animal’s prior experience in order to understand exactly
ing the banana toward him, the two sticks separated. That what it knows when it manipulates objects. Let’s consider
Sultan quickly recovered the sticks and rejoined them was another example of apparently insightful behavior in a
evidence to Köhler that the chimp understood that fit- bird, this time in the common raven (Corvus corax).
ting two bamboo poles together was an effective way to Ravens are known to pull up ice fishing lines to steal fish,
increase his reach far enough to obtain the fruit. Because as shown in the illustration at the start of this chapter.
of the suddenness of Sultan’s solution, Köhler called his Bernd Heinrich (1995) presented hand-reared ravens
behavior insight. He documented other similar situations with meat suspended from string, a problem he knew
where a flash of understanding seemed to occur, such as they had never encountered before. To reach the sus-
when chimps stacked boxes and climbed on top in order pended food, a bird had to pull up a loop of string, step
to knock down a hanging banana with a stick. on the loop to hold it in place, and then reach down and
Perhaps the chimps were able to see new relation- pull up another loop. The bird had to repeat this cycle
ships among events and were able to consider the prob- six to eight times to obtain the food. At least ten species
lem as a whole. Perhaps they even formed a mental of birds can be taught by operant conditioning to pull
representation of the problem and then mentally worked up food dangling on a string if the distance between the
through solutions to it. However, we must be careful food and the perch is gradually lengthened. However, a
here. All the details of the chimps’ prior experience were few of the ravens in Heinrich’s study solved the problem
not known. Perhaps chimps that moved boxes and then immediately without any indication of going through a
climbed on them to reach a banana had previously learning process. In fact, one bird went through the
learned two separate behaviors—moving boxes toward entire sequence of 30 steps and obtained the food the
 Other Evidence of Cognitive Abilities in Animals 93

first time it approached the string, even though no other were given a puzzle consisting of a clear vertical tube
bird in the group had previously shown the behavior. with a small bucket in it. The bucket contained food and
Ravens can correctly solve more complicated versions of had a handle that could be reached from above, but only
this task. When given two strings, one with a rock and with the aid of a tool. One crow was given a straight
one with a piece of meat, that are crossed over, ravens piece of wire, a substance she had no experience with.
can pull on the correct string on their first trial. She quickly bent it into a hook and retrieved the bucket
In nature, ravens don’t normally pull on one object (Weir et al. 2002).
to obtain another one. When they pull on food, such as New Caledonian crows also use “metatools”: they
the entrails of a dead animal, they eat the food while can use one tool on another. In this task, the crows
pulling. Thus, Heinrich argues that it is unlikely that this needed to use a short stick in order to retrieve a longer
complicated behavior was learned, was genetically pro- stick from a “toolbox.” They could then use the longer
grammed, or occurred by chance. The ravens apparently stick to retrieve a piece of meat. Six of seven crows cor-
have the ability to find insightful solutions to new prob- rectly tried to extract the long tool with the short tool,
lems, using string as a tool. and four crows successfully solved the problem on their
Insight alone may not always be enough. A recent first attempt. In a follow-up experiment, the positions of
extension of the research on ravens demonstrated that the tools were reversed—the small stick was in the “tool-
familiarity with a simpler task may be required to suc- box,” and the long tool was given directly to the crows.
ceed on a new task. Here, the string was looped up and In this setup, the crows did not need to use the
through the cage, then down again. The birds had to pull “metatool” approach—they simply needed to pick up the
the string down in order to raise the meat. Ravens that long tool and extract the food. All six crows that were
were familiar with the pull-up task could quickly do the tested briefly attempted the unnecessary step of using the
pull-down task, but naive birds could not (Heinrich and long tool to access the short tool in the “toolbox” but
Bugnyar 2005). quickly rectified their mistake and began going straight
On the South Pacific island of New Caledonia, to the food with the long tool (Taylor et al. 2007). The
native crows are especially adept at making and using authors argue that the crows may well have understood
tools. In the wild, crows craft tools of several different the more general causal relationship that tools can be
varieties out of twigs and leaves (Hunt 1996; Hunt and used to access out-of-reach objects, even other tools.
Gray 2004; Hunt et al. 2006). For example, to make a As an interesting note, scientists have recently been
hook, they snap off a twig, strip off the leaves, and then able to attach tiny video cameras to wild New
use their bill to sculpt the end of the twig. The crows Caledonian crows, enabling us to see them use tools in
then poke the hooked twigs into holes in order to extract nature when no experimenters are nearby. These videos
insects and other small prey (Figure 5.12). They can even have revealed that crows appear to keep particularly
make hooks out of unfamiliar material. In the lab, crows good tools for future use (Rutz et al. 2007).

FIGURE 5.12 A New Caledonian crow

using a stick as a tool.
94 Chapter 5 / Learning and Cognition

These experiments on ravens and crows seem to tives. As we have already seen, chimpanzees are accom-
reinforce the idea that animals might fully understand plished tool users. In the wild, they use sticks to forage
tool-using tasks, but let’s look at a case where this is for termites and rocks as a hammer and anvil to pound
clearly not so (see Shettleworth 1998 for a more open nuts. Recently, chimpanzees have been observed
detailed review of this literature). Capuchin monkeys fashioning spears out of sticks, trimming branches off
were shown a Plexiglas tube mounted horizontally on a and using their incisors to sharpen the end. The chimps
stand with a peanut in the middle. The tube was too nar- then jabbed the spears into hollow trees. One was seen
row for the monkeys to reach in with their arms. When extracting a bushbaby (a small primate) from a hole in a
they were provided with sticks, monkeys could quickly tree after jabbing with the spear (Pruetz and Bertolani
learn to poke a stick into the tube to push the peanut out 2007). It is likely that foraging and hunting were the first
(Visalberghi and Trinca 1989). But then an interesting contexts in which our ancestors used tools.
twist was introduced into the design: a trap was placed As we see, the investigation of tool use can be tricky.
into the tube (Figure 5.13). If a monkey inserted the stick The experimenter can easily miss, or misinterpret, a
in the wrong end of the tube, it would push the reward moment of sudden insight. In addition, small changes in
into the trap. Of four monkeys tested, only a single mon- the task can greatly alter the outcome. Carefully con-
key consistently inserted the stick in the correct end of trolled experiments are the key, with attention to poten-
the tube, and only after 90 trials (Visalberghi and tial alternative interpretations.
Limongelli 1994). Further tests demonstrated that even
this individual didn’t fully grasp the task. The researchers
suspected that she might be following the rule, learned DETOURS
by trial and error, of “Push the stick in the side of the Detouring is the ability to identify an alternative route
tube furthest from the treat.” Indeed, when the tube was to a reward when the direct route is blocked. Although
rotated so that the trap was on top, the monkey followed animals can often improve on detour tests with experi-
this rule, even though it was no longer necessary. ence, here we are most interested in how they respond
Chimpanzees, in contrast, showed more understanding the first time they are confronted with a test. Nearly
of the task (Limongelli et al. 1995). Human children everyone who has taken an unruly dog for a walk has
under three years of age behaved more like capuchins had a demonstration that not all animals understand
than chimpanzees (Visalberghi and Limongelli 1996). how to solve detour problems. A dog that has wrapped
Tools are, of course, of great importance in human its leash around the legs of its owner while trying to get
evolution, and perhaps we can gain some insight into our to a squirrel does not comprehend that sometimes the
own past by examining the behavior of our close rela- best route from point A to point B is not necessarily the
most direct. In controlled experiments, dogs were
clearly motivated to reach a toy or food on the other
side of a wire-mesh fence, but they often tried digging
under the fence rather than detouring around it
(Pongrácz et al. 2001). In contrast, if you have watched
tree squirrels, you know that other species are very good
at detour problems. Squirrels seem to immediately “see”
that to get from tree to tree, they must choose the
branch that reaches between the trees. Another com-
parative study of detour ability shows differences even
among more closely related species. Quail and herring
gulls were easily able to solve a detour task in which they
were required to walk around a barrier, but canaries
could not (Zucca et al. 2005). The difference might be
because canaries don’t walk much in the wild—when
they face a similar sort of detour task in daily life, they
can solve it by flying.
How exactly do animals solve detour tasks? One
species that excels at detours might come as a surprise:
FIGURE 5.13 The tube task but with a trap added. The jumping spiders in the genus Portia. Like other jumping
subject must push a treat out of the tube with a stick. spiders, Portia has large anterior eyes specialized for
Because of the trap in the center of the tube, the stick acute vision (Figure 5.14). However, whereas most jump-
must be inserted in a particular end of the tube. Here, ing spiders attack insect prey by stalking and tackling it,
a capuchin monkey is about to make an error. Portia prefers to hunt other spiders. To do it, Portia
 Other Evidence of Cognitive Abilities in Animals 95

a Back Prey
rampway pole
Forward Forward
Lure rampway
rampway

Support
Support
pole
pole

Starting
platform

b Back Prey
rampway pole
Forward
Lure rampway
Gap

FIGURE 5.14 A jumping spider of the genus Portia,

showing its large specialized anterior eyes.

c Back Prey
climbs right into the webs of other spiders and lures rampway pole
them in with a series of vibratory signals similar to those
Forward Lure
given by struggling prey. When the web owners get rampway Gap
close, Portia grabs them (Jackson and Wilcox 1993a).
Portia can spot spider webs from some distance away, but
often, in order to reach them in their complicated three-
dimensional environments, it must perform a detour
(Jackson and Wilcox 1993b). In the lab, it can, before set-
ting out, choose correctly between detours that lead to
prey versus those that do not (Tarsitano and Jackson
1997), and detour routes that are complete versus those FIGURE 5.15 A detour test for a jumping spider, Portia.
that have gaps (Tarsitano and Andrew 1999) (Figure A lure is placed on the prey pole. Portia must leave the
5.15). Portia solves a detour problem by looking at the starting platform and climb the supports to reach the
lure, then slowly scanning along the horizontal features prey. In (a), both support poles lead to the prey, but in
of the potential route. If the route ends, the spider turns (b) and (c) only one side of the route is complete, and
back to look at the lure again and then begins once more. the other side has a gap. Spiders can correctly choose
This example demonstrates how fairly simple behavioral the complete path at the start of their route. (From
Tarsitano and Andrew 1999.)
rules can give rise to behaviors that appear to be quite
complex.
of objects. The ability to count things, no matter what
they are, demonstrates some understanding of the
UNDERSTANDING NUMBERS AND
abstract concept of numbers.
OTHER ABSTRACT CONCEPTS
Research into animal counting had an inauspicious
It seems quite valuable for animals to have a sense of beginning (described in Shettleworth 1998). Clever
quantity: Is there more fruit on this tree or that tree? Are Hans was a horse in the early 1900s who would answer
there more ducks in the pond to my left or to my right? arithmetic questions extremely accurately by tapping
The ability to discriminate these sorts of relative num- with his hoof. He caused quite a sensation, until a young
bers is quite common and does not seem to require dif- psychologist tested him when he could not see any peo-
ficult mental gymnastics. ple who knew the answer. It turned out that Clever Hans
What is more difficult is the ability to count—to really was clever, but at reading very subtle signals of
assign a tag such as “1, 2, 3,” to individual quantities. An humans rather than at math. As Hans tapped his way
animal that can count can apply tags to different kinds toward the correct answer, the body posture of the
96 Chapter 5 / Learning and Cognition

humans in the room would subtly change, and Hans

would stop.
Clever Hans unwittingly taught researchers quite a
bit about the importance of designing careful experi-
ments to test the ability to count. We now have good evi-
dence that several nonhuman animals can count. One
that demonstrated this talent vocally was an African gray
parrot (Psittacus erithacus) named Alex (Figure 5.16), who,
sadly, died as this book was being revised. We all know
that parrots can be trained to talk, but most of us would
guess that they are mimicking their trainers. This was
certainly not true of Alex. Irene Pepperberg (2000)
detailed more than 20 years of research on Alex in The
Alex Studies. He learned labels (names) for over 35 dif-
ferent objects. By combining labels, he could identify,
request, refuse, or comment on more than 100 different FIGURE 5.16 Alex, an African gray parrot who learned
objects. Furthermore, he used language to show that he several concepts. Alex knew the concept of same/differ-
understood certain abstract concepts. One such concept ent, an idea once thought to characterize only humans
was quantity. He could say how many items were in a and their closest primate relatives.
group for collections of up to six items, even if the
objects were scattered around a tray (Pepperberg 1987a). at a time. The objects would differ in one of three qual-
Even more remarkable was Alex’s ability to accurately ities: color, shape, or material. He might be shown a yel-
count specific items in what is called a confounded num- low, rawhide pentagon and a gray, wooden pentagon or
ber set, which are items that vary in more than one char- a green, wooden triangle and a blue, wooden triangle.
acteristic. For example, a set might consist of two types Then Alex would be asked, “What’s same?” or “What’s
of objects, say balls and keys, that appear in two colors, different?” A correct answer to the first question is to
red or blue. When presented with all these objects mixed name the category of the similar shared characteristic.
together on a tray, Alex could say the number of items When he saw the first of the previous examples, Alex
of a specific type and color, such as the number of blue would have to answer “shape,” not “pentagon.” In the
keys. He responded correctly to these types of questions second example, a correct response to “What’s differ-
83% of the time (Pepperberg 1994). He could even add ent?” would be “color,” not “green.” When shown
up the total from two sequentially presented collections objects he had seen before, Alex correctly identified the
(Pepperberg 2006). Alex also may have had a limited characteristic that was the same or different 76% of the
understanding of the concept of zero. He spontaneously time. He was also shown pairs of objects that he had
began to say “none” in response to the absence of objects never seen before, and 85% of the time he correctly
on the tray (Pepperberg and Gordon 2005). However, identified the characteristic that was the same or differ-
in a follow-up experiment, when asked explicitly how ent (Pepperberg 1987b).
many objects were underneath an empty cup, he either Alex’s abilities were particularly impressive given
refused to answer or said “one” (Pepperberg 2006). Alex that his brain was the size of a walnut. However, as
starred in many nature and science television shows, and Srinivasan and Zhang (2003) argue, brain size does not
it is well worth watching him in action. reliably indicate the ability of animals to do tasks such
Another impressive demonstration of counting abil- as concept learning. Even honeybees can learn to dis-
ity comes from Sheba, a chimpanzee. Using a touch tinguish between same and different. For example, they
screen, Sheba can indicate the Arabic numeral that can learn that if they see a particular pattern when they
describes a group of objects. She can also add numbers: enter a testing apparatus, they must then choose the
if three small groups of objects are put in three separate matching pattern (a delayed matching-to-sample proce-
places around the room, she can visit them in turn and dure). Once they learn this task, they can immediately
then correctly choose the numeral that represents the correctly perform a similar task with colors, without fur-
sum. Finally, if the three groups of objects are replaced ther training. Thus, they appear to learn the concept of
with numeral cards, she can still choose the numeral that matching (reviewed in Srinivasan and Zhang 2003).
represents the correct total (Boysen and Berntson 1989). Some very interesting experiments on pigeons
Counting is one abstract concept, but there are oth- (Columba livia) show that they are able to form concepts
ers. Alex the parrot understood other abstract concepts— such as “tree” or “water” or “human.” The typical pro-
the concepts of same and different. He demonstrated this tocol is a training session with a series of slides, gener-
ability in experiments in which he was shown two objects ally a variety of photographs. Pigeons are rewarded when
 Other Evidence of Cognitive Abilities in Animals 97

they peck at a slide that has an example of a particular individuals, and to understand what others know and do
category, such as a person. For instance, they might see not know. To clarify this idea, let’s look at an example.
40 photographs with people pictured from different Hare and Tomasello (2000) set up a competitive situa-
angles, in partial view, and so on, as well as 40 pho- tion over two pieces of food between a subordinate and
tographs with no people. After they have learned to clas- a dominant chimpanzee. If a dominant chimp sees a sub-
sify the photos correctly, they are then given new slides ordinate eating food, it will take the food away. In the
they have never seen before, and they are asked to clas- experiment, the dominant chimp could only see one
sify them. Pigeons can recognize water, for instance, in piece of food because the other was hidden by a barrier.
various forms—a droplet, a river, a lake (Herrnstein et The subordinate chimp could see both pieces of food
al. 1976; Mallot and Siddall 1972; Siegel and Honig and could also see the dominant chimp. The question
1970). They can even learn to distinguish paintings by was whether the subordinate chimp was aware of which
Monet from those of Picasso (Watanabe et al. 1995). piece of food the dominant could see—could the subor-
dinate understand the perspective of the dominant? In
SELF-RECOGNITION fact, this seemed to be the case: when given a choice,
subordinates selected the piece that was not visible to the
AND PERSPECTIVE TAKING
dominant individual. Capuchins, in contrast, did not
How do animals perceive themselves? Do they see their show evidence of perspective taking (Hare et al. 2003).
bodies as unique entities, separate from the rest of the There is a growing literature on “seeing and knowing”
world? Can they adopt the point of view of other ani- and the attribution of knowledge and mental states, and
mals? On the surface, these questions seem to be many species have been tested with a range of clever
untestable—but research on a variety of fronts has shed experiments.
some light on them. A compelling idea about the evolution of the
The “mark test” was devised nearly 40 years ago “theory of mind” is that it is driven by social complex-
(Gallup 1970). It is still in use, and its interpretation is ity: the social environment creates new selection pres-
still controversial. Here, the subject is given a mirror and sures for the evolution of “social intelligence.” For
is given time to adapt to it. Initially, animals often treat example, the ability to learn and keep track of relation-
the mirror as a conspecific, making threats to it or greet- ships among other individuals may well be evolutionar-
ing it. After some time, some species, such as chimps, ily advantageous. We’ll return to the topic of sociality in
begin to use the mirror to groom otherwise out-of-sight Chapters 18 and 19.
areas, pick their teeth, and the like. At this point, the ani-
mal is given general anesthesia, and a harmless, odorless
dye is applied to some area of the face. After the subject SUMMARY
recovers from anesthesia, its behavior is observed with-
out the mirror for a baseline sample. Then the animal is Types of Learning
shown the mirror. If the subject recognizes itself, it Learning is a change in behavior as a result of experi-
should see that it now has an odd new mark, and direct ence, excluding changes as a result of maturation of the
touches and grooming toward the marked area. The nervous system, fatigue, or sensory adaptation. Learning
best-controlled versions of this experiment compare is traditionally divided into categories:
these touches to the marked area with touches toward
Habituation. The animal learns not to respond to a spe-
predefined control areas on the head that have no mark
cific stimulus because it has been encountered frequently
(Povinelli et al. 1997; Shettleworth 1998). Species that
without important consequences. Habituation is adap-
have “passed” the mark test include chimpanzees (Pan
tive because it conserves energy and leaves more time for
troglodytes) (Povinelli et al. 1997), dolphins (Tursiops trun-
other important activities.
catus) (which of course have no way to touch marks but
do turn their bodies to inspect marks in the mirror; Reiss Classical Conditioning. Classical conditioning is a type of
and Marino 2001), and Asian elephants (Elephas max- associative learning. The animal learns to give a response
imus) (which touch marks with their trunks; Plotnik et normally elicited by one stimulus (the unconditioned
al. 2006). Even in these species, often only some of the stimulus, or US) to a new stimulus (the conditioned stim-
individuals tested are successful. ulus, or CS) because the two are repeatedly paired.
But what does this mean? Certainly this helps to Conditioning is most effective if the CS reliably precedes
demonstrate that an animal has self-perception and a the US. If the CS is presented many times without the US,
knowledge of its physical body, but does it mean the ani- the response to the new stimulus will be gradually lost.
mal has a concept of self in the same way we do? A dif- This is called extinction.
ferent approach to the question of self versus nonself is Operant Conditioning. This is another type of associa-
to ask whether animals can take the perspective of other tive learning. Here, the frequency of a behavior is
98 Chapter 5 / Learning and Cognition

increased because it is reinforced. Novel behaviors can that rely heavily on being able to find stored seeds have
be introduced into the repertoire through shaping. better spatial memory than species that do not.
During shaping, the reward is made contingent upon
closer and closer approximation to the desired action.
Sometimes not every response is reinforced. The fre-
Other Evidence of Cognitive Abilities
quency with which the reward is given is called the rein- in Animals
forcement schedule. Animals demonstrate cognitive skills in other tasks
Latent Learning. Latent learning occurs without any besides learning. A common theme is that the design of
obvious reinforcement, and is not obvious until some- experiments is crucial: subtle differences can produce
time later in life. The information gained through explo- profoundly different outcomes.
ration is an example. Tool Use. Tools are objects that an animal uses to reach
Social Learning. The animal learns from others. Types a goal. In some cases, animals seem to understand how a
of social learning include stimulus and local enhance- tool works; in other cases, their understanding is more
ment (where animals are attracted to an object or loca- limited.
tion by conspecifics), observational conditioning (a type
Detour Behavior. An animal takes an indirect route to a
of classical conditioning that occurs in social contexts),
goal. Species differences in the ability to detour are well
goal-directed emulation (where an animal learns the goal
documented.
of a task by watching another animal but not exactly how
to perform it), and imitation (where an animal copies Understanding Numbers and Other Abstract Concepts.
another’s action). Traditions spread through a group and Many species understand relative numbers, but only a
are stable over time. few have been shown to be able to “tag” particular quan-
tities in the way that we do. Many animals understand
other concepts, such as “same” vs. “different,” and can
Species Differences in Learning: classify objects into various categories.
Comparative Studies Self-Recognition and Perspective Taking. Do animals
The ability to learn has a genetic basis, as artificial selec- perceive themselves as separate from others? The “mark
tion experiments show. Thus, differences in learning test,” whereby animals are given a mark without their
ability may be due to natural selection. Comparative knowledge and then allowed to examine themselves in a
studies of learning have addressed the question of mirror, provides evidence that some animals can recog-
whether learning is adaptively specialized across differ- nize themselves. Experiments on perspective-taking
ent species. The most complete case study is that of suggest that some animals can understand that other
food-storing birds, where evidence suggests that species individuals do not have the same knowledge that they do.
 6
Physiological Analysis—Nerve Cells
and Behavior
Concepts from Cellular Neurobiology from the cockroach when the intended victim turned its
Types of Neurons and Their Jobs body away from the cat and ran. Indeed, if you have tried
The Message of a Neuron to step on a cockroach, you may have been unpleasantly
Ions, Membrane Permeability, and Behavior surprised at its speed, 70 to 80 cm per second, and its
ability to change direction rapidly enough to avoid your
Behavioral Change and Synaptic Transmission
well-aimed foot. Such a quick response depends on the
The Structure of the Synapse nervous system; hormones, the subject of the next chap-
Integration ter, could not trigger such a speedy response. Although
Specializations for Perception of Biologically the cockroach escape response may not impress you, it
Relevant Stimuli—Sensory Processing has inspired some researchers’ efforts to build antimis-
Processing of Sensory Information for Sound sile technology and crash devices for automobiles. Our
Localization goals in this chapter will not be that lofty. We will con-
Predators and Prey: The Neuroethology sider the escape response of the cockroach as an intro-
of Life-and-Death Struggles duction to the types of neurons and the concept that
Processing in the Central Nervous System neurons are organized into networks.
Brain Changes Underlying Behavioral Change Adaptive behavior depends on interactions among
Social Behavior Network the components of the nervous system, the body, and the
Responding—Motor Systems environment. Sensory receptors must detect critical
stimuli, and sensory input must be filtered to extract the
Neural Control in Motor Systems
most biologically relevant information. Based on this
Locust Flight
information, the nervous system must then produce
adaptive responses. For instance, a moth’s simple nervous
Just as the house cat raises its paw to strike, the cock- system must process information from a hunting bat’s
roach (Periplaneta americana) dashes across the floor and calls to avoid predation. The nervous system of a barn
disappears into a tiny crevice. If we were to film this owl processes auditory information so that the sounds of
sequence and then replay it in slow motion, we would a scurrying mouse can be precisely located and a direct
see that the cat’s paw was still several centimeters away strike executed in complete darkness. The brain

99
100 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

responds to incoming stimuli in a dynamic fashion,

which involves changes in connections between neurons, Cercus
the growth of new neurons, and interactions between
brain regions. The movements involved in an animal’s
response are also sculpted by interactions among nerve Cercus
cells. One of the most interesting questions in the field
of motor control is how rhythmic motor patterns, such
as locust flight, are generated by groups of neurons called
central pattern generators.
These are just a few of the issues we will address in
this chapter. Although we will focus on mechanism here,
keep in mind that mechanism cannot be considered apart FIGURE 6.1 A cockroach, showing the two cerci, each
with approximately 200 wind-sensitive hairs.
from evolution. Nervous systems are the product of evo-
lution, and they in turn affect the direction of evolution.
Before we can understand the behavioral responses
of animals, we must first learn how the nervous system interneurons process sensory input and determine the
is put together. We will present a series of examples motor response.
demonstrating that nervous systems have evolved so that Look at these three types of neurons in the cock-
animals can quickly (1) detect pertinent events in their roach as we consider the role each plays in the escape
environment, (2) choose appropriate responses to such response (Figure 6.2). At the base of each of the wind-
events, and (3) coordinate the parts of their bodies nec- sensitive hairs on the cercus is a single sensory neuron
essary to execute the responses. that relays pertinent information from the external envi-
ronment into the central nervous system. In the central
nervous system, the sensory neuron makes contact with
an interneuron; in this case, the interneuron is described
CONCEPTS FROM CELLULAR as a giant because of its exceptionally large diameter.
NEUROBIOLOGY This giant interneuron ascends the nerve cord to the
head. Before reaching the head, however, the giant
TYPES OF NEURONS AND THEIR JOBS interneuron makes contact with an interneuron in the
How did the cockroach in the opening scenario detect thoracic area, which in turn connects with motor neu-
the predator in time to take evasive action? Kenneth rons that relay messages to the hind leg muscles
Roeder, and later Jeffrey Camhi and his colleagues, stud- (Schaefer et al. 1994). (An advantage of studying the
ied the escape response of cockroaches and discovered neural basis of behavior in an invertebrate animal such
that these unlovable house guests respond to gusts of air as the cockroach is that it is sometimes possible to iden-
that are created by even the slightest movements of their tify the individual neurons involved in a specific behav-
enemies (Camhi 1984, 1988; Camhi et al. 1978). ior, particularly a pattern of behavior associated with
Cockroaches, it turns out, have numerous hairlike escape. Because escape requires fast action, the neurons
receptors that are sensitive to wind, and these receptors involved in escape responses are often large in diameter
are located on two posterior appendages called cerci (sin- to permit the rapid conduction of messages. The result
gular, cercus; Figure 6.1). When these wind-sensitive is that these large neurons are somewhat easier to identify
receptors are stimulated, they alert the nervous system than their smaller counterparts are.)
of the cockroach, and within a matter of milliseconds, How does the cockroach determine the direction of
the cockroach turns away from the direction of the wind a wind gust, so that it can run away from cats rather than
and starts to run. straight into them? Most of the segments on each cer-
The escape of the cockroach is orchestrated by the cus have a row of sensory hairs that can be deflected
interactions among nerve cells, which are also called slightly by wind or touch. Each wind-sensitive hair
neurons. Neurons can be classified into three groups responds differently to a gust of wind from a particular
based on their function. Neurons that carry signals from direction. Thus, the pattern of output from the sensory
a receptor organ at the periphery toward the central ner- hairs encodes information about the direction of the
vous system (in vertebrates, the brain and spinal cord, wind. This information is sent to the giant interneurons.
and in invertebrates, the brain and nerve cord) are called Seven giant interneurons run along each side of the
sensory or afferent neurons. Those that carry signals cockroach’s ventral nerve cord (Levi and Camhi 1995).
away from the central nervous system to muscles and The pattern of output from the sensory hairs caused by
glands are called efferent or motor neurons. Inter- a wind gust will stimulate each giant interneuron differ-
neurons, found within the central nervous system, con- ently. The firing rates are summed to determine the
nect neurons to each other. Interactions among direction of the gust of wind, and the cockroach will turn
 Concepts from Cellular Neurobiology 101

Third Thin axon

Nerve cord
thoracic of giant interneuron
Last ganglion
Abdominal segments
Cercus abdominal
ganglion

Cercus
Sensory Lambda
neuron Giant interneuron Leg cell
Filiform hairs motor
(wind receptors) neuron

Right hind leg

FIGURE 6.2 Some of the cells in the underlying neural circuitry of the cockroach escape response.
Note that the sensory neuron, also called the wind-receptor neuron, that is leaving the cercus makes
contact with a giant interneuron in the central nervous system, which in turn makes contact with
another interneuron that synapses with a motor neuron in the leg. (Modified from Camhi 1980 with
new information from Ritzmann 1986.)

and run in the opposite direction (Levi and Camhi 2000). nervous system that become wrapped around the axon
If the cockroach were to run directly away from the many times. Since a single glial cell encloses only a small
threat each time it escaped, the predator could predict region, about 1 mm, of an axon, the myelin sheath is not
the escape direction and capture the roach. Instead, the continuous. The regions along an axon between adjacent
cockroach escapes in one of a preferred set of paths away glial cells are exposed to the extracellular environment.
from the threat—usually about 90°, 120°, 150°, or 180° This arrangement is important to the speed at which the
away from the threat. This unpredictability in the escape nerve cell conducts messages. The message “jumps” suc-
path keeps the predator guessing (Domenici et al 2008). cessively from one exposed region to the next, increas-
Although there is no such thing as a typical neuron, ing the rate of transmission as much as 100 times. For
it is possible to identify characteristics common to some this reason, axons that conduct signals over long
neurons. We will use a motor neuron (Figure 6.3), in this distances are usually myelinated.
case from a mammal, as our example. The nucleus of a
motor neuron is contained in the cell body (soma), from
which small-diameter processes (neurites) typically extend.
THE MESSAGE OF A NEURON
In the traditional view, information enters a neuron via Let us now delve more deeply into the details of the how
a collection of branching neurites and then travels down and why of ion movements that are responsible for a
a single, long neurite to be passed on to other neurons. neuron’s message, called an action potential. An action
The neurites that receive the information are called potential is an electrochemical signal caused by electri-
dendrites; the single, long, cable-like neurite that trans- cally charged atoms, called ions, moving across the mem-
mits the information to other neurons is called an axon. brane. Ions can cross the membrane of a nerve cell by
In most vertebrates, some axons have a fatty wrapping means of either the sodium-potassium pump or ion
called the myelin sheath. In our example of the motor channels. The pump uses cellular energy to move three
neuron, the axon ends on a muscle or a gland (an effec- sodium ions (Na+) outward while transporting two potas-
tor), which brings about the animal’s behavioral sium ions (K+) inward. An ion channel, on the other
response. Although the terms dendrite and axon are well hand, is a small pore that extends through the membrane
established in the literature, it is now recognized that the of a nerve cell. There are different types of channels,
flow of information through a neuron is often not so each type forming a specific passageway for only one or
neatly divided into separate receiving and transmitting a few kinds of ions. Whereas some are passive ion chan-
processes. We will continue to use the terms, keeping in nels that are always open, others are active ion channels
mind that in many cases the specific direction of the (also called gated channels) that open in response to a
informational flow has not actually been demonstrated. specific triggering signal. Triggering signals may include
The myelin sheath is formed by the plasma mem- the presence of chemicals (neurotransmitters) in the
branes of glial cells, which are supporting cells in the space between the membranes of neurons, changes in the
102 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

(see discussion of synaptic transmission later in this chap-

Nucleus ter). Once an ion channel is open, the ion it admits may
move across the membrane in response to either a con-
centration gradient (ions tend to move from an area
where they are highly concentrated toward an area of
lesser concentration) or an electrical gradient (because
ions are charged atoms, they tend to move away from an
area with a similar charge and toward an area of the
opposite charge).
Cell body
Dendrites
The Resting Potential
In a resting nerve cell, one that is not relaying a message,
the area just inside the membrane is about 60 millivolts
Axon
(mV) more negative than the fluid immediately outside
the membrane. This charge difference across the mem-
brane is called the resting potential of the neuron. A
membrane in this resting state is described as polarized
Direction of (Figure 6.4).
impulse The resting potential results from the unequal dis-
tribution of certain ions across the membrane. The con-
centration of Na+ is much greater outside the neuron
than within. The concentration of K+ shows just the
opposite pattern, and is greater inside the cell than out-
Myelin sheath
side. Certain large, negatively charged proteins are held
within the neuron either because the membrane is
impermeable to them or because they are bound to intra-
cellular structures. These proteins are primarily respon-
sible for the negative charge within the neuron.
Most of the active ion channels in the membrane of
a resting neuron are closed, but passive channels are, of
course, open. Because most of the passive channels are
specific for K+, the membrane is much more permeable
to K+ than it is to other ions. Drawn by the negative
charge within, positively charged K+ will enter the neu-
ron and accumulate there. At some point, when there is
roughly 20 to 30 times more K+ inside than outside, the
concentration gradient counteracts the electrical gradi-
ent. When the two forces—an electrical gradient that
Muscle
draws K+ inward and a concentration gradient that
pushes K+ outward—are equally balanced and there is no
further net movement of K+; the cell has reached its rest-
ing potential.
Why is Na+ more concentrated outside the neuron?
FIGURE 6.3 A motor neuron. The soma or cell body Although Na+, like K+, is attracted by the negative charge
maintains the cell. The dendrites are extensions special- inside the neuron, the membrane is relatively imperme-
ized for receiving input from other cells. The axon is able to it, and so only a few Na+ can leak through.
specialized to conduct the message away from the cell Furthermore, the sodium-potassium pump actively
and to release a chemical that will communicate with removes Na+ from within the cell, transporting it out-
another cell. In vertebrates, some axons are covered by ward against electrical and concentration gradients.
a fatty myelin sheath.

The Action Potential

charge difference across the membrane (the membrane The action potential (nerve impulse) is an electro-
potential), changes in the concentration of intracellular chemical event that lasts about 1 millisecond. The action
calcium ions (Ca++), or any combination of these factors potential consists of a wave of depolarization followed by
 Concepts from Cellular Neurobiology 103

Resting potential

a Polarity b Distribution of ions

+40 mV
0 mV
–60 mV

+ + + + + + + + + + Na+ K+ Na+ K+ Na+ Na+ K+ Na+

– – – – – – – – – – K+ Na+ K+ Na+ K+ K+ Na+ K+
– – – –
– – – – – – – – – – K+ K+ Na+ K+ K+ Na+ K + K+
+ + + + + + + + + + Na Na+ K+ Na+ Na+ K+ Na + Na+
+

Organic ions

FIGURE 6.4 The resting potential. (a) In the resting state, the inside of an axon is more
negative than the outside. (b) This charge difference is caused by the unequal distribu-
tion of ions inside and outside the cell. There are more sodium ions outside and more
potassium ions inside. In addition, there are large, negatively charged proteins
(organic ions) held inside the cell, giving the interior an overall negative charge.

repolarization that spreads along the axon. The depo- to K+. Potassium ions then leave the cell, driven by the
larization, or loss of the negative charge within, is caused temporary positive charge within and by the concen-
by the inward movement of Na+. However, the repolar- tration gradient. The exodus of K+ restores the nega-
ization, or restoration of the negative charge within the tive charge to the inner boundary of the membrane. In
neuron, is caused by K+ leaving the cell (Figure 6.5). fact, enough potassium ions may leave to temporarily
Let’s see how depolarization and repolarization make the cell’s interior even more negatively charged
occur. The membrane becomes slightly depolarized than usual, a condition called hyperpolarization.
when some of the active Na+ channels open and Na+ Notice that although the original resting potential is
enters the cell, drawn by both electrical and concen- eventually restored, the distribution of ions is differ-
tration gradients. The positive charge on Na+ slightly ent. This situation is corrected by the sodium-potas-
offsets the negative charge inside the cell, and the sium pump, which moves K+ back in and Na+ back out
membrane becomes slightly depolarized. If the depo- of the cell.
larization is great enough, that is, if threshold is This depolarization and repolarization of the neu-
reached, voltage-sensitive sodium channels open and ronal membrane spreads rapidly along the axon, gener-
Na+ ions rush to the interior of the cell. At roughly the ated at each spot in the same manner in which it was
peak of the depolarization, about 0.5 millisecond after started. The local depolarization at one point of the
the voltage-sensitive sodium gates open, they close and membrane opens the voltage-sensitive sodium channels
cannot reopen again for a few milliseconds. Almost in the neighboring region of membrane, thereby trig-
simultaneously, voltage-sensitive potassium channels gering its depolarization. The net result is that a wave
open, greatly increasing the membrane’s permeability of excitation travels down the axon.

Action potential

Polarity Distribution of ions

+40 mV
0 mV +40 mV +40 mV
0 mV 0 mV
–60 mV
–60 mV –60 mV
+ + – + + + + + + + +
Na Na Na + +
Na+ Na+
– – +
– – – – – – – – K+ K+ K+ K+ K+ K+
– Na+ – Na+
– – + – – – – – – – K+ K+ K+ K K K+
+ +

+ + – + + + + + + +
Na Na+ Na+
+
Na+ Na+
Direction of Depolarization Repolarization
impulse

FIGURE 6.5 The action potential of a neuron. Depolarization is caused by sodium ions
entering the cell, and repolarization is caused by potassium ions exiting the cell.
104 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

Immediately after an action potential, the neuron contractions. Furthermore, the muscle contractions
cannot be stimulated to fire again for 0.5 to 2 milliseconds were uncoordinated because the release of the transmit-
because the sodium channels cannot be reopened right ter from different neurons was asynchronous (Jan et al.
away. At the start of this absolute refractory period, no 1977). Recall that a neuron’s message is an electro-
amount of stimulation can generate an action potential. chemical signal consisting of a wave of depolarization
During the latter part of this interval, the relative refrac- caused by sodium ions (Na+) entering the nerve cell fol-
tory period, stimulation must be greater than the usual lowed by a wave of repolarization caused by potassium
threshold value to generate an action potential. Although ions (K+) leaving the nerve cell. By recording from the
the refractory period is brief, it is biologically significant neurons of adult Shaker flies, it was shown that the
because it determines the maximum rate of firing. mutant neurons do not repolarize as quickly as normal
Because action potentials are generated anew as they neurons (Tanouye et al. 1981). As a result, an excess of
travel down the neuron, there is no loss in their magnitude: calcium ions enters the neuron and causes the release of
an action potential has the same amplitude at its point of neurotransmitter to continue longer than is typical.
initiation as at every other point along the neuron. We Why doesn’t the cell repolarize normally? Apparently,
therefore describe the action potential as an all-or-none the Shaker gene codes for a protein that forms part of the
phenomenon: it either occurs or it doesn’t, and doesn’t potassium channels involved in repolarization; a mutation
occur halfway. This characteristic makes action potentials at the Shaker locus results in certain potassium channels
well-suited to be the nervous system’s long-distance signal. not being formed, and this disrupts the process of repo-
If an action potential is always the same, how can larization (Kaplan and Trout 1969; Molina et al. 1997).
differences in the intensity of stimuli be sensed? The That the behavioral defect, that is, shaking, results from
intensity is encoded in the firing rate of the neuron and the absence of K+ channels was elegantly shown by exper-
by the number of neurons responding. For example, if iments in which a functional Shaker gene was inserted into
we place our hand on a hot stove rather than on a warm mutant flies. This experimental procedure resulted in a
one, the firing rates of neurons in our hand that respond normal flow of potassium across the membranes of nerve
to heat or pain may be increased. Also, touching a very cells and an end to the jittery behavior caused by the
hot stove will activate more neurons than will touching mutations (Zagotta et al. 1989).
a warm stovetop because the thresholds of neurons that
register heat vary—more neurons reach their thresholds
and fire at higher temperatures. BEHAVIORAL CHANGE AND
SYNAPTIC TRANSMISSION
IONS, MEMBRANE PERMEABILITY, Synapses are important structures because they are deci-
AND BEHAVIOR sion and integration points within the nervous system.
Although ions and their movements through the chan- The molecular events that occur at synapses determine
nels of nerve cell membranes may seem, at best, to be whether the message of one neuron will generate an
only remotely related to an animal’s behavior, we will see action potential in the next neuron. Typically, a neuron
that this is not the case. Let’s consider how ions and receives input through thousands of synapses. That input
changes in membrane permeability relate directly to is integrated in ways that make complex behaviors pos-
what we see an animal doing. Here we describe an exam- sible. Changes in the functioning of synapses or in the
ple of how changes in the membrane permeability caused number of synaptic connections often explain why
by mutation in the Shaker gene of the fruit fly (Drosophila behavior can change because of experience or matura-
melanogaster) produce atypical behavior. tion. As we will see shortly, the gill-withdrawal reflex of
When fruit flies are anesthetized with ether, one the sea hare Aplysia can be modified through various
occasionally sees a mutant fly that shakes its legs, wings, forms of learning, each involving changes in how neu-
or abdomen. Among the mutations that result in shak- rons communicate with one another at synapses. We will
ing under ether anesthesia are Shaker, hyperkinetic, and also learn how the modification of synaptic function
ether-a-go-go. All this shaking is apparently a result of caused by chemicals called neuromodulators changes leg
neurons with mutations that make them exceptionally movements during swimming and courtship in the blue
excitable. More is known about what makes the Shaker crab. But before exploring the cellular mechanisms for
mutants so jittery, so we will concentrate on them. behavioral change, we should become more familiar with
It was first shown that the Shaker larvae were jittery the structure of synapses.
because an excessive amount of neurotransmitter, a
chemical released by a neuron that allows communica-
THE STRUCTURE OF THE SYNAPSE
tion with another neuron or a muscle cell (discussed
shortly), was released at the junction between a motor The gap between neurons is called a synapse, and at a
neuron and a muscle cell, causing extreme muscle specific synapse information is typically transferred in
 Behavioral Change and Synaptic Transmission 105

Presynaptic

2 nm

b Presynaptic
membrane
Terminal
Presynaptic bouton

Neurotransmitter
molecules

Receptor
20 nm macromolecules
Synaptic
cleft
Postsynaptic
membrane
Postsynaptic

FIGURE 6.6 The gap between neurons is called a synapse. There are two types of
synapses: (a) electrical and (b) chemical. In an electrical synapse, the distance between
neurons is very small, and ions can flow directly from one neuron to the next. In a
chemical synapse, however, the distance between neurons is greater, and one neuron
affects the activity of the other by releasing chemicals into the gap between them.
The original distribution of ions is later restored by the sodium-potassium pump.

one direction, from the presynaptic neuron to the potential difference between the inside of this cell and
postsynaptic neuron. (The descriptor presynaptic or the postsynaptic cell. Because of the difference, positively
postsynaptic refers to the direction of information flow at charged ions, mostly K+, move from the presynaptic cell
a specific synapse.) There are two major categories of through the tiny tubular connections into the postsy-
synapses: electrical and chemical (Figure 6.6). naptic neuron. These newly arriving ions may suffi-
ciently depolarize the postsynaptic cell to induce an
action potential.
Electrical Synapses
We will discuss electrical synapses first because they are
much less common and the behavioral examples that fol- Chemical Synapses
low pertain to chemical synapses. Electrical synapses are Chemical synapses are characterized by a larger space
known for their speed of transmission. Whereas a signal between the membranes of the two neurons (typically
may cross an electrical synapse in about 0.1 millisecond, 20–30 nm) than is found in electrical synapses. Rather
durations about 0.5 or 1 millisecond are typical at chemi- than information being transmitted from one neuron to
cal synapses. Not surprisingly, then, electrical synapses are the next via direct electrical connections, it is transmit-
often part of the neural circuitry that underlies patterns of ted across a chemical synapse in the form of a chemical
behavior when sheer speed is essential, such as the escape called a neurotransmitter. There are several steps in
responses exhibited by animals confronted with a predator. this process, and these steps account for the slower speed
In an electrical synapse, the gap between the neu- of transmission at a chemical synapse than at an electri-
rons is small, only about 2 nanometers (1 nm = 10-9 m), cal one. First, the action potential travels down the axon
and is bridged by tiny connecting tubes that allow ions to small swellings called terminal boutons at the end of
to flow directly from one neuron to the other. When an the axon. Second, at the terminal boutons, the action
action potential arrives at the axon terminal of the potential causes the neuron to release a neurotransmit-
presynaptic neuron, Na+ enters this terminal, causing a ter from small storage sacs called synaptic vesicles.
106 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

Third, the release of neurotransmitter occurs because K+

the action potential opens voltage-sensitive calcium ion Na+ Ca++
channels. Fourth, the calcium ions that flood to the
inside initiate events that cause the synaptic vesicles to
move toward the membrane of the terminal bouton.
Fifth, once there, the vesicles fuse with the membrane
and dump their contents into the gap between the cells
K+
(the synaptic cleft). Finally, the neurotransmitter then
diffuses the short distance across the cleft and binds to 15 mV
receptors on either another neuron or a muscle cell.
5 msec
When the neurotransmitter binds to a receptor on
the postsynaptic cell, it either directly opens ion gates or FIGURE 6.8 The events at an inhibitory synapse. The
indirectly affects ion gates through biochemical mecha- neurotransmitter binds to a receptor and causes the
nisms. In either case, the neurotransmitter will increase opening of channels that allow potassium ions to leave
the permeability of the membrane of the postsynaptic the postsynaptic cell. Thus it becomes hyperpolarized.
cell to specific ions. This results in either excitation or This hyperpolarization is called an IPSP.
inhibition of the postsynaptic neuron, depending on the
particular ions involved. In the case of excitation, the
neurotransmitter causes the opening of channels that or the influx of negatively charged chloride ions makes
allow both Na+ and K+ to pass through. Although some the inside of the postsynaptic membrane even more
K+ ions move out, they are outnumbered by the Na+ ions negative than usual. In other words, an inhibitory
moving in. This causes a slight temporary depolarization neurotransmitter momentarily hyperpolarizes the mem-
of the membrane of the postsynaptic cell. This depolar- brane. The hyperpolarization is called an inhibitory
ization is called an excitatory postsynaptic potential, or postsynaptic potential, or IPSP (Figure 6.8).
EPSP (Figure 6.7). If this depolarization reaches a As long as the neurotransmitter remains in the
certain point, the threshold, an action potential is synapse, it will continue to excite or inhibit the postsy-
generated in the postsynaptic cell. naptic cell. However, the effect of a neurotransmitter can
On the other hand, a neurotransmitter may act in be halted in a variety of ways. In some cases, the neuro-
an inhibitory fashion, making it less likely that an action transmitter is broken down by an enzyme, and its com-
potential will be generated in the postsynaptic neuron. ponent parts are absorbed for resynthesis. In other
When the neurotransmitter binds to the receptors in an instances, however, the molecules of the neurotransmit-
inhibitory synapse, either K+ channels or K+/Cl- chan- ter are released intact after acting on the postsynaptic
nels open. The exit of positively charged potassium ions cell, absorbed by the presynaptic cell, and repackaged for
subsequent release.

a Na+
INTEGRATION
Na+ Ca++
Most neurons receive input from many other neurons.
In fact, a given neuron may communicate with hundreds
or thousands of other cells. The slight depolarization
(EPSPs) and hyperpolarization (IPSPs) that result from
input from all the synapses are summed on the postsy-
Na+ naptic membrane, either because one neuron sends a
repeated signal or because many neurons send messages
to one postsynaptic cell. In other words, the EPSPs and
b c Action
potential
IPSPs combine with one another as they arrive at the cell
EPSP
body. If these interacting changes in membrane poten-
15 mV EPSP
tial combine to produce a large enough depolarization,
5 msec voltage-sensitive Na+ gates are opened and an action
potential is triggered.
FIGURE 6.7 The events at an excitatory synapse. (a) The
neurotransmitter binds to a receptor and causes the
Neuromodulators are chemicals that cause voltage
opening of channels that allow sodium ions to enter the changes that occur over a slower time course than those
cell, thereby slightly depolarizing the postsynaptic cell. caused by “traditional” neurotransmitters—seconds, min-
(b) This slight depolarization is called an EPSP. (c) If the utes, hours, and perhaps even days. The fast changes are
depolarization reaches the threshold, an action potential brought about by traditional neurotransmitters—those
is generated. chemicals, previously discussed, that open ion gates,
 Specializations for Perception of Biologically Relevant Stimuli—Sensory Processing 107

causing EPSPs or IPSPs. In contrast, neuromodulators Neuromodulators, combined with the proper
alter neuronal activity slowly, by biochemical means. olfactory stimulation, affect whether the crab will per-
The effects of neuromodulators appear to be mediated form the courtship display instead of the two swimming
by substances within the postsynaptic neuron called sec- behaviors. When the neuromodulators proctolin,
ond messengers (Breedlove et al. 2007). These second dopamine, octopamine, serotonin, and norepinephrine
messengers (e.g., calcium and the cyclic nucleotides are separately injected into blue crabs, each drug pro-
cAMP and cGMP couple the membrane receptors of the duces a unique posture or combination of limb move-
postsynaptic cell with the movements of ions through one ments. The postures and limb movements are the same
or more enzymatic steps. Neuromodulators may, for exam- as those observed in freely moving, untreated crabs.
ple, upon reaching the receptor on the postsynaptic neu- Injection of dopamine produces the posture of the
ron, trigger the formation of the second messenger cAMP courtship display, in which the male stands high on the
within the neuron, which in turn activates an enzyme that tips of his walking legs; injection of proctolin produces
changes the shape of proteins in certain ion channels. Once the rhythmic leg movements characteristic of courtship
the ion channels have been altered in this manner, the per- that serve to fan chemicals in his urine (pheromones)
meability of the membrane to specific ions is also changed, toward a female to attract her (Wood 1995b).
thereby affecting the activity of the neuron. It is the rela- Electrical stimulation of specific neurons under dif-
tively slow pace of the enzymatic activities that produces ferent conditions has identified interneurons in the
the typically prolonged effects of neuromodulators. esophageal connectives (neural connections between the
Functionally, neuromodulators appear to be inter- brain and ventral nerve cord) that trigger rhythmic wav-
mediate to classic neurotransmitters and hormones. ing of the swimming legs. Some of these interneurons
Whereas neurotransmitters are released at specific trigger rhythmic leg waving when sex pheromone is
synaptic clefts and hormones are broadcast throughout applied to the antennule of the crab. Under these con-
the body via the bloodstream, neuromodulators are ditions, the leg waving is not distinctly characteristic of
released in the general vicinity of their target tissue. It any of the three rhythmic behavior patterns. However,
is, however, difficult to establish the precise point at when proctolin is applied while these interneurons are
which a neurotransmitter becomes a neuromodulator, being stimulated, the motor output changes to the rhyth-
and a neuromodulator a hormone. In fact, the same mic waving of the courtship display (Wood 1995b). The
chemical may have different functions in different natural source of the proctolin that initiates courtship leg
places. Some chemicals, dopamine, for instance, act as movements is thought to be a cluster of nerve cells in the
neurotransmitters at some synapses and as neuromodu- subesophageal ganglia (Wood et al. 1996).
lators at others. Similarly, a chemical may act locally in
the nervous system as a neurotransmitter, whereas in
other places in the body it is released into the blood-
stream and acts at a distant site as a hormone does (see SPECIALIZATIONS FOR
Chapter 7 for a discussion of hormones). Despite some PERCEPTION OF BIOLOGICALLY
fuzziness in definition, there is no question that neuro- RELEVANT STIMULI—SENSORY
modulators, through their actions on neurons, glands, PROCESSING
and muscles, can produce profound effects on behavior.
Consider the behavior of the male blue crab In spite of the wide variety of stimuli bombarding an
(Callinectes sapidus) as an example of neuromodulation of animal from its environment, it is able to detect only a
rhythmic movements of the swimming legs. Just before limited range, and of those that it detects, it may ignore
a female blue crab matures, she releases a pheromone, a all but a few key stimuli. The job of an animal’s sensory
chemical used to communicate with other members of system is not to transmit all available information but
the species, in her urine. When a male blue crab senses rather to be selective and provide only information that
the pheromone, he begins his courtship display. He is vital to the animal’s lifestyle or, more to the point,
spreads his claws apart in front, extends his walking legs, information that influences its reproductive success.
and raises his swimming legs in the rear. He then waves In some cases, the receptors themselves are “tuned”
his paddle-shaped swimming legs from side to side above to detect biologically relevant stimuli. Consider, for
the carapace (Kamio et al. 2008). Besides courtship, two example, the relationship between the sensitivity of pho-
other distinct stereotyped behaviors, sideway swimming toreceptors and the dominant wavelengths of light in the
and backward swimming, involve the rhythmic move- habitat of certain teleost fish. The wavelengths of light
ment of swimming legs. In each behavior, the swimming that actually reach the eye of a fish will depend on many
legs are waved in a slightly different way and the crab factors. The color of the water is one such factor, and it
assumes a different posture. However, because these varies among habitats. When the sensitivity of the eyes
three behaviors are so similar, it is likely that they share of fish from different habitats is compared, we see that
common neural elements (Wood 1995a). the sensitivity to different wavelengths (colors) of light
108 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

has been adjusted by natural selection so that peak What properties of sound enable its source to be
sensitivity occurs at the most common wavelength of located? Actually, part of the answer is remarkably sim-
light in a given habitat (Bowmaker et al. 1994; Lythgoe ple: Sounds can be localized by how loud they are, that
et al. 1994). We will see more examples of sensory tun- is, by their intensity. A simple rule might be that sound
ing when we explore the evolution of communication in seems louder when the receptor is closer to the sound.
Chapter 17. If only one ear is involved in locating the sound source,
Besides detecting stimuli critical to survival, animals however, the rule may not hold (Camhi 1984). Let’s say
must be able to pick out that stimulus from a back- that the left ear hears a soft sound; was the sound soft
ground of “noise” in the same sensory modality. For because it was produced by a weak source on the left side
example, recall that the cockroach runs when exposed or by a strong source on the right side? To eliminate such
to a gust of wind from a predator. Yet, it ignores non- confusion, both ears must be used in the sound local-
threatening or irrelevant sources of wind, such as the ization process—this is called binaural comparison.
wind that it creates itself while walking. How is such Some animals use binaural comparison of sound inten-
selectivity possible? In the laboratory, cockroaches run sity to locate the source of sound.
when exposed to wind with a peak velocity of only 12 Timing is also important in locating the source of
mm per second, the approximate velocity of wind cre- a sound. Two differences in timing could be of poten-
ated by the lunge of a predator such as a toad. tial use, and both rely on binaural comparison. The
Cockroaches, however, do not respond to the 100-mm- first occurs at the onset of sound—sound begins and
per-second wind that they create by normal walking. ends sooner in the ear that is closest to the source. The
How is it that cockroaches manage not to respond to the second difference in timing occurs during an ongoing
relatively strong wind created by walking and yet run sound. During a continuing sound, there are differ-
when exposed to much softer wind signals? It turns out ences in the phase (the point in the wave of compres-
that it is not the velocity of the wind that is the critical sion or rarefaction) of the sound wave reaching each
factor but the acceleration (rate of change of wind ear. The extent of the phase difference will depend
speed), and a strike by a predator delivers wind with both on the wavelength of the sound and on the dis-
greater acceleration than the wind produced by the step- tance between the ears. When the wavelength of the
ping legs of a cockroach. In fact, when cockroaches were sound is twice the width of the head, the peak of a
tested with wind puffs that had the same peak velocity sound wave arrives at one ear and the trough arrives
but differed in acceleration, they ran more frequently at the other. Under these conditions, the sound is eas-
when exposed to winds of higher acceleration (Plummer ily localized. In contrast, when the wavelength of the
and Camhi 1981). Winds with low acceleration typically sound equals the head width, the phase of the sound
produced no response. Thus, cockroaches appear to pay wave is the same in each ear, and the sound is difficult
particular attention to the acceleration of the wind stim- to localize (Figure 6.9).
ulus, and this allows them to ignore irrelevant wind sig-
nals and to focus on important information in their
environment.

PROCESSING OF SENSORY
INFORMATION FOR SOUND
LOCALIZATION
We will consider the mechanisms of sound localization
as an example of stimulus processing of biologically
important stimuli. It is often important for an organ-
ism to locate the source of sound. For example, a
potential mate may be producing the sound. A male Wavelength = Wavelength = Wavelength = less
mosquito finds a female by the sound of her beating two times head width head width than two
times head width
wings. It would do a male little good to know that a
female was present and be unable to locate her. Most useful for No phase Confusion
Similarly, many predators determine their prey’s posi- phase information difference
tion by localizing sounds generated by the prey. FIGURE 6.9 Binaural comparison of phase. When the
Locating the source of a sound has obvious importance sound is prolonged, differences in the phase of the
to prey animals as well—the crunching sound of brush sound wave at each ear may indicate the direction of
under a leopard’s foot has fixed its position for many the source. The usefulness of this cue depends on the
a wary baboon. wavelength and the distance between the ears.
 Specializations for Perception of Biologically Relevant Stimuli—Sensory Processing 109

PREDATORS AND PREY: same way that your car’s radar detector alerts you of a
THE NEUROETHOLOGY police radar trap.
OF LIFE-AND-DEATH STRUGGLES How does the moth’s nervous system analyze the
available information and direct effective evasive
Now let’s consider how nervous systems gather and maneuvers? The sensitive A1 cell responds to the
process information about the source of sounds to pro- sounds of a distant bat, and its input reveals the direc-
duce adaptive behaviors—escape behavior by prey and tion and distance of the bat (Figure 6.10). If the bat, for
prey localization by a predator. We will first consider example, is on the left side, the left A1 cell is exposed
how sound information is processed by the relatively to louder sounds because the A1 cell on the right is
simple nervous system of a night-flying (noctuid) moth, somewhat shielded by the moth’s body. Therefore, the
allowing it to escape from an echolocating bat. Then we left receptor fires sooner and more frequently upon
will consider how a barn owl obtains and processes sound receiving each sound of the bat. When the bat is directly
information to locate its prey. behind or in front of the moth, both neurons will fire
simultaneously. A slight turn of the moth’s body will
then result in differences in the right and left receptors,
Escape Responses of Noctuid Moths which will reveal whether the bat is approaching from
Noctuid moths are a favorite prey of certain bats. Indeed, the front or rear. What about its altitude? If the bat is
moths typically make up more than half of a bat’s diet. above the moth, the bat’s sounds are louder during the
The bats capture their prey on the wing, locating these upward beat of the moth’s wings when the moth’s ears
flying insects by echolocation—that is, by emitting high- are uncovered than when the moth’s wings are down,
frequency sounds that bounce back to the bat from any covering the ears and muffling the bat’s cries. However,
structure in the environment. Here we will focus on how if the bat is beneath the moth, the bat’s echolocation
moths escape predation. cries will reach the moth’s ears unimpeded regardless of
Kenneth Roeder (1967) has provided a fascinating the position of the moth’s wings. Therefore, the moth’s
account of how the relatively simple auditory apparatus wingbeats will have no effect on the pattern of neural
of the moth is used to detect an approaching bat and how firing. The moth, then, is able to decode the incoming
the moth then takes evasive action. When the bat’s ultra- data, so that it detects both the presence and precise
sonic echolocation pulses are soft, indicating that the bat location of the bat.
is still at a distance, the moth turns and flies directly How is this information processed to produce an
away. However, loud ultrasonic pulses mean that the bat appropriate escape pattern? If the bat is passing some
is very close, and emergency actions are needed—erratic, distance away, the A1 cell begins to fire. Its firing rate
unpredictable looping and wingfolding to produce a free will increase as the bat gets closer and its cries become
fall. Moths that hear a bat’s approach and take evasive louder. Up to a certain firing rate of the A1 cell, the dis-
action are about 40% less likely to be eaten. tance between predator and prey is too great for the bat
Roeder found that these moths have two ears, one to detect the moth. Therefore, the most adaptive
on either side of the thorax (the insect’s midsection), and response of the moth would be to turn and fly directly
that each ear has only two auditory receptor cells away, thus decreasing the likelihood of detection by
(Roeder and Treat 1957). The receptors are tuned to the increasing the intervening distance and by exposing less
frequencies of the echolocation calls of species of bats surface area to the bat. This escape pattern results when
living in their vicinity, which is generally between 20 and the moth turns its body until the A1 firing from each ear
50 kHz. One receptor, called the A1 cell, is about ten is equalized. When the bat changes direction, so does the
times more sensitive than the other cell, the A2 cell. The moth (Roeder and Treat 1961).
A1 cell begins to respond when the sound is soft, indi- Bats fly faster than moths, though, and if the bat gets
cating that the bat is still at a distance. The sensitivity of too close, then the moth’s evasive maneuver switches to
this cell is important because it will determine how much an erratic flight pattern. The moth’s wings begin to beat
time the moth will have to take appropriate evasive in either peculiar, irregular patterns or not at all. The
action. The A2 cell responds only to loud sounds (Pérez insect itself probably has no way of knowing where it is
and Coro 1984; Roeder and Payne 1966), as would come going as it begins a series of loops, rolls, and dives. But
from a nearby bat. it is also very difficult for the bat to pilot a course to
The moth responds to bat sounds long before the intercept the moth. If the moth crashes into the ground,
bat can detect the moth (Roeder and Payne 1966). North so much the better. It is safe here because the earth will
American moths can detect a hunting big brown bat mask its echoes.
(Eptescicus fuscus) from a distance of nearly 100 ft, whereas How does the moth determine whether the bat is
the bat must be within about 15 ft to detect a moth-size gaining on it? One clue is that the sound of an
target (Fenton 1992). The A1 cell, then, warns the moth approaching bat grows louder. Recall that the A2 cell is
that there is a hunting bat in the vicinity, in much the less sensitive than the A1 cell and doesn’t begin to fire
110 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

a b c Wings up Wings down

Sound pulses
A1 cell firing

A1 cell firing Sound pulses A1 cell firing

Sound stimulus
A1 cell firing A1 cell firing

FIGURE 6.10 The relationship of sound pulses from a hunting bat and an auditory neural firing in the
hunted moth. (a) When a hunting bat, emitting its high-pitched sounds, approaches a noctuid moth
from the side, the receptors on that side fire slightly sooner and more rapidly than those on the
shielded side. (b) When the bat is behind the moth, the moth’s receptors on both sides fire with a simi-
lar rapid pattern. (c) When the bat is above the moth, the moth’s auditory receptors fire when its wings
are up but not when its wings cover the receptors on the down stroke. (Redrawn from Alcock 2001.)

until the bat is close by. Based on these differences in when the intensity of the bat’s call increased.
threshold, Roeder suggested that the A1 cell functions Notodontid moths have a single type of A cell, but they
as an “early warning” cell and the A2 cell as an “emer- display both types of evasive behavior (Surlykke 1984).
gency” neuron that switches the moth’s evasive response Thus in noctuid moths, which have two auditory cells,
to an erratic flight pattern. As reasonable as the hypoth- the A2 cell does begin firing when the bat is nearby, but
esis seems, it is not consistent with the data. One would this activity may not be responsible for the change to
predict, for instance, that if the activity of the A2 cell erratic flight.
was the switch that changes the evasive response from Another clue to the bat’s proximity is provided by
flying directly away to erratic flight, then a moth with the type of echolocation sounds the bat produces
only one type of A cell would not switch to erratic flight because these change during the hunt. While the bat is
 Specializations for Perception of Biologically Relevant Stimuli—Sensory Processing 111

searching for prey, its pulses are relatively long (about locate the source of sound, this nocturnal predator can
10 ms) and are repeated slowly (about 10 per second). pinpoint its prey by the rustlings the prey makes, and it
When prey has been detected, the bat switches to the can precisely determine not only the prey’s location
approach phase of the hunt. The sound pulses get along the ground but also its own angle of elevation
shorter (about 5 ms) and are repeated more rapidly above the prey.
(about 20 per second). In the final approach, which How do we know that the hunting owl uses the
begins when the bat is within a meter of its prey, the bat prey’s sound? For one thing, we know that barn owls can
begins a feeding buzz, consisting of short pulses (0.5 to catch a mouse in a completely darkened room (Payne
2 ms) repeated rapidly (100 to 200 per second) (Boyan 1962). In experiments, a barn owl was able to capture a
and Miller 1991). skittering leaf pulled along the floor by a string in a dark
The response of the A1 cell can follow the bat’s call room (indicating that sight and smell are not involved),
rates at all phases of the hunt up to about 150 ms before and if unable to see, it will leap into the middle of an
the bat would capture the moth (Fullard et al. 2003). expensive loudspeaker from which mouse sounds
Since the call rate changes as the bat gets closer to its emanate.
prey, the output of the A1 cell provides information To locate its prey by using sound cues, the barn owl
about the distance of the bat. The A1 cell sends this must place the source of the sound on a horizontal plane
information directly to two interneurons, called 501 and from left to right (i.e., its azimuth), as well as on a ver-
504. These interneurons respond differently to the same tical plane (i.e., its elevation). We now know that a barn
input from A1. The differences in interneuron responses owl uses different cues for locating sound cues in hori-
somehow encode information about the distance of the zontal and vertical planes.
bat and direct the appropriate escape response (Boyan The owl uses time differences in the arrival of
and Fullard 1986; Boyan and Miller 1991). sound in each ear to place it on a horizontal plane and
differences in intensity between the two ears to deter-
mine the elevation of the sound source (reviewed in
Prey Localization by Barn Owls
Konishi 2003). Masakazu (Mark) Konishi (1993a, b)
Silently and suddenly, a barn owl (Tyto alba) sweeps from learned this by playing sound in a barn owl’s ear through
the sky to strike its prey with astonishing accuracy small earphones. An owl turns its entire head to face the
(Figure 6.11). How does it find its prey? Although in direction from which it perceives the sound source,
nature the barn owl’s keen night vision is important in because its eyes are fixed in their sockets. When the
locating prey, the sounds of a scurrying mouse are sound in one ear preceded that in the other, the owl
sufficient for the owl to strike with deadly precision. turned its head in the direction of the leading ear. The
Laboratory tests have revealed that birds such as the barn longer the time difference, the further the owl turned
owl are able to locate the source of sounds within 1° or its head.
2° in both the horizontal and vertical planes (1° is The intensity differences in the two ears vary with
approximately the width of your little finger held at arm’s the elevation of the sound source largely because of the
length). Because of its astounding ability to detect and arrangement of the ear canals and facial feathers (von
Campenhausen and Wagner 2006). The two ear canals
that channel the sound toward the inner ears are, oddly
enough, situated asymmetrically, with the right one
higher than the left. Because of this difference in ear
placement, each ear responds differently to a sound at
a given elevation. This helps the owl determine its own
elevation above the sound source, information critical
to an aerial predator. Also, the face of the barn owl is
composed of rows of densely packed feathers, called the
facial ruff, that act as a focusing apparatus for sound
(Figure 6.12). Troughs in the facial ruff, like a hand
cupped behind the ear, both amplify the sound and
make the ear more sensitive to sound from certain
directions.
The facial ruff assists the owl in localizing sounds by
creating differences in intensity of the sound in both
ears. Loudness is a cue to localizing the sound in both
the horizontal and the vertical dimensions. Sound is gen-
FIGURE 6.11 A hunting barn owl. A barn owl can locate erally louder in the ear closer to the source. Because of
its prey by using sound cues alone. the structure of the facial ruff, the left ear collects
112 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

to both of these nuclei. Whereas the branch of the audi-

tory nerve that goes to the magnocellular nuclei conveys
timing information, the branch to the angular nuclei
transmits intensity information. Thus, the timing data
that place the sound on a horizontal plane are processed
separately from the intensity data that place the sound on
a vertical plane. These different features of sound are
processed in parallel along nearly independent pathways
to higher processing stations, where a map of auditory
space is eventually formed (Konishi 1993b).
The map of auditory space is formed in the exter-
nal nucleus of the inferior colliculus of the midbrain.
Within the inferior colliculus are certain neurons that
respond selectively to specific degrees of binaural dif-
ferences in sound (Figure 6.13). For example, one neu-
ron may respond maximally to differences that
correspond to a sound originating 30° to the right of the
owl. The sound would arrive a certain amount sooner
and be a certain degree louder in the right ear than in
0
30 L 30 R
+90
+60
60 L 60 R

1
+30

90 L 1 0 90 R
FIGURE 6.12 The barn owl, a night hunter, has a facial 1
1
disc of densely packed feathers that may gather sounds 2 2
2
and aid in detecting their source. 3
–30
3 2
3
4
4
3 –60
low-frequency sounds primarily from the left side, and –90
the right ear collects low-frequency sounds from the Lateral
right side. A comparison of the intensity of low-fre- Posterior Anterior
quency sounds in each ear helps the owl determine from
Medial
which side of the head the sound originates. However,
the facial ruff channels high-frequency sound to each ear
differently, depending on the elevation of the sound Optic
10L 5L 0
20L15L tectum
source. As a result, the right ear is more sensitive to high- 25L 10R5R
40L 15R
frequency sounds that originate above the head, and the Left MLD
left ear is more sensitive to high-frequency sounds from
below the head. The owl compares the loudness of high-
frequency sounds in each ear to determine its position FIGURE 6.13 Auditory neurons in the midbrain area (the
above or below the sound source. As a sound source inferior colliculus) of a barn owl. The top figure shows a
moves upward from below the bird to a position above hemisphere of space in front of the owl’s head. Neurons
the owl’s head, the high-frequency sounds would first be in the inferior colliculus respond to sounds that origi-
loudest in the left ear and then gradually become louder nate at different points. The numbered rectangles
in the right ear (Knudsen 1981). indicate 14 areas to which specific inferior colliculus
neurons are tuned. The lower figure indicates the
Information on the timing and loudness of sounds in
manner in which the auditory space of the owl is
each ear is then sent to the owl’s central nervous system
represented in the inferior colliculus. A horizontal
over the auditory nerve in a pattern of nerve impulses. section of the inferior colliculus is shown with bars,
The information is first sent to the cochlear nuclei. Each indicating the position of specific neurons. The point
side of the brain (cerebral hemisphere) has two cochlear in space to which that neuron responds is indicated.
nuclei: the magnocellular nucleus and the angular Notice that the neurons in the inferior colliculus are
nucleus. Every axon in the auditory nerve sends a branch spatially organized. (From Knudsen and Konishi 1978.)
 Processing in the Central Nervous System 113

regions. The pattern of activity among the brain regions

in this network determines social behavior.

BRAIN CHANGES UNDERLYING

BEHAVIORAL CHANGE
When behavior changes because of the animal’s experi-
ence, we usually say that the animal has learned some-
thing. However, we cannot be sure that learning has
occurred unless we can elicit a memory. There are sev-
eral stages of memory formation. Short-term memo-
ries can be retrieved for minutes, hours, or perhaps a
day. Short-term memories may then be strengthened to
form intermediate-term memories, which may last a
FIGURE 6.14 A young owl wears goggles that shift its
visual field about 10° to the left. When reaching toward day or so. We will discuss habituation and sensitization
a visual target, the bird first misses the target by about in the sea slug Aplysia as examples of intermediate-term
10° to the left but gradually adjusts. The owl’s auditory memory formation. Long-term memories last weeks,
spatial map is calibrated to its visual spatial map, so months, or years.
once it adjusts to the goggles, sounds from a particular What are memories made of? A philosopher might
location are also excited by visual stimuli presented 10° have a poetic answer, but from our point of view, mem-
to the left of the target. ories are made from synaptic modifications—changes in
the strength or number of synapses. Different molecu-
the left. Those exact differences in timing and loudness lar mechanisms are responsible for the formation of
stimulate that particular neuron in the inferior collicu- memories of different durations. Nonetheless, this state-
lus. The degree of binaural difference varies with the ment holds true for all memory formation: Following an
location of the sound, and the binaural difference that experience, some synapses will get stronger, and others
stimulates cells of the inferior colliculus varies from will get weaker. The pattern of the changes in synapses
neuron to neuron (Knudsen 1982). The resulting audi- throughout the nervous system underlies memories
tory space map is then transmitted to the optic tectum (Beer 1999). Possible changes that leap to mind gener-
(an area of the brain involved in localizing and orienting ally include functional changes, such as an increase or
to visual information). decrease in the amount of neurotransmitter released at
Auditory maps can be formed without visual input, a synapse and changes in the number of receptors on the
but the precision of the map is increased by visual expe- postsynaptic membrane that alter the responsiveness of
rience. Indeed, when the information from the auditory the postsynaptic cell. Structural changes in neurons
map conflicts with visual input, owls trust their vision could alter the number of synapses that a particular neu-
over their hearing. Normally, the auditory space map in ron makes. The growth of new neurons may also under-
the inferior colliculus and the visual space map in the lie learning and memory (Bruel-Jungerman et al. 2007).
optic tectum are aligned. When the two maps are mis-
aligned by blocking one ear, the owl initially mislocates Intermediate-Term Learning in Aplysia
sound in the direction of the open ear. After many weeks
experience with an earplug, a young owl learns new Let’s consider the molecular events that occur at
associations between auditory and visual cues and ori- synapses when behavior can change with experience with
ents itself correctly. The auditory and visual space maps a single stimulus (nonassociative learning), as occurs
also become misaligned when an owl wears goggles that when the sea hare Aplysia learns. While Aplysia moves
shift the visual field by 10o. A young owl gradually across the ocean bottom eating seaweed, its siphon is
adjusts the location of sound localization to match its extended, and its gills, the respiratory organs, are spread
distorted visual map (reviewed in Knudsen 2002) out on the dorsal side. The gills are partly covered by a
(Figure 6.14). protective sheet called the mantle shelf, which termi-
nates in the siphon, a fleshy spout through which Aplysia
can squirt out excess seawater and wastes. When the
PROCESSING IN THE CENTRAL siphon is touched, the gills, the siphon, and the mantle
NERVOUS SYSTEM shelf withdraw into the mantle cavity. This defensive
response, called the gill-withdrawal reflex, protects the
The brain is a dynamic structure that changes with expe- gills from predators (Figure 6.15).
rience. Brain regions operate in networks. For example, The gill-withdrawal response can be modified by
there is a social behavior network consisting of six brain experience, that is, through learning. One form of
114 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

a Siphon Mantle shelf

Head
Sensory
neurons

Facilitating interneurons

Motor
neurons
Gill

Excitatory
Gill
interneurons

Siphon
Sensory
neurons

FIGURE 6.16 Neural circuitry for habituation and sensi-

FIGURE 6.15 The gill-withdrawal reflex in the sea hare, tization of the gill-withdrawal reflex in Aplysia.
Aplysia. The gills, mantle shelf, and siphon are drawn
here as if the animal were transparent. Normally, the
gills are spread out and are only partially protected by neural circuitry for the gill-withdrawal reflex is shown in
the mantle shelf. (a) The siphon, through which water Figure 6.16. Although there are actually 24 sensory neu-
is drawn in over the gills and excess water is expelled, is rons serving the siphon skin that terminate on 6 motor
extended so that just the tip is visible when the animal neurons to the muscle for gill retraction, this simplified
is seen from the side. (b) If the siphon is touched, the
diagram shows only one of each type of neuron. You can
gills, mantle shelf, and siphon are withdrawn into
the mantle cavity. The gill-withdrawal reflex can be
see that the sensory neuron from the siphon skin
modified by learning. synapses directly on a motor neuron for gill withdrawal.
Beginning in the late 1960s, neurobiologists began
to work out the details of the neural mechanism of habit-
learning, called habituation, occurs when an animal uation. One hypothesis was that the sensory neuron was
learns not to respond to a repeated stimulus that proves giving a weaker response to repeated stimuli. This idea
to be harmless (also discussed in Chapter 5). In other was shown to be incorrect by inserting a microelectrode
words, Aplysia learns to ignore an irrelevant stimulus. into a sensory neuron to measure its electrical responses
Habituation is adaptive because it saves energy. We can to stimulation. After repeated stimulation, the sensory
demonstrate habituation by disturbing an Aplysia’s neuron still responds normally. However, it fails to excite
siphon many times, by touch or by a brief jet of sea- the motor neuron as it initially did. Could habituation
water, for instance. After 15 such stimuli administered be due to a decrease in the motor neuron’s responsive-
ten minutes apart, the reflexive responses are only half ness? No. If a motor neuron is repeatedly stimulated
of their initial value. In sensitization, a second form of directly with an electrical current, it remains fully
learning in Aplysia, the withdrawal reflex becomes responsive. Another guess was muscle fatigue, but this,
stronger when a stimulus that elicits withdrawal is pre- too, was ruled out. Even after habituation, direct stim-
ceded by a strong, noxious stimulus, such as an electric ulation of the motor neuron causes full contraction of
shock, administered almost anywhere on Aplysia’s sur- the gill muscle (Kupfermann et al. 1970).
face. Depending on the number and strength of the nox- It turns out that habituation occurs because the sen-
ious stimuli, sensitization can last seconds to days (Carew sory neuron releases less neurotransmitter because of
et al. 1972; Kandel 1979; Pinsker et al. 1970). repeated stimulation (Figure 6.17). This, in turn, results
Through years of intensive research, Eric Kandel in fewer action potentials in the motor neuron for gill
and other investigators have determined the neural cir- withdrawal. Recall that many molecules of neurotrans-
cuits and many of the molecular mechanisms that under- mitter are released from a synaptic vesicle. This packet
lie these forms of learning in Aplysia (Kandel 1976; of many molecules of neurotransmitter is called a quan-
Kandel 1979a, b). Aplysia is an ideal organism for a neu- tum. The change in the EPSP caused by the release of
robiologist because it has just 20,000 neurons and the neurotransmitter from one synaptic vesicle is called a
neurons are large—10 to 50 times larger than neurons quantum of response. During habituation, the EPSP
in a mammalian brain. As a result, neurobiologists know decreases with repeated stimulation of the sensory neu-
many of these nerve cells by name. A diagram of the ron, and it does so in integral multiples of a quantum of
 Processing in the Central Nervous System 115

Habituation Sensitization

Repeated stimulation of the sensory neuron from siphon skin Strong, noxious stimulation to skin stimulates a sensory
neuron from skin

Less effective calcium channels in axon terminal Stimulates a facilitating interneuron that synapses on the
of sensory neuron axon terminal of the sensory neuron from siphon skin

Decreased Ca⫹⫹ inflow Release of serotonin from facilitating neuron

Decreased neurotransmitter released by sensory neuron Increased cAMP levels within axon terminal of sensory
neuron from siphon skin

Decreased firing rate of gill motor neuron cAMP causes closing of a certain type of K⫹ channel and
K⫹ increases in cell, prolonging depolarization of sensory
neuron

More Ca⫹⫹ than usual entering the axon terminal

Increased release of neurotransmitter by sensory neuron

Increased firing rate of gill motor neuron

FIGURE 6.17 Changes in synaptic functioning that accompany habituation and sensitization.

response (Castellucci and Kandel 1974). In other words, els cause more neurotransmitter to be released (Kandel
during habituation, successive action potentials cause and Schwartz 1982).
increasingly fewer synaptic vesicles to fuse with the
membrane and release their contents into the synaptic
Long-Term Memory Formation
cleft. The reason is that calcium channels become less
effective because of repeated stimulation, and so they We look in the fridge and notice we need milk. The next
allow less Ca++ into the axon terminal (Byrne 1987; afternoon, we stop at the store to pick up a carton. How
Hochner 1986). Calcium ions are needed for the synap- is this memory stored? We now know quite a bit about
tic vesicles to fuse with the presynaptic membrane. the molecular basis of memory. Memory involves
Sensitization also involves changes in the function- changes in synaptic connections, which involves four
ing of synapses, but in this case, the amount of neuro- processes: long-term potentiation, long-term depression,
transmitter released by the sensory neuron from the synaptic remodeling, and neurogenesis (Bruel-
siphon onto the gill motor neuron is increased, thereby Jungerman et al. 2007). These mechanisms alter the
increasing the motor neuron’s rate of firing. Sensitization strength of existing synapses, add new synapses, or
requires a facilitating interneuron. The process begins remove old synapses. We will consider each mechanism,
when strong stimulation to the body surface of Aplysia in turn.
stimulates a sensory neuron, which in turn stimulates a
facilitating interneuron. These interneurons release Long-Term Potentiation Repeated stimuli, as might
serotonin onto the axon terminal of the sensory neuron occur when an animal is learning a task, result in long-
from the siphon skin. Serotonin increases intracellular term potentiation, or LTP, which strengthens the
concentration of a second messenger, in this case cyclic connections between the adjacent neurons. LTP occurs
adenosine monophosphate (cAMP), which causes the at many synapses, possibly every excitatory synapse,
closing of a certain type of K+ channel of the cell mem- in the mammalian brain. It is the molecular mechanism
brane1.This keeps K+ inside the cell and keeps the sen- that underlies the acquisition and storage of memories
sory neuron depolarized longer than normal, allowing (Malenka and Bear 2004). In the laboratory, we can
additional Ca++ to enter the cell. The elevated Ca++ lev- simulate the events that occur at a synapse during learn-
ing by electrically stimulating a presynaptic neuron.
1
At some synapses, such as this one, the neurotransmitter opens or If the neuron is stimulated once a second, the response
closes ion channels through indirect biochemical mechanisms. These of the postsynaptic cell to each stimulus remains the
indirect mechanisms are similar to those used by neuromodulators. same. However, if the neuron is stimulated rapidly and
116 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

400 FIGURE 6.18 A rapid train of stimuli enhances

the responsiveness of the postsynaptic cell. This
phenomenon, called long-term potentiation, is
EPSP slope (% of control)

300 induced by learning and underlies memory

Late LTP (4 trains)
formation. (From Kandel 2001.)

200

Early LTP (1 train)

100

0
–30 0 30 60 90 120 150 180 210
Time (min)

repeatedly for a second or two, the response of the post- within seconds and last for days or weeks. Furthermore,
synaptic neuron is elevated and remains elevated (poten- long-term memory formation and the later stages of
tiated) for minutes or hours (Figure 6.18). LTP require protein synthesis. However, these features
Here is how LTP works (Figure 6.19). LTP is could be coincidental.
believed to involve two proteins embedded in the mem- Jonathan Whitlock and colleagues (2006) tested the
brane of nerve cells, called AMPA receptors2 and hypothesis that learning induces LTP, which underlies
NMDA3 receptors. These proteins open molecular gates long-term memory formation. They trained rats to asso-
that allow ions to pass in and out of the nerve cell, sim- ciate a painful stimulus with a location. Rats prefer dark
ilar to the sodium and potassium channels we discussed areas to light areas but will avoid a dark area where they
earlier. In the early stage of LTP, while an animal is have received a foot shock. Whitlock and colleagues
learning to associate a stimulus with a response, the reasoned that if LTP underlies learning, then the respon-
stimulus causes the presynaptic neuron to release a siveness of neurons in a region of the hippocampus (CA1)
neurotransmitter called glutamate into the synapse. known to be essential for spatial learning would be
Glutamate binds to both AMPA and NMDA receptors, increased by training. They measured the responsiveness
but at this point NMDA receptors are blocked by mag- of neurons in the area of the hippocampus before and after
nesium ions and cannot respond to glutamate. When training. We would not expect that all the synapses in this
glutamate binds to AMPA receptors, sodium channels brain region would be altered by training. Whitlock’s team
open. Sodium ions enter the nerve cell, and the postsy- used a biochemical marker to identify the synapses that
naptic cell becomes depolarized. If depolarization had been altered by learning. When they looked at these
reaches threshold, an action potential is generated in the synapses specifically, they found enhanced responsiveness,
postsynaptic cell. Because of the depolarization, the mag- which indicates LTP. The altered synapses occurred in
nesium ions blocking the NMDA channel are driven out. about 25% of the neurons in the CA1 area.
The unblocked NMDA receptors can now respond to
the glutamate, and calcium ions enter along with addi- Long-Term Depression When we learn something
tional sodium ions. The calcium ions activate several new, the pattern of synaptic excitability of many neurons
proteins within the postsynaptic neuron. Some of these changes. Some synapses become stronger, and others
activated proteins further increase the sensitivity of exist- become weaker. Long-term depression is a mechanism
ing AMPA receptors, and others cause additional AMPA that weakens the effectiveness of a synapse, decreasing
receptors to move to the postsynaptic membrane. Thus, the magnitude of a response by the postsynaptic cell. It
the cell becomes even more responsive to glutamate. As occurs after a slow train of stimuli (in contrast to the
we will see shortly, during the late phase of LTP, another rapid train of stimuli that causes LTP) has activated the
protein in the postsynaptic cell, CREB, is activated and presynaptic cell. Thus, long-term depression may also
turns on certain genes, resulting in long-lasting play a role in memory formation.
structural changes in the connections between neurons.
Does LTP underlie long-term memory formation? Synaptic Remodeling Long-term memory requires
They do have common features. Both can be induced structural changes in synapses. Several kinds of struc-
tural changes in the brain are reviewed by Julie
2
A glutamate receptor that also binds AMPA (␣-amino-3-hydroxy-5- Markham and William Greenough (2004). One change
methyl-4-isoxazolepropionic acid). is in dendritic spines—short extensions on dendrites
3
A glutamate receptor that also binds NMDA (N-methyl-D-aspartate). that form half of a synapse. Changes in the size, shape,
 Processing in the Central Nervous System 117

1. The neurotransmitter glutamate is released

from the presynaptic cell.

2. Glutamate binds to NMDA receptors

and AMPA receptors.

Mg++ blocks the NMDA receptor 3. AMPA receptors allow Na

+

and prevents it from working. to enter the postsynaptic cell

and begins depolarization.

Repeated stimulation of AMPA receptors

causes the postsynaptic membrane
to depolarize more.

4. Depolarization of the postsynaptic membrane,

drives Mg++ out of the NMDA receptor,
removing blockage and allowing Ca++ ions to enter
the postsynaptic cell.

++
5. Ca activates protein kinases, which in turn
activate other proteins, such as CaMK
(calcium-calmodulin kinase, PKA and PKC).

6. Activated CaMK. Activated CaMK, PKC, and PKA.

Makes AMPD receptors more Promotes movement of additional Activates the protein CREB,
responsive to glutamate, allowing AMPD receptors which turns on many
+ +
Na and K to move across from inside the cell different genes.
the postsynaptic membrane more easily. to the postsynaptic membrane.

Changing levels of various proteins

cause a signal from the postsynaptic cell
to the presynaptic cell that increases
the release of glutamate.

FIGURE 6.19 The molecular events of long-term potentiation. These changes lead to a
long-lasting increase in the responsiveness of the synapse, making communication
between the neurons easier.

and number of dendritic spines enhance communication CREB is to turn on sets of genes that synthesize
between neurons, thereby reinforcing a particular additional proteins. These proteins then bring about
neural pathway. Other structural changes involve glial long-lasting structural changes in the synapses.
cells, which produce growth factors that guide axons to
form synapses in the correct destinations and influence Neurogenesis At one time, we believed that the
communication at synapses by removing neurotrans- growth of new neurons was limited to development and
mitters from the synapse. that no new neurons formed in adulthood. This belief
Structural changes such as these require protein syn- implied that adult learning and memory were due only
thesis. Recall that during long-term potentiation, cal- to changes in synapses, as described above. However,
cium ions enter the postsynaptic cell and activate certain recent evidence shows that the birth of new neurons, or
proteins. CREB is among these. One of the functions of neurogenesis, can occur even in older animals.
118 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

Song Learning in Birds Fernando Nottebohm was Control

BrdU+/Hu+ cells per mm2 HVC

80 Contra
one of the first to suggest that brain changes that occur
70 Ipsi
with behavioral changes might include the birth of new
60 *
neurons, along with the death of old neurons.
50
Nottebohm studied the neurobiology of bird songs
40
(2005). As we will see in Chapter 8, discrete but inter-
30
connected centers in the songbird’s brain, collectively
20
called the song system, make the learning and produc-
10 Song learning *
tion of song possible. Nottebohm (1981) discovered that
0 Stable song
the volume of two song centers, the HVC and the RA 21–25 35–39 61–65 120–124
(robust nucleus of the arcopallium), increases and Days (post-hatch) of injections
decreases seasonally, along with the amount of song
learning. What causes these song centers to change in **Significant difference from the ipsilateral hemisphere and controls.
size? Researchers can test whether new neurons are FIGURE 6.20 The addition of new neurons to one of
added to an area by adding marker chemicals. New neu- the song control centers (HVC) of male zebra finches
rons incorporate these chemicals when they make DNA changes during the course of song learning. It is high as
in preparation for cell division; old neurons do not syn- song learning begins, and it slowly declines as the song
thesize DNA and therefore do not incorporate these is learned. It is lowest after a stable song has developed.
markers. As it turns out, the increase in the size of the (From Wilbrecht et al. 2002.)
RA is caused by an increase in both the size of neurons
and the space between them, but in the HVC it is due
to the formation of new neurons (Sherry 2005; Goldman learn their songs. They predicted that if song learning
and Nottebohm 1983). The neurons are born in the causes the addition of new neurons to the HVC, then
lateral ventricles of the forebrain and differentiate into limiting the opportunity for song learning will reduce
specific types of neurons as they migrate to the HVC. the number of neurons added.
Most of the new neurons die within a few weeks. The Let’s consider how the opportunity for song learn-
survivors are those that become connected to existing ing has been manipulated in male zebra finches. One can
neural networks, and they replace cells that died. The deafen the males toward the beginning of song learning
number of neurons dying in the HVC peaks in late sum- (26 days) and can prevent them from singing by cutting
mer, and the rate at which new neurons are added to the the nerves to the syrinx (vocal organ of birds). There
HVC peaks in the autumn. Through the fall and win- were no significant differences in the number of new
ter, the rate of neuron birth and death must be equal neurons in the HVC between hearing and deaf birds or
because the total number of neurons in the HVC does birds in which the nerves to both sides of the syrinx had
not increase until spring. Thus, neurogenesis regularly been cut. Compared with controls, there were twice as
replaces neurons in a bird’s brain (Sherry 2005; Strand many new neurons in the HVC of birds in which the
and Deviche 2007). nerve on one side of the syrinx had been cut (Figure 6.20).
How does neurogenesis relate to song learning? The total number of neurons in the HVC of all adult
Most of the time, song learning by young male zebra birds was not affected by any procedures. Thus, song
finches (Taeniopygia guttata) and the addition of new neu- learning doesn’t alter the total number of cells in the
rons to the HVC are correlated. When reared with adult adult HVC, but it may affect the addition of new neu-
males, young zebra finches learn their song by imitating rons during a limited amount of time and special cir-
the songs of their adult companions from age 30 through cumstances, such as having the nerve to the syrinx cut
65 days, eventually perfecting their stereotyped song on only one side (Wilbrecht et al. 2002).
between days 65 and 90. As you can see in the control A second approach to manipulating the opportunity
group of Figure 6.20, neurogenesis is highest as song for song learning is to extend the sensitive period (a
learning begins, declines as the song is learned, and is developmental stage when key experiences alter behav-
lowest after 60 days, around the time that the song ior more than at other ages) during which a young male
stabilizes (Wilbrecht et al. 2002). imitates and perfects his song to match the songs of adult
We see, then, that a correlation exists between the males. The sensitive period is extended when young
timing of the additional new neurons to the HVC and males are raised alone without an opportunity to see or
song learning. But which event causes the other? Do the hear adult males. Males raised in isolation improvise a
new neurons allow song learning to take place, or does song that, even after day 80, they can modify to match
song learning cause the addition of new neurons to the the song of an adult male if one becomes available. If
HVC? Researchers have explored the question of which learning causes the addition of neurons to the HVC,
comes first by manipulating the opportunity for birds to then we predict that extending the sensitive period
 Processing in the Central Nervous System 119

FIGURE 6.21 A section through the dentate

gyrus of the hippocampus showing new neu-
rons. The new neurons are darker in color
because they are labeled with BrdU, which is
taken into cells that are synthesizing DNA in
preparation for cell division.

would increase the number of new neurons in the HVC. a region of the hippocampus called the dentate gyrus and
This is indeed the case. Isolated males added about 1.6 the olfactory bulb (Figure 6.21).
times more neurons than control males added during What is the role of neurogenesis in the hippocam-
posthatching age 65 to 150 days (Wilbrecht et al. 2006). pus? Because the hippocampus is especially important in
The alternative hypothesis—that new neurons are spatial learning, we might expect spatial learning tasks
needed for song learning—appears to be correct. The to be more affected by blocking neurogenesis than learn-
current hypothesis is that seasonal changes in photo- ing tasks that are independent of the hippocampus.
period cause changes in blood testosterone levels in male Tracey Shors and colleagues (2001) investigated the
songbirds. As the days get longer during the spring, effect of a reduction in neuronal cell division on mem-
testosterone levels increase. Testosterone then causes an ory formation in rats. They treated rats with saline or
increase in singing, which increases the production of with a chemical that suppresses cell division (methylaz-
brain-derived neurotrophic factor, a protein that oxymethanol acetate; MAM) and found that the
enhances the survival of new neurons in the HVC. In production of new neurons is important in memory
this way, older, underused neurons will be replaced by formation that requires the hippocampus (Figure 6.22).
new, heavily used neurons (Nottebohm 2002). This David Dupret and his colleagues (2007) concluded
hypothesis is supported by data from male house finches that cell death is as important as the birth of new neu-
(Carpodacus mexianus) (Strand and Deviche 2007) and rons during spatial learning. A common test of spatial
male Gambel’s white-crowned sparrows (Zonotrichia leu- learning in mice and rats is the water maze. In this test,
cophrys) (Brenowitz et al. 2007). the animal is placed in a pool of opaque liquid. It must
swim until it finds a platform on which it can rest. The
Spatial Learning in Mammals As in birds, we also platform is hidden beneath the liquid’s surface, and so it
see the appearance of new neurons and the death of old is not visible to the animal. The animals must learn the
ones. However, in mammals neurogenesis occurs only in location of the platform using visual cues placed around
Conditioned responses (%)

a 10,000 b 100
BrdU-labeled cells

Saline
8,000 80 MAM

6,000 60

4,000 40

2,000 20
0 0
Saline MAM SP 20 100 200 300 400 500 600 700 800
Trials

FIGURE 6.22 Adult neurogenesis is involved in conditioning that depends on the hippocampus.
Before training, the animals were treated with saline or with a chemical that suppresses cell
division (MAM). New neurons were labeled by administering a chemical that is incorporated
into DNA during DNA synthesis. (a) The animals treated with MAM produce fewer new
neurons and did not learn the task. (b) The control animals (treated with saline) continued to
produce new neurons, and their performance on the task improved. (From Shors et al. 2001.)
120 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

the water tank. The hidden platform water maze is a hand-reared marsh tits (Parus palustris) that were allowed
form of spatial learning, and it depends on the to store and retrieve sunflower seeds had more new neu-
hippocampus. rons in the ventricular zone border of the hippocampus
When cell death was inhibited by treating the ani- than did age-matched control birds that were not per-
mals before training with a chemical, spatial learning was mitted to store food. The birth of new neurons in the
impaired. The rats could learn the position of the hid- hippocampus of black-capped chickadees (Parus atri-
den platform within a training session but could not capillus) increases beginning in February and March and
remember it the next day. peaks in October (Barnea and Nottebohm 1994). It is
Dupret suggests that spatial learning causes a cas- also at this time of year that food storing is thought to
cade of events similar to the formation of functional peak, but see Pravosudov (2006) who suggests that there
neural networks during development. Thousands of new may be an additional peak of food storing in the spring
neurons are formed in the hippocampus each day. that is not accompanied by an increase in new neurons
During the early phase of learning in the hidden-plat- in the hippocampus.
form water maze, when the animal’s performance is Does spatial learning increase neurogenesis and alter
improving quickly, the survival of neurons that are about the anatomy of the hippocampus in humans as well?
one week old is increased (Figure 6.23; Dupret et al. Licensed London taxi drivers provided a pool of subjects
2007). After the animal has begun to master the task, on which to explore this possibility because they must
additional new neurons—those that have not yet made complete two years of training, colloquially known as
connections with other neurons—die. Cell death is nec- “being on The Knowledge,” to learn to navigate thou-
essary for the survival of the earlier born neurons and for sands of streets in London. Eleanor Maguire and her
an increase in neurogenesis, which increases the pool of colleagues (2000) hypothesized that the anatomy of the
new neurons. hippocampus would be altered by years of navigational
Changes in the number of new neurons in the training. They compared structural MRI brain scans of
hippocampus are also associated with spatial learning 16 healthy, right-handed licensed taxi drivers with scans
during food storing and retrieval in birds. For example, of 50 right-handed males of similar age but without
experience driving taxis. The only structural differences
in the brains of men in these two groups were in the
Early phase regions of the hippocampus. Compared with control
males, the posterior hippocampus of taxi drivers was
Performances

larger. In contrast, the anterior region of the hippocam-

pus was smaller in taxi drivers than in control males.
Furthermore, the volume of the posterior hippocampus
Late phase
increased, and the volume of the anterior hippocampus
decreased with the amount of time spent as a taxi cab
Days driver (Figure 6.24). These results would be predicted if
the posterior hippocampus stores a “mental map” of the
Old
newborn city and its volume can expand with the amount of
neurons information encoded in this mental map.
Survival

SOCIAL BEHAVIOR NETWORK

Death
Now that we have considered changes in the brain that
Young underlie learning, we will explore interactions among
newborn
neurons different regions of the brain involved in social behav-
Proliferation ior. Sarah Newman (1999) originally suggested that
mammalian social behavior is controlled by six regions
FIGURE 6.23 Spatial learning depends on the production (nodes) of the brain: (1) the medial extended amygdala,
of new neurons, the selective survival of new neurons, including the medial bed nucleus and the nucleus of the
and selective cell death. During the early phase of learn-
stria terminalis (BSTm), (2) the lateral septum (LS), (3)
ing in the hidden-platform water maze, new neurons
begin to make synaptic connections with existing neu-
the preoptic area (POA), (4) the anterior hypothalamus
rons. Learning increases the survival of these neurons. (AH), (5) the ventromedial hypothalamus (VMH), and
In the late phase of learning, slightly younger neurons (6) the midbrain. These regions of the brain are thought
die by apoptosis. This cell death enhances the survival to make up a social behavior network for several reasons.
of the new neurons that have made synaptic connections Each of these regions plays a role in multiple social
and the rate of neurogenesis. (From Dupret et al. 2007.) behaviors, including aggression, sexual behavior, social
 Processing in the Central Nervous System 121

a 6 FIGURE 6.24 The volume of both the

posterior and the anterior hippocampus
posterior hippocampus
Measure of volume of 4
correlates with the number of years
2 experience as a taxi driver in London.
Taxi drivers require years of spatial train-
0 ing to navigate the streets of London.
(a) The volume of the posterior hip-
–2
pocampus increases with the number of
–4 months experience driving a taxi in
London. (b) The volume of the anterior
–6 hippocampus decreases with the number
0 50 100 150 200 250 300 350 400
of months experience driving a taxi in
Time as a taxi driver (months)
London. (From Maguire et al. 2000.)
130
b
anterior hippocampus
Measure of volume of

120
110
100
90
80
70
60
0 50 100 150 200 250 300 350 400
Time as a taxi driver (months)

recognition, affiliation, bonding, parental behavior, and biparental care, and live in similar habitats. The finches
response to stress. We have looked at examples of included the zebra finch (Taeniopygia guttata) and the
involvement of these areas in social behavior in Chapter spice finch (Lonchura punctulata), which live in colonies
3, and we will consider other examples in Chapter 7. of from 90 to 300. Two species of waxbills were also
Interactions between these regions are possible because
they are bidirectionally connected. Furthermore, each of Male aggression
these regions contains sex steroid receptors, which are
needed for sexual differentiation and coordination of
social behavior. Each brain region is responsive to a
variety of stimuli. The observed social behavior is not the
result of independent responses of brain regions but
rather the pattern of response across all regions
(Figure 6.25) (Goodson 2005).
Male sex behavior
James Goodson proposes that this social behavior
network can be found in the brains of all vertebrates; nat-
ural selection produces diversity in social behavior by
acting on the responses of each region to steroid hor-
mones and neuropeptides. Goodson’s work on the mid-
shipman fish (Porichthys notatus) demonstrates that this
social behavior network was present in the earliest ver-
tebrates. In addition, many of the functional, structural,
and neuroendocrine responses of the social behavior net- Ext.MeA LS
work of fish and birds are similar to those of mammals
Mid
(Goodson 2005; Goodson et al. 2005). POA
Goodson and his colleagues hypothesized that the VMH
pattern of activity across the brain regions forming the AH
social behavior network would vary in species of birds
that differ in the typical size of their social group. They FIGURE 6.25 Six regions of the brain compose a social
identified four species of birds that differ in their degree behavior network that is active in many types of social
of sociality. The subjects included four species of estril- behavior. The pattern of activity of these regions varies
did finches, two of which are monogamous, show with the type of behavior. (From Newman 1999.)
122 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

subjects. The waxbills included the moderately gregari- a Gregarious/colonial

ous Angolan blue waxbill (Uraeginthus angolensis), which Modestly gregarious
lives in groups of from 8 to 40 individuals, and the ter-
ritorial violet-eared waxbill (Uraeginthus granatina),
which lives in small groups containing both parents and
their dependent young.
Goodson’s team (2005) examined the activity of the
immediate early genes, c-fos and zenk, in the brain
regions connected in the social behavior network to b Territorial LS
measure the brain’s response to a social stimulus—expo-
Ventrolateral
sure to a same-sex conspecific. (Immediate early genes Other
signal changes in neural activity.) Indeed, the activity of
two brain regions does differ among species of birds that
Lateral
differ in their degree of sociality. The medial extended
amygdala functions in social arousal and avoidance. The
activity of this region of the brain (as measured by the
activity of immediate-early genes) increased from the VMH-medial
colonial, gregarious species of finches to the moderately Med. Ext. LS
gregarious Angolan blue waxbill to the territorial violet- Amyg.
Mid
eared waxbill. There were also changes in activity in POA
brain regions involved in social stress and dominance- VMH
related behaviors. We might predict that exposure to a AH
same-sex conspecific bird would be more stressful to a
FIGURE 6.26 The pattern of activity across the brain
territorial bird than to a colonial bird. This prediction regions is different in a gregarious species and a territo-
held true when activities in the brain regions involved in rial species of waxbills. At the base of the figure are
stress and dominance were compared. The immediate- each of the six brain regions making up the social
early gene responses were clearly higher in territorial behavior network. (a) The pattern of activity across the
species than in colonial species. As you can see in brain regions in a gregarious species, the Angolan blue
Figure 6.26, the pattern of activity across the brain waxbill (group size 8–40). (b) The pattern of activity
regions is different in a gregarious species and in a ter- across the same brain regions in a territorial species,
ritorial species of waxbills. the violet-eared waxbill. (Goodson et al. 2005.)

is ultimately controlled by the activity of a neural circuit.

There are three major ways in which neural circuits con-
RESPONDING—MOTOR SYSTEMS trol and coordinate movement: (1) the sensory reflex, (2)
Animals receive information about their environment via the central pattern generator, and (3) motor command.
their sensory systems and then respond by means of their In sensory reflexes, sensory neurons initiate activity
motor systems. Two principal components of motor sys- in motor neurons, sometimes through direct synaptic
tems are motor organs (typically muscles) and the neural connections but more typically through connections
circuits that control them. Like sensory receptors, mus- with a small number of interneurons. We have already
cles are often described as biological transducers. There considered the gill-withdrawal reflex in Aplysia, so here
are, however, some important distinctions between the we will consider the ways in which a central pattern gen-
two. Whereas sensory receptors transduce environmen- erator and motor command control the locust’s flight.
tal energy such as light into the electrochemical signals
of the nervous system, muscles convert the signals of the
LOCUST FLIGHT
nervous system into the movements of the body.
Furthermore, whereas sensory receptors are concerned Locusts, those species of “short-horned” grasshoppers
with input, muscles are involved in output. found in the family Acrididae, exhibit legendary mass
migrations. In fact, accounts of locust plagues date back
NEURAL CONTROL to the Book of Exodus, written in about 1500 B.C.
(Williams 1965). Swarms of locusts have also been
IN MOTOR SYSTEMS
recorded in recent times. Although representatives of the
A particular movement, or behavior, is produced by mus- family are found throughout the world, the migratory
cles, whose activity is controlled by motor neurons. species are found in the tropics or subtropics, typically
Recall that motor neurons, in turn, usually receive their in the drier regions of these areas. As evidence of the
information from interneurons. Thus, each movement amazing flight behavior of locusts, consider this account
 Responding—Motor Systems 123

two pairs of wings do not move in a precisely synchro-

nous manner, but rather the hind wings lead the
forewings. The entire cycle of movement lasts about 50
milliseconds.
Two sets of muscles act on each wing. One set, the
elevators, raises the wing; the other set, the depressors,
lowers it. Because these two sets of muscles have oppo-
site effects on the wing, they are called antagonists (mus-
cles with the same effects on a specific structure are
called synergists).
Patterns of activity in the flight muscles of a teth-
ered locust can be recorded by inserting tiny wire elec-
trodes into the muscles (see Figure 6.28a). The signals
(a recording of muscle action potentials) from the ele-
vator and depressor muscles of the flying insect are then
FIGURE 6.27 A desert locust. displayed on an oscilloscope; this type of recording is
called a myogram (Figure 6.28b). Myograms have shown
that the depressors are activated when the wings are up,
written by the entomologist C. B. Williams (1965, and the elevators are activated when the wings are down
pp. 77–78) while working in East Africa: (Figure 6.28c). The relative timing of the activation of
these muscles is critical, and we will now examine how
A most spectacular flight occurred on 29th January, such timing may be controlled by the central nervous
1929, at Amani, in the Usambra Hills in north-east- system.
ern Tanganyika. . . . We received a telephone warning
shortly after breakfast that an immense swarm of
locusts was passing in our direction over an estate about The Neural Basis of Locust Flight
six miles to the north. . . . An hour or so later the first There are two general hypotheses for the neural basis of
outfliers began to appear—gigantic grasshoppers about rhythmic behavior such as locust flight. The peripheral-
six inches across the wings, and of a deep purple-brown. control hypothesis is that each movement stimulates
Minute by minute the numbers increased, like a brown sensory receptors, which in turn trigger the next move-
mist over the tops of the trees. When they settled they ment in the sequence. The second movement stimulates
changed the colour of the forest; by the weight of their other sensory receptors that trigger the first component.
numbers they broke branches of trees up to three inches Thus, sensory feedback is necessary for this hypothesis.
in diameter; the noise of their slipping up and down on An alternative hypothesis is that a central pattern gen-
the corrugated iron roofs of the houses made conversa- erator controls locust flight. A central pattern gener-
tion difficult. . . . The swarm was over a mile wide, ator is a neuron or network of neurons that is capable
over a hundred feet deep, and passed for nine hours at of generating patterned activity in motor neurons, even
a speed of about six miles per hour. when all sensory input has been removed from the sys-
The species of locusts described in Williams’s tem (Carew 2000).
account was the desert locust, Schistocerca gregaria Donald Wilson’s paper in 1961 was pivotal in our
(Figure 6.27). Although this species has been used in understanding of rhythmic behavior (reviewed in
some laboratory studies of flight behavior, the migratory Edwards 2006). Wilson reasoned that if the peripheral-
locust (Locusta migratoria) is more commonly studied. control hypothesis were correct, then elimination of sen-
Let us now look at the neural control of flight behavior sory feedback would eliminate the pattern of activity in
in these remarkable insects. the motor neurons that raise and lower the wings. In
contrast, the central pattern generator hypothesis pre-
dicts that the elimination of sensory feedback would not
The Motor Pattern of Locust Flight prevent the pattern of activity in the motor neurons con-
Locusts have two pairs of wings: the forewings, located trolling the wing movements. Locusts have mechano-
on the second thoracic segment, and the hind wings, receptors on their wings that send information on wing
located on the third thoracic segment. The wings of free- position, via sensory neurons, to the central nervous sys-
flying locusts move up and down, in a rhythmic manner, tem. This control system can be eliminated by section-
about 20 times per second. Because locusts maintained ing the sensory nerve or by injecting the chemical
in the laboratory and tethered to a holder exhibit close phentolamine, which blocks the activation of
to normal flight, it is possible to obtain a detailed analy- mechanoreceptors without affecting the central nervous
sis of the motor pattern (Figure 6.28). During flight, the system (Ramirez and Pearson 1990). The removal of
124 Chapter 6 / Physiological Analysis—Nerve Cells and Behavior

a Electrodes b 150

Up Hind wing

120

Angle of wing (degrees)

Forewing

90
Wind

60

Down

30

10 20 30 40 50
Time (ms)
c
= Muscle D 112
activated Hind wing D 128
E = Elevator muscles
= Depressor D 129
D
E 113, 118, 119

D 81
Forewing D 98
muscles
D 99
E 83, 89, 90

FIGURE 6.28 Analysis of the flight pattern of locusts. (a) If wind is directed at the head, a tethered
locust can be induced to fly. The activity of flight muscles of the second and third thoracic segments
can then be measured by inserting small wire electrodes in them. (b) The pattern of wing movements
during flight in a tethered locust. (c) The pattern of activity in individual flight muscles as recorded
with microelectrodes. (a: Modified from Horsmannet al. 1983. b and c: Modified from Young 1989.
Original data from Wilson 1961 and Wilson and Weis-Fogh 1962.)

sensory input by either technique is called deafferenta- the elevator and depressor muscles. We see, then, that a
tion. (Recall that sensory neurons are also called affer- central pattern generator controls locust flight.
ent neurons.) Although some changes in the flight Although the basic rhythm of flight is generated by
pattern are observed after the surgery (e.g., the fre- a central pattern generator in the thoracic ganglia, infor-
quency of wing beats drops from 20 to about 10 per sec- mation gathered by sense organs located on the wings
ond), the basic pattern of wing movements is still enables the locust to adjust its flight pattern on a cycle-
present. The removal of sensory feedback from the by-cycle basis. Sensory information thus imparts some
wings can be taken one step further than simply sec- flexibility to the system of motor control, allowing the
tioning the sensory nerves. The wings and thoracic mus- animal to respond to air turbulence and other environ-
cles can be completely removed. Even after such mental uncertainties. The system of motor command
extensive dissection, a pattern of neural activity similar plays a particularly important role in that the interneu-
to that seen during flight occurs in the motor nerves that rons that descend from the brain initiate (and modulate)
emerge from the thoracic ganglia. Taken together, these activity in the central pattern generator. These interneu-
results suggest that the basic flight pattern of the locust rons appear to play a role in correctional steering, when
is not dependent on sensory feedback but is generated the locust compensates for deviations in the flight course.
in the central nervous system. In other words, a neuronal The combination of central and peripheral control
circuit in the central nervous system generates the alter- observed in locust flight probably characterizes most pat-
nating bursts of activity in motor neurons that innervate terns of rhythmic behavior in animals (Delcomyn 1980).
 Summary 125

work more slowly than neurotransmitters and bring

SUMMARY about changes by biochemical means. Neuromodulators
can have profound effects on behavior, as demonstrated
The basic unit of the nervous system is the neuron, a cell by the effects of the neuromodulators on rhythmic leg
that usually has three parts: a cell body, or soma, which movements in the blue crab.
maintains the cell; dendrites, which receive information Animals receive information from the environment
and conduct it toward the soma; and an axon, which con- at their sense organs. Sense organs, such as eyes and ears,
ducts the nerve impulse away from the soma. There are are selective, containing specific receptor cells that
three functional classifications of neurons. A sensory neu- respond to a particular form of environmental energy.
ron is specialized to detect stimuli or to receive infor- Noctuid moths have only a single receptor cell (A1 cell)
mation from sensory receptors and conduct it to the that is involved in detecting a hunting bat and evoking
central nervous system. Interneurons, located in the cen- an escape response. Barn owls have an auditory space
tral nervous system, link one neuron to another. A motor map that allows them to accurately locate sounds made
neuron carries the information from the central nervous by their prey in complete darkness.
system to a muscle or gland. When the siphon of Aplysia is touched, the animal
When a neuron is resting, that is, not conducting an withdraws its gills, mantle shelf, and siphon into the
impulse, it is more negative inside than outside. Whereas mantle cavity. This is called the gill-withdrawal reflex.
potassium ions (K+) are more concentrated inside the This reflex shows habituation, that is, a decrease in
neuron’s membrane than outside, sodium ions (Na+) are responsiveness because of repeated stimulation. It also
more concentrated outside the membrane. shows sensitization, in which a strong stimulus anywhere
The message of a neuron is called an action poten- on Aplysia’s surface will cause an exaggerated gill-with-
tial or nerve impulse. The action potential is a wave of drawal reflex from a light touch on the siphon.
depolarization followed rapidly by repolarization that Short-term memory involves changes in the
travels to the end of the axon with no loss in strength. strength of synaptic connections. In Aplysia, habituation
Should a stimulus open ion channels that will allow occurs because the sensory neuron releases less neuro-
sodium ions to cross the membrane, sodium ions will transmitter when it is repeatedly stimulated. As a result,
enter the neuron and depolarize the membrane. the gill motor neuron is less likely to fire. Sensitization
Potassium ions will then move out of the neuron, repo- occurs when a strong stimulus causes a sensory neuron
larizing the membrane. to activate a facilitating interneuron to release serotonin
Information travels from one neuron to the next by on its synaptic ending near the axon terminal of the sen-
crossing the gap, or synapse, between them. The two sory neuron from the siphon skin. Serotonin then causes
types of synapses are electrical and chemical. Electrical biochemical changes that lead the sensory neuron to
synapses involve such tight connections between the release more neurotransmitter than usual. This increases
neurons that the impulse spreads directly from cell to the likelihood that the gill motor neuron will fire.
cell. Several steps are involved in transferring informa- Long-term memory involves long-lasting changes in
tion across a chemical synapse: (1) The action potential synapses. These changes occur through four mecha-
travels down the axon to small swellings called terminal nisms: long-term potentiation, long-term depression,
boutons. (2) Here, the action potential causes a neuro- synaptic remodeling, and neurogenesis. Neurogenesis in
transmitter to be released from storage sacs (synaptic a song center (HVC) in the brain underlies song learn-
vesicles) into the gap between cells (synaptic cleft). (3) ing in birds. Neurogenesis in the hippocampus under-
The neurotransmitter diffuses across the gap and binds lies spatial learning.
to a special receptor. (4) When this occurs, ion channels Despite the diverse types of stimuli that bombard an
open, changing the charge difference across the mem- animal from its environment, it is able to detect only a
brane. (5) If the synapse is excitatory, the postsynaptic limited range, and of those, it may ignore all but a few
cell is slightly depolarized. However, if the synapse is key stimuli.
inhibitory, the postsynaptic cell is slightly hyperpolar- Animals use their motor systems to respond to infor-
ized, and no new action potential will be generated. mation picked up by their sensory systems. Central pat-
The membrane potential of the postsynaptic cell can tern generators can activate motor neurons in the
also be changed by neuromodulators. These substances absence of all sensory feedback.
This page intentionally left blank
 7
Physiological Analysis of Behavior—
The Endocrine System
The Endocrine System Interactions Between Hormones, Behavior,
Endocrine Glands and Hormones and the Environment
Hormonal versus Neural Communication Adjusting to the Harshness and Predictability of the
Physical Environment
Types of Hormones and Their Modes of Action
Adjusting to Onlookers in the Social Environment
How Hormones Influence Behavior
A Detailed Look at the Hormonal Basis
Effects on Sensation and Perception
of Selected Behaviors
Effects on Development and Activity of the Central
Nervous System Helping at the Nest
Effects on Muscles Scent-Marking
Migrating
Methods of Studying Hormone–Behavior
Relationships
Interventional Studies
Correlational Studies
After nightfall, from late spring through summer, along
Organizational and Activational Effects
the western coast of North America, one can hear the
of Hormones
male plainfin midshipman fish (Porichthys notatus)
Defining the Dichotomy
“singing” to attract females to his nest. The song,
Sex Differences in the Behavior of Norway Rats which is more of a droning hum than a melodic love-
Individual Differences in the Behavior of Male Tree song, is produced when the male contracts a pair of
Lizards
vocal muscles against the swim bladder (a buoyancy-
Questioning the Dichotomy regulating organ). The nests are built under rocky shel-
The Dynamic Relationship Between Hormones ters in the intertidal zone. Although a female mates
and Behavior with only one male each breeding season, a male may
A Reciprocal Relationship mate with five or six females. He then guards 1000 to
Hormonal Suppression of Behavior 1200 eggs.

127
128 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

There are actually two types of male plainfin mid- song. The size of the motor neurons to the vocal mus-
shipman fish, and they display different reproductive cles and the brain center controlling those muscles are
strategies. The males we’ve described, those that build also larger in a type I male. Surprisingly, however, the
nests and sing to court females, are type I males. In con- ratio of testis to body weight of a type II male is nine
trast, type II males don’t build nests or hum to attract times that of a type I. Because of his bulging gonads, a
females. Instead, they either sneak into the type I’s nest type II male resembles a gravid female (Figure 7.1). The
and spawn or they lie outside the entrance and deposit coloration of a type II male is also more similar to that
sperm there while fanning water toward the nest’s open- of a female than a type I male. These similarities to
ing. The sperm of the type II male will then compete females probably make it easier for type II males to lurk
with that of the type I male, who did all the work (Bass around the nests of type I males.
et al. 1997). Type II males are sneaky cuckolders (males Sneaky cuckoldry by type II males seems to be a
that steal matings from other males) whose success “fixed” behavioral strategy because these males will not
depends on avoiding detection by territorial males. hold territories or court females even when both options
There are other differences between type I and type are offered to them (Brantley and Bass 1994). The mor-
II males. Type I males are larger and take longer to reach phology of type II males—small bodies, vocal muscles,
sexual maturity. Although type II males become sexually and underlying neural circuitry—seems to prevent
mature sooner, they can’t attract mates. The vocal mus- behavioral plasticity, the ability to switch to an alterna-
cles of type I males used to produce the droning hum tive behavior. In contrast, field and laboratory observa-
that is so attractive to females are much more developed tions of type I males show that these males will cuckold
than those of a type II male. The brain of a type I male other type I males when given the chance (Lee and Bass
is also specialized to allow him to hum his courtship 2004). The morphology of type I males apparently does

Vocal muscle Swim bladder Fin

Vocal muscle

Testis

Swim bladder Ovary

Male type I Male type II Female (gravid)

FIGURE 7.1 In plainfin midshipman fish, there are two types of males, each with a distinct mating strategy. Type I
males build nests, sing to attract females, and guard nests. To sing, a type I male contracts his vocal muscles against
the swim bladder to create the droning hum of his courtship song (top). A type II male waits for a type I male to
attract a female and then sneaks into the nest and spawns or releases sperm just outside the nest and fans water into
the nest opening. The vocal muscles of a type I male (bottom left) are much more developed than those of a type II
male (bottom center) or a female (bottom right). The ratio of vocal muscle to body weight is six times greater in a
type I male than in a type II. Type I males are larger, but the ratio of gonads to body weight is nine times smaller than
that of a type II male. The large testes make a type II male look like a gravid female. (Modified from Bass 1996.)
 The Endocrine System 129

not preclude behavioral plasticity. Indeed, type I cuck- 20HE), also functions as a feeding deterrent between
olders seem to take advantage of their large size and individual shore crabs (Carcinus maenas) (Hayden et al.
aggressively resist ejection by resident type I males. 2007). Mating in shore crabs is a delicate affair that coin-
Thus, whereas type II males engage in sneaky cuckoldry, cides with molting in females. Prior to a female molting,
type I males exhibit aggressive cuckoldry when not hum- a male will guard her from rival males and predators,
ming and defending a nest. cradling her beneath his abdomen for a few days. Once
Hormonal differences are responsible for the phys- a female molts, copulation soon follows. A soft-shelled
ical and behavioral differences among type I males, type female is extremely vulnerable to predation by fish, octo-
II males, and females. What are hormones, and how do pus, and other crustaceans, including members of her
they produce such dramatic effects on the nervous sys- own species. It turns out that 20HE released by the
tem, muscles, and behavior? What role do hormones female at this time effectively deters her mate from eat-
play in the development and display of behavioral dif- ing her. The feeding deterrent effect of 20HE is sex-
ferences between the sexes and between individuals of specific, deterring C. maenas males but not females.
the same sex? Do hormonal effects on behavior vary as Here, then, 20HE is acting not as a hormone (as it does
a function of the physical or social environment? Can when regulating molting within the individual) but as a
behavior, in turn, influence the levels of hormones? It is pheromone or chemical substance that functions in com-
to these issues that we now turn. munication between individuals of the same species (see
Chapter 16 for a further discussion of pheromones). It
is not yet known whether the 20HE released by newly
THE ENDOCRINE SYSTEM molted C. maenas females deters feeding in other species
of predatory crustaceans. It is known, however, that
ENDOCRINE GLANDS AND HORMONES
20HE released by Pycnogonum litorale, another marine
We begin with a definition. Hormones are chemical sub- arthropod, deters feeding by C. maenas, one of its preda-
stances secreted in one part of the body that cause tors (Hoffmann et al. 2006; Tomaschko 1994). Thus, the
changes in other parts of the body. Hormones are chemical signal 20HE functions within individual
secreted either by endocrine glands or by neurons. arthropods (as molting hormone), between individuals of
Endocrine glands, unlike exocrine glands (e.g., sweat, the same species (as a sex-specific feeding deterrent dur-
salivary, or scent glands), which have specialized ducts ing mating interactions in C. maenas), and between indi-
for secretion of products, lack ducts and secrete their viduals of different species (as a feeding deterrent in
products into the spaces between cells, from which the P. litorale against predatory crustaceans such as C. mae-
hormones diffuse into the bloodstream. Once in the nas). Having hopefully impressed you with the varied
blood, hormones travel along the vast network of vessels roles chemical signals play in communication, we now
to virtually every part of the body. Hormones secreted turn our attention exclusively to how such signals func-
by nerve cells are called neurohormones or neurosecre- tion within individuals.
tions. These are produced in the cell body of the nerve
cell, travel along the axon, and are released at the axon
HORMONAL VERSUS NEURAL
tip. Functioning as chemical messengers in an elaborate
COMMUNICATION
system of internal communication, hormones and neu-
rohormones exert their effects at the cellular level by The endocrine system of an animal is closely associated
altering metabolic activity or by inducing growth and with its nervous system. As mentioned, some hormones,
differentiation. Changes at the cellular level can even- in fact, are made by nerve cells. In addition, neurons and
tually influence behavior. hormones often work together to control a single
In this chapter we will focus on how chemical sig- process. For example, in some interactions neurons
nals function within an individual to influence its behav- respond to hormones, whereas in others endocrine
ior, morphology, and physiology. We will see that a given glands receive information and directions from the brain.
hormone can have diverse effects within an individual. Nervous and endocrine systems are so closely associated
It is also worth mentioning, however, that some chem- that they are often discussed as a single system, the neu-
ical signals that function in communication within indi- roendocrine system. Despite this close association,
viduals also function in communication between neural and hormonal modes of information transfer have
individuals. This is the case for some chemical signals in different purposes within the body, and each system is
arthropods, a large group of invertebrates that includes essential in its own right.
insects, crustaceans, spiders, and scorpions. Molting hor- In comparing communication through nervous or
mones in arthropods, known as ecdysteroids, help endocrine pathways, we should first briefly review how
orchestrate the periodic shedding of the exoskeleton in neurons transfer information. In the nervous system,
these animals. One of these molting hormones in crus- information is transmitted along distinct pathways
taceans, 20-hydroxyecdysone (fortunately abbreviated (chains of neurons) at speeds of up to 100 meters per
130 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

second. After the impulse arrives at its destination in the Nucleus

body, neural information is transmitted via a series of Hormone-
electrical events that culminate in the release of neuro- producing cell
transmitter near its target, such as an effector tissue—a
muscle, for example. Neurotransmitters are rapidly Hormone (first messenger)
destroyed after they are secreted. As a result, informa-
tion delivered by the nervous system usually produces
Blood vessel
a response that is rapid in onset, short in duration, and
highly localized.
Plasma membrane of target cell
In contrast to neural communication, hormonal
transfer of information tends to occur in a more
Membrane
leisurely, persistent manner. Typically, hormones are receptor Adenylcyclase
secreted slowly and remain in the bloodstream for some
Cyclic AMP
time. Rather than traveling to a precise location, these (second
ATP messenger)
chemical messengers contact virtually all cells in the
body, although only some cells are able to respond to the
particular hormonal stimulus. The cells that respond,
called target cells, have receptor molecules that recog-
nize and bind to specific hormones. This binding acti-
vates the receptor and initiates the cell’s response to the
hormone, perhaps by turning on certain genes or by
altering the cell’s secretory activity or the properties of Intracellular receptor
its plasma membrane. The precise nature of the response
depends on the type of target tissue because different
types of cells are specialized to perform specific functions Activates specific enzymes
in the body. Only cells with receptors for a particular
hormone can respond to it. The concentration of recep-
tors for a particular hormone determines the cell’s sen- Action indicated by hormone
(change in cell function)
sitivity to it. In short, transfer of information by the
endocrine system often occurs more slowly than that of
the nervous system, and it usually produces effects that FIGURE 7.2 Mechanism of action of some peptide
are more general and long lasting. hormones.

follicle-stimulating hormone (FSH) produced by the

TYPES OF HORMONES
anterior pituitary gland.
AND THEIR MODES OF ACTION
Animals produce two major classes of hormones: pep-
Steroid Hormones
tides and steroids. Although differing in structure and
mode of action, both types cause changes within the cell Steroid hormones are a group of closely related hor-
that eventually influence behavior. mones secreted primarily by the gonads and adrenal glands
in vertebrates. The four major classes of steroids include
progestogens, androgens, estrogens, and corticosteroids.
Peptide Hormones and Amino Acid The first three classes are secreted primarily by the gonads
Derivatives and are often referred to as the sex steroids. All steroid hor-
Peptide hormones are amino acid chains, ranging in mones are chemically derived from cholesterol and hence
size from about 3 to 300 amino acids. These hor- are highly fat-soluble. As a result of their solubility in lipids,
mones, along with amino acid derivatives, are water- steroid hormones move easily through the lipid boundaries
soluble and usually affect cells by binding to receptor of cells and into the cell interior, or cytoplasm (Figure 7.3).
molecules on the cell surface (Figure 7.2). Through a Once inside a cell, steroids combine with receptor mole-
complex sequence of molecular interactions, often cules in the nucleus and, in some cases, in the cytoplasm.
including the use of a secondary messenger (such as If binding occurs in the cytoplasm, then the steroid-recep-
cyclic adenosine monophosphate, or cAMP), peptide tor complex moves to the nucleus. In the nucleus, the com-
hormones create short-term changes in cell mem- plex attaches to DNA and affects subsequent gene
brane properties and long-term changes in protein expression and protein synthesis, a process that may take
function, often by activating enzymes. Examples of several hours to days and one that produces relatively long-
peptide hormones are luteinizing hormone (LH) and lasting effects on behavior.
 The Endocrine System 131

Because we will be discussing the role of steroid hor-

mones in reproductive behavior, two points about the sex Hormone-
steroids should be emphasized here. First, hormonal out- producing cell
put is not rigidly determined by sex. Females generally
have small amounts of “male hormones” such as testos- Hormone
terone, and males typically have low levels of “female
hormones” including estrogens. A second point to Blood vessel
emphasize is that although different hormones produce
different effects, the sex steroids are chemically very sim- Plasma membrane of target cell
ilar (Figure 7.4). Some hormones lie along the pathway
of synthesis of other hormones. Testosterone, for exam-
ple, is an intermediate step in the synthesis of estradiol.
As a result of the common structure of these two steroid
hormones, some behavior patterns may be activated by
injections of either testosterone or estradiol. In other
Intracellular
words, a certain degree of substitutability is associated receptor
with steroid hormones.
We have presented the traditional dichotomy of
peptide versus steroid hormones and their different DNA (genes)
Hormone-
mechanisms of action, but things are never so simple. receptor
For example, we describe steroids as binding to recep- complex
tors inside the cell, and typically modifying gene expres-
sion and protein synthesis over several hours or days to Nucleus
produce relatively long-lasting effects on behavior; Specific enzymes
steroid effects such as these are described as genomic.

Action indicated by hormone

(change in cell function)

FIGURE 7.3 Mechanism of action of steroid hormones in

which gene expression is altered.
Cholesterol

Cholesterol Cholesterol Cholesterol Cholesterol

Pregnenolone Pregnenolone Pregnenolone Pregnenolone

Progesterone Progesterone Progesterone Progesterone

17 OH Progesterone 17 OH Progesterone 17 OH Progesterone 17 OH Progesterone

17 OH Desoxy Androstenedione Androstenedione Androstenedione Androstenedione

corticosterone
Testosterone Testosterone Testosterone Testosterone
Cortisol
19 Nortestosterone 19 Nortestosterone 19 Nortestosterone 19 Nortestosterone

Estradiol 17β Estradiol 17β Estradiol 17β Estradiol 17β

Adrenal cortex cell Ovarian follicle cell Corpus luteum cell Testicular Leydig cell

CH2OH CH3
CO OH CO OH
HO OH

O HO O O
Cortisol Estradiol 17β Progesterone Testosterone

FIGURE 7.4 Biochemical pathways by which steroid hormones are synthesized. Note that many of the steroid
hormones are chemically very similar. (Modified from Daly and Wilson 1983; Tepperman 1980.)
132 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

In recent years, steroids have been found to work in of the boar pheromone, 16-androsterone, than are
additional ways. It is now known that steroids, acting on males (Dorries et al. 1995). Females are attracted to this
the brain through diverse pathways that include inter- chemical and assume a sexually receptive posture in
acting with membrane receptors and altering cell- response to it. Males are not attracted by the boar
signaling pathways, can produce rapid behavioral pheromone. However, if a male is castrated to remove
responses, on the order of seconds or a few minutes the source of testosterone before the age of five months
(reviewed by Moore and Evans 1999; Remage-Healey and and is given the female hormone estradiol as an adult,
Bass 2006). Steroid effects that occur rapidly after steroid he shows the usual female responses to a boar (Adkins-
application are described as nongenomic because the short Regan et al. 1989).
latency indicates that modifications to gene expression and Sometimes hormones mediate behavioral responses
protein synthesis are probably not involved. to particular stimuli. As an example, we will consider the
Finally, we have also described steroids as being pro- effects of vasotocin, a peptide hormone found in non-
duced primarily by the gonads and adrenal glands of ver- mammalian vertebrates that is very similar to the hor-
tebrates. It is important for us to mention that steroid mone vasopressin in mammals. Vasotocin affects how
hormones are also produced by the nervous system male roughskin newts (Taricha granulosa) react to visual,
(Compagnone and Mellon 2000; Plassart-Schiess and chemical, and tactile stimuli from females (Rose and
Baulieu 2001). These steroids, called neurosteroids, dif- Moore 2002). When identifying prospective mates,
fer from those produced by the gonads and adrenal male newts first rely on visual cues such as size, shape,
glands (sometimes called peripheral steroids) in at least and color, and then switch to olfactory cues for close-
one major way. Whereas a peripheral steroid travels in up confirmation of species, gender, and reproductive
the bloodstream and acts on target cells some distance state. If all checks out, then the male clasps the gravid
from its gland of origin, a neurosteroid acts in the ner- female for several hours in a posture known as amplexus
vous system on either nearby cells or the same cell that (Figure 7.5). During this time, the female becomes sex-
produced it. In terms of mechanism of action, some neu- ually receptive and will pick up the spermatophore
rosteroids act in the classic way we described for periph- (package of sperm) that the male subsequently deposits.
eral steroids (binding to receptors inside the target cell Vasotocin has been shown to influence the early stages
and affecting gene expression); others influence recep- of mate choice in males—identification of potential
tors on the membranes of neurons; and still others act mates based on visual and chemical stimuli—as well as
on microtubules within neurons (microtubules are com- the clasping response of amplexus. Males injected with
ponents of the cytoskeleton involved in transport within vasotocin spend more time in the vicinity of female
cells and cell division and movement). Thus, we now rec- visual and olfactory cues than do control males injected
ognize that steroid hormones are produced by more with saline, and they show enhanced responsiveness to
locations within the vertebrate body and have more tactile stimuli from females during amplexus. Although
diverse mechanisms of action than previously thought. the precise location(s) at which vasotocin exerts its
effects along the neural pathways of sensory input and
motor output are not yet known, it is clear that vaso-
tocin influences reproductive behavior by enhancing
HOW HORMONES INFLUENCE male responsiveness to female stimuli.
BEHAVIOR Hormones can also cause a change in preference in
Hormones can influence behavior through several path- animals. Young animals that receive parental care typi-
ways. Generally speaking, hormones modify behavior by cally interact almost exclusively with parents and sib-
affecting one or more of the following: (1) sensory or lings, and when given a choice of social partner, most
perceptual mechanisms, (2) development or activity of youngsters prefer to be around members of their fam-
the central nervous system, and (3) muscles important in ily. However, as the young mature, their social prefer-
the execution of behavior. ence tends to switch to nonfamily members (as any
parent of an adolescent knows), especially those of the
opposite sex. Sex steroids have been implicated in this
EFFECTS ON SENSATION developmental change in social preference (Adkins-
Regan and Leung 2006).
AND PERCEPTION
Hormone-mediated changes in social preference
Hormones influence the ability to detect certain stim- are not limited to maturing animals. Adult female
uli, as well as the responses to and preferences for par- meadow voles, Microtus pennsylvanicus, exhibit an adap-
ticular stimuli. In some species, mate choice is at least tive seasonal change in odor and social preference.
partially based on hormone-mediated differences in the During the winter, female meadow voles nest commu-
ability to detect stimuli. In domestic pigs (Sus scrofa), nally with other females and, at this time, prefer female
for instance, females are able to detect lower quantities odors to male odors. However, in the spring and sum-
 How Hormones Influence Behavior 133

FIGURE 7.5 A male roughskin newt (on

top) in amplexus with a female. The hor-
mone vasotocin enhances male responsive-
ness to visual, chemical, and tactile stimuli
from the female.

mer, when they defend territories against other females the hormonal milieu (environment) soon after hatch-
and mate with males, they prefer the scent of a male to ing. The steroid hormones involved in the early mas-
that of a female. Thus, changing odor preferences help culinization of the zebra finch brain (and indeed in the
female voles choose their company so they can success- general organization of the developing brain in this
fully raise as many offspring as possible. The reversal in species) appear to be neurosteroids rather than gonadal
odor preference is caused by changes in the amount of steroids (London and Schlinger 2007; Wade and
estrogen the female produces. Estrogen levels fluctuate Arnold 1996). (See Chapter 8 for a detailed discussion
in response to seasonal changes in the length of day- of bird song.)
light, with higher levels associated with longer days
(Ferkin and Zucker 1991).
EFFECTS ON MUSCLES

EFFECTS ON DEVELOPMENT AND Hormones can influence behavior by affecting muscles

ACTIVITY OF THE CENTRAL and motor neurons. Consider, for example, two cases of
sexually dimorphic patterns of behavior—calling behav-
NERVOUS SYSTEM
ior in frogs and copulatory movements in rodents—that
Circulating hormones can affect behavior by influenc- illustrate hormonal influences on muscles.
ing the central nervous system. In fact, hormones have Our first example concerns the calling behavior of
been found to influence several characteristics of dif- the South African clawed frog, Xenopus laevis (reviewed
ferent regions of the brain, including (1) the volume of in Kelley 1996). Clawed frogs occur in sub-Saharan
brain tissue, (2) the number of cells in brain tissue, (3) Africa, where they inhabit shallow, and often murky,
the size of cell bodies of neurons, (4) the extent of den- bodies of water. Males of this species emit six different
dritic branching of neurons, (5) the percentage of neu- calls, the most common of which is the advertisement
rons sensitive to particular hormones, and (6) the call (Tobias et al. 2004). This metallic-sounding call con-
survival of neurons. sists of alternating fast and slow trills; during fast trills,
An example in which hormones influence the cen- there may be a progressive increase in volume. The
tral nervous system concerns the development of advertisement call allows females to find males in their
singing behavior in birds. In the zebra finch typically soupy locations. Sexually receptive females
(Taeniopygia guttata), sex differences in the brain nuclei approach calling males and produce a rapping call, which
that control song are established around the time of stimulates the males to answer, and a duet ensues (Tobias
hatching. Early exposure to steroid hormones regulates et al. 1998). Receptive females permit males to clasp
the size of song nuclei, the size and number of neurons them around the waist for several hours while their eggs
within these brain areas, the extent of dendritic branch- are released and fertilized. Whereas receptive females
ing, and the number of androgen receptors (reviewed rap, females that are not sexually ready tick. The ticking
in Wade and Arnold 2004). Thus in the zebra finch, sex call consists of slow, monotonous clicks with no change
differences in adult singing behavior (males sing and in intensity; the rapping call is similar but has a some-
females do not) are linked to differences that are estab- what shorter interval between clicks. The advertisement
lished in the brains of males and females as a result of call of males and the ticking call of unreceptive females
134 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

have been most completely studied and are shown in times per second in females). Adult males have 8 times
Figure 7.6. as many muscle fibers in their larynx as do females. Also,
How does it come about that male clawed frogs pro- male muscle cells are of the fast-twitch, fatigue-resistant
duce advertisement calls of rapid trills, whereas unre- type, whereas most muscle cells in the larynx of females
ceptive females produce the slow ticking call? Darcy are slow-twitch and fatigue-prone.
Kelley and co-workers demonstrated that characteristics What causes these differences in laryngeal muscles?
of the muscles and neuromuscular junctions of the lar- At the time of metamorphosis (the change from tadpole
ynx are responsible for sex differences in the rate at to frog), the number of muscle fibers in the larynx of
which calls are produced (in males the muscles of the lar- males and females is identical to the number in the lar-
ynx contract and relax 71 times per second and only 6 ynx of adult females. Apparently, as males mature and
their levels of androgens rise, new fibers are added. In
addition to increasing the number of muscle fibers,
androgens also influence the type of fiber, promoting
a expression of the fast-twitch cells. Hormones other than
androgens, including prolactin and secretions from the
thyroid gland, also appear to play a role in masculiniz-
ing the larynx. In short, sex differences in the calling
behavior of X. laevis can be traced, in part, to hormone-
kHz induced changes in the muscles of the larynx.
Slow Fast Single Our second example of sex differences in behavior
7 click
6 resulting from hormonal influences on muscles concerns
5
the perineal muscles involved in mating (the perineum
is the area between the urogenital and anal openings).
4
The levator ani/bulbocavernosus muscles control copu-
3
latory reflexes in male Norway rats (Hart 1980).
2 Although these muscles are present in the perineum of
1 both sexes at birth, they are completely absent in adult
0 females. Breedlove and Arnold (1983) have shown that
0 1 the levator ani/bulbocavernosus muscles shrink and fold
1 Second
inward in females unless supplied with androgen. The
lack of androgen in females during the perinatal period
b
(the time surrounding birth) also results in the death of
the motor neurons that supply these muscles. Thus, sex
differences in the copulatory movements of adult rats
result, in part, from early hormonal influences on the
growth and maintenance of the specific muscles and
kHz Single motor neurons involved in mating.
click Most mammals studied to date are similar to
7
Norway rats in having sexually dimorphic perineal mus-
6
cles and motor neurons; that is, the muscles and motor
5 neurons that innervate them are absent or reduced in
4 females. An interesting exception to this pattern has
3 emerged in the case of the naked mole rat (Heterocephalus
2 glaber). We begin by providing some necessary back-
1 ground on this unique species. Naked mole rats are
0
cooperatively breeding rodents that live in large colonies
0
1 Second
1 (Jarvis 1981). Each colony contains a queen (the only
breeding female), one or a few breeding males, and
FIGURE 7.6 Differences in the calling behavior of male numerous nonreproductive subordinates, almost all of
and female clawed frogs result from the effects of hor-
which will never breed (a subordinate can attain breed-
mones on the muscles of the larynx. (a) Male clawed
ing status only when a breeder dies or upon dispersing
frog clasping a sexually receptive female that responded
to his mating call (shown below pair). The male’s call from the colony and encountering a potential mate).
consists of slow and fast trills. (b) Male clasping a sexu- Male and female subordinates are remarkably similar in
ally unreceptive female that responds to his clasp by their behavior and body size, and even in their external
emitting the ticking call (shown below pair). (From genitalia, although internal reproductive organs are nor-
Kelley and Gorlick 1990.) mally sexually differentiated (Jarvis 1981; Lacey and
 Methods of Studying Hormone–Behavior Relationships 135

Sherman 1991). Consistent with these observations of

few sex differences in subordinate naked mole rats are
data showing that the perineal muscles and their motor
neurons are also sexually monomorphic (i.e., the same in
males and females) (Peroulakis et al. 2002). It seems that
sexual differentiation in behavior and reproductive
structures is limited in subordinate naked mole rats, and
this makes sense given the similar nonbreeding roles
played by low-ranking males and females in colony life.
Among breeders, the levator ani muscle is actually larger
in the queen than in breeding males, while other perineal
muscles are similar in size (Seney et al. 2006). The results Significantly different
for size comparisons of the levator ani muscles of breed- (queen > breeder male)
ers and nonbreeders are shown in Figure 7.7. In species

Levator ani muscle volume (mm3)

10
of mammals studied to date, the perineal muscles con- Not significantly
trol reflexes of the penis; their role, if any, in females with 8 different
reduced muscles and motor neurons is unknown. The
suggestion for naked mole rats is that the enlarged lev- 6
ator ani muscle helps the queen deliver the enormous
number of offspring that she will produce over the 4
course of her lifetime (a queen can produce litters every
2
80 or so days and can have up to 28 young in a litter!).
Although much work remains to be done on the 0
endocrinology of naked mole rats and other coopera- Female Male Queen Male
tively breeding mammals, the current data demonstrate subordinate subordinate breeder
that hormonal influences on perineal muscles reflect FIGURE 7.7 Muscles of the perineum are sexually
reproductive life history. dimorphic in most mammals, being larger in males than
in females as a result of the early effects of androgens.
In naked mole rats, however, perineal muscles do not
differ in size between subordinate males and females,
METHODS OF STUDYING perhaps reflecting their similar nonbreeding roles in
HORMONE–BEHAVIOR colony life (results for one perineal muscle, the levator
RELATIONSHIPS ani, are shown here). The levator ani muscle of the
queen is actually larger than that of breeding males.
Several techniques are available for the study of hor- (From Seney et al. 2006.)
monal influences on behavior. Here we examine two
general approaches to questions about behavioral
endocrinology. The first approach might be called inter-
ventional because the experimenter manipulates the hor-
mones of the animal. This often involves the removal of David Crews (1974, 1979a) used gland removal (cas-
a gland, followed by hormone replacement therapy. In tration) and hormone (androgen) replacement therapy in
the second approach, researchers look for changes in the study of hormonal control of sexual and aggressive
behavior that parallel fluctuations in hormone levels. behavior in lizards. Among his favorite subjects is Anolis
Studies that use this second approach are called corre- carolinensis, the green anole. This small iguanid lizard
lational studies. inhabits the southeastern United States and displays a
social system in which males fiercely defend their terri-
tories against male intruders. The territory of a single
INTERVENTIONAL STUDIES
male often encompasses the home ranges of two or three
Fairly conclusive evidence of the function of a hormone females. As you might expect from these living arrange-
can be gained by removing its source, that is, the ments, male anoles have an interesting repertoire of
endocrine gland, and recording the subsequent effects. aggressive and sexual behaviors (Figure 7.8).
The hormone is then replaced by implanting a new Both agonistic and sexual displays of male A. caro-
gland or by administering the hormone. If the effects of linensis share a species-typical bobbing movement,
gland removal are reversed by replacing the hormone, which is made even more dramatic by extension of the
we conclude that the hormone was responsible for the red throat fan, or dewlap. When confronted by a male
changes. intruder, a resident male anole immediately begins to
136 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

a match ensues as the resident and intruder circle, with

locked jaws, in an attempt to dislodge the other from
the prized perch. Engaging in aggressive behavior is
apparently quite rewarding to males, for they prefer to
spend time at the sites of their previous aggressive
encounters (Farrell and Wilczynski 2006). The
courtship behavior of male green anoles is very similar
to their aggressive behavior. Typically, however, the
b body is not laterally compressed, and the bobbing
dewlap display is less stereotyped in courtship. In effect,
each male has his own version of how best to attract
females.
Once Crews had documented the display reper-
toire of the feisty A. carolinensis, he set out to examine
hormonal control of male aggressive and sexual behav-
ior through castration and androgen replacement ther-
apy. Removal of the testes led to a sharp decline in
FIGURE 7.8 Displays of the male green anole. sexual behavior, but administration of testosterone
(a) Aggressive posturing between two males often implants reinstated this behavior to precastration lev-
includes extension of the dewlap and stereotyped els (Figure 7.9) (Crews 1974; Crews et al. 1978). Thus,
bobbing. (b) Courtship displays directed by a male to Crews and his co-workers concluded that testosterone
a female (smaller individual) are similar to aggressive
regulates courtship and copulation in the male green
displays in appearance, and both displays are mediated
by testosterone.
anole.
The relationship between testosterone and aggres-
sive behavior was not so simple. If a male was castrated
display—usually by compressing his body and adjust- and returned to his home cage, he continued to be
ing his posture in such a way as to present the intruder aggressive toward intruders. However, if the male was
with a lateral view of his impressive physique. As if this castrated and placed in a new cage, his aggressive behav-
were not enough, the resident then lowers his hyoid ior declined in a manner similar to that noted for sexual
apparatus (a structure in the back of the throat that is behavior. Thus, unlike sexual behavior, aggressive behav-
responsible for movements of the tongue) and exhibits ior appears to be only partially dependent on gonadal
a highly stereotyped bobbing pattern. The display ends hormones and subject to influence by social factors such
at this point if the intruder rapidly nods his head, as residence status.
thereby acknowledging his subservient position. Although simple in concept, interventional studies
However, if the intruder fails to display the submissive have become quite sophisticated as a result of major
posture, the display of the resident male escalates to advances in techniques for manipulating hormone lev-
ever-increasing frequencies. In the heat of confronta- els. For example, cannulation techniques now allow
tion, the two combatants acquire a crest along the back administration of minute amounts of hormone to spe-
and neck and a black spot behind each eye. A wrestling cific brain regions. Other advances utilize techniques

Androgen
Castration replacement

2.0
Mean number of displays per min

1.0

FIGURE 7.9 Effect of castration and testos-

0 terone replacement therapy on the courtship
1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 9 behavior of the male green anole. (Modified
Behavioral test from Crews 1979b.)
 Methods of Studying Hormone–Behavior Relationships 137

a b
100 50
First litter

Percent pups attacked

Second litter 40
Percent infanticide

75

30
50
20

25
10

0 0
C57BL/6 PRKO Oil Progesterone RU486

FIGURE 7.10 Several different interventions demonstrate that progesterone regulates aggression shown by male mice
toward infants. (a) Control males from the C57BL/6 laboratory strain kill offspring in their first litters about 74% of the
time; instances of infanticide decline somewhat with second litters but are still high (58%). PRKO males are insensitive to
progesterone because they lack progesterone receptors; these males show no infanticide toward young in either first or
second litters. (b) Treatment with progesterone increases attacks by males on young relative to control males receiving
oil, and treatment with RU486, an antihormone that blocks progesterone receptors, decreases attacks. Males used in
the experiments shown in (b) had never mated, so were tested with pups from other pairs. (From Schneider et al. 2003.)

whereby hormones are labeled with radioactivity and and now we see that progesterone mediates aggression
their paths traced through the body. The discovery of directed at infants.
antihormones, drugs that can temporarily and
reversibly suppress the actions of specific hormones, has
also aided investigation of hormonal influences on CORRELATIONAL STUDIES
behavior. We can also study hormonal influences on behavior with
Genetic “knockout” mice also offer new opportuni- correlational studies. In using this approach, researchers
ties to manipulate hormone levels to study the relation- look for changes in behavior that parallel fluctuations in
ship between hormones and behavior. A knockout hormone levels. Correlational studies are useful, but they
mouse is one in which a specific gene is targeted and are not as conclusive as experimental work because there
inactivated to eliminate the gene product. In this case, is no evidence of causation. Consider, for example, a cor-
the gene product may be a hormone or a receptor for a relational study that revealed the relationship between the
hormone. For example, there are progesterone receptor level of testosterone and aggressive behavior in a songbird.
knockout (PRKO) mice; these mice do not respond to John Wingfield has examined the behavioral
progesterone because they lack the appropriate recep- endocrinology of birds under natural conditions (for a
tors. Whereas males of most strains of laboratory mice review, see Wingfield and Moore 1987). In one study of
behave aggressively toward infants and often kill them, song sparrows, Melospiza melodia (Figure 7.11), Wingfield
PRKO males are not aggressive to infants (Figure 7.10a)
(Schneider et al. 2003). Like PRKO males, male mice
whose progesterone receptors have been temporarily
blocked by administration of the antihormone RU486
rarely attack pups. Finally, male mice that have received
progesterone implants (silastic implants filled with prog-
esterone and sesame oil) are more aggressive toward
infants than control mice receiving implants filled only
with sesame oil (Figure 7.10b). Taken together, these
findings indicate that progesterone mediates the aggres-
sion shown by adult male mice toward infants. Aside
from demonstrating several techniques used in research
in behavioral endocrinology, this study is interesting for
at least two more reasons. First, progesterone has tradi-
tionally been viewed as a hormone that influences female
behavior and physiology, and here we have an effect of
progesterone on male behavior. Second, different hor-
mones seem to mediate different forms of aggression in
adult male mice; it has long been known that testos-
terone mediates aggression directed at other adult males, FIGURE 7.11 Male song sparrow.
138 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

(1984) captured males in mist nets or traps baited with clutch. Wingfield (1984) speculates that high levels of
seed, collected a small blood sample from the wing vein, testosterone and the resulting heightened levels of male
and marked each individual with a unique combination aggression would interfere with paternal behavior. Field
of leg bands. Birds were released at the site of capture and studies of the song sparrow demonstrate that circulating
seemed relatively unperturbed by the sampling proce- levels of testosterone wax and wane in parallel with
dure; in fact, some individuals sang within 15 to 30 min- changing patterns of male territorial aggression. This
utes of release. A given male was sampled from five to ten correlational evidence strongly suggests that testosterone
times during a single breeding season, and each sample mediates aggressive behavior during the breeding season
was analyzed for testosterone. Wingfield also observed in this species. Finally, we have the challenge
their behavior during this period. hypothesis (Wingfield et al. 1990), which states that lev-
As shown in Figure 7.12, Wingfield found a close els of hormones, such as testosterone, that regulate
correlation between peak levels of male territorial and aggression and dominance are influenced by the social
aggressive behavior and maximum levels of testosterone. environment, and that their levels rise during times of
A male song sparrow defends his territory most intensely social challenge or instability (e.g., during the initial
during its initial establishment and when his mate is lay- period of territory establishment in the case of song spar-
ing the first clutch. At the time of egg laying, females are rows). This hypothesis is now being tested in other
sexually receptive and males aggressively guard them species of vertebrates (Hirschenhauser and Oliveira
from other would-be suitors. Testosterone reaches peak 2006) and also in insects (Trumbo 2007). The relation-
levels during the initial period of territory establishment ship between testosterone and aggressive behavior is also
and during the laying of eggs for the first brood. discussed in Chapter 18.
Interestingly, not only does testosterone increase aggres- In male song sparrows in sedentary (nonmigratory)
sive behavior in song sparrows at these times, but the populations, territorial aggression is not limited to the
reverse is also true; that is, aggressive interactions can breeding season. Indeed, such males also exhibit territo-
increase plasma levels of testosterone. So, we see that a rial aggression during the nonbreeding season, after they
hormone may influence a particular behavior, and that have completed molting in late summer. Does testos-
behavior, in turn, may influence levels of the hormone. terone regulate aggression during the nonbreeding season
It is also interesting that testosterone does not peak dur- when testes have regressed? Apparently not, because lev-
ing the period when the female is sexually receptive and els of testosterone are undetectable in the plasma of non-
laying the second clutch. Aggressiveness is correlated breeding males and do not increase following aggressive
with testosterone levels. A male guards his mate with less interactions. Instead, the culprit appears to be estradiol.
enthusiasm during the second laying period than during It seems that dehydroepiandrosterone (DHEA), a pre-
the first period. This pattern may be related to the fact cursor that can be converted into active sex steroids in
that while a female is laying the second clutch, a male is appropriate tissues, is metabolized in the brain of males
often responsible for feeding fledglings from their first to form estradiol, which fuels aggression during the non-
breeding period (Wingfield and Soma 2002). DHEA may
originate in the adrenal glands or the regressed testes. It
is also possible that estradiol forms directly in the brain
from cholesterol (recall that the nervous system can also
Brood 1 Brood 2 make steroid hormones). Why might a different hormone
Establishing territory

mediate territorial aggression during breeding and non-

Feeding fledglings

Feeding fledglings
Feeding nestlings

Feeding nestlings

Postnuptial molt

breeding periods? Several studies have shown that high

Plasma level of testosterone (ng/ml)

Pair formation

levels of testosterone can be costly to birds in terms of

Egg-laying

Egg-laying
Incubation

Incubation
Migration

increased metabolic rate and decreased body mass, fat

stores, and immune function (reviewed in Soma 2006).
The energetic costs of testosterone would be most criti-
cal during the nonbreeding season when birds experience
lower temperatures and reduced food supplies.
Researchers therefore suggest that the mechanism of
DHEA-estradiol mediated aggression may have evolved
so that males could avoid the costs of high testosterone
during the nonbreeding season.
F M A M J J A S O In recent years, it has become possible to monitor
Calendar month hormone levels through the analysis of urine and feces
FIGURE 7.12 Changes in circulating levels of testos- rather than blood. These less invasive procedures are
terone in free-living male song sparrows as a function of often used in field studies and when repeated sampling
the stage of the breeding cycle. (From Wingfield 1984.) is necessary (Whitten et al. 1998).
 Organizational and Activational Effects of Hormones 139

permanent, and form the basis for sex differences in mat-

STOP AND THINK ing behavior.
Corticosterone is a steroid hormone released from the Steroid hormones also affect behavior by activating
adrenal glands in response to stress. If you were charged neural systems responsible for mediating specific pat-
with measuring levels of corticosterone in free-living
terns of behavior. In contrast to organizational effects,
squirrels, how would you proceed? Specifically, would you
activational effects usually occur in adulthood and tend
collect blood, urine, or feces? What would you consider
when making your choice? Would your choice differ if to be transient, lasting only as long as the hormone is
you were measuring a different hormone? present at relatively high levels. In keeping with their
impermanence, activational effects are thought to
involve subtle changes in previously established con-
nections (such as slight changes in neurotransmitter pro-
ORGANIZATIONAL duction or release along established pathways) rather
AND ACTIVATIONAL EFFECTS than gross reorganization of neural pathways. At this
OF HORMONES point, we will consider the organizational and activa-
tional effects of steroid hormones as they relate to the
The modes by which steroid hormones influence behav- development and display of sexual behavior in the
ior may be classified as organizational and activational Norway rat.
(Phoenix et al. 1959). We first define the dichotomy, and
then we provide two examples of these different effects,
the first focusing on sex differences in the mating behav- SEX DIFFERENCES IN THE BEHAVIOR
ior of rats and the second on individual differences in OF NORWAY RATS
aggressive behavior of male tree lizards. We conclude Not surprisingly, adult male and female rats differ in their
this section by presenting some questions raised by sexual behavior (Figure 7.13). Whereas social investiga-
researchers about the usefulness of the dichotomy. tion, mounting, intromission, and ejaculation typify mat-
ing in males, behavioral patterns associated with
solicitation and acceptance characterize the sexual behav-
DEFINING THE DICHOTOMY
ior of females. The lordosis posture, for example, is a cop-
In organizational effects, steroids organize neural path- ulatory position that female rats assume when grasped on
ways responsible for certain patterns of behavior. the flanks by an interested male. The intensity of the lor-
Organizational effects occur early in life, usually just dosis response varies across the ovulatory cycle, being
before or after birth, and tend to be permanent. This most pronounced when mature eggs are ready to be fer-
permanence implies structural changes in the brain or tilized. The sexual behavior of female rats also includes
other long-term cellular changes, such as in the respon- a variety of solicitation behaviors, such as ear wiggling
siveness of neurons to steroid hormones (Arnold and and a hopping and darting gait, that typically precede dis-
Breedlove 1985). Steroid hormones can also have orga- play of the lordosis posture (Beach 1976). Although
nizational effects on nonneural systems. For example, mounting is almost always associated with males and lor-
testosterone during the late embryonic period organizes dosis with females, these behavior patterns occasionally
development of the anal fin and its skeletal support struc- occur in the other sex. Every once in a while, females will
tures in western mosquitofish, a species in which the mount other females and, similarly, males will occasion-
male’s modified anal fin is used to fertilize eggs inside the ally accept mounts from their cagemates. However, by
female (Rosa-Molinar et al. 1996). These changes to the and large, males display mounting and females assume
male fin and its skeletal supports occur early in life, are the lordosis posture. These differences in patterns of

FIGURE 7.13 Male and female Norway

rats differ in their sexual behavior.
Whereas mounting is characteristic of
males, the acceptance posture, called
lordosis, is characteristic of females.
These sex differences in adult behavior
are established through the action of
steroid hormones around the time
of birth.
140 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

adult copulatory behavior are due to differences in the Treatment Adult behavior
brains of male and female rats, differences that were Sex of patterns
newborn
induced by the irreversible actions of androgens in late
fetal and early neonatal life. Let’s now consider in some
detail the organizational effects of gonadal steroids on None
sexual behavior.
It is clear that testosterone in the bloodstream of Castrated at No further
age 1 day treatment
neonatal rats produces organizational effects. During
perinatal life, male and female rats have the potential to Testosterone injection
develop neural control mechanisms for both masculine before age 5 days
and feminine sexual behavior. Certain neurons in the
brains of both males and females have the capacity to
bind sex hormones. During a brief period, starting about
two weeks after conception and extending until approx-
None
imately four to five days after birth, however, testos-
terone secreted by the testes of developing males is Testosterone injection
bound to receptors in the target neurons (testosterone before age 5 days
also binds to target muscles at this time). Once there,
testosterone initiates the production of enzymes that will
switch development onto the “male track.” The neonatal
FIGURE 7.14 Pattern of sexual differentiation in the
testosterone causes males to (1) develop the capacity to
brain and behavior of the Norway rat.
express masculine sexual behavior and (2) lose the capa-
city to express feminine copulatory behavior.
Experiments involving castration and hormone estradiol, a hormone usually associated with females, that
replacement techniques have demonstrated the organi- actually directs development along the masculine track.
zational effects of early secretion of testosterone. Testosterone enters neurons in specific regions of the
Removal of the testes in a rat soon after birth results in brain and is converted intracellularly to estradiol, which
an adult with a reduced capacity to display masculine in turn causes masculinization. Look again at Figure 7.4
patterns of sexual behavior and an enhanced capacity to to see that steroid hormones are chemically very similar
display feminine patterns. These males are capable of and that testosterone lies along the pathway of synthe-
high levels of female solicitation and lordosis as adults. sis of estradiol. For the record, androgens, acting
However, if removal of the testes is followed by an exper- through androgen receptors, still have a role in copula-
imental injection of testosterone before five days of age tory behavior; recall that androgens masculinize the
and the proper male hormones are administered in muscles of the penis (levator ani/bulbocavernosus) and
adulthood, the rat will display normal male sexual behav- their associated motor neurons.
ior. Normal female fetuses produce low levels of testos- The main question that arises from estradiol’s role
terone, so the male developmental pattern is not in the masculinization process is this: Why doesn’t estra-
initiated. A single injection of testosterone into a female diol have the same effect in young female rats? To begin
rat soon after birth, however, produces irreversible with, the levels of estradiol in young females are very
effects on her adult sexual behavior. The testosterone- low. In addition, during this critical period of brain
treated female shows fewer feminine and more mascu- development, an estrogen-binding protein, called alpha-
line patterns of copulatory behavior than does a normal fetoprotein, is produced in the livers of the fetuses. This
female. Thus, the development of a “male” brain protein, found in the cerebrospinal fluid of newborn
requires the presence of testosterone around the time of males and females, persists in ever-decreasing amounts
birth. In the absence of testosterone, a “female” brain during the first three weeks of life. During this time,
develops. The effects of perinatal testosterone secretion alpha-fetoprotein prevents estradiol from reaching tar-
on adult sexual behavior are organizational in that they get neurons in the brain. In female rats, then, alpha-feto-
occur early in life and involve permanent structural protein binds any circulating estradiol and thereby
changes in the brain. Figure 7.14 summarizes sexual dif- prevents it from initiating the male pattern (McEwen
ferentiation in the brain and behavior of the young rat. 1976). Alpha-fetoprotein does not, however, bind testos-
Before moving to the activational effects of sex terone. Thus, in male rats, testosterone produced by the
steroids, we should mention that masculinization of the testes can reach the brain, be converted to estradiol, and
brain and behavior is somewhat more complex than just result in sexual differentiation.
described. In laboratory rats, testosterone appears to be In adulthood, steroid hormones produce activa-
only an intermediate chemical in the process, and it is tional effects on sexual behavior in male and female
 Organizational and Activational Effects of Hormones 141

rats. Female rats with high blood levels of estrogen and dewlaps (Moore et al. 1998). Males are one phenotype
progesterone display feminine sexual behaviors in the or the other and remain so for life; such alternative phe-
presence of a sexually active male, but these patterns notypes are said to be developmentally fixed. The two types
rarely occur when levels of these ovarian hormones are of males have similar hormone profiles in adulthood, so
low. In fact, an adult female whose ovaries have been activational effects are not indicated. It has been found
removed will not copulate unless she receives injections that differences between the male phenotypes are orga-
of estrogen and progesterone. Similarly, removal of the nized by steroid hormones during the neonatal period.
testes in an adult male eventually eliminates copulatory Further work with tree lizards has shown that the nonag-
behavior, unless he is given injections of testosterone. gressive males are flexible in their territorial behavior,
In these cases, the effects of steroid hormones on sex- switching between sedentary and nomadic behavior in
ual behavior are described as activational because estro- response to environmental conditions (Knapp et al.
gen and progesterone in females and testosterone in 2003); these alternative phenotypes are thus said to be
males presumably exert their effects by activating exist- developmentally plastic. When conditions become stress-
ing neural pathways. High levels of the gonadal ful, the hormone corticosterone rises and ultimately
steroids activate specific patterns of sexual behavior. (through its effects on testosterone) reduces site fidelity,
Thus, in contrast to permanent changes in sexual triggering nomadic behavior in nonaggressive males.
behavior caused by administration of testosterone dur- This switch in territorial behavior is prompted by
ing the neonatal period, only a transient activational changes in hormone levels in adulthood and thus
effect on copulatory behavior is produced by sex reflects activational effects of steroid hormones. Males
steroids in adulthood. of the aggressive phenotype do not show plasticity in
One final point will help to distinguish organiza- their territorial behavior; they are always territorial. The
tional and activational effects of steroid hormones on findings for tree lizards support the relative plasticity
sexual behavior. Males and females that have had their hypothesis proposed by Michael Moore concerning the
reproductive organs removed in adulthood generally actions of steroid hormones and alternative male phe-
cannot be induced to behave like members of the notypes. Moore (1991) hypothesized that alternative
opposite sex. For example, a female rat whose ovaries phenotypes that are developmentally fixed rely on orga-
have been removed in adulthood cannot, through nizational effects of steroid hormones, whereas alterna-
injections of testosterone, be induced to show mount- tive phenotypes that are developmentally plastic rely on
ing behavior. By adulthood, the nervous systems of activational effects.
adult males and females have already differentiated
(i.e., the organizational effects of early steroid secre-
QUESTIONING THE DICHOTOMY
tion have long since occurred), and the mature brains
are not capable of responding to hormonal signals of Although many researchers employ the distinction of
the opposite sex. organizational and activational effects of steroid hor-
mones on behavior, Arnold and Breedlove (1985) ques-
tioned the usefulness of this dichotomy. They reviewed
INDIVIDUAL DIFFERENCES IN THE
experimental findings from the previous decade and con-
BEHAVIOR OF MALE TREE LIZARDS
cluded that the organizational–activational distinction
We have focused on how organizational and activa- was too restrictive. How would one classify, for exam-
tional effects of steroid hormones explain differences ple, effects produced by steroid hormones that were both
between the sexes, using sexual behavior of the Norway organizational and activational in nature, such as the pro-
rat as our example. More recently, the organiza- duction of permanent effects in adulthood? Also, while
tional/activational dichotomy has been used to under- acknowledging the wealth of behavioral evidence sup-
stand differences between individuals of the same sex. porting the organizational–activational dichotomy, their
Because individual differences are most pronounced in attempts to uncover biochemical, anatomical, or physi-
species with naturally occurring alternative male phe- ological evidence of two fundamentally different ways in
notypes, these species have been the focus of this line which steroid hormones act on the nervous system were
of research (Crews and Moore 2005). In the chapter unsuccessful. In their opinion, failure to find specific cel-
opener we described the alternative male phenotypes of lular processes uniquely associated with each type of
plainfin midshipman fish; now we consider those of the effect further blurs the organizational–activational dis-
tree lizard. tinction. Although it is important to keep such concerns
The tree lizard (Urosaurus ornatus) has alternative in mind when discussing steroid influences, we believe
male phenotypes that differ in aggressive behavior and that the traditional distinction of organizational and acti-
color of the dewlap. Aggressive males have orange-blue vational effects is still useful in categorizing hormonal
dewlaps, whereas nonaggressive males have orange effects on behavior.
142 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

60
THE DYNAMIC RELATIONSHIP
BETWEEN HORMONES AND

Androgen concentration (ng/ml plasma)

50
BEHAVIOR
The interaction between hormones and behavior is a 40
dynamic one. As an illustration, we will consider the rec-
iprocal relationship between hormones and behavior,
30
and also show how hormones can rapidly and adaptively
suppress a behavior.
20

A RECIPROCAL RELATIONSHIP
10
Whereas hormones can activate specific forms of behav-
ior, behavioral stimuli can, in turn, induce rapid changes
0
in the levels of those hormones. We already have men-
Amplexus Amplexus Amplexus Amplexus
tioned that aggressive interactions in male song sparrows time 0 1 hour 2 hours 3 hours
cause levels of testosterone to rise. Sexual stimuli have
also been shown to trigger rapid increases in androgen FIGURE 7.15 Changes in plasma androgen in male
marine toads as a function of the time spent in
levels. The marine toad (Bufo marinus), a native of
amplexus. The rise in androgens during amplexus
Central and South America, is an explosive breeder and suggests that mating behavior induced the hormonal
the first amphibian species in which it was shown that response rather than vice versa. (From Orchinik
sexual behavior could affect hormonal state. et al. 1988.)
Orchinik, Licht, and Crews (1988) studied two pop-
ulations of marine toads in Hawaii, where the species with amplexus and “hope for the best,” or does he ter-
breeds year-round, with bursts of mating activity fol- minate the behavior and seek a safe hiding place?
lowing heavy rainfall. During these breeding explosions, Researchers have found that the hormone corticosterone
males typically compete to clasp the limited number of rapidly suppresses amplectic-clasping behavior in male
females, and mating involves prolonged amplexus. roughskin newts. This was discovered by administering
When male toads were allowed to clasp stimulus females corticosterone to amplexing males and also by exposing
for zero, one, two, or three hours, concentrations of males to stressful conditions known to prompt corticos-
androgens (testosterone and a form of testosterone called terone secretion (Moore and Miller 1984). In addition
5-alpha-dihydrotestosterone, or 5-DHT) increased with to its observable suppressive effects on clasping behav-
the number of hours spent in amplexus (Figure 7.15). In ior, corticosterone also reduces the activity of certain
addition, in field-sampled males, androgen concentra- neurons that are typically active when tactile stimuli trig-
tions were higher in amplexing males than in unpaired, ger amplexus (Rose and Moore 1999). The actions of
“bachelor” males. The apparent rise in androgens dur- corticosterone on neurons and clasping behavior occur
ing amplexus suggests that mating behavior induced the within minutes of its administration; these rapid
hormonal response rather than vice versa. Although a responses indicate that corticosterone is acting via a
similar relationship between amplexus and androgen receptor in the membrane of neurons rather than by
level has been found in several other species of frogs and binding to an intracellular receptor and altering gene
toads, the pattern is not found in all species examined to expression and protein synthesis (Moore and Evans
date (Moore et al. 2005). 1999; Orchinik et al. 1991). We see, then, that the
dynamic interactions between hormones, behavior, and
HORMONAL SUPPRESSION the nervous system allow the behavioral state of an ani-
mal to shift rapidly and adaptively (if not always conve-
OF BEHAVIOR
niently!) to changing environmental circumstances.
Hormones can also rapidly and selectively suppress a
behavior, when such suppression is appropriate. We
return to the roughskin newt (Taricha granulosa), whose INTERACTIONS BETWEEN
mate choice behavior we described earlier in the chap- HORMONES, BEHAVIOR,
ter. Androgens and vasotocin mediate amplexus in AND THE ENVIRONMENT
roughskin newts, and undisturbed males will clasp a
gravid female for several hours while she becomes sex- Our example of the roughskin newt shows that hor-
ually receptive. But what happens when a male in mones provide a mechanism through which an animal
amplexus detects a predator nearby? Does he continue can adjust its behavior so that it is appropriate for the sit-
 Interactions Between Hormones, Behavior, and the Environment 143

uation at hand. Next, we take a look at how hormones, a Mating

behavior, and the environment interact to generate adap-
tive behavior in the long term. With respect to the role
of the environment, we first consider the physical envi-
ronment and then the social environment.

ADJUSTING TO THE HARSHNESS AND

Associated
PREDICTABILITY OF THE PHYSICAL
ENVIRONMENT
The habitats of different species vary in the number of b
mating opportunities they provide. As a result, the asso-

Gonadal activity
ciation between gonadal hormones and sexual behavior
varies among species in ways that allow animals to pro-
duce the greatest number of surviving offspring.
David Crews (1984, 1987) compared patterns of
reproduction in a wide variety of vertebrates and
found numerous exceptions to the “rule” of hormone Dissociated
dependence of mating behavior that we observed in c
the Norway rat. In his survey, Crews considered rela-
tionships among the following three components of
the reproductive process: (1) production of gametes, Constant
(opportunistic)
(2) secretion of sex steroids by the gonads, and (3) tim-
ing of mating behavior. Amid the diversity of repro-
ductive tactics of vertebrates, the following three
general patterns of reproduction emerged: associated,
dissociated, and constant (Figure 7.16). As we describe
these three general patterns, keep in mind that even Time
these three categories do not cover all of the repro-
ductive patterns exhibited by vertebrates. Indeed, as FIGURE 7.16 Vertebrates exhibit three general patterns
data on additional species accumulate, it is clear that of reproduction. (a) In species exhibiting the associated
some species fall between these three categories and reproductive pattern, mating occurs at the time of
maximum gonadal activity as measured by the matura-
that the diversity of reproductive patterns is better
tion of gametes and peak levels of sex steroids. (b) In
described as a continuum rather than as consisting of
species exhibiting the dissociated pattern, mating
several discrete categories (Woolley et al. 2004). occurs at a time of minimal gonadal activity. (c) In
Some animals, such as the Norway rat, exhibit a species exhibiting the constant (also called the oppor-
close temporal association between gonadal activity tunistic) reproductive pattern, gonadal activity is main-
and mating; specifically, gonadal growth and an tained at or near maximum levels at all times. (From
increase in circulating levels of sex steroids activate Crews 1987.)
mating behavior. This pattern of gonadal activity in
relation to mating has been termed an associated
reproductive pattern (Figure 7.16a) and has been predictable, but narrow, window of opportunity to breed,
found in most vertebrates studied to date (Crews and and a specific physical or behavioral cue triggers mating
Moore 2005). behavior. Consider, for example, the red-sided garter
Some species, however, exhibit a dramatically dif- snake (Thamnophis sirtalis parietalis), a species that ranges
ferent pattern of reproduction in which mating behav- farther north than any other reptile in the Western
ior is completely uncoupled from gamete maturation and Hemisphere. The window of mating opportunity for this
secretion of sex steroids. In species that exhibit the dis- snake is from one to four weeks, and courtship behavior
sociated reproductive pattern (Figure 7.16b), gonadal of adult male garter snakes is activated by an increase in
activity occurs only after all breeding activity for the cur- ambient temperature following winter dormancy, rather
rent season has ceased, and gametes are thus produced than by a surge in testicular hormones (reviewed by
and stored for the next breeding season. Gonadal hor- Crews 1983; Crews and Moore 2005).
mones may not play any role in the activation of sexual Garter snakes in western Canada emerge in early
behavior in species that display the dissociated pattern. spring from subterranean limestone caverns, where they
Typically, species with a dissociated reproductive have hibernated in aggregations of up to 10,000 indi-
pattern inhabit harsh environments in which there is a viduals. Males emerge first, en masse, and congregate at
144 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

the den opening. Females emerge singly or in small monal control or that sex steroids do not play an orga-
groups over the next one to three weeks and mate with nizational role in the development of sexual behavior,
males that are hanging out at the entrance. Because of but rather that mating does not occur at the time of
this timing difference in the emergence of males and maximum gonadal activity.
females, males greatly outnumber females at the den Like the male, the female red-sided garter snake
opening, sometimes 500 to 1 (Figure 7.17). In view of mates when her gonads are small, gametes immature,
these odds, it is not unusual for a writhing mass of and circulating levels of sex steroids low. In the case of
snakes, called a “mating ball,” to form, in which over 100 the female, however, changes in sexual attractivity and
males attempt to mate with a single female. Against all receptivity are mediated by physiological changes that
odds, females usually mate with only one male and occur as a consequence of mating. Thus, although both
immediately disperse to summer feeding grounds, where male and female red-sided garter snakes display a disso-
they give birth to live young in August. Males, on the ciated pattern of reproduction, they differ in the type of
other hand, remain at the den opening and move to feed- stimulus that triggers breeding behavior. Whereas a
ing grounds only after all the females have emerged. change in ambient temperature triggers courtship behav-
Testicular activity is minimal in male garter snakes ior in males, stimuli associated with mating appear to
during the period of emerging and mating. In fact, it is activate physiological changes in females.
five to ten weeks later, after the males have left the den The third type of reproductive tactic, described by
site and will no longer court females, that the testes grow Crews (1987) as the constant or opportunistic repro-
and androgen levels increase. Sperm produced during ductive pattern (Figure 7.16c), is characteristic of species
this time is stored for use during the next spring. that inhabit harsh environments, such as certain deserts,
Male red-sided garter snakes use environmental where suitable breeding conditions occur suddenly and
cues instead of circulating levels of sex hormones to unpredictably. In the case of desert-dwelling animals,
determine the appropriate season for mating. reproduction is often initiated by rainfall. While waiting
Numerous experiments, utilizing castration and for suitable circumstances in which to breed, these species
replacement techniques and destruction of either the maintain large gonads, mature gametes, and high circu-
temperature-sensing areas of the brain or the pineal lating levels of sex steroids for prolonged periods of time.
gland, have revealed that, rather than relying on surges In Chapter 6 and earlier in this chapter we intro-
of sex steroids, the neural mechanisms that activate sex- duced you to the singing behavior and underlying
ual behavior in male garter snakes are triggered by a changes in the brain of male zebra finches, Taeniopygia
shift in temperature (Crews et al. 1988; Krohmer and guttata. Here we focus on the reproductive behavior and
Crews 1987). More specifically, it is the increase in physiology of male and female zebra finches, with an
temperature in the spring that follows a long period of emphasis on life history. Zebra finches live in the deserts
cold temperatures and dormancy. This is not to say that of Australia, where rainfall occurs rarely and unpre-
courtship behavior is completely independent of hor- dictably. Through droughts that can last for years, males

FIGURE 7.17 Male red-sided garter

snakes wait at den openings for emerg-
ing females. Dense mating aggregations
form as females emerge singly or in
small groups. The activation of sexual
behavior in this species is independent
of sex steroids.
 Interactions Between Hormones, Behavior, and the Environment 145

and females maintain their reproductive systems in a Teresa Dzieweczynski and colleagues hypothesized that
constant state of readiness. No matter how long the wait, social environment (in this case, presence or absence of
each sex is poised, prepared for the opportunity to breed an audience) and territory status (here, presence or
(Serventy 1971). The connection between reproduction absence of a nest) would influence aggressive behavior
and rainfall is based on food; the rains produce flushes and hormone levels in males of this species
of grass seeds that the adults feed to their young. (Dzieweczynski et al. 2005, 2006). The experimental
Courtship among males and females in desert pop- setup consisted of three abutting tanks that were sepa-
ulations begins shortly after the rain starts to fall; copu- rated by opaque partitions prior to testing; the two males
lation occurs within hours, and nest building can begin to be scored for aggressive behavior were each placed in
as early as the next day. To maintain this accelerated a tank (Figure 7.18). For the audience conditions, either
pace, both males and females carry material to the nest. a male or a female was confined in a small transparent
It is interesting that in more humid areas of their range, container in the third tank, or the container in the third
where the reproductive needs are not so immediate, tank was left empty. Once all fish were in their respec-
males and females exhibit the division of labor charac- tive tanks, they were given 24 hours to adjust to their
teristic of finches—that is, the male alone carries grass new surroundings. At the time of testing the next day,
to the nest, and the female waits at the nest for each first the opaque partition that separated the two test fish
delivery and arranges the new material as it arrives from the audience fish was removed and test fish were
(Immelmann 1963). Consistent with these early obser- given five minutes to view either the audience or the
vations of behavioral differences in zebra finches living empty container in the third tank. Then, the opaque par-
in climatically different environments are more recent tition between the two test fish was removed and their
data showing that degree of breeding readiness—as mea- behavior scored for 20 minutes. In a subset of the males,
sured by size of testes in males and ovarian follicles in the authors also measured levels of 11-ketotestosterone
females—also varies with habitat predictability. Zebra (11 KT), a major androgen in fish known to mediate
finches living in arid rangelands of central Australia, a aggressive behavior. This hormone was extracted from
habitat with highly unpredictable rainfall patterns, water into which each test fish was moved and housed
maintain higher levels of reproductive readiness than do for two hours after the behavioral test was completed.
those living in a seasonal, more predictable habitat in (This noninvasive method of hormone collection is
southeastern Australia (Perfito et al. 2007). This com- another example of the efforts being made to measure
parison of populations of the same species occupying hormones without stressing or killing the test subjects.)
habitats that differ in predictability and harshness pro- Aggressive behavior and levels of 11KT were influ-
vides strong evidence of the close tie between environ- enced by male territory status and audience. When
mental conditions and the reproductive behavior and neither test male had a nest, they were less aggressive
physiology of a species.

ADJUSTING TO ONLOOKERS
IN THE SOCIAL ENVIRONMENT Nest
Like the physical environment, the social environment of Audience
an animal can influence hormone–behavior relationships. female
We will consider how the behavior and hormone levels of
individuals can change when conspecifics are watching.
Male Siamese fighting fish (Betta splendens) are spec- Test male
9 cm
tacularly colorful creatures with long flowing fins and a OR
propensity for building nests at the water surface. These
nests, made of mucus-covered bubbles blown by the
males, attract females and serve as home for eggs and
newly hatched fry. But don’t let the beauty and home-
Audience
making tendencies of male Siamese fighting fish fool male
you, for as the common name of this fish suggests, they Test male
Nest
are also known for their pugnacity. During aggressive
contests, males flare their gills, beat their tails, and bite
one another. Under natural conditions, males defend ter-
ritories centered on their nest, and such territories may FIGURE 7.18 Experimental setup for testing the response
be closely spaced in the environment. of male Siamese fighting fish to presence and type (male
Given the apparent importance of the nest and or female) of audience. (Modified from Dzieweczynski et
neighbors in the natural history of Siamese fighting fish, al. 2005.)
 146 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

0.40 duties, including providing food and protecting the

0.35 young. Helping has been reported in over 200 species of
birds and more than 100 species of mammals.
11KT concentration (ng)

0.30 One of the favorite species to study is the Florida

0.25 scrub jay (Aphelocoma coerulescens), a bird that lives in
south-central Florida, usually in the dry oak scrub.
0.20 Although a group can range from two to six members,
0.15 it usually has three. As we will discuss in Chapter 19, the
helpers are offspring of the breeding pair from a previ-
0.10 ous year who remain on the territory and help raise their
0.05 siblings. Because the siblings share some of the helpers’
genes, the helpers manage to get some of their genes into
0.00
the population, even though the lack of a territory pre-
ts p ience,

no n nce,

rese e,

no n nce,
rese e,

no n nce,

nt

ests

nt

ests
nt

ests

vents them from breeding. In this way, the helpers make

c
c

rese

nes dien
nes audien

ie

die
ie

the best of a bad situation. Thus, we can see the evolu-

d

aud
aud

ts p
ts p

e au

e au
a

tionary causes of helping. But what about the physio-

No

No
ale

ale

nes

Mal

Mal

logical basis of helping behavior?

Fem

Fem

Treatment
An obvious first question to ask is whether helpers
help because they are physiologically incapable of
FIGURE 7.19 In male Siamese fighting fish, levels of breeding themselves. Ronald Mumme and colleagues
11-ketotestosterone (11KT) are influenced by presence
have tested this hypothesis in a population of Florida
of a nest and audience. Generally, levels of 11KT were
lower in males with nests and when a female audience
scrub jays at the Archbold Biological Station. Their data
was present. A similar pattern was found for aggressive suggest that there are no physiological reasons that cause
behavior. (From Dzieweczynski et al. 2006.) helpers to delay reproduction (Schoech et al. 1996;
Schoech 1998). Helpers do produce the hormones
important to reproduction. In both males and females,
when a female audience or no audience was present than
breeders and nonbreeders have the same levels of
when a male audience was present. In addition, in the
luteinizing hormone (LH), the hormone from the ante-
presence of a male audience, males were less aggressive
rior pituitary gland that stimulates the growth and devel-
when both had nests as compared to when neither had
opment of the ovaries and testes at the start of each
a nest. The results for hormone levels generally paral-
breeding season. Although male breeders have somewhat
leled those for behavior; levels of 11KT were lower in
higher levels of testosterone than do male helpers, the
test males with nests as compared to those without nests
seasonal pattern of testosterone production is essentially
and lower with a female audience (Figure 7.19). We see,
the same. In females, breeders and helpers have the same
then, that in Siamese fighting fish, levels of aggressive
level of estradiol, their primary sex steroid. However, the
behavior during male–male encounters and the mediat-
seasonal pattern of estradiol production is different in
ing androgen 11KT are influenced by nesting status of
female breeders and female helpers. Nevertheless, these
the male and who, if anyone, is watching.
observations reveal that the testes and ovaries of helpers
are functional, at least for hormone production.
Another hypothesis is that helpers might be physi-
A DETAILED LOOK ologically capable of breeding, but that they delay
AT THE HORMONAL BASIS reproduction because they are unable to gather enough
OF SELECTED BEHAVIORS food in order to breed successfully. This, too, does not
appear to be the case. Helpers are indeed slightly
As you have undoubtedly noticed while reading this
smaller than breeders, but this probably is not because
chapter, much of the work in the field of behavioral
they are less successful foragers, but because they are
endocrinology focuses on the hormonal bases of sexual
younger. When the size difference is taken into account,
behavior and aggressive behavior. However, hormones
the weights of male helpers and breeders are equivalent.
also influence many other types of behavior. Here we
Female helpers are apparently able to forage as well as
take an in-depth look at the hormonal bases of three
female breeders during the winter months because they
behaviors that we have not yet considered: helping at the
weigh about the same at the start of the breeding sea-
nest, scent-marking, and migrating.
son. If the helpers delay breeding because they cannot
gather enough food, we would expect that supplying
HELPING AT THE NEST
supplemental food to the population would allow more
Helpers—nonbreeding animals that assist the breeding helpers than usual to switch to being breeders. When
pair in rearing their young—have assorted parental this was tried, however, the additional food did not
 A Detailed Look at the Hormonal Basis of Selected Behaviors 147

increase the number of helpers that became breeders. not help. It is interesting that there is no correlation
Failure of food supplementation to prompt helpers to between the prolactin levels of breeders and their feed-
become breeders is even more striking, given recent ing score. Prolactin is known to affect many aspects of
data indicating that food supplementation has positive parental behavior besides feeding the young. A rela-
effects on reproduction in scrub jays: breeders in sup- tionship between prolactin and parental behavior among
plemented groups of jays initiated clutches earlier than breeders may not be seen in these data because parental
did those in control groups that were not supplemented behaviors other than feeding were not measured
(Schoech et al. 2004). (Schoech et al. 1996).
Another hypothesis that might explain why helpers
delay reproduction is stress: perhaps the presence of the
SCENT–MARKING
dominant breeding pair stimulates the helper’s adrenal
gland to produce the stress hormone corticosterone, Scent-marking is the act of strategically placing a chem-
which is known to suppress the production of the hor- ical mark in the environment; many mammals apply
mones needed for reproduction. This doesn’t seem to urine, feces, or secretions from special scent glands.
occur. Helpers and breeders have equivalent amounts of (Recall that scent glands are exocrine glands and, as
corticosterone throughout the breeding season. such, differ from endocrine glands in having ducts that
Thus, it appears that helpers are physiologically release their products to the body surface. Scent glands
capable of reproduction, but simply lack the opportunity are found in many locations on the body and, depend-
to do so. The next proximate question about helping ing on the species, can occur in such locations as
behavior concerns the role of hormones in initiating between the digits, on the legs, chest, or belly, on the
parental behaviors, such as feeding the nestlings. head, or in the anal canal. Many mammals have scent
Parental behaviors may be caused by the pituitary hor- glands at multiple locations.) Scent marks likely convey
mone prolactin in both helpers and breeders. Prolactin information about individual identification, age, and
production is stimulated by cues from the nest, eggs, and reproductive state, and function to establish and main-
nestlings. For this reason, prolactin levels increase tain territories and breeding relationships (see Chapter
throughout the breeding season. Birds that spend the 16 for more information on scent-marking). Our next
most time caring for the eggs and young produce the example concerns a behavior that probably everyone has
most prolactin. In general, females produce more pro- observed—urine-marking by domestic dogs. Urine-
lactin than males, and breeders of either sex produce marking in dogs is distinguished from simple elimina-
more than helpers. Prolactin levels are lower in helpers tion by the fact that urine is directed at a specific object
because breeders won’t allow them near the nest until the in the environment, such as a tuft of long grass or a fire
young have hatched. But both breeders and helpers feed hydrant.
the young. There is a direct relationship between a The urinary behavior of domestic dogs (Canis lupus;
helper’s level of prolactin and the feeding score (a mea- yes, dogs now have the same scientific name as wolves
sure of how much a bird fed the nestlings) (Figure 7.20). and no longer go by Canis familiaris) is sexually dimor-
Notice that some helpers didn’t actually help; they have phic. Adult males urinate more frequently than do adult
feeding scores of zero. The prolactin levels of the helpers females and are more likely to direct their urine at
that did help are much higher than those of birds that did objects in the environment (i.e., to urine-mark). Even
urinary posture is sexually dimorphic; whereas males lift
a leg to urinate, females typically squat. Sex differences
100
in urinary posture are organized by sex steroids (testos-
Breeders
Non-breeders terone) around the time of birth (Beach 1974; Ranson
80 Combined and Beach 1985). Testosterone is not, however, required
to activate the leg-lifting posture in adulthood. As any
Prolactin (ng/ml)

60 owner of a castrated dog can attest to, even though the

source of testosterone has been removed, a neutered
40 male still lifts his leg to urinate. We see from this exam-
ple that whereas some sexually dimorphic patterns of
20 behavior, such as sexual behavior in Norway rats, are
organized and activated by sex steroids, others are sim-
0 ply organized by these hormones.
0 2 4 6 8 10 12 As we have said several times in this chapter, things
Feeding score are often more complicated than they first appear, and
FIGURE 7.20 In Florida scrub jays, there is a direct rela- further research often reveals new details about partic-
tionship between how much a non-breeder feeds nestlings ular behaviors and their hormonal control. Recent stud-
and its prolactin levels. (Data from Schoech et al. 1996.) ies with female Jack Russell terriers have shown that
148 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

marking with urine is more common than previously tain stimuli or perhaps influencing the nervous system
described for female dogs (about 60% of urinations by or muscles. We have also seen that the relationship
females are directed at objects in the environment) and between hormones and behavior is dynamic, allowing
that not all female dogs squat (Wirant and McGuire animals to adjust to their physical and social environ-
2004). Indeed, there is considerable variation among ments. We end the chapter by considering hormone-
females in the postures used while urinating, with some induced changes in the behavior, physiology, and
females displaying the traditional squatting posture, morphology of the Atlantic salmon (Salmo salar), a
while others use squat-raises (one leg is raised while in species whose complex and fascinating life history cen-
a squatting posture) and even handstands (Figure 7.21). ters on making repeated movements between freshwater
Furthermore, the propensity to mark with urine varies and saltwater. We begin with a brief summary of the
across the estrous cycle, being most common just before species’ life cycle.
and during estrus. These data suggest that female dogs Adult Atlantic salmon leave the ocean and return to
mark with urine to convey information about their their natal stream to spawn. Spawning typically occurs
reproductive state (Wirant et al. 2007). Whether these in the fall, and the adults bury thousands of large, fer-
patterns of urinary behavior in female Jack Russell ter- tilized eggs in a gravel depression called a redd. In the
riers will generalize to other breeds of dogs or mixed spring, the eggs hatch into yolk-sac larvae (alevins) that
breeds remains to be seen. The precise physiological remain in the nest, living off their yolk for about six
bases for changes across the estrous cycle and for varia- weeks. Eventually, when the yolk runs out, young salmon
tion among females in urinary behavior also remain to emerge from the nest as fry and begin to feed indepen-
be determined. dently. As time passes and the salmon reach about 5 cm
in total length, they develop vertical marks on the sides
of their bodies. Salmon at this stage are known as parr,
MIGRATING
and the marks, called parr marks, function as camouflage.
We have seen that hormones influence behavior in When parr reach about 15 cm in length and environ-
diverse ways, sometimes altering a preference for cer- mental conditions are right, they undergo a process

a b

FIGURE 7.21 Hormonal control of urinary behavior has

been well studied in male dogs but less so in female
dogs. Contrary to early descriptions that female dogs
typically squat to urinate, recent observations indicate
that female Jack Russell terriers exhibit diverse urinary
postures. Shown here are the (a) squat, (b) squat-raise,
and (c) handstand.
 A Detailed Look at the Hormonal Basis of Selected Behaviors 149

called smoltification after which they are known as by the reflective silver color so characteristic of fishes
smolts. Smoltification is a critical developmental process evading predators in pelagic environments. Even the
that prepares young salmon for migrating to the ocean way young salmon position themselves in the current
and living there for one or more years before returning changes, and this undoubtedly involves adjustments in
to their natal stream as adults to spawn and begin the their nervous and sensory systems. Parr face into the
cycle anew. Because Atlantic salmon are iteroparous current (a behavior called positive rheotaxis), and this
(reproducing repeatedly rather than dying after spawn- position allows them to see food coming their way from
ing as is characteristic of Pacific salmon), they may make upstream. Smolts, on the other hand, face downstream
multiple forays between freshwater and saltwater over (negative rheotaxis), an orientation necessary for their
the course of their lives. Our focus will be on the hor- impending migration to the sea. We also see changes in
mone-mediated changes associated with the first trip social behavior. Whereas parr aggressively defend indi-
from their natal stream to the sea. vidual feeding territories, smolts exhibit decreased ter-
Dramatic changes in physiology, morphology, and ritorial and agonistic behavior and eventually form
behavior occur during the developmental change from schools.
parr to smolt. The osmoregulatory system of parr is set What environmental factors initiate and which hor-
up for life in freshwater and must change for life in salt- mones mediate the changes associated with becoming a
water, so we see changes in the cells of the gills, gut, and smolt and preparing for life at sea? Although we do not
kidneys in preparation for seaward migration and life. have all of the answers yet, we are beginning to under-
Also during smoltification, the parr marks that func- stand key aspects of this transformation (Figure 7.22).
tioned as camouflage for young bottom-dwelling Environmental factors, such as high water flow and
salmon in the shallow stream environment are replaced increasing photoperiod and temperature, appear to

Egg

Yolk-sac larva (alevin)

Adult male

Fry
Adult female

Parr
Smolt

Smoltification is triggered by:

• Growth
• Environment
• Hormones

FIGURE 7.22 Environmental cues (high water and increasing temperature and day length) prompt the developmental
change from parr to smolt in Atlantic salmon that have reached a critical body length. Several hormones mediate the
necessary changes in physiology, morphology, and behavior associated with the move from freshwater to saltwater.
150 Chapter 7 / Physiological Analysis of Behavior—The Endocrine System

prompt smoltification in fish of sufficient size cell boundaries and bind to receptors inside the cell. The
(McCormick et al. 1998). Several hormones have also hormone-receptor complex then enters the nucleus and
been identified as important in mediating the necessary turns on certain genes. Recent evidence indicates that
physiological and morphological changes. Whereas pro- steroid hormones also interact with membrane recep-
lactin promotes osmoregulation in freshwater and tors. This pathway produces more rapid changes in
decreases during smoltification, growth hormone behavior than is possible in the pathway involving gene
increases and works antagonistically to increase toler- expression.
ance for saltwater (Sakamoto and McCormick 2006). The mechanisms by which hormones influence
Cortisol appears to interact with both hormones to pro- behavior include alterations in (1) sensation or percep-
mote acclimation to a particular environment tion, (2) development and activity of the central nervous
(McCormick and Bradshaw 2006). Thyroid hormones system, or (3) muscles responsible for the execution of
are responsible for the replacement of parr marks by sil- behavior.
ver coloration (Hoar 1988). The hormones involved in Traditionally, the effects of steroid hormones on
many of the behavioral changes of smoltification have behavior have been divided into organizational and acti-
yet to be identified. vational effects. Organizational effects occur early in life
This example shows the intricate interactions of the and tend to be permanent. In contrast, activational
developing organism and its external environment, and effects occur in adulthood and tend to be transient, last-
how hormones help to orchestrate adjustments in behav- ing only as long as the hormone is present in relatively
ior, physiology, and morphology in different environ- high concentrations. In activational effects, steroids serve
ments. Development is the focus of our next chapter. only to activate existing neural pathways responsible for
a specific behavior rather than to organize neural path-
ways. Though still useful, the traditional distinction
SUMMARY between organizational and activational effects has been
questioned because of the lack of biochemical, anatom-
Animals have two closely associated, yet different, sys- ical, and physiological evidence for two fundamentally
tems of internal communication: the nervous system and different ways in which steroids produce their effects.
the endocrine system. Typically, transfer of information Hormonal effects on behavior can be studied by the
occurs more slowly by the endocrine system than by the removal of the gland and hormone replacement (so-
nervous system, and the effects produced are more gen- called interventional studies) or by correlational studies.
eral and long lasting. Whereas neural information is In the latter, researchers look for changes in behavior
transmitted via a series of electrical events, communica- that parallel fluctuations in hormone levels.
tion by the endocrine system occurs through hormones, The interaction between hormones and behavior is
chemical substances that are secreted by either endocrine a dynamic one. In some instances, hormones initiate
glands or neurons. changes in behavior; in others, behavior causes changes
Hormones and neurohormones produce changes at in the levels of circulating hormones. Hormones can
the cellular level that ultimately influence behavior. also rapidly suppress a behavior when appropriate. The
Peptide hormones are water-soluble amino acid chains interactions between hormones and behavior are sen-
that bind to receptors at the cell surface, which activates sitive to aspects of the physical and social environment.
a cascade of chemical reactions within the cell. In con- Behaviors mediated by hormones include aggression,
trast, steroid hormones are derived from cholesterol, and courtship, mating, caring for young, scent-marking,
because they are fat-soluble, they can move through the and migrating.
 8
The Development of Behavior

Influences on Behavioral Development A mallard duckling (Anas platyrhynchos), nestled inside its
Development of the Nervous System egg, hears the muffled voice of its mother overhead. After
Development of Nonneural Structures spending approximately four weeks in the egg, the
youngster pecks through the shell and frees itself (Figure
Hormonal Milieu
8.1). It then spends one more day beneath its mother.
Physical Characteristics of the Environment
Should a predator, sometime during the first day
Experience through Play posthatching, wander into the area around the nest, the
The Concept of Sensitive Periods duckling, in concert with its eight or so siblings, responds
Changing Terminology—From Critical Periods to rapidly to the alarm call of its mother by “freezing”—
Sensitive Periods crouching low and ceasing all movement and vocalization
Timing of Sensitive Periods (Miller 1980). If the duckling goes unnoticed by the
Multiple Sensitive Periods
predator, the very next morning it responds promptly to
yet another of its mother’s calls, this time the assembly
Some Examples of Sensitive Periods in Behavioral
Development call, by following her from the nest through the long
grass to a nearby pond (Miller and Gottlieb 1978). Here,
Pulling It All Together—The Development the duckling will string along behind its mother and sib-
of Bird Song lings for some time to come. As the weeks pass and the
Genetic, Hormonal, and Neural Control of Song young bird continues to associate with family members,
Role of Learning in Song Development it learns the characteristics of an appropriate mate (Schutz
Sensitive Periods in Song Learning 1965). This information, though obtained early in life,
Own-species Bias in Song Learning will not come in handy until the first breeding season.
Indeed, the duckling learns, soon after hatching, many of
Social Factors and Song Development
the things it needs to survive and reproduce. We see,
A Diversity of Song Learning Strategies
then, that the changes that occur during behavioral devel-
Developmental Homeostasis opment may contribute to fitness immediately (as in the
Rehabilitation of Chimpanzees after Long-Term duckling’s freeze response to its mother’s alarm call) and
Isolation in adulthood (as in mate preference).

151
152 Chapter 8 / The Development of Behavior

DEVELOPMENT OF
THE NERVOUS SYSTEM
Behavior is primarily controlled by the nervous system,
and so is intimately linked with the development of this
system. This is especially obvious early in an animal’s
life when development of the nervous system is rapid
and dramatic. Consider, for example, the neural and
behavioral development of embryonic Atlantic salmon,
Salmo salar (Abu-Gideiri 1966; Huntingford 1986).
Stages in the development of this fish are depicted in
Figure 8.2.
The first movements of the embryo are seen in the
feeble twitches of the heart, soon followed by move-
FIGURE 8.1 Newly hatched mallard ducklings. ments in the dorsal musculature. Interestingly, these
Experiences prior to and soon after hatching will movements begin before the nervous system has formed
profoundly influence each duckling’s behavioral and are thus myogenic, or muscular, in origin. The
development. impulse begins in the muscle itself. Approximately
halfway through embryonic life, the major motor sys-
tems appear in the spinal cord. A short time later, the
Several questions arise from this brief description of motor neurons make contact with anterior muscles, giv-
the early behavioral development of a mallard duck. ing the embryo the ability to flex its body. Soon, with the
How does the genetic makeup of the duckling interact development of neural connections at different points
with its internal and external environment to produce and on both sides of the body, the embryo displays the
such behavior? How do the nervous and endocrine sys- first undulating movements associated with swimming.
tems influence behavioral development? What role do Development of the sensory system of the trunk and its
visual, auditory, or social stimuli play in the development connection to the skin occurs a short time later; after
of freezing, following, and sexual behaviors? Are expe- this, the embryo can move in response to tactile stimu-
riences prior to hatching important to the development lation. Finally, the neural circuits that underlie both fin
of posthatching behavior? What happens when a behav- and jaw movements are complete, allowing independent
ior, such as the following response, ceases to be a part of and coordinated movement of these structures. Neural
the individual’s behavioral repertoire? If we look beyond and behavioral development continues (in fact, the
the single duckling in our example and consider the young salmon has not yet even hatched). We can see
species as a whole, does behavioral development always from this example that development of key parts of the
proceed in a predictable and reliable fashion? In this nervous system underlies the appearance of new patterns
chapter we address these questions about the behavioral of behavior.
development (change in behavior over time) of mallard When a behavior disappears from an animal’s reper-
ducks and other animals. Recall that Niko Tinbergen toire, does this mean that the underlying neural circuits
(1963) considered questions about development to be have also disappeared? In some cases, the neural circuits
one of the four main types of questions that should be are dismantled or permanently altered in concert with
asked about behavior. cessation of the behavior. As an example of this situation,
we will consider changes that occur in the tobacco horn-
worm (Manduca sexta) during complete metamorphosis,
INFLUENCES ON BEHAVIORAL when the insect changes from caterpillar to pupa to
DEVELOPMENT moth. Much of this work has been carried out in the lab-
oratory of Janis Weeks (Hazelett and Weeks 2005;
Patterns of behavior come and go throughout devel- Weeks et al. 1989; Weeks 2003).
opment. A behavior may appear in an animal’s reper- Complete metamorphosis entails dramatic changes,
toire, only to disappear or change shortly thereafter. not only in the animal’s morphology, but also in its
Here we consider some of the factors that influence the behavioral repertoire (during incomplete metamorpho-
development of behavior. Keep in mind, however, that sis, as occurs in insects such as grasshoppers, juveniles
these factors are not mutually exclusive and that they look much like adults). During the remarkable trans-
likely interact with one another throughout develop- formation of complete metamorphosis, the nervous sys-
ment. We will not focus on the role of genes in behav- tem of the tobacco hornworm must sequentially control
ioral development because this topic is covered in three very different stages: the larva or caterpillar, the
Chapter 3. pupa, and the adult moth (Figure 8.3). The animal is
 Influences on Behavioral Development 153

FIGURE 8.2 Behavioral development in the

Atlantic salmon. Patterns of behavior
Nip at conspecifics
emerge in parallel with the development
Snap at food of neural structures necessary for their
Upward
performance. (Modified from Huntingford
jumps 1986; drawn from the data of Abu-Gideiri
Tail flick to mechanical stimuli
1966; Dill 1977.)

On side Horizontal Supported by pectoral fins

Swimming movements

Move in response to light and gravity

Jaw and gill cover movements

Fin movements

Move away from tactile stimuli

Body flexion

Dorsal muscle contractions

Heart beats

0 10 20 30 40 50 60 70 80 90 100 10 20 30 40

Hatching Emergence
Embryonic life (%) Days after hatching

a b

FIGURE 8.3 Metamorphosis in the tobacco hornworm,

Manduca sexta: (a) larva or caterpillar, (b) pupa, and
(c) adult.
154 Chapter 8 / The Development of Behavior

transformed from a crawling, eating machine to a fly- a Head

ing, reproducing machine. Although some patterns of
behavior are exhibited in all three stages (e.g., behavior
b Larval muscles of prolegs
associated with shedding of the cuticle), many behaviors
are restricted to a single stage (e.g., crawling in the larva
and flight in the adult). Much of the neural circuitry
controlling stage-specific patterns of behavior is assem-
bled and dismantled during development. Let’s see what Changes in structure of motor neurons
happens when crawling behavior is lost at the larval–pupal c
transformation.
Caterpillars of the tobacco hornworm have abdom-
inal prolegs, stumplike appendages that are not devel-
opmentally related to the legs of the adult moth (Figure
8.4a). The prolegs act in simple withdrawal reflexes, as Larva Pupa
well as in more complex behaviors such as crawling and d
helping the animal grasp the substrate. Although these

Hormone titer
behaviors are important to the caterpillar, they are not
to the pupa; and the proleg behaviors gradually disap-
pear during the larval–pupal transformation. The ques-
tion is, then, what causes their disappearance?
While we often associate the formation of new neu-
rons with behavioral development, sometimes the death Time (days)
of old neurons is key to changes in behavior over time.
Most proleg movements are accomplished by retractor
muscles (Figure 8.4b) that are innervated by motor neu- Larva Pupa Adult
rons with densely branching arbors, or dendrites.
During the larval–pupal transformation, substantial FIGURE 8.4 During metamorphosis in the tobacco horn-
regression of the dendrites of many of the motor neu- worm, modifications to the nervous system (induced by
changing levels of hormones) eliminate patterns of
rons occurs (Figure 8.4c). These motor neurons die, and
behavior. (a) The abdominal prolegs of the caterpillar
their associated proleg muscles degenerate and become
are involved in withdrawal reflexes, crawling, and grasp-
nonfunctional. Specifically, the proleg muscles and ing the substrate. These behaviors are important to the
motor neurons in abdominal segments 1, 4, 5, and 6 larva but disappear from the animal’s repertoire once it
degenerate; we will focus on what causes the death of the reaches the pupal stage. (b) Cutaway view showing the
motor neurons and muscles in these segments. retractor muscles of the prolegs. (c) The dendrites of
The demise of the proleg neuromuscular system, the motor neurons that innervate the retractor muscles
and hence proleg behaviors, is prompted by a peak in regress during the larval–pupal transformation. (d) The
ecdysteroid hormones secreted by the prothoracic peak in ecdysteroid hormones just before transforma-
gland at the back of the head. This peak occurs just tion to the pupa triggers dendritic regression. (Modified
before the transition to the pupal stage (Figure 8.4d). At from Weeks et al. 1989; data from Bollenbacher et al.
1981; Weeks and Truman 1984.)
this time, high levels of ecdysteroids trigger regression
of the dendrites of the motor neurons that innervate the
proleg muscles. As a result, the motor neurons are
removed from behavioral circuits, and proleg behaviors method was an interesting one. Rather than asking the
are lost in the pupa. age-old question, “Which came first, the chicken or the
Does the underlying neural circuitry always disap- egg?” they asked, “What happens when you put a
pear when a behavior is lost from an animal’s repertoire? chicken back into an egg?” Bekoff and Kauer placed
Not always. Chickens (Gallus gallus domesticus) typically posthatching chicks up to 61 days of age (chicks at this
hatch over a 45- to 90-minute period at the end of incu- age have molted their fuzzy down, are fully feathered,
bation. During hatching, the chick escapes from the con- and basically resemble small chickens) in artificial glass
fines of its shell through a highly stereotyped series of eggs and recorded their behavior and muscle move-
movements, rotating its upper body and thrusting its ments. Each chick was gently folded into the hatching
head and legs. Because these patterns of behavior asso- position and placed into a ventilated glass egg of the
ciated with hatching later disappear from the repertoire appropriate size. Within two minutes of being placed in
of chickens, Anne Bekoff and Julie Kauer (1984) became the artificial eggs, chicks of all ages began to produce a
interested in the fate of the neural circuitry underlying, behavior that qualitatively and quantitatively resembled
in particular, the leg movements of hatching. Their that of normal hatching. Rather than being dismantled
 Influences on Behavioral Development 155

or permanently altered after hatching, the neural cir- adult pattern of indiscriminate filter feeding. Here, then,
cuitry for the leg movements of hatching clearly remains we have an example of how changes in behavior are
functional in older chickens. Additional studies suggest coordinated with the development of specific morpho-
that a basic neural circuit for leg movements is built early logical structures.
in embryonic development and that it is modified to pro-
duce movements associated with hatching and later
walking (Bekoff 1992). Thus, we see that in some cases HORMONAL MILIEU
the disappearance of a behavior is not associated with the We have seen that hormones prompt regression in parts
complete dismantling of its underlying neural circuitry; of the neural circuitry of the tobacco hornworm cater-
instead, the circuitry is modified to serve other patterns pillar, ultimately resulting in loss of crawling behaviors
of behavior. in the pupa. Now we’ll consider two additional examples
of how developing animals are influenced by the hor-
monal milieu (environment) they experience early in life.
DEVELOPMENT OF
Both examples relate to the effects of exposure to andro-
NONNEURAL STRUCTURES gens, such as testosterone, during the prenatal period;
Sometimes behavioral change is driven by morphologi- the first example concerns house mice and the second,
cal changes that are not neural. Obviously, a behavior black-headed gulls.
can’t be performed unless the animal has developed the Like peas in a pod, house mouse fetuses (Mus mus-
morphological structures necessary for its performance. culus) line the uterine horns of their mother (Figure 8.6).
Consider the changes in feeding behavior that occur in Each fetus has its own personal placenta (vascular con-
the paddlefish Polyodon spathula. This fish, from the nection to the mother) and floats within a fluid-filled
Mississippi and Ohio River drainages of North America, compartment called the amniotic sac. Even before birth,
has a bizarre, paddle-shaped snout, which can be almost the endocrine glands of these tiny individuals are pro-
half the length of the body. Although larval paddlefish ducing hormones that may permanently alter not only
feed by chasing and selectively plucking individual zoo-
plankton from the water, adult paddlefish are indiscrim-
inate filter feeders, dropping their lower jaw and Ovary
consuming all material strained from the water as they
plow through their environment (Figure 8.5). Changes Oviduct
in the feeding behavior of paddlefish parallel the devel-
opment of gill rakers (Rosen and Hales 1981). These 0M 1M
bony structures, comb-like in appearance, project from
the gill arches into the oral cavity and strain food parti-
cles from the water. Absent from larval paddlefish, gill
rakers begin to develop when young fish are about 100 2M
1M
mm long (about 4 in.). Gill rakers are well-developed
when the young paddlers reach 300 mm in length, and
it is at this stage that feeding behavior takes on the fully
Amniotic Uterine
sac Cervix horn
Placenta

Vagina

FIGURE 8.6 Mouse fetuses line the uterine horns of a

pregnant female. Because the fetuses are in such close
quarters, hormones from one fetus can influence behav-
ioral development of contiguous fetuses. Female fetuses
can occupy the following three positions relative to
FIGURE 8.5 A paddlefish exhibiting the adult feeding male fetuses: 2M, between two males; 1M, next to one
pattern, filter feeding. Rather than indiscriminate filter male; 0M, not next to a male. The 2M females differ
feeding, larval paddlefish selectively pick zooplankton substantially from 0M females in their adult behavior
out of the water column. The eventual development of and physiology as a result of proximity to male fetuses
the adult mode of feeding parallels the development in the intrauterine environment. (Modified from
of gill rakers in the young fish. McLaren and Michie 1960.)
156 Chapter 8 / The Development of Behavior

their own but also their neighbor’s morphology, physi- imental birds (those hatched from first-laid eggs injected
ology, and behavior. In many fascinating experiments, with androgens and sesame oil) and control birds (those
Frederick vom Saal, John Vandenbergh, and others have hatched from first-laid eggs injected only with sesame
shown that development of mouse fetuses can be mod- oil). Such comparisons for black-headed gulls have shown
ified by exposure to hormones secreted by contiguous that experimental birds exhibit the following effects dur-
littermates (studies reviewed in Ryan and Vandenbergh ing the prefledging period (period before leaving the
2002). As shown in Figure 8.6, fetuses can occupy three nest): (1) earlier hatching, (2) enhanced growth, (3)
intrauterine positions relative to siblings of the opposite enhanced begging for food, and (4) suppressed immune
sex. Females can be positioned between two male fetuses function. Maintaining immune function is energetically
(2M females), next to one male (1M females), or not next expensive; the latter effect is therefore consistent with the
to a male (0M females). But what does intrauterine posi- overall pattern of enhanced androgen exposure resulting
tion have to do with adult patterns of behavior? It turns in chicks that direct energy to growth and competitive-
out that by day 17 of gestation, levels of testosterone are ness rather than to immune function. Less is known
three times higher in the blood of male fetuses than in about the long-term effects of enhanced yolk androgens,
the blood of female fetuses. Even more intriguing is the but the data so far indicate that experimental birds show
finding that on this same day, 2M female fetuses (i.e., more frequent courtship and threat displays, win more
those females nestled between two male littermates) have aggressive encounters, and have lower adult survival
significantly higher concentrations of testosterone in when compared to control birds.
their blood and amniotic fluid than do female fetuses not The study of avian maternal hormones is relatively
next to males (i.e., 0M females). Apparently, hormones new, and much remains to be learned about the effects
pass through the amniotic fluid and possibly via uterine of differing levels of androgens on developing young and
blood vessels to contiguous littermates. As a result of the possible adaptive value of androgen deposition. Early
prenatal exposure to testosterone, adult 2M females dis- indications are that deposition of androgens may be a
play traits that distinguish them from 0M females. way for mothers to adjust the developmental trajectories
Specifically, 2M females (1) are less attractive to males, of their offspring to current environmental conditions.
(2) are more aggressive to female intruders, (3) mark a For example, black-headed gulls begin to incubate their
novel environment at a higher rate, (4) maintain larger eggs when the first egg is laid rather than waiting until
home ranges, (5) experience their first estrus at a later age, the clutch of three eggs is complete. As a result, the
and (6) produce fewer viable litters. These differences in chicks hatch asynchronously, with the first-laid egg
physiology and adult behavior exist despite the fact that hatching first, the second-laid egg hatching next, and the
after birth, testosterone levels do not differ between the last-laid egg hatching last. Chicks that hatch last are
two groups of female mice. Thus, behavioral differences smaller and not as strong as their older siblings. Perhaps,
in adulthood can be traced to differential exposure to hor- then, the greater deposition of androgens in last-laid
mones in the intrauterine environment. Position in the eggs helps to lessen the effects of being the last to hatch
uterus also affects the behavioral development of male by enhancing the growth and competitiveness of these
mice, but males are somewhat less sensitive than females chicks. Avian mothers also adjust androgen deposition
to intrauterine effects. from one clutch to the next. There is some evidence that
Prenatal exposure to androgens also influences the female black-headed gulls exposed to intense competi-
growth, physiology, and behavioral development of tion deposit higher levels of androgens in their eggs. It
young birds (Groothuis et al. 2005). While developing is possible, then, that these mothers are preparing their
within its egg, a bird is exposed to androgens (testos- offspring for the highly competitive environment they
terone, dihydrotestosterone, and androstenedione) may face once they are free from the confines of their
deposited into the yolk by its mother while the egg shell (Groothuis et al. 2005).
formed in her ovaries. The precise mechanism by which
the mother deposits androgens is not yet known. In
black-headed gulls (Larus ridibundus), yolk levels of testos-
terone and androstenedione increase with each egg laid. STOP AND THINK
Because first-laid eggs have the lowest androgen levels, In this section we describe the effects on behavioral devel-
they are used in experimental manipulations. Androgens opment of intrauterine position in rodents and maternal
can be injected into the yolk of first-laid eggs without hormones in bird eggs. What might be some advantages
exceeding the levels typically found in later-laid eggs, thus and disadvantages of each model system for exploring hor-
keeping experimental hormone levels within the normal mone-mediated maternal effects on developing offspring?
Also, for those of you who have already read Chapter 7,
limits experienced by the species. Injections are usually
are the effects of intrauterine position in rodents and
given at the start of incubation, with sesame oil serving
maternal hormones in bird eggs organizational or activa-
as the vehicle (substance in which the hormone is dis- tional? Explain your answer.
solved). Thus, comparisons can be made between exper-
 Influences on Behavioral Development 157

PHYSICAL CHARACTERISTICS gest that incubation temperature influences brain devel-

OF THE ENVIRONMENT opment and leads to different perceptions of individuals
of the opposite sex. Here, then, we have an example of a
Imagine a situation in which the mate choice of an adult physical characteristic of the prenatal environment influ-
animal is shaped by the environmental temperature it encing the development of adult behavior.
experienced during embryonic development. This is Physical characteristics of the postnatal environ-
precisely the case for the leopard gecko (Eublepharis ment also influence development of the brain and
macularius), a lizard with temperature-dependent sex behavior. This can be seen in the results of numerous
determination. In this species (as well as in several other studies comparing neural and behavioral development
species of lizards, many turtles, and all crocodilians), in captive animals reared either in enriched laboratory
whether an individual is a male or a female is determined environments or in standard laboratory housing con-
not by sex chromosomes, but by the temperature experi- ditions. Enriched environments typically mean larger
enced during a window of time about midway through its cages, a more complex and variable physical environ-
incubation period. More specifically, in leopard geckos, ment with nesting material, foraging devices, toys, hid-
females are typically produced at low and high incubation ing places, and the opportunity for voluntary exercise,
temperatures (26º and 34ºC, respectively), and different as well as more complex social groups (we will consider
sex ratios are produced at intermediate temperatures (for the specific effects of social isolation later in the chap-
example, mostly females at 30ºC and mostly males at ter). In rodents, environmental enrichment results in
32.5ºC). Incubation temperature, however, determines several structural changes in the brain, which include
more than an individual’s gender—it also influences the increased numbers of neurons, synapses, and dendritic
individual’s adult aggressive and sexual behavior in com- branches, and increased brain weight and size (van
parison to others of its sex. Here we will focus on the Praag et al. 2000; Würbel 2001). Rodents from
behavior of mate choice as studied by Oliver Putz and enriched environments also exhibit enhanced learning
David Crews (2006). When males reared at 30ºC were and memory. In captive nonhuman primates, individ-
placed in a Y-maze and given a choice between spending uals housed in enriched environments exhibit a more
time near a female reared at 30ºC and one reared at 34ºC, balanced repertoire of natural behaviors and display
they preferred the female from a 34ºC incubation tem- these behaviors at species-typical frequencies and
perature (Figure 8.7a). In contrast, when males reared at intensities. Primates from enriched environments also
32.5ºC were placed in the Y-maze and given a choice exhibit fewer abnormal behaviors such as self-injurious
between spending time near a female reared at 30ºC and behaviors (e.g., self-biting and head-banging) and
one reared at 34ºC, they preferred the female from a 30ºC stereotypic behaviors (repetitive behavior, such as pac-
incubation temperature (Figure 8.7b). The authors sug- ing and flipping, with no apparent biological function).

a 300 b 300

250 250

200 200
Time (sec)

Time (sec)

150 150 FIGURE 8.7 Characteristics

of the physical environment
100 100 influence behavioral develop-
50 50 ment in leopard geckos.
Shown here are the effects of
0 0 incubation temperature on
the mate preferences of adult
males, as measured by the
30.0°C Female 30.0°C Female
time that they spend near
individual females in a
34.0°C Female
34.0°C Female Y-maze. (a) Males from a
30ºC incubation temperature
prefer females from a 34ºC
incubation temperature.
(b) In contrast, males from a
32.5ºC incubation tempera-
30.0°C Male 32.5°C Male ture prefer females from a
30ºC incubation tempera-
ture. (From Crews and
Groothuis 2005.)
158 Chapter 8 / The Development of Behavior

These abnormal behaviors appear to develop in by whether the behavior is spontaneous and voluntary
response to unavoidable stress or fear and to the inac- and exhibited by healthy individuals (Burghardt 2005).
tivity, boredom, and frustration experienced in envi- Play is also difficult to define because there are
ronments lacking physical and social complexity several types. First is social play, which includes play
(Honess and Marin 2006; Lutz and Novak 2005). fighting or play chasing, as well as sexual play. We have
Indeed, environmental enrichment is now advocated as all been amused by the friendly tussles of kittens and
the preferred means to treat diverse zoo animals that puppies as they chase, wrestle, and pounce on one
have developed stereotypic behaviors in captivity another. Sexual play includes playful mounting by
(Mason et al. 2007). gazelles and precocious courtship in some turtles. A
second form is locomotor play—that is, exercise. Foals
EXPERIENCE THROUGH PLAY kick up their heels and gallop. Young primates, includ-
ing human children, may swing and roll and slide.
Play is thought to be important in the normal develop-
Polar bear cubs climb ashore only to leap back into the
ment of behavior in many mammals, as well as in some
water. Object play is the third form. In this form of
birds and even a few reptiles. Play behavior also has been
play, objects are manipulated. When first presented
described in octopuses, thus extending its occurrence to
with a novel object, a young animal typically explores
invertebrates (Kuba et al. 2006; Mather and Anderson
it by touching, sniffing, or viewing it from different
1999). We know play when we do it and when we see
angles. After the initial sensory investigation, the
other animals doing it (Figure 8.8). During a visit to the
object may become a toy (Fagan 1981). Sometimes a
zoo, we even recognize play in species we have never
young animal will flit among the types of play in rapid
seen before. The attribute that pervades all play, and by
succession, and the predominant form of play may
which we most commonly identify it, is our subjective
change as the animal matures (Burghardt 1998).
judgment of a lack of purposefulness.
But why do animals play? In other words, what func-
Although play is easy to spot, it is difficult to define,
tion does this frolicking serve? What role does play have
mainly because no specific behavior pattern or series of
in the development of behavior? The hypotheses for the
activities exclusively characterizes it. Play borrows pieces
long-term significance of play can be grouped into three
of other behavior patterns, usually incomplete sequences
categories (Thompson 1996):
and often in an exaggerated form. It consists of elements
drawn from other, functionally different behavior pat- 1. Physical: training for strength, endurance, and
terns juxtaposed in new sequences. Some species may muscular coordination, particularly the skills relat-
rapidly alternate prey-catching movements and aggres- ing to intraspecific fighting and prey capture.
sive behavior, while others mix components from hunt-
2. Social: practice of social skills such as grooming and
ing and sexual behavior. The movements of play are often
sexual behavior; establishment and maintenance of
repeated more often than during nonplay interactions,
social bonds. Play also helps develop an animal’s
and play may be slightly modified from another behav-
ability to read and send signals to communicate with
ior. When a dog is aggressive, it bares its teeth and
other members of its species.
growls. Its hair stands on end, adding to its ferocious
appearance. However, during a play fight, the growl is not 3. Cognitive: learning specific skills or improving over-
accompanied by the raising of hair. Play is also defined all perceptual abilities.
As we explore these hypotheses further, we will see
that some examples of play fit into more than one cate-
gory. One of the physical benefits of play may be that it
helps form connections between neurons in the brain,
especially in the cerebellum, a brain region important in
motor coordination and memory of motor patterns.
There is a limited time period during which synapses are
being formed between neurons in the cerebellum. During
this time, experience affects the number and pattern of the
synapses. In house mice, Norway rats, and domestic cats,
locomotor play coincides with the period when the cere-
bellar architecture is being shaped. In other words, loco-
motor play begins just when experience can modify the
connections within the cerebellum and ends when the
cerebellar structure is set (Byers and Walker 1995).
FIGURE 8.8 Lion cubs playing. Play is thought to be vital Another physical benefit of play is that it affords ani-
to the development of behavior in mammals. mals the opportunity to practice skills that will be essen-
 The Concept of Sensitive Periods 159

tial to later survival (Caro 1988). Hunting games, for observed in some provisioned troops; it has never been
instance, may help perfect the movements of catching observed in nonprovisioned troops. Stone handling pre-
prey such as stalking and shaking. Some examples of sumably occurs in provisioned troops because troop
these actions are familiar. Kittens stalk leaves, and pup- members have substantial free time and don’t have to
pies often “shake the life” from toys as they would a prey search for food. The behavior is only acquired by young
animal. During object play, fledgling American kestrels individuals, but once acquired, it is practiced into adult-
(Falco sparverius) prefer objects that resemble their nat- hood and old age. The performance of stone handling
ural prey (Negro et al. 1996). Play may have some imme- in older individuals makes the behavior different from
diate benefits for cheetah (Acinonyx jubatus) cubs by many other forms of play in primates, which disappear
honing their predatory skills. Cubs that playfully from the behavioral repertoire with age.
crouched and stalked littermates also crouched and Charmalie Nahallage and Michael Huffman (2007)
stalked prey more often than less playful companions. conducted daily observations of a provisioned troop of
And when the mother released live prey for her cubs to Japanese macaques known to exhibit stone handling.
catch, those that showed the highest rates of object play The researchers indicate that young Japanese macaques
and contact social play were more likely to be success- engage in frequent, short-duration bouts of stone han-
ful (Caro 1995). dling, often accompanied by locomotion and energetic
The play fighting of young animals may serve as prac- movements. In contrast, adults stone-handle less fre-
tice for the battles of adults that establish dominance hier- quently and for longer durations. Adults also exhibit
archies and defend territories. In the fury of a play fight, more complex manipulative patterns and stone-handle
no serious biting and no threat behavior take place. In when stationary, often in a favorite location. Based on
some species, larger, older, and more dominant animals these and other differences, the authors hypothesize that
seem to handicap themselves in tussles with weaker play- the functions of stone handling may differ for young and
mates. Strength and skill are often matched to those of the adult macaques. Nahallage and Huffman propose that
partner. Some animals seem to practice territory defense stone handling functions in the development of motor
as well. Young deer and goats vie for possession of an area and perceptual skills in young macaques, and in the
in a game reminiscent of King of the Mountain. maintenance and regeneration of neural pathways in
It has also been suggested, however, that rather than adults. They further suggest that stone handling may
being training for serious adult fighting, play fighting slow the deterioration of cognitive function associated
develops beneficial cognitive and social skills. During with aging in this long-lived species. Testing their
play fights among squirrel monkeys (Saimiri sciureus), the hypotheses will require additional behavioral observa-
young males may reverse dominant and submissive roles tions, cognitive testing, and neuroanatomical data.
(Biben 1998). Without experience in a dominant role,
young monkeys may grow up to be overly submissive,
and without experience in the submissive role, they may
grow up to be bullies. Play fighting may also help a juve- THE CONCEPT
nile learn to read the intentions of others. Is the oppo- OF SENSITIVE PERIODS
nent bluffing? How motivated is this opponent? These
We see, then, that factors such as the experience of play,
social and cognitive skills may, in fact, prove to be more
characteristics of the physical environment, or changes
important than physical skills.
in the animal’s nervous system or hormonal state can
Although we typically associate behavioral develop-
cause behavioral change during development. Now we
ment with youth, changes in behavior continue through-
will see that often these factors have their effects on
out adulthood. After all, the nervous and endocrine
development during particular windows of time, called
systems of animals change with age, as do the conditions
sensitive periods.
individuals experience in their physical and social envi-
ronments. Though less playful than juveniles, some
adults do play. For example, spontaneous play is com- CHANGING TERMINOLOGY—
mon among adult dogs (Bauer and Smuts 2007) and their FROM CRITICAL PERIODS
ancestors, wolves (Mech 1970). Does play behavior dif-
TO SENSITIVE PERIODS
fer between adults and young? And if engaging in play
helps young animals to hone their physical, social, and Early on, windows of opportunity for learning were
cognitive skills for use in adulthood, then what function called critical periods. Konrad Lorenz (1935) borrowed
might play have in adulthood? Let’s take a look at stone this term from embryology, where it was used to describe
handling, a type of solitary object play exhibited by some times in early development that were characterized by
adult, juvenile, and infant Japanese macaques (Macaca rapid changes in organization. During these brief, well-
fuscata). The behavior consists of gathering, manipulat- defined periods, an experimental interruption of the nor-
ing, and scattering stones. Stone handling has been mal sequence of events produced profound and
160 Chapter 8 / The Development of Behavior

irreversible effects on the developing embryo. Thus, as specifics during a brief sensitive period before leaving the
first used by Lorenz, a critical period was a phase of sus- nest. Otherwise, choosing an appropriate mate could
ceptibility to environmental stimuli that was brief, well later be a confusing exercise indeed because birds that
defined, and within which exposure to certain stimuli waited too long to learn the defining qualities of their
produced irreversible effects on subsequent behavior. species might very well learn the plumage and song char-
Recently, terms such as sensitive period, sensitive acteristics of another species.
phase, susceptible period, and optimal period have been used Some animals have little or no contact with their
in place of critical period. The newer terms indeed parents or other close relatives after birth or hatching.
reflect certain modifications in the definition of this One might wonder, then, would sensitive periods occur
period in light of more recent research (reviewed in early in development in these species? And is early
Michel and Tyler 2005). In fact, many of Lorenz’s (1935) learning limited to acquiring knowledge about appro-
basic precepts have now been modified. Specifically, we priate social partners, or do animals also learn charac-
now know that such periods (1) are fairly extended, (2) teristics of appropriate places to live or breed? Consider
are not sharply defined but gradual in their onset and the case of Pacific salmon in the genus Oncorhynchus.
termination, (3) differ in duration among species, indi- Adult salmon spawn in freshwater, usually streams, and
viduals, and functional systems, and (4) depend on the depending on the species and population, they may or
nature and intensity of environmental stimuli both may not die after spawning. When the eggs hatch and
before and during the sensitive period. Moreover, most the fry eventually emerge from their gravel nests, they
phenomena based on sensitive periods are not irre- typically pass through several developmental stages in
versible. Instead, patterns of behavior developed during their home streams, the last of which, called the smolt
sensitive periods can usually be altered or suppressed stage, prepares them for migrating thousands of kilo-
under certain conditions, especially those associated meters downstream to enter oceanic feeding grounds.
with high levels of stress. Deprivation (e.g., rearing an After a time at sea, virtually all surviving adults return
animal in isolation or in darkness) can reverse or destroy to their natal stream to spawn. It is a remarkable feat of
a pattern of behavior established during a sensitive navigation, and when they reach the freshwater inlets,
period. It is important, however, not to overemphasize they unerringly swim up the appropriate tributaries,
the reversibility of patterns of behavior established dur- making all the correct decisions at every fork until they
ing sensitive periods. Conditions such as rearing in com- reach the very stream where they were spawned—and
plete isolation or total darkness are unlikely to be they seem to do it by smell. Apparently, before their
encountered by most animals outside the laboratory migration to the sea and during a sensitive period, juve-
environment. Furthermore, even in the laboratory, nile salmon learn the odor of water at the site where
behaviors established during sensitive periods are usu- they were spawned. The water at the natal spawning site
ally more resistant to change than those learned at other has a unique chemical composition known as the “home
times (Immelmann and Suomi 1981). We will use the stream olfactory bouquet,” or HSOB. The sensitive
term sensitive period because it is now commonly found period for learning the HSOB seems to coincide with
in the literature. Our working definition of sensitive smoltification, the developmental transformation of
period is a time during development when certain expe- young salmon from parr (freshwater residents) to
riences have a greater influence on the characteristics of smolts ready for seaward migration and life (see
an individual than at other stages. Chapter 7) (Carruth et al. 2002). It is also possible that
the sensitive period begins somewhat earlier (Dittman
and Quinn 1996). In any case, upon returning to their
natal stream, adult salmon are stimulated to swim
TIMING OF SENSITIVE PERIODS
upstream by the familiar odor. Why is it important that
In most animals, sensitive periods occur early in life. salmon learn the precise location of their natal stream?
Why is this so? We usually assume that this is the time The answer is that each population is finely adapted to
when animals have the greatest opportunity to gain its home water, so much so that salmon experimentally
knowledge from parents and close relatives, knowledge introduced into other streams show a higher mortality
that is particularly important in species recognition. rate than locally adapted individuals (Quinn and
Later, they might not interact with them so intimately, Dittman 1990). The period of early learning thus
and in some cases they will, in fact, be exposed to intense ensures that it is the odor of the fish’s own spawning
stimuli from other species (Immelmann and Suomi place that is remembered (for a more detailed discussion
1981). For example, in some species of birds, the young of salmon homing, see Chapter 10). We see, then, that
remain in the nest for only a few weeks after hatching during sensitive periods, animals may learn the appro-
and then leave to join mixed-species flocks. It would not priate cues, not only of conspecifics, but also of the local
be surprising if these young learn to recognize con- physical environment.
 The Concept of Sensitive Periods 161

ple, why learning a second language later in life is often

STOP AND THINK more difficult than learning the first language.
A common strategy for restoring or enhancing popula-
tions of anadromous salmonids (i.e., those that return
from the sea to breed in freshwater) is to artificially rear MULTIPLE SENSITIVE PERIODS
young in hatcheries and then release them into streams
with the expectation that they would eventually migrate Individuals typically experience multiple sensitive peri-
to the sea. Given what you know about olfactory imprint- ods during their development (Bischof 2007). We will
ing in salmon, how might the release be orchestrated and see that a young male songbird experiences sensitive
its timing planned to result in good return rates to fresh- periods during song learning. The same young bird will
water spawning grounds? experience sensitive periods associated with sexual
imprinting (learning the characteristics of an appropri-
ate mate; see later discussion). Another example of mul-
Onset of Sensitivity tiple sensitive periods within individuals concerns visual
It is clear that young animals have a heightened sensitiv- development in humans (Lewis and Maurer 2005).
ity to certain environmental stimuli such as the physical Researchers comparing visual development in visually
appearance of a parent or the smell of a home stream. normal children with that of children deprived of early
One might ask, then, what causes this increased sensitiv- visual experience because they were either born with
ity to certain cues? One suggestion is that the onset of cataracts or developed them later have discovered that
sensitivity may be due both to endogenous (internal) and there are different sensitive periods for different aspects
exogenous (external) factors (Bateson 1979). Increases in of vision (e.g., visual acuity, peripheral vision, motion
sensitivity generally begin as soon as the relevant motor detection), and even within each of these aspects there
and sensory capacities of the young animal are developed. may be more than one sensitive period.
Changes in the internal state, such as fluctuations in hor-
mone levels, may also influence sensitivity. Then,
endogenous factors interact with environmental variables
SOME EXAMPLES OF SENSITIVE
to produce the start of the sensitive period. For example, PERIODS IN BEHAVIORAL
although the visual component of filial imprinting in DEVELOPMENT
birds (the response of newly hatched young to follow Now that we have discussed the definition and timing of
their mother; see later discussion) begins once hatchlings sensitive periods, and the fact that a given individual—
are able to perceive and process optical stimuli, experi- be it a songbird or a human—will experience multiple
ence with light appears to interact with the internal con- sensitive periods, let’s consider some examples of behav-
ditions to initiate this particular sensitive period (e.g., ioral development that depend, to varying degrees, on
Bateson 1976). Here, then, we have endogenous factors specific experiences during a window of time.
(ability of the nervous system to process optical stimuli)
interacting with environmental factors (exposure to
Filial Imprinting
light) to produce the start of the sensitive period.
Anyone who has ever watched chicks in a farmyard or
ducklings and goslings on a pond knows that the young
Decline in Sensitivity generally follow their mother wherever she goes
Several explanations have been proposed for the termi- (Figure 8.9). How does such following behavior develop?
nation of sensitive periods (reviewed in Johnson 2005); Konrad Lorenz (1935), working with newly hatched
we consider two of them here. One explanation assumes goslings, was the first to systematically study this behav-
that the decline in sensitivity is under endogenous con- ior. In one experiment, he divided a clutch of eggs laid
trol, perhaps influenced by a maturational timetable. by a greylag goose (Anser anser) into two groups. One
According to this idea, some physiological factor, intrin- group was hatched by the mother, and as expected, these
sic to the animal, ends the period of receptivity to exter- goslings trailed behind her. The second group was
nal stimulation once a specific stage of maturation has hatched in an incubator. The first moving object these
been reached. Another explanation is based on the idea goslings encountered was Lorenz, and they responded
that learning is a self-terminating process. In essence, to him as they normally would to their mother. Lorenz
learning causes neurobiological changes that reduce marked the goslings so that he could determine in which
plasticity and ultimately terminate the sensitive period. group they belonged and placed them all under a box.
An example of this idea is the proposition that learning When the box was lifted, liberating all the goslings
a first language causes neurobiological changes in simultaneously, they streamed toward their respective
humans that effectively bring the sensitive period for lan- “parents,” normally reared goslings toward their mother
guage learning to a close. This might explain, for exam- and incubator-reared ones toward Lorenz. The goslings
162 Chapter 8 / The Development of Behavior

FIGURE 8.9 Young Canada geese follow-

ing their mother. The following response
results from filial imprinting, the process
by which young precocial birds learn the
characteristics of their mother and then
preferentially follow her.

had developed a preference for characteristics associated For example, chicks without any previous exposure to
with their “mother” and expressed this preference an adult conspecific or a red box preferentially approach
through their following behavior. The attachment was a stuffed conspecific when given a choice between it and
unfailing, and from that point on Lorenz had goslings the red box. (This is not to say that experience is unim-
following in his footsteps (Figure 8.10). portant in the development of the predisposition for
Because social attachment evidenced by following conspecific characteristics. It turns out that other non-
seemed to be immediate and irreversible, Lorenz named specific experiences, such as being handled or placed in
the process Pragung, which means, “stamping.” The a running wheel, can induce the predisposition as long
English translation is “imprinting.” Used in this context, as these experiences occur during the sensitive period.)
the term suggests that during the first encounter with a In the second process, called filial imprinting, the
moving object, its image is somehow permanently young bird learns, through exposure to its mother, her
stamped on the nervous system of the young animal. particular characteristics and then preferentially follows
We now know that at least two distinct processes her. The biological function of filial imprinting is prob-
are involved in the development by young birds such as ably to allow young birds to recognize close relatives
chicks, ducklings, and goslings of a preference for fol- and thereby distinguish their parents from other adults
lowing their mother (studies reviewed by Bolhuis and that might attack them (Bateson 1990). The two
Honey 1998; Hogan and Bolhuis 2005). In one process, processes—development of the predisposition and filial
a predisposition to approach objects with the general imprinting—seem to be localized in different regions of
characteristics of conspecifics emerges in the young the brain. We will focus our discussion on the develop-
bird, even without previous exposure to a conspecific. ment of the following response in mallard ducklings.

FIGURE 8.10 Goslings following their

“mother,” Konrad Lorenz. Lorenz was
one of the first scientists to study
imprinting experimentally.
 The Concept of Sensitive Periods 163

Mallard ducklings, like most young in the orders cues in the development of the following response and
Anseriformes (ducks, geese, and swans) and Galliformes (2) whether the parental call of a duckling’s own species
(chickens, turkeys, and quail), are precocial, that is, quite would be more effective than that of other species in
capable of moving about and feeding on their own just inducing and maintaining attachment behavior. As part
a short time after hatching. Filial imprinting is usually of the study, 224 eggs were hatched in an incubator; thus,
studied in species with precocial young. The following the ducklings never came in contact with their mothers.
response is nonexistent—or much less evident—in The ducklings were divided into four groups and tested
species such as the songbirds discussed later, whose for their following response. In all four groups, the duck-
young are altricial, that is, virtually helpless and incapable lings were tested with a stuffed replica of a Peking hen
of feeding on their own or following their parents for the as it moved about a circular runway. Individuals in one
first few weeks after hatching. We begin with a brief of the groups, however, were tested with a silent hen,
description of reproduction and early development in whereas individuals in the other three groups were tested
mallards. with hens that emitted assembly calls through a speaker
Upon finding a suitable nesting site, typically a shal- concealed on their undersides. Of the ducklings that
low crevice in the ground, the mallard hen begins to lay heard assembly calls, one group was exposed to mallard
her eggs, at the rate of one egg per day (Miller and calls (Peking calls), one group to wood duck calls, and
Gottlieb 1978). The average clutch size is eight to ten, one group to domestic chicken calls. Each duckling was
and after the last egg is laid the hen begins incubation, given a 20-minute test to determine whether it would
a process that lasts approximately 26 days. Encouraged follow the stuffed model around the circular arena.
by the warmth of the mother’s body, the embryo inside The results are presented in Figure 8.11. As you can
each egg begins to develop. Two to three days before see, the auditory stimulus of the maternal call is impor-
hatching, each embryo moves its head into the air space tant in filial imprinting: all conditions with calls were
within its egg and begins to vocalize; these vocalizations much more effective than no call at all. Furthermore,
are called contentment calls. About 24 hours later, the even though the ducklings had never before heard the
embryos pip the outer shell and then take another day maternal call of their species, they responded selectively
to break through the rest of the shell and hatch. Most to it. The maternal call of the mallard was far more effec-
of the ducklings in a clutch will hatch within an inter- tive than that of the wood duck or chicken in inducing
val of ten hours. The hen broods her young for a day following by the ducklings.
and then leaves the nest and emits calls to encourage the
ducklings to follow. Although she vocalizes during incu-
bation and brooding, the frequency increases dramati- 100
cally at the time of the nest exodus. Prompted by their
mother’s assembly calls, the young leave the nest and
follow her to a nearby pond or lake, where they will pad-
Percentage following

dle behind her.

Observations such as these stimulate many ques-
tions. What characteristics of the mother form the basis 50
for the ducklings’ attachment? Do the ducklings imprint
on the mother’s call, physical appearance, or some com-
bination of the two? What role might siblings play in
the development of the following response? Finally, is
there a sensitive period during which exposure to cer-
tain cues must occur for the normal development of fil- 0
ial behavior? Mallard Wood duck Chicken Silence
Many of the answers to these questions came from
Experimental condition
the laboratory of Gilbert Gottlieb, a pioneer in behav-
ioral development who passed away in 2006. For several FIGURE 8.11 A stuffed Peking hen emitting mallard calls
decades, Gottlieb and co-workers examined the devel- was more effective in eliciting the following response in
opment of the following response in Peking ducks, a incubator-reared Peking ducklings than a hen emitting
either wood duck or chicken calls (Peking ducks are a
domestic form of the mallard (despite their domestica-
domestic form of the mallard duck). All hens with calls
tion, Peking ducks are quite similar to their wild coun- were more effective than a silent hen. Thus, auditory
terparts in their behavior). Here we consider some of the cues from the mother are important in controlling the
work of Gottlieb and his colleagues. early behavior of ducklings, and ducklings respond
In one experiment with ducklings that had never had selectively to the call of their own species without
contact with the mother, Gottlieb (1978) examined (1) previous exposure to it. (Drawn from the data of
the relative importance of the hen’s auditory and visual Gottlieb 1965.)
164 Chapter 8 / The Development of Behavior

In a second experiment, incubator-reared ducklings embryonic auditory experience always choose the mal-
were given a choice of following either a stuffed Peking lard’s call over the chicken’s (Table 8.1). Thus, auditory
hen that was emitting mallard calls or a stuffed Peking experience before hatching is important to the devel-
hen that was emitting chicken calls. When placed in this opment of the following response in mallard ducklings
simultaneous choice situation approximately one day because it induces a predisposition to approach the mal-
after hatching, the majority of ducklings (76%) followed lard assembly call.
the model that was emitting the mallard call. Is there a sensitive period during which exposure to
Taken together, these experiments demonstrate that contentment calls must occur for ducklings to exhibit a
auditory stimuli from the mother are an important preference for the call of the mallard hen? If ducklings
influence on the behavior of newly hatched ducklings. heard contentment calls after hatching, would they dis-
Furthermore, ducklings respond selectively to the play the normal preference for mallard calls? Gottlieb
maternal assembly call of their own species without any (1985) set out to answer these questions. He began by
previous exposure to it (remember that all ducklings raising ducklings in isolation and making them mute
were reared in incubators and therefore had no contact before they began to vocalize within the egg (again, these
with hens). This preferential response by the ducklings manipulations ensure that the ducklings are not exposed
to the assembly call of their species is an example of the to their own calls or the calls of their siblings). One
predisposition described earlier. Recall that young pre- group was exposed to contentment calls during the
cocial birds may develop a preference for stimuli from embryonic period (approximately 24 hours prior to
conspecifics without prior exposure to the particular hatching), and another group was exposed to content-
stimuli. ment calls during the postnatal period (approximately 24
These findings do not, however, suggest that expe- hours after hatching). Ducklings in each group were then
rience is unimportant in the development of a prefer- tested for their preference for the mallard hen’s call. This
ence for the assembly call. Experience, as it turns out, time, however, they were given the choice of approach-
is critical, but it occurs prenatally (before hatching). For ing a speaker that emitted normal calls of a mallard hen
Peking ducklings to exhibit a preference for the mallard or a speaker that emitted artificially slowed calls of a mal-
hen’s call, they must hear their own contentment calls lard hen. As shown in Table 8.2, ducklings must be
or those of their siblings before hatching (Gottlieb exposed to contentment calls before hatching if they are
1978). If ducklings are reared in isolation (and therefore to show a preference for the normal call; exposure after
are not exposed to the calls of their siblings) and made hatching is ineffective in producing the preference for
mute just before they begin to vocalize within the egg the normal call of their species. Here, then, we have an
(and therefore are not exposed to their own calls), they example of a sensitive period occurring during embry-
no longer display their highly selective response to the onic development.
maternal call of their species. When these ducklings What role, if any, might visual stimuli play in the
without normal embryonic auditory experience are development of the following response? Several experi-
placed in a test apparatus equidistant between two ments have shown that two conditions must be simulta-
speakers, one speaker emitting a mallard’s maternal call neously met if visual imprinting on a mallard hen is to
and the other a chicken’s maternal call, they choose the occur in the ducklings. The ducklings must be reared
latter almost as often as they choose the former 48 hours with other ducklings and allowed to actively follow a
after hatching. In contrast, ducklings with normal mallard hen, or a model, if they are to prefer the appear-

TABLE 8.1 The Effects of Embryonic Auditory Experience on Call Preferences

of Peking Ducklings 48 Hours after Hatching
Preference

N Mallard call Chicken call Both

Vocal-communal 24 24 0 0
Mute-isolated
First experiment 22 12 9 1
Replication 21 14 6 1

Total 43 26 15 2
Ducklings raised in the vocal-communal group could hear themselves and the calls of siblings prior to hatching; mute-isolated ducklings
had no such auditory experience. N ⫽ number of ducklings that responded to calls.
Source: Data from Gottlieb (1978).
 The Concept of Sensitive Periods 165

TABLE 8.2 Effects on Preferences of Mute-Isolated Ducklings of Embryonic Versus Postnatal

Exposure to Contentment Calls
Preference

Time of exposure to
contentment calls N Normal mallard Slowed mallard Both

Embryonic 32 21 8 3
Postnatal 37 14 20 3
Source: Data from Gottlieb (1985).

ance of a hen of their own species to that of a hen of round the meadow, in figures of eight, glancing con-
another species at later testing (e.g., Dyer et al. 1989; stantly over his shoulder and quacking—but the duck-
Lickliter and Gottlieb 1988). Ducklings that are reared lings, the all-revealing and all-explaining ducklings
in isolation and given passive exposure to a stuffed mal- were hidden in the tall spring grass from the view of
lard hen (i.e., housed with a stationary model) do not the astonished crowd!
develop a preference for the mallard hen. These results
To summarize, the experiments of Gilbert Gottlieb
suggest that under natural conditions, ducklings learn
and co-workers have demonstrated that the following
the visual characteristics of their mother after leaving the
response in Peking ducklings results from a complex
nest and following her around.
interaction of auditory, visual, and social stimuli pro-
We see, then, that both auditory and visual stimuli
vided by the hen and siblings. Several important gen-
are important in development of the following response.
eralizations about early behavioral development have
Auditory stimulation from the mother appears to be
arisen from their work. First, we can no longer think of
largely responsible for prompting the ducklings to leave
experience as occurring only after birth or hatching;
the nest and for influencing their earliest following
embryonic experience can also influence behavior. In
behavior. However, the hen’s appearance becomes
the case of Peking ducklings, listening to the content-
important after the nest exodus.
ment calls of siblings before hatching is critical to devel-
The importance in filial imprinting of auditory cues,
opment of their preference for the maternal call of their
visual cues, and active following was well summarized by
own species after hatching. Second, the experiments
Lorenz (1952, pp. 42–43). In an experiment with Peking
with ducklings demonstrate that a variety of stimuli may
ducklings, he found himself in a rather embarrassing
be involved in developing a single pattern of behavior
position for one destined to become a Nobel Laureate.
and that the relative importance of different stimuli may
In his words,
change as the young animal matures. Although the fol-
The freshly hatched ducklings have an inborn reaction lowing behavior of ducklings soon after hatching is
to the call-note, but not to the optical picture of the influenced largely by auditory cues from their mother,
mother. Anything that emits the right quack note will only a few days later visual stimuli (in combination with
be considered as mother, whether it is a fat white auditory stimuli) become important in the following
Peking duck or a still fatter man. However, the substi- response. The relative priorities of different cues match
tuted object must not exceed a certain height. At the the timing for development of the auditory and visual
beginning of these experiments, I had sat myself down systems; in ducklings, as in all birds, the auditory sys-
in the grass amongst the ducklings and, in order to tem develops before the visual system (Gottlieb 1968).
make them follow me, had dragged myself, sitting, These results emphasize the close interaction between
away from them. So it came about, on a certain Whit- physical maturation and experience in early behavioral
Sunday, that, in company with my ducklings, I was development. Finally, the study of filial imprinting in
wandering about, squatting and quacking, in a May- Peking ducklings illustrates that we must be open-
green meadow at the upper part of our garden. I was minded when trying to sort out just which experiences
congratulating myself on the obedience and exactitude affect a given behavior. Who would have thought that
with which my ducklings came waddling after me, listening to siblings before hatching or interacting with
when I suddenly looked up and saw the garden fence siblings after hatching would be critical in the develop-
framed by a row of dead-white faces: a group of ment of the ducklings’ attachment to their mother? We
tourists was standing at the fence and staring horrified now know that such nonobvious experiential factors are
in my direction. Forgivable! For all they could see was indeed essential to the development of following behav-
a big man with a beard dragging himself, crouching, ior in this species.
166 Chapter 8 / The Development of Behavior

Sexual Imprinting (sometimes called the stabilization stage). This stage

We have seen that early experience influences a duck- occurs when the male actually courts a female for the
ling’s attachment to its mother, an attachment that the first time. During courtship, the social preference for the
young bird demonstrates by trailing behind her in the parental species becomes linked to sexual behavior and
days and weeks following the exodus from the nest. Early is consolidated or stabilized. The precise timing of the
experience also has important consequences for the consolidation stage requires further research, but it
development of mate preferences in birds. In many appears to begin when males become sexually mature
species, experience with parents and siblings early in life around 70 days of age and to end at about 100 to 150
influences sexual preferences in adulthood. The learn- days of age. Once the consolidation stage ends, no new
ing process in this case is called sexual imprinting. preferences can be established.
Typically, sexual preferences develop after filial prefer- The acquisition and consolidation phases are dif-
ences, although the sensitive periods may overlap to ferent processes, and each can be modified by different
some degree (Bateson 1979). Whereas filial imprinting factors. The strength of the social preference for the
is indicated by the following response of young birds, parental species that develops during the acquisition
sexual imprinting is typically shown in the preferences phase is influenced by the amount of food that the young
of sexually mature birds for individuals of the opposite male is given by his parents. However, consolidation, the
sex. It is important to note that unlike filial imprinting, linking of the social preference for the parental species
sexual imprinting occurs in both altricial and precocial to sexual behavior, is most affected by the degree to
birds. Our first example concerns the altricial zebra which the male is aroused at the time of courtship
finch, and our second, the precocial Japanese quail. (Oetting et al. 1995).
One early and dramatic demonstration of the At first glance, it might seem curious that an indi-
importance of early experience to subsequent mate pref- vidual has to learn to identify an appropriate mate.
erence came from cross-fostering experiments with Wouldn’t it be a safer evolutionary strategy to have a
finches. Klaus Immelmann (1969) placed eggs of zebra mating preference that cannot be modified by early
finches (Taeniopygia guttata) in clutches belonging to social experience? Apparently not. For now, however, we
Bengalese finches (Lonchura striata). The Bengalese fos- can only hypothesize about the importance of early
ter parents raised the entire brood until the young could learning in choosing mates.
feed themselves. From then on, young zebra finch males One idea, put forth by Patrick Bateson (1983), pro-
were reared in isolation until sexually mature. When vides an interesting explanation for the functional sig-
they were later given a choice between a zebra finch nificance of sexual imprinting. Bateson suggested that
female and a Bengalese finch female, they courted animals learn to identify and selectively respond to kin.
Bengalese females almost exclusively. Armed with information on what their relatives look like,
A second study demonstrated that brief contact with individuals then choose mates similar but not identical to
foster parents early in life could exert a more powerful, their family members. Given that both extreme inbreed-
longer-lasting influence on mate preference than long- ing and outbreeding may have costs (see Chapter 14), sex-
term social contact in adulthood. Cross-fostered zebra ual imprinting provides information that allows animals
finch males were again separated from their foster to strike a balance between the two. Bateson used evi-
Bengalese parents, but this time they were provided dence from studies of quail to support his argument.
with a conspecific female and nesting supplies. Most of Japanese quail (Coturnix japonica) prefer to mate with
these males eventually mated with conspecific females individuals that are similar to, yet slightly different from,
and successfully produced young. When they were members of their immediate family (Bateson 1982). In
tested several months or years later, however, the males one study, chicks were reared with siblings for the first
still displayed a preference for Bengalese females 30 days after hatching and then socially isolated until they
(Immelmann 1972). became sexually mature. At 60 days of age, males and
More recent experiments have shown that sexual females were tested for mate preference in an apparatus
imprinting is a two-stage process (Bischof 1994, 2003). that permitted viewing of several other Japanese quail
The first stage is the acquisition stage. This stage begins (Figure 8.12a). The birds that were viewed were of the
around ten days after hatching (when the visual system opposite sex of the test animals and belonged to one of
of a zebra finch has matured sufficiently to detect struc- the following five groups: (1) familiar sibling, (2) novel
tured visual information) and ends at about 40 to 60 days (unfamiliar) sibling, (3) novel first cousin, (4) novel third
of age. It is during this time that a young male forms a cousin, or (5) novel unrelated individual. Similarity in
social bond to its parents. Because of this bond, the male plumage between Japanese quail is considered to be pro-
prefers to socialize with members of its parents’ species, portional to genetic relatedness, and thus test animals
and this social preference guides his first courtship could presumably judge genetic distance on the basis of
attempts. The second stage is the consolidation stage plumage characteristics. As we see in Figure 8.12b, both
 The Concept of Sensitive Periods 167

a himself points out, we must be careful in generalizing

results from laboratory studies to animals in their natural
environment. Clearly, laboratory conditions of rearing
and testing animals are vastly different from the natural
conditions under which animals live and choose mates.
We should also be aware that our choice of measures, such
as the number of approaches an animal makes toward
another, might not reflect mate preference or actual mat-
ing in nature. In the best of all possible worlds, we would
compare results from the controlled environment of the
laboratory with field observations on the impact of early
experience on subsequent mate choice. In the field, how-
ever, it is often difficult to know the precise genetic rela-
tionships of animals and to track animals from birth to
adulthood as would be necessary to chronicle their early
experiences and subsequent mating behavior.
Although many factors make the study of sexual
imprinting and later mate choice in free-living animals
difficult, Tore Slagsvold and colleagues have performed
b 40 cross-fostering experiments with wild birds and then
followed the birds’ subsequent survival and mating suc-
Percentage time near

cess (Slagsvold 2004; Slagsvold et al. 2002). In one

30
experiment, they cross-fostered whole broods of great
tits (Parus major) to nests of blue tits (Cyanistes caeruleus)
and found that the young experienced low mating suc-
20
cess because they strongly preferred their host species.
Thus, sexual imprinting had occurred in this field exper-
10 iment. However, when the researchers performed the
reciprocal cross, that is, cross-fostered blue tits to great
tits, most young mated with a conspecific, indicating
that they had not sexually imprinted on their host
Sibling Sibling 1st 3rd Unrelated
Cousin Cousin species. Similarly, when pied flycatchers (Ficedula
hypoleuca) were cross-fostered to nests of blue tits and
Familiar Novel great tits, there was no evidence that the young had sex-
ually imprinted on either host species. The cross-
FIGURE 8.12 Japanese quail prefer to spend time near
fostered flycatchers did not differ from control birds
individuals that are similar to, yet slightly different from,
members of their immediate family. (a) Apparatus used by
(flycatchers from unmanipulated nests) in their mating
Patrick Bateson to test preferences of adult quail. (b) Both success or breeding success (e.g., clutch size, number of
male (shown as triangles) and female quail (shown as young hatched and fledged). In addition, cross-fostered
circles) prefer first cousins, possibly striking an optimal male flycatchers were similar to control male flycatch-
balance between inbreeding and outbreeding. (a: From ers in their response to intruders presented in a cage
Bateson 1982. b: Modified from Bateson 1982.) close to their nest box. The cross-fostered males
reacted aggressively when the cage contained a conspe-
males and females preferred to spend time near first cific male, displayed at their nest when it contained a
cousins. The sexual preferences displayed by quail reflect conspecific female, and ignored the cage when it con-
a choice slightly displaced from the familiar characteris- tained a female of their host species. Taken together,
tics of immediate family members: siblings are too famil- these experiments with free-living birds suggest that
iar, novel unrelated animals are too different, but first species vary in their sensitivity to sexual imprinting.
cousins are the perfect mix of familiarity and novelty. Although the causes of this variation are not yet known,
The results from studies of Japanese quail certainly Slagsvold and co-workers hypothesize that several
seem to support Bateson’s contention that through sexual factors may influence the sensitivity of a species to sex-
imprinting some young animals learn the characteristics ual imprinting. They predict decreased sensitivity to
of their close relatives and then, in adulthood, choose a sexual imprinting in species that are solitary rather than
mate similar but not identical to their family members. A social and polygynous rather than monogamous. The
note of caution, however, is in order. As Bateson (1983) researchers also predict that species will be less sensi-
168 Chapter 8 / The Development of Behavior

tive to sexual imprinting when cross-fostered to a host at one year of age, cross-fostered males exhibited strong
species that is either socially dominant or distantly social and mating preferences for females of their mater-
related. The study of additional species in the field is nal species, spending about 89% of their time with these
necessary to test these hypotheses. females. In contrast, normally reared male sheep and
So far, we have limited our discussion of sexual goats preferred females of their genetic species, spend-
imprinting to birds. Is there evidence of a similar process ing 96% of their time with them. For the next three
in mammals? Keith Kendrick and colleagues reciprocally years, the cross-fostered males were housed exclusively
cross-fostered offspring of domestic sheep (Ovis aries) with their genetic species, and mating preferences were
and goats (Capra hircus) at birth (Kendrick et al. 1998, assessed each year. Even after three years of sole expo-
2001). The young grew up with their heterospecific sure to their genetic species, cross-fostered males still
mothers (designated the maternal species), but also had strongly preferred to mate with females of their mater-
social contact with their genetic species throughout nal species (Figure 8.13a). Cross-fostered females pre-
development (e.g., lambs raised by goat mothers were ferred social contact with their maternal species,
allowed social contact with other sheep). When tested spending about 69% of their time with females of their

a Cross-fostered male goats Cross-fostered male sheep

100 100
Mean % mating choice

Mean % mating choice

80 80

60 60

40 40

20 20

0 0
1 2 3 4 1 2 3 4 Year 1 2 3 4 1 2 3 4 Year
Female Female Female Female
sheep goat sheep goat

b Cross-fostered female goats Cross-fostered female sheep

100 100
Mean % mating choice

Mean % mating choice

80 80

60 60

40 40

20 20

0 0
1 1 2 3 4 Year 1 2 3 4 1 2 3 4 Year
Male Male Male Male
sheep goat sheep goat

FIGURE 8.13 Mate preferences in adult goats and sheep are influenced by exposure to their mother early in life.
(a) Male mating preferences. Cross-fostered male goats and sheep prefer to mate with females of their maternal
species, and this preference is maintained even after several years of sole exposure to their genetic species. (b) Female
mating preferences. Cross-fostered female goats and sheep exhibit weaker mating preferences for the maternal species
than do males, and female preference can be reversed over time by exposure to their genetic species. (Modified from
Kendrick et al. 2001.)
 The Concept of Sensitive Periods 169

maternal species, but their preferences were weaker than

those shown by cross-fostered males. Cross-fostered
females also exhibited weaker mating preferences for
males of the maternal species, and such preferences were
reversible within one to two years of housing with their
genetic species (Figure 8.13b). Thus, sheep and goats
appear to learn the characteristics of an appropriate mate
during early exposure to their mother, but males are
more strongly influenced than females by such exposure.
What features of the mother (genetic or foster)
might young goats and sheep attend to? Further exper-
iments demonstrated that patterns of preference identi-
cal to those found when sheep and goats were tested with
live stimulus animals could be obtained by showing test FIGURE 8.14 A mother goat rejecting an unfamiliar off-
animals pictures of sheep and goat faces. These data sug- spring. Mother goats develop an attachment to their own
gest that features of the face may be critical in the for- young shortly after birth and butt away unfamiliar young.
mation of social and mating preferences in the two
species. Interestingly, another report from this research
group has shown that sheep have a remarkable ability to ment entails two processes, each of which requires
remember the faces of other sheep; perhaps to some peo- maternal contact with the offspring during sensitive peri-
ple’s surprise, sheep can remember 50 different sheep ods in the immediate hours after parturition. The sen-
faces for two or more years (Kendrick et al. 2007)! Also sitive periods for maternal attachment in sheep and goats
of interest are findings showing the importance of thus serve as an example of sensitive periods occurring
parental facial features in the development of mate pref- in adult animals.
erences in our own species (Bereczkei et al. 2004). The development of maternal attachment in sheep
and goats consists of two interrelated processes: (1) the
activation of maternal responsiveness (i.e., the display of
Maternal Attachment maternal behavior when in the presence of young) and
Another imprinting-like process that occurs in mammals (2) the establishment of maternal selectivity (i.e., direct-
is maternal attachment. In some species of ungulates, a ing maternal behavior only toward those young with
lasting bond between the mother and young is estab- which the mother has bonded) (studies reviewed by
lished rapidly after birth and results in the mother direct- Poindron et al. 2007). The existence of two processes
ing her care exclusively toward her own offspring. was initially suggested by findings showing that right
Domestic sheep and goats have been extensively studied after parturition mothers will accept any young; in other
in this regard, so we will return to them for our discus- words, there is a period of time during which mothers
sion of the development of maternal attachment. Like are maternally responsive but not selective. Here is a
many ungulates, sheep and goats live in large social description of the two processes in more detail, begin-
groups in which females tend to give birth synchro- ning with maternal responsiveness. Undisturbed female
nously. Young of both species are precocial, capable of sheep and goats display maternal behavior within min-
wandering away from their mother soon after birth and utes of giving birth. However, if their offspring is
mixing with other members of the group. Thus, it is removed at birth before they have had any contact with
important that a bond between mother and offspring be it, then maternal responsiveness fades within hours;
established early. Furthermore, young sheep and goats these females fail to show maternal behavior when
will initially approach any mother to nurse, so it also pays reunited with their offspring 4 to 12 hours later. Even a
for a mother to be able to tell her own young from oth- few hours of contact with their offspring, or a few min-
ers. Mothers do not provide care for one another’s off- utes in some cases, are sufficient to maintain maternal
spring. Indeed, rejection of the young of other mothers responsiveness. Thus, the sensitive period for maternal
is a no-nonsense affair that can be quite violent, involv- responsiveness seems to extend from parturition to no
ing head-butts, high-pitched vocalizations, and with- more than 12 hours postpartum in most mothers.
drawal from the unrelated youngster attempting to Experiments involving separation and subsequent test-
suckle (Figure 8.14). ing for acceptance of own or unrelated offspring suggest
Although the formation of mother–young relation- that the sensitive period for development of selectivity
ships in sheep and goats depends on mutual recognition extends from parturition to perhaps 1 or 2 hours post-
and bonding between the mother and her offspring, we partum; thus, this sensitive period appears to lie within
focus our discussion on the development of the mother’s the sensitive period for maternal responsiveness. Only 30
attachment to her offspring. Formation of this attach- minutes of contact with their offspring during the
170 Chapter 8 / The Development of Behavior

sensitive period appears sufficient for most mothers to duced into their respective colonies. These were the con-
display selective maternal behavior. trol workers. Other ants, the experimental workers, were
During the sensitive periods, what sensory cues from labeled in the same manner and then isolated in a glass
offspring are most important in the development of tube equipped with food and water. In one colony the
maternal responsiveness and selectivity? In both isolation period began at emergence; in the second,
processes, the olfactory system of the mother plays a third, and fourth colonies, the period of isolation began
prominent role, and olfactory cues from amniotic fluid when workers were two, four, or eight days old, respec-
seem particularly important. However, the two processes tively. Like the controls, there were 15 workers in each
differ in their ability to compensate for loss of olfaction of the four isolation groups. Champalbert and Lachaud
induced through experimental procedures such as cut- recorded the behavior of control and experimental ants
ting the main olfactory nerves, destroying the olfactory for 45 days after emergence.
bulbs, or blocking olfaction at the nostrils. Whereas all The question was, what effects, if any, would social
of these experimental treatments performed individually isolation have on the behavioral development of work-
on mothers prevent the development of maternal selec- ers in the experimental groups? Also, would the effects
tivity, they do not interfere with maternal responsive- of isolation depend on when they were isolated? During
ness. In the absence of olfactory cues, mothers appear to the first week after emergence, the control workers spent
compensate and rely on other sensory cues from most of their time feeding or being groomed by other
neonates to display maternal behavior at parturition and colony members. By the second week, however, work
beyond. However, they do not selectively respond to began, and the ants started to specialize in nursing activ-
their own offspring. ities, first caring for larvae, then cocoons, and finally
eggs. Sometime during the third week, the workers
changed their job status again and began to explore the
Brood Care nest and engage in domestic tasks. Finally, about a month
Until now we have been considering the effects of expe- after emergence, the control workers focused on activi-
rience during specific windows of time on the develop- ties related to their new careers as either guards at the
ment of social behavior in birds and mammals. We can nest or foragers outside the nest. Thus, over the course
now ask, what evidence have we of critical periods in of only a few weeks, E. tuberculatum workers typically
behavioral development in other groups of animals? Our underwent several changes in behavioral specialization,
next example comes from the insects. and in a very specific sequence.
The life of an ant can often be quite complicated. So, how did the behavior of ants in the experimen-
In many species, individuals live in complex societies in tal groups compare with that of controls? What were
which labor is divided among colony members, and their the effects of isolation immediately at emergence, or
roles can change dramatically and repeatedly over the two, four, or eight days later? It is interesting that the
course of a few weeks. Given their intricate social sys- workers isolated at two days after emergence showed the
tem and frequent changes in job status, it is not surpris- most abnormal behavioral development. They differed
ing that ants have become popular subjects for from controls in the order of appearance of their activ-
researchers interested in behavioral development. Of ities and in the level of performance of their tasks, being
particular interest is the Neotropical ant Ectatomma particularly lax in brood care (Figure 8.15). In contrast,
tuberculatum; in this species, an individual’s job changes the behavioral development of workers isolated either
with age. at emergence or four or eight days later was more sim-
Annette Champalbert and Jean-Paul Lachaud ilar to that of the controls. In the case of workers iso-
(1990) examined the role of early social experience in the lated at emergence, behavioral development was simply
development of behavior of E. tuberculatum workers. The delayed. The reintroduction of these workers into soci-
two researchers were interested in what effects a ten-day ety seemed to correspond to a second emergence;
period of social isolation would have on behavioral although 11 days old at the time of reintroduction, these
development in workers of different ages. The workers behaved like newly emerged ants. Soon after
researchers began their task on a coffee plantation in reintroduction into the colony, however, the various
southern Mexico, where they collected four ant colonies specializations in activities appeared progressively in the
from the bases of different coffee trees, each colony con- same sequence as that of the control workers. Only
taining 200 to 300 individuals. Each colony was placed minor abnormalities were noted in the workers isolated
into an artificial nest made of plaster that had several at four days, and the behavioral development of those
interconnected chambers and a foraging area. A glass isolated at eight days was almost identical to that of the
pane permitted the ants to be observed. A few hours after controls. In summary, the extent of behavioral abnor-
their emergence from cocoons, 15 workers from each of malities observed in E. tuberculatum workers as a result
the four colonies were individually labeled with a small of a ten-day period of social isolation depended on the
numbered tag glued to their thorax and then reintro- age of the workers at the time of isolation, with the most
 Pulling It All Together—The Development of Bird Song 171

12 to the colony. In view of the poor performance of work-

Average frequency of brood care activities ers isolated at two days, it also appears that the entire
sensitive period must occur in an uninterrupted fashion
to ensure normal behavioral development. Because
these workers received some portion of the necessary
social stimulation during their first two days following
emergence, their developmental system could not be put
6
on hold until reintroduction into the colony.
So far we have described the factors that influence
behavioral development and the fact that many of these
factors exert their influence during sensitive periods.
Now we will describe the development of singing behav-
ior in birds, an example that illustrates the combined
0 influences of genes, experience, and neural and hor-
Control Emergence 2 Days 4 Days 8 Days
monal control during particular sensitive periods.
Age at social isolation

FIGURE 8.15 A sensitive period for the development of

brood care behavior in worker ants. Here, a ten-day PULLING IT ALL TOGETHER—
period of social isolation beginning two days after
emergence produces more severe disruptions in the
THE DEVELOPMENT
performance of brood care activities by Ectatomma OF BIRD SONG
tuberculatum workers than do periods of isolation
beginning either at emergence or four or eight days All of us have probably, at one time or another, been
later. (Drawn from the data of Champalbert and struck by the beautiful and often complex songs pro-
Lachaud 1990.) duced by the birds around us. Consider, for example, the
rich and melodious soliloquies of the canary (Serinus
canaria) or the odd mix of rattles, whistles, and clicks
serious abnormalities occurring in ants isolated two days woven throughout the songs of the European starling
after emergence. (Sturnus vulgaris), a talented bird, indeed, but one that
Champalbert and Lachaud (1990) concluded from is largely unappreciated by North American birdwatch-
their observations of E. tuberculatum worker ants that ers because of its status as an introduced species that
there is a sensitive period during the first four days after aggressively preempts native cavity-nesting species from
emergence in which exposure to social stimuli affects the choice nest sites. Canaries and European starlings are
establishment of behavior, especially that related to members of the order Passeriformes. Some species
brood care. One might ask, then, why would workers within this order learn to sing (those within the oscine
isolated at emergence (and therefore removed from suborder, the so-called songbirds), while other species do
social stimuli during the four-day critical period) display not (those within the suboscine suborder, but there may
fairly normal behavioral development? The authors sug- be some exceptions; Kroodsma 2005; Saranathan et al.
gest that when workers are isolated immediately after 2007). Of the 22 remaining orders of birds, song learn-
emergence, their development is simply put on hold. ing is found in parrots (order Psittaciformes) and hum-
The reintroduction of these workers into the colony mingbirds (order Apodiformes). We will focus on the
after their period of social isolation mimics an emer- species that learn their song, and our examples will come
gence. During the four days following reintroduction, from the order Passeriformes. Nevertheless, it is impor-
these workers receive the social stimuli important in the tant to realize that even if species do not learn to sing,
development of brood care activities. In other words, it they may still produce fairly complex vocalizations. Let’s
is not the physiologically defined first four days of life take a look at how bird song develops.
that is important for the development of brood care Song development has been examined from the
behavior but rather the first four days of social contact standpoint of evolution and ecology, as well as from the
with colony members. These four days may follow either more mechanistic approach of assessing genetic, neural,
natural emergence or the artificial second emergence hormonal, and social influences on behavior. Together,
produced by the reintroduction of workers into the these approaches have revealed the continuous interplay
colony after isolation. Here, again, we see an example of between the developing bird and its internal and exter-
flexibility associated with sensitive periods in behavioral nal environment. As we will see, the factors that influence
development; in ants isolated at emergence, the lack of the development of song range from interactions between
exposure to appropriate social stimuli delays closure of cells to those between individuals. We begin by describ-
the sensitive period until a few days after reintroduction ing the development of singing behavior in the zebra
172 Chapter 8 / The Development of Behavior

FIGURE 8.16 Male zebra finches sing,

and females are attracted to their
song. Here a female zebra finch (left)
perches near a male (right).

finch, an extremely social species native to Australia of the song repertoire (number of songs sung by a bird).
(Figure 8.16). Zebra finches do very well in captivity (in For example, only male zebra finches sing, and song
fact, many of you have probably seen them in pet stores nuclei are larger in males than in females of this species
and heard the “beeps” in their song), so they are well- (Figure 8.17b) (Nottebohm and Arnold 1976). In zebra
suited to behavioral studies in the laboratory. Along the finches, sex differences in neuroanatomy are also evident
way we will also discuss other species that differ from at the cellular level; male song nuclei have more and
zebra finches to emphasize the variation across species. larger neurons than do the same regions in females
(Gurney 1981).
The apparent association in zebra finches between
GENETIC, HORMONAL, AND NEURAL
sex differences in singing behavior and sex differences in
CONTROL OF SONG
the size of song nuclei in the brain is not found in all
In zebra finches, as in many songbirds, only males sing species of songbird. As an example of a disassociation
and females are attracted to their song (e.g., Holveck and between singing behavior and the volume of song nuclei,
Riebel 2007). Males of many bird species use song not we turn to one of several tropical species of birds in
only to attract females, but also to advertise the bound- which both sexes sing. Male and female bush shrikes
aries of their territories. Zebra finches, however, breed (Laniarius funebris) sing duets; they sing equally complex
in colonies, and males seem quite tolerant of other birds songs and have similar song repertoire sizes. Based on
except in the immediate vicinity of their nests, so song this information, we might predict that the song nuclei
in this species seems to function largely in mate attrac- would be the same size in male and female bush shrikes.
tion. We’ll look first at what is known about the neural In contrast to our prediction, however, males still have
substrates of song. larger song nuclei (HVC and RA) than females and have
Researchers have identified several areas of the brain more neurons in these areas (Gahr et al. 1998). Examples
involved in song. These areas are called song nuclei (a such as this one suggest that the relationship between
nucleus is a collection of neurons that is anatomically dis- singing behavior and morphology of the song system in
tinct), and they are linked to one another in neural path- the brain may be species-specific or at least more com-
ways (Figure 8.17a). Two major pathways have been plicated than the zebra finch example suggests (Bolhuis
identified. The posterior pathway (shown by white and Gahr 2006).
arrows in Figure 8.17a) seems to be involved in the pro- How do sex differences in song nuclei arise? The
duction of song, and the anterior pathway (shown by chromosomal difference between the sexes is thought to
black arrows in Figure 8.17a) in the acquisition of song. initiate the dramatic differences in male and female brain
Still other areas of the brain function in the perception structure. In birds, males have two Z chromosomes, and
(hearing) of song and in the storage of memories of song females have one Z and one W chromosome. The dif-
heard when young (Bolhuis and Eda-Fujiwara 2003; ferential action of genes encoded on the sex chromo-
Gobes and Bolhuis 2007). The overall size of song nuclei somes is thought to prompt sexual differentiation of the
in the two major pathways is sometimes linked to the size song system in zebra finches. However, much remains
 Pulling It All Together—The Development of Bird Song 173

a HVC

NCM CLM

FIGURE 8.17 (a) The songbird CMM

RA L
brain showing the nuclei LMAN
involved in song. Although all
structures are found on both
sides of the brain, those on only
one side are shown for simplic-
ity. White arrows show the pos-
terior pathway thought to be
involved in song production;
black arrows show the anterior DLM
pathway thought to be involved
in song acquisition. For com-
pleteness, areas involved in
auditory processing and percep-
tion are also shown (field L,
NCM, CLM, and CMM). nXllts Area X
Abbreviations: CLM, caudal lat-
eral mesopallium; CMM, caudal
medial mesopallium; DLM, nucleus dorsolateralis ante- b 2.5
rior, pars medialis; HVC, a letter-based name; field L;
LMAN, lateral magnocellular nucleus of the anterior
nidopallium; NCM, caudal medial nidopallium; nXIIts, 2.2
tracheosyringeal portion of the nucleus hypoglossus;
Volume in mm3

RA, robust nucleus of the arcopallium. The nomencla- 1.0

ture of the avian brain has been revised (Reiner et al.
2004; see Reiner, Perkel, Mello, and Jarvis 2004 for spe-
cific reference to songbirds), so don’t be surprised when
the names used in this figure differ somewhat from 0.5
those in earlier publications. (b) Sexual dimorphism in
the brains of zebra finches. Note that four of the brain
regions involved in song (Area X, HVC, RA, and nXIIts)
are substantially larger in males than in females. No
Area X HVC RA nXIIts Rt SpM
such difference is found in two regions (Rt and SpM)
that are not involved in song. (a: Modified from Bolhuis Neural regions involved Neural regions not
2008. b: Nottebohm and Arnold 1976.) in song involved in song

to be learned about which genes are expressed at which responding areas in the normal male brain (Figure 8.18).
times in development. Hormones also play a role in Thus, estrogen appears to masculinize the brain. What
sexual differentiation of the song system. Estrogen, in is the source of the estrogen that masculinizes the brain?
particular, seems critical. A remarkable group of exper- Most evidence points to the brain itself. (Recall from
iments with zebra finches illustrates the powerful effects Chapter 7 that steroid hormones are produced by the
of estrogen on parts of the neural system that control gonads, adrenal glands, and the brain.) One hypothesis
singing behavior (Gurney and Konishi 1980). If pellets for sexual differentiation of the zebra finch brain suggests
containing estrogen are implanted in newly hatched that genes on the sex chromosomes cause differences in
male zebra finches, they have no effect (i.e., in adulthood the brain’s synthesis of estrogen, and the higher levels of
the brains and songs of estrogen-treated males are sim- estrogen in male brain tissue masculinize the song sys-
ilar to those of untreated males). In contrast, treatment tem (Wade and Arnold 2004).
of nestling females with estrogen results in enlarged song Hormones, we will see, have both organizational
areas in the brain. These areas are larger than those of and activational effects on singing behavior. Indeed,
untreated females, though they are not as large as cor- experiments have revealed that the effects of hormones
174 Chapter 8 / The Development of Behavior

els of steroid hormones occur during the breeding sea-

Volume of song area HVC (mm3)
0.3
son because increasing day length induces increases in
the size and activity of the gonads). Increased levels of
0.2 testosterone and its metabolites prompt seasonal
changes in the structure and physiology of song nuclei,
which in turn promote singing behavior (Brenowitz
0.1 2004; Brenowitz et al. 2007). Indeed, during the breed-
ing season, the volumes of song nuclei (HVC, RA, X,
and nXIIts) may increase by up to 200%! Beyond over-
0.0
Normal male Female treated Normal female all changes in volume, cellular changes in song nuclei
(N = 4) with estrogen (N = 4) also occur in seasonally breeding birds. These changes
at hatching
(N = 3)
include increases during the breeding season in neuron
number, density, metabolic capacity, and spontaneous
FIGURE 8.18 Treatment with estrogen at hatching neurophysiological activity. (Recall from Chapter 6 that
results in enlarged song areas in the brains of female we once thought that new neurons could not form in
zebra finches. Here, the volume of the song area HVC adult animals.) Gonadal testosterone and its metabolites
is substantially larger in estrogen-treated females than may act directly on the song nucleus called the HVC,
in normal females, but it is still somewhat smaller than
which then influences other song nuclei.
that of normal males. Estrogen is thus involved in the
What happens with the approach of the nonbreed-
masculinization of the brains of songbirds. (Modified
from Gurney and Konishi 1980.) ing season? As the hours of daylight shorten and levels
of testosterone and its metabolites decline, the song sys-
tem in the brain regresses. Regression of song nuclei can
on the brain and singing behavior of zebra finches do not occur quite rapidly, revealing the amazing plasticity in
end around the time of hatching. Although the typical structure (and function) of the avian brain. For exam-
male has a masculinized brain, elevated levels of testos- ple, when adult male Gambel’s white-crowned sparrows
terone in the bloodstream are necessary to stimulate (Zonotrichia leucophrys gambeli) were maintained in the
singing in adulthood. Similarly, those females that laboratory under conditions consistent with the breed-
receive estrogen implants soon after hatching, and thus ing season (long day length and implants containing
develop enlarged song areas, do not sing in adulthood testosterone) and then castrated and shifted to short day
unless testosterone is administered (Gurney and Konishi lengths (testosterone implants were removed at the time
1980). Females that do not receive estrogen implants at of castration), the volume of the HVC regressed 22%
hatching, but do receive testosterone in adulthood, do within 12 hours of the withdrawal of testosterone
not sing. Apparently, early exposure to estrogen estab- (Thompson et al. 2007). With regression of the song
lishes a sensitivity to testosterone in the brains of exper- system, singing declines or ceases altogether. In species
imental females (and normal males), and exposure to in which singing declines but does not stop completely,
testosterone in adulthood stimulates song. In other the structure of song in nonbreeding birds may become
words, whereas estrogen early in life organizes the devel- more variable than song during the breeding season.
opment of a male song system, a high circulating level Regression of the song system during the nonbreeding
of testosterone in adulthood appears necessary to activate season has been interpreted as a way to reduce the ener-
singing behavior. Finally, the brains of birds are not getic costs of maintaining the system during the fall and
always sensitive to the organizing effects of estrogen. winter when many birds face significant energetic chal-
There appears to be a sensitive period, around the time lenges, including those associated with migration, low
of hatching, in which the brain is particularly sensitive temperatures, and low food availability.
to hormonal influences. During this window of time, We have seen that genetic information, the hor-
estrogen exerts its powerful effects on the developing monal milieu, and development of the nervous system all
nervous system. After that, the neural pathways that con- influence singing behavior. We can now ask what role
trol song cannot be switched to the male track. (See learning plays in the development of song.
Chapter 7 for a further discussion of the organizational
and activational effects of steroid hormones.) ROLE OF LEARNING
For many species of songbirds, the hormonal acti-
IN SONG DEVELOPMENT
vation of song in free-living adult birds occurs on a sea-
sonal basis. Song occurs most often or only during the Whereas zebra finches have frequently been used to
breeding season when levels of testosterone and its study the physiological bases for song development,
metabolites (the products of the breakdown of testos- white-crowned sparrows (Zonotrichia leucophrys), song-
terone, estrogen, and 5α-dihydrotestosterone) are birds from North America, have been a favorite subject
higher than during the nonbreeding season (higher lev- for those curious about the role of learning (Figure 8.19).
 Pulling It All Together—The Development of Bird Song 175

sensitive period can be extended beyond 50 days when

white-crowned males are exposed to live white-crowned
tutors rather than taped songs (Baptista and Petrinovich
1984). Thus, we see that environmental conditions can
influence the sensitive period for song learning (we
return to this topic shortly). The important point now,
however, is that, for song to develop normally, male
white-crowned sparrows must be exposed early in life to
the songs of adult males of their species.
In addition to hearing adult conspecific song,
young male white-crowned sparrows must also hear
themselves sing if they are to produce normal song. A
series of classic experiments revealed the critical role of
auditory feedback in song development (Konishi 1965).
Young male white-crowned sparrows that are exposed
to the songs of adult male conspecifics early in life and
then deafened before the onset of subsong produce a
rambling and variable song. Apparently, to develop full
song, a sparrow must be able to hear his own voice and
compare his vocal output to songs that he memorized
FIGURE 8.19 White-crowned sparrows, such as the male months previously. Further studies revealed that iso-
shown here, have often been used in studies of learning lated white-crowned males with intact hearing produce
and song development. songs that are somewhat more normal in structure than
those produced by males deafened prior to subsong
(Refer again to Figure 8.20.) Thus, if a white-crowned
The importance of learning to song development in sparrow can hear himself sing, he can produce a song
adult male white-crowned sparrows was demonstrated in with a few of the normal species-specific qualities.
isolation experiments conducted by Peter Marler and his Overall, however, the song of an isolated male is still
colleagues (Marler 1970; Marler and Tamura 1964).
Under natural conditions, a young male sparrow hears
the songs of his father and other adult males around him a kHZ Wild bird
5
during his first summer and autumn. During these first 4
few months of life, the young male produces only sub- 3
song, a highly variable, rambling series of sounds with 2
1
none of the syllables typical of the full song of adult
males. As the male’s first breeding season approaches, his
song begins to contain elements recognizable as white-
crowned sparrow syllables. This vocalization, which b Isolated bird
occurs in late winter or early spring of the male’s first 5
4
year, is called plastic song. The male continues to refine 3
his song and eventually his song crystallizes. Thus, at the 2
start of his first breeding season, he begins to produce 1
the full song characteristic of adult male white-crowned
sparrows. The songs of the male are copies (imitations)
of those he heard many months before.
c Deaf bird
Young male white-crowned sparrows reared in iso- 5
lation develop abnormal song (Figure 8.20). However, 4
if isolated males hear tapes of white-crown song during 3
2
the period from 10 to 50 days of age, they will develop 1
their normal species song. Young white-crowned spar-
0.5 1.0 1.5 2.0
rows that hear taped white-crown songs either before 10
Time (sec)
days of age or after 50 days do not copy it. Thus, under
the conditions of social isolation and “tape tutoring,” the FIGURE 8.20 Sonograms of songs that were produced by
sensitive period for song learning occurs between 10 and white-crowned sparrows (a) reared under natural condi-
50 days posthatching. There is some flexibility, however, tions, (b) reared in social isolation, and (c) deafened at
in the timing and duration of this sensitive period: the an early age. (Modified from Konishi 1965.)
176 Chapter 8 / The Development of Behavior

quite abnormal. Finally, although auditory feedback is TABLE 8.3 Phases of Song Development in
important for song development, it is not essential for Sparrows
the maintenance of song in male white-crowned spar-
rows. Males deafened after their songs have crystallized 1. Sensory phase
continue to produce relatively normal songs for at least a. Acquisition—song learned
one year after deafening (Konishi 1965). There are, b. Storage—song retained in memory
however, species in which auditory feedback appears 2. Sensorimotor phase
essential to the maintenance of normal song in adult-
a. Retrieval and production—motor rehearsal of learned
hood. For example, deafening of adult males results in song
the deterioration of song in zebra finches (Nordeen and
b. Motor stabilization—song crystallized into final adult
Nordeen 1992), Bengalese finches (Woolley and Rubel
song
1997), and canaries (Nottebohm et al. 1976). Species
differences in the importance of auditory feedback to Source: Marler (1987).
maintenance of adult song may relate to species differ-
ences in song complexity or the ability to modify songs
in adulthood (Brainard and Doupe 2000). It is also pos-
sible that the conflicting results relate more to the age communication, song sharing may allow neighboring
at which males were deafened, with auditory feedback males to avoid territorial battles that are costly in time
being less important to song maintenance in older males and energy. In support of this idea, there is evidence
(Konishi 2004; Lombardino and Nottebohm 2000). that sharing songs is beneficial to male song sparrows:
The results of these and other laboratory studies sug- degree of song sharing is positively correlated with
gest that song development in sparrows consists of two male survival (Wilson et al. 2000) and territory tenure
phases—a sensory phase, during which songs are learned in this species (Beecher et al. 2000). Interestingly, a
and stored in memory, and a sensorimotor stage, when comparison of plastic and crystallized versions of
singing a learned song actually begins. In the sensory retained songs revealed that some of the retained songs
phase, sounds that are heard during the first few months are modified to make them less similar to the songs of
of life are stored in memory for months without rehearsal. neighbors, perhaps promoting individual recognition.
During this time, young males produce only subsong, Thus, song learning in young male song sparrows
which does not involve retrieval or rehearsal of previously seems to involve a delicate balance between two
learned material. In the species of sparrows studied to opposing forces—the tendency to copy and keep the
date, the sensorimotor phase begins at about seven or songs of neighboring males and the need for each male
eight months of age. At this time, birds retrieve a learned to make at least some of his songs individually distinc-
song from memory and rehearse it, constantly matching tive (Nordby et al. 2007).
their sounds to the sounds they memorized months ear- So far, we have confined our discussion of song
lier. Over the next few months song patterns crystallize, learning to the well-studied sparrows. However, a great
and males begin to produce full song at the start of their deal of variation exists across species. This variation is
first breeding season. The final adult song will remain vir- especially evident when it comes to sensitive periods in
tually unchanged for the rest of the sparrow’s life. The song learning, our next topic.
phases of song development are outlined in Table 8.3.
As young males move from the plastic phase of SENSITIVE PERIODS
song to the crystallized phase, they sometimes drop
IN SONG LEARNING
songs from their plastic song repertoire. For example,
in the months leading up to their first breeding season, We learned from experiments on male white-crowned
young male song sparrows (Melospiza melodia) drop one sparrows raised in isolation and tutored with tape record-
to three songs from their plastic song repertoire, even- ings of white-crown songs that the sensitive period for
tually ending up with a crystallized song repertoire of song learning extends from 10 to 50 days of age. We also
8 to 11 songs. The young males do not drop songs at learned that the sensitive period could be extended by
random. Indeed, the discarded songs match the songs exposing young males to live conspecific tutors instead
of fewer neighboring males or match them less well of taped song. These results suggest some degree of flex-
than the songs that are retained. Sharing songs may ibility or variability in sensitive periods and indicate that
facilitate communication among neighboring males. sensitive periods can be influenced by environmental
For example, shared songs are a reliable signal of a conditions. It has been found that there is considerable
young male being a “local” and perhaps less likely to variation both among and within species of songbirds in
trespass or to attempt to take over another male’s ter- the time course for song learning. Here, we consider
ritory (Beecher and Brenowitz 2005). By facilitating some of that variation and what it might mean.
 Pulling It All Together—The Development of Bird Song 177

The length and timing of the sensitive period for song of their foster parents. But field and laboratory obser-
learning vary greatly across species. Canaries, for exam- vations indicate that this is not the case at all; when male
ple, have been called lifelong learners (or an open-ended cowbirds become sexually mature, they sing the cowbird
species) as a result of their ability to continually revise song. This raises an interesting question: when does the
their songs throughout adulthood. In contrast, white- sensitive period for song learning occur in brood para-
crowned sparrows and zebra finches have a more sites such as cowbirds? After fledging from the nests of
restricted period of learning that occurs during the first their foster parents in late summer, young brown-
weeks or months of life; these species are called age- headed cowbirds join flocks containing other
limited learners (or closed-ended species). The discov- cowbirds—is this when they learn their song? But this
ery of some degree of song plasticity in adulthood in too can be a difficult time for song learning because by
some age-limited learners has prompted the suggestion this time, many adults have already left breeding
that these two developmental patterns—lifelong learn- grounds, and those males that remain have ceased
ers and age-limited learners—are not dichotomous (i.e., singing. Given these details of the cowbird life history,
they are not the only two possible options), but instead it should come as no surprise that song learning is
points on a continuum (Brenowitz 2004). delayed in brown-headed cowbirds relative to song spar-
Species differences in length and timing of sensitive rows: male cowbirds do not perfect their song until their
periods for song learning often reflect species differ- second winter. This delay gives them sufficient time to
ences in life history. For example, we have described hear as yearlings the songs of adult male conspecifics
how young male song sparrows learn to sing by listen- (O’Loghlen and Rothstein 1993).
ing to and memorizing the songs of adult males that Sensitive periods for song learning also vary within
they hear during the first few months of life. These a species. Individual differences in the timing of song
young males then prune their plastic song repertoires learning are commonly reported in both field and lab-
to produce full adult song by their first breeding season, oratory studies, and for some well-studied species, the
when they are not quite one year old. Now consider the variation can be understood in the context of their life
odd life history of brown-headed cowbirds (Molothrus history. In the marsh wren, Cistothorus palustris, the time
ater). These birds are brood parasites. Females of this course for vocal learning is influenced, in part, by a
species lay their eggs in the nests of other bird species, parameter of the physical environment—photoperiod
thereby relinquishing all parental duties to foster par- (Kroodsma and Pickert 1980). In the northeastern
ents (Figure 8.21). If young male cowbirds relied on United States, young wrens hatch from mid-June to
learning characteristics of their songs by listening to late August. Males that hatch in June experience long
male birds in the vicinity of the nest during their first day lengths, whereas those that hatch in August expe-
few weeks of life, they would most likely sing the song rience relatively shorter day lengths. Does exposure to
different photoperiods after hatching affect the time
course for song learning in this species? Laboratory
experiments revealed that males raised under day
lengths that simulate hatching late in the breeding sea-
son (August) learned songs the following spring,
whereas males raised under conditions that simulate
early hatching (June) did not. Thus, the sensitive period
for late-hatched males appears to extend beyond that
for early-hatched males. A flexible sensitive period
determined, in part, by environmental conditions may
make sense for some songbirds, especially those species
in which late-hatched young tend to disperse further
than early-hatched young. Such flexibility permits the
learning of song the following spring in the area in
which the late hatched males have settled (recall that
singing the songs of neighbors is a successful strategy
for many songbirds).
Thus, we see that variation in the length and tim-
ing of sensitive periods for song learning both among
FIGURE 8.21 Brown-headed cowbirds lay their eggs in and within species is perhaps best viewed in the context
the nests of other species, leaving their offspring in the of life history. Furthermore, such variation probably
care of the host parents. Here, a host parent, a yellow serves to maximize the chances that each male develops
warbler, feeds a cowbird nestling. species-appropriate singing behavior.
178 Chapter 8 / The Development of Behavior

OWN-SPECIES BIAS a
IN SONG LEARNING
When given a choice, most young male birds learn their
songs from members of their own species. For example,
male white-crowned sparrows copied only the songs of
their own species when tutored with tapes of the song
sparrow and white-crowned sparrow (Marler 1970). This
preference is called own-species bias, and it also influ-
ences the speed and accuracy with which young males
learn. Typically, birds learn the songs of their own
species more rapidly and accurately than those of a dif-
ferent species, although this tendency varies across
species and with method of song presentation. In light
of evidence that own-species bias characterizes song
learning, researchers are now investigating the neural
basis for this bias. These research efforts often involve b 100
recording responses from neurons in the auditory area
known as field L (refer, again, to Figure 8.17a) while
exposing anesthetized male birds to a variety of sounds,
Percentage of songs that release the
including conspecific song (e.g., Grace et al. 2003).
copulatory posture in females
When a young male is exposed to tape recordings
of conspecifics, live tutors of another species can some-
times override the bias for learning conspecific songs.
When young white-crowned sparrows are presented 50
with visible, singing song sparrow tutors and can hear
the songs of white-crowned sparrows in the background
only, they learn the song of the song sparrow (Baptista
and Petrinovich 1986). That young male sparrows learn
the song of the adult that they are allowed to interact
with, even if this adult belongs to another species,
emphasizes the importance of social interactions in the
song-learning process. Next we examine how social fac- 0
tors influence the development of song. Auditory and Visual Visual Direct
visual isolation isolation contact contact

Rearing condition of males with respect

SOCIAL FACTORS to male conspecifics
AND SONG DEVELOPMENT
FIGURE 8.22 Female cowbirds prefer the songs of males
As described previously, young cowbirds are reared by reared in isolation to those of males reared socially.
parents of different species, and they begin to associate (a) Females display their preference for songs by assum-
with conspecifics after fledging, when they join cowbird ing the copulatory posture. (b) Males without visual or
flocks. Several studies have revealed that interactions direct contact with dominant males produce more
with conspecific males and females in flocks on breed- potent songs than do males that have visual or direct
ing and wintering grounds influence song development contact. In cowbirds, then, young males learn to modify
in cowbirds. their songs to avoid attack by adult males. (Drawn from
the data of West et al. 1981.)
Interest in the role of social interactions in the devel-
opment of song in cowbirds grew from the finding that
the songs of male cowbirds reared in isolation from male
conspecifics were substantially more appealing to female
cowbirds than the songs of males reared in groups, as an established colony of cowbirds. Group-reared males
measured by the willingness of females to display a cop- kept a low profile in their new environment; they did not
ulatory posture (Figure 8.22) (West et al. 1981). Why are sing in the presence of resident males and were not
the songs of isolates more potent than those of normal attacked. In contrast, the isolates repeatedly sang their
males? The answer comes from observations of what high-potency song in the presence of resident males and
occurred when group-reared males and isolation-reared females, and the resident males responded by attacking
males were individually introduced into an aviary with and sometimes killing the energetic songsters.
 Pulling It All Together—The Development of Bird Song 179

Apparently, males living in mixed-sex groups, such as the repertoires that are biased toward the preferences of
flocks that occur in nature, learn to suppress their song their female companions. More recently, it has been
and thereby avoid attacks from dominant males. shown that female cowbirds also influence the rate of
Subsequent observations of young males maintained in song development in young males: males housed with
flocks with and without adult males reveal that adult females from their own population progress from sub-
males influence not only the potency of song eventually song to plastic song to stereotyped (full) song more
produced by young males, but also the rate at which song rapidly than males housed with females from a different
develops: specifically, presence of adult males slows the population (Smith et al. 2000).
rate of song development in young male cowbirds Because female cowbirds do not sing, their influence
(White et al. 2002). on male song development cannot be through song.
What role might female cowbirds play in shaping Instead, they may encourage the male’s singing by a sim-
the songs of males? To examine this question, King and ple display called the wing stroke, in which one or both
West (1983) used two different subspecies of cowbirds, wings are moved rapidly away from the body (West and
one found in the area of North Carolina (M. ater ater) King 1988). These wing strokes are not doled out indis-
and the other around Texas (M. ater obscurus). We will criminately: approximately 94% of the males’ songs
refer to these cowbirds as subspecies A and subspecies elicit no visible change in the behavior of female cow-
O, respectively. Male cowbirds of subspecies A were birds, and on average, a single wing stroke occurs for
hand-reared in acoustic isolation until they were 50 days every 100 songs. It seems, then, that male cowbirds
old. At this time they were housed for an entire year with adjust their songs to the whims of their audience, mold-
(1) individuals of another species (canaries or starlings), ing their song structure to avoid conflict with dominant
(2) adult female cowbirds of subspecies A, or (3) adult males and to stimulate available females. Just how they
female cowbirds of subspecies O. At the end of this achieve this delicate balance in natural flocks is an inter-
period, the researchers examined the songs of the males. esting question.
The three groups of males of subspecies A had developed Social factors are also important in shaping the
remarkably different songs (Figure 8.23). Males that songs of other songbird species. As mentioned previ-
were housed with canaries or starlings had diverse reper- ously, young male songbirds often drop a few songs from
toires, singing a mix of songs from the two cowbird sub- their plastic song repertoire along the way to reaching
species, along with imitations of the songs of other their final song repertoire. This so-called selective attri-
species in their cage. Whereas males housed with cow- tion phase is thought to occur at the start of the male’s
bird females of subspecies A sang all A songs, those first breeding season when he seeks to establish a terri-
housed with cowbird females of subspecies O sang pre- tory and engages in countersinging interactions with
dominantly O songs. Male cowbirds thus develop song neighboring males (Nelson and Marler 1994). During
these interactions, the young male attempts to match the
songs of his neighbors. Over time, he drops those songs
that are the poorest matches. These direct interactions
Percentage of subspecies O song

100
with other males at the start of the breeding season
therefore shape song development. More recent evi-
dence indicates that some young songbirds choose the
songs to keep in their final repertoire not only based on
50 the singing interactions in which they directly partici-
pate, but also by eavesdropping on the singing interac-
tions of other males. Indeed, at least under laboratory
conditions, young male song sparrows seem to learn
more by eavesdropping (indirect social interactions) than
0 by direct social interactions, possibly because overheard
Other species Subspecies A Subspecies O
females females interactions are less threatening (Beecher et al. 2007;
Companions of singers Burt et al. 2007). We see, then, that song development
can be influenced by direct and indirect social factors.
FIGURE 8.23 Female cowbirds influence the develop-
ment of song in males. When males of subspecies A are
housed with either subspecies A or subspecies O A DIVERSITY OF SONG
females, they mold their song to the preferences of LEARNING STRATEGIES
their female companion. Note that subspecies A males
housed with females of their own subspecies do not sing As studies of different species of songbirds have accu-
any subspecies O song, but those housed with sub- mulated over the years, it has become clear that song-
species O females sing subspecies O songs predomi- birds exhibit diverse song learning strategies (reviewed
nantly. (Drawn from the data of King and West 1983.) by Beecher and Brenowitz 2005). We have seen, for
180 Chapter 8 / The Development of Behavior

example, that the period within which young males learn mallard ducklings follow their mother, and most worker
song ranges from a relatively brief sensitive period early ants perform their brood care responsibilities in an
in life to learning throughout a lifetime. We described admirable fashion. Why do such complex patterns of
how social isolation produces abnormal song in white- behavior develop so reliably when there seems to be so
crowned sparrows; however, while this is true for many much room for error?
other species of songbirds, it is not true for all species The developmental processes appear to be capable
(Kroodsma et al. 1997; Leitner et al. 2002). Diversity of buffering themselves against potentially harmful
also exists in the number of songs that birds learn. While influences to produce functional adults. This buffering
the song repertoires of most species are small (less than capacity is called developmental homeostasis. A buffer
five songs) to moderate (about ten songs), some reper- is something that dampens drastic changes. We can look
toires are truly astronomical in size (over 2000 songs). around us and see that despite diverse experiences,
There also is variation in how closely young males copy behavioral development in most individuals proceeds in
tutor songs. While males in some species faithfully copy a very predictable and reliable manner. In effect, the
(imitate) tutor songs, males in other species modify tutor developmental process demonstrates a certain stability
songs or invent species-typical songs that bear no resem- and resilience in the face of a host of constantly chang-
blance to tutor song. Song learning strategies among ing variables, including, as we have seen, many of those
songbirds clearly differ along several dimensions. One we introduce experimentally. Our final example, meant
challenge today is to understand the diversity of song to illustrate the resilience of the developmental process
learning programs in the context of songbird phylogeny and the efforts being made today to improve the lives of
with the aim of determining whether certain programs captive nonhuman primates, concerns the recovery of
have evolved within certain lineages (Beecher and chimpanzees from years of profound social isolation.
Brenowitz 2005).

REHABILITATION OF CHIMPANZEES
STOP AND THINK AFTER LONG-TERM ISOLATION
Spencer et al. (2007) raise the issue of how best to raise Michaela Reimers and colleagues (2007) studied 13 male
young songbirds brought to wildlife rehabilitation hospi- chimpanzees (Pan troglodytes) that had been separated as
tals whose aim is to ultimately return the birds to their infants from their mothers and conspecifics and
natural habitats. We have discussed the effects of rearing
imported from Africa to a laboratory in Europe between
conditions on the development of song and the impor-
1976 and 1986 for use in hepatitis and HIV research.
tance of song to subsequent survival and breeding success
(e.g., song functions in territory maintenance and mate When being used in biomedical research, the chim-
attraction). If you were charged with developing guide- panzees were individually housed in indoor cages at the
lines for housing and tutoring young songbirds brought laboratory. Although all 13 males experienced early
to your rehabilitation facility, what factors would you con- maternal separation, spatial confinement, and lack of
sider in making your recommendations? control over their daily lives, they differed in the tim-
ing at which social deprivation began and the total dura-
tion of social isolation. Six males arrived at the research
laboratory when they were between one and two years
DEVELOPMENTAL of age, and did not experience peer housing before
HOMEOSTASIS placement into individual cages. These males were cat-
egorized as early deprived (ED). The remaining seven
Throughout this chapter we have seen that experience males arrived at the laboratory when they were between
plays an important role in behavioral development. three and four years of age, and then spent one year in
Having said this, we might also note that the manipula- a peer group before being separated into individual
tions used to demonstrate the sensitivity of the devel- cages. These males were categorized as late deprived
oping animal to external influence are often quite severe. (LD). At the time of the study by Reimers and col-
Such manipulations typically involve rearing in isolation leagues, ED males were about 22 years old and had
or rearing by another species; neither condition is likely spent, on average, 21 years in isolation, and LD males
to be encountered by most animals under natural con- were about 19 years old and had spent, on average, 17
ditions. Indeed, despite the fact that in their natural envi- years in isolation. The study consisted of four phases:
ronments, individuals within a species often develop laboratory (three months); transport to the new facility
under a diverse array of physical and social conditions, (seven days, which included the three-hour trip and sub-
the vast majority of adults display species-typical, nor- sequent adjustment period); habituation to the new
mal behavior. Under field conditions, after all, most facility (six weeks); and resocialization (one year). The
white-crowned sparrows sing white-crown songs, most new facility contained a spacious indoor enclosure with
 Developmental Homeostasis 181

tree trunks and other wooden structures. During the the food drawer and to enter the novel environment
habituation phase, males were individually and repeat- (Figure 8.24a). In addition, whereas chimpanzees in the
edly introduced to the indoor enclosure. During the ED group explored the new toy and environment in a
resocialization phase, males were introduced into the cautious manner, those in the LD group exhibited
enclosure with a small number of other males for a few exploratory behavior that was active and bold. ED and
hours; the precise length of a session was determined by LD chimpanzees also differed in their social behavior,
the well-being and progress of the participants, and thus with ED males initiating and receiving fewer social inter-
was effectively controlled by them. Socialization ses- actions than LD males. When considering all of the
sions were interspersed with resting periods in individ- study chimpanzees, stress hormone levels increased dur-
ual cages, and the composition of the social groups ing the transport phase, declined during the habituation
changed over days. Eventually, ten of the males became phase, and declined still further during the resocializa-
members of one all-male group, and the remaining tion phase (Figure 8.24b). Overall, levels of stress hor-
three males became members of mixed-sex groups. mones were higher in ED males than in LD males, and
Reimers and colleagues monitored levels of glucocorti- the patterns of change across phases also differed
coids (stress hormones) in fecal samples across the four between the two groups. Glucocorticoid levels in ED
phases of the study, conducted two standardized novelty males remained fairly high during habituation and then
tests, and observed social interactions in the all-male declined significantly during resocialization to below
group during the first year of group living. The first habituation levels. In contrast, glucocorticoid levels in
novelty test was conducted during the laboratory phase, LD males were already lower during habituation and
and consisted of scoring the response of each chimp to then declined still further during resocialization to lev-
a novel plush toy placed in the food drawer of its indi- els below those for the laboratory phase. Taken together,
vidual cage. The second novelty test was conducted dur- the behavioral and hormonal data indicate that chim-
ing the habituation phase and consisted of scoring the panzees separated from their mothers and conspecifics
response of each chimp when first introduced to the new at a younger age and kept in isolation for more years are
and spacious indoor enclosure. less explorative, less social, and more susceptible to stress
The timing and duration of social deprivation influ- than chimpanzees that experienced somewhat less severe
enced the behavioral and hormonal responses of the deprivation. In addition, the patterns of change in lev-
chimpanzees. Chimpanzees in the ED group took longer els of stress hormones across the study suggest that
than those in the LD group to take the novel toy out of chimps are better able to cope with social stressors (as
Log cortisol metabolites (ng/g feces)

a 750 b
ED 2.4 ED
LD LD
Latency (sec)

500 2.2

250 2.0

0 1.0
Laboratory

Transport

Habituation

Resocialization

Novel toy Novel environment

Study phases

FIGURE 8.24 Timing and duration of social isolation influence the behavioral and hormonal profiles of male chim-
panzees previously used in biomedical research. (a) Chimpanzees in the early deprivation group (separated from
mother and conspecifics at 1 to 2 years of age; ED) take longer then those in the late deprivation group (separated
from mothers and conspecifics at 3 to 4 years of age; LD) to explore a novel toy and a novel environment. (b) Levels of
stress hormones (as measured by metabolites of the hormone cortisol in feces) also differ between ED and LD groups,
although both groups showed a decreased stress response to resocialization. Thus, chimpanzees isolated for many
years can be rehabilitated to live less stressful, social lives, thereby illustrating the resilience of developmental
processes. (From Reimers et al. 2007.)
182 Chapter 8 / The Development of Behavior

during the resocialization phase) than environmental tors is perhaps best illustrated by the development of sex-
stressors (as during the habituation phase)(refer again to ually dimorphic patterns of behavior, such as singing in
Figure 8.24b); the authors suggest that this may reflect birds. In general, only male birds sing, and in some
the chimps’ perceived greater control over social situa- species this is reflected in their neuroanatomy. The
tions than environmental situations. The main point for developmental basis for sex differences in the brains and
our discussion of developmental homeostasis, however, vocal behavior of songbirds is the chromosomal differ-
is that chimpanzees that have experienced profound ence between the sexes. This genetic difference dictates
social isolation, in some cases lasting two decades, can patterns of secretion of steroid hormones that, in males,
recover and live less stressful lives with conspecifics. lead to the growth and differentiation of the regions of
Here, then, we have an example of careful and thera- the brain involved in song. In addition to neural and hor-
peutic rehabilitation revealing the resilience of develop- monal influences, experience plays an important role in
mental processes in our closest relatives. the development of singing behavior. If they are to pro-
duce normal songs in adulthood, young males of many
species must listen to the songs of adult conspecifics and
SUMMARY must be able to hear themselves sing. Social interactions,
whether direct (e.g., countersinging) or indirect (e.g.,
Patterns of behavior appear, disappear, and alter in form eavesdropping), also influence the development of song.
as animals develop. There are several causes of behav- In male songbirds, the pulse of steroid hormones that
ioral change, and one of the most significant is the devel- affects the developing nervous system and the experience
opment of the nervous system. During embryonic life, of hearing conspecific songs are not always capable of
patterns of behavior often emerge in parallel with devel- influencing the development of singing behavior. Indeed,
opment of the sensory and neural structures necessary to be effective, the change in hormone level and the expe-
for their performance. Sometimes neuronal growth, rience of listening to conspecifics often must occur dur-
death, or alterations in structure can be linked to devel- ing somewhat restricted periods of time, when the young
opmental changes in a particular behavior. Behavioral bird is particularly sensitive to such influences. These
change may also come about through the development periods of enhanced sensitivity to environmental stimuli
of specific morphological structures. In developing pad- are called sensitive periods, and they characterize diverse
dlefish, the change in feeding behavior from picking behavioral phenomena, including filial imprinting (the
select items out of the water column to indiscriminate response of young precocial birds to follow their mother),
filter feeding parallels the development of gill rakers. sexual imprinting (learning the characteristics of an
Changes in hormonal state can also trigger behavioral appropriate mate), and maternal attachment (the devel-
change, as in intrauterine effects in mice and the depo- opment of a bond between a mother and her offspring
sition by avian mothers of varying amounts of hormones such that she selectively cares for her own young). There
in their eggs. Finally, experience also affects the devel- are tremendous differences among species in the timing
opment of behavior. For example, play is an experience of sensitive periods for the development of specific behav-
that is vital to behavioral development in many species. iors, and even within a species there is flexibility in the
It is expressed in several ways: social play; locomotor play time course of sensitive periods.
(exercise); and object manipulation. Hypotheses for the We have said that experience, especially that occur-
function of play include the following: it improves phys- ring during a sensitive period in the young animal’s life,
ical condition, it is important in developing social skills can profoundly influence behavioral development.
and bonds, and it helps develop and maintain cognitive Under normal circumstances, however, animals
skills. Neural, hormonal, morphological, and experien- encounter a wide variety of physical and social condi-
tial causes of behavioral change often interact during the tions, and most develop normally. The ability of devel-
continuous interplay between the developing animal and opment to proceed in a normal fashion in the face of
its internal and external environment. environmental perturbation is called developmental
Genetic factors also enter into the complex interac- homeostasis. The successful rehabilitation of research
tion between the developing animal and its environment. chimpanzees subjected to long-term social isolation
The interaction between genetic and environmental fac- speaks to the resilience of behavioral development.
PART TWO

Survival

183
This page intentionally left blank
 9
Biological Clocks

Defining Properties of Clock-Controlled Rhythms Imagine for a moment that you have a pet hamster, a
Persistence in Constant Conditions friendly fellow who quietly shares your bedroom.
Entrainment by Environmental Cycles However, late one Thursday night you are studying for
Temperature Compensation a test but finding it hard to concentrate because a ham-
ster is most active during night hours, and his running
Rhythmic Behavior wheel is squeaking. So, you place the hamster and his
Daily Rhythms cage in your closet and continue to study. The next
Lunar Day Rhythms morning, you take the test and then leave for the week-
Semilunar Rhythms end. On Monday night you are back, and you notice a
Monthly Rhythms squeaking in your closet. You have forgotten about the
Annual Rhythms hamster. He has had plenty of food and water, but he has
been in the dark for three days. As you retrieve him, you
The Clock Versus the Hands of the Clock
notice that he begins to run on his wheel at about the
Advantages of Clock-Controlled Behavior same time as he normally did.
Anticipation of Environmental Change How could he know what time it was? We attribute
Synchronization of a Behavior with an Event That the ability to measure time without any obvious environ-
Cannot Be Sensed Directly mental cues to an internal, living clock. When any ham-
Continuous Measurement of Time ster is experimentally sequestered in the constant darkness
Adaptiveness of Biological Clocks and temperature of the laboratory so that each turn of its
running wheel can be recorded automatically for months
Organization of Circadian Systems or even years, a record similar to the one shown in Figure
Multiple Clocks 9.1 usually results. Notice that the hamster woke up
Coordination of Circadian Timing almost exactly 12 minutes later each day during the entire
Human Implications of Circadian Rhythms study. Its bouts of activity alternate with rest with such
regularity that it is often described as an activity rhythm.
Jet Lag
The ability to measure time is common not just in ham-
Human Health
sters but also in most animals. In fact, biological clocks

185
186 Chapter 9 / Biological Clocks

a variations in light intensity, temperature, relative

humidity, barometric pressure, geomagnetism, cosmic
radiation, and the electrostatic field. The earth also
rotates relative to the moon once every lunar day
(24.8 hours). The moon’s gravitational pull draws the
water on the earth’s surface toward it, causing it to “pile
up” and thus resulting in high tide. These tidal cycles
cause dramatic changes in the environment of intertidal
organisms—flooding followed by desiccation when
exposed to air. The relative positions of the earth, moon,
and sun result in the fortnightly (biweekly) alternation
between spring and neap tides, as will be explained
shortly. The moon revolves about the earth once every
lunar month (29.5 days), generating changes in the
b 1
intensity of nocturnal illumination and causing fluctu-
ations in the earth’s magnetic field. Finally, the earth,
5 tilted on its axis, circles the sun, causing the progression
Days of experiment

of the seasons, with its sometimes dramatic alterations

in photoperiod and temperature.
10 Although the environmental modifications may be
extreme, they are generally predictable. Often it is
advantageous to gear an activity to occur at a specific
15
time relative to some rhythmic aspect of the environ-
ment. So, biological clocks may have evolved as adapta-
tions to these environmental cycles. Biological clocks
12 14 16 18 20 22 24 2 4 6 8 10 12
also provide a mechanism to synchronize various inter-
Time of day (hours)
nal processes with other internal processes. Thus, inter-
FIGURE 9.1 (a) A hamster in a running wheel equipped nal synchronization may be another adaptive value
to record each turn of the wheel shows periodic bouts (Sharma 2003).
of activity that alternate with rest. (b) In constant dim In this chapter, we will see examples of the various
light, the cycle length of this activity rhythm is slightly approaches to the study of behavior discussed in pre-
longer than 24 hours. For each rotation of the running vious chapters: Rhythms are so pervasive that they have
wheel, a vertical line is automatically made on a chart at piqued the interest of scientists studying their adaptive
the time of day when the activity occurred. In this
value and evolution, genetic underpinnings, the hor-
record, the bouts of activity were so intense that the
vertical lines appear to have fused, forming dark hori-
monal control, and the neural control. We are now
zontal bands. Notice that although the animal had no beginning to understand the connections among these
light or temperature cycle as a cue to the time, it would levels of analyses. First, we will describe some rhyth-
awaken about 12 minutes later each day. mic behaviors and then the properties of the clock that
drives them.

have been found in every eukaryotic organism tested, as

well as in cyanobacteria (Paranjpe and Sharma 2005; DEFINING PROPERTIES
Woelfle and Johnson 2006). OF CLOCK-CONTROLLED
The rhythmical nature makes sense in the light of RHYTHMS
evolutionary principles. Life evolved under cyclical con-
ditions, and the ecological conditions under which ani- Like any good clock, biological clocks measure time at
mals find themselves differ tremendously at different the same rate under nearly all conditions, and they have
times of the cycle. As we discuss in more detail later in mechanisms that reset them as needed to keep them syn-
the chapter, it is adaptive to predict upcoming changes in chronized with environmental cycles.
a cycle—for example, that darkness is coming or that
winter is nearly upon us—rather than just respond to
PERSISTENCE IN CONSTANT
these events as they occur.
Every living thing is subjected to the regularly vary-
CONDITIONS
ing environmental conditions on earth orchestrated by A defining property of clock-controlled rhythms is that
the relative movements of the earth, moon, and sun. As cycles continue in the absence of environmental cues
the earth spins on its axis, life is exposed to rhythmic such as light-dark and temperature cycles. This means
 Defining Properties of Clock-Controlled Rhythms 187

a b c Sometimes this point is emphasized by describing the

circadian period length in constant conditions as free-
running, implying that it is no longer manipulated by
environmental cycles.
If any clock is to be useful, it must be precise, and
Consecutive days

the biological clock is no exception. When an animal is

cloistered in unvarying conditions and the free-running
period length of its activity rhythm is determined on suc-
cessive days of several months, the measurements are
usually found to be extremely consistent. For some ani-
mals the precision is astounding. For example, the bio-
logical clock of the flying squirrel measures a day to
within minutes without external time cues (Mistlberger
and Rusak 2005).
D L

24 hours 24 hours 24 hours ENTRAINMENT BY ENVIRONMENTAL

CYCLES
FIGURE 9.2 Diagram illustrating a biological rhythm in
the entrained and free-running state. (a) The clock is By our definition, biological rhythms are generally not
entrained to the light–dark cycle indicated by the bars at exactly the same length as the natural cycles—a circadian
the bottom of the column. Entrainment is the establish- rhythm, for example, has a period of approximately one
ment of a stable phase relationship between the rhythm day (“circa” means “about”), which we see when animals
and the light–dark cycle, thus ensuring that the activi- are held under constant conditions. To keep from get-
ties programmed by the clock occur at the appropriate ting wildly out of synchrony with the natural cycle, bio-
times. When an organism is placed in constant condi- logical clocks need to be reset, or entrained, to the cycle.
tions, the period length of its rhythms is seldom exactly
Let’s focus on some examples of the environmental
24 hours. Depending on the organism, the light inten-
sity, and the temperature, the period length may be
cycles that reset circadian clocks.
(b) slightly longer than 24 hours or (c) slightly shorter For circadian rhythms, the most powerful phase-
than 24 hours. The adjective circadian is used to setting agent is a light–dark cycle (Johnson et al. 2003).
describe this change in period. (From Brown et al. In the laboratory, the biological clock can be reset at will
1970.) by manipulating the light–dark cycle. If a hamster is kept
in a light–dark cycle with 12 hours of light alternating
that the external day-night cycles in light or temperature with 12 hours of darkness such that the light is turned
are not causing the rhythms. Instead, we attribute the off at 6 P.M., its activity begins shortly after 6 each
ability to keep time without external cues to an internal, evening. If the light–dark cycle is be changed abruptly
(endogenous) biological clock. so that darkness begins at midnight, over the next few
However, in the constancy of the laboratory, the days the hamster’s clock would be gradually reset so that
period (the interval between two identical points in the at the end of about five days, activity would begin shortly
cycle) of the rhythm rarely is exactly what it was in after midnight real time. If, after several weeks of this
nature; that is, it becomes slightly longer or shorter. This lighting regime, the hamster is returned to constant con-
change in the period is described with the prefix circa. So, ditions, its activity rhythm would have a period that
a daily rhythm, one that is 24 hours in nature, is approximates 24 hours, and, more important, the
described as being circadian—circa, “about;” diem, “a rhythm would initially be in phase with the second
day” (Figure 9.2). A lunar day (tidal) rhythm is described light–dark cycle.
as being circalunidian; a monthly rhythm, circa- Phase resetting occurs because a cue, a change from
monthly; and an annual rhythm, circannual. In other dark to light for instance, affects the clock differently
words, a laboratory hamster kept in constant conditions depending on when in the clock’s cycle it occurs.
may begin to run a little later every night. If it starts to Although the rhythms are separate from the clock itself,
run 10 minutes later in each cycle, after two weeks its we assume that they indicate what time the clock is sig-
activity will be about 2.5 hours out of phase with the naling. If an organism is kept in constant dark, its
actual daily cycle. rhythms will free-run, and we refer to points in the cycle
When an animal is kept in constant conditions in the as circadian time. For example, a hamster is active at
laboratory, the period length of its rhythms generally night. So, when the hamster is kept in constant darkness,
deviates from that observed in nature. We assume that we refer to the time when it begins activity as early cir-
the period length of the rhythm in constant conditions cadian nighttime. If a brief light pulse interrupts the
is a reflection of the rate at which the clock is running. darkness during early circadian night, it causes a phase
188 Chapter 9 / Biological Clocks

delay. In other words, it resets the clock so that the ham- cycle with only a few minutes of light exposure each
ster will become active later than expected the next day. week. In nature, this pattern of phase adjustment also
But if a brief light pulse interrupts darkness during the causes the squirrel’s onset of activity to follow sundown
late circadian night, it causes a phase advance; that is, the as it changes through the seasons.
animal becomes active sooner than expected in the next
cycle. In most animals, a brief light pulse occurring dur-
ing circadian daytime has little or no effect. In nature, STOP AND THINK
the clock is reset by light at dawn and dusk each day so A stable phase relationship between a 24-hour environ-
that it keeps accurate time and is set to local time. mental cycle and a rhythmic activity is not enough to
Although entrainment is important for all organ- demonstrate that the environmental cycle is entraining the
isms because it adjusts biological rhythms to prevailing biological clock. Instead, it is necessary to show that when
environmental cycles, it is especially important for ani- the environmental cycle is eliminated and the animal is in
mals living in temperate regions. Here, the length of constant conditions, the clock begins to free-run with a
daylight within each 24-hour day varies throughout the phase determined by the environmental cycle (Johnson et
year. In the northern temperate zone, the interval of al. 2003). Why is this necessary? What could be an alter-
daylight gets progressively shorter each day from the native explanation for a stable phase relationship between
an environmental cycle and a rhythmic activity?
summer solstice to the winter solstice. Then, the inter-
val of daylight gets progressively longer each day until
the summer solstice. Therefore, if it were advantageous
TEMPERATURE COMPENSATION
for an animal to become active at dawn or at twilight,
the onset of activity would have to change each day to The biological clock remains accurate in spite of large
stay appropriately synchronized with environmental changes in environmental temperature. This is some-
light–dark cycle. what surprising if one assumes that the timing mecha-
Consider the activity of a flying squirrel (Glaucomys nism is rooted in the cell’s biochemistry. As a rule,
volans). Flying squirrels nest in deep holes in trees where chemical reactions double or triple in rate for each 10ºC
they cannot see whether it is light or dark outside. Flying change in temperature. However, the effect of an equal
squirrels are nocturnal, and each has a circadian clock temperature rise on the rate at which the clock runs is
that serves as an alarm clock to awaken them at twilight, usually minor, rarely as large as 20%. This insensitivity
which allows the squirrel to synchronize its activity to a to the effects of temperature suggests that the clock
specific light intensity shortly after sunset, regardless of somehow compensates for them (reviewed in Mistlberger
the time of year (DeCoursey 2004a). and Rusak 2005). It should be apparent that if the clock
When awakened by its circadian clock, the squir- were as sensitive as most other chemical reactions are to
rel goes to the den opening and briefly samples the temperature changes, it would function as a thermome-
prevailing light. If it is still light outside, the squirrel ter, indicating the ambient temperature by its rate of
returns to its nest to sleep a bit longer and its circadian running, rather than as a timepiece.
clock is reset so that activity begins slightly later the next Consider some possible consequences if the biolog-
night. Patricia DeCoursey (1986) demonstrated that ical clock were affected by changes in temperature. First,
light sampling behavior resets the circadian clock. She the biological clocks of some animals would run at dif-
captured flying squirrels and kept them in a light-proof ferent rates at different times of day. This would obvi-
nest box connected by a tube to an open area with food ously affect poikilothermic animals, but it could also be
and water. The outside area had a standard light–dark a problem for certain mammals. Consider, for example,
cycle. On most days, the squirrels did not see light: they the cave-dwelling, insectivorous bats of the temperate
slept in their light-proof nest box, leaving only after the zone. Many species roost in cool, deep caves during the
lights went off and returning before the lights came on. daytime. They allow their body temperature to drop to
The period of the activity rhythm of a flying squirrel is that of the cave while they are resting, which conserves
shorter than 24 hours in constant conditions. Therefore, metabolic energy. However, when they leave the cave to
it would awaken a few minutes earlier each evening. forage at night, their body temperature rises to 37oC. If
After a few nights, it encountered light when it reached the biological clock were not temperature compensated,
the exit of the nest box. Upon seeing the light, it imme- how could it remain accurate when faced with such dras-
diately returned to its dark nest box and rested before tic changes in body temperature (DeCoursey 2004a)?
making a second attempt to exit. This brief exposure to The phase relationship between two oscillators (in this
light reset the squirrel’s circadian clock; the next night case, the biological clock and the LD cycle) varies with
the squirrel woke up 30 minutes later, and when it left the period length of the oscillations. Thus, the second
the nest box, it encountered darkness once again. This consequence of the clock running at different rates due
pattern was repeated every four to five days. In this way, to daily or seasonal changes in environmental tempera-
the squirrel’s activity rhythm entrained to the light–dark ture would be a change in the phase relationship between
 Rhythmic Behavior 189

the rhythmic activity and the light–dark cycle, possibly

causing the activity to occur at an inappropriate time of Day 1

day. For example, activity might begin earlier or later

Day 2
than it usually does. The change in the phase relation-
ship would occur because the free-running period Day 3
length is an important factor determining the phase rela-
tionship between the rhythmic activity and the Day 4

light–dark cycle (Gunawan and Doyle 2007). Day 5

Level of activity
Day 6

RHYTHMIC BEHAVIOR Day 7

Rhythmicity in behavior and physiology is so common Day 8

that it must be considered by anyone studying animal
behavior. An animal is not perpetually the same. Rather, Day 9

its behavior may fluctuate so that it is appropriate to the Day 10

time of day or the state of the tides or the phase of the Day 11
moon or the season of the year. We will begin by Day 12
Day 13
describing a variety of biological rhythms, each of which Day 14
is synchronized with a geophysical cycle in nature. For Day 15

a biological rhythm to be attributed to a biological clock, Day 16

the rhythm must be shown to continue in constant con- Time of day

ditions but with a slight change in period length. FIGURE 9.3 An activity rhythm of a fiddler crab (Uca
pugnax). Although the crab was maintained in constant
darkness and temperature (20ºC), the animal was active
DAILY RHYTHMS at approximately the times of low tide at its home
The predominant geophysical cycle is the daily beach. (From Palmer 1970.)
light–dark cycle, caused by the rotation of the earth rel-
ative to the sun. The activity of most animals is syn-
chronized with the daily light–dark cycle, with most ratory, away from tidal changes, its behavior remains
animals restricting their activity to a specific portion of rhythmic. Periods of activity alternate with quiescence
the day. Nocturnal animals, including the familiar ham- every 12.4 hours, the usual interval between high tides
sters, cockroaches, bats, mice, and rats, are busiest at (reviewed by Palmer 1995, 2000; Figure 9.3). (Palmer
night. Diurnal animals, such as most songbirds and [1995; 2000] presents evidence that rhythms synchro-
humans, are active during the day. Crepuscular animals nized with tides are lunar day rhythms with two peaks
are active primarily at dawn and dusk. each lunar day.)

LUNAR DAY RHYTHMS SEMILUNAR RHYTHMS

As the moon passes over the surface of the earth, its grav- The height of the tides is also influenced by the gravi-
itational field draws up a bulge in the ocean waters. One tational field of the sun. In fact, the highest tides are
bulge occurs beneath the moon and another on the caused when the gravitational fields of the moon and the
opposite side of the earth. These bulges sweep across the sun are operating together. At new moon, when the
seas as the earth rotates beneath the moon, thus causing moon appears as a dark disk, and full moon, the earth,
high tides when they reach the shoreline. Since there are the moon, and the sun are in line, causing the gravita-
two “heaps” of water, there are usually two high tides tional fields of the sun and the moon to augment each
each lunar day, one every 12.4 hours. The tides may other (Figure 9.4). Thus, the earth experiences the high-
cause some rather dramatic changes in the environment, est high tides and lowest low tides at new and full moons.
particularly for organisms living on the seashore. These periods of greatest tidal exchange are referred to
The activity of the fiddler crab, Uca pugnax, a resi- as the spring tides. At the quarters of the moon, the grav-
dent of the intertidal zone, is synchronized with the tidal itational fields of the moon and the sun are at right
changes. Fiddler crabs can be seen scurrying along the angles to each other. Because their pulls are now antag-
marsh during low tide in search of food and mates. onistic, the tidal exchange is smaller than at other times
Before the sea floods the area, the crabs return to their of the month. These periods of lowest high tides and
burrows to wait out the inundation. When a fiddler crab highest low tides are called the neap tides. Some organ-
is removed from the beach and sequestered in the labo- isms possess a biological clock that allows them to
190 Chapter 9 / Biological Clocks

a b a few hours once every two weeks, during each spring

low tide. When the tide recedes, the males are first to
break free of their puparia, the cases in which they devel-
Sun Sun oped. Each one locates a female and assists in her emer-
gence. They have little time to waste, for their habitat
will soon be submerged again, so copulation follows
quickly. Then the winged male carries his mate to where
she will lay her eggs. All these activities must be precisely
timed so that they occur during the short, two-hour
New
moon period during which the habitat is exposed.
Dietrich Neumann (1976) found that if a population
is brought into the laboratory and maintained in a
light–dark cycle in which 12 hours of light alternate with
Earth Earth 12 hours of darkness, emergence from the puparium is
First Last random. If, however, one simulates the light of the full
quarter quarter moon by leaving a dim light on for four consecutive
nights, the emergence of adults from the puparia
Full becomes synchronized. Under these conditions, just as
moon in nature, emergence occurs at approximately fortnightly
FIGURE 9.4 The effect of the relative positions of the intervals for about two months.
earth, moon, and sun on the amplitude of tidal
exchange. (a) At the times of new and full moons, the
MONTHLY RHYTHMS
gravitational fields of the moon and the sun assist each
other, causing the spring tides. (b) During the first and The interval from full moon to full moon, a synodic lunar
last quarters of the moon, the gravitational fields of the month (29.5 days), corresponds to the length of time it
moon and the sun are perpendicular to each other. This takes the moon to revolve once around the earth. Some
results in the smallest tidal exchange, the neap tides. organisms have a clock that allows them to program their
activities to occur at specific times during this cycle.
predict the times of spring tides or neap tides and gear The ant lion (Myrmeleon obscurus) shows a monthly
their activities to these regular changes. rhythm in the size of the pit it builds. A lazy hunter, it
An example of a fortnightly rhythm is seen in the builds a steep-sided conical pit in the sand and then lies
tiny chironomid midge, Clunio marinus. In Clunio, the in ambush at the base, with all but its immense mandibles
end of development, the emergence of adults from their covered with sand, waiting for some small arthropod,
pupal cases, is programmed to coincide with tidal such as an ant, to slide into the pit toward its outstretched
changes. These insects live at the lowest extreme of the jaws. The ant lion then sucks out the prey’s body fluids
intertidal zone so that they are exposed to the air for only (Figure 9.5). Like a werewolf, the ant lion changes its

a b 11

10
Mean daily pit volume, ml

Full moon Full moon

FIGURE 9.5 A monthly rhythm in the pit size of the predatory ant lion. (a) The ant lion waits at the bottom of a self-
constructed pit with only its pincers exposed. When a small arthropod, such as this ant, slips into the pit, the ant lion
sucks the prey’s body juices. (b) Monthly rhythms in the pit size of 50 ant lions maintained in constant conditions in the
laboratory. Each of the predators was fed one ant a day. Larger pits were constructed at full moon than at new moon,
even though the ant lions did not experience changes in the phase of the moon. (From Youthed and Moran 1969.)
 Rhythmic Behavior 191

behavior at full moon. It constructs larger pits at the time Body mass (g)
of full moon than at new moon. Careful daily measure- 16 24 32
ments of the size of the pits of ant lions cloistered in con-
stant conditions in the laboratory have revealed that this
is a clock-controlled rhythm and not a simple response
to some aspect of the environment such as the amount

Sept.
of moonlight (Youthed and Moran 1969).

Oct.
ANNUAL RHYTHMS
The seasonal changes in the environment can be dra-

Nov.
matic, especially in the temperate zone. As the days
shorten and the temperature drops, plants and animals
prepare themselves for severe and frigid weather. Some

Dec.
species avoid the cold and limited food of winter by
migrating. An annual biological clock is important in
timing migration. We see this, for example, in the activ-

Jan.
ity of garden warblers, Sylvia borin. The bird’s activity
can be monitored by using microswitches mounted
beneath its perch. The bird whose activity is shown in
Figure 9.6 was maintained in the laboratory at a constant

Feb.
temperature and with an unvarying length of day
(12 hours of light alternating with 12 hours of darkness),
so that it was deprived of the most obvious cues for the
Mar.
onset of winter or of spring. Notice that during the sum-
mer and winter months, its activity was limited to the
daylight hours. However, during the autumn and spring,
Apr.

when it would be migrating in nature, the caged bird also

became somewhat active at night. This nocturnal activ-
ity, called Zugunruhe, or migratory restlessness, is a cage-
May

adapted form of migratory activity.

This timing function of the annual clock is particularly
important for birds that winter close to the equator, where
there are few cues to the changing season. At the equator,
the photoperiod is constant throughout the year, just as it 0 24 48 1 3 5
is in the laboratory, and rainfall and food abundance are Time Testicular width
too variable from year to year to serve as reliable cues sig- (h) (mm)
naling the appropriate time to begin migrating.
FIGURE 9.6 Annual cycles in migratory restlessness,
An annual clock also physiologically readies birds for body weight, testis size, and molting in a garden warbler
migration and reproduction. The bird gets fatter (indi- held in a constant light–dark cycle (12 hours of light
cated by body mass) during the winter, which helps pro- alternating with 12 hours of darkness) and at a constant
vide fuel for the spring migration; it molts during the temperature. Activity was measured with a microswitch
winter; and its testes enlarge for summer reproductive mounted under the perch. Successive days are mounted
activity. These cycles are free-running for many years in underneath each other. The original record (0–24) is
constant conditions, and the length of the cycle is gen- repeated on the right (24–48). Most of the bird’s perch-
erally slightly longer or shorter than a year (Gwinner hopping activity occurred during the day. When birds in
1996). (The mechanisms of orientation during migration nature are migrating, in the autumn and spring, the
are discussed in Chapter 10. The costs and benefits of caged bird showed increased activity (migratory rest-
lessness) during the night. The body weight changes
migration are discussed in Chapter 11.)
throughout the year such that the bird fattens during
One must be cautious in describing a behavior or the winter. These energy stores will increase the
physiological process that fluctuates annually as one that chances of successful spring migration. The testes
is controlled by an annual clock. Seasonal changes in enlarge during the spring in preparation for summer
behavior may be controlled not by an annual clock, but breeding. The molt (indicated by the vertical colored
by a response to the changing photoperiod, the shorten- bar) occurred in late February to March. (Data from
ing of days during the winter months and the increasing Gwinner 1996.)
192 Chapter 9 / Biological Clocks

8 FIGURE 9.7 The rhythm in spontaneous

electrical activity in isolated neurons from
rat SCN. Sodium-dependent action poten-
Firing rate (Hz)

tials are blocked for 2.5 cycles by

tetrodotoxin (TTX). When the inhibitor is
4
washed out, the rhythm in spontaneous
wash nerve firing returns with a phase predicted
TTX
by the initial cycles. This demonstrates
that the biological clock is separate from
0 the rhythmic process it drives. (From
0 3 6 9
Welsh et al. 1995.)
Circadian cycles

daylight of the spring and summer. To measure a change action potentials that require sodium (Na+) ions.
in day length requires only a circadian clock. Unlike a However, 2.5 days later, when the tetrodotoxin was
response governed by photoperiod alone, a rhythm that washed out of the cells, the rhythm reappeared with a
is controlled by an annual clock will continue to be rhyth- phase predicted by the initial cycles (Figure 9.7). This
mic even in the absence of changing day length. suggests that although nerve firing had been halted, the
clock was running accurately the entire time. Therefore,
like the hands of a clock, the rhythmic process—nerve
STOP AND THINK firing in this case—is separate from the clock mecha-
When one wants to determine whether a daily, tidal, or nism. Processes are made to be rhythmic because they
lunar rhythm is controlled by an endogenous clock, the are coupled to and driven by a biological clock.
organism is placed in constant light or constant darkness.
When one wants to determine whether an annual rhythm
is controlled by an annual clock, the animal is kept in a
constant photoperiod. Why are the procedures different? ADVANTAGES OF CLOCK-
CONTROLLED BEHAVIOR
We have seen that many behavioral rhythms match the
prominent geophysical cycles—a day, a lunar day, a lunar
THE CLOCK VERSUS THE HANDS
month, and a year. The geophysical cycles generate rhyth-
OF THE CLOCK mic changes in environmental conditions. One might
When we study biological rhythms, we actually look at wonder, then, why biological clocks exist at all. If the
the rhythmic processes and make inferences about the clocks cause changes that are correlated with environ-
clock itself. However, it is important to remember that mental cues, why not just respond to the cues themselves?
the biological clock is separate from the processes it
drives. Perhaps an analogy to a more familiar timepiece ANTICIPATION OF ENVIRONMENTAL
will emphasize this important point. Think of a clock
CHANGE
with hands. If you tear the hands from its face, the inter-
nal gears will continue to run undaunted. And so it is One reason for timing an event with a biological clock
with biological clocks. You can alter the behavior that is rather than responding directly to periodic environmen-
controlled by a clock without necessarily stopping the tal fluctuations is that it lets an animal anticipate the
clock mechanism itself. change and allow adequate time for behavioral prepara-
David Welsh and his colleagues (1995) performed tion. For example, in nature, adult fruit flies (Drosophila)
the biological equivalent of tearing the hands from the emerge from their pupal cases during a short interval
clock. The suprachiasmatic nuclei (SCN) in the brain of around dawn. At this time, the atmosphere is cool and
a mammal are a “master” biological clock that drives moist, allowing the flies an opportunity to expand their
rhythms in other processes, as we will discuss in detail wings with a minimal loss of water through the still per-
later. Welsh removed neurons from the SCN of a new- meable cuticle. This procedure takes several hours to
born rat and grew them in tissue culture. The sponta- complete. However, the relative humidity drops rapidly
neous rate of firing of these single neurons varies after the sun rises. If the flies waited until there was a
regularly during each day, even in tissue culture, and so change in light intensity, temperature, or relative humid-
we can assume that the rhythm is driven by an internal ity before beginning the preparations for emergence, they
cellular clock. This nerve firing was completely stopped would emerge later in the day, when the water loss to the
by the addition of tetrodotoxin, a chemical that prevents arid air could prevent the wings from expanding properly.
 Advantages of Clock-Controlled Behavior 193

SYNCHRONIZATION OF A BEHAVIOR most of the bees frequented the feeding station only dur-
WITH AN EVENT THAT CANNOT BE ing the learned hours. The real test, however, was on sub-
SENSED DIRECTLY sequent days, when no food was present at the feeding
station. As seen in Figure 9.8c, the greatest number of bees
Another advantage of the clock’s control of a behavior is returned to the empty feeding station only at the time at
that it allows the behavior to be synchronized with a fac- which food had been previously available (Beling 1929). In
tor in the environment that the animal cannot sense subsequent tests, it was found that the bees’ time sense is
directly. An example is the timing of bee flights to patches astonishingly accurate. Bees can be trained to go to nine dif-
of flowers that the bees have learned are open only during ferent feeding stations at nine different times of the day.
restricted times of the day (Figure 9.8). The flowers visited They are able to distinguish points in time separated by as
for nectar may be far away from the hive, and so the bee few as 20 minutes (Koltermann 1971). The adaptiveness of
could not use vision or olfaction to determine whether the such abilities for bees is clear. Flowers have a rhythm in nec-
flowers were open. Their time sense was experimentally tar secretion, producing more at some times of the day than
demonstrated during the early part of this century by indi- at others. The biological clock allows bees to time their vis-
vidually marking them and offering them sugar water at a its to flowers so that they arrive when the flower is secret-
feeding station during a restricted time each day, between ing nectar. This means that the bees can gather the
10 A.M. and noon. After six to eight days of this training, maximum amount of food with the minimum effort.

a b

c 35

30

25
Number of visiting bees

20

15

FIGURE 9.8 The time sense in bees. (a) Honeybees can use their
10 biological clock to time their visits to distant patches of flowers
so that they arrive when the flowers are open and nectar is
available. (b) Bees were marked for individual recognition and
5
trained to come to a feeding dish only at the specific times at
which food was made available. (c) After six days of training, the
0 feeding dishes were left empty and the number of bees arriving
5 6 7 8 9 10 11 12 13 14 15 16 17 1819 h throughout the day was recorded. The bees arrived at the feed-
Time ing station only when food had been previously present.
Training time Training time Training time (Modified from Beling 1929.)
194 Chapter 9 / Biological Clocks

CONTINUOUS MEASUREMENT ence in nighttime activity had unfortunate consequences

OF TIME for the SCN-lesioned animals. One night, when seven
control animals and five SCN-lesioned animals had been
Sometimes an animal may consult its clock to deter- introduced to the enclosure, a feral cat treated the enclo-
mine what time it is. As we have seen, this information sure as a kitty-convenience store. The videotape
is necessary to anticipate periodic environmental recorded the cat picking off ground squirrels that were
changes or to synchronize behavior with other events. active that night. As a result, the cat killed 60% of the
However, at other times a clock is consulted to mea- SCN-lesioned animals but only 29% of the intact con-
sure an interval of time. This, then, is a third benefit trols (DeCoursey et al. 1997). Thus, it seems that an
of a biological clock. important function of the clock for these ground squir-
The ability to measure the passage of time contin- rels may be to reduce activity at dangerous times, such
uously is crucial to an animal’s time-compensated as nighttime.
orientation. For example, a worker honeybee (Apis spp.) DeCoursey then asked whether a biological clock
indicates the direction to a nectar source to recruit bees enhances survival in eastern chipmunks, Tamias striatus.
through a dance that tells them of the proper flight bear- Animals were captured in the wild and taken to the lab-
ing relative to the sun. Because the sun is a moving ref- oratory for surgery. The SCN was destroyed in 30 ani-
erence point, the honeybee must know not only the time mals, and 24 others were given sham lesions; that is,
of day when it discovered the nectar but also how much they were anesthetized and their skulls were opened,
time has passed since then. The biological clock provides but they were not lesioned. The sham-lesioned animals
this information. The use of the sun as a compass will served as surgical controls because they experienced the
be explored in more detail in Chapter 10. same surgical procedures except for lesioning and they
suffered the consequences of removal from their habi-
tat, such as the possible takeover of their dens by other
animals, but did not undergo lesioning. Both groups of
ADAPTIVENESS OF squirrels were equipped with radio telemetry collars.
BIOLOGICAL CLOCKS The survival of these two groups of animals was com-
pared with that of 20 intact controls (animals that did
Keeping these advantages of the clock in mind, we may
not undergo surgery or removal from their habitat).
wonder whether there is any evidence that a biological
Although none of the chipmunks was active outside of
clock actually does increase fitness. Surprisingly few peo-
their burrows after dark, some chipmunks were active
ple have addressed this question. There is evidence that
within the den. SCN-lesioned chipmunks were much
the clock enhances fitness in cyanobacteria (Woelfle and
more active than chipmunks in either of the control
Johnson 2006), but that may be of marginal interest to
groups. The primary cause of mortality was predation,
those interested in animal behavior. However, Patricia
probably by a weasel. There were significant differ-
DeCoursey and her colleagues have gathered evidence
ences in mortality rates between treatment groups and
that the clock is indeed adaptive for antelope ground
both control groups. The largest differences in mor-
squirrels (Ammospermophilus leucurus). In their study,
tality were seen on days 15–80 of the study. The chip-
these researchers destroyed the SCN of some animals
munks in the SCN-lesioned group died at the rate of
and compared their survival rate outdoors to that of
0.798% per day. In contrast, those in the sham-oper-
intact animals. Because the SCN is the master biologi-
ated group died at the rate of 0.169% per day, and those
cal clock in mammals (as we will see shortly), this pro-
in the intact control group at the rate of 0.276% per
cedure allowed them to compare the survival of animals
day. The authors hypothesize that the movements of
with and without clocks. In their initial study, 12 intact
the SCN-lesioned chipmunks within their dens at night
control animals and 10 SCN-lesioned squirrels were
alerted a predator to their locations (DeCoursey et al.
monitored in a desert enclosure. Their activity was con-
2000).
tinuously monitored in several ways, including a motion
detector and a video camera. All the ground squirrels
were primarily active during daylight.
However, an important difference in the behavior of ORGANIZATION
the two groups is that the SCN-lesioned animals were OF CIRCADIAN SYSTEMS
more likely to be active on the ground surface of the
enclosure during the nighttime than were intact animals. Single cells may contain the necessary equipment for
Whereas the amount of activity occurring during the biological timing. Unicellular organisms have biolog-
night in SCN-lesioned animals ranged from 16% to ical clocks, and the cells that make up tissues and
52.1%, nighttime activity represented no more than organs often have their own independent clocks. Thus,
1.3% of the activity of intact animals. Nine of the 12 a complex nervous system or endocrine system is not
control animals were active only in the day. This differ- an essential component of the biological clock.
 Organization of Circadian Systems 195

MULTIPLE CLOCKS a b
There is no such thing as “the” biological clock. Instead,
there are clocks scattered throughout an animal’s body
(Reppert and Weaver 2002). Fruit flies (Drosophila) have a
multitude of independent clocks located throughout their
bodies, and these clocks respond to changes in light–dark
cycles without any help from the head (Plautz et al. 1997a,
b). This was shown by using an interesting technique that
caused cells with clocks to glow. The technique has since
proved to be a valuable tool; it has advanced the study of c
rhythms because it allows researchers to observe the mol-

Bioluminescence (CPS x 10)

ecular activity of important clock genes in a single, living,
intact animal. Before this, clock gene activity had to be stud-
ied by synchronizing the clocks of members of a large pop-
ulation of fruit flies with a light–dark or temperature cycle
and then periodically selecting a group of flies from the pop-
ulation, grinding them up, and testing for gene activity.
Many researchers investigating the fruit fly’s clock
focus their efforts on the period (per) gene, which is an inte-
gral part of the clock mechanism. To monitor the clock’s d
activity, the research groups headed by Jeffrey Hall and
by Steve Kay genetically engineered fruit flies to contain
the firefly luciferase gene. Luciferase is an enzyme that
acts on luciferin to produce light, allowing the firefly to
glow. Whenever the per gene was turned on, the luciferase
gene was also switched on. Because the fruit flies’ diets
were laced with luciferin, the fruit flies glowed whenever
luciferase was present. Thus, whenever the fly glowed, it
meant that the per gene was turned on. Special cameras
and video equipment measured the glow, and computers Days
traced and recorded the glow pattern. FIGURE 9.9 Biological clocks are found throughout fruit
These glow rhythms that indicate per gene activity flies, not just in the brain. The period (per) gene is
will synchronize with light–dark cycles and will continue thought be an integral part of the clock’s mechanism.
in constant darkness with a free-running period length. To measure per activity, the firefly gene for luciferase
Not only do intact flies glow rhythmically, but so will cul- was linked to the per promoter, which turns on the
tures of head, thorax, or abdomen. Furthermore, separate gene. Luciferase is the enzyme that causes a firefly to
cultures of body parts exposed to the same light–dark cycle glow in the presence of luciferin. As a result, (a) the
fruit fly glowed with an eerie green color whenever the
will glow in unison, showing that each piece of cultured
per gene was turned on. Computers measured and
tissue has its own independent clocks and that these clocks
recorded the pattern of glow. The glow rhythm per-
have their own photoreceptors. Moreover, this raises the sisted in constant darkness for several cycles and could
possibility that the insect’s brain is not required as a mas- be synchronized with light–dark cycles. When parts of
ter clock to synchronize rhythms throughout the body. the fly were cultured separately, the cultured segments
The glow rhythm of a fly kept in constant darkness continued to glow rhythmically and could still be set
gradually decreases in amplitude because the clocks in dif- by light–dark cycles. Rhythmic glow can be seen in
ferent cells run at slightly different rates without a separately cultured (b) heads, (c) thoraxes, and
light–dark cycle; thus, the independent clocks gradually (d ) abdomens. Thus, these peripheral clocks do not
become asynchronous. It is interesting to note that the require input from the brain. (From Plautz et al. 1997.)
head is the only body part in which the clocks remain syn-
chronized in the prolonged absence of light. However, COORDINATION
when exposed to a new light–dark cycle, the clocks
OF CIRCADIAN TIMING
throughout the fly entrain within one cycle and the glow
becomes rhythmic again. In nature, asynchrony among Most rhythmic animals have a multitude of independent
peripheral clocks is not a problem: Fruit flies usually have peripheral clocks in cells throughout the body.
an environmental light cycle that is able to synchronize However, information about environmental cycles may
their many independent clocks because each has its own not reach each clock directly. How, then, are an
photoreceptor, as we will see shortly (Figure 9.9). individual’s many clocks synchronized so that all the
196 Chapter 9 / Biological Clocks

rhythmic processes occur at the appropriate time rela- Photoreceptors for Entrainment
tive to one another and the environment’s cycles? It The eyes contain the photoreceptors for light entrain-
appears that there is at least one “master” clock in the ment in mammals. However, the photoreceptors for
brain that is entrained by the light–dark cycle and that entrainment are in a different part of the retina than
regulates other clocks through the nervous and/or those involved in vision. Instead, the photoreceptors for
endocrine system. Therefore, we can consider four entrainment are ganglion cells in the retina that contain
questions: (1) What photoreceptors are responsible for the photopigment melanopsin. (The photopigment
entrainment? (2) Where is the master clock? (3) What cryptochrome may also play a role in entrainment.) The
is the genetic basis of the clock? (4) How does the mas- information about the lighting conditions reaches the
ter clock regulate the other clocks in the body? clock through the retinohypothalamic tract (RHT), a
The general scheme of circadian organization is that bundle of nerve fibers connecting the retina with the
one clock or several interacting clocks function as mas- hypothalamus (Reppert and Weaver 2002).
ter clocks to synchronize peripheral clocks. The output
from the master clock(s) can be neural or hormonal. The
clocks are set to the right time because photoreceptors Master Clock
convey information on the light–dark cycle to the In mammals, the circadian system is arranged as a hier-
clock(s). (There may be additional entraining input.) archy of clocks with the SCN as the master biological
The peripheral clocks generate the rhythmic output, clock (Reppert and Weaver 2002) (Figure 9.10). What
which may feed back on and affect the master clock(s). is the evidence that the SCN is the master clock? The
We will use circadian timing in mammals as an example. activity of the SCN remains rhythmic in tissue culture,
Information on circadian organization in other species confirming that it is an independent clock. Indeed, when
can be found in DeCoursey (2004b). cultured SCN neurons are separated, the spontaneous

a Single SCN neuron b

Firing rate

0 1 2 3 4 5 6 7 8 9 SCN
Day

SCN explant
Blolurninescence

RHT

0 1 2 3 4 5 6 29 30 31 32
Day

Liver explant
Blolurninescence

Medium change

Kidney Liver

0 1 2 3 4 5 6 7 8 9 10 11 12 13
Day

FIGURE 9.10 The master biological clock of mammals is the suprachiasmatic nucleus (SCN) in the brain. (a) The firing
rate of a single neuron from the SCN continues to fire rhythmically is tissue culture. Neurons from the SCN that are
modified to glow when a clock gene turns on also remain rhythmic in tissue. There are also circadian oscillators in
other regions of the brain and in other tissues throughout the body. Here, the rhythm in liver tissue in culture damp-
ened after a few days but was restored by changing the medium. (b) Photoreceptors in the mammalian circadian sys-
tem reach the SCN via the retinohypothalmic tract (RHT). The SCN synchronizes the oscillators throughout the
body. (From Reppert and Weaver 2002.)
 Organization of Circadian Systems 197

electrical firing of individual neurons is rhythmic in con- mation and not just a component needed to make the
stant conditions, each of them with a slightly different host’s clock function (discussed in Mistlberger and Rusak
period (Welsh et al. 1995). In another experiment, the 2005).
SCN was isolated from neural input by a knife cut that
created an island of hypothalamic tissue. Following this
Genetic Basis of Mammalian Circadian Timing
treatment, the neural activity in the hypothalamic island
remained rhythmic in constant darkness, but activity in What are the molecular gears that make the clock tick?
other brain regions was continuous (Figure 9.11). This Rhythmic gene activity seems to be involved in the clock
strongly suggests that the SCN is a self-sustaining oscil- mechanism. The products of one gene or set of genes
lator that instills rhythmicity in other brain regions activate or inhibit the activity of other genes, which in
through neural connections (Inouye and Kawamura turn affect the activity of the first genes. This creates a
1979). self-regulated feedback loop of gene activity that mea-
The SCN was finally established as the primary sures an approximately 24-hour interval.
clock in mammals by transplantation studies. When an We will begin discussing the genetic basis of the cir-
SCN from a conspecific is transplanted into the brains cadian cycle with two proteins—Clock and Bmal 1—that
of rats or hamsters that have been made arrhythmic by bind together, forming a complex that enters the
destroying their own SCN, their activity becomes rhyth- nucleus (Figure 9.12). The Clock/Bmal 1 complex turns
mic once again. Importantly, the period length of the on the activity of both the period (per) and the cryp-
restored activity rhythm matches that of the transplanted tochrome (cry) genes. The protein products of these genes
SCN rather than the period length previously displayed (Per and Cry) bind together along with the protein prod-
by the recipient. This result is what one would expect if uct of the tau gene (casein kinase I episilon) to form a
the SCN were the clock that was providing timing infor- complex. The Per/Cry/Tau complex suppresses the
action of the Clock/Bmal 1 complex, resulting in less
a Intact activity of per and cry. With less Per and Cry being pro-
3 Caudate
duced and the degradation of Per, Cry, and Tau, the level
2 of the Per/Cry/Tau complex declines. Now, with less
1 inhibition of their activity, per and cry are turned on
0 again. This cycle takes about 24 hours to complete.
21 9 21 9 21

2
Peripheral Clocks
Discharges (105/30min)

1 SCN
Although the SCN may be the master biological clock,
0
21 9 21 9 21 other circadian clocks tick throughout the body, keep-
ing their own internal time (Figure 9.13). These so-
b Island called peripheral oscillators have been demonstrated
3 Caudate
with a technique described earlier—using the luciferase
2
1
gene to indicate the activity of genes involved in the
0 clock mechanism. We now know that rhythms in biolu-
21 9 21 9 21 minescence persist for more than 20 days in cultures of
4 cells from the SCN, liver, and lung (Yoo et al. 2004).
Indeed, robust bioluminescence rhythms persisted in
2 SCN individual fibroblasts that were maintained in tissue cul-
0 ture for decades. Fibroblasts are “generic” cells found
21 9 21 9 21
throughout the body, so if fibroblasts have a personal
Time clock perhaps all cells do (Welsh et al. 2004).
FIGURE 9.11 Nerve activity in specific brain regions (the
caudate and the suprachiasmatic nuclei) of a rat before Clock Output
and after isolating an “island” of brain tissue containing The role of the SCN is to entrain the peripheral circa-
the suprachiasmatic nuclei. (a) The activity in a normal,
dian oscillators so that they are correctly set to environ-
intact rat is rhythmic in both regions of the brain.
mental time. The phase relationship between rhythmic
(b) After a region of the hypothalamus that contains the
suprachiasmatic nuclei was isolated as an island, the output of the SCN and rhythmic clock genes in periph-
nerve activity was rhythmic within the SCN but not out- eral tissues varies. Peak clock gene expression occurs at dis-
side the island in the caudate region. This supports the tinct times of the day and varies in different tissues,
idea that the suprachiasmatic nuclei are self-sustaining suggesting that the clock is not directly causing the rhyth-
oscillators. (Courtesy of S. T. Inouye.) mic output of peripheral tissues (Herzog and Tosini 2001).
198 Chapter 9 / Biological Clocks

1. Clock and Bmal 1

proteins bind
together, and the
complex moves into
the nucleus. 2. The Clock/Bmal 1 complex 3. Per and Cry proteins
turns on per and cry genes. leave the nucleus and
form a complex with
Clock another protein (the
product of tau).

Bmal 1
Nucleus
per
Per Tau

DNA

cry

Cry
4. The Per/Cry/Tau
complex suppresses the
action of the Clock/Bmal 1
complex, which suppresses
transcription of per and cry
genes. The levels of Per
and Cry proteins decline.

Per and Cry gradually break

down, allowing the
Clock/Bmal 1 complex to
become active again.

FIGURE 9.12 The genetic basis of circadian timing in mammals consists of two feedback loops in gene activity. This is
a simplified diagram. There are actually two per genes and two cry genes. The protein product of tau (Tau in diagram)
is casein kinase 1 episilon.

Light Eye RHT SCN Pineal

Circadian rhythms
(activity, drinking, hormonal level)

Food

Liver Skeletal muscle Lung

FIGURE 9.13 A model of circadian organization in mammals. Ganglion cells in the retina of the eye are the photo-
receptors for the circadian system. Light information travels to the SCN, which is the master clock. Neural output
(indicated by arrows) and hormonal output (indicated by dots) from the SCN entrain peripheral circadian clocks
located throughout the body. (From Herzog and Tosini 2001.)
 Organization of Circadian Systems 199

Some signals from the SCN are sent over neural the brain without the help of neural connections. Rae
pathways, and others are hormonal. Most of the neural Silver and her colleagues (1996) demonstrated this
connections from the SCN are to other regions of the through transplant experiments similar to those
hypothalamus and therefore to the autonomic nervous described earlier, but with one important difference. The
system, which influences the level of many hormones. donor SCN tissue was enclosed within a capsule that
The SCN has at least two neural output pathways that allowed nutrients and diffusible molecules to flow
affect rhythms. One of these is a pathway to the preop- between the host and graft tissue but did not allow neural
tic nucleus of the hypothalamus, and this seems to con- processes to grow (Figure 9.14a). As in previous trans-
trol the rhythm in ovulation but does not affect the plant experiments, both SCN of the host were destroyed
activity rhythm. The second neural pathway leads to the prior to transplant, making the animal arrhythmic. The
paraventricular nucleus in the hypothalamus, an impor- transplants were made between hamsters whose clocks
tant brain center that integrates neuroendocrine and ran at different rates because some carried the tau muta-
autonomic functions. This neural pathway then leads to tion, which alters the period length observed in constant
the pineal, which rhythmically produces melatonin. Both conditions. The encapsulated grafts restored the activ-
reproductive responses to the length of day, which ity rhythm with the period length characteristic of the
depend on melatonin from the pineal, and rhythms in donor SCN (Figure 9.14b).
hormones generally require neural connections, but In mammals, the rhythmic production of two
some may also be influenced by small molecules that dif- hormones—melatonin from the pineal gland and gluco-
fuse to their target (Tousson and Meissl 2004). corticoids (cortisol) from the adrenal glands—are
Activity rhythms, on the other hand, are caused by thought to be important in entraining peripheral oscil-
the SCN’s release of chemical signals to other parts of lators (Herzog and Tosini 2001). Neural connections

a b B25-G88
0 24 48
DD 1
Intact 21.4 Intact
4 .01
SCN-X
8

24 12 6 4

.01 Lesioned
Implant
37
SCN capsule

24 12 6 4

.02 Implanted

24.0
71

24 12 6 4
101 Period in hours

FIGURE 9.14 Encapsulated donor SCN can reinstill rhythmic activity in SCN-lesioned hamsters. (a) The capsule,
shown here in white, surrounds the donor SCN. Although the capsule allowed nutrients and small molecules to dif-
fuse between the host and graft tissue, it prevented the formation of neural connections. (b) The activity rhythm of
the intact host hamster (wild-type/tau) had a period length of 21.4 hours in constant darkness. After the SCN was
destroyed, the animal was arrhythmic. Implantation of a capsule containing wild-type SCN (period length of 24 hours)
restored the activity rhythm with a period characteristic of the implant. (From Silver et al. 1996.)
200 Chapter 9 / Biological Clocks

from the SCN cause the pineal to produce more mela-

tonin at night. Melatonin then amplifies the body tem- HUMAN IMPLICATIONS
perature rhythm, facilitates sleep, and controls OF CIRCADIAN RHYTHMS
photoperiodic responses. Neural connections from the
SCN to the anterior hypothalamus set into action a cas- JET LAG
cade of hormonal events that result in the rhythmic pro- Perhaps the most familiar way that circadian clocks affect
duction of glucocorticoids by the adrenal gland. humans is jet lag, a syndrome of effects that frequently
Glucocorticoids are steroid hormones that control many includes a decrease in mental alertness and an increase
physiological functions (DeCoursey 2004b). in gastric distress. Jet lag is caused by a disruption of cir-
cadian timing. We have seen that one function of the
STOP AND THINK biological clock is to time certain activities so that they
Light cannot reset the clock in the SCN of people who occur at the best point of some predictable cycle in the
are totally blind. Consequently, their sleep–wakefulness environment. To be useful, then, they must be set to local
rhythms drift out of phase with the day–night cycle. As time. Light resets the master clock in the SCN, which
a result, they are often sleepy during the day or wide then resets peripheral clocks so that each peaks at the
awake at night. In one experiment, blind people were able most adaptive time relative to other clocks (Yamazaki
to set their clocks by taking a dose (10 mg) of melatonin et al. 2000).
at bedtime (Sack et al. 2000). Why do you think this is Traveling across time zones is a problem for humans
possible?
because of the speed with which we can do so. If you
were to fly from Cape Cod, Massachusetts, across sev-
What is the molecular basis of circadian clock out- eral time zones to Big Sur, California, the first thing that
put? The clock controls the pattern of gene activity in you would want to do is to set your watch to the local
cells, and that pattern is different in different types of tis- time. It is obvious that if your biological clock is to gear
sue. As you may recall from Chapter 3, a gene encodes your activities to the appropriate time of day in the new
the information needed to make a protein. When a gene locale, it too must be reset.
is active, that protein is produced. The protein either As you disembark from a plane after traveling across
forms part of a structure in the cell or plays a functional time zones, your biological clocks are still set to the local
role, such as turning other genes on or off or altering time of your home. The clocks will gradually adjust to
metabolic processes in the cells. Thus, if the pattern of the day–night cycle in the new locale (Figure 9.15).
gene activity is different in different types of tissue, the However, this shift cannot occur immediately; it may
response of different tissues also differs. However, since take several days. The length of time required for the
the clock is rhythmically controlling the pattern of gene biological clock to be reset to the new local time
activity, the response of the cell is also rhythmic. increases with the number of time zones traversed. To
DNA microarray analysis, a technique that can make matters worse, not all your body functions adjust
reveal the activity of thousands of genes at a time (see at the same rate, so the normal phase relationship among
Chapter 3), shows us the pattern of gene expression. Of physiological processes is upset. Therefore, for a few
the genes examined in rodents, 0.5 to 9% of those in the days after longitudinal travel, your body time is out of
SCN, pineal gland, heart, liver, and kidney are clock- phase with local time, and your rhythms may be peak-
controlled. The pattern of gene expression differs in dif- ing at inappropriate times relative to one another.
ferent types of tissue. Only about 10% of the During this time, you often suffer psychological and
clock-controlled genes are similar in two or more tissues, physiological disturbances associated with jet lag.
and these are usually genes involved directly in the
mechanism of the clock. The remainder of the active
HUMAN HEALTH
clock-controlled genes play roles in a variety of cellular
pathways, including cell-signaling pathways, regulation Nearly every physiological process in humans is rhyth-
of the cell cycle, and protein metabolism (Duffield 2003). mic, with each process peaking at the appropriate time
Jonathan Arnold and his colleagues (2008) discovered of day. It should not be surprising, then, that certain
that about 25% of the genes in the mold Neurospora acute medical conditions are more likely to occur at a
crassa are clock-controlled. Many of those are involved particular time of day. Heart attacks and strokes, for
in the assembly of ribosomes (the structures on which example, are most likely to occur between 6 A.M. and
proteins are constructed). Many of these proteins are noon. It is during these morning hours that blood
enzymes that control key metabolic processes that are pressure rises, platelets become stickier and more likely
important in a specific cell type. Thus, the clock controls to form blood clots, and the mechanism that breaks
when and where specific biochemical reactions will down blood clots is least active. On the other hand,
occur, and those reactions are specific to a particular type asthma attacks are most likely to occur at night when
of tissue (Hastings et al. 2003). the level of epinephrine, a hormone that causes the air
 Summary 201

Bullfinch Human FIGURE 9.15 The resetting of the biologi-

1 cal clock by light–dark cycles. Immediately
following a trip across time zones, the bio-
logical clock is still set to home time
rather than that of the vacation locale. It
5
may take several days for the clock to be
reset so that it has the proper phase rela-
tionship with the new light–dark cycle.
Time (days)

In addition, various rhythms (clocks) may

10
rephase at different rates. During the
interval of readjustment, the individual
suffers from jet lag and may not feel well.
(From Aschoff 1967.)
15
Body-
temperature
(minima)
20

0 6 12 18 24 0 6 12 18 24 6
Time (hours)

tubules to dilate, and cortisol, a hormone that sup- Paolo Sassone-Corsi and his colleagues (Asher et al.
presses the immune system, are low (Waterhouse and 2008) have found the molecular link between the circa-
DeCoursey 2004). dian clock and metabolism, opening new possibilities for
Because of the current epidemics of obesity and sleep treating diabetes and obesity. The activity of a gene
loss, it is interesting to note the relationships between cir- called Sirt1 is clock controlled, and it is modulated by
cadian rhythms, sleep, and energy metabolism. In one of how many nutrients a cell is consuming. Sirt1 responds
the earliest studies showing this relationship, four healthy to the energy state of a cell and transmits that informa-
males were allowed to sleep only four hours each night tion on the energy state to the clock by binding to the
for six consecutive nights. Following breakfast on the last Clock/Bmal 1 complex (see earlier discussion). Thus,
day of sleep restriction, the males had a high blood level these findings help explain why lack of sleep can
of glucose and reduced sensitivity to insulin. The extent increase hunger and lead to obesity and diabetes.
of the changes in glucose and insulin was comparable to
changes that accompany aging or diabetes. A second
study looked at the level of the hormones leptin, which STOP AND THINK
decreases appetite, and ghrelin, which increases appetite. If you were traveling from Tampa, Florida, to San Diego,
Following two consecutive nights with only four hours California, to compete in an important athletic event,
what steps could you take before you left to minimize jet
of sleep each night, the subjects’ leptin levels were down
lag? If you could choose the time of the event, would you
and ghelin levels were up. The males also reported that
choose morning or afternoon? Why?
they had a heartier appetite than usual. A much larger
study, the Wisconsin Sleep Cohort Study, included over
1000 subjects. The subjects who reported they spent lit-
tle time sleeping each night had lower leptin levels and SUMMARY
higher ghrelin levels than did those who reported sleep-
ing longer (Laposky et al. 2008). Life evolved in a cyclic environment caused by the rela-
Recall that the clock gene is an important part of the tive movements of the earth, sun, and moon. Often it is
circadian mechanism in mammals. Within six weeks of advantageous to gear an activity to occur at a specific time
birth, homozygous clock mutant mice eat more than nor- relative to some rhythmic aspect of the environment.
mal mice and gain more weight. The mutant mice are Thus clocks evolved as adaptations to these environ-
also active much longer than normal mice, and they eat mental cycles. Clock-controlled biological rhythms have
frequently throughout their active period. In contrast, three defining properties: persistence in constant condi-
normal mice eat at the beginning and end of their active tions, entrainment to environmental cycles, and temper-
period. As the mutant clock mice aged, they developed ature compensation. In the constancy of the laboratory,
high cholesterol, high triglycerides, high blood sugar, the period length of biological rhythms may deviate
low insulin, and bloated fat cells (Turek et al. 2005). slightly from the one displayed in nature. For this reason,
202 Chapter 9 / Biological Clocks

periods are described with the prefix circa, meaning enough time to prepare for the behavior, (2) synchro-
“about,” and are called circadian, circalunidian, or cir- nization of the behavior with some event that cannot be
cannual. The period length in constant conditions is sensed directly, and (3) continuous measurement of time
described as the free-running period and is assumed to so that time-compensated orientation is possible. There
reflect the rate at which the clock is running. The free- is now some evidence that a functional clock enhances
running period is generally kept constant, which indicates survival. Lesioning of the SCN, the master biological
that the biological clock is very accurate. clock in mammals, increases the mortality of certain
Although the period length of a biological rhythm free-living rodents.
is “circa” in the constancy of the laboratory, in nature it Biological clocks exist in single cells, and there are
matches that of the geophysical cycle exactly because the many clocks in a single individual. It seems that there is
clock is entrained to (locked onto) an environmental a hierarchy of clocks, with one or more master clocks
cycle. Entrainment adjusts both the period length and regulating the activities of other, peripheral clocks.
the phase of the rhythm. Daily rhythms can be entrained We considered the organization of circadian timing
to light–dark cycles and, in some species, to temperature in mammals. The photoreceptors for entrainment in
cycles. mammals are the eyes. The information reaches the
Environmental temperature has only a slight effect suprachiasmatic nucleus via nerves (the retinohypothal-
on the rate at which the clock runs. This property is amic tract). The master clock control for certain
called temperature compensation. rhythms is in the SCN, which regulates activity through
There are many examples of rhythmic processes in nerves and chemicals. The genetic basis of circadian
animals that match the basic geophysical periods: a day clocks involves self-regulated feedback loops in which
(24 hours), the tides (12.4 hours), a lunar day (24.8 hours), the products of one or more genes affect the activity of
a fortnight (14 days), a lunar month (29.5 days), and a other genes. The master circadian clock is the SCN,
year (365 days). Many of these processes remain rhyth- which exerts control over peripheral oscillators through
mic when the individual is isolated from the obvious nerves and hormones.
environmental cycles that might be thought to provide The genetic basis of mammalian clocks involves
time cues. For instance, many daily rhythms persist when feedback loops. The activity of Bmal 1 and Clock is coun-
the individual is kept in the laboratory without terbalanced by per and cry activity.
light–dark or temperature cycles. Therefore, we say that Many mammalian tissues have their own circadian
the rhythms are caused by an internal biological clock. clocks. These peripheral clocks persist in cell culture.
The biological clock is separate from the rhythms it Within the body, the peripheral clocks are entrained
drives. Processes become rhythmic when they are cou- by neural and humoral output of the SCN. In these
pled to the biological clock. ways, the SCN controls the pattern of gene expression
There are several reasons why it may be advanta- in tissues.
geous to have a biological clock to measure time rather The human circadian clock is related to health in
than responding directly to environmental changes: several ways: jet lag, onset of acute medical conditions,
(1) anticipation of the environmental changes with and energy metabolism.
 10
Mechanisms of Orientation
and Navigation
Levels of Navigational Ability Many of us have been moved by a crisp autumn day,
Piloting enveloped in the reds, yellows, and browns of the season
Compass Orientation
and watching formations of ducks or geese fly against a
steely sky. We might have noticed that if it is early in the
True Navigation
day, the flocks may be heading almost due south. If it is
Multiplicity of Orientation Cues nearing dusk or if fields of grain are nearby, they may be
Visual Cues temporarily diverted to resting or feeding areas. But when
they resume their flight, they will head southward again.
Landmarks
In the following spring, we may stand beside a swift-
Sun Compass moving river in the Pacific Northwest and watch salmon
Star Compass below a dam or a fish ladder. As they lie in deeper pools,
Polarized Light and Orientation resting before the next powerful drive that will carry
Magnetic Cues them one step nearer the spawning ground, they all face
one way—upstream.
Cues from the Earth’s Magnetic Field
Both the birds and the fish are responding to a com-
Directional Information from the Earth’s Magnetic plex and changing environment by positioning them-
Field: A Magnetic Compass
selves correctly in it and by moving from one part of it
Positional Information from the Earth’s Magnetic Field:
to another. Although the feats of migration are astound-
A Magnetic Map?
ing, they are no more crucial to survival than are mun-
Magnetoreception
dane daily activities such as seeking a suitable habitat,
Chemical Cues looking for food and returning home again, searching for
Olfaction and Salmon Homing a mate, or identifying offspring. These actions also
Olfaction and Pigeon Homing depend on the proper orientation to key aspects of the
environment. Indeed, an animal’s life depends on ori-
Electrical Cues and Electrolocation ented movements both within and between habitats.

203
204 Chapter 10 / Mechanisms of Orientation and Navigation

In this chapter we will explore some of the mechanisms Although we usually think of landmarks as visual, the
by which animals orient themselves in space. (The costs guidepost may be in any sensory modality. As we will see
and benefits of dispersal, habitat selection, and migration shortly, magnetic cues guide sea turtles during their
are covered in Chapter 11.) oceanic travels, and olfactory cues guide salmon during
their upstream migration.
LEVELS OF NAVIGATIONAL
ABILITY COMPASS ORIENTATION
Many animals often travel between home and a goal, but A second level, called compass orientation, is the
they do not all accomplish this feat in the same manner. ability to head in a geographical direction without the
We group animal strategies for finding their way into use of landmarks. The sun, the stars, and even the
three levels of ability (Bingman and Cheng 2005; earth’s magnetic field may be used as compasses by
Ronacher 2008). many different species. One way to demonstrate that
an animal is using compass orientation is to move it to
PILOTING a distant location and determine whether it continues
in the same direction or compensates for the displace-
One level is piloting, the ability to find a goal by refer- ment. If it does not compensate for the relocation,
ring to familiar landmarks. The animal may search either compass orientation is indicated (Figure 10.1). When
randomly or systematically for the relevant landmarks. immature birds of certain migratory species, such as
Relocated
European starlings, were displaced experimentally,
150 km to east they flew in the same direction as the parent group that
Start x
had not been moved, and they flew for the same dis-
tance (Perdeck 1967). In other words, they migrated
in a path parallel to their original migratory direction.
Normal Actual
course course However, because they had been experimentally dis-
placed before beginning their migration, they did not
reach their normal destination. In some cases, this
meant that they ended up in ecologically unsatisfactory
Goal Ends up places (Figure 10.2).
150 km east
of goal

FIGURE 10.1 Experimental relocation of an animal that

Uses for Compass Orientation
is using compass orientation causes it to miss the goal Compass orientation can be used in different ways—in
by the amount of its displacement. both short-distance and long-distance navigation.

Breeding
area

Migration route

Netherlands
Migration interrupted
Normal
and birds displaced FIGURE 10.2 Immature starlings captured
wintering
area in the Netherlands and released in
Switzerland
Switzerland did not compensate for the
relocation during their autumn migration.
Instead, they traveled southwest, their
Juveniles migrated normal migratory direction, and ended up
here in incorrect wintering areas. (Modified
from Perdeck 1958.)
 Levels of Navigational Ability 205

Migratory Direction of Juvenile Birds Most first navigation [Ronacher 2008].) The estimates of distance
time migrant birds reach their destination without know- and direction are often adjusted for any displacement
ing where that goal is located. They are guided by an due to current or wind. Once close to home, landmarks
inherited program that tells the juveniles in which direc- may be used to pinpoint the exact location of home.
tion to fly and how long to fly. This innate program is Many types of animals use path integration to find
sometimes called vector navigation (Berthold 2001; their way around. Consider, for example, the desert ant
Bingman et al. 2006). (Cataglyphis bicolor). During its foraging forays, this insect
What observations have supported the idea of vec- wanders far from its nest over almost featureless terrain.
tor navigation? Individual birds held in the laboratory After prey is located, sometimes 100 meters away from
flutter in the direction in which they would be flying if the nest, roughly the distance of a football field, the ant
they were free. When their cousins in nature have com- turns and heads directly toward home. It appears that the
pleted their migratory journey, the captive birds also ant knows its position relative to its nest by taking into
cease their directional activity. Furthermore, many account each turn and the distance traveled on each leg
species, particularly those that fly from Central Europe of its outward trip. If a researcher captures an ant as it
to Africa, change compass bearing during their flight. is leaving a feeding station headed for home and relo-
Garden warblers (Sylvia borin) and blackcaps (S. atri- cates the ant to a distant site, the ant’s path is in a direc-
capilla) held in the laboratory change the direction in tion that would have led it home if it had not been
which they flutter in their cages at the time that free- experimentally moved (Wehner and Srinivasan 1981).
flying members of their population change direction How does a desert ant determine the direction and
(Gwinner and Wiltschko 1978; Helbig et al. 1989). distance of its outward route? The direction is deter-
Cross-breeding studies have also shown the inheritance mined using the pattern of polarization of skylight. Ants
of migratory direction. Andreas Helbig (1991) cross- determine their direction by using the pattern of skylight
bred members of two populations of blackcaps that had polarization, which is caused by the sun’s position (dis-
very different migratory directions. The orientation of cussed shortly) (Müller and Wehner 2007). Desert ants
the offspring was intermediate between those of the par- determine the distance they travel using a mechanism
ents. Indeed, migratory direction is inherited by the that integrates the number of strides required to reach
additive effects of a number of genes (Berthold 2001). the goal with stride length. Matthias Wittlinger and col-
leagues (2007) demonstrated this internal pedometer in
Path Integration Besides their use in long-distance a very clever way. As we all know, a person with longer
navigation, compasses can be used to improve in another legs requires fewer steps to reach a goal than does a per-
type of navigation, called path integration or dead son with short legs. Therefore, the researchers predicted
reckoning. In path integration, the animal integrates that manipulating the length of ant’s legs would cause the
information on the sequence of direction and distance ants to misestimate the distance to the nest. The
traveled during each leg of the outward journey (Figure researchers collected ants at an experimental feeder and
10.3). Then, knowing its location relative to home, the manipulated the length of the ants’ legs. They length-
animal can head directly there, using its compass(es). A ened the legs of some ants by attaching pig’s bristles to
compass may also be used to determine the direction the ant’s legs, creating stilts. They shortened the legs of
traveled on each leg of the outward journey, or the direc- other ants by partial amputation. The ants walking on
tion may be estimated from the twists and turns taken, stilts overestimated the distance to the nest, whereas the
sounds, smells, or even the earth’s magnetic field. ants with stubby legs underestimated the distance. An
Information from the outward journey is used to calcu- added complication to this means of calculating the dis-
late the homeward direction (vector). (Thus, some tance traveled from home is that stride length varies with
authors consider path integration to be a type of vector rate of travel. Thus, as remarkable as this stride count-
ing might seem, the actual mechanism of distance deter-
mination also includes an estimation of stride length.
10 km
x Home Once at home, cues from inside the nest reset the path
integrator to zero, so that it can be set again by the next
outward journey (Knaden and Wehner 2006).
15 km
12 km Map and Compass A compass may also be used with
a map to calculate a homeward path. Imagine yourself
FIGURE 10.3 Navigation by path integration. This abandoned in an unfamiliar place with only a compass to
involves determining one’s position by using the direc- guide your homeward journey. Before you could head
tion and distance of each successive leg of the outward home, you would also need a map so that you could know
trip. A compass can then be used to steer a course where you were relative to home. Only then could you
directly toward home. use your compass and orient yourself correctly.
206 Chapter 10 / Mechanisms of Orientation and Navigation

Relocated ent mechanisms, and any given species usually has sev-
150 km to east eral navigational mechanisms available. Indeed, com-
Start x
mon themes in orientation systems are the use of
multiple cues, a hierarchy of systems, and transfer of
information among various systems (Berthold 2001;
Bingman and Cheng 2005; Walcott 2005). When one
Normal mechanism becomes temporarily inoperative, a backup
course Actual is used. Furthermore, a navigational system may
course involve more than one sensory system. These interac-
tions can be quite complex, but we will simplify mat-
ters by considering each sensory mechanism separately.

Goal

FIGURE 10.4 An animal that finds its way by using true VISUAL CUES
navigation can compensate for experimental relocation
and travel toward the goal. This implies that the animal
Visual mechanisms of orientation include the use of
cannot directly sense its goal and that it is not using visual landmarks and celestial cues such as the sun, stars,
familiar landmarks to direct its journey. and polarized light.

LANDMARKS
TRUE NAVIGATION
A landmark is an easily recognizable cue along a route
A third level of orientation, sometimes called true nav- that can be quickly stored in memory to guide a later
igation1 is the ability to maintain or establish reference journey. Although landmarks can be based on any sen-
to a goal, regardless of its location, without the use of sory modality, we most commonly think of visual land-
landmarks (Bingman and Cheng 2005; Ronacher 2008). marks. Indeed, landmark recognition is perhaps the
Generally, this implies that the animal cannot directly most obvious way that vision may be used for orienta-
sense its goal and that if it is displaced while en route, it tion or navigation. Humans use landmarks frequently
compensates by changing direction, thereby heading when giving directions: “turn left before the bank” or
once again toward the goal (Figure 10.4). “make a right just after the gas station.” Because the use
Only a few species, most notably the homing pigeon of landmarks is so familiar to us, it is probably not too
(Columba livia), have been shown to have true naviga- surprising to learn that many animals also use them to
tional ability. Certain other groups of birds, including find their way.
oceanic seabirds and swallows, are also known to home
with great accuracy (Able 1980; Emlen 1975), as do sea
turtles (Lohmann and Lohmann 2006). Interestingly, an Demonstrating Landmark Use
invertebrate, the spiny lobster (Panulirus argus), also There are various ways to show that landmarks play a
seems to have true navigation abilities (Boles and role in orientation. One way is to move the landmark and
Lohmann 2003). see whether this alters the orientation of the animal. In
a classic study, Niko Tinbergen demonstrated that the
digger wasp, Philanthus triangulum, relies on landmarks
MULTIPLICITY OF to relocate its nest after a foraging flight. While a female
wasp was inside the nest, a ring of 20 pine cones was
ORIENTATION CUES placed around the opening. When she left the nest, she
The feats of migration are indeed astounding—an arc- flew around the area, apparently noting local landmarks,
tic tern circumnavigating the globe, a monarch butter- and then flew off in search of prey. During her absence,
fly fluttering thousands of miles to winter in Mexico, a the ring of pine cones was moved a short distance (1 foot)
salmon returning to the stream in which it hatched away. On each of 13 observed trips, the returning wasp
after years in the open sea. How do they do it? There searched the middle of the pine cone ring for the nest
is no simple answer. Different species may use differ- opening. However, she did not find it until the pine
cones were returned to their original position
(Tinbergen and Kruyt 1938).
1
True navigation is an unfortunate term since it carries with it the impli-
Animals can also be prevented from using landmarks
cation that other means of finding one’s was from place to place are
not real methods of navigating. This is certainly not true. Nevertheless, by clouding their vision. Consider, for example, the inge-
we will use the term simply to distinguish this method of maintaining nious way that Klaus Schmidt-Koenig and Hans
a course from the others. Schlichte (1972) demonstrated that homing pigeons do
 Visual Cues 207

FIGURE 10.5 Homing pigeons that are wearing frosted

contact lenses are unable to use landmarks for
navigation. However, these pigeons head home just as
accurately as those with normal vision do. Therefore, FIGURE 10.6 The desert ant uses a remembered
although pigeons may use landmarks if they are sequence of landmark images to find its way home in a
available, they do not require them to home. familiar area.

not require landmarks to return to the vicinity of their images of surrounding objects with a series of memory
home loft: they created frosted contact lenses for the snapshots of the landmarks along a familiar route.
pigeons (Figure 10.5). Through these lenses, pigeons One animal that appears to use memory snapshots
could only vaguely see nearby objects and distant ones of landmarks is the desert ant (Figure 10.6). As previ-
not at all. Nonetheless, the flight paths of these pigeons ously mentioned, desert ants are able to plot a course
were oriented toward home just as accurately as those of back to the nest by path integration; that is, they inte-
control pigeons. Thus, the pigeons cannot be depend- grate the directions and distances traveled on all legs of
ing on familiar landmarks to guide their journey home. the journey away from the nest to plot a direct course
Note that this does not mean that they do not use land- back. However, they also use landmarks, especially when
marks when they are available, just that they can deter- they have almost reached the nest on their return from
mine the homeward direction without them. Also, the foraging site (Åkesson and Wehner 2002). Once the
although pigeons with frosted lenses get to the general ants are close to the nest entrance, they use a systematic
area of their home loft, they often cannot find the loft search to find the opening of the burrow. The search
itself. Landmarks, then, may be important in pinpoint- strategy varies with the species of desert ant and the
ing the exact loft location but are not necessary for deter- number of natural landmarks in their native habitat
mining the direction of home. (Narendra et al. 2008). Desert ants tend to follow famil-
iar routes. In fact, if landmarks are available, desert ants
often use landmarks instead of path integration. If the
Models of Landmark Use most direct path is an unfamiliar route, it could lead over
Knowing that an animal uses landmarks to find its way rocks or be blocked by scrub, and so landmarks are
does not tell us how those landmarks are used. Do other favored. Nonetheless, if the ant comes across a clearing,
animals use landmarks as humans do, as part of a men- it can use path integration to take the most direct course
tal map of the area? Perhaps some species do, but oth- home (Collett et al. 1998).
ers might use landmarks in different ways. A simple
model of landmark use is that the animal stores the
SUN COMPASS
image of a group of landmarks in its memory, almost like
a photograph of the scene. Then it moves about the Many animals use the sun as a celestial compass. In other
environment until its view of nearby objects matches the words, these animals can determine compass direction
remembered “snapshot” (Emery and Clayton 2005). from the position of the sun. Because of the earth’s rota-
Rüdiger Wehner (1981) suggested that a whole series of tion, the sun appears to move through the sky at an aver-
memory snapshots might be filed in the order in which age rate of 15º per hour. The sun rises in the east and
they are encountered. He added that invertebrates might moves across the sky to set on the western horizon. The
be able to use landmarks by comparing the successive specific course that the sun appears to take varies with
208 Chapter 10 / Mechanisms of Orientation and Navigation

the latitude of the observer and the season of the year, move at an average rate of 15º an hour. Therefore, an ani-
but it is predictable (Figure 10.7). Therefore, if the sun’s mal heading straight for its goal and navigating by keep-
path and the time of day are known, the sun can be used ing a constant angle between its path and the sun would,
as a compass. after one hour, be following a path that would be off by
Knowledge of one compass bearing is all that is nec- 15º. Some species take only short trips, so errors due to the
essary for orientation in any direction. Consider this sim- sun’s apparent motion are inconsequential. These species
plified example. Suppose you decided to camp in the do not adjust their course with the sun’s. But if the sun is
woods a short distance north of your home. As you to be used as an orientation cue for a prolonged period,
headed for your campsite at 9 A.M., the sun would be in the animal must compensate for the sun’s movement. To
the east, so you would keep the sun on your right to do so, it must be able to measure the passage of time and
travel north. However, during your homeward trek the correctly adjust its angle with the position of the sun. At
next morning, you would keep the sun on your left to 9 A.M. an animal wishing to travel south might keep the
travel south. sun at an angle of 45º to its left. By 3 P.M., however, the
The use of the sun for orientation is complicated by sun will have moved approximately 90º at an average rate
its apparent motion through the sky. The sun appears to of 15º an hour. To maintain the same southward bearing,

South West South West South West

East North East North East North

Summer Fall–Spring Winter

11 A.M. Noon 11 A.M. Noon

Sun Sun
10 A.M. 's p 10 A.M. 's p
at at
h h
9 A.M. Sun's 9 A.M.
South altitude 14° 15° South
13°
8 A.M. Hor 8 A.M.
izo 11°
n
7 A.M. Sunrise Sunset
24° 34°
16°
6 A.M. 13°
11°
East West East West
n

10° o
ri z
Ho
Sunrise Sunset

North
North

Summer Winter

FIGURE 10.7 The sun follows a predictable path through the sky that varies with latitude and season. If the sun’s
course and the time of day are known, the sun’s bearing (azimuth) provides a compass bearing. The sun appears to
move across the sky at an average rate of 15º an hour. Therefore, if the sun is to be used as a compass for a long time,
the animal must compensate for its movement.
 Visual Cues 209

the animal must now assume a 45º angle, with the sun on when the sun was blocked from view. Kramer (1951) set
its right. Time is measured by using a biological clock (dis- up outdoor experiments with caged starlings, Sturnus
cussed in Chapter 9; time-compensated orientation of bee vulgaris (Figure 10.8), which are daytime migrators, and
dances is discussed in Chapter 16). found that they oriented in the normal migratory direc-
The first work on sun compass orientation was done tion unless the sky was overcast, in which case they lost
on birds and bees in the laboratories of Gustav Kramer their directional ability and moved about randomly.
(1950) and Karl von Frisch (1950), respectively. When the sun reappeared, they oriented correctly again,
Although these two investigative groups worked at the suggesting that they were using the sun as a compass.
same time, neither knew of the other’s work. Then Kramer devised experiments in which the sun was
Nevertheless, they often used similar experimental blocked from view and a mirror was used to change the
designs to reveal the details of sun compass orientation. apparent position of the sun. The birds reoriented
We will take a closer look at the experiments of Gustav according to the direction of the new “sun.”
Kramer here, but if you want to compare these studies Because migration occurs during limited periods in
to those of von Frisch, consult von Frisch’s (1967) fasci- the fall and spring, experiments using migratory rest-
nating book, The Dance Language and Orientation of Bees lessness to study orientation mechanisms are limited to
or the discussion of bee dances in Chapter 16. two brief intervals a year. To eliminate this problem,
Gustav Kramer (1949) began his studies by trapping Kramer (1951) devised an orientation cage in which
migrant birds and caring for them in cages. He then there were 12 identical food boxes encircling a central
noticed that they became restless during their normal birdcage (Figure 10.9). Kramer and his students trained
migration season. Furthermore, most of their activity birds to expect food in a box that lay in a particular com-
took place on the side of the cage corresponding to the pass direction. This ring of food boxes could be rotated
direction in which the birds would be flying if they were so that a bird trained to get food in a given compass
free to migrate. This activity has been aptly named direction would not always be going to the same food
migratory restlessness. In noting these tendencies, box. This eliminated the possibility that the bird might
Kramer set the stage for a series of experiments that learn to recognize the food dish by some characteristic,
would yield valuable evidence in the quest for the navi-
gational mechanisms of birds.
The indication that birds migrating during the day
use the sun as a navigational cue was that the orienta-
tion (directionality) of migratory restlessness was lost
Revolving cage

Revolving walls

Food boxes

FIGURE 10.9 Kramer’s orientation cage. The bird can

see the sky through the glass roof but is prevented from
FIGURE 10.8 Starlings are daytime migrators and were seeing the surrounding landscape. It is trained to look
the subject of Gustav Kramer’s pioneering work on bird for food in a food box that is placed in a particular com-
navigation. pass direction.
210 Chapter 10 / Mechanisms of Orientation and Navigation

such as a dent. As long as the birds could see the sun, should be off by the amount that the sun had moved dur-
they would approach the proper food box. However, on ing that interval. In this example, orientation should be
overcast days the birds were often disoriented, as would shifted 90º (6 x 15º) clockwise, for example, west instead
be expected if they were using a sun compass. of south (Figure 10.10).
The results of experiments with birds in Kramer’s
orientation cages not only confirm those on migratory STOP AND THINK
restlessness (Kramer 1951), but also indicate that the
How would orientation change if the light–dark cycle was
birds compensate for the movement of the sun. Actually, changed so that the lights came on at midnight instead of
the idea of time-compensated sun compass orientation 6 A.M.?
began when Kramer noticed that the birds in his orien-
tation cages were able to orient in the proper direction
even as the sun moved across the sky. When the real sun One of Kramer’s students, Klaus Hoffmann (1954),
was replaced with a stationary light source, the birds con- was the first to use the clock-shift experiment to
tinually adjusted their orientation with the stationary sun demonstrate the involvement of the biological clock in
as though it were moving. The orientation with the arti- sun compass orientation. After resetting the internal
ficial sun changed at a rate of about 15º an hour, just as clock of starlings by keeping them in an artificial
it would to maintain a constant compass bearing using light–dark cycle for several days, the birds’ orientation
the real sun. was shifted by the predicted amount.
The birds are able to compensate for the sun’s Using experiments similar to the classic studies
apparent movement; therefore, they must possess some described above, we have confirmed that a time-
sort of independent timing mechanism. As we saw in compensated sun compass exists in a wide variety of
Chapter 9, the biological clock that allows birds to com- organisms (Åkesson and Hedenström 2007, Bingman
pensate for the movement of the sun can be reset by arti- 2005; Rozhok 2008). We also know more details about
ficially altering the light–dark regime. Initially, the birds time-compensated sun orientation. For example, even
are placed in an artificial light–dark cycle that corre- with limited exposure to the sun (experience with a par-
sponds to the natural lighting conditions outside; the tial arc), many animals develop a sun compass that can
lights are on from 6 A.M. to 6 P.M. The light period is be used all day (discussed in Rozhok 2008). Furthermore,
then shifted so that it begins earlier or later than the the apparent movement of the sun through the sky varies
actual time of dawn. For example, if the animal is with the time of day; it appears to move faster at noon
exposed to a light–dark cycle that is shifted so that the than at sunrise or sunset. The internal clock of birds
lights come on at noon instead of 6 A.M., the animal’s compensates for daily variation in the rate of the sun’s
biological clock is gradually reset. In this case, the ani- apparent movement (Wiltschcko et al. 2000).
mal’s body time would be set six hours later than real Importantly, the compasses used by animals—sun, the
time. Therefore, if the biological clock is used to com- stars, and the earth’s magnetic field—interact in some
pensate for the movement of the sun, orientation interesting ways, as we will see shortly.

a b c

45°

90°
45° 45°

Sun 3:00 P.M.

9:00 A.M. 3:00 P.M. Under natural sky
Under Under after biological clock was
natural sky natural sky retarded 6 hours

FIGURE 10.10 A clock-shift experiment demonstrates time-compensated sun compass orientation. (a) The flight path of
a bird flying south at 9 A.M. might be at an angle of 45º to the right of the sun. (b) By 3 P.M., the sun would have moved
roughly 90º, so to continue flying in the same direction, the bird’s flight path might be at an angle of 45º to the left of
the sun. (c) If the bird’s biological clock were delayed by six hours and the bird’s orientation tested at 3 P.M. (when the
bird’s body time was 9 A.M.), it would orient to the west. The flight path of the bird would be determined by the bird’s
biological clock. The flight path would, therefore, be appropriate for 9 A.M., and orientation would be shifted by 90º
clockwise. (From Palmer 1966.)
 Visual Cues 211

STAR COMPASS a
Many species of bird migrants travel at night. Even if
they set their bearings by the position of the setting sun,
how do they steer their course throughout the night?
One important cue is the stars. This was first demon-
strated by Franz and Eleonore Sauer (Sauer 1957, 1961;
Sauer and Sauer 1960). Using several species of sylviid
warblers, they performed a series of experiments aimed
at discovering just which objects in the nighttime sky the
birds use as cues. The Sauers kept their caged warblers
inside a planetarium so that the nighttime sky could be
controlled. They first lined up the planetarium sky with
the sky outside and found that the birds oriented
themselves in the proper migratory direction for that
time of year. Then the lights were turned out, and the b
star pattern of the sky was rotated. The birds continued
to orient according to the new direction of the plane-
tarium sky. When the dome was diffusely lit, the birds
were disoriented and moved about randomly. In some
experiments, even though the moon and planets were
not projected, the birds oriented correctly, apparently
taking their bearings from the stars.
We know the most about the mechanism of star
compass orientation in the indigo bunting (Passerina
cyanea). Our knowledge has been gained primarily
through Stephen Emlen’s systematic planetarium stud-
ies. These indicate that the indigo bunting relies on the
region of the sky within 35º of Polaris (Figure 10.11). FIGURE 10.11 (a) Star compass orientation was explored
Since Polaris is the pole star, it shows little apparent by exposing nocturnal migrants, indigo buntings, to a
movement and, therefore, provides the most stationary planetarium sky. During the normal time of migration,
reference point in the northern sky. The other constel- caged birds will flutter in the proper migratory direction
lations rotate around this point (Figure 10.12). The stars if the stars are visible. (b) In some studies, a bird’s feet
nearer Polaris move through smaller arcs than do those were inked, thus creating a record of its activity on the
farther away, closer to the celestial equator. The birds sides of a funnel-shaped cage.
learn that the center of rotation of the stars is in the
north, information that is used to guide their migration shown that it is not necessary for all these constellations
either northward or southward. The major constellations to be visible at once. If one constellation is blocked by
in this region are the Big Dipper, the Little Dipper, cloud cover, the bird simply relies on an alternative
Draco, Cepheus, and Cassiopeia. Experiments have constellation (Emlen 1967a, b).

Capella
Vega

Big
Dipper

FIGURE 10.12 The stars rotate around

Polaris Polaris, the North Star. The center of
* rotation of the stars tells birds which way
is north. The positions of stars in the
Cassiopeia Little
Dipper northern sky during the spring are shown
here. The closed circles indicate star posi-
tions during the early evening, and the
open circles indicate the positions of the
W N E same stars six hours later.
212 Chapter 10 / Mechanisms of Orientation and Navigation

a Former rotational north b Former rotational north

(Polaris) (Betelgeuse)

FIGURE 10.13 The orientation of indigo buntings to a stationary planetarium sky after exposure to different celestial
rotations. During their first summer, indigo buntings learn that the center of celestial rotation is north. This was
demonstrated by exposing a group of young birds to a planetarium sky that rotated (a) around Polaris (the North Star)
or (b) around Betelgeuse. During their first autumn, when they would be migrating south, they were exposed to a
stationary planetarium sky. Each dot is the mean direction of activity for a single test. The arrow on the periphery of
the circle is the overall mean direction of activity. Each group oriented away from the star that had been the center
of rotation. (Modified from data of Able and Bingman 1987; Emlen 1970.)

Young birds learn that the center of rotation of stars when their view of most of the sky is blocked. How is
is north. The axis of rotation then gives directional this possible? For at least some of these animals, another
meaning to the configuration of constellations. Once celestial orientation cue is available in patches of blue
their star compass has been set in this way, the birds do sky—polarized light. Before considering how animals
not need to see the constellations rotate. Simply viewing orient to polarized light, let’s examine the nature of
certain constellations is sufficient for orientation. This polarized light and how the pattern of skylight
was first demonstrated by exposing groups of young polarization depends on the position of the sun.
indigo buntings to normal star patterns in a planetarium
sky. One group saw a normal pattern of rotation, one that
The Nature of Polarized Light
rotated around Polaris. The other group viewed the nor-
mal pattern of stars, but instead of rotating around Light consists of many electromagnetic waves, all vibrat-
Polaris, these stars rotated around Betelgeuse, a bright ing perpendicularly to the direction of propagation
star closer to the equator. When the birds came into a (Figure 10.14). As a crude analogy, think of a rope held
migratory condition, their orientation was tested under loosely between two people as a light beam. The rope
a stationary sky. Although each group was headed in a dif- itself would define the direction of propagation of the
ferent geographic direction, both groups were well ori- light beam. If one person repeatedly flicked his or her
ented in the appropriate migratory direction relative to wrist, the rope would begin to wave or oscillate. These
the center of rotation they had experienced, either oscillations would also be perpendicular to the length of
Betelgeuse or Polaris (Figure 10.13). In other words, in the rope, but they could be vertical, horizontal, or any
the autumn, when the birds would be heading south for angle in between, depending on how she flicked her
the winter, those that had experienced Betelgeuse as the wrist. The same is true of light waves. Most light con-
center of rotation interpreted the position of that star as
north and headed away from it (Emlen 1969, 1970, 1972).
The development of the star compass has been studied a b c
in only a few species other than the indigo bunting.
Garden warblers (Wiltschko 1982; Wiltschko et al. 1987)
and pied flycatchers, Ficedula hypoleuca (Bingman 1984) also
learn that the center of celestial rotation indicates north.
Unpolarized Vertically Horizontally
polarized polarized
POLARIZED LIGHT AND ORIENTATION
FIGURE 10.14 Unpolarized and polarized light. The
One of the puzzling facets of sun compass orientation is arrows show the planes of vibration of a light beam that
that many animals continue to orient correctly even is coming straight out of the page.
 Visual Cues 213

sists of a great many waves that are vibrating in all pos- dicular to the direction in which the light beam is trav-
sible planes perpendicular to the direction in which the eling. If you were to draw imaginary lines of latitude on
wave is traveling. Such light is described as unpolarized. the celestial sphere so that they formed concentric circles
In fully polarized light, however, all waves vibrate in only around the sun and antisun, these lines would indicate the
one plane. Our rope light beam, for instance, would plane of polarization at any point in the sky. Since the
become vertically polarized if the person’s wrist were entire pattern of polarization of light in the sky is deter-
flicked only up and down. In this case, the rope might mined by the sun’s position, the pattern moves westward
oscillate vertically in the spaces between the boards of a as the sun moves through the sky (Waterman 1989).
picket fence.
As sunlight passes through the atmosphere, it
becomes polarized by air molecules and particles in the Uses of Polarized Light in Orientation
air, but the degree and direction of polarization in a given Polarized light reflected from shiny surfaces, such as
region of the sky depend on the position of the sun. In water or a moist substrate, is used by some aquatic
other words, there is a pattern of polarized light in the insects to detect suitable habitat. Indeed, polarized light
sky that is directly related to the sun’s position (Figure may actually attract them (Schwind 1991). For the back-
10.15). One aspect of this pattern is the degree of polar- swimmer, Notonecta glauca (Figure 10.16), not only is the
ization. To picture the pattern of polarization, think of horizontally polarized light that is reflected from the sur-
the sky as a celestial sphere with the sun at one pole and face of a pond a beacon that helps the insect, as it flies
an “antisun” at the other. The light at the poles is unpo- overhead, locate a new body of water during dispersal,
larized, but it becomes gradually more strongly polarized but it also triggers a plunge reaction that brings the
with increasing distance from the poles. Thus, between insect closer to a new home (Schwind 1983).
the sun and the antisun, there is a band where the light The plane of polarization of the light in the sky is
in the sky is more highly polarized than in other regions. used as an orientation cue in two possible ways. First,
This region is described as the band of maximum polarized light is used as an axis for orientation. In other
polarization. But there is more to the pattern than this: words, an animal might move at some angle with respect
the direction of the plane of polarization (called the to the plane of polarization. Many animals use polarized
e-vector) also varies according to the position of the sun. light in this way. Salamanders living near a shoreline, for
The plane of polarization of sunlight is always perpen- instance, can use the plane of polarization to direct their

a b c

d e f

Sun at horizon (dawn) Sun at 45 degree elevation Sun at zenith (noon)

(perhaps 9 A.M.)
FIGURE 10.15 The sky viewed through a polarizing filter to show the pattern of skylight polarization at (a) 9 A.M., (b) noon,
and (c) 3 P.M. The darker region of the sky is the band of maximum polarization. The diagrams below show the pattern of
polarization (d) with the sun on the horizon, (e) at 45º elevation, and (f) at zenith. The arrows indicate the direction of the
plane of polarization. The small circle denotes the position of the sun. The pattern of polarization depends on the position
of the sun. The blue sky provides an orientation cue for animals that can perceive the plane of polarization.
214 Chapter 10 / Mechanisms of Orientation and Navigation

FIGURE 10.16 Many aquatic insects, such as

this backswimmer, use polarized light
reflected from water or a moist surface
to locate an appropriate habitat. A back-
swimmer spends almost its entire life
underwater. These insects are commonly
seen in ponds, suspended beneath the
water surface, as this one is. Adults can fly,
however, and may disperse to a new pond
before laying the second batch of eggs of
the season.

movements toward land or water (Adler 1976). Second, cues are limited or absent, such as a roosting cave, under-
the pattern of polarization of sunlight might be used to ground tunnel, or the depths of an ocean. And, unlike
determine the sun’s position when it is blocked from celestial cues, it is constant year round, night and day.
view. The polarization of light in the sky could also pro-
vide an orientation cue at dawn and dusk, when the sun CUES FROM THE EARTH’S
is below the horizon. Many birds that migrate at night MAGNETIC FIELD
set their bearings at sunset. Apparently, the pattern of
To picture the geomagnetic field around the earth, imag-
skylight polarization at sunset (Able 1982) and at sunrise
ine an immense bar magnet through the earth’s core
(Moore 1986) assists the orientation of birds migrating
from north to south. However, this bar magnet is tilted
at these times because some experiments have shown
slightly from the geographic north-south axis, and the
that the birds’ directional tendencies are altered when
magnetic poles are shifted slightly from the geographic,
the plane of polarized light to which they are exposed is
or rotational, poles (Figure 10.17). The difference
experimentally shifted by rotating polarizing filters.
Indeed, when a bird is setting its bearings for the night, Rotational Magnetic
polarized light is a more important orientation cue than North Pole North Pole
the sun’s position along the horizon at dusk or the
geomagnetic field (Able 1993; Able and Able 1996).

MAGNETIC CUES N
Magneti
Many organisms, ranging from bacteria to certain c Geographic
equator
vertebrates, orient their activities relative to the earth’s equator
magnetic field. These activities include direction finding
and navigation over long and short distances—the long-
distance migrations of birds (reviewed in Wiltschko and S
Wiltschko 2005) to the nightly foraging forays of spiny
lobsters (Lohmann et al. 2007). Magnetic sense may also
help an organism locate a preferred direction, as when
bacteria swim downward, toward the muddy bottom Magnetic
they call home (Blakemore and Frankel 1981). The South Pole
Rotational
earth’s magnetic field may also orient nest building, as South Pole
in the Ansell’s mole rat, a rodent that lives underground
(Marhold et al. 1997), or roosting place of bats (Wang FIGURE 10.17 The earth’s magnetic field. The lines of
et al. 2007). Indeed, Wolfgang and Roswitha Wiltschko force leave magnetic south vertically; curve around the
(Wiltschko and Wiltschko 2007) suggest that, in birds at earth’s surface; and enter magnetic north, heading
straight down. The geomagnetic field provides several
least, a magnetic compass evolved in nonmigratory
possible cues for navigation: polarity, the north-south
species first. These species probably used the magnetic
axis of the lines of force, and the inclination of the lines
compass for optimizing paths to and from various goals, of force. The magnetic compass of most animals
such as nest sites, feeding sites, and drinking sites. Later, appears to be an inclination compass. They determine
when some species began to migrate, the migrants use the north-south axis from the orientation of the lines of
the magnetic compass to orient during migration. force but assign direction to this by the inclination of
The ability to use the earth’s magnetic field as a com- the force lines. In the northern hemisphere, north is the
pass has its advantages. It can be used in places where visual direction in which the force lines dip toward the earth.
 Magnetic Cues 215

between the magnetic pole and the geographic pole is well as vertebrates, including some fish and birds; the
called the declination of the earth’s magnetic field. mole rat, a rodent that lives underground (Wiltschko and
Because the declination is small in most places, usually Wiltschko 2006); and a bat (Wang et al. 2007). We know
less than 20º, magnetic north is usually a reasonably good that an animal responds to polarity when its orientation
indicator of geographic north. (On maps and nautical changes in response to an experimental shift in the direc-
charts used for precise navigation, both geographic north tion of magnetic north.
and magnetic north are indicated, so that a navigator or Other animals, including most birds and sea turtles,
backpacker can adjust her compass readings for declina- appear to use the magnetic field inclination. Instead of
tion.) The declination is, of course, greatest near the north or south, they distinguish between “poleward,”
poles. where the lines of force are steepest, and “equator-
Several aspects of the earth’s magnetic field vary in ward,” where the lines of force are parallel to the earth’s
a predictable manner and could, therefore, provide surface. Although the horizontal component of the
directional cues. One aspect is polarity. The magnetic earth’s field (the direction of magnetic north), which
north pole is called the positive pole, and the magnetic runs between magnetic north and magnetic south, indi-
south, the negative pole. The second aspect is the angle cates to the animal the north-south axis, the vertical
of the lines of force with respect to earth’s surface. These component of the earth’s magnetic field (the inclination
leave the magnetic south pole vertically; curve around of the field) is the cue that tells the animal whether it
the surface of the earth; become level with the surface is going toward the pole or toward the equator
at the magnetic equator; and reenter the magnetic north (Wiltschko and Wiltschko 2006).
pole, going straight down. The angle of inclination, or We can determine whether an animal is using the
dip, of the magnetic field is the angle that the line of polarity or the angle of inclination of the pole by sepa-
force makes with the horizon. The angle of inclination rately altering the horizontal and the vertical compo-
is steepest (vertical) near the poles and near zero (hori- nents of the experimental magnetic field and observing
zontal) near the equator. The third aspect that varies pre- the effect of the animal’s orientation. If an animal uses a
dictably is the intensity (or strength) of the geomagnetic polarity compass, it will shift its orientation when the
field. It is greatest at the poles and least at the equator. horizontal component of the field is shifted. In contrast,
Thus, we see that the polarity, inclination, and an animal using an inclination compass will shift its ori-
intensity of the earth’s magnetic field vary systematically entation when the vertical component of the experi-
with latitude, providing three potential orientation cues. mental field is altered.
Which of these are used? Our own experience with com- Ansell’s mole rats (Cryptomys anselli) orient using the
passes immediately brings polarity to mind. When the polarity of the magnetic field. These small rodents nor-
needle on a compass points north, it is responding to the mally live in darkness in subterranean colonies. When
polarity of the earth’s field. Indeed, some species of ani- housed in circular arenas in captivity, they reliably and
mals seem to respond to polarity (Table 10.1). This list spontaneously build their nests in the southeastern
includes invertebrates, the spiny lobster, for instance, as region of the arena. Researchers placed mole rats of the

TABLE 10.1 Animals Demonstrated to Use a Magnetic Compass

Systematic group Type of compass

Molluscs
Snails 1 order 1 family 1 species ???

Arthropods
Crustaceans 3 orders 3 families 5 species Polarity compass
Insects 6 orders 7 families 9 species Polarity compass

Vertebrata
Cartilageous fish 1 order 1 family 1 species ???
Bony fish 2 orders 2 families 4 species Polarity compass
Amphibians 1 order 2 families 2 species Inclination compass
Reptilians 1 order 2 families 2 species Inclination compass
Birds 4 orders 12 families 21 species Inclination compass
Mammals 2 orders 2 families 3 species Polarity compass
216 Chapter 10 / Mechanisms of Orientation and Navigation

same family group into a circular test arena. Within ented if the inclination in the experimental field was
hours, the animals gathered nesting materials and built altered (Figure 10.18b). It is interesting that these birds
a nest in the southeast sector of the arena. Then were not able to orient according to magnetic field lines
researchers used a Helmholtz coil, a device that gener- that were horizontal to the earth’s surface. Horizontal
ates a magnetic field when an electric current runs field lines occur around the equator. A bird could deter-
through it, to alter the magnetic field experienced by the mine the north-south axis in a horizontal field, but with-
mole rats. The magnetic field experienced by the birds out the inclination it would not know which direction
can be altered by reversing the direction of current flow is north or south (Wiltschko and Wiltschko 1972).
through the coil. When researchers reversed the hori- The results of an experiment on free-flying hom-
zontal component (the polarity) of the magnetic field, ing pigeons are also consistent with the idea that a bird’s
the mole rats began to build nests in the northwest sec- magnetic compass is based on the inclination of the
tor of the arena. However, when researchers inverted the magnetic lines of force. Small Helmholtz coil hats were
vertical component (the angle of inclination) of the mag- fitted onto the heads of homing pigeons (Figure
netic field, the mole rats continued nesting in the south- 10.19a). A Helmholtz coil is a device that generates a
east sector (Figure 10.18a) (Marthold et al. 1997). magnetic field when an electric current runs through
In contrast, birds use the inclination angle of the it. The magnetic field experienced by the birds can be
earth’s magnetic field for orientation. For example, in altered by reversing the direction of current flow
the laboratory the migratory restlessness of European through the coil. On cloudy days, when the pigeons
robins remains oriented in the proper direction even relied on magnetic cues rather than their sun compass,
when the birds have no visual cues. When the magnetic they oriented as if they considered north to be the
world that the birds experienced was reversed by switch- direction in which the magnetic lines of force dip into
ing the polarity of an experimental field, there was no the earth. Those birds that experienced the greatest dip
effect on their orientation. However, the birds reori- in the magnetic field in the north, as it is in the nor-

a N N
mN N mN

W E W E W E

S S S
mN
Local geomagnetic field Horizontal component Vertical component
reversed inverted

b
mN N
N mN

W E W W

mN

S
Local geomagnetic field Horizontal component Vertical component
deflected by 120° inverted

FIGURE 10.18 The earth’s magnetic field can serve as a compass. (a) Mole rats respond to the polarity (horizontal
component) of the ambient magnetic field. They build their nests in the southeast portion of a circular arena. If the
magnetic field is experimentally reversed, mole rats build their nests in the northwest portion of the arena. However,
if the vertical component of the ambient magnetic field is reversed, mole rats do not change their orientation. (b) Birds
use the inclination of the lines of force (vertical component of the earth's magnetic field) as a compass. The lines of
force are steepest at the poles and horizontal at the equator. Birds reverse their orientation when the inclination of
the magnetic field is reversed, but they do not alter their orientation if the polarity of the magnetic field is changed.
(From Wiltschko and Wiltschko 2005.)
 Magnetic Cues 217

a mal geomagnetic field, headed home. In contrast, the

birds that experienced the greatest dip in the magnetic
field in the south were misdirected by the reversed
magnetic information and headed directly away from
home (Figure 10.19b) (Visalberghi and Alleva 1979;
Walcott and Green 1974).
There are also some indications that several species
respond to the small differences in the intensity of the
geomagnetic field. Among these animals are bees
(Kirschvink et al. 1997; Walker and Bitterman 1989),
homing pigeons (Dennis et al. 2007; Keeton et al. 1974;
Kowalski et al. 1988), sea turtles (Lohmann and
Lohmann 1996a), and the American alligator (Rodda
1984). If changes in magnetic intensity can be sensed, the
gradual increase in strength between the equator and the
poles could also serve as a crude compass.

DIRECTIONAL INFORMATION
FROM THE EARTH’S MAGNETIC FIELD:
b Orientation on overcast day A MAGNETIC COMPASS
If we keep in mind that orientation is essential to the sur-
vival of migrating or homing animals, it should not come
as a surprise that orientation is affected by the interac-
South-seeking
tion of many cues, as well as many variables, including
pole of compass experience, species differences, and amount of stored
in induced field energy. We will separate some of these interacting vari-
pointing up
ables to try to understand just how animals remain ori-
ented when faced with the real problems of navigating.
Many animals can obtain directional information from
the earth’s magnetic field; that is, the earth’s magnetic
field can serve as a magnetic compass.

The Magnetic Compass and Bird Navigation

As we have seen, birds use the earth’s magnetic field as
a compass. They determine whether they are headed
Current flow
reversed–north-seeking toward the pole or the equator by the angle of inclina-
pole up tion of the magnetic lines of force.

Inherited Migratory Program Migratory birds

inherit a program that tells them to travel in a certain
geographical direction, based on magnetic cues, for a
certain amount of time. Because the magnetic compass
of birds is an inclination compass, migrants from either
FIGURE 10.19 (a) A pigeon with a Helmholtz coil, a the northern or the southern hemisphere might use the
device that generates a magnetic field, on its head. same migratory program—fly toward the equator (where
(b) The magnetic field experienced by the pigeon can be the lines of force are more horizontal) in the fall and
altered by changing the direction in which the electric toward the pole (where the lines of force are more ver-
current runs through the coil. On overcast days, when tical) in the spring (Wiltschko and Wiltschko 1996).
the birds could not use the sun as a compass, the mag-
Some birds, however, cross the equator during
netic field influenced their orientation. They oriented as
if they interpreted north as the direction in which the
migration and then keep going. We might wonder, then,
magnetic lines of force dip toward the earth. Each dot how a bird from northern regions that crosses the equa-
indicates the direction in which a bird vanished from tor can continue to fly south in the southern hemisphere.
sight after being released. The arrow in the center indi- To continue flying in the same geographical direction
cates the mean vanishing bearing. (Modified from data when the equator is crossed, the birds must reverse their
of Walcott and Green 1974.) migratory direction with respect to the inclination
218 Chapter 10 / Mechanisms of Orientation and Navigation

compass: they must now fly “poleward” instead of “equa- compass of sea turtles has many of the characteristics of
torward.” Experience with the horizontal magnetic field the avian magnetic compass.
around the equator is the switch that causes the birds to A sea turtle begins its journey immediately after
begin flying “poleward” (Wiltschko and Wiltschko 1996). hatching. It uses local cues to head toward the ocean.
The sensitivity of the magnetic compass of birds When sea turtle hatchlings first enter the ocean, they
corresponds to the strength of the earth’s magnetic field. simply swim into the waves to maintain an offshore
A bird generally does not respond to magnetic fields that heading. Near the shore, the waves come directly toward
are much stronger or weaker than that which is typical land, so swimming into the waves takes the turtles out
in the area where it has been living. In fact, the range of to sea. The course that is initiated by swimming into the
intensities to which a bird may respond on a given day waves is later transferred to the magnetic compass.
is usually narrower than those that it might experience In the open ocean, waves can no longer serve as a
during migration. However, it seems that the range of navigational cue because they can come from any direc-
sensitivity may be adjusted by exposure to a field of a new tion. Here, sea turtles maintain the same angle with the
strength for a period of time. Thus, responsiveness may magnetic field that they assumed while swimming into
be fine-tuned during migration (Wiltschko 1978; the waves. In this way, they stay on course. Simultaneous
Wiltschko and Wiltschko 1999). experience with both cues seems to be important. This
was revealed in an experiment in which hatchling log-
The Magnetic Compass of Sea Turtles gerhead sea turtles swam into surface waves in tanks for
either 15 or 30 minutes. Their orientation was then
Some sea turtles travel tens of thousands of kilometers
tested in still water and in a magnetic field. Only those
during their lifetimes, a feat that can require continuous
hatchlings with 30 minutes of experience swimming into
swimming for periods of several weeks, with no land in
waves in a magnetic field were able to maintain their ori-
sight. As a loggerhead sea turtle, Caretta caretta (Figure
entation in still water (Goff et al. 1998).
10.20), makes its way across the featureless Atlantic
Ocean from the coast of Florida (perhaps to the
Sargasso Sea and back), it is guided by the earth’s mag- POSITIONAL INFORMATION
netic field (Lohmann and Lohmann 1992). The hatch- FROM THE EARTH’S MAGNETIC FIELD:
lings swim toward magnetic northeast in the normal A MAGNETIC MAP?
geomagnetic field and continue to do so when the field As we have seen, true navigation requires not only a
is experimentally reversed (Figure 10.21) (Lohmann compass but also a map. The map is necessary to know
1991). And, similar to a bird’s magnetic compass, that of one’s position relative to the goal, and then a compass is
the sea turtle is based on the inclination of the magnetic needed to guide the journey in a homeward direction.
lines of force (Light et al. 1993). Indeed, the magnetic Kenneth and Catherine Lohmann (2006; Lohmann et al.
2007) caution that the magnetic maps of animals have
not been fully characterized and may function in a very
different way than human maps do. Investigation of
magnetic maps has been hampered because there is no
standard definition of the term map among researchers.
For some researchers, a map requires a mental image—
an internal spatial representation—of the region, but
that view is increasingly giving way to a broader view of
a map. For example, by the Lohmanns’ definition, an
animal has a magnetic map if it can obtain positional
information from the earth’s magnetic field, that is, if the
animal can use the earth’s magnetic field to determine its
position relative to a target or goal. In this construct of
a magnetic map, the map may be inherited or learned
and specific or very general. We will use the Lohmanns’
definition of a magnetic map in this text.
What features of the earth’s magnetic field could
provide positional information? As we have seen, the
angle of inclination varies predictably with latitude, so
an animal that could detect this feature could determine
whether its position is north or south of the goal. If an
FIGURE 10.20 A hatchling loggerhead sea turtle. These animal could detect the intensity of the total magnetic
turtles may use the earth’s magnetic field to guide their field, the horizontal component of the field and/or the
travels through the open ocean. vertical component of the field, it could determine its
 Magnetic Cues 219

a Magnetic Signposts
The magnetic “maps” of some animals may consist of
inherited responses to magnetic landmarks, or signposts,
that trigger changes in direction. We see such magnetic
triggers along the migratory pathways of certain birds, for
instance, the pied flycatcher. The Central European pop-
ulation of pied flycatchers first flies southwest to Iberia
and then southeast. The change in migratory direction
allows the birds to avoid the Alps, Mediterranean Sea, and
the central Sahara (Figure 10.22).The birds have an inher-
ited program that causes them to change migratory direc-
tion when they experience a magnetic field characteristic
of key geographical locations at the appropriate time.
Flycatchers held in captivity will flutter their wings and
head in the correct migratory direction when they are
exposed to a magnetic field characteristic of Frankfurt,
b Geomagnetic field Germany, where their free-flying comrades begin their
North migration. If captive flycatchers are then exposed to the
magnetic field characteristic of Iberia, where the migrat-
ing flycatchers change direction, the captive flycatchers
shift the direction of their fluttering to southeast. Captive
flycatchers who continue to experience the same magnetic
West East
field throughout the migratory time period or who expe-
rience the magnetic field characteristic of the end point
do not appropriately shift direction. Thus, the local mag-
netic field of Iberia acts as a signpost telling the migrat-
South ing birds to shift flight direction slightly to the left (Beck
Reversed field
and Wiltschcko 1988; Wiltschko and Wiltschko 2005).
Magnetic south Magnetic field of Magnetically
Frankfurt, Germany simulated migration
Leg 1 N N
(until mid-October)
Magnetic Magnetic
east west
W E W E

Magnetic north
S S
FIGURE 10.21 A demonstration of the ability of logger-
head sea turtle hatchlings to orient to magnetic fields. Leg 2 N N
(after mid-October)
(a) A sea turtle is harnessed in a small tank so that its
swimming direction can be determined. A coil that can
alter the magnetic field experienced by the turtles sur-
rounds the tank. (b) When exposed to the earth’s mag- W E W E
netic field, the turtles orient toward magnetic northeast.
When the field is reversed, the hatchlings still orient to
magnetic northeast, even though this is in the opposite
geographic direction. (From Lohmann 1991.) S S

FIGURE 10.22 Orientation of young pied flycatchers

position relative to the goal. Declination (the difference
held in captivity during their first migration and
between geographic north and magnetic north) also
exposed to magnetic fields typical of those along the
varies in a regular pattern and could potentially be used route. Only the birds exposed to the correct magneti-
as a clue to position. We will see that animals can use cally simulated journey oriented properly. Each triangle
cues from the earth’s magnetic field to navigate, but the represents the direction in which a bird oriented.
cues used may differ among animals or as an animal ages The arrow indicates the mean direction of all birds.
(Lohmann and Lohmann 2006; Lohmann et al. 2007). (Modified from data of Beck and Wiltschko 1988.)
220 Chapter 10 / Mechanisms of Orientation and Navigation

Magnetic signposts also trigger changes in swim- These inherited orientation responses to magnetic
ming direction during the open-sea navigation of sea fields help to keep the young loggerheads from straying
turtles. When loggerhead hatchlings are exposed to a out of the gyre (reviewed in Lohmann et al. 2008). This
magnetic field typical of northern Florida, they swim was demonstrated by recording the preferred swimming
east-southeast using the earth’s magnetic field as a direction of hatchling loggerheads that had never been
compass. This heading will bring the sea turtles to the in the ocean. The turtles were exposed to magnetic fields
Gulf Stream, which will lead them to the North characteristic of three widely separated regions along the
Atlantic gyre, a circular current that flows clockwise migratory route of the North Atlantic gyre. The young
around the Sargasso Sea. Young loggerheads remain loggerheads oriented to each field by swimming in a
in the warm, rich water of this gyre for five to ten direction that would keep them in the favorable waters
years. of the gyre if they had been migrating (Figure 10.23).

Northern Florida Northeastern gyre

0° 0°

270° 90°
270° 90°

Dangerous zone
Geomagnetic north border
180° Corrective 180°
navigation

Safe zone

Corrective
navigation er
bord
south
c
neti one
mag us z
0°
Geo e ro
g
Dan

270° 90°

180°
Southern gyre

FIGURE 10.23 Magnetic signposts in the earth’s magnetic field may direct juvenile sea turtles in the proper direction to
remain within the North Atlantic gyre, a circular current in the Sargasso Sea. The arrows in the ocean indicate the
direction of the major currents of the gyre. Juvenile sea turtles normally swim within the gyre for several years. In the
laboratory, juvenile sea turtles exposed to magnetic fields characteristic of three locations along the migratory route
preferred to swim in the direction that would keep them swimming within the gyre if they had been migrating. The
arrows leading to each circle show the location of the magnetic field to which the turtles were exposed. Each dot indi-
cates the direction in which a harnessed juvenile sea turtle swam. The arrow in the center indicates the mean swim-
ming bearing. (Modified from Lohmann et al. 2001.)
 Magnetic Cues 221

Thus, hatchling loggerheads are programmed to swim In a more recent study, Todd Dennis and his col-
in a particular direction when they encounter magnetic leagues (2007) equipped homing pigeons with GPS-
fields found in critical regions of the gyre—places where based tracking devices and tracked their flight paths near
leaving the gyre would lead the juveniles to unfavorable places with magnetic anomalies. Regardless of the
waters. Regional differences in earth’s magnetic field direction to home, the pigeons flew either parallel or
serve as navigational beacons that guide the open-sea perpendicular to the local lines with similar intensity of
migration of young loggerheads, without the turtles hav- the geomagnetic field. The alignment of flight paths
ing a conception of their geographic position or their with magnetic intensity lines is interpreted as an indica-
position relative to a goal (Lohmann et al. 2001). tion that the pigeons can detect and respond to spatial
variability of the geomagnetic field.
Position Relative to Goal As a sea turtle matures, it learns the geomagnetic
Certain animals may use an aspect or aspects of earth’s topography of specific areas and uses that information as
magnetic field as a map to locate their position relative to at least part of the map it uses to locate an isolated tar-
a goal. We do know some animals can detect both the incli- get, such as a nesting beach (Lohmann and Lohmann
nation and the intensity of earth’s magnetic field. Both of 1996a, b). After spending several years swimming in the
these features vary across the earth’s surface, and they vary North Atlantic gyre, juvenile loggerhead turtles and
in different directions. Thus, animals could use either of green turtles (Chelonia mydas) that hatched along the east-
these features to “know” the direction to the goal. ern coast of the United States move toward the coastline
Some of the magnetic effects on pigeon homing to feeding sites. Certain sea turtles migrate along the east
seem to be more than interference with the magnetic coast between summer feeding grounds in temperate
compass and, therefore, may support the idea of a mag- regions and winter feeding grounds in the south. These
netic map. One example is the disorientation of juvenile turtles migrate to the same specific feeding loca-
pigeons released in magnetic anomalies, places where tions each autumn and spring (Avens and Lohmann
the earth’s magnetic field is extremely irregular. 2004). Every few years, adult sea turtles of nearly all
Pigeons relying on the predictable changes in the geo- species migrate from their feeding locations to nesting
magnetic field would become confused in areas where areas and back again. Adults of many populations return
the field is abnormal. Some magnetic anomalies dis- to nest on the same beaches where they hatched
orient pigeons even under sunny skies, when presum- (reviewed in Lohmann et al. 2008).
ably they would be using the sun as a compass (Frei How do sea turtles migrate with such precision?
1982; Frei and Wagner 1976; Wagner 1976; Walcott The earth’s magnetic field provides a global positioning
1978). A perfect compass (the sun) cannot help if the system that tells them their position relative to a goal.
map is messed up. This suggests that the geomagnetic Kenneth Lohmann and colleagues (2004) demonstrated
field may be more than just a compass. As you can see that juvenile and adult sea turtles use the geomagnetic
in Figure 10.24, some birds released at magnetic anom- field as a navigational map—a more complex use than
alies appear to follow the magnetic topography, usually hatchlings. The researchers captured juvenile green tur-
preferring the magnetic valleys, where the lower field tles from their feeding grounds located at about the
strength is closer to home values. midpoint of the eastern coast of Florida. The swimming

Home Home
N direction N direction

FIGURE 10.24 The flight paths of pigeons

in magnetic anomalies. In some places the
geomagnetic field is highly irregular.
Pigeons released in these areas may be
completely disoriented, even on sunny
2k 2k days. The paths of these pigeons seem to
gammas

gammas

follow the magnetic valleys, where the

field strength is closer to the value at the
3 km 3 km home loft. (From Gould 1980.)
222 Chapter 10 / Mechanisms of Orientation and Navigation

0° tion also provides information about the local magnetic

field. The second hypothesis involves magnetite, a mag-
netic mineral found in many animals that orient to the
geomagnetic field. In this model, the magnetite responds
270° 90° to the earth’s magnetic field. This response could then
affect other sensory receptors, perhaps mechanoreceptors,
open ion channels, or act on the cell physically.
180°
0°
Light-Dependent Magnetoreception
Test site
Florida Because birds are the best-studied group, we will tell their
story. We must add, however, that similar observations of
a relationship between photoreception and magnetorecep-
270° 90° tion have been discovered in other animals (Rozhok 2008).
What initial observations suggest that photorecep-
tion and magnetoreception are linked in birds? First, the
magnetoreceptor is located in the eye, specifically the
180°
right eye. Second, birds cannot remain oriented to a
FIGURE 10.25 As sea turtles mature, they use the earth’s magnetic field in darkness. Not only is light required, but
magnetic field to determine their location relative to it must be light of specific wavelengths. Birds usually
home. Sea turtles return to the same feeding grounds require blue light to remain oriented to a magnetic field
every year. Researchers captured sea turtles at their but may be able to orient in red light if they are given
feeding ground along the east coast of Florida. The pre- time to adjust (Wiltschko and Wiltschko 2006).
ferred direction of swimming of each turtle (indicated Cryptochrome, a photopigment involved in magne-
by a black dot in the circle) was determined as previ- toreception, stimulates the photoreceptors differently
ously described. The turtles were exposed to a magnetic depending on the orientation of the magnetic field. Thus,
field similar to the field that exists north of the site or
it seems that migratory birds sense the magnetic field as
to the field that exists south of the site. The sites are
a visual pattern (Figure 10.26) (Ritz et al. 2000). Unlike
indicated by stars. The turtles swam in a direction that
would return them to their feeding grounds (the test some photopigments, which change shape when they
site) if they actually had been displaced. (From absorb light, cryptochrome uses photons to transfer elec-
Lohmann et al. 2004.) trons forming radical pairs (pairs or triplets of spinning
electrons). The radical pairs lead to further reactions in
direction of tethered turtles was monitored as in previ- a cascading pathway, and magnetic fields alter the func-
ous experiments. Turtles were then exposed to either a tioning of radical pairs. Cryptochromes absorb blue-
geomagnetic field that would be found 337 km north of green light—the wavelengths important for magnetic
the test site or a magnetic field that would be found 337 orientation. In migratory birds, cryptochromes are pro-
km south of the test site. Turtles exposed to a northern
magnetic field swam approximately southward; those N
exposed to a southern magnetic field swam northward
(Figure 10.25). The magnetic field may tell the turtle NW NE
whether it is north or south of its goal. The turtle might
then move in the appropriate direction until it encoun-
ters other cues, perhaps chemical, that identify the feed-
ing grounds (reviewed in Lohmann and Lohmann 2006).
W E

MAGNETORECEPTION
Humans do not sense magnetic fields—at least not con-
sciously. We might wonder, then, how animals sense the SW SE
earth’s magnetic field. There are at least two types of mag-
netoreceptors. One type involves specialized photorecep- S
tors and is light dependent. Thus, certain animals may FIGURE 10.26 Seeing the earth’s magnetic field. The
“see” the earth’s magnetic field. The basic idea of this visual field of a bird flying parallel to the horizon in
light-dependent model of magnetoreception is that pho- Urbana-Champaign, Illinois, would be modulated
toreceptor molecules absorb light better under certain depending on the direction in which the bird was
magnetic conditions. Thus, the amount of light absorp- looking. (From Ritz et al. 2000.)
 Magnetic Cues 223

duced (the genes for cryptochromes are active) at night, to earth-strength changes in the direction of the mag-
when many migrants are flying. Nonmigratory birds pro- netic field (Semm and Beason 1990).
duce cryptochromes only during the day. The difference A popular way to demonstrate a role of magnetite
in the times of cryptochrome production suggests that all in magnetoreception is to remagnetize the magnetite
birds may need magnetic information during the day, but using a brief, strong magnetic pulse to the region of the
only night-flying migrants also need it at night. Notably, animal where magnetite is located. If a strong magnetic
cryptochromes are found in the ganglion cells of a migra- pulse alters orientation, the conclusion is that magnetite
tory garden warbler’s retina and in large displaced gan- is a part of the magnetoreceptor. In this way, researchers
glion cells, which project to brain areas where have demonstrated that the polarity compass of bats is
magnetically sensitive neurons have been reported and based on magnetite (Holland et al. 2008).
these areas show high levels of neuronal activity during In addition to their light-dependent inclination com-
magnetic orientation (Mouritsen et al. 2004). pass, birds have a magnetite receptor. Experiments on
The cryptochrome-containing cells of the retina con- Australian silvereyes (Zosterops l. lateralis) provide an exam-
nect to neurons in a brain region called Cluster N, where ple (Holland et al. 2008). When adult silvereyes that were
neurons are especially active when night-flying migratory flying northward from Tasmania during their autumn
birds are orienting to a magnetic field. The retina and migration were exposed to a strong magnetic pulse, their
Cluster N are connected via pathways through the thal- orientation was shifted clockwise by about 90º toward the
amus, a brain region important for vision. Dominik east. Similar results were obtained when adult silvereyes
Heyers and his colleagues (2007) demonstrated this con- were exposed to a strong magnetic pulse during the spring
nection using special dyes that can be traced as they travel migration. These observations support the idea that a
along nerve fibers. They injected one type of tracer dye magnetite-based receptor plays a role in orientation, but
into the cryptochrome-containing cells of the retina and they don’t indicate whether it is involved in the compass
another type in the neurons in Cluster N. The tracers met sense or the map sense. However, when juvenile silvereyes
in the thalamus, which supports the hypothesis that birds are exposed to a magnetic pulse shortly after fledging,
use their visual system to sense magnetic fields. before they begin to migrate, the pulse had little effect on
their orientation. The juveniles continued to orient in
Magnetite their normal autumnal migratory direction. Unlike adult
Many organisms known to have the ability to sense migrants, which have established a navigational map dur-
geomagnetic fields, including honeybees, trout, salmon, ing previous migrations, the juveniles rely on an innate
birds, and sea turtles, have deposits of magnetic mate- migratory program that heads them in the appropriate
rial, magnetite, which often forms chains or clumps. In compass direction for their first migration. Because a mag-
vertebrates, these deposits are commonly found in the netic pulse disrupts orientation in adults but not in juve-
head or skull. The magnetite crystals can twist into niles, it is thought that the earth’s magnetic field is part
alignment with the earth’s magnetic field if permitted of the navigational map of adults (reviewed in Wiltschko
to do so. Such movement might stimulate a stretch et al. 2005; Wiltschko and Wiltschko 2006, 2007).
receptor.
If the magnetite deposits function as magnetore- Two Magnetoreceptor Systems
ceptors in larger organisms, the information they pro- Recent studies aimed at exploring the physiological basis
vide would have to be transmitted to the nervous system. for magnetoreception support the idea that animals
Therefore, associations between magnetite and the ner- might have more than one type of magnetic sensitivity.
vous system are of particular interest. So far, the closest As we have seen, there are two proposed mechanisms for
we have come to identifying the actual magnetorecep- magnetoreception, one light-dependent and the other
tor cells is in the rainbow trout (Oncorhynchus mykiss). based on magnetite. Table 10.2 presents mechanisms and
Michael Walker and his colleagues first confirmed that their proposed functions.
the ophthalmic branch of the trigeminal nerve (a cranial Certain species seem to have both types of magne-
nerve that carries sensory information from the front of toreception systems, each serving a different purpose. For
the head) contains fibers that respond to magnetic fields. example, the eastern red-spotted newt (Notophthalmus viri-
Then they used a special dye to trace these fibers both descens) uses a magnetic compass based on the inclination
to the brain and to cells in the olfactory epithelium in of the magnetic lines of force when orienting toward the
the nose of the trout. These cells, the candidate magne- shore. We know this because their orientation was shifted
toreceptor cells, contain small amounts of a material by about 180o when the vertical component of the mag-
thought to be magnetite (Walker et al. 1997). netic field was inverted. These newts are also able to home,
In birds, magnetite deposits are found in the area of that is, to return to the point of origin after being moved
the upper beak. Interestingly, branches of the bobolink’s to an unfamiliar location. During homing, the newt’s ori-
trigeminal nerve appear to innervate the region in which entation is unaffected by an inversion of the vertical com-
magnetite deposits are found. These branches respond ponent of the magnetic field (Phillips 1986), but is shifted
224 Chapter 10 / Mechanisms of Orientation and Navigation

TABLE 10.2
Photopigment-dependent
magnetoreceptor Magnetite magnetoreceptor

Feature of geomagnetic field detected Inclination or polarity Intensity

Tasks in which it is used in birds Compass (direction finding) Map (position, signpost, or trigger)
Site of reception Retina of right eye Upper beak and/or ethmoid region
Nerve Optic nerve Ophthalmic branch of trigeminal nerve
Brain structures involved Nucleus of the basal optic root Trigeminal ganglion
(nBOR); optic tectum

by a change in polarity (Phillips 1987). Thus, these initial yet formed a magnetic map. Instead, their orientation
observations suggest that, in the newt at least, the mech- was based on an innate migratory program. They use
anism(s) for magnetoreception involved in homing differs their magnetic compass, which is based on a light-depen-
from the one involved in shoreward compass orientation. dent magnetoreception process, to head in the appro-
The magnetic compass used by the eastern newt priate direction according to their inherited migratory
when orienting toward the shore is light-dependent program (reviewed in Wiltschko et al. 2005; Wiltschko
(Phillips and Borland 1992). The orientation of newts and Wiltschko 2006, 2007).
during homing is also affected by exposure to different
wavelengths of light. However, the effects of long wave-
lengths on homing are different from those on shoreward CHEMICAL CUES
orientation. Furthermore, light-dependent processes are
not expected to respond to the polarity of a magnetic field, In this section, we will focus on the use of olfactory cues
and we know that a newt’s homing ability is sensitive to for orientation during homing. We will discover that
polarity changes. This again suggests two magnetorecep- salmon are guided to the stream where they hatched by
tion mechanisms in newts (Phillips and Borland 1994). chemical landmarks, and we will examine the more
Migratory birds may also have two mechanisms of recent suggestion that pigeons also use olfactory cues
magnetoreception that serve different functions. The when homing.
light-dependent mechanism serves as a magnetic com-
pass (Ritz et al. 2009; Rodgers and Hore 2009). Because
OLFACTION AND SALMON HOMING
a magnetite-based mechanism is theoretically capable of
detecting minute variations in the earth’s magnetic field, One of the most remarkable stories in the annals of ani-
it may be part of the magnetic “map” receptor. To use mal behavior concerns the travels of the salmon. Salmon
the geomagnetic field as a map, an animal might merely hatch in the cold, clear freshwater of rivers or lakes and
compare the local intensity of the field with that at the then descend from the streams that flow from those areas
goal. A receptor system used in a map sense, then, would and swim to sea, fanning out in all directions. Once they
not have to respond to the direction of the field, but it reach the ocean, depending on the species, they may
would be expected to respond to slight variations, less spend one to five years there until they reach their breed-
than 0.1%, in the intensity of the magnetic field experi- ing condition. Now large, glistening, beautifully colored
enced. The amount of magnetic material typically creatures, they head from their feeding grounds back
found in pigeons’ skulls could comprise a receptor that through the trackless sea to the very river from which
would provide enough sensitivity to small differences in they came. When they reach the river, they swim
magnetic field to fit the bill. A comparison of the effects upstream, turning up the correct tributary until they
of a strong magnetic pulse on the orientation of juvenile reach the very one where they spent their youth.
and adult Australian silvereyes supports the idea that a Wild salmon return to the specific location of the
magnetite-based receptor system is part of a “map.” It is natal stream in which they were born with remarkable
commonly believed that whereas adult migrants have precision. Thomas Quinn and his colleagues (2006)
established a navigational map, juveniles have not. As we demonstrated this site fidelity by using temperature
have seen, the orientation of adult silvereyes is shifted by changes during incubation of prehatch sockeye salmon
a magnetic pulse, presumably because their navigational embryos to cause banding patterns on the ear bones of
map was affected. In contrast, the juvenile silvereyes the fish. These banding patterns marked fish for later
remained oriented in the appropriate migratory direc- identification. The researchers chose a pond associated
tion after a magnetic pulse. The magnetic pulse may not with Hansen Creek in southwestern Alaska as the site
affect the orientation of juveniles because they have not where the embryos would emerge and buried the
 Chemical Cues 225

Only odor of upper reach Only odor of pond

Pond
Upper
Hansen Creek
reach

1 km

Odor of upper reach,

50 m pond, and lower reach
Lower
reach
Lake Aleknagik

FIGURE 10.27 A map of Hansen Creek, Alaska, showing the distribution of olfactory
cues in different regions of the creek area.

embryos at the bottom of the pond (see the map in another) for many years, but it has only recently been
Figure 10.27). The embryos emerged, migrated to the verified. Radio-tagged spring-summer Chinook salmon
sea along with unmarked fish from the Hansen Creek (Oncorhynchus tshawyscha) tracked in the Columbia River
area, and then migrated back to the creek. Sockeye system that chose the wrong branch of the river
salmon die after spawning. The carcasses of salmon returned to the fork and swam up another branch
along the creek and in the pond were examined for band- (Keefer et al. 2008).
ing patterns on the ear bones. Of the 324 salmon car- Sensory deprivation experiments have demonstrated
casses in the pond, 12 were marked, but none of the 138 the importance of olfaction in salmon homing. Blinding
carcasses found in the creek were marked. Thus, the the fish had no effect, but plugging their nasal cavities
marked salmon returned to the site of their incubation— impaired their ability to home correctly. Coho salmon
the pond associated with Hansen Creek. (Oncorhynchus kisutch) were trapped shortly after they had
Although navigation in the open seas appears to made their choice of forks in a Y-shaped stream. The
depend on the integration of several sensory cues, nasal cavities of half of those caught in each branch were
including magnetism (Lohmann et al. 2008), sun com- plugged. The other half were untreated. All the fish were
pass, polarized light, and perhaps even odors, navigation then released downstream from the fork and allowed to
up the rivers is based primarily on olfactory cues repeat their upstream migration. Whereas 89% of the
(reviewed in Dittman and Quinn 1996). According to the control fish returned to the branch where they were orig-
olfactory hypothesis of salmon homing, young salmon inally captured, only 60% of the fish with nose plugs
learn the odors of the home stream before they begin made the correct choice (Wisby and Hasler 1954). In
their downstream migration (Hasler and Wisby 1951). another study, a fish with its nose plugged swam with oth-
The odor of the home stream is most likely the partic- ers of its kind to the opening of its home pond. However,
ular mixture of amino acids in the water (Shoji et al. unable to smell the special characteristics of its home
2000; Shoji et al. 2003; Yamamoto and Ueda 2007). After waters, it did not enter the pond (Cooper et al. 1976).
spending time at sea, the salmon return to the coast and Olfactory cues, not qualities of the habitat, guide
use olfactory cues to locate the mouth of the river in salmon to their birthplace. These conclusions are con-
which they hatched. During their upstream migration, sistent with a study done on sockeye salmon in Hansen
the salmon follow a chemical trail back to the tributary Creek in Alaska. As you can see in Figure 10.27, Hansen
where they hatched. When they come to a fork in the Creek has an upper and a lower reach (a reach is the
river, they may swim back and forth across the two region of a river or creek between two bends). It is also
branches. If they mistakenly swim up the wrong branch fed by water from a pond (the pond in which marked fish
and lose the scent of the home stream, they retreat hatched in the study illustrating site fidelity discussed ear-
downstream until the scent is encountered again. Then, lier). During the spawning season, salmon were collected
they usually take the correct route. Researchers have and tagged from both reaches of Hansen Creek and from
hypothesized such nondirect homing (choosing the the pond. The olfactory cues available at these sites dif-
wrong tributary and returning to the fork to choose fered. The upper creek had only the odor of the upper
226 Chapter 10 / Mechanisms of Orientation and Navigation

creek, and the pond had only the odor of the pond. The needles at its release site, it would assume that it was in
pond was a better quality area because its characteristics the forest south of its loft and would use one of its com-
made predation on salmon less likely. The control salmon passes, perhaps the sun or the earth’s magnetic field, to
were released at their capture site, and they remained in fly north.
that immediate area. The experimental fish were released Hans G. Wallraff (1980, 1981) has suggested a
at a site other than their capture site. Salmon from the “gradient” model of olfactory navigation that assumes
pond that were released in the lower reach, where more that there are stable gradients in the intensity of one or
olfactory cues were available, were more likely to return more environmental odors. Then, wherever it was, the
to the pond than were salmon released in the upper reach. bird would determine the strength of the odor and com-
But recall that the pond is a more suitable habitat. How pare it to the remembered intensity at the home loft.
do we know whether the fish displaced from the pond Unlike the mosaic model, which requires only that the
returned because of odor cues or habitat cues? Consider bird make qualitative discriminations among odors, the
the behavior of fish captured from the upper reach and gradient model demands that the bird make both qual-
released either in the lower reach or in the pond. As the itative and quantitative discriminations. Reconsider the
displaced upper-reach fish swam upstream, they had previous example. The smell of the ocean might form an
olfactory information from the upper reach, as well as east-west gradient, and the fragrance of the evergreen
from the pond. During the upstream journey, the fish also forest might generate a north-south gradient. If the bird
assessed habitat-quality cues. Most of the homing salmon in the previous example smelled the air at a release site
bypassed the habitat- quality cues from the pond and fol- and determined that the scent of the sea was stronger but
lowed olfactory cues to the upper reach. Most upper- the smell of the forest was weaker than at the home loft,
reach fish displaced to the pond stayed in the pond; they it would determine that its current position was north-
did not have olfactory cues to guide them back to the west of home.
upper reach (Stewart et al. 2004).

Tests of the Models

STOP AND THINK
These models of olfactory navigation have stimulated
What would you have concluded if fish from the upper intensive research, and it is becoming clear that odors
reach that were released in the pond had stayed in the
are important in the navigation of homing pigeons. Let
pond?
us see how different researchers have approached
the question.

OLFACTION AND PIGEON HOMING Distorting the Olfactory Map A method of test-
No one denies that olfactory cues are of paramount impor- ing olfactory hypotheses is to manipulate olfactory
tance during the upstream migration of salmon, but the information to distort the bird’s olfactory map. This has
role olfaction plays in pigeon homing has been contro- been done by deflecting the natural winds to make it
versial (Wallraff 2004, 2005). Let’s look at the evidence. seem that odors are coming from another direction.
The deflector lofts used in these experiments typically
have wooden baffles that shift wind flow in a pre-
Models of Avian Olfactory Navigation dictable manner (Figure 10.28). For instance, wind
Two models for olfactory navigation have been sug- from the south might be deflected so that it seemed to
gested. According to Floriani Papi’s “mosaic” model, come from the east. A pigeon in this loft would form
pigeons form a mosaic map of environmental odors an olfactory map that was shifted counterclockwise by
within a radius of 70 to 100 km of their home loft. Some 45º. When it was released south of its loft, we would
of this map would take shape as the young birds expe- expect it to interpret the local odors as being east of its
rienced odors at specific locations during exercise and loft and fly west to get home.
training flights. More distant features of the map would Deflector loft experiments have shown consistent
be filled in as wind carried faraway odors to the loft. shifts in the orientation of homing pigeons (Baldaccini
One odor might be brought by wind from the north and et al. 1975; Kiepenheuer 1978; Waldvogel et al. 1978).
another by wind from the east. The bird would associ- However, there are reasons to believe that the shift in
ate each odor with the direction of the wind carrying it. orientation observed in pigeons from deflector lofts
When the wind shifted direction, the odors that arrived might be due to something other than a distorted olfac-
first would be closer than those that took longer to tory map. We would expect pigeons that were tem-
arrive (Papi et al. 1972). For instance, a hypothetical porarily prevented from smelling at the time of their
pigeon might learn that the sea is to the west, an ever- release to be unable to read their olfactory map and to
green forest is south, a large city is north, and a garbage orient randomly. But this is not the case: the orientation
dump is east. If the bird in this example smelled pine of smell-blind (anosmic) pigeons from deflector lofts is
 Chemical Cues 227

a
Wind
Vane

Wind

b N N

N
264°
122°

208° 208°
9.0 km 9.0 km
Clockwise–birds Counter
Home Home clockwise–birds

205°

208°
9.0 km
Control birds
Home

FIGURE 10.28 Deflector lofts shift the orientation of pigeons. (a) Deflector lofts have baffles that shift the apparent
direction of the wind by 90º. Pigeons living in deflector lofts should form shifted olfactory maps. (b) The vanishing
directions of these pigeons are shifted by about 90º. The dots at the periphery of the circle denote the direction in
which the pigeon flew out of sight. The arrow within the circle indicates the mean bearing of all birds. Although a
shift in orientation is reported in all deflector loft experiments, it may be due to the deflection of light rather than
a shift in the olfactory map. (Data from Baldaccini et al. 1975.)

still shifted (Kiepenheuer 1979). Accordingly, it has been olfactory hypothesis. For example, the orientation of
concluded that the baffles in these lofts also deflect sun- pigeons was influenced by their experience with an
light and that the consistent shift in pigeon orientation unnatural odor, benzaldehyde (Figure 10.29). Pigeons
is caused by an alteration in the sun compass (Phillips et were kept in lofts where they were fully exposed to the
al. 2006). wind. The experimental birds were exposed to an air
current coming from a specific direction and carrying
Manipulating Olfactory Information Although the the odor of benzaldehyde in addition to the natural
interpretation of olfactory deprivation and deflector loft breezes. We would expect these pigeons to incorporate
experiments is quite controversial, the experiments in this information into their olfactory maps. The control
which olfactory information predictably alters the ori- birds were exposed to only the natural winds, so they
entation of pigeons remain as unshaken support for an would not have an area with the odor of benzaldehyde
228 Chapter 10 / Mechanisms of Orientation and Navigation

a Geographic orientation of apparatus Depriving Birds of Their Sense of Smell Another

approach in testing olfactory hypotheses is to deprive the
340° pigeon of its sense of smell and observe the effect on its
Benzaldehyde orientation and homing success. These anosmic pigeons
N are less accurate in their initial orientation, and fewer
return home from an unfamiliar, but not from a famil-
iar, release site. Regardless of its effect on orientation,
olfactory deprivation always delays the bird’s departure
from the release site (Able 1996). These results are con-
sistent with the idea that olfaction plays an important
role in pigeon homing.
Besides its effect on the pigeon’s sense of smell, per-
haps olfactory deprivation affects another behavior, one
Benzaldehyde not primarily controlled by olfaction, and this other
behavior alters homing performance. Suppose the pro-
cedures that impair the sense of smell also affect the
b North
pigeons’ motivation or their ability to process informa-
Control birds tion. Though possible, the evidence does not support
Experimental birds
these possibilities. Anosmic pigeons home as well as con-
Home 262° 0 1
trol pigeons when they are released from familiar sites.
264°
50.2 km 168° Thus, the procedures do not seem to affect the birds’
motivation to return home. Furthermore, pigeons whose
sense of smell is temporarily blocked by an application
of zinc sulfate to the olfactory epithelium have problems
FIGURE 10.29 The results of an experiment that manip- in returning home from unfamiliar locations, but they
ulated a pigeon’s olfactory information. (a) The experi- perform as well as controls in a spatial memory task that
mental pigeons were kept in a loft that was exposed to does not involve homing (Budzynski et al. 1998).
natural odors, as well as to a breeze carrying the odor Could it be that some other sense, say sensitivity to
of benzaldehyde from a source northwest of the loft. magnetism, is blocked along with olfaction? The dis-
Control birds were exposed to only natural odors. covery of magnetite deposits in the beaks of homing
While they were transported to the release site, all birds pigeons, which are thought to be magnetoreceptors,
were exposed to the odor of benzaldehyde. (b) The ori-
makes this an intriguing possibility (Tian et al. 2007).
entation of the experimental birds, but not the control
birds, was altered by exposure to benzaldehyde. The
Recall that information from the magnetite magne-
initial orientation of control birds was homeward. toreceptors travels to the brain over the trigeminal
However, the initial orientation of experimental birds nerve. Information about odors travels to the brain over
was toward the southeast, as would be expected if they the olfactory nerve. To evaluate the relative importance
had interpreted the odor of benzaldehyde as an indica- of magnetic and olfactory information, Anna Gagliardo
tion that the release site was northwest of the loft. The and her colleagues (2006) severed the trigeminal nerves
experimental birds oriented as if they formed an olfac- of one group of pigeons to deprive the pigeons of mag-
tory map containing an area with the odor of benzalde- netic information and severed the olfactory nerves of
hyde. (Data from Ioalé et al. 1990.) another group of pigeons to deprive the pigeons of
olfactory information. A control group of pigeons
in their olfactory map. All the birds were exposed to underwent sham surgery, in which the pigeons under-
benzaldehyde while they were transported to the release went similar surgical procedures as the experimental
site and at that site. The experimental birds took off in birds but the nerves were not severed. None of the
a direction opposite to that from which they had expe- pigeons had experience outside of its loft. The pigeons
rienced benzaldehyde at the loft. In other words, they were released more than 50 km from home. As you can
oriented as if they used an olfactory map that contained see in Figure 10.30, the initial orientation of the released
an area scented with benzaldehyde. If the release site did pigeons of the sham-operated control group and the
not smell of benzaldehyde, the experimental birds were group that had the trigeminal nerve severed was in the
homeward oriented. The control birds were not con- general direction of home. In contrast, the initial ori-
fused by the smell of benzaldehyde at the release site entation of pigeons with severed olfactory nerves was in
and flew home. Since benzaldehyde was not part of their the opposite direction. Furthermore, the number of
olfactory map, they did not associate it with a particu- pigeons that returned home within 24 hours (23 out of
lar direction. They used other cues to guide them home 24) was the same in the sham-operated control group
(Ioalé et al. 1990). and the experimental group without input from their
 Electrical Cues and Electrolocation 229

Control Trigeminal Olfactory

(sham operated) nerves severed nerves severed
H H H

FIGURE 10.30 Information from olfactory receptors is necessary for homing pigeons to return from unfamiliar loca-
tions, but input from magnetoreceptors is neither necessary nor sufficient for homing ability. The dots within the
circles represent the vanishing direction of each pigeon in each. The arrow within the circle indicates the mean vec-
tor of the group’s vanishing direction. The arrow outside the circle pointing to H indicates the home direction.
Sham-operated pigeons and pigeons with the trigeminal nerve cut headed in the direction of home, but pigeons with
the olfactory nerve cut headed away from home. (Gagliardo et al. 2006.)

magnetite receptors. However, only 4 of the 24 pigeons numerous electroreceptors in the lateral lines along the
lacking olfactory information made it home. These sides of the fish. A weakly electric fish generally keeps
results are consistent with the hypothesis that olfactory its body rigid, a posture that simplifies the analysis of
cues are more important than magnetic cues in a hom- the electrical signals.
ing pigeon’s navigational map. These fish examine their surroundings by using their
electrical sense. Since they live in muddy water, where
vision is limited, and since they are active at night, elec-
ELECTRICAL CUES trolocation is quite useful. Objects whose electrical con-
AND ELECTROLOCATION ductivity differs from that of water disturb this electrical
field. An object with greater conductivity than that of
Electrical cues have a variety of potential uses for those water—another animal, for instance—directs current
organisms that can sense them. As we will see in toward itself. Objects that are less conductive than water,
Chapter 12, certain predators use the electrical cues such as a rock jutting into its path, deflect the current
given off by living organisms to detect their prey. In (Figure 10.31). Thus, the fish can distinguish between liv-
addition, electrical fields generated by nonliving ing and nonliving objects in its environment.
sources, such as the motion of great ocean currents, The distortions in the electrical field create an elec-
waves and tides, and rivers, could provide cues for nav- trical image of objects that can tell the fish a great deal
igation. Although there is currently no evidence that about its environment. The distortion varies according
migrating fish such as salmon, shad, herring, or tuna to the location of the object relative to the fish, so the
are electroreceptive, there is some evidence that elec- location of the image on its skin tells the fish where in
trical features of the ocean floor may help guide the relation to its own body the object is located. If the dis-
movements of bottom-feeding species such as the dog- tortion is greatest on the right, the object is located on
fish shark (Waterman 1989). the right. An object near the fish’s head creates the great-
Although most living organisms generate weak est distortion near the head (Caputi et al. 1998). The
electrical fields in water, only a few species have elec- degree to which the electrical field is distorted by an
tric organs that generate pulses, creating electrical object (the amplitude of the image) is greater in the cen-
fields that can be used in communication (discussed in ter of the image than at the periphery. The fish often
Chapter 16) and orientation (reviewed in Caputi and performs a series of movements close to the object under
Budell 2006). The electric organs of weak electric fish investigation. These actions might provide sensory
(mormyriforms and gymnotidforms), located near input that helps the fish determine the object’s size or
their tail, for instance, generate a continuous stream of shape (von der Emde 1999).
brief electrical pulses. The result is an electrical field Electric fish can even measure the distance of most
around the fish in which the head acts as the positive objects accurately, regardless of the object’s size, shape,
pole and the tail as the negative pole. Nearby objects or material of which it is composed. In contrast to a
distort the field, and the distortions are detected by visual image, the size (width) of an electrical image
230 Chapter 10 / Mechanisms of Orientation and Navigation

Conductor Nonconductor

FIGURE 10.31 Electroreception. The electrical field generated by this fish is distorted by nearby objects. A good
conductor, such as another living organism, draws the lines of force together. A nonconductor, such as a rock, spreads
them out. Using electroreceptors distributed over its body surface, the fish senses the changes in the electrical field to
“picture” its environment. (From von der Emde 1999.)

increases with distance. In addition, the amplitude dif- image. The electrical images of a 2-cm cube of metal
ferences between the center and the edges of an electri- or plastic presented at different distances and measured
cal image become smaller with the increasing distance along the midline of an electric fish are shown in
of the object (Figure 10.32). The fish uses both of these Figure 10.33 (von der Emde 1999).
features—size and amplitude—together to determine the
distance of an object. A large, nearby object might cast
the same-sized image as a smaller, distant object, but SUMMARY
the more distant object would have smaller amplitude
differences between the central and outer areas of the Navigational strategies can be grouped into three levels.
One level of orientation, called piloting, is the ability to
locate a goal by referring to landmarks. A second level
is compass orientation, in which an animal orients in a
Electric organ particular compass direction without referring to land-
marks. This is the type of navigation used by most bird
migrants. A young bird migrant uses its compasses for
vector navigation, an inherited program that tells the
bird to fly in a given direction for a certain length of
time. Some animals use a compass in path integration:
they memorize the sequence of direction and distance on
the outward journey to determine their location relative
to home, and then they use a compass to travel directly
Electric organ
home. A third level of navigational skill describes an ani-
mal’s ability to locate the goal without the use of land-
marks, even if it is released in an unfamiliar location.
True navigation requires a map to determine location
and a compass to guide the journey.
FIGURE 10.32 The electrical image of a metal sphere at
different distances from the fish. The size (width) of
Animals have access to and use many different cues
the electrical image increases with distance. The ampli- for orientation and navigation. The sensory modality of
tude differences in the degree of distortion of the elec- the primary cue varies among species, and many species
trical field between the center and the periphery of the have a hierarchy of cues. Although the interactions
electrical image decrease with increasing distance. among cues can be complex, we have considered each
(From von der Emde 1999.) sensory basis separately.
 Summary 231

a pensate for the sun’s movement. Animals must learn to

use the sun as a compass. The point of sunset is also an
Object Electric organ
orientation cue that some nocturnal migrants use to
select their flight direction, which is then maintained
20 mm
throughout the night by using other cues.
The stars provide an orientation cue for some noc-
turnal avian migrants. Birds such as the indigo bunting
b 1.4
learn that the center of celestial rotation is north. This
gives directional meaning to the constellations in the cir-
Metal cube
cumpolar area. Since the spatial relationship among
Modulation of EOD amplitude

1.2 these constellations is constant, if one is blocked by cloud

cover, the birds can use the others to determine the
direction of north.
1.0 Sunlight becomes polarized as it passes through the
atmosphere. The pattern of polarization of light in the
sky varies with the position of the sun. Polarized light
0.8 Distance 12 mm
may provide an axis for orientation, or it may allow ani-
Plastic cube Distance 17 mm
mals to locate the sun from a patch of blue sky even when
Distance 22 mm their view of the sun is blocked.
0.6 The earth’s magnetic field provides several cues that
4 6 8 10 could be used for orientation: polarity, inclination, and
Distance from mouth (cm) intensity. Some animals use a polarity compass, but most
FIGURE 10.33 (a) A weakly electric fish, Gnathonemus animals use an inclination compass, which distinguishes
petersii, with a 2-cm cube positioned for electrical image between equatorward (where the magnetic lines of force
measurement. (b) The electrical images of a metal or are horizontal) and poleward (where the lines of force
plastic cube at three distances from the fish’s surface, dip toward the earth’s surface).
measured at the midline. The electrical image of the Birds and sea turtles use an inclination magnetic
metal cube is shown as a peak, and the image of the compass for directional information. Birds use their
plastic cube is shown as a trough because metal (a con- compass to follow an inherited migratory program based
ductor) pulls the lines of force together, and plastic (a on magnetic cues. Hatchling sea turtles use a magnetic
nonconductor) spreads them out. Regardless of the compass while migrating across the Atlantic Ocean.
composition of the cube, the width of the electrical
They calibrate their magnetic compass relative to the
image increases with increasing distance. The difference
in amplitude between the center and the periphery of
direction of the surface waves that they experienced as
the image gets smaller with increasing distance. The they initially swam offshore.
fish uses the ratio of two features of the image—size A magnetic map provides information from earth’s
and the amplitude differences between the core and the magnetic field that an animal can use to determine its
rim—to determine the distance of an object. (From von position relative to a goal or target. Migrating birds and
der Emde 1999.) sea turtles may have a general magnetic map consisting
of an inherited program of changes in direction of travel
Visual cues include landmarks; the sun, stars, or in response to magnetic signposts (magnetic fields char-
moon; and the pattern of skylight polarization. acteristic of specific locations). Homing pigeons and sea
Methods of demonstrating that an animal uses land- turtles develop a more detailed magnetic map with expe-
marks in navigation include moving the landmark to see rience living in a region. For example, with experience
whether the animal reorients or becomes disoriented swimming in specific regions, sea turtles form a magnetic
and impairing the animal’s vision so that it cannot use map based on the learned topography of the geomag-
landmarks. Some species use landmarks by matching netic field. This information helps a turtle navigate to a
the objects viewed with the remembered image of the specific target area.
array of landmarks. When landmarks are used in this There are (at least) two types of magnetoreceptors.
way, the animal must always follow the familiar path. One is light dependent. In birds, the magnetoreceptor,
The sun may be used as a point of reference by assum- located in the right eye, contains the photopigment cryp-
ing some angle relative to it during the journey and tochrome. Cryptochrome absorbs light differently
then reversing the angle to get home. Alternatively, depending on the orientation of the magnetic field.
since the sun follows a predictable path through the sky, Information from these receptors connects to a part of
if the time of day is known, the sun’s position provides the brain called Cluster N, which connects to a region
a compass bearing. If the sun is used as an orientation of the brain that analyzes visual information. A second
cue over a long interval of time, the animal must com- type of magnetoreceptor contains deposits of a magnetic
232 Chapter 10 / Mechanisms of Orientation and Navigation

material called magnetite. The crystals of magnetite tent, they may not be due to a shifted olfactory map.
twist in alignment of the magnetic field. This twisting However, the results of experiments in which olfactory
could stimulate a stretch receptor. Salamanders and birds information is manipulated are consistent with an olfac-
are among the animals that have two magnetoreceptor tory basis for pigeons’ navigation. The role of olfaction
mechanisms, each serving a different function. in their homing remains controversial.
Chemical cues are also used for orientation. Salmon Some aquatic species can detect electrical fields.
are guided to their natal stream by chemical cues. Young These could be of use in navigation. A few species have
salmon learn (imprint on) the characteristic odors of electric organs that can generate electrical fields, which
their natal stream and then follow the odor trail back to can be used in communication and navigation. The weak
that place. electric fish generate a stream of electrical pulses and
Homing pigeons may also rely on olfactory naviga- then sense objects by the disturbance created in this sym-
tion. Although the results of deflector lofts are consis- metrical field.
 11
The Ecology and Evolution
of Spatial Distribution
Remaining at Home Versus Leaving far upstream in the Yangtze River (Yang et al. 2006).
Costs and Benefits of Natal Philopatry From there, larvae would drift or swim downstream,
Costs and Benefits of Natal Dispersal eventually reaching the sea at Shanghai as juveniles. The
Sex Biases in Natal Dispersal young sturgeon spend years at sea and finally begin their
Natal Dispersal and Conservation Biology
first upstream spawning migration about ten years later.
The Yangtze River is the third longest river in the world;
Habitat Selection in addition to being lengthy, it has powerful currents, so
Indicators of Habitat Quality it took about 18 months for the sturgeon to migrate 3000
Search Tactics km to their traditional spawning grounds. All this
Effects of Natal Experience changed in the 1980s, however, with construction of the
Habitat Selection and Conservation Biology Gezhouba Dam on the Yangtze, a project that blocked
Migration Chinese sturgeon halfway along their migration route to
Costs of Migration the traditional spawning grounds.
Benefits of Migration
Chinese sturgeon responded by breeding below the
Gezhouba Dam, in a much smaller stretch of the
Migration and Conservation Biology
Yangtze. The number of spawning adults dropped from
2000 before construction of the dam to a few hundred,
Sturgeons and their ancestors have spawned in rivers for reflecting disruption by the dam, water pollution, over-
millions of years. Today, however, nearly all of the 25 or fishing, and heavy boat traffic (boat propellers kill several
so living species of sturgeons are endangered, including sturgeon every year). Efforts are under way to spawn
the Chinese sturgeon (Acipenser sinensis). This magnifi- Chinese sturgeon in captivity for release in the Yangtze.
cent fish can reach 4 or 5 m in length, weigh more than In addition, because characteristics of preferred spawn-
550 kg (1000 lb), and live for a century (Figure 11.1). ing habitats are known (e.g., rocky substrate and mod-
Like many other species of sturgeons, A. sinensis is erate water velocity), some researchers propose
anadromous, spending most of its life in the sea but constructing artificial spawning grounds. It is still
returning to freshwater to breed (Bemis and Kynard unclear whether these efforts will save the now endan-
1997). In better times, Chinese sturgeon spawned at sites gered Chinese sturgeon.

233
234 Chapter 11 / The Ecology and Evolution of Spatial Distribution

FIGURE 11.1 The endangered

Chinese sturgeon. In the past,
these fish migrated 3000 km up
the Yangtze River to their tradi-
tional spawning grounds. A dam
now blocks their migratory
route, forcing them to spawn in
lower reaches of the river. Here,
researchers are shown measuring
a Chinese sturgeon. They will
also attach an acoustic tag to the
fish, which will allow them to
track its movements after
release.

The story of Acipenser sinensis raises several questions. COSTS AND BENEFITS
Why do animals undergo long-distance migrations? OF NATAL PHILOPATRY
What are the costs and benefits of this behavior? And if
we know what features endangered animals seek in a What determines whether a juvenile should remain in
breeding habitat, then how can we best use this knowl- the area of its birth or disperse? There is no simple
edge to conserve them? These are some of the questions answer to that question. Indeed, there are probably mul-
that we will consider in this chapter as we describe the tiple influences on dispersal and philopatry (Boinski et
movements that animals make when searching for places al. 2005; Dobson and Jones 1985; Solomon 2003).
to live and breed. Furthermore, such factors differ in their importance
between species, sexes, and individuals. Here we will
consider some of those factors as they relate to the costs
and benefits of natal philopatry and dispersal.
REMAINING AT HOME One potential cost of remaining in the birth area is
VERSUS LEAVING inbreeding—mating between relatives. Extreme
inbreeding involves mating between parents and off-
Some animals are born in one place and then move to spring or between siblings. Inbreeding is costly because
another location, where they breed, never to return to it reduces variation among offspring and increases the
their birthplace. This behavior is called natal dispersal, risk of producing offspring that are homozygous for
and it has come to be defined in many ways. Here, we harmful or lethal recessive alleles. All organisms proba-
will use Clobert et al.’s (2001, p. xvii) definition of natal bly carry some harmful alleles that are recessive and
dispersal as “the movement between the natal area or therefore not expressed. Close relatives are more likely
social group and the area or social group where breed- to have inherited the same versions of these alleles than
ing first takes place.” Dispersal can also occur after are other members of the population. If close relatives
reproduction. This second type of dispersal, called mate, then their offspring are at increased risk of inher-
breeding dispersal (or sometimes postbreeding disper- iting a copy of the deleterious allele from each parent.
sal), is defined by Clobert et al. (2001, p. xvii) as “the These offspring will then display the harmful trait
movement between two successive breeding areas or (Shields 1982).
social groups.” For example, some female rodents, upon It can be difficult to determine the fitness costs of
weaning their litters at one location, leave the nest and inbreeding because the frequency of inbreeding is usu-
litter and move to a new site where they will give birth ally low. However, a long-term study of the Mexican jay
to their next litter. In this chapter, we will focus on natal, (Aphelocoma ultramarina) did find fitness costs associated
rather than breeding, dispersal. with inbreeding. The brood sizes of inbred pairs were
Natal dispersal can be contrasted with natal smaller than those of outbred pairs, which suggests
philopatry, in which offspring remain at their natal area hatching failure. Furthermore, compared with outbred
and share the home range or territory with their parents nestlings, significantly fewer of the inbred nestlings sur-
(Waser and Jones 1983). vived to the next year (Brown and Brown 1998).
 Remaining at Home Versus Leaving 235

A second potential cost of staying at home is repro- provide a fairly constant physical environment and safety
ductive suppression. Adult breeders may suppress the from some types of predators (for example, hawks and
reproductive development of philopatric young through owls). Long-distance above-ground dispersal by indi-
chemical means (e.g., pheromones) or behavioral meth- viduals of fossorial or semifossorial species exposes them
ods (e.g., aggression). Although suppressed young may to potentially harsh weather conditions and a suite of
not themselves reproduce, they may still be successful predators unable to enter their burrow systems. These
(from the standpoint of having their genes represented costs may be compounded by a lack of familiarity with
in future generations) by helping their parents rear sub- the terrain and high levels of aggression from residents
sequent broods or litters. We will have more to say of the population in which they are attempting to estab-
about group living and helping by older offspring in lish a new home. Male prairie voles released at unfamil-
Chapter 19. iar locations move greater distances and take longer to
A third possible cost of philopatry is increased com- find refuge than those released at familiar locations
petition. If conditions are crowded at home, then young (Jacquot and Solomon 1997). In addition, nonresident
that remain in their birth area may have to compete with woodland voles (Back et al. 2002) and Norway rats
relatives for food, nest sites, or mates. Sometimes the (McGuire et al. 2006) face harsh treatment by residents
competition among relatives for breeding opportunities when attempting to settle in an area. Indeed, all in all,
within the natal group is so intense that battles become estimates of dispersal-induced mortality in animals can
deadly (reviewed by Griffin and West 2002). Thus, lim- exceed 50% (e.g., Daniels and Walters 2000; Johnson
ited access to critical resources could result in lower sur- and Gaines 1990). At the opposite extreme, dispersal
vival and/or reproductive success than could be achieved costs appear to be negligible in some species (Gillis and
by dispersing (Shields 1987). In other words, sometimes Krebs 2000).
the grass really is greener on the other side of the street, Dispersal also appears to have several benefits. One
so juveniles are better off leaving home than competing potential benefit is that dispersers avoid competition
with relatives for limited resources. with kin for critical resources in their natal area (e.g.,
What benefits might be associated with philopatry? Perrin and Lehmann 2001). Some evidence to support
Populations of animals can become adapted to local con- this benefit comes from wolf spiders (Pardosa monticola),
ditions. Combinations of genes that work well together a species in which mothers help their offspring disperse.
under these local conditions might be favored by selec- Dries Bonte and colleagues (2007) studied natal disper-
tion. An animal that disperses from its birth site and set- sal in wolf spiders in the laboratory and in coastal dune
tles elsewhere, in an area with different ecological grasslands of Belgium. Before describing their findings,
conditions, may not be as well adapted to its new home. we should mention a few things about the natural his-
Another benefit of remaining near the birthplace tory of wolf spiders. Wolf spiders do not live in webs;
is familiarity with the local physical and social setting. instead, they employ a combination of active foraging for
Such familiarity may enable philopatric young to be prey and sit and wait strategies, depending on energy
efficient not only at finding and controlling food but demands. Young hatch from an egg sac that the female
also in escaping from predators. Familiarity with fam- carries, and then climb onto her abdomen where they
ily and neighbors is also likely to reduce the levels of live off their own stored energy reserves (i.e., they do not
aggression and stress associated with social interactions. feed at this stage). Observations of newly independent
In short, philopatric young may live longer and leave spiderlings suggest that they display limited mobility
more offspring because of the relatively low risks and with respect to ballooning (using strands of silk to travel
energy use associated with living in familiar surround- by wind) and moving on the ground. Given their lim-
ings (Shields 1982). ited mobility, Bonte et al. (2007) predicted that spider-
Sometimes offspring remain at home due to con- lings would face strong kin competition if they all
straints, such as a shortage of potential mates or lack of simultaneously left their mother’s abdomen to occupy
suitable territories in which they might settle. the same general location (brood size, they discovered,
Constraints on dispersal are discussed in more detail in can be as large as 72 spiderlings). It turns out that spi-
Chapter 19. derlings do not simultaneously disperse from mom’s
abdomen; instead, broodmates dismount gradually over
COSTS AND BENEFITS the course of about 185 hours following hatching (Figure
11.2a). Equally interesting is the observation that
OF NATAL DISPERSAL
females carrying spiderlings are much more mobile than
Dispersing also has costs and benefits. Dispersers often either females carrying egg sacs or females without egg
face high energy costs and increased predation risk sacs or spiderlings (Figure 11.2b). Bonte et al. (2007) sug-
(Metzgar 1967; Nunes and Holekamp 1996). This is par- gest that through their increased movement at this par-
ticularly true for small mammals that inhabit under- ticular time, wolf spider mothers actively spread their
ground burrows (a lifestyle described as fossorial) that offspring throughout the environment. According to the
236 Chapter 11 / The Ecology and Evolution of Spatial Distribution

a b
1.0 12

Distance of movement bout (cm)

spiderlings that leave the mother

0.9 11
10
Cumulative proportion of

0.8
9
0.7 8
0.6 7
0.5 6
0.4 5
4
0.3
3
0.2 2
0.1 1
0 0
0 20 40 60 80 100 120 140 160 180 200 +Eggsac +Spiderlings None
Time after hatching (hours)

FIGURE 11.2 Patterns of natal dispersal and maternal movement in wolf spiders ensure that siblings do not compete
with one another. (a) Spiderlings dismount from their mother’s abdomen gradually over a period of about 185 hours
after hatching, rather than leaving all at once. (b) Female spiders carrying spiderlings move greater distances than
females carrying egg sacs or females without egg sacs or spiderlings (labeled as “None” in the graph). (Modified from
Bonte et al. 2007.)

Bonte study, this behavior, combined with the asyn- avoidance of inbreeding remains a possible explanation
chronous departures of individual young from their for this pattern (Holekamp and Sherman 1989).
abdomen, promotes the avoidance of competition with Similarly, observations of brown bears in two different
kin for resources. study sites in Sweden revealed that nearly all males leave
In other species, avoiding competition with kin can- home (94% of those studied), regardless of ecological
not explain patterns of natal dispersal. For example, if factors such as population density and sex ratio. In con-
competition for resources in the natal area is an impor- trast, only 41% of female brown bears dispersed from
tant factor leading to dispersal, then we might predict their natal site (Zedrosser et al. 2007). Predominant (or
that within a species, dispersal would increase with more distant) dispersal by one sex is predicted if indi-
increases in litter size. However, in brown bears viduals disperse to avoid inbreeding. For the most effec-
(Zedrosser et al. 2007) and prairie voles (McGuire et al. tive inbreeding avoidance, we would also expect
1993), the likelihood of natal dispersal is not associated dispersers to leave home before becoming reproductive,
with litter size. In fact, natal dispersal is more common and many do (Dobson 1982).
from small than large groups of prairie voles. This find- There has been considerable debate over the role of
ing is the precise opposite of what we would predict if inbreeding in natal dispersal, with some scientists argu-
avoidance of competition for resources at home was an ing that it plays a significant role and others describing
important function of natal dispersal in this species. its role as irrelevant (reviewed in Perrin and Goudet
Competition for mates also does not explain natal dis- 2001). Indeed, some authors have argued that animals
persal in prairie voles, for dispersal occurs at similar lev- should rely on mate choice rather than a risky behavior
els from natal groups having high levels of mate like dispersal to avoid inbreeding (Moore and Ali 1984).
competition (number of competitors 7 number of Young female mammals, for example, may avoid select-
potential mates) and low levels of mate competition ing their brothers as mates, and in response, their broth-
(number of competitors 6 number of potential mates). ers disperse from the natal site to find females willing to
Avoidance of inbreeding is another potential bene- mate with them. According to this argument, individu-
fit of natal dispersal. Evidence that some young animals als do not disperse from home to avoid breeding with
leave home to avoid breeding with close relatives comes relatives; instead, natal dispersal is the result of females
in several forms. For example, some evidence involves not choosing male relatives as mates.
observations that young disperse in response to the pres- Spotted hyenas (Crocuta crocuta) seem to be a
ence of an opposite-sex parent (e.g., Gundersen and species in which female mate choice drives natal disper-
Andreassen 1998; Le Galliard et al. 2003; Wolff 1992). sal by males. In this species, dispersal by females from
Other evidence comes in the form of all members of one the natal clan is very rare; in contrast, most, but not all,
sex dispersing no matter what the ecological or social males leave their natal clan. Thus, predominant disper-
conditions. In Belding’s ground squirrels, for example, sal by one sex occurs in spotted hyenas, but does this
all males leave home (and very few females do) regard- reflect female mate choice or inbreeding avoidance? In
less of the levels of competition for mates or resources; an effort to sort out the causes of natal dispersal in male
 Remaining at Home Versus Leaving 237

spotted hyenas, Höner and colleagues (2007) studied males that were members of your clan when you were
eight clans in northern Tanzania from April 1996 to born and select as mates those males that arrived in your
April 2006. Using behavioral observations, demographic clan (through birth or immigration) after your birth.”
data, and genetic analyses of paternity, the researchers This rule would reduce the chances that females would
examined patterns of female mate choice and tested four mate with their father and older brothers. The paternity
hypotheses for natal dispersal by males. Specifically, they analyses showed that nearly 90% of 134 litters examined
examined whether males disperse in order to avoid (1) were sired by males that were born into or immigrated
competition with other males for mates, (2) breeding into the female’s clan after she was born (Figure 11.3a).
with close female relatives, or (3) competition for food Paternity analyses also revealed another aspect of female
resources. The fourth hypothesis tested was that males choice: young females were more apt than older females
disperse in response to patterns of mate choice by to select males with shorter residency times in their clan
females. To test these four hypotheses, Höner et al. (refer again to Figure 11.3a). Given these patterns of
(2007) examined which of the following four variables female mate preferences, the researchers predicted that
(keyed to each hypothesis) predicted the clan in which males would begin their reproductive career in the clan
males began their reproductive career: (1) intensity of (natal or otherwise) that had the greatest number of
male–male competition in the clan (estimated by the young females. This was, in fact, what they found. After
number of reproductively active natal males and immi- making short-term forays into the territories of nearby
grant males), (2) number of unrelated adult females in clans (presumably to evaluate dispersal options), males
the clan (defined as those with a coefficient of related- select as their place to begin reproduction the clan with
ness < 0.5), (3) number of main prey animals in the clan the highest number of young females (remember, these
territory per adult or yearling hyena at the time of clan are the females most likely to choose them as mates).
selection, or (4) number of young females most likely to Importantly, such a choice results in long-term fitness
breed with males. Clan selection by males was indepen- benefits for the males (Figure 11.3b). We see from this
dent of the first three variables (intensity of male–male example that patterns of mate choice can be key to
competition, number of unrelated females, and per understanding patterns of natal dispersal. We will have
capita number of available prey). These findings suggest more to say about mate choice in Chapter 14.
that competition for mates, inbreeding avoidance, and We have focused on the ultimate causes of natal dis-
competition for food do not cause natal dispersal by persal, such as avoidance of inbreeding and competition
males. What about patterns of mate choice by females? with kin for resources or mates. There are also proximate
Using paternity analyses of litters, Höner et al. (2007) causes of natal dispersal (reviewed in Lawson Handley
found that females adhered to the following general rule and Perrin 2007; Nunes 2007). Factors suggested to trig-
to avoid mating with close relatives: “avoid mating with ger natal dispersal include attainment of sufficient body

a Litters sired by males that arrived in the b 2.5

clan through birth or immigration after
the mother’s birth
Long-term reproductive success

Litters sired by males that were present

in the clan when the mother was born
2.0
9
8
Tenure of father (years)

7 1.5
6
5 1.0
4
3
2 0.5

1
0 0.0
0 1 2 3 4 5 6 7 8 9 10 11 Most Fewer
Age of mother (years) Number of young females in clan selected by males

FIGURE 11.3 Natal dispersal by male spotted hyenas is driven by patterns of female mate choice. (a) Shown here are
data from 134 litters (each circle represents a litter). Most litters are sired by males that arrive (through birth or
immigration) in the mother’s clan after her birth. In addition, young females are more apt than older females to select
males with shorter residency times in their clan. (b) Males that begin their reproductive careers in clans with the
highest number of young females have higher long-term reproductive success than males that select clans with
fewer young females. (From Höner et al. 2007.)
238 Chapter 11 / The Ecology and Evolution of Spatial Distribution

size or fat reserves, aggression from other group mem- following factors, sometimes in combination: (1) inbreed-
bers, shortage of food at the natal site, attraction to ing avoidance; (2) local resource competition; (3) local
opposite sex individuals in other groups, and weakening mate competition; and (4) cooperative behavior among
of social bonds with members of the natal group. There kin. Mammals have been best studied with respect to
is evidence from some species that androgens organize natal dispersal (Lawson Handley and Perrin 2007), and
dispersal behavior (see Chapter 7 for a discussion of the some explanations focus on them. The first hypothesis we
organizational effects of steroid hormones on behavior). discuss, however, was developed specifically to explain the
Natal dispersal has also been linked to individual dif- different dispersal patterns in mammals and birds.
ferences in personality traits such as “boldness” or
exploratory behavior. In great tits, for example, post-
Local Resource Competition and Inbreeding
fledging movement distances of individual females in a
wild population correlate with their scores on a labora- One hypothesis is that a sex bias in dispersal evolved as
tory test of exploratory behavior. More specifically, a way to avoid the genetic costs of inbreeding while
females that score high on the laboratory test (a score enjoying the benefits of familiarity with local physical
equals the number of movements made during a two- and social conditions. A sex bias in dispersal seems to be
minute test in a novel environment) move further dur- the perfect compromise: extreme inbreeding is pre-
ing postfledging movements than females with low vented because members of one sex disperse, and indi-
exploratory test scores (Dingemanse et al. 2003). The viduals of the other sex experience the benefits of
positive correlation between exploratory behavior and philopatry (Greenwood 1980).
distance dispersed does not characterize all species, how- Although this hypothesis explains why sex biases in
ever. Radiotracking of Siberian flying squirrels before dispersal tendencies occur, it does not explain which sex
and after dispersal indicates that long-distance dispersers leaves home and why the direction of the bias differs in
explore less than short-distance dispersers (Selonen and birds and mammals. Two explanations have been sug-
Hanski 2006). gested for the direction of the sex bias.
The first idea is that the sex most involved in terri-
tory acquisition and defense should stay home because
SEX BIASES IN NATAL DISPERSAL it benefits most from familiarity with the natal territory
As we have seen, males and females of a particular (Greenwood 1980). Most birds are socially monogamous
species often differ in whether or not they disperse from (i.e., they live in male-female pairs), and males usually
their birthplace. Even more striking is the observation compete for territories that attract females, rather than
that the direction of the sex bias in natal dispersal dif- competing for females directly. This is called a resource-
fers between birds and mammals (Table 11.1). In the defense mating system. Under conditions such as these,
majority of bird species that have been studied, females familiarity with a particular area might be more impor-
are more likely to disperse than males. In mammals, tant to males than to females; thus males should be
however, just the reverse is true—males are more likely philopatric. Female birds might disperse to avoid the
to disperse than females. genetic costs of close inbreeding and to choose territo-
We can ask what drives sex biases in natal dispersal, ries with the best resources.
and why does the bias differ in direction between birds In contrast, most mammals exhibit mate-defense
and mammals? Several hypotheses have been proposed polygyny; that is, a single male defends a group of
to explain sex-biased dispersal (reviewed by Lawson females. In this mating system, males directly compete
Handley and Perrin 2007). Those that we will consider with one another for females rather than territories with
suggest that sex differences in natal dispersal reflect the resources. Young or subordinate males, unable to com-
pete successfully for access to females, may disperse to
increase their chances of mating. Also, female mammals
often live in matrilineal social groups (groups of moth-
TABLE 11.1 Number of Species of Mammals ers, daughters, and granddaughters) in which the bene-
and Birds in Which Natal Dispersal Is fits of living with kin may be quite high. Because of this
Male-Biased, Is Female-Biased, or in Which
social system, females benefit most by staying home and
Offspring of Both Sexes Disperse
males may benefit by dispersing to avoid the genetic
Predominant dispersing sex costs of extreme inbreeding. Thus, female-biased dis-
persal in birds seems to be linked to resource-defense
Male Female Both mating systems, and male-biased dispersal in mammals
Mammals 45 5 15 seems to be linked to mate-defense mating systems
(reviewed in Clarke et al. 1997).
Birds 3 21 16
What about the exceptions, for instance, mammals
Source: Data from Greenwood (1980). that display resource-defense (rather than mate defense)
 Remaining at Home Versus Leaving 239

polygyny? What is the direction of sex bias in natal dis- TABLE 11.2 Number of Species of Mammals
persal in these species? In some mammals with resource- in Which Natal Dispersal Is Male-Biased, Is
defense mating strategies, natal dispersal is female Female-Biased, or in Which Offspring
biased, as would be predicted by Greenwood (1980). of Both Sexes Disperse, as a Function of Type
This is the case, for example, for sac-winged bats (Nagy of Mating System
et al. 2007). However, European roe deer also defend
Predominant dispersing sex
resources rather than mates and dispersal is not female
biased in this species, illustrating that a resource-defense Mating system Male Female Both
mating system does not necessarily lead to female-biased
natal dispersal in mammals (Coulon et al. 2006). Monogamous 0 1 11
The second idea to explain the direction of the sex Polygynous
bias is that the sex that gets first choice of breeding sites or promiscuous 46 2 9
is the one that remains in the natal area; the other sex Source: Data from Dobson (1982).
disperses. This model was first developed to explain sex-
biased dispersal patterns in mammals (Clutton-Brock
1989), but it was later extended to birds (Wolff and
Plissner 1998). In either case, the model assumes that
philopatry is more desirable than dispersal. According to When dispersal data for species of mammals with
this model, mating systems affect the dispersal patterns different mating systems were examined, they revealed
of mammals indirectly by influencing whether the remarkable agreement with this hypothesis (Table 11.2).
father will be present when his daughters are old enough However, avoidance of inbreeding is probably a signifi-
to breed. If he is not, females have first choice of the cant influence on natal dispersal in many species because
breeding site, and they choose to stay at home. However, reduction in competition for mates can’t explain all sex
if the father is still around when his daughters reach differences in natal dispersal. (For instance, it doesn’t
sexual maturity, he has first choice of the breeding site, explain why females are more likely than males to dis-
and so females disperse to avoid inbreeding. perse in monogamous birds.)
Because female mammals nurse their young, a role
for males in early parental care is necessarily limited,
Cooperative Behavior Among Kin
and so in most species males have little involvement in
offspring care. This allows males to avoid long-term When discussing the benefits of natal dispersal, we men-
pair bonds, and they are free to wander over large areas. tioned that leaving home allows young animals to avoid
When competition over mates is intense, as among ele- competing with kin (for resources or mates) and to avoid
phant seals or red deer, a male’s opportunity to breed breeding with kin. However, if kin exhibit cooperative
may be limited. As a result, he is likely to be gone before behavior, then it might be beneficial to stay at home.
his daughters are old enough to reproduce. Thus, Furthermore, if cooperation benefits one sex more than
daughters don’t have to disperse to avoid inbreeding. the other, then we would expect the cooperative sex to
However, in those species in which a male’s reproduc- be philopatric. We would expect the other sex to disperse
tive life span is long and he is present when his daugh- to avoid inbreeding because inbreeding costs for this sex
ters are old enough to breed, as in chimpanzees, the are not counterbalanced by advantages gained through
females usually disperse (Clutton-Brock 1989; Wolff kin cooperation (Perrin and Goudet 2001). In this way,
1994). cooperative behavior can contribute to sex biases in natal
dispersal.
One prediction that arises from the kin cooperation
hypothesis is that the magnitude of sex-biased dispersal
Local Mate Competition should increase with increases in social complexity.
Another hypothesis is that differences between males and Consistent with this prediction is the observation that
females in levels of competition for mates might be particularly dramatic sex differences in natal dispersal
involved in sex differences in the dispersal tendency in often characterize highly social polygynous mammals
mammals (Dobson 1982). Because most mammals are (Pusey 1987; Smale et al. 1997). Other evidence comes
polygynous, competition for mates would be more from polygynous ground-dwelling sciurids, a group of
intense among males than among females, and thus dis- rodents that includes the ground squirrels (Spermophilus
persal should be more common in males. Furthermore, spp.), marmots (Marmota spp.), and prairie dogs
in species with monogamous mating systems, levels of (Cynomys spp.). This group is particularly well suited for
competition for mates would be more equal between the testing the relationship between degree of sex bias in
sexes, and males and females should disperse in similar natal dispersal and social complexity because social
proportions. structures range from solitary to large social groups with
240 Chapter 11 / The Ecology and Evolution of Spatial Distribution

FIGURE 11.4 In polygynous ground-dwelling

Cynomys gunnisoni sciurids (members of the squirrel family), the
60 degree of sex bias in natal dispersal (calculated
Cynomys ludovicianus
as the difference between male and female dis-
persal rates) increases with social complexity.
Sex-biased dispersal

Spermophilus columbianus Marmota flaviventris

Shown here are data from two species of
40 marmots (Marmota spp.), two species of
Spermophilus towsendii
prairie dogs (Cynomys spp.), and seven species
Spermophilus tereticaudus of ground squirrels (Spermophilus spp.).
Marmota monax (From Devillard et al. 2004.)
Spermophilus beldingi
20
Spermophilus richardsonii

Spermophilus elegans

Spermophilus variegatus
0

0.2 0.4 0.6 0.8 1.0 1.2

Social complexity index

individuals from several generations. Sébastien NATAL DISPERSAL AND

Devillard and colleagues (2004) reviewed the literature CONSERVATION BIOLOGY
on sciurids and extracted for 11 species natal dispersal
rates (number of animals that leave their natal site or Before moving to the topic of habitat selection, we
social group before their first reproduction/total num- should mention that natal dispersal has important con-
ber of potential dispersers in the population). They used sequences for the genetic structure of populations. Small
an index of social complexity previously developed for populations cut off from other populations may lose
ground-dwelling sciurids. The index, thought to better genetic diversity through genetic drift, but immigration
describe degree of social complexity than group size brings in new genes. Not surprisingly, then, natal dis-
alone, considers the types of social interactions possible persal has important implications for the conservation of
within groups (e.g., the extent of interaction between populations and species. Habitat destruction by humans
individuals of different age and sex classes) (Blumstein has fragmented populations of many species and has
and Armitage 1997). When Devillard et al. (2004) ana- made inhospitable the areas between habitat fragments.
lyzed data from the literature, they found that male- Such changes hinder dispersal. As part of our efforts to
biased natal dispersal characterized all 11 species. In conserve threatened populations, it may be critically
addition, plots of sex bias in natal dispersal (calculated important to preserve dispersal corridors (narrow areas
as the difference between male and female dispersal that connect other, larger areas of habitat) that promote
rates) versus index of social complexity revealed that gene flow. Knowledge of dispersal patterns will also help
male-biased dispersal increased with social complexity to predict range expansions of introduced predators and
(Figure 11.4). However, the increased bias resulted pests (Macdonald and Johnson 2001). We see, then, that
solely from increases in male dispersal and not from understanding a behavior such as natal dispersal can have
expected concomitant decreases in female dispersal. So, important conservation implications.
in ground-dwelling sciurids, data for males, but not
females, support Perrin and Goudet’s (2001) model STOP AND THINK
regarding the relationship between sex-biased dispersal In the field, natal dispersal and philopatry are often mea-
and social complexity. sured by mark and recapture methods (e.g., catch an ani-
Sex biases in natal dispersal are likely caused by sev- mal, mark it for individual identification, release it, and
eral interacting factors, including mating system, then subsequently recapture it at the same or a different
inbreeding avoidance, kin competition, and kin cooper- location) (Figure 11.5). Sometimes genetic methods are
ation. Although mammals have been most studied in this used in combination with mark-recapture approaches to
regard, studies with birds are increasing, and sex biases examine how dispersal translates into gene flow. Given
in dispersal have also been documented in other groups what you know about natal dispersal, what are some
of vertebrates, including fishes and amphibians. potential problems in using mark-recapture methods to
study dispersal in natural populations? For example, how
Increased study of natal dispersal in these groups, as well
easy would it be to know whether a particular animal dis-
as in invertebrates, will permit testing the generality of
persed from home or died? How might you design a study
the hypotheses put forth largely from mammalian data to minimize this issue?
(Lawson Handley and Perrin 2007).
 Habitat Selection 241

a question, “What features of the habitat might dispersers

use to judge habitat quality?”

INDICATORS OF HABITAT QUALITY

Animals seem to have clear habitat preferences. For
example, the forest buffalo (Syncerus caffer nanus) of
western and central Africa prefers grassy clearings and
open stands of forest with large trees and open canopy
to other types of forest (e.g., mixed, riparian, and for-
est dominated by a single tree species) (Melletti et al.
2007). This preference is particularly striking given that
clearings are quite rare, representing only 1% of habi-
tats in some areas. Because forest buffalo are social
creatures, they may prefer clearings, particularly as
resting spots, because members of the herd can easily
b maintain visual and physical contact. Let’s take a look
at what general features animals might evaluate when
selecting a habitat.
Indicators of habitat quality include the presence of
resources (e.g., food, nest, and rest sites), conspecifics,
and heterospecifics (members of another species)
(Stamps 2001). As you might imagine, dispersers may
lack the time and energy needed to complete in-depth
assessments of critical resources in the habitats they
encounter in their search for a new home, so how do
they evaluate real estate? Evidence suggests that some
dispersers rely on “quick and dirty” cues to assess the rel-
ative quality of prospective settlement sites. For exam-
ple, young lizards (Anolis aeneus) that leave their natal site
to search for feeding territories spend only about six
hours evaluating a particular location. Given their var-
FIGURE 11.5 Mark-recapture methods are often used to ied diet of arthropods, this time period is probably too
monitor natural populations of small mammals. short to permit a detailed assessment of prey availabil-
Repeated live-trapping provides data on behavior ity. So, rather than assessing arthropod availability at
(e.g., dispersal and social organization) and demography each site, the lizards seem to assess habitat characteris-
(e.g., changes in population density). (a) A student at tics such as light intensity and amount of leaf litter.
the University of Illinois checking a live trap. These characteristics correlate with prey availability and
(b) Another student handling a prairie vole removed
lend themselves to more rapid evaluation than the
from a live trap at the same study site.
painstaking task of evaluating the local availability of sev-
eral different prey species (Stamps 1994).
What about conspecifics? Should dispersers evalu-
ate their presence or absence in the vicinity of a prospec-
HABITAT SELECTION tive home? Absence of conspecifics from a particular site
might be a good thing. After all, if an individual settles
Animals that disperse from their natal site or breeding into an unoccupied site, then it would avoid all that nasty
site must eventually select a new location in which to set- intraspecific competition for resources. There are two
tle. This process of habitat selection can be divided into main ideas concerning how fitness might change with
three phases: (1) search (animal searches for a new habi- number of conspecifics in an area. One model, called the
tat), (2) settlement (animal arrives in a new habitat and ideal free distribution, predicts that individual fitness will
begins to establish a home range or territory), and decline as the number of conspecifics in a patch increases
(3) residency (animal lives in the new habitat) (Stamps (Fretwell 1972). Another idea suggests that individual fit-
2001). Generally speaking, the phases of search and set- ness increases with the number of conspecifics at low to
tlement are costly times for dispersers, while residency moderate densities, and then declines from moderate to
is when the benefits of habitat selection accrue. We will high densities (Allee 1951). According to this “Allee
focus largely on the search phase, and we begin with the effect,” having a few neighbors is beneficial, especially
242 Chapter 11 / The Ecology and Evolution of Spatial Distribution

when it comes to defense against predators and access to one species to resident birds of other species. There are
potential mates. several possible explanations for why migrants might be
Presence of conspecifics in an area could also serve attracted to heterospecific residents. First, residents
as a source of information about habitat quality (Stamps could indicate high-quality habitat. Second, migrants
1988). Juvenile lizards (A. aenus), for example, seem to may experience food or safety benefits from grouping
use conspecific presence as an indirect cue to habitat with heterospecific residents (as described for mixed-
quality. Juveniles were allowed to view two sites of species flocks). Third, using presence of heterospecific
equivalent quality; a territory owner was present on one residents as an indicator of habitat quality may be a faster
site but not on the other. After ten days, the territory and more accurate method of habitat assessment than
owner was removed and the juveniles were allowed to independent sampling upon arrival at potential breeding
select between the two sites. The young lizards pre- sites (residents, after all, have all year to assess the qual-
ferred the previously occupied site to the equivalent, ity of various habitat patches as compared to migrants
unoccupied site (Stamps 1987). To summarize, then, who have a much shorter time period).
when it comes to habitat selection, there are two gen- Jukka T. Forsman and colleagues (2002) examined
eral explanations for the phenomenon of conspecific whether the presence of resident titmice (Parus major,
attraction: (1) Allee effects in which individuals benefit P. montanus, and Cyanistes caeruleus) influences the set-
from the presence of conspecifics after arriving at a site tlement and fitness of migrant pied flycatchers (Ficedula
(e.g., through enhanced detection of predators or hypoleuca). Titmice and pied flycatchers overlap in their
access to mates), and (2) individuals’ use of conspecifics use of resources; all species nest in cavities and forage for
as an indicator of habitat quality. arboreal arthropods using hovering and foliage gleaning.
Beyond the mere presence of conspecifics, might Forsman et al. (2002) conducted their experiments on
individuals looking for a place to settle evaluate certain two spatial scales: landscape and nest site. In the land-
characteristics of resident conspecifics? The answer scape scale experiment, they removed all titmice from
appears to be yes. There is good evidence, for example, some forest patches (to achieve a density of zero) and
that breeding birds monitor the reproductive success of added titmice to other forest patches (to achieve a den-
conspecifics in their local area and then use this infor- sity of about five pairs per 10 ha); these density manip-
mation to decide where to nest during future breeding ulations were performed prior to the arrival of
efforts (Doligez et al. 2002; Parejo et al. 2007). flycatchers in the spring, and the achieved densities were
Information concerning local conspecifics is described as maintained throughout the breeding season. In the nest-
“public information” in order to differentiate it from site scale experiment, flycatchers were allowed to choose
“personal information.” Personal information is the focal between a nest box close to an active titmouse nest (dis-
bird’s own breeding success at a particular site, which tance of 25 m) or farther away from an active titmouse
also affects whether it will remain at that site or leave. nest (distance of 100 m). In both experiments, the
The presence of heterospecifics in a particular habi- researchers recorded the arrival dates of flycatchers and
tat can have costs and benefits for an animal consider- several measures of fitness for flycatcher broods pro-
ing whether to settle there. For example, interactions duced at the different sites. At both landscape and nest-
between species that share mutual resources could be site scales, flycatchers were attracted to the vicinity of
negative due to interspecific competition. Interactions titmice. Flycatchers tended to arrive earlier on forest
between heterospecific individuals can also be beneficial, patches where titmice numbers were increased than on
however, such as when birds form mixed species flocks patches where titmice had been removed, and they pre-
and experience the benefits of enhanced food acquisition ferred nest boxes placed near an active titmouse nest to
and antipredator behavior. Thus, moving in where there those located farther away. With regard to the fitness
are neighbors of other species may be beneficial under effects of habitat selection, brood sizes of flycatchers
certain circumstances. This idea has been formalized as were larger in patches where numbers of titmice had
the heterospecific attraction hypothesis, which states that been experimentally increased than in patches from
individuals choose habitat patches based on the presence which titmice had been removed (landscape scale exper-
of established residents of another species (Mönkkönen iment), and flycatchers breeding closer to titmice had
et al. 1999). This model predicts that individuals search- larger nestlings than those breeding farther away (nest-
ing for a new home will display the strongest attraction site scale experiment) (Figure 11.6). These results sup-
to heterospecifics when the benefits of social aggregation port the heterospecific attraction hypothesis. (Note also
outweigh the costs of competition and when the costs of that the results are just the opposite of what we would
independent sampling of habitats (i.e., evaluating habi- expect if interspecific competition were the defining
tat quality on one’s own rather than using the presence interaction between titmice and flycatchers.)
of heterospecifics as a cue) are high. We see that individuals searching for a home may
An interesting test of the heterospecific attraction evaluate multiple and diverse characteristics of potential
hypothesis concerns the attraction of migrant birds of settlement sites, ranging from physical features of the
 Habitat Selection 243

a b tances in relatively straight lines before establishing res-

7 16 idence in an area. Decisions by dispersers using sequen-

nestling flycatchers (g)

tial search are influenced by factors such as the total time
nestling flycatchers

15
available for the search, quality of available habitats, and

Body mass of
Number of

6 14 how often high-quality habitats are encountered. In

13 addition, dispersers using sequential search tactics may
5 have an acceptance threshold, and this threshold may
12
decline as the search continues (i.e., an animal may be
11
4 highly selective at the start of the search and less selec-
Add Rem Near Far
tive over time). Comparative and sequential search tac-
FIGURE 11.6 In support of the heterospecific attraction tics do not sort out by taxon; indeed, both tactics can
hypothesis, migrant pied flycatchers are attracted to sites occur within a single population of a species (e.g., Byrom
with resident titmice, and this attraction results in higher and Krebs 1999).
fitness for the flycatchers. (a) Numbers of flycatcher
nestlings 13 days post hatching were greater in patches
where numbers of titmice were experimentally increased EFFECTS OF NATAL EXPERIENCE
(Add) than in patches from which titmice had been
Experience in the natal environment appears to be an
removed (Rem). (b) Flycatchers breeding near titmice
important influence on habitat selection. As an exam-
had larger nestlings than those breeding farther away.
ple, let’s consider the host preferences exhibited by the
parasitoid wasp (Pachycrepoideus vindemia). A parasitoid
is an organism whose offspring develop on or within a
habitat to details of the social scene. We next consider the host, eventually killing the host; in our discussion of
strategies used by animals when searching for a new home. habitat selection, then, the host of a parasitoid is the
equivalent of a habitat. In other words, just as other ani-
mals select habitats within which to reproduce, female
SEARCH TACTICS
parasitoids select hosts on (or within) which their off-
Animals use two general search strategies during habi- spring will develop. Young P. vindemia develop on the
tat selection: comparison and sequential search (Figure pupae of several different fly species. R. J. Morris and
11.7) (reviewed in Stamps and Davis 2006). The first tac- M. D. E. Fellowes (2002) conducted laboratory exper-
tic, comparison, entails visiting several areas, revisiting iments with female P. vindemia to determine the effects
some, and then choosing the area judged to be the high- of natal host and experience on subsequent host pref-
est quality. This tactic is often described as a best-of-N erence. In terms of natal host, female wasps were
strategy because the animal visits N habitats and then allowed to develop on either fruit flies or houseflies.
selects the best one. In contrast, animals using sequen- Following development on their host, the experience of
tial search tactics arrive at a location, decide whether to females was manipulated by placing females in vials (for
accept or reject it as a place to live, and in the event they 24 hours) either alone, with two fruit fly pupae, or with
reject it, continue their search. Animals using sequential two housefly pupae. Thus, there were three different
search tactics do not return to areas they have already groups regarding experience: (1) no opportunity to
visited (except by chance), and they often travel long dis- attack either fruit flies or houseflies; (2) opportunity

a b

FIGURE 11.7 Animals searching for a new place to settle employ two general strategies called comparison and sequen-
tial search. (a) In comparison (best of N) strategies, an animal visits several areas, revisits some, and then chooses the
area judged to be the highest quality. (b) During sequential search, an animal does not return to sites except by
chance, and tends to travel long distances in relatively straight lines.
244 Chapter 11 / The Ecology and Evolution of Spatial Distribution

to attack fruit flies; and (3) opportunity to attack house-

flies. At the end of their 24-hour experience, females
were tested for host preference by placing them in a
Petri dish for two hours with ten fruit fly pupae and ten
housefly pupae. To assess host preference, the
researchers monitored the number of emerged parisi-
toids, emerged flies, and dead pupae in each dish. The
results indicate that host preference in P. vindemia is
influenced by an interaction between natal host and
subsequent experience. Female wasps reared on fruit
flies preferred fruit flies as hosts, but only when allowed
to gain experience attacking fruit flies, which they did
preferentially. Similarly, female wasps reared on house-
flies preferred houseflies as hosts, but only when
allowed to gain experience attacking houseflies, which FIGURE 11.8 Many translocated animals travel long dis-
they, too, did preferentially. The authors suggest that tances from the site of release. In cougars, this is espe-
cues from the natal host prime females to respond to cially true for adult males.
these cues should they be encountered again, and when
reencountered, the cues are learned by females, estab-
lishing host preference. We see, then, that the natal and then releasing them to the wild) often fail because
environment (here, the natal host) can influence selec- animals move rapidly away from the site of release, often
tion of a site for reproduction. traveling long distances (Figure 11.8) (reviewed in
The general phenomenon whereby an animal’s Stamps and Swaisgood 2007). Long-distance travel is
experience in its natal habitat induces a preference for a associated with high mortality, and those individuals that
postdispersal habitat with similar qualities is called natal escape death and eventually settle may exhibit declines
habitat preference induction (NHPI) (Davis and Stamps in condition and depressed reproduction. In addition,
2004). Recall that the search phase of habitat selection rapid, long-distance movements from the release site
is thought to be costly in terms of time and energy, not make postrelease provisioning and monitoring extremely
to mention the risk of predation associated with dis- difficult, further compromising conservation efforts.
persing through unfamiliar terrain. Thus, it is possible Indeed, translocations and captive-release programs tend
that NHPI helps a disperser to more quickly and effi- to have low success rates (11–75%).
ciently recognize a suitable habitat, thereby minimizing Rapid, long-distance movement from a release site
the costs of the search phase. Also, it seems reasonable suggests that the animal rejects the habitat at the release
to expect that a habitat similar to the one in which the site and is searching for suitable habitat in which to set-
disperser grew up would be of sufficient quality to set- tle (Stamps and Swaisgood 2007). An understanding of
tle in because the dispersing individual itself has survived habitat selection, in general, and natal habitat preference
to leave home. Another explanation for NHPI is that dis- induction, in particular, may inspire modifications to
persers will have greater fitness if they settle in a habi- these programs that could help animals find the release
tat similar to their natal habitat because their particular site a more acceptable place to settle. For example, if ani-
phenotype has been shaped by this type of habitat. For mals developing in captivity were provided with con-
example, while living at home, individuals may develop spicuous stimuli and cues (for example, odors or sounds)
specific methods for finding and capturing prey, and similar to those at the intended release site, then perhaps
these methods might work best in postdispersal habitats they would be more inclined to stay in the vicinity of the
similar to their natal habitat and less well in habitats dif- release site. In translocation efforts, placing stimuli and
ferent from their natal habitat. NHPI has been docu- cues from the original habitat at the release site might
mented in diverse taxa, including insects, fishes, reduce the disparity between the two habitats, making
amphibians, birds, and mammals (Davis and Stamps the release site a more acceptable place to settle. If taken,
2004) and may have significant implications for conser- these suggestions and others by Stamps and Swaisgood
vation biology, as we see next. (2007) have the potential to increase the success of
translocation and captive-release programs.
HABITAT SELECTION Another area in which habitat selection by animals
is relevant to conservation biology concerns ecological
AND CONSERVATION BIOLOGY
traps. An ecological trap is a low-quality habitat that ani-
Conservation efforts using translocation (moving ani- mals prefer over a high-quality habitat (Battin 2004).
mals from one part of their natural range to another) and One of the best examples concerns Cooper’s hawks
captive-release programs (breeding animals in captivity (Accipiter cooperii) living in Tucson, Arizona (Boal and
 Migration 245

Mannan 1999). Cooper’s hawks seem to select the city

as a home because of its plentiful nest sites and prey; MIGRATION
indeed, urban hawks nest earlier and have larger clutches
In some species, spatial distribution varies over time,
than hawks nesting in less urban areas outside Tucson.
often with the seasons. Indeed, there may be dramatic
However, nestling mortality is substantially higher in
mass movements of animals, and some of these are
Tucson (7 50%) than in areas surrounding the city
migrations. The term migration has been defined in
(6 5%), making Tucson an ecological trap for Cooper’s
many ways (reviewed in Dingle and Drake 2007). The
hawks. The very pigeons and doves that attract Cooper’s
definition that we will use compares migration to other
hawks to the city because of their abundance (these prey
movements made by animals that take them beyond their
species comprise 84% of the diet of city hawks) carry tri-
current home range for a significant period of time (e.g.,
chomoniasis; this disease is the primary cause of nestling
natal and breeding dispersal). During dispersal, individ-
mortality in the city-dwelling Cooper’s hawks. Thus, the
uals cease moving and settle into a new home range once
cues apparently used by Cooper’s hawks to select a habi-
a suitable location has been found. In contrast, migra-
tat (availability of nest sites and prey) lead them to select
tion is movement away from the home range that does
an inferior habitat in which disease causes dramatically
not stop upon encountering the first suitable location.
reduced survival of their offspring. Ecological models
Instead, migrating animals continue to move until they
suggest that populations living in an ecological trap
eventually become responsive to the presence of
move toward extinction, especially if the initial popula-
resources, such as nest sites and food, and then they stop
tion size is small (Battin 2004). The Cooper’s hawks in
(Dingle and Drake 2007). From this definition, we see
Tucson are not disappearing, however. Instead, the num-
that within a particular species, migratory movements
ber of hawks in the city is at least stable, probably
occur over greater distances than dispersal movements.
because birds continue to move into the city from the
In some species, migration involves the movement
surrounding areas. Unfortunately, we know little about
of animals away from an area and the subsequent return
the capacity of (and time required for) animals to evolve
to that area. In these cases, animals usually migrate
new habitat preferences or to adapt to new environ-
between breeding areas and overwintering, or feeding,
mental conditions. This lack of knowledge makes it dif-
areas (Figure 11.9). In other species, however, migration
ficult for us to predict whether species can escape from
is a one-way affair (i.e., it is not a round trip); many
ecological traps.
migratory insects, for example, permanently abandon
their site of origin. In general, round-trip migrations are
STOP AND THINK
associated with long-lived species (e.g., vertebrates) and
Many examples of ecological traps are associated with one-way migrations with short-lived species (e.g.,
human activities. However, ecological traps can occur
insects) (Dingle and Drake 2007). There is also variation
without human influence. Can you think of a situation
in the distance moved during migration. For example, a
without human involvement that would qualify as an eco-
logical trap? salamander may travel less than a kilometer from its
woodland home to the pond where it breeds. In other

FIGURE 11.9 Migrating caribou marching through Alaska in July. During the spring, many caribou
breed in the tundra. Beginning in July, they migrate south, where food will be available through
the winter.
246 Chapter 11 / The Ecology and Evolution of Spatial Distribution

species, the distances are truly astounding. Northern ele- breeding grounds (Fisher 1979). The mortality rate of
phant seals migrate twice a year—once to breed and black-throated blue warblers is at least 15 times higher
again mostly to eat—from beaches in Southern during spring and autumn migration than during peri-
California and Baja California, Mexico, to northern feed- ods when individuals are not migrating (Sillett and
ing grounds in the Aleutian Islands. Thus they migrate Holmes 2002).
about 8000 km each year, and that is just the horizontal
distance. They make frequent, deep dives that can add Energy
another 3000 km of vertical distance to their journey One reason for the enormous losses associated with
(Tennesen 1999). The arctic tern migrates even further, migration is that traveling such long distances requires
about 20,000 km one way, between its southern and its a great deal of energy. For instance, a bird uses about six
northern breeding area (Baker 1980). There are also to eight times more energy when flying than when rest-
long-distance migrants among the insects, perhaps best ing. Imagine yourself running 4-minute miles continu-
represented by desert locusts and monarch butterflies, ously for 80 hours. This would require roughly the same
with one-way migration distances of about 5000 km amount of energy per kilogram of body mass as a black-
(Waloff 1959) and 3600 km (Brower 1996), respectively. poll’s nonstop transoceanic flight of 105 to 155 hours
To further complicate matters (or to make things (Williams and Williams 1978). It is a good thing, how-
more interesting we should say!), migration can take sev- ever, that the blackpoll is not walking. It turns out that
eral forms (Berthold 2001; Dingle and Drake 2007). the energy costs of traveling a certain distance vary with
Migration can be obligate (an individual always migrates) the mode of locomotion. If each vertebrate animal in our
or facultative (an individual migrates if local conditions comparison obtained the same amount of energy from
deteriorate but stays put if conditions remain suitable). each gram of fat, then using 1 gram, a mammal could
There is also differential migration, in which the migra- walk 15 km, a bird could fly 54 km, and a fish could swim
tory patterns of individuals within a population differ by 154 km (Aidley 1981). However, even if it is easier for
age class or gender. In some small passerines (perching birds to travel great distances than it is for you, migra-
birds), individuals migrate in their first year but remain tion is still metabolically demanding.
on breeding grounds in subsequent years (the species Natural selection favors behaviors that reduce the
may even be described as resident in the breeding area), risk of starvation during migration. One way to do this
while in others, females tend to be more migratory than is to store fat before the journey begins. Gram for gram,
males. Sometimes both age and gender influence who fat provides more than twice the energy of carbohydrate
migrates; in blue tits, for example, there is a predomi- or protein. Thus it is an extremely important energy
nance of juveniles and females among migrants. These source during migration. It should not be surprising,
age and gender differences are usually attributed to juve- then, that migratory animals as diverse as insects, fish,
niles and females being less able to compete for food on birds, and mammals put on fat reserves prior to migra-
the breeding grounds when it becomes scarce during the tion (Berthold 2001). Indeed, the body mass of birds that
nonbreeding period (Berthold 2001). migrate over long distances may more than double
Why should an animal bother to travel hundreds before migrating (Klaassen 1996). (Increasing body mass
or thousands of kilometers to one location only to to avoid starvation, however, must be balanced against
return to its starting point half a year later? This ques- the potential negative effects of a heavy fuel load on
tion probably has many answers, and no single one flight performance.)
could apply to the diverse species that migrate. Even with premigratory fat deposition, some ani-
However, the simplicity of one explanation hides its mals do not have sufficient fat stores to complete their
profundity: those animals that migrate do so because journey without stopping. Certain small birds reduce
they produce more offspring this way. Although the their risk of starvation by refueling along the way.
actual costs and benefits of migration vary among Rather than making a nonstop trip, their migration con-
species, the benefits must result in the production of sists of alternating periods of flight and stopover. These
more offspring than would be possible if the individual birds tend to spend much more time in stopover than
stayed put. To explore this question further, we will in flight, making time spent in stopover the principal
consider some of the possible costs and benefits that determinant of the total duration of their migration
might accompany migration. (Alerstam 2003; Hedenström and Alerstam 1997).
Dividing migration into short episodes also allows birds
to store smaller energy reserves that do not compromise
COSTS OF MIGRATION
flight performance. How do they know where to stop
Migration takes a tremendous toll. Only half of the song- for food along the migratory route? Some evidence for
birds that leave the coast of Massachusetts each year ever insectivorous migrants moving between the Neotropical
return, and less than half of the waterfowl in North (Central and South America) and Nearctic regions
America that migrate south each fall return to their (North America) suggests that they use landscape fea-
 Migration 247

tures, such as the amount of hardwood forest, when of a long journey, but also during the journey itself.
deciding where to take a refueling break (Buler et al. Some migratory species experience heavy predation
2007). Amount of hardwood forest, it has been found, because their predators follow their seasonal move-
is positively correlated with arthropod abundance in the ments. Lions, cheetahs, and hyenas often track the
understory, something that should be of interest to hun- movements of African ungulates (Schaller 1972); wolves
gry insect-eating birds. follow North American caribou (Sinclair 1983); and
Not all migratory birds, however, make extended water pythons migrate seasonally to exploit their migra-
stopovers. Osprey (Pandion haliaetus) often exhibit fly- tory prey, the dusky rat (Madsen and Shine 1996).
and-forage migration, a strategy that combines for- Songbirds leaving North America in the autumn for
aging with covering migration distance (Strandberg southern overwintering grounds face some 5 million
and Alerstam 2007). Individual ospreys deviate some- raptors (birds of prey) that are also making the trip and
what from their migratory route to forage for fish in feeding along the way. And, of course, there are non-
nearby bodies of water but tend to move on in less migratory predators at stopovers. We see, then, that the
than 12 hours (Figure 11.10). This strategy may be predator landscape for migrants has two components:
used by other birds that fly extensively while foraging, one related to nonmigratory predators located along the
such as falcons and seabirds. route and at the end points, and a second consisting of
migratory predators. It has been suggested that for birds
the timing and routing of migration have been shaped,
Risk of Predation
at least in part, by predation risk (Ydenberg et al. 2007).
Many weary migrants fall to predators. For example,
songbirds are “fast food” for the Eleanora’s falcon. The
songbirds, worn out by their flight across the Risk of Inclement Weather
Mediterranean, land in the nesting area of this falcon, Migration generally occurs during the spring and fall,
where they become easy prey for the predatory birds in times of notoriously unstable weather, which can
the midst of feeding newly hatched young (Walter drastically raise the cost of migration. Severe rain-
1979). Risk of predation is present not only at the end storms and snowstorms kill millions of migrating

80

70 Autumn
Spring

60
Percentage (%) of ospreys

50

40

30

20

10

0
No deviation Deviation <12 Hours 1 Night 2–4 Nights >5 Nights
without foraging Deviation with foraging for varying lengths of time

FIGURE 11.10 Osprey often exhibit fly-and-forage migration, a strategy that combines foraging with
covering migration distance. Ospreys employing this strategy deviate from their migratory route to
forage for fish in nearby bodies of water, but typically stay for relatively short periods of time
(P12 hours). In contrast, many other migratory birds alternate periods of flight with relatively
lengthy stopovers. (Modified from Strandberg and Alerstam 2007.)
248 Chapter 11 / The Ecology and Evolution of Spatial Distribution

monarch butterflies. It is not uncommon to see thou-

sands of dead or dying monarchs on the shores of Lake
Ontario or Lake Erie following a severe storm
(Urquhart 1987). In birds, the most devastating mor-
tality occurs when land species encounter storms over
water, where they cannot take shelter (reviewed in
Newton 2007). Many of these birds are probably lost
without a trace, some wash up on shore, and others die
upon reaching shore. A flock of Lapland longspurs
migrating through Minnesota encountered a sudden
snowstorm one night. The next morning, 750,000 of
these small birds were found dead on the ice of two
lakes, each lake about 1 square mile in size; carcasses
were reported from a much larger area, however, lead-
ing to a mortality estimate of 1.5 million individuals
FIGURE 11.11 Wind turbines that produce electricity,
from this one storm (Roberts 1907a,b). Mortality in such as this one at the Shoals Marine Laboratory on
migratory birds also results from unseasonably cold Appledore Island, Maine, generate little or no pollution
temperatures soon after arriving at breeding areas or and do not expel greenhouse gases. They can, however,
before departing from such areas in late summer or pose problems for bats and nocturnally active birds,
autumn (reviewed in Newton 2007). especially when operated as part of large scale wind
farms.

Obstacles
tory each year and compete vigorously to become
The cost of migration is high in many cases because
reestablished the following year. So what rewards could
large areas of inhospitable terrain must be crossed
possibly override such disadvantages?
(e.g., deserts in some cases and water for land birds).
And then there is the matter of obstacles. Birds often
crash into tall structures such as lighthouses, sky- BENEFITS OF MIGRATION
scrapers, and TV towers. In a single night, seven tow-
ers in Illinois felled 3200 birds (Fisher 1979). In Energy Profit
addition, in the push to develop alternative and We can intuitively understand the advantages of moving
renewable sources of energy, wind-powered turbines from approaching arctic winters to the sunny tropics.
that generate electricity have been erected in many Even a cursory familiarity with the elements of nature
locations, sometimes on a small scale (e.g., a single tur- could also convince us of the advantages of simply mov-
bine) (Figure 11.11) but also on the much larger scale ing from a mountaintop to a valley every winter. In each
of wind farms. Although wind turbines generate little case, the animals are trading a less hospitable habitat for
or no pollution and do not expel greenhouse gases, a more hospitable one.
they can pose problems for nocturnally active birds and The severe weather during the northern winters
bats (Kunz et al. 2007; Kuvlesky et al. 2007). These has favored migration. Each fall, millions of monarch
problems can be categorized as direct or indirect. butterflies migrate southward from the central and
Direct problems refer to fatalities resulting from eastern regions of Canada and the United States to fir
night-flying birds and bats colliding with wind turbine forests in central Mexico—sites with particular char-
rotors and monopoles. Indirect problems result from acteristics that enable the survival of the butterflies
the alteration of the landscape associated with the (Figure 11.12a). These forests are found about 3000 m
development of wind farms (e.g., the construction of (nearly 2 mi) above sea level on the southwest slopes
roads, buildings, and electrical transmission lines). of a very small area of mountaintops. An important
Monitoring bird and bat fatalities at wind facilities can characteristic of these forests is that their temperature
be challenging owing to issues of searcher efficiency is cool but not freezing. Exposure to freezing tem-
(most searches are done by humans, but in some cases peratures would kill the monarchs. However, warmer
trained dogs are used) and removal of carcasses by temperatures would unnecessarily elevate their meta-
scavengers before they can be counted. Nevertheless, bolic rates and waste energy reserves (Calvert and
the data available suggest that migrating passerines Brower 1986).
make up a significant portion of the fatalities, as do Another characteristic of these overwintering sites
tree-dwelling migratory species of bats. that enhances the survival of monarchs is the tall trees,
In addition to such dangers, territorial animals, such primarily oyamel firs. Besides providing branches on
as birds, must relinquish the rights to a hard-won terri- which the butterflies can roost in large numbers, the
 Migration 249

FIGURE 11.12 The migration of monarch butterflies is truly phenomenal. (a) The monarchs travel in large groups
from northeastern North America to overwinter in Mexico or from central California to the coast. (b) Monarch
butterflies roost on trees in their overwintering sites in the mountains west of Mexico City.

trees form a thick, protective canopy over the butter- months (Urquhart 1987). Certain species of insectivo-
flies (Figure 11.12b). The canopy serves as an rous bats may migrate in response to the size of the
umbrella, shielding the butterflies from rain, snow, or insect supply. Mexican free-tailed bats, for instance, leave
hail. This increases survival because a dry monarch can the southwestern United States as the harsh winter cli-
withstand colder temperatures better than can one mate causes the insect supply to dwindle. They migrate
with water on its surface (Figure 11.13). If monarch to regions of Mexico where insects are available through-
butterflies are wet, 50% of the population will freeze out the winter (Fenton 1983).
at –4.2ºC. If they are dry, however, the temperature As winter approaches, increasing the animals’
can dip to –7.7ºC before 50% of the population energy needs, the food supply drops and forces any res-
freezes. The canopy also serves as a blanket that keeps ident species into more severe competition for such
the butterflies warm. Openings in the forest canopy commodities. So, in spite of the energy required for
increase radiational cooling, which can lower body migration, it may result in an overall energetic savings.
temperatures to as much as 4ºC below the ambient air For example, a study of the dickcissel revealed that
temperature. The body temperature of monarchs despite the energy costs of migration, this bird enjoys
under a dense canopy is approximately the same as the an energetic advantage from both its southward autumn
air temperature. However, body temperature drops in migration and its northward spring migration. Studies
proportion to the degree of exposure, increasing the of the junco, white-throated sparrow, and American tree
chances that the butterflies will freeze to death sparrow show that by avoiding the temperature stresses
(Anderson and Brower 1996). of northern winters, these species compensate for at
Seasonal changes in climate also affect food supply. least some of the energy spent on migration (reviewed
In some species, migration is an adaptation that permits in Dingle 1980).
the exploitation of temporary or moving resources. The The question arises, then, if there is so much food
larvae of monarch butterflies feed only on milkweed. In in the warmer winter habitats, why do species migrate
regions of the eastern United States, however, the milk- from such areas? Why do they return to their summer
weed plants grow only during the spring and summer homes at all?
250 Chapter 11 / The Ecology and Evolution of Spatial Distribution

Forest thinned: exposed and wet

butterflies lose freeze protection

0 Low

–2

Ambient
Freeze
temperature
–4 protection
at which
50% of the
butterflies die
(°C)

–6

High
50
–8
100 40
80 30
60
20 Wetness
40
Exposure to 20 10 (mg H2O per butterfly)
open sky (%)
0 0

Forest intact: butterflies protected

from freezing

FIGURE 11.13 The intact forest canopy of the monarch butterfly’s overwintering site normally serves
as both an umbrella and a blanket, which keeps the monarchs from freezing. The vertical axis indi-
cates the ambient air temperature at which 50% of a butterfly sample will freeze to death. A wet
butterfly freezes at a warmer ambient air temperature than a dry one. Because of radiational cooling,
exposure to open sky also allows a butterfly to freeze at a higher ambient temperature than if it were
protected by the forest canopy. When the forest is thinned because local farmers need the wood,
monarchs are vulnerable to freezing, especially on clear nights after wet winter storms. Thus,
thinning of the forests in the monarch’s overwintering area in Mexico is endangering the monarch
population. (Data from Anderson and Brower 1996.)

Reproductive Benefits Some species migrate to areas that provide the

One answer might be that there are important advan- necessary conditions for breeding or that offer some
tages in rearing broods in the summer habitats. For protection from predators. Gray and humpback
example, days in the far north are long, and the birds’ whales, for example, breed in coastal bays and lagoons
working day can be extended—they can bring more that provide the warmer temperatures needed for calv-
food to their offspring in a given period of time and ing and help protect the calves from predation. The
perhaps rear the brood faster. Another result of long need for protected rookery sites may prompt seal, sea
days is that more food is available for offspring and lion, and walrus migrations, as they come ashore on
more young can be raised (Figure 11.14). Although fac- their traditional beaches after months at sea. Sea tur-
tors other than food availability may also play a role, tles also regularly migrate thousands of kilometers
generally the farther north from the tropics a species between feeding grounds and breeding areas, usually
breeds, the larger is its brood (Welty 1962). the same beaches on which they hatched. For exam-
 Migration 251

Reduction in Competition
Another advantage in returning to the temperate zone
is that of escaping the high level of competition that
exists in a warmer, more densely populated area. The
annual flush of life in the temperate zones provides a
predictable supply of food that can be exploited read-
ily by certain species without competition from the
large number of nonmigrants that inhabit the tropics
(Lack 1968).

Reduction in Predation and Parasitism

A third advantage in returning to temperate zones to
breed lies in escaping predation. If predators are unable
to follow herds of migratory ungulates, for instance,
each individual’s chance of survival is enhanced. Thus,
escape from predation has been suggested as the reason
that the number of migratory ungulates is so much
greater than the number of nonmigratory ungulates
FIGURE 11.14 The arctic tern, a champion of migration,
(Fryxell et al. 1988).
moves between its breeding grounds in the Arctic and
its winter home in the Antarctic. By breeding in north- In the far north, breeding periods are very short
ern regions, it can take advantage of long days to gather because of the weather cycles. This can be an advantage
more food for its young. to nesting birds, which are in danger from predators.
The short season results in a great number of birds nest-
ing simultaneously, thereby reducing the likelihood of
ple, female green turtles (Figure 11.15) feed in the any single individual being taken by a predator. Also,
warm marine pastures off the coast of Brazil and then since there is no extended period of food availability for
swim roughly 1800 km to the sandy shores of predators, their numbers are kept low. By leaving certain
Ascension Island, where they lay about 100 eggs, each geographical areas each year, migratory species deprive
the size of a golf ball (Lohmann 1992). The beaches many parasites and microorganisms of permanent hosts
where sea turtles lay their eggs are on isolated to which they can closely adapt. Long, harsh winters in
stretches of continental shores or small remote islands the frozen north further reduce the number of these
(Lohmann and Lohmann 1996b). Because of their iso- threats.
lation, these beaches might have fewer predators.
MIGRATION AND
CONSERVATION BIOLOGY
Long-distance migrations are one of the most spectac-
ular of biological phenomena; sadly, they are becoming
increasingly rare events. For example, in the Greater
Yellowstone Ecosystem, a 19-million-acre temperate
ecosystem that spans parts of Idaho, Wyoming, and
Montana, many of the historic and current migration
routes of ungulates have already been lost. More specif-
ically, all 14 migration routes used by bison have been
lost, as well as 78% of the routes used by pronghorn and
58% of those used by elk. The causes for these losses
include increases in the human population and associated
losses of habitat; in many cases, fences, highways, and
housing subdivisions block migration routes, and petro-
leum development threatens wintering areas. A possible
landscape-level solution to this problem lies in creating
FIGURE 11.15 Some green turtles migrate from their a protected network of national wildlife migration cor-
feeding ground off the coast of Brazil to Ascension ridors; in the Yellowstone region, this could protect
Island to breed on sheltered beaches, where it is safe long-distance migrations by all three ungulate species
from predators. (Berger 2004).
252 Chapter 11 / The Ecology and Evolution of Spatial Distribution

than females. At least three hypotheses have been sug-

SUMMARY gested to explain patterns of dispersal in birds and
mammals: (1) inbreeding avoidance/local resource com-
Natal dispersal involves permanent movement away petition; (2) local mate competition; and (3) cooperative
from the natal area or social group to where breeding behavior among kin.
first occurs. In natal philopatry, offspring remain at their Animals that disperse from their natal site or
birthplace. Dispersal can also occur after reproduction; breeding site must eventually select a new location in
the movement between two successive breeding areas or which to settle. This process of habitat selection can be
social groups is called breeding dispersal. divided into search, settlement, and residency. Animals
Costs and benefits are associated with philopatry and searching for a new place to settle employ comparison
natal dispersal. Potential costs of philopatry include (best of N) strategies and sequential search strategies,
those of inbreeding, reproductive suppression, and and likely evaluate resources, and presence of con-
competition with relatives for mates or resources. On the specifics and heterospecifics when assessing the qual-
other hand, a certain level of inbreeding may actually be ity of a site. There is evidence that natal experience can
beneficial due to the maintenance of gene complexes that influence habitat selection.
are particularly well suited to local conditions. Another Migration occurs over greater distances than disper-
advantage of staying home is familiarity with the local sal, and also has costs and benefits. Among the costs are
physical and social setting—animals can find food and increased energy expenditure, predation, and exposure to
escape predators more easily in a familiar area. severe weather. A major advantage concerns attaining a
Dispersers, in contrast, are thought to face high energy favorable net energy balance. In spite of the energetic cost
costs and risks of predation as a result of increased move- of migration, an animal may experience a net energy
ment and lack of familiarity with the physical and social profit. For example, North American songbirds that
environment. migrate south each fall gain by escaping the metabolically
The direction of sex differences in natal dispersal draining harsh winter temperatures and by avoiding the
differs between birds and mammals. In most birds, increase in competition that accompanies a reduction in
females are more likely to disperse than males. In most winter food supplies. Migratory individuals may also gain
mammals, however, males are more likely to disperse by a reduction in predation or parasitism.
 12
Foraging Behavior

Obtaining Food back to the tree, he sits next to the table and eats it.
Suspension Feeding He continues searching, and finds a half a cookie.
Omnivory Lifting his head high to hold it somewhat awkwardly
Herbivory off the ground, he runs to the nearest tree and climbs
Carnivory
up. Not a moment too soon—a jogger, with her dog
ranging in front of her, appears at the edge of the lawn
Adaptations for Detecting Prey
and soon heads past the picnic table.
Optimal Foraging If you attend college on a North American campus,
Diet Selection: A Simple Model you’ve probably seen something similar to what we just
Deciding When to Leave a Patch: The Marginal Value described. What you may not realize is that you’ve wit-
Theorem nessed a series of decisions about foraging—finding,
Adding Complexity and Realism processing, and eating food—that might surprise you in
The Utility of Models their complexity. Squirrels base their foraging decisions
on a food item’s size, its energetic value, how easy it is
to carry, and its distance from cover—and how these
On a college campus on a warm spring day, a gray squir-
variables interact (Lima et al. 1985, Lima and Valone
rel peers down from a tree and makes a chittering sound.
1986). In this chapter, we will explore how animals make
A group of students sitting under the tree finish takeout
decisions such as these.
coffees and snacks and get to their feet, brushing crumbs
from their clothes. As they walk away, the squirrel dashes
down the tree and seizes a grape-sized piece of muffin
from the ground. Without pausing, it turns and runs OBTAINING FOOD
back up the tree, where it sits on a branch and finishes
off the muffin in less than a minute. Animals must consume either plants or other animals
Across campus, another squirrel is sniffing the in order to live. Animals can neither capture energy
ground near a picnic table on a grassy lawn. There is directly from the sun, as plants do, nor get carbon or
a line of maple trees about 10 m away. This squirrel nitrogen (components of biologically important mole-
also finds a small piece of food, but instead of racing cules) directly from the environment. Instead, animals

253
254 Chapter 12 / Foraging Behavior

must acquire these essentials by feeding on other a

organisms. Although we are all intimately familiar with
foraging, animals have a surprising range of adaptations
for acquiring food. We’ll explore the diversity of for-
aging techniques in this section.

b
SUSPENSION FEEDING
Many aquatic species feed by removing small suspended
food particles from the surrounding water. Species vary
in the techniques they use. A few species sieve water and
strain out the food. Others trap particles on sticky sur-
faces of mucus. For example, the annelid worm
Chaetopterus sits in a U-shaped burrow, where it creates FIGURE 12.2 The filtering apparatus of dabbling ducks. (a)
a mucous net. As it pumps water through its burrow, Duck beak, with area of closeup in (b). From Gurd (2007).
food particles are trapped in the mucus, which the worm
periodically rolls into a ball and swallows (Figure 12.1). familiar mallards. Dabbling ducks strain food from
You might think that suspension feeding would only water through filters that fringe the edge of their bills.
be a feasible way to make a living for small animals, but They can adjust this filtering apparatus by changing the
the largest animal species on earth feeds this way. Blue position of their bill while feeding (Figure 12.2; Gurd
whales (Balaenoptera musculus), a species of baleen whale, 2007). When they feed with their bill in a more closed
can often weigh over 100 tons, yet survive wholly on tiny position, they can filter out small particles. When they
shrimp-like creatures called krill. Instead of teeth, feed with their bill in a more open position, only larger
baleen whales have rows of fringed plates made of ker- particles are retained. However, because closing the bill
atin (similar to human fingernails). These whales take in during feeding helps force more water out of the bill,
enormous mouthfuls of water, then expel the water foraging in a more open position reduces the rate at
through the plates, leaving behind the krill to be swal- which water, and therefore food, is pumped through the
lowed. It has been estimated that blue whales can filter bill. Reducing the foraging rate may be advantageous to
out up to four tons of krill per day. ducks foraging in muddy wetlands. Gurd (2007) found
Suspension feeders are often sessile (stationary) and that when ducks were given water that had small,
must take the food that comes their way, but others (like unwanted debris particles in it, as well as larger food
the whales) can move about in order to select good for- particles, they only consumed the food, but when they
aging areas. Some species can even adjust their filtration were given water with only food in it, they filtered the
pattern to choose particular types of particles at differ- food faster. These results show that the ducks found it
ent times. Among these selective species are dabbling better to accept a lower foraging rate, but avoid the
ducks, a group that includes northern shovelers and the detritus, than to consume both food and detritus at a
higher rate.
Water flow
OMNIVORY
We will spend most of our time in this chapter treating
herbivory (plant eating) and carnivory (meat eating) sep-
arately, for this simplifies our discussion a bit. However,
keep in mind that many animals exhibit omnivory and
eat both plants and animals. Animals may be omnivorous
for a variety of reasons, including the limited availability
Mucous net of a preferred food, a need for nutritional variety, and to
minimize exposure to risks associated with a particular
food type (e.g., predators or toxins that may have cumu-
lative effects) (reviewed in Singer and Bernays 2003).

HERBIVORY
FIGURE 12.1 Chaetopterus, an annelid worm that lives in Plants have a variety of parts—roots, leaves, stems,
U-shaped tunnels in the sand, filter feeding with a fruits, and flowers—each of which may be consumed
mucous net. (From Brusca and Brusca 1990.) by different species of animals in countless interest-
 Obtaining Food 255

ing ways. To understand the diversity of behaviors repeatedly into the midrib of the milkweed leaf. The
associated with herbivory, we need to understand the sap leaks out through the bite holes (Figure 12.3)
plant’s perspective: a plant may well suffer a loss of fit- before it reaches the tips of the leaves, where the bee-
ness if an animal eats its roots or leaves, but it bene- tle can then feed without gumming up its mouthparts
fits if its pollen is transferred to another plant by a (Dussourd 1999). This isn’t good for the plant because
pollinator, or if the seeds in its fruit are carried off to midrib cutting and feeding reduce the plant’s ability to
a new germination site. Thus, the challenges facing an photosynthesize (Delaney and Higley 2006). Perhaps
herbivore vary depending on whether the plant is beetles are, for the moment, winning this particular
marshalling a defense that must be overcome or arms race. (See Chapter 13 for a look at how milkweed
encouraging an animal’s attentions. Let’s look at some and monarch butterflies interact.)
examples. While roots, stems, and leaves are often defended,
Spiny cacti, poisonous hemlock, thorny rose- fruits and flowers have evolved to attract animals. Most
bushes, and prickly thistles are all familiar examples of ripe fruit is red (think of apples and raspberries) or black
defended plants. A more unusual and particularly (think blackberries). Is this because these colors have
unpleasant defense is used by plants in the milkweed evolved to attract frugivores (fruit eaters) that will dis-
family, which get their common name because their perse the seeds? Birds, one of the main frugivores, have
leaves and stems, when cut, ooze a white sap. However, four types of retinal cones that give them excellent color
only in appearance is the sap milky—it is sticky and vision. Foraging birds do prefer red and black fruits—but
thick, and, in many species, it has chemical irritants that not because they are attracted to the color itself. Instead,
make it even more noxious to touch or taste. red and black fruits contrast well with the typical green
Nevertheless, many insects are able to feed on milk- background of foliage, and it is this contrast that attracts
weed. Milkweed beetles (Tetraopes tetrophthalmus) over- birds (Schmidt et al. 2004).
come this defense in an innovative way: by biting Flowers have also evolved visual cues attractive to
pollinators. As you probably know, many plants rely on
animals for sexual reproduction. A bee, hummingbird,
bat, or other pollinator that visits a flower is dusted with
pollen. As it moves from flower to flower, it transfers the
pollen and flowers are fertilized. Many plants have
evolved means of enticing pollinators to visit, such as by
offering nectar rewards. Some flowers have markings
called nectar guides that act like runway lights, direct-
ing pollinators to the part of the flower where the nec-
tar is stored. Even flowers that appear white to us may
have nectar guides: bees and many other insects can see
in the ultraviolet, and many flowers have markings that
are visible only under UV. Other flowers offer special-
ized rewards. Flowers that are pollinated by carrion flies
(which feed on dead flesh) are not recommended for
household decoration—their putrid odor has evolved to
attract their pollinators, not humans (e.g., Burgess et al.
2004). An especially clear example of a coadaptation
between plants and pollinators comes from a species of
hummingbird (Elampis jugularis) that feeds on two
species of Heliconia. Female birds have a long, curved bill
that matches the long, curved flowers of their preferred
flower species, H. bihai. The short, straight bill of the
male matches their preferred flower, H. caribaea (Temeles
et al. 2000).
A different approach to herbivory is taken by species
that cultivate some or all of the food they need.
Agriculture has evolved independently in three insect
orders: ants, termites, and ambrosia beetles. Leaf cutter
FIGURE 12.3 Vein-cutting behavior of milkweed beetles ants cut fresh leaves and carry the pieces back to the nest
(Tetraopes tetrophthalmus). After a beetle cuts the midrib under the ground (Figure 12.4). There they encourage
of the leave, the sticky noxious sap oozes out. The the growth of a special fungus on the leaves. This fungus,
beetle can then feed at the tips of the leaves. whose existence is unknown outside the ant nests, is the
256 Chapter 12 / Foraging Behavior

FIGURE 12.4 Leaf cutter ants are transporting leaves to fertilize their fungus gardens. These ants
cultivate a special fungus that serves as their primary source for food.

primary food source for these ants (Weber 1972). The CARNIVORY
ants prepare their fungus gardens by licking the leaves
Whereas plants sometimes benefit from herbivores, car-
on both sides to remove the waxy layer covering the leaf
nivores must capture food that never benefits from being
and reduce the population of microorganisms that might
eaten. Here, we often see evidence of an arms race: prey
compete with the desired fungus. The leaf fragments are
species have evolved myriad defenses against predation,
then chewed to a pulp, placed in the fungus garden, and
and predators have evolved to overcome those defenses.
inoculated with hyphae of fungus. This preparation
In this section, we will examine the strategies a suc-
makes the leaves a richer source of nourishment for the
cessful predator might use, and we will devote Chapter
fungus. The ants are attentive farmers. When they tidy
13 to the perspective of the prey. To quote Jerry Seinfeld,
their gardens, they collect debris and compress it into a
when we watch a nature show about antelope, we root
pocket inside their mouth. This pocket functions as a
“Run, antelope, run! Use your speed, get away!”—but
sterilization area that kills the spores of a garden para-
the very next week, watching a show about lions, we
site. Sterilization is caused by bacteria that live inside the
cheer “Get the antelope, eat him, bite his head! Trap
pocket and produce antibiotics (Little et al. 2006).
him, don’t let him use his speed!” In this section, pre-
Fungus gardens in abandoned nests degenerate quickly
pare to cheer for the predators.
as they become overrun with other microbes (Quinlan
and Cherrett 1977).
Other species also modify their own food supplies. Pursuit
Limpets, which are related to snails, leave behind The classic nature-show scene like the one we have just
mucous trails as they crawl along on underwater rocks described is what generally comes to mind when we con-
and feed on microalgae. In several species, the mucus not sider predation. It’s no wonder: no one can forget the
only traps microalgae but stimulates its growth. Limpets sight of a cheetah in full pursuit of a gazelle. Adapted for
then revisit their old slime trails to harvest the crop speed, with long, slender legs and a flexible spine that
(Connor 1986). Gorillas are herbivorous and will rip allows for an even longer stride, cheetahs can reach a top
down large plants, resulting in a surge in the growth of speed of 70 mph (113 km/h) (Caro 1994).
young, fast-growing vegetation, which is encouraged by Other species that chase their prey include several
the new light (Watts 1987). seabirds. Northern gannets (Sula bassana) feed by spec-
tacular plunge dives: they plummet down into the ocean
to grab unsuspecting fish. By attaching data-logging
units to the gannets that measured depth, researchers
STOP AND THINK were able to characterize the dives in detail. Gannets
perform both long, deep (more than 8 m) U-shaped
Which is likely to be stronger, selection on the predator
to be better at capturing prey, or selection on the prey to dives, during which they use their wings to propel them-
be better at escaping the predator? selves, as well as short, shallower V-shaped dives (Garthe
et al. 2000).
 Obtaining Food 257

a Tracking A common strategy of animals that engage in pur-

suit is to hunt in groups. We’ll investigate this behavior
in more detail in Chapter 19.

Stealth
Full-out pursuit is exhausting and dangerous. Even
predators built for speed like cheetahs do not always
engage in a long chase, but will approach their prey
slowly until quite close, using a hunting method called
stalk-and-rush. They also flush prey concealed in vege-
b Interception
tation and pursue them only over short distances
(reviewed in Caro 1994). To facilitate surprise attacks,
predators often choose hunting sites that provide good
cover. Hopcraft et al. (2005) tested two alternative
hypotheses: do lions spend more time hunting in habi-
tat where there is good cover or in habitat with a higher
density of prey? In the field, lions followed large groups
of prey (e.g., migrating herds), but on a smaller scale,
they more frequently chose to hunt where there was
FIGURE 12.5 Different methods of pursuing a target.
(a) Tracking behavior, where the predator steers to
cover, even when fewer prey were present.
keep the moving prey, represented by dots, in front. In some cases the predator’s ability to draw near is
(b) Interception, where the predator aims in front of enhanced by camouflage, rendering the predator more
the target. From Olberg et al. (2000). difficult to detect. Often this takes the form of markings,
colors, or even behaviors that make the predator blend
Many insects also pursue prey. You might imagine that into the background. Jumping spiders of the genus Portia
it could be very tricky for one insect to capture another fly- use environmental disturbances as a smokescreen to
ing at high speed. Insects use several alternative mecha- camouflage their approach to the prey. Portia preys on
nisms to accomplish this feat (Collett and Land 1978; other spiders, and often does so by climbing right into
reviewed in Olberg et al. 2000). First, predators can steer their webs. The movement of Portia shakes the web
so as to keep the moving image directly in front of them. slightly, and this can alert the prey spider to the
This results in tracking behavior (Figure 12.5a), and if the approaching danger. In the field, researchers noticed that
predator is faster than the prey, the prey will lose the race. Portia seemed to move more when the wind blew the
A number of insects (e.g., houseflies and beetles) use track- web, and to pause when the air was still. In the lab,
ing to pursue objects, be they mates or food. Other insects Wilcox et al. (1996) used a fan to create a breeze to blow
use the strategy of interception. Instead of aiming at the tar- on the web. By turning it on and off, they demonstrated
get’s current position, the pursuer intercepts the prey’s flight that Portia does indeed use environmental disturbance to
path by aiming at a place in front of it (Figure 12.5b). This mask its approach, a behavior called smokescreening
is the strategy used by dragonflies. Dragonflies have excel- (Wilcox et al. 1996).
lent vision and typically perch on vegetation until a prey
happens by; then they pursue it and intercept it. Dragonflies
are extremely effective predators. In one study, they cap- Aggressive Mimicry
tured 97% of the prey they pursued (Olberg et al. 2000). In aggressive mimicry, a predator gets close to its
How do dragonflies, or, for that matter, other animals, prey because it mimics a signal that is not avoided by
calculate the necessary course to intercept a prey? It seems the prey and may even be attractive to it. Some preda-
like a tricky problem from physics class, but it can be tors have specialized structures that are used as lures.
solved by a simple rule. A dragonfly will intercept its prey The alligator snapping turtle lies on the muddy bot-
if the dragonfly moves so that the position of the target tom of streams and lakes and holds its mouth wide
on the retina remains at a single point (Olberg et al. 2000). open. Its tongue has a wormlike outgrowth that, when
Both fielders catching baseballs (McBeath et al. 1995) and wiggled enticingly, attracts fish that the turtle then
dogs catching frisbees (Shaffer et al. 2004) use a similar snaps up. The deep-sea angler fish lives at extreme
strategy. Fielders and dogs, unlike dragonflies, are con- ocean depths, where virtually total darkness prevails.
fined to a two-dimensional field, so they can’t restrict the Nonetheless, the female has a long, fleshy appendage
targets on the retina to a single point. Instead, the fielder attached to the top of the head that is luminous and
or dog runs so that the target appears to rise in a straight acts as a lure, bringing curious fish within reach of its
line. This strategy results in a curved path to the target. toothy jaws (Wickler 1972).
258 Chapter 12 / Foraging Behavior

a b

FIGURE 12.6 (a) A cleaner wrasse is removing parasites from another fish. The service of cleaner fish is beneficial
because it removes external parasites, diseased tissue, fungi, and bacteria. The cleaner wrasse has distinctive markings
and behavior patterns that advertise its services. (b) Aspidontus is disguised as a cleaner wrasse and mimics its behavior.
As a result, fish in need of the cleaning service allow the phony cleaner to approach. Aspidontus then takes a bite out of
the would-be customer.

Some predatory species draw within striking dis- get more than just calories from this deception.
tance of their unsuspecting prey by mimicking benefi- Photinus fireflies of both sexes produce defensive com-
cial species. For example, some blennid fish look and act pounds. Photuris females can only get these compounds
like beneficial cleaner fish to lure their prey. The clean- by preying on Photinus (Eisner et al. 1978, 1997). The
ing wrasse (Labroides dimidiatus) removes parasites, dis- severity of this selection pressure is demonstrated by
eased tissue, fungi, and bacteria from other fish (Figure Lucidota atra, another firefly that also has defensive
12.6a), even swimming into their mouths. Indeed, par- chemicals and that is also eagerly consumed by
asite-laden fish line up at coral reef “cleaning stations” Photuris. In contrast to Photinus, L. atra is diurnal, with
much like cars at a car wash to avail themselves of the only vestigial bioluminescent organs, suggesting that
services of the cleaning wrasse. Some cleaning stations it may have evolved to be day-active in order to escape
may have hundreds of patrons each day. The cleaning predation by Photuris (Gronquist et al. 2006).
wrasses generally advertise their services with distinctive We will discuss more examples of deception in pre-
swimming motions performed above the cleaning sta- dation and in other contexts in Chapters 13 and 19.
tion. Customers show their willingness to be cleaned by
postures that are characteristic of the species. A phony
cleaner, a blennid fish (Aspidontus taeniatus), looks and Traps
behaves like the cleaning wrasse (Figure 12.6b), but when Some predators are able to trap their prey: they manip-
it is invited to approach, it takes a bite out of gills or ulate objects or alter their environment in such a way as
other soft parts of the customer (Wickler 1972). to capture, or at least restrain, prey. For example, hump-
Many predators attract their prey by sending sig- back whales (Megaptera novaeangliae) build bubble “nets”
nals that mimic the mate of the prey species. One of to trap, or more accurately to corral, their prey.
the most famous examples is found in fireflies. Not Beginning about 15 m deep, the whale blows bubbles by
flies at all, these beetles have bioluminescent organs on forcing bursts of air from its blowhole while swimming
the ventral side of their abdomens. In northeastern in an upward spiral. The bubbles form a cylindrical net
North America, a number of firefly species live in the that concentrates krill and small fish (Figure 12.7). The
same habitat. At night, males fly about and flash in whale then swims upward through the center of the bub-
species-specific patterns. A receptive female, waiting in ble net with an open mouth and devours the prey in a
the vegetation below, flashes back to a conspecific single gulp (Earle 1979). Researchers created artificial
male. The male flies down to land near her, whereupon bubble nets so they could observe the response of prey
they signal back and forth in a “lengthy courtship dia- more closely. Herring were very reluctant to swim
logue” (Woods et al. 2007). However, this does not through bubble nets, although they were more willing
always work out as a male might wish: sometimes a to do so in order to reach a larger group of conspecifics
Photuris female signals back to a Photinus male. When on the other side (Sharpe and Dill 1997).
the Photinus male lands to find his mate, the Photuris Some of the most familiar trap builders are spiders.
female eats him (Lloyd 1975, 1986). Photuris females Spider webs come in a variety of forms. The familiar
 Obtaining Food 259

FIGURE 12.7 The humpback whale traps

krill in cylindrical bubble “nets” that it
constructs by forcing air out of its blow-
hole. In this photo, the bubble net is
seen as a ring of bubbles on the surface
of the water. Using a bubble net, a whale
can corral the fast-swimming krill and
consume them in a single gulp.

“Charlotte’s web”-style orb web incorporates as many as of glue, and reels it in (Figure 12.8) (reviewed in Yeargan
seven different kinds of silk to make the radii, the cen- 1994). Some species of bolas spiders lure moths to them
ter hubs, and the sticky spiral (Foelix 1996). Sheet by emitting female sex pheromones (Stowe et al. 1987).
weavers make flat sheets that can be quite plain—just a How spider webs are designed for catching prey has
flat or slightly curved sheet—or much more elaborate. been the subject of a great deal of research—it’s gener-
For example, you can probably make a good guess as to ally easy to measure webs and to quantify prey capture
what a bowl-and-doily spider web (Frontinella communis) success in the field, so it is often possible to quantify the
looks like: a flat sheet, the “doily,” with a bowl-shaped relationship between web architecture and fitness. For
structure sitting over the top of it, all surrounded by example, Venner and Casas (2005) observed the orb webs
support strands. The spider hangs upside down under of Zygiella x-notata and found that most of the prey the
the web, grabbing insects through the silk that have hit spiders caught were quite small (less than 2 mm). Large
the support strands and tumbled into the bowl above the prey (more than 10 mm long) were quite rare and are
spider. Other spiders have very reduced webs. Deinopsis, captured only about every 20 days. However, these large
for example, stretches a small net of silk between the tips prey are crucial to fitness: spiders cannot survive and
of its feet and waits patiently with its arms outspread. reproduce eggs without catching at least some large prey.
When an insect happens past, it claps its legs together Venner and Casas suggest that webs are designed to take
and traps its victim. The prize for the most minimalist advantage of these rare events.
web must go to the bolas spider, which uses only a silk
line with a drop of glue at the end. It swings the line
around with its leg, smacks a passing insect with the drop ADAPTATIONS FOR DETECTING PREY
Sensory Specializations
One of the recurring themes in this book is that we must
continually be aware of the sensory systems of our study
organisms. As we look around the animal kingdom, we
find many sensory specializations for prey detection that
are poorly developed or lacking in humans, and thus not
immediately obvious to researchers. Here, we’ll make
that point by looking at a few examples.
The snakes in two large families, the Crotalidae,
or pit vipers (e.g., rattlesnake, water moccasin, and
copperhead), and the Boidae (e.g., boa constrictor,
python, and anaconda), use their prey’s body heat to
help guide their hunt. They have special receptors that
are so sensitive to infrared radiation (heat) that these
snakes can locate their warm-blooded prey even in the
FIGURE 12.8 A bolas spider, which catches insects with darkness of night (Figure 12.9). It’s been estimated
a drop of glue swung on the end of a silk line. that the sensory endings must be able to respond to
260 Chapter 12 / Foraging Behavior

FIGURE 12.9 (a) The heat sensors of the black-tailed rattlesnake, located in pit organs between each eye and nostril
(the dark circular structures in this photo), allow it to detect its warm-blooded prey in complete darkness. Snakes of
the families Crotalidae and Boidae have infrared heat sensors that help them find their prey. (b) A rattlesnake’s
“picture” of a mouse facing toward it.

contrasts in temperature of 0.001ºC (Bakken and The star-nosed mole (Condylura cristata) finds prey
Krochmal 2007). In spite of the ability of snakes to by using its sense of touch. Critical to that sense is its
strike precisely at their prey (Figure 12.10), the image unusual star-shaped nose, consisting of 22 fleshy, mobile
on the pit organ is apparently quite poorly focused, appendages. Aside from its snout, this mole looks like
which suggests that snakes must select environments most moles. It is about 6 inches (15 cm) long and has
to hunt that provide good thermal contrast (Bakken powerful digging claws and poorly developed eyes.
and Krochmal 2007). Rarely coming to the surface, the star-nosed mole lives
In the quiet of the Mojave Desert, running insects in the extensive tunnel system it digs in the wetlands. It
transmit vibrations through the sand. Predatory sand searches these tunnels for earthworms and other small
scorpions (Paruoctonus mesaensis) can detect these distur- prey by touching the walls with its nose as it moves for-
bances from up to 30 cm away. Scorpions sense vibra- ward. Each second, the nose makes an average of 13
tions by structures called slit sensilla located on each of touches. The movements are controlled by muscles that
their eight legs. Having so many legs is a convenience: are attached to the skull and then connected to the base
by comparing the timing of the arrival of a vibration at of each appendage. Each appendage is sheathed with
each leg, the scorpion can precisely determine the direc- touch receptors, called Eimer’s organs, that communi-
tion of nearby vibrational sources, and unerringly and cate with the somatosensory area of the cortex of the
swiftly attack (Brownell 1984; Brownell and van mole’s brain. One particular pair of appendages is used
Hemmen 2001). for detailed investigation, analogous to the foveal area of
the human eye that is tightly packed with sensory recep-
tors (Catania and Kaas 1996; Catania and Remple 2004).
The mole’s Eimar organs are so sensitive and fast-mov-
ing that it’s been estimated that a mole can examine 300
m of tunnel floor every day (Catania and Remple 2004).
10° 0° 10°
20° 20° High-speed video recording has revealed that star-nosed
30° 30°
40° 40° moles can also detect odors underwater, a feat previously
50° 50° thought impossible. Moles accomplish this by exhaling
60° 60° air bubbles onto objects and then sniffing the bubbles
back up (Catania 2006).
70° 70°
Procellariiform seabirds, including the petrels and
80° 80° albatrosses, seem to sniff out seafood patches scattered
90° 90° over vast expanses of open ocean (Nevitt 1999a). These
FIGURE 12.10 A rattlesnake’s heat sensors can direct an
are sometimes called tube-nose seabirds because of their
amazingly accurate strike. Each circle indicates the long tubular nostrils, used to excrete excess salt as well
angular error of a strike at a warm soldering iron by a as to smell. What can they be smelling as they soar over
blindfolded rattlesnake. The average error of the strikes the featureless ocean, hundreds of miles from land? Like
was less than 5º. (Redrawn from Newman and Hartline the humpback whales discussed earlier, tube-nose
1982.) seabirds feed on krill. As predators eat krill, the macer-
 Obtaining Food 261

ated krill bodies release volatile compounds that seabirds The first hint of something unusual came in 1678,
can smell. Krill themselves feed on phytoplankton, when the Italian anatomist Stefano Lorenzini described
which releases an odorous gas, dimethyl sulfide (DMS). strange pores speckling the head of sharks near their
DMS can also lead the seabirds to krill. mouths. By peeling back the skin, he found that the
Gabrielle Nevitt and her colleagues conducted pores led to gel-filled tubes inside the sharks’ heads.
experiments out at sea demonstrating that the odors of What could these be? By the late nineteenth century,
macerated krill and of DMS attract seabirds. They cre- using improved microscopy, scientists could trace nerves
ated vegetable-oil slicks, some of which were perfumed leading from these tubes to the brain, suggesting a sen-
with krill extract; plain vegetable-oil slicks served as sory function. Other biologists during the 1900s
controls. Observers who did not know which of the described that the firing rate of these nerves varied with
slicks were krill-scented counted the number of changes in pressure, touch, temperature, and salinity.
seabirds attracted to the slicks and measured how One scientist, R. W. Murray, happened to switch on a
quickly they appeared. Cape petrels, southern giant magnetic field near a shark neuron preparation and
petrels, and black-browed albatrosses showed up noticed that the nerve’s firing rate skyrocketed. As Fields
within minutes. Five times more of these birds (2007) describes: “Astonishingly, Murray determined
appeared at the krill-scented slicks than at the control that the organs could respond to fields as weak as one
slicks. The odor did not attract storm petrels or millionth of a volt applied across a centimeter of seawa-
Antarctic fulmars. Equal numbers of these birds ter. This effect is equivalent to the intensity of the volt-
showed up at control and krill-scented slicks (Nevitt age gradient that would be produced in the sea by
1999b). Similar experiments compared the attractive- connecting up a 1.5-volt AA battery with one pole
ness of DMS-scented slicks with plain vegetable-oil dipped in the Long Island Sound and the other pole in
slicks. A number of bird species, including prions, the waters off Jacksonville, Fla. Theoretically, a shark
white-chinned petrels, and two species of storm swimming between these points could easily tell when
petrels showed up at DMS-scented slicks twice as often the battery was switched on or off.”
as at control slicks. But three species of albatrosses and What could this electromagnetic sense be for?
Cape petrels were just as attracted to the control slicks Neural and muscular activity in animals generates elec-
as to the scented ones (Nevitt et al. 1995). Thus, it tromagnetic signals: their pumping hearts and brain
seems that the odors of krill or DMS can provide an waves, for example, produce short, weak pulses.
immediate and direct way to assess the potential pro- However, these are not the signals that the sharks’ sense
ductivity of an area but that different species of organs are tuned to detect. They detect only slow-
seabirds use these cues in different ways. changing fields, such as those generated by batteries.
Recently, researchers have begun investigating the This is exactly the sort of signal produced by a fish’s body
development of the seabirds’ ability to follow their noses. in seawater: the salt solution in the fish’s body differs
Even young blue petrel chicks, one to six days before from that of seawater.
they fledge, prefer the scent of DMS over a control odor By the 1970s, Adrianus Kalmijn (1966, 1971) was
in a choice test. Thus, chicks do not have to learn the the first to demonstrate that sharks can use this sense to
scent of DMS through their own foraging experience, locate prey by electrical fields. He found that captive
but instead embark on their first foraging trip ready to sharks (Scyliorhinus canicula) would attack electrodes
follow the odor to food (Bonadonna et al. 2006). buried in the sand of an aquarium. In the ocean, Fields
Other remarkable sensory specializations for detect- and his team used a fiberglass (nonmetallic) boat. They
ing prey are found in those most infamous of predators, lowered pairs of electrodes into the water but activated
sharks. Many senses are involved in the shark’s astound- only one. Observers were not told which was active.
ing ability to detect and track down prey. Sharks can hear Then the biologists dumped ground-up fish into the
prey from a distance of over nine football fields and will water and waited for the sharks to come. Time and again
turn and swim toward it. When the shark is within a few sharks attacked the “live” electrode, demonstrating that
hundred meters of its prey, its nose can help direct its in natural conditions, sharks respond to electrical cues.
search. As the shark nears its prey, its lateral line organs The use of bioelectric cues is not restricted to sharks
detect small disturbances in the water caused by the but is also found in rays, skates, lungfish (Figure 12.11),
swimming motion of prey, and so the shark knows the and some larval amphibians. An example with unex-
location of the prey even if it cannot see it. Within close pectedly similar talents is the platypus (Ornithorhynchus
range, the shark can see its prey. However, an even more anatinus), an oddball egg-laying Australian mammal. The
surprising sense is perhaps the most impressive. The platypus has both electroreceptors and mechanorecep-
story of how this bizarre sense came to be understood is tors on its ducklike bill that help in locating prey (Figure
described by Fields (2007) and illustrates how tricky it 12.12) (Griffiths 1988; Scheich et al. 1986). (Electro-
can be to figure out a behavior that is beyond the nor- reception was lost with the move to land and reevolved
mal sensory experience of humans. in monotremes, the group of mammals that includes the
262 Chapter 12 / Foraging Behavior

a 6

Target
2

0
Difference index

–2
–4

Distractor
–6
–8

–10
–12
–14
FIGURE 12.12 The duck-billed platypus is among the
predators that can locate prey by detecting their electric
b 0.4
fields.
0.3
Target
0.2 slightly below it and almost completely to its side (80º
0.1 lateral from the tip of the bill). A platypus can even
detect a battery lying on the bottom of the creek. A hunt-
Difference index

0.0
ing platypus will often dive to the bottom of the body of
–0.1
water and dig in the silt with its paws and bill. As it
–0.2 probes, it stirs up small animals, including crayfish, insect
Distractor

–0.3 larvae, and freshwater shrimp, which dart away. As the

prey swim off, they generate electrical fields sensed by
–0.4
the platypus’s electroreceptors, while the mechanical
–0.5 waves in the water are sensed by mechanoreceptors on
–0.6 the platypus’s bill. Together these sensory organs give the
Live Live Shielded Dead Empty platypus a complete, three-dimensional fix on the posi-
prey prey prey prey
+ + + tion of its underwater prey (Manger and Pettigrew 1995;
displaced displaced displaced Pettigrew et al. 1998).
odor odor odor The lungfish and the platypus use electrical cues
FIGURE 12.11 The results of experiments demonstrating given off by animals to locate prey, but other species
that Australian lungfish can locate their prey by using locate prey by generating their own electrical field and
the electrical field generated by a living organism in sensing disturbances in that field created by the prey.
seawater. The results are shown as the difference The black ghost knifefish (Apteronotus albifrons) is a
between the amount of foraging activity above the tar- nocturnal predator that uses its electrical sense to feed
get, the hidden chamber where the prey was located, on small prey such as the insect larvae and small crus-
and the distractor, the region of the aquarium to which taceans in the freshwater rivers of South America. This
the prey’s odor was displaced. A positive score indicates
fish has two types of electroreceptor organs distributed
that the fish were foraging above the target; a negative
over its entire body. One type, the tuberous electrore-
score indicates that the fish were foraging over the
distractor. The difference indices are shown for (a) the ceptor organ, is specialized for detecting alterations in
foraging intensity and (b) the foraging accuracy of the electrical field that the fish generates itself and uses
each treatment. Although the lungfish can and do use for orientation (electrolocation is discussed further in
chemical cues to locate prey, bioelectric cues are more Chapter 10). In contrast, the ampullary organs are spe-
important. (From Watt et al. 1999.) cialized to detect electrical fields generated by external
sources, such as prey animals. These receptors, in coor-
platypus.) The platypus hunts at night in murky streams. dination with the mechanosensory lateral line system,
It swims slowly, patrolling for food with its eyes, ears, help the fish hunt for prey at night or in muddy water.
and nostrils closed. As it swims, it makes intermittent An agile swimmer, the black ghost kitefish can swim back-
head sweeps, scanning the area for electrical signals. ward as easily as forward and hover in place. It can even
When it detects an electrical stimulus, it swims two to swim upside down or horizontally while making sweep-
three times faster and makes larger, regular head sweeps ing searches for prey. By changing its position, velocity,
at an average rate of two sweeps per second. The platy- and orientation, it can influence the pattern of incoming
pus is most sensitive to electrical stimuli that originate electrical signals, making it a formidable predator.
 Obtaining Food 263

Search Image Formation a

So far we’ve been talking about adaptations of the sen-
sory organs that enable animals to detect prey. Now let’s
consider a more subtle change that happens as an ani-
mal forages. Luuk Tinbergen (1960) was watching birds
bringing insects to their chicks in the Dutch pinewoods
when he became aware of an interesting pattern. As the
season progressed, the relative abundance of different
insect species varied—some species became more plen-
tiful, while others hit a population peak and then faded
away. One might expect that birds would capture the dif-
ferent species in proportion to their abundance. Instead,
when a species was in low numbers, it was taken less fre-
quently than would be expected. As it became more
common, it was taken more frequently than expected.
What could be going on? b
Perhaps the underlying process is similar to what we
commonly experience when we are searching for an
object. As you walk along a pond’s edge, you may find
that frogs seem to materialize in front of you and leap
into the water with an “Eep!” However, if you take time
to scan the area for a while, you can begin to pick out
froggy shapes, and soon you will see many frogs where
previously there had appeared to be none. The height-
ened ability to detect a target is called forming a search
image. Perhaps this is what drove the pattern Tinbergen
saw: after the birds got experience with a particular
species of prey, they began to focus their attention on it,
and they ignored other prey that did not fit their search
image. Since Tinbergen’s observation, this phenomenon
of “overselection” has been found under controlled con-
ditions in the laboratory (e.g., Bond 1983; Langley 1996;
Langley et al. 1996). In a typical experiment, birds are c
given a chance to forage for food, which can be manip-
ulated to either stand out from or blend in with the back-
ground. Overselection is more likely to be exhibited
when prey are camouflaged and difficult to pick out
against the background—obvious prey are generally
eaten in the expected proportions.
An assumption of the search image hypothesis is that
a search image is specific. If animals are really forming
a search image, we would expect that they improve in
their ability to detect only a particular kind of prey, not
all camouflaged prey. Pietrewicz and Kamil (1979, 1981)
used an elegant approach to simulate predator–prey
interactions between blue jays (Cyanocitta cristata) and
one of their normal prey types, underwing moths
(Catocala spp.). Hungry blue jays were given a slide show
(Figure 12.13). Some slides contained a moth; some did
FIGURE 12.13 (a) A blue jay is “foraging” for cryptic
not. If a moth was present and the bird saw it, it could
prey. The blue jay was shown slides, some of which con-
peck at the slide and receive a mealworm reward. After tained a cryptic moth. Among the cryptic moths shown
a short interval, the next trial began. When no moth was to the blue jays on slides were (b) Catocala retecta and
present, the blue jay could abandon its search of that area (c) C. relicta. If the bird was shown a slide that contained
and look elsewhere by pecking at a key that would then a moth and pecked the appropriate key after spotting it,
advance the projector to the next slide. The bird could the bird received a mealworm reward.
264 Chapter 12 / Foraging Behavior

therefore make two types of mistakes. One was a false What exactly comprises a search image? If you are
alarm—incorrectly responding as if a moth were present. looking for a book with a green cover and white letter-
The other was a miss, in which the bird failed to see the ing among the papers, tape dispenser, tissue box, and
cryptic moth and pecked at the advance key. Either of empty coffee cups on your desk, are you trying to com-
these mistakes resulted in a delay before the hungry pare a mental image of the book against each of the
predator could search for prey in the next slide. This objects on your desk? Or, instead, are you looking for a
experimental design, though it seems quite artificial, has particular feature (e.g., green items)? An interesting
the advantage over experiments using free-living birds experiment suggests that, at least sometimes, animals are
because it allows the experimenters to control the order looking for particular features. Photographs of wheat
in which the birds encounter prey. and beans were digitally manipulated so that either color,
The search image hypothesis predicts that recent shape, or both were changed. After experience with a run
experience with a prey type would allow the birds to of normal stimuli, pigeons were given one of these
learn its key characteristics and prime them to look for manipulated stimuli. Pigeons attended to both color and
others of that type. Encounters with one species of cam- shape of beans, but only to color of wheat—and this is
ouflaged moth would help jays find that species more the feature that best allowed them to discriminate the
accurately and quickly, but would not improve their abil- grain from the background (Langley 1996).
ity to find a second camouflaged species. The jays were The study of search image formation has new prac-
shown a series of slides, half of which contained a moth. tical implications. The detection of hard-to-find objects
They were tested under three conditions: (1) the moth- is exactly the problem faced by airport personnel look-
containing slides showed C. retecta only, (2) the ing for weapons on X-ray screens or in bag searches.
moth-containing slides showed C. relicta only, and One particular type of airport employee has been shown
(3) the two moth species were shown in random order. to use search images: sniffer dogs trained to detect
The results, shown in Figure 12.14, are exactly what explosives in luggage. When dogs were exposed to a
would be predicted by the search image hypothesis. The high percentage of TNT containers relative to other
jays’ ability to detect one prey type improved with con- explosives and were thus given the opportunity to form
secutive encounters. However, prey detection did not an olfactory search image, they were better able to
improve when the two species were encountered in ran- detect TNT (Gazit et al. 2005). (Be reassured—the
dom order. Thus, we see that experience with one type dogs’ ability to detect explosives was very high even
of cryptic prey improved the predator’s ability to find without the chance to develop a search image.) The snif-
that type of prey but not other kinds of cryptic prey. fer dogs also illustrate that search images need not be
visual.

100

OPTIMAL FORAGING
90
Are there any rules to make foraging as efficient as pos-
Percentage correct

80 sible? In this part of the chapter, we shift away from the

details of specific examples and expand to a broader per-
70 spective. We will use optimality theory to understand the
decisions that foraging animals make. In Chapter 4, we
60 introduced the technique of optimality modeling.
Catocala retecta only Animals have different behavioral options (strategies)
50
Catocala relicta only available to them, and we use models to weigh the costs
Randomized
and benefits of each strategy. A model is a mathematical
40 expression of all the costs and benefits of each strategy.
0 1 2 3 4 5 6 7 8 These costs and benefits are measured by a common
Position in sequence currency that represents some measure of fitness. In
other words, animals that make the best choice are
FIGURE 12.14 The percentage of correct responses by favored by natural selection. Remember that we do not
blue jays when shown moth slides in sequences of the
need to assume that the animal is able to work out com-
same species (runs) or a random sequence of both
plicated solutions in its head—natural selection can, in
species. After experience with one cryptic species of
moth, blue jays became better able to detect that a sense, do the hard work of giving animals the appro-
species. However, if the species were shown in random priate abilities for solving the optimality problem.
order, the jays’ performances did not improve. This is Foraging has been a favorite subject for testing opti-
consistent with the search image hypothesis. (Data from mality theory because it is relatively easy to fit into a
Pietrewicz and Kamil 1981.) modeling framework. First, we can often break down
 Optimal Foraging 265

foraging into a series of decisions and then focus on one is a very simple model. It’s best to read this sample model
type of decision at a time. Examples of some of these slowly and try to follow each step. (If you are one of the
decisions are what to eat, where to look for food, how many readers who tends to skip over equations, this is a
long to search one area before moving on, and what sort good chance to practice translating math into English.)
of path to take through an area. Here is the scenario that we are modeling. A forager
Second, often we can identify a logical currency, or is searching for food. Two kinds of food are available in
common measure, by which to compare these decisions. the environment, and the forager only finds one piece
For example, for many species it is reasonable to assume of food at a time. Thus, when the forager encounters a
that it is useful to maximize the rate of energy gain over piece of food, it has a decision to make: it could eat it,
a particular time period. Increased food intake increases or it could ignore it and keep looking (and perhaps find
survival and fecundity (number of offspring) in many a better piece of food). In modeling jargon, two strate-
species. (To take just two of many examples, both spiders gies are available to the forager.
and songbirds produce more offspring when given more Next, in order to decide which strategy is best, we
food; Frey-Roos et al. 1995; Sherman 1994.) Thus, we must compare them using a common currency. In this
are often reasonably comfortable in using short-term model, we will assume that the best decision that a for-
measures of foraging success as an indicator of long-term ager can make is the one that produces the fastest rate
fitness. Energy consumption (caloric intake) has the of energy gain. Thus, our currency is the rate of
additional advantage that it can often be measured energy gain.
directly. Thus, energy-based models are a reasonable Next, we must identify the constraints in our model.
place to start. However, we should always think critically One constraint is the amount of time it takes to process
about, and try to test, whether our study system meets food: a nut, for example, must be shelled, a crab must
the assumption that the foraging behavior we are inter- have its carapace removed, but a juicy caterpillar need
ested in has evolutionarily meaningful consequences. only be swallowed. We will call this processing time the
Finally, we can often identify limitations, or handling time. We will also assume that an animal can-
constraints, on foraging behavior. For example, an ani- not eat food and look for food at the same time.
mal’s ability to gather food may be constrained by its gut It also takes time to find food, and different types of
capacity, its ability to detect food, or the presence of food may be easier or more difficult to find. For exam-
predators in the environment. ple, moths on trees may be difficult to discover, and
In the next section, we will describe two models that other prey items may be relatively rare. We call this the
illustrate different types of foraging decisions: diet selec- search time. These are simple ideas, and it is easy to
tion and movement from one patch of food to another. describe them using variables:
These are both simple models with only a handful of
E1 ⫽ the amount of energy gained by eating
variables and some restrictive assumptions. We will then
prey type 1 (in calories)
explore ways to improve their realism.
E2 ⫽ the amount of energy gained by eating
prey type 2 (in calories)
DIET SELECTION: A SIMPLE MODEL
Foraging animals often encounter many items that are For handling food:
possible to eat. The question is whether they should
include all these different types of food in their diet, or h1 ⫽ the time it takes to eat prey type 1
whether they should instead focus on just some of them (in seconds)
and ignore the others. You can imagine that, in real life, h2 ⫽ the time it takes to eat prey type 2
all sorts of considerations might go into this decision— (in seconds)
how hard it is to gather particular items, how rare they
are, how tasty, their nutritional value, if they are dan- For searching:
gerous to catch, and so on. All of these are variables that
describe some aspect of each food item, and each might S1 ⫽ the amount of time it takes to find prey
well be important for a given species. However, it would type 1 (in seconds)
be very difficult to account for all of them at once.
S2 ⫽ the amount of time it takes to find prey
Instead, we are going to begin by stripping this
type 2 (in seconds)
problem down to its bare bones and focusing on only a
few variables. In general, models oversimplify nature. As To see what the next step in constructing our model
a rule of thumb, simpler models have more general con- should be, consider the currency. We want to measure
clusions that apply to a wider range of examples, but the rate of energy gain from each prey type. A good clue
more detailed models tend to make more precise and can be found in the units: rate of energy gain is measured
accurate predictions for a given situation. This example in calories/second.
266 Chapter 12 / Foraging Behavior

Once a forager has found a prey, what is the rate of type 2, then the forager should always ignore it, even if
energy gain? We call this the profitability of each prey. prey type 2 is piled up to the forager’s chin!
E1 calories
profitability of prey type 1 ⫽ in units of
h1 second Tests of the Diet Model
E2 How does one test the optimal diet model? Let’s begin
calories
profitability of prey type 2 ⫽ in units of with a few specific examples. Observations that are
h1 second qualitatively consistent with these predictions were
To make this model easier to discuss, let’s define prey found in a field study of redshanks (Tringa totanus),
type 1 as the type with the highest profitability. Now we shorebirds that feed on worms (Figure 12.15).
are ready to make our first prediction based on this Redshanks feed on both large and small worms. Large
model. Imagine a forager has just found prey type 1. worms provide more energy than small ones. In some
Should it eat it? (Would you? Think about this for a locations, large worms are common (search time, S1, is
moment before you continue reading.) short) and in other locations, they are rare (S1 is long).
In our model, a forager should always eat prey type At sites where large worms were common, the birds
1. It has the highest profitability, so the forager can never were indeed more selective and ignored the small
do better than to eat it. This is the first prediction from worms (Goss-Custard et al. 1998).
our model. In field studies such as this one, the rate at which the
Now consider a more challenging question. Imagine subject encounters different prey types is largely out of
a forager has just found prey type 2. Should it eat it, or the experimenter’s control. To better control for
should it ignore it and keep looking? encounter rate, John Krebs and his co-workers (1977)
To answer this question, compare the following rates designed an apparatus that would allow them to control
of energy gain. The first is the rate of energy gained by the abundance of two different food items presented to
eating prey type 2. The second is the rate of energy great tits (Parus major). The birds were allowed to pick
gained by ignoring prey type 2 and continuing to search. mealworms off of a conveyor belt that ran past their
(Take a minute to write out what these would be in sym- cage. This allowed Krebs to control both the type of prey
bols before you look ahead.) encountered and the abundance of the different prey.
The profitability of prey type 2, once a forager has The prey items were small and large pieces of meal-
found it, is E2/h2 The rate of gain of finding and eating worms. The energy value of large mealworms was
prey type 1 is a little different because the forager must greater than that of small ones, but the handling time
find it first, and then it can eat it. Thus, the two rates we was equal for the two prey types. As predicted by the
must compare are: model, birds always ate the large mealworms. As the
large mealworms were made more plentiful, the birds
E2 E1 became selective and generally ignored the small ones.
7 However, the shift between including and excluding
h 2 S1 + h 1
small worms in the diet was not sudden, as predicted by
The left side of the equation says “Now that I have the model, but gradual. A number of explanations have
found prey type 2, what is the energy per second that I been proposed for what is termed partial preferences for
will gain if I eat it?” The right side of the equation says high-ranking prey, including discrimination errors (con-
“What is the energy I will gain per second if I look for fusion between large and small prey) and the time
and eat prey type 1 instead?” When the rate on the left required to learn the values of S, the rates of encounter
side of the equation is greater than the right side, the for- of different prey types (Krebs and McCleery1984). Even
ager should eat prey type 2. This is the second predic- when Krebs’s experiment was repeated with modifica-
tion of the model: the predator should switch tions, including giving the birds time to estimate S, birds
instantaneously between including prey type 2 in the diet still never excluded the smaller prey from their diet
or not, depending on which side of the equation gives (Berec et al. 2003).
the higher rate of energy gain. So, how successful is optimal diet theory? Sih and
Take a minute to examine this equation a bit further. Christensen (2001) reviewed 134 studies from 1986 to
Which of the variables is missing? You will notice that 1995 and scored each on the degree of fit of the data to
S2 is not in the equation. It may seem counterintuitive, predictions. They found a great deal of variation in the
but the third prediction of the model is that the forager performance success of the model. Interestingly, the
should not take into account the search time for prey model did a good job in explaining the diets of foragers
type 2 when making this choice. Another way to say this that consume immobile or essentially immobile prey
is that the number of the less profitable prey in the habi- (such as leaves, seeds, nectar, mealworms, and clams) but
tat (type 2) should not influence the choice. If the equa- did not do well when prey were active. This makes sense:
tion above says that the forager should not eat prey the optimal diet theory, as originally formulated, does
 Optimal Foraging 267

gained as it stays in a patch. To return to our berry-

picking example, you can imagine that the y-axis is the
total number of berries that you have accumulated in
your bucket. The curve flattens and nears an asymptote
as the berries run out. So, when should you give up and
move to another patch?

STOP AND THINK

It’s a bad year for berries. Each patch of bushes has fewer
total berries, but the travel time between bushes remains
the same. Should you stay in each patch for a longer,
shorter, or the same amount of time than in a normal year?

To answer this question, you need to know more

information: how far away is the next patch? If it’s a few
steps away, you might be more likely to leave and move
to a better patch as soon as berries start to become dif-
ficult to find. However, if it’s a 15-minute walk, you
might be more likely to strip every last berry off the
FIGURE 12.15 The redshank is a shorebird that feeds on bushes. We can add the travel time to the next patch as
worms. Large worms are more profitable than small a point on the left-hand side of the x-axis. It represents
ones. When redshanks forage in areas where large the time it will take before you can start getting energy
worms are abundant, they are more selective and eat from the next patch. By drawing a tangent from this
more large worms than they do in areas where large point to the gain curve, and dropping a line straight
worms are rare.
down, we can determine the best time to leave your cur-
rent patch. This is called the marginal value. If an ani-
mal leaves a patch before this time, it will be traveling
not take into account the behavior of prey. Two types of
when it could still be profitably foraging, but if it stays
prey may be of equal abundance, but if one is better at
too long, it will waste time searching for food in a
hiding or is more likely to escape successfully, then the
depleted area.
predator’s diet will reflect that.

Tests of the Marginal Value Theorem

DECIDING WHEN TO LEAVE A PATCH:
The marginal value theorem has been tested in a wide
THE MARGINAL VALUE THEOREM
range of organisms, both vertebrates and invertebrates.
As an animal forages within a particular location, food Many studies mimic the example above almost exactly:
in that patch may become more difficult to obtain. the experiment measures the duration that foragers
Imagine picking blueberries along a roadside. There spend in either natural or artificially constructed patches.
are patches where blueberries are common, separated The marginal value theorem can also be applied to
by areas without berry bushes. As you pick berries slightly different questions, such as predicting how much
from a particular patch, your success rate will change food an animal should carry in a single load back to a
over time. The longer you stay, the slower your bucket central place, such as a nest or burrow. For example, east-
will fill, as the berries get more and more rare. ern chipmunks carry food stuffed in their cheeks. It gets
Similarly, prey may become increasingly rare as a harder and harder to add more food as their cheek
predator hunts because they take evasive action. pouches get full, resulting in a gain curve just like that
Regardless of the cause, at some point it will become in Figure 12.16. The amount of time that chipmunks
advantageous for the forager to move to a new patch, spent at a seed tray (patch time) increased with the dis-
where food will be easier to find. Here again we can tance between the seed tray and the burrow (Giraldeau
model the decision of the forager: when should it stay and Kramer 1982).
in a patch, and when should it go? This problem was Reviews of the performance of the marginal value
modeled by Charnov (1976) in the marginal value theorem (Stephens and Krebs 1986, Nonacs 2001) show
theorem. that there is good qualitative support for it. In general,
This model is exceptionally easy to present as a animals prefer rich patches to poor ones, patch residence
graph. The curved line in Figure 12.16 is the gain curve. times correlate with patch quality, and increased travel
It is the cumulative amount of energy that a forager has time leads to longer time in patches. However, foragers
268 Chapter 12 / Foraging Behavior

Cumulative energy gain

Optimal
time in patch

Travel time Time in patch

FIGURE 12.16 The marginal value theorem. The curved

line is the gain curve. To the right side of the x-axis,
the time that an animal stays in the patch increases.
To the left side, travel time increases. A line drawn from
the travel time and tangent to the gain curve indicates
the optimal time to spend in a patch. A shorter travel
time means a shorter optimal time in the patch. FIGURE 12.17 Moose must obtain enough energy for
growth and maintenance, and the leaves of deciduous
consistently stay longer than predicted (Nonacs 2001), trees contain more energy than aquatic plants.
suggesting that this simple version of the patch model However, moose must also obtain a minimal amount of
sodium. So, although sodium-containing aquatic plants
might be too simple. Better results are obtained if, for
provide less energy, moose must eat some low-calorie
example, the nutritional state of the forager is considered
plants. This is an example of how nutrient requirements
(Nonacs 2001). We’ll return to the role of nutritional can influence optimal foraging.
state later in this chapter.
have a higher concentration of sodium. A moose bal-
ances these needs by eating a mixture of plants so that
ADDING COMPLEXITY AND REALISM
energy intake is as great as it can be while sufficient
These two simple models give us a starting point for sodium is still obtained. When sodium needs are added
thinking about foraging, but they do not make accurate to the model, moose behavior is more accurately pre-
predictions for every case. Now let’s look at some added dicted (Belovsky 1978). Another forager that attends
variations that can help us better predict the foraging to nutrients is the wild stripe-tailed hummingbird.
strategies of animals. Hummingbirds spent more time at feeders that had
vitamin tablets dissolved in the sugar solution than at
control feeders with only sugar water (Carroll and
Energy Alone Is not Enough
Moore 1993).
So far we have assumed that the only thing a forager has Insects also regulate their nutrient intake. For
to consider is maximizing its energy intake, but we instance, when locusts (Schistocerca gregaria) that have
should not be surprised that animals often have specific been kept on a diet low in protein or carbohydrate are
nutritional requirements. The commonly touted nutri- given a choice of food, they select the food that redresses
tional guideline for humans, “Eat a variety of foods,” their nutritional deficiencies (reviewed in Simpson et al.
may also apply to other animals. For instance, nestlings 2004). Similarly, tent caterpillars (Malacosoma disstria) can
of the European bee-eater (Merops apiaster) convert food select nutritionally balanced food over unbalanced food.
to body weight more efficiently if they are fed a mixture However, there is an odd twist to this story: these cater-
of bees and dragonflies than if they eat only bees or only pillars are extremely gregarious, and they steadfastly fol-
dragonflies (Krebs and Avery 1984). low the silk lines of their nestmates. If the first caterpillar
A classic example of nutritional constraints comes chooses the wrong food type, the rest follow the trail and
from the moose (Alces alces) that live on Isle Royale in become “trapped,” unable to reverse the suboptimal
Lake Superior. They must obtain enough energy for choice (Dussutour et al. 2007).
the growth and maintenance of their huge bodies
(Figure 12.17), but they also have a minimum daily
requirement for sodium. The leaves of deciduous trees Incomplete Information
on the shore contain more calories than aquatic plants, You may have noticed something very unrealistic about
so to maximize energy intake the moose should eat the models we have discussed: they assume a great deal
only land plants. However, land plants have a low of knowledge on the part of the forager. In the diet
sodium content. In contrast, low-calorie aquatic plants model, for example, our calculations assume that the ani-
 Optimal Foraging 269

carefully assess whether the animal has all the informa-

tion at its disposal that the model assumes it does.

Rules of Thumb
We know that animals do not actually make the complex
calculations that models do, any more than you solve
equations before selecting a snack from the refrigerator.
Animals, however, may be able to get close enough to an
optimal behavior by following an approximation called
a rule of thumb. Let’s look at an example.
Northwestern crows (Corvus caurinus) search along
the waterline during low tide for whelks, which are large
snails. A crow will pick up a whelk, carry it over the
FIGURE 12.18 Chipmunks generally forage on seeds rocks, fly almost vertically upward, and then drop the
beneath deciduous trees. Thus, their food is found in whelk. If the whelk smashes open, the crow eats the
patches of fluctuating abundance. As the quality of their meat. If not, the bird retrieves the whelk, flies upward,
current feeding location declines, they spend more time and drops it again, repeating the procedure until it does
sampling the food abundance at other locations. break.
Reto Zach wondered whether the crows were selec-
mal knows the search time for each of the two types of tive about which whelks they dropped. He also won-
prey. This is not necessarily true: a forager may be able dered whether there were differences in the ease of
to learn that with experience, but it takes some time. handling whelks of different sizes. One might expect that
Similarly, with the marginal value theorem we assume flying upward would be the most costly part of eating
that the forager knows average travel time between whelks in terms of energy. Then Zach asked whether
patches, and what the gain curve of other patches looks crows adjust their cost according to the expected caloric
like. What if you don’t know where all the blueberry value of the meal. How many times would a crow con-
bushes are? To gain this knowledge, the animal must be tinue to drop a whelk that was difficult to crack open?
familiar with the area and must perhaps even periodically Zach (1978) found that the crows preferentially prey
sample other patches. on large whelks. He demonstrated this finding in several
Some foragers, in fact, do monitor their environ- ways. First, he collected broken pieces of whelk shells,
ment in this way. For instance, chipmunks feed on and by comparing the size of the base of the shells to
seeds from deciduous trees, found in patches of fluc- those of living whelks, he estimated the size of the whelks
tuating abundance (Figure 12.18). The supply of seeds that had been eaten. The whelks selected by the crows
below a particular tree may be here today and gone appeared to be among the largest and heaviest on the
tomorrow, varying with the schedule of ripening, beach. Larger whelks provide more energy, so this obser-
amount of wind, and activities of other animals. As a vation was consistent with the hypothesis that the crows
result, chipmunks must decide how often to check the could distinguish profitable prey. However, Zach was
seed supply at other trees to determine whether it cautious in his interpretation. Perhaps, he reasoned, the
might be beneficial to switch foraging locations. Of pieces of smaller shells were more easily washed out to
course, time spent checking other trees is time lost sea by wave action, leaving the larger pieces overrepre-
feeding, so there is an optimal amount of sampling. In sented in this sample. So, Zach offered each of three
one experiment, the chipmunks fed from artificial pairs of crows equal numbers of small, medium, and
patches—trays of sunflower seeds—and the value of a large whelks and, at hourly intervals, recorded the num-
given patch was manipulated by varying the number of ber of each size taken. Crows selected large whelks
seeds on the tray. The chipmunks spent more time (Table 12.1). Even after the crows had eaten most of the
sampling the food density at other locations as the large whelks, they continued to ignore smaller ones. Was
quality of the patch being exploited decreased (Kramer this because large whelks are more palatable? No. By
and Weary 1991). removing whelks from their shells and presenting equal
Although chipmunks seem to successfully monitor numbers of each size class to crows, Zach demonstrated
their environments, other animals may not, or may not that all size classes were equally palatable.
yet have had time to accurately sample. In a few species, Why, then, do crows accept only the largest and
animals are able to assess their environment by watch- heaviest whelks on the beach? One reason seemed sim-
ing conspecifics; we will return to this issue in Chapter ple enough—larger whelks have a higher caloric value.
19. When evaluating the predictions of a model, we must However, Zach wondered whether size affects the ease
270 Chapter 12 / Foraging Behavior

TABLE 12.1 Numbers of Small, Medium, and Large Whelks Laid Out
and Cumulative Numbers of Whelks Taken Over the Subsequent 5 Hours*
Size of whelk

Small Medium Large

Total laid out 75 75 75

Taken after 1 hour 0 2 28
2 hours 0 3 56
3 hours 0 4 65
4 hours 0 4 68
5 hours 0 6 71

*Results from three pairs of crows were homogeneous (replicated goodness-of-fit test) and therefore combined.
Right from the start whelks were taken nonrandomly (p <.005; single classification goodness–of-fit test).
(From Zach 1978.)

of breaking the shell as well. To answer this question, he Given this information, Zach (1979) was able to cal-
collected whelks of different sizes and dropped them culate a crow’s expected energy profit from whelks of dif-
onto rocks from different heights. He found that large ferent sizes. He found that crows are likely to gain 2.04
whelks were more likely to break than the medium and kilocalories from a large whelk, but they use 0.55 kilo-
small whelks; the large ones required fewer drops from calories to obtain it. Thus, the net energy gain for a large
any given height. whelk is 1.49 kilocalories per whelk. Medium and small
We see in Figure 12.19 the total height to which a whelks, on the other hand, require more energy to han-
crow would have to fly, on average, to break whelks of dle because they are harder to break. They also contain
the three different sizes at different heights. Notice that fewer calories. Zach calculated that medium whelks actu-
the probability that a whelk will break increases with the ally cost energy to eat: the net energy gain for a medium
height of the drop. The lower the dropping height, the whelk is –0.30 kilocalories. Obviously, small whelks,
more drops are required to break it. The total vertical which are even harder to break and contain fewer calo-
height needed to break the shell can be determined by ries, would be even more costly to eat. Crows thus
multiplying the number of drops by the height of the attempt to feed upon only profitable whelks, which are
drop. There is a height—slightly more than 5 m—at large whelks.
which the total height required to break a whelk is min- Crows make similar decisions when feeding on lit-
imized. Zach (1979) found that the crows dropped tleneck clams (Tapes philippinarum). Again, the largest
whelks from a mean height of 5.23 m (⫾0.07 m). clams are the most profitable, and again, crows choose the
largest clams to drop (Richardson and Verbeek 1986).
100 So far, the crows’ behavior is just as predicted by the
models. Let’s add another twist. If crows are given a
Total height per whelk (m)

80 choice between clams and whelks, what should they do?

Small
Clams, per unit of weight, offer more calories than
60 whelks, so even a slightly smaller clam should be pre-
Medium ferred to a larger whelk. However, crows did not always
40
pick the most profitable prey. Instead, they followed the
Large
20 rule of thumb, “Take the heaviest prey item.” This rule
led them, on some occasions, to pick a hefty whelk over
0 a not-quite-so-heavy clam, and thus not get as many
1 3 5 7 9 11 13 15
Height of drop (m) calories as they might have. However, these losses were
not huge, and the rule of thumb provides a reasonably
FIGURE 12.19 The total height of the drop required to
good approximation of optimal behavior (O’Brien et al.
break the shells of different-sized whelks. Northwestern
2005).
crows choose large whelks and drop them from a height
of about 5 m (indicated by the arrow). This minimizes We often see rules of thumb such of these. It is very
the energy used in obtaining food and yields an energy often impossible for an animal to gather the data it would
profit. The crows would use more energy than they need to follow an optimal strategy, as behavioral, phys-
obtain by feeding on small or medium whelks. (From iological and time constraints limit its ability to assess its
Zach 1979.) environment (Stephens and Krebs 1986). One may
 Optimal Foraging 271

a b

River bed Woodland Plains Hills

Swakop River Namib Desert

FIGURE 12.20 (a) Chacma baboons make a trade-off between predation risk and energy gain. They feed on leaves,
flowers, fruits, and pods. They sleep in trees or on cliffs, but they spend most of the day on the ground, where they
are exposed to predators. The primary predators are leopards and lions. (b) The four habitats available to the baboons
in the study are riverbed, woodland, plains, and hills. Both the risk of predation and the abundance of food differ in
each habitat. The baboons do not spend much time foraging in the woodland, where food is most abundant. Instead,
they feed mostly in the riverbed habitat. Although less food is available, this habitat is safer from predators. (Diagram
from Cowlishaw 1997.)

approach this problem by comparing the performance of have a simple diet. During the time span of this study
these rules of thumb against the performance of an opti- (late winter), 92 to 97% of the baboons’ feeding time
mal strategy, as we did for the crows, or by modifying was spent gathering leaves, flowers, fruits, and pods
the model to include more realistic constraints. Either from only five plant species. Over 90% of the food
approach gives us insight into exactly what determines energy was found in the woodland and almost all of the
animal decisions. rest in the riverbed. The most common predators of
baboons are leopards and lions, which hunt by stalk-
and-rush methods. They have better luck capturing prey
Avoiding Predators
from shorter ambush distances and are, therefore, more
Animals foraging for food are potential meals them- successful in habitats with cover that hides their
selves and would do well to take that into account. approach. Cowlishaw estimated the predation risk in
Numerous studies confirm that animals do indeed each habitat. The woodland had the highest estimated
assess predation risk (Brown and Kotler 2004; Lima and risk of predation, the riverbed and the plains had a mod-
Dill 1990). est risk, and the hills had the lowest risk. The woodland,
unfortunately for the baboons, had the most food.
How Foragers Minimize Predation Risk Chan- When foraging, the baboons spent more time in the
ges in foraging behavior in the face of predation risk safer, riverbed area, even though much less food could
are manifested in a number of ways, including: (1) be found there. Other activities, such as resting and
avoiding dangerous places, (2) avoiding dangerous grooming, were usually conducted in the hills, where
times of day, (3) increasing vigilance when foraging, the predation risk was lowest.
and (4) selecting portable foods. Let’s look at each of The threat of predation can also affect when an ani-
these in turn. mal chooses to forage. Scorpions, fearsome predators
Species from a range of taxa forage in less profitable
but safer sites, rather than in more profitable but more
dangerous areas. To take an example, the foraging TABLE 12.2 Relative Food Abundance and
behavior of a desert population of chacma baboons, Predation Risk in Four Habitats Available
Papio cynocephalus ursinus (Figure 12.20a) was studied by to Baboons in Tsaobis Leopard Park
Cowlishaw (1997). The site was Tsaobis Leopard Park, Food abundance Predation risk
Namibia, located in southwest Africa. The park’s envi-
ronment is rugged, with mountains and ravines, as well Riverbed Modest Modest
as gravel and alluvial plains. Four different habitats can Woodland High High
be identified in the reserve—riverbed, woodland, plains, Plains Negligible Modest
and hills—each differing in food availability and preda-
Hills Negligible Low
tion risk (Figure 12.20b; Table 12.2). These baboons
272 Chapter 12 / Foraging Behavior

themselves, can also be victims. One species forages less to normal spiders—by decreasing their movement. The
on moonlit nights when they are likely to be more visi- researchers set up various types of enclosures with
ble to predators such as owls (Skutelsky 1996). Because plants, grasshoppers, and glued or normal spiders. The
scorpions detect their own prey by its vibrations, as presence of spiders—glued or unglued—reduced the
described above, their own foraging success is not com- amount of plant damage by grasshoppers. The spiders’
promised by this cautious behavior. mere presence changed the behavior of the grasshoppers
If an animal is foraging in a dangerous place, it can enough to have a significant, measurable impact on
reduce its risk by being more vigilant: for example, it can plants. Thus, this example illustrates how antipredator
lift its head and look around more, or be quicker to flee behavior can have a cascade of consequences for an eco-
if it hears a noise. Vigilance may come with an energetic logical community.
cost, as it is often difficult or impossible to be vigilant
while foraging. For example, when wolves were reintro-
The Presence of Competitors
duced into Yellowstone National Park, greater elk
(Cervus elaphe) spent more time being vigilant and less American crows (Corvus brachyrhynchos) drop walnuts to
time feeding (Laundre et al. 2001). break them open in the same way that northwestern
Perhaps less obviously, predation can also affect diet crows drop whelks (Figure 12.21). Recall that the upward
choice. Remember the example that opened the chap- flight to drop the food item is energetically costly. Cristol
ter. Under risky conditions, foraging gray squirrels will and Switzer (1999) have shown that American crows
sometimes reject items with a higher profitability adjust the height from which they drop a walnut accord-
(energy gained per second of handling time), in favor of ing to the circumstances. English walnuts break more
less profitable items that are easier to carry to a safe place easily than black walnuts, and the crows drop the English
(Lima and Valone 1986). walnuts from lower heights than black ones. The crows
adjust the height of the drop to account for substrate
hardness. American crows often feed in large flocks, and
Quantifying How an Animal Perceives Risk An so there is always the threat that a dropped nut will be
interesting tack that many researchers have taken is to stolen by another bird. The crows also adjust the height
“ask” animals exactly how risky they perceive a partic- to minimize the chances of theft. The risk of theft was
ular site to be. For example, gerbils (Gerbillus a. allen- determined by using an index that combined the num-
byi) were given a choice between foraging for seeds in
a safe plot or a “risky” plot. “Risky” plots were either
exposed to the simulated light of a full moon or were
visited by a trained owl flying overhead. As you might
expect, gerbils preferred the safe plot over the risky plot
when both held the same number of seeds. However,
gerbils were willing to accept increased risk when
the price was right: if enough seeds were added to the
risky plot, gerbils would forage there (Abramsky et
al. 2002).

The Cascading Effect of Risk Avoidance The fol-

lowing experiment hammers home the wide-reaching
ecological consequences of antipredator behavior by
foragers. Beckerman et al. (1997) studied an ecosystem
of spiders, their grasshopper prey, and plants that the
grasshoppers fed on. Grasshoppers decrease their move-
ments when in the presence of spiders, presumably to FIGURE 12.21 American crows adjust the manner in
avoid detection. The question was whether this which they handle food so that energy gain is maxi-
antipredator behavior affected the amount of plant dam- mized. The crows eat both English and black walnuts,
age. The researchers devised a clever way to distinguish which they break open by flying up to 30 m upward
and dropping them onto the ground. The energetic
between the effect of the spiders consuming the
cost is the upward flight to drop the walnut. American
grasshoppers and the effect of the antipredator behav-
crows adjust the drop height to account for the hard-
ior adopted by the grasshoppers in response to the spi- ness of both the walnut and the ground. Since they
ders. The research team created “safe” spiders by gluing feed in flocks, there is a risk that a dropped walnut will
their mouthparts together. These spiders could move be stolen by another bird. The crows lower the drop
around normally but could not kill prey. Grasshoppers height when the threat that the walnut would be stolen
responded to glued spiders in the same way as they did exceeds a certain threshold.
 Optimal Foraging 273

ber and proximity of conspecifics. When the risk which symbolizes one of its changeable attributes (its
exceeded a certain threshold, the crows lowered the drop state). So, for example, a modeler might represent
height so that they could recover the nut before it was hunger with one state variable, and size with another.
snatched by another bird. Thus, even given the same set of ecological circum-
You can see that some crows do all the hard work of stances, different individuals may make different deci-
flying up, whereas others get something for nothing and sions. The decisions an animal makes on one day
grab the nut that someone else has discovered and influence its state, which in turn influences the deci-
dropped. This sort of pilfering is not at all unusual, and sions it makes the next day, making the system change
a whole class of models has been created to explore it. dynamically. These models thus capture more of the
These are called producer/scrounger models, where a complexity of real-life foraging but are more computa-
producer is the animal that makes the resource available tionally complex.
and the scrounger is the one that steals it.
Risk Sensitivity: Response to Variability
The Role of Internal State In the models we’ve considered so far, we’ve been
The simplest foraging models assume that all animals assuming that foragers are maximizing the long-term
behave the same way. For example, the diet selection rate of energy gain. However, foragers may also
model took into account only energy, search time, and respond to the variability in food availability. That is,
handling time. However, individual foragers vary. an individual may have to choose between a site that
Some might be more experienced foragers than others, reliably supplies a moderate amount of food and one
in better condition and thus able to move more quickly, that fluctuates between a rich and poor food supply.
closer to reproduction, or hungrier. These variations The individual that chooses a variable site could get
across individuals may mean that the optimal decisions lucky and find plenty of food quite easily, but there is
of foragers differ. For example, the colonial spider always the risk that food will be scarce. In the jargon
(Metepeira incrassata) lives in large colonies with a of foraging theory, the term risk in these models refers
shared frame web. Each spider puts up an individual to variability in food abundance. Some animals are
orb within the frame. The spider’s position within the gamblers and choose the variable site. They are called
colony influences foraging success and the risk of pre- risk-prone. Others, those who are risk-averse, tend to
dation or parasitism. Individuals on the periphery of the choose reliable sites where they are more or less guar-
colony are more successful at capturing prey, getting 24 anteed of finding at least some food (Stephens and
to 42% more flying insects than individuals in the Krebs 1986).
colony’s core. Unfortunately, because predators (wasps Risk sensitivity has been documented in a variety of
and birds) approach from the edges of the colony, taxa, including insects, fish, birds, and mammals
peripheral spiders are also more likely to be eaten. (Kacelnik and Bateson 1996). In a typical experiment, an
Thus, as we have seen in other species, there is a trade- animal is offered a choice between a constant option that
off between foraging success and predation risk. This always offers the same amount of food, and a variable
trade-off is especially pronounced for large females option that offers the same average amount of food as the
because predators prefer them. Furthermore, if the constant option, but with variability. The animal is given
eggs are left unguarded at the periphery, there is an the chance to learn about both these options before mak-
increased chance of egg-mass parasitism. As a result, ing its choice.
most individuals hatch and begin life in the central One interesting pattern in many studies is that an
regions of the colony. Younger, smaller spiders may animal’s hunger level often determines whether it is
take a chance and build their orbs on the edges of the risk-prone or risk-averse (Bateson 2002; Kacelnik and
colony, where they can obtain more food and grow Bateson 1996). Why might this be so? The reasoning is
faster, thereby increasing the odds of reaching sexual as follows: an animal that fails to find a certain minimal
maturity. But as the spiderlings mature, the balance of amount of energy each day will die of starvation. If
risks changes and safety becomes more important than enough food can be found at the site that provides a sta-
foraging success. So, the larger spiders that have ble food supply, there is no benefit in gambling on find-
reached sexual maturity prefer the core positions ing sufficient food at the variable location. However, if
(Rayor and Uetz 1990). the stable site does not provide enough food to prevent
Any situation where each individual’s traits influ- starvation, the only chance for survival is to forage in
ence the decisions that are optimal requires a much the location where the food supply is variable and hope
more sophisticated model than we have presented so far. for the best. In general, because animals that are full are
One way to approach this problem is with dynamic less likely to starve, they should be risk-averse, whereas
state-variable models (Clark and Mangel 2000). Every hungry animals should be risk-prone (Stephens and
individual is described by a set of variables, each of Charnov 1982).
274 Chapter 12 / Foraging Behavior

THE UTILITY OF MODELS its costs by the greatest amount. Several simple models
form the basis for making predictions about optimal for-
Now that we have seen that few animals behave exactly
aging. The first model is the diet selection model, which
as predicted by our simplest models, it’s worth asking
assumes that foragers should attempt to maximize their
whether it is useful to develop them at all. The per-
rate of energy gain. The profitability of a food item is
spective of models, as we described in Chapter 4, is that
defined as the energy it provides divided by the time it
modeling provides a chance for us to clarify our assump-
takes to find, capture, prepare, and digest that item. The
tions about our study animal, such as how its behavior
model predicts that when the most profitable items are
relates to evolutionary fitness and what information it
abundant and search time for them is short, less prof-
has about its environment. Models are really just a way
itable items will be eliminated from the diet. In general,
to formally state a hypothesis about a behavior. Based
tests of the model have shown that it does a good job
on that hypothesis, we can generate testable and often
when predicting the behavior of foragers that consume
quantitative predictions. If we can successfully predict
immobile or essentially immobile prey.
animal behavior, we are more confident—though rarely
A second decision that animals must make when
positive—that we understand it.
they are foraging is whether or not to leave a patch of
food and move onto the next patch. Again, we assume
that animals are maximizing their rate of energy gain.
SUMMARY The marginal value theorem, a graphical model, uses the
energy gain within a patch and travel time to the next
There are many ways in which an animal can obtain patch to predict when an animal should leave the patch.
food. Many aquatic animals filter food from the When travel time increases, animals should stay longer
surrounding water. Some animals are omnivores, eating in their patch. The predictions of the marginal value
both plants and animals. Herbivores forage on plants. theorem are generally supported in empirical tests, but
Many behaviors have evolved to overcome plant defenses. foragers consistently stay longer than predicted.
Others foragers are drawn in by plants, which benefit There are numerous reasons why these simple mod-
by having their seeds dispersed or flowers pollinated. els may not be adequate to predict behavior. Animals
Some animals work as farmers and cultivate their own may choose less profitable food items because of specific
food food. The leaf cutter ant, for instance, maintains nutritional requirements. They may not know everything
fungus gardens. that the model assumes they know; they may have to
Carnivores are often engaged in arms races with their spend time sampling their environment. They may be
prey. Some carnivores rely primarily on speed to capture using “rules of thumb” to approximate the optimal
their prey, whereas others use stealth to approach with- behavior. They may be taking the possible presence of
out detection. Some lure their prey with specialized struc- predators or competitors into consideration.
tures or behaviors that mimic the courtship displays of the Our basic models assume that all individuals behave
prey species. Other predators use traps. the same. However, this may not be the case. Individual
Many predators have sensory specializations that characteristics of foragers, such as hunger and age, may
improve their ability to detect their prey. For example, mean that different behaviors are optimal for different
some species of snakes detect the body heat that individuals. More complex models are needed to deal
emanates from their warm-blooded victims. The sand with this situation.
scorpion, on the other hand, is exceedingly sensitive to Some animals are sensitive not only to rate of energy
vibrations in the sand that are created by its prey. Star- gain, but also to variability in the availability of food.
nosed moles have bizarre nasal appendages that are Risk-sensitive animals have preferences when given a
highly sensitive to touch. Petrels and albatrosses have an choice between a feeding location that supplies a mod-
extraordinary sense of smell. From a human perspective, erate, but constant, supply of food and one that fluctu-
perhaps the most unusual sensory specialization is the ates between a rich and a poor food supply. Often,
ability of some animals, including sharks, to detect the animals that are hungry tend to be risk-prone, choosing
electrical fields that living organisms generate when they the variable site, whereas satiated animals tend to be risk-
are in seawater. averse, choosing the constant site.
Another adaptation for finding cryptic prey is the use The simplest models often cannot predict exactly
of search images. Here, a predator learns the key features how animals forage. However, models force us to clar-
of particular prey and focuses on them while hunting. ify our assumptions, generate testable predictions about
According to optimality theory, natural selection behavior, and increase our confidence that we under-
favors the behavioral alternative whose benefits outweigh stand the factors that influence behavior.
 13
Antipredator Behavior

Camouflage What could be better to eat than a butterfly? Soft and

Coloration Matching the Visual Background juicy, with no teeth to nip you or claws to scratch you,
Disruptive Coloration butterflies are preyed on by birds, mammals, and spiders
Countershading alike. In light of the constant threat of predation, but-
Transparency terflies have developed an impressive array of devices to
Masquerade outsmart their enemies, and their protective strategies
Other Functions of Color appear to work, at least some of the time. Here we con-
Polymorphism sider the antipredator strategies employed by the
monarch butterfly (Danaus plexippus).
Warning Coloration
Like many animals, monarch butterflies use a com-
Batesian Mimicry bination of color pattern and behavior to avoid being
Diverting Coloration, Structures, and Behavior eaten. Their boldly patterned, orange, black, and white
False Heads wings warn potential predators that they taste bad, their
Autotomy unpalatability being due to their assimilation of noxious
Feigning Injury or Death chemicals from food plants. In particular, monarch lar-
Intimidation and Fighting Back vae feed on milkweed plants (Asclepiadaceae) and
Enhancement of Body Size and Display of Weaponry incorporate toxins, called cardiac glycosides, into their
Eyespots own tissues (Brower et al. 1968). Predatory birds that
Chemical Repellents eat one of these insects, even as adults, have severe vom-
iting and tend to avoid butterflies of similar appearance
Pronouncement of Vigilance
in the future. However, one might wonder what good
Group Defense it is to be filled with toxins if the individual must be
Alarm Signals eaten before the poisons will work. The advantage is
Improved Detection that many predators release, unharmed, prey that are
Dilution Effect brightly colored and bad tasting (some predators may
Selfish Herd even have an innate aversion to bright colors). Poisons,
Confusion Effect stolen from plants, may thus deter some predators.
Mobbing However, not all milkweed plants contain the same
Maintenance of Antipredator Behavior amount of cardiac glycosides. Butterflies reared on

275
276 Chapter 13 / Antipredator Behavior

plants that do not contain substantial amounts of the tering site. Forming dense aggregations, of course, can
poison are quite palatable, although they may still be have benefits in addition to avoiding being eaten (see
avoided by predators who have had experience with Chapter 19), and monarchs may gain additional bene-
more noxious members of the species. fits from overwintering in large groups.
No defense system works all the time. Even for the Our example of the monarch butterfly illustrates a
monarch butterfly, the effectiveness of protective strate- general point about antipredator behavior: animals typ-
gies varies with the season, species of predator, and con- ically have multiple devices to avoid being eaten.
text of the predator–prey encounter. For example, in the Monarchs also illustrate a second point that we will dis-
late summer and autumn, monarch butterflies from east- cuss next: the colors that function in antipredator devices
ern North America migrate to the mountains of central often have other functions as well. In monarch butter-
Mexico (Brower 1996). The months spent in Mexico, flies, the orange color that deters predation when paired
however, are far from a winter vacation (Figure 13.1). with black and white in bold patterns also functions in
Birds of two species, the black-backed oriole and the reproduction. Males with deeper orange on their wings
black-headed grosbeak, have penetrated the monarch’s mate more often than those with a lighter shade of
chemical defense system; these two species eat an esti- orange (Davis et al. 2007). In monarchs, then, the intense
mated 4,550 to 34,300 butterflies per day in some over- orange color works in concert in two different behavioral
wintering colonies (Brower and Calvert 1985). The contexts—it helps to deter predators and it enhances
oriole selectively strips off relatively palatable portions male mating success. It is not hard to imagine, however,
of the butterflies’ bodies (e.g., the thoracic muscle and a scenario in which a color or pattern that functions as
abdominal contents), and the grosbeak appears insensi- an antipredator device might not be ideal for mating suc-
tive to the cardiac glycosides (Fink and Brower 1981). cess. For example, a species that relies on having dull col-
However, all is not lost for the monarch. Facing the ors that blend with the environment might face
prospects of an avian feeding frenzy each winter, the but- problems when trying to attract a mate, a time when
terflies reinforce their antipredator system by converg- bright, conspicuous colors are usually best.
ing in enormous numbers at their overwintering Because predation is such a pervasive theme in the
colonies; some of these colonies have tens of millions of pageant of life, we can ask how animals cope with its con-
individuals (Brower et al. 2004; Calvert et al. 1979). By stant threat. Which devices aid in escaping detection by
forming dense aggregations, it is likely that the monarchs a predator, and which come into play once a prey ani-
dramatically dilute the predation risk to any one individ- mal has been detected and capture seems all too immi-
ual. Also, because predation is most intense at the nent? Does membership in a group always confer
periphery of the colony, central positions are highly antipredator privileges? And what compromises do ani-
sought after and are quickly assumed by the first indi- mals reach when avoiding being eaten conflicts with
viduals to arrive. In the life of a monarch butterfly, it does other critical behaviors such as feeding and reproducing?
not pay to be fashionably late in arriving at the overwin- We begin with camouflage.

FIGURE 13.1 Avian predators penetrate the

chemical defense system of monarch butterflies
at their overwintering sites in Mexico.
 Camouflage 277

use it in a more general sense to mean any coloration

CAMOUFLAGE aimed at avoiding detection. In its more general usage,
cryptic coloration would include other types of col-
Animals have several forms of camouflage (Endler 1981,
oration, such as disruptive coloration (see below). So, we
2006). Some forms, such as disruptive coloration, counter-
will use coloration matching the visual background
shading, transparency, and coloration matching the visual
because it clearly describes the phenomenon.
background, are thought to help prey avoid detection by
Probably all of us, at one time or another, have mar-
visually hunting predators. Another type, termed masquer-
veled at the ability of certain animals to blend with the
ade, relies on prey animals appearing to be inedible to
background on which they are resting. While walking
predators. Thus, while such prey are detected as distinct
through the forest, we might struggle to see a moth
from their background, their uncanny resemblance to a
against the bark of a tree or a grouse in its nest on the for-
leaf, twig, or bird dropping makes them of little interest
est floor (Figure 13.2). We might even describe their col-
to a hungry predator searching for another animal to eat.
oration as “incredible camouflage” to our hiking partner.
Although camouflage comes in several forms, each with its
But how do researchers study this phenomenon under
own unique twist, its general message is simple:
controlled experimental conditions? Can the degree to
“I am not here.” In the sections that follow, we will define
which animals match the visual background be quantified?
and provide examples of each type of camouflage and illus-
One way to assess such coloration stems from a particu-
trate how researchers test hypotheses concerning the dis-
lar definition of it. John Endler (1978) defined it as con-
guises of animals. Most forms of animal camouflage were
cealment that results from an animal’s resemblance to a
described more than a century ago. Surprisingly, direct
random sample of the visual background. Thus, in some
experimental evidence to support their roles as antipreda-
studies, degree of background matching is assessed by
tor devices is often quite limited. The good news, however,
quantifying the similarity between prey coloration and its
is that studies of camouflage are now flourishing and yield-
background, and often the researcher chooses the relevant
ing incredible insights into the interactions between
aspects to compare (perhaps the density or distribution of
predators and prey, and the evolution of prey coloration.
elements in the color pattern). Another option is to let
predators “tell” us about the effectiveness of different prey
colorations. For example, by monitoring the search times
COLORATION MATCHING THE VISUAL
of predators presented with different prey, researchers can
BACKGROUND
gauge the effectiveness of the prey’s camouflage; the
The coloration of some animals resembles their back- longer it takes the predator to find the prey, the closer the
ground and appears to reduce the risk of detection by match to the visual background. We will consider a study
visually hunting predators. It is sometimes called cryp- that used the second method.
tic coloration, and the strategy is called background Sami Merilaita and Johan Lind (2005) used artificial
matching. We will avoid using the term cryptic coloration prey and an artificial background, but real avian preda-
because it means different things to different people; tors, in an experiment designed to test the hypothesis
some people use it in a narrow sense to refer only to col- that background matching is maximized when coloration
oration matching the visual background, whereas others visually matches a random sample of the background.

FIGURE 13.2 Animals, such as this grouse,

that blend with their background are
often described as cryptic. Technically,
the grouse displays coloration matching
the visual background.
278 Chapter 13 / Antipredator Behavior

a b 18

Effective search time (sec)

16
Difficult Easy
14
1 cm 12

10

FIGURE 13.3 Testing the effectiveness of camouflage. 4

Difficult Easy
(a) The background used to test whether background
matching is maximized when prey coloration matches a FIGURE 13.4 Search times of great tits were longer for
random sample of the background. (b) Artificial prey the prey classified as difficult to detect than for prey
were classified as either difficult to detect or easy to classified as easy to detect. These results indicate that
detect. (Modified from Merilaita and Lind 2005.) different samples of a background may provide different
degrees of camouflage through background matching.
(Modified from Merilaita and Lind 2005.)
The predators were great tits (Parus major) that were
captured in mist nets and individually housed indoors at ground matching than simply color and pattern. Many
the research station in Sweden where the work was car- animals appear to select “correct” backgrounds, and once
ried out. Merilaita and Lind designed paper T-shaped there they exhibit behavior that maximizes their camou-
prey that the birds could learn to recognize. They also flage. The California yellow-legged frog (Rana muscosa)
designed a background of sufficient complexity to make inhabits swift-flowing streams in the woodlands of
detection of the prey relatively difficult (Figure 13.3a). southern California. The light gray granite boulders that
The paper background was mounted on cardboard. Prey line the streams seem conspicuous resting spots for the
patterns were random samples from the black and white yellow-brown frog. Below the water, however, these
background pattern and were of two types—those same boulders are covered by a yellow-brown layer of
judged difficult to detect and those easy to detect (Figure algae. At a moment’s notice, R. muscosa leaps into the
13.3b). Ease of detection for the two prey types was water and lies motionless against a background to which
judged by the researchers and by a small number of birds it is perfectly matched (Norris and Lowe 1964).
used in a pilot experiment, but not in the “real” experi- Is the combination of matching and selecting the
ment described next. In the “real” experiment, a bird was appropriate background adaptive? If it is, then prey
moved to a new cage and given about an hour to adjust should experience less predation when sitting on the sub-
to its new surroundings before testing began. At testing, strates that they tend to select as resting spots than they
each prey item was presented singly to the bird; the do when sitting on other surfaces. Blair Feltmate and
paper prey item had been lightly glued to the back- D. Dudley Williams (1989) tested this idea by using rain-
ground to cover a hole that contained a food treat, a bow trout (Oncorhynchus mykiss) as predators and stone-
peanut chip (the birds had been previously trained to fly nymphs (Paragnetina media) as prey. Background color
associate the paper prey with food). The researchers preferences of stoneflies, stream insects that are dark
measured with a stopwatch the time that a bird spent on brown to black in color, were first tested by placing each
the background searching for the prey. Merilaita and of 24 nymphs into its own aquarium along with one dark
Lind found that search time was significantly longer for brown and one light gray commercial tile on the bottom
the prey expected to be more difficult to detect than for of the aquarium. Nymphs were left to settle for 24 hours,
the prey expected to be easier to detect (Figure 13.4). and then at 1400 hours (2 P.M.) the researchers recorded
These results indicate that all samples of a background whether the nymphs rested on the dark brown or light
do not provide equally camouflaged prey coloration. The gray substrate. The experiment was repeated with
authors suggest that prey coloration matching a random recordings of nymphal position at 0200, 0600, 0800, and
visual sample of the background may only maximize 2100 hours to test whether selection of substrate varied
background matching on very simple backgrounds. This as a function of time of day (lights in the laboratory were
study speaks to the benefits of designing experimental on timers and were off from 1900 to 0700 hours). Thus,
conditions that mimic, to some extent, the conditions independent replicates of the experiment were run at five
faced by prey animals in their natural habitats (i.e., real different times of the day, three in the dark (2100, 0200,
predators and complex backgrounds). and 0600 hours) and two in the light (0800 and 1400
Up until now we have focused on the physical hours). The results, depicted in Figure 13.5, demonstrate
appearance of prey. However, there is more to back- that stoneflies selected the dark brown substrate rather
 Camouflage 279

than the light gray one at 0800, 1400, and 2100 hours; suggest that the choice of dark resting spots by stoneflies
no selection was observed at 0200 or 0600 hours. has been favored by natural selection, at least in part,
Although stonefly nymphs selected the dark over the light because it reduces the risk of being found and eaten by
substrate, this selection ceased approximately two hours visually hunting fish. The breakdown in substrate color
after the lights in the laboratory went off and resumed selection during the hours of darkness links visual preda-
within one hour of their being turned on. tion to the distribution of nymphs. After all, animals need
In the next experiment, Feltmate and Williams (1989) to be cryptic only when they are most vulnerable to pre-
examined whether stoneflies resting on the light substrate dation by visual hunters (Endler 1978). Note, however,
were more vulnerable to predation by rainbow trout. As that we do not know how the visual abilities of stoneflies
before, each stonefly was introduced into its own aquar- change under dark conditions and what role this might
ium. This time, however, the tank contained either light play in the observed breakdown of substrate selection.
or dark tiles (not both, as in the first experiment). A trout The choice of substrate by stoneflies may also conceal
was released into each tank after the nymphs had two them from their own prey, as has been shown for other
hours to adjust to their new surroundings. Twenty-four aquatic insects (Moum and Baker 1990).
hours after releasing the nymphs, the authors recorded the Usually, animals that employ background matching
number of stoneflies consumed in tanks containing either are camouflaged in some habitats but not in others, and
the light or dark substrate. The consumption of nymphs thus their occurrence is often restricted to those par-
by trout was lower in tanks that contained the dark sub- ticular areas where they are best concealed. One way
strate (3 of 24 nymphs eaten) than in tanks that contained some species get around this restriction is by changing
the light substrate (19 of 24 nymphs eaten). These data color as they change backgrounds. The cuttlefish (Sepia
officinalis), which is not a fish at all but a cephalopod
mollusk related to such creatures as squid, octopus, and
1900 0700
Lights on Lights off Lights on nautilus, is perhaps the true master of color change,
known for its swift and dramatic changes in body color
and pattern (Messenger 2001). (Chances are that you
24 have actually seen part of the white, internal shell sup-
port of this animal hanging in a birdcage to help para-
keets and other birds keep their beaks sharp. Called
Number of nymphs on dark substrate

20 cuttlebone, it is not bone at all.) When resting on the

* *
* bottom, Sepia adjusts its color to that of the substrate
at hand. Indeed, within a matter of seconds of settling
16 on a sandy bottom, the dorsal color can change from
gray (Figure 13.6a) to sandy brown, the latter pattern
12
rendering it virtually invisible to predators (Figure
13.6b). These changes in body pattern are considered
examples of coloration matching the visual back-
8 ground. On certain backgrounds, however, the cuttle-
fish displays color patterns with bold contrasting
elements (Figure 13.6c-d). Is this another form of cam-
4
ouflage, and if so, how could it make the cuttlefish less
likely to be detected by visually hunting predators? It
0 turns out that the bold elements of the cuttlefish are
1400 2100 0200 0600 0800 part of disruptive coloration, the form of camouflage to
Time (hours)
which we turn next.
FIGURE 13.5 Substrate selection in stonefly nymphs at
various times of the day (white bars represent data when
lights are on in the laboratory and colored bars when lights STOP AND THINK
are off) when given the option of resting on dark brown or Martin Stevens (2007) strongly suggests that when study-
light gray tiles. During the lights-on period (0800 and 1400 ing the effectiveness of prey camouflage (as well as other
hours) and shortly after the lights go off (2100 hours), a protective markings) efforts be made to consider the visual
larger number of nymphs were observed on dark tiles, a and cognitive abilities of the prey’s predators. Early stud-
background on which they were camouflaged, than on light ies often relied on assessment by humans. Why might it
tiles. Substrate selection was not apparent during the two be unwise to rely solely on human assessment to deter-
remaining dark observations (0200 and 0600 hours). The mine the effectiveness of a particular example of camou-
dashed line represents expected results if no selection flage? What would you want to know about a predator to
occurred; the asterisks indicate selection for dark brown gauge the effectiveness of its prey’s camouflage?
tiles. (From Feltmate and Williams 1989.)
280 Chapter 13 / Antipredator Behavior

a b FIGURE 13.6 Many animals that employ

background matching are restricted to
portions of their habitat in which they are
well concealed. The cuttlefish gets around
this restriction by changing its color to
match the particular background on which
it rests. Shown here is a cuttlefish resting on
(a) a uniformly gray artificial background;
(b) a natural sand background; (c) an
artificial checkerboard background; and
(d) a natural background of dark and light
rocks. The coloration of the cuttlefish can
match the background as in (b) or be
c d disruptive as in (c) and (d).

DISRUPTIVE COLORATION like triangles and the edible body was a dead mealworm
(killed by freezing the night before and then thawed).
Many animals avoid being seen by matching their back-
They then pinned the artificial prey to oak trees in a
ground, but sometimes such matching is not enough forest and monitored “survival” of the prey 2, 4, 6, and
because visually hunting predators may recognize prey 24 hours later. Predation by birds could be distin-
by their body outline. Some animals seem to break up guished from that by slugs and spiders. Whereas birds
their body outline by developing bizarre projections; simply took most or all of the mealworm, slugs left tell-
other species appear to do so with bold contrasting tale slime trails and spiders sucked out the fluids of the
markings. These bold patches on prey are thought to mealworm, leaving the empty exoskeleton. In both
function in preventing, or at least delaying, visual recog- experiments, numerous replicates were run in different
nition of the prey by a predator. How do the patches areas of the forest over a six-month period.
achieve this? Bold patches may catch the eye of a preda- In experiment 1, the markings of the prey (printed
tor, thereby drawing attention away from the outline of patterns on the paper) either overlapped the edges of the
the prey’s body; in addition, patches at the periphery of wings (the “Edge” treatment) or were located toward the
the prey’s body may break up the continuity of the body inside of the prey, away from the edges (the “Inside”
outline (Cott 1940; Merilaita 1998). Coloration designed treatment), or the prey was a single color (Figure 13.7a).
to prevent perception of a prey animal’s form is called There were two different examples in the Inside treat-
disruptive coloration. Although disruptive coloration is ment, Inside 1 (the same markings used in Edge were
often described as a widespread device in animal cam- moved inward so that they did not overlap the edges) and
ouflage, direct experimental evidence for it is quite lim- Inside 2 (other randomly selected markings were placed
ited. Here we describe a study in which scientists inside the prey; this example was included because move-
conducted direct field tests for disruptive coloration, ment of the pattern elements from the periphery to the
using birds as predators and artificial moths as prey. inside for Inside 1 created straight lines that could increase
Innes Cuthill and colleagues (2005) conducted a conspicuousness). In these three bicolored treatments, the
field study designed to test the following two hypothe- markings were black on a dark brown background
ses about disruptive coloration: (1) color patterns at the designed to match the ridge patterns of mature oak trees.
periphery of an animal should provide better conceal- Importantly, because these three prey treatments had pat-
ment than those placed randomly (tested in experiment terns derived from photographs of tree bark, they were
1), and (2) colors of high contrast should provide bet- expected to be equally camouflaged from the standpoint
ter concealment than those of low contrast (tested in of background matching. Thus, only disruptive coloration
experiment 2). The researchers constructed artificial would predict that the treatment Edge would survive bet-
moth-like prey in which the paper wings were shaped ter than the two Inside treatments. The researchers also
 Camouflage 281

a form of camouflage (Figure 13.7b). As you can see, prey

in the Edge treatment survived better than those in the
Edge Inside 1 Inside 2 Black Brown two Inside treatments, and prey in the three bicolored
treatments, in turn, survived better than those in the two
monochrome treatments.
b 1.0 In experiment 2, artificial moths were designed to test
Edge
whether high-contrast colors provide better concealment
Probability of survival

0.8 than low-contrast colors. Cuthill and colleagues used edge,

Inside 2
inside, and monochrome patterns as in experiment 1, but
0.6 Inside 1 this time included high- and low-contrast prey in each
type. Thus, the following six treatments were run: (1)
Black
0.4 Edge, high contrast; (2) Edge, low contrast; (3) Inside, high
Brown contrast; (4) Inside, low contrast; (5) Average color of the
0.2 high-contrast color pair; and (6) Average color of the low-
contrast color pair. As predicted only by the hypothesis of
0 200 400 600 800 1,000 1,200 1,400 disruptive coloration, prey in the Edge, high-contrast
Time (min) treatment survived best (Figure 13.8). Taken together, the
results from experiments 1 and 2 indicate that disruptive
FIGURE 13.7 Disruptive coloration. (a) Artificial moth- coloration is an effective camouflage device against birds,
like prey designed to test whether color patterns at the
above and beyond that of background matching.
periphery provide better concealment than either patterns
at the interior or a single color. (b) Survival over time of
What, then, is the relationship between disruptive
artificial moth-like prey pinned to oak trees in a forest. coloration and coloration matching the visual back-
Patterns at the edge of the body increase survival of ground, and how might these forms of camouflage inter-
artificial prey. (From Cuthill et al. 2005.) act with other aspects of the lives of prey animals? John
Endler (2006) suggests that there may be a three-way
relationship (or trade-off) between the two types of col-
include prey with wings that were monochrome brown or oration and habitat specialization. Species that rely
monochrome black; according to background matching, purely on background matching may successfully evade
the prey in these two treatments (“Black” or “Brown”) detection by predators as long as they are living in a
should be less camouflaged than those in the three treat- particular habitat; such species may thus be habitat spe-
ments with bicolored prey (Edge, Inside 1, and Inside 2). cialists. Disruptive coloration, in comparison, works on
The results support disruptive coloration as an effective a greater variety of visual backgrounds and may be a

Edge, high Inside, high Average, high

contrast (EH) contrast (IH) contrast (AH)

Edge, low Inside, low Average, low

contrast (EL) contrast (IL) contrast (AL)

b 1.0
EH
0.9
Probability of survival

EL
0.8 IH

0.7 IL
FIGURE 13.8 Disruptive coloration. (a) Artificial
AH moth-like prey designed to test whether high-
0.6
AL contrast color patterns provide better conceal-
0.5 ment than low-contrast patterns. (b) Survival
over time of the artificial moth-like prey pinned
0.4 to oak trees in a forest. High-contrast disruptive
0 200 400 600 800 1,000 1,200 1,400 patterns increase survival of artificial prey. (From
Time (min) Cuthill et al. 2005.)
282 Chapter 13 / Antipredator Behavior

strategy employed by habitat generalists. The existence viewed from below by a predator. Similarly, a dark back
of a trade-off between the different forms of camouflage would help an animal match the background of dark,
and degree of habitat specialization warrants further deep waters when viewed from above by a predator.
investigation (and we urge you to do so!). Although many animals display countershading,
direct evidence that this pattern achieves camouflage
COUNTERSHADING through self-shadow concealment is embarrassingly
meager (Ruxton et al. 2004). Indeed, the alternative
Many animals have dark backs and light bellies, a pattern mechanism of background matching cannot be ruled out
called countershading. Over a century ago, the painter in most purported cases. Furthermore, for many exam-
and naturalist A. H. Thayer (1896) suggested that coun- ples of countershading, we cannot even discount the pos-
tershading makes animals difficult to detect because it sibility that the combination of dark backs and light
allows them to obscure their own shadows. As the argu- bellies is completely unrelated to camouflage and instead
ment goes, because light normally comes from above, the functions in thermoregulation or protection from ultra-
ventral surface of the body is typically in shadow and violet radiation (Kiltie 1988). We will consider a study
predators could cue in on darkened bellies (Figure 13.9). that examined the possible functions of countershading
Thayer suggested that by being darker dorsally and paler in the naked mole rat, a small mammal with a big repu-
ventrally, animals could obscure the ventral shadow. This tation for its highly social lifestyle. The study is impor-
mechanism by which countershading is proposed to tant because it examines multiple potential functions of
achieve camouflage is called self-shadow concealment the countershading pattern.
(Kiltie 1988; Ruxton et al. 2004). An alternative mecha- Naked mole rats (Heterocephalus glaber) are fossorial,
nism through which countershading could make animals spending most of their time in extensive underground
difficult to detect is one that we have already discussed, burrow systems in the dry areas of East Africa. They are
background matching. Background matching might be also eusocial, living in large colonies that contain a sin-
particularly common in aquatic animals because they are gle breeding female (the queen), a few breeding males,
likely to be viewed from above and from below by preda- and numerous nonbreeding workers that care for the
tors. With light coming from above, a light belly would queen’s offspring, maintain the burrow system, and feed
help an animal match the bright background when and defend the colony. (Eusociality is discussed in more
detail in Chapter 19.) Stanton Braude and colleagues
a (2001) have studied the behavior of naked mole rats in
captive colonies at the University of Michigan; they have
also monitored the demography and behavior of wild
colonies in East Africa for many years. Their early, anec-
dotal observations on the color of field and laboratory
animals suggested that most colony members display
b countershading, having a darker dorsal coloration

FIGURE 13.9 The effects of countershading on conspic-

uousness. (a) Light normally comes from above, and
under these circumstances fish that are uniformly col-
ored have a conspicuous ventral outline. (b) A counter-
shaded fish is darker dorsally than ventrally (as shown
here, illuminated from all sides). (c) Thus, its body
outline is obscured when light comes only from above.
This mechanism by which countershading is proposed FIGURE 13.10 Although most members of naked mole
to achieve camouflage is called self-shadow conceal- rat colonies exhibit countershading (darker backs than
ment. (After Cott 1940.) bellies), the queen does not.
 Camouflage 283

(purple-brown-gray) than ventral coloration (pink). when searching for the functions of a particular color
They also noticed, however, that queens were interesting pattern and testing alternative hypotheses.
(and all pink) exceptions (Figure 13.10). A few other
colony members also lacked the dark dorsum. On the
basis of these early observations, Braude and co-workers
STOP AND THINK
decided to look more systematically at color in naked mole The study by Braude et al. (2001) is nonmanipulative and
rats. As part of their long-term field study, they began to thus represents an indirect test of the hypothesis that
countershading functions in camouflage. How would you
meticulously record the occurrence of live-trapped naked
design a more direct test of the hypothesis that counter-
mole rats with extreme pink dorsal coloration. They also shading contributes to camouflage in naked mole rats?
began to carefully quantify the color of known age ani- What experimental manipulations would you propose?
mals in their laboratory colonies. They discovered that Would your manipulations allow you to distinguish
most naked mole rats are indeed countershaded; excep- between self-shadow concealment and background match-
tions to the countershaded pattern included newborns, ing as the mechanism by which camouflage is achieved?
queens, most breeding males, and very old individuals, all
of whom are uniformly pink. In naked mole rats, coun-
tershading begins to develop a few weeks after birth and
is fully developed by 3 months of age; interestingly, it
TRANSPARENCY
begins to disappear at about 7 years of age. Some animals are camouflaged simply by being transpar-
What is the function of countershading in naked ent. Although no animal is completely transparent,
mole rats, especially given that these animals spend most organisms such as cnidarians (e.g., hydroids and jelly-
of their time in dark underground burrow systems? Can fish), ctenophores (e.g., comb jellies), and the pelagic
the distribution of countershading among colony mem- (open ocean), larval stages of many fish achieve near
bers or the timing of its development and its loss tell us transparency by such means as high water content of tis-
anything about its function(s)? Braude et al. (2001) used sues, small size, and reduced number of light-absorbing
the field and laboratory data they had collected to test molecules or pigments (McFall-Ngai 1990).
five hypotheses for the existence of countershading in Frequently neglected in discussions of camouflage
naked mole rats. The first four hypotheses are not mechanisms, transparency is probably the dominant
mutually exclusive; they state that countershading (1) form of camouflage in aquatic environments, particularly
protects individuals from ultraviolet light, (2) facilitates in pelagic habitats where organisms have no surfaces to
thermoregulation, (3) protects against abrasion because match or places to hide (Johnsen 2001). Transparency is
the pigment melanin may strengthen skin, and (4) pro- extremely rare in terrestrial habitats for at least two rea-
vides camouflage for individuals dispersing above sons. The first reason concerns refractive indexes (the
ground. These hypotheses can be contrasted with the angle at which light bends when passing from one
fifth and final hypothesis that countershading has no medium into another) of water and air. Let’s first con-
current function in naked mole rats and is simply a rem- sider the situation for aquatic organisms, and then we’ll
nant of adaptive countershading in their surface- look at that of terrestrial ones. Because animals’ bodies
dwelling ancestors. We will not review all of the are largely water, when light travels from the surround-
evidence for or against each hypothesis here; we will ing water into the tissues of an aquatic animal, the angle
simply say that Braude and co-workers concluded that of light is virtually unchanged; in the absence of light-
their data were most consistent with the camouflage scattering or light-absorbing elements, the animal
hypothesis. Known dispersers at their field site were, on appears to be transparent (light, then, is basically pass-
average, 2.3 years of age, with the oldest recorded dis- ing from water into water). In contrast, in a terrestrial
perser being 5 years old. Thus, the age of dispersal coin- environment, light must pass from air into the water-
cides with the time when naked mole rats have filled tissues of an animal. The difference in the refrac-
well-developed countershading. Other observations at tive indexes of air and the terrestrial animal’s tissues
their field site indicate that naked mole rats tend to dis- creates an obvious body outline, greatly diminishing
perse on the surface at night rather than during the day, transparency. The second reason transparency is rarely
and often on moonlit nights. Thus, a darkened dorsum used as a camouflaging mechanism by terrestrial animals
could provide protection against nocturnal avian preda- has to do with the deleterious effects of ultraviolet radi-
tors (and would not be needed at night for protection ation on land. Animals on land need protective pigments,
against ultraviolet radiation). Although this study does making transparency difficult, if not impossible. In
not distinguish between self-shadow concealment and aquatic habitats, much of the ultraviolet radiation is fil-
background matching as possible mechanisms through tered out within a few meters of the water’s surface, and
which countershading might achieve its camouflage thus animals living beyond this distance are not subject
function, it is a good example of keeping an open mind to the same radiation damage as terrestrial organisms.
284 Chapter 13 / Antipredator Behavior

MASQUERADE ous scientist), the catfish, with no discernible move-

ment of its fins, drifts slowly downstream like a water-
As mentioned at the beginning of our discussion of cam-
logged leaf.
ouflage, masquerade differs from the other forms
because the prey may be detected but deemed inedible
by predators searching for animals to eat (Endler 1981, OTHER FUNCTIONS OF COLOR
2006). Leaf resemblance is a particularly common dis-
guise. Among insects that resemble leaves, we find green Evasion of predators is not the only function of color
or brown coloration, leaf-like patterns of venation on pattern in animals. Color affects heat balance and thus
their bodies, and flattened shapes. These morphological plays a role in thermoregulation. Color and pattern are
specializations are often accompanied by behavioral also important in many aspects of communication,
ones, ranging from remaining still during daylight hours including mate recognition, courtship, male–male com-
to swaying like a leaf in the wind. petition, and territorial defense.
Leaf resemblance also occurs in some small verte- As mentioned at the start of this chapter, the vari-
brates, including amphibians that normally inhabit the ous functions of animal color and pattern may act in
leaf litter of the forest floor (Figure 13.11a). concert or in opposition. Let us consider a case in
Resemblance to dead leaves, however, is not restricted which they act in opposition. If color and pattern are
to terrestrial species. Indeed, Tetranematichthys wallacei, adjusted for thermoregulation, how can animals com-
a nocturnal catfish that inhabits small Amazonian municate effectively with mates and competitors and at
streams, displays a remarkable resemblance to dead the same time be inconspicuous to visually hunting
leaves, and its disguise includes aspects of its physical predators? Although some animals that employ back-
appearance and its behavior (Sazima et al. 2006). The ground matching have evolved alternative means of
body and fins of T. wallacei are the color of dead leaves, exchanging information (e.g., relying on auditory or
and the body is laterally compressed or flattened, giv- olfactory signals to communicate with conspecifics),
ing it a leaf-like shape. A nocturnal forager, the catfish many still rely on visual cues. As we will see, the color
spends daylight hours lying on its side amid dead leaves pattern displayed by a particular animal may be a com-
on the stream bottom (Figure 13.11b). If disturbed promise between factors that favor camouflage and
from its resting place by a potential predator (or a curi- those that favor conspicuousness.
John Endler’s (1978) work with wild populations of
guppies (Poecilia reticulata) in northeastern Venezuela and
Trinidad provides an excellent example of how color pat-
a
terns may represent a balance between mate acquisition
and camouflage. Whereas a female’s choice of mate and
competition among males favors brighter colors and
more visible patterns in guppies, selection by diurnal
visual predators (at least six species of fish and one fresh-
water prawn) favors less colorful and less conspicuous
patterns. It is interesting that as predation risk increases
across communities, the colors and patterns of guppies
become less obvious because of (1) shifts to less conspic-
uous colors, (2) reductions in the number of spots, (3)
reductions in the size of spots, and (4) slight reductions
in the diversity of colors and patterns (Figure 13.12). In
areas in which guppies encounter low-predation pres-
b
sure, however, the balance shifts toward attracting mates,
and colors and patterns become more conspicuous.

FIGURE 13.11 Some prey masquerade as objects that

appear inedible to predators searching for another
animal to eat. Here, (a) frogs from Malaysia and (b) a
catfish from Amazonian streams resemble dead leaves.
 Polymorphism 285

consume remaining individuals (a search image is the

heightened ability to detect a prey species; see Chapter 12).
If individuals of the prey species are widely spaced, however,
predators will rarely encounter them and will soon forget
the search image. Indeed, individuals of many species that
employ background matching occur at widely spaced loca-
a 4 11 tions throughout their environment.
All spots 10 Other species get around the problem of search
images by occurring in several different shapes and/or
Number of spots per fish

9
3 Red color forms, that is, by exhibiting polymorphism. We
8
will consider an example of color polymorphism that
7
Black Iridescent concerns fox squirrels (Sciurus niger) in the eastern
2
United States. Fox squirrels have been described as the
most variable in color of all mammals in North America
Blue (e.g., Cahalane 1961). Color varies both among and
1
Green within populations. Dorsal coloration may range from
Yellow gray or tan to black, and coloration on the head and ear
0 region is often distinctive (Figure 13.13). Even within a
Risk of predation single litter, both melanistic (black) and nonmelanistic
young can be found. Intrigued by the variation in coat
Length of color patch (mm)

b color of fox squirrels, Richard Kiltie (1989) examined

2.0
All spots
close to 2000 museum specimens of this species. He
1.5
determined the percentage of dorsal black for each skin
1.5 and compiled information on the occurrence of wildfires
Red Iridescent 1.0 in the eastern United States. Taken together, his data on
Blue coat color and fires show that the incidence of melanis-
1.0 Black tic individuals is correlated with the frequency of wild-
fires over the total range of the fox squirrel. Both
Risk of predation
wildfires and melanistic squirrels are more common in
c the southeastern United States (Figure 13.14).
Percentage of fish with

In fox squirrels, the melanistic polymorphism in coat

0.6 colors may thus be maintained by the periodic blacken-
unique patterns

ing of the ground and lower portions of tree trunks by

0.4 wildfires. One would imagine that dark squirrels are less
conspicuous to hawks than are light or variably colored
0.2 individuals against a blackened background. However, the
advantage does not remain with the black squirrels for
0.0
Risk of predation

FIGURE 13.12 An animal’s color is often a compromise

solution to the problem of selection for conspicuousness
in courtship displays and selection for inconspicuousness
to visually hunting predators. Changes in the color and
pattern of guppies as a function of predation pressure
reflect the fine balance between these two selective
forces. As predation risk increases, the (a) number of
spots, (b) length of color patches, and (c) diversity of
patterns decrease. (Modified from Endler 1978.)

POLYMORPHISM
Like most things, background matching is not foolproof. FIGURE 13.13 Some color morphs of the fox squirrel.
Although individuals may blend with their background, Note the variation in the percentages of dorsal black
predators in a given area may develop a search image for and the pattern around the head and ear region.
that particular species and systematically search out and (After Kiltie 1989.)
286 Chapter 13 / Antipredator Behavior

a long. As rainfall and new plant growth convert a charred

area into a less uniformly black substrate, fox squirrels
with variable amounts and patterns of black dorsal col-
oration would be more difficult to see than uniformly
black individuals against the patches of light and dark
underground. Finally, when the period of regrowth of the
pine and oak forest is almost complete, the advantage may
shift to squirrels that are uniformly light in coloration.
All specimens
Thus, variable coat color in fox squirrels may result from
b the alternating superiority of light and dark individuals in
matching the background of an environment that period-
ically burns and regenerates.
In some cases, polymorphic species do not have col-
oration that matches the visual background and rely
solely on their diverse appearance to evade detection by
predators. Whether camouflaged or not, by being differ-
100% dorsal black ent, individuals of prey species can occur at higher den-
c sities without suffering increased mortality from
predators searching for individuals with a specific appear-
ance. Some species that occur at very high densities
exhibit extreme polymorphism, making it almost impos-
sible to find two individuals that look alike (Figure 13.15).
Gairdner Moment (1962) described the phenome-
non in which members of a population look as little like
one another as possible. In such populations the proba-
1–99% dorsal black bility of an individual’s having a certain appearance is
d 0 inversely related to the number of other individuals in
7 the population that have that appearance. If one morph
1 5 in a polymorphic population is much more common
7
1
4 than another morph, predators are likely to develop a
28 37 13
2 3
29
search image for the more common, rather than the rare,
1 28 18
35
60 47
54 32 129
82 63 87
25 81

58

FIGURE 13.14 In the fox squirrel, the incidence of

melanism is correlated with the frequency of wildfires.
(a) Most of the counties in the eastern United States
from which Kiltie (1989) examined museum specimens
of fox squirrels. (b) Counties from which specimens
with 100% dorsal black were noted. (c) Counties from
which specimens with intermediate levels of dorsal
black (1–99%) were recorded. (d) Average number of
wildfires per state in protected forestlands during the
years 1978–1982; values have been normalized to take
into account the area of land under wildfire surveillance
(dashed lines depict limits of the fox squirrel’s range).
Note that melanistic fox squirrels occur primarily (b) in
the southern portions of the species range and that
individuals with intermediate levels of dark coloration
are limited to (c) the eastern Gulf and Atlantic coastal FIGURE 13.15 Although all the same species, these four
plains. These areas in the southeastern United States brittle stars are dramatically different in appearance,
are also the areas in which wildfires are most common thereby inhibiting the formation of search images in
(d). (From Kiltie 1989.) predators. (Drawn from photograph in Moment 1962.)
 Warning Coloration 287

morph. The end result is that predators take more of the noxious qualities, and through this warning discourage
common form relative to its frequency in the population. an attack. For example, the dramatic black and white
Thus, for example, when two morphs are equally cam- markings of spotted and striped skunks may serve, in
ouflaged and are exposed to predators that use search part, to warn predators of the foul-smelling repellent
images when hunting, the rare morph will have a selec- that may, upon further harassment, be released from the
tive advantage over the common morph. We described skunks’ anal scent glands (Figure 13.16). Many insects,
this form of frequency-dependent selection in Chapter 4. such as the social wasps, have a boldly patterned yellow
This form of selection has been called apostatic selection and black body thought to warn of their painful sting.
(Clarke 1969). Its strength varies as a function of factors The phenomenon by which a conspicuous appearance
such as density, palatability, and conspicuousness of prey (often coloration) serves to advertise dangerous or
(Allen 1988). Furthermore, Jeremy Greenwood (1984) unpleasant attributes is called aposematism; sometimes,
indicates that predators need not hunt by search image it is simply called warning coloration. Although it is not
to cause apostatic selection in prey. Some predators, for difficult to find apparent examples of aposematism in
example, may simply have an aversion to prey that are nature, by now you know that we are also interested in
rare or unfamiliar to them. direct tests of presumed antipredator devices. Here we
What experimental evidence do we have that being consider some of the direct evidence for warning
different pays off? Croze (1970), working on a sandy coloration in dendrobatid frogs.
peninsula in England, placed 27 painted mussel shells Frogs within the family Dendrobatidae are best
with pieces of meat under them on the ground and known for their bright coloration and toxic skin secre-
exposed them to predation by carrion crows (Corvus tions, although there is substantial variation among and
corone). In some of the 14 trials, the shells were within species in both of these characteristics. Some
monomorphic (i.e., all the same color), whereas in oth- species are red, yellow, blue, or some combination of
ers they were trimorphic (9 red, 9 yellow, and 9 black). these colors, and the colors may contrast with black
The results, summarized in Table 13.1, show that the markings. Perhaps the most notorious of the toxic species
crows took fewer of the trimorphic than the monomor- is Phyllobates terribilis, a single individual of which has
phic prey. The percentage of survival for each of the enough toxin in its skin to kill about 20,000 mice or, in
three morphs in a trimorphic population was two to more familiar currency, 100 humans. The Choco Indians
three times higher than in monomorphic populations. of western Colombia make deadly weapons by simply
Thus, a morph had a twofold to threefold selective wiping their blowgun darts across the back of one of these
advantage when occurring as part of a trimorphic pop- metallic yellow frogs; so lethal is the poison in the frog’s
ulation. Croze’s results demonstrate that when prey skin that a dart poisoned in this manner can remain
populations occur at the same density, individuals in deadly for more than a year. The bright coloration
polymorphic populations experience less predation than of dendrobatid frogs has been widely viewed as an exam-
those in monomorphic populations. ple of aposematic coloration, a warning to potential

TABLE 13.1 Percent Survival of Painted Mussel

Shells in Either Monomorphic or Trimorphic
Populations When Exposed to Predation by
Carrion Crows
Type of population

Shell color Monomorphic Trimorphic

Yellow 10 31
Black 12 40
Red 19 45

Source: Data from Croze (1970).

WARNING COLORATION
Many animals that have dangerous or unpleasant attrib- FIGURE 13.16 The bold black and white patterns of
utes appear to advertise this fact with bright colors and skunks are thought to be aposematic, warning potential
contrasting patterns. Bold markings, typically in black, predators of the foul liquid that may be sprayed from
white, red, or yellow, warn the predator of the prey’s glands beneath the tail.
288 Chapter 13 / Antipredator Behavior

predators of the poisons in their skin. What evidence do How can we reconcile the divergent findings of the
we have that their coloration serves an aposematic func- two studies? Although without further studies we can’t be
tion? Kyle Summers and Mark Clough (2001) reasoned sure of the precise role of methodological differences in
that if the coloration of dendrobatid frogs is indeed yielding the different findings, there is at least one other
aposematic, then we would predict that the more toxic possibility. A recently developed theoretical model of the
species will have brighter, more extensive coloration than costs and benefits of defensive traits suggests that the
the less toxic or nontoxic species. They set out to test the relationship between conspicuousness and toxicity will
hypothesis that color will evolve in tandem with toxicity vary under different ecological conditions, being posi-
in dendrobatid frogs using a phylogenetically controlled tively correlated under some conditions and even nega-
comparative analysis. (This type of analysis controls for tively correlated under others (Speed and Ruxton 2007).
associations between toxicity and coloration caused by A negative relationship between conspicuousness and
shared ancestry. See Chapter 4 for discussion of the com- toxicity might occur if the costs of conspicuousness
parative approach). Summers and Clough obtained tox- increase to the point that prey are better off decreasing
icity data for 21 species from the literature, and using their investment in bright colors and increasing their
information on toxin diversity, amount, and lethality, investment in toxins to compensate. The studies of apose-
determined a total toxicity score for each species. They matism in dendrobatid frogs thus highlight the diversity
assessed the brightness and extent of coloration of each that can exist even within a particular form of defense.
of the 21 species using two methods. In the first method, Animals with warning coloration often enhance their
human observers were presented with color photographs conspicuousness behaviorally. Many are active during the
of each species and asked to rank them. In the second, daytime, and individuals of some species form dense,
color photographs of each species were scanned into a obvious aggregations. Although rare forms in aposematic
computer where a program measured the brightness of animals are typically selected against (predators will not
colors and the proportion of the frog covered by each be as familiar with the rare form as they are with the com-
color. The same measurements were made on scanned mon form and may attack), they are at less of a disadvan-
color photographs of leaf litter to produce an overall tage when they occur in clusters (Greenwood et al. 1989).
measure of contrast of each species of frog against a leaf Thus, dense aggregations of aposematic prey not only
litter background. Summers and Clough found a signif- emphasize the warning but also function as areas in which
icant association between toxicity and coloration rare forms may arise and survive.
(whether assessed by humans or the computer): the more The response of predators to aposematic coloration
toxic species were the most colorful. This finding is con- may be learned or innate. In the first case, predators sam-
sistent with the hypothesis that bright coloration in den- ple some of the prey, discover their unpleasantness, and
drobatids serves an aposematic function. learn to avoid animals of similar appearance when
Additional data on the color and toxicity of dendro- searching for subsequent meals. For example, garter
batid frogs have been collected in the years since the snakes (Thamnophis radix) develop a much stronger
study by Summers and Clough (2001). What have these avoidance of conspicuously colored noxious prey than of
data shown? Results from one study conducted by nonaposematic prey, even though olfaction plays an
Catherine Darst, Molly Cummings, and David important role in detection and ingestion of prey. Two
Cannatella (2006) indicate that degree of toxicity and different types of prey, earthworms and fish, were offered
brightness can be decoupled in some dendrobatid frogs. to garter snakes on forceps that were aposematically col-
As part of a larger study, these researchers evaluated the ored (yellow and black) or nonaposematically colored
conspicuousness and toxicity of three closely related (green). After they had consumed fish that was presented
species within the genus Epipedobates. Their methods on forceps of either type of coloration, snakes in the
differed in several ways from those of Summers and experimental group were injected with lithium chloride
Clough (2001). For example, they assessed toxicity by to induce illness. Immediately following the induced ill-
monitoring the recovery times of mice injected with ness, the garter snakes avoided all fish, regardless of the
minute quantities of extracts from the skin of the three color of the forceps. However, the garter snakes that had
frog species (the quantities were chosen so as not to kill been offered fish on aposematically colored forceps had
the mice but to cause mild irritation). Despite method- a much longer-lasting aversion to fish than those who
ological differences, we might still expect to see a pat- had been offered fish on green forceps (Figure 13.17).
tern similar to that found by Summers and Clough Seven days after the induced illness, only one of five
(2001), that is, that toxicity and conspicuousness are pos- snakes in the aposematic treatment group ingested any
itively associated in the three frog species studied. fish during the 120-second test interval. In contrast, all
Instead, Darst and colleagues found that the most toxic five of the snakes in the nonaposematic treatment group
species (E. parvulus) was not the most conspicuous, and eventually attacked fish. Thus we see that predators learn
the most conspicuous species (E. bilinguis) was only mod- to avoid unpalatable prey more readily if the prey are
erately toxic. conspicuously colored (Terrick et al. 1995).
 Batesian Mimicry 289

a 140 advantage in survival and reproduction, it will continue

Attack latency (sec) 120 Aposematic in the population.
Control
100
Non-
80 aposematic
60
BATESIAN MIMICRY
40 Batesian mimicry is named after the nineteenth-century
20 English naturalist Henry Walter Bates, and it refers to a
0 palatable species that has adopted the warning character-
istics of a noxious or harmful species. The harmless species
30 is called the mimic and the noxious one, the model. By
b
resembling a noxious species, the mimic gains protection
Attack latency (sec)

from predators. The evolution of mimicry has been the

20
center of lively discussion (e.g., Holmgren and Enquist
1999; Joron and Mallet 1998; Mappes and Alatalo 1997).
10 The degree of protection experienced by the mimic
varies as a function of numerous factors. The ratio of
models to mimic is important, for example. The mimic
0 does better when it is rare and therefore less likely to be
Pre- Day 7 Day 12 Day 17 Day 22
treatment detected by the predator than the noxious model (Turner
1977). And the more distasteful the model, the better the
FIGURE 13.17 Predators learn more quickly to avoid mimic fares. These predictions have been experimentally
distasteful prey that are conspicuous. Here, garter tested, using great tits (Parus major) as the predators. The
snakes were first offered pieces of fish on aposematic birds were offered model and mimic prey one at a time,
(yellow and black) or on nonaposematic (green) forceps. as prey would be encountered in nature. Both prey types
The snakes were then made ill by an injection of
survived better when there were fewer mimics. The mod-
lithium chloride. The post-treatment attack latencies to
els survived significantly better the more distasteful they
(a) pieces of fish or (b) earthworms indicate that snakes
in the aposematic treatment group had a stronger were, and the degree of unpalatability also affected the
aversion to fish than did snakes in the nonaposematic survival of the mimics (Lindstrom et al. 1997). The mem-
treatment group. (Data from Terrick et al. 1995.) ory of predators, availability of alternate prey, and
whether mimics and models are encountered simultane-
ously or separately may also play a role (Speed and Turner
1999). As a result, a mimic may gain most if its habits and
Sometimes two warningly colored species look alike. daily activity overlap those of its model species.
Apparently, two noxious species can benefit from a This increase in benefits was demonstrated by using
shared pattern because predators consume fewer of each naive birds, brown-eared bulbuls (Hypsipetes amaurotis
species in the process of learning to avoid all animals of pryeri), that were trained to take food from two feeders in
that general appearance. This phenomenon is called captivity. The model prey was Pachliopta aristolochiae, a
Müllerian mimicry. butterfly that sequesters alkaloids as a larva and is, there-
Although some predators learn through memorable fore, distasteful. A noxious model butterfly was placed at
experiences to avoid aposematic prey, others display one of the feeders. After an unpleasant experience eating
innate avoidance. An innate response to warning col- the model, the bird took less of the palatable food from
oration might be favored over a learned response when the feeder. This suggests that the bird associated the
the secondary defense of the prey has the potential of unpleasantness not just with the model but also with the
being fatal to the predator. Learning at the moment of place where it was experienced. The birds were then
death is of little value. offered female swallowtail butterflies (Papilio polytes),
Sometimes, like advice, warning coloration is which come in mimic and nonmimic forms. The birds
ignored. A predator that is starving might tackle a nox- avoided the mimetic forms of swallowtails (Uesugi 1996).
ious prey that it would normally pass up during better Although in some instances the resemblance
times. Wolves will attack skunks and porcupines when between model and mimic seems almost exact, the like-
other prey is scarce. In addition, some predators are spe- ness usually does not have to be perfect because preda-
cialists and are able to eat certain aposematic animals, tors appear to generalize conspicuous features of noxious
or the least noxious parts of them, as we saw for the prey. In some cases, poor mimics may exist because they
black-headed grosbeak and black-backed oriole that exploit some constraint in the predator’s visual or learn-
were preying on the unpalatable monarch butterfly. ing mechanisms. An example is provided by hoverflies,
However, as long as an antipredator device confers a net which mimic certain wasps. Studies have shown that
290 Chapter 13 / Antipredator Behavior

pigeons rank hoverflies according to their similarity to to be unoccupied. In the experiments, blennies were
the wasp model. To human eyes, the two most common housed alone in an aquarium and were presented with
types of hoverflies show the least resemblance to wasps, juvenile snails on either an algal frond with polychaete
yet the pigeons rank them as being very similar to wasps. tubes or an algal frond without tubes. At each presenta-
It is thought that these wasps have some key feature that tion, three juvenile snails (one white-spiral, one yellow,
is used by pigeons in pattern recognition (Dittrich et al. and one brown) were randomly positioned on the frond
1993). This is a good reminder that when studying the and the frond was lowered to the bottom of the tank.
defenses of prey animals, it is important to consider the Once blennies detected a snail, they plucked it off the
cognitive and perceptual abilities of their predators. frond and swallowed the shell whole. Reimchen
Numerous mimetic resemblances are recounted in recorded the first snail taken at each trial. Overall, white-
the literature, but only a few studies demonstrate that the spiral snails suffered the lowest number of attacks, and
purported mimics actually gain protection from their attacks were less common on fronds with polychaete
natural enemies. T. E. Reimchen (1989) first described tubes (9.4%) than on fronds devoid of tubes (22.9%).
a system of Batesian mimicry involving the juvenile stage Thus, in this unusual system of snail-polychaete mim-
of a snail (the mimic) and the tubes of a polychaete worm icry, resemblance to the model does appear to confer a
(the model) and then provided evidence that the resem- protective advantage to the mimic.
blance actually conferred some degree of protection to
the young snails. The snail, Littorina mariae, lives in the
intertidal zone of the North Atlantic. The shells of some DIVERTING COLORATION,
juveniles have a conspicuous white spiral, and the shells STRUCTURES, AND BEHAVIOR
of others are yellow or brown. When adult, the snails are
either yellow or brown, and the white spiral possessed Many animals have evolved colors, structures, and pat-
by some is visible only as a white apex on the shell. Egg terns of behavior that seem to divert a predator’s atten-
masses of the snail are deposited directly on algal fronds, tion, while they, and in some cases their offspring, escape
and once the juveniles hatch they disperse on the fronds. with little or no damage. Whereas many of the
Snails with white-spiral coloration were observed only antipredator devices discussed so far help prey to avoid
in habitats where the polychaete Spirobis was present. In an encounter with a predator, distraction devices come
these habitats, white-spiral phase juveniles are virtually into play once a prey animal has been discovered or
indistinguishable from the tubes of Spirobis that are when discovery seems all too imminent. We next con-
cemented to the fronds (Figure 13.18). sider some of the devices of prey animals that operate
Reimchen collected the intertidal fish Blennius pho- during an encounter with a predator.
lis, an important predator on juvenile snails, and con-
ducted predation experiments in aquaria in the
FALSE HEADS
laboratory. Although the polychaete tubes are not nox-
ious to the fish, they represent a substantial investment Many predators direct their initial attack at the head of
in time and energy because they are difficult to remove the prey. Some prey species have taken advantage of this
from the substrate, and once removed, they may prove tendency by evolving false heads that are located at their
posterior end, a safe distance from their true heads.
Lycaenid butterflies (Lepidoptera: Lycaenidae) display
patterns of color, structure, and behavior that are con-
sistent with deflecting predator attacks toward a false
head (e.g., Robbins 1981). Individuals of the species
Thecla togarna, for example, have a false head, complete
with dummy antennae, at the tips of their hind wings
(Figure 13.19). These butterflies enhance the structural
illusion of a head at their hind end by performing two
rather convincing behavioral displays. First, upon land-
ing, the butterfly jerks its hind wings, thereby moving
the dummy antennae up and down while keeping the
true antennae motionless. The second ploy occurs at the
instant of landing, when the butterfly quickly turns so
FIGURE 13.18 The white phase juvenile of a snail
that its false head points in the direction of previous
(shown by the arrow) is a Batesian mimic of the tubes flight. An approaching predator is thus confronted with
of certain polychaetes. By resembling the tubes, young a prey that flutters off in the direction opposite to that
snails may gain protection from fish that are searching expected. Experimental tests have demonstrated that
for food on algal fronds. markings associated with false heads misdirect the
 Diverting Coloration, Structures, and Behavior 291

postautotomy tail movement varies among species, in

some lizards the tail may thrash for as long as five min-
utes. The effectiveness of tail autotomy is underscored
by the presence of tails in the stomachs of predators, as
well as the occurrence of tailless lizards and lizards with
regenerated tails in natural populations.
Direct experimental evidence for the importance of
tail autotomy as an antipredator device comes from a lab-
oratory study by Benjamin Dial and Lloyd Fitzpatrick
(1983). These researchers tested the effectiveness of tail
autotomy and postautotomy tail movements in permitting
the escape of lizards from mammalian and snake preda-
tors. In the first study, staged encounters were conducted
between a feral cat and two species of lizards, Scincella
lateralis (a species with vigorous postautotomy tail thrash-
ing) and Anolis carolinensis (a species with less vigorous
thrashing). Dial and Fitzpatrick recorded the cat’s reac-
tion to lizards of both species under two conditions: (1)
FIGURE 13.19 The false head of a butterfly. Note the
thrashing tail trials—lizards and their autotomized tails
pattern of markings that tends to focus attention on the
posterior end of the butterfly and the dummy antennae were placed in front of the cat immediately after auto-
and eyes at the tips of the hind wings. Markings and tomy, and (2) exhausted tail trials—tails were allowed to
structure combine with behavior (e.g., movement of thrash to exhaustion and then lizards and their auto-
dummy, rather than true, antennae) to divert the tomized tails were placed in front of the cat. In both types
attention of a predator away from the true head. of trials, autotomy was induced by the experimenters.
(Redrawn from Wickler 1968.) They gripped the lizards’ tails at the caudal fracture plane
with forceps (in many species of lizards, tail breakage takes
place at preformed areas of weakness). The results, sum-
attacks of avian predators and, in particular, increase the marized in Table 13.2, show that the dramatic postauto-
possibility of escape if the prey is caught to begin with tomy tail thrashing of S. lateralis is an effective escape
(Wourms and Wasserman 1985). tactic, whereas the more subdued tail movement of
A. carolinensis is not. Note that in all of the thrashing-tail
trials with S. lateralis, the cat attacked the tail, rather than
AUTOTOMY
the lizard, and in all cases the lizard escaped. In 100%
Rather than simply diverting a predator’s attack toward of the exhausted-tail trials with this species, however,
a nonvital portion of the anatomy, some prey actually the cat attacked and captured the lizards. The results for
hand over a “disposable” body part to their attacker, A. carolinensis were quite different: the cat attacked the
almost as a consolation prize. Autotomy, the ability to lizards and ignored the tails in all trials.
break off a body part when attacked, has evolved as a
defense mechanism in both vertebrates and inverte-
brates. Tail autotomy in lizards, for example, is com- TABLE 13.2 Responses of a Feral Cat to
monly reported, as well as in some salamanders, a few Simultaneous Presentation of Autotomized Tails
snakes, and even some rodents. A more dramatic auto- (Either Thrashing or Exhausted) and Live Tailless
tomy, however, is seen in sea cucumbers (members of the Bodies of Two Species of Lizards
phylum Echinodermata), which, upon being attacked,
Number of responses
forcefully expel their visceral organs (guts, in the vernac-
ular) through a rupture in the cloacal region or body Attack Attack Escape
wall. The predator may then begin to feed on the sea Tail to tail to body of lizard
cucumber’s offering as it makes its slow escape. In most
Anolis carolinensis
autotomy cases, the disposable body part is subsequently
regenerated. As an example of the phenomenon, we will Exhausted 0 8 3
focus on tail autotomy in lizards. Thrashing 0 6 1
Tail autotomy benefits the lizard in two ways: first, Scincella lateralis
it allows the lizard to break away from its attacker, and, Exhausted 0 6 0
second, if the detached tail continues to thrash and
Thrashing 7 0 7
writhe, the attacker is distracted as the lizard runs away
(Arnold 1988). Although the vigor and duration of Source: Data from Dial and Fitzpatrick (1983).
292 Chapter 13 / Antipredator Behavior

In the second experiment, Dial and Fitzpatrick

(1983) examined whether postautotomy tail movement
influenced the predator’s handling time. The authors
staged encounters between S. lateralis and the snake
Lampropeltis triangulum, again using autotomized tails that
were either thrashing or exhausted. On average, the snakes
required 37 seconds longer to handle thrashing tails than
exhausted tails, providing the tailless lizard with more time
to escape. Thus, for the lizard S. lateralis, postautotomy tail
movement supplements the simple mechanism of break-
ing away from the predator’s grasp and, depending on the
type of predator, may either attract the predator’s atten-
tion (as in the case of the cat) or increase the time required
to handle the autotomized tail (as in the case of the snake).
Either way, postautotomy tail movement enhances the
opportunity for the lizard to escape.
FIGURE 13.20 An opossum playing dead.
Until this point we have focused on the benefits of
tail autotomy without mentioning potential costs.
Depending on the species of lizard, tail loss may lead to Hognose snakes (Heterodon platirhinos) have a com-
reductions in speed, balance, swimming, climbing, or plex repertoire of antipredator mechanisms, and feigning
mating ability, and when the tail is used as a display, it death is one option. These fairly large nonvenomous or
may even lead to declines in social status (Cooper et al. slightly venomous snakes occur in sandy habitats in the
2004; Fox and Rostker 1982; Langkilde et al. 2005). eastern United States. When first disturbed, the hognose
Furthermore, regeneration of the tail must certainly opts for bluffing—it flattens and expands the front third
entail costs in energy and materials. Many lizards, after of its body and head, forming a hood and causing it to
all, have substantial fat deposits in their tails that are also look larger. It then curls into an exaggerated S-coil and
lost with the tail. Finally, unlike many other antipreda- hisses, occasionally making false strikes at its tormentor.
tor devices, once used, autotomy cannot be employed When further provoked, however, it drops the bluff and
again, at least until the tail regenerates. begins to writhe violently and to defecate. Then it rolls
over, belly up, with its mouth open and tongue lolling. If
the predator loses interest in the now immobile “corpse”
FEIGNING INJURY OR DEATH
and moves away, the snake slowly rights itself and crawls
In ground-nesting birds such as the killdeer (Charadrius off. Natural encounters between predators and prey are
vociferus), a parent may feign injury in an elaborate effort rarely witnessed events, so most research on death feign-
to divert the attention of an approaching predator away ing has been conducted under laboratory conditions with
from its nest and young, particularly soon after hatch- staged predator–prey encounters. We will consider two
ing, when offspring are most vulnerable (Brunton 1990). such studies, but keep in mind that there is a general need
Upon spying a predator, an adult may suddenly begin for field observations of death feigning during natural
dragging its wing, as it flutters away from its nest. The predator–prey encounters (Gregory et al. 2007).
predator follows, and as it closes in the killdeer suddenly The complete repertoire of antipredator mecha-
recovers and flies away, giving a loud call. If all goes as nisms occurs in young hognose snakes, and Gordon
planned, the predator will continue to wander off. Burghardt and Harry Greene (1988) have shown that
Some animals rely not on diverting the attention of newborn snakes are capable of making very subtle assess-
a predator but on causing the predator to lose interest. ments of the degree of threat posed by a particular
Because some predators kill only when their prey is mov- predator. The researchers conducted two experiments in
ing, an animal that feigns death may fail to release the which they monitored the recovery from feigning death
predator’s killing behavior, and with any luck the preda- (i.e., crawling away) of newly hatched snakes under var-
tor will lose interest and move along in search of a live- ious conditions. In experiment 1, the recovery of snakes
lier victim. Perhaps the most familiar feigner of death is was monitored in the presence or absence of a stuffed
the opossum, Didelphis virginiana (Figure 13.20). Hence screech owl (Otus asio) mounted on a tripod 1 meter from
the phrase “playing possum” has come to be synonymous the belly-up snake. In experiment 2, the snake recovered
with “playing dead.” Although their performance is less (1) in the presence of a human who was staring at the
well publicized than that of opossums, juvenile caimans snake from a distance of 1 m, (2) in the presence of the
(Caiman crocodilus) react aggressively toward humans when same person in the same location but whose eyes were
approached on land but feign death when handled in averted, and (3) in a control condition in which no
water (Gorzula 1978). The response of an individual to a human being was visible. Both the presence of the owl
particular predator may thus vary as a function of context. (experiment 1) and the direct human gaze (experiment
 Diverting Coloration, Structures, and Behavior 293

2) resulted in longer recovery times than the respective TABLE 13.3 Results from Predation
control conditions (Figure 13.21). When the human Experiments Involving Pygmy Grasshoppers
being averted his or her eyes, the recovery time was (as Prey) and Four Different Predators
intermediate. Thus, young snakes are capable of using
Number of Death feigning Predation
rather subtle cues from predators to make adjustments
Predator attacks occurred successful
in their antipredator behavior.
From our discussion so far, you can see that the Frog 20 17 4
antipredator behavior typically described as death Bird 10 0 2
feigning eventually involves immobility elicited by the Mantis 9 0 8
presence (or grasp) of a predator. Some observers have Spider 20 0 0
noted, however, that the rigid postures assumed by some
supposedly death-feigning animals barely resemble Source: Data from Honma et al. (2006).
those of dead animals. Could immobility in these cases
serve an antipredator function other than mimicking
death? The answer appears to be yes. Atsushi Honma experiments are presented in Table 13.3. When grasped
and colleagues (2006) conducted predation experiments by a frog, a grasshopper immediately assumes a rigid
in which pygmy grasshoppers (Criotettix japonicus) were T-shape (formed by firmly bending its hind legs
exposed to four different potential predators found in downward), which it maintains for several minutes, even
the grasshopper’s habitat: a frog, a bird, a praying after release by the frogs, most of which could not swal-
mantis, and a wolf spider. The results of the predation low the immobile grasshopper. During encounters with
other potential predators, however, the grasshopper
never becomes immobile. Instead, it exhibits different
behaviors, such as struggling to free itself from the grasp
Mean recovery time (log sec)

a 4.5
of a praying mantis, hopping away if the initial attack
4.0 by the bird was unsuccessful, or swaying back and forth
in an apparent attempt to threaten an approaching spi-
3.5 der. In a second experiment, Honma et al. (2006) tied
the hind legs of grasshoppers with fishing line so they
3.0 could not assume the T-shape, and presented them to
frogs (the life of a grasshopper destined for use in pre-
0.0 dation experiments can be harsh indeed!). Videotape
Control Owl
analysis revealed that the frogs easily swallowed the
bound grasshoppers; each frog could readily adjust the
5.5
position of the grasshopper in its mouth. On the basis
b
of their results, Honma and co-workers suggest that the
Mean recovery time (log sec)

5.0 rigid posture assumed by pygmy grasshoppers is a spe-

cialized adaptation for avoiding frog predation. They
4.5
further propose that the immobile posture deters pre-
dation not by mimicking death, but by enlarging the
4.0 functional body size of the grasshopper, making it dif-
ficult to swallow. Praying mantises are obviously quite
3.5 effective predators of grasshoppers (refer again to
Table 13.3), so why don’t grasshoppers assume the rigid
0.0 posture when captured by them? According to the
Control Human Human authors, the size-enhancing rigid posture of grasshop-
being being
(averted (direct
pers would be ineffective against praying mantises
eyes) gaze) because these predators gnaw on captured prey rather
than swallowing it whole. Findings such as these
FIGURE 13.21 Mean time to recovery from feigning death remind us that when evaluating the defensive behavior
in neonatal hognose snakes exposed to various recovery of a prey animal, it is always good to know something
conditions. (a) In experiment 1, snakes recovered in the about the foraging mode of its predators. The results
presence of a stuffed owl or in the absence of a stuffed owl
with pygmy grasshoppers also suggest a need to revise
(control condition). (b) In experiment 2, snakes recovered
the way in which we view examples of predator-induced
in the presence of a human being (with eyes staring at the
snake or with eyes averted) or in the absence of a human immobility in prey: in some cases, prey immobility may
being (control condition). Because the recovery times were function to mimic death, while in others it may serve to
skewed, the data were transformed into logarithms of enhance body size, a strategy that also seems to intim-
seconds. (From Burghardt and Greene 1988.) idate predators, as we discuss next.
294 Chapter 13 / Antipredator Behavior

INTIMIDATION
AND FIGHTING BACK
Prey animals have many ways of communicating “I am
formidable” to a predator. Presumably, when a predator
encounters a large, threatening, well-armed prey, it will
continue on its way, searching for a less challenging meal.
Here are some ways that prey can intimidate predators
and even fight back under certain circumstances.

ENHANCEMENT OF BODY SIZE

AND DISPLAY OF WEAPONRY
When dealing with potential predators, some animals
employ the size-maximization principle. A cat hunches
its back and erects its fur in the presence of a dog. Some
toads and fishes inflate themselves when disturbed. In
each case, the animal increases its size and appears more
formidable or unswallowable. Several displays of
intimidation through an increase in size are shown in FIGURE 13.22 Intimidation displays in several species of
Figure 13.22, but threat maximization need not always animals. The displays make the animal appear larger and
be visual. Loud calls, hisses, or growls may also cause a more formidable to predators. The animals shown here
predator to look elsewhere for its next meal. are (counterclockwise) the frilled lizard, cat, short-eared
Some animals display their weapons when con- owl, and porcupine fish. (Modified from Johnsgard 1967.)
fronting a predator. Ungulates often display their horns
and paw at the ground, perhaps to draw attention to their emitting a hissing sound. (The sound is produced when
dangerous hooves. Porcupines erect their spines and cats the butterfly rubs the veins on the base of its forewings
display their teeth. All these postures are probably meant against those on the top of its hindwings.) The butter-
to intimidate a predator. fly repeatedly performs this sequence of wing flicking
and hissing, all the while appearing to track changes in
predator position. The suddenly flashed eyespots and the
EYESPOTS hissing sounds certainly seem like devices to intimidate
Some animals have spots that resemble eyes in the odd- potential predators, but what experimental evidence do
est of places (Figure 13.23). What possible function could we have to support this function for the two traits?
such markings have? It turns out that eyespots can serve Adrian Vallin and colleagues (2005), armed with a
at least two defensive functions (Owen 1980). First, if the black permanent marker and a pair of scissors, set out to
spots are relatively small, then they may serve as targets test whether the eyespot display and hissing sound of the
to misdirect a predator in a manner similar to that harmless peacock butterfly increase its chances of surviv-
described for false heads. Such eyespots are typically ing an encounter with an avian predator, the blue tit
located on nonvital portions of the body, and thus prey can (Cyanistes caeruleus; previously Parus caeruleus). The
often escape a predatory encounter with less than fatal
damage. Second, if the spots are large, few in number, and
suddenly flashed, they may startle or frighten a predator;
it is this function of eyespots that will be our focus here.
What evidence is there to support the claim that by
flashing eyespots a harmless prey animal can increase its
chances of surviving an encounter with a potential preda-
tor? We will consider the case of the peacock butterfly
(Inachis io). At rest, peacock butterflies are dark-colored
leaf mimics. The margins of their wings are irregular,
and they keep their wings closed so that only the dark
ventral side of each is displayed. When disturbed, how-
ever, the butterfly transforms into a very different crea-
ture, suddenly flicking open its wings to expose bright FIGURE 13.23 A toad directs its backside toward an
colors and four large eyespots, and at the same time attacker, revealing a pair of eyespots.
 Intimidation and Fighting Back 295

researchers presented wild caught blue tits with butterflies group to determine the effects of sound production); (4)
from one of the following six treatments (Figure 13.24): butterflies whose sound production capabilities were
(1) butterflies whose eyespots had been colored over with intact but who had a small part of the base of their hind-
a black marker (experimental group to determine the wings removed (control group for the sound production
effects of eyespots); (2) butterflies whose eyespots were manipulation); (5) butterflies whose eyespots had been
intact but whose wings on the dorsal side, close to the colored over and whose sound production had been
body, had been colored over with a black marker (con- stopped (experimental group to determine the combined
trol group for eyespot manipulation); (3) butterflies effects of eyespots and sound production); and (6) but-
whose sound production abilities had been stopped terflies whose wing bases had been colored with a black
through removal of a small part of the forewings marker and who had a small part of the base of their
involved in producing the hissing sound (experimental hindwings removed (control group for eyespot and

Experimental groups Control groups

Effect of eyespots
(1) (2)

Effect of sound production

(3) (4)

Effect of eyespots and

sound production
(5) (6)

FIGURE 13.24 Six different treatment groups of peacock butterflies used to test the effectiveness of eyespots and
sound in deterring an avian predator. Upper row: (1) butterflies whose eyespots had been colored over with a black
marker (experimental group to determine the effects of eyespots) and (2) butterflies whose eyespots were intact but
whose wings on the dorsal side, close to the body, had been colored over with a black marker (control group for eye-
spot manipulation). Middle row: (3) butterflies whose sound production abilities had been stopped through removal of
a small part of the forewings involved in producing the hissing sound (experimental group to determine the effects of
sound production) and (4) butterflies whose sound production capabilities were intact but who had a small part of the
base of their hindwings removed (control group for the sound production manipulation). Bottom row: (5) butterflies
whose eyespots had been colored over and whose sound production had been stopped (experimental group to deter-
mine the combined effects of eyespots and sound production) and (6) butterflies whose wing bases had been colored
with a black marker and who had a small part of the base of their hindwings removed (control group for eyespot and
sound production manipulations). (Modified from Vallin et al. 2005.)
296 Chapter 13 / Antipredator Behavior

sound production manipulations). During predation tri-

als, a butterfly from one of the six groups was placed on
a willow log in the middle of a small room, and a blue
tit was released into the room. Each trial lasted 30 min-
utes, and during this time Vallin and co-workers directly
observed and videotaped the bird–butterfly interactions.
They conducted from eight to ten trials in each of the
six treatments. No individual bird or butterfly was used
more than once, and birds were released at their site of
capture after use in a trial. All birds were banded prior
to release, so their individual identity would be known
to the researchers, and they would not be used again. All
butterflies were discovered by the birds, and all that were
seized were eaten without hesitation. The latter obser-
vations suggest that the butterflies were palatable to the
FIGURE 13.25 The bombardier beetle ejects a hot,
birds. The results, summarized by treatment in Table
irritating spray at its attackers. This beetle, tethered
13.4, show that butterflies survived better when they had to a wire fastened to its back with wax, responds with
eyespots than when the eyespots were colored over. In excellent aim to a forceps pinch on its left foreleg.
contrast, there was no difference in the survival of but-
terflies in the sound and no sound treatments. Finally,
butterflies survived better when they had eyespots and Other masters of chemical warfare are the bom-
sound than when they lacked eyespots and sound, bardier beetles, which deter predators by emitting a defen-
although there was no evidence that eyespots and sound sive spray that contains substances stored in two glands
were any better than eyespots alone (compare “eyespots that open at the tip of the abdomen (Dean et al. 1990;
and sound” with the “no sound” treatment). Taken Eisner 1958). Because the tip of the abdomen acts as a
together, these results indicate that eyespot displays by revolving turret, the spray can be aimed in all directions
a harmless butterfly can be very effective in deterring an (Figure 13.25). The chemical reactants from the two
avian predator. Although sound production did not seem glands are mixed just before they are discharged, produc-
critical to the peacock butterfly surviving its encounters ing a sudden increase in temperature of the mixture. The
with blue tits, the authors suggest that the hissing sound hot spray is ejected, accompanied by audible pops, in quick
might prove effective in deterring other predators. pulses. The effect has been likened to that of the German
V-1 “buzz” bomb of World War II (Dean et al. 1990).
TABLE 13.4 Survival of Peacock Butterflies Chemical deterrents are by no means limited to
during Predation Trials with Blue Tits arthropods, as anyone who has had the misfortune of
surprising a skunk or who owns a dog that has enjoyed
Number % Survival the same experience must surely know. Although the
Treatment of trials of butterflies defensive response of the Texas horned lizard
No eyespots 10 50.0 (Phrynosoma cornutum) is perhaps less well known than
Eyespots 9 100.0 that of the skunk, it is certainly no less spectacular.
No sound 8 100.0 When disturbed by canid predators such as kit foxes and
Sound 8 87.5 coyotes, this small, spine-covered lizard of the south-
No eyespots or sound 10 20.0 western United States typically spatters its attacker with
Eyespots and sound 9 100.0 a stream of blood ejected from the sinus surrounding its
eyes (Sherbrooke and Mason 2005; Sherbrooke and
Source: Data from Vallin et al. (2005). Middendorf 2004). When squirted, coyotes and kit
foxes respond with exaggerated mouth movements and
lateral head shaking. Blood squirting in P. cornutum is
CHEMICAL REPELLENTS
rarely directed at predators other than canids. At the
A wide variety of insects can discharge noxious chemi- turn of the twentieth century, Charles Holder (1901)
cals when they are captured. Some of these chemicals are examined blood squirting behavior and suggested, on
powerful toxins or irritants, and in some species they can the basis of trials in which his fox terrier posed as a
be shot with considerable accuracy in several directions. predator, that the ejected blood contained noxious com-
The assassin bug (Platymeris rhadamantus) reacts to a dis- ponents. Apparently, contact between the ejected blood
turbance by spitting copious amounts of fluid in the and nasal membranes of the dog was particularly irri-
direction of the attacker. The saliva is rich in enzymes tating, and only a single encounter was required to pro-
and causes intense local pain when it comes in contact duce “a wholesome dread” in the lizard’s canine
with membranes of the eyes or nose. tormentor. More recent data suggest that the squirted
 Pronouncement of Vigilance 297

blood primarily affects oral receptors (Sherbrooke and TABLE 13.5 Hypotheses to Explain the
Mason 2005) and that the source of the defensive com- Function of Stotting
pounds in the lizard’s blood may be its diet of ants
(Sherbrooke and Middendorf 2004). Benefits to the individual
Signaling to the predator
1. Pursuit invitation
2. Predator detection
PRONOUNCEMENT
3. Pursuit deterrence
OF VIGILANCE 4. Prey is healthy
Prey that scan their surroundings for potential predators 5. Startle
are said to be vigilant. You might be surprised to learn 6. Confusion effect
that having detected a predator, some prey actually Signaling to conspecifics
approach it rather than head as quickly as possible in the 7. Social cohesion
opposite direction. Other prey may leave, but not before 8. Attract mother’s attention
performing a display seemingly directed at the predator. Signaling not involved
What could be the function(s) of these prey reactions? 9. Antiambush behaviora
One function might be in communicating to the preda- 10. Play
tor that it is dealing with an alert and aware prey. Sit- Benefits to other individuals
and-wait predators might leave their ambush sites once 11. Warn conspecifics
prey communicate to them that they have been spotted. a
Height gained through stotting allows fleeing individuals a better
An excellent example concerns the interactions between view of their surroundings.
prey and timber rattlesnakes (Crotalus horridus), preda- Source: After Caro (1986a).
tors that may spend hours or even days at an ambush site
waiting for the appropriate moment to strike. Rulon
Clark (2005) captured timber rattlesnakes at a nature The displays had visual components (for example, tail-
preserve in New York, surgically implanted radiotrans- flagging by the rodents) and may have had auditory com-
mitters in their body cavities, and returned the snakes to ponents, but this could not be determined because the
their site of capture. He then located individual snakes videocameras were not equipped with microphones.
on a daily basis and trained videocameras on those that Hunting snakes that received the displays were 4.3 times
were either basking (snake in the sunlight, with loose more likely to abandon their site after harassment than
coils) or hunting (snake with tight coils and head ori- hunting snakes that were not harassed. Also, hunting
ented in the direction of a runway used by small mam- snakes that were harassed tended to move greater dis-
mals). Using this combination of radiotracking and tances before resuming foraging than those that were not
videotaping, Clark recorded natural encounters between harassed and were simply changing ambush sites. Snakes
the snakes and their prey, which included chipmunks, that were basking showed no increased tendency to
gray squirrels, and a wood thrush. In 12 instances, indi- abandon their site following harassment. These data pro-
vidual prey were recorded harassing a snake; six of the vide evidence from natural encounters between preda-
snakes were basking and six hunting. During harassment tors and their prey to support the notion that prey may
displays, prey made repeated approaches to the snake. benefit by communicating to an actively hunting sit-and-
wait predator that it has been spotted. In this case, the
specific benefit to the prey is that the predator leaves the
general area following harassment.
Stalking predators may also abandon the hunt once
they receive the signal that they have been detected.
Stotting, a stiff-legged bounding display performed by
pronghorn, and many species of deer and antelope
(Figure 13.26), appears to be just such a signal. The stot-
ting display has attracted the attention of many investi-
gators, and arrival at pronouncement of vigilance (i.e.,
announcing to a predator that it has been detected) as a
plausible function has involved testing predictions from
several hypotheses. At least 11 hypotheses have been
proposed for the function of stotting (Caro 1986a; Table
13.5). Although not mutually exclusive, the hypotheses
range from the interpretation of stotting as a signal given
by a hunted animal to either a predator or a conspecific
FIGURE 13.26 Stotting by a Thomson’s gazelle. to the interpretation that stotting has no signal value at
298 Chapter 13 / Antipredator Behavior

TABLE 13.6 Outcome of 31 Cheetah Hunts the antiambush hypothesis for Thomson’s gazelles
Involving Thomson’s Gazelles That Did or Did (recall that this hypothesis states that stotting is not a
Not Stott signal but simply a way to gain a better view of the sur-
roundings), Stankowich and Coss (2007a) found some
Chase occurred
empirical support for this hypothesis for Columbian
Chase Chase Hunt black-tailed deer (Odocoileus hemionus columbianus).
successful unsuccessful abandoned Total When approached by a single human observer, black-
tailed deer stotted more often in taller vegetation, a
Gazelle stotts 0 2 5 7 finding consistent with one prediction of the anti-
Gazelle does ambush hypothesis. Stankowich and Coss predict,
not stott 5 7 12 24 however, that stotting in black-tailed deer likely has
Source: Modified from Caro (1986b). multiple functions, as it does in Thomson’s gazelles.
Finally, although often performed in the presence of
predators, stotting also occurs during intraspecific
all and is simply a form of play or, alternatively, a means encounters in many species, and we can only guess what
to get a better view of the surrounding vegetation (the its function is under these circumstances.
antiambush hypothesis). In the first true effort to
distinguish among the hypotheses, Tim Caro (1986b)
recorded the response of Thomson’s gazelles (Gazella GROUP DEFENSE
thomsoni) to naturally occurring predators, usually chee-
tahs (Acinonyx jubatus), in the Serengeti National Park of Until now we have focused almost exclusively on strate-
Tanzania. He analyzed prey behavior, cheetah behavior, gies employed by individual animals to avoid being
and the outcome of hunts and found that cheetahs were eaten. Some animals, however, are social, and member-
more likely to abandon hunts when their prey stotted ship in a group makes accessible several antipredator tac-
than when they did not (Table 13.6). These results, com- tics that are not available to solitary individuals. We next
bined with other data that refuted many of the remain- consider some examples of how social animals cope with
ing hypotheses, suggested that stotting typically predators. Although we discuss each tactic separately,
functioned to inform the predator that it had been some may interact for any given species. Finally, as you
detected. Two other functions for stotting were sup- consider these antipredator mechanisms, keep in mind
ported by Caro’s observations. First, mothers may stott that group living has many advantages, including those
to distract a predator from their fawn, a function much totally unrelated to protection from predators (see
like the broken wing displays described for killdeer. Chapter 19).
Second, fawns appear to stott to inform their mother
that they have been disturbed at their hiding place.
ALARM SIGNALS
A subsequent study suggests that the context of the
cheetah–gazelle encounter and age of the performing When a predator approaches a group of prey, one or more
gazelle are not the only factors that influence the function individuals within the aggregation may give a signal that
of stotting. The type of predator is another consideration. alerts other members of the group to the predator’s pres-
When hunted by coursing predators that rely on stamina ence. Alarm signals may be visual, auditory, or chemical,
to outrun their prey, gazelles appear to use stotting as an and they often inspire retreat by prey to a safe location.
honest signal of their ability to outrun predators In some cases, the alarm may aid the signaler or its rela-
(FitzGibbon and Fanshawe 1988). Coursing predators tives. In other instances, an alarm appears to benefit all
such as African wild dogs (Lycaon pictus) concentrate their those exposed to the signal, including in some cases indi-
chases on those individuals within a group that stott at viduals of different species and even different taxonomic
lower rates. Thus they appear to use information conveyed groups. Eurasian red squirrels, for example, respond with
in stotting to select their prey. In the study by FitzGibbon flight or increased vigilance to the alarm calls of Eurasian
and Fanshawe, the mean rate of stotting by gazelles that jays (Randler 2006). The proposed selective advantages of
were chased was 1.64 stotts per second, and for those not signaling alarm are covered in more detail in Chapter 19.
chased, 1.86 stotts per second. By signaling their ability We will focus our discussion here on the chemical alarm
to escape at the start of a hunt, those gazelles with high system of an amphibian, the western toad.
stamina and/or running speeds may not have to prove Injured tadpoles of the western toad, Bufo boreas,
their physical prowess by outrunning wild dogs in long, produce an alarm substance, and there is experimental
exhausting, and potentially dangerous chases. evidence that it functions as an effective antipredator
If the function of stotting varies with the species of device. Individuals of this species live in the ponds and
predator, then we should not be surprised if future lakes of western North America, where the tadpoles
studies reveal that the function varies with the species form dense aggregations. Diana Hews (1988) first doc-
of prey as well. Indeed, although Caro (1986b) discarded umented the response of toad tadpoles to release of the
 Group Defense 299

Alarm substance solution, N = 10 The benefits of increased predator-detecting ability

Control solution, N = 10
accrue to members of many different kinds of groups.
Florida scrub jays (Aphelocoma coerulescens) form single-

Number of captures per attempt (mean ± SE)

species groupings, usually consisting of two to eight fam-
ily members. Because these birds live in small, permanent
0.80 groups of stable composition, it is possible for individuals
to coordinate their vigilance in a highly structured sentinel
0.65 system (McGowan and Woolfenden 1989). At any given
time, only one family member typically sits on an exposed
0.50 perch and continually scans the surroundings for preda-
tors. If a predator is spotted, the sentinel sounds the alarm,
0.35 and family members respond by either mobbing a ground
predator or fleeing, or monitoring the movements of an
0.20
aerial attacker. Periodic exchanges among family members
occur to relieve the sentinel bird of its duties. Sentinel sys-
0.05
tems have also been reported for some mammals that live
Treatment in family-based social groups, including dwarf mongooses
FIGURE 13.27 Effects of alarm substance of toad tadpoles (Rasa 1986) and meerkats (Moran 1984). The benefit of
on the attack success rate of dragonfly naiads. Tadpoles improved detection, however, is not reserved solely for
exposed to the alarm substance were less vulnerable to family-based social groups with relatively permanent
predation by naiads than those exposed to the control membership. Even members of temporary groupings, such
substance, water. (Modified from Hews 1988.) as Iberian green frogs that fortuitously come together at
favorable foraging locations, appear to benefit from
alarm substance and then tested whether tadpoles improved detection of predators (Martín et al. 2006).
alerted by the substance had higher survival rates than Finally, members of mixed-species groups also benefit
those not exposed. Two natural predators of western toad from improved detection, providing that they are on the
tadpoles, giant water bugs and dragonfly naiads, were lookout for the same species of predators and that they
used in the experiments. communicate detection to other group members.
When tested in aquaria, western toad tadpoles
avoided the side of the tank that contained a giant water DILUTION EFFECT
bug that was feeding on a conspecific tadpole (in a visu-
ally isolated but interconnected container) and increased Individuals in groups are safer not only because of their
their activity. Tadpoles did not avoid the predator’s side enhanced ability to detect predators but also because
of the tank when the water bug was feeding on a tadpole each individual has a smaller chance of becoming the
of another species, though they did show a slight, short- next victim. Called the dilution effect, this advantage
term increase in activity. Regarding the latter results for for grouped prey operates if predators encounter single
activity, Hews suggests that there might be a small and individuals or small groups as often as large groups and
nonspecific activity response to substances released from if there is a limit to the number of prey killed per
injured tadpoles of any species, but that the response to encounter. As group size increases, the dilution effect
an injured conspecific is much stronger and longer becomes more effective, and improved vigilance appears
lasting. It is important to note that toad tadpoles alerted to provide relatively less benefit (Dehn 1990).
by the conspecific alarm substance were less vulnerable Although this notion of safety in numbers has intu-
to predation. Dragonfly naiads had fewer captures per itive appeal, in some cases, predators aggregate in areas
attack in tests with tadpoles exposed to the toad extract where prey are abundant. As a result, some grouped prey
containing the alarm substance than with tadpoles may actually suffer higher predation rates. In an exami-
exposed to the control extract, water (Figure 13.27). nation of the balance between the forces of the dilution
effect and the aggregating response of predators,
Turchin and Kareiva (1989) studied grouping in aphids
IMPROVED DETECTION
(Aphis varians). These small insects form dense clusters
Early detection of a predator can often translate into on the flowerheads of fireweed, and it is here that they
escape for prey, and groups are typically superior to lone are preyed on by ladybird beetles, typically Hippodamia
animals in their ability to spot predators. Increases in the convergens. In one experiment, the researchers quantified
number of group members (and hence the number of per capita population growth rates (a measure of indi-
eyes, ears, noses, etc.) often result in increases in the vidual survivorship) for aphids living singly and for those
immediacy with which approaching predators are living in colonies of over 1000 individuals and found that
detected, and the escape response of a vigilant individ- individual aphids benefited by forming groups.
ual can alert others to the approaching danger. Grouping was only advantageous, however, in the
300 Chapter 13 / Antipredator Behavior
Number of ladybird beetles attacking

edges. By obtaining a central position, animals can

>1000 Aphids
100 Aphids decrease their chances of being attacked and increase the
10 Aphids probability that one of their more peripheral colleagues
will be eaten instead. This antipredator mechanism,
an aphid colony

4
often referred to as the selfish herd (Hamilton 1971),
3 emphasizes that although a given group appears to con-
sist of members that coordinate their escape efforts, it is
2
actually composed of selfish individuals, each trying to
1 position as many others as possible between itself and the
predator. The selfish herd hypothesis differs from the
0 dilution effect and improved detection because it con-
3 6 24 48 72
siders the spatial arrangement of individuals within a
Time (hours)
group (papers reviewed by Beauchamp 2007).
FIGURE 13.28 The tendency for prey individuals to form According to the selfish herd hypothesis, individu-
large groups and thereby dilute their chances of becom- als at the center of a group are safer than those at the
ing the next victim may be countered by the tendency periphery when attacked by a predator from outside the
by their predators to aggregate where prey are most group. Thus, during an attack, we might expect mem-
common. Here, predatory ladybird beetles gather in
bers of a group to seek the safer, more central positions,
larger numbers at aphid colonies that contain the most
individuals. (Modified from Turchin and Kareiva 1989.)
and this would lead to a reduction in nearest neighbor
distances. Here we consider one study that examined the
way in which individuals exposed to an alarm signal place
presence of predators; when ladybird beetles were themselves in a group as compared to individuals who
excluded from fireweed plants, the individual survivor- were habituated to the signal. Schreckstoff is an alarm
ship of aphids did not increase with the colony’s size. chemical produced by some species of fish when physi-
The next question, then, is how do ladybird beetles cally attacked. Jens Krause (1993) habituated 14 dace
respond to the grouping of their prey? Turchin and (Leuciscus leuciscus) to the odor of Schreckstoff, so that
Kareiva found that beetles exhibited a strong aggregation they no longer reacted to its presence. A lone minnow
response: more than four times as many beetles were (Phoxinus phoxinus), which was still responsive to
found at aphid colonies of over 1000 individuals than at Schreckstoff, was added to the group. Before Schreckstoff
small and medium-sized colonies (Figure 13.28). In addi- was added to the water, the minnow randomly inter-
tion to gathering at large colonies, ladybird beetles also mingled in the shoal of dace. In repeated tests, when
increased their feeding rate as the aphids’ density Schreckstoff was added, however, the minnow moved
increased. On average, beetles consumed 0.9 aphids per closer to the other fish and positioned itself so that it
ten minutes in colonies of 10 individuals and 2.4 aphids was surrounded by them. Only the alarmed fish chose
per ten minutes in colonies of 1000 individuals. Thus, the a central location in the group, thus supporting predic-
group size of aphids appears to affect the per capita growth tions of the selfish herd hypothesis.
rate of the aphid colony, the number of predators attracted One might ask, then, are central locations within the
to the colony, and the rate at which predators feed. Given group always the best? The answer is no. In fact, a study
all these factors, does grouping reduce predation risk for on the antipredator advantages of schooling in fish sug-
aphids? Apparently so. When the researchers calculated gests that the center is sometimes the most dangerous
the instantaneous risk of predation to an individual aphid place to be. When in the company of a predatory seabass
in a ten-minute period, they obtained values of 0.05 for (Centropristis striata), silversides (Menidia menidia) at the
colonies of 10 and 0.008 for colonies of 1000 or more. center of a school suffered the most attacks (Parrish 1989).
Thus, in the aphid-ladybird system, the dilution effect still Rather than assaulting the margins, seabass swim toward
occurs despite the strong tendency of predators to aggre- the center of the school, split the school into two groups,
gate at large colonies of prey. Turchin and Kareiva are and then strike at the tail end of one of the groups, where
quick to point out, however, that predators are not the individuals that were in the center now find themselves.
only enemies of aphids. Parasitoids and pathogens may The relative safety of a location within a group thus
increase rapidly in large groups of aphids and may pro- depends on the predator’s method of attack. Because
foundly affect mortality, perhaps even eliminating the schools of fish undoubtedly cope with a number of preda-
antipredator advantages of the dilution effect. tors, each possibly using a different attack strategy, the rel-
ative advantage of central versus peripheral locations may
change. In addition, factors such as foraging efficiency
SELFISH HERD
(those in the front see the food first) and the energetics
In most groups, as we saw with those of overwintering of locomotion (fish in the front of a school may experi-
monarch butterflies at the start of this chapter, centrally ence more “drag” than those at the back) probably also
located animals appear to be safer than those at the influence optimal positions within the school.
 Group Defense 301

CONFUSION EFFECT of the school of prey increased (Figure 13.29). In the

three ambush predators, the increased size of the prey’s
Predators that direct their attacks at a single animal
school appeared to produce hesitation and behavior
in a group may hesitate or become confused when
characteristic of conflict (such as alternating between
confronted with several potential meals at once. No mat-
approach and avoidance). Perch, however, switched tar-
ter how brief, any delay in the attack will operate in favor
gets more frequently as the size of the school increased
of the prey. The confusion effect describes the situa-
and, with each switch, reverted to an earlier stage of the
tion in which predators are less successful in attacking
hunting sequence. Under natural conditions, predators
prey because they are unable to single out and attack
of fish may achieve hunting success by restricting their
individual prey (Krause and Ruxton 2002; Miller 1922).
attacks to individuals that have either strayed from the
Few experiments have investigated whether specific
school or have a conspicuous appearance. In both cases,
predators are susceptible to the confusion effect, but the
the predator can concentrate on the odd target.
limited data available suggest that tactile predators
(including nematodes and the larvae of certain insects)
are particularly susceptible, as are visual predators with MOBBING
highly agile prey (Jeschke and Tollrian 2007). Our exam- Sometimes prey attack predators. Approaching, gathering
ple concerns visual predators. around, and harassing one’s enemies is called mobbing
The confusion effect is thought to be one of the pri- (Curio 1978). This antipredator strategy typically involves
mary antipredator advantages of schooling in fish. visual and vocal displays, as well as swoops, runs, and
When the fish in the school scatter, it makes it difficult direct hits on the predator. Mobbing is usually initiated
for visual predators to focus on a single one. If you’ve by a single individual, and then conspecifics, or members
ever tried to use a little net to catch a single member of of another species, join in the fracas. The possible func-
a school of fish in an aquarium, this may seem especially tions of mobbing include, but are not limited to, (1) con-
believable to you. Neill and Cullen (1974) examined the fusing the predator; (2) discouraging the predator either
effects of the size of the school on the hunting success through harassment or by the announcement that it has
of two cephalopod predators (squid, Loligo vulgaris, and been spotted early in its hunting sequence; (3) alerting
cuttlefish, S. officinalis) and two fish predators (pike, Esox others, particularly relatives, of the danger; and (4) pro-
lucius, and perch, Perca fluviatilis). Whereas squid, cut- viding an opportunity for others, again particularly rela-
tlefish, and pike are ambush predators, perch typically tives, to learn to recognize and fear the object that is being
chase their intended victims. In most cases, predators mobbed (Curio 1978). Most evidence suggests that mob-
were tested with fish of their natural prey species in bing is not an act performed by a cooperative group of
schools of 1, 6, and 20 individuals. For all four preda- individuals that is attempting to protect the group as a
tors, attack success per encounter decreased as the size whole, but rather it is the selfish act of individuals that are
attempting to protect only those that will benefit them
directly, that is, themselves and their mates, offspring, and
Cuttlefish Squid Pike Perch
relatives (Ostreiher 2003; Shields 1984; Tamura 1989).
Is mobbing costly? Mobbing certainly takes time and
energy away from other activities. In addition, when in
0·8 hot pursuit of a predator, mobbers appear to have a
greater chance of being preyed on than nonmobbers,
Capture/contact ratio

0·6 although there is some disagreement over whether mob-

bing actually entails a deadly risk to those that participate
(Curio and Regelmann 1986; Hennessy 1986). One fac-
0·4
tor influencing the degree of risk might be the size of the
mob. We can imagine, for example, that as more individ-
0·2 uals join the mob, the risk to any one of them could be
reduced through dilution or confusion effects. But does
risk arise only from the predator being mobbed? At least
for pied flycatchers (Ficedula hypoleuca), the answer
1 6 20 1 6 20 1 6 1 6 20 appears to be no. Tatjana Krama and Indrikis Krams
School sizes (2005) played tapes of flycatcher mobbing calls or blank
tapes from the top of nest boxes that contained parts of
FIGURE 13.29 The confusion effect. As the size of the
school of prey increases, four predators have reduced recently abandoned flycatcher nests and a quail egg. The
hunting success because of either hesitation and conflict tapes were played during daylight hours when flycatch-
behavior (in the case of the cuttlefish, squid, and pike) ers give mobbing calls in response to predators in the
or frequent switching of targets (perch). (Modified from vicinity of their nest. The researchers were interested in
Neill and Cullen 1974.) whether martens (Martes martes), weasel-like predators
302 Chapter 13 / Antipredator Behavior

that significantly decrease the reproductive success of model of their current predator, a puma, than to a model
pied flycatchers through nocturnal raids on nests, might bearing a spotted pattern similar to the jaguar that
eavesdrop on mobbing calls while resting nearby during roamed the area up until about 600,000 years ago
the day. Krama and Krams found that more nest boxes (Stankowich and Coss 2007b).
with mobbing calls were depredated by martens (identi- Populations do not lose antipredator behaviors at
fied by their tracks) than nest boxes with blank tapes the same rate once a particular predator disappears. The
(13/56 as compared to 2/56, respectively). These data time course of the loss of antipredator behavior depends
suggest that beyond the potential risk from the predator on, among other things, the costliness of fleeing from the
being mobbed, prey also face risks from other predators predator (Blumstein et al. 2006), and whether all or just
that may be eavesdropping on their mobbing displays. some predators are lost (Blumstein 2006).
In this chapter we have described the many ways in
which prey species respond to predators. We will end our
coverage with a discussion of what happens to these pat- SUMMARY
terns of behavior when prey species lose their predators;
loss of predators can occur naturally (say, when the prey Some antipredator mechanisms decrease the probability
expands its range into an area where the predators do not of an encounter with a potential predator. The prey may
occur) or unnaturally (for example, if the prey species is go undetected if it blends with its background (col-
brought into captivity). When predators are lost, do oration matching the visual background) or if the out-
antipredator behaviors remain in the repertoire of the line of its body is broken up by bold contrasting patterns
prey species, or do the behaviors eventually disappear? (disruptive coloration). Other prey occur in several
And if they do disappear, what factors influence the shapes and colors (polymorphism), presumably to pre-
speed with which they depart? vent the formation of search images by predators.
Alternatively, the prey may be detected by a predator and
either be recognized as inedible (warning coloration) or
MAINTENANCE go unrecognized as a potentially tasty meal (masquerade
OF ANTIPREDATOR BEHAVIOR and Batesian mimicry). Although many of these defenses
involve colors and patterns, the behavior of a prey ani-
If an animal freezes or flees when it detects a predator, mal is critical to the success of these mechanisms.
that means it’s not doing something else: for instance, it’s Other defenses operate during an encounter with a
not foraging, looking for mates, or resting. So, respond- predator and increase an animal’s chances of surviving
ing willy-nilly to everything in the environment as if it the encounter. Amid the many options available, an indi-
were dangerous may well decrease an animal’s fitness. vidual may divert the predator’s attention, inform the
If antipredator behavior is costly, then we expect that predator that it has been spotted early in the hunt, or
it should be lost when it is no longer needed. One case turn the tables and fight back.
in which animals may suddenly find themselves free of Membership in a group has antipredator advantages.
predators is on islands. We’ve already described how Generally, group-living prey are better than solitary prey
birds on the island nation of New Zealand often exhibit at detecting, confusing, and discouraging predators.
remarkable naiveté when it comes to potentially danger- In addition, during any given attack, an individual in a
ous mammals, which are not naturally found there large group has a lower probability of being the one
(Chapter 4). Marine iguanas (Amblyrhynchus cristatus) on selected by the predator (dilution effect) and may use
the Galapagos Islands have also been free of predators for other group members as a shield between itself and the
the last 5 to 15 million years, until feral cats and dogs enemy (selfish herd).
were introduced about 150 years ago. Although iguanas Although for convenience we have discussed
become more wary following experience with a perceived antipredator behavior as distinct defensive mechanisms,
threat of predation, they still are not cautious enough to our intent is not to imply that a given individual or
escape from predators (Rodl et al. 2007). Marsupials (kan- species is characterized by a single protective strategy.
garoos, wallabies, and their relatives) also lose some (but Indeed, most animals face a multitude of predators with
not all) of their antipredator behaviors after they are iso- diverse methods for detecting and capturing prey, and
lated on islands (Blumstein and Daniel 2005). thus having several defensive tactics is crucial. The
Relaxed selection for antipredator behavior is not behavior and color patterns of animals must be inter-
confined to islands. Predators that were once historically preted in the context of several selective forces; after all,
important might become locally extinct. Ted Stankowich animals must not only avoid being eaten but also must
hid behind cover, then popped up life-sized pho- feed and reproduce. Finally, antipredator behaviors have
tographs of large, predatory cats to elicit alarm responses costs, as is nicely illustrated by the loss of such behav-
from Columbian black-tailed deer (Odocoileus hemionus iors from the repertoire of a prey species when its preda-
columbianus). Deer responded more strongly to the tors disappear.
 PART THREE

Interactions Between Individuals

303
This page intentionally left blank
 14
Reproductive Behavior

Sexual Selection: Historical and Theoretical females, but still females—with their two feeding
Background claws—gain energy more quickly (Weissburg 1992,
Explanations for Sex Differences in Reproductive 1993). Males incur other energetic costs associated with
Behavior their greatly enlarged claw. In sand fiddler crabs (Uca
Revisiting the Ideas of Bateman pugilator), for example, males with their major claw have
Intrasexual Selection—Competition for Mates higher metabolic rates than males without their major
Adaptations That Help a Male Secure Copulations claw (Allen and Levinton 2007). In this same study, when
Adaptations That Favor the Use of a Male’s Sperm
males were matched for body size and exercised on a
Sexual Interference: Decreasing the Reproductive
Success of Rival Males
Intersexual Selection—Mate Choice
Criteria by Which Females Choose Mates
Origin and Maintenance of Mate-Choice Preferences
Cryptic Female Choice
Sexual Conflict

Male fiddler crabs of the genus Uca are spectacular crea-

tures (Figure 14.1). Unlike females that have two small
claws for feeding, males have one small feeding claw and
one supersized claw (often called the major claw) that can
make up nearly half of their total body mass. When feed-
ing, fiddler crabs use the feeding claw(s) to scoop mud FIGURE 14.1 A male fiddler crab has one enormous claw
or sand into the mouth where specialized mouthparts and one small feeding claw (a female, not shown, has
remove algae and detritus from the sediment. To par- two small feeding claws). The greatly enlarged claw is
tially compensate for having only one feeding claw, males expensive to produce and maintain, so why do males
feed at faster rates and for longer periods than do have such a structure?

305
306 Chapter 14 / Reproductive Behavior

1.0 Without major claw certain preferred characteristics; this form of sexual
With major claw selection is called intersexual selection. Thus, males
0.8
not only fight with each other for access to females but
also compete to attract females through the elaboration
Proportion still running

of structures or behavior patterns. Male fiddler crabs also

0.6 use their major claw to attract females, holding it at cer-
tain angles and executing precision movements. In fid-
0.4 dler crabs then, the claw has dual functions, male–male
combat and mate attraction. In other animals, a spec-
tacular structure or behavior may function in both capac-
0.2
ities, or only in one.
Our focus in this chapter will be on reproductive
0.0 behavior, with a specific emphasis on sexual selection.
0 15 30 45 60 75 90
Time (min) Although Darwin (1871) envisioned sexual selection
occurring before copulation, we now know that sexual
FIGURE 14.2 Endurance capacity of male fiddler crabs selection operates after copulation as well (see discus-
with their major claw and without their major claw sions of sperm competition and cryptic female choice).
when run on a treadmill. Males without their major
claw exhibit greater endurance than those with their
major claw, indicating that the greatly enlarged claw is EXPLANATIONS FOR SEX DIFFERENCES
costly. (Modified from Allen and Levinton 2007.) IN REPRODUCTIVE BEHAVIOR
Throughout the animal kingdom, males typically com-
treadmill, those with their major claw had lower pete for females and females actively choose their mates.
endurance capacities than those without their major claw Why? Some people think that these differences in mat-
(Figure 14.2). Finally, observations of males that lose ing strategies between the sexes are related to differences
their major claw and then regenerate it indicate that sig- in investment in gamete production and parental care.
nificant production costs are associated with the claw At the core of the gamete argument are differences
(Backwell et al. 2000; Hopkins 1982). in the size and number of male and female gametes.
This costly claw presents a puzzle. What is the func- More specifically, whereas females produce a small
tion (or functions) of the enormous claw of male fiddler number of large, energetically expensive eggs, males
crabs? Given all the documented costs, how did the claw produce millions of small, relatively inexpensive sperm.
evolve? How can we reconcile the production of such a A. J. Bateman (1948), working with fruit flies, was the
bizarre structure with natural selection? first to suggest that as a result of differences in the num-
ber of gametes produced by the sexes, females (or more
correctly, their eggs) become a limited resource for
SEXUAL SELECTION: which males compete. A female, he reasoned, is likely
HISTORICAL AND THEORETICAL to maximize her reproductive success by finding the
BACKGROUND best-quality male to fertilize the limited number of eggs
she produces. Whereas a male’s reproductive success is
Darwin (1871) was the first to suggest that spectacular predicted to increase with the number of mates he
structures (such as the enormous claw of male fiddler acquires, a female’s reproductive success is not expected
crabs) could arise and be maintained through the to increase by mating with more than one male. Also,
process of sexual selection. According to Darwin, sexual among his fruit flies, some males appeared quite unsuc-
selection can occur through two mechanisms: male com- cessful in their attempts to acquire mates, while others
petition for access to mates and female choice of mates. garnered a disproportionate share of matings. Females,
In the first, called intrasexual selection, individuals of on the other hand, were usually similar in their number
one sex, usually males, gain a competitive edge by fight- of matings, probably because they could quickly obtain
ing with each other, and winners could claim the spoils sufficient sperm to fertilize their limited supply of eggs.
of victory—females. Intense fighting and competition for In short, Bateman (1948) found that variation in repro-
mates could lead to selection for increased size and elab- ductive success was greater among males than among
orate weapons. Male fiddler crabs, it turns out, use their females. Taken together, Bateman’s experiments with
major claw in battles with other males to secure breed- fruit flies led him to suggest that the sex difference in
ing burrows in preferred locations. Sexual selection can gametes may help to explain the “undiscriminating
also involve selective mate choice, in which individuals eagerness” of males and the “discriminating passivity”
of the sex in demand, usually females, choose mates with of females.
 Sexual Selection: Historical and Theoretical Background 307

Later, Robert Trivers (1972) suggested that, in addi- The gift is especially important to females when food is
tion to differential investment by the sexes in gametes, scarce, and it is during these times that females compete
there was differential investment in offspring, and this for males. In contrast, when food is plentiful, the rela-
too was responsible for competition and mate choice. In tive value of the gift declines, and males compete for
Trivers’s view, the sex that provides more parental invest- females (Gwynne and Simmons 1990). We see, then,
ment for offspring (usually the female) becomes a limit- that a change in courtship roles within a species of katy-
ing resource for which the sex that invests less (usually did results from variation in the relative importance of
the male) competes. As a result, males compete for access male parental investment.
to females and females have the luxury of choosing Trivers’s theory can also be tested by examining the
among available suitors. select group of nontraditional species—those in which
Within-species variation in the courtship roles of males invest more than females in the care of offspring.
some insects supports the idea that the degree of If Trivers were correct, then we would predict that
parental investment is important in determining which females of these species would compete for males, who
sex competes for access to individuals of the other sex. in turn would be quite discriminating in their choice of
For example, courtship roles are far from fixed in cer- mates. As we will see in the following chapter on parental
tain species of katydids (relatives of the grasshoppers). care and mating systems, Trivers’s theory is often, but
Indeed, in one species (as yet not formally classified by not always, supported by these exceptions to the rule. In
scientists), whether males compete for females or many species with sex-role reversals, males are choosy
females compete for males varies with the relative and females actively compete for mates, as predicted by
importance of male parental investment. In katydids, Trivers’s theory. In others, however, females also are
male parental investment consists of a nuptial meal. At choosy (even though they do not provide parental care),
the time of mating, a male katydid transfers his sper- and males still seem to compete with one another for
matophore (a packet of sperm and fluids) to the female. access to mates.
Following separation of the couple, the female bends Findings that are inconsistent with Trivers’s pre-
around and eats part of the spermatophore that is still dictions have caused some scientists to challenge the
attached to her abdomen (Figure 14.3). This protein- notion that patterns of parental investment are prime
rich meal is important in successful reproduction determinants of the nature and strength of sexual selec-
because both the number and fitness of her offspring are tion (Gowaty 2003). Others have suggested that while
enhanced by the male’s gift. The relative importance of patterns of parental investment may explain which sex
the male’s gift, however, varies with food availability. competes for mates in most animals (i.e., the sex that
invests less in parental care competes more intensely),
it may not explain patterns of reproductive competition
in all animals. Meerkats (Suricata suricatta), for example,
are cooperatively breeding African mammals in which
a single dominant female within a group monopolizes
reproduction. Despite displaying the typical mammalian
pattern of greater parental investment by females than
males (females, after all, are responsible for gestation
and lactation), competition for reproductive opportu-
nities is more intense among females than among males
(Clutton-Brock et al. 2006). Interestingly, female
meerkats that achieve breeder status show changes in
their behavior and morphology that are either not
shown by males or shown to a lesser extent, including
increases in body mass, levels of testosterone, and
aggression. Here, then, is an example of a species in
which patterns of parental investment do not allow us
to correctly predict patterns of intrasexual competition
for breeding opportunities.
Clearly, there is more variation in the reproductive
FIGURE 14.3 A female katydid is eating part of the sper-
matophore deposited by her mate. This protein-rich roles of males and females than previously imagined or
gift constitutes the male katydid’s parental investment. acknowledged. It should not be too surprising, then, that
When food is scarce, the gift has relatively greater with the discovery of diverse mating strategies also came
value, and rather than females choosing mates, males questions about the ideas originally put forth by Bateman
choose females. (1948).
308 Chapter 14 / Reproductive Behavior

REVISITING THE IDEAS OF BATEMAN such as searching for and courting potential mates, we
would be wrong when it comes to the snake Vipera berus.
In recent years, some of Bateman’s ideas and their near
In this species, sperm production and overt reproductive
uncritical acceptance and use by scientists have been
behavior occur sequentially (rather than simultaneously)
challenged (Tang-Martinez and Ryder 2005). The chal-
and are separated in time by an obvious indicator,
lenges include methodological problems with Bateman’s
sloughing of the skin. Importantly, males do not eat dur-
(1948) experiments, subsequent simplification and selec-
ing either period, so the rate at which they lose body
tive use of his data by other scientists, and evidence col-
mass can serve as an estimate of the energetic costs of
lected in the years following publication of his paper that
each period. The rate of mass loss while immobile and
contradicts his ideas on male mating costs and female
producing sperm is at least as great as that during the
reproductive success and behavior. Our focus here will
subsequent period of active reproductive behaviors
be on some of this new evidence.
(searching for females, competing with other males,
courtship, and copulation), and is sometimes greater
(Olsson et al. 1997). There is also evidence from other
Male Mating Costs and Selectivity
vertebrate species that males cannot produce unlimited
Some researchers have questioned the long-held notion numbers of sperm and may, in fact, face sperm depletion.
that males incur trivial costs when producing gametes In male deer mice, sperm counts decrease in successive
(Dewsbury 1982a; Simmons 1988). They are quick to ejaculates, and counts remain depressed for at least one
point out that although, gamete for gamete, sperm are day (Dewsbury and Sawrey 1984). Dominant male Soay
vastly smaller and cheaper to produce than eggs, sperm sheep mate frequently and as a result ejaculate insuffi-
are probably never passed along one at a time. Instead, cient sperm for fertilization (Preston et al. 2001).
millions of sperm are transferred in groups along with Data such as these for invertebrates and vertebrates,
accessory fluids in ejaculates (unpackaged fluids) or sper- collected in the years following the publication of
matophores (packaged fluids). Although the cost of pro- Bateman’s (1948) paper, indicate that there can be sig-
ducing a single gamete may be minuscule, the costs of nificant costs associated with producing sperm and secre-
producing sperm groups and accessory fluids may limit tions from accessory glands, and sperm supplies may be
the reproductive potentials of males. In field crickets limited. Given these findings, we should not be surprised
(Gryllus bimaculatus), for example, the costs of sper- to learn that males are somewhat prudent in their allo-
matophore production (calculated by determining the cation of ejaculates (Wedell et al. 2002), not to mention
percentage of the donor male’s body weight made up by selective in their choice of mates. Indeed, male selectiv-
his spermatophore) are relatively greater for small ity has been reported for many species (e.g., fruit flies,
males than for larger ones. Small males appear to cope Gowaty et al. 2003; great snipe, Saether et al. 2001;
with these higher costs not by producing smaller sper- house mice, Drickamer et al. 2003; and savannah
matophores but by increasing the refractory period baboons, Alberts et al. 2006). We will continue to eval-
between matings (time from when the male attaches a uate Bateman’s ideas in light of new evidence that has
spermatophore to a female until the onset of the next emerged in recent years and will return to mate choice
attempt at courtship, when another spermatophore is later in the chapter.
ready). Thus, for small male field crickets, the costs of
spermatophore production appear to limit their number
of matings (Simmons 1988). Female Behavior and Reproductive Success
Spermatophores also appear costly to males of other As mentioned previously, Bateman (1948) described
invertebrate species. In some butterflies, for example, the females as somewhat passive participants in courtship
costs are reflected in the decreased size and increased water and mating. However, more recent data indicate that
content of successive spermatophores (Ferkau and Fischer females are active participants in reproductive activities,
2006). In one species of stalk-eyed fly, the high costs of often soliciting copulations from one male or several
spermatophore production, particularly those related to different males. Even though glass laboratory vials
secretions from accessory glands, reduce life span in males might not seem like a place that would inspire courtship,
with high mating rates (Pomiankowski et al. 2005). Finally, female fruit flies (Drosophila spp.) often approach and
in some male crayfish, ejaculate size decreases with suc- pursue males when placed together in such vials, and
cessive fertilizations (Rubolini et al. 2007). some show interest in particular males that is equal to
Males of some vertebrate species also appear to or greater than the interest that these males show in
experience costs related to sperm production. Although them (Gowaty et al. 2003). Also, Drosophila females may
most of us might assume that remaining immobile and signal their enthusiasm for a male and willingness to
building up sperm supplies would be less energetically mate with him by becoming immobile, but this in no
expensive than engaging in active reproductive behaviors way should be interpreted as “passive” behavior on the
 Intrasexual Selection—Competition for Mates 309

Number of eggs laid 80

Females that
Finally, although Bateman (1948) found lower vari-
mated with 3 males ance in reproductive success among female than male
60
fruit flies, there is now evidence that variance in the
Females that reproductive success of females is not always lower than
40 mated with 1 male
that of males. For example, variance in reproductive suc-
20 cess is greater in females than in males for red sea urchins
(Levitan 2005) and meadow voles (Sheridan and Tamarin
0
10 15 20 25 30 1988). It has also been suggested that we might expect
Female wet weight (mg) greater variance in female than male reproductive suc-
cess in cooperatively breeding mammals in which social
FIGURE 14.4 Fecundity of female fireflies increases with groups contain several nonbreeding adults and females
number of mates, when controlled for female body experience greater reproductive suppression than males
weight. Here, females that mated with three males (Hauber and Lacey 2005). These examples show that
produced more eggs than females that mated with one
there are exceptions to the rule that variance in female
male. (Modified from Rooney and Lewis 2002.)
reproductive success is lower than variance in male
reproductive success, and of course the interesting ques-
tion for such exceptions is “Why?”
part of females. Bonobos (Pan paniscus), sometimes With this introduction to some of the historical and
described as “our other closest relatives” because they theoretical background for sexual selection, and recent
are in the same genus as chimpanzees (Pan troglodytes), challenges to some of these ideas, we will consider
exhibit interesting social and sexual dynamics, to say the intrasexual selection, and then intersexual selection, in
least. Female bonobos are dominant to males and mate more detail.
promiscuously (and publicly) with males from within
their community and with males from other communi-
ties (Parish and de Waal 2000). Thus, more recent
observations from species as diverse as fruit flies and INTRASEXUAL SELECTION—
bonobos indicate that females are anything but passive COMPETITION FOR MATES
in mating interactions.
Bateman’s (1948) prediction that female reproduc- Intrasexual selection has led males to evolve a broad
tive success should not increase with number of mates spectrum of attributes related to intense competition for
has also been challenged. We first consider an example mates. Some mechanisms operate prior to copulation,
in which female reproductive success increases with whereas others have their effect once mating has
number of mates. In fireflies (Photinus ignitus), females occurred. We will now consider a few of the many
that mate with three males have higher fecundity (num- tactics males use to gain a competitive edge in the mat-
ber of eggs) than those that mate with a single male ing game.
(Figure 14.4) (Lewis et al. 2004; Rooney and Lewis
2002). Why should this be the case? During mating,
males transfer a spermatophore to the female. Sperm at ADAPTATIONS THAT HELP A MALE
the anterior end of the spermatophore are released into SECURE COPULATIONS
the female’s spermatheca (sperm storage organ) to be
used later. The remainder of the spermatophore passes Dominance Behavior
to the spermatophore-digesting organ where it will Males in some species may secure copulations by dom-
break down over a few days, freeing nutrients for use by inating other males and thereby excluding competitors
the female. (Thus, unlike the katydids discussed earlier, from females. This behavior, of course, places a certain
females can get nutrients from the spermatophore sim- premium on greater male strength and more effective
ply by absorption rather than eating it.) Photinus ignitus weaponry. Because the female makes no such invest-
females use the protein from the digested spermatophore ments, males tend to diverge from females in both
to fuel development of their eggs (specifically the pro- appearance and aggressiveness.
duction of yolk). Because these fireflies feed as larvae but In many birds and mammals, males are larger than
not as adults, the nutrients gained from the sper- females, presumably because large body size improves
matophore supplement the female’s larval energy the fighting ability and hence the reproductive success
reserves. Thus, we see that P. ignitus females benefit from of males. It can be expected that the greatest levels of
additional matings because such matings supplement sexual dimorphism (difference in the appearance and
their declining larval energy reserves and lead to greater behavior of the sexes) will be found in species in which
egg production. the competition between males is most intense.
310 Chapter 14 / Reproductive Behavior

This is, in fact, often the case, but sexual dimor- SSD results from an ecological constraint on male body
phism can result from other factors as well (Isaac 2005). size (the ability to transport shells depends on large body
In many organisms, the effect of female body size on size), and (4) SSD results from an ecological constraint
female reproductive success is equally important in on female body size (the ability to enter and breed in
determining which sex is more “built to last” than the shells depends on small body size). The specific predic-
other. Also, although ecological factors usually affect tions associated with each hypothesis, general testing
males and females in a similar manner, sexual dimor- methods, and the results are shown in Table 14.1. As you
phism in body size sometimes results from sex differ- can see from the table, Schütz and Taborsky found
ences in predation pressures or resource use. Indeed, strong support for hypotheses 3 and 4. Thus, it appears
some sexually dimorphic traits may be caused by diver- that the ecological constraint of shell size influences the
gence in food use. Some of the best examples of this body sizes of males and females in a divergent manner,
concern feeding structures rather than overall body size. favoring males large enough to carry shells and females
Males and females of certain species of mosquito have small enough to fit into shells. The findings also sug-
different mouthparts, and this dimorphism is related to gest that intrasexual selection and intersexual selection
the nectar diet of males and the blood diet of females have relatively minor and indirect effects on the evolu-
(Proctor et al. 1996). Similarly, sexual differences in bill tion and maintenance of sexual size dimorphism in
lengths and bill curvatures of the purple-throated L. callipterus. For example, field observations revealed that
carib, a species of hummingbird from the West Indies, large males do not have larger nests than small males,
also are caused by a divergence in food use; males and and laboratory data indicated that large and small males
females sip nectar from flowers of different lengths and do not differ in the speed with which they obtain nests.
curvatures on different plant species (Temeles et al. However, large males tend to hold nests longer and in
2000). Thus sex differences in appearance are some- the end sire more offspring. We see from this study that
times caused by sexual selection and other times by eco- carefully conducted field observations and laboratory
logical factors. experiments can tease apart the influences on body size
How would you determine whether a difference in of sexual selection and ecological factors, and that some-
body size between the sexes is caused by sexual selec- times several factors can influence body size, and in dif-
tion or ecological factors? As an example of how you ferent ways.
might go about making such a determination, we’ll con- In species in which males form dominance rela-
sider the case of the cichlid fish Lamprologus callipterus, tionships, high-ranking males typically engage in more
a species endemic to Lake Tanganyika, East Africa. sexual activity than low-ranking males and experience
Males of this species are a whopping 12 times heavier greater reproductive success (Dewsbury 1982b; Ellis
than the females with which they spawn. To understand 1995). In their attempts to monopolize sexual access to
the factors that influence the extreme sexual size females, dominant males sometimes interfere directly
dimorphism in L. callipterus, we must first know some- with the copulations of subordinate males. Daniel Estep
thing about the species’ life history. Territorial males and colleagues (1988) examined interference by domi-
guard nests of empty snail shells; they inherit shells from nant males and the consequent inhibition of sexual
previous nest owners, collect some unclaimed shells behavior among subordinate males of the stumptail
on their own, and steal shells from the nests of neigh- macaque (Macaca arctoides). Long-term observations of
bors. At first glance, this male obsession with empty a captive colony of this Old World primate revealed the
snail shells might seem strange, but it is perfectly existence of a stable, linear dominance hierarchy
understandable in light of the observation that female among adult males. Dominant males monopolize
L. callipterus will only breed in empty shells. Females access to females by inhibiting subordinate males
arrive on the territory of a male, enter an empty shell, through aggressive threats and disruption of mating
spawn, and then care for their brood within the shell for attempts. Within the dominance hierarchy, copulation
about two weeks. In an effort to determine the influence frequencies were rank-related; the alpha (highest-
of sexual selection and ecological constraints on the ranking) male had almost as many copulations as the
extreme sexual size dimorphism in L. callipterus, Dolores other 17 adult males combined (Table 14.2). Within
Schütz and Michael Taborsky (2005) directly observed this group, the three lowest-ranking males failed to
the behavior of free-living males and females in Lake mate even once during the 34-week study. If low-
Tanganyika and conducted experiments in aquaria in the ranking males mated at all, they usually did so surrep-
laboratory. The researchers tested the following four titiously, out of view of dominant males. When tested
nonexclusive hypotheses concerning sexual size dimor- in the absence of dominant males, low-ranking males
phism (SSD) in L. callipterus: (1) SSD results from intra- copulated far more frequently than they did in their pres-
sexual selection (male–male competition), (2) SSD ence. (Dominance relationships are discussed further in
results from intersexual selection (female choice), (3) Chapter 18.)
 Intrasexual Selection—Competition for Mates 311

TABLE 14.1 Testing Whether Sexual Size Dimorphism in a Cichlid Fish Results from Sexual Selection
or Ecological Constraints
Hypothesis Prediction Testing condition Prediction supported?

(1) Intrasexual selection (1a) Because of enhanced Field observations No

competitive abilities, large
males have larger nests than
small males

(1b) Large males acquire nests Laboratory experiment in No/Yes: both large and small
more rapidly and hold onto which two different sized males acquire nests rapidly, but
nests longer than do small males directly competed for large males are more aggressive,
males nests and females hold nests longer, and as a
result sire more offspring

(2) Intersexual selection (2a) Females prefer larger or Field observations No

heavier males Laboratory experiment in No
which females were allowed
to choose between two males
that differed in standard
length and condition
(weight/standard length3)

(2b) Females choose males Field observations Yes

with larger nests Laboratory experiment in No
which females were allowed
to choose between two males
that differed in the number
of shells in their nest (5 shells
versus 15 shells)

(3) Ecological constraints of (3a) Males must reach a Laboratory experiment in Yes
male body size minimum size to carry shells which males were presented
with shells of different sizes
and observed to see whether
they could carry the shells
to their nest

(3b) Larger males carry shells Same experiment as directly Yes

more efficiently above

(4) Ecological constraints of (4a) Females prefer large Laboratory experiment in Yes
female body size shells for spawning which 3–4 females were
placed in a tank with a
territorial male and 10 shells
of different sizes

(4b) Female reproductive Same experiment as directly Yes

success increases with relative above
shell size

(4c) Females adjust their Laboratory experiment in Yes

growth to available shell sizes which two groups of 6 small,
wild-caught females were
housed with a territorial male
and either 10 large shells or
10 small shells

Modified from Schütz and Taborsky (2005).

312 Chapter 14 / Reproductive Behavior

TABLE 14.2 Patterns of Copulation in Relation or cuckholders (Gross 1982; Neff et al. 2003; Stoltz and
to Dominance in a Captive Colony of Stumptail Neff 2006). Parental males become sexually mature around
Macaques 7 years of age. During the breeding season, these males
compete for sites at which to build their nest, guard females
Male
that enter their nest to spawn, and then provide sole care
dominance Type of Total
to the developing young. Cuckholders, on the other hand,
rank copulation copulationsa
are precocial males, maturing at 2 years of age to begin
Alpha male Visible 424 their life of stealing fertilizations from parental males.
Surreptitious — Cuckholders do not build and defend nests, nor do they
14 Other adult males b
Visible 69
provide parental care. When small (2 to 3 years old), they
are termed “sneakers” because they hide in vegetation at
Surreptitious 393
the edge of a parental male’s nest and then dart into the nest
a
Per 1050 hours of observation. to release sperm at the time the parental male is spawning
b
Of 18 males in the colony, 3 failed to secure a copulation. with a female. Although parental males actively defend
Source: Estep et al. (1988). their nests against sneakers, they are not always successful
in keeping the small males out. Once cuckholders reach a
substantial body size at about 4 to 5 years of age, they
STOP AND THINK become female mimics (sometimes called “satellites”), tak-
Spritzer et al. (2006) studied the relationship between ing on the coloration and behavior of mature females.
social dominance of male meadow voles (Microtus pennsyl- Female mimics succeed in entering the nests of parental
vanicus) and reproductive success. They first tested males males, who presumably are deceived by their feminine ways
for dominance in neutral arenas in the laboratory and then and attire. Once inside the nest, the female mimic positions
released them into field enclosures. Each of eight enclo- himself between a mature female and the parental male, all
sures received four males (two with the highest dominance the time engaging in behavior typical of an egg-bearing
ranks and two with the lowest ranks from the original female (e.g., slow dipping movements and exaggerated rub-
groups of eight males) along with four females, each of bing of the side of the male). At the time of spawning, both
which was confined to a wooden nest box. The nest boxes
the parental male and the female mimic release sperm.
had one-way doors so that males could enter and mate, but
females could not leave; the boxes were checked twice daily
Alternative male mating tactics have also been doc-
to release males, and females were checked weekly for umented in species of amphibians in which males employ
pregnancy. Pregnant females were brought into the labo- loud (and energetically expensive) calls to attract females.
ratory to give birth and to check the paternity of their off- Although all males may be sexually mature and capable
spring. Females removed to give birth in captivity were of calling, some remain silent and associate closely with
replaced with new females; males that disappeared from a calling male, ready to intercept females attracted to the
enclosures were replaced with males of known dominance calls of the other male. The silent males are typically
rank from their original group of eight. Over the course called satellite males. We’ll take a look at callers and
of several weeks, Spritzer and colleagues monitored voles satellites in the natterjack toad.
in the enclosures through live trapping. Their data from On spring evenings, male natterjack toads (Bufo
trapping and paternity analyses revealed that meadow vole
calamita) gather at the edges of temporary ponds and call
males with high-dominance rank had larger home ranges
but sired significantly fewer litters than males with low-
dominance rank. How might you explain these results?
What questions would you want to ask the researchers
about the details of their study methods and results?

Alternative Reproductive Strategies

In keeping with the old adage “out of sight, out of
mind,” subordinate stumptail macaques cope with
dominant males by copulating on the sly. This question
arises, then: how do males of other species deal with
large, dominant males? Males of many species use alter-
native reproductive strategies. We begin by describing
the breeding antics of bluegill sunfish, a species with
external fertilization (fertilization occurs in the external FIGURE 14.5 A male natterjack toad calls for females at
environment and not the female reproductive tract). the edge of a pond. Calling males compete with other
Male bluegill sunfish (Lepomis macrochirus) display a males, called satellite males, which crouch silently
discrete life history polymorphism and are either parentals nearby and intercept females on their way to the pond.
 Intrasexual Selection—Competition for Mates 313

loudly to attract females. Callers typically adopt a head-

up posture and energetically belt out their call (Figure
14.5). Satellites, on the other hand, keep a low profile
and remain stationary and silent in a crouched position
next to a calling male. If a female is attracted to the caller,
a satellite male will make every effort to intercept and
clasp her. In natterjack toads, body size and call inten-
sity are highly correlated with mating success; in general,
small males with weak calls are at a reproductive disad-
vantage. Not surprisingly, the consistency with which
males in a population adopt the calling or satellite tac-
tic is correlated with their body length: small males tend
to be satellites; large males tend to be callers; and inter-
mediate males are switch hitters, alternating oppor-
tunistically between the two strategies (Arak 1988).
FIGURE 14.6 Male amphipods guard females during the
precopulatory period, just before the female molts. By
maintaining constant contact with a premolt female, the
ADAPTATIONS THAT FAVOR
male ensures that his sperm rather than the sperm of a
THE USE OF A MALE’S SPERM rival are used in fertilization.
Mate Guarding
In addition to the option of adopting alternative repro- have gone a long time without female social contact will
ductive strategies, males display a host of adaptations accept them (Dunham and Hurshman 1990). Once in the
that increase the probability that their sperm, and not the precopula position, the two swim as a single unit until the
sperm of a competitor, will fertilize the eggs of a partic- female molts and sperm transfer is complete. Thus, the
ular female. An obvious strategy is to be present when male has ensured that his sperm, rather than the sperm of
eggs can be fertilized and prevent other males from gain- a competitor, are used in fertilization.
ing access to the female during this time. Depending on In some cases, a male may protect his reproductive
whether the female stores sperm between copulation and interests by guarding his mate after copulation as well.
fertilization, mate guarding can occur before or after If a female were to mate with more than one male, the
copulation, or both. first male to copulate with her would presumably have
In some species, including many crustaceans, mating a reduced chance of fertilizing her eggs, as a result of
is often restricted to a short period of time after the female sperm competition with the second male (see later dis-
molts, and females don’t store sperm. Consequently, the cussion). Male blue crabs (Callinectes sapidus) guard
first male present fertilizes all the eggs. In such species, females before and after mating. A male guards a female
we generally find precopulatory mate guarding (Birkhead before she molts so that he will be present when her eggs
and Parker 1997). The tiny amphipod Gammarus lawren- can be fertilized. For the record, a male blue crab’s ver-
cianus (Figure 14.6), for example, lives in estuaries along sion of guarding his mate is to carry her beneath him,
the coast of North America. Typically, a free-swimming with his sternum against her dorsal carapace; once she
male grabs a passing female before she molts and draws has completed molting, he flips her over such that her
her to his ventral surface. Once the male has a firm hold abdomen is against his and copulation can begin. But
on the female, he moves her in such a way that the long about 12% of the females mate again within a few days
axis of her body is at right angles to his. In this posture of the final molt. A female that mates more than once
the male palpates the female with various appendages, and stores all the sperm from both males. Thus, sperm from
after a minute or so, rotates her into the precopula posi- both males have equal access to her eggs. So, to protect
tion in which her body is now underneath and parallel to his interests, a male blue crab also guards his mate after
his. It is during the period of palpation and inspection that copulation. If many other males are present, a male
mate-guarding decisions are made. One factor in the deci- guards a female for a longer time after copulation than
sion is whether the female has already mated. Males are he does if few males are present (Jivoff 1997).
more likely to guard females with an empty brood pouch Male Gammarus lawrencianus and Callinectes sapidus
than those with a brood pouch full of juveniles or recently guard females by holding onto them. In other animals,
fertilized eggs (Dunham 1986). A second factor is the however, guarding males stay close to females but do not
length of time the male will have to guard the female maintain direct physical contact with them. Especially
before mating. Although females early in their reproduc- in species in which males employ the “stay close by”
tive cycle are also typically rejected (presumably because method, how can we distinguish mate guarding from
of the greater time investment in guarding), males that some other behavior involving close association, such as
314 Chapter 14 / Reproductive Behavior

when individuals stay close to one another to detect or a 120

deter predators? Pamela Willis and Lawrence Dill 100

duration (min)
Association
(2007) followed Dall’s porpoises (Phocoenoides dalli) in 80
the coastal waters off southwestern Canada to determine 60
whether mate guarding occurs in this species. Female 40
Dall’s porpoises give birth to a single offspring each sum- 20
0
mer and shortly thereafter enter estrus. Most females
with newborn calves are accompanied by a male, but does

distance (body lengths)

this association represent mate guarding by the male? b 10

Mean inter-animal
Willis and Dill observed 18 male–female pairs during the 8
weeks following parturition and compared their behav- 6
ior while at the surface with that of 24 associations involv- 4
ing two males (hereafter called male–male pairs). The
2
researchers made numerous predictions, including the 0
following: if the association between males and females
does indeed represent mate guarding, then when com-
pared to members of male–male pairs, members of c 0.9
male–female pairs will (1) maintain shorter distances

Proportion of
synchronous
surfacings
between one another, (2) remain in proximity to one 0.6
another for longer durations, (3) be more likely to sur-
0.3
face in synchrony, and (4) be less likely to join other indi-
viduals or groups. Perhaps most importantly, Willis and
0
Dill predicted that if males were guarding females, then Male–male Male–female
such males would respond more aggressively than males Type of association
traveling with another male to approaches by adult males.
Their results provide strong behavioral evidence that FIGURE 14.7 Behavioral evidence that male–female
male Dall’s porpoises guard females during the period fol- associations (pairs) of Dall’s porpoises during the period
lowing parturition when females are in estrus. As you can of female receptivity represent mate guarding. Here,
see in Figure 14.7, members of male–female pairs stayed (a) the duration of such associations, (b) the distance
between members of the associations, and (c) the
together longer, maintained shorter distances between
proportion of synchronous surfacings by association
one another, and were more likely to surface at the same
members are compared with those of male–male
time than were members of male–male pairs. associations. (Modified from Willis and Dill 2007.)
Male–female pairs were also less likely than male–male
pairs to join other individuals. Finally, also as predicted,
when paired with a female, males were more likely to clutch with a single egg, and males invest heavily in the
respond aggressively to approaches by other adult males. care of the long-dependent youngster. With restricted
Although all research has its challenges, working with opportunities to reproduce and substantial parental invest-
cetaceans (whales, porpoises, and dolphins) is extremely ment, it presumably pays for a male to keep a close eye on
difficult given their aquatic existence, wide dispersal, high his mate to ensure that the nestling he will be feeding for
mobility, and typically large body size. Because of these the next four months is his own. Previous studies on the
challenges, behavioral observations of wild individuals are effectiveness of mate guarding by male birds have typically
often limited to surface sightings, and genetic analysis of removed the male from a pair and then monitored the
parentage to determine the effectiveness of mate guard- occurrence of extra-pair copulations or extra-pair fertil-
ing seems extremely difficult, if not logistically impossi- izations involving the experimentally “widowed” female.
ble, in natural populations. So, we will turn to a smaller, One problem with this approach is that the female may
more terrestrial creature in an effort to answer the ques- dramatically change her behavior in response to her mate’s
tion of whether mate guarding is effective. More specifi- disappearance, perhaps stepping up efforts to pair with a
cally, we will examine a study that addressed the question, new male or males. So, instead of removing the male from
“Does mate guarding reduce extra-pair fertilizations?” a pair, Komdeur and colleagues left males with their mates
Jan Komdeur et al. (2007) examined the effectiveness but induced them to stop mate guarding. The researchers
of mate guarding in reducing extra-pair fertilizations in the took advantage of the fact that a male Seychelles warbler
Seychelles warbler (Acrocephalus sechellensis). Males of this abruptly stops guarding his mate once he spots an egg in
species closely follow their mate during the five or so days their nest. The researchers placed a model egg in the nest
prior to egg laying, the time when the female is most of 20 experimental pairs during the fertile period of the
receptive. Such guarding seems to make sense given that female, but before egg laying, and thereby induced males
each breeding season a pair typically produces a single in these pairs to stop mate guarding (the cessation of mate
 Intrasexual Selection—Competition for Mates 315

guarding was confirmed by behavioral observations). a

These nests and adults were monitored along with
another 20 control nests without model eggs. Of the 20
control nests, 9 eggs hatched, 1 failed to hatch, and preda-
tors took 10. Of the 20 experimental nests, 8 eggs hatched,
2 failed to hatch, and predators took the remaining 10
eggs. Parentage analysis revealed that of those eggs that
hatched in the two groups, 44% of control nestlings were
not sired by the female’s social mate (male with which she
shared a territory and nest) as compared to 75% of
nestlings in the experimental group in which males were
induced to stop mate guarding. Thus, in the Seychelles
warbler, male mate guarding reduces the risk of cuckoldry.
We’ll have more to say about the generally high level of
extra-pair fertilizations in birds in Chapter 15. b

Mechanisms to Displace or Inactivate

Rival Sperm
Although physical struggles among males for females are
often conspicuous, perhaps the fiercest clashes related to
mating occur more quietly—in the female reproductive
tract. If the female has mated with more than one male,
their sperm must compete for the opportunity to
fertilize her eggs; this phenomenon is called sperm
competition. For example, the quality of a male’s
sperm—measured, perhaps, by their motility or the num-
ber of viable sperm per ejaculate—may determine the
FIGURE 14.8 Male damselflies remove the sperm left by
proportion of young that he sires. Another factor that
rivals. (a) A copulating pair of damselflies. (b) The penis
may come into play is his position in the line of suitors of a male damselfly serves not only to transfer sperm to
to mate with a specific female. In some species, the advan- the female but also to remove sperm previously left by
tage goes to the first male to mate with the female, in oth- competitors; backward-pointing hairs on the horns of
ers the last male, and in still others mating order does not the penis remove clumps of rival sperm.
seem to be an important factor in determining patterns
of paternity. Geoff A. Parker (1970, 1984) pioneered
research in the area of sperm competition and described a male of another species of damselfly stimulates the
the phenomenon as a “push-pull” relationship between female sensory system that controls egg laying and fer-
two evolutionary forces—one that acts on males to dis- tilization, causing her to release sperm from her sperm
place previous ejaculates left by rivals, and one that acts storage structure. This allows him to gain access to rival
on early males to prevent such displacement. sperm that would otherwise be unreachable (Cordoba-
In some cases, the interference is rather crude—the Aguilar 1999). A similar strategy is used by male dun-
male simply removes rival sperm. For example, some male nocks, which are small songbirds. If the male sees his
damselflies use their penis not only to transfer sperm but mate near another male, he pecks at the female’s cloaca,
also to remove sperm previously deposited by competitors causing her to eject sperm-containing fluid (Davies 1983).
(Waage 1979). Backward-pointing hairs on the horns of Another phenomenon that is often linked to sperm
the damselfly’s penis appear to aid in scooping out clumps competition is sperm heteromorphism (the simultaneous
of entangled sperm left by earlier rivals (Figure 14.8). production by a single male of at least two types of sperm
Some male crustaceans employ equally subtle tactics. in the same ejaculate) (Holman and Snook 2006). In
Rather than scooping out their rival’s sperm, male spider most cases of heteromorphism, there are two sperm
crabs push the ejaculates of earlier males to the top of the morphs, one that can fertilize eggs and one that cannot.
female’s sperm storage receptacle; seal them off in this new The two morphs are called eusperm and parasperm,
location with a gel that hardens; and then place their own respectively. Sperm heteromorphism occurs in diverse
sperm near the female’s oviduct, the prime location for animals, including many species of mollusks, insects, and
fertilization to occur (Diesel 1990). fishes (see papers reviewed by Hayakawa 2007). But what
Males of other species stimulate the female into is the function of infertile sperm, and what is their con-
ejecting the sperm of another male. During copulation, nection with sperm competition?
316 Chapter 14 / Reproductive Behavior

TABLE 14.3 Hypotheses for the Evolution of Parasperm (Infertile Sperm)

Hypothesis Description

Nonadaptive Parasperm have no adaptive function

Sperm competition Parasperm help eusperm compete to fertilize eggs
Offensive Parasperm displace or kill rival sperm
Defensive Parasperm prevent rival sperm from reaching the eggs or inhibit remating by the female
Provisioning Parasperm, upon degeneration, provide nutrition to the female, her eggs, or eusperm
Facilitation Parasperm facilitate transport or readiness of eusperm
Sacrificial sperm Parasperm dilute the effects of spermicides, thereby increasing survival of eusperm
Cryptic female choice Parasperm increase the chances that eusperm will be preferred during cryptic female choice

Modified from Holman and Snook (2006).

There are several hypotheses for the evolution of ther matings (Cook and Wedell 1999; Silberglied et al.
infertile sperm (Table 14.3). First, it is possible that 1984). Finally, in some fish that employ external fertil-
parasperm have no adaptive function. Most consider this ization, there is evidence that parasperm function in at
unlikely given the large number of parasperm produced least two ways (Hayakawa 2007). Parasperm increase the
(parasperm may make up half of an ejaculate) and the distance that semen travels at spawning, thereby increas-
presumed costs of their production. Alternatively, ing the number of eusperm reaching an egg mass. Then,
parasperm could have one or more adaptive functions. upon arrival at the egg mass, parasperm form defensive
For example, parasperm may play a role in sperm com- lumps at the outskirts of the mass that prevent sperm
petition. Such a role could be offensive (for example, from other males from reaching the eggs.
killing or displacing sperm from another male) or defen-
sive (for example, inhibiting remating by the female or
blocking sperm from another male). Other potential Mechanisms to Avoid Sperm Displacement
adaptive functions of parasperm include their ability to How can males ensure that their sperm are not displaced
facilitate the transport or readiness of eusperm or to pro- by those of rival males? We’ll consider the case of Rocky
vide (upon their degeneration in the female reproduc- Mountain bighorn sheep (Ovis canadensis canadensis), a
tive tract) nutrients to the female, her eggs, or the species in which females usually mate with several males
eusperm. Holman and Snook (2006) provide two addi- during a single period of estrus. Sperm competition in
tional adaptive hypotheses: (1) parasperm may perform this species is intense because subordinate males (called
a sacrificial function by diluting the effects of spermicides coursing rams) are remarkably successful in forcing cop-
produced by females and thereby promoting survival of ulations on ewes guarded by tending rams (Hogg 1988;
eusperm (the authors define spermicides as anything that Hogg and Forbes 1997). Tending and coursing rams cop-
kills, disables, or dumps sperm), or (2) parasperm could ulate with estrous ewes at extremely high rates. Frequent
increase the chances that eusperm will be selected dur- copulations by individual male bighorns presumably
ing cryptic female choice (choice by a female who has function to increase the proportion of their own sperm
mated with more than one male regarding which sperm in the female reproductive tract. Tending rams copulate
will fertilize her eggs). In terms of the cryptic female at especially high rates immediately after successful cop-
choice hypothesis, it is possible, for example, that females ulation by a coursing ram. Such “retaliatory” copulations
assess males by the number or quality of parasperm they by dominant males (Figure 14.9) may serve a function
produce. We will discuss cryptic female choice in more akin to that of the damselfly penis—mechanical dis-
detail later in the chapter. placement of rival sperm. Alternatively, retaliation may
At present, the functions of parasperm remain be advantageous to the dominant male because sperm
unknown for most species. Notable exceptions include from the last mating are more likely to fertilize eggs.
some species of Lepidoptera (moths and butterflies) in Some males mate longer than is necessary for the
which parasperm facilitate readiness of eusperm for par- release of sperm. Males may use prolonged mating to
ticipation in fertilization (Sahara and Takemura 2003). avoid or reduce competition with the sperm of other
Parasperm also appear to function in sperm competition males. Two hypotheses have been suggested to explain
in Lepidoptera. The functional equivalent of “cheap the function of prolonged mating. The first, called the
filler” parasperm may delay further matings by females. extended mate-guarding hypothesis, suggests that pro-
More specifically, distention of a female’s sperm storage longed mating reduces the chances that the female will
organ by parasperm could signal a successful insemina- be inseminated by another male. The second, the ejac-
tion and thereby reduce the female’s receptivity to fur- ulate transfer hypothesis, suggests that prolonged mat-
 Intrasexual Selection—Competition for Mates 317

allowed to mate did not, however, influence the number

of sperm found in the female’s spermathecae. These data
on sperm counts thus fail to support the ejaculate trans-
fer hypothesis. In summary then, prolonged copulation
in Nephila clavipes does not result in the transfer of more
sperm, but it does function to reduce the likelihood of
the female mating with another male.

Repellents and Copulatory Plugs

In an effort to reduce the likelihood of future matings
by competitors, males of many taxa apply a repellent
odor to their mates. In the case of a neotropical butter-
fly, a male transfers an “antiaphrodisiac” pheromone to
FIGURE 14.9 Competition for mates is intense among his mate during copulation, which makes her repulsive
males of Rocky Mountain bighorn sheep. Although to future suitors (Gilbert 1988).
dominance and access to mates are established by fierce In other species, a male may deposit a copulatory
physical clashes, male competition continues in the plug made of thick, viscous secretions in the female’s
reproductive tracts of females. Dominant males copu- reproductive tract. Copulatory plugs occur in many ver-
late at high rates immediately after successful copula- tebrates, including snakes, lizards, marsupials, rodents,
tion by a subordinate male. Such retaliatory copulations
bats, and primates. Several functions have been suggested
may displace the sperm of the subordinate male.
for copulatory plugs, among them: (1) “enforcing
chastity,” with the plug acting as a barrier to subsequent
ing results in more of the male’s sperm being transferred inseminations (Martan and Shepherd 1976; Voss 1979);
to the female. Christopher Linn and colleagues (2007) (2) ensuring the retention of sperm in the female repro-
tested these two alternative hypotheses in Nephila ductive tract (Dewsbury 1988); (3) aiding the transport
clavipes, an orb-weaving spider with a penchant for cop- of sperm within the female reproductive tract
ulation. In the two days following a female’s molt to (Carballada and Esponda 1992); (4) providing for the
adulthood, the largest male on her web copulates with gradual release of sperm as the plug disintegrates
her several hundred times, with individual copulations (Asdell 1946); and (5) providing a means by which a male
lasting up to 14 minutes. Smaller, peripheral males on can scent-mark a female’s body and convey information
the web hang around the preoccupied couple, but rarely regarding his identity and dominance status (Moreira et
are successful in gaining copulations. At the end of two al. 2006). For this last proposed function, the informa-
days, the mating comes to an end and the female even- tion conveyed in the plug might be used by the male
tually abandons the web. To ascertain the advantage of himself (for example, to identify females with which he
prolonged copulation in this species, Linn et al. (2007) has already mated), by rival males, or by the female (for
allowed members of male–female dyads to mate for one example, during cryptic female choice). Unfortunately,
of the following three durations: (1) two natural copu- for the vast majority of species the precise function of
lations, each lasting about 14 minutes, (2) 2 hours, dur- copulatory plugs remains a mystery (Ramm et al. 2005;
ing which numerous copulations occurred, or (3) 48 Reeder 2003). Nevertheless, researchers continue to
hours (the natural mating period following the female’s make intriguing observations—for example, males and
molt). In each of the three treatments, some females females of some species remove copulatory plugs—that
were used to quantify the number of sperm that had been may someday shed light on copulatory plug function(s).
transferred to them by the male; sperm storage organs Placing plugs in the female reproductive tract is not
(spermathecae) were removed from females, and the confined to male vertebrates. Indeed, some male spiders
sperm within each organ were counted. The remaining intent on protecting their paternity employ quite dra-
females in each treatment were presented with a second matic techniques to plug the reproductive tracts of
male four days after the molt to determine their likeli- females with which they have mated. A male spider, by
hood of mating again (termed remating by the authors). the way, copulates by inserting his pedipalps (paired
The data show that the longer a female was allowed to extremities posterior to the fangs that serve as copula-
copulate with the first male, the lesser the likelihood of tory organs) into the genital opening(s) of a female (in
her mating with a second male (percent of encounters in some groups of spiders, the genital openings of females
which a female mated with a second male: 93% in the are paired, while in others they are not). In Argiope
two-copulation treatment; 90% in the 2-hour treatment; aurantia, within seconds of inserting his second pedipalp
and 28% in the 48-hour treatment). These data on like- into the female’s genital opening, the male becomes
lihood of remating support the mate guarding hypoth- unresponsive and dies shortly thereafter, becoming, in
esis. The duration that the female and first male were essence, a whole-body mating plug (Foellmer and
318 Chapter 14 / Reproductive Behavior

SEXUAL INTERFERENCE: DECREASING

THE REPRODUCTIVE SUCCESS OF
RIVAL MALES
As we have seen, mating success is often measured by how
well one advances one’s own reproductive efforts and how
effectively one interferes with a competitor’s efforts. We
will now concentrate on the latter. Any behavior that
reduces a rival’s fitness by decreasing his mating success
is called sexual interference (Arnold 1976).
FIGURE 14.10 Some male spiders leave parts of their Some of the most effective animals in sexual inter-
copulatory organs in the genital tracts of females with
ference are male newts. Adrianne Massey (1988) stud-
which they have mated. Such parts appear to obstruct
matings by subsequent males. This image from a scan-
ied sexual interactions in field populations of
ning electron microscope shows parts of a male’s pedi- red-spotted newts (Notophthalmus viridescens) and noted
palps (indicated by the letters E, for embolus, and C, for three tactics of sexual interference: (1) spermatophore
conductor) that were left in each of the two genital transfer interference, (2) pseudofemale behavior, and (3)
openings of a female. amplexus interference. These are all methods through
which male newts decrease the reproductive success of
competing males.
Fairbairn 2003). In a slightly less dramatic manner, cop- The first method, spermatophore transfer interfer-
ulating males of the golden orb web spider Nephila fen- ence, occurs when a rival male inserts himself between
estrata protect their paternity by leaving parts of their a female and the courting male that has just dismounted.
pedipalps in the genital openings of females (Figure The rival male not only induces spermatophore deposi-
14.10). These pedipalp fragments obstruct copulations tion by the first male but also slips his own sper-
by subsequent suitors (Fromhage and Schneider 2006). matophore into position so that the female picks that up
So far in our discussion of sperm competition, we rather than the spermatophore of the first male. This
have focused on interactions between males. Do females form of sexual interference reduces the first male’s sup-
benefit by mating with more than one male? Multiple mat- ply of spermatophores for future inseminations and pre-
ings may, in fact, be to a female’s advantage. Mating with vents him from inseminating the female (as well as
several males may (1) increase the probability of fertiliza- permitting the intruding male to inseminate her). The
tion, (2) increase the genetic diversity of offspring, (3) second method, pseudofemale behavior, also causes
result in the accumulation of material benefits if males pro- courting males to waste spermatophores, but this time
vide nutritional gifts at copulation, or (4) ensure that a in the context of male–male pairings (Figure 14.11). Male
female’s sons are good at the game of sperm competition, red-spotted newts often clasp other males, although such
if the trait is heritable. These and other suggestions are pairings are usually brief because clasped males give a
reviewed in more detail by Møller and Birkhead (1989). head-down display that elicits release. In some cases,

FIGURE 14.11 There are few differences in heterosexual and

homosexual courtship in newts. (a) A receptive female has
straddled the tail of a male to stimulate the male to deposit a
spermatophore. (b) A male mimics female behavior to cause
the courting male to deposit a spermatophore. Pseudofemale
behavior is one form of sexual interference used by male
newts to decrease the reproductive success of competitors.
 Intrasexual Selection—Competition for Mates 319

reported for diverse invertebrate and vertebrate taxa

(Ebensperger and Blumstein 2007; Hausfater and Hrdy
1984; Trumbo 2006; Van Schaik and Janson 2000). In
some mammals, sexually selected infanticide occurs
when one or more males from outside the social group
usurp the resident male. In Hanuman langurs, the pri-
mate species in which infanticidal male intruders were
first reported, infanticide occurs when individuals from
all-male bands invade harems (Sugiyama 1965). With
her offspring gone, the female quickly returns to estrus;
in Yukimaru Sugiyama’s (1984) words, infants are
attacked because they are “little more than an obstacle
to activation of the mother’s receptivity.”
Although most primatologists accept the hypoth-
esis that male infanticide in Hanuman langurs is sex-
ually selected, acceptance is not universal (Brown
1996; Dixson 1998). Nonetheless, supporting evidence
is slowly accumulating. One criticism was that the acts
FIGURE 14.12 Although infanticide may serve several
of killing were not actually observed and were only
functions, it is a dramatic example of sexual interfer-
assumed to have been committed by the males that
ence. Here, a lion has killed the cub of a rival male.
took over the groups. However, many instances have
now been observed (Sommer 1993). Other critics
however, the clasped male does not signal his maleness argue that infanticide is a result of crowded living con-
to the clasping male. Furthermore, once the clasping ditions, but it turns out that male infanticide is equally
male dismounts, the clasped male may nudge him, in the common in low-density and high-density populations
manner of females, and get him to uselessly deposit his (Dixson 1998; Newton 1986). Furthermore, two pre-
spermatophore. Males of the red-spotted newt occa- dictions of the sexual selection hypothesis for male
sionally engage in a third method of interference, infanticide—that the infanticidal males will kill only
amplexus interference. Here, an intruder simply inspects infants that are unrelated to them and that infanticide
a pair in amplexus at close range, usually just before the will increase a male’s chance of siring the next
mating male dismounts to deposit a spermatophore. The infants—have now been supported by DNA analysis
presence of a voyeur leads the amplexing male to pause. (Borries et al. 1999).
When he resumes his mating behavior, he usually picks We have focused on infanticide by males and the
it up at an earlier stage. The interference, then, causes hypothesis that the killing of infants is adaptive
the mating male to waste time and energy without because intruding males gain more rapid access to
increasing his probability of fertilizing the eggs. mates. There are, however, alternative hypotheses for
Other species exhibit more dramatic forms of sex- infanticidal behavior (Table 14.4), and predictions
ual interference, such as infanticide, the killing of a com- from these hypotheses should also be tested when
petitor’s offspring (Figure 14.12). Infanticide has been searching to explain the killing of conspecific young.

TABLE 14.4 Hypotheses for Infanticide by Males

Hypothesis Predictions

Nonadaptive Infanticide is associated with severe disturbance to the physical environment (e.g., habitat reduction)
Infanticide is associated with severe disturbance to the social environment (e.g., overcrowding)
Performance and context of infanticide should resemble that of an accident
Acquisition of mates Infanticidal males kill only offspring they do not sire
Killing infants shortens the interbirth interval of victimized females
Infanticidal males mate with the female whose offspring they killed
Acquisition of food Males consume infants they have killed
Infanticide and eating of young will increase when food supply is limited
Performance and context of infanticide should resemble that of a predatory attack
Acquisition of space Infanticide will increase when territories or nest sites are limited
Infanticidal males take over the nests or territories used by the individual whose offspring they killed

Based on table and text of Ebensperger and Blumstein (2007).

320 Chapter 14 / Reproductive Behavior

Females of many species also commit infanticide; were each allowed to mate with four small females, one
explanations for female infanticide include all of those female on each of four successive days; these males were
listed in Table 14.4, with the exception of the access considered “depleted males.” (The researchers had
to mates hypothesis. males mate with four females because previous data had
shown that the number of ejaculated sperm found in
females decreases with a male’s successive matings, and
INTERSEXUAL SELECTION— this is especially true for the male’s fifth mate [Sato and
MATE CHOICE Goshima 2006].) Another 30 large males were catego-
rized as “unmated males” because they were not given
Intersexual selection occurs when one sex is in the posi- any prior access to females. In mate-choice experiments
tion of choosing individuals of the other sex as mates. conducted one to three days after the depleted males’
Although there are exceptions, females often do the fourth day of mating (female stone crabs only mate right
choosing while males compete among themselves to be after molting, so this window of time was necessary to
chosen. We will consider this general case, realizing that ensure the availability of newly molted females), the
sometimes males choose and females compete. First, we researchers placed a depleted male and an unmated male
consider some criteria on which females may base their each into an opaque plastic pipe in the test aquarium
choice of mate. (Figure 14.13a). Each plastic pipe had six very narrow
openings to allow the exchange of seawater (and chem-
ical cues) between the inside and outside of each pipe;
CRITERIA BY WHICH FEMALES the openings in the pipes were narrow enough, however,
CHOOSE MATES to prevent a crab on the outside from seeing the indi-
vidual within the pipe and vice versa (these pipes had
Characteristics used by females to select a mate should
been developed for use in the earlier experiments to
affect female fitness, be assessable, and vary among males
eliminate visual cues and test for the importance of
(Searcy 1979). Given these criteria, females are thought
chemical cues in female choice). Fifteen minutes after
to choose mates on the basis of their ability to provide
the males were introduced into the test aquarium, a large
sufficient sperm, useful resources, parental care, and
female was introduced, and she too was given time to
genes that are good or compatible. Sufficient sperm, use-
acclimate to her surroundings. Sato and Goshima used
ful resources (e.g., food or nest sites), and parental care
large females for the testing because large females spawn
are considered direct material benefits; these benefits are
more eggs than small females, thus it would be particu-
obtained by the female doing the choosing. Good or
larly challenging for males to fertilize all the eggs of a
compatible genes, on the other hand, are considered
large female. Once the female acclimation period had
indirect because they benefit the female’s offspring in the
ended, the researchers recorded the total time out of the
next generation (Andersson 1994). Although some
40-minute test that the female spent in the circular area
females may evaluate potential mates on only one char-
surrounding each pipe (refer, again, to Figure 14.13a).
acteristic, others may base their choice on multiple cri-
Thirty such choice tests, each with different crabs, were
teria. Here we consider some of the characteristics that
conducted. On average, females spent more time near
females might evaluate when searching for “Mr. Right.”
unmated males than near depleted males (Figure 14.13b).
Female stone crabs prefer unmated males and
Ability to Provide Sufficient Sperm appear to use chemical cues to assess male ability to
Earlier we stated that the sperm supplies of males are no transfer sufficient sperm. The authors suggest that such
longer viewed as limitless and that successive matings a preference makes sense in this species because males
may deplete a male’s sperm supplies. But do females in experience sperm depletion with successive matings
the market for a mate consider the number of sperm a (there are other species of crabs in which males can
particular male has available? In the case of female stone replenish sperm between matings), and females typically
crabs (Hapalogaster dentata), females appear to do just mate with one male during a reproductive season, so
that. Taku Sato and Seiji Goshima (2007) collected stone their choice is critical. Finally, males guard newly molted
crabs off the coast of Japan and transported them to a females by holding onto them, so the use of chemical
marine station where they conducted several experi- cues by females allows them to assess at a distance the
ments in aquaria filled with seawater. In the first three ability of males to provide sufficient sperm and to only
experiments, they determined that female stone crabs approach a male that passes the test. If females assessed
prefer large males to small males and that females use males by visual cues, then this presumably would
chemical cues, rather than visual ones, when making require closer inspection by the female and increase the
their selection. In the fourth experiment they examined chances that she would be grabbed and guarded, and
the influence of sperm limitation. Thirty large males prevented from exercising mate choice.
 Intersexual Selection—Mate Choice 321

b 1700

Time spent by females in circular areas (sec)

a 1500
Opaque plastic pipes
1300
Circular areas
1100
32 cm

Depleted Unmated 900

male male
700
Openings
500
Test
female Plastic bell 300
on release point
100
42 cm Unmated Depleted
males males
Type of male
FIGURE 14.13 Female stone crabs evaluate whether males have sufficient sperm. (a) The experimental aquarium in
which female stone crabs could choose between spending time near a depleted male (a male that had mated with four
other females) or a male that had not mated. Each male was placed in an opaque plastic pipe with narrow openings
that allowed the passage of chemical cues. Females were initially placed under a plastic bell and allowed to acclimate
to testing conditions. Female preference was assessed by the amount of time they spent in the circular area surround-
ing each pipe. (Modified from Sato and Goshima 2007.) (b) Female stone crabs prefer unmated males to depleted
males. (Drawn from data in Sato and Goshima 2007.)

Ability to Provide Useful Resources to the female, by making it easier to copulate, or by max-
Females of some species may base their choice of a mate imizing the amount of sperm transferred. Second, they
on the quality of resources provided by males. By so may serve as paternal investment by increasing the num-
doing, they could receive either immediate gains from ber or fitness of his offspring (Vahed 1998).
gifts presented during the courtship period or more In some species, for example, the katydid men-
long-term benefits from access to valuable resources, tioned previously, the spermatophore that the male pre-
such as food or nest sites, that are controlled by males. sents as a nuptial gift provides nutrients for the female
Females that exchange mating for material goods could or the eggs. Depending on her own nutritional state, a
place themselves at an advantage. The increased com- female can either put a large portion of the male’s con-
modities could obviously enhance reproductive output tribution into the eggs or use the materials from the
by enabling the female to live longer by being well fed spermatophore for herself, increasing the chances that
or having access to a protected nest site. Not only would she will live long enough to breed again (Gwynne and
she have a competitive advantage over females without Brown 1994).
mates, but also high-quality resources would probably The benefit obtained from the nuptial gift isn’t
improve the survivorship and competitive ability of her always energy. There are many other valuable sub-
offspring. Here we consider a few of the diverse mater- stances. Larvae of the arctiid moth Utetheisa ornatrix
ial goods that males may provide to females and upon sequester alkaloids from the plants on which they feed.
which females may evaluate potential mates. These alkaloids make them quite distasteful to certain
predatory spiders. Males transmit some of these alkaloids
to the female with their spermatophores. The female
Nuptial Gifts and Cannibalism The males of many bestows some of the alkaloids on her eggs, giving them
birds and some species of insects offer nutrition or other protection from predators (Eisner and Meinwald 1995).
valuable substances to the female during courtship. The female is also protected by this gift. Almost imme-
These nuptial gifts may take various forms, including diately, she becomes unacceptable prey to these spiders
prey, glandular secretions, and the spermatophore. (Gonzalez et al. 1999). In other moths, limited micronu-
Indeed, a male may even offer parts or all of his body for trients are passed to the female or her eggs in the nup-
the female to feed on (Andersson 1994). The functions tial gift (Smedley and Eisner 1996).
of the gifts generally fall into two categories, which are Males of some insect species present a gift of prey
not mutually exclusive. First, they may increase the that the female feeds on during copulation, buying them
male’s chances of mating by making him more attractive time to transfer sperm. Male hangingflies (Bittacidae)
322 Chapter 14 / Reproductive Behavior

usually offer a large prey item, one that will take at least copulation because he copulates about twice as long.
20 minutes to consume. It takes the male about 20 min- Another advantage is that the female is about 17 times
utes to completely transfer enough sperm and fluids to less likely to mate again after consuming the first mate
make the female unreceptive to other males and begin (Andrade 1996).
laying her eggs (Thornhill 1976). In other insects, the
male’s nuptial gift is a part or all of his body. During Territory In many species, females selecting mates
copulation, a female sagebrush cricket feeds on the appear to assess the quality of territories held by males.
male’s fleshy hind wings and the hemolymph that oozes For example, female house wrens (Troglodytes aedon)
from the wounds. This courtship feeding keeps the select a mate based, at least in part, on the number of
female mounted while the male transfers his sperm. nest sites in his territory. House wrens are secondary cav-
Males whose hind wings have been surgically removed ity nesters (i.e., they are not capable of excavating their
transfer significantly less sperm than do intact males own nest cavities, so they instead use nest holes made by
(Figure 14.14) (Eggert and Sakaluk 1994). other species, such as woodpeckers) that will readily use
Finally, in some species of spiders, scorpions, man- nest boxes. By experimentally adding from one to three
tids, and diptera, the male makes the ultimate sacrifice nest boxes to male territories already containing a sin-
during copulation. He gives his body to the cannibalis- gle nest box, Kevin Eckerle and Charles Thompson
tic female. Oddly enough, under some circumstances this (2006) found that time to pairing for males was associ-
ultimate sacrifice is adaptive. The male redback spider ated with the number of nest boxes on their territory:
(Latrodectus hasselti) stores his sperm in a tightly coiled males with the most nest boxes were selected first by
structure on his head, called a palp. To transfer sperm, females, leading to shorter times to pairing for these
he scrapes the palp across the female’s genital opening, males. Extra nest sites might decrease the chances that
which causes the coil to unspring. The open coil is then the female’s own nest would be found by predators or
inserted into the genital opening and sperm is trans- provide alternative sites should re-nesting be necessary.
ferred. A few seconds later the male does a somersault, In other species, females might evaluate the amount or
placing his abdomen directly above the female’s jaws. quality of food on a male’s territory.
About 65% of the time, the female eats him while he’s
in the somersault position. His chances of being con- Ability to Provide Parental Care Females may also
sumed are much greater if the female is hungry. The assess a male’s parental abilities. This seems even more
males gain two paternity advantages from this suicidal challenging than estimating the value of material
behavior. One advantage is that a cannibalized male fer- resources, but it appears to occur nonetheless. Indeed,
tilizes about twice as many eggs as a male who survives evidence from diverse species indicates that females may
use physical or behavioral features of males to predict
parental quality. In some species of birds, for example,
females may judge male parental ability on the basis of
Hind wings removed
the quality of nutritional gifts provided during the period
8 Hind wings intact
of courtship (Figure 14.15). A large number of high-
quality gifts may signal a male’s superb foraging skill and
willingness to feed his mate and offspring during incu-
Number of pairs

6 bation and posthatching stages. Some females may use

the success of previous nesting attempts to judge the
parental ability of males (Coulson 1966). We will focus
4
on how certain female fish use the speed of water cur-
rents to judge the parental abilities of potential mates.
Gobies of the Rhinogobius brunneus species complex
2
inhabit streams with steep gradients in Japan. Male gob-
ies build nests, provide sole care to the eggs, and engage
in aggressive displays (Figure 14.16a). Although males
0
0 1 2 3 and females occur in pools, they only court in currents.
Number of spermatophores transferred At the start of courtship, a female remains stationary on
the stream bottom; the male, on the other hand, rises off
FIGURE 14.14 Male sagebrush crickets with intact hind
the bottom and begins to dance. If not carried away by
wings transfer significantly more sperm packets to females
in a 12-hour period than do males whose hind wings have
the swift-flowing water, the male approaches the female
been surgically removed. A female sagebrush cricket feeds and shakes his head from side to side. If the female bends
on the male’s fleshy hind wings during copulation. This her body in response, then he tries repeatedly to lead her
helps to keep her mounted while the male transfers his to his nest, all the while spreading his fins and vibrating
sperm. (Modified from Eggert and Sakaluk 1994.) his body. At this stage, many females reject males, and
 Intersexual Selection—Mate Choice 323

b
2
1.0

Egg survival rate

0.8

0.6

0.4

0.2
3
0.0
0 10 20 30
Maximum water velocity of male courtship (cm/sec)

FIGURE 14.15 A male European crossbill passes regurgi-

FIGURE 14.16 Female gobies choose males that can
tated seeds to a female. Female birds may judge the
court in fast currents because such males are in superior
parental ability of a male by the quality and quantity
physical condition and will not eat their offspring when
of gifts provided during courtship feeding.
guarding them. (a) A male goby in an aggressive posture.
(b) Shown here is the relationship between maximum
current speed at which a male can court and the survival
rejected males leave the swift-flowing water and return of his eggs. (Modified from Takahashi and Kohda 2004.)
to the pool. If the female does not reject the male and
instead follows him to his nest, she then enters the nest
and deposits her eggs in a single layer on the ceiling and these species? There is evidence that females evaluate
the male fertilizes them. Following her departure, the male traits that signal genetic quality.
male arranges the substrate to close off the nest, effec-
tively walling himself in with the kids. He remains in the
nest, fanning the eggs for about two weeks. While con- Ability to Provide Good Genes
fined, his eggs are the only food available, and some Barring direct examination of a male’s genotype, how
males cannibalize some or all of their eggs. could a female evaluate variation in genetic quality
On what basis does a female goby accept or reject a among suitors? Females might judge genetic quality by
male? Daisuke Takahashi and Masanori Kohda (2004) examining a male’s general physical well-being, capacity
conducted mate-choice experiments in the laboratory to dominate rival males, or capacity for prolonged sur-
and discovered that female gobies favor males that per- vival. We’ll first describe how a showy secondary sex
form their courtship dance in the fastest water currents. characteristic can be a reliable, honest indicator of a
But why? It turns out that only males in exceptional male’s health and good genes. (Honest signals are dis-
physical condition can dance in fast currents and not be cussed further in Chapter 17.)
swept away. In addition, the maximum water velocity at William D. Hamilton and Marlene Zuk (1982) were
which a male can dance is positively correlated with the the first to propose that the elaborate ornaments of males
survival of his eggs (Figure 14.16b). Males unable to represent reliable signals of health and nutritional sta-
court in fast currents are in less than exceptional physi- tus. They examined plumage coloration in North
cal condition and more likely to consume their eggs dur- American birds and suggested that only males in top
ing the two-week guarding period. So, female gobies physical condition would be able to maintain bright,
appear to evaluate the velocity of the current in which a showy plumage. Because bird species vary in their sus-
male displays, and such information indicates his condi- ceptibility to parasitic infection, Hamilton and Zuk pre-
tion and ability to abstain from eating his offspring. dicted that the degree of male brightness would be
Thus, courting in fast current is an “honest signal” of a correlated with the risk of attack by parasites.
male’s ability to offer paternal care. Accordingly, if males of a species vary substantially in
In some species, males do not provide direct mate- their parasite load, it would behoove a female to choose
rial benefits to their mates or parental care to their off- a male that honestly signals his good health, for the off-
spring. How might females evaluate potential mates in spring of this male may experience increased viability if
324 Chapter 14 / Reproductive Behavior

they inherit his resistance to parasites. Because the bars may differ between sides. In nature males differ in the
brightness of plumage is closely tied to a male’s general presence of bars, number of bars, and symmetry in the
health, it is a reliable signal that cannot be faked by a par- number of bars. In laboratory choice tests, female sailfin
asite-laden male. mollies prefer males with vertical bars to those without
Based on this hypothesis, Hamilton and Zuk rea- bars. Furthermore, when bars are present, females prefer
soned that in species in which the risk of infection by symmetrical males—those with the same number of bars
parasites is minimal, information regarding parasite load on each side of the body (Schluter et al. 1998).
has little value to females in the market for a mate, and Studies on barn swallows suggest a way in which
thus females should not display a preference for males trait symmetry might reflect good genes. Female barn
with showy features. Hamilton and Zuk therefore pre- swallows (Hirundo rustica) prefer to mate with males with
dicted that in species with low risk of parasite infection, long tails (Møller 1988). Mite infestations stunt the
males would not be brightly colored. growth of tail feathers, and so tail length advertises
In testing their hypothesis, Hamilton and Zuk sur- whether a male has been previously infected with mites.
veyed the literature on avian parasites and determined Mite infestations also increase asymmetry in tail length.
the risk of infection for each bird species. They then Besides preferring males with longer tails, female swal-
ranked each species from 1 (very dull) to 6 (very strik- lows prefer males with tails that are symmetrical in
ing) on a plumage showiness scale. In support of their length (Møller 1990). So, by choosing males with long,
ideas, there was a significant association between showi- symmetrical tails, females are choosing males with genes
ness and the risk of parasitic infection: those species with for parasite resistance (Møller 1992).
the highest risk of infection from blood parasites had the We mentioned at the start of this section that
showiest males. females might also evaluate male genetic quality by
In the years since Hamilton and Zuk (1982) published assessing a male’s physical condition and ability to dom-
their findings, the influence of parasites on mate choice inate other males. Our next two examples focus on male
has been examined in many species of birds (e.g., Borgia physical condition and dominance as indicators of
1986; Møller 1991; Spencer et al. 2005), as well as in a host genetic quality. The first of the two examples involves
of other animals, including amphibians (Hausfater et al. elaborate male ornaments; the second example does not.
1990); fish (Barbosa and Magurran 2006); and humans Male satin bowerbirds (Ptilonorhynchus violaceus)
(Low 1990). Some studies support the idea that females build unique stick structures called bowers in which they
choose mates by using cues that will lead to more viable display to females in attempts to secure copulations
offspring; others do not. Taken together, however, the (Figure 14.17). Males decorate their bowers with artifi-
overall pattern that emerges is that showy males have
fewer parasites and are better able than dull males to
mount a stronger immune response against a wide vari-
ety of parasites. These observations support the idea that
females choose showy males to obtain good, parasite-
resistance genes for their offspring (Møller et al. 1999).
The extent of asymmetry, called fluctuating asym-
metry, in otherwise bilaterally symmetrical traits is
another potential indicator of overall genetic quality.
Because the growth and development of both sides of the
body are controlled by the same set of genes, we would
expect the right and left sides to be identical. This is not
always the case, however. Environmental insults or
genetic defects can cause one side of the body to develop
in a slightly different manner from the other, causing the
traits to be asymmetrical. Females in many species prefer
males with symmetrical traits, presumably because sym-
metry signals a healthy condition and good genes (Møller
and Pomiankowski 1993; Møller and Swaddle 1997).
Consider the sailfin molly Poecilia latipinna, a small
FIGURE 14.17 Male satin bowerbirds decorate their
tropical fish commonly found in brackish coastal marshes. bowers with natural and artificial materials. The num-
It is so named because the males have a large dorsal fin ber and rarity of decorations on a bower are important
that resembles a sail. The fin is displayed to the female determinants of mating success. Female bowerbirds may
during courtship. Males are brightly colored and may have judge the genetic quality of a male by the attractiveness
vertical bars on the sides of their bodies. The number of of his bower.
 Intersexual Selection—Mate Choice 325

Female choice based on good genes need not

always involve elaborate ornaments, whether worn by
the male himself or meticulously arranged by him on
his display arena. Our next example of female choice
for good genes concerns pronghorn (Antilocapra amer-
icana), a species of ungulate endemic to the Nearctic
region (Figure 14.18). Pronghorn are the fastest ter-
restrial mammals in the New World, reaching speeds
near those of cheetahs, the record holders of the Old
World. Endurance and vigor are important for male
and female pronghorn when escaping predators; these
traits are also evaluated by females when selecting a
mate. Like male satin bowerbirds, male pronghorn do
not monopolize resources critical to females and do
not provide paternal care, yet females still actively
choose their mate. All this suggests that females might
be basing their choice on good genes. John Byers and
colleagues have been studying a population of prong-
horn in Montana for over two decades. Their obser-
vations indicate that females copulate once per estrous
period and that most use an energetically expensive
FIGURE 14.18 Pronghorn females may choose males sampling strategy to select a mate (Byers et al. 1994,
based on good genes, as assessed by the male’s ability to 2006). For the two weeks prior to estrus, females move
attract and defend females from other males during the independently and make short-duration visits to widely
breeding season. spaced harem-holding males (harems are one or more
females defended by a single male). Although harem-
holding males make every effort to prevent visiting
cial materials (e.g., bottle caps), snail shells, feathers, and females from leaving their site, females always leave a
flowers. They display strong preferences for inflores- male that fails to maintain a sufficient zone of tran-
cences of certain colors; blue and purple flowers are rel- quility (an area free of other males trying to gain access
atively rare in the environment of satin bowerbirds and to the females) around his harem. After visiting
are preferred over yellow and white flowers, whereas numerous males, females make their decision, and
orange, red, and pink inflorescences are completely return and mate with the chosen male, leaving him
unacceptable (Borgia et al. 1987). Males also masticate after about a day. Most females in the population con-
leaves and then paint the resulting mixture on the walls verge on a small subset of harem-holding males, and
of their bower. Bowers and their decorations appear to these males achieve the majority of the copulations.
have no intrinsic value to either sex outside the context What is it about these select males that leads to their
of sexual display and probably serve primarily as indica- choice by most females?
tors of male quality. Indeed, the number and type of dec- Male pronghorn mating success is not correlated
orations on a bower are important determinants of a with body size, horn size, age, or rate of display (during
male’s mating success, as are characteristics of his behav- displays, males scent-mark, parallel-walk, and emit
ioral display (Coleman et al. 2004; Patricelli et al. 2003). snort-wheezes). However, a male’s mating success is
Not surprisingly, males go to great lengths to steal rare strongly correlated with his ability to attract and retain
decorations from the bowers of competitors (Wojcieszek females, as measured by haremdays, the sum of a male’s
et al. 2007). Females may favor males who exhibit exotic mean harem size for each day across all days of the
decorations because the ability to accumulate and hold breeding period (Figure 14.19) (Byers et al. 1994).
these decorations indicates that a male is in top physical Number of haremdays, in essence, is a measure of a
condition. Because the number of inflorescences on male’s vigor, especially as it pertains to his ability to
bowers is correlated with age, females could use the maintain a large zone of tranquility for his harem. But
number of flowers, and specifically the number of rare is male vigor a good predictor of offspring quality?
decorations, to assess the experience of a male. Thus, Apparently so. Offspring of vigorous males are more
although female bowerbirds cannot directly examine the likely to survive to weaning and beyond than offspring
genotype of potential mates, they may evaluate a male’s of less vigorous males (Byers and Waits 2006). Thus, in
genetic quality based, at least in part, on the attractive- pronghorn, females evaluate male vigor, and male vigor
ness of his bower. indicates good genes.
326 Chapter 14 / Reproductive Behavior

10 in the anterior roof of the mouth of many mammals and

1 some other terrestrial vertebrates.)
8 We still aren’t sure why MHC-dependent mating
Number of copulations

1 preferences exist, but two hypotheses have been sug-

6 2 gested. The first is that they evaluate the degree of
4 genetic relatedness of potential mates to avoid inbreed-
4
4 ing. Inbreeding caused by mating with close relatives
7 increases homozygosity and hence the risk of producing
2 10 offspring that are homozygous for deleterious or lethal
21 recessive alleles. On the other hand, extreme outbreed-
0 145 ing may cause the breakup of successful parental com-
plexes of genes. Given the potential costs associated with
0 20 40 60 80 100 120 140
mating with either very close relatives or complete
Mean haremdays
strangers, females may strike a balance between extreme
FIGURE 14.19 Number of copulations per breeding sea- inbreeding and outbreeding when choosing a mate. The
son in relation to haremdays, defined as the sum of a second hypothesis is that MHC-dependent mating pref-
male’s mean harem size for each day across all days of erences allow a female to increase her offspring’s resis-
the breeding season. Number beside each point indi- tance to disease. Recall that the genes of the MHC region
cates sample size. (Modified from Byers et al. 1994.) are important in protecting against disease-causing
organisms. By choosing a male whose MHC alleles dif-
fer from hers, she increases the variability of the MHC
Ability to Provide Compatible Genes region in her offspring. This may make them resistant to
Sometimes the fitness consequences of mating are not a wider variety of disease-causing organisms.
simply a function of the male’s genetic quality; instead, There are additional questions to be answered about
the fitness consequences may depend on the “fit” the MHC and mate preferences. For example, preference
between the genes of the female and those of the male for an MHC-dissimilar mate is reported in some, but not
(Zeh and Zeh 1996; Mays and Hill 2004). In other all, studies of mate choice in house mice (reviewed in
words, females may evaluate the ability of males to pro- Solomon and Keane 2007). How can we explain the dif-
vide genes compatible with their own. Sometimes the ferent findings of the different studies? Also, it remains
most compatible genes are those that are dissimilar to to be determined just how MHC-based odors compete
the female’s genes. We’ll consider an example involving with other genetically based odors known to be impor-
genes of the major histocompatibility complex (MHC). tant in individual and kin recognition in house mice (for
The MHC is a large chromosomal region that varies example, those associated with major urinary proteins)
tremendously among individuals. It is important in the (Carroll and Potts 2007; Cheetham et al. 2007).
immune responses that protect animals from disease-
causing organisms. (The role of the MHC in kin dis-
crimination is discussed in Chapter 19.) We will focus STOP AND THINK
on the role of the MHC in mate choice in house mice, Several methodological problems are commonly encoun-
the species in which the phenomenon was first described tered when studying female choice. Aside from the fact that
by Yamazaki et al. (1976). Although the initial discovery mate choice may be very subtle, it is often masked or con-
concerned male mice preferring to mate with females founded by extraneous effects, such as motivation of the
female at the time of assessment. If a female does not dis-
carrying dissimilar MHC genes, some subsequent stud-
play a preference for a male during a choice test, is it because
ies have tested the preferences of females, and we will
she has rejected him as a potential mate, or is it simply
focus on these investigations. because she is not receptive? Also, as we have seen, female
Several laboratory studies have shown that female choice in the laboratory is often measured by orientation
house mice prefer males that differ from themselves in toward or time spent near a particular male; yet in some
the makeup of the MHC region (studies reviewed in cases, these behavioral responses may not relate to actual
Penn and Potts 1999; Solomon and Keane 2007). Also, mating inclination. And then there is the problem of deter-
studies in seminatural enclosures show that female mice mining the precise roles of male competition and female
leave their territories to engage in extra-pair matings with choice within a single mating system. Although females may
males that differ from them in the MHC region (Carroll appear to choose large males over small males, is it really
and Potts 2007). Mice use odor cues to determine the mate choice or is it because large males effectively exclude
small males from courtship activities? What factors would
degree of similarity of the potential mate’s MHC region
you take into consideration when designing a test of female
to their own. Peptides, bound to MHC molecules, are the
choice? What would you want to know about your study
source of the odors, and they are detected by the mouse’s species before designing such a test? How would you sepa-
vomeronasal organ (Leinders-Zufall et al. 2004). (The rate the effects of female choice and male–male competition?
vomeronasal organ is an accessory olfactory organ found
 Intersexual Selection—Mate Choice 327

ORIGIN AND MAINTENANCE

OF MATE-CHOICE PREFERENCES
It isn’t too difficult to see why a female should be
choosy or why she would choose the male that offers
her the most in material benefits, such as nuptial gifts,
a good territory, or high-quality parental care. In
these instances, a female’s preference evolves because
it increases her fitness. But how do female mate pref-
erences originate when there are no direct material
benefits, that is, when the female’s preference affects the
fitness of her offspring rather than her own fitness? We’ll
consider two explanations for the origin of preferences
that do not involve direct material benefits: indicator and
runaway mechanisms. Some people prefer not to make
a distinction between these two explanations (Kokko et
al. 2003), but we find it useful to discuss them separately.
We end with a third explanation for the origin of female
preferences called sensory bias.

Indicator Mechanisms
Indicator mechanisms are often called good genes
models. These models assume that a particular trait in
males indicates viability and that both the trait and via- FIGURE 14.20 A peacock displays his train. Peahens
bility have a genetic basis. If a female preference (also choose males with the most elaborate trains. Elaborate
genetically based) for the male trait should arise, then trains appear to indicate good genes; thus, females that
these females mate with males carrying genes for the select highly ornamented males increase the viability of
trait and enhanced viability. In this way, genes for the their offspring.
male trait, high viability, and the female preference
become associated. High viability is favored by natural
selection; because the genes for the male trait and spring, adult males gather to court females on a com-
female preference are associated with viability, they too munal breeding ground called a lek. An adult peacock is
are favored. a thing of beauty (Figure 14.20). His chief glory is a train
An example of a good genes model is the handicap of long, beautifully marked feathers, each tipped with an
principle proposed by Amotz Zahavi (1975). According iridescent “eye.” A courting male lifts his tail, which lies
to this principle, females prefer a male with a trait that under the train. This elevates the train and spreads it out
reduces his chances of survival but announces his supe- like a fan. He then struts around on his small territory
rior genetic quality precisely because he has managed to on the lek, vibrating his tail rapidly. In turn, the tail
survive despite his “handicap.” In short, male secondary vibrations cause the plumes in his train to rattle audibly.
sexual characteristics act as honest signals, indicating He attempts to copulate with a “hoot-dash,” in which he
high fitness, and females choose males with the greatest begins to lower his train and rushes toward the female
handicaps because their superior genes may help pro- while giving the “hoot” call. If his copulation attempt is
duce viable offspring. successful, the female will squat in front of him, allow-
The hypothesis that females choose males with good ing him to mount her.
genes predicts that the mating preferences of females Peahens are choosy, preferring males with elaborate
should increase the viability of their offspring. Some sup- trains. When it comes to peacock mating success, the
porting evidence comes from studies that show a rela- “eyes” have it; that is, mating success is significantly cor-
tionship between exaggerated male traits and offspring related with the number of eyespots in the male’s train.
viability. One of the best known examples of an exag- On one lek, which consisted of ten courting males, the
gerated male trait is the ostentatious tail of the peacock. most successful male copulated 12 times, but the least suc-
Peafowl (Pavo cristatus) are also an example of a species cessful males never did. A female never accepted the first
in which there is good evidence that the females (known suitor she saw. On average, she visited about three males
as peahens) get genetic benefits that increase survival of before copulating. In 10 out of 11 observed courtship
their offspring through mate-choice preferences. Marion displays that ended with successful copulation, the female
Petrie has studied a free-ranging population of peafowl chose the male with the highest number of eyespots
in Whipsnade Park in England. During the early in his train of those she visited (Petrie et al. 1991).
328 Chapter 14 / Reproductive Behavior

When 20 eyespots were experimentally cut out of the have sons with the trait, providing that the attractive
trains of some peacocks, the male’s mating success was sig- character of the male is inherited, and she will have
nificantly less than his success during the previous year. daughters who show a preference for that trait. The
The attractiveness of control males, whose tails were left attractive sons will acquire more mates with a prefer-
intact after being captured and handled, remained the ence for that characteristic than other males and thus
same (Petrie and Halliday 1994). will leave more progeny. In this way, runaway selection
Besides boosting a peacock’s mating success, the can produce increasingly exaggerated male traits and a
extent of sexual ornamentation appears to be corre- stronger female preference for them. So, whereas indi-
lated with survivorship. In the spring of 1990, of 33 cator mechanisms require that females acquire good
displaying males in the same population, 22 copulated genes that increase the viability of their sons and
successfully at least once and 11 were never success- daughters, a runaway selection mechanism requires that
ful. During the following winter, two foxes managed females acquire genes that make their sons particularly
to enter the park, and they killed five peacocks. Four attractive.
of the birds that were killed were among those that had Lande (1981) and Kirkpatrick (1982) developed
been unsuccessful in gaining any copulations in the mathematical models of Fisher’s (1930) ideas for the
previous season. The fifth bird had copulated only evolution of female choice. These models demonstrate
twice. The males that were killed also had shorter that runaway selection can indeed result in mate choice
trains with fewer eyespots than surviving males had. for characteristics that are arbitrary or even disadvan-
This observation is consistent with the hypothesis that tageous to the health and survival of individuals, pro-
only the healthiest males can develop long, elaborate viding that females prefer to mate with males that
trains. Thus females are choosing males with good possess them. Runaway selection would favor ever more
genes (Petrie 1992). exaggerated male characteristics and females that find
It also seems that the offspring of females who them attractive. So when does it all stop? The process
choose mates with elaborate trains benefit from their will be stabilized only when natural selection balances
father’s good genes. Petrie (1994) demonstrated this by sexual selection (Fisher 1930). In other words, when the
pairing males with females chosen at random. In each male trait becomes too energetically costly to produce
large cage, she placed one male and four females. The or when it makes males less likely to escape from preda-
mated males varied in attractiveness to females, as mea- tors, selection will no longer favor further exaggeration
sured by the mean area of eyespots in the train. All the of the trait.
offspring were raised under common conditions. When Experimental support for the runaway selection
they were 84 days old, the offspring were weighed. The hypothesis comes from sandflies of the Lutzomyia longi-
offspring of males with more elaborate trains weighed palpis species complex. At night, male sandflies aggregate
more than those with less showy fathers. After two on the back of a vertebrate host, perhaps a chicken,
years, more of the offspring of the highly ornamented where they defend small (radius of 2 cm), mobile terri-
males were still alive than were those of less attractive tories by bumping one another with their abdomens.
males. Because the matings were arranged by Petrie, Females visit the host and evaluate several males but cop-
the differences in offspring viability cannot be due to ulate with just one. Mating success among males is
differences in the quality of females. By ruling out highly skewed, with some males copulating with many
maternal effects, Petrie has provided strong evidence females and others unsuccessful, reflecting the consid-
that a female can enhance the survival chances of her erable agreement displayed by females in their choice of
offspring through mate-choice preferences. Here, then, mate. Therésa M. Jones and colleagues (1998) conducted
is an example in which females choose males on the laboratory experiments designed to evaluate the fitness
basis of a trait that indicates male genetic quality, and consequences of mate choice for female sandflies and
this choice, in turn, results in more viable sons their offspring. Female sandflies were allowed to choose
and daughters. among several males. Females and their offspring were
then monitored for survival and fecundity. The results
indicated that females do not gain direct benefits by mat-
Runaway Mechanisms ing with attractive males (here, attractiveness is defined
Runaway mechanisms are also called Fisher or as the number of females with which a male mates)
Fisherian mechanisms, after R. A. Fisher (1930), who because females that mated with attractive males did not
first described the process. The basic idea is that the themselves have higher survival or fecundity. But what
male trait is correlated with the female preference for about indirect benefits, that is, benefits to their off-
that trait. It begins when females evolve a preference spring? Jones and colleagues found no evidence that a
for a particular male characteristic. A female that mates female’s choice of mate influenced the viability of her off-
with a male that has the attractive characteristic will spring; the survival of sons and daughters, and the fecun-
 Intersexual Selection—Mate Choice 329

0.50 Rice 1998). It emphasizes conflict between males and

Mean mating success of sons

females, a topic that we consider in more detail at the
0.45
end of this chapter.
0.40 Recall that male satin bowerbirds decorate their
bowers with natural and artificial objects, and that they
0.35
are especially picky about the color of their decorations.
0.30 Recall also that a male satin bowerbird prefers blue and
that the color of his decorations is an important deter-
0.25
minant of his mating success. J. R. Madden and K.
0.20 Tanner (2003) investigated a possible example of sensory
Low Average High
bias in bowerbirds. They found that when given a choice
Attractiveness of fathers
of grapes of different colors (commercial food dyes were
FIGURE 14.21 In sandflies, the mating success of sons used to make brown, yellow, red, green, or blue grapes),
increases with the attractiveness of their fathers. These female regent bowerbirds and female satin bowerbirds
data support runaway selection models for the origin of (to a somewhat lesser extent) preferred to eat blue grapes
female preferences. and males of both species preferred to decorate with blue
grapes. These data are thus consistent with the sensory
bias model—choice of bower decoration could have
dity of daughters, did not increase with attractiveness of evolved to exploit a sensory bias in females that was orig-
their father. These findings thus do not support good inally related to foraging. See Chapter 17 for additional
genes models. However, the mating success of sons examples of sensory bias.
increased with the attractiveness of their father (Figure
14.21). Thus, female sandflies that select an attractive CRYPTIC FEMALE CHOICE
male appear to benefit by having attractive sons; these
findings support runaway selection. The observation that “It ain’t over ‘til it’s over” may ring
true for female mate choice. Females of certain species
appear able to choose the sperm that will fertilize their
Sensory Bias eggs after copulating with several males. This is called
According to the sensory bias model, female prefer- cryptic female choice because it is a hidden, internal
ences for certain traits in males could evolve because decision made after copulation (Eberhard 1996). In the
male traits stimulate an existing bias in the female’s sen- sperm wars, this choice is the female equivalent of male
sory system. The original bias might relate to feeding sperm competition.
or avoiding predators. For example, females might have So far, evidence for cryptic female choice is some-
a sensory bias to help them find food of a particular what limited because it is difficult to distinguish from
color; males can then exploit this bias by using the same sperm competition (Andersson and Simmons 2006).
color to attract females during courtship. Consider how Nevertheless, cryptic female choice has been demon-
such a scenario might work. By chance, a male has a strated in several species, most notably insects. The
mutation that produces a trait that exploits a sensory mechanisms by which females choose sperm after cop-
bias of females. This male, though he provides neither ulation are diverse and include (but are not limited to)
the resources nor the good genes that other males pro- the following: premature termination of copulation; pre-
vide, is successful in convincing females to mate with mature removal of spermatophores; failure to store
him. Over time, males with such a trait would become sperm that have been transferred; removal of stored
more common in the population. Eventually, however, sperm; and differential use of stored sperm in fertiliza-
there would be selection for females to resist mating tion. Females of a given species may use one or more
with such males because they receive no benefits from mechanisms. We will focus on the red flour beetle
selecting them. In response to increasing female resis- (Tribolium castaneum), a species in which females mate
tance, selection would then favor males with mutations with two or more males in quick succession.
that further exaggerate the trait. The cycle of increas- In a somewhat grisly experiment, Tatyana Fedina
ing female resistance to mating with males that possess (2007) investigated whether female flour beetles con-
the trait and male exaggeration of the trait leads to trol the quantity of sperm transferred to them per sper-
expensive male traits of no use to females but needed matophore. In this species, sperm are not packaged into
by males to obtain matings. This specific explanation spermatophores before copulation (as in insects like
for the evolution of extreme traits in males that exploit crickets and katydids), but during copulation. She
a sensory bias of females and provide no benefits to manipulated male phenotypic quality (body condition)
females is called the chase-away model (Holland and by using fed males (each male was housed in a vial that
330 Chapter 14 / Reproductive Behavior

3.5 Controlling the amount of sperm transferred per sper-

Fed males
Starved males
matophore is not the only mechanism of cryptic female
Number of sperm transferred*105

3.0 choice in flour beetles. Females also selectively prevent

the transfer of spermatophores at copulation, expel
2.5 spermatophores soon after mating, and may control the
amount of sperm stored in the spermatheca (Bloch
2.0 Qazi et al. 1998; Fedina and Lewis 2004, 2006). Indeed,
female flour beetles appear able to make rather fine
1.5
adjustments during sperm transfer and storage that
allow them to manipulate the sperm representation of
1.0
different males.
0.5

0.0 SEXUAL CONFLICT

Dead Live
Females Parker (1979, 2006) defines sexual conflict as a con-
flict between the evolutionary interests of males and
FIGURE 14.22 Female flour beetles exhibit cryptic female
choice. In this experiment, fed males transferred signifi-
females. It sometimes is generated by sexual selection.
cantly more sperm than did starved males, but only For example, males compete with one another to mate
when mating with live females. Thus, live female flour with females. That competition may result in the evo-
beetles control the number of sperm transferred to them lution of reproductive behaviors and adaptations that
by males, favoring males in good physical condition. are harmful to females. For example, male fruit flies
produce accessory gland proteins that promote a male’s
success in sperm competition, but decrease the
longevity and reproductive success of frequently mat-
ing females (Chapman et al. 1995, 2000). Sexual con-
contained wheat flour for seven days) and starved males flict takes several forms, but the two main ones
(each male was housed in a vial without wheat flour for concern mating/fertilization and parental investment.
seven days). She also manipulated whether females These situations involve interactions between males
could control the amount of sperm transferred to them and females, during which each individual’s fitness
during copulation by using live females (which can depends on its own strategy as well as that of its part-
exert control over copulations) or dead females (which ner. Our focus here will be on sexual conflict during
cannot exert control) (yes, male flour beetles will mate mating/fertilization.
with recently killed females). The experimental design As mentioned, sexual conflict can lead to traits
thus contained the following treatments, with 10 to evolving in males that are beneficial to them but dam-
12 male–female pairs in each: (1) live females and fed aging to females. Females, in turn, may evolve a counter-
males, (2) live females and starved males, (3) dead adaptation that reduces or overcomes the harmful
females and fed males, (4) dead females and starved effects of male sexual behavior. Coevolution between
males. The design allowed female influence to be dis- males and females that is propelled by sexual conflict is
tinguished from male influence on sperm transfer. called sexually antagonistic coevolution. We will briefly
Immediately after members of male-female pairs finished describe two examples of this phenomenon—one con-
copulating, Fedina dissected out the reproductive tract cerns bedbugs and the other, seed beetles.
of the female and estimated the number of sperm pre- Rather than mating in the traditional manner, male
sent. As you can see in Figure 14.22, fed males trans- bedbugs (Heteroptera: Cimicidae) engage in traumatic
ferred significantly more sperm than starved males, but insemination whereby they stab their intromittent
only when mating with live females. Thus, when female organ through the abdomen of a female, and inject
flour beetles have the ability to do so (i.e., when alive), sperm and accessory gland fluids directly into her body
they control the number of sperm transferred to them cavity (Morrow and Arnqvist 2003; Stutt and Siva-Jothy
by males, favoring males in good physical condition. 2001). This form of male sexual behavior might have
Although the precise physiological mechanism by evolved because it allowed males to inseminate females
which females exert such control remains to be deter- that had already mated (in the usual way) and were resis-
mined, Fedina (2007) suggests that females may con- tant to further inseminations. While beneficial to
tract or relax certain muscles in their reproductive tract males, traumatic insemination is harmful to females;
(bursal muscles), thereby controlling the flow of sperm. females with high mating rates have shorter life spans.
 Summary 331

a b c

FIGURE 14.23 Sometimes sexual conflict leads to sexually antagonistic coevolution, in which an adaptation in one sex is
met with a counteradaptation in the other sex. Shown here are scanning electron micrographs of male genitalia and
cross sections of female copulatory ducts (insets) for three species of seed beetles. Note that the degree of robustness
of female copulatory ducts parallels the degree of spininess of male genitalia: both are greatest in (a), intermediate in
(b), and least in (c). Comparative data like these suggest that in seed beetles, the male adaptation of spiny genitalia has
been met with a female counteradaptation, robust copulatory ducts.

Female bedbugs, however, have responded with a

counteradaptation. Just beneath their cuticle, at the SUMMARY
exact spot on their abdomen where the male typically
Sexual selection results from (1) competition within one
inserts his intromittent organ, the females have a
sex for mates (intrasexual selection) and (2) preferences
spermalege. This special organ somehow reduces the
exhibited by one sex for certain traits in the opposite sex
trauma associated with the piercing of the female’s
(intersexual selection). Early observations regarding sex-
abdomen. Indeed, experimental piercing of the female’s
ual selection suggested that because females invest more
abdomen outside the spermalege causes dramatic
in gametes and parental care they are more selective than
decreases in female life span and lifetime egg produc-
tion (Morrow and Arnqvist 2003). Generally, males males in their choice of mates and are usually a limited
should benefit by harming their mates as little as possi- resource for which males compete. There have been
ble, and so they typically inseminate females at the site recent challenges to these ideas. Nevertheless, competi-
of the spermalege. In the case of bedbugs, then, male tion among males for access to females appears respon-
competition may have led to the evolution of traumatic sible for the evolution in males of a broad spectrum of
insemination by males, which in turn led to the evolu- physical and behavioral attributes that enhance fighting
tion of the spermalege in females. This provides us with prowess. Weapons, large size, and physical strength have
an example of how sexual selection can generate sexual evolved in males of many species. In addition, males
conflict, which in turn can lead to sexually antagonistic devote much time and energy to ensuring that their
coevolution. sperm and not the sperm of a competitor are used to fer-
Our second example of sexually antagonistic tilize a female’s eggs. In some species this may take the
coevolution concerns seed beetles (Coleoptera: form of dominance behavior, mate guarding, or alterna-
Bruchidae). Males of some species have genitalia tive reproductive strategies; in others, repellents or cop-
armed with elaborate spines that help to anchor them ulatory plugs are used. Because reproductive success is
during copulation. These spines, however, are harm- measured in relative terms, males may also enhance their
ful to females, penetrating their copulatory duct dur- position by employing tactics of sexual interference that
ing mating and leaving scars where the damaged tissue decrease the reproductive success of other males.
heals (Rönn et al. 2007). The degree of harm experi- Females may base their choice of mate on the
enced by females increases with the spininess of male male’s ability to provide material benefits such as suf-
genitalia. Evidence for sexually antagonistic coevolu- ficient sperm, nuptial gifts, food or nest sites, or
tion comes from comparative data showing that the parental care. Females may also consider whether the
male adaptation of spiny genitalia appears to have male has good or compatible genes. In some species
been met by a female counteradaptation, robust cop- females evaluate males on a single criterion, while in
ulatory ducts (Figure 14.23). We’ll have much more others females consider multiple criteria. Explanations
to say about the battles between the sexes in Chapter for the origin and maintenance of female mate-choice
15 when we describe sexual conflict in relation to preferences when the male does not provide material
parental investment. benefits include indicator mechanisms (good genes) and
332 Chapter 14 / Reproductive Behavior

runaway selection. A third explanation, sensory bias, Sexual selection may lead to sexual conflict, defined
suggests that female preferences for certain traits in as a conflict between the evolutionary interests of males
males could evolve because male traits stimulate an and females. Such conflict is often associated with mat-
existing bias in the female’s sensory system. Female ing and fertilization, and can lead to sexually antagonis-
choice that occurs after copulation is described as cryp- tic coevolution in which an adaptation in one sex is met
tic because it is a hidden, internal choice. with a counteradaptation in the other sex.
 15
Parental Care and Mating Systems

Parental Care early 1800s) reveals an intergenerational cost to bear-

Conflicts Among Family Members over Parental ing sons. Specifically, after controlling for ecological
Investment conditions (different locations) and social class, Ian
Some Factors That Influence the Allocation of Parental Rickard and colleagues (2007) found that offspring
Resources born after elder brothers had similar survival but lower
Overall Patterns of Parental Care lifetime reproductive success than offspring born after
Dispensing with Parental Care—Brood Parasitism elder sisters (Figure 15.1). The reduced lifetime
Mating Systems reproductive success was due largely to lower lifetime
fecundity (number of offspring produced); lower sur-
Classifying Mating Systems
vival rates of offspring played a lesser role. The authors
Monogamy
suggest that mothers who produced sons experienced
Polygyny
a substantial reproductive cost, which made them less
Polyandry able to invest in their next child. Their lower invest-
ment in the next child was then manifested in the lower
In several species of mammals, including humans, lifetime reproductive success of that child.
mothers that produce sons incur greater costs than Reproduction certainly seems to be a complicated
those that produce daughters. The costs experienced affair, with both short- and long-term ramifications for
by mammalian mothers that bear sons vary with family members. In this chapter, we focus on the costs
species, but include such things as higher parasite and benefits associated with reproduction. We begin
loads, delay in the next reproductive effort, reduced with the parent–offspring relationship and examine con-
likelihood of future reproduction, and reduced flicts among family members and how parents allocate
longevity. The costs, however, do not stop there. resources to their offspring. We also examine the diver-
Indeed, a study of human pedigree data gleaned from sity of animal caregivers, including mothers, fathers, and
population registers containing records of births, nonbiological parents. Our second focus is on the mat-
marriages, and deaths from areas of preindustrial ing relationship and the evolution of different mating
Finland (roughly from the early 1700s through the systems.

333
334 Chapter 15 / Parental Care and Mating Systems

a 0.8 survival, though perhaps not immediately. In mammals,

indirect forms of parental care include acquiring and
Probability of surviving to age 15

0.7
defending critical resources, building and maintaining
nests or dens, defending offspring against predators or
infanticidal conspecifics, and caring for pregnant or lac-
0.6
tating females. With respect to the latter, pregnancy and
lactation are energetically demanding times for females,
0.5
so food deliveries to females by mates are especially
helpful and qualify as parental investment. Male owl
0.4
monkeys, for example, feed lactating females and this
provisioning is thought to increase the quantity and/or
0.3
Female Male quality of milk produced for offspring. An added bene-
Sex of elder offspring fit of provisioning in owl monkeys is that it reduces the
interval between births of offspring. Thus, parents
b 3.0 appear to benefit by producing well-fed offspring and
more of them over the course of a lifetime (Wolovich
Lifetime reproductive success

et al. 2008).
2.5 As a rule, patterns of parental investment should
maximize an individual adult’s lifetime reproductive suc-
2.0 cess and not necessarily each reproductive event. Thus,
all parents must make two important decisions. First,
they must decide how much of their own resources to
1.5 devote to reproduction instead of to their own growth
and survival. Second, they must decide how to allocate
the available resources among their offspring (Clutton-
1.0
Female Male Brock and Godfray 1991). It is easy to see how these
Sex of elder offspring decisions can lead to conflicts of interest between par-
ents and offspring and among siblings. We first consider
FIGURE 15.1 In a preindustrial human population in the nature of these conflicts and then look at some of the
Finland, producing sons does not affect the survival
specific factors that influence parental decisions regard-
of subsequent offspring but does affect the lifetime
reproductive success of subsequent offspring. (a) The
ing resource provision.
probability of surviving to 15 years of age in relation to
the sex of the elder offspring. (b) Lifetime reproductive
success, defined as number of children raised to CONFLICTS AMONG FAMILY MEMBERS
15 years of age, in relation to the sex of the elder OVER PARENTAL INVESTMENT
offspring. (From Rickard et al. 2007.)
Evolutionary conflicts over parental investment include
sexual conflict, intrabrood conflict, and interbrood con-
flict (Mock and Parker 1997; Trivers 1974). Recall from
PARENTAL CARE Chapter 14 that sexual conflict is conflict between the
evolutionary interests of males and females (Parker 1979,
Trivers (1972) defined parental investment as any 2006). This conflict takes several forms, but the two
investment by parents in an offspring that increases the main ones concern mating (discussed in Chapter 14) and
survival of that offspring while decreasing the ability of parental investment (our focus here). Both of these sit-
the parent to invest in other offspring. Parental behav- uations involve interactions between males and females
iors are characterized as either direct or indirect during which each individual’s fitness depends on its own
(Kleiman and Malcolm 1981). Direct parental care strategy as well as the strategy of its partner. For parental
includes behaviors that have an immediate physical care, conflict emerges because the costs of providing care
impact on offspring and their survival. In mammals, for are paid separately by each parent, whereas the benefits
example, direct parental care includes behaviors such as accrue to both parents regardless of which one provides
nursing (and feeding), grooming, transporting, and hud- the care (Wedell et al. 2006). As a result of this arrange-
dling with young. Indirect parental care includes behav- ment, each parent should prefer that the other do the
iors that parents may perform while away from the lion’s share of the work when it comes to taking care of
young, so these behaviors may not involve direct phys- the kids. During intrabrood conflict, young try to obtain
ical contact with offspring. Nevertheless, behaviors clas- resources that parents prefer to distribute to other
sified as indirect parental care still affect offspring members of the current brood, and during interbrood
 Parental Care 335

conflict, young try to obtain resources that parents prefer unusually low birth weight, a condition that puts these
to save for future offspring. As an example of conflict piglets at a severe disadvantage for the intense compe-
among family members over parental investment, we tition that characterizes the early postnatal period
consider intrabrood conflict. (Figure 15.2a). Shortly after birth, piglets compete for
It is easy to see how differences in the distribution teats at which to suckle; large piglets typically locate
of resources by parents can lead to sibling rivalry. Bluntly and retain possession of a particular teat or pair of teats,
put, each youngster derives a greater fitness benefit from while smaller piglets fight with littermates to secure a
the parental care it personally receives than from the care location but may be continually displaced, especially if
its siblings receive. In some species, sibling rivalry there are many piglets in the litter. Displaced piglets fail
involves overt, substantial aggression and results in the to obtain milk and colostrum (the fluid released from
death of one or more siblings; such fatal sibling rivalry the teats shortly after parturition that boosts the new-
is called siblicide. In other species, the rivalry is some- born immune system). Most piglet deaths occur in the
what subtler, with lower levels of fighting or siblings rac- first few days after birth, and many are due to starva-
ing to outcompete each other for parental resources (the tion because small piglets expend energy fighting for
latter is called scramble competition; Mock and Parker teat locations only to be routinely displaced before
1997). We first consider an example of sibling rivalry obtaining a meal (studies reviewed by Drake et al.
with relatively low levels of fighting, and then we 2008). Battles among piglets for teats involve frantic
describe the conditions under which siblicide, the most shoving and can include wounding because newborns
extreme form of sibling rivalry, typically occurs. are armed with slashing teeth (Figure 15.2b). The canines
In domestic piglets (Sus scrofa), sibling competition and third incisors of piglets are fully erupted at birth
begins before birth when certain parts of the uterus are and angled in such a way that quick sideways move-
not spacious enough to support maximum growth of ments of the head can lacerate the faces of adjacent sib-
embryos. Some embryos in these areas die early in lings (farmers routinely clip these teeth to prevent piglet
pregnancy while others survive but are characterized by injuries). These slashing teeth seem to function solely

a b

FIGURE 15.2 Sibling rivalry occurs among piglets as they

c compete for access to the teats of the mother. (a) Rivalry
is most intense in the first few days following birth when
piglets compete for a particular teat or pair of teats to which
they return at each nursing bout. (b) At birth, piglets have
outwardly angled canines and third incisors, which they use
as weapons during the early competition for teats. This jaw
is from a newborn piglet; the canine and third incisor on the
right side of the lower jaw have been labeled “c” and “i”,
respectively. (c) Canines and third incisors become increas-
ingly insignificant as the piglet grows, other teeth erupt,
and the orientation of the anterior part of the jaw changes.
Shown here are jaws from a newborn piglet, a 21-day-old
piglet, and an 84-day-old piglet (the typical age of weaning).
The dashed line illustrates how the orientation of the ante-
rior part of the jaw changes with age, so that the third incisor
assumes a more forward orientation typical of incisors.
336 Chapter 15 / Parental Care and Mating Systems

in early sibling competition because they become much If both eggs hatch, the older, stronger chick generally
less significant as other teeth erupt and the jaws grow kills its younger nestmate (Mock et al. 1990). Another
with age (Figure 15.2c). David Fraser has studied possible explanation is that the overproduction of young
maternal care and sibling rivalry in domestic pigs for is an adaptation to a variable food supply. In years when
many years. Here, we include a poem that served as the food is plentiful, all the young may reach adulthood.
abstract for one of his papers on sibling competition However, in years of food scarcity, when sibling compe-
among piglets (From: D. Fraser, and B. K. Thompson. tition for resources is severe, the weaker siblings will be
1991. Armed sibling rivalry among suckling piglets. killed (Clutton-Brock and Godfray 1991). In other cases,
Behavioral Ecology and Sociobiology 29:9–15). Can you extra offspring may be produced to benefit the stronger
determine the experimental design and results of the siblings, either by helping them raise offspring when
study? For the record, the term farrowing refers to the they become adults or by serving as critical meals to pro-
time of parturition in pigs. vide nourishment when conditions are particularly harsh
(Mock and Parker 1998).
A piglet’s most precious possession In view of the many conflicting and competing
Is the teat that he fattens his flesh on. interests associated with parental care, we can ask to
He fights for his teat with tenacity what extent do parents respond to other family members
when deciding how much care to provide? The data
Against any sibling’s audacity.
available suggest great variation in the responsiveness of
The piglet, to arm for this mission, parents to the actions and needs of family members. For
Is born with a warlike dentition example, some parents appear oblivious to the behavior
Of eight tiny tusks, sharp as sabres, of their mate (e.g., house sparrows; Schwagmeyer et al.
2002), while others are extremely responsive to the level
Which help in impressing the neighbors;
of parental effort put forth by their partner and adjust
But to render these weapons less harrowing, their own effort accordingly (e.g., burying beetles;
Most farmers remove them at farrowing. Smiseth and Moore 2004). Similarly, some parents seem
We studied pig sisters and brothers remarkably insensitive to the needs of their offspring
(e.g., guinea pigs; Laurien-Kehnen and Trillmich 2003),
When some had their teeth, but not others.
while others respond to signals of offspring need by
We found that when siblings aren’t many, adjusting the amount of care provided (e.g., canary;
The weapons help little if any, Kilner 1995). In addition, the responsiveness of parents
But when there are many per litter, to other family members varies not only across species,
but also within species (Hinde and Kilner 2007). Let’s
The teeth help their owner grow fitter.
examine some of the specific factors that influence
But how did selection begin parental allocation of resources.
To make weapons to use against kin?

As mentioned above, in some species, sibling rivalry SOME FACTORS THAT INFLUENCE
leads to one offspring attacking and killing its brother or THE ALLOCATION OF PARENTAL
sister. Siblicide is most common in species in which RESOURCES
resources are limited and parents deposit eggs or young Many factors influence the level of care that parents pro-
in a “nursery” with limited space (Mock and Parker vide. Some of these factors pertain to the parents and
1998). The nursery can take various forms, including a some to the young. Here we consider how life history,
uterus, a brood pouch, a parent’s back, a nest, or a den. certainty of paternity, and gender of offspring influence
Although it seems odd at first, siblicide may be advan- the amount of care provided by parents.
tageous to the parents of some species. For example,
when more young are produced than can be raised suc-
cessfully, siblicide can save the parents time and energy Life History
by eliminating the young that are least likely to reach An obvious factor influencing how much effort parents
adulthood. invest in current offspring is the likelihood that the par-
Why would a parent produce more young than it ents will have future opportunities to breed. This, in
can raise successfully? Several answers to this question turn, will be affected by the parents’ age (Trivers 1974)
have been suggested. One idea is that overproduction of and the life span of individuals of that species (Linden
young is insurance in case some eggs or offspring fail to and Møller 1989). On one hand, it might be expected that
develop. In some species of eagles, for instance, the in short-lived species with little hope of producing addi-
female typically lays two eggs, but only one chick reaches tional young in the future, parents would invest more
fledgling age. The eggs are usually laid a few days apart. heavily in the present young. On the other hand, parents
 Parental Care 337

of long-lived species might spend more of their resources fertilizing species run the risk of investing time and
on their continued growth and survival because they energy in raising another male’s offspring, the odds run
might have the opportunity to breed again. against the evolution of paternal behavior. Reliability of
This hypothesis has been tested experimentally with paternity is assumed to be greater when eggs are fertil-
several bird species by handicapping the parents so that ized externally instead of inside the female. Nevertheless,
more parental effort was required to raise the young and certainty of paternity can still be an issue for externally
then determining whether the parents would bear the fertilizing species (see below).
increased costs of reproduction themselves or pass the Although appealing to our sense of intrigue, the idea
costs on to their young. Leach’s storm-petrel that certainty of paternity should influence paternal care
(Oceanodroma leucorhoa) is a relatively long-lived seabird. has a mixed history of support from the scientific com-
Adult petrels make long journeys to ephemeral food munity. Early models, such as those developed by
patches to gather planktonic crustaceans, drops of oil, Maynard Smith (1977) and Werren et al. (1980) raised
and small fish to feed their chicks. A foraging trip usu- questions about the usefulness of the paternity hypothe-
ally lasts two to three days. About 30% of that time is sis as a general explanation for the evolution of patterns
spent airborne, so the cost of flight for a parent that is of parental care. Later models, using different assump-
provisioning chicks is significant. Shortening the wing tions, showed that paternity could influence paternal care,
span by clipping feathers increases the energetic cost of under certain conditions (Houston 1995; Westneat and
flight, raising the cost of reproduction. When parent Sherman 1993). Finally, in a more recent evaluation of this
petrels were handicapped in this way, they passed the issue, Sheldon (2002) concludes that testing models relat-
increased reproductive costs to their offspring and ing paternal care to paternity is extremely difficult (for
maintained their own nutritional condition. Feather example, how does one measure or manipulate “cer-
growth, which is a measure of a bird’s nutritional state, tainty”?) and that we are a long way from understanding
did not differ between adult birds whose wings were whether males of most species adjust parental care in rela-
clipped and untreated control birds. However, the tion to certainty of paternity. Keeping this history in mind,
chicks whose parents’ wings had been clipped grew let’s look at one species for which there is strong evidence
more slowly and spent more nights without food than that certainty of paternity influences level of paternal care.
chicks with untreated parents (Mauck and Grubb As described in Chapter 14, male bluegill sunfish
1995). In contrast, when parents of short-lived species, (Lepomis macrochirus) display a discrete life history poly-
such as starlings (Wright and Cuthill 1990), flycatchers morphism in that they are either parentals or cuckhold-
(Slagsvold and Lifjeld 1988), and tits (Slagsvold and ers (Gross 1982; Neff et al. 2003; Stoltz and Neff 2006).
Lifjeld 1990), were handicapped in a way that increased Parental males compete for sites at which to build their
their reproductive costs, they bore at least part of the nest, guard females that enter their nest to spawn, and
increased costs themselves and continued to allocate then provide sole care to the developing young (care
nearly the same amount of resources to their chicks. entails guarding and fanning the eggs for two to three
Whereas the single chick raised each year by long-lived days until they hatch and then defending the fry from
petrels represents only a small part of the parent’s life- predators for another week or so). Cuckholders, on the
time reproductive success, in these short-lived species, other hand, steal fertilizations from parental males. When
each clutch represents a large proportion of the parent’s small, cuckholders are termed “sneakers” because they
lifetime reproductive success. Thus, in these studies, we hide in vegetation at the edge of a parental male’s nest and
do find that expected life span influences a parent’s allo- then dart into the nest to release sperm at the time the
cation of resources in a way that maximizes lifetime parental male is spawning with a female. Once cuck-
reproductive success. holders reach a large body size, they become female mim-
ics that gain entry into the nests of parental males by their
physical and behavioral resemblance to females. Once
Certainty of Paternity inside the nest, a female mimic releases sperm along with
Trivers (1972) was one of the first to suggest that the parental male at the time of spawning. Sneakers seem
parental solicitude toward young should be correlated to be particularly effective in fertilizing eggs.
with the likelihood of genetic relatedness. Females can Given the interesting and well-studied reproductive
usually be fairly certain that they are related to their off- lives of bluegill sunfish, Bryan Neff (2003) decided to
spring. Certainty of maternity guarantees that 50% of a manipulate the perceived paternity of parental males by
mother’s genes are present in each of her progeny. Males, using two cues. The first cue was the presence of sneaker
especially of species with internal fertilization, cannot be males during spawning; this is an indirect cue of
so confident. Recall from Chapter 14 that even though reduced paternity. The second cue was a water-borne
a male copulates with a female, he has no guarantee that chemical released by newly hatched fry, possibly in their
his sperm, rather than the sperm of a competitor, will urine; parental males use this direct cue to distinguish
fertilize her eggs. In short, because males of internally their offspring from those of sneakers and female mimics
 338 Chapter 15 / Parental Care and Mating Systems

(Neff and Sherman 2003). The chemical cue is not was no difference between treatment and control males
released by eggs, only by fry. in the level with which they defended fry (Figure 15.3a;
Neff performed two experiments using a population compare results for the fry phase). Viewing the overall
of bluegills in Lake Opinicon Canada; this particular pop- results from the first experiment, we see that treatment
ulation has been studied for more than 30 years. In the males increased their level of defense from the egg to the
first experiment, parental males in the midst of spawning fry stage more than control males. Experiment 1 thus
were exposed to four sneaker males held in clear con- provides evidence that male bluegills increase their care
tainers to prevent them from fertilizing any eggs. in response to increases in their certainty of paternity.
Parental males exposed to sneaker males were “treatment In the second experiment, Neff removed one-third of
males.” Parental males in the control group (“control a clutch of eggs from parental males in the treatment
males”) were exposed to empty containers near their group and replaced them with unrelated eggs from other
nests. The day after spawning, treatment males and con- parental males. Parental males in the control group were
trol males were tested for their willingness to defend their exposed to sham swaps of eggs (i.e., Neff went through
brood against an egg predator: a pumpkinseed sunfish the motions of swapping eggs, which entailed taking the
held in a clear bag at the outskirts of each parental male’s eggs, swimming away from the nest, but then returning
nest. Treatment males displayed lower levels of egg the very same eggs to the nest). He assessed the willing-
defense than control males (Figure 15.3a; compare results ness of parental males to defend young before the egg
for the egg phase). The eggs were allowed to hatch, and swap (baseline), the day of the swap, and the day after the
then males were tested again, this time for their willing- eggs hatched. Note that in this experiment, only the sec-
ness to defend their fry. Note that for this test, parental ond cue (the chemical released by fry) was available to
males had available to them chemical cues from fry; such influence male behavior. There was no difference between
cues should allow males to reassess their paternity. There treatment and control males in their willingness to defend
young either before the swap (baseline) or the day after
the swap (during the egg phase) (Figure 15.3b). However,
a b after the eggs hatched (and the chemical cues of fry were
120 120
available), treatment males decreased their level of defense
more than controls. Experiment 2 provides evidence that
male bluegills decrease their care in response to decreases
100 100
in their certainty of paternity. Taken together, experiments
Parental care

Parental care

1 and 2 show that male bluegill sunfish adjust their level

80 80
of care in response to changes in certainty of paternity.

Gender of Offspring
60 60 Sometimes patterns of parental investment are influ-
enced by gender of the offspring. The manner in which
parents distribute resources between the production of
40 40 sons and daughters is called sex allocation. Parents can
Egg Fry Egg Fry Base Egg Fry Base Egg Fry bias their allocation of resources in two main ways: they
Control Treatment Control Treatment can either produce more offspring of one sex, or they
can provide more (or better) resources to offspring of
FIGURE 15.3 Certainty of paternity influences level of
parental care by male bluegill sunfish. (a) In experiment one sex. We focus on the second option concerning
1, males exposed to sneaker males during spawning resources. Although in most animals resources are
(treatment group) display less parental care toward eggs divided about equally between sons and daughters, there
than males not exposed to sneaker males (control group). are some species in which parents distribute resources
Treatment and control groups do not differ during the in a most biased fashion. Consider the case of the brown
fry phase, when males can reassess their paternity using songlark (Cinclorhamphus cruralis) (Magrath et al. 2004,
chemical cues from the fry. (b) In experiment 2, males 2007). Songlarks are polygynous warblers endemic to
whose clutches had been manipulated (one-third of the Australia (polygyny is a mating system in which one
eggs were exchanged with eggs from another male) do male mates with more than one female during the
not differ from control males either before the manipula-
breeding season; see later discussion of mating systems
tion (baseline) or the day after the manipulation (egg
in this chapter). This species exhibits extreme sexual size
phase). However, treatment males do differ from control
males after the eggs hatch and males can assess paternity dimorphism, with adult males being more than twice as
using chemical cues from fry (fry phase). These data indi- heavy as adult females. Mother songlarks are largely
cate that male bluegill sunfish increase (as in experiment responsible for feeding the young. At hatching, nestling
1) or decrease (as in experiment 2) their parental care in males and females do not differ in body mass, but this
response to certainty of paternity. (From Neff 2003.) changes dramatically over the next three weeks, when
 Parental Care 339

b Prey types Mean

proportion
in diet
Spiders (0.12)
a 60

Butterfly and (0.22)

50 moth larvae
Male nestlings
Female nestlings Other (0.19)
40
Body mass (g)

30 Cicadas (0.04)

20
Grasshoppers (0.43)
Age when chicks
10 typically leave the nest
30 20 10 0 10 20 30 40 50
0
5 10 15 20 25 Female bias Male bias
Age (days) Deviation from expected (%)

FIGURE 15.4 Songlark mothers bias their parental investment toward their sons. (a) Although male and female
nestlings have similar body masses at hatching, males become increasingly larger than females in the next few
weeks. These data represent daily means (with standard deviations) for 180 male nestlings and 134 female nestlings.
(b) Mothers not only deliver prey at higher rates to broods with more males, they feed male nestlings a higher quality
diet than they feed female nestlings. For songlarks, a high-quality diet consists of more spiders and fewer grasshop-
pers. In the graph, the prey category “Other” represents prey that occurred very rarely in the diet or that could not be
identified from the videotapes. (Modified from Magrath et al. 2004, 2007.)

males become much heavier than females (Figure 15.4a).

Indeed, by 10 days of age, male chicks are almost 50% STOP AND THINK
heavier than their sisters! It turns out that male Given the above explanation for male-biased care in
nestlings not only receive more prey than their smaller polygynous species, what patterns of parental investment
sisters, but they also receive higher quality prey. The would you expect in monogamous species (monogamy
occurs when an individual male or female mates with only
diet fed to male nestlings contains more spiders and
one partner per breeding season)? Would you predict that
fewer grasshoppers than that fed to female nestlings
parental investment would be biased, and if so, in which
(Figure 15.4b). This is important because spiders have direction? Explain your answer.
certain amino acids that are important for early growth
and development in birds. Also, compared with
grasshoppers, spiders have less chitin (the indigestible
carbohydrate of the exoskeleton). As you might imag-
OVERALL PATTERNS
ine, raising male offspring is costly to mother songlarks;
OF PARENTAL CARE
when experimental all-male and all-female broods were
established, mothers raising all-male broods expended Given the varied life histories of animals, we should
27% more energy than those raising all-female broods. not be surprised to find differences among taxa in the
Given this cost in energy, what is the benefit of raising extent and pattern of parental care. Within vertebrates,
large male offspring? In a polygynous mating system, for example, most teleost fishes (79% of families), frogs
large body size is an important determinant of male and toads (92% of genera), and lizards and snakes
reproductive success; it is typically less important for (97% of genera) show no parental care at all, yet all
female reproductive success (Andersson 1994). Thus, by crocodilians and mammals display some form of
producing especially large sons who, when adult, will parental care (Table 15.1) (Reynolds et al. 2002).
likely be successful in attracting and competing Parental care is also typical of most birds. Indeed, only
for mates, mothers ensure that their genes are well- a small percentage of species (1%) dispenses with
represented in future generations. Producing large parental care by laying their eggs in the nests of oth-
females would not yield as large a fitness benefit. We ers and relinquishing all care to the “host” parents (see
now turn our attention to the basic patterns of parental discussion of brood parasitism later in this chapter)
care displayed by animals. (Cockburn 2006).
340 Chapter 15 / Parental Care and Mating Systems

TABLE 15.1 Patterns of Parental Care Exhibited by Some Groups of Vertebrates. Data Pertain to the
Taxonomic Level Indicated in Parentheses After Each Group
Group No care Male-only care Female-only care Biparental care

Teleost fishes (422 families) 79% 10% 1% 3%

a
Frogs and toads (315 genera) 92% 9% 9% 1%
Lizards and snakes (938 genera) 97% 0% 3% 0%
Crocodilians (21 species) 0% 0% 62% 38%
Birds (9456 species)b 1% 1% 8% 81%
c
Mammals (1117 genera) 0% 0% 91% 9%
a
Percentages add up to more than 100% because some genera show more than one form of care.
b
Percentages add up to 91% because 9% of birds have three or more individuals providing care to young. This pattern, called cooperative
breeding, is discussed in Chapter 19.
c
Cooperative breeding also occurs in about 120 species of mammals.
Modified from Reynolds et al. (2002); data for birds from Cockburn (2006).

Who Provides the Care for feeding offspring, and thus biparental care evolves.
In those taxa for which parental care is the norm, there Fish and amphibian parents, on the other hand, rarely
are differences with regard to who provides the care feed their offspring, and parental duties consist largely
(refer again to Table 15.1). For example, whereas of guarding, a task that may be performed almost as well
female-only care is the most common form of care in by one parent as by two (Clutton-Brock 1991). Because
mammals (91% of genera; Kleiman and Malcolm male mammals cannot feed their offspring right after
1981), biparental care is the most common form of birth when the young are dependent on milk, early
care in birds (81% of species; Cockburn 2006). Why paternal care in mammals is always in conjunction with
the difference? In mammals, internal gestation and lac- maternal care.
tation necessitate a major parental role for the female Having stated the broad generalization of the ade-
and restrict the ability of the male to help in early off- quacy of solitary male care in fishes and amphibians, we
spring care. Male mammals cannot take over the duties should point out that there are exceptions to this rule,
of pregnancy and lactation for their mates, and rather including some cases of biparental care in fishes. For
than hang around during the period of early develop- example, cichlids, a lineage of freshwater teleosts in
ment, males, in most cases, seek mating opportunities Africa, South and Central America, and India, display
elsewhere. In contrast to the extended period of inter- elaborate parental behavior that includes biparental care.
nal development in mammals, birds develop outside Indeed, about 40% of cichlid genera display biparental
the mother’s body. Embryos, along with food in the care; for comparison, among teleosts overall, only about
form of yolk, are packed in eggs that develop largely 3% of genera show biparental care (Goodwin et al. 1998;
in the external environment. Because male birds are Reynolds et al. 2002). What is it about cichlids that has
just as capable as their mates at providing care, parental led to the evolution of biparental care? Biparental care
duties such as incubation, feeding, and guarding can be may have evolved in some cichlids because their broods
divided more equally between the sexes. Here we see face intense predation pressure, and two parents may be
that basic biological attributes of a lineage can con- necessary to successfully rear the young (McKaye 1977).
strain evolutionary possibilities. Another explanation seems possible for the biparental
A second example of taxon differences among ver- discus (Symphysodon discus). Two days after fertilization,
tebrates concerns whether male care occurs alone (male- the fry of this cichlid hatch, aided by both parents who
only care) or in conjunction with female care (biparental chew open the egg cases. The two parents then deposit
care). When male care occurs in fishes and amphibians, their youngsters on aquatic vegetation, where the wrig-
it usually takes the form of solitary male care rather than glers dangle from threads as they live off their yolk sup-
shared male and female responsibilities. In contrast, ply. Within two or three more days, however, the brood
paternal investment in birds almost always occurs in can swim freely, and they attach themselves to their par-
addition to maternal care. The higher frequency of ents and feed off parental skin secretions (Figure 15.5).
biparental care in birds probably reflects the fact that Biparental care may have evolved in this species because
parental care in this group usually involves both feeding young that can feed off two parents may grow and
and guarding the young. Two parents are better than one develop more rapidly than those that have only a single
 Parental Care 341

only care; and biparental care (Reynolds et al. 2002).

(Note that in some animals young are cared for by
more than two individuals, a system called coopera-
tive breeding; see Chapter 19). Given their diverse
forms of parental behavior, teleosts and anurans have
often been the focus of studies on the evolution of
parental care patterns. Both groups also have some
members with external fertilization (fertilization in the
external environment) and some with internal fertil-
ization (fertilization within the female reproductive
tract). This situation allows an examination of how
mode of fertilization influences the evolution of patterns
of parental care. We examine this relationship in ray-
finned fishes, the larger group of bony fishes to which
teleosts belong.
Several authors have suggested an association
between external fertilization and male-only care and
between internal fertilization and female-only care
(Gross and Sargent 1985; Ridley 1978; Trivers 1972).
To test this association in ray-finned fishes, Judith
Mank and colleagues (2005) surveyed the literature and
constructed a comprehensive phylogeny using phylo-
genies based largely on molecular data, though they
also included some morphology-based phylogenies.
The researchers then used the structure of the resulting
phylogeny to examine the evolution of different modes
of parental care (female-only, male-only, or biparental)
in relation to several factors, including mode of
fertilization. The results indicated that female-only,
male-only, and biparental care have arisen repeatedly
and independently in ray-finned fishes (Figure 15.6).

FIGURE 15.5 Young cichlids of the biparental species Female-only care

Symphysodon discus graze on parental skin secretions.

Internal External Biparental Male-only

parent (Skipper and Skipper 1957). We see that in some gestation guarding care care
fish—such as those exposed to extreme levels of preda-
tion or those whose care involves the feeding of young—
biparental care may be necessary to ensure the survival 8-13
4
and growth of the offspring. 22-27
Patterns of parental care among vertebrates are cer- 3-5
tainly diverse, and differences exist both among and Internal
4-5
within taxonomic groups. As we have seen, some of the fertilization
3-6
differences among groups reflect basic biological differ-
ences, such as where the young develop and how they
13-18
are fed. We have also seen that differences within groups
can sometimes be understood in light of particular eco-
logical conditions, such as intensity of predation. We External fertilization,
no parental care
continue our consideration of factors that influence the
extent and type of parental care displayed by animals by
FIGURE 15.6 Diagram showing the major independent
looking at how mode of fertilization affects patterns of evolutionary transitions among modes of parental care
parental care. in ray-finned fishes. Size of arrows reflects the relative
Teleost fishes and anuran amphibians (frogs and numbers of evolutionary transitions, and numbers next
toads) display four of the general categories of parental to arrows indicate the range of evolutionary transitions
care in vertebrates: no care; male-only care; female- estimated. (Modified from Mank et al. 2005.)
342 Chapter 15 / Parental Care and Mating Systems

The most common pathway was from external fer- ual selection in Chapter 14, that patterns of parental
tilization and no care leading directly to male-only investment are probably not the sole determinant of
care. The second most common pathway was from sexual selection.)
external fertilization and no care to internal fer- Sometimes, however, the sex roles are reversed, and
tilization to female-only care. Occurring much less the burden of parental care falls largely or entirely on
frequently were transitions from external fertiliza- the male. When parental investment by males is greater
tion and no care to either female-only care or than that of females, we would expect males to be
biparental care. choosy and females to be competitive. We describe two
But what is the source of the association between species of birds known as jacanas that exhibit the phe-
mode of fertilization and mode of parental care? One nomenon of sex-role reversal in parental care. These
possibility is that it simply relates to the proximity of species exhibit full sex-role reversal (i.e., males provide
adults and offspring (Williams 1975). According to this almost all the parental care and are selective in their
hypothesis, external fertilization, particularly when it choice of mate, and females compete for access to
occurs in a territory defended by a male, would be asso- males). There are other animal species that exhibit par-
ciated with parental care by males. Because territorial tial sex-role reversal whereby males are choosy and
males are almost always in the neighborhood of the eggs competitive or both sexes are choosy. With the notable
they fertilize, paternal behavior is likely to evolve. In exception of mammalian species, some level of sex-role
contrast, with internal fertilization, it is usually the reversal has been reported in insects and birds, as well
female that carries the embryos, and thus she is in the as in crustaceans, fishes, and amphibians (see reviews by
best position to care for the young when they enter the Eens and Pinxten 2000; Bonduriansky 2001). Overall,
world. Paternal behavior is less feasible because fathers however, sex-role-reversed species make up a minority
may not be in the vicinity when the eggs are laid or the of animal species.
young are born. The northern jacana (Jacana spinosa) is a small bird
of tropical marshes best identified by its spindly green-
ish legs and toes. Donald Jenni and Gerald Collier
Sex-Role Reversals (1972) conducted an extensive study of the population
In many animals, females provide more parental dynamics, behavior, and social organization of individ-
investment than males, and this is thought to explain ually marked jacanas in Costa Rica. They found that
sex differences in mating competition. Trivers (1972) northern jacanas have a polyandrous mating system in
suggested that the sex with greater parental investment which a given female is simultaneously paired with sev-
(in this case females) becomes a limiting resource, eral males. At their study site, individuals breed year-
essentially an object of competition among individuals round, and a female defends a large territory that
of the sex investing less (here, males). The greater encompasses from one to four male territories. In addi-
investment by females in their offspring also means tion to helping her mates defend their individual terri-
that females are not available for other reproductive tories, the female independently repels intruders from
opportunities. In contrast, males, unencumbered by her entire territory. A female’s critical role in territorial
parental care, are able to take advantage of such oppor- defense is matched by her dominant role in courtship.
tunities. This results in two things. First, the opera- Because female northern jacanas keep harems of males,
tional sex ratio, defined as the ratio of potentially many females are excluded from breeding. The result is
mating males to fertilizable females, becomes male- heightened competition among females for males.
biased; from this we might also expect that males, as Morphological correlates appear to accompany the
the more abundant sex, would be the more competi- female jacana’s male-like role in competition for males
tive sex (Emlen and Oring 1977). And second, the during courtship: breeding females weigh approxi-
potential reproductive rate, defined as the maximum mately 145 grams; adult males typically weigh in at a
number of independent offspring that each parent can trim 89 grams.
produce per unit of time, becomes greater for males Parental activities are the province of the male
than for females. Clutton-Brock and Vincent (1991) (Jenni and Betts 1978). Nest building and incubation
predicted that mating competition would be more are male duties exclusively, a division of labor under-
intense in the sex with the higher potential reproduc- scored by the fact that only male jacanas have incuba-
tive rate. Taken together, the end result of greater tion patches (highly vascularized bare patches of skin on
investment by females in parental care is that males the belly). Once the precocial young have hatched,
compete among themselves for access to females (a sit- males provide most of the care to the chicks, including
uation that favors large body size and aggressiveness in brooding and defending them. When necessary, females
males), and females in turn are selective in their choice back up males in confronting potential predators;
of mates. (Recall, however, from our discussion of sex- indeed, the female’s larger body size seems to make her
 Parental Care 343

Other brood parasites lay their eggs in the nests of

other species and are called interspecific brood par-
asites; these species are also described as obligate brood
parasites because they have no other reproductive
option (i.e., they never build nests in which to lay eggs
and raise their own young). The term parasite is used
because of the apparent benefit experienced by the true
parents and the harm that befalls the host parents, who
typically experience a reduction in reproductive suc-
cess (see below). We focus on interspecific (obligate)
brood parasitism.
Approximately 1% of all species of birds lay their
eggs in the nests of another species (Cockburn 2006).
Obligate brood parasitism has arisen independently sev-
eral times within birds, occurring in honeyguides, Old
FIGURE 15.7 Many jacana species exhibit sex-role World cuckoos, New World cuckoos, viduine finches,
reversal. Here a male African jacana cares for his chick cowbirds, and the black-headed duck (Sorenson and
and eggs. Payne 2002).
We mentioned that raising the young of brood par-
more effective than the male at predator deterrence asites is costly to host parents. Damage to the host or
(Stephens 1984). its young may be directly inflicted by either the para-
The pattern of sex-role reversal described for sitic adult or its offspring (reviewed by Davies 2000).
northern jacanas is also typical of several other jacana The adult, for example, must place its egg in the host’s
species (Figure 15.7). Observations of the wattled jacana nest when the host is beginning to incubate its own
(Jacana jacana) suggest that females no longer have incu- eggs. Upon discovering a host’s nest after incubation or
bation as part of their behavioral repertoire, but that they hatching has begun, a female cuckoo may eat the eggs
retain the ability to perform all of the behaviors involved or kill the young of the host, causing the potential host
in the care of chicks, though they rarely perform them. to nest again. If the cuckoo is on time, however, she
Indeed, during seven field seasons in Panama, Stephen may throw out a host’s egg before laying one of her
Emlen and Peter Wrege (2004) never observed incuba- own in the nest. To add insult to injury, cuckoos often
tion by females, even when their mates disappeared lay their eggs from a perch above the nest, and their
when incubating eggs. Also, in 97% of the 242 observed
broods, females did not associate at all with their depen-
dent chicks.

DISPENSING WITH PARENTAL CARE—

BROOD PARASITISM
Some birds exploit the parental behavior of other
birds. These birds, known as brood parasites, lay their
eggs in the nests of other birds and leave all subsequent
parental care to the foster (“host”) parents. Some
brood parasites lay their eggs in the nests of con-
specifics; these species are called intraspecific brood
parasites. Included in this category are species such as
cliff swallows, red-fronted coots, and wood ducks. In
some cases of intraspecific brood parasitism, the
brood parasite occasionally lays eggs in the nests of
conspecifics, while still laying eggs in her own nest. In
other instances, the brood parasite lays eggs in the
nests of conspecifics and does not maintain a nest of
her own; females in this group may resort to para-
sitizing nests because independent reproduction is FIGURE 15.8 A nestling parasitic cuckoo is evicting a
impossible due to a shortage of nest sites or territories. host’s egg from the nest.
 344 Chapter 15 / Parental Care and Mating Systems

thick-shelled eggs break the host’s eggs when they two or more chicks, even when nests have not been
strike them. Nestling cuckoos are renowned for parasitized (Broom et al. 2008).
methodically evicting eggs or young from the nest of In response to the devastating effects of brood par-
their foster parents. By positioning itself under a nearby asitism, host species have developed ways to avoid
egg or nestling, a young cuckoo may lift a nestmate being parasitized. The most common defenses are
onto its back, slowly work its way to the edge of the those used to reduce predation—the host species sim-
nest, and nudge the host’s egg or nestling to its death ply conceal their nests and defend them when they
below (Figure 15.8). Nestling honeyguides employ an are discovered. Some hosts identify and remove the
even more gruesome tactic to ensure full attention from eggs (or young) of parasites from their nests.
their foster parents. At hatching, young honeyguides Recognition and rejection of parasitic eggs is a tricky
use their hooked bills to kill young with whom they business because the host could make a recognition
share the nest. mistake and erroneously throw out its own egg when
An alternative strategy to the outright killing of parasitism has not occurred. Consistent with this view
foster siblings is simply to monopolize parental care. that egg recognition can be a challenging endeavor is
Brood parasites usually mature more rapidly than a the observation that hosts take longer to reject eggs
host’s young, thus gaining a critical head start in that more closely resemble their own eggs (Antonov et
growth and development. Also, their huge mouths and al. 2008).
persistent begging often elicit preferential and pro- In response to the antiparasite devices of hosts,
longed feeding by foster parents. The host’s young may many aspects of the laying behavior of brood parasites
be no match for their larger, more aggressive foster are designed for better deception of hosts. For example,
sibling and may die from starvation, crowding, or cuckoos time their laying for the late afternoon, when
trampling. Sometimes, however, parasitic young ben- hosts are less attentive, and whereas some passerines
efit by keeping a few of the host’s young around. spend at least 20 minutes laying an egg in their own
Parasitic brown-headed cowbird nestlings exhibit nests, female cuckoos can deposit an egg in a host’s nest
increased growth and survival when they share the nest in less than 10 seconds! Parasitic eggs or young often
with one or two host young as compared to when they resemble those of the host species, a ploy that may
are the sole occupant of the host’s nest. Host parents reduce the chances of host parents recognizing them as
apparently increase the rate at which they feed larger different and rejecting them (Figure 15.9). Coevolution
broods, and the young cowbird gets more than its fair between the brood parasite and host may produce adap-
share of the stepped-up feedings (Kilner 2003; Kilner tations and counteradaptations of increasing complex-
et al. 2004). Also, killing host young might be a risky ity (Davies and Brooke 1988; Rothstein and Robinson
strategy for brood parasitic young if host parents, as 1998).
part of their life history strategy, are more likely to We turn now from the interactions between parents
desert single chick broods as compared to broods with and offspring to those between mates.

a b

FIGURE 15.9 Parasitic eggs or young often resemble those of host species. Presumably such resemblance
reduces the chances that hosts will reject different-looking eggs or young from their nest. (a) The egg of the
brood parasitic common cuckoo (upper left) closely resembles those of its host, the great reed-warbler, but is
somewhat larger. (b) A nestling of the brood parasitic pin-tailed whydah (left) has elaborate mouth markings and
begging behavior that resemble those of the young of its host, the common waxbill. The two smaller nestlings
are common waxbills.
 Mating Systems 345

breeding season. Accordingly, monogamy occurs when

STOP AND THINK a male and female have only a single mating partner per
Viduine finches (whydahs and indigobirds; genus Vidua) are breeding season; polygyny when some males copulate
obligate brood parasites within the Family Estrildidae that with more than one female during the breeding season;
typically parasitize a single host species. Hosts, such as wax- and polyandry when some females mate with more than
bills and firefinches, are also members of Family Estrildidae,
one male during the breeding season. Polygynandry,
and all nestlings in this family have striking mouth mark-
sometimes called promiscuity, occurs when both males
ings. The mouth markings of parasitic nestlings precisely
match those of host nestlings (look again at Figure 15.9b). and females mate with multiple individuals. Ecological
Traditionally, the resemblance in mouth markings between factors, such as predation, resource quality and distrib-
parasite and host has been viewed as the outcome of a ution, and availability of receptive mates, affect the need
coevolutionary arms race, with host rejection of offspring for parental care, the ability of males to monopolize
that look different from their own as the driving force. females, and the ability of females to choose among
Mark Hauber and Rebecca Kilner (2007) suggest a differ- potential suitors (Emlen and Oring 1977). Because such
ent scenario: perhaps the elaborate mouth markings have ecological conditions often vary within and between
evolved through nestling competition for parental care. In locations, considerable flexibility is usually associated
a reversal of the traditional view that selection acts on par- with the mating patterns of a given species. For exam-
asitic young to look like their hosts, they further suggest
ple, black howler monkeys, best known for their loud
that selection might be acting on host young to look like
calls that can travel up to 3 miles through dense forest,
the more competitive parasitic young. How might you
design an experiment to test whether the elaborate mouth were found to be primarily polygynous in a deciduous
markings of nestlings in Estrildidae function as signals of habitat and polygynandrous in a riparian (alongside a
species identity (the traditional view) or as structures to river) habitat, when studied in the Central American
attract parental provisioning? country of Belize (Jones et al. 2008). The lush riparian
habitat, characterized as more productive and pre-
dictable than the deciduous habitat, also supported
higher population densities of howler monkeys overall
MATING SYSTEMS and larger groups of females, perhaps making the
monopolization of females by a single male less likely.
We have seen that parents and offspring often disagree Our coverage of mating systems will focus on
over the details of their interactions. As we will see in the monogamy, polygyny, and polyandry.
pages that follow, such conflict is not restricted to the At the start of our discussion we should say that
parent–offspring relationship. In fact, adult males and regardless of mating system, sexual fidelity is hard to find
females are often at odds over what constitutes an ideal among animals. Genetic analyses of parents and off-
mating relationship. Although the ultimate goal of spring often reveal that apparent social partners mate
reproduction for both sexes is to maximize fitness (the with other individuals: the social father is not always the
relative number of offspring that survive and reproduce), genetic father of all the offspring. This has led
the reproductive success of males and females is con- researchers to distinguish between social monogamy
strained by different factors. In many species, a male’s (an exclusive living arrangement between one male and
success is limited by access to females, while a female’s one female that makes no assumptions about mating
is limited by access to resources (Davies 1991). exclusivity or biparental care) from genetic monogamy
Consequently, a male can often boost his reproductive (an exclusive mating relationship between one male and
success by mating with more than one female. In con- one female). DNA studies of the offspring of 180 species
trast, a female increases her reproductive success by of socially monogamous songbirds showed that only
gathering more resources, including male parental care 10% of these species were genetically monogamous
and access to a high-quality territory. Generally speak- (Morell 1998). Although less studied than avian species
ing, then, males focus on mating effort, and females tend in this regard, some species of socially monogamous
to emphasize parental effort. Therefore, we should not fishes and mammals also engage in extra-pair fertiliza-
expect perfect parental harmony during the production tions (Clutton-Brock and Isvaran 2006; Ophir et al.
of offspring. We would predict instead that each parent 2008; Sefc et al. 2008). Cuckoldry can be a problem for
will attempt to maximize its own reproductive success, polygynous males as well. Indeed, extra-pair matings
even if this is costly to the other (Trivers 1972). seem to be the rule rather than the exception.
Extra-pair matings have costs and benefits, espe-
cially if they result in fertilizations. A male’s costs include
CLASSIFYING MATING SYSTEMS
the time and energy used in searching for receptive
We begin our consideration of mating systems with females other than his mate. While he’s away, his primary
some definitions. Mating systems can be defined based mate may copulate with another male, which reduces his
on the number of copulatory partners per individual per reproductive success. However, if he is successful in
346 Chapter 15 / Parental Care and Mating Systems

inseminating mates of other males, he can boost his the dominant pair increased, so too did the proportion
reproductive success substantially. of extra-pair young in broods (Figure 15.10). In fact,
Hypotheses for female benefits of extra-pair matings most females that obtained extra-pair fertilizations did
can be grouped into two categories: material or genetic. not obtain any fertilizations from their partner
By mating with several males, a female may gain added (Blackmore and Heinsohn 2008). Now we describe the
assistance in raising her offspring. In some species, such major mating systems and the circumstances under
as red-winged blackbirds, extra-pair males help defend which each is likely to evolve.
the female’s nest from predators and thereby improve
her fledging success (Gray 1997). Females may also
exchange copulations for a valuable resource, food, for MONOGAMY
instance. (In Chapter 14, the material benefits of nup- Monogamy results when a male and female have only a
tial gifts presented to the female at the time of copula- single mating partner per breeding season. Because sperm
tion were discussed.) Extra-pair matings may provide from one male is often sufficient to fertilize a female’s lim-
sufficient sperm to fertilize all the eggs of a given female ited number of eggs, monogamy is often sufficient from
(Birkhead 1996). the female perspective, but for males, confining copula-
There is also evidence that females seek extra-pair tion to a single female is a rather conservative means of
matings for genetic reasons, such as obtaining “good ensuring genetic representation in the next generation.
genes” for their offspring. Female reed warblers seek We might wonder, then, why a male would be monoga-
extra-pair matings with males with a larger song reper- mous and what ecological circumstances might favor
toire than their mate’s. Although they don’t get mater- monogamy over polygyny. Several hypotheses, which are
ial benefits from the extra-pair mate, they apparently do not mutually exclusive, have been proposed (Kleiman
get good genes for their offspring: postfledgling sur- 1977; Komers and Brotherton 1997; Whiteman and Côté
vivorship of the young is related to the genetic father 2004). Here, we consider a few of them.
(Hasselquist 1998; Hasselquist et al. 1996). Finally,
females of some cooperatively breeding bird species
engage in extra-pair matings to avoid mating with close Necessity of Biparental Care
relatives (recall from Chapter 14 that extreme inbreed- When biparental care is necessary or at least important
ing can be costly). Gray-crowned babblers live in social for offspring survival, monogamy may be favored. As we
groups that consist of a dominant breeding pair and non- mentioned previously, biparental care is more common
breeding helpers (typically offspring from previous sea- among birds than mammals. Nevertheless, males of
sons). Sometimes, members of the dominant pair are some mammalian species do have important parental
related, and it is under these circumstances that extra- responsibilities, and the fitness of both mates depends on
pair young are found in the nest. A recent field study in the male’s parental investment. Monogamy is rare
which DNA samples were taken from young and adult among mammals; it has been reported for only about 5%
members of social groups showed that as relatedness of of species (Kleiman 1977). Some rodents are monoga-
mous, and our examples come from this group.
Male California mice (Peromyscus californicus) pro-
1.0 vide extensive care to their young. Indeed, with the
exception of nursing, fathers participate in all parental
Proportion of young sired

0.8 activities and to the same extent as mothers. The need

by extra-pair male

for paternal care seems to have played a role in the evo-

0.6 lution of monogamy in this species. Genetic analyses
have shown that once paired, these mice never stray
0.4 (Ribble 1991). Pups are typically born at the coldest time
of the year, and because they cannot maintain their own
0.2 body temperature until about 2 weeks of age, they need
their parents’ body heat to survive. Both parents take
0.0 turns huddling over the pups to keep them warm. The
−0.2 −0.05 0.1 0.25 importance of father presence to pup survival in
Relatedness of dominant pair California mice has been documented in the field and
FIGURE 15.10 In the cooperatively breeding gray-
under challenging conditions in the laboratory (cold tem-
crowned babbler, the proportion of young sired by peratures and when parents must wheel run for food); in
extra-pair males increases with relatedness of the both settings, experimental removal of fathers resulted in
dominant pair, suggesting that females of this species lower pup survival compared to control pairs (Table 15.2).
engage in extra-pair matings to avoid inbreeding. Thus, in California mice, paternal assistance may improve
(From Blackmore and Heinsohn 2008.) offspring survival to such an extent that it is better for a
 Mating Systems 347

TABLE 15.2 Survival of Young California Mice a 32

When Fathers Are Present or Absent under Male removal
Laboratory and Field Conditions. 31 Control
Pup survival (%)
30
Father Father
Conditions present absent

Days
29
Laboratory
Cold temperaturesa 90 55 28
b
Parents wheel run for food 83 45
Field 81 26 27

a
Mice were maintained in environmental chambers at 8.5–10.5°C;
26
this temperature matched temperatures encountered in the field 1st and 2nd Litters 2nd and 3rd Litters
during the coldest months of the breeding season.
Interlitter intervals
b
Mice were maintained in foraging chambers equipped with a run-
ning wheel; adults were required to run 250 revolutions in a wheel
for each pellet of food. b 34

Interlitter interval (days)

Laboratory data from Gubernick et al. (1993); field data from 32
Gubernick and Teferi (2000).
30

28
male to remain with one female and invest in offspring
than it is to seek additional mating opportunities 26
(Gubernick et al. 1993; Gubernick and Teferi 2000).
24
Presence of the father can increase reproductive suc-
cess in ways other than affecting survival of young—for 22
90 95 100 105 110 115 120 125
example, it can shorten the interval between litters. A
Time spent in the nest by males (hours)
shorter interlitter interval can be critical to lifetime
reproductive success, especially in short-lived species. FIGURE 15.11 Presence of fathers in mound-building
Consider the case of the mound-building mouse (Mus mice leads to shorter intervals between litters.
spicilegus), a monogamous rodent found in Central and (a) Length of interval (mean ± SE, in days) between
Eastern Europe. (By the way, the name reflects the delivery of first and second litters and delivery of second
mouse’s habit of building enormous mud and vegetation and third litters for control pairs in which mates were
mounds in which to spend the winter; once spring left together and for pairs in which males were removed
arrives, individuals disperse from the mounds to repro- one day after the second litter was born. (b) Data for
duce.) Christophe Féron and Patrick Gouat (2007) seven males videotaped over six days reveal a negative
correlation between time spent by males in the nest and
maintained pairs of mound-building mice in the labora-
length of interval between litters: the more time that
tory and removed males from some pairs one day after males spent in the nest with offspring, the shorter the
delivery of their second litter (males and females mate interval until birth of their next litter. (From Féron and
shortly after the female gives birth, so even these pairs Gouat 2007.)
had time to mate). Males and females of other pairs were
left together, and the researchers monitored production
of third litters in the two groups. All females produced for six days to determine the amount of time spent in the
a third litter, but those whose mates were left with them nest by individual males (taping began after postpartum
produced their litter after a shorter interval than those sexual behavior had ended to ensure that time spent in
whose mates were removed (Figure 15.11a). The the nest by males was devoted to care of young and not
extended pregnancies of females whose mates had been to mating activities). The data from this experiment show
removed presumably were caused by delayed implanta- that male mound-building mice vary in the amount of
tion of embryos, a response that may reflect the high time they spend in the nest with young: among the seven
energetic costs of concurrent pregnancy and lactation, males, total time in the nest for six days ranged from
along with the absence of help from their mate. Litter about 94 to 120 hours. Especially interesting was the
size at birth and survival of pups to weaning did not dif- finding of a negative correlation between male time in
fer between the two groups. nest and interlitter interval: the more time a male spent
In another experiment, Féron and Gouat (2007) con- in the nest, the shorter the interval until the next litter
tinuously videotaped seven male–female pairs with young (Figure 15.11b). Although the specific mechanism by
348 Chapter 15 / Parental Care and Mating Systems

which males shorten the interval between litters remains a

to be identified, this example shows that when male
mound-building mice stay with their mate and provide
paternal care to their offspring, both partners benefit by
having shorter intervals between litters.

Distribution of Females
The distribution of females throughout the habitat is
another factor that influences mating systems (Clutton-
Brock and Isvaran 2006; Emlen and Oring 1977).
Circumstances that make it difficult for a male to monop-
olize multiple mates will favor monogamy over polygyny.
For example, when receptive females are uniformly dis-
tributed, perhaps because they defend exclusive territo-
ries, monogamy may evolve. Also, if females are widely b 60
dispersed, it might be more beneficial for a male to
remain with a given female than to search endlessly for

Numer of hosts
additional mates. With monogamy, a male is at least 40
assured of access to one mate.
Consider the case of the symbiotic shrimp Pontonia
margarita, which lives inside the mantle cavity of the pearl 20
oyster. (The mantle cavity of oysters and other mollusks
is a semi-internal cavity containing the gills and into which
digestive, excretory, and reproductive systems discharge
0
their products.) Pearl oysters are found at low densities in 0 1 2 3 4 5
the Pacific Ocean from the Gulf of California to off the Number of shrimp per host
coast of northern Peru, and as permanent living quarters,
they are quite small, even for shrimp. The maximum shell FIGURE 15.12 Shrimp that live within oysters may be
height of pearl oysters is about 14 cm, and the average monogamous because their hosts are small and scarce.
carapace lengths of adult male and female shrimp are (a) A pair of symbiotic shrimp retrieved from their host,
a pearl oyster. (b) A survey of 68 oysters revealed that
around 11 and 14 mm, respectively (the carapace is the
most contained two shrimp, and only a few had one or
hard covering of chitin on the back of the shrimp). Based no shrimp. All pairs were male–female pairs. (From
on the relatively small size and scarcity of their hosts, J. Baeza 2008.)
Antonio Baeza (2008) predicted that P. margarita would
be monogamous. He collected 68 oysters and recorded
the number of shrimp per host, sex of each shrimp, and ter their original oyster. Here, then, we have an example
the reproductive condition of all females. His data where several factors—scarcity of host, small size of host,
revealed that the number of shrimp per host varied from and potentially high costs of searching for additional
zero to two (most oysters with one or no shrimp were mates—seem to make monogamy the best option.
small), with two being most common (Figure 15.12). All
shrimp pairs consisted of a male and a female, and females Mate Guarding
were either close to spawning (eggs visible in ovaries) or Sometimes monogamy evolves because a male can effec-
brooding embryos of various stages. Body sizes of shrimp tively guard only one female from other males. Kirk’s
and host oysters were closely correlated, suggesting the dik-dik (Madoqua kirkii) is a small antelope (only about
particular associations are long lasting (frequent switch- 15 inches tall) that lives in dry scrub in many parts of
ing among hosts, as occurs in some symbiotic crustaceans, Africa. Dik-diks form monogamous bonds that last for
would result in a loose relationship between shrimp size several years, if not for an entire lifetime. Unlike many
and oyster size). These data are consistent with social species, dik-diks seem to be faithful to their mates:
monogamy for P. margarita. Although genetic data are not genetic analyses revealed no evidence of extra-pair pater-
yet available, Baeza suggests that genetic monogamy may nity (Brotherton et al. 1997). Given the rarity of
occur in P. margarita for at least two reasons. First, mov- monogamy among mammals, we might wonder why
ing between hosts in search of extra-pair copulations such devotion has evolved in dik-diks.
would likely be rare in these shrimp because they live in Several hypotheses can be ruled out. The need for
predator-rich waters and are especially vulnerable when biparental care can’t be important in the evolution of
away from their hosts. Second, shrimp that temporarily monogamy in this species because paternal care is
leave their host may encounter difficulties trying to reen- absent—the male does not defend resources or reduce
 Mating Systems 349

predation risk, and there isn’t a risk of infanticide such as nest sites or territories. Also, the activity around
(Brotherton and Rhodes 1996). Furthermore, at least these areas may attract predators, and other receptive
some monogamous males defend territories that could females may increase the competition for commodities
support more than one female. In a rare population in such as food.
which some males were polygynous, the territories of the
polygynous males were not of higher quality than those
Hypotheses for Females’ Acceptance
of monogamous males. And it seems unlikely that
of Polygyny
monogamy evolved because constant male attention is
necessary to keep females from wandering off the terri- Why would a female accept the costs of polygyny? There
tory and mating with another male. Pair mates spend is no simple answer to this question. Indeed, the answers
only about 64% of their time together, and so there may vary among species and even within a species. Of
would be ample opportunity to wander away. the several hypotheses proposed to explain why females
Instead, it seems that monogamy evolved because the accept the costs of polygyny, we consider two—the
costs of guarding more than one female from intruding polygyny threshold hypothesis and the “sexy son”
males would be too great. The male guards his female by hypothesis.
preventing other males from knowing when she is in
estrus. He does this by covering up the scent of his Polygyny Threshold Hypothesis Polygynous matings
female’s territorial markers—piles of dung—by scratch- will be advantageous to females when the benefits
ing dirt over them and then defecating on top of them. achieved by mating with a high-quality male and gain-
It is also essential that a male advertise his territorial own- ing access to his resources more than compensate for
ership because rivals bent on filling vacant territories are costs. In other words, a female may reproduce more suc-
never far off. This is accomplished by marking his terri- cessfully as a secondary mate on a high-quality territory
torial borders with the scent from glands under his eyes than as a monogamous mate on a low-quality territory.
(Komers 1997). If a male were to try to overmark the Jared Verner and Mary Willson (1966) coined the phrase
scent of two females on separate territories, he might fail polygyny threshold to describe the difference in a territory’s
to mark his territory sufficiently. As a result, he could lose quality needed to make secondary status a better repro-
ownership in the intense competition for vacant territo- ductive option for females than primary status.
ries. In the few instances in which a male dik-dik suc- Gordon Orians (1969) elaborated on Verner and
ceeded in polygyny, the females shared a territory. The Willson’s ideas and reasoned that if polygyny is always
female accepts being guarded because an extra-pair mat- to a male’s advantage and yet does not always occur, then
ing might incite a fight between her male and the rival,
which could prove harmful to her or her offspring
(Brotherton and Manser 1997). HB HA

F1

POLYGYNY
F2
Recall that polygyny occurs when one male mates with
more than one female during a breeding season. Natural
Female fitness

selection’s ledger sheet shows that polygyny has both

costs and benefits for males and females. Because a male
can generally fertilize more eggs than a female can pro-
duce, a male usually benefits from polygyny by produc-
ing more offspring (if paternal care is not required). As
males maximize their reproductive output through mul- Polygyny
tiple matings, polygyny results. Possible costs to a male threshold
include an increased chance of cuckoldry, because he is
not guarding each female from other males, and the costs
associated with achieving dominance or defending Environmental quality
resources or territories.
FIGURE 15.13 The polygyny threshold model. The
Polygyny has several significant costs for females.
polygyny threshold is the difference in quality between
For example, males in such cases usually do not help HA and HB that would favor polygynous matings.
rear the young. In species in which the males do pro- F1 ⫽ fitness curve for monogamous females
vide some parental care, it is divided among the off- F2 ⫽ fitness curve for secondary females
spring from more than one female or sometimes HA ⫽ highest quality breeding habitat
provided only to the first female, the so-called primary HB ⫽ marginal breeding habitat
female. Females must also share essential resources, (From Wittenberger 1979; modified from Orians 1969.)
350 Chapter 15 / Parental Care and Mating Systems

the circumstances under which it does occur must be of new females on the territories. Sixteen of the 19 ter-
those in which there is some advantage to the female. He ritory pairs were settled by new females. Preference in
reasoned that females should join a harem when this 2 of the 16 settled pairs could not be ascertained because
decision confers greater reproductive success than the territories of the mated and unmated males were set-
monogamous alternatives. According to this argument, tled simultaneously (i.e., an observer arrived to find both
then, the average reproductive success of females should territories occupied). However, in 12 of the remaining
not decrease as the harem’s size increases. Figure 15.13 14 territory pairs, a female settled first on the territory
illustrates this model’s way of relating differences in a of the mated male. Thus, in support of the polygyny
territory’s quality to a female’s choice of already mated threshold model, superior territory quality was shown to
versus unmated males. reverse the basic preference for pairing with an unmated
Experimental field evidence in support of the polyg- male in this population of red-winged blackbirds.
yny threshold model has been obtained for a population
of red-winged blackbirds in Ontario, Canada. First, one “Sexy Son” Hypothesis According to this hypothesis,
study established that all things being equal, females in access to good genes for her offspring compensates a
this population prefer unmated males to already mated female for the costs of polygyny. A female may benefit
males when settling to breed in early spring (Pribil and from mating with an already mated male if her sons
Picman 1996). In this study, the researchers established inherit the genes that made that male attractive. Her sexy
16 pairs of adjacent territories; the territories within pairs sons will presumably provide her with many grandchil-
were similar in size, water depth (a critical determinant dren, so the female’s lifetime reproductive success may
of territory quality; see below), and harem size the pre- be enhanced by choosing to mate with a male that is
vious breeding season. Within each pair, one territory attractive to many females (Weatherhead and Roberston
was defended by a monogamously mated male and the 1979, 1981). According to this hypothesis, then, a female
other by an unmated (bachelor) male. The researchers that chooses an already mated male may benefit indirectly
maintained these distinctions by removing females if the good genes she acquires for her offspring boost
when necessary (e.g., females that had already settled in their survival and reproductive success.
the bachelor territory before the start of the experiment Thomas Huk and Wolfgang Winkel (2006) exam-
were removed, as were any additional females that set- ined their 31 years of data on a German population of
tled on the territory of the monogamously mated male). pied flycatchers (Ficedula hypoleuca) to see if there was evi-
They then recorded where newly arriving females set- dence supporting the “sexy son” hypothesis. Their analy-
tled. Within about two weeks all territory pairs had been ses showed that direct reproductive success (number of
settled by new females. Female red-winged blackbirds fledglings) was lower in females mated to polygynous
were unanimous in their choice of where to settle: in all males (here, males with two mates) than in females
16 pairs of territories, the first new female settled on the mated to monogamous males. This was especially true
bachelor territory. Thus, females in this population pre- for secondary females without male assistance.
fer unmated males to mated males. A second study, using Polygynous males of this species give priority to the
a similar experimental design, revealed that this basic young of primary females (females whose young hatch
preference could be reversed if the territories of mated first) and show either reduced or no care to young of sec-
males were made superior to those of unmated males. ondary females (females whose young hatch later than
For female red-winged blackbirds, reproductive success those of the primary female).
is most closely correlated with water depth, with nests However, maybe it is not so bad being mated to a
over deep water being more successful (presumably polygynous male: perhaps the sons of polygynous males
because such nests are harder for ground predators to are particularly attractive and produce many offspring of
detect and reach). Stanislaw Pribil and William Searcy their own. In other words, might female pied flycatch-
(2001) manipulated pairs of adjacent territories. Within ers mated to polygynous males experience delayed com-
a pair of territories, one was designated to be that of a pensation, if the sons they produced were especially
monogamously mated male (one female present) and the attractive and produced many offspring? In order to test
other to be that of an unmated male (no female present). whether the “sexy son” hypothesis applied to their study
As before, the pairing status of each male was maintained population, the researchers compared the reproductive
by trapping and removing any females that arrived on success of males hatched in monogamous broods with
the territory before the start of the experiment. The that of males hatched from polygynous broods; polygy-
researchers placed nesting platforms in open water in the nous broods were further classified as primary or sec-
territories of mated males and on dry land in the terri- ondary. Males hatched in primary polygynous broods
tories of unmated males; in this way, the territories of produced more fledglings over the course of their life-
mated males were made superior and those of unmated times than did males from monogamous broods.
males inferior. Each of the 19 territory pairs was then However, males hatched in secondary polygynous
visited twice daily by the researchers to monitor presence broods produced no more fledglings than those hatched
 Mating Systems 351

Number of fledglings (mean ± SE) 18

during a male’s lifetime
16
14
12
10
8
6
4
Primary Secondary Monogamous
broods broods broods

FIGURE 15.14 Data from long-term monitoring of a

population of pied flycatchers reveal that sons of
secondary females do not differ from sons of monoga-
mous females in number of fledglings produced over a
lifetime. These data do not support the sexy son
hypothesis. (From Huk and Winkel 2006.)
FIGURE 15.15 A male elephant seal. Males of this
species compete for access to, and defend, large num-
in monogamous broods (Figure 15.14). These results bers of females at birthing sites. This is an example of
indicate that good genes from fathers can increase the female defense polygyny.
fitness of their sons, but only if fathers do not reduce
paternal care (such as in the broods of primary females).
The data for males hatched in secondary broods indicate (LeBoeuf 1974). This male defends his harem against
that females do not benefit by pairing with an already all other male intruders in bloody, and sometimes
mated male, even if he is sexy and his sexiness is herita- lethal, fighting.
ble. For secondary females, then, choosing an already
mated male with good genes does not compensate for Resource Defense Polygyny Resource defense polyg-
the negative impact of his reduced paternal care. These yny occurs when males defend resources essential to
data for secondary females do not support the “sexy son” female reproduction (e.g., nest sites or food) rather than
hypothesis. defending females themselves. In other words, even
though receptive females do not live together, a male can
monopolize a number of mates by controlling critical
Types of Polygyny resources. Conditions typical of resource defense polyg-
There are three types of polygyny—female defense, yny include the following: (1) quality of the monopolized
resource defense, and lek—distinguished on the basis of resource reflects male quality (i.e., higher quality
what males are defending. We describe each in turn. resources should be defended by higher quality males);
(2) females prefer males with resources to those without
Female Defense Polygyny In female defense polyg- resources; and (3) males with resources have higher mat-
yny, a male defends a harem of females. This type of ing and reproductive success than those without
polygyny occurs when females live in groups that a resources (Emlen and Oring 1977; Thornhill 1981).
male can easily defend. In some species, female gre- In some cases, there is good evidence that a female’s
gariousness may be related to cooperative hunting or choice is based on the quality of resources controlled by
increased predator detection; in other species, clump- a male. An example is provided by several American
ing is more directly related to reproduction. For exam- species of scorpionflies (Mecoptera: Panorpidae), in
ple, female elephant seals (Mirounga angustirostris) which males fiercely defend the area around a dead
become sexually receptive less than one month after arthropod (Thornhill 1981). Standing next to his nutri-
giving birth. Each year pregnant females haul them- tious find, a male will disperse a sex attractant and dis-
selves onto remote beaches of Año Nuevo Island, play with rapid wing movements and abdominal
California. Female gregariousness, a shortage of suit- vibrations (Figure 15.16). If a female approaches, copu-
able birth sites, and a tendency to return annually to lation is the fee that she must pay to gain access to this
traditional locations result in the formation of dense food. Thus, while males enjoy the benefits of mating
aggregations of receptive females. Under these with several females, females enjoy a hearty meal that
crowded conditions, a single dominant male, weighing may enhance their reproductive effort. In these
in the vicinity of 8000 pounds and sporting an enor- American species, larger males obtain larger arthropods,
mous overhanging proboscis (Figure 15.15), can while small males that are unable to capture large
monopolize sexual access to 40 or more females prey resort to less successful tactics such as stealing
352 Chapter 15 / Parental Care and Mating Systems

chicken, where they defend small (radius of 2 cm),

mobile territories by bumping and jostling one another
with their abdomens. Females visit these nocturnal leks,
evaluate several males, but copulate with just one. There
is considerable agreement among females in their choice
of mate, and as a result, male mating success is highly
skewed, with some males copulating with many females
and others being unsuccessful. Sandflies do not remain
on leks during daylight hours but may return to leks over
consecutive nights.
How might lek behavior such as that seen in sand-
flies have evolved? Several hypotheses have been sug-
gested. These have been divided into two groups, based
on potential benefits for males or females. From the male
FIGURE 15.16 A male scorpionfly. American scorpion- perspective, group, rather than individual, displays may
flies display resource defense polygyny in that a male increase the signal range or the amount of time that sig-
will defend the area around a dead insect and females
nals are emitted (e.g., Lack 1939; Snow 1963). Or males
must mate with him to gain access to the meal.
may aggregate because they require specific display habi-
tats that are limited and patchily distributed (e.g., Snow
1974). Leks may provide protection from predators
copulations or presenting salivary secretions (another through increased vigilance—with more eyes watching,
option for a nuptial gift). a predator would have a harder time sneaking up on
Based on these findings for several American them (e.g., Wiley 1974). Leks may also serve as infor-
species of scorpionfly, we might expect resource defense mation centers, where males exchange the latest news on
polygyny to characterize all species of Panorpa, but this good foraging sites (e.g., Vos 1979). Males may gather
is not the case. In experiments designed to be similar, if near “hot spots,” areas in which females are most likely
not identical, to those performed with American species, to be encountered (e.g., Bradbury et al. 1986; Ryder et
Merle Missoweit and Klaus Sauer (2007) obtained the al. 2006). Finally, leks may arise because less successful
following results for two European species: (1) males males generally have better mating chances in the vicin-
and females were equally adept at monopolizing food ity of highly successful males. Because certain males are
items, (2) males did not preferentially occupy large food extremely successful at attracting females, other, less suc-
items, (3) there was no relationship between male qual- cessful males gather around these “hotshots” and obtain
ity and resource quality, and (4) males that provided sali- more copulations than they would have had they dis-
vary secretions obtained more matings than those that played by themselves (Beehler and Foster 1988).
provided dead arthropods. Thus, we see that even From the female perspective, large groups of males
closely related species may differ dramatically in pat- may facilitate mate choice (Alexander 1975; Bradbury
terns of mating. 1981). After all, it might be easier to distinguish
between superior and inferior males when comparison
Lek Polygyny The third category of polygyny, lek
polygyny, occurs when males defend “symbolic” territo-
ries that are often located at traditional display sites
called leks (Figure 15.17). Males of lek species do not
provide parental care and defend only their small terri-
tory on the lek, not groups of females that happen to be
living together nor resources associated with specific
areas. Females visit these display arenas, select a mate,
copulate, and leave. This extreme form of polygyny
occurs when environmental factors make it difficult for
males to monopolize females either directly, as in female
defense polygyny, or indirectly, as in resource defense
polygyny.
In Chapter 14 we introduced you to sandflies of the
Lutzomyia longipalpis species complex, and we return to
them for an example of lekking behavior (reviewed in
Jones and Quinnell 2002). Recall that at night, male FIGURE 15.17 Two black grouse males displaying on
sandflies gather on the back of a vertebrate host, often a a lek.
 Mating Systems 353

shopping is possible. Mating within a group of males genetic mechanisms such as increased genetic diversity
may reduce the vulnerability of females to predation among progeny and reduced risk of fertilization by
since any predator might be distracted by so many dis- genetically incompatible sperm (Simmons 2005; Yasui
playing individuals (Wittenberger 1978); or if males 2001; Zeh and Zeh 2001). Let’s take a closer look at pro-
aggregate in a less desirable habitat, lek mating may duction of diverse offspring as a possible explanation for
reduce competition between the sexes for resources the evolution of polyandry.
(Wrangham 1980). Finally, females living in large, Honeybees (genus Apis) are social insects that
unstable, mixed-sex social groups may visit leks to avoid exhibit polyandry: the reproductive female of a colony
harassment from males in their group (Clutton-Brock et (known as the queen) mates with multiple males.
al. 1992). All of these hypotheses could be useful as Reproductive males are called drones, and infertile
explanations for the development of leks but for now are females, known as workers, labor on behalf of the queen.
best regarded as only possible explanations. Workers are closely related to the queen, so their help-
ful (altruistic) behavior can be understood from the
standpoint of kin selection (see Chapter 19 for further
POLYANDRY
discussion of social insects). This brief introduction to
Polyandry occurs when a female has more than one honeybee society raises an interesting question: if high
mate during the breeding season. Recall from Chapter levels of relatedness are the basis for altruistic behavior
14 that Bateman’s principle (1948) suggested that male, within honeybee colonies, then why would a queen mate
but not female, reproductive success exhibits substantial with multiple males, a behavior that generates genetic
increases with number of mates. As the thinking went, diversity? Are there significant benefits associated with
because most males offer only sperm in return for cop- a genetically diverse workforce? Heather Mattila and
ulation and because the sperm from one male is often Thomas Seeley (2007) set out to answer this question
sufficient to fertilize all the eggs of a given female, then by establishing founding colonies of honeybees that
females would have little to gain by mating with more were either genetically diverse (queens artificially
than one male, especially given the costs associated with inseminated, via an instrument, with sperm from 15
mating (exposure to predators, disease, and aggressive drones) or genetically uniform (queens inseminated
males). We now know, however, that mating with mul-
tiple males is widespread across taxonomic groups
(Jennions and Petrie 2000; Rivas and Burghardt 2005; a Genetically diverse
Zeh and Zeh 2001). We also have evidence from an 12000
Genetically uniform
Mean area of comb (cm2)

increasing number of species that female reproductive

success, like male reproductive success, can significantly 9000
increase with number of mates. For example, if copula-
tion opens the door to critical resources or male
parental assistance, then mating with several males may 6000
result in reproductive benefits for females. In insects,
polyandry has been shown to increase the number of 3000
eggs laid and the hatching success of those eggs (studies
reviewed by Arnqvist and Nilsson 2000). Explanations
for clutch size effects include increased nutrients passed 0
13 June
16 June
20 June
23 June
26 June
30 June
4 July
7 July
10 July
17 July
25 July
2 Aug.
7 Aug.
14 Aug.
21 Aug.
30 Aug.
to females of species in which males provide nuptial gifts
and increased receipt by females of hormonal stimulants
in male ejaculates. Possible explanations for hatching
success effects include avoidance of sperm depletion and b 12
(bees, comb, brood, food) (kg)

Genetically diverse
Mean weight per colony

Genetically uniform
FIGURE 15.18 Female honeybees mate with multiple 9
males, and this results in genetic diversity within
colonies. The benefits of such diversity have been 6
demonstrated by experimentally creating founding
colonies that were either genetically diverse or geneti-
cally uniform and by monitoring their progress. 3
(a) Diverse colonies built more comb than did uniform
colonies. (b) Diverse colonies also weighed more and 0
13 Aug.
2 Sept.
22 Sept.
12 Oct.
1 Nov.
21 Nov.
11 Dec.
31 Dec.
20 Jan.
9 Feb.
1 Mar.
21 Mar.
10 Apr.
30 Apr.
11 June
4 July
24 July

20 May

survived longer than uniform colonies. Each arrow

represents death of a colony. (From Mattila and
Seeley 2007.)
354 Chapter 15 / Parental Care and Mating Systems

with the same volume of sperm but from one drone). sexes). Parental behavior is not displayed by all species.
They then monitored several features of colony devel- For example, avian brood parasites dispense with all
opment during the very challenging time of colony parental responsibilities by laying their eggs in the nests
founding. Genetically diverse colonies were more effi- of other species.
cient than genetically uniform colonies at building a Conflicts of interest are not restricted to parent–
comb (the latticework of cells that form the internal young interactions, and in fact such conflicts charac-
structure of the nest) (Figure 15.18a). Genetically terize most social behavior. In the case of mating
diverse colonies also weighed more than genetically uni- relationships, males usually produce more offspring by
form colonies, and whereas all uniform colonies died by seeking additional mates. In contrast, females tend to
mid-December, several genetically diverse colonies emphasize parental effort rather than mating effort and
survived the winter (Figure 15.18b). When compared can usually produce more offspring by gaining male
with genetically uniform colonies, genetically diverse parental investment. The disparity between the sexes in
colonies also foraged at higher rates, produced more parental investment interacts with ecological factors
workers, and reared more drones. Here, we see that the such as predation, resource quality and distribution,
genetic diversity generated by polyandry results in many and the availability of receptive mates to shape the
benefits for honeybees. mating system of a species.
Mating systems can be defined on the basis of the
number of individuals that a male or female copulates
SUMMARY with during a breeding season. According to this system,
monogamy occurs when a male and female have only a
Parental care is one component of the overall life his- single mating partner per breeding season; polygyny
tory of a species. Because animals have limited time, when some males copulate with more than one female
energy, and resources to devote to reproduction, evo- during the breeding season; and polyandry when some
lutionary “decisions” must be made about the amount females mate with more than one male during the breed-
of care and who should assume parental responsibili- ing season. Polygynandry occurs when both males and
ties. Evolutionary conflicts over parental investment females mate with multiple partners.
include sexual conflict, intrabrood conflict, and inter- Extra-pair matings are common regardless of the
brood conflict. The issue of who provides the care mating system. These have different costs and benefits
appears to be related to factors such as mode of fertil- for males and females and will affect the evolution of
ization (male care being associated with external fertil- mating systems. Potential benefits to males include an
ization and female care with internal fertilization) and increased number of offspring. Females gain material
phylogenetic history (in mammals parental care usually benefits, such as food or help from the extra-pair male
rests with the female, whereas in birds parental duties in raising offspring, or genetic benefits, such as fertility
are typically partitioned somewhat equally between the insurance or high-quality genes.
 16
Communication: Channels and Functions

The Definition of Communication Sometimes important messages are whispered rather

Channels for Communication than shouted. This seems to be the case for Asian corn
borer moths (Ostrinia furnacalis). Males of this species,
Vision when very close to a female (< 2.75 cm away), produce
Audition extremely low-intensity ultrasonic courtship songs
Substrate Vibrations (sounds with frequencies above the range of human hear-
Chemical Senses ing are described as ultrasonic; see section on audition).
Touch The songs are produced during wing strokes when males
rub specialized scales on their forewings against those on
Electrical Fields
their thorax (Figure 16.1). The scales are not found on
Multimodal Communication females. These low-intensity courtship songs suppress
Functions of Communication escape behavior of the female, thus facilitating mating.
Species Recognition
The private nature of the courtship songs may reduce
the risks of eavesdropping by rival males and predators
Mate Attraction
(Nakano et al. 2008).
Courtship and Mating This brief foray into the confidential sexual com-
Maintaining Social Bonds munications of moths raises some general questions
Alarm about animal communication. How do we define com-
Aggregation munication? What sensory channels do animals use
when communicating? How are signals shaped by com-
Agonistic Encounters
petitors, predators, and the environment? And what
Communication about Resources: A Case Study messages are conveyed?

355
356 Chapter 16 / Communication: Channels and Functions

FIGURE 16.1 Male Asian corn borer moths produce low-

[

a
intensity ultrasonic courtship songs when close to females.
The songs are produced by rubbing specialized scales on
the forewings against those on the thorax. (a) In this dorsal
view, the scales are indicated by arrow heads within the box
on the right side. The scales are normally hidden from view
by a small leaf-like structure that overlaps the base of
b 6

Distance between wings (mm)

Relative amplitude of sound

the forewing. (b) The production of ultrasonic song is
1 2 3 4 5 6
synchronized with wing strokes. Shown in light orange are 5
six pulses of sound recorded with a special microphone
4
during courtship. The black line traces the distances
between distal edges of the right and left forewings; 3
distances decrease during the upstroke and increase during
2
the downstroke. These data show that the pattern in which
pulses are generated corresponds with the up and down 1
movements of the forewings (pulses 1, 3, and 5 coincide
0
with upstrokes and pulses 2, 4, and 6 correspond with 0 10 20 30 40 50 60
downstrokes). (Modified from Nakano et al. 2008.) Time (msec)

ignored signals, and eventually senders would stop pro-

THE DEFINITION ducing them. Therefore, some authors include in their
OF COMMUNICATION definition of communication the idea that receivers also
benefit from the transmission of information. The prob-
Defining communication is surprisingly difficult. One lem that arises with this definition is that receivers do not
broad definition holds that communication occurs when always benefit from receiving a signal. Sometimes
a sender produces a signal that contains information senders manipulate receivers by sending dishonest sig-
(Batteau 1968). This signal is transmitted through the nals, as we will explore in detail in the next chapter.
environment and detected by a receiver, who then inter- For our purposes, we’ll use the definition that
prets the signal and decides how to respond. communication occurs when information is transferred
This definition covers most of what we would nor- from the sender to the receiver, and that it benefits the
mally consider communication—the leaping courtship sender, on average. Keep in mind, however, that this
dance of a male crane, the growl and bared teeth of a definition is not ideal; as Bradbury and Vehrencamp
wolf—but it also encompasses interactions that we might (1998) point out, “Evolution hates definitions.”
be more reluctant to include. For instance, the sound of
a mouse rustling in a runway under dried leaves on its
way to its food stores may be heard by an owl. The owl
makes use of the information about the location of the CHANNELS FOR
mouse by swooping down to grab the sender. Should we COMMUNICATION
call this communication? By the above definition it is,
but this strikes many behaviorists as problematic. Communication can involve any of a variety of sensory
Clearly, the function of the noise, from the mouse’s channels—vision, audition, chemical, touch, and elec-
perspective, was not to alert the owl. trical fields. Each channel has both advantages and lim-
Many researchers add another element to the defi- itations (Table 16.1). As we will see, the channel used for
nition: the sender benefits from the transmission by a particular signal will depend on the biology and habi-
altering the behavior of the receiver. This definition tat of the species, as well as the function of the signal.
excludes examples such as the rustling mouse; instead, we
would call the rustling a cue. In contrast, a courtship
VISION
dance, a song, a brightly colored feather crest, or some
other feature that does fit this more precise definition is Visual signals have two obvious properties. The first is
called a signal. A stereotyped sequence of behaviors that ease of localization. If the signal can be seen, the loca-
has a signaling function is called a display. tion of the sender is known. For example, when a male
Must the receiver also benefit from the information bird is engaging in a visual display to attract a mate, there
being transferred? Logically, on average, receivers must is never any doubt about his precise location. The
benefit from paying attention to a particular signal. If receiver can see him and, therefore, respond to him in
they did not, selection would favor receivers that terms of his exact location, as well as his general pres-
 Channels for Communication 357

TABLE 16.1 Characteristics of Different Sensory Channels for Communication

Type of signal

Feature Visual Auditory Chemical Tactile Electrical

Effective distance Medium Long Long Short Short

Localization of sender High Medium Variable High High
Ability to go around obstacles Poor Good Good Good Good
Speed of transmission Fast Fast Slow Fast Fast
Complexity High High Low Medium Low
Durability Variable Low High Low Low

ence. The second property is rapid transmission and a

rapid fade-out time. The message is sent literally at the
speed of light, and as soon as the sender stops display-
ing the signal is gone. For instance, if the displaying bird
suddenly spots a hawk and takes cover in a nearby hedge,
its original position will not be revealed by any linger-
ing images.
Visual systems can provide a rich variety of signals. b
This diversity is possible because of the number of stim-
ulus variables that can be perceived by most animals.
These include brightness and color, as well as spatial and
temporal pattern that can be altered by the animal’s
movements and posturing. The presence and relative
importance of one type of visual stimulus or another can
be influenced by environmental conditions. For exam-
ple, during agonistic displays aimed at conspecifics (or FIGURE 16.2 In many sharks, agonistic displays involve
divers who invade their personal space), many shark symmetrical depression of the pectoral fins. (a) A
species strongly depress both pectoral fins and hold them Galapagos shark using this visual signal of agitation.
for prolonged periods in this symmetrical downward (b) Species that inhabit clear water habitats, such as the
position (Martin 2007). Many species of sharks inhabit- blacktip reef shark, often have markings on their
ing clear water environments, such as near shore and reef pectoral fins that may enhance the agonistic display.
habitats, have conspicuous markings on their pectoral (Drawn from photographs in Martin 2007.)
fins, including black or white tips and margins that
enhance the visibility of this postural display. In contrast,
sharks living in habitats where light is scarcer rely sim- territorial male’s presence to the interloper, who then
ply on the posture (Figure 16.2). retreats. (Females, by the way, move freely through male
Visual signals have disadvantages, of course. The territories, without territorial males even batting an eye.)
most obvious is, quite simply, that if the sender cannot Spring runoff is a time when sediment increases dra-
be seen, then its signals are useless, and vision is easily matically in the Colorado River, making the transmission
blocked by all sorts of obstructions, from vegetation to of visual signals very difficult. It is probably no mistake,
fog to sediment in water. Sometimes, however, the visual then, that the spawning season of razorbacks is timed to
displays of animals can be timed to avoid these obstruc- occur before spring runoff, when eye flashes are still vis-
tions. Consider the case of the razorback sucker ible in the less turbid water. Thus, we see that two
(Xyrauchen texanus), an endangered species of fish found aspects of razorback life history—their move from deep
in the Colorado River in the western United States. As to shallow water for spawning where light from the
the spawning season approaches, razorback suckers move water’s surface can penetrate to the river bottom, reflect
from deep water to more shallow water where the males off eyes, and be seen by conspecifics and the timing of
rest within territories along the river bottom. Should a spawning before spring runoff—make possible the use of
roving male approach a territorial male, the territorial visual signals in territorial communication (Flamarique
male quickly rolls his eyes downward, exposing the et al. 2007).
whites of his eyes to light coming from the water sur- At night or in dark places, light-producing species
face. These two quick flashes of reflected light signal the use visual signals. Most animals, however, cannot produce
358 Chapter 16 / Communication: Channels and Functions

light, which leads us to our next question: do nocturnal presence of a receiver, a courting male turns and faces
species that cannot produce light ever use visual signals? the approaching female and directs his display toward
The answer is yes, but as with the razorback sucker, tim- her. Males, it turns out, do more than simply face the
ing is everything. It turns out that many nocturnal species female at show time. Indeed, males adjust the tempo-
are most active at dawn and dusk, when some light is ral and structural elements of their claw-waving display
available, so visual displays are still an option at these in relation to female distance (How et al. 2008). As dis-
particular times of day. Nevertheless, colors are difficult tance to the female decreases, the interval between claw
to distinguish under conditions of low light, so we might waves decreases, as does the duration of each wave
expect visual signals at dawn and dusk to focus on con- (Figure 16.4b and 16.4c, respectively). These changes in
trast, rather than color, and to involve white. And that is timing translate into a more intense display as the
exactly what we find for eagle owls (Bubo bubo). Males and female nears, and may reflect a male’s increasing will-
females of this nocturnal species have a white badge of ingness to invest energy in displaying as his chances of
feathers on the throat (Figure 16.3). This badge is par- a successful mating increase. Changes in timing could
ticularly visible when the throat is alternately inflated and also reflect a change in signal function. For example,
deflated during vocal displays, which peak at dawn and rather than serving as a conspicuous beacon (as it does
dusk (Delgado and Penteriani 2007). Breeding eagle owls for broadcast displays), the claw-waving display may
also appear to use white material to mark the area around become a signal of male quality. The structure of the
their nest; they deposit white feces at obvious defecation claw-waving display also changes as distance to the
posts and display the white feathers of prey species at con- female decreases. For example, the horizontal distance
spicuous plucking sites (Penteriani and Delgado 2008). swept by the tip of the enlarged claw decreases with
So we see that nocturnal species can use visual signals, decreasing distance to the female (Figure 16.4d, left
providing that such displays focus on contrast rather than side). This last result indicates that when claw-waving
color, and occur when some light is available. displays are used as long-range signals, the horizontal
Finally, the apparent size of visual signals (their distance swept by the tip of the enlarged claw is at its
conspicuousness) diminishes with distance. Given this greatest and displays are at their most conspicuous (see
limitation, we might expect animals to adjust their Figure 16.4d, right side). Thus, these long-range signals
visual signals with respect to receiver distance. Courting may serve as beacons to call in distant females. All of
male fiddler crabs are well known for their conspicu- these adjustments in the timing and structure of the
ous and energetically expensive claw-waving displays claw-waving display are quite remarkable because they
(see Chapter 14). During such displays, a male repeat- require these small crustaceans, living on mudflats, to
edly waves his single, greatly enlarged claw. We will accurately measure the distance between themselves
focus on Uca perplexa, a species of fiddler crab found in and an approaching female.
Australia, whose display is shown in Figure 16.4a. In the
absence of a receiver (a female conspecific), males of
this species broadcast their courtship displays. In the
AUDITION
Sound signals have several advantages. They can be trans-
mitted over long distances, especially in water. Although
sound is transmitted more slowly than light, it still can
be a rapid means of sending a message, particularly at
close range. The transitory nature of sound makes
possible a rapid exchange and immediate modification,
but it does not permit the signal to linger. After the mes-
sage has been sent, the signal disappears without a trace.
Sound signals have an additional advantage of being able
to convey a message when there is limited visibility, such
as at night or in deep water or dense vegetation. Finally,
sound signals can be complex. Every music lover is aware
of the tremendous variety of sound that can be produced
by the temporal variation of just two of its parameters:
frequency (pitch) and amplitude (loudness). Depending
on the species, animals may vary either or both of these
FIGURE 16.3 Although eagle owls are nocturnal, they aspects of sound, or they may, like a drummer, simply
use visual signals at dawn and dusk when some light is alter the pattern of presentation.
available.The white throat patch used in these displays The types of sounds used by a particular species in
contrasts sharply with the surrounding darker plumage communication are determined by how the animal pro-
and is exposed during calls. duces them. Sounds may be generated by respiratory
 Channels for Communication 359

Start position Stage 1 Stage 2 Stage 3

b broadcast c
5 signal
1.5
4

Time (sec)
Time (sec)

3
1.0
2

1
0.5

0
0 20 40 60 0 20 40 60
Distance between Distance between
sender and receiver (cm) sender and receiver (cm)

d
2 cm

0–7.5 7.5–10.4 10.4–16.3 16.3–26.2 26.2–69.2

Distance between sender and receiver (cm)

FIGURE 16.4 Male fiddler crabs adjust the timing and structure of their claw-waving displays in relation to receiver
distance. (a) Stages in the claw-waving display of male Uca perplexa. (b) The interval between claw waves, (c) the
duration of claw waves, and (d) the horizontal sweep of the claw tip decrease with decreasing receiver distance. Thus,
as distance to receiver decreases, male displays increase in intensity but become less conspicuous. Being conspicuous
is more important when receivers are far away. (Modified from How et al. 2008.)

structures, striking objects in the environment, or rub- called a scraper, that rubs against a row of ridges, the
bing appendages together. Many structures specialized file, on the underside of the other wing cover.
for sound production have evolved in association with Whenever the scraper is moved across the file, a pulse
respiratory structures. For example, mammals use their of sound is produced. This method of generating sounds
larynx and birds use their syrinx to produce sound. The is called stridulation.
anatomical structure and location of these organs are Stridulation is not confined to insects; it is also used
different, but both allow production of complex sounds. by males of the club-winged manakin (Machaeropterus
The environment may also be used to produce auditory deliciosus), a small bird of the Neotropics (Bostwick and
signals. Humans often tap their toes when listening to Prum 2005). During courtship displays, a perching male
music, but for some other animals, such as rabbits and creates a tick-tick-ting sound, which is produced by spe-
deer, foot stamping itself is the signal. This musical cific movements of the wings and highly modified sec-
theme has several variations—beavers slap the water ondary feathers. (Secondary feathers arise from the ulna,
with their tails and woodpeckers drum on trees with a bone in the forearm, whereas primary feathers arise
their bills. Insects most commonly produce auditory sig- from bones of the hand). The “tick” sound is created
nals by rubbing together parts of their exoskeleton. when the male rotates his wings forward, rapidly flips the
Crickets, for example, produce sound by opening and feathers above his back, and then forcefully brings the
closing their wings. Each wing has a thickened edge, wings together, causing the secondary feathers to collide.
 360 Chapter 16 / Communication: Channels and Functions

The “ting” sound is produced in much the same man- Some animals make sounds that humans cannot
ner except that after the secondary feathers collide, the hear. For example, several groups of mammals, includ-
male shivers his wings, causing oscillations of the sec- ing cetaceans, bats, and rodents, produce and detect
ondary feathers. The end result of these movements ultrasounds as part of echolocation (see Chapter 10) or
is that one highly modified feather (the fifth secondary communication systems. Ultrasounds are sounds whose
feather, known as the “pick”) rubs back and forth frequencies are above those audible to humans, which
against the ribbed surface of the adjacent feather (the means frequencies greater than about 20 kHz. Ultrasonic
sixth secondary feather, known as the “file”), exciting communication, however, is not restricted to mammals,
resonance in the enlarged rachis (central shaft of the and it recently was described in an amphibian. Consider
feather to which the vanes attach), and putting the the fascinating case of the concave-eared torrent frog
finishing touches on this unique auditory signal (Odorrana tormota), a nocturnal species that lives along
(Figure 16.5). noisy streams and waterfalls in Huangshan Hot Springs,
central China. Its common name refers to the recessed
a ears of males. Unlike other frogs who typically lack ear
canals and whose eardrums are located at the skin sur-
face (and hence are visible), males in this species have ear
canals and recessed eardrums, similar to mammals. Field
experiments reveal that audible and ultrasonic compo-
nents of male calls evoke vocal responses by male con-
specifics (Feng et al. 2006). This ultrasonic hearing
capacity was confirmed in the laboratory by electro-
physiological recordings from an auditory processing
center of the midbrain when frogs were exposed to
bursts of ultrasound. More recent work showed that
females, just before ovulation, also produce calls with
ultrasonic components (Shen et al. 2008). When female
calls were played in the field and in the laboratory, they
evoked calls from nearby males, who then rapidly and
precisely approached the loudspeaker, a behavior called
positive phonotaxis. Researchers working with concave-
eared torrent frogs suggest that ultrasonic communica-
tion allows this unusual species to avoid the masking
b Sixth secondary (the file) effects of the constant low-frequency background noise
Fifth secondary of the local streams and waterfalls in their habitat.
(the pick) We mentioned that humans cannot hear sounds
whose frequencies are above 20 kHz (20,000 Hz) and that
these sounds are called ultrasounds. Human hearing also
has a lower limit: we typically cannot hear sounds whose
frequencies are less than about 20 Hz, so sounds below
this limit are called infrasound. When we think of ele-
phants, we often think of the audible call known as the
trumpet blast. In reality, however, most elephant calls are
infrasonic, with fundamental frequencies of 15 to 25 Hz
(Payne et al. 1986; Poole et al. 1988; note that some of
these calls cross into the range of human hearing, and
indeed humans can hear some of these low-frequency
calls). The source of infrasonic calls is the same as that of
audible calls: elephants drive air from the lungs to set the
FIGURE 16.5 Male club-winged manakins produce non- vocal folds of the larynx in motion (Garstang 2004).
vocal courtship sounds, described as a tick-tick-ting Elephants are very social animals that live in matri-
sound. (a) The tick sound is created by hitting the
lineal family groups in which daughters remain with
secondary feathers of the wing together, above the back.
(b) The ting sound is generated by highly modified
mothers and sons leave to live in bachelor groups or as
secondary feathers of the wing. Following the collision lone bulls. Family groups range over large areas in search
of the feathers, the male shivers his wings, causing the of food and water, so long-distance communication
fifth secondary feather (the pick) to scrape across ridges between family members and between different family
on the sixth secondary feather (the file). groups is critical, as is communication between sexually
 Channels for Communication 361

receptive females and males. (Estrus in elephants occurs

every four years and lasts only five days, so males must be
able to reliably, and quickly, find females over enormous
distances.) When compared with high-frequency sounds,
low-frequency sounds are much less degraded through
refraction (the bending of sound waves as they pass from
one medium to another of different density), reflection
(the bouncing off of a new medium), and absorption (the
conversion of sound energy to heat as it passes through a
medium, resulting in decreased intensity). Therefore
infrasound works especially well for long-distance
communication (Garstang 2004; Langbauer 2000).
FIGURE 16.6 Water striders communicate by tapping
out messages on the surface of the water, thereby creat-
STOP AND THINK ing surface waves. The pattern of the wave determines
The ultrasonic calls of laboratory rats have been studied the message. Here, a male taps out a repel signal.
in detail for many years, and three distinctly different
ultrasonic vocalizations have been identified: (1) a 40- to serve several functions, including sex identification, mate
50-kHz call given by pups when separated from their attraction, courtship, and territorial defense (Jablonski
mother, which can induce retrieval of the pup by the
and Wilcox 1996).
mother; (2) a 22-kHz call given by adolescent and adult
Seismic signals also can be generated when an air-
rats during aversive situations (e.g., exposure to predators,
handling, social isolation, or electric shock), which may borne vocalization couples with the ground. For example,
serve to deter predators or signal alarm; and (3) a 50-kHz the infrasonic calls of elephants travel not only through the
call given by adolescent and adult rats during encounters air as auditory signals, but also through the ground as seis-
that are playful or sexual, which can stimulate social mic signals (O’Connell-Rodwell 2007). Elephants detect
approach (Brudzynski 2005; Portfors 2007; Wøhr and and respond to seismic signals experimentally generated
Schwarting 2007, 2008). The first two calls reflect a neg- by researchers directing previously recorded vocalizations
ative affective (emotional) state, while the 50-kHz call of into the ground. Elephants also can distinguish subtle dif-
adolescents and adults indicates a positive state. Because ferences between seismic calls. For example, they can dis-
the ultrasonic calls of rats reflect their affective state, some criminate between the seismic alarm calls of familiar and
scientists suggest that these calls be routinely monitored
unfamiliar conspecifics (O’Connell-Rodwell et al. 2007).
(using ultrasonic microphones) by laboratory personnel to
How do elephants detect seismic signals? Two pathways
facilitate best care practices for rats housed in animal facil-
ities. Do you think it would be worthwhile to use audi- have been suggested. The first is bone conduction. In this
tory signals to assess the well-being of laboratory rodents? pathway, seismic signals travel from the ground through
Why or why not? If such a system were in place, how the feet, up the front legs to the shoulders, and on to the
might you use the information from ultrasonic calls to middle ear. The second potential pathway involves
organize laboratory practices? For example, what proce- mechanoreceptors, such as Pacinian corpuscles, which are
dures do you think should not be performed in rooms found in the skin of the trunk and feet. When compared
housing other rats? What activities might be okay? with airborne signals, seismic signals generally travel far-
ther and maintain their integrity longer, so vocalizations
transmitted seismically may be detectable at greater dis-
tances, a feature that would be especially important to far-
SUBSTRATE VIBRATIONS ranging species like elephants. It is also possible for
Some animals communicate using seismic signals (Hill elephants to simultaneously monitor groundborne signals
2001). These signals are encoded in the pattern of vibra- and airborne signals, which travel at different speeds and
tions of the environmental substrate, such as the ground therefore arrive at the receiving elephant at slightly dif-
or the water surface, and can be produced through per- ferent times, to determine the distance of the vocalizing
cussion on the environmental substrate. For example, individual (O’Connell-Rodwell 2007).
territory ownership is declared through foot drumming
on the ground in kangaroo rats (Randall and Lewis 1997)
CHEMICAL SENSES
and through head banging against the burrow ceiling in
blind mole rats (Rado et al. 1987; Nevo et al. 1991). The chemical senses, smell and taste, are another channel for
Tapping on the surface of water creates seismic signals communication. These senses are based on the movement
of a slightly different sort—ripples, or surface waves. of odor molecules from signaler to receiver (Wyatt 2003).
Ripple signals have been described in several species of Information may be carried by chemicals over long
water striders (Figure 16.6). These vibrational signals distances, especially when assisted by currents of air or
362 Chapter 16 / Communication: Channels and Functions

water. The rates of transmission and fade-out time are Some signals are complex blends of chemicals. In
slower than for visual or auditory signals. Depending on these situations, it may be the relative proportions of dif-
the function of the signal, this may be an advantage. In ferent chemicals that produce the effect on a sender
the delineation of territorial boundaries, for example, a rather than the presence of a particular chemical in the
durable signal is more efficient because it remains after mixture. In other words, effects may be produced by the
the signaler has gone (Figure 16.7). Some mammals full chemical “image” of the signal, sometimes called the
increase the signal life of chemicals used to mark terri- “odor mosaic” (Johnston 2000). Recall from Chapter 7
tories by secreting them with oily carrier substances, such that scent marking is the act of strategically placing a
as those from sebaceous glands, or by associating them chemical mark in the environment. Here we consider the
with certain urinary proteins that slow the release of the chemical image left by common marmosets (Callithrix
signals (reviewed in Wyatt 2003). Another benefit is that jacchus), small primates endemic to the forests of Brazil,
these long-lasting chemical signals do not require con- and easily recognized by their white ear tufts and fore-
tinued energy expenditure by the sender. Furthermore, head blaze (Figure 16.8). Females of this species mark
chemical signals can be used where visibility is limited. their environment by depositing a complex mixture of
The ease with which the sender of a chemical signal can scent from several sources, including urine, feces, secre-
be located varies with the chemical emitted, but it is usu- tions from the reproductive tract and nearby oil and
ally more difficult to locate a signaler that is using apocrine glands (apocrine glands are restricted to certain
chemicals than one using visual or auditory signals. body regions, such as the anogenital area, where they
secrete odorous substances onto hair follicles). In the lab-
a oratory, female common marmosets were trained (using
positive reinforcement) to deposit a scent mark on a spe-
cial collection device. Detailed analysis of the scent marks
revealed the presence of 162 distinct chemicals! While
the scent marks of most females contained all of the
chemicals, the marks of individual females differed in the
amounts of specific chemicals. Previous behavioral test-
ing had shown that females could discriminate between
the scent marks of familiar conspecifics and those of unfa-
miliar conspecifics. Thus, it appears that each female has
a unique scent signature (or odor mosaic), which is based
on the ratios of chemicals—particularly the volatile
ones—in the scent mark (Smith 2006).
Sometimes the meaning of a particular signal may
b vary with the context in which it is given. This is the case
with a chemical signal sent by a queen honeybee. The

FIGURE 16.7 Muskrats mark their territories with

secretions whose musky scent is long-lasting. (a) These
relatively large, chunky rodents are found in wetlands
throughout much of North America. (b) Paired glands
near the genitalia emit an odorous secretion that is
deposited with urine and feces on defecation posts.
These posts are found on rocks along runways (shown
here) and on their homes (lodges). The durable scent
may serve to advertise territory boundaries and to FIGURE 16.8 Common marmosets deposit scent marks
attract mates. that are complex combinations of many chemicals.
 Channels for Communication 363

chemical trans-9-keto-2-decenoic acid is picked up

from the queen as the workers groom her and is dis-
tributed throughout the hive, along with the food that
is shared by the workers. When attained in this manner,
the chemical prevents the rearing of any additional
queens. However, the queen also exudes the chemical as
she soars skyward on her nuptial flight. In this context,
the same chemical causes males to gather around her. In
other words, it serves as a queen inhibitor or as a sex
attractant, depending on the context (Robinson 1996).
The detection of chemical cues may occur at a dis-
tance, such as when volatile chemical cues become air-
borne and reach mammal noses or insect antennae; this
is called remote chemoreception. Alternatively, chemi-
cal cues may be detected through direct contact with FIGURE 16.9 Flehmen is a characteristic posture in
the chemical, which (not surprisingly) is called contact which the head is raised and the upper lip is curled
chemoreception. For example, you have probably seen back. It serves to deliver nonvolatile communicatory
one ant rapidly and repeatedly touch the body of another chemicals, such as those found in urine or glandular
ant with its antennae. What you may not have appreci- secretions, to the vomeronasal organ.
ated at the time is that the touchy-feely individual is eval-
uating the mixture of chemicals on the outside of the other
ant’s body to determine whether the individual is a colony any males in the vicinity, binds to the receptor hairs on
member or an intruder. Contact chemoreception is asso- the male’s antennae. As few as 200 molecules of bombykol
ciated with relatively nonvolatile chemical cues. have an immediate effect on the male’s behavior—
Some species of amphibians, reptiles, and mammals he turns and flies upwind in search of the emitting female
have a vomeronasal (or Jacobson’s) organ that is impor- (Schneider 1974). The highly specific olfactory receptors
tant in chemical communication between mates, parents on the antennae of male silk moths have been identified
and offspring, and rivals. It is anatomically separate from and characterized (Sakurai et al. 2004). Other examples
other chemosensory structures, and its neural wiring goes of releaser pheromones in insects are trail pheromones,
to brain regions different from those associated with the which direct the foraging efforts of others, and alarm sub-
main olfactory system (Halpern and Martinez-Marcos stances, which warn others of danger. Vertebrates also
2003). The vomeronasal organ is located in the roof of the produce releaser pheromones. For example, lactating rab-
mouth or between the nasal cavity and the mouth, so com- bits produce mammary pheromone, which stimulates
municative chemicals must reach it through the nose, their pups to search for and grasp onto a nipple (Hudson
mouth, or both. Because the chemicals that the and Distel 1995; Moncomble et al. 2005; Schaal et al.
vomeronasal organ is specialized to detect are primarily 2003). Quick attachment by pups to a nipple is critical
nonvolatile, they must be brought to the organ. In a snake, because mother rabbits return to their nest to nurse pups
for example, the vomeronasal organ is on the roof of the only once a day for about three to five minutes!
mouth and the chemicals are delivered to it by the tongue. Primer pheromones exert their effect more slowly, by
A mammal, however, must lick or touch its nose to the altering the physiology and subsequent behavior of the
chemicals, which are usually found in urine or special recipient. In insect societies, queens control the repro-
body secretions left on surfaces by conspecifics. Following ductive activities of nest mates largely through primer
this contact, many mammals make a characteristic facial pheromones. For example, a queen honeybee produces
grimace, known as flehmen, which helps transfer the several compounds from her mandibular gland that
chemicals to the vomeronasal organ. In flehmen, the head ensure that she will remain the only reproductive indi-
is raised and the lips are curled back (Figure 16.9). vidual in the colony (Robinson 1996; Winston and
Chemicals produced to convey information to other Slessor 1992). This pheromone coats the queen’s entire
members of the same species are called pheromones. body surface but is most concentrated on her head and
Some of these, releaser pheromones, have an immediate feet. Most of the pheromone spreads through the colony
effect on the recipient’s behavior. A good example of a by the activities of the workers that are attending the
releaser pheromone is a sex attractant. The most famous queen, but some is spread through the wax of the comb
sex attractant is probably that of the female silk moth, (Naumann et al. 1991). The pheromone prevents the
Bombyx mori. She emits a minuscule amount, only about workers from feeding larvae the special diet that would
0.01 microgram, of her powerful sex attractant, bom- cause them to develop into rival queens. When the queen
bykol, from a small sac at the tip of her abdomen (Figure dies, the inhibiting substance is no longer produced, and
16.10). This pheromone, which is carried by the wind to new queens can be reared (Wilson 1968).
364 Chapter 16 / Communication: Channels and Functions

a b

FIGURE 16.10 Female silk moths produce the sex attractant bombykol. (a) The tip of the female’s abdomen has a pair
of glands, shown in an expanded active state here. Within these glands is a small quantity of bombykol. (b) The feath-
ery appearance of the male’s antennae is due to numerous branches, each containing many odor-receptor hairs. Half of
the odor receptors on the male’s antennae are tuned to bombykol, which is why the male is sensitive to such minute
quantities of it. On detecting bombykol, he turns and flies upwind in search of the female.

Vertebrates also produce primer pheromones that organ by the flicking tongue. Chemicals from prey
influence reproductive activity in various ways. They species clearly do not fit the definition of a pheromone
may help regulate reproductive activities so that repro- (which requires communication to be within a species)
duction occurs in the appropriate social or physical set- and the behavior of interest is foraging, not communi-
ting. Table 16.2 lists some of the effects of primer cation. In addition, some chemical signals that fit the
pheromones known in mice. definition of pheromone, such as the rabbit mammary
Before leaving the topic of chemical communica- pheromone discussed earlier, appear to act via the main
tion, we want to make sure that we have not left you olfactory system and not through the vomeronasal
with the impression that the vomeronasal organ is for organ. Consider, also, the scent marks left by female
pheromones and the main olfactory system (i.e., the hamsters around their territories (Johnston 1998;
olfactory epithelium, which is the layer of cells lining Meredith 1998; Swann et al. 2001). These marks con-
the nasal passages that contains olfactory receptors) is tain vaginal fluid pheromones that a male detects via his
for smelling general odorants. In fact, there are no clear- main olfactory system and that prompt him to locate the
cut functional differences between the vomeronasal female. Once the male has found the female, another
organ and the main olfactory system (Baxi et al. 2006). component of the vaginal secretion, this time perceived
The vomeronasal organ, it has been found, can be stim- through his vomeronasal organ, prompts him to inves-
ulated by substances other than pheromones. For tigate and mount her. This is how it works for sexually
example, the vomeronasal organ of a hunting snake inexperienced males. Sexually experienced males, how-
responds to chemical cues of prey species brought to the ever, have learned the odor cues of receptive females and
no longer need the input via the vomeronasal organ to
stimulate mounting. Thus, pheromones may be per-
TABLE 16.2 Effects of Some Primer ceived by either the vomeronasal organ or the main
Pheromones in Mice. olfactory system. In some situations, the two systems
work together, although this relationship may be mod-
Origin Recipient Effect ified by experience.
Female urine Female Inhibits estrous cycling
and ovulation TOUCH
Male urine Female Induces estrous cycling Animals also communicate by touch (Figure 16.11).
and ovulation Tactile messages can be sent quickly, and it is easy to
locate the sender, even in the dark. Honeybee scouts, for
Female urine Male Prompts release of
example, inform nest mates of the location of a food
testosterone and
luteinizing hormone
source by dancing. The recruits cannot see the chore-
ography because the hive is so dark, but they follow the
 Channels for Communication 365

ELECTRICAL FIELDS
Two distantly related groups of tropical freshwater fishes
produce electrical signals used in both orientation
(electrolocation, discussed in Chapter 10) and communi-
cation. These groups are the knifefishes (gymnotiforms)
of South America and the elephant-nose fishes (mormyri-
forms) of Africa. Gymnotiforms and mormyriforms are
described as “weakly electric” to contrast them with fishes,
such as torpedo rays or electric eels, which generate very
strong electric discharges to stun prey or predators.
The electrical signals are generated by electric
organs derived from muscle. When a normal muscle cell
contracts, a weak electrical current is generated. The
modified muscle cells in an electric organ also generate
FIGURE 16.11 Social grooming, a form of tactile com-
munication that builds and maintains social bonds, is a weak electrical current, but because they are arranged
displayed by many mammals, including horses. in stacks their currents are added, resulting in a stronger
current. When an electric organ of a gymnotiform or
mormyriform discharges, the tail end of the fish, where
the electric organ is located, becomes momentarily neg-
dancers’ movements by touching them with their anten- ative with respect to the head. Thus, an electrical field
nae (discussed shortly). Tactile signals are obviously only is created around the fish (Figure 16.12a). This electri-
effective over short distances and are not effective cal field is the basis of the signal. Diverse signals can be
around barriers. A message sent via touch may be var- created by varying the shape of the electrical field, the
ied in several ways, including how the recipient is discharge frequency, and the timing patterns between
touched (e.g., rubbing, patting, pinching), where the signals from the sender and receiver, as well as by stop-
recipient is touched, the frequency and duration of ping the electrical discharge. Electric organ discharges
touching, and the extent of surface area touched are detected by special sensory receptors in the skin
(Hertenstein et al. 2006a). called electroreceptors (Bullock et al. 2005).
Humans also skillfully send and decode tactile sig- Two general patterns of electric organ discharge occur
nals. Consider one study in which two strangers were in weakly electric fish: wave-type and pulse (Hopkins 1974;
asked to sit at a table, where they were separated by a Zakon et al. 2008). These two distinct patterns have inde-
barrier (an opaque black curtain). One person was pendently evolved in both gymnotiforms and mormyri-
charged with sending a tactile signal that conveyed one forms. Those species whose discharges are classified as the
of 12 different emotions and the other person was to wave-type pattern produce signals continuously (whether
receive the signal and decode it. The person to receive active or inactive) and the waveform of the signal is
the tactile signal positioned his or her bare arm (from monophasic, resembling a sine wave (Figure 16.12b). In con-
elbow to fingertips) on the sender’s side of the curtain; trast, species whose electric organ discharges are emitted as
this ensured that the receiver could not see any part of pulses produce discharges at higher rates when active and
the tactile signal. The participants were asked not to lower rates when resting, and the waveform of the signal has
speak to one another during the experiment. Twelve a complex multiphasic structure (Figure 16.12c).
emotion words were presented serially to the sender on What are the characteristics of electrical signals?
sheets of paper in random order; once presented with a When the electric organ discharges, an electrical field is
sheet, the sender was asked to convey that particular created instantaneously. It also disappears at the instant
emotion through touching the receiver’s arm. The recip- the discharge stops. As a result, electrical signals are
ient of the touch was then asked to choose which of the ideally suited for transmitting information that fluctuates
12 emotions the sender was trying to communicate. quickly, such as aggressive tendencies (Hagedorn 1995).
Study participants decoded fear, anger, love, gratitude, An electrical signal does not propagate away from the
sympathy, and disgust at levels above chance, but per- sender but instead exists as an electrical field around the
formed less well with happiness, sadness, surprise, sender. Because an electrical signal is not propagated, its
embarrassment, envy, and pride. In addition, specific waveform is not distorted during transmission. As a
types of touch were associated with specific emotions: for result, the waveform of the electrical signal may be a reli-
example, trembling with fear, hitting and squeezing with able indicator of the sender’s identity (Hopkins 1986a).
anger, stroking with love, shaking the hand with grati- Although the waveform is generally constant for a par-
tude, patting with sympathy, and pushing with disgust ticular individual, it is different in males and females and
(Hertenstein et al. 2006b). among different species (Stoddard et al. 2006).
366 Chapter 16 / Communication: Channels and Functions

a b Wave-type discharges

Sternopygus macrurus

Eigenmannia virescens

Sternarchorhamphus macrostomus

Apteronotus albifrons
FIGURE 16.12 Some tropical freshwater
fishes communicate via weak electrical
5 msec signals. (a) Discharges from an electric
organ create an electrical field around
c Pulse discharges a weakly electric fish that is used as a
Rhamphichthys rostratus
communication signal. The signal can
be varied by altering the shape of the
electrical field, the waveform of the
Gymnorhamphichthys hypostomus electrical discharge, the discharge
frequency, and the timing patterns
between signals from the sender and
Hypopygus lepturus receiver, as well as by stopping the
electrical discharge. (b) Some species
Hypopomus artedi of weakly electric fish produce elec-
trical signals continuously with
monophasic waveforms. (c) Other
Hypopomus brevirostris
species produce electrical signals in a
pulse pattern, often with multiphasic
Gymnotus carapo
waveforms. These so-called pulse-fish
discharge at high rates when active
and low rates when at rest. (From
1 msec 10 msec Hopkins 1974.)

Electrical signals are well suited for communication Hopkins 2007). Electrical signals are also used during ago-
in the environments in which the weakly electric fish live. nistic encounters, where certain patterns of discharge have
Both gymnotiforms and mormyriforms are active at been associated with aggression, dominance, and submis-
night and generally live in muddy tropical rivers and sion (Hupé and Lewis 2008; Triefenbach and Zakon 2008).
streams, or at depth in tropical rivers, where visibility is Finally, there is some evidence that parents and offspring
poor. Electrical signals can move around obstacles and may communicate via electrical signals to maintain prox-
are undisturbed by the suspended matter that creates imity to one another (Crampton and Hopkins 2005).
murky water. However, they are effective only over short
distances, about 1 to 2 m, depending on the depth of the
water and the relative positions of the sender and MULTIMODAL
receiver (Hopkins 1986b, 1999). The shortness of their COMMUNICATION
effective distance may actually be an advantage. Different
weakly electric species may coexist in an area, so the Animals do not always use a single channel when
short effective distance of the signal may reduce electri- communicating. In fact, the displays of many animals con-
cal “noise” when many individuals are signaling at once. tain signals from two or more sensory modalities. This
Weakly electric fish use electrical signals to convey the type of communication is called multimodal communi-
same messages that other organisms send by other chan- cation, and the signaling in different channels can occur
nels. For example, males of some species not only adver- either simultaneously or sequentially (Partan and Marler
tise their sex and species by electrical signals, but also court 2005). Here we focus on simultaneously produced multi-
females by “singing” an electrical courtship song modal signals. For example, the courtship display of a
(Hagedorn 1986). Indeed, in some species, a courting male male bird may simultaneously contain visual and auditory
and female engage in electrical duetting, a coordinated pat- signals. During allogrooming, a monkey may touch the
tern of communication in which the signals of one indi- skin of its companion while at the same time making a cer-
vidual alternate with those of the other (Wong and tain facial expression and vocalization. We know from
 Functions of Communication 367

experience that when humans communicate through

speech, information is often simultaneously passed
through the visual channel. Indeed, it is often much eas-
ier to understand a spoken message if we have both visual
and auditory information from the sender (especially if the
information from the two channels corresponds and does
not conflict). Finally, we mentioned previously that an ele-
phant vocalization could have seismic (ground borne) as
well as auditory (airborne in this case) components.
During multimodal communication, the messages
conveyed in different signaling channels can be redun-
dant (convey the same thing) or nonredundant (convey
different things). Whether the messages are redundant
or nonredundant can be determined by presenting each
component of the display individually to a recipient and
seeing how the recipient responds (Partan and Marler
2005). In the case of redundant messages, the recipient
should respond in the same way to each component pre-
sented individually. In the case of nonredundant mes-
sages, the recipient should respond differently to the FIGURE 16.13 The courtship display of a male brush-
different components. legged wolf spider is multimodal, involving both visual
As an example of a multimodal communication, con- and seismic components. The visual signal consists of
sider the courtship display of male brush-legged wolf spi- rapidly raising and lowering the forelegs, which have
ders (Schizocosa ocreata) (Gibson and Uetz 2008; Uetz and conspicuous tufts.
Roberts 2002; Uetz et al. 2002). Males of this species
court females using a complex “jerky-tapping” display
that simultaneously (and sequentially) contains both tions, having redundant signals may be necessary to pre-
visual and seismic signals. The visual part of the display vent miscommunication between males and females,
consists of the male rapidly raising and lowering his first which could lead to the male becoming the female’s next
pair of legs; each of these legs is adorned with a conspic- meal rather than her mate. Miscommunication could also
uous tuft of bristles (Figure 16.13). The seismic compo- lead to mating with the wrong species, which would also
nents of the male courtship display include: (1) be costly to all involved (Gibson and Uetz 2008).
stridulation produced by organs on the pedipalps Multimodal communication also has costs for
(appendages modified as copulatory organs in male spi- senders and receivers (Partan and Marler 2005).
ders), (2) up and down bouncing of the entire body, and Signaling in multiple sensory modalities may require
(3) striking the substrate with chelicerae (mouthparts). more of the sender’s energy, and recipients may need
Experimental presentation of isolated visual and seismic more energy to receive and process multiple signals.
components reveals that the signals in both channels elicit Furthermore, signaling in multiple channels may make
receptivity in females. In fact, the information content of senders more susceptible to predation. Even receivers
visual and seismic signals appears quite similar; both allow could be more susceptible to predation as a result of
a female to assess the quality (size and physical condition) devoting more senses to detecting signals from con-
of the displaying male. Thus, visual and seismic signals specifics and fewer to detecting predators.
appear to contain largely redundant information in the
male wolf spider’s display. If females responded differ-
ently to individual presentations of visual and seismic sig- FUNCTIONS OF
nals, then nonredundant messages would be more likely. COMMUNICATION
Multimodal communication has benefits for senders
and receivers (Hebets and Papaj 2005; Partan and Marler Communication is integral to much of animal behav-
2005). For nonredundant multimodal signals, more infor- ior. Some signals bring individuals together, and oth-
mation can be sent per unit time. For redundant multi- ers help keep them apart. Some signals settle conflicts,
modal signals, a major benefit is insurance that the whereas others incite them. Some cause alarm and oth-
message will be received and accurately recognized, even ers are pacifying. Some signals attract suitors and others
if one sensory channel is especially noisy. In the case of repel them. Not surprisingly, descriptions of commu-
the brush-legged wolf spiders, the leaf litter in which they nication are woven throughout many chapters in this
live is a structurally complex habitat where communica- book. Here we will highlight just a few of the messages
tion is difficult at best. Under these challenging condi- that signals convey.
368 Chapter 16 / Communication: Channels and Functions

SPECIES RECOGNITION with the bright yellow cheliped characteristic of the

species, even if it is a male of another species with its claw
It’s striking how attuned many species of animals are to the
painted yellow (Detto et al. 2006).
presence of conspecifics. For example, a male chickadee
We expect that selection for species to have distinctly
may give little or no observable response to the song of a
different signals will be strongest where species occur
gray-crowned rosy finch but immediately increases its call
together in the same place, a prediction that is borne out
rate when a speaker broadcasts the song of another chick-
in frogs. Male frogs and toads usually attract their mates
adee (Charrier and Sturdy 2005). It’s no wonder.
by calling at night (Figure 16.14). Judging by the
Conspecifics are likely to be competitors: they may want
response of nearby females, these calls seem to be the
the same food, shelter, and mates, and thus should be dealt
amphibian equivalent of “I’m over here. Come find me.”
with accordingly. Other conspecifics might be friendly
The calls of different species are impressive in their vari-
members of a social group, or potential mates that should
ety, ranging from high-pitched peeps to trills to bass-drum
be wooed. It’s also adaptive not to mistake heterospecifics
booms. Males of several species often serenade together
for conspecifics, as it is a waste to squander time and energy
in a chorus, and a female must choose one of her own kind
courting an animal with whom it is impossible to produce
from the variety of callers at the local pond. The female’s
viable offspring, or to defend a territory from an individ-
ability to discriminate is important, as males generally are
ual that is not competing for resources or mates.
unselective and will grab any female (or male) of approx-
Virtually all sensory channels are used for species
imately the right size that happens by (Gerhardt 2001). As
recognition in one species or another (reviewed in
a result, selection has favored clear species differences in
Bradbury and Vehrencamp 1998). Birds generally use
calls, and a reasonably skilled person can determine the
song characteristics such as frequency (which notes are
species of many frogs by their calls alone. The ability of
sung) and syntax (how the notes are strung together).
green tree frog females to distinguish conspecific calls
Some animals, such as crickets that make sound by scrap-
from other calls has been shaped by selection. In parts of
ing together a file and rasper on their wings, don’t have
its range, Hyla cinerea shares its breeding ponds with the
the ability to modulate the notes of their songs, so instead
closely related H. gratiosa. In other regions, H. gratiosa is
they rely on differences in their rhythmic patterns. Other
not present, and H. cinerea lives alone. Female H. cinerea
species, particularly insects, use olfactory cues. For exam-
that share the same area with H. gratiosa prefer calls with
ple, some species use species-specific pheromones to
the acoustic properties that distinguish them from H. gra-
attract mates. Still others rely on visual cues, such as dis-
tiosa calls (Höbel and Gerhardt 2003).
plays, color patterns, or other aspects of appearance.
It is not always true that animals can unfailingly dis-
Researchers can pinpoint exactly which part of a sig-
tinguish conspecifics from others. As you might predict
nal conveys information about species identification
from sexual selection theory (Chapter 14), males of
through experimental manipulation. Bird songs, for
many species indiscriminately court females. Australian
instance, can be easily digitized and manipulated so that
Julodimorpha bakewelli beetles will even attempt to
components are omitted, added, or played in a different
copulate with discarded beer bottles. Apparently, the fact
order or at a different speed. Researchers can then play
that the bottles are, like females, shiny brown and bumpy
back the modified calls to watch the response of free-
is enough to overcome their noticeable lack of ability to
living birds. By this technique, we know that the chick-
adees described earlier pay attention to a number of
acoustic features when deciding whether a strange call
is produced by a conspecific (Charrier and Sturdy 2005).
In some cases, experimental manipulation leads to coun-
terintuitive results. For example, let’s take another look
at fiddler crabs (genus Uca), which we mentioned earlier
in the section on visual cues. In some places, many
species can be found packed into close quarters on the
beach; in one study, there were 12 species in an area
slightly smaller than a quarter of a tennis court. Crabs
signal by waving their large front claws, the chelipeds,
and different Uca species wave in different patterns: U.
rhizopharae waves up and down, U. annulipes in large cir-
cles, and U. pugilator in small circles (Crane 1941). These
display patterns have long been thought to be important
in species recognition. However, in a test of another pair FIGURE 16.14 A male American toad (Bufo americanus)
of Uca species, U. mjoebergi and U. capricornis, species is calling to attract a mate. The male amplifies the sound
recognition depended not on display behavior but on the produced by his vocal cords by inflating the large vocal
color of the cheliped. Female U. mjoebergi prefer a male sac beneath his chin.
 Functions of Communication 369

give behavioral signals (Gwynne and Rentz 1983). Even

females, which are generally the more selective sex, some- STOP AND THINK
times make erroneous choices (reviewed in Gröning and When males signal to attract prospective mates, they usu-
Hochkirch 2008). In some cases, such as recently intro- ally give auditory or visual signals, whereas females that
duced invasive species that share some traits with natives, signal usually use the olfactory channel (Bradbury and
Vehrencamp 1998). Why might this be so? Think in par-
natural selection has not had time to favor those indi-
ticular of the duration of receptivity, the costs of signals,
viduals that can successfully make the distinctions.
and the dangers of signaling.

MATE ATTRACTION
Animals that spend most of their time alone have the COURTSHIP AND MATING
challenge of locating each other when it is time to breed.
Besides being species-specific, the signals that attract a Once the individuals are close enough to interact, they
mate must be easy to locate and effective over long dis- court before committing themselves to mating.
tances so that males and females can find each other even Communication has several functions: identification of
if the species members are widely distributed. For this a partner of the appropriate species and sex, assessment
reason, chemical and auditory signals are used commonly, of mate quality, coordination of the mates’ behavior and
but not exclusively, for attraction. For example, we have physiology, and, in some species, maintenance of pair
already described the sex attractant pheromone of the bonds after mating.
female silkmoth (Bombyx mori) in this chapter. Under the
right wind conditions, the pheromone may attract males Identification of the Opposite Sex
from perhaps 100 m away. The sex attractant pheromone
We’ve already discussed how animals might signal their
of another species, Actias selene, is also potent. In one
species identity. Animals might also communicate their
experiment, males that had been experimentally displaced
sex. In many cases, the differences between males and
46 km away were still able to relocate newly hatched
females are readily apparent and are indicated by the
females at the original site (Immelmann 1980).
development of antlers and other secondary sexual char-
Auditory signals also carry well, especially when
acteristics. Some species have displays that seem to
amplified by communal displaying or by special anatom-
showcase the aspects of their body that indicate their sex.
ical or environmental structures. We’ve already discussed
When a female stickleback (Gasterosteus aculeatus) enters
the courtship songs of birds, frogs, and crickets, and how
a male’s territory, she reduces the probability of attack
they can attract mates from long distances. As you can
by assuming a head-up position that displays her egg-
see in Figure 16.15, female crickets will approach a loud-
swollen abdomen and distinguishes her from an intrud-
speaker that broadcasts the courtship song of a male of
ing male (Tinbergen 1952).
its species.
In other species, differences are more subtle. For
example, male blue-ring octopuses (Hapalochlaena
lunulata) apparently cannot distinguish males from
females until quite late in the courtship sequence.
Octopuses mate by inserting their modified third right
arm into the mantle cavity of the female and releasing a
spermatophore, or sperm packet. Male blue-ring
octopuses insert their arms indiscriminately into both
males and females, but only release spermatophores
when inserted into females (Cheng and Caldwell 2000).
Male–male interactions are brief and not aggressive;
apparently the fitness costs of making an insertion into
a male are low.

Mate Assessment
Courtship may allow a female to judge the qualities of
her suitor so that she can choose the one most likely to
enhance her own reproductive success (or, more rarely,
it may allow the male to choose the characteristics of an
FIGURE 16.15 A female cricket will be attracted to a appropriate female). This function of communication
loudspeaker that is broadcasting the courtship song has been widely studied. We provide many examples and
of a conspecific male. more details about the selection pressures on mate
370 Chapter 16 / Communication: Channels and Functions

choice in both Chapters 14 and 17, and only briefly dis- assessing their weight (Soler et al. 1996, 1999). Male
cuss some examples here. wheatears that carried heavier stones scored better on a
Courtship displays may provide a means for evalu- test of immunocompetence, an indicator of male health
ating a suitor’s qualities, including his physical prowess, (Soler et al. 1999).
ability to provide food for the offspring, or even the
extent of his commitment. For example, male common Coordination of Behavior and Physiology
terns (Sterna hirundo) catch fish and offer them to the Males and females have very different reproductive sys-
female. She compares the quantity of fish provided by tems that may not always be synchronized. Courtship
her various suitors and usually chooses the best fisher- displays can function to coordinate the couple’s behav-
man. The number of fish a male provides during ior and physiology. Ring doves (Streptopelia risoria) are a
courtship is correlated with the quantity he later provides well-studied example. Each step in the mating sequence,
to his chicks. The quality of the courtship offering, then, from initial courtship through nest construction, copu-
is a reliable indicator of the male’s ability to provide for lation, egglaying, and feeding nestlings, is choreo-
the pair’s offspring (Wiggins and Morris 1986). graphed by the behavior and hormonal state of each
An odder example is the wheatear, a small bird from partner (Figure 16.16) (reviewed in Fusani 2008).
Spain and Morocco. Males demonstrate their devotion Because displays used for coordinating receptivity
to their mates by collecting stones in their beaks, one by are generally between partners near each other many are
one, and carrying them to cavities that serve as poten- visual or tactile. Some, however, rely on pheromones
tial nest sites. In one study, males carried an average of that are delivered at close range: a male mountain dusky
277 stones weighing 1.8 kg over the course of a single salamander (Desmognathus ochrophaeus) applies a
week (Morena et al. 1994)—impressive for a 35- to courtship pheromone to the female by pulling his lower
40-g bird! Females watch the males carry stones, and jaw across the female’s back and angling his snout so that
even occasionally heft stones that have been carried as if his premaxillary teeth scrape the female’s skin, thereby

Adult male dove ale Adult female dove

f fem
c eo
Hormones Behavior en Behavior Hormones
es
Pr

Responses
Courtship
to courtship
Testosterone

Nest Nest
Estrogen

building building

FIGURE 16.16 The mating behavior of

Progesterone

Copulation Copulation a pair of ring doves is choreographed

by signals. The sight of the female
causes the male to increase his testos-
terone production, and he begins to
display. In response, the female coos,
Egg laying and her own vocalization stimulates
estrogen production. Estrogen stimu-
lates the development of the oviduct.
Both begin nest building. In the
female, estrogen declines and proges-
Incubation Incubation terone increases. The pair copulates
and the female lays two eggs. The
sight of the eggs induces incubation
and suppresses courtship, testosterone,
and progesterone. An increase in
Prolactin

Prolactin

Feeding Feeding prolactin leads to incubation.

squabs squabs
(From Bradbury and Vehrencamp
1998, based on Nelson 1995.)
 Functions of Communication 371

“injecting” the pheromone into the female’s circulatory a female received a pheromone treatment, she assumed
system. The female then indicates her receptivity by a tail-straddle position, indicating receptivity, 43 minutes
assuming a tail-straddling position, and the male deposits (26%) sooner and mated 59 minutes (28%) sooner than
a packet of sperm called a spermatophore (Figure she did after receiving a saline injection (Figure 16.17b).
16.17a). Houck and Reagan (1990) demonstrated that
the male’s courtship pheromone makes the female more Maintenance of Pair Bonds
receptive. They staged a total of 200 courtship encoun- A final function of communication in the context of
ters between 50 pairs of salamanders on each of four reproduction is the formation of bonds between (rela-
nights. By removing the mental gland, which produces tively) monogamous pairs of animals. As with displays
the pheromone, from each male, they prevented the that coordinate mating, pair-bond displays often occur
males from delivering any pheromone during courtship. at close range and are thus often visual or tactile, such
These glands were then used to create an elixir con- as dusky titi monkeys that sit with their tails intertwined
taining the courtship pheromone. Thirty minutes before (Figure 16.18).
some encounters, each female was treated with this Let’s examine a particular pair bond in detail (Sogabe
pheromone-containing elixir. Before other encounters, and Yanagisawa 2007, 2008). Pipefish (Corythoichthys
the same female was treated with a saline solution. After haematopterus) are long skinny fish related to seahorses.
Like seahorses, males have a brood pouch into which
females deposit eggs. The males care for the offspring
a for one to eight weeks. Pairs are monogamous and only
form bonds with other individuals when their mate dis-
appears. Male and female pipefish conduct a greeting
ceremony every morning. They approach one another,

b
Courtship pheromone
Saline
200
Time (min)

100

0
Courtship Tail Spermatophore
straddling transfer
Courtship latencies

FIGURE 16.17 During courtship, a male mountain dusky

salamander injects a female with his courtship pheromone.
(a) The male alternately scrapes the female’s back with his
teeth and swabs her with the pheromone, which is produced
by a gland beneath his chin. The female signals her readi-
ness to mate by placing her chin on the base of the male’s
tail and straddling his tail. (b) The courtship pheromone
makes the female receptive. Tail straddling and mating
occur more quickly when a female has been treated with FIGURE 16.18 Dusky titi monkeys maintain their pair
a courtship pheromone compared to a control treatment bond with physical contact. (From Bradbury and
with saline. (From Houck and Reagan 1990.) Vehrencamp 1998, based on Moynihan 1966.)
372 Chapter 16 / Communication: Channels and Functions

swim in parallel side-by-side, cross abruptly over each

other’s back, arch their bodies, or rise into a vertical posi-
tion. Interestingly, although the home ranges of pairs
overlap with those of other individuals, greeting cere-
monies are only performed by members of a pair.
Greetings are even carried out during the nonrepro-
ductive season. Members of a pair meet at a particular
site every morning, exchange greetings for several min-
utes, and then have no further contact for the rest of the
day. Because no other benefits seem obvious to this off-
season greeting—the pairs do not cooperate in any other
behavior, for example—it is thought that its function is
solely to maintain the bond with the partner in
preparation for the next breeding season.

FIGURE 16.19 Allogrooming between crab-eating

MAINTAINING SOCIAL BONDS macaques. Besides the benefits of skin and fur care and
In some species, social group members, and not just parasite removal, allogrooming is important in social
mated pairs, use communication to maintain their bonds. bonding.
These communicatory signals are generally based on
contact: resting together, nuzzling, and touching in gen- part of the body was groomed. The responses to the five
eral tend to firm social bonds (Eibl-Eibesfeldt 1975). body regions typically groomed may reflect a continuum,
Many of the greeting signals exchanged by animals from a tendency to maintain an affiliation at one extreme
as they encounter one another serve as an assurance of to a tendency to terminate the interaction at the other.
nonaggression. Chimpanzees often greet by touching Grooming can also facilitate a transition from one type
hands or sometimes by placing a hand on the compan- of social interaction to another. For example, in pri-
ion’s thigh (Goodall 1965). Sea lions rub noses, and lions mates, grooming can smooth over tension and restore
rub cheeks. African wild dogs greet one another by relationships after conflicts (e.g., Watts 2006).
pushing their muzzles into the corners of each other’s
mouths (Schaller 1972). You may have been head-
bumped by your cat—you should take this as a compli- ALARM
ment. (It is more appealing than engaging in the Alarm signals warn another animal of danger. Usually,
greeting ceremony of the domestic dog, which includes danger presents itself in the form of predators, but indi-
anal sniffing.) viduals may also have to guard against other members
In many mammals, social bonds appear to be built of their species bent on infanticide or some other form
and maintained through social grooming, also called of aggression.
allogrooming (Figure 16.19). If we assume that self-
grooming has skin care as its primary function, what evi-
dence do we have to support a communicative function Alarms That Cause Animals to Flee
for social grooming? Maria Boccia compared several Alarm signals have different functions, and their char-
aspects of social grooming and self-grooming in rhesus acteristics depend on their function (Bradbury and
monkeys: body site preference, duration (both overall Vehrencamp 1998). Many alarm signals cause those who
and to specific areas of the body), and method (stroking hear them to flee or take cover. We can predict that
or picking). She reasoned that if the primary function of “flee” signals should share several characteristics. These
both social grooming and self-grooming was hygiene, signals should be easy to make quickly, before it is too
then these physical aspects of grooming would be the late. From the signaler’s point of view, it’s also best if the
same in both. However, social grooming was found to signal is difficult to locate. Alarm signals from several
be different from self-grooming in each of these respects. sensory channels fit the bill: examples include rapid
Therefore, she concluded that skin care is not the most visual signals such as the flash of a deer’s tail; highly
important factor in molding the form of social groom- volatile pheromones that diffuse quickly; and high-
ing (Boccia 1983). Furthermore, she showed that the pitched sounds that are hard to localize. As you might
message of the tactile signal varies according to the body expect given the similar selective pressures, species often
site being groomed (Boccia 1986). In other words, the share similar alarm signals. For example, a number of
recipient monkey’s response depended on which part of species of passerine birds have the characteristics that
its body was groomed. The animal being groomed was would be expected to make them difficult to locate—they
likely to move away from the groomer when the posterior begin and end gradually and employ only a few wave-
 Functions of Communication 373

9 predator, they emit staccato grunts that are loud and

Reed bunting
7
low-pitched. These features allow the grunts to be eas-
ily located and to be transmitted over long distances,
5 thereby broadcasting the position of the predator.
9 Other monkeys run into thickets, where the dense brush
Blackbird makes it difficult for a swooping eagle to catch them
7 (Struhsaker 1967). The responses just described are also
Frequency, kc per sec

5
typical of responses to playback tapes of these three types
9
of call. The responses, then, are specific to the nature of
Great titmouse the alarm call and not to the appearance of the particu-
7 lar type of predator (Seyfarth et al. 1980).
Other calls vary according to context rather than the
5 species of the predator. For example, California ground
9 squirrels whistle when a predator arrives suddenly and
Blue titmouse
7
there is little time to escape, and give a chatter-chat call
when predators are at a distance (Owings and Hennesy
5 1984; for other examples see Seyfarth and Cheney 2003).
9 We’ll talk more about the evolution of alarm calls in
Chaffinch
Chapter 19.
7

5 Alarms That Cause Animals to Assemble

0 Time, seconds 10
Other types of alarm signals cause those who hear them
FIGURE 16.20 Sonograms of alarm calls of several to congregate into a group in order to defend a resource
species of small song birds. Note the similarity of the or drive off predators (Bradbury and Vehrencamp 1998).
calls. As a result, an alarm call will alert birds of many As you might expect, compared to signals that cause oth-
species. (Modified from Marler 1959.)
ers to flee, assembly signals generally need to be easier
to localize (so conspecifics can find the signaler) and need
lengths centered on 8 kHz (Figure 16.20). As mentioned to be longer-lasting. Often they are repetitive. You may,
in Chapter 13, some species even respond to the signals for example, have heard a flock of crows cawing loudly
of other species: Eurasion red squirrels flee or increase and repeatedly as they mob a hawk.
their vigilance when they hear the alarm calls of Responses to assembly alarms can be behaviorally
Eurasian jays (Randler 2006). quite complex. Consider the general pattern of the
Most species use the same signal to indicate any response of ants to alarm pheromones (summarized in
source of danger, although some use specific calls to Yamagata et al. 2007). First they freeze, then they raise
designate the type of threat. For example, vervet their heads and wave their antennae. They next move
monkeys (Cercopithecus aethiops) classify their most com- toward the source of the pheromone and perhaps release
mon predators into one of at least three groups—snakes pheromone themselves. Finally, they begin biting the
(e.g., pythons), mammals (e.g., leopards), or birds potential enemy. (If you have ever been bitten by fire ants,
(e.g., eagles). Considering the hunting strategy of each you can appreciate how effective a defense this is.) This
type of predator, the characteristics of the alarm call and complicated response is mediated by numerous neurons
the response of conspecifics within hearing distance that are sensitive to alarm pheromone components.
seem to be adaptive. The low-amplitude alarm call emit-
ted when a snake is encountered captures the attention
AGGREGATION
of individuals near the caller that might be in danger
from the slow-moving reptile without attracting other Besides assembling in response to alarm signals, animals
predators in the area. Other monkeys respond by look- often aggregate for other reasons—for example, to hiber-
ing at the ground, the most likely place to find a snake. nate, to share a resting place or a roost, or to prepare for
However, when they see a large mammalian predator migration. To select a particularly unpleasant example,
such as a leopard, monkeys emit very loud, low-pitched, consider bedbugs (Cimex lectularis). Bedbugs are house-
and abrupt chirps. These properties make the call audi- hold pests that come out from hiding places, such as in
ble from a great distance and make it easy for con- cracks in a bedframe, in the middle of the night and bite
specifics to locate the caller. The most common response sleeping humans for a blood meal. They have recently
to the chirp is to scatter and run for cover in the trees, been in the news because there has been a resurgence in
a relatively safe haven from the ambush style of attack bedbugs, even in expensive hotels. Bugs may benefit
characteristic of a leopard. When monkeys spot an avian from aggregating for a number of reasons, including
374 Chapter 16 / Communication: Channels and Functions

decreased sensitivity to desiccation, protection from some communicatory role (Munz 2005). Von Frisch was
predators, and ease of finding mates. To find one an ethologist who had worked on several different
another, bedbugs release an aggregation pheromone species. (Munz reports that one of von Frisch’s earlier
(Siljander et al. 2008). Perhaps humans will be able to papers had the intriguing title of “A Catfish that Comes
exploit this pheromone to control bedbugs. When Summoned by Whistling,” in which he described
training a blinded catfish to associate whistling with a
food reward, thereby settling an argument over whether
AGONISTIC ENCOUNTERS catfish can hear airborne sound.) Von Frisch’s study of
the dance language of honeybees spanned 50 years. The
Animals can be in conflict with conspecifics over food,
remarkable story that unfolded from his research even-
territory, mates, and their places in the dominance hier-
tually earned him the Nobel Prize in 1973 (see Chapter
archy. Agonistic behaviors are the actions involved in
2). Research on bee communication has been, at times,
conflict, including both aggressive behaviors, such as
fraught with controversy, and there is still much to learn,
threats and attacks, and submissive behaviors, such as
but it is inextricably linked with von Frisch’s name.
appeasement or avoidance. Familiar examples include
In his research, von Frisch used very simple tech-
bighorn sheep butting heads, cats hissing, and dogs
niques, many of which are still employed today. One
rolling on their backs to offer their vulnerable bellies to
technique was to put out feeders of sugar water that bees
a stronger rival. We will talk about these behaviors
readily visit. Another was to individually mark bees by
extensively in Chapters 17 and 18.
painstakingly applying unique patterns of paint (these
days you can buy premade numbered tags, ready to glue
on your bee). Glass-walled observation hives allowed
COMMUNICATION ABOUT
him to observe bee behavior inside the hive. With these
RESOURCES: A CASE STUDY
techniques, von Frisch could follow individuals as they
Group-living animals sometimes communicate the loca- foraged at the feeders, watch them dance in the hive, and
tion of food and other resources to one another. For then see whether marked bees who attended a dancing
example, some animals are central-place foragers, leav- bee then visited the appropriate feeder.
ing from a shared nest to collect food. In this section, we
describe one remarkable and well-studied communica-
tion system, that of the honeybee (Apis mellifera). Two Types of Dances
Honeybees live in large colonies in carefully con- Von Frisch classified the dances he saw into two cate-
structed hives. We explore honeybee sociality in more gories. The round dance is a circling dance, just as the
detail in Chapters 15 and 19; for now, what is most rel- name implies: the bee runs in a circle, then reverses
evant is that all the labor required to maintain a colony direction and circles again (Figure 16.21a). The waggle
is done by female worker bees. Workers divide up the dance is in the shape of a figure 8. The bee runs in a
tasks of caring for the hives and brood. Older bees are straight line through the center, circles to the right, runs
foragers and search for flowers that provide nectar and straight through the center again, and then circles to the
pollen. Depending on the neighboring habitat, flowers left (Figure 16.21b). During the central run, called the
can be widely scattered, and their distribution changes “waggle run,” the bee waggles her abdomen about 15
rapidly over time as different species bloom and fade. It times per second and buzzes her wings at about 250 beats
may take a few days for a newly opened flower to be dis- per second.
covered. However, once a bee finds the resource, many Von Frisch originally suggested that these different
additional recruits soon appear. Scientists have known types of dances indicate different food sources: he
for a very long time that the initial discoverer does not noticed that bees performed a round dance after visiting
simply lead the others to the food, because new bees are a feeder and a waggle dance after visiting a flower.
recruited even if the discoverer is captured as she leaves However, he soon realized a potential confound in his
the hive on a return trip (Maeterlinck 1901). Thus, suc- experiment. He had initially placed all of his feeders
cessful foragers must be communicating something close to the hive for convenience’s sake, and flowers were
about the new food source to other bees. further away. Instead of indicating the type of the food
In the darkness of a hive, returning foragers do a source, he wondered whether instead the two types of
very obvious and characteristic dance, which we describe dances indicated distance information. Let’s examine
in detail below. Other bees very attentively follow the some of his experiments.
dancers, touching them with their antennae and heads. Here is one of the protocols that von Frisch used
Although many naturalists had noted these curious (and that is often used today). First, bees are trained to
dances, their function was unknown. In the late 1910s, come to a feeding station by setting out a weak,
a scientist named Karl von Frisch was captivated by the unscented sugar solution. Because the solution is weak,
mystery of the dances and began to suspect that it had the foragers do not dance upon their return to the hive.
 Functions of Communication 375

a a b

Number of bees arriving at scent plates

10m 100m

278

27 37 20 19 6 2 1
N E S W N E S W
b Direction of scent plates Direction of scent plates

FIGURE 16.22 Number of bees arriving at scent plates

positioned to the north, south, east, and west of the hive
when the training station was (a) 10 m to the east of the
hive and (b) 100 m to the north of the hive. When the
food is close to the hive, the recruits search randomly
within a certain radius. When the food source is distant,
they search primarily in the direction of the training
station. (Modified from von Frisch 1971.)

the scout returned to the hive and did a round dance.

While the hive was closed, the feeding station was emp-
tied and scent plates were positioned to the north, south,
east, and west. As you can see in Figure 16.22a, the
FIGURE 16.21 (a) The round dance of a honeybee, per- recruits arrived at each scent plate in almost equal num-
formed after finding food near the hive. The bee circles bers. Apparently, they did not know the direction to the
alternately to the left and right, while recruits follow her. nectar. This result suggests that the round dance does
(b) The waggle dance, performed after finding food not convey direction information, but only tells recruits
greater than about 50 m from the hive. During the to “search for nearby food.” It also suggests that bees
waggle run through the center of the figure 8, the may pick up the lavender scent as a cue. We return to
bee waggles her abdomen and buzzes her wings. these points with updated information below.
When von Frisch moved the same feeding station
100 m north of the hive, the returning discoverer did a
On the test day, a stronger sucrose solution, with a scent waggle dance. This time almost all the bees arrived at the
such as lavender oil, is set out at the feeding station. northern scent plate, indicating that they had direction
Foragers feeding on this solution return to the hive and information (Figure 16.22b). The accuracy of the direc-
dance. During the dancing, the hive is closed to prevent tion information was further demonstrated in fan exper-
bees from exiting. Meanwhile, food is removed from the iments. Following his usual procedures, von Frisch
feeding station, and scent plates containing lavender trained bees to a feeding station and then placed scent
fragrance but no food reward are set out. Different plates so that they were arrayed like a fan. Each plate was
arrangements of these plates are used to test different 550 m from the hive and separated from its neighbors
hypotheses. When the hive is open, recruits leave and by 150 m. The overwhelming majority of recruits
search for the food. The number of bees arriving at each appeared at the scent plate nearest to the original feed-
scent plate indicates the area in which the bees are ing station (Figure 16.23). Then von Frisch performed
searching. a step experiment to examine the accuracy of the distance
To test whether the round dance and the waggle information. After training bees to a feeding station,
dance are used to indicate different distances, bees were scent plates were placed in line with the empty feeding
trained to a feeding station 10 m to the east of the hive. station at intervals closer or farther away. Again, the
After sipping the strong scented sucrose on the test day, recruits appeared at the scent plates closest to the
376 Chapter 16 / Communication: Channels and Functions

0 hive, and the flower, it would know which way to go

(Figure 16.25). This angle is exactly the information that
3 dancing bees convey. Honeybee dances are performed on
the vertical comb of the hive, so it’s not possible for a bee
550 m
to simply waggle run in the direction of the food. Instead,
5
Shed the bees use the angle of the run with respect to gravity.
The angle between the waggle run and “up” on the comb
42 is the same as the angle formed between the flower, the
Hive 600 m
Feeding station hive and the azimuth. If a bee needs to fly toward the sun
during training in order to reach the flower, the waggle run is oriented
2 straight up; if the bee must fly directly away from the sun,
the waggle run is oriented straight down. Likewise, a food
source 20º to the right of the sun would be indicated by
1
a waggle run directed 20º to the right of vertical.

FIGURE 16.23 The setup of a fan experiment to deter-

mine whether bees use directional information. The
solid square shows the position of the feeding station a
during training, and squares show the position of the
scent plates, which had no food. After following a
waggle dance, most recruits arrive at the scent plate
nearest the site of the feeding station. The number of c
bees arriving at each station is indicated.
20

original feeding station (Figure 16.24). Thus, the wag-

gle dance not only says “Food is far away,” but also
encodes information about direction and distance. As
further experiments have shown, direction and distance
Hive
are encoded by different aspects of the waggle dance.

How Dances Encode Information

About Direction
Picture a forager poised at the exit of a hive, ready to set
off for a particular flower. One piece of information that b
would help a bee determine its departure direction
involves the sun. If a bee knows the angle formed by the
sun’s azimuth (the point on the horizon below it), the a b c 20°

0
Hive 0
24
22
3
100m N
300m TRA 0
CK
500m
0
850m

1200m FIGURE 16.25 Bees dance inside a dark hive on a vertical

comb. During a waggle dance, the dancer indicates the
FIGURE 16.24 The setup of an experiment to determine direction to the food by the orientation of the waggle run
whether bees use distance information. The solid square relative to a vertical line. (a) If the food source is in the
shows the position of the feeding station during training, direction of the sun, the waggle run is oriented straight
and circles show the position of the scent plates. After up. (b) If a bee should fly away from the sun to reach
following a waggle dance, most recruits arrive at the the food, the waggle run is oriented straight down. (c) If
scent plate nearest the site of the feeding station. The the food is 20º to the right of the sun, as in the diagram,
number of bees arriving at each station is indicated. the waggle run is oriented 20º to the right of vertical.
 Functions of Communication 377

One of the challenges with using the sun as a direc- 150

tional cue is that, as the earth rotates, the sun appears to
move across the sky. Bees often dance for a long time in 50
Hive Feeder
the darkness of the hive without going back outside to
update their idea of where the sun lies. However, they
–50

Meters
don’t need to, as the sun’s apparent movement is very
predictable. Dancing bees adjust the angle of their
–150 Release points
dances by approximately 15º an hour to compensate for
the apparent movement of the sun.
How precise are these directions? Not that precise –250
at all. Dancing bees repeat their dances over and over,
and the direction of the waggle run is not exactly the –350
same each time. It appears that observing bees take an –350 –250 –150 –50 50 150 250 350
average of the runs they’ve followed in order to select a Meters
direction in which to fly (Tanner and Visscher 2008). FIGURE 16.26 The flight paths of bees that had attended
The directions encoded in the dance get bees to the gen- a waggle dance, mapped by harmonic radar. The hive is
eral area of the resource, and from there they use scent at the center point of a large field. Bees recruited by a
to pinpoint the location of the flower. dance generally went straight to the feeder from the
hive. Bees captured as they were leaving the hive and
moved to a new release point 250 to 300 m away flew
How Dances Encode Information to the location where they expected the feeder to be.
About Distance (From Riley et al. 2005.)
Distance to the food source is correlated with two other
features of the waggle run. The more waggles, the searched where the dancer had led them to expect the
greater the distance to the food (von Frisch 1967). The feeder to be relative to their release point (Figure 16.26).
distance to the food is also correlated with the duration It has even been possible to “talk” to bees in their
of buzzing during the straight run. As the distance to the own language by using a mechanical model of a danc-
food increases, the sound trains of buzzes are longer ing bee (Figure 16.27). A computer controlled the
(Wenner 1964). model’s movements and sound production. The danc-
A recent analysis of waggle and round dances sug- ing model bee fed recruits scented sugar water through
gests that perhaps these two types of dances are not as a syringe, mimicking the manner in which a real bee
discrete as they first appeared to be (Gardner et al. 2008). would feed the recruits upon returning to the hive.
Round dances often contain very brief waggle phases
with acoustic signals that contain distance information.
In addition, round dances also can contain directional
information. However, there is a large amount of “noise”
in round dances, thus making these dances less precise
than waggle dances.

New Technology Offers Additional Insights

Most studies of bee navigation rely on simply seeing
which bees show up at particular feeders. New technol-
ogy has allowed researchers to map the paths of recruits.
Riley et al. (2005) attached tiny harmonic transponders,
weighing only a few milligrams, to bees. Unlike much
heavier radio collars that are used in tracking larger ani-
mals, these passive transponders return signals to a radar. FIGURE 16.27 A mechanical bee that can communicate to
live bees is constructed of brass covered in beeswax. On
By positioning the radar at the edge of a large mowed
its back is a single wing made from part of a razor blade.
field, researchers could accurately map the paths of indi-
An electromagnet causes the wing to vibrate so that it
vidual bees. Bees that had followed a dance generally mimics the sound patterns produced by real dancing bees.
went straight to the feeder. Bees did not use odor cues A tiny plastic tube connected to a syringe releases
to find the feeder, as no scented foods were used, and the droplets of scented sugar water, simulating regurgitation
prevailing breeze came from a different direction. If bees by live bee scouts. Computer software choreographed the
leaving the hive were captured and displaced, they did robot’s dance so that it directed the bees to a food source
not fly in the true direction of the feeder, but rather in a location chosen by the experimenters.
378 Chapter 16 / Communication: Channels and Functions

The model danced, indicating a food source 250 m to How Do Bees Measure Distance?
the south of the hive. Scent plates without food were In order to communicate information about distance, bees
placed 250 m to the north, south, southeast, and south- must have some way of measuring it. There are a number
west of the hive. Although the model’s dance was not of different possibilities, such as keeping track of the effort
as effective as a live bee’s dance in recruiting others to it takes to fly, or by keeping track of the landscape going
look for food or in accurately directing them to the by. In an elegant experiment, Srinivasan et al. (2000) dis-
correct location, most of the recruits showed up at the covered that foragers rely on visual cues to estimate how
feeding station indicated by the dance. Two critical far they have traveled. These researchers trained bees to
components of the dance are the wagging movements fly into tunnels (6.4 m long, 11 cm wide, and 20 cm high)
and the buzzing. If either was missing from the model’s in order to reach a feeder. The researchers manipulated
dance, the recruits showed no preference for the the amount of image motion that the bees experienced,
direction indicated (Michelson et al. 1989). also called the optic flow, by manipulating the patterns on
the inside walls of the tunnels (Figure 16.28a). Some tun-
nels had horizontal stripes lining the inside, parallel to the
bees’ movement. This design provided very little infor-
a mation to the bees about the distance they traveled. Other
Hive
tunnels had vertical stripes, which provide the bees with
the perception of a great deal of image motion, especially
6m in the narrow confines of the tunnel. The tunnels were
placed outdoors (Figure 16.28a), either 6 m or 35 m from
35 m
4
the hive, so that part of the bees’ journeys to the feeders
was through the tunnel.
1 In order to determine how the bees perceived direc-
2
tion, the researchers took advantage of the two types of
3
6m dances. The round dance is generally given when
resources are within 50 m of the hive, and waggle dances
b 1.0
W for longer distances. All the tunnels were well within the
R W
Probability of waggle (W)

R
50-m mark. Researchers watched how bees danced when
or round (R) dance

0.8
they returned to the hive after visiting a particular feeder.
0.6 The results are shown in Figure 16.28b: When bees vis-
ited a feeder at the opening of a tunnel, or visited a
0.4 feeder with horizontal stripes, they performed a round
dance. However, when bees flew through a tunnel with
0.2 W W R vertical stripes, they acted as if they had flown a long way
R
and performed a waggle dance, even when the tunnel
0.0 was only 6 m from the hive. By analyzing the duration
1 2 3 4
Experiment of the waggle runs, the researchers could determine by
how much bees were overestimating the distance they
FIGURE 16.28 (a) Layout for four experiments using tun-
nels with visual cues. Tunnels 1, 2, and 4 had vertical
flew. The tunnel increased the dancers’ perception of the
stripes that provide the bee with the perception of a distance they had flown by a factor of 31! (Srinivasan
great deal of image motion. Tunnel 3 had horizontal 2000). The bees that were recruited by foragers confused
stripes that provided little perception of motion. The by vertically striped tunnels searched for the food well
small circles inside the tunnels indicated the placement beyond the tunnel, rather than looking inside the tun-
of the feeders. (b) The probability that bees did waggle nel (Esch et al. 2001).
(W) dances, indicating that they perceived the feeder to Thus, foraging bees learn about distance by using
be far away, or round (R) dances, indicating that they optic flow, and they convey that information to other
perceived the feeder to be nearby, for each experiment. bees. Normally this system works, as food sources are
In experiment 1, the feeder was placed near the tunnel not hidden inside tunnels, and recruits are likely to per-
entrance, and bees perceived it to be close by. In experi-
ceive the same optic flow patterns as a dancing bee,
ment 2, the feeder was placed at the far end of a tunnel
with vertical stripes, and bees perceived it to be far away.
given that the direction information is accurate. For
In experiment 3, the feeder was placed at the far end of example, a bee will perceive equally long flights over
a tunnel with horizontal stripes, and the bees perceived water or through a forest as very different distances, but
it to be close by. In experiment 4, the tunnel with verti- if the recruits also take the same path, they will suc-
cal stripes was placed close to the hive, and bees per- cessfully find the food (reviewed in Dacke and
ceived it to be far away. (From Srinivasan et al. 2000.) Srinivasan 2007).
 Summary 379

Other Cues Used in Foraging Auditory signals can be transmitted over long dis-
As alluded to earlier, whether waggle dances truly con- tances. The rate of transmission and fade-out is rapid.
vey information has been the subject of controversy They do not require light and, in fact, work well under
(reviewed in Munz 2005). Adrian Wenner (e.g., Wenner water. The sounds may be generated by respiratory
2002 and references therein) has been the major advo- structures, the rubbing of appendages, or beating on
cate of the idea that bees rely on odor cues and that the parts of the environment. Some animals produce sounds
dance language is not used by recruits to gain informa- above (ultrasound) or below (infrasound) the limits of
tion. Although most bee researchers accept the validity human hearing.
of the waggle dance as a communicatory system, Seismic signals are those caused by vibration of the
Wenner’s work and that of others helped focus attention environment. They are well suited to communication
on other cues used by bees, particularly odor cues. over long distances. Seismic signals can be produced
The waggle dance directs bees to a particular area, when an airborne vocalization couples with the ground.
but not that precisely, as we have discussed. The loca- Chemicals can convey messages over great distances,
tion of the flower is pinpointed by odor cues. Dance particularly when assisted by currents of air or water.
followers can detect food scents on the dancers and They are transmitted more slowly, are more durable, and
even contact dancers in the appropriate place: hive- are usually more difficult to locate than visual or audi-
mates that foraged for pollen were more likely to con- tory signals. Furthermore, they are effective in environ-
tact the dancers’ legs (where pollen is stored in ments with limited visibility. Chemicals used to convey
“baskets”), whereas dancers that had collected sugar information to conspecifics are called pheromones.
solutions were more likely to have head-to-head con- When the chemical has an immediate effect on the
tacts (Díaz et al. 2007). behavior of the recipient, as occurs with sex attractants
Besides the chemicals they pick up from food, danc- and trail and alarm substances, it is called a releaser
ing foragers also produce other chemicals from their pheromone. Primer pheromones act slowly, exerting
abdomens. By comparing air samples taken from danc- their effect by altering the physiology of the recipient.
ing and nondancing bees, and analyzing them with gas Some species of amphibians, reptiles, and mammals have
chromatography and mass spectrometry, Thom et al. a vomeronasal organ, used (in addition to the olfactory
(2007) identified four chemicals released by the dancers. epithelium) to sense chemical cues.
When the researchers injected these chemicals into the Tactile signals are effective only over short distances,
air inside the hive, more foragers exited the hive. Thus, and it is easy to locate the sender. These signals are
it seems that these chemicals cause the bees to become rapidly transmitted. Allogrooming, a special form of
primed to look for food. tactile communication, is practiced by many species but
is especially prominent among primates.
Electrical fields are used for communication among
the mormyriform fishes of Africa and the gymnotiform
SUMMARY fishes of South America. Transmission and fade-out are
Different definitions of communication have been used in almost instantaneous, but the signals do not travel far.
the literature. Here, we will use the definition that com- They are effective when visibility is limited.
munication occurs when information is transferred Many animals communicate using displays that con-
from the sender to the receiver and that the sender ben- tain signals from more than one sensory channel; this is
efits, on average. A behavior, such as a call, or a feature, called multimodal communication. The messages sent
such as a colorful feathered crest, that transmits infor- by different channels can be redundant or nonredundant.
mation is called a signal. Signals are thus distinct from Communication has many functions, many of which
cues, which may provide information to another animal we explore in more detail in other chapters. Many animals
but do not benefit the sender, such as the rustle of leaves use communicatory signals to differentiate conspecifics
that attracts an owl to a mouse. from other animals. This distinction is clearly valuable, as
Any sensory channel may be used for communication. conspecifics are likely to be competitors or potential
Signals within each channel have characteristic properties mates. Virtually all sensory channels are used for species
that make them more or less useful, depending on the identification in one species or another. In regions where
species, the environment, and the function of the signal. two closely related species overlap, we sometimes see
Visual signals are easy to locate, are transmitted greater differentiation between the signals of the two
quickly, and disappear just as fast. However, visual sig- species than in regions where there is no overlap. This
nals must be seen and are, therefore, useful only when suggests that selection against making a mistake is partic-
there is enough light and where there are few obstacles ularly strong. In other examples, such as in invasive species
to obscure them. Visual signals include color and pattern, that do not have a long evolutionary history with native
as well as movements and postures. species, we see many errors in species identification.
380 Chapter 16 / Communication: Channels and Functions

Communication is used throughout the reproductive Other alarm signals cause animals to assemble in
process. First, many solitary animals have the challenge order to defend a resource or drive off predators. This
of locating potential mates and use communicatory sig- type of alarm signal tends to be easy to localize and
nals to attract them. Because chemical and auditory chan- longer-lasting than the signal that causes receivers
nels work particularly well over long ranges, they are to flee.
often (but not always) used for mate attraction. Second, Animals must aggregate for other reasons besides
animals might signal their identity, both their species and alarm, such as to hibernate, roost, or prepare for migra-
sex. Third, animals might advertise their qualities as a tion. Many species use aggregation pheromones in order
mate, as discussed in Chapters 14 and 17. Fourth, animals to attract conspecifics to a particular spot.
might use displays to coordinate their physiology, as illus- Animals are often in conflict with conspecifics over
trated by the well-studied case of the ring dove. Finally, food, territory, mates, and their places in the domi-
some animals communicate in order to maintain their nance hierarchy. Many signals have evolved in the con-
pair bond, even when it is not mating season. text of agonistic behavior. We will delay discussion of
Group-living animals maintain social bonds as well, these to Chapters 17 and 18, where they are treated
generally through touch. Many animals lie or sit in extensively.
contact with each other or have greeting ceremonies. Animals also communicate about the location of
Many species allogroom, or groom one another, which food resources. We end the chapter with a detailed look
has the benefit of improving the health of skin, fur, or at a specific case study, the dance of the honeybee.
feathers, and also cements social bonds. Honeybee scouts return to the hive and communicate
Alarm signals warn others of danger. Some alarm the direction of food with either a round dance (for
signals cause receivers to flee. These signals are gener- nearby resources) or a waggle dance (for distant
ally quick to produce and difficult to locate, which ben- resources). The waggle dance encodes information
efits the signaler that is trying not to attract the attention about the direction of the resource (via the angle of the
of a predator. Alarm signals tend to be similar across “waggle run” section of the dance) and the distance of
species, and some animals respond to alarm calls of other the resource (via the duration of the “waggle run” and
species. Some alarm signals specify the nature of the dan- its accompanying buzzing). Bees measure information
ger, such as a predator approaching on the ground ver- about distance via the amount of optic flow they perceive
sus in the air. as they fly.
 17
The Evolution of Communication

The Changing Views of Communication tionary costs and benefits of the signal, as well as the
Sharing Information morphological and physiological features that influence
Manipulating Others the signal’s production.
We will begin by examining in more detail some-
Signals and Honesty
thing that we alluded to in the last chapter—the chang-
When Are Honest Signals Likely?
ing way that animal behaviorists have viewed
When Are Dishonest Signals Likely? communication. We’ll consider when we expect com-
Can Honest and Dishonest Signals Coexist? munication to be honest, and when we expect it to be
The Evolutionary Origins of Signals unreliable. Then we will discuss two hypotheses about
Ritualization the evolutionary origin of signals. Next we will look at
Receiver-Bias Mechanisms some of the selective forces that influence signal design.
Selective Forces That Shape Signals We will examine language—what it is and whether it is
uniquely human. Finally, we will discuss what the study
Characteristics of the Sender
of communication can tell us about animal cognition.
Characteristics of the Environment
Characteristics of the Receiver
Language and Apes
What Is Language? THE CHANGING VIEWS
Ape Language Studies OF COMMUNICATION
Communication and Animal Cognition
SHARING INFORMATION
In the last chapter, we described how communication
As we saw in the last chapter, we can often answer the can be broadly defined as the transmission of informa-
who, what, where, and when of the communication tion from one animal to others. Many researchers
process, but more difficult are questions of how and why through the 1970s emphasized a cooperative view of
have signals taken the various forms that they have. To communication, where both sender and receiver ben-
answer these questions, we must consider the evolu- efit from the accurate transfer of information (reviewed

381
382 Chapter 17 / The Evolution of Communication

in Dawkins and Krebs 1978). Under this scenario, signals, even if it is sometimes deceived. After all, not all
selection should act to make signals efficient, reliable, signals are dishonest. Often a signaler benefits from con-
and unambiguous. Dishonest, or inaccurate, signaling veying accurate information, such as in the case of the
was thought to be unlikely in animals; after all, even in female moth alerting males that she is ready to mate.
humans, lying is socially risky and hard to do convinc- Thus, signals are normally reliable but occasionally
ingly (Smith 1977). deceptive. In the next section, we will examine the cir-
The idea that communication benefits both partners cumstances that favor honest signaling and those that
does seem reasonable at first glance. Consider, for allow for deception.
instance, the complex role of communication in coordi-
nating mating activities. First, potential partners must
locate each other. As described in Chapter 16, many SIGNALS AND HONESTY
species do this via species-specific pheromone signals. It
is clear that females are adapted to broadcast news of WHEN ARE HONEST SIGNALS LIKELY?
their availability, while males are adapted to detect the
We will review four common circumstances under which
particular chemical signal of prospective mates. After
we expect to see honest signals: (1) when senders and
potential partners find each other, they may engage in a
complex give-and-take of displays that culminate in mat- receivers share overlapping goals, (2) when signals indi-
ing. Surely, both partners receive fitness benefits by cate something about the sender that cannot be faked,
clearly communicating their intentions to each other. (3) when signals are costly to produce, and (4) when dis-
Thus, many familiar examples of communication—and honest signalers can be identified.
indeed much of our own experience—seem to uphold
the view that its function is to share information. If Senders and Receivers Share
Overlapping Goals
Sometimes both the sender and receiver share the same
MANIPULATING OTHERS
goal: the sender benefits by accurately transmitting
This cooperative view of communication does not fit information, and the receiver benefits by accurately
every situation, however. There are some situations in interpreting it. The coordination of behaviors between
which the sender might gain by sending an inaccurate animals that have chosen to mate provides an example.
signal. In a territorial dispute, for example, the sender In cases like these, we expect that natural selection will
might bluff by sending signals that exaggerate its will- favor unambiguous, honest signals.
ingness to escalate the contest. Similarly, a male com- When can we expect to see overlapping goals? One
peting for a female’s attention might increase his mating obvious place to look is the relationship between parent
success by exaggerating his qualities. In these and many and offspring, although as described in Chapter 15, even
other cases, an animal that gave a dishonest signal would this relationship can entail significant conflict.
gain an advantage over animals that honestly communi- Nevertheless, if you’ve been around infants, you will
cated about their abilities or intentions. In fact, we expect have no trouble believing that they have been selected
deceptive signals will evolve on a regular basis. (Note to convey their needs for food and comfort to their par-
that the terms honesty, dishonesty, and deceptive imply ents. Of course, human parents are not the only ones at
nothing about the cognitive abilities or the intentions of the beck and call of their offspring. The human infant’s
the sender, but are generally used in the literature as need for near-constant care is rivaled by altricial baby
descriptive shorthand terms.) birds. Naked and helpless, one of their few behavioral
How can we incorporate the possibility of deceptive options is to beg their parents to feed them, and they do
signals into our understanding of communication? a lot of this. (Beware the undergraduate research assis-
Richard Dawkins and John Krebs (1978) suggest that tant position that entails raising baby birds!) Nestlings
animals communicate not to convey information, but to beg with a hugely gaping mouth. As their begging inten-
manipulate the behavior of others to their own advan- sifies, they add vocalizations. Begging calls are them-
tage. Thus, an animal will produce a signal when, on selves graded in intensity, and can be emitted with great
average, it results in the increase of its own reproductive persistence and at a noise level that is hard to ignore. At
success by influencing the behavior of others. We the highest level of intensity, baby birds stand, gape, flap
emphasize “on average” because not every signal an ani- their wings, and call all at once (Figure 17.1).
mal gives will necessarily be to its benefit. But for sig- Experiments have shown that in many bird species, the
nals to evolve, they must benefit senders overall. vigor of begging increases with the degree of hunger
In this view, senders may not be trustworthy. So why (e.g., Redondo and Castro 1992). Begging chicks are
would a receiver ever respond to a signal? If signals are quite obvious to predators as well as parents, so begging
potentially dishonest, why not just ignore them? On too much, when food is not needed, is a poor strategy
average, the receiver must benefit from responding to (Haskell 1994). Parents usually respond to the signal by
 Signals and Honesty 383

FIGURE 17.1 Levels of begging display in

birds. (Modified from Redondo and
Castro 1992.)

providing food to those chicks that beg most vigorously each other head-to-head, with their forelegs spread
(reviewed in Searcy and Nowicki 2005). Parents are outward and parallel to the eyestalks (Figure 17.2). This
under selection to respond because their own fitness pose allows each competitor to compare the length of
depends on the survival of their offspring. Thus, in its eyestalks to that of its rival. Eyestalk length increases
general, communication between baby birds and their with body size, and males with shorter eyestalks usually
parents appears to be honest. retreat without a fight (Burkhardt and de la Motte 1983;
Many other examples exist in which the goals of the de la Motte and Burkhardt 1983). Thus, eyestalk length
sender and receiver of a signal are aligned. For example, is an accurate index of size that cannot be faked.
as we will describe in more detail in Chapter 19, Belding’s Other signals are honest because they are linked very
ground squirrels call to warn relatives of the presence of strongly to an animal’s health and physiological well-
a predator. Because genes are shared between relatives, being. For example, the bright reds and yellows of feath-
both the sender and receiver of the call reap a fitness ben- ers, scales, or fleshy necks or combs of some birds depend
efit. In other species, members of a group share a com- on chemicals called carotenoids. These chemicals cannot
mon goal: every bee benefits when scouts convey be synthesized by vertebrates but must be obtained in the
accurately the location of a patch of flowers in full bloom. diet. This makes these bright colors a good candidate as
In such cases, honest communication is likely. an honest signal of foraging ability and health. In a series
of studies, Geoff Hill examined this signal in house
finches, common backyard visitors throughout the United
If Signals Cannot Be Faked
States (Hill 1990, 1991; Hill and Montgomerie 1994).
Sometimes signals are honest because they are tightly Female house finches are a rather bland, streaky gray, but
linked to a trait of the sender. Although the sender might the head, throat, and breast of male house finches varies
benefit if it could lie, it simply is not possible to fake the from pale yellow to bright red within a population. With
signal. For instance, size is usually a good predictor of controlled feeding experiments, Hill determined that the
fighting success, and many displays allow opponents to brightness of the color depends on carotenoids in the diet.
judge one another’s size. In some species, combatants can Females prefer the brighter, redder males. The females
enhance their apparent size by puffing up their feathers, benefit from their choice: brighter males are better par-
fluffing out their fur, or assuming an upright posture, but ents, bringing more food to the nestlings.
in other species, size is not so easily faked. In the threat
display of male stalk-eyed flies (Diopsidae), flies face
If Signals Are Costly to Produce
Other signals seem to be more plastic or labile in their
expression. For example, plumage colors may be bright
or dull, and courtship dances may be given vigorously or
lethargically. When signals are labile, and the sender and
receiver do not share the same goals, the situation is ripe
for deceit. For example, imagine a male attempting to
convince a female that he, above all others, is the right
one for her. It is to the male’s advantage to convince the
female to mate; however, it is to the female’s advantage
to accurately assess the quality of as many males as fea-
sible in order to pick the best father for her offspring.
What then is to prevent the male from exaggerating his
FIGURE 17.2 An aggressive display of male stalk-eyed
qualities? How would selection favor honesty?
flies. In this pose, each male can determine the distance One hypothesis, proposed by Amotz Zahavi, is that
between the eyes of its opponent and thereby the rival’s reliable signals will be favored in a population when sig-
size. Size is correlated with fighting success. Since there nals are costly to the sender (Zahavi 1975, 1977; Zahavi
is little a combatant can do to alter its body size, this is and Zahavi 1997). This is the handicap principle
an example of honest signaling. that you met in Chapter 14. Originally developed for
384 Chapter 17 / The Evolution of Communication

signaling in the context of mate choice, it is also applic- species, the long tail feathers that males sport as sexual
able to other types of communication. The idea is per- ornaments impair their ability to fly (Jennions 1993).
haps best described by an analogy. Imagine a person who Researchers have now accumulated many examples
buys expensive cars and $2000 wristwatches. These are in which higher-quality senders are better able to pay
hardly necessary to the owner: in fact, they divert the cost of producing higher-quality signals (brighter
resources from other things that might be more impor- colors, more vigorous displays) than are lower-quality
tant. However, cars and watches might act as a signal to senders. Among red deer, for example, only the males
others, including prospective mates, that the owner has in top physical condition can continue roaring long
money to burn. Similarly, Zahavi suggested that some enough to win the vocal duel. Roaring uses many of the
signals confer an advantage on their owner, not because same muscles and behavior patterns involved in fight-
they are useful themselves, but because their very extrav- ing, and so it serves as an honest signal of a male’s
agance indicates the owner’s qualities. These signals are fighting ability (Bradbury and Vehrencamp 1998).
referred to as handicaps because their owners are per- Sometimes, a particular quality of the signal rather than
ceived as doing well in spite of the handicap of invest- its repetition is condition dependent. For example,
ing in the signal. Zahavi’s idea sparked lively debate and when a rival encroaches on his territory, a male common
was followed by tests of its plausibility via mathematical loon (Gavia immer) gives a “yodel” (a different call than
models (e.g., Grafen 1990; Johnstone 1997). Eventually, the eerie wail that may be more familiar to you).
several variations of the original formulation were Individuals in better condition (as measured by body
introduced and are now generally regarded as plausible mass adjusted for overall body size) produce lower-
ways by which signals may evolve. frequency (Hz) yodels. Over five years of study, the
For a signal to evolve as a handicap, three criteria yodels of loons that gained mass dropped in frequency,
must be met: the signal must be costly, it must relate to and those that lost mass rose in frequency. Yodels were
the quality of the sender, and the receiver must be inter- artificially changed in frequency and played back, and
ested in the quality of the sender that is being signaled, lower frequency yodels were greeted by livelier
and thus must benefit from attending to an honest sig- responses from other loons (Mager et al. 2007).
nal. We’ve already seen a number of examples of how A particularly charming example of honest signaling
receivers benefit from correctly assessing the quality of is in the blue-footed booby (Sula nebouxii), seabirds with
a signaler, especially in the context of sexual selection brightly colored feet. As part of their courtship ritual,
(see Chapter 14), so let’s look at some examples of sig- boobies stand facing one another as the male lifts his feet
nals that have been shown to meet the first two criteria. in alternation to display them for the female’s inspection
Signals may be costly in several ways, but energetic (Figure 17.3a). Males in good condition have bright
costs are the most commonly measured. A way to moni- blue-green feet, but when held without food for 48 hours
tor energy use is by measuring the amount of oxygen con- (which routinely happens to boobies in nature), their feet
sumed during an activity. In insects and frogs, oxygen quickly fade to dull blue (Figure 17.3b, c). Females
consumption increases 5 to 30 times when they call (Ryan respond to this rapid change in signal, preferring males
1988). Male red deer lose weight during the rut (mating with brightly colored feet. In fact, in one research study
season) because they constantly roar at one another as they females adjusted their own investment in offspring based
compete for mates (Clutton-Brock et al. 1982). Another on this signal: they laid smaller second eggs when
type of cost is a decrease in the ability to move about, researchers used makeup to color their mates’ feet a dull
escape predators, or forage. For instance, in several bird blue (Velando et al. 2006).

a b c

FIGURE 17.3 (a) Blue-footed boobies lift their feet for

examination by their partner during a courtship dance.
Foot color changes with recent feeding, as shown by the
feet of (b) food-deprived and (c) supplemented boobies.
 Signals and Honesty 385

The penguin encounters invariably begin with low-

risk displays and escalate until one opponent retreats or
a fight occurs. The process is somewhat analogous to
human actions at an auction—the bids begin low and
gradually increase until bidders unwilling to pay the
price drop out of the process. The price that little blue
penguins must pay is the risk of injury. As a territorial
contest escalates and the price of the property increases,
one “bidder” usually decides that the territory is not
worth that great a risk. For these penguins, the signals
remain honest because they are potentially costly, and
the cost for a given level of signal is greater for weaker
FIGURE 17.4 These eye and head injuries in a little blue than for stronger individuals.
penguin resulted from fights. Little blue penguins have
an extensive repertoire of aggressive displays, which If Dishonest Signalers Can Be Identified
vary in cost (the risk of injury). By its choice of display,
a penguin indicates the extent of its willingness to fight. A stable social unit also favors honest communication.
Signals remain honest because they are costly. One reason to expect honesty is that individuals will both
send and receive signals at different times. The advan-
Costs of signals may also come in the form of the tages of sending dishonest signals will be reversed when
receivers’ response to the signal. Let’s examine a situation the animal is the receiver. Therefore, the advantages of
that illustrates the dynamic relationship between the cost receiving honest signals might outweigh the advantages
of a signal, the quality of the sender, and honesty. Little of sending dishonest ones, and honesty might come to
blue penguins (Eudyptula minor) have a repertoire of predominate in the population.
aggressive displays that differ in cost (risk of injury), effec- Members of social units may be able to recognize
tiveness in deterring an opponent, and ability to predict one another and remember previous interactions, and
an attack. Although two displays may involve the same rit- thus learn whether a particular individual is honest.
ualized posture, one that involves moving within the rival’s Working with vervet monkeys in Kenya’s Amboseli
striking distance is riskier to perform than one that is per- National Park, Dorothy Cheney and Robert Seyfarth
formed while stationary and out of the opponent’s reach. (1988) tested the idea that members of a social group
By its choice of display, a penguin conveys information would stop believing an individual that routinely gave
about both its willingness to sustain injury while per- unreliable signals. Vervet monkeys utter different vocal-
forming the display and its willingness to fight. It chooses izations during different situations. They have two
a display with costs that represent the value it places on acoustically different calls that warn of the encroachment
the resource. Roughly 10% of penguin interactions are of another group of monkeys on their territory. One call,
not settled by displays and end in fighting. During bat- the “wrr,” is given when another group is spotted in the
tles, losing penguins commonly suffer flesh wounds and distance, perhaps as far away as 200 m. The other, a
sometimes eye loss (Figure 17.4). These injuries could “chutter,” is given if the other group comes so close that
make it more difficult to obtain sufficient food or to breed there are threats, chases, or actual contact (Figure 17.5).
successfully, or the wounds might become infected and Vervet monkeys also have a stable social group in which
cause illness or death. Thus, attempting to intimidate an the animals recognize one another, including their calls.
opponent into retreat by bluffing a strong motivation to What would happen if one individual falsely signaled the
attack could be quite costly, if the rival called the bluff and approach of another group? Would other monkeys
a fight ensued (Waas 1991). believe the “liar” again?

8 8
7 7
Frequency kHz

6 6
FIGURE 17.5 Vervet monkey alarm calls that 5 5
alert group members of the approach of a 4 4
neighboring troop are of two types. The 3 3
“chutter” warns that another group is 2 2
nearby. The “wrr” is given when another 1 1
group is spotted in the distance. The x-axis 0 0
0.5 1.0
indicates time, and the horizontal lines
Time (sec) Time (sec)
show frequency in intervals of 1 kHz.
(From Cheney and Seyfarth 1991.) Chutter Wrr
386 Chapter 17 / The Evolution of Communication

To test this question, Cheney and Seyfarth broad- and shoot forward in a manner similar to that of the
cast tape recordings of the calls of a member of the group praying mantis (Figure 17.6a). The raptorial appendages,
to create the illusion that it was giving inaccurate calls. which are adapted either for spearing or smashing, are
The calls were broadcast from a speaker that was con- used both in prey capture and territorial defense. Mantis
cealed where the group might expect that individual to shrimp are strong enough to easily break a human fin-
be. First, an individual’s “chutter” call was played to see ger or the side of a glass aquarium. Combatants may be
what the baseline response of the others would be. The seriously injured or even killed during the battles over
next day, when no other group was in sight, they played burrows. Readiness to attack is signaled by a threat dis-
that individual’s “wrr” call eight times, at approximately play, called a meral spread, in which the raptorial
20-minute intervals. They found that the other monkeys appendages are splayed out (Figure 17.6b) (Caldwell and
gradually stopped responding to that individual’s warn- Dingle 1976; Dingle and Caldwell 1969).
ing that another group was in sight. The monkeys no So far, this seems like a typical case of signaling in
longer believed the liar. However, if another monkey aggressive contests. However, there is a twist: a newly
uttered the “wrr” call, the group still believed the warn- molted stomatopod is virtually defenseless. Arthropods
ing and responded appropriately. So we see that in a sta- molt by pulling their bodies out of a break in their old
ble social group in which individuals are recognized, exoskeleton, and in order to do so their new exoskele-
others may soon learn not to believe a dishonest signaler. tons are very soft and pliable. A newly molted stomato-
pod gives the meral spread display even though it is
unable to back up this signal with force. Its opponent is
WHEN ARE DISHONEST SIGNALS deceived and responds to the signal by retreating. A
LIKELY? newly molted stomatopod can get away with the bluff
Just as we can predict the conditions under which hon- because the receiver might pay dearly if it chose to test
est signals might be likely to occur, we can also predict the honesty of the signal and the signaler was not newly
when we expect to see dishonest signals. molted (Adams and Caldwell 1990). Thus, the signal is
stable because, on average, it is honest.
If Senders and Receivers Have Different Goals
Just as overlapping goals between senders and receivers CAN HONEST AND DISHONEST
favor honesty in signaling, different goals set the stage SIGNALS COEXIST?
for deception. Note, however, that this is not usually a
We see, then, that signals may be honest or dishonest,
black-and-white distinction. For example, a father bird
and often coexist in the same population. Exactly how
and his chick share the general goal of the chick’s sur-
coexistence occurs has been a bit of a puzzle. Consider
vival. However, the chick may want to induce the father
the analysis of animal communication by John Maynard
to give slightly more parental investment than he would
Smith and his colleagues (e.g., Maynard Smith 1974). In
like, perhaps at the cost of his other chicks or his future
Chapter 4, we introduced the idea of an evolutionarily
reproduction. Thus, we expect to see signals between
stable strategy (ESS). An ESS is a behavioral strategy
parents and chicks that are essentially honest, but with
that, when adopted by nearly all members of a popula-
some attempt by the chick at manipulation and deceit.
tion, cannot be beaten by a different strategy. Imagine,
for example, a population in which every animal always
If Signals Are Costly gives an honest signal. Now imagine that a new mutant
to Assess or Challenge appears in the population that gives dishonest signals and
Assessing signals can itself be costly. For example, a that giving dishonest signals offers a fitness benefit. Over
cricket assessing the quality of a singing male must invest time, the strategy of being dishonest will spread through
time in listening and may risk attack by parasites or the population until every sender is dishonest. Receivers,
predators that are attracted to the male’s song. Under in turn, will have greater fitness if they ignore the
conditions such as these, there may be more opportunity senders. Eventually “everyone would be lying and no one
for senders to get away with deceit. listening” (Rowell et al. 2006).
A receiver may pay dearly for challenging the hon- If this is the case, then why do we see populations
esty of a signal. If receivers are unwilling to accept the cost with both honest and dishonest signals? There are sev-
of a test, honest signals may be corrupted by occasional eral ideas. One idea, as described earlier, is that the
dishonest bluffs. For example, stomatopods (Gonodactylus costliness of signals can reinforce honesty. When sig-
bredini), commonly known as mantis shrimps, are marine nals are generally honest, a low level of deceit in a pop-
crustaceans that ferociously defend the burrows and cav- ulation can be stable and is, in fact, often expected to
ities in which they live. Stomatopods have two large evolve. Another idea is that populations are in flux and
forelimbs, called raptorial appendages, that can unfold that at any given point in time we may be seeing the
 The Evolutionary Origins of Signals 387

a
Eyestalk

Antennule
Telson Antenna Spear

Dactyl
Maxillipeds
Raptorial appendage

Hit

FIGURE 17.6 (a) The mantid shrimp, a

stomatopod, has two raptorial appendages
that can be used in prey capture or combat.
(b) The threat display of a stomatopod,
called a meral spread, is a good predictor
of attack. Here the display is seen from
the front.

spread of dishonesty—eventually it will take over the and the receiver. However, these signals did not just
population, and the signal will fade from use. A third spring fully formed into being. How did they originate?
hypothesis for the coexistence of both honest and dis- How does an incipient signal acquire meaning?
honest signals takes into account something we haven’t Behaviorists have taken two different approaches to
explicitly considered before: that any given signal may thinking about the evolutionary origins of signals. The
be perceived by different receivers, some with the same first focuses on identifying the behaviors of the senders
goals as the sender, and some with different goals. For that form the raw material for signals. This, the study of
example, the same signal may be perceived simultane- ritualization, was a focus of those meticulous observers
ously by a prospective mate and by a competitor. For of animal behavior, the ethologists. The second approach
a signaler under these conditions, a mixed signaling focuses on how signals exploit the receiver’s sensory
strategy, sometimes honest and sometimes deceptive, biases, or ability to detect some kinds of information bet-
may be best (Rowell et al. 2006). These explanations ter than others. These two evolutionary pathways are not
are not mutually exclusive, nor are they the only mutually exclusive alternatives to one another: both may
scenarios possible. have played a role in a given system.

RITUALIZATION
THE EVOLUTIONARY ORIGINS
OF SIGNALS The ethologists proposed that many signals get their
start as part of another behavior or as a physiological
So far, we’ve been discussing established communication response, and only later take on a signaling function.
systems: a loon yodels, and another loon swims away. Then, evolution favors modification of the incipient sig-
The signal apparently has meaning for both the sender nal so that it becomes more stereotyped and more
388 Chapter 17 / The Evolution of Communication

unmistakable, thus facilitating the process of communi- take off from this ritualized pose. The movement has
cation. Julian Huxley (1923) called this process ritual- changed during the ritualization process.
ization. The study of ritualization was at its peak with
the early ethologists, and we will illustrate our discussion
Displacement Activities
with classic examples.
We will consider three sources of raw material for Ethologists defined displacement activities as irrelevant
signals: intention movements, displacement activities, actions performed in situations in which an animal has
and autonomic responses. conflicting motivations and is thus indecisive. For
example, when faced with an aggressor, an animal may
have conflicting motivations to fight and to flee, and
Intention Movements may instead preen (groom) itself. Similar to intention
Animals often begin behavior patterns with some char- movements, displacement activities are often incom-
acteristic movements that prepare them for action. plete actions.
Because it is often possible to judge from these activities Courtship is a time of conflicting tendencies.
just what the animal intends to do, they have been named Sexual partners must come together to mate in spite of
intention movements (Heinroth 1910). For example, the aggressive tendencies that often tend to keep them
wolves pull back their lips and bare their teeth before apart. Lorenz (1972) suggested that the mock preening
biting. It’s likely to improve a wolf’s fitness if he correctly of the courtship displays of males of many duck species
interprets the bared teeth of an aggressive rival rather (Figure 17.8), including the familiar mallard, originated
than waiting to feel jaws clamp down. in displacement preening (Lorenz 1972). However, as
Numerous avian displays originated with intention Smith (1977) points out, it can be tricky to decide
movements for flight or walking (Daanje 1950). A bird whether a display is really a result of displacement or
about to take flight goes through a sequence of prepara- whether it has a function that is unclear to the
tory motions that are helpful in achieving takeoff. It will researcher. For example, duck courtship includes other
usually begin by crouching, pointing its beak upward, vigorous displays, such as rearing up and splashing down
raising its tail, and spreading its wings slightly. into the water, which may well ruffle a drake’s feathers.
Components of the takeoff leap have been ritualized Preening during courtship may not be a display at all,
into communicative signals in different species. The but rather strictly functional.
blue-footed booby, along with exhibiting its brightly
colored feet to its prospective mate, also incorporates a Autonomic Responses
posture called “sky pointing” into its courtship dance
(Figure 17.7). Notice that the wings are spread and that The autonomic nervous system regulates many of the
the tip of the beak and the tail point upward. Although basic body functions—digestion, circulatory activities
this display most likely had its origins in flight intention such as heart rate and diameter of the blood vessels, and
movements, it is difficult to see how a bird could ever thermoregulation, to name a few. Many displays seem to
have originated from autonomic functions (Morris
1956). For example, at times of stress or conflict, there
is often a change in the distribution of blood through-

Mallard Garganey

Shelduck Mandarin

FIGURE 17.8 Mock preening by courting male ducks, a

FIGURE 17.7 Sky pointing by a blue-footed booby. This display thought to have evolved from displacement
display is part of courtship and probably evolved from preening. The movements emphasize the bright mark-
the intention movements of flight. ings on the wings. (Modified from N. Tinbergen 1951.)
 The Evolutionary Origins of Signals 389

When all the hair is erect, the tamarin is likely to attack

or behave indecisively. However, when only the tail hair
is erected, it will probably flee (Moynihan 1970).

Other Behaviors as Raw Materials

for Displays
Although these three classes of behaviors have tradi-
tionally been considered sources of displays, E. O.
Wilson (1980) points out that “ritualization is a perva-
sive, highly opportunistic evolutionary process that can
be launched from almost any convenient behavior pat-
tern, anatomical structure or physiological change.” For
example, predatory behaviors have been ritualized in the
male gray heron. During courtship, he erects his crest
and certain other body feathers, points his head down-
FIGURE 17.9 A courting male frigate bird inflates his ward, as if to strike at an object, and snaps his mandibles
huge, brilliant red throat pouch to capture the interest closed, movements similar to those used during fishing
of females. (Verwey 1930). Food exchange has also been ritualized.
Billing, the touching of bills together, is likely to be
derived from the parental feeding of young and has taken
out the body. In birds such as the turkey, jungle fowl, and on a variety of meanings in different species of birds. It
bateleur eagle, the naked head or neck skin may flush and is common in courtship and appeasement display, func-
fleshy appendages may swell because of vasodilation. tioning to establish or maintain bonds. Mated pairs of
These changes now are part of signaling during masked lovebirds bill during greetings and after a spat
courtship and aggressive encounters. (Figure 17.10).
The respiratory system also has provided fodder for
the evolution of signals. In air-breathing vertebrates,
modifications of the respiratory system are used for the
production of sound signals. Visual signals can also
evolve from the respiratory system. Perhaps the best
examples are the inflation displays of birds, during which
the males fill pouches on their body with air to attract
mates. For instance, the male frigate bird has a pouch on
its throat that is inconspicuous when deflated, but when
inflated its enormous size and brilliant scarlet color
attract passing females (Figure 17.9).
Piloerection, or the erection of feathers and hair,
has a thermoregulatory function: fluffed feathers or hair
trap heat. Piloerection is frequently a part of aggressive
and appeasement displays (Smith 1977). For example,
a zebra finch that cannot escape from a dominant indi-
vidual fluffs its feathers as an appeasement signal
(Morris 1954). Feather position can also be part of
courtship displays (Morris 1956). Sometimes only
feathers in some regions are erected, such as the tail-
raising courtship display of the peacock. In other
species, the feathers used in signaling form a crest, as
in the sulfur-crested cockatoo; an ear tuft, as in several
species of owls; a throat plume as in some herons; or
eye tufts, such as the shiny green patch between the
eyes of the bird of paradise. In some mammals, such as
the rufous-naped tamarin, a squirrel-sized South FIGURE 17.10 Lovebirds are billing. This display may be
American monkey, the meaning of the message varies derived from parental feeding behavior. It now signals
with the part of the body on which hair is erected. nonaggression during greetings and after conflicts.
390 Chapter 17 / The Evolution of Communication

Normal flight Display flight The Ritualization Process

Ethologists hypothesized that during the evolutionary
Red-winged
process of ritualization, behaviors that take on a signal-
blackbird ing function become more stereotyped. They may also
change from the original movement, perhaps slowing
down, speeding up, or becoming more exaggerated.
Anatomical features, such as bright colors, large claws,
Normal flight Display flight and the like, might evolve in order to draw attention to
the display.
Yellow-headed
In a highly ritualized signal, ethologists hypothe-
blackbird sized that behaviors may become freed from the inter-
nal and external factors that originally caused them, in
a process called emancipation (Tinbergen 1952b). In
other words, the original proximate triggers no longer
cause the behavior to occur, and the behavior has only a
FIGURE 17.11 Ritualized flight in blackbirds. Almost any communicative function. Few experimental tests of
behavior can be a starting point for ritualization. Notice emancipation exist (Sebbel et al. 1998). One of the dis-
the exaggeration of motion that developed during ritu- plays explored most thoroughly from this perspective is
alization. (Modified from Orians and Christman 1968.)
the whistle-shake given by a species of duck (Tadorna
tadorna). The hypothesized evolutionary precursor of the
whistle-shake is the body-shake. Body-shakes, which are
Locomotion can also be ritualized for communica- just what they sound like, have the function of drying and
tion. In a number of bird species, male courtship includes rearranging the feathers. Ducks give body-shakes when
aerial displays of ritualized flight patterns. As you can see they are exposed to rain—or, in these experiments,
in Figure 17.11, ritualized flight is more conspicuous sprayed with water. Whistle-shakes look much like body-
than normal flight because some movements are accen- shakes but are finished off by the duck tilting back its
tuated and because the movements reveal special head and emitting a trill. Whistle-shakes are given in
plumage patterns (Orians and Christman 1968). Portions social situations, whereas body-shakes are not. However,
of the courtship display of the fiddler crab Uca beebei may like body-shakes, whistle-shakes are given in response to
have evolved from the movement of the male while water sprays. Thus, they are not completely emancipated
entering his burrow. When a female approaches him and from their original proximate causes (Sebbel et al. 1998).
his burrow, he raises his body and flexed front claw to
an almost vertical position, an exaggeration of his
movements into his burrow during the final stages of RECEIVER-BIAS MECHANISMS
courtship. This exposes his dark underside, which Whereas ritualization focuses on characteristics of the
probably functions to guide a female to his burrow sender that influence signal evolution, other hypotheses
(Christy and Salmon 1991). about the origin of signals focus on characteristics of the
A final, very odd example of a ritualized display is receiver. One hypothesis that specifically concerns sig-
pupil dilation in parrots. In birds, light regulation is per- nal origins is sensory exploitation. (We’ll save the
formed by a layer of the retina rather than by dilation discussion of another receiver-bias hypothesis, sensory
and constriction of the iris, as it is in mammals. Thus, drive, for the next section.)
pupil dilation is “freed up” for use in displays. Some According to the sensory exploitation hypothesis,
species of parrots, when aggressive or excited, rapidly the receiver has a preexisting preference for a particular
contract and expand their pupils (Bradbury and signal. In other words, features of the receiver’s nervous
Vehrencamp 1998). system—either its peripheral sense organs or how its
brain processes the signal—makes it more responsive to
a particular form of stimulus. The sender takes advan-
tage of the receiver’s preexisting sensory biases when new
STOP AND THINK signals are evolving (Ryan 1990; Ryan and Rand
Many of our examples are drawn from observations of 1993a, b).
behavior and reasonable guesses about the origins of sig- We already mentioned one example of sensory
nals. Is it possible to test whether a particular behavior has
exploitation—that of the evolution of swordtails in fish
been ritualized from another behavior? How would you
do so?
(Chapter 4). Another example of sensory bias comes
from túngara frogs (Physalaemus pustulosus). Like other
 Selective Forces That Shape Signals 391

display by moving his legs. The leg movements cause

vibrations that mimic those of the prey, leading the
female to grab him as she would a prey item. Food-
deprived females are more likely than satiated ones to
approach and clutch males. When the female detects
that she has grabbed a male and not a meal, she releases
the male. He then deposits his spermatophores (pack-
ets of sperm) in front of her and fans a pheromone con-
tained in the spermatophore toward the female. The
pheromone causes the female to pick up the packets and
place them in her genital opening. It is to the male’s
advantage to elicit the predatory clutch because it
allows him to orient to the female so that the sper-
matophores can be placed directly in front of her
FIGURE 17.12 A male túngara frog is calling to attract a (Proctor 1991). Analysis of the evolutionary relation-
mate. It is thought that “chuck” notes, added to the end
ships among water mites suggests that males evolved
of the “whine” part of the call, evolved to take advan-
courtship trembling after the sensitivity to water vibra-
tage of the female’s greater sensitivity to low-frequency
sounds. Thus, it is an example of the sender’s exploita- tions for prey detection. Thus, it seems that males
tion of a sensory bias in the receiver. exploit the prey-detecting mechanisms of the female to
their own advantage (Proctor 1992). This exploitation
works because females generally benefit from respond-
frogs, males call to attract females (Figure 17.12). The
ing to this vibratory cue. Other examples of sensory
calls consist of a descending whine, followed by a short
bias are described in Chapter 14 and later in this chap-
“chuck” call. Females prefer males with lower frequency
ter in the section on environmental influences on
chucks. In fact, one of the membranes in the ears of
signal design.
females is tuned so that it is more sensitive to low-
frequency chucks than to sounds of other frequencies.
Females benefit from their preference: low-frequency
chucks are produced by larger males, and mating with SELECTIVE FORCES THAT
larger males results in more fertilized eggs (reviewed in SHAPE SIGNALS
Ryan 2005).
Comparative studies shed light on the evolution of As you saw in the last chapter, the variety of signals is
the chuck call. Males of a closely related species of tún- astonishing: whiffs of chemicals, ear-splitting screeches,
gara frog, P. coloradorum, do not have chuck calls but sim- vigorous arm waves, electrical jolts. Even within a single
ply give the whine portion of the signal. Strangely, channel of communication, the form of signals varies
female P. coloradorum, although they have never heard a tremendously—think of the cacophony of different bird
chuck, prefer calls with chucks artificially added to the songs during a spring dawn. What causes this diversity of
end over the normal calls of their own species (Ryan and signal form? We’ll organize our presentation by discussing
Rand 1993a, b). The data suggest that the female pref- in turn the influence of characteristics of the sender, the
erence for chuck calls existed prior to the evolution of receiver, and the environment. However, these three
chucks. (The alternate hypothesis, that P. coloradorum influences do not work in isolation from one another, as
males once gave a chuck call but lost it over evolution- illustrated by many of the examples presented here.
ary time, is possible but less likely.)
There are also examples of male courtship signals
CHARACTERISTICS OF THE SENDER
that appear to have exploited the females’ sensory adap-
tations for prey detection. An interesting example is the It is obvious that anatomical structures form the foun-
water mite (Neumania papillator). The eyes of water dation for producing signals, as illustrated by the many
mites don’t form images, and so they must find food examples of signaling described in the last chapter. Some
and mates by other means. This species of water mite signals are produced by specialized structures, such as the
hunts by ambush. A mite waits on aquatic vegetation human larynx or the sensory organs of electric fish, and
with its first four legs raised. This position allows it to it’s clear that signal production is wedded to the evolu-
detect and orient toward passing copepods by the char- tion of these structures.
acteristic water vibrations caused by the prey’s swim- Even without highly specialized structure, body
ming movements. When a courting male detects a form influences signal design. For instance, consider four
female in the hunting posture, he performs a trembling ways that the form of a display might be influenced by
392 Chapter 17 / The Evolution of Communication

body size. (1) Small species might not be visible from CHARACTERISTICS OF THE
great distances, and other sensory modalities might be ENVIRONMENT
favored for long-distance communication. (2) Body size
may influence the evolution of physical characteristics As described in Chapter 16, a species’ habitat is key in deter-
that enhance visual signals. For example, in addition to mining which channel of communication (sound, chemi-
whole-body movements and posturing, two forms of cal, visual, etc.) is likely to be used. For example, unless you
visual display are common among the primates of can produce your own light like lantern fish do, visual sig-
Central and South America: facial expression and the nals are not much use in the ocean’s depths. For this rea-
erection of hair on parts of the body. Which form is son, whales rely on sound for long-distance communication
favored by evolution relates to body size, probably (Payne and Webb 1971). Even within a sensory channel,
because it is easier to see hair raising on a small body the structure of the signal can be influenced by the envi-
than a change in expression on a diminutive face ronment. Let’s look next at a few examples of signal evolu-
(Moynihan 1967). Squirrel-sized tamarins and mar- tion within sensory channels among closely related species.
mosets have long, silky fur that makes their displays
more effective. In contrast, larger species such as The Structure of Bird Song
capuchins, spider monkeys, and woolly monkeys have a On a weekend night, walking near a college dormitory, you
richer variety of facial expressions, although they tend to can hear for yourself some of the problems that face singing
be poker-faced compared to their Old World cousins. birds. As you pass very close by a dorm where a party is
(3) Body size influences agility, which in turn might taking place, you might be able to hear the sounds of music,
influence the type of signal that can be given. Just as voices, and laughter. You might be able to make out words
gymnasts are usually more petite than weight lifters, or even whole conversations, and you should be able to dis-
smaller animals are often better acrobats. For example, tinguish different elements of the music, perhaps the vocals,
many bird species give impressive aerial displays to their guitar, bass, and drums. As you walk further away, some of
mates. Small species of herons (Meyerriecs 1960) give these elements will start to fade out and become less audi-
more acrobatic displays than do larger species, as do ble. If your path takes you through trees, the high-pitched
smaller males within the same species of dunlin sounds of laughter may become inaudible more quickly
(Blomqvist et al. 1997). (4) Body size, at least in verte- than the thumping of the drums; if you are walking
brates, influences some characteristics of vocalizations. between buildings, the sounds may bounce and echo and
One such characteristic is the production of formants, become distorted; and if you are walking across an open
particular frequencies produced by resonances of the quad, all the sounds of the party may follow you for a long
vocal tract (reviewed in Searcy and Nowicki 2005). Vocal way. We describe characteristics of sound transmission as
tract length influences formant production and also gen- attenuation (weakening), or how far the sound will carry,
erally correlates with body size (Fitch 1997; Fitch and and degradation, or how distorted the signal becomes dur-
Hauser 2003). Formant production correlates with ing transmission. Both of these are affected by the charac-
body size in several species, including rhesus monkeys teristics of the habitat. If your goal was to make the noise
(Fitch 1997). This relationship also holds for domestic of your party transmit as far as possible, you might con-
dog breeds, with their 100-fold range in mass from chi- sider the surrounding environment before making your
huahua to St. Bernard (Riede and Fitch 1999). choice of playing particular styles of music or whether to
These examples bring home a general point: the invite guests most apt to emit high-pitched squeals.
evolution of animal signals does not happen in a vac- By playing recordings of bird songs and re-record-
uum. Many signals are produced by structures that have ing them at measured distances from the source, we can
other functions. Body size, hair, beaks, respiratory measure how songs attenuate and degrade in different
tracts—these traits are certain to be under selection for environments. As you might expect, dense foliage means
other reasons besides their role in producing signals. that songs are both more attenuated and degraded than
Thus, a complete understanding of how characteristics if the habitat is open (e.g., Blumenrath and Dabelsteen
of the sender influences signal design must account for 2004). However, not all songs are the same: for exam-
other selective pressures besides those directly on sig- ple, lower frequencies are less likely to become distorted
nals. This idea should be familiar. In Chapter 4, we in habitats with lots of foliage than are higher frequen-
described how beak shape in Darwin’s finches is impor- cies. The acoustic adaptation hypothesis proposes that
tant in both feeding behavior and song production. the acoustic properties of bird song are shaped by habi-
Selection on feeding behavior drives the evolution of tat structure (Morton 1975). Specifically, the hypothesis
beak shape. Changes in beak shape then have secondary predicts that in habitats with complex vegetational struc-
consequences for song production, which in turn influ- ture, songs should have low frequencies, narrow band-
ences female choice (Huber and Podos 2006; Podos and widths (the bandwidth is the difference between the
Nowicki 2004). Thus, a myriad of different selective highest and lowest frequency), whistles, and long notes.
forces act to shape a signal. In open habitat, like grasslands, songs should have high
 Selective Forces That Shape Signals 393

frequencies, broad bandwidths, trills, and short notes dietary supplements. So what can we say about the gen-
(Morton 1975; reviewed in Boncoraglio and Saino 2007). eral validity of the acoustic adaptation hypothesis? One
How do these predictions bear out? In some studies, approach is to do a meta-analysis. In this technique, one
both across species and across populations of the same performs a comprehensive literature review to find all the
species, the predictions hold up quite well. For instance, available studies that test the hypothesis. Then, the
the song of the great tit (Parus major) is ideal for this type results of all the studies are combined and tested
of analysis because the species has such a large geographic together using statistical techniques especially developed
distribution. Forest dwellers were found to have songs for such a purpose (e.g., Gurevitch and Hedges 1999).
with a lower pitch, a narrower range of frequencies, and Boncoraglio and Saino (2007) carried out a meta-
fewer notes per phrase than those of open woodland birds analysis for the acoustic adaptation hypothesis and found
(Figure 17.13). In fact, birds from similar habitats in that habitat structure does significantly influence the
Oxfordshire, England, and Iran, separated by 5000 km, acoustic properties of bird songs, though weakly.
sing more similar melodies than two English populations
occupying different habitats (Hunter and Krebs 1979). Lizard Displays and Background Motion
However, in other species, no differences in the predicted Environmental characteristics influence not only the trans-
direction were found between birds in different habitat mission of acoustic signals, but also that of other channels
types (e.g., Hylton and Godard 2001). of communication. For instance, visual signals are easier to
So, here we have a challenge: some studies support pick up when there is not so much “clutter” in the back-
the prediction of the hypothesis, whereas others do not. ground, so one might predict that displays should differ in
A mixture of different results is not unusual, not only in visually “noisy” environments versus environments with
behavioral studies, but also in other areas of scientific plain backgrounds. This hypothesis has been tested in
research such as tests of the efficacy of different drugs or lizards. Many species rely heavily on visual displays such as

a 8 b

4
Sweden Sweden Spain Spain
0
8

4
Norway Norway Iran Iran
0
Frequency (kHz)

4
England England Greece Greece
0
8

4
Poland Poland England England
0
8

4
Morocco Morocco Morocco Morocco
0 0 0 0
0.5 1.0 0.5 1.0 0.5 1.0 0.5 1.0
Time (sec)

FIGURE 17.13 Song structure of populations of great tits in many countries: (a) forest dwellers, (b) woodland inhabitants. The
left column in each section shows songs that are typical of each area, and the column on the right shows songs that are most
different from songs of the other habitat. Notice that the songs of the species that live in forests have a lower pitch, fewer
frequencies, and fewer notes per phrase than those of the woodland species. (From Hunter and Krebs 1979.)
394 Chapter 17 / The Evolution of Communication

a b

Vertical motion (degrees/sec)

c
40

–40
–40 0 40 –40 0 40
Horizontal motion (degrees/sec)

d
80

Duration (sec)
60

40

FIGURE 17.14 (a) A Jacky lizard from Australia in its

complex habitat. (b) and (c) show the movement of the 20
vegetation in the background under calm and windy
conditions. (d) Signaling Jacky lizards flick their tail
0
significantly longer under windy conditions. Calm Windy

head bobs, push-ups, back arching, extension of the dewlap Habitat Changes Caused by Humans
(a flap of skin under the chin) and tail flicks. Some species Human activities can alter the environment in which ani-
live in trees, so it might already be challenging for a receiver mals communicate and alter the selection pressures on
to see the head bob of a rival through the twigs and leaves. signal form. For instance, let’s look at another study of
Matters get worse when the wind is blowing and the veg- the birds that so nicely fit the predictions of the acoustic
etation is tossed about. Ord et al. (2007) observed two adaptation hypothesis, the great tits. This species is quite
species of Anolis lizards in nature and found that they dis- willing to live in close contact with humans and does well
played faster when it was windy and the background veg- in both urban and rural environments. Slabbekoorn and
etation moved more. Peters et al. (2007) studied Jacky den Boer-Visser (2006) recorded great tit songs in ten
lizards (Amphibolurus muricatus) in outdoor enclosures, a less big European cities, and compared them to ten forest
natural environment but one that allowed the researchers sites near each city. In all ten urban/forest comparisons,
to manipulate the “wind” by turning fans on and off. Jacky songs sung by urban birds were shorter, sung faster, and
lizards flicked their tails more when it was windy (Figure had higher minimum frequencies (Figure 17.15), most
17.14). Thus, the senders of signals may change their likely due to competition with the cacophony of the
behavior because of changes in environmental conditions. urban background.

a Paired t test: n = 10, t = 7.86, p < 0.001 b Paired t test: n = 10, t = –4.98, p = 0.001
3700 125
Duration of 1st note (msec)
Minimum frequency (Hz)

3600 115
3500
105
3400
95
3300

3200 85

3100 75
Urban Forest Urban Forest

FIGURE 17.15 The song of great tits (Parus major) in urban habitats and in nearby forest habitats. Ten city-forest pairs
were studied. Data from members of each pair are connected by a line. (a) The minimum frequency of songs was
higher in the urban habitats in each pair. (b) The duration of the first note of the song was significantly shorter in
urban habitats.
 Selective Forces That Shape Signals 395

Visual signals can be impeded by human-induced

environmental changes. For example, in the Baltic Sea,
the water has grown turbid (cloudy) because of phyto-
plankton growth, caused by the introduction of excess
nitrogen and phosphorus into the water. This is the
habitat of stickleback fish (Gasterosteus aculeatus), a
species in which males woo females with a courtship
dance and bright coloration. Turbidity makes it harder
for females to see males, and males in turbid water have
to court much more vigorously in order to get the
attention of the female (Engström-Öst and Candolin
2006). Males also can’t see each other as well when the
water is turbid. This is important from a female’s per-
spective because male sticklebacks keep each other hon-
est. Males in poor physical condition signal vigorously
but make poor mates—for one thing, they are prone to
cannibalize eggs, hardly the characteristic of an ideal
father (Candolin 2000a, b). In clear water, males pre-
vent other males in poor condition from dishonestly
signaling their condition to females. In turbid condi-
tions, males are not as likely to see each other. Thus,
in turbid conditions, the honesty of male signals is
reduced (Wong et al. 2007).
Research on the human impact on animal signaling
systems is in its infancy. No doubt some species, like the
great tits, will adapt successfully to changing conditions,
whereas others will not.

CHARACTERISTICS OF THE RECEIVER

Sensory Drive
We delayed discussing characteristics of the receiver that
affect signal evolution until after we discussed the role
of the environment because receiver characteristics can
themselves be shaped by the environment. This is called
sensory drive. The sensory drive hypothesis predicts
that if the constraints on a signal can be identified, then
we can make testable predictions about the direction of
signal evolution (Endler 1992).
Here is an example of sensory drive that is supported
by phylogenetic, physiological, and behavioral evidence
(Cummings 2004, 2007; Cummings and Partridge 2001).
Surfperch (Family Embiotocidae) are marine fish that live
FIGURE 17.16 Different optical habitats of related
in a variety of habitats differing in both overall light inten- surfperch species. Rainbow surfperch (Hypsurus caryi)
sity and spectral qualities (color) (Figure 17.16). The reti- (top) live in an environment that is highly variable in
nas of surfperch species differ. For some, the visual light intensity. Striped surfperch (Embiotoca lateralis)
pigments are best at detecting differences in color contrast. (bottom) live in deeper water in the kelp forest where
In others, the visual pigments are tuned to detect differ- the background light is more even.
ences in brightness. Because of physiological limits asso-
ciated with having only two classes of retinal cones, it is Cummings made over 250 scuba dives in order to
possible to be good at detecting differences in color con- measure the light characteristics of each species’ environ-
trast, or at differences in brightness, but it’s not possible ment. By combining these data with what she knew about
to be good at both. The first question Molly Cummings the performance of the photoreceptors under different
(2004) tested was whether surfperch species are adapted conditions, and the visual characteristics of their food, she
to detect prey in their particular light environment. could predict the optimal “tuning” of the visual pigments
396 Chapter 17 / The Evolution of Communication

for each environment. The species matched these pre- followed by chucks. Fringe-lipped bats prey on frogs and
dictions well. For example, deepwater surfperch are bet- are more likely to prey on frogs that give multiple chucks
ter at detecting brightness differences, whereas species (Ryan et al. 1982). Thus, the presence of unintended
living in environments with lots of brightness variation receivers may also impose selection pressures on the sig-
(e.g. shallow water with lots of light flickering by waves) nal. The effect of unintended receivers, both within and
are better able to see color differences (Figure 17.17a). across species, on signal design is currently a lively area
Thus, it seems that natural selection to improve the abil- of research in communication.
ity to detect prey has produced fish with different biases
in their visual systems.
Now, let’s connect this with communicatory signals. LANGUAGE AND APES
Surfperch species vary in their color patterns, which males
use in courting females. Male color patterns have evolved Humans seem to be very interested in what separates us
to take advantage of their conspecific female’s biases: in from other species. We mentioned in Chapter 5 that tool
species that are good at detecting brightness, male colors use was long considered to be confined to humans, but
are more detectable in the brightness “channel,” and in now we know that this is not the case—it’s not even nec-
species that are good at distinguishing colors, males are essary to be a primate to be adept at using tools.
more detectable in the color “channel” (Figure 17.17b) Language, in contrast, seems to be a talent confined to
(Cummings 2007; Cummings and Partridge 2001). Thus, humans. Or is it? A better question is a more specific one:
changes in habitat led to changes in the visual system, which elements of language are unique to humans, and
which in turn led to changes in color patterns. which are more broadly shared? Are differences between
humans and other species a matter of kind or a matter of
degree? We won’t be able to answer all these questions
Unintended Receivers fully, but we can at least begin to shed some light on them.
Unfortunately for signalers, it’s not always just the
intended target that senses a signal. Examples abound
WHAT IS LANGUAGE?
where predators, parasites, or conspecifics other than the
intended receiver detect a signal, often to the detriment There is no one accepted definition of language. Here,
of the signaler. For example, recall the túngara frogs we will consider that true language contains the follow-
described earlier in this chapter, whose call is a whine ing four elements. First, words or signs must be used as

a 0.2
Difference in performance between
each species and the ancestor

0.15 H.caryi

0.1
M.aurora
0.05
Sensory
0
bias
–0.05
E.jacksoni D.vacca
–0.1
E.lateralis
–0.15
–0.2
FIGURE 17.17 (a) The sensory bias of the
ability of five surfperch species to detect dif-
b 0.3 ferences in brightness (dark bars) and color
Difference in detectability between

H.caryi
each species and the average

(light bars). The y-axis is the difference in

0.2 performance between each modern-day
species and that estimated for their shared
0.1
ancestor. (b) The difference in detectability
M.aurora of male color patterns (signals) in the
Signal brightness (dark bars) and color (light bars)
0 channels. The y-axis represents the differ-
bias
E.jacksoni ence in reflectance detectability between a
D.vacca
–0.1 species’ measured reflectance and that of the
average reflectance of species with similar
–0.2 E.lateralis colors. (From Cummings 2007.)
 Language and Apes 397

true symbols that can stand for, or take the place of, a real
object, event, person, action, or relationship. Second,
symbols should permit reference to objects or events that
are not present. Third, there should be some elements of
grammar, or rules that determine the relationship
between words. For example, true language necessitates
knowledge of syntax because a change in the order of
symbols can alter the meaning of the message. Finally,
words or signs should be combined to form novel phrases
or sentences that are understandable to others.

APE LANGUAGE STUDIES

Investigators have studied the ability of great apes—
chimpanzees (Pan troglodytes); bonobos, a species within
the same genus as chimpanzees (P. paniscus); western
(lowland) gorillas (Gorilla gorilla); and orangutans
(Pongo pygmaeus)—to learn language. As you will see, FIGURE 17.18 The chimpanzee Washoe using American
the approaches of different investigators have varied Sign Language.
tremendously.
Early studies were designed to teach chimpanzees to an unseen dog. She also invented combinations of signs
talk. In the longest and most thorough of these attempts, to denote objects for which she had no name. Classic
Keith and Cathy Hayes (1951) managed to teach a chimp examples are her signing “water bird” for a swan on a
named Viki to say three words—“mama,” “papa,” and lake and “rock berry” for a Brazil nut (Fouts 1974). By
“cup”—in a voiceless aspiration. In retrospect, these the time she knew eight to ten signs, Washoe had begun
attempts to teach chimps to speak were doomed to fail- to string them together. Examples of typical early com-
ure because chimpanzees lack the necessary vocal appa- binations are “please tickle,” “gimme food,” and “go in”
ratus to make the range of sounds of human speech (Gardner and Gardner 1969).
(Lenneberg 1967). Our precise motor control over our
face and mouth is linked to a particular gene called
FOXP2 (Marcus and Fisher 2003), which became estab- STOP AND THINK
lished in human populations at the relatively recent date Do you agree that the best way to test whether apes have
of about 200,000 years ago (Enard et al. 2002). linguistic skills is to incorporate them into human society
A more fruitful approach to demonstrating the abil- as much as possible? What are the advantages and disad-
ity of apes to acquire language skills has been to use non- vantages of this approach?
verbal languages. A well-known nonverbal language, and
the first to be taught to a chimpanzee, is American Sign
Language for the Deaf, ASL. In the 1960s, Allen and Other ape-signing projects were begun by other
Beatrice Gardner trained Washoe, a young chimpanzee, teams. Roger Fouts (1973) and Herbert Terrace and his
to communicate using ASL (Figure 17.18). The co-workers (Terrace et al. 1979) continued working with
Gardners believed that an interesting and intellectually chimpanzees, Francine Patterson (1978, 1990) worked
stimulating environment would assist the development with a gorilla named Koko, and Lynn Miles (1990)
of language skills. For this reason, Washoe and other extended the studies to an orangutan named Chantek.
chimps were reared as much as possible like human chil- The techniques employed in these signing studies were
dren, even living inside the house with the Gardners. similar to those originally used by the Gardners, except
However, spoken English was not permitted around that spoken English was permitted in the presence of the
Washoe because it was feared that it might encourage apes. The emphasis of these studies was on the produc-
her to ignore signs. After four years of training, Washoe tion of language (the use of signs) and not on the com-
had a reported vocabulary of 132 signs. Her signs were prehension of language (understanding the meaning of
not restricted to requests. She used the signs to refer to signs) (Rumbaugh and Savage-Rumbaugh 1994).
more than just the original referent; she applied them Then a more critical voice was heard. Herbert
correctly to a wide variety of referents. For example, Terrace and his co-workers (1979) also taught ASL sign-
Washoe extended the use of the sign for dog from the ing to chimpanzees and initially felt they had some
particular picture of a dog from which she learned it to success. Most of their work was done with Nim
all pictures of dogs, living dogs, and even the barking of Chimpsky, a young male chimp named in “honor” of
398 Chapter 17 / The Evolution of Communication

Noam Chomsky, the famous linguist who has argued

that language is uniquely human. Terrace was writing a
book on Nim’s successes when, as he was reviewing slow-
motion videotapes of the signing chimp, he came to the Sarah No
conclusion that Nim was not spontaneously signing after
all. Their most important criticism of the work with
“talking” chimps was that the animals were simply imi- Jam Honey
tating their trainers. Terrace argued that the ape’s signs
were imitations of what the trainer had just signed and
that the trainers were too liberal in their interpretation
of the signs. Terrace’s team then analyzed the tapes of the Bread Cracker
Gardners’ work with Washoe and concluded that their
analyses were plagued with similar problems.
Take
Needless to say, Terrace’s conclusions were not left
unchallenged. Fouts, Patterson, and the Gardners FIGURE 17.19 Symbols used as “words” by Sarah, a
argued that Nim’s language abilities were stunted by the chimpanzee. Sarah learned to communicate using these
operant-conditioning procedures used in his training. plastic shapes.
Allen Gardner backed his claim “that you can turn it
[imitation] on and off, depending on the type of train-
ing you give” with a videotape of a chimp who showed her use of language consisted of using one word from a
little or no imitation of his trainer’s signs until the last choice of several to complete a preformed statement or
third of the tape, when operant-conditioning techniques arranging four to five words into a sentence of a specific
were begun. During this last section of the tape, 70% of word order. Premack established certain criteria for
the chimp’s signs were imitative. The Rumbaughs accepting that Sarah was using a particular chip as a
argued that because of the way Nim was trained, he word. Sarah had to be able to use the plastic chip to
never understood the meaning of words and that is why request the object it stood for, to select the proper chip
he was unable to create a sentence. In addition, Nim’s when asked the name of the referent, and to “describe”
trainers changed so often that he may not have had the the referent of a particular chip by using other chips.
opportunity to form the relationships claimed to be Premack’s strategy was to break down linguistic rules
essential for language development (Marx 1980). into simple units and to teach them to the chimp one at
Nonetheless, it became widely accepted that chim- a time. In this way, Sarah was taught not only to name
panzees could not learn language, and the later successes many objects but also to use more complicated rela-
of Nim and other chimps received little attention. tionships such as if-then and same-different.
Project Nim was discontinued, but other trainers began The more successful aspects of the pioneering stud-
to work with him. His language skills improved impres- ies with Washoe and Sarah were combined in the LANA
sively and no longer depended on imitation (O’Sullivan Project (Rumbaugh 1977; Rumbaugh et al. 1973). The
and Yeager 1989). The signs of the orangutan, Chantek, chimpanzee Lana was trained to use a computer to com-
were more spontaneous than those of Nim and could not municate. This computerized language system elimi-
be attributed to imitation (Miles 1990). Washoe and the nated social cueing and the difficulty of interpreting the
other language-trained chimps signed to other animals symbols, problems that plagued the sign language stud-
and objects (Gardner and Gardner 1989) and frequently ies. Lana communicated in a symbolic language, Yerkish,
to themselves (Bodamer et al. 1994). After Washoe’s bio- which was invented for the purpose. Yerkish words,
logical infant died, she adopted a ten-month-old infant called lexigrams, are geometric figures built from com-
named Loulis. For the next five years, humans avoided binations of nine simple design elements such as lines,
using any sign language in the presence of Loulis. circles, and dots (Figure 17.20). Lana chose words by
Nonetheless, Loulis learned his first 55 signs during this pressing a computer key labeled with the lexigram.
time by observing other chimps (Gardner and Gardner When a key was depressed, it lit up and the lexigram
1989). Washoe and her family signed to one another simultaneously appeared on a projection screen. It is sig-
during all their daily activities, including playing and eat- nificant to note that Lana was required to use lexigrams
ing and even family fights, until Washoe’s death in 2007 in an appropriate order. That is, she had to learn syntax,
(Fouts and Mills 1997). the rules governing word order in a sentence. For exam-
At the same time that the Gardners were working ple, she learned that pressing lexigram keys to say,
with Washoe, David Premack (1976) was training “Please machine give . . .” might be rewarded but that
another chimpanzee, Sarah, to use plastic chips of vari- pressing “Please give machine . . .” was not an accept-
ous shapes and colors as words (Figure 17.19). Most of able way to make a request. Unlike Sarah, who was given
 Language and Apes 399

Design elements the assumption is often made that it is using language

and not simply performing a complicated behavior for a
reward. This conclusion is especially tempting when the
animal strings words together in a sequence for a reward.
1 2 3 4 5 How could it be determined whether the animals were
using the words as symbols or just mastering a complex
conditioned response? It has been argued (Savage-
Rumbaugh et al. 1980) that in the studies just described,
the apes were not required to do anything that eliminates
6 7 8 9
the possibility that they were simply using words as labels
rather than as symbols.
Examples of lexigrams
Additional experiments were designed to test
whether chimpanzees were really using “words” as true
symbols (e.g., Savage-Rumbaugh et al. 1978a). Two
chimpanzees, Sherman and Austin (Figure 17.21) com-
1,7 purple 1,2 black 1,5,6,9 red 1,4 black municated information to each other through the use of
Machine Name-of M&M Out-of
symbols, information that could not have been commu-
nicated without the symbols. They were trained, as Lana
was, to communicate by pressing computer keys
embossed with lexigrams. The emphasis in this program
6,8 purple 3,5,8 blue 2,3,5,9 blue 5,7 black was on interanimal communication, so the animals were
Lana Eat Tickle Into not taught to produce strings of lexigrams or to adhere
to grammatical rules. In addition, they were raised in a
FIGURE 17.20 Symbols of the Yerkish language. Each
social, preschool-style setting. The animals were first
“word” is a combination of a few geometric shapes
embossed on keys of a computer keyboard. taught to name foods by pressing lexigram keys. It is
Chimpanzees have learned to communicate by important to note that they were taught to distinguish
pressing the appropriate keys. between the use of a food name as its name and the use
of the food name as a request because they were never
allowed to eat the same item that they named.
a limited choice of words to use at one time, Lana After this training, the animals’ ability to commu-
always had her complete vocabulary available to her. nicate with each other symbolically was investigated. In
Lana developed a large vocabulary and mastered one test, Sherman and Austin were able to specify foods
Yerkish grammar. She also coined new words. For to one another using lexigrams. The first chimpanzee
example, she called an orange soda a “Coke-which-is- was taken to a different room, where he watched the
orange” (Rumbaugh 1977).
The evidence presented up to this point suggests
that apes do quite well on the checklist of language char-
acteristics. The projects with Washoe, Sarah, and Lana
demonstrated that apes are relatively adept at learning
“words” in various forms, use them to refer to objects not
present, readily string words together in short sequences
so that the strings adhere to rules of grammar if that is
required, and have the ability to coin new words.
However, let’s look more deeply.
Are these apes really using words as symbols? When
a chimp uses a “word” (sign, plastic chip, or lexigram) to
name an object, is it used as a symbol that stands for the
object, one that can be used to refer to the object even
when it is not present, or is the animal producing the
“word” because it has been associated with the object
through a reward system? When a pigeon pecks a red FIGURE 17.21 Sherman and Austin, two chimps, have
key for food and a green key for water, no one assumes shown that apes can cooperate with each other to solve
that the animal is using the key to represent the item. problems by using the symbolic language Yerkish. They
However, when an animal as intelligent as an ape presses can communicate only when they have access to the
a key that is labeled, not with a color, but with a “word,” computer keyboard.
400 Chapter 17 / The Evolution of Communication

experimenter bait a container with one of 11 different lexigrams, he also knew the English words that the lex-
foods or drinks. That animal was then led back to the igrams represented.
keyboard and asked to name the food in the container. Kanzi had begun to learn communication skills sim-
After observing the response of the first chimp, the sec- ply by observing his mother’s training. He was never
ond chimp was permitted to use the keyboard to request trained to use lexigrams. Instead, he picked up the use
the food. If both animals were correct, they were given of language in much the same way as a child would. Once
the food or drink. Sherman and Austin were able to com- the researchers recognized this, reward-based language
municate with one another in this manner whether or training was stopped and replaced with conversation.
not they used the same keyboard and even if the exper- Kanzi’s constant human companions used lexigrams, ges-
imenter was ignorant of the identity of the food in the tures, and speech to communicate with one another and
container. Also, the animals communicated regardless of with him. In this way they served as communicative
which of them was the observer. The animal that did the models. Once he learned to use lexigrams, he began to
requesting based on the information provided by the use them to refer to items like food or objects or to loca-
knowledgeable chimp could demonstrate that he knew tions that were not in sight (Savage-Rumbaugh 1986;
which item he was asking for by selecting its picture from Savage-Rumbaugh and Lewin 1994). Kanzi became a
a group. However, when the chimp who knew the iden- language star able to communicate on a board with 256
tity of the food was prevented from using the keyboard lexigrams (Figure 17.22).
to describe the contents of the container, he could not Besides being able to produce language, Kanzi and
transmit the information to his buddy (Savage- several other apes have demonstrated that they understand
Rumbaugh et al. 1978a). spoken English. In these tests, the words were presented
The chimps also passed the next test—using symbols through headphones or from behind a one-way mirror to
to inform each other of the appropriate tool to use to avoid inadvertently cueing by gestures or facial expres-
solve a problem. The animals were kept in separate sions. The sentences were usually commands to perform
rooms. One chimp had to decide which one of six tools some action with one or more objects or people. The per-
he needed to obtain hidden food and then ask the other son evaluating the response did not know what had been
one for that tool via the keyboard. They could success- requested. Many of the requests were so unusual that it
fully cooperate in this manner only when the keyboard would be impossible to have carried them out without
was turned on (Savage-Rumbaugh et al. 1978b). Clearly, actually understanding the language. Consider for exam-
Sherman and Austin were using words as symbols and ple, the directive, “Put the raisins in the shoe.” Kanzi
not simply labeling objects. responded correctly to 72% of over 600 requests (Savage-
The work with Sherman and Austin was important
for reasons other than demonstrating that chimpanzees
can use symbols in communication. It marked the begin-
ning of a shift in emphasis from demonstrating that apes
can produce language to showing that they can understand
the symbols or words of language. In addition, it showed
how important the learning environment is in the devel-
opment of language comprehension.
The change in emphasis and learning environment
led to great progress in the study of apes’ language abil-
ities. Consider the remarkable abilities of Kanzi and his
half-sister, Panbanisha, bonobo or pygmy chimpanzees
(Pan paniscus). Kanzi, born in 1980, was adopted and
raised by Matata, who was part of a language study by
Sue Savage-Rumbaugh. For two years, trainers futilely
tried to teach Matata to use lexigrams to communicate.
Kanzi, who was 6 months old at the beginning of the
study, was always present during Matata’s training ses-
sions. But other than occasionally chasing the symbols
projected over the keyboard, he showed little interest. FIGURE 17.22 Kanzi, a bonobo, has demonstrated the
When Kanzi was about 2.5 years old, he was separated most advanced language skills so far. He communicates
from Matata so that she could be bred at another site. with a computer keyboard that has over 250 lexigrams.
Much to the amazement of the experimenters, Kanzi He was not trained by operant-conditioning techniques.
began to use the symbols of the keyboard that they had Instead, he observed and interacted with humans who
tried to teach to Matata. Not only did he know the used gestures and lexigrams to communicate.
 Communication and Animal Cognition 401

Rumbaugh et al. 1998). Panbanisha, Kanzi’s half-sister,

was reared in the same type of learning environment as COMMUNICATION AND
Kanzi had been. She, too, shows remarkable comprehen- ANIMAL COGNITION
sion of spoken English, responding correctly to 77% of
145 sentences (Williams et al. 1997). Many people have wondered what it is like to be an ani-
A different approach to understanding Kanzi’s lan- mal—specifically, whether nonhuman animals have
guage skills comes from an analysis of his vocalizations. thoughts or subjective feelings and whether they are
Kanzi was not trained to vocalize, but he spontaneously aware of other animals’ feelings. Such musings have led
makes sounds many thousands of times per day. The some investigators to consider whether nonhuman ani-
question was whether the structure of these vocalizations mals are cognitive, conscious, aware beings. We exam-
was related to anything in the environment. Taglialatela ined some evidence for cognition in Chapter 5, including
et al. (2003) analyzed hundreds of hours of videotape and self-recognition and perspective-taking. Here, we will
categorized vocalizations according to what Kanzi was examine animal cognition and communication.
doing at the time. For example, if Kanzi gestured to a Donald Griffin (1978, 1981, 1984) has suggested
grape or pointed to the lexigram representing a grape, that tapping animals’ communication lines is a way to
his vocalizations were coded with “grape.” Then, the find out whether animals have conscious thoughts or
acoustic structure of the vocalizations was analyzed with feelings. After all, the only way we know about the
a computer. Vocalizations given within a particular con- thoughts or feelings of other people is when they tell us,
text were statistically similar to each other, but different through either verbal or nonverbal communication. So,
from those given in other contexts. Thus, Kanzi seemed if nonhuman animals also have thoughts and feelings,
to be spontaneously uttering different sounds that have they probably communicate them to others through
different content (Savage-Rumbaugh et al. 2001). their communication signals. If we could learn to speak
Slocombe and Zuberbühler (2005) studied natural their language, we could eavesdrop and thereby get a
sounds produced by chimpanzees when they find food. glimpse into the animal mind. We might also learn about
Chimps produce acoustically distinct “rough grunts” the animal mind through interspecies communication—
when encountering different foods. In the chimps’ enclo- teach the animal a language that we understand, such as
sure, bread was always in one location and apples in in the great ape language studies, and then ask it how and
another. When a chimp heard a recording of the “bread what it thinks (Pepperberg 1993).
call,” it looked in the bread location, and when it heard Most people agree that one sign of cognition is the
an “apple call,” it looked in the apple location, suggest- ability to form mental representations of objects or
ing that it correctly interpreted the calls to refer to each events that are out of sight. So, one way to look for cog-
food type. This skill is found in other animals besides nition is to ask whether animal signals are symbolic, that
apes, including vervet monkeys (Seyfarth et al. 1980) and is, whether they refer to things that are not present
even chickens (Evans et al. 1993). (Smith 1991). We have seen that certain apes can learn
If we return to our definition, we see that there is a language that uses symbols, and they can use it to “talk”
some evidence that ape communication can fulfill each about things they don’t currently see and things that
of the four requirements of language. However, their occurred in the past (Savage-Rumbaugh 1986; Savage-
skills in some areas, especially using grammar and cre- Rumbaugh et al. 1998).
ating novel words or signs, are very limited compared to In Chapter 5, we discussed Alex, the African gray
humans. Consider that from age 18 months to 4 years a parrot who was able to vocally request more than 80 dif-
human child’s vocabulary grows from a few words to ferent items, even if they were out of sight. In addition,
thousands, reaching an estimated rate of one new word he could quantify and categorize these objects. He
per hour (Terrace 2005). showed an understanding of the concepts of color, shape,
Many researchers have argued that it is more infor- and same versus different for both familiar and novel
mative to study the natural communication systems of objects (Pepperberg 1991). Louis Herman and his col-
animals rather than artificially constructed systems in the leagues have shown that bottlenosed dolphins (Tursiops
laboratory. At this point, studies on the natural vocal- truncatus) can also learn to understand symbolic lan-
izations of great apes in the wild are still few; monkeys guages. In one of these languages, the “words” are ges-
are much better studied (Zuberbühler 2005). Still oth- tures, as in sign language. In the other, the words are
ers have ethical objections to keeping primates captive sounds generated by a computer. The words can refer to
in order to study their communication. Nonetheless, the objects, actions, and relationships among actions, and
experiments with Kanzi and other great apes have cer- many other things. In tests of language comprehension,
tainly provided food for thought and have gone a long the dolphins show that they understand the experi-
way in defining both the capabilities and limits of menter’s references to objects that are not present
language learning in apes. (Herman and Forestell 1985; Herman et al. 1993).
402 Chapter 17 / The Evolution of Communication

We can also learn about the way animals think from ferred from a harmonic arch to a warble or vice versa, two
their natural communication systems. The use of lan- calls that advertise high-quality food items. But it did not
guage involves the ability to relate words to meanings. transfer between calls that have different meanings, from
In some species, individuals classify signals by their a grunt to a warble or harmonic arch.
meaning and not by some obvious physical property, Additional evidence of cognition is the ability to
such as the way the signal sounds. Consider, for exam- understand the mental states of other individuals. Does
ple, the way rhesus monkeys respond to food calls. a communicating animal really understand what poten-
When a rhesus monkey finds food, it announces this to tial receivers know? Vervet monkeys apparently do not.
others with one or more of five different food calls. If Mother vervet monkeys give alarm calls when they sense
the item is really good—of high quality or rare—the a predator, but they do not appear to take into account
monkey will issue a warble, harmonic arch, or chirp. whether their offspring are ignorant or knowledgeable
However, a low-quality, common food item is about the predator (Seyfarth and Cheney 2003).
announced with a coo or a grunt. Thus, there are two Some intriguing data come from Thomas langurs
categories of food calls, and the calls sound distinctively (Presbytis thomasi) (Figure 17.24). Males give alarm calls,
different (Hauser 1998). but only if there is an audience. In one study, researchers
You may recall from Chapter 5 that animals will provoked alarm calls in wild langurs in the rainforest by
habituate (gradually stop responding) to a stimulus that having an observer carry a fake tiger skin over his shoul-
is repeated many times without consequence. Marc ders. After spotting the “tiger,” males called persistently,
Hauser (1998) used habituation to determine how the stopping only when every other member in the group
monkeys classify food calls. The reasoning is that if the also gave an alarm call (Wich and de Vries 2006). Thus,
monkeys classify the call by its sound, after habituating callers seemed to keep track of all the other members
to one type of call they will still respond to any other type in their groups. Whether langurs understand each
of call because it sounds different. However, if they clas- other’s state of knowledge is not yet known, but this
sify the call by its meaning, they will be unresponsive to species is a good one in which to test such questions
a call in the same category but will remain responsive to (Zuberbühler 2006).
a call in a different category. The calls were broadcast The question of animal awareness is difficult to
through a speaker, and a monkey was said to have answer scientifically, and we only have space here to skim
responded if it turned its head toward the speaker and the surface of the relevant literature. The answer, how-
stared at it. As you can see in Figure 17.23, rhesus mon- ever, has ethical ramifications, especially for researchers
keys classify calls by their meaning. Habituation trans- who study animal behavior. If the line between animals

a c
7 7 Warble/Harmonic arch
6 6
5 5
Mean response duration (s)

Mean response duration (s)

4 Harmonic arch 4
3 3 Grunt
2 2
1 Warble 1
0 0

b 7
d
7
6 6 Warble/Harmonic arch
5 5
4 Warble 4
3 3
2 Harmonic Grunt
2
1 arch 1
0 0
1 2 3 4 5 1 2 3 4 5
Habituation trials Test Habituation trials Test

Same meaning Different meaning

FIGURE 17.23 Rhesus monkeys relate specific food calls to their meaning. A rhesus monkey that finds food announces
this to troop members with a food call. Whereas warbles and harmonic arches announce high-quality, rare food items,
grunts are used for low-quality, common items. (a, b) When a call is broadcast repeatedly through a loudspeaker, mon-
keys gradually stop responding; that is, they habituate. Although the calls sound different, habituation transfers from a
warble to a harmonic arch or vice versa because the calls have the same meaning (only 1 of 12 monkeys responded to a
test call with the same meaning). (c, d) However, habituation does not transfer from a grunt to a warble or harmonic
arch because the calls have different meanings (13 of 14 monkeys responded to a test call with a different meaning).
(Data from Hauser 1998.)
 Summary 403

monkeys, for instance, learn to ignore animals that give

inappropriate signals.
We then examined two conditions that favor dis-
honest signals. (1) In some cases, senders and receivers
have different goals. For instance, a male might only be
interested in convincing a female to mate with him, but
a female might be interested in assessing the quality of
all males and choosing the best. (2) Signals might also
be dishonest if they are costly to assess or to challenge.
A stomatopod that challenges a conspecific that gives a
threat display will be attacked with great force if the
threat was not a bluff. Finally, we outlined conditions
under which honest and dishonest signals can coexist in
the same population.
How do signals originate? Earlier studies focused on
FIGURE 17.24 A Thomas langur. Males give alarm calls,
but only when there is an audience. identifying the behaviors of the sender that serve as raw
material for the evolution of displays in a process called
ritualization. The study of sensory exploitation focuses
on how some senders take advantage of a preexisting
and humans is erased or even smudged a bit, should we sensory bias of the receiver.
rethink the way we treat animals? Should we keep them In ritualization, three evolutionary sources for dis-
in zoos? Should great apes be used for language studies? plays have been recognized: (1) intention movements,
What about dolphins? those behavior fragments that may precede a functional
So what’s your opinion? Are animals aware, cogni- action, (2) displacement activities, or behaviors that
tive beings? All of them? Where do we draw the line? occur out of context, and (3) autonomic responses, such
as vasodilation and piloerection. Although these seem to
be among the most common sources for signals, almost
SUMMARY any behavior can be ritualized.
During the evolutionary process of ritualization, the
The view of communication that was most commonly form of the ancestral behavior is modified into a stereo-
held through the 1970s was that it is beneficial for animals typed signal. If one is reasonably sure of the precursor
to share information. Signals should thus be under selec- behavior, it can be compared to the current display to
tion to be honest, unambiguous, and informative. determine the changes in the original form. Typically, all
However, this view of communication is not always appro- or a part of the ancestral behavior pattern is exaggerated
priate. Sometimes the sender of a signal benefits from by changes in the duration or extent of movement, by
being dishonest. In this view of communication, a sender alterations in the rate of performance, or by repetition.
should send a signal because, on average, it manipulates In addition, the original actions may be combined in a
the behavior of others to its own advantage. Receivers new order, or some parts may be deleted. As the signal
should respond to signals when they benefit from them, becomes “emancipated” from the factors that originally
on average, even if they are sometimes deceived. caused it, the behavior may be shown in a new context
We next examined four conditions under which we and/or be motivated by different factors. Frequently,
expect signals to be honest. (1) Honest signals are likely these changes in behavior are accentuated by anatomical
when senders and receivers share overlapping goals. For modifications such as bright colors, antlers, and manes.
example, a baby bird and its parents both benefit when In sensory exploitation, the preexisting biases of the
the baby honestly signals its need for food. (2) Some sense organs of the receivers drive the evolution of a sig-
signals are honest because they cannot be faked. For nal. By using phylogenies to reconstruct the history of a
example, the distance between the eyes of a stalk-eyed group of species, it is sometimes possible to determine
fly is an honest signal of its size that it cannot change. that the bias of the sensory system was in place before a
(3) Signals are also likely to be honest when they are particular signal evolved.
costly to produce. This is the basis of the handicap The design of the signal may be influenced by a vari-
hypothesis for the evolution of signaling. For example, ety of factors. The anatomy and physiology of the sender
an animal that is in good condition might be better able are important in directing the evolution of their signals.
to pay the price of signal production than an animal in Characteristics of the habitat also play a key role in
poor condition. Thus, signals will be honest indicators determining both which sensory channel is favored for
of condition. (4) Signals are also likely to be honest communication, and exactly how signals within a sensory
when dishonest signalers can be identified. Vervet channel are structured. For example, the environment
404 Chapter 17 / The Evolution of Communication

influences how well bird song is transmitted, and one that can take the place of their referent and because they
can generate and test predictions about which sorts do not string signals together to form novel sentences.
of songs will be found in different environments. Visual Researchers have used an array of approaches in their
signals also differ in their detectability in different attempts to teach apes language. Their findings are
environments. Wind, for instance, can make visual sig- fascinating but the interpretation is controversial.
nals harder to see, and several species of lizards modify Some people have suggested that knowledge about
their signals under windy conditions. Human activities the communication systems of animals may provide an
are changing the conditions for signal transmission for insight into the question of animal cognition. Some ani-
many species. Finally, characteristics of the receiver also mals understand signals that represent items that are out
influence signal design. Receivers’ sensory systems and of sight. The natural communication systems of some
the psychology of how they interpret signals have been species, such as rhesus monkeys, reveal that they can
modified by natural selection. relate meaning to signals. Researchers have also looked
Most animal communication signals are not true for evidence that senders of signals understand the men-
language because animals do not use signals as symbols tal states of those with whom they are communicating.
 18
Conflict

Aggression and Conflict seek each other out in the barren mountains of the
Why Do Animals Fight? Middle East and fight for access to females on near-
An Evolutionary View of Conflict vertical slopes.
The Evolution of Fighting Behavior This particular confrontation begins with an assess-
Using Game Theory to Understand the Evolution ment. The two males approach each other, lower their
of Conflict heads, and interlace their horns so that the horn of one
Asymmetries in Contests individual is inside the horn of another. The horns make
Conflict Among Group Members contact with a loud clack, and then the ibex begin to push
How Dominance Is Determined one another, shoving each other’s heads to the side.
The Benefits of Being Dominant The fight escalates. The ibex back away from each
The Benefits of Being Subordinate other. One rears up on its hind legs, walks a step or two
Conflict Over Space toward its opponent, and crashes downward with its head
turned to slam its horns against its opponent’s horns.
Home Ranges, Core Areas, and Territories
The Ideal Free Distribution and Space Use
This sequence of shoving, then “rear clashing” (Alvarez
The Economics of Holding a Territory
1990), repeats. Each time, the clash increases in feroc-
The Economics of Territory Size
ity. Finally, the fight ranges across a steep, rubble-strewn
Strategies for Reducing the Cost of Territorial
slope that provides only treacherous footing, and rocks
Defense and small boulders clatter and bounce down into the dis-
A Proximate View of Conflict tant valley below. Sometimes similar fights end with the
loser plunging down a cliff to a messy death.
Aggression and Testosterone
You may see a very different battle in a British wood.
Stress, Aggression, and Dominance
Speckled wood butterflies, meeting each other in a sunny
patch in the forest, spiral up and up, until finally one flies
If you’ve seen the Trials of Life, the PBS nature series off and yields the spot to its rival. No blows are
with David Attenborough, you probably remember the exchanged, and the loser flies off unharmed (Davies 1978).
ibex fight. Ibex are mountain goats with long, curved, Why are these fights so different? It’s obvious, you
ridged horns. At the start of the breeding season, males may say—butterflies have no weapons, and ibex do.

405
406 Chapter 18 / Conflict

But even when animals do have weapons, contestants or its own survival. As we add more examples to our list
often exhibit restraint. Many ibex fights—in fact, most— in this chapter, keep in mind that we judge the value of
are low-risk shoving matches. Besides locking horns and a resource from an evolutionary perspective: that is,
pushing, ibex shove with their necks and shoulders in whether it increases the likelihood that an animal passes
apparent tests of strength of their opponent. Many con- on its genes.
tests end at this point, with the weaker ibex galloping off. Conflict is not ubiquitous across species, but is
In this chapter, we will address the nature of con- extremely common indeed. It manifests itself in a vari-
flict. Why are some fights settled peaceably and quickly, ety of ways: sea anemones lash each other with their
whereas others are more dangerous? Why do some ani- stinging tentacles, baboons charge at one another and
mals negotiate dominance status or a territorial bound- slash with their alarmingly large canine teeth, and even
ary, and then honor the agreement? limpets (shelled molluscs, resembling flattened snails)
have slow-motion battles in which they try to pry each
other off rocks into the swirling surf. The variation in
AGGRESSION AND CONFLICT the manner and intensity of conflict in different species
is enormous. But can we find and apply broad principles
Given the drama of many conflicts between animals, it’s that help us understand the evolution of conflict in a
no wonder that behaviorists have been motivated to study more general way? Let’s first take a big-picture view of
them, both empirically, with field and laboratory exper- conflict. We’ll then move to two special cases: conflict
iments, and by developing theoretical models. Several within groups, and conflict over territory. Finally, we’ll
terms have been used to describe conflict. Aggression briefly touch on proximate causes of conflict.
has an everyday connotation for us but has been formally
defined in the behavioral literature as a behavior that
appears to be intended to inflict noxious stimulation or AN EVOLUTIONARY VIEW
destruction on another organism (Moyer 1976). OF CONFLICT
Interspecific interactions such as predation are often
included under the umbrella of this term, but generally THE EVOLUTION OF FIGHTING
not included are behaviors that occur in response to BEHAVIOR
aggression, such as fleeing. Another term that is meant
to encompass the behavior of both the aggressor and the At first glance, there seems to be an evolutionary puzzle
animal that is the focus of the aggression is agonistic in animal conflict. In the last chapter, we saw many
behavior (note that this differs from agnostic behavior!). examples of displays in which animals signal to one
Agonistic behavior includes all conflict between con- another rather than engaging in physical aggression. But
specifics, including threats, submissive behavior, chasing, if a resource is in limited supply, why don’t animals
and physical combat. Agonistic behavior only includes always fight with their maximum effort? After all, an
interactions between conspecifics and excludes aggres- animal that only stands erect and fluffs its feathers to
sive acts between species, such as predation. communicate its aggressive intentions will surely lose if
its rival attacks violently.
Early ethologists thought that restraint in fighting
WHY DO ANIMALS FIGHT? evolved “for the good of the species,” but we know now
Conflict is potentially dangerous. It is also metabolically that this explanation is not plausible. Natural selection
costly: displays and attacks require energy. In many favors the individual that passes on more of its genes,
species, oxygen consumption increases and lactic acid not the individual that behaves for the sake of other
accumulates in the blood or hemolymph (the fluid sur- individuals that are unrelated to it. Thus, we must seek
rounding the organs in many invertebrates), requiring other explanations for the variation we see in the inten-
recovery time (Briffa and Sneddon 2007). If there were sity of animal fights. To do this, we turn to game theory.
enough resources—such as food, shelter, and mates—for
all animals, conflict would be far less common. USING GAME THEORY TO
Unfortunately, resources are frequently limited. We’ve UNDERSTAND THE EVOLUTION
already seen numerous examples of conflict over limited
OF CONFLICT
resources. In Chapter 12, you read examples of conflicts
over food, and in Chapter 14, conflict over mates. To construct hypotheses about the evolution of conflict
Chapter 15 described conflict between parents and off- behavior, researchers have used mathematical models.
spring over the amount of care that a parent should Models are logic-based tools to help us understand
provide—here, the limiting resource is the amount of more precisely the costs and benefits of behavioral
energy and time that a parent has available to lavish on strategies. Recall that in Chapter 12 we used a mathe-
a particular offspring versus investing in other offspring matical model to predict which foods of a range of
 An Evolutionary View of Conflict 407

choices a forager should eat. For the problem of animal The average payoff for a hawk vs. hawk interaction
conflict, we need a different type of modeling approach is the payoff for the winning hawk, plus the payoff for
than we used for foraging. Because the optimal strat- the losing hawk, divided by 2 to get the average.
egy of one individual depends on what its opponent is
doing, we need game theory. Game theory predicts an value of the resource cost of being wounded V W

animal’s optimal behavior while taking into account the 2 2
behavior of other animals. Game theory was developed
by economists to predict human behavior in economic We can add this value to the payoff matrix in
markets, and was borrowed and modified by animal Table 18.1.
behaviorists (notably by John Maynard Smith, especially Now, see if you can figure out, without looking at the
1974, 1976). next paragraph, the payoffs for the other three interac-
Game theory shares some vocabulary with the other tions: hawk against dove, dove against hawk, and dove
models we’ve discussed but adds some new terms. In a against dove.
game-theory model, the combatants are called players. If a hawk meets a dove, the hawk immediately
Players have available to them different decisions, or attacks, and the dove flees. The hawk wins the resource,
strategies (a term also used in foraging models). We so its payoff is V. If a dove meets a hawk, the dove imme-
measure the costs and benefits, or payoff, of each strat- diately flees. The dove does not get injured, nor does it
egy using a currency. In economics, the currency of win anything. So, its payoff is 0.
game-theory models is money. As in the other biologi- If a dove meets a dove, both display. One eventually
cal models we’ve examined, in behavior, the currency is wins the resource, and the other walks away without get-
something that relates to fitness, such as the number of ting injured. Both pay the cost of display. Thus, the
offspring produced or the number of calories acquired. payoff for the winning dove is V – D, and for the losing
Strategies are assumed to be heritable, and thus suc- dove it is just –D. To get the average, sum these and
cessful strategies will increase in the population. A table divide by 2. We can rearrange these to make it a little
called a payoff matrix is used to organize the values of easier on the eye:
the payoffs of each strategy against each of the other V - D - D V - 2D V
strategies. = = - D
2 2 2
The simplest game-theory model of aggression is
called the hawk-dove game. Here is the scenario: two
players are fighting over a resource. Each opponent has TABLE 18.1 The Payoff Matrix for the
the option of playing one of two strategies, called hawk Hawk–Dove Game. Hawks always attack.
and dove. The hawk strategy is to immediately attack its Doves flee when attacked but otherwise display.
opponent. The dove strategy is to flee immediately if The contested resource provides an increase
confronted by a hawk and display if confronted by in fitness of V, the cost of being wounded is a
another dove. (Thus, “hawk” and “dove” do not refer to decrease in fitness of W, and the cost of
real hawks or doves, but are simply meant to evoke a pic- displaying is a decrease in fitness of D
ture of the “personalities” of these strategies.) If a hawk When opponent is:
meets a hawk, or a dove meets a dove, each opponent has
a 50% chance of winning. Payoff to: Hawk Dove
It’s very useful to look at the details of this simple
V - W
game, as the predictions from this model drive the ratio- Hawk V
2
nale for a great deal of empirical work described later in
the chapter. First, let’s work out the payoff matrix for this V
Dove 0 D
game. We must define three variables, each measured in 2
a currency that relates to fitness:
V  the value of the resource being contested Check the payoff matrix in Table 18.1 to see that the
W  the cost of being wounded in a fight values make sense to you. Now let’s get more specific,
D  the cost of displaying to an opponent and put some numbers into the payoff matrix. In this
What happens if an animal playing the hawk strat- example, the cost of injury is high.
egy meets another hawk? Both will attack immediately. V  30
One hawk wins the resource, so its payoff is V. The W  60
other hawk will be wounded, so its payoff will be -W. D 5
So, what is the average payoff for a hawk vs. hawk inter-
action? (Take a moment here to calculate this before Based on these numbers, fill in the payoff matrix.
reading ahead.) Check your answers in Table 18.2.
408 Chapter 18 / Conflict

TABLE 18.2 Values of Payoff Matrix When

V ⴝ 30, W ⴝ 60, and D ⴝ 5 BOX 18.1: HOW MANY HAWKS
AND DOVES?
When opponent is:
Often, neither hawk nor dove is an evolutionarily stable
Payoff to: Hawk Dove strategy. Rather, the stable, equilibrium composition of the
population is some combination of hawks and doves in a
30 - 60 mixed ESS. The stable proportion of hawks and doves
Hawk = - 15 30
2 occurs when the average payoff for the hawk strategy equals
30 the average payoff for the dove strategy. We can calculate
Dove 0  5  10 these proportions with an equation with just a bit more math:
2
p  the proportion of hawks in a population
1  p  the proportion of doves (because the propor-
Now what do we do with this information? tions of hawks and doves must add up to 1)
Remember that the currency of these models is in units
The average payoff for the hawk strategy can be put into
of fitness and that we assume that these strategies are
words:
heritable. Successful doves will have offspring that also
play the dove strategy, and hawks will give rise to hawks. (chance it will meet another hawk)  (payoff when
Thus, we can use game-theory models to predict meeting a hawk)
whether strategies in a population will increase in fre- 
quency, remain stable over generations, or disappear. (chance it will meet a dove)  (payoff when meeting a
In Chapter 4, we introduced the concept of an evo- dove)
lutionarily stable strategy, or ESS. An ESS is a strat- We assume that animals encounter each other randomly,
egy that, when played by most members of the population, so the chance an animal will meet a hawk is equal to the
cannot be invaded by another strategy. Let’s look at our proportion of hawks in the population, or p. Similarly, the
example with the numbers above, and ask if the dove chance of meeting a dove is (1 – p). We’ve already figured
strategy is an ESS. If the population were comprised out the payoff matrix in Table 18.1. Substituting variables
entirely of animals playing the dove strategy, the aver- for the words, we find that the average payment for the
age payoff would be 10. If an animal playing hawk hawk strategy is
entered the population, all of its opponents would be (V - W )
doves. The payoff to the hawk against a dove is 30, so p + (1 - p)V
2
the hawk strategy will do well. Because the units in the
Now figure out what the equation would be for doves
payoff matrix represent fitness, the hawk’s genes
before you continue reading.
increase in the population, and the hawk strategy The average payoff for the dove strategy is:
increases in frequency.
So, will the population eventually become all hawks? (chance it will meet a hawk)  (payoff when meeting a
In other words, is hawk an ESS? If the population is hawk)
comprised of all hawks, the average payoff is –15. If a 
dove were to enter the population, it would not do too (chance it will meet another dove)  (payoff when
badly in comparison to the hawks—it won’t win, but meeting a dove)
unlike the hawks, at least it won’t be wounded during Thus, the average payment for the dove strategy
half its battles. So, the frequency of the dove strategy
V
would increase. = p(0) + (1 - p) a - Db
2
Thus, with these values of W, V, and D, neither a
“pure hawk” strategy nor a “pure dove” strategy is an At equilibrium, the average payment for hawks equals
ESS. However, there is a combination of hawk and dove the average payment for doves. Thus, for any particular
values of V, W, and D, we can solve for p and figure out
strategies that is stable. This is called a mixed ESS. The
the stable proportion of hawks and doves. Let’s try it for
stable proportion of hawks and doves occurs when the
V  30, W  60, and D  5, the same values we used
average payoff for the hawk strategy equals the average to figure out the payoff matrix in Table 18.2.
payoff for the dove strategy. A mixed ESS can come
about either by a certain proportion of animals always (V - W )
Payoff for dove strategy = p + (1 - p)V
playing hawk and another proportion always playing 2
dove, or by all animals playing both hawk and dove with V
particular probabilities. See Box 18.1 for an example of Payoff for hawk strategy = p(0) + (1 - p) a - Db
2
how to calculate the proportions of hawks and doves at
a stable equilibrium.
 An Evolutionary View of Conflict 409

At equilibrium,
(V - W ) V
p + (1 - p)V = p(0) + (1 - p)a - Db
2 2
(30 - 60) 30
p + (1 - p)30 = p(0) + (1 + p)a - 5b
2 2
Rearranging to solve for p gives p  0.57. This is the
proportion of hawks. The proportion of doves is 1 – p,
or 0.43.
A mixed ESS can occur in two ways. First, different indi-
viduals can play either hawk or dove. At equilibrium, there
is a stable mixture of the different types of strategists in
the population. Alternatively, every individual can play
both strategies at the calculated frequency. In this exam- FIGURE 18.1 Male elephant seals are scarred and bloody
ple, an individual’s optimal strategy would be to play hawk from battle. Fights between bulls are brutal and often
57% of the time and dove 43% of the time. result in injury. Such battles are predicted by game
theory when the value of the resource is very high,
as in this example. Only the most dominant bull will
be harem master and leave offspring.
Are either hawk or dove stable strategies for any
values of V and W? If the value of a resource, V , is
and LeBoeuf 1977). The commotion attracts the attention
greater than the cost of being wounded, W , a pure
of other males, who attempt to interfere. As a result, gen-
hawk strategy is an ESS. If V  W, then a mixed ESS
erally only the largest and strongest males are able to mate.
will result. A pure dove strategy is never an ESS.
In other species, we see intensive hawk-like fighting,
but for a different reason: not because the value of the
Testing the Predictions of Game Theory resource is so enormously high as in elephant seals, but
Game-theory models generate a number of testable pre- because fight costs are low. For example, toads have no
dictions. One prediction is that the ferocity of a contest real weapons to use against conspecifics, and fights rarely
depends on the value of the resource relative to the cost end in serious injury or death (Davies and Halliday
of injury. This makes intuitive sense—you would prob- 1978). Because fight costs are low, toads are especially
ably put up more of a struggle with a mugger if your life willing to engage in lively battles over females (Figure
savings were at stake rather than a pocketful of change 18.2). In contrast, in other species that have serious
(in other words, if the value of the resource is relatively weapons, battles are often generally restricted to displays
higher). You would probably even let a thief take every- (Figure 18.3). Game theory reminds us that it is the cost
thing you owned if you thought you would be killed if
you put up a struggle (that is, if the cost of fighting were
high). As you might expect, this prediction bears out in
animal species as well. Let’s take a look at a few examples.
In some species, the prize for winning a fight might
be incredibly valuable, perhaps even a lifetime’s worth of
reproductive success. In these cases, we predict that ani-
mals should risk everything, even fighting to the death.
A classic example of brutal and bloody fights are those
among male elephant seals (Mirounga angustirostris and
M. leonina) for the right to mate (Figure 18.1). Adult
males bear many scars that serve as silent testimony to
the intensity of combat. After a fight, when a male enters
the sea, the water is often reddened with his blood. All
matings are performed by a few dominant males who
defend harems of females. The duels between males are
so strenuous that a male can usually be harem master for
only a year or two before he dies (LeBoeuf 1974; FIGURE 18.2 A toad fight. Game theory predicts that
McCann 1981). The rivalry is intensified by the females, fights are more likely to escalate when the costs (risk of
who vocalize loudly when a male tries to copulate (Cox injury) are low.
410 Chapter 18 / Conflict

ASYMMETRIES IN CONTESTS
In the basic hawk–dove model, we assume that all ani-
mals value the contested resource to the same extent and
that all individuals have the same ability to fight. But how
often is this actually the case? In real life, rivals are rarely
true equals, but generally differ in some quality. Thus,
contests are usually asymmetric.
Inequalities, or asymmetries, among rivals can be
grouped into categories (reviewed in Gherardi 2006).
Here, we examine asymmetries in: (1) the ability of
each contestant to defend the resource, (2) the expe-
riences of each contestant in previous fights, (3) the
value of the resource to each contestant, and (4) arbi-
trary asymmetries unrelated to either resource value or
the ability to defend the resource (Maynard Smith and
Parker 1976).

Asymmetry in Fighting Ability

One combatant may be larger or heavier, have bigger
FIGURE 18.3 A rattlesnake fight. Each of these males
could kill the other, but they do not bite. The males weapons, or be a more skilled fighter (reviewed in Hsu
press against one another, belly to belly. Finally, the et al. 2006). Characteristics such as these that bear on
weaker individual yields, and his head is pushed to the an opponent’s ability to defend a resource describe its
ground by the stronger animal. resource-holding potential (RHP). It seems intuitive
that contestants might increase their fitness by assess-
ing their opponent’s RHP relative to their own, and
of battle relative to the benefit of winning that drives adjusting their fighting strategy accordingly (Archer
fight intensity. 1988; Parker 1974). This is an example of a conditional
Compelling evidence of the influence of resource strategy, adjusted based on the conditions of the par-
value on fight intensity comes from species in which the ticular fight.
value of the resource changes over time or in different How do animals assess RHP? We’ve already seen in
places, so that we can test within a single species whether Chapters 16 and 17 that many displays are designed to
fighting intensity correlates with resource value. For convey an impression of size and strength to an oppo-
instance, the value of a female to a male is not always the nent. Thus, we expect to see animals evolve to assess one
same. In mammals, for example, the likelihood that a another accurately and to bluff convincingly when pos-
mated female conceives varies over time. Red deer stags sible. Some traits are difficult to bluff, and such traits are
fight most fiercely and, as a result, are wounded most often used in assessment. For example, a male mountain
frequently during the period when most calves are con- sheep with small horns will defer to a competitor with
ceived (Figure 18.4) (Clutton-Brock et al. 1982). larger horns (Geist 1971). Similarly, male red deer judge

Fights per 4 days

Conceptions per 4 days
Injuries per 8 days
16
14
conceptions, and injuries

12
Number of fights,

FIGURE 18.4 The number of fights among male red deer

10 (open circles) is compared to the number of conceptions
8 during the rutting season. The number of fights and
conceptions is indicated for four-day intervals. The
6
number of injuries per eight-day interval during the
4 rut is indicated by triangles. The number of fights and
2 injuries peaks during the interval when conception is
most likely. As predicted by game theory, male red deer
23 27 1 5 9 13 17 21 25 29 2 6 fight harder when the value of the resource is greater.
Sept. Oct. Nov. (From Clutton-Brock et al. 1982.)
 An Evolutionary View of Conflict 411

each other’s size by interlocking their antlers and push- did give some challenge displays but significantly fewer
ing, and the one that is outclassed retreats (Clutton- than their opponent. No fighting occurred. Prior losers
Brock et al. 1982). The size of the claw that one male just gave up and were chased away.
shore crab (Carcinus maenus) presents to his opponent
during an agonistic contest is more important than his
overall body size in determining the outcome of the con- Asymmetry in the Value of the Resource
flict (Sneddon et al. 1997). to Each Individual
Some species may not be able to judge their oppo- It is not difficult to imagine conflicts in which the con-
nent’s RHP, but only their own. Combatants that assess tested resource is more valuable to one contestant than
themselves as being weak may give up more quickly than to the other. Food is more precious to a starving animal
combatants that assess themselves as being strong. than to a well-fed one, so we might expect the hungry
Careful experimental design and analysis is needed to contestant to fight harder for it. This prediction is borne
determine whether a particular species is able to deter- out in northern harriers (Circus cyaneus). These hawks
mine the RHP of its rivals, or only its own RHP (Taylor catch prey that requires some time to digest before
and Elwood 2003). another one can be consumed. Once the owners of a ter-
ritory have eaten, the value of the remaining prey on the
territory temporarily decreases. During this interval,
Asymmetry in Experience harriers are not as aggressive toward territorial intrud-
After the initial contest, the experience of either winning ers that might steal food (Temeles 1989).
or losing can influence the outcome of subsequent Asymmetry in resource value can even tip the out-
encounters. Fighting experience is important in many come of fights in favor of a weaker opponent that has
species, including insects, spiders, molluscs, fishes, birds, lower RHP. For instance, like many species that endure
and mammals (Hsu et al. 2006). Depending on the a long and rigorous migration, small birds called
species, winners become more likely to win subsequent bluethroats (Luscinia svecica) prepare for departure by
fights, losers become more likely to lose, or both. In an storing away as much body fat as they can. A fatter bird
example that illustrates a typical experimental design, has a better chance of survival, and every bit of added fat
Mary Whitehouse set Argyrodes antipodiana spiders is an extra safeguard. When experimenters offered food
against each other. These tiny spiders make their living to groups of bluethroats, individuals with low body fat
primarily by stealing food from the webs of other spi- won access to the food more frequently than did those
ders, and they will fight fiercely for packages of silk- that were close to their maximum weight. Leaner birds
wrapped prey. Because size is important in determining were even able to chase away larger birds that were
the outcome of a spider fight, spiders were paired by size already positioned at the feeding bowl, presumably
so that neither had a natural advantage over the other. because the food was more important to the lean birds,
Each member of the pair was assigned randomly to fight and so they were more highly motivated to win
against either larger spiders (so they generally lost their (Lindstrom et al. 1990).
fights) or smaller opponents (so they generally won their An animal that knows a great deal about a resource
fights). After these “training fights,” the sized-matched may value it more than an animal that knows less about
spiders were pitted against each other. The results were it. For instance, a territory might be more important to
very convincing: in 15 of 17 pairs tested, spiders that had the resident than to an intruder because a resident has
experience as winners trounced their size-matched learned the location of food sources, escape routes, and
opponents (Whitehouse 1997). refuges (e.g., Stamps 1995). Thus, one might predict that
Loser effects can be quite long-lasting, as Gordon an individual should grow to value a territory more as it
Schuett (1997) demonstrated in male copperhead snakes becomes familiar with it. This expectation is supported
(Agkistrodon contortrix) fighting for access to a female. in removal studies of male red-winged blackbirds
First, for each trial, the experimenters placed two males (Agelaius phoeniceus). Territory owners were removed
with no fighting experience during the previous 6 to 12 until replacement pairs moved in. When the original
months in an arena with a female snake (the prize). One owners were released, they fought to reclaim their ter-
of the males was 8% to 10% longer than the other, mea- ritory. When the original territory holders were released
sured as the distance between the snout and the vent (the after being held in captivity for up to 49 hours, they
genital opening), and had greater body mass. In all cases, nearly always won back their territories (Beletsky and
the larger male won the fight and the female. When the Orians 1987). However, when the original owners were
pairs were rematched 24 hours later, prior losers gave up retained for up to a week, the new residents usually
without even challenging the competitor. The next day, defeated the former owners. The released owners were
prior losers were paired with unknown males that closely just as persistent in their attempts to recover their ter-
matched their own length but that had no recent fight- ritories after they had been removed for seven days as
ing experience. This time the losers from the first day they were after two days. The difference in the outcome
412 Chapter 18 / Conflict

of the contests, therefore, seems to be due to a change prior ownership: play hawk if you had possession first;
in the behavior of the new residents. They were more otherwise, play dove. If the bourgeois strategy is added
willing to escalate contests as the territory became more to hawk and dove strategies in a population, it does bet-
valuable to them (Beletsky and Orians 1989). ter than either, so it is an ESS. If all animals are playing
We see other examples in which knowledge about a the bourgeois strategy, the owner always wins with barely
resource influences fighting behavior. Hermit crabs, for a squabble, and the outcome of any dispute can be
example, use empty snail shells as shelters, carrying them reversed by switching ownership.
around on their abdomens. Much shell-swapping takes Is the bourgeois strategy common in nature? This
place as crabs attempt to steal shells from conspecifics. question is more difficult to answer than it first appears,
Researchers studying Pagurus bernhardus crabs made as illustrated by the following example.
some shells less desirable by gluing sand deep inside the
interior. Only the owner could tell that a particular shell
The Difficulty of Teasing Apart the Rules
had this unpleasantly scratchy surface; competitors for
of Animal Fights: A Case Study
the shell could not. Owners of the poor shells changed
their fighting strategies, fighting harder if they were in We return to the male speckled wood butterflies
the role of attacker, but giving up more quickly if defend- (Pararge aegeria), which defend spots of sunlight that
ing their shells (Arnott and Elwood 2007). serve as mating territories. Wood butterflies fight by fly-
ing upward together in a spiraling pattern. At the top of
the spiral, one flies away. When Nick Davies netted res-
Arbitrary Asymmetry ident butterflies, he found that the resident’s sunspot was
The differences between contestants that we have dis- almost instantly claimed by another male. When the ini-
cussed so far have a fairly straightforward relationship to tial resident was released again, it always lost the fight
the fighting strategy of each contestant. We have seen to regain its spot. Only if two males were experimentally
that a competitor might follow a conditional strategy tricked into joint “ownership” did an escalated contest
such as “play hawk if larger; play dove if smaller” or “per- involving spiral flight occur. Wood butterflies thus
sist longer when the resource is more valuable.” The seemed to be following the rule that “owner wins,” and
association between a resource’s value and resource hold- they became a classic example of an uncorrelated asym-
ing potential is thus called a correlated asymmetry. A metry (Davies 1978).
contestant maximizes its fitness by paying attention to It is very difficult, however, to document conclu-
correlated asymmetries because these are likely to pre- sively that the outcome of a contest is truly uncorrelated
dict the outcome of a battle. with the resource-holding potential of the winner. Might
Other differences between contestants are not log- there be some other reason that some butterflies are bet-
ically connected to the fighting strategy of the opponents ter fighters? One hypothesis is that a butterfly that is in
but do affect the outcome of the dispute. These arbitrary a sunspot, even for a few minutes, gets warmer and is
asymmetries (also called uncorrelated asymmetries) are, thus able to fly longer. This would produce the same
in essence, rules or conventions used to settle conflicts. result as Davies saw: the resident male would have a
Examples of arbitrary rules for settling differences chance to stay warm, while the male waiting it out in
among humans are flipping a coin or pulling straws. Davies’s net might cool off and then be at a disadvantage
There is no reason that “heads” should win or that the in the next contest.
short straw should lose. These are simply rules that are To test the thermal hypothesis, Stutt and Willmer
mutually agreed upon. (1998) experimentally manipulated the body tempera-
A potential example of an arbitrary asymmetry is ture of speckled wood butterflies. For each trial, two but-
that of prior ownership (or residency). Animals often terflies were caught and marked, and each was placed
appear to adhere to the principle that “possession is nine- into one of two clear plastic boxes. One box was insu-
tenths of the law.” For example, a hamadryas baboon lated with polystyrene, and the other was covered in
(Papio hamadryas) male permitted to associate with a black plastic. A 30-mm area at the top of each box was
female for as little as 20 minutes was perceived as the left transparent, allowing light to enter. The boxes were
“owner” by a second, newly introduced male. That the then placed next to each other and moved into a sunspot.
second male was deferring ownership was revealed when, The temperature in each box was monitored, and it
several weeks later, he was permitted to associate with quickly climbed higher in the black box than in the insu-
the female. When the tables were turned, the first male lated one. Within five minutes the temperature in the
did not challenge the second male’s ownership (Kummer black box rose to that previously measured in sunspots.
et al. 1974). However, the temperature in the insulated box still
To add the possibility of prior ownership, a variation approximated temperatures in shady locations. Then the
of the hawk–dove model includes a third strategy, bour- butterflies were released, each considering itself to be the
geois. The bourgeois strategy sets rules for dealing with resident owner of the sunspot. The duration of the esca-
 Conflict Among Group Members 413

lated flight was recorded, and the winner was noted as missive animal predictably yields to a dominant one
the one who returned to the sunspot where the boxes (Kaufmann 1983).
had been placed. Dominance hierarchies vary among species, and
When escalated flight occurred, the warmer male within a species they can vary with conditions and over
won significantly more often. This is consistent with the time. The simplest form of a dominance hierarchy is
idea that the winner of the spiral contest will be the male despotism, in which one individual rules over all others
with the higher RHP as measured by body temperature. in the group and the subordinates are equal in rank.
Note that the result is the same as in Davies’s experi- Hierarchies may also be linear. In this case, A is domi-
ments—the owner wins—but the underlying reason is nant over all other animals, B is dominant over all but
different. A, and C is subordinate to A and B but dominant over
The matter is still not settled, however. Kemp and the rest, as follows:
Wiklund (2004) argued that the experimental protocols
A ––> B ––> C ––> D ––> E
used in previous studies were unnatural and probably
traumatized the butterflies—and thus not all contestants This is often called a pecking order because it was
were motivated to hold territories. Kemp and Wicklund first described in chickens. Chickens commonly demon-
used enclosures in which butterflies were allowed to dis- strate their dominance by pecking lower-ranking animals.
cover each other naturally. The butterflies were allowed Sometimes dominance hierarchies are even more
to have a contest over a sunspot. The initial loser was complicated, with A dominant over B, B dominant over
then granted sole ownership of the sunspot, while win- C, but C dominant over A, or with hierarchies shifting
ners were temporarily stored in a cooler. Winners were as circumstances change.
then allowed to bask and initiate flights at their leisure.
In another experiment, Kemp and Wiklund also manip- HOW DOMINANCE IS DETERMINED
ulated temperature by chilling butterflies in a cooler. By
tossing bark chips up into the sunspot, investigators All the determinants of fight outcome that are important
could coax chilled butterflies out to investigate before in single encounters—size, strength, and experience as
they warmed up, and thus trigger a fight. a winner or loser—are important in social groups as well.
Contrary to what other researchers found, in In addition, in social groups, dominance may also be
Kemp and Wiklund’s studies, the individual that won attained through an association with a high-ranking indi-
the first fight was highly likely to win the second fight vidual. For example, when two flocks of dark-eyed jun-
when it was an intruder. Thus, there is no evidence of cos merge, all the birds of one flock tend to rank above
an uncorrelated asymmetry. In addition, they found no those of the other. It is as if the subordinate birds ride
effect of temperature on winning. Instead, Kemp and the coattails of the highest ranking bird to achieve dom-
Wiklund suggest that butterfly fights may be deter- inance in the combined flock. A possible explanation for
mined by intrinsic aggression or fighting ability, or by the subordinates’ rise in dominance is that the highest
prior experience. ranking individual behaves differently toward birds that
This example holds a few lessons. First, alternative are familiar because they were members of its original
hypotheses that produce the same results must be care- flock, than toward unfamiliar flockmates (Cristol 1995).
fully identified and parsed out. Second, subtleties in In other species, dominance may be a birthright
experimental design can lead to dramatically different based on the status of one’s parents, just as it is in many
results—even very elegant experimental designs may human societies. This is the case among rhesus mon-
have hidden problems. This species is sure to be the keys. Adult females have a linear dominance hierarchy,
focus of additional research. and offspring assume a dominance position just below
their mother (de Waal 1991).

CONFLICT AMONG GROUP THE BENEFITS OF BEING DOMINANT

MEMBERS
It seems entirely reasonable to expect that being domi-
Next let’s consider a special case of conflict: conflict nant will translate into increased fitness. Two fitness-
among animals in the same social group. In a stable social related benefits have been especially well studied: access
group, the same two individuals are likely to repeatedly to food and access to mates.
encounter one another in competitive situations. In such Dominant individuals get more food than subordi-
cases, the animals don’t usually fight each time they nates across a wide range of species. For instance, food
meet. Instead, relationships develop among them. is an important benefit of dominance among brown hye-
Dominance refers to the ability of one animal to nas (Hyaena brunnea) of the central Kalahari (Figure
assert itself over others in acquiring access to a limited 18.5). Each sex has a clear linear dominance hierarchy,
resource, such as food, a mate, or a nesting site. A sub- and the male and female at the top have equal rank.
414 Chapter 18 / Conflict

likely lose (e.g., Fournier and Festa-Bianchet 1995). In

fact, in many groups, both subordinates and dominants
suffer from a shake-up in the hierarchy. When the dom-
inance hierarchy in chickens is stable, hens fight less and
lay more eggs than when dominance relationships are
still being established (Pusey and Packer 1997).
Sometimes dominance has its costs—and subordi-
nates avoid incurring that cost (reviewed in Huntingford
and Turner 1987). For instance, dominant great tits or
pied flycatchers have higher resting metabolic rates than
do subordinate birds and require more food (Røskaft et
al. 1986). In some species (but not others), dominant indi-
viduals show increased glucocorticoid levels, a steroid hor-
FIGURE 18.5 Top-ranking brown hyenas enjoy two mone associated with stress (reviewed in Sands and Creel
important benefits of dominance—enhanced reproductive 2004). We’ll discuss these hormonal data later in this
success and increased access to food. Dominant hyenas chapter when we turn to a proximate view of dominance.
have more feeding time at carcasses than do subordinates.
In some cases, subordinates have the option of leav-
ing and joining another group. This, too, can be risky.
Although brown hyenas live in clans, they forage alone. For example, although subordinate red foxes have little
During the rainy season, the primary component of their hope of living long enough to become dominant in their
diet is the remains of kills, such as giraffe, gemsbok, and natal group, the mortality rate of those who disperse is
wildebeest, made by other predators. As many as six hye- also very high (Baker et al. 1998). So, although the ben-
nas may arrive at a carcass together, but only one or two efits of group living may not be as great for a subordi-
will feed together. The top-ranking animals of a clan nate animal as they are for a dominant one, they may still
have more feeding time at carcasses than subordinates. outweigh the costs of leaving the group. Besides, the sit-
In addition, subordinate males and females are signifi- uation in the group could get better for a subordinate.
cantly more likely to leave the carcass without feeding if The dominant animal could die or become weak enough
a dominant animal is present (Owens and Owens 1996). to be displaced. In the meantime, since many groups
In some social groups, all or nearly all the reproduc- consist of family members, a subordinate animal may
tive benefits in a group go to a single individual that is the gain some fitness through kin selection by helping to
best competitor. In some species, the dominant female raise its siblings. (A more detailed discussion of helping
(and in some cases the dominant male) suppresses repro- is found in Chapter 19.)
duction by other members of the group. The most dra- Sometimes an animal that is generally subordinate can
matic examples of reproductive suppression occur in the occasionally win a fight to briefly gain access to a resource.
eusocial species, such as social insects and naked mole rats, Among Rocky Mountain bighorn sheep, for instance, sub-
in which only a single female—called the queen—repro- ordinate rams occasionally manage to win a fight and get
duces. Eusociality is discussed further in Chapter 19. a few seconds with a female—all that is necessary to cop-
More commonly, however, the dominant animals ulate. When lambs are born, the male most likely to be the
have a clear reproductive advantage, but they don’t com- father is identified by using a combination of genetic and
pletely suppress reproduction by subordinates. For behavioral data. The subordinate rams fathered 44% of the
example, each pack of African wild dogs (Lycaon pictus) 142 lambs born in two natural populations (Hogg and
has a clear dominance hierarchy in each sex. In two study Forbes 1997), so subordinates do quite well.
sites, top-ranking alpha females produced 76% and 81% Sometimes subordinate members of a group employ
of the litters. Although 82% of the dominant females more subversive techniques, such as alternative repro-
gave birth each year, only 6% to 17% of the subordinate ductive strategies. Sneaky males generally mimic female
females did so (Creel et al. 1997). characteristics to get past a dominant male and copulate
with his female. Satellite males generally position them-
THE BENEFITS OF BEING selves so that they can intercept females who are
attracted to a dominant male and copulate with them.
SUBORDINATE
We describe examples of these alternative reproductive
With the benefits of a dominance hierarchy clearly strategies elsewhere in this book: sneaky male side-
stacked on the side of the high-ranking members, we blotched lizards (Chapter 4), sneaky male plainfin mid-
may wonder what’s in the relationship for a subordinate. shipman fish (Chapter 7), and satellite male natterjack
If subordinates have no other choice but to stay in a toads (Chapter 14). In all cases, the underlying strategy
group, repeatedly challenging other individuals to fights is to avoid the costs of achieving and maintaining dom-
would lead to a risk of injury in conflicts they would most inance and still enjoy some reproductive success.
 Conflict Over Space 415

Subordinates may sometimes band together to chal- type of data collected, the lack of consistent terminology
lenge dominant individuals. For example, in a troop of can make comparisons among studies difficult (Maher
savanna baboons with eight adult males, the three lowest- and Lott 1995). Here, we will define a territory as a
ranking males regularly formed alliances to oppose a sin- defended space.
gle higher ranking male. The alliances gained reproductive Territories may have different uses, depending on
access to the female on 18 of 28 attempts (Noë and Sluijter the resource being contested. They may be used solely
1990). Alliances do not always overturn the current hier- for feeding, mating (recall our discussion of leks in
archy, however. In some Old World monkeys, females Chapter 15), or raising young, or they may be used for
band together and “gang up” on other females, and they a variety of purposes, in which case they are called mul-
generally target lower ranking monkeys (reviewed in tipurpose territories.
Wittig et al. 2007). Alliances occur in other animals besides
primates. For instance, in the ground-dwelling tropical THE IDEAL FREE DISTRIBUTION
bird called the white-winged trumpeter, which lives in
AND SPACE USE
social groups, subordinate males appear to collaborate in
interrupting the copulation attempts of the dominant male Before discussing how and why animals exclude each
(Eason and Sherman 1995). other from territories, it’s useful for comparison pur-
poses to think about how we might expect space to be
divided if animals do not defend territories. You can
CONFLICT OVER SPACE see an example if you have two bags of bread, a friend,
and a duck pond (or a flock of pigeons). Stand a short
A special category in the study of conflict is conflict over distance away from your friend. Both of you throw
space. Whereas some animals peaceably coexist, even bread pieces into the water. You throw a piece of
forming groups (Chapter 19), many species exclude con- bread every 15 seconds, and your friend throws his
specifics from particular areas. bread three times faster than you do, at the rate of a
piece every 5 seconds. What will the ducks do? The
first duck to arrive at the feeding area should go to your
HOME RANGES, CORE AREAS,
friend, where the bread is coming at a faster pace. The
AND TERRITORIES second duck should also go to your friend—even if it
We’ve already been talking about territories and space, only gets half the bread that your friend is throwing,
but let’s take a moment to formally define some terms. it’s still averaging one piece every 10 seconds, better
The home range of an individual animal is the area in than what you are providing. The third duck will get,
which it carries out its normal activities. It includes on average, a piece of bread every 15 seconds no mat-
space it defends from others, as well as space that is ter where it goes. If the third duck chooses your friend,
used by others. Within the home range there is often where will the fourth duck go? See Fig. 18.6 for a gen-
an area in which most activities are concentrated—the eral case.
core area. In some cases, the core area may be the
area immediately surrounding the nest site or perhaps
a food or water source. Animals may have a home
range and core area whether or not they share space
with conspecifics.
Surprisingly, the definition of a territory generates
little agreement. A survey of the literature on territori-
ality revealed no fewer than 48 different definitions
(Maher and Lott 1995). The most common one, “a
defended area” (Noble 1939), was used in only 50% of
the papers. Although this definition of territory empha-
sizes active defense of an area, other definitions down-
play defense and emphasize instead the exclusive use of
space (Schoener 1968). This latter definition is often
FIGURE 18.6 The ideal free distribution. Illustrated is a
more practical for describing the space use patterns of
case with two habitats, with habitat A of higher quality
animals when aggressive interactions over territory than habitat B. Thus, as animals arrive in the area, they
boundaries are difficult to observe in the field. Because should select habitat A. As the number of animals in
of the secretive habits of many small mammals, for exam- habitat A increases, the number of resources available
ple, it is virtually impossible to state with any certainty to each animal declines. Finally, the quality of A declines
that the exclusive use of an area is maintained by active to the point that a new arrival will get the same benefits
defense (Ostfeld 1990). Since the definition affects the regardless of which habitat it selects.
416 Chapter 18 / Conflict

The pattern by which these ducks settle into different Resource Abundance Generally speaking, territori-
areas is an example of the ideal free distribution (Fretwell ality is favored when resources are moderately abundant.
1972): it’s “ideal” because the animals know the value of If resources are scarce, an individual may not gain
each habitat and can instantly choose the best one, and enough benefits to pay the defense bill, and it is eco-
“free” because every duck is free to choose its location nomically wiser just to let other animals enter the area.
without interference. Of course, animals are not always For instance, the golden-winged sunbird will abandon a
either “ideal” or “free.” An “ideal” animal would have per- territory when it no longer contains enough food to meet
fect knowledge of the quality of a site, but of course that’s the energy costs of daily activities, including defense (Gill
not always true—we already have seen in Chapter 12 that and Wolf 1975). Chickadees living in habitats disturbed
animals often need time to gather information before they by logging and thus low in quality were less likely to
can assess the quality of their environment. In addition, defend their territories against intruders than were birds
animals are not always “free”; instead, others constrain in untouched habitats (Fort and Otter 2004). At the other
them from behaving optimally. For example, some ducks extreme, if there are more than enough resources to go
may be better competitors than others and grab more around, it is unnecessary to defend a territory. Water
than their fair share of the food. However, in spite of striders, insects that skate on top of ponds and streams,
the fact that deviations from the assumptions of the ideal are among the species that will cease defending territo-
free distribution are common, numerous species have ries if supplied with abundant food (Wilcox and
been shown to behave in accord with its predictions Ruckdeschel 1982). Similarly, female marine iguanas do
(Giraldeau 2006). not bother defending territories with nest sites on most
of the Galápagos Islands. They defend territories only on
Hood Island, the only Galápagos island where nest sites
THE ECONOMICS OF HOLDING are in short supply (Eibl-Eibesfeldt 1966).
A TERRITORY Resource Distribution All else being equal, we pre-
What influences the decision of an animal to hold a ter- dict that animals are more likely to be territorial if
ritory versus simply share its space with other animals? resources are moderately clumped. A pile of food, for
Having a territory all to oneself has clear benefits, such instance, is easier to defend than food that is spread
as exclusive access to resources, be they shelter, food, thinly over a large expanse, as long as the number of
areas of safety, mates, a high-quality site to raise off- competitors anxious to contest ownership is not too
spring, or some combination of these things. However, great (see discussion in Maher and Lott 2000).
holding a territory also has its costs. Energy is needed Intruder Pressure The number of other individuals
to patrol territory boundaries and display to or forcibly that are willing to compete for a territory is one of the fac-
evict intruders, and sometimes boundary fights can be tors determining territory cost. The more competitors,
dangerous. Territory acquisition and defense also take the greater the cost of defense. Male fruit flies (Drosophila
time away from other essential activities such as forag- melanogaster) defend small patches of food that are suit-
ing, an example of a “lost opportunity” cost. To take a able for oviposition, particularly if there are females in the
single example, great tits, a bird we have met before, vicinity. As would be predicted, however, flies are less
generally feed within 3 m of the ground, but most ter- likely to hold territories when there is a higher density of
ritorial defense, particularly singing, takes place high in males (Hoffmann and Cacoyianni 1990).
trees, about 10 m above the ground. Obviously, then, at
any given time these birds must choose between defense
THE ECONOMICS OF TERRITORY SIZE
and foraging (Ydenberg and Krebs 1987).
We predict, then, that territoriality will occur only Costs and benefits influence not only whether a territory
if the fitness benefits from enhanced access to the is held at all, but how large a territory should be. We can
resource exceed the fitness costs of defending the visualize what the optimal territory size might be by
resource—that is, when the territory is economically using a graph in which costs and benefits are plotted with
defendable (Brown 1964). What factors set the relative separate lines. Examine Figure 18.7. Here, benefits ini-
costs and benefits of holding a territory? Luckily for tially increase as territory size increases, but the line
researchers, we can often answer this question very begins to flatten out: there are just so many resources an
clearly because many species are flexible in their territo- animal can use, and any additional resources are not nec-
rial behavior, defending territories under some conditions essary. The exact shape and placement of the benefit
but not others (e.g., Lott 1991). By observing animals curve depends on resource quality and distribution. For
under different conditions, or even manipulating the con- example, if resources are plentiful, the curve might be
ditions experimentally, we can identify exactly what leads more like the top curve rather than the bottom. As you
them to hold territories or give them up. Let’s look at a can see, costs of defense also increase with territory size.
few variables that have been studied. Larger territories mean a longer border to patrol and
 Conflict Over Space 417

Territory size (number of flowers)

3000

2000

1000

Weight gain (grams)

0.3

0.2
FIGURE 18.7 The hypothetical relationship between the
costs and benefits of territoriality. Both costs and bene- 0.1
fits increase with territory size. The shape of the bene-
fits curve varies with the quality of the territory. It is
1 2 3 4 5
profitable to defend the territory as long as the benefits
Day
exceed the costs, between points A and B. If the animal
is to maximize its net gain, the optimal territory size is FIGURE 18.8 The relationship between territory size
at point X or X ¿, depending on the exact placement of and weight gain for one rufous hummingbird. It is
the cost and benefit curves. important for these birds to gain weight maximally dur-
ing their stopovers along their migratory route. These
more intruders to drive off. Territories will only be worth data indicate the weight gained by a single territorial
defending when benefits exceed costs. The optimal ter- bird on five successive days. This individual adjusted its
ritory size is the size at which the benefits outweigh the territory size so that its weight gain was maximal.
costs by the greatest amount, marked by the dashed lines (From Carpenter et al. 1983.)
on the graph (Davies and Houston 1984).
cutting the energy costs of defense, and maximizing its
STOP AND THINK weight gain (Carpenter et al.1983).
What factors might affect the shape and placement of the The exact shape of the benefit and cost curves in
cost curve in Figure 18.7? In general, do you expect the cost Figure 18.7 and their relative placement will vary among
and benefit curves will vary together, or independently? species and environments. The precise placements of
these curves may alter or even reverse predictions of
We might predict, therefore, that in some species optimal territory size (Schoener 1983).
the size of an individual’s territory would be adjusted to
maximize energy gains. At least some individual rufous STRATEGIES FOR REDUCING THE
hummingbirds (Selasphorus rufus) appear to do this.
COST OF TERRITORIAL DEFENSE
During their southward migration, these birds pause for
a few days in the mountain meadows of California to Territory holders can reduce their defense costs in a
build the fat reserves needed to fuel the next leg of their number of ways. One way is to band together and share
journey. During this interval they feed on the nectar of a territory. We’ll explore this issue a bit more in the next
the flowers of the Indian paintbrush. Each bird defends chapter.
a group of flowers as a territory. The territory size and Animals can also reduce their defense costs by
weight gain for a single individual are shown in Figure selecting their territories wisely, as Perri Eason and her
18.8. As you can see, this bird adjusted the size of its ter- collaborators have shown in a variety of taxa. In general,
ritory so that it could gain weight as quickly as possible. the faster an intruder is detected, the easier it is for a res-
This individual began with a small territory that con- ident to drive it away. Territories that include good van-
tained few flowers, so its weight gain was minimal. It tage points, such as high trees, can improve the resident’s
increased the territory size greatly on the third day. This chances of detecting an intruder quickly. Conversely, ter-
territory had more flowers and the bird could obtain ritories with obstacles can decrease the ability to detect
more energy, but it had to invest more energy in defense. intruders. Eason and Stamps (1992) tested the effect of
Nonetheless, it gained somewhat more weight than pos- visibility on territory settlement in juvenile Anolis
sible on the smaller territory. It reduced the size of its lizards. Lizards were released into two types of habitats
territory slightly on the fourth and fifth days, thereby in the laboratory, identical except for the presence of a
418 Chapter 18 / Conflict

visual barrier. Lizards in habitats without a barrier sive, in both time and energy, to renegotiate a territor-
defended compact territories similar to those in the field. ial boundary time and again, and selecting a territory
Lizards in habitats with a barrier avoided including the with an obvious boundary can save trouble. (A look at
barrier within their territories, adjusting the shape of maps of land ownership in humans also makes this
their territories dramatically. Some birds have also been point—features such as ponds, rivers, or roads often
shown to pay attention to visibility as they choose their serve as property boundaries.) The use of boundary
territories. For instance, red-capped cardinals (Paroaria markers was nicely demonstrated in cicada killer wasps
gularis) defend territories along rivers and lakes in Peru. (Sphecius speciosus) (Figure 18.9a). The larvae of these
They feed on insects that they find in the vegetation. wasps mature underground, and then adults emerge.
Strangely, although they do not feed over the water, they Male wasps mature slightly before females and defend
defend territories consisting of two parts on opposite areas where females are likely to emerge. Eason et al.
shores. The benefit of this strategy is that they can more (1999) studied the behavior of males on a grassy, mowed,
easily see intruders on the opposite shore than on the featureless lawn. The researchers began by mapping the
same shore and are more likely to detect and evict them territories of the males by watching them patrol their
immediately (Eason 1992). In addition, cardinals spend boundaries (Figure 18.9b). Then they added landmark
more time in areas of their territories that provide good cues: dowels (wooden sticks) laid flat on the grass so that
visibility (Eason and Stamps 2001). they did not align with any territory boundary (Figure
The costs of territorial defense can be reduced if the 18.9c). These dowels provided visual cues only, and not
territory has natural boundary markers. It can be expen- perching sites. The next day, all of the wasps had shifted

a b C D E F G H I

c C D E F G H I d C D E F G H I

9 9

8 8

7 7

6 6

5 5

4 4

3 3

FIGURE 18.9 (a) An adult cicada killer wasp. (b) The original placement of the territories. (c) Wooden dowels were
laid on the ground so that none of them aligned with territory boundaries. (d) The next day, the wasps shifted their
territory boundaries so that they aligned with the dowels. (Eason et al. 1999.)
 A Proximate View of Conflict 419

genetic basis (Chapter 4) and that hormones influence

aggression (Chapter 7). Here we will revisit several
Non- concepts in the context of some interesting work on
landmark
neighbor aggression. This is a large literature, and our hope is
to entice you to explore it further (e.g., see texts by
Adkins-Regan 2005; Nelson 2005) rather than to attempt
a complete coverage.

Landmark Focal Landmark AGGRESSION AND TESTOSTERONE

neighbor wasp neighbor
One of the best known examples of the influence of a
hormone on aggressive behavior is that of testosterone.
In Chapter 7, we discussed several examples from birds
that illustrate that point. Let’s begin with an example
from another taxon, lizards, which illustrates the nega-
Non- tive as well as the positive aspects of having high testos-
landmark
neighbor terone. In male spiny lizards (Scleroporus jarrovi), seasonal
changes in testosterone concentrations are tightly cor-
related with the intensity of territorial aggression
FIGURE 18.10 In this experiment, pairs of dowels were
(Moore and Marler 1987). In winter, testosterone levels
placed parallel to another. Wasps established territories are low and territorial defense is lethargic at best. When
between the dowels. Thus, each experimental territory males were implanted with testosterone-laden capsules
had two borders that were defined by a landmark, and in winter, they became very aggressive, spending so
two that were not. Defense costs were lower at the much time in territorial defense and vigorous displays
borders defined by a landmark. that they failed to eat or rest enough, and died sooner
than other males (Marler and Moore 1988, 1991; Marler
et al. 1995).
Let’s return to birds for another example that illus-
their territories so that the boundaries aligned with the trates how testosterone affects several behaviors. Ellen
dowels (Figure 18.9d). The benefits of this shift were Ketterson and her colleagues have many years of data on
documented in a second experiment. Two dowels were dark-eyed juncos (Junco hyemalis), a common visitor to
laid down parallel to one another, and wasps established bird feeders in North America. As in other species,
territories between them. Thus, these more-or-less testosterone fluctuates quickly in response to changing
square-shaped territories had two boundaries defined by social situations. It is thought that it is advantageous to
landmarks and two boundaries that had no landmarks be able to raise and lower testosterone levels as needed
(Figure 18.10). Focal wasps fought longer and more to avoid carrying the costs of constantly high levels. To
frequently with nonlandmark neighbors than with land- examine the ability of individual juncos to increase
mark neighbors. testosterone levels, researchers injected them with
Animals may also reduce the cost of territorial GnRH, or gonadotrophin-releasing hormone, which
defense by paying attention to the early-warning system stimulates the production of transient (short term)
provided by neighboring territory holders. The Peruvian increases in testosterone. The peak testosterone levels
red-capped cardinals again provide an example. When a produced in response to this “GnRH challenge” pre-
cardinal detects an intruder, it evicts it with a flurry of dicted how strongly juncos responded to a simulated
activity, including chasing and calling. This raucous intruder on their territory. However, there is a trade-off:
behavior alerts neighbors that there is an intruder lurk- birds that had the highest response to the GnRH chal-
ing nearby. Territory holders are more likely to imme- lenge were slackers when it came to delivering food to
diately detect and evict an intruder if their neighbor has their nestlings (McGlothlin et al. 2007).
just evicted it (Eason and Stamps 1993). So, is testosterone’s main function to make an indi-
vidual more aggressive and thus better able to fend off
an intruder? Interestingly, in nearly all species exam-
A PROXIMATE VIEW ined, testosterone increases after there is an aggressive
OF CONFLICT response to an intruder (reviewed in Wingfield 2005).
This suggests that testosterone allows a territory
In previous chapters, we’ve looked in detail at the prox- holder to maintain its high aggression levels, especially
imate mechanisms underlying many behaviors. We’ve in the face of a persistent intruder (Wingfield
already seen that aggressive tendencies can have a et al. 1990).
420 Chapter 18 / Conflict

STRESS, AGGRESSION, AND uations. Increased heart rate, rapid breathing, and sweat-
DOMINANCE ing are all part of the rapidly induced fight-or-flight
response that prepares the body to take immediate action.
The experiments just described indicate that there is a In addition, digestion, growth, and reproduction—all
link between aggression and testosterone. However, we useful functions but not of utmost importance when you
are not yet done unraveling the proximate control of are about to be attacked—are shut down. These
aggression. Aggressive behavior is regulated by both the responses are under the control of short-term bursts of
endocrine system and the nervous system. That straight- glucocorticoids. In addition, long-term, baseline patterns
forward sentence glosses over a complicated truth: hor- of glucocorticoid levels, present even before an animal
monal and nervous control are interconnected with fights, influence its predisposition toward aggression.
many feedback loops. It doesn’t help matters that the Summers and Winberg (2006) suggest a model to
endocrine system and the nervous system are themselves describe what generally happens to serotonin and glu-
sometimes hard to distinguish. For example, the brain cocorticoids over the course of a fight, especially among
not only responds to hormones, but also produces them. fish and lizards, shown in Figure 18.11 (note that testos-
In addition, the presence of a hormone does not neces- terone also primes aggression, but here we are omitting
sarily mean a particular behavior will follow: an animal it for simplicity’s sake). First, baseline levels of serotonin
that is primed to fight by testosterone will probably not and glucocorticoids establish an animal’s tendency to be
initiate a battle if a predator is nearby. As Adkins-Regan aggressive: in dominant animals, glucocorticoids are high
(2005) says, hormones are not like light switches turn- and serotonin is low, while the reverse is true in subor-
ing light bulbs on and off, but rather regulate or prime dinate animals. At this stage, serotonin’s function is to
(increase the likelihood of) a behavior. hold aggressive tendencies in check. However, the rela-
To illustrate this complexity, we’ll briefly describe tionship between serotonin and aggression is not that
another area of intense research: the relationship among simple. When a social interaction begins, and signals are
stress, aggression, and dominance. We are now begin- first exchanged, dominant individuals show an increase
ning to understand the neurochemical and hormonal in glucocorticoids, serotonin, and another neurotrans-
events that occur over the course of an aggressive inter- mitter called dopamine. As the fight increases in inten-
action. Interestingly, in vertebrates, there is a tight rela- sity and becomes more stressful, both dominant and
tionship between the nervous and hormonal control of subordinate animals show an increase in serotonin and
aggressive behavior, and the stress response. In particu- corticosterone. When dominance is established and the
lar, aggression and stress are linked by the neurotrans- fight ends, serotonin remains high in subordinate ani-
mitter serotonin and by the hormones called mals, which are then less likely to initiate fights. This
glucocorticoids (corticosterone or cortisol, depending on may remind you of the winner effect we discussed ear-
the species), made by the adrenal cortex. lier in the chapter: winners are more likely to win sub-
You doubtless know from experience that physio- sequent fights, and losers are more likely to lose them.
logical changes accompany conflict and other stressful sit- This is a mechanism by which this pattern may occur.

Social interaction Aggressive Social rank

begins interaction established

Before
interaction
Variable Seconds Minutes to hours Hours, days, weeks, months
Predispositions Dominant: Stress responses in both: Chronically elevated serotonin
toward rapid increases rapid increases in cort, and cort in subordinate animal
aggression: in cort, serotonin, serotonin, dopamine
dopamine
Dominant:
cort high
serotonin low FIGURE 18.11 A model of the possible interactions between glucocorticoids, serotonin,
Subordinate:
and dopamine over the course of an interaction. Before the interaction begins, dominant
serotonin high individuals are more likely to have higher levels of corticosterone (abbreviated cort), and
cort low lower levels of serotonin. This predisposes dominant individuals to behave aggressively.
During the interaction, both dominant and subordinate individuals show a stress response,
and both cort and serotonin increase. After the interaction, chronic increases in serotonin
and cort inhibit aggression. (From Summers and Winberg 2006.)
 Summary 421

So is being subordinate always more stressful than essary to evaluate the selection pressures on dominant
being dominant? Not necessarily. In fact, the relation- and subordinate animals.
ship between status and stress, as measured by gluco-
corticoid levels, varies across species. We now have data
on glucocorticoid levels in several species of wild ani- SUMMARY
mals. These data can be quite tricky to collect! Here is
Robert Sapolsky (1994) on the challenges of darting Conflict is a part of the lives of many animals, from sea
baboons in the wild: anemones to primates. Conflict occurs when resources,
such as mates, shelter, or food, are limited. Terms used
You have to dart every animal at the same time of
in the study of conflict include aggression, which is
day to control for daily fluctuations in hormone lev-
behavior intended to inflict noxious stimulation or
els. If you want to get a first blood sample in which
destruction on another organism (which, according to
hormone levels reflect basal, nonstressed conditions,
some authors, includes predation), and agonistic behav-
you can’t dart someone who is sick or injured or who
ior, which refers to the behavior of both the aggressor
has had a fight or intercourse that day. . . . If you
and the object of the aggression.
are trying to measure resting hormone levels, you
Animals generally do not engage in full-out fights to
can’t spend all morning making the same animal
the limit of their abilities. Instead, as we have seen in pre-
nervous as you repeatedly try to dart him; instead
vious chapters, fights are often limited to displays and
you get one shot, and you can’t let him see it coming.
other low-cost actions. Researchers have used a type of
Finally, once you dart him, you have to obtain the
modeling called game theory in order to understand
first blood sample rapidly, before hormone levels
when conflicts escalate and when they do not. Like the
change in response to the dart. Quite a thing to do
foraging models we studied in Chapter 12, we use game-
with your college education.
theory models to evaluate the relative fitness benefits
(Luckily, we now have less invasive ways of mea- that animals might acquire by pursuing different behav-
suring hormone levels using feces and hair, so you can ioral strategies. Game-theory models differ from forag-
become a behavioral endocrinologist even if your aim ing models because the payoff to playing a particular
is poor.) strategy depends on what other animals are doing.
Sapolsky’s (1992) long-term field studies show that One of the simplest game-theory models of conflict
subordinate baboons have higher levels of glucocorti- is the hawk–dove model. Here, animals can play one of
coids, but that glucocorticoids in dominant animals also two strategies. The hawk strategy is the strategy of esca-
spike when dominance hierarchies are unstable. lation: immediately attack the opponent. The dove strat-
However, the pattern of glucorticoids across dominant egy is to display if one meets another dove, but to flee if
and subordinate individuals varies across species. In encountering a hawk. The payoff of each strategy
recent reviews of field studies of several vertebrate taxa, depends on the value of the resource, the cost of being
including birds, primates, canids, and others, glucocor- wounded in a fight, and the cost of display. If the value
ticoid levels were sometimes positively correlated, neg- of the resource is greater than the cost of being
atively correlated, or uncorrelated with rank (Creel wounded, then a pure hawk strategy is an evolutionarily
2005; Muller and Wrangham 2004; Sands and Creel stable strategy: that is, a population of hawks cannot be
2004), or even varied in their correlation with rank invaded by a dove. This situation occurs in species where
throughout the day (Muller and Wrangham 2004). A most or all of an animal’s lifetime reproductive success
comparative study of seven primate species showed that depends on the outcome of a fight, such as in elephant
cortisol level varied according to the social situation: seals fighting over access to a harem of females. More
subordinates showed higher levels when they did not commonly, the value of the resource is less than the cost
have close social support (Abbot et al. 2003). Currently, of being wounded. In that situation, the evolutionarily
researchers are collecting data on dominance rank and stable strategy is a mixture of hawk and dove strategies.
hormone levels for an even wider variety of species, so It is possible to calculate what that mixture is for partic-
perhaps we will better understand this relationship in ular values.
the next few years. In the basic hawk–dove game, it is assumed that all
Stress responses are adaptive in the short term, but players are equal and can choose either strategy.
persistent, long-term stress responses can lead to health However, this is not often the case. Instead, contests are
problems (reviewed in Sapolsky et al. 2000; Sands and often asymmetrical. Sometimes one animal is a stronger
Creel 2004), such as increased susceptibility to disease fighter than the other—larger, heavier, more aggressive.
and a shorter lifespan. Thus, understanding the rela- Experience can also create asymmetries: animals that
tionship between rank and long-term stress will be nec- have won a fight are more likely to win their next fight.
422 Chapter 18 / Conflict

Resources may also differ in value to the contests: a hun- knowledge of the location of resources) and “free” (not
gry animal will value food more, and thus fight harder prevented from going where they wish by other animals).
for it, than a well-fed animal. Finally, arbitrary asym- In contrast, territory holders prevent others from settling
metries are not correlated with either the value of the in their space. It is not always economically feasible to
resource or the ability of an animal to defend it, but are defend a territory: borders can be energetically costly
simply conventions that fighters follow. An example of and even dangerous to patrol. Territorial defense is
an arbitrary asymmetry is “the current owner of the favored when resources are moderately abundant (too
resource wins the fight.” An example with speckled wood scarce and the territory would need to be too large; too
butterflies illustrates how difficult it can be to ascertain abundant and there is no need to defend it), when
the rules of fights. resources are moderately clumped (too clumped and too
Next we turn to a special case of conflict: conflict many intruders would flood in to try to wrest them away;
among group members. Relationships develop among too dispersed would mean boundaries are too long), and
group members. Often some group members are dom- when the number of intruders is not too high. A similar
inant over others and routinely assert themselves in logic can be applied to the optimal size of territories.
access to resources. Submissive animals are those that Territory holders can reduce the cost of defense by
predictably yield to dominant ones. The structure of selecting territories with good visibility so that intrud-
dominance relationships can be very simple straight-line ers can be easily spotted, with landmarks that help
hierarchies, with A dominant over B, who is dominant to define boundaries, and by attending to the early-
over C, and so on, or they can be more complex. warning system provided by neighbors that alert them
Dominance may be determined by fights or, in some to the presence of intruders.
cases, as a birthright. Dominant animals often get more Proximate causes of aggression have been discussed
resources and access to mates than do subordinate ani- in other chapters. Here, we describe the relationship
mals. Subordinate animals often stay in a group because between dominance, stress, and neurotransmitters and
no other options are available. However, subordinates hormones, and we describe one proposed model for the
may occasionally win a fight, engage in sneaky alterna- hormonal pattern that occurs during an aggressive
tive strategies, or band together with others. interaction. However, evidence is accumulating that not
The second special case that has received a great all species respond in the same way. In some species,
deal of attention is conflict over space. Territorial species dominant animals appear to exhibit the highest levels
defend space from one another (or, by another defini- of stress, and in others, subordinate animals do. Finally,
tion, have exclusive use of space). The ideal free distri- we look at an example of a species in which hormone
bution describes how animals should distribute levels and aggressiveness in territorial interactions are
themselves in space if they are “ideal” (with perfect intertwined.
 19
Group Living, Altruism, and Cooperation

Living in Groups: From Aggregations crawl inside rock crevices where they are sheltered from
to Structured Societies cold. When houses are available, they can find their way
Benefits of Group Living inside through vents and cracks. The species plaguing
Costs of Living in Groups homes in eastern North America is a particularly unwel-
Balancing Costs and Benefits come guest. Originally introduced from Asia in order to
control invasive pests, Harmonia axyridi has become
The Puzzle of Altruism
wildly successful and its numbers have soared. In some
Individual Selection and “Altruism”
years, like this past one, H. axyridi find their way into
Kin Selection houses in such numbers that they accumulate in drifts on
Reciprocal Altruism the windowsills, flying around on warmer days and star-
Manipulation tling the unwary. They have more serious consequences
Examples of Cooperation Among Animals too, as they outcompete and even eat native ladybug
Alarm Calls species, and sometimes cause allergic reactions in their
Cooperation in Acquiring a Mate human hosts (Goetz 2008; Snyder et al. 2004).
Cooperative Breeding and Helping If we array animal groups from simple to complex,
Eusociality
ladybug groups are at the “simpler” end of the spectrum.
Their groups are seasonal rather than lifelong—the lady-
bugs will disperse in spring and forage and reproduce on
their own. There’s no evidence that ladybugs interact in
LIVING IN GROUPS: FROM complex ways, recognize each other as individuals, or
AGGREGATIONS TO cooperate, although they seem to be mutually attracted
STRUCTURED SOCIETIES to one another via chemical cues. Some animal species
that form groups do not even show mutual attraction, but
Brushing her teeth one night in her Massachusetts home simply respond to the same features of the physical envi-
in the dead of a snowy winter, one of the authors was ronment and thus independently end up in the same spot.
smacked in the side of the head with . . . a ladybug? Sure Now consider this example. Two adult male chim-
enough. Some species of ladybug beetles overwinter in panzees have a noisy, aggressive conflict. The challenged
aggregations of thousands of individuals. In nature, they male flees into a tree. The opponent slowly approaches,

423
424 Chapter 19 / Group Living, Altruism, and Cooperation

FIGURE 19.1 A chimpanzee offers an

open hand to another chimp with
whom he has just had a conflict.

offering him an open hand (Figure 19.1). Seconds later, States, Harris’s hawks (Parabuteo unicinctus) live and hunt
the chimpanzees touch, kiss, and climb to the ground to in family groups (Bednarz 1988). In the early morning,
groom one another (de Waal 2005). Later, it is likely that family members typically gather at one perch site, from
these chimpanzees will reciprocally exchange favors, which the group then splits into smaller subgroups of
such as food for grooming. Chimpanzees, and other one to three individuals. The subgroups take turns mak-
species, are members of structured groups called soci- ing short flights and “leapfrog” through their family’s
eties. Interaction with conspecifics is deeply ingrained area. Upon discovering a rabbit, hawks employ one of
in every part of life, so much so that Roger Fouts said, three hunting tactics, or more typically a combination of
“One chimpanzee is no chimpanzee.” the three. The most common, the surprise pounce,
Not all animals live in groups, but group living occurs when several hawks arrive from different direc-
occurs in nearly every major taxon. In this chapter, we’ll tions and converge on a cottontail or jackrabbit unfor-
first explore some selective advantages—and disadvan- tunate enough to be out in the open. Even if the rabbit
tages—of living in groups. Next we’ll discuss a particu- escapes under vegetation, however, safety may only be
larly interesting aspect of many animal interactions temporary. At this point, the hawks employ their flush-
called altruism, or helping behavior. Finally, we’ll select and-ambush tactic, a strategy in which one or two hawks
some well-studied examples of animal cooperation to flush the rabbit from the cover, and then family mem-
explore in more detail. bers perching nearby pounce on it. Relay attack is the
third, and least common, hunting tactic. Here, family
members constantly chase the prey, with a new lead bird
BENEFITS OF GROUP LIVING taking over each time there is a missed attempt to kill.
What fitness benefits accrue to animals living in groups? Regardless of the tactic used, the dead prey is shared by
Several themes recur in many taxa: improved foraging, all members of the hunting party.
decreased risk of predation, conservation of water and In order for cooperative hunting to be favored by
heat, and decreased energetic costs of movement selection, individuals must average at least the same
(reviewed in Krause and Ruxton 2002, whose organiza- amount of food as they would get by hunting alone.
tion guides us here). Not every species profits in each Because the food is generally shared among all partici-
way outlined below, but many do. pants in the hunt, this must mean that hunting success
must be increased a great deal in groups. In Harris’s
hawks, the chance of killing a prey correlates with group
Improved Foraging size—hunting parties of five to six individuals do better
Foraging success in groups can be improved through than smaller parties (Figure 19.2a). Even with more indi-
several mechanisms. For example, some animals coordi- viduals sharing the prey, the average energy intake per
nate their foraging, sometimes with great precision, in individual from rabbit kills is higher in groups of five or
cooperative hunting. In the southwestern United six members than in smaller groups (Figure 19.2b).
 Living in Groups: From Aggregations to Structured Societies 425

a 4 FIGURE 19.2 In Harris’s hawks, members

of family groups engage in cooperative
hunting. (a) Groups of five or six individu-
3
als are most successful at killing cotton-
jackrabbit kills
Cottontail and

tails and jackrabbits, and (b) individuals in

2 such groups have a higher average energy
intake than individuals in smaller groups,
1 and are more likely to meet their basic
energetic needs. (From Bednarz 1988.)
0

Energetic
b 150
(kcal per hawk per day)

maintenance

100
Energy

50

0
2-3 4 5 6
Group size

Another means by which animals improve their for- can take advantage of the “confusion effect” and decrease
aging while in groups is by information sharing. Many a predator’s ability to track and kill any one individual.
species pay attention when their conspecifics discover And finally, group members can band together to drive
food, and use this information to guide their own forag- a predator away by mobbing.
ing. For example, geese are more likely to land near arti-
ficial geese that have their heads down in a feeding
position than to land near those standing erect (Drent Conserving Heat and Water
and Swierstra 1977). By huddling together, animals can reduce the surface
Communal roosts or colonies can act as informa- area exposed to the environment and thus reduce the loss
tion centers. If successful foragers return to the roost of heat or water. For instance, many mammals sleep or
or colony, and then return to the food site, others might overwinter together in communal burrows. Many birds
be able to identify them as successful foragers and fol- perch snuggled up next to one another when they sleep.
low them. Scientists debate how common this phenom- A male emperor penguin, balancing its egg on top of its
enon is (Krause and Ruxton 2002), but it has been feet and tucked under its belly, could not survive the long
demonstrated in some species, including cliff swallows Antarctic winter without huddling with other males. The
(Brown 1986) and the honeybees that we discussed metabolic rates of penguins in small groups are reduced
in Chapter 16. by 39% compared to isolated birds, and those of pen-
guins in larger groups are reduced by another 21%
(Gilbert et al. 2008). Even some not-very-cuddly animals
Decreasing Predation Risk huddle: some slugs rest in contact with one another in
Recall from our discussion of antipredator behavior in order to reduce water loss (Cook 1981).
Chapter 13 that group membership allows animals to
employ several antipredator tactics that are not available
to solitary individuals. To recap briefly: groups may be Conserving Energy by Moving Together
more likely than a single individual to detect a predator Bicyclists know the value of drafting, or riding close
(the “many eyes” hypothesis). Animals in groups often behind in the slipstream of another bike: it reduces the
give alarm signals when they detect a predator. (We’ll amount of energy needed to pedal. Researchers have
examine alarm signals in more detail later in the chap- explored whether the same holds true for animal groups
ter.) When a group is detected and attacked, an individ- that travel together, such as schools of fish and flocks of
ual within a group has a smaller chance of becoming the birds flying in formation (reviewed in Krause and Ruxton
next victim (the dilution effect), and animals in the cen- 2002). It can be technically tricky to get these measure-
ter of the group often have a lower chance of being ments and to tease apart the role of energetic savings
preyed upon than do animals at the edge (the selfish from other potential benefits of traveling in a group,
herd). By fleeing in different directions, group members such as avoiding predators or using conspecifics as
426 Chapter 19 / Group Living, Altruism, and Cooperation

retracting their nets more quickly when a conspecific is

nearby (Gagern et al. 2008).
Group-living animals might often lose food to
thieves—in fact, stealing can be considered a strategy.
Giraldeau and his colleagues (e.g., Giraldeau et al. 1994)
have extensively modeled and tested producer–scrounger
wing beat frequency

105 situations in which an individual might either look for

90 food (and be a “producer”) or steal food that others have
(beats/min)

75 found (a “scrounger”). As you might predict, there is a

Mean

60 1 limit on how many “scroungers” a population can sup-

45 2 port before the strategy becomes unrewarding.
3 4 5 Last
12
0
Increased Risk of Disease and Parasites
heart rate (beats/min)

200
190 As anyone who has been in a crowded classroom during
180 cold and flu season can attest, another potential risk of
Mean

170 social living is increased exposure to disease and parasites.

160 For example, cliff swallows nest in large colonies, and
150 colony size is correlated with the number of blood-
Gliding Alone at Alone at In sucking swallow bugs (Oeciacus vicarius). Swallow bugs are
50 m 1m formation harmful to swallows in many ways: they reduce nestling
mass, decrease survival of birds of all ages, increase white
FIGURE 19.3 Heart rates of pelicans flying in formation
versus those flying solo. Compare pelicans that are glid- cell counts, and even cause asymmetrical growth of feath-
ing, flying solo at two different heights above the water, ers (Brown and Brown 1986, 2002, 2004; Brown et
or in formation. The numbers next to the circles indi- al. 1995). Colony size is correlated with higher levels of
cate the position of the birds in the group, with number glucocorticoid hormones, which we have already seen are
1 being the leader. (From Weimerskirch et al. 2001.) released in response to stress (Chapter 18). Raouf et al.
(2006) hypothesized that the stress response was related
to the presence of the swallow bugs rather than to other
navigational cues. To illustrate what a daunting task this aspects of group living, such as competition for food. The
can be, consider Weimerskirch et al.’s (2001) study of researchers tested this hypothesis by fumigating some
why pelicans (Pelecanus onocrotalus) fly in a V formation. colonies in order to kill the parasites. Fumigation low-
The effort expended by birds in flight can be measured ered corticosteroid levels, indicating that exposure to the
by strapping heart monitors to their backs. But how to parasites caused the stress response.
keep pelicans that are so equipped from simply flying Many group-living species have evolved behaviors
away? The researchers trained eight pelicans to fly after that help fend off disease and parasites. Allogrooming
a moving motor boat and an ultralight airplane. They animals lick each other and pick off ectoparasites, often
could thus measure the heart rates of pelicans in a vari- focusing on those hard-to-reach places. Social insects,
ety of group sizes. Birds flying in formation had lower such as honeybees, ants, and termites, have several
heart rates than did birds flying solo (Figure 19.3). defenses against disease. They remove corpses and other
waste from the colony, and they even wall off or remove
infected individuals (Cremer et al. 2007). In fact, leaf-
COSTS OF LIVING IN GROUPS cutting ants are so effective in defending against a fatal
Increased Competition fungus via grooming and antibiotic secretions that group
living confers a net benefit in resisting disease (Hughes
As you might expect, individuals that live in groups often
et al. 2002). For most species, however, proximity to con-
compete with each other for mates, nest sites, or food.
specifics presents an increased risk of infection.
Consider the snail Dendropoma maxima, which forages in
a slightly disgusting way: it secretes a sticky mucous net
that floats in the water and traps plankton. The snail then Interference With Reproduction
draws the net back in, eating both the mucus and its Animals in groups sometimes face interference with their
catch. When snails are grouped together, their mucous own reproductive efforts. For instance, extra-pair copu-
nets frequently overlap and stick together, so that snails lations—copulations that occur outside of the pair
end up consuming the nets of their neighbors. The snail bond—are extremely common, even among “monoga-
that is slow to retract its net loses food—and valuable mous” birds. Animals that live in groups have more
slime!—to its neighbor, and thus grows more slowly. opportunities for such dalliances than animals that do
Snails seem to adjust for the presence of neighbors by not. Thus, group-living males may be more likely than
 The Puzzle of Altruism 427

males of solitary species to invest resources in caring for

young that are not theirs. Similarly, in the hustle and
bustle of a large group, such as a colony of nesting birds,
a female may fail to notice the arrival of another female
at her nest and the quick deposit of an egg by the
intruder. (For more extensive discussion of intraspecific
brood parasitism, see Chapter 15.)
Mammalian mothers of 68 species have been docu-
mented to allonurse, or nurse offspring that are not their
own (reviewed in Roulin 2002). Allonursing is common
in animals that share roosts and reproduce communally,
or where all the females reproduce in a small space (such
as seals confined to suitable rocky ledges). Several
hypotheses for the occurrence of allonursing have been FIGURE 19.5 Whirligig beetles aggregate on the surface
of a pond. Those on the outside of the group get more
proposed, some of which include fitness benefits to the
food but are also more vulnerable to predators, such as
nursing mother (e.g., feeding young that are related, or
the fish pictured here. In this experiment, beetles were
feeding the offspring of a mother that will later recipro- confined to a ring.
cate). Another hypothesis is that nursing the young of
others is costly misdirected parental care. For example,
Mexican free-tailed bats spend their days in colonies that other prides. Group living also appears to be important
exceed several million individuals, and it is possible that in protecting cubs from nomadic males that commit
allonursing occurs primarily because of mistaken iden- infanticide and in the defense of the pride’s home area
tity (McCracken and Gustin 1991). against intrusion by neighboring prides. Thus, for lions,
the benefits of group living include the defense of food,
young, and living areas against conspecifics.
BALANCING COSTS AND BENEFITS The costs and benefits of group living may differ
Of course, any particular species may face a combination across individuals. For instance, the aptly named
of these costs and benefits. Although ideas on patterns whirligig beetles form aggregations on the surface of
of grouping in lions initially focused on hunting success freshwater, each spinning in circles (Figure 19.5). Beetles
(Figure 19.4), data suggest that this factor alone cannot at the outside edge of the group are more likely to get
explain the formation of prides in this species (Packer bits of food dropped into the water, but are also more
1986; Packer et al. 1990). Scavenging may be a more effi- likely to get attacked by fish (Romey 1995; Romey et al.
cient means of obtaining food than hunting, and groups 2008). How beetles weight the trade-off between preda-
may be necessary to defend carcasses against lions from tion risk and food availability depends on how hungry
they are: hungry beetles move to the outside of the
group, in spite of the risk (Romey 1995).

THE PUZZLE OF ALTRUISM

By now, you should be very comfortable with the idea
that natural selection will cause a particular behavior to
increase in a population if its fitness benefits outweigh
its costs. What are we to make of the fact, then, that
some animals seem to help other members of their
species? For example, a Belding’s ground squirrel
(Spermophilus beldingi) increases its risk of being spotted
by an approaching predator when it barks an alarm.
Nevertheless, it barks at the sight of a badger, and all
those in the area scurry to safety (Sherman 1977). In wild
turkeys (Meleagris gallopavo), males form coalitions that
court females and defend them against other groups
FIGURE 19.4 A group of lions shares a wildebeest kill. (Figure 19.6). However, only dominant males father off-
Studies of group living in lions have often focused on spring: subordinate males expend effort in display, but
hunting success, but additional functions include the to no benefit to themselves (Krakauer 2005; Watts and
defense of cubs, space, and scavenged food. Stokes 1971). Members of still other species cooperate
428 Chapter 19 / Group Living, Altruism, and Cooperation

inspection it seems that the parents gain because the

presence of the adopted young reduce the risk that their
own young will be picked off by a predator (McKaye and
McKaye 1977). Thus, before assuming that a behavior
is truly altruistic, it is important to investigate carefully
whether the actor benefits directly from the behavior.

KIN SELECTION
In more puzzling cases, the costs of a behavior to the actor’s
survival and reproductive success outweigh any benefit it
might accrue. If an individual behaves so as to decrease its
own reproductive success, how could natural selection pos-
sibly act to increase the frequency of its altruistic trait?
FIGURE 19.6 Wild turkey brothers strut together in W. D. Hamilton’s groundbreaking paper in 1964
order to attract mates. In general, only the dominant revolutionized our thinking about altruism. Hamilton’s
brother mates. key insight was that individuals can improve their fitness
not only through their own offspring, also called their
descendant kin, but also through the reproductive suc-
in rearing offspring that do not belong to them. For cess of their other relatives, or nondescendant kin. Here
example, dwarf mongooses (Helogale parvula) bring food is the logic: family members other than offspring also
to the young of others and guard the den from preda- possess copies of some of the same alleles because they
tors. Perhaps the most striking examples of helping inherited the alleles from the same ancestor. Therefore,
behavior are found in eusocial insects (ants, termites, and if family members are assisted in a way that increases
some wasp and bee species). The workers toil tirelessly their reproductive success, the alleles that the altruist
to care for their colony. They may even die in defense has in common with them are also passed on, just as
of the nest. However, the young they help rear are not they would be if the altruist reproduced personally.
their own: the workers are sterile. Selection that works through relatives in this manner is
Animal behaviorists call examples such as these called kin selection.
altruism. An altruistic behavior appears to be costly to Of course, not all relatives have the same likelihood
the altruist and beneficial to another member of its of sharing a particular allele. Intuitively, you can imagine
species. Although it is difficult to measure, biologists that closer relatives, such as siblings, are more likely to
define altruism in terms of fitness: it is a behavior that share alleles than are more distant relatives, such as
raises the fitness (number of offspring produced that live cousins. We can be even more precise: we can calculate
to breed) of another individual at the expense of the the probability that particular pairs of relatives share the
altruist’s direct fitness, as measured by the number of off- same allele through common descent. This probability
spring it leaves (Hamilton 1964). is called the coefficient of relatedness, or r. You may
How could altruism possibly evolve? Shouldn’t recall from our discussions in Chapter 3 that sexually
alleles that promote selfish behavior multiply more reproducing animals have two alleles for each gene and
quickly in the population than alleles that promote that these separate during the formation of gametes
altruism? Not surprisingly, many researchers, including (eggs or sperm). There is, therefore, a 50–50 chance
Darwin, have puzzled over these questions. (a probability of 0.5) that any particular allele will be
found in an egg or a sperm produced by the parent. In
INDIVIDUAL SELECTION other words, an individual shares 50% of its alleles with
its parent, so a parent and offspring have a coefficient of
AND “ALTRUISM”
relatedness of 0.5. The value of r ranges from 0 (non-
In some cases, behavior appears to benefit others at the relatives) to 1 (identical twins or clones). (As an interest-
cost to the actor, but in fact, the actor might directly ben- ing aside, recent findings show that even identical twins
efit from its behavior. For example, an animal that gives might have small differences in DNA; Bruder et al. 2008.)
an alarm call may appear to be alerting others at its own An easy way to calculate the coefficient of related-
expense, but it may actually improve its own survival by ness between more distant relatives is by using a family
alerting the predator that it has been seen. Similarly, in tree, as illustrated in Figure 19.7. By this method, we
some species of cichlid fish, adults adopt unrelated young determine that, on average, an animal shares 50% (1/2)
into their own brood, caring for them and defending of its genes with a full sibling (r ⫽ 0.5); 25% (1/4) with
them from predators as if they were family. Although this a half sibling or grandparent (r ⫽ 0.25), and only 12.5%
may at first seem to be quite a generous act, on closer (1/8) with a first cousin (r ⫽ 0.125).
 The Puzzle of Altruism 429

a Half sibs FIGURE 19.7 A method of calculating

Mother One path with 2 steps:
relatedness (r) with a family tree. First,
r = (1/2)2
draw in the family members that connect
=1/4 two individuals whose relationship you
You Half sib
wish to know. Draw lines between all
parents and each of their offspring. For
b Full sibs Mother Father Two paths with 2 steps each: each path between the two individuals of
r = (1/2)2 + (1/2)2 interest, count n, or the number of steps,
= 1/2 and raise 1/2 to the power of n. Do this
You Full sib for each possible path, and sum the result
to get r. (a) Only one path with two steps
c Cousins connects half sibs. (b) Two paths (one
Grandmother Grandfather Two paths with 4 steps each:
through the father and one through the
r = (1/2)4 + (1/2)4 mother), each with two steps, connect full
= 1/8 sibs. (c) Two longer paths connect
Mother Father Uncle
cousins.

You Cousin

Now let us see how the coefficient of relatedness

affects altruism. When will the gene for an altruistic STOP AND THINK
behavior spread in a population? To figure this out, we A bird will have two offspring if she raises them without
can employ Hamilton’s rule: help. However, if her altruistic sister helps her by bring-
ing food to the nest and driving off predators, she will have
five offspring. This behavior has a cost to the potential
B 1 altruist: if she helps, she will not have any of her own off-
>
C r spring, whereas if she does not help, she will have one.
Should she help?
where
B is the benefit to the recipient in terms of extra off-
spring that the relative produces because of the altruis-
tic act; We are now prepared to expand the definition of fit-
C is the cost to the actor, measured as the number ness that we have been using throughout the book. Thus
of offspring it does not produce because of the altruistic far, we have primarily considered direct fitness, mea-
act; and sured by the number of offspring that an individual has
r is the coefficient of relatedness between the recip- as a result of its own efforts. Hamilton proposed that we
ient and the actor. Note that 1/r has a value of 1 or add to this another component of fitness, indirect fit-
greater, because r is a fraction ranging from 0–1. ness. To calculate indirect fitness, we count the number
To take an example: when should an individual forgo of extra offspring that an individual gains by helping a
reproduction to help its sister reproduce? For siblings, relative, devalued by the genetic distance between the
r ⫽ 0.5, so 1/r ⫽ 2. Therefore, in this example, the ben- individual and the relative who was helped (in other
efits of acting altruistically must outweigh the costs by words, multiplied by r). Inclusive fitness is the sum of
2:1 for an individual to help its sister reproduce. direct and indirect fitness.
Actually applying Hamilton’s rule to real animals is
quite tricky. We cannot just count the total number of
offspring that a relative produces: this will inflate the Discriminating Between Kin and Nonkin
value of helping. Instead, in order to calculate B in the If animals have the ability to aid each other, we expect
equation above, we must count up the number of extra that it should be evolutionarily advantageous to discrim-
offspring that are produced only because of the help of the inate kin from nonkin. Here, we’ll discuss four ways in
altruist. It is, of course, difficult to know what might have which animals may accomplish this.
been. However, sometimes we can estimate this quan-
tity by comparing the number of offspring produced by Location As you may recall from Chapter 11, individ-
animals that followed one course of action with the num- ual animals are often found in fairly predictable loca-
ber produced by those that did not. There are reasons tions. In mammals, for instance, males generally disperse
that this estimate may be inaccurate (Grafen 1984), but from home as they mature, whereas females tend to
it is a beginning. Later in the chapter, we’ll see an exam- remain in their natal area. Thus, female mammals that
ple of how this might be accomplished. help conspecifics located near their home are likely to be
430 Chapter 19 / Group Living, Altruism, and Cooperation

helping relatives, even if they do not have the ability to Familiarity If kin don’t always encounter each other
recognize them individually as kin. Simply following the in the same place, they must use some other method
rule of thumb of “If you are a female, help those near besides location to reliably identify one another. One
home” can lead to increased inclusive fitness. Thus, frequently used method is familiarity. Here, the young
when relatives are distributed predictably in the habitat, learn to recognize the individuals with which they are
kin selection can work if the altruistic deeds are directed raised through their experiences during early develop-
toward those individuals in areas where relatives are most ment, and then, later in life, they treat familiar and unfa-
likely to be found. miliar animals differently. The ideal setting for this
Another example of how location is used as a mech- learning is a rearing environment such as a nest or
anism for kin-biased behavior is when a parent identi- burrow that excludes unrelated individuals (reviewed
fies its offspring as those young in its nest or burrow. In in Mateo 2004).
many species of birds, parents will feed any young they Familiarity is apparently a mechanism used by a
find in their nest. We see this among bank swallows young spiny mouse when attempting to distinguish its
(Riparia riparia), which live in holes that they excavate siblings from strangers. When weanling pups are released
in banks (Figure 19.8). Each pair of swallows lives in its into a test arena, they often huddle together in pairs, and
own hole, but holes are clustered together. The parent the members of the pair are generally siblings. However,
bank swallow learns its nest-hole location and feeds any the spiny mice do not identify their siblings per se, but
chicks inside, including any neighbor’s chicks placed rather they prefer to huddle with familiar pups, their lit-
inside by experimenters. A parent will ignore its own termates. We know this because siblings separated soon
chicks if they are moved to a nearby nest hole. After after birth and raised apart treat one another as nonsib-
about two weeks, at the time the young begin to leave lings. However, if unrelated young are raised together,
the nest and fledglings unrelated to the parent might they respond to one another as siblings. So we see that
enter the nest, the parent begins to recognize its young kin-biased behavior among spiny mice seems to develop
by its distinctive calls and also begins to reject foreign as a result of familiarity (Porter et al. 1981).
young experimentally placed in the nest (Beecher et Like using location cues as a label of kinship, rely-
al. 1981). At this point, location is no longer a reliable ing on learning about characteristics of conspecifics does
cue to relatedness. Interestingly, in barn swallows, not suit every occasion. For example, if nonkin are
where chick intermingling is rare, there is no evidence encountered during the learning process, they may be
that parents can recognize the voices even of older mistakenly classified as kin (reviewed in Mateo 2004).
offspring (Beecher et al. 1989).
The trouble, of course, is that this very simple sys- Phenotype Matching Phenotype matching allows
tem for discriminating kin from nonkin breaks down animals to identify kin even if they have never met them
when animals are not where they are “supposed” to be. before (Alexander 1979; Holmes and Sherman 1982;
For kin discrimination to work in a variety of locations, Lacy and Sherman 1983). As a reminder, an animal’s
we need a different mechanism. phenotype is its physical, behavioral, and physiological
appearance (Chapter 3). As you know, family members
often resemble one another, so phenotypic similarities
can be a useful clue to relatedness.
Animals can learn the “kin phenotype” either by
learning about the phenotype of familiar individuals (in
informal shorthand, the equivalent to saying, “This is
what my brothers look like”) or by learning their own
phenotype (“This is what I look like”). This template is
then compared against strangers (“He looks like my
brother” or “He looks like me”). Of course, we don’t
need to hypothesize any astonishing cognitive powers in
order to invoke phenotype matching—animals may sim-
ply learn to respond favorably to a familiar cue that is
also exhibited by a stranger.
Several species have been shown to learn about the
phenotype of familiar individuals to develop a template,
and then match new individuals to that template. Let’s
look at an example that uses both familiarity and pheno-
type matching. As we will see in more detail shortly,
FIGURE 19.8 Colonial bank swallows will feed any young Belding’s ground squirrels (Spermophilus beldingi) give
they find in their nest. alarm calls to warn their mothers, daughters, and sisters
 The Puzzle of Altruism 431

of an approaching predator. Pups apparently identify group of alleles that enable it to recognize others with
their siblings because they learn one another’s odors the same allele(s). The postulated allele would have three
while still in the same nest burrow. When pups from dif- simultaneous effects: it would endow its bearer with a
ferent nests are experimentally switched, unrelated pups recognizable label, endow the bearer with the ability to
that are raised together will treat one another as siblings perceive that label in others, and cause the bearer to
later in life (Sherman et al. 1997). Experimental designs behave preferentially toward others with the label. This
like these, where infants are reared by unrelated foster recognition system has been named the “green beard
parents, are called “cross-fostering” experiments, and are effect” to indicate that the label could be any conspicu-
very useful in teasing apart which mechanisms of kin ous trait, such as a green beard, as long as the allele
recognition might be operating in a given species (Mateo responsible for it also causes its owner to behave appro-
and Holmes 2004). Later, juveniles and adults can use priately to other labeled individuals (Dawkins 1976,
phenotype matching to discriminate relatedness among 1982; Hamilton 1964). Demonstrating the existence of
individuals that they have never encountered before recognition alleles has proven quite difficult, primarily
(Holmes 1986a,b). because it is so hard to eliminate all the possible oppor-
A slight variation on this sort of system comes from tunities for learning recognition cues during an animal’s
paper wasps (Polistes fuscatus), which discriminate lifetime. In addition, theoretical models predict that the
between kin and nonkin by using the odor of hydrocar- high degree of genetic polymorphism required for this
bons that become locked into the insect’s cuticle before mechanism to function may be unstable and disappear
it hardens. The odor comes from the nest. Nest odor dif- (Gardner and West 2007; Rousset and Roze 2007).
fers among colonies because it depends on the type of Laurent Keller and Kenneth Ross (1998) may have
plant fiber used to build the nest, as well as on secretions identified a green beard allele in the red fire ant
produced by wasps that built the nest. Each colony uses (Solenopsis invincta). Originally from South America, the
a unique combination of plants to construct the nest, so fire ant is a recently introduced pest in the southern
each has a distinctive odor. The odoriferous hydrocar- United States. Its social organization is controlled by the
bons are transferred from the nest to the workers as they protein-encoding gene, Gp-9, with two alleles, B and b.
emerge from the pupal case. Even when a wasp meets a Workers that encounter individuals carrying the b allele
nestmate away from the nest, it can recognize it. Since form a template that they later use to determine
a colony consists of a queen and her worker daughters, whether to attack particular queens. Workers that came
the nest odor is generally a reliable label of colony mem- into contact with ants that bear b alleles when they were
bers as relatives (Breed 1998; Pfennig et al. 1983). forming their template accept only b-bearing queens
More challenging to demonstrate conclusively is the (bb and Bb), whereas workers that contacted only BB
second form of phenotype matching, in which an animal individuals during template formation accept only BB
uses its own phenotype as a template against which to queens (Gotzek and Ross 2007). Thus, kin discrimina-
compare strangers. This process is descriptively nick- tion in fire ants is very closely tied to genetics, but mem-
named the “armpit effect”—imagine an individual sniff- ory (in the formation of a template) also plays a role.
ing its own armpit and then that of a stranger. Several Insects are not the only animals with potential for
examples of self-referent phenotype matching have been genetically based recognition mechanisms. For example,
proposed. For example, Mateo and Johnston (2000) cross- a particular region of DNA called the major histocom-
fostered golden hamster females (Mesocricetus auratus) on patibility complex (MHC) may be important in recog-
the day of birth. Because only one hamster was transferred nition. The MHC region codes for molecules on the
into its new foster family, it had little opportunity to learn surface of cells that allow the body to distinguish
about the phenotypes of its relatives. Nevertheless, as between “self” and “nonself.” The MHC region can
adults, hamsters could discriminate between the odors of cause grafted tissue to be rejected and triggers protec-
unfamiliar kin and unfamilar nonkin. This finding sug- tive immune responses when disease-causing organisms
gests that they were using their own odor to form a tem- enter the body. These same genes may serve as direct
plate. It is still possible that hamsters acquired some cues of relatedness, allowing individuals to identify their
knowledge of kin in the brief period before they were kin (reviewed in Brown and Eklund 1994; Penn and
transferred (Hare et al. 2003), although Mateo and Potts 1999).
Johnston (2003) argue that this seems unlikely given the The larvae of the sea squirt Botryllus schlosseri use
developmental stage of their discrimination abilities. This MHC to discriminate kin. These larvae, which superfi-
work points out the challenge in controlling for every pos- cially resemble a frog tadpole, float in the water column
sible source of information available from kin. for a short time and then settle, attach to the sea bot-
tom, and develop to the adult form. When groups of sib-
Recognition Alleles Our fourth and final mechanism lings settle, they tend to clump together, but groups of
of kin discrimination is genetically based. In this mech- unrelated larvae settle randomly. If the siblings do not
anism, an individual inherits a “recognition allele” or share an allele in the MHC region, they do not settle
432 Chapter 19 / Group Living, Altruism, and Cooperation

together. Unrelated larvae that happen to share an allele may recall from Chapter 18 that evolutionary game
are just as likely to settle together as are siblings that theory is designed to handle situations such as this, in
share an allele (Grosberg and Quinn 1986). MHC has which the best course of action depends on what others
been reported to play a role in kin recognition in a vari- are doing.
ety of other taxa, including amphibians (Villinger and The name of the game comes from an imaginary
Waldman 2008) and mammals, including humans story in which two suspects are arrested for a crime and
(reviewed in Johnston 2003). Most effects of MHC- kept in separate jail cells to prevent them from commu-
based recognition have been found in the context of nicating. Certain that one of them is guilty, but lacking
mate choice and incest avoidance. To date there is no evi- sufficient evidence for a conviction, the prosecutor offers
dence that MHC genes influence the perception of odor, each a deal. Each prisoner is told that there is enough
which is necessary for MHC to be a “recognition allele” incriminating evidence to guarantee a short jail term, but
under our strict definition above (Mateo 2004). freedom can be obtained by providing enough evidence
Nonetheless, the MHC system remains an interesting to send the other to jail for a long time. However, if each
case of a close link between genes and kin recognition. informs on the other, they both go to jail for an inter-
mediate length of time. We can construct a payoff matrix
(Table 19.1), just as we did for hawk–dove games of con-
RECIPROCAL ALTRUISM flict in Chapter 18. The possible strategies available to
As we have seen, helping relatives is favored by natural each player are to cooperate (don’t squeal on your part-
selection when the inclusive fitness of the helper is ner) or to defect (squeal). The best that you can do is to
increased. However, in everyday speech, we typically defect while your partner cooperates (getting a payoff of
use the word “altruist” not to describe a person help- T, which stands for the temptation to defect). If both you
ing a relative, but a person helping a nonrelative. Can and your partner cooperate, the payoff is R (for reward
altruism toward nonrelatives be favored by natural for cooperation). If both partners defect, the payoff is P
selection? (for punishment for defection). If you cooperate and
The answer is yes, but only under a narrow set of your partner defects, you get the lowest possible payoff,
circumstances. Altruism between nonrelatives can S (for sucker’s payoff). For a game to be a Prisoner’s
evolve if there is an opportunity for payback in the Dilemma, the payoffs in the four cells illustrated in Table
future. We call this reciprocal altruism, evolution’s 19.1 must be in the order T > R > P > S.
version of “you scratch my back and I’ll scratch yours” In order to determine the best strategy, examine the
(Trivers 1971). payoff matrix. If you are Player A, what is your best strat-
In order to think about the circumstances that egy if your opponent cooperates? Because T > R, it is bet-
might favor reciprocal altruism, it’s helpful to follow the ter to defect—you will go free. Similarly, if your
lead of many other behaviorists and think about a opponent squeals, compare the payoffs of your two
particular scenario: the Prisoner’s Dilemma. In recip- options. Because P > S, it is again better to defect, rather
rocal altruism, the costs and benefits to the altruist than taking the rap for both of you. Therefore, in a sin-
depend on whether the recipient returns the favor. You gle round of playing the game, it is always better to

TABLE 19.1 Fitness Payoffs for Player A in the Prisoner’s Dilemma. The payoffs are ordered T > R > P > S

Strategy of Player B

Cooperate Defect

R S
Cooperate
Reward for mutual Sucker’s payoff
cooperation
Strategy of Player A

T P
Defect
Temptation to defect Punishment for mutual
defection
 The Puzzle of Altruism 433

defect. If both players follow this logic, they will both Vampire bats meet the conditions that Trivers laid
defect and will each do worse than if they cooperated. out that are favorable to the evolution of reciprocal altru-
If we translate jail sentences into fitness payoffs, the ism. First, the benefit to the recipient of the blood gift
solution to the Prisoner’s Dilemma seems to imply that is greater than the cost to the donor. Since a bat’s body
reciprocal altruism cannot evolve. Indeed, this may be weight decays exponentially after a meal, the recipient
true if the prisoners will never meet again. In real life, may gain 12 hours of life and, therefore, another chance
however, individuals often interact repeatedly, and when to find food. However, the donor loses fewer than 12 hours
they do, reciprocal altruism may evolve. of time until starvation and usually has about 36 hours,
One strategy, called “tit-for-tat,” can be a winner in another two nights of hunting, before it would starve.
repeated games of Prisoner’s Dilemma. In this strategy,
an individual begins by being cooperative and in all sub- a
sequent interactions matches the other party’s previous
action. This strategy, then, has the characteristics of
being “nice” (it begins with cooperation), it is retaliatory
(if the partner defects, it defects in return), and forgiv-
ing (it immediately “forgets” a defection and cooperates
if the partner later cooperates). If a population of indi-
viduals adopts this strategy, it cannot be overrun by a
selfish mutant that always defects (Axelrod 1984). So,
when the individuals have repeated encounters, recipro-
cal altruism can be an evolutionarily stable strategy
(ESS), one that cannot be invaded by another strategy
(see Chapter 4).
Given these results from game theory, let’s step back
and think more generally about when reciprocal altru- b
ism is likely to evolve. Trivers (1971) outlined several
conditions that favor reciprocal altruism: (1) the benefit
of the act to the recipient is greater than the cost to the
actor, (2) the opportunity for repayment is likely to
occur, and (3) the altruist and the recipient are able to
recognize each other. These factors are most likely to
occur in a highly social species with a good memory, long
life span, and low dispersal rate (Trivers 1971).
One of the best-known examples of reciprocal altru-
ism among nonhuman animals, partly because it is such
a startling species in which to find this trait, is vampire
bats (Desmodus rotundus). Vampire bats fly out at night
to find their favorite prey, large mammals such as cattle c
or unsuspecting tropical biologists. The bats land near
the sleeping victim and then skitter silently toward it
along the ground. They then climb aboard and bite the
prey with razor-sharp teeth specialized for slicing
through skin, lapping up the blood. Anticoagulants in
the bats’ saliva keep the blood flowing. Successful bats
can have giant meals—they can consume up to 50% of
their body mass.
At the end of the night, vampire bats return to their
communal roosts. Here is where altruistic acts take place:
an unlucky bat that did not obtain a blood meal begs for
food by licking one of its roostmates under the wings and
FIGURE 19.9 A vampire bat that was unsuccessful in
on its lips. A receptive donor will then regurgitate blood obtaining a meal during a night’s hunt begs for food
(Figure 19.9). The regurgitated food is enough to sus- from a roostmate. (a) First it grooms the roostmate by
tain the hungry bat until the next night, when it may find licking it under the wings, (b) then it licks it on the lips.
its own meal. Donors may give blood to recipients that (c) If receptive, a well-fed roostmate will respond by
are not related to them (DeNault and McFarlane 1995; regurgitating blood to the hungry partner. (From
Wilkinson 1984, 1990). Wilkinson 1990.)
434 Chapter 19 / Group Living, Altruism, and Cooperation

Physiological studies of metabolism suggest that vampire and offspring are not genetically identical, their interests
bats are unusually susceptible to the effects of starvation are not always perfectly aligned. Parents have an advan-
(Freitas et al. 2003, 2005). tage in power struggles, as they are generally larger and
Second, bats are likely to have the opportunity to more experienced than their offspring. Because offspring
repay favors and to get favors repaid. Most bats are likely are related to their parents, their motivation to resist
to be in the position of needing blood from roostmates, coercion is reduced—if, for example, they forgo their
because on any given night, roughly 33% of the juveniles own breeding in order to help their parents reproduce,
that are less than two years old and 7% of older bats fail they at least gain indirect fitness.
to feed. Bats are also likely to encounter the same individ-
uals time and again. Vampire bats roost in somewhat sta-
ble groups of both related and unrelated members. A EXAMPLES OF COOPERATION
typical group consists of 8 to 12 adult females and their AMONG ANIMALS
pups, a dominant male, and perhaps a few subordinate
males. Males leave their mothers when they are about 12 As we consider various forms of cooperation among ani-
to 18 months old, but females usually remain well past mals, we will note many similarities among distantly
reproductive maturity. New females occasionally join the related groups. We will also notice that the selective
group. Although the groups may change slightly over time, forces leading to similar forms of cooperation may be
there are numerous opportunities to share food. Females quite different.
may live as long as 18 years, and in one study, two tagged
females shared the same roost for more than 12 years.
Trivers’s third condition is that individuals that rec- ALARM CALLS
ognize each other can thus direct help appropriately. Some of the classic long-term field studies of coopera-
Generally, only bats that have had a prior association share tion come from studies of alarm calls—warnings that
food. In an experiment, a group of bats was formed in the animals give to alert others of danger. Let’s examine a
laboratory. Aside from a grandmother and granddaugh- particularly well-studied taxon, ground squirrels and
ters, all the bats were unrelated. The bats were fed nightly their relatives.
from plastic measuring bottles so that the amount of blood
consumed by each bat could be determined. Then, every
night one bat was chosen at random, removed from the Belding’s Ground Squirrels
cage, and deprived of food. When it was reunited with its High in the mountains of the Sierra Nevada in California,
cagemates the following morning, the hungry bat would in Tioga Pass in Yosemite National Park, lives a very
beg for food. In almost every instance, blood was shared well-studied population of Belding’s ground squirrels
by a bat that came from the starving bat’s population in (Spermophilus beldingi). During most of the year, the alpine
nature. Furthermore, there seemed to be pairs of unre- meadows are covered in deep snow, but in the short sum-
lated bats that regurgitated almost exclusively to each mer season, the meadows are awash in flowers and grasses.
other, suggesting a system of reciprocal exchange. It is then that the ground squirrels become active, forag-
If Trivers’s three characteristics of species likely to ing in the fields during the day and returning to under-
show reciprocal altruism bring our own species to mind, ground burrows to sleep and care for their offspring.
it is not surprising. Robert Trivers (1971) points out that It’s not only ground squirrels that call these lush
reciprocal altruism is particularly common among humans. meadows home. Squirrels are attacked both from the
Not only do humans help the needy through social pro- ground (badgers, snakes, weasels, coyotes) and the air
grams (with the expectation that they, too, might someday (hawks and eagles). When squirrels spot a predator, they
benefit from such a program), but they also help one give an alarm call—a series of short sounds for a terres-
another in times of danger and they share food, tools, and trial predator, and a high-pitched whistle for an aerial
knowledge. Trivers even argues that our feelings of envy, predator (Figure 19.10).
guilt, gratitude, and sympathy have evolved to affect our Paul Sherman (1977; 1980a,b; 1985) studied the
ability to cheat, spot cheaters, or avoid being thought of function of these calls: are they directed at the predator
as a cheater ourselves. We’ll see more examples of recip- to let it know it has been detected? Or are they directed
rocal altruism among our primate kin later in this chapter. at kin? The answers seem to differ depending on the
circumstances.
MANIPULATION
Individual Selection When a hawk is spotted over-
In some cases, animals can coerce others to help them. head or when an alarm whistle is heard, near pandemo-
This is especially likely to occur between parents and off- nium breaks out in the colony. Following the first
spring (Alexander 1974, Trivers 1974; reviewed in warning, others also whistle an alarm and all scurry to
Crespi and Ragsdale 2000). Recall that because parents shelter. When a hawk is successful, the victim is most
 Examples of Cooperation Among Animals 435

seems that the alarm whistles given at the sight of a

predatory bird directly benefit the caller by increasing its
chances of escaping predation in the ensuing chaos
(Sherman 1985).

Kin Selection In contrast, individual selection does

not seem to be behind the evolution of the ground
squirrels’ alarm trills, given in response to terrestrial
predators. In this case, the caller is truly assuming a risk;
we know this because significantly more callers than
noncallers are attacked. As can be seen in Table 19.2, 8%
of the ground squirrels that called in response to terres-
trial predators were captured, whereas only 4% of the
noncallers were caught. The predators, even coyotes
whose hunting success often relies on the element of sur-
prise, did not give up when an alarm call was sounded.
Furthermore, the caller was not manipulating its neigh-
bors to its own advantage. Generally, the reaction of
other ground squirrels was to sit up and look in the
direction of the predator or to run to a rock. Their reac-
tion did not create the chaos that might confuse a preda-
tor. Nor did the caller seek safety in the midst of
aggregating conspecifics (Sherman 1977).
The evidence that kin selection is the basis for alarm
trills to terrestrial predators by ground squirrels is
strengthened by information on the structure of their
FIGURE 19.10 A female Belding’s ground squirrel emits society. Daughters tend to settle and breed near their
an alarm call. birthplace, so the females within any small area usually
are genetically related to one another. The sons, on the
other hand, set off independently before the first winter
likely to be a noncaller. In one study, only 2% of the hibernation, never to return to their natal burrow.
callers but 28% of the noncallers were caught (Table Sherman knew the relationships and the identity of the
19.2). The most frequent callers were those that were in squirrels (and could identify them from a distance
exposed positions and close to the hawk, regardless of because he marked them by painting their flanks with
their sex or relationship to those around them. Thus, it Lady Clairol hair dye).

TABLE 19.2 Alarm Calling and Survival in Belding’s Ground Squirrels at Tioga Pass, California. All Data
Are from Observations Made During Attacks by Hawks (n ⴝ 58) and Predatory Mammals
(n ⴝ 198) That Occurred Naturally During 1974–1982 (Data from Sherman 1985)
Number of Ground Squirrels

Captured Escaped % Captured P value

Aerial predators
Callers 1 41 2%
Noncallers 11 28 28% <0.01
Total 12 69 15%
Terrestrial predators
Callers 12 141 8%
Noncallers 6 143 4% <0.05
Total 18 284 6%
436 Chapter 19 / Group Living, Altruism, and Cooperation

Whistles to aerial predators nonoffspring relatives does not significantly affect their
Expected to call Observed to call likelihood to call (Blumstein et al. 1997). Other rodent
Category 80 60 40 20 0 20 40 60 80 species (e.g., mice of the genus Peromyscus) apparently do
––––––––
Adult females not call at all. Shelley and Blumstein (2005) carried out
Adult males a comparative analysis (see Chapter 4 for more details on
1–year females this research technique) in which they plotted traits onto
1–year males a phylogenetic tree to see which traits were likely to have
Juvenile females evolved together. Social species were more likely to call
Juvenile males
than nonsocial species. However, an even stronger rela-
(n = 210)
G = 4.8, p > 0.50
tionship was that diurnal (day-active) rodent species were
more likely to call than nocturnal (night-active) species
Trills to terrestrial predators and that the evolution of diurnality precedes the evolu-
Expected to call Observed to call tion of alarm calling. Shelley and Blumstein argue that,
80 60 40 20 0 20 40 60 80 in general across rodent species, alarm calling serves pri-
Adult females marily to communicate with predators and that benefits
Adult males arising from kin selection are secondary. Belding’s
1–year females ground squirrels may benefit more from kin selection
1–year males
than some other rodent species because they live in high-
Juvenile females
density meadows where many relatives are likely to be
Juvenile males
within earshot (Blumstein 2007).
(n = 127)
G = 73.5, p > 0.001

FIGURE 19.11 Expected and observed frequencies of COOPERATION IN ACQUIRING A MATE

alarm calls by Belding’s ground squirrels in response to
aerial and terrestrial predators. Expected frequencies Males of some species cooperate in attracting a mate.
are those that would be predicted if the animals called Some even relinquish the opportunity to pass their
randomly. The calls in response to aerial predators are alleles into the future generation personally, at least tem-
close to the expected frequencies. However, the calls in porarily. Indeed, these males concentrate their efforts on
response to terrestrial predators are more likely to be making another male more attractive to females. This
given by females with relatives nearby than would be seems like a guaranteed route to lower fitness, but in
predicted if the animals called randomly. (From some cases it can be advantageous. Let’s look at four
Sherman 1985.) examples, two from birds and two from mammals. As
you will see, sometimes the benefits to cooperators can
arise via individual selection, kin selection, or reciprocal
Sherman’s data (Figure 19.11) suggest that the altruism, and sometimes a combination.
ground squirrels practice nepotism, or favoritism for
family members. Notice in the figure that when a ter-
Wild Turkeys
restrial predator appears, females are more likely than
males to sound an alarm. This is consistent with kinship Strangely, most male wild turkeys (Meleagris gallopavo) in
theory because it is females that are more likely to have some Texas populations never mate. Toward the end of a
nearby relatives that would benefit from the warning. In young male’s first autumn, when he is about six to seven
addition, reproductive females are more likely than non- months old, he and his brothers forsake the others in their
reproductive females to call. An even finer distinction family and form a sibling group that will be an inseparable
can be made: reproductive females with living relatives unit until death. This sibling group and all other juvenile
call more frequently than reproductive females with no male sibling units in the area flock together for the winter.
living family members. Only the dominant male in each group mates. During
the first winter, each male’s status within this fraternity is
decided by the outcome of two contests. One competition
Alarm Calls in Other Rodents is for dominance within the sibling group. Brothers battle
Many rodent species, not only Belding’s ground squir- by wrestling, spurring, striking with their wings, and peck-
rels, give alarm calls, providing us with a powerful tool ing at each other’s heads and necks. Endurance is the key
for comparative analysis. Species vary in whether alarm to success: the turkeys fight until they are exhausted. When
calling increases the caller’s own chance of survival, pri- only one is able to do battle, however weakly, he is the win-
marily aids the caller’s own offspring, or helps other ani- ner. The second contest is between rival sibling groups.
mals. For example, yellow-bellied marmots are most The groups challenge and fight one another until a dom-
likely to call when their own offspring are nearby, and inance hierarchy is established. The sibling group with the
(unlike Belding’s ground squirrels) the presence of most members is usually victorious. Renegotiation of rank
 Examples of Cooperation Among Animals 437

is rare; the dominance hierarchy within and between sib- every 15 minutes around the clock. Any of the males in
ling groups is stable. the coalition may be the first to find her, mate with her,
When the breeding season begins, females inter- and keep others away by his presence. A female may
ested in mating visit the open meadows, where the males change mates during this period but generally not more
congregate. Two to four siblings form display partner- than once a day (Bertram 1975, 1976).
ships within larger aggregations of males. The brothers In turkeys, we saw that males helped their brothers
of each unit court the hens by strutting in unison, even gain mates. Could kin selection also underlie the forma-
though only the dominant male in the highest-ranking tion of coalitions in lions? Coalitions of more than three
sibling group will mate. Of 170 tagged males displaying individuals usually consist of close relatives that left their
at four grounds, not more than 6 males accounted for all natal pride as a group (Packer et al. 1991). They remain
59 observed matings. If a subordinate male is presump- together, and after one to three years of traveling
tuous enough to attempt a mating, the dominant male nomadically, they challenge the males of other prides.
chases him away and then mates with the hen. Thus, a male in a coalition with relatives has the chance
Watts and Stokes (1971) were the first to suggest that to gain reproductive success indirectly (by helping his
kin selection is the major driving force behind the evo- male relatives mate with the female).
lution of this behavior, and this was confirmed recently Kin selection can’t be the entire story behind coop-
by genetic analysis (Krakauer 2005). A subordinate male erative groups of male lions, or even most of it. It is now
gains inclusive fitness by helping his brother to perpetu- known that roughly half of male coalitions contain at
ate his alleles. On the other hand, without his assistance, least one unrelated male, and coalitions of two or three
the brother could not be successful. The cooperative usually consist of unrelated males (Packer 1986; Packer
efforts of siblings are necessary for their unit to become et al. 1991; Packer and Pusey 1982). When presented
dominant, and the synchronous strutting of siblings with a stuffed “intruder” lion accompanied by playback
makes the dominant male more attractive to the hens. of recorded roars, lions attacked it (Figure 19.13),
Thus, the subordinate brother reproduces by proxy.

Lions a
Male lions (Panthera leo) also cooperate in acquiring
mates. They form coalitions, or partnerships, that chal-
lenge the males of other prides. Coalitions may take
over a pride by slowly driving out the resident males, or
it may be a hostile takeover, involving serious fighting
(Figure 19.12). In such contests, the larger coalition
usually wins.
The reward for the victors is a harem of lionesses.
When the females come into reproductive condition
(which is sometimes hastened if the new males kill cubs
that were sired by other males), they often do so simul-
taneously. During the two- to four-day period when a
female is in reproductive condition, she mates about
b

FIGURE 19.12 Two male lions fight for control of a FIGURE 19.13 A male lion attacks a stuffed lion after
pride. hearing the recorded roar of a male intruder.
438 Chapter 19 / Group Living, Altruism, and Cooperation

Long-Tailed Manakins
The Costa Rican rainforest is filled with fascinating ani-
mals, but a standout is a small bird called the long-tailed
manakin (Chiroxiphia linearis). Walking through the for-
est near Monteverde, one can often hear in the distance
their calls of “To-lay-do! To-lay-do!” A quiet approach
6 yields a remarkable sight: in a small clearing, two or even
Expected reproductive

three males work together to attract a mate. They begin

success per male

with the call, emitted as many as 19 times a minute and

4
5000 times a day. Once a female arrives, they begin a
visual display. In one of the most common variations of
2 the display, called the up-down jump display, the birds
perch side by side on a branch. One male jumps into the
air, emitting a wheezy buzzee call, and hangs there
1 2 3 4 5–7 momentarily; just as he lands, the other male jumps up.
Coalition size
Alternately jumping up and down, the males are remi-
FIGURE 19.14 Males in larger coalitions have greater niscent of children on a see-saw. Even more spectacular
reproductive success. (Data from Packer and Ruttan is the cartwheel. Here, males again begin by perching
1988.) next to each other. Then, one male jumps upward and
backward over the second male. Meanwhile, the second
male hops along the branch to take over the first male’s
regardless of whether they were related to other spot. Now this male takes a turn at jumping up and mov-
members of their coalition (Grinnell et al. 1995). ing to the rear. The dance looks as if the two birds were
Lions attacked the model regardless of the behavior of balls being juggled. The courtship sequence may be
other coalition members and did not appear to moni- repeated only once or as many as a hundred times in suc-
tor the actions of others, suggesting that reciprocal cession. When the display bout is over, one male leaves
altruism is not playing an important role (Grinnell et the display branch and watches while the remaining male
al. 1995). does a solo performance. If his gymnastics have
Why, then, do unrelated males gather into coali- impressed the female, she mates with him.
tions? The answer turns out to be quite simple—the Although the males take turns jumping during their
larger the coalition, the greater a male’s reproductive tandem courtship display, the same male always mates.
success (Figure 19.14). Larger coalitions have a better The benefit of this elaborate dance to the mating male
chance of ousting the current coalition in a pride, main- is obvious, but what about the other male? It is not indi-
taining control of that pride, and perhaps even gaining rect fitness because it seems unlikely that the two males
residence in a succession of prides. A solitary male has are related. A typical brood consists of only one or two
little chance of reproducing and, therefore, much to offspring, and there is no reason to assume that the siblings
gain by joining a coalition. A small coalition may also are necessarily the same sex. Furthermore, just before
benefit by accepting an unrelated male because the and after each breeding season, the young, particularly the
extra member may help it take over prides. Indeed, subadult males, disperse. It seems unlikely, therefore,
coalitions accept unrelated companions only while that male relatives would stay in proximity for the three
coalition members are not yet resident in a pride to four years it takes them to acquire adult plumage. If
(Packer and Pusey 1987). the partners are not genetically related, the nonbreeder
Larger coalitions also remain in control of a pride is not increasing his indirect fitness. So why does a sub-
longer than smaller ones. A coalition of three to six ordinate bird stay?
males may remain in control as long as two to three A subordinate bird’s chances of mating would not be
years. A coalition of two might be in possession for over increased by deserting his partner. Solitary males cannot
a year. If a lone male manages to gain control of a pride, mate. They cannot even perform the courtship display.
which happens infrequently, his tenure generally lasts If a male cannot dominate his current partner, his
only a few months (Bertram 1975). As a result, the life- chances of becoming the dominant member of another
time success of a male lion increases by cooperating pair will be low. However, if the subordinate male out-
with other males in taking over a pride, even if all the lives his partner, it is likely that a younger male, one that
males are not related. Packer et al. (1988) estimated can be dominated, will become his new associate. Then
that each additional member of a coalition increases it will be his turn to mate and raise his direct fitness
individual reproductive success by 0.64 surviving cubs (Foster 1977; McDonald and Potts 1994). In addition,
per male. singing—in particular the ability to match a partner’s
 Examples of Cooperation Among Animals 439

FIGURE 19.15 An alliance between two male olive

baboons. The two males on the right are cooperating to
challenge the male on the left. At a later time, the male
that was assisted will have to reciprocate to maintain the
alliance.

song, which impresses the females—improves with age

and practice (Trainer and McDonald 1995; Trainer at al.
2002). Thus, the most likely reason for cooperative
courtship in this species is that it increases the chances
of obtaining direct fitness benefits in the future.

Olive Baboons FIGURE 19.16 An acorn woodpecker places an acorn in a

hole drilled in a dead tree. Food stores must last the
Male olive baboons (Papio anubis, Figure 19.15) also
communal group through winter.
cooperate in attracting a mate, but here we see yet
another evolutionary mechanism at work. A male who
lacks a female consort sometimes enlists the help of a
What leads us to begin this section with acorn wood-
friend to win another male’s mate. The following scenario
peckers, however, is not their bizarre food-storing habits
is typical of what often occurs: Male A is associating
but their extensive social networks. These birds live in
(called consorting) with an estrous female. Male B, who
family groups of up to 15 or so, containing one to four
has no female of his own, solicits the help of male C, and
breeding males, one or two breeding females, and from
the two form an alliance and challenge male A. While the
zero to ten nonbreeding helpers. It is these nonbreeding
battle is in progress, male B gets away with the female.
helpers that are of particular interest. They are adult birds,
Male C has acted altruistically; he risked injury while
physiologically capable of breeding themselves, that
assisting another to acquire a mate. However, at some
instead stay to help their parents raise additional young.
time in the future, he will enlist the help of male B in win-
Acorn woodpeckers thus fit the definition of cooperative
ning a consort of his own (Packer 1977). You will recog-
breeding: some individuals (helpers) assist in the care and
nize this as an example of reciprocal altruism.
rearing of another’s young rather than producing offspring
of their own. Cooperative breeding was first described in
COOPERATIVE BREEDING birds (Skutch 1935) but has since been documented in
AND HELPING many taxa, including birds, mammals, insects, and spiders,
although it is quite rare (e.g., it is found in only about 3%
When hiking in the California woodlands, keep alert for of birds and mammal species; Emlen 1997).
the unmistakable signs of the acorn woodpecker There are many variations on the theme of coop-
(Melanerpes formicivorus). If you don’t recognize it by its erative breeding. Instead of limiting ourselves to the
harsh “waka-waka-waka” call, or by its black-and-white woodpecker example, we will organize the remainder of
plumage topped off by a snazzy red cap, the dead give- this section by the questions that have intrigued
away that you are in its habitat is a dead tree or even a researchers.
telephone pole peppered with a multitude of holes, each
with an acorn inserted snugly in it. Acorn woodpeckers
drill these holes to make their granaries, and they store How Do Helpers Help?
their acorns for leaner times (Figure 19.16). (One group In many species, helpers pitch in to feed offspring. The
of woodpeckers got a bit carried away and stored 485 time-intensive task of rearing baby birds is easily shared
pounds (220 kg) of acorns in a water tank in Arizona!) because anybody, not just the parents, can collect and
440 Chapter 19 / Group Living, Altruism, and Cooperation

a helpers mainly contribute by protecting offspring; they

can contribute little to the nourishment of the young
(Dugatkin 1997).
Provisioning of food and extra protection are the
most common services offered by helpers, but other tasks
also lend themselves to sharing. In some bird species,
helpers may build and clean nests or incubate and brood
the nestlings (Skutch 1987). In saddle-backed tamarins
(Saguinus fuscicollis), a small primate, male helpers lug
around the offspring. At birth a tamarin is almost 20%
of its adult weight, and litters typically consist of twins.
Thus, carrying these youngsters is burdensome, and if
the duty were not shared the mother might not be able
b to obtain enough nourishment for herself and to ensure
an adequate milk supply (Terborgh and Goldizen 1985).

Is “Helping” Really Helpful?

Given all the useful tasks helpers can perform, this may
seem a silly question. However, not every study has
found a relationship between the presence or number of
helpers and the reproductive success of the breeding
individuals (reviewed in Clutton-Brock 2002), so it is
wise not to make assumptions.
In many species, researchers have shown that the
number of helpers correlates with the survival of the
young, the survival of the breeders, or both. For exam-
FIGURE 19.17 A jackal helper (a) prepares to regurgitate ple, we have already mentioned the Florida scrub jay
food to a pup and (b) chases away an intruder.
(Woolfenden 1975; Woolfenden and Fitzpatrick 1990).
These birds have been studied for many years in the scrub
carry food to the nest and pop it in a nestling’s mouth. habitat of Florida, currently threatened by development.
Acorn woodpeckers provide nestlings with food, both The jays have proven to be easily tamed, allowing for
insects and acorn fragments (Koenig and Mumme 1987). close observation (Figure 19.18). The jays form territo-
Helpers in the well-studied Florida scrub jay (Aphelocoma ries that contain one breeding pair and a varying num-
coeruslescens) deliver about 30% of the food consumed by ber of helpers—from none to as many as six. The
nestlings. The parents’ job is thus reduced, and, as a breeding success of pairs with helpers clearly exceeds that
result, they enjoy better health. In one study, 87% of the of pairs without helpers (Woolfenden 1975). Figure 19.19
breeders with helpers survived to the next year, com-
pared to 80% of breeders without helpers (Stallcup and
Woolfenden 1978).
Although mammalian mothers are uniquely
equipped to provide milk to their offspring, helpers can
deliver other kinds of food by carrying it in their jaws or
in their stomachs. Blackbacked jackals (Canis mesomelas)
regurgitate food to eager pups (Figure 19.17). Helpers
not only contribute 18 to 32% of all regurgitations to
pups, but they sometimes also regurgitate to the lactat-
ing mother, allowing her more time to remain with the
pups instead of hunting (Moehlman 1979).
Help can also come in the form of extra protection
for the young. Florida scrub jays give alarm calls to
predators such as snakes and even help to drive them
away by mobbing them (Woolfenden 1975). Jackal fam-
ilies with helpers always have an adult on guard to drive
away predators, whereas groups lacking helpers may FIGURE 19.18 Florida scrub jays are not difficult to
have to leave the pups unattended while hunting. In fish, observe.
 Examples of Cooperation Among Animals 441

Eggs Number of Number of Number of lives in year-round territorial groups of 1 to 13 birds in

5 nestlings fledglings independent the open woodland of Queensland, Australia. Parents are
young
usually assisted by a variable number of their offspring
4 from previous broods, but in this experiment, nine of the
breeding groups were reduced to a single helper. These
Average number of young

groups then raised an average of 0.8 young, less than half

3 the number of fledglings produced by the 11 control
groups, which had more assistance. Therefore, the pos-
itive relationship between breeding success and the num-
2
ber of helpers found in gray-crowned babblers does seem
to be a result of the presence of helpers (Brown et al.
1 1982). However, a problem with this experimental
approach is that removal or addition of helpers has other
unintended effects: it can be very disruptive to the social
0 1 2 3–5 0 1 2 3–5 0 1 2 3–5 0 1 2 3–5
structure of a group, and it also changes group size,
Number of helpers which in turn can affect the group’s success (reviewed in
Wright and Russell 2008).
FIGURE 19.19 The relationship between the number of Another experimental approach to measuring the
Florida scrub jay helpers and the breeding success of the benefits of helping is to change the ratio of helpers to
experienced parents. Helpers do not increase the number young in a different way, by temporarily removing or
of eggs laid. They do, however, increase the chances that adding young while the number of helpers remains con-
the eggs will hatch and that the young will survive to stant. When Clutton-Brock et al. (2001) performed this
become independent. (Data from Woolfenden 1975.)
manipulation in meerkats, which are (quite adorable)
mammals (Figure 19.20), pup weight gain was increased
when the number of pups was reduced, and decreased
shows the breeding success of experienced pairs with and when the number of pups was increased. Extra food
without helpers during one five-season study. Notice that helps the pups by reducing the age at which they first
the presence of helpers has no effect on the number of
eggs laid but does increase the chances that the young will
hatch, leave the nest, and become independent birds.
Similarly, breeding success increases with the pres-
ence of helpers in some mammalian and fish species. For
example, pairs of blackbacked jackals are joined by
between one and three young from previous litters, who
help them rear the next pups. The reproductive success
of a pair of blackbacked jackals increases with the num-
ber of helpers (Moehlman 1979).
But we must be cautious: finding that there is a cor-
relation between breeding success and the presence of
helpers is not by itself sufficient to demonstrate that
helping causes increased breeding success (Brown et al.
1982). It’s quite possible that another factor, such as ter-
ritory quality, causes both an increase in the number of
helpers and increased breeding success, thus creating the
correlation. In addition, helpers are often the offspring
of the breeders they are assisting—perhaps there is a
correlation between breeding success and the number of
helpers simply because the breeding pair is of consis-
tently high quality from year to year. How might we be
sure that it is really the helpers that caused increased
reproductive success of the breeders?
One way to increase our certainty that helping
causes increased reproductive success in breeders is by
removing helpers and measuring the consequences. For
example, helpers were removed from the nests of gray-
crowned babblers (Pomatostomus temporalis), a bird that FIGURE 19.20 A meerkat helper with pups.
442 Chapter 19 / Group Living, Altruism, and Cooperation

reproduce and by increasing the chance that they can Is It Costly to Help?
successfully compete for a high-ranking spot in the social Helping behavior becomes a more interesting evolu-
hierarchy (Russell et al. 2007). tionary puzzle if it has a cost, and you won’t be sur-
A third way to approach the problem of teasing apart prised to learn that it often does. For example,
correlation from causation, at least in long-term studies mongoose (Suricata suricatta) helpers forgo feeding and
with plenty of data, is with statistical techniques by which stay at the burrow to baby-sit for the young pups and
different variables can be controlled (Cockburn et al. guard them from predators for an entire day while the
2008; Wright and Russell 2008). Cockburn et al. (2008) parents and others forage. During a 24-hour shift, the
used elegant statistics to analyze 19 years of data on the baby-sitter loses 1.3% of its body weight. In contrast,
superb fairy-wren. They found that helpers did not the foraging group members gain roughly 1.9% of their
increase the survival of offspring but instead increased body weight (Heinsohn and Legge 1999). Helping may
the future survival of breeding females. even reduce survival. For example, the helpers among
Do breeders always want help? Not always. After stripe-backed wrens (Campylorhynchus nuchalis) that
all, an extra helper around might mean competition for bring the most food die more quickly than other birds
resources or increase the chance that your mate has an (Rabenoid 1990).
extra-pair copulation. In the pied kingfisher (Ceryle
rudis), helpers are tolerated only when their services are
needed. These birds usually have primary helpers, STOP AND THINK
which are older offspring, but may also have secondary Imagine you are spending your graduate school years
helpers, which are unrelated. Heinz-Ulrich Reyer studying helping behavior in a little-known bird species.
(1980) compared two colonies of pied kingfishers in East You find that birds that remain behind on their natal ter-
Africa. Breeding pairs at Lake Naivasha typically have ritory to help their parents are far more likely to die than
only one primary helper. When males apply for a job as those that go off to breed on their own. In your disserta-
secondary helpers, they are persistently chased away by tion defense, you would like to make the argument that
the male territory holder. In contrast, at Lake Victoria, helping is costly. How confident are you? What experi-
secondary helpers are eventually tolerated and permit- ment(s) would you like to perform in order to increase
your confidence?
ted to stay and feed the young. Why? The answer is that
the services of secondary helpers are needed to raise off-
spring at Lake Victoria but not at Lake Naivasha. These If it is costly to help, why do it? We can break this ques-
birds fish for a living, and Lake Victoria is a harder lake tion down into two parts: (1) Why would an offspring delay
to fish. Victoria’s rougher waters increase the time it dispersal and stay near home? (2) Why would it help? It
takes to catch a fish, and the fish are smaller. makes sense to address these as separate questions because
Furthermore, the fishing grounds are farther from the not every offspring that delays dispersal offers assistance,
colony. With the additional fish provided by secondary and sometimes offspring that have dispersed to nearby areas
helpers, the breeding pair can raise more offspring return to their parents’ area in order to provide care
(Table 19.3). (reviewed in Pruett-Jones 2004; Ekman et al. 2004).

TABLE 19.3 The Effect of Helpers on the Reproductive Success of Pairs of Pied Kingfishers. Shown Are
Mean, Standard Deviation, and Sample Size
Lake Victoria Lake Naivasha

Mean SD n Mean SD n

Clutch size 4.9 0.6 22 5.0 0.6 8

Young hatched 4.6 0.5 14 4.5 0.7 2

Young fledged

No helpers 1.8 0.6 14 3.7 0.9 9

1 helper 3.6 0.5 12 4.3 0.5 4

2 helpers 4.7 1.0 6 — — —

 Examples of Cooperation Among Animals 443

Why Should an Offspring Delay Dispersal? breeding units had helpers. In short, the frequency of
In Chapter 11, we discussed the costs and benefits of dis- helping in populations of acorn woodpeckers varies
persal and philopatry (staying near home). For example, directly with the scarcity of open territories.
animals may disperse in order to avoid inbreeding, Not all territories are equal in quality, and it might
reproductive suppression by their relatives, and compe- be better to stay home if the only other choice is a poor-
tition. On the other hand, animals may exhibit philopa- quality territory. The interaction between habitat satu-
try because they are adapted to the local conditions, and ration and territory quality has been nicely demonstrated
familiar with the physical and social settings of home. by Jan Komdeur (Komdeur 1992; Komdeur et al. 1995),
Dispersal may be risky: for example, small individuals of who transplanted Seychelles warblers to unoccupied
a species of cichlid fish (with the evocative name of the islands. At first, these small birds bred independently.
Princess of Burundi) are very likely to be eaten by preda- When territories began to fill up, some birds stayed to
tors when they venture off their territory. Even when help their parents rather than move to low-quality ter-
breeders are experimentally removed from a territory, ritories that had relatively few insects. When breeders on
potential helpers choose to stay at home instead of mov- low-quality territories were removed, the territories were
ing to the unoccupied territory (Taborsky 1985). filled only by helpers from territories that were equiva-
Animals might also delay dispersal because other lent or worse in quality.
options may be limited. Let’s look at two major ways in
which this might happen. Lack of Mates In addition to habitat, mates can also
be a limited resource (Emlen and Vehrencamp 1983).
Habitat Saturation Florida scrub jays live only in a For example, splendid fairy-wren females (Figure 19.21)
special scrub habitat, comprised of dry-adapted shrubs suffer much greater annual mortality than do males
and understory plants in sandy soils, that is already lim- (57% and 29%, respectively; Rowley 1981), and thus
ited in its distribution and growing ever scarcer because females are frequently in short supply. Helpers tend to
of human development. Habitat availability is so limited be males that are awaiting an available mate. For some
that virtually every territory is filled. Once a scrub jay is individuals, the wait can be as long as five years. As pat-
lucky enough to acquire a territory, it generally keeps it terns of mortality and the resultant sex ratios vary, so
for life (Woolfenden 1975). The most common way for does the percentage of groups with helpers: when
a male to acquire a territory is by inheriting a portion of females are scarce, male helpers are plentiful (Russell and
his parents’ property, either by replacing his father after Rowley 1988; Rowley and Russell 1990).
his death or by subdivision of his father’s territory. If
there is more than one son helping, the dominant one
is favored in the property settlement (Woolfenden and
Fitzpatrick 1978). Otherwise, a scrub jay can only claim
a territory of its own if it defeats a breeder or success-
fully competes for the territory of a breeder that has
died. Thus, we believe that a major reason that scrub jays
help their parents is that they are making the best of a
bad set of circumstances.
Support for the idea that the availability of territo-
ries is important in determining helping has also been
found in acorn woodpeckers. Study sites in California,
New Mexico, and Arizona vary with respect to wood-
pecker density, territory turnover rate, and territory
fidelity. In California, the habitat is extremely saturated:
not a single territory became vacant during a three-year
study (MacRoberts and MacRoberts 1976). Forty-nine
percent of the juveniles remained at home, and 70% of
the groups had helpers. Young acorn woodpeckers in the
Magalena Mountains of New Mexico face somewhat
better odds in their quest for suitable territories than do
those on the West Coast: 19% of the territories in New
Mexico became vacant over a three-year study (Stacey
1979). Here, 29% of the youngsters stayed at home, and FIGURE 19.21 A male splendid wren carries an insect to
59% of the groups had helpers. In the Huachuca feed its young. When adult females are in short supply
Mountains of southeastern Arizona, there is no shortage in the population, this breeding male can count on his
of territories (Stacey and Bock 1978). Only 16% of the sons to help rear the next brood.
444 Chapter 19 / Group Living, Altruism, and Cooperation

Why Should a Helper Help?

Even if an animal does not have the opportunity to
breed, that does not automatically mean it will become
a helper. Many animals become “floaters” and wander
around without a territory. Even if an individual remains
on its own territory, it may not help (as we know from
television sitcoms, if not our own experience), so there
must be other reasons besides lack of other opportuni-
ties that underlie helping behavior.
As with other forms of cooperative behavior, we can
categorize fitness benefits as direct benefits, in which a
helper’s own lifetime reproductive success is increased by
its actions, and as indirect benefits, in which a helper
benefits by increasing the production of relatives. The
relative importance of direct and indirect fitness bene- FIGURE 19.22 On the left is a nest built by a 4-year-old
fits depends on the species. female Seychelles warbler with no experience. Note that
it is built between a tree branch and a thin leaf stalk. On
Helpers May Get Indirect Fitness Benefits As we the right is a nest built by a 4-year-old female that had
have seen, in most species, helpers are older offspring experience in being a helper. Note its placement in a
sturdy tree fork.
who help their parents raise their younger siblings. Thus,
in cases where helpers improved the survival of nestlings,
helpers may gain indirect fitness benefits by increasing tion between helping behavior and relatedness. As a
the number of their younger siblings. result, the literature on cooperative breeding that once
Some long-term studies suggest that kin selection focused largely on studying indirect fitness benefits has
can be important. For example, researchers collected shifted emphasis to a study of direct benefits (Cockburn
five years of data on white-fronted bee-eaters to test 1998; Clutton-Brock 2002; Dickinson and Hatchwell
alternate hypotheses for helping behavior (Emlen 1991; 2004). To see how animals might improve their own
Emlen and Wrege 1989). They found no evidence for reproductive success by helping, we’ll look at several
direct benefits to the helpers (increased survival to the case studies.
next breeding season, increased chance of mating in Our first example comes from the Seychelles war-
subsequent years, or increased success in rearing young), blers we met earlier. Although these birds normally
but concluded that helping led to increased production remain on their natal territory to help their mothers raise
of related young, suggesting that indirect fitness bene- additional offspring, they are not as related to the off-
fits are important. spring they help as you might expect. In this species,
Another line of evidence supporting the hypothesis extra-pair copulation is common, so often a helper will
that indirect benefits can be important comes from cases not have the same father as the offspring it helps. In
in which animals have a choice of whom to help. In pied addition, sometimes eggs are deposited in a nest by other
kingfishers, there are two kinds of helpers: as we defined birds. Indirect benefits are thus relatively lower in this
previously, secondary helpers help unrelated offspring, species; instead, helpers gain direct benefits in several
and primary offspring help related individuals. ways. Genetic analyses of offspring show that helpers
Kingfishers become secondary helpers only when both may add their own eggs to a nest (Richardson et
their parents are dead; otherwise they are primary al. 2002). Second, subordinates may be able to take over
helpers (Reyer 1984, 1990). In long-tailed tits, helpers a territory budded off from the main territory (Komdeur
are failed breeders. When nests failed either naturally or and Edelaar 2001). Finally, helpers have a chance to
because of the interference of researchers, the failed practice their parenting skills. When they finally acquire
breeders helped at the nests of relatives rather than at their own territory, they are superior breeders (Figure
equidistant nests of nonrelatives. In fact, if close relatives 19.22; Komdeur 1996).
were not available, failed breeders didn’t help at all When mates are scarce, a helper may increase its
(Russell and Hatchwell 2001). own reproductive success by mating with the original
breeder in a later year or if the original mate dies.
Helpers May Get Direct Fitness Benefits Many Remember that in pied kingfishers, the secondary
social groups are comprised of relatives, so it is not sur- helpers are unrelated to the breeders. More than half of
prising that indirect fitness benefits have drawn the these helpers return to the same area the following year.
attention of many researchers. However, many groups Of these, half succeed in mating with the female they had
have unrelated members, and many studies find no rela- assisted (Reyer 1980, 1984, 1986).
 Examples of Cooperation Among Animals 445

guards and will quickly sting you, losing their lives in the
STOP AND THINK process—as they pull away, their stingers rip out of their
Helping behavior in humans has also been studied from an abdomens and remain in your flesh, into which venom
evolutionary perspective. Because human children are very continues to pump. There is a dearth of males in the
dependent for years, and mothers can give birth to a num- hive—all those busy workers are female. Only one indi-
ber of children in rapid succession, most mothers require vidual in the hive lays eggs, a bloated queen.
help from others to successfully raise their children. In a Bees are an example of a eusocial (literally “truly
review of 45 studies done across many cultures, Sear and social”) species. Eusocial species are defined by three
Mace (2008) found that the presence of maternal grand-
characteristics: reproductive division of labor (some indi-
mothers (the mother’s mother) and sibling helpers improves
viduals have offspring, and others do not), cooperation in
child survival, but surprisingly, fathers improved child sur-
vival in only one-third of studies. What is a hypothesis the care of young, and overlap of at least two generations
based on kin selection that might explain these data? capable of sharing in the colony’s labor (Michener 1969;
Wilson 1971). The astute reader may realize that these
characteristics also apply to some of the cooperative
breeders that we have just discussed. Species range along
EUSOCIALITY
a eusociality continuum according to how evenly repro-
If you’ve spent any time watching a glass-walled obser- duction is shared among group members (Figure 19.23)
vation hive of honeybees, you were probably impressed (Lacey and Sherman 1997; Sherman et al. 1995). At one
by the sense of purpose in the colony. The packed-in bees end, all or many of the group members breed. At the
bustle around, busy at their tasks. Watch for longer, and other end, breeding is restricted to one or several group
you’ll notice that younger workers stay inside the hive members. One way that biologists have described the
and feed the queen’s helpless larval offspring, each tucked degree of eusociality is by measuring “reproductive skew,”
carefully away in its own cell. Other workers maintain the or the proportion of individuals that give up reproduction.
hive and carry out the dead. Older workers forage at flow- Eusociality is rare. For many years, the only species
ers and, upon returning to the hive, communicate the known to be eusocial were insects from only two taxonomic
location of these resources to their hive mates in an elab- groups: either hymenopterans (ants, bees, and wasps) or
orate dance (Chapter 16). If you are foolish enough to isopterans (termites). More recently, the list of eusocial
threaten the hive, you will discover that some bees are invertebrates has been expanded to include species of

The Eusociality Continuum

Index of Reproductive Skew

0 1
Groove-billed anis Florida scrub jays Dwarf mongooses Fungus ants
Acorn woodpeckers Stripe-backed wrens Naked mole-rats e.g., Atta app.)
Black-tailed prairie Golden jackals African wild dogs Yellow-jacket wasps
dogs Stenogastrine wasps Halictid bees (e.g., Vespula spp.)
Spotted hyenas (e.g., Parischnogaster (e.g., Augochlorella Mound-building termites
nigricans striata) (e.g., Macrotermes spp.)
Carpenter bees Paper wasps Pemphigid aphids
(e.g., Xylocopa sulcatipes) (e.g., Polistes fuscatus) (e.g., Pemphigus spp.)
Social spiders Damp-wood termites Honey bees
(e.g., Anelosimus eximius) (e.g., Zootermopsis (e.g., Apis spp.)
nevadensis)

FIGURE 19.23 The eusociality continuum blurs the distinction between cooperative breeders and eusocial species.
The traditional definition of a eusocial society is one with reproductive division of labor, cooperation in caring for the
young, and an overlap of adult generations. Cooperative breeding species share these characteristics with eusocial
species, and so it has been suggested that they form a continuum of social systems. The primary differences among
the social systems is the degree to which reproduction is shared among group members. When reproduction is
restricted to a single individual, the reproductive skew is 1. A reproductive skew of 0 indicates that the lifetime
reproductive success of all group members is equal. This diagram shows predicted locations of a variety of coopera-
tively breeding species along the eusociality continuum. (From Lacey and Sherman 1997.)
446 Chapter 19 / Group Living, Altruism, and Cooperation

aphids (e.g., Aoki 1972, 1979, 1982), an ambrosia beetle good of another organism, but members of eusocial
(Kent and Simpson 1992), tiny insects called thrips species may relinquish all chances of reproduction. It is
(Crespi 1992), spiders (Vollrath 1986), and snapping no wonder that some researchers describe colonies such
shrimp (Duffy 1996). In each case, the colonies meet the as these as “superorganisms”—much as cells and tissues
criteria described above but with variations across taxa. function together for the survival and reproductive suc-
For example, some ant species are well known for their cess of the body, members of a eusocial colony function
striking morphological castes. Whereas all worker hon- efficiently together to ensure the survival and reproduc-
eybees look more or less the same, in some ant species tive success of the colony.
individuals vary tremendously according to their job, such In the remainder of this section, we’ll examine euso-
as intimidating soldier ants with their giant heads and cial species from two perspectives: first, what favors the
strong jaws that deliver a memorable bite. evolution of eusociality? Second, how do eusocial
Eusociality is extremely rare in mammals, and in fact colonies function on a day-to-day basis?
the only well-accepted examples are species in a group
called the mole rats. Of mole rats, the most studied is the
The Evolutionary Origins of Eusociality
naked mole rat, Heterocephalus glaber (Jarvis 1981). This
bizarre creature, furless and with long buck teeth, con- Haplodiploidy and the Hymenoptera Perhaps the
structs massive burrows where it feeds on giant under- most common feature of eusocial societies is that they
ground tubers. Naked mole rats fit the classical definition are family groups (Strassman and Queller 2007). As we
of eusociality originally applied to social insects. Breeding have seen, relatedness among individuals can favor the
is restricted to a single female, the queen, even in groups evolution of altruism, so the role of relatedness has long
with almost 300 members (Figure 19.24). Other adult been a focus of research.
females are smaller than the queen and neither ovulate We’ll begin by examining a special case of related-
nor breed. Only one to three males breed with the queen, ness that has generated a great deal of ink. As we’ve
although most adult males do produce sperm. Mole rat already seen, eusociality is found only in a handful of
colonies contain overlapping generations of offspring, species. What attracted the attention of researchers
which are communally cared for, and there is division of (especially before the recent expansion of the list of euso-
labor among individuals within the colony. The duties cial species) is that many of these species have an unusual
assumed by the nonbreeding members seem to depend genetic system called haplodiploidy. The females are
on their size and age. Smaller members generally gather diploid, with two sets of chromosomes, as are most
food and transport nest material. As they grow, they begin familiar animals. The males, however, are haploid, with
to clear the elaborate tunnel system of obstructions and a single set of chromosomes. Males grow from an unfer-
debris. Larger members dig tunnels and defend the tilized egg and produce sperm that are genetically iden-
colony (Honeycutt 1992; Lacey and Sherman 1991; tical to themselves. When a male mates with a female,
Sherman et al. 1992). the female offspring get a sampling of 50% of their
Eusociality presents an obvious puzzle. As we have mother’s DNA but all of the father’s DNA.
seen in our earlier discussions, it is rare that an animal Haploidiploidy changes the coefficients of related-
gives up even a portion of its reproductive success for the ness among family members. Full sisters (that share both
a mother and a father) get half of their DNA from their
mothers. Each sister gets a sample of 50% of mom’s
DNA, so on average, full sisters share 50% of their
maternally derived DNA. Full sisters also get half their
DNA from their fathers, but because fathers are haploid
and all of a male’s sperm is identical, this paternally
derived DNA is identical for every sister. When we con-
sider together the paternally and maternally derived
DNA, full sisters share 75% of their DNA with one
another (r = 0.75). Thus, a female of a haplodiploid
species shares more DNA with her full sister than with
her mother or daughter (r = 0.5) (Figure 19.25).
Hamilton (1964), whose work we have already met
earlier in this chapter (recall Hamilton’s Rule), realized
the potential implications of haplodiploidy for the evo-
lution of eusociality. Because sisters are more related to
FIGURE 19.24 A queen naked mole rat, the only repro- one another than to their own daughters, they pass on
ductive female of the colony, is resting on the workers more genes by rearing reproductively capable siblings
that feed her and help care for the young. than they would if they produced their own offspring.
 Examples of Cooperation Among Animals 447

chromosomes chromosomes FIGURE 19.25 The genetic contributions of a

A B A B male and female of a haplodiploid species to
copy copy
their offspring. Males are haploid and have
#1 #1 only one set of chromosomes. They arise from
#2 unfertilized eggs of the female. All their sperm
are identical. Females are diploid, and each of
Queen Drone their haploid eggs contains half their genes.
(Female parent) (Male parent)
Females have half of their mother’s genes and
all of their father’s genes. On average, two full
sisters share all the genes they got from their
A B A B A B A B father and half the genes they got from their
mother.

A B A B A B A B

Male offspring Female offspring

Degrees of relatedness in haplodiploid species

Daughter Son Mother Father Sister Brother

Female to her: 0.5 0.5 0.5 0.5 0.75 0.25
Male to his: 1 N/A 1 N/A 0.5 0.5

This insight quickly captured the imagination of animal working diligently as nursemaids and performing other
behaviorists. Could haplodiploidy be key to understand- tasks. Because the nursemaids that care for the brood are
ing eusociality? unrelated to the colony, their fitness is not increased by
Counterintuitively, some of the strongest evidence altering the 1:1 sex ratio that is optimal for the queen.
for the importance of relatedness in eusocial groups is not Although sex ratio data were available for only two slave-
cooperation but conflict. The flip side to the argument making species, Trivers and Hare (1976) found that the
that individuals should cooperate most with their closest investment in females and males was about 1:1.
relatives is that they should have the most conflict with These results, as well as other evidence of the
those that are least related. Favoritism of close over dis- importance of relatedness in colony function, led
tant relatives is demonstrated in a number of ways in researchers to hypothesize that there might be a tight
eusocial colonies. For an example, let’s look at whether relationship between haplodiploidy and eusociality. But
individuals should favor rearing male or female larvae. As how well has the haplodiploidy hypothesis stood the test
Figure 19.25 illustrates, a queen shares, on average, an of time? Unfortunately, not nearly as well as was initially
equal percentage of alleles with her sons and her daugh- hoped (reviewed in Linksvayer and Wade 2005, Wilson
ters. Thus, from a queen’s perspective, she should invest 2008). First, as noted earlier, we now know of a number
equally in both sexes and should favor a 1:1 sex ratio. In of eusocial species, such as termites, that are not hap-
contrast, as we have seen, worker bees (all females) share lodiploid. Thus, haplodiploidy is not necessary for the
an average of only 25% of their genes with their broth- evolution of eusociality. Second, there are many hap-
ers, but 75% with sisters. Thus, from a worker’s point of lodiploid species that are not eusocial. Thus, hap-
view, the optimal sex ratio is 3:1 in favor of females. lodiploidy alone is not sufficient for the evolution of
So who gets to decide the sex ratio in a colony? eusociality. Finally, even in haplodiploid species that are
Workers care for the brood and thus have an opportu- eusocial, the picture is complicated by the fact that
nity to manipulate the sex ratio. The investment in off- females very often mate with more than one male (e.g.,
spring is indeed nearly 3:1 in favor of females, in line with Laidlaw and Page 1984; Page and Metcalf 1982; Ross
the idea that workers are manipulating the sex ratio for 1986; Strassmann et al. 1989). Multiple mating reduces
their own benefit (Trivers and Hare 1976). For compar- the relatedness among sisters and also reduces the ben-
ison, consider the slave-making ants. Soldiers of slave- efit of helping sisters compared to producing offspring.
making species wage war against colonies of other ant However, a recent comparative study shows that mating
species and drag back the pupae or larvae. When the cap- with a single female is ancestral in all eight eusocial lin-
tives grow into adults, they serve their colony as slaves, eages that were examined. The authors suggest that
448 Chapter 19 / Group Living, Altruism, and Cooperation

monogamy was critical to the origin of eusociality and

that multiple mating evolves only after workers have lost
the ability to breed (Hughes et al. 2008).
The extent to which haplodiploidy makes the evo-
lution of eusociality more likely is still under debate.
There may be an indirect route: haplodiploidy may make
the evolution of maternal care more likely, which in turn
is necessary for the evolution of eusociality (Linksvayer
and Wade 2005). In any case, it is clear that we need to
look beyond haplodiploidy for a complete explanation
of eusociality.

Extended Parental Care and Long-Lasting Sibling

Associations The existence of parental care may well
be a prerequisite for the evolution of eusociality, as has
long been noted (reviewed in Linksvayer and Wade
FIGURE 19.26 In some species of ants, all individuals
2005). Long-lasting associations between parents and
look very similar, whereas in others there are strikingly
offspring allow for kin selection and inclusive fitness, as
different morphological castes. In this species of
we saw for other social species. In addition, long-lasting Acanthomyrmex, there are minor workers (top) and
associations allow the opportunity for parents to manip- major workers (bottom). The latter have giant heads,
ulate their offspring by, for example, restricting food to specialized for milling seeds and colony defense. (From
make the offspring smaller, or by aggressively harassing Hölldobler and Wilson 1990.)
offspring in order to make them cooperate (Crespi and
Ragsdale 2000). flict is often quite vicious and may even result in death
(reviewed in Strassman and Queller 2007).
Sharing a Defendable Resource Most eusocial
species rely on some sort of defendable resource, such Coordination Despite the existence of conflict, what
as a nest, burrows, or tunnels bored in wood. For exam- is most impressive to the casual observer of eusocial
ple, eusociality has arisen at least seven times in the groups is the level of cooperation. Although all eusocial
50,000–60,000 species of nest-building aculeate wasps, colonies show coordinated behavior, perhaps the most
but never (to our knowledge) in the 70,000 species of impressive examples are found among the ants. In some
wasps that do not build nests but instead lay their eggs species, different functions are carried out by morpho-
on prey (Wilson and Hölldobler 2005). Our one exam- logically distinct castes, as illustrated in Figure 19.26.
ple of a eusocial mammal, the mole rat, also vigorously Some castes have giant heads useful in milling grains and
defends its system of tunnels. There is strength in num- in defending the colony against intruders, whereas small
bers: for example, Gamboa (1978) found that a paper castes specialize in foraging and nest maintenance.
wasp that founded a nest by herself more often lost it In other species, the workers are all morphologically
to a challenger than did wasps that founded nests with similar but divide up tasks according to the needs of the
other females. Thus, defense of a resource can favor colony. If we saw such organization on a factory floor,
group living and cooperation. However, many animals we would assume that there was some sort of central
share communal resources but do not have reproduc- command with a manager in charge of assigning tasks.
tive division of labor, so this alone cannot be the sole But how can such complex behavior be carried out by
explanation for eusociality. individuals with brains smaller than pinheads? Is there
some sort of a central organizational structure?
The short answer is no. Instead of a central com-
Day-to-Day Functioning of Eusocial Groups
mand, ants make their decisions about which tasks to
Conflict Individuals within eusocial colonies are gen- perform based on simple rules. These rules depend on
erally not genetically identical and thus may have con- the information they gain from their immediate environ-
flicting goals. We described above how workers and ment and their interactions with others (reviewed in
queens in haplodiploid species might favor different sex Bourke and Franks 1995). For example, red harvester
ratios of the colony’s offspring. Conflict occurs in other ants (Pogonomyrmex barbatus) live in the deserts of the
contexts as well: for example, some wasp colonies are ini- southwestern United States and Mexico. They rely on
tiated by multiple cofoundresses, often sisters. Given patchily distributed seeds that vary over time. After a
that the cofoundresses are not genetically identical, we heavy rain, many buried seeds are exposed, so it makes
predict—and generally see—conflict between them over sense for foraging efforts to increase. When food is
who gets to produce more offspring. In fact, this con- depleted, colony members should not bother going out
 Summary 449

to forage. Deborah Gordon and her colleagues (2008) reproductive success they enjoy will also perpetuate the
tested whether ants inside a colony make decisions about alleles that the altruist shares with them by virtue of their
whether to forage based on the success of other foragers. common descent. The altruist’s relatives are more likely
By the simple strategy of capturing and briefly detain- than nonrelatives to carry the alleles that lead to altru-
ing ants carrying seeds back to the nest, the researchers ism. Another way of saying this is that an individual may
were able to slow down the apparent rate of foraging raise its indirect fitness by helping relatives raise more
success. A decrease in the forager return rate of only offspring than they could without help. We make this
three to five minutes led to an almost immediate reduc- more explicit with Hamilton’s Rule. Hamilton’s Rule
tion in the number of new foragers that left the colony. states that an individual should help its relative raise off-
Instead, those ants could then turn their attention to spring if B/C > 1/r, where B is the benefit to the recipi-
other tasks. We see similar phenomena in other social ent in terms of offspring produced because of the
insects: a bee’s decision whether to collect nectar altruistic act, C is the cost to the actor in terms of off-
depends on the amount of nectar already stored, and a spring it does not produce because of the altruistic act,
wasp’s decision whether to collect more wood pulp for and r is the coefficient of relatedness between the recip-
building depends on how long she had to wait since her ient and the actor.
last load was accepted at the nest (reviewed in Gordon How might an animal identify its relatives? There are
1996). The take-home message is that behaviors that several possibilities. One way might be to use location as
appear to be very complex and highly coordinated can a cue: the individuals that share one’s home are likely to
be generated by a handful of simple decision rules. be kin. Individuals might also be identified as kin because
they are familiar. Animals that have never been encountered
before may be recognized as kin through phenotype
SUMMARY matching. In other words, an individual compares a
stranger’s traits to those of a known family member or
A wide variety of animal species live in groups, from sim- even to itself. Finally, recognition may be genetically
ple aggregations to highly coordinated societies. As with based. Perhaps there are alleles that in addition to label-
many other behaviors we discussed in this book, it’s valu- ing relatives with a noticeable characteristic cause the
able to consider the costs and benefits that group living altruist to assist others that bear the label.
may provide. Benefits can include increased foraging suc-
cess, decreased predation risk, conservation of heat and Reciprocal Altruism Altruism might also evolve in
water by huddling together and conservation of energy spite of the initial cost to the altruist if the service is
by moving together. Costs include increased competition, repaid with interest. In other words, altruism will be
risk of disease and parasites, and interference with repro- favored if the final gain to the altruist exceeds its initial
duction. How costs and benefits are balanced vary cost. However, for reciprocal altruism to work, individ-
according to the species, the ecological circumstances, uals that fail to make restitution must be discriminated
and even across individuals within a group. against. Because of this requirement, three factors make
Altruism is the performance of a service that bene- reciprocal altruism more likely: the benefit of the act to
fits a conspecific at a cost to the one that does the deed, the recipient is greater than the cost to the actor; there
all measured in terms of units of fitness. The occurrence should be a good chance that an opportunity for future
of altruism is puzzling. If aiding a conspecific costs the repayment will arise; and the individuals involved must
altruist, the altruist should leave fewer offspring than do be able to recognize one another.
the beneficiaries of its services. As a result, the alleles for
altruism would be expected to decrease in the popula- Manipulation Sometimes animals can coerce others
tion over generations. Hypotheses for the evolution of into helping them by threatening even greater harm.
altruism can be classified into four overlapping classes. No single hypothesis applies to every example of
altruistic behavior. In addition, these evolutionary mech-
Individual Selection The general thrust of these anisms are not mutually exclusive; more than one may
hypotheses is that when the interaction is examined be responsible for a single example of altruism. To com-
closely enough, the altruist will be found to be gaining, plicate matters even further, similar behaviors may
rather than losing, by its actions. The benefit may not evolve by different mechanisms in different species.
be immediate; sometimes the gain is in the individual’s Members of some social species emit alarm calls to
future reproductive potential. Another way of saying this warn their neighbors of a predator’s approach. Belding’s
is that an individual may receive direct benefits over the ground squirrels emit two types of alarm calls—one in
course of its life by being altruistic. response to terrestrial predators and one in response to
aerial predators. The calls seem to have been selected in
Kin Selection When the beneficiaries of the good different ways. Individual selection seems to be the best
deeds are genetically related to the altruist, the enhanced explanation for the evolution of alarm calls in response to
450 Chapter 19 / Group Living, Altruism, and Cooperation

aerial predators. However, kin selection seems to be the asking two questions. First, we can ask why an offspring
most likely mechanism for the evolution of ground squir- might delay dispersal. As we saw in Chapter 11, disper-
rels’ alarm calls in response to terrestrial predators. sal may be dangerous or entail other costs. In addition,
Belding’s ground squirrels may be something of an excep- other options, such as empty habitat or available mates,
tion: in a comparative study across all rodents, calls gen- may be unavailable. However, just because an offspring
erally serve to communicate with predators rather than kin. delays dispersal does not mean it will help. In a number
Many species cooperate in acquiring a mate. We of species, animals gain indirect fitness by helping to
reviewed four different examples. (1) In wild turkeys, raise kin. In other species, help goes to nonrelatives, so
males display in a group, but only one mates. Generally direct benefits are important. These may include inher-
males display with their siblings, so they receive indirect iting a territory or a mate.
benefits from helping. (2) The situation is more compli- Eusocial species present an extreme example of
cated in lions. Male lions form coalitions to challenge cooperative breeding. Eusocial species have reproductive
other prides. When members of a coalition are related, division of labor (some animals have no offspring), coop-
males can benefit from helping their brothers mate. eration in the care of young, and overlap of at least two
Males also benefit directly from being in larger groups, generations that care for the colony. Eusociality was
even if they are composed of unrelated males. (3) Long- originally described in insects, but the list of eusocial
tailed manakins perform a carefully choreographed species now includes representatives from other taxa,
courtship dance in which displaying males fly around including mammals. Early researchers were intrigued by
each other. Only one of the displaying males mates with a correlation between eusociality and haplodiploidy
the females that are attracted to the show. However, the (a genetic system in which males are haploid and females
subordinate male may become a dominant male if it outlives are diploid). This genetic system means that sisters have,
his partner. Males improve their dancing with practice, on average, a higher level of relatedness to one another
so a long apprenticeship may pay off. (4) Olive baboons than to their own mothers or daughters. By Hamilton’s
form alliances to challenge another male in the troop. In Rule, a higher level of relatedness means that kin selec-
return, they may receive help later, in an example of rec- tion is more likely to evolve. However, haplodiploidy is
iprocal altruism. neither necessary nor sufficient for eusociality to evolve.
Another form of altruism is cooperative breeding. A Other factors that favor the evolution of eusociality are
helper is an individual that assists in the rearing of off- extended parental care, long-lasting sibling associations,
spring that are not its own, usually by providing food or and sharing a central resource, such as a wasps’ nest or
protecting or carrying the young. In most cases, helping the giant tubers fed upon by naked mole rats.
appears to be beneficial, but it is a good idea to confirm Recently, a great deal of work has focused on how
this experimentally. For example, pairs on high-quality eusocial groups function from day to day. It is common
territories may attract additional helpers, and their to see eusocial individuals favor colony-mates that are
increased reproductive output may be due to the help more closely related over those that are more distantly
they have received or to the territory quality. An exper- related. An impressive level of coordinated behavior
iment is needed to tease these hypotheses apart. among hundreds of individuals can arise from simple
Helping is generally costly to the helpers. We can rules, and does not require any particular individual to
ask about the conditions under which helping evolves by have oversight over the colony’s operations.
 Glossary

Acoustic adaptation hypothesis The idea that the acoustic Altricial Describes young that are virtually helpless and inca-
properties, such as pitch, of auditory signals are shaped by pable of feeding on their own or following their parents for
habitat structure. For example, songs in open habitat are the first few weeks after birth or hatching. Compare with
predicted to have higher frequencies than those in heavily precocial.
vegetated habitat. Altruism One animal’s apparent aid to another animal at its
Action potential A nerve impulse. An electrochemical signal own expense.
conducted along an axon. A wave of depolarization caused Amplexus The reproductive embrace in which a male
by the inward flow of sodium ions followed by repolariza- amphibian grasps a female with his front legs from a dor-
tion resulting from the outward flow of potassium ions. sal position.
Activational effects Effects of steroid hormones that typi- Antihormones Drugs that can temporarily and reversibly
cally occur in adulthood and tend to be transient, lasting suppress the actions of specific hormones.
only as long as the hormone is present at relatively high Aposematism See warning coloration.
levels. For example, increases in circulating levels of sex Apostatic selection A form of frequency-dependent selec-
steroids activate mating behavior. Activational effects may tion that occurs when one morph in a polymorphic popu-
involve subtle changes in previously established connec- lation is much more common than another morph.
tions rather than gross reorganization of neural pathways. Predators may develop a search image for the more com-
Compare with organizational effects. mon, rather than the rare, morph, and take more of the
Adaptation (1) The process of natural selection in which common form relative to its frequency in the population.
evolutionary modification occurs in response to selection Predators need not hunt by search image to cause apostat-
pressures. (2) The result of natural selection; that is, ic selection in prey; some predators may simply have an
behavioral, morphological, developmental, or physiologi- aversion to rare or unfamiliar prey.
Artificial selection The process by which humans change
cal changes that have been preserved because they have
the frequency of traits within populations by intentional
had a selective advantage. (3) The immediate physiological
selective breeding. Dog breeds provide an example of arti-
response of an organism to a change in conditions, such as
ficial selection.
an increase in temperature. In this book, we are generally
Associated reproductive pattern Pattern shown by some
concerned with one of the first two definitions.
vertebrates in which there is a close temporal association
Afferent neuron A sensory neuron. A nerve cell that carries
between gonadal activity and mating. Gonadal growth and
information from the peripheral receptors toward the cen-
an increase in circulating levels of sex steroids activate mat-
tral nervous system.
ing behavior.
Aggression A behavior apparently intended to inflict noxious
Associative learning The formation of some sort of mental
effects or destruction on another organism.
connection between representations of two stimuli.
Aggressive mimicry Phenomenon whereby a predator can Audience effect The impact that passive onlookers have on
get close to its prey because it imitates a signal that is not the behavior and physiology of an individual performing a
avoided by the prey or that even attracts it. task.
Agonistic behavior Behavior that encompasses all conflict Autotomy The ability to break off a body part when attacked.
between conspecifics, including threats, submissive behav- Axon A long extension from the cell body of a neuron that
ior, chasing, and physical combat. It is not used to describe carries an electrochemical message away from the cell
aggressive acts between species, such as predation. body toward another neuron or effector (muscle or gland).
Allee effect Phenomenon whereby having a few neighbors is The tips of the axon release a chemical called a neuro-
beneficial, especially for gaining access to potential mates transmitter that can affect the activity of the receiving cell.
and defense against predators. According to this idea, indi- Typically, there is one long axon on a neuron.
vidual fitness increases with number of conspecifics at low Batesian mimicry Phenomenon whereby a palatable species
to moderate densities, and then declines from moderate to adopts the warning characteristics of a noxious or harmful
high densities. With regard to habitat selection, compare species. The harmless species is called the mimic and the
with ideal free distribution. noxious one, the model. By resembling a noxious species,
Allele Alternative form of a gene. One of two or more slight- the mimic gains protection from predators.
ly different versions of a gene that code for different forms Behavioral ecology A discipline that examines the evolu-
of the same trait. tionary and ecological bases of the behavior of animals. It
Allogrooming Caring for the fur, skin, or feathers of a con- grew out of the ethological approach to studying animal
specific; it is distinguished from autogrooming (self- behavior. Behavioral ecologists examine the costs and ben-
grooming). Allogrooming functions in hygiene and also efits of behaviors, such as foraging, mating, group living,
builds and maintains social bonds. and communicating.

451
452 Glossary

Breeding dispersal The movement an animal makes background and thereby reduces risk of detection by visu-
between two successive breeding areas or social groups. It ally hunting predators. It is sometimes called cryptic col-
is sometimes called postbreeding dispersal. oration.
Camouflage A collective term used to describe several Communication The process of transferring information
antipredator devices, such as disruptive coloration, coun- from sender to receiver to the benefit of the sender, on
tershading, and transparency, whose general message is “I average.
am not here.” Comparative method Using the evolutionary relationships
Candidate gene A gene that is hypothesized to be involved in among species in order to gain insights into adaptation.
a particular behavior based on a search of the literature for For example, unrelated species that share similar environ-
genes known to be involved in producing a similar behav- ments may evolve similar traits, while related species that
ior in another organism or by comparing the sequence of live in different environments may evolve different traits.
the gene to the sequences of genes in other organisms using Comparative psychology A branch of psychology that stud-
genome-sequencing data. ies the behavior of animals, with a focus on physiology,
Carnivory Feeding on other animals. learning, and development. The experiments are quantita-
Central pattern generator A neuron or network of neurons tive and often laboratory-oriented.
that is capable of generating a rhythmic pattern of activity in Compass orientation The ability to find one’s way without
motor neurons, even when all sensory input has been using landmarks. The animal keeps a certain angle toward
removed from the system. an external reference system used as a compass. External
Chain of reactions A behavioral repertoire built from a reference systems include the sun, the stars, and the earth’s
sequence of fixed action patterns. Each fixed action pattern magnetic field.
brings the animal into a situation that triggers the next Conditional strategy In agonistic encounters, a strategy that
fixed action pattern in the sequence. is adjusted according to the conditions of a particular fight.
Challenge hypothesis Hypothesis that levels of hormones Conditioned response In the study of classical condition-
involved in dominance and aggression rise during times of ing, the learned behavioral response elicited by the condi-
social challenge or instability, such as during the initial tioned stimulus.
period of territory establishment. Conditioned stimulus In classical conditioning, an innocu-
Chase-away model An explanation for the evolution of ous stimulus, such as a tone or a light, is paired with an
extreme traits in males that exploit a sensory bias of unconditioned stimulus that is rewarding to an animal,
females. Females receive no benefits from being selective such as food. Animals that have learned the pairing
in their choice of mate. The model is based on sexual con- respond to the tone or light, which is then called the con-
flict. ditioned stimulus.
Circadian rhythm Biological rhythms that are about a day in Confusion effect Situation in which predators are less suc-
length (circa, about; diem, a day). Used formally, it refers to cessful in attacking grouped prey because they are unable
a solar-day rhythm that persists in constant conditions of to single out and attack an individual prey.
light and temperature with a period length that is slightly Conspecific attraction hypothesis Hypothesis proposing
longer or shorter than 24 hours. Its period length is rela- that individuals choose habitat patches based on the pres-
tively independent of temperature, and its phase can be ence of established residents of their own species.
reset with pulses of light. Constant reproductive pattern Pattern shown by some
Circalunadian rhythm A basic lunar-day rhythm that per- vertebrates inhabiting harsh environments where suitable
sists in constant conditions with a period that is slightly breeding conditions occur suddenly and unpredictably.
longer or shorter than 24.8 hours. A circalunadian rhythm While waiting for suitable circumstances in which to
is usually bimodal. breed, these species maintain large gonads, mature
Circamonthly rhythm A synodic monthly rhythm that per- gametes, and high circulating levels of sex steroids for pro-
sists in constant conditions with a period length that differs longed periods of time. Sometimes called the opportunis-
slightly from 29.5 days. tic reproductive pattern.
Circannual rhythm An annual rhythm that persists in a con- Constraint In mathematical modeling, a limitation that we
stant light-dark cycle with a period length that differs place on a behavior. For example, we may create a model
slightly from 365 days. in which the amount of food an animal can eat is con-
Circatidal rhythm A rhythm that persists in constant condi- strained by gut capacity.
tions with a period length that differs slightly from 12.4 Continuous reinforcement schedule In the study of learn-
hours. It is the same as a circalunadian rhythm. ing, a reinforcement schedule in which each occurrence of
Classical conditioning In the study of learning, the pairing a desired behavior is rewarded. This reinforcement sched-
of an innocuous stimulus with a rewarding stimulus, so ule is effective when initially training an animal to perform
that the animal learns to respond to the innocuous stimu- a behavior.
lus. This procedure was made famous by Pavlov and his Cooperative hunting Coordinated foraging efforts among
study of dogs. predators.
Coefficient of relatedness (r) The probability that a partic- Copulatory plug A plug made of thick, viscous secretions
ular pair of animals share the same allele through common deposited by males in the reproductive tract of females at
descent. mating. Copulatory plugs occur in many vertebrates,
Coloration matching the visual background An antipreda- including snakes, lizards, marsupials, rodents, bats, and
tor device in which a prey animal’s coloration resembles its primates. Several functions have been suggested for copu-
 Glossary 453

latory plugs, among them: (1) “enforcing chastity” in parental care includes behaviors such as nursing (and feed-
which the plug acts as a barrier to subsequent insemina- ing), grooming, transporting, and huddling with young.
tions; (2) ensuring the retention of sperm in the female Displacement activities Irrelevant actions performed in sit-
reproductive tract; (3) aiding the transport of sperm with- uations in which an animal has conflicting motivations.
in the female reproductive tract; (4) providing for the grad- Display A stereotyped sequence of behaviors that has a sig-
ual release of sperm as the plug disintegrates; and (5) pro- naling function.
viding a means by which a male can scent-mark a female’s Disruptive coloration Coloration designed to prevent per-
body and convey information regarding his identity and ception of a prey animal’s form.
dominance status. Dissociated reproductive pattern Pattern shown by some
Core area The area of a home range in which the activities vertebrates in which mating behavior is completely uncou-
of an animal are concentrated. pled from gamete maturation and secretion of sex steroids.
Countershading Pattern of coloration characterized by dark Gonadal activity occurs only after all breeding activity for
backs and light bellies, which may achieve camouflage the current season has ceased, and gametes are thus pro-
through either self-shadow concealment (obscuring the duced and stored for the next breeding season.
ventral shadow) or background matching. Dominance The ability of an animal to assert itself to others
Counting The process of assigning a tag such as “1, 2, 3” to in acquiring access to a resource such as food, a mate, or a
individual numbers. display or nesting site.
Cross-fostering A technique for detecting maternal influ- Dominant allele The allele that is fully expressed in a het-
ences by transferring shortly after birth the offspring from erozygous individual.
one strain (or species) to the mother of another strain (or Dove In game theory, the strategy of immediately fleeing if
species). confronted by an opponent that is playing the hawk strat-
Cryptic female choice Selection by females who have mated egy, and displaying if confronted by an opponent that is
with several males of the sperm that will fertilize their also playing a dove strategy. See also hawk.
eggs. This ability is described as cryptic because it is a hid- Ecological trap A low-quality habitat that animals prefer
den, internal decision made after copulation.
over a high-quality habitat.
Cue Information transmitted from one animal to another,
Efferent neuron A motor neuron specialized to carry
but not necessarily to the benefit of the sender, such as the
information away from the nervous system to an effector
rustling noise made by a mouse and detected by an owl.
(muscle or a gland).
Currency In mathematical modeling, the common unit in
Emancipation Over evolutionary time, a behavior that has
which costs and benefits of different strategies are measured.
lost its original function and now serves only a commu-
Examples are the number of offspring produced or the
nicative function.
number of calories gained.
Entrainment The process by which an environmental
Dead reckoning See path integration.
rhythm, such as a light-dark cycle, regulates the period and
Dendrite A process of a neuron specialized to pick up mes-
phase of a biological rhythm.
sages and transmit them toward the cell body. There are
Environmental enrichment Efforts made to enrich the lives
typically many short branching dendrites on a neuron.
Deoxyribonucleic acid (DNA) The molecular basis of of captive animals. Such efforts usually involve social hous-
genetic inheritance. A category of nucleic acids that usual- ing and the provision of larger cages, a more complex and
ly consists of a double helix of two nucleotide strands. The variable physical environment with nesting material, for-
sequence of nucleotides carries the instructions for assem- aging devices, toys, hiding places, and the opportunity for
bling proteins. voluntary exercise.
Depolarization A change in the difference in electrical Epigenetics A stable change in a gene that does not involve
charge across a membrane that moves it from a negative changes in nucleotide sequence. Genes become active or
value toward 0 mV. During a nerve impulse (action poten- are shut down due to unwinding or winding of DNA.
tial), depolarization is caused by the inward flow of posi- Ethology The study of animal behavior that focuses largely
tively charged sodium ions. on the evolution and function of behavior. Many early
Detouring The ability to identify an alternative route to a ethological studies were comparative and took place in the
reward when the direct route is blocked. field.
Developmental homeostasis Ability of developmental Evolution A change in the frequencies of alleles in a popula-
processes to buffer themselves against potentially harmful tion of organisms over generations.
influences to produce functional adults. Evolutionary stable strategy (ESS) A behavioral strategy
Dilution effect Phenomenon whereby individuals living in that when adopted by all members of the population can-
groups are safer from predators because each has a smaller not be replaced by a different strategy. A pure ESS con-
chance of becoming the next victim. sists of a single strategy, while a mixed ESS consists of sev-
Direct fitness Fitness gained by an individual through its eral strategies in a stable equilibrium.
own reproduction, generally measured as the number of Excitatory postsynaptic potential (EPSP) A temporary
surviving offspring it produces. Compare with indirect electrical change (depolarization) in the membrane of the
fitness. postsynaptic neuron that is caused by the binding of an
Direct parental care Patterns of behavior performed by par- excitatory neurotransmitter. An EPSP makes it more like-
ents that have an immediate physical impact on offspring ly that the postsynaptic neuron will generate an action
and their survival. In mammals, for example, direct potential.
454 Glossary

Extinction In classical conditioning, presenting a condi- viability have a genetic basis. If a female preference (also
tioned, or learned, stimulus without the reinforcer, leading genetically based) for the male trait should arise, then
to a loss of the learned response over time. these females mate with males carrying genes for the trait
Female defense polygyny Form of polygyny in which a and enhanced viability. In this way, genes for the male
male defends a harem of females. This type of polygyny trait, high viability, and the female preference become
occurs when females live in groups that a male can easily associated.
defend. Habitat selection Process by which animals that disperse
Filial imprinting Process by which a young bird learns, from their natal site or breeding site eventually select a
through exposure to its mother, her particular characteris- new location in which to settle. Habitat selection has three
tics and then preferentially follows her. Filial imprinting phases: (1) search (animal searches for a new habitat); (2)
may function to allow young birds to recognize close rela- settlement (animal arrives in a new habitat and begins to
tives and thereby distinguish their parents from other establish a home range or territory); and (3) residency (ani-
adults that might attack them. mal lives in the new habitat).
Fitness The reproductive success of an allele or an individual Habituation In learning, the waning of a behavioral response
compared to other alleles or individuals in the same popu- to a stimulus because during repeated presentations of that
lation. See also direct fitness, indirect fitness, and inclusive stimulus it was shown to be harmless.
fitness. Hamilton’s rule A rule designed to determine whether an
Fixed action pattern An innate stereotyped motor response animal should behave altruistically toward another animal.
that is initiated by a stimulus but that can continue to com- It includes three variables: the coefficient of relatedness,
pletion without ongoing stimulation. the benefit to the receiver in terms of the number of addi-
Fixed ratio schedule In the study of learning, a reinforce- tional offspring it produces because of the altruistic act,
ment schedule in which the animal must respond a set and the cost to the altruist in terms of the number of off-
number of times before a reward is given. This schedule spring it does not produce because of the altruistic act.
generally results in very high response rates. Handicap principle An example of a good genes model for
Foraging Finding, processing, and eating food.
mate choice. According to this principle, females prefer a
Free-running period The period length of a biological
male with a trait that reduces his chances of survival but
clock when it is not being influenced by external time cues.
announces his superior genetic quality precisely because he
Frequency-dependent selection A type of natural selection
has managed to survive despite his “handicap.” In short,
in which an allele has a greater selective advantage when it
male secondary sexual characteristics act as honest signals,
is rare in a population and a smaller selective advantage
indicating high fitness, and females choose males with the
when it is common. For example, one type of frequency-
greatest handicaps because their superior genes may help
dependent selection occurs when females prefer to mate
produce viable offspring.
with males of a rare phenotype. Frequency-dependent
Handling time The time required to process a food item.
selection results in the fluctuation of allele frequencies
Hawk In game theory, the strategy of immediately attacking
over time.
an opponent. See also dove.
Full song In songbirds, the final adult song that in many
species will remain virtually unchanged for the rest of the Herbivory Feeding on plant material.
male’s life. Sometimes called crystallized song. Heritability The portion of the variability in a specific trait
Game theory A type of mathematical model in which the in a population that is due to genetic factors; the ratio of
optimal strategy of one animal depends on the behavior of the phenotypic variance due to genetic factors to total phe-
its opponent. notypic variance.
Gene flow The movement of genes from one population to Heterospecific attraction hypothesis Hypothesis propos-
another. Gene flow causes populations to become more ing that individuals choose habitat patches based on the
similar to one another. presence of established residents of another species.
Genetic drift Evolutionary change in a population due to Heterozygous The condition of having two different alleles
chance events. Genetic drift is more likely to be an impor- for a particular gene.
tant evolutionary force in small populations than in large Home range The area in which an animal carries out its nor-
populations. mal activities.
Genetic monogamy An exclusive mating relationship Homing See true navigation.
between one male and one female. Compare with social Homozygous The condition of having two identical alleles
monogamy. for a particular gene.
Genomics The study of all the genes in an organism. Honest signals Signals that accurately convey information
Genotype The genetic makeup of an individual. It refers to about the sender. Not all signals are honest, and honest
the precise alleles present. signals are likely to evolve only under particular condi-
Goal-directed emulation Behavior whereby an observer tions.
seems to learn from observation what goal is to be achieved Hot spots Areas in which males gather because they are most
but does not precisely copy what the demonstrator does. likely to encounter receptive females.
Compare with imitation. Hotshots Males that are extremely successful at attracting
Good genes models Models that assume that a particular females. Less successful males may increase their chances
trait in males indicates viability and that both the trait and of mating by staying near these highly successful males.
 Glossary 455

Hypothesis A logical, testable explanation for a specific set Interspecific brood parasites In birds, species that lay their
of observations that serves as the basis for experimentation. eggs in the nests of other species. Interspecific brood par-
Ideal free distribution A method to describe how animals asites never build nests in which to lay eggs and raise their
distribute themselves in space if they know the value of own young, so they are described as obligate brood para-
each habitat and are free to go where they choose. sites.
Imitation An observer’s exact copying of what a demonstra- Intrasexual selection Mechanism of sexual selection where-
tor does. Compare with goal-directed emulation. by members of one sex (usually males) compete with one
Inbred lines Domestic animals that have very little genetic another for access to the other sex. Intense fighting and
variability. They are created by mating close family mem- competition for mates can lead to selection for increased
bers with one another. size and elaborate weapons.
Inclusive fitness The sum of both direct fitness (fitness Intraspecific brood parasites In birds, species that lay their
gained through offspring) and indirect fitness (fitness eggs in the nests of conspecifics. In some cases of intraspe-
gained by helping relatives raise additional offspring they cific brood parasitism, the brood parasite occasionally lays
would not be able to raise without help). eggs in the nests of conspecifics, while still laying eggs in
Indirect fitness Fitness gained by helping relatives raise her own nest. In other instances, the brood parasite lays
additional offspring that they would not have been able to eggs in the nests of conspecifics and does not maintain a
raise on their own. Compare with direct fitness. nest of her own.
Indirect parental care Patterns of behavior performed by Intrauterine position effects In rodents, effects on mor-
parents that may not involve direct physical contact with phology, physiology, and behavior caused by exposure to
offspring, but still affect offspring survival. In mammals, hormones secreted by contiguous littermates during gesta-
for example, indirect forms of parental care include acquir- tion.
ing and defending critical resources, building and main- Ion An atom or group of atoms that carries an electric charge
taining nests or dens, defending offspring against preda- resulting from the loss or gain of electrons.
tors or infanticidal conspecifics, and caring for pregnant or Ion channel A protein-lined pore or channel in a plasma
lactating females. membrane through which one type or a few types of ions
Infanticide The killing of a conspecific infant to acquire cop- can pass. Ion channels in nerve cells are important in the
ulations or resources. Can also occur in response to severe generation and propagation of nerve impulses.
disturbance of the physical or social environment. Kin selection A type of natural selection in which animals
Information center Communal roosts or colonies from which help their relatives reproduce, thereby helping pass the
animals can follow successful conspecifics to food sites. alleles that they share to the next generation. See also coef-
Information sharing Improving foraging by paying atten- ficient of relatedness and direct, indirect, and inclusive
tion to the behavior of group members. fitness.
Infrasound Sounds whose frequencies are below those audible Landmark An easily recognizable cue along a route that can
to humans, which means frequencies lower than about 20 be quickly stored in memory to guide a later journey.
Hz. Some animals, such as elephants, use infrasound to com- Language A term that has four elements: first, words or signs
municate over long distances. Other animals may use infra- must be used as true symbols that can stand for, or take the
sound during orientation. place of, a real object, event, person, action, or relation-
Inhibitory postsynaptic potential (IPSP) A temporary ship. Second, symbols should permit reference to objects
electrical change (hyperpolarization) in the membrane of or events that are not present. Third, there should be some
the postsynaptic neuron that is caused by the binding of an elements of grammar, or rules that determine the relation-
inhibitory neurotransmitter. An IPSP makes it less likely ship between words. Fourth, words or signs should be
that the postsynaptic neuron will generate an action poten- combined to form novel phrases or sentences that are
tial. understandable to others.
Insight An example of animal cognition in which under- Latent learning Learning that occurs without any obvious
standing seems to occur suddenly and without practice. immediate reward. For instance, an animal can learn
Intention movements A behavior, such as spreading the important characteristics of its environment during unre-
wings for flight, that signals an animal’s behavior to an warded explorations and then use this information later.
observer. Intention movements are thought to be a route by Learning A change in the capacity for behavior as a result of
which some signals evolve. experience, excluding the effects of fatigue, sensory adap-
Intermediate-term memory Memory of new information tation, or maturation of the nervous system.
that may last a day or so. Habituation and sensitization are Lek polygyny Form of polygyny in which males defend
examples of intermediate-term memory formation. “symbolic” territories that are often located at traditional
Interneuron An association neuron. Neurons located within display sites called leks. Males of lek species do not provide
the central nervous system between sensory and motor parental care and defend only their small territory on the
neurons that integrate information. lek, not groups of females that happen to be living togeth-
Intersexual selection Mechanism of sexual selection where- er nor resources associated with specific areas. Females
by members of the sex in demand (usually females) choose visit these display arenas, select a mate, copulate, and leave.
mates with certain preferred characteristics. Thus, males Local enhancement A type of social learning in which an
compete to attract females through the elaboration of animal is attracted to a particular location because a con-
structures or behavior patterns. specific is there.
456 Glossary

Long-term depression (LTD) A lasting decrease in respon- that it has been spotted early in its hunting sequence; (3)
siveness of postsynaptic neurons after sensory neurons alerting others, particularly relatives, of the danger; and
have received a slow train of stimuli. LTD is a mechanism (4) providing an opportunity for others, again particularly
that weakens the effectiveness of a synapse, decreasing the relatives, to learn to recognize and fear the object that is
magnitude of a response by the postsynaptic cell. It may being mobbed.
play a role in memory formation. Model organism A species chosen to study particular bio-
Long-term memory Memory that lasts weeks, months, or logical principles with the expectation that the information
years. Changes in synapses underlie long-term memory. learned from the model organism can be applied to other
Long-term potentiation (LTP) A stable and long-lasting species.
responsiveness to an action potential by the receiving neu- Monogamy Mating system in which a male and female have
ron caused by rapidly repeated strong stimulation. LTP only a single mating partner per breeding season. Further
strengthens the connections between the adjacent neurons. classified as genetic or social monogamy.
It is the molecular mechanism that underlies the acquisition Motor neuron A neuron specialized to carry information
and storage of memories. away from the nervous system to an effector (muscle or
Lordosis The copulatory posture that some female mammals gland).
assume when ready to mate. The posture typically involves Müllerian mimicry Antipredator strategy in which two
ventral curvature of the vertebral column. warningly colored species look alike. Two noxious species
Macroevolution Large-scale evolutionary changes in lin- may benefit from a shared pattern because predators con-
eages, generally observed over geological time. Compare sume fewer of each species in the process of learning to
with microevolution. avoid all animals of that general appearance.
Major histocompatibility complex (MHC) A large chro- Multimodal communication Communication that contains
mosomal region that varies tremendously among individu- signals from two or more sensory modalities (e.g., vision,
als and is important in the immune responses that protect audition, and touch). Signaling in different channels can
against disease-causing organisms. Some animals may occur either simultaneously or sequentially and the mes-
identify kin and choose mates based on MHC genes. sages can be redundant or nonredundant.
Marginal value theorem A model that predicts when a for- Mutation A change in the DNA sequence of an organism,
aging animal should leave a patch of food. It is generally such as an addition or a deletion of nucleotides.
depicted in a graph. Myelin sheath An insulating layer around axons of nerve cells
Mate guarding Strategies employed by males to increase the that carry action potentials (nerve impulses) over relatively
probability that their sperm, and not the sperm of a com- long distances. It is composed of multiple wrappings of the
petitor, will fertilize the eggs of a particular female. Mate plasma membrane of certain glial cells. The myelin sheath
guarding can occur before or after copulation or both. greatly increases the speed at which impulses travel. The cells
Meta-analysis A statistical approach in which one collects all that form the myelin sheath are separated from one another
the studies that test a particular hypothesis and then com- by short regions of exposed axon. The nerve impulse “jumps”
bine the results in order to look for significance in the from one exposed region of axon to the next.
overall pattern. Natal dispersal The movement an animal makes from its
Microarray analysis An analysis that reveals which of hun- natal area or social group to the area or social group where
dreds or thousands of genes are active at any moment. it first breeds.
Thousands of genes are stamped on a solid surface, and Natal habitat preference induction (NHPI) Phenomenon
molecular tags are used to identify the messenger RNA whereby an animal’s experience in its natal habitat induces a
(mRNA) produced by each gene. The more active a gene, preference for a postdispersal habitat with similar qualities.
the more mRNA is produced. Genes that are active only Natal philopatry Occurs when offspring remain at their
during a particular behavior may play a role in producing natal area and share the home range or territory with their
that behavior. parents.
Microevolution Minor evolutionary changes within a Natural selection Process by which organisms with favor-
species. Compare with macroevolution. able characteristics are more likely to survive and repro-
Migration Movement away from the home range that does duce than those with unfavorable characteristics.
not stop upon encountering the first suitable location. Negative-assortative mating A mating pattern in which
Migrating animals continue to move until they eventually animals preferentially mate with those with phenotypes
become responsive to the presence of resources, such as different from their own.
nest sites and food, and then they stop. Within a particular Neurite A small-diameter process extending from a neuron’s
species, migratory movements occur over greater distances cell body. A neurite can be either an axon or a dendrite.
than dispersal movements. Neurogenesis The creation of new neurons.
Mobbing An antipredator strategy whereby prey approach, Neuromodulators Chemicals that cause voltage changes
gather around, and harass their predators. Mobbing is that occur over seconds, minutes, hours, and perhaps even
usually initiated by a single individual, and then con- days. Neuromodulators alter neuronal activity slowly, by
specifics, or members of another species, join. The possi- biochemical means. Different neuromodulators can act on
ble functions of mobbing include, but are not limited to, the same neural elements to produce different behaviors.
(1) confusing the predator; (2) discouraging the predator Neurons Nerve cells involved in intercellular communi-
either through harassment or through the announcement cation.
 Glossary 457

Neurosteroids Steroid hormones produced by the nervous Peptide hormones Water-soluble hormones that cannot
system that act in the nervous system on either nearby cells pass through the plasma membrane of target cells on their
or the same cell that produced the hormone. This mode of own, so they influence cells indirectly through second
action contrasts with peripheral steroids produced by the messenger systems. These hormones activate existing
gonads or adrenal glands that travel in the bloodstream enzymes. Examples of peptide hormones are luteinizing
and act on target cells some distance from their gland of hormone (LH) and follicle-stimulating hormone (FSH)
origin. produced by the anterior pituitary gland.
Neurotransmitter A chemical released from the axon tip of Period The time interval of one complete cycle of a biologi-
a neuron that affects the activity of another cell (usually a cal rhythm.
nerve, muscle, or gland cell) by altering the electrical Peripheral-control hypothesis Hypothesis proposing that
potential difference across the membrane of the receiving rhythmic behavior occurs because the first movement
cell. stimulates sensory receptors, which in turn trigger the next
movement in the sequence. The second movement stimu-
Nuptial gifts Food or other valuable substances offered by
lates other sensory receptors that trigger the first compo-
males to females during courtship or mating.
nent. Thus, sensory feedback is necessary for this hypoth-
Observational conditioning A type of classical conditioning
esis.
that takes place when an animal observes another animal
Phenotype The observable physical and physiological traits
rather than going through a conditioning experience
of an individual. Phenotype results from the inherited alle-
directly. les and their interactions with the environment.
Omnivory Feeding on both plants and animals. Phenotype matching A type of kin recognition accomplished
Operant conditioning A form of associative learning in which by assessing the degree of similarity between one’s pheno-
the outcome (positive or negative) depends on the animal’s type and that of another individual, or by learning the phe-
behavior, such as an animal that learns to run a maze in order notypes of relatives.
to receive a reward. Also called trial-and-error learning and Pheromones Chemicals that convey information to other
instrumental conditioning. members of the same species. Some pheromones, called
Operational sex ratio The ratio of potentially mating males releaser pheromones, have an immediate effect on the
to fertilizable females. recipient’s behavior. Other pheromones, known as primer
Optimality modeling The use of mathematical procedures pheromones, exert their effect more slowly, by altering the
to weigh the costs and benefits of different strategies and physiology and subsequent behavior of the recipient.
determine which strategy provides the maximum award Piloting The ability to find one’s way using landmarks.
under a specified set of assumptions. Play Behavior that borrows pieces of other behavior patterns,
Organizational effects Effects of steroid hormones that usually incomplete sequences and often in an exaggerated
occur early in life, usually just before or after birth or form. It consists of elements drawn from other, functional-
hatching, and tend to be permanent. This permanence ly different behavior patterns juxtaposed in new sequences.
implies structural changes in the brain or nonneural sys- Play includes social play, locomotor play, and object play.
tems. For example, steroid hormones around hatching Players Individuals whose strategies are modeled in game
organize the brain of a male songbird to make him capable theory.
of singing in adulthood. Compare with activational effects. Polyandry Mating system in which a female has more than
Own-species bias In studies of bird song, the preference of one mate during a breeding season.
young male songbirds to learn their songs from members of Polygyny Mating system in which one male mates with more
their own species. than one female during a breeding season.
Polygyny threshold hypothesis Hypothesis proposing that
Parental investment Any investment by parents in an off-
polygynous matings will be advantageous to females when
spring that increases the survival of that offspring while
the benefits achieved by mating with a high-quality male
decreasing the ability of the parent to invest in other off-
and gaining access to his resources more than compensate
spring.
for costs. Thus, a female may reproduce more successfully
Path integration Process by which an animal integrates
as a secondary mate on a high-quality territory than as a
information on the sequence of direction and distance
monogamous mate on a low-quality territory. The term
traveled during each leg of the outward journey. Then, polygyny threshold describes the difference in a territory’s
knowing its location relative to home, the animal can head quality needed to make secondary status a better repro-
directly there, using its compass(es). A compass may also ductive option for females than primary status.
be used to determine the direction traveled on each leg of Polymorphism Phenomenon whereby a species occurs in
the outward journey, or the direction may be estimated several different shapes and/or color forms. This may pre-
from the twists and turns taken, sounds, smells, or even the vent predators from forming search images.
earth’s magnetic field. Information from the outward jour- Potential reproductive rate The maximum number of
ney is used to calculate the homeward direction (vector). independent offspring that each parent can produce per
Also called dead reckoning. unit of time.
Payoff matrix In game theory, a table used to organize the Precocial Describes young that are capable of moving about
values of each strategy when it is played against each of the and feeding on their own just a short time after birth or
other strategies. hatching. Compare with altricial.
458 Glossary

Quantitative trait A trait that is influenced by several genes Ritualization Phenomenon whereby over evolutionary
acting together, rather than by any one gene acting alone. time, incipient signals become more stereotyped and
Receiver-bias mechanism Hypotheses about the evolution unambiguous.
of communication that focus on the characteristics of the Round dance A circling dance performed by honeybees that
receiver of signals rather than those of the sender. indicates a food source is nearby.
Recessive allele The allele whose effects are usually masked Rule of thumb A term indicating that instead of following
in the heterozygous condition. precisely the behavioral strategy that would be optimal, an
Reciprocal altruism One animal helps another, which in animal may follow a rule of thumb that yields adequate
turn helps the first animal at a later time. results under most circumstances.
Recognition allele An allele or group of alleles that enable Runaway selection A form of sexual selection in which a
an animal to recognize others with the same genotype. positive feedback loop is created when genes for mate
Recombination The production of a new combination of choice in the female become genetically linked to genes
genes in offspring that differ from either of the parental for the preferred traits in males. In this way, runaway
genotypes, generally through the process of crossing over selection can produce increasingly exaggerated male traits
during meiosis. and a stronger female preference for them.
Regulatory gene A gene that influences the activity of other Satellite male A male that remains silent and associates
genes. Regulatory genes often produce transcription factors. closely with a signaling male, ready to intercept females
Reinforcement schedule In the study of learning, the fre- that are attracted to the other male’s signals or resources.
quency and timing with which an experimenter rewards a Scent-marking The act of strategically placing a chemical
particular behavior. mark in the environment.
Reinforcer A stimulus that changes the probability that an Schreckstoff An alarm chemical produced by some species
animal will repeat its behavior. Reinforcers can be positive of fish when physically attacked.
or negative. Scientific method A procedure underlying most scientific
Relative plasticity hypothesis The hypothesis relating investigations that involves observation, formulating a
alternative phenotypes to the different effects of steroid hypothesis, making predictions, experimenting to test the
hormones. It states that developmentally fixed alternative predictions, and drawing conclusions. Experimentation usu-
phenotypes (i.e., individuals are one phenotype or the ally includes a control group and an experimental group that
other and remain so for life) rely on organizational effects differ in one or very few factors (variables). New hypotheses
of steroid hormones, whereas developmentally plastic may be generated from the results of experimentation.
alternative phenotypes (i.e., individuals switch between Search image The heightened ability to detect a particular
phenotypes in response to environmental conditions) rely target with experience.
on activational effects of steroid hormones. Search time The time it takes a forager to find food.
Releaser A sign stimulus that is emitted by a member of the Seismic signals Signals that are encoded in the pattern of
same species. vibrations of the environmental substrate. These signals
Repolarization The return of the membrane potential to can be produced through percussion on the environmental
approximately its resting value. Repolarization of the nerve substrate, such as when rodents drum their feet on the
cell membrane during an action potential occurs because ground or insects tap the surface of water to create ripples.
of the outflow of potassium ions. Seismic signals can also be generated when an airborne
Resource defense polygyny Form of polygyny in which vocalization couples with the ground, such as when ele-
males defend resources essential to female reproduction phants produce infrasonic calls.
(e.g., nest sites or food) rather than defending females Selfish herd An antipredator mechanism that considers the
themselves. spatial arrangement of individuals within a group. In most
Resource-holding potential The ability of an animal to groups, centrally located animals are safer than those at the
defend a resource from competitors. edges. By obtaining a central position, animals can
Resting potential The separation of charge across the plas- decrease their chances of being attacked and increase the
ma membrane of a neuron when the neuron is not trans- probability that one of their more peripheral colleagues
mitting an action potential. It is primarily caused by the will be eaten instead. This mechanism emphasizes that
unequal distribution of sodium ions, potassium ions, and although a given group appears to consist of members that
large negatively charged proteins on either side of the plas- coordinate their escape efforts, it is actually composed of
ma membrane. The resting potential of a neuron is about selfish individuals, each trying to position as many others
–70 mV. as possible between itself and the predator.
Ribonucleic acid (RNA) A single-stranded nucleic acid that Self-shadow concealment The mechanism by which coun-
plays several roles in protein synthesis. tershading can achieve camouflage by obscuring the ven-
Risk sensitivity When an animal’s decisions depend on the tral shadow. An alternative mechanism is background
variance, or riskiness, of an option, it is said to be risk sen- matching.
sitive. An example of a risky option is a foraging patch that Sensitive period A time during development when certain
has a fluctuating amount of food. A risk-prone animal experiences have a greater influence on the characteristics
chooses a risky option because there is a chance it will pro- of an individual than at other stages.
duce a high payoff. A risk-averse animal chooses a safe Sensitization A simple type of learning that involves an
option that has a low variance. enhanced response to repeated stimuli.
 Glossary 459

Sensorimotor stage In songbirds, phase when singing Short-term memory Memory of new information that lasts
learned song actually begins. Birds retrieve a learned song for a few seconds or minutes.
from memory and rehearse it, constantly matching their Siblicide The killing of a sibling. Siblicide is most common
sounds to those they memorized months earlier during the in those species in which parents face limited resources and
sensory phase. deposit eggs or young in a “nursery” with limited space;
Sensory bias model A model for mate choice which states the nursery could be a uterus, a brood pouch, a parent’s
that female preferences for certain traits in males could back, a nest, or a den. Siblicide can be advantageous to par-
evolve because male traits stimulate an existing bias in the ents when more young are produced than can be raised
female’s sensory system. The original bias might relate to
successfully.
feeding or avoiding predators. For example, females might
Sign stimulus A stimulus that triggers a fixed action
have a sensory bias to help them find food of a particular
pattern.
color; males can then exploit this bias by using the same
Social learning Learning in which animals acquire informa-
color to attract females during courtship.
Sensory drive In the study of communication, the hypothe- tion from other animals.
sis suggesting that receiver characteristics are shaped by Social monogamy An exclusive living arrangement between
the environment and will thus affect signal evolution. one male and one female that makes no assumptions about
Sensory exploitation In communication, when a receiver mating exclusivity or biparental care. Compare with genet-
has a preexisting bias for a particular stimulus and a ic monogamy.
sender’s signal evolves to exploit that bias. Societies Structured animal groups.
Sensory neuron A nerve cell that carries information from Sociobiology A discipline that applies the principles of
the peripheral receptors toward the central nervous system. evolution to social behavior.
Sensory phase In songbirds, phase during which songs are Sodium-potassium pump A molecular mechanism in a
learned and stored in memory for months without plasma membrane that uses cellular energy in the form of
rehearsal. adenosine triphosphate (ATP) to pump ions against their
Sex allocation The manner in which parents distribute concentration gradients. Typically, each pump ejects
resources between the production of sons and daughters. three sodium ions from the cell while bringing in two
Parents can bias their allocation of resources in two main potassium ions.
ways: they can either produce more offspring of one sex or Sperm competition Competition that results when two or
they can provide more (or better) resources to offspring of
more males have deposited sperm in the reproductive tract
one sex.
of one female.
Sex-role reversal Pattern shown by some species in which
Sperm heteromorphism The simultaneous production by a
parental investment by males exceeds that of females. In
these species, females often compete for access to males single male of at least two types of sperm in the same ejac-
and males are selective in their choice of mates. ulate.
Sexual conflict A conflict between the evolutionary interests Spermatophore A packet used by many species to transfer
of males and females. It takes several forms, but the two sperm from the male to the female.
main ones concern mating/fertilization and parental Star compass orientation The ability to use the star con-
investment. stellations in the northern part of the sky to determine
Sexual dimorphism A difference, for example in behavior or compass directions.
appearance, between the sexes. Steroid hormones A group of closely related hormones
Sexual imprinting Learning process through which experi- chemically derived from cholesterol and secreted by the
ence with parents and siblings early in life influences sexu- gonads and adrenal glands in vertebrates. The four major
al preferences in adulthood. classes of steroids include progestogens, androgens, estro-
Sexual interference Any behavior that reduces a rival’s fit- gens, and corticosteroids. Steroid hormones are fat-soluble,
ness by decreasing his mating success. so they move easily through the plasma membranes of tar-
Sexual selection A form of natural selection that occurs get cells into the cell interior, where they affect gene expres-
through male competition for access to mates and female sion and protein synthesis. See neurosteroids for steroid
choice of mates.
hormones produced by the nervous system.
Sexy son hypothesis Hypothesis proposing that access to
Stimulus enhancement A type of social learning in which an
good genes for offspring compensates a female for the
animal is attracted to a particular object because a conspe-
costs of polygyny. A female may benefit from mating with
cific is near it or is interacting with it.
an already mated male if her sons inherit the genes that
made that male attractive. Her sexy sons will presumably Stotting A stiff-legged bounding display performed by many
provide her with many grandchildren, so the female’s life- species of deer and antelope which appears to have several
time reproductive success may be enhanced by choosing to functions, including announcing to a predator that it has
mate with a male that is attractive to many females. been detected.
Shaping In learning, changing the behavior of an animal by Strategy A behavioral option available to an animal, often
rewarding increasingly closer approximations of the used in the context of mathematical models.
desired behavior. This procedure often is used in animal Submissive individual An animal that predictably yields to a
training. dominant animal.
460 Glossary

Subsong Vocalizations produced by young songbirds during cetaceans, bats, and rodents, produce and detect ultra-
the sensory phase that do not involve retrieval or rehearsal sounds as part of echolocation or communication systems.
of previously learned material. Ultrasound is not restricted to mammals.
Sun compass orientation The ability to use the sun as a Unconditioned stimulus In classical conditioning, a
visual cue to determine compass directions. Because the rewarding stimulus (such as food) that is paired with an
sun appears to move across the sky throughout the day, an innocuous stimulus (such as a tone or light).
animal must constantly change the angle it maintains with Variable ratio schedule In learning, a reinforcement sched-
the sun to remain headed in the same compass direction. ule in which the number of responses required to receive a
Thus, sun compass orientation is usually time compensat- reward varies randomly. Because of the variability, the ani-
ed. mal cannot easily detect when reinforcement has stopped,
Suspension feeding Removing small food particles suspend- so the response tends to persist. This is the reinforcement
ed in the water by means of several techniques. schedule used in slot machines.
Synapse The site of communication between a neuron and Vector navigation An inherited program that tells an animal
another cell, such as another neuron or a muscle cell. the compass direction to head in and for how long. Certain
Synaptic remodeling A refinement of synaptic connections species of migratory birds use this type of navigation on
caused by the development of new synapses and the loss of their first migration.
others that often occurs during development. Vomeronasal organ An accessory olfactory sense organ
Syntax In communication, a change in the order of symbols found in some species of amphibians, reptiles, and mam-
that alters the meaning of the message. mals. Located in the roof of the mouth or between the
Territory As used here, a defended space. Many other defin- nasal cavity and the mouth, the vomeronasal organ is
itions exist. anatomically separate from other chemosensory structures,
Tool use Use of an object in order to obtain a goal. and its neural wiring goes to brain regions other than those
Traditions Learned behaviors that are stable in a group over associated with the main olfactory system. Sometimes
time. called Jacobson’s organ, the vomeronasal organ can be
True navigation The ability to maintain or establish refer- stimulated by either pheromones or general odorants.
ence to a goal, regardless of its location, without the use of Waggle dance A figure-eight dance performed by honeybees
landmarks. Sometimes called homing. that indicates the direction and distance of food resources.
Ultrasound Sounds whose frequencies are above those audi- Warning coloration The phenomenon by which conspicuous
ble to humans, which means frequencies greater than coloration advertises dangerous or unpleasant attributes.
about 20 kHz. Several groups of mammals, including
 References

Abbott, D. H., E. B. Keverne, F. B. Bercovitch, C. A. Shively, Aidley, D. J. 1981. Questions about migration. In Animal
S. P. Mendoza, W. Saltzman, C. T. Snowdon, T. E. Ziegler, Migration, edited by D. J. Aidley. New York: Cambridge
M. Banjevic, T. Garland, Jr., and R. M. Sapolsky. 2003. Are University Press, pp. 1–8.
subordinates always stressed? A comparative analysis of rank Åkesson, S., and A. Hedenström. 2007. How migrants get
differences in cortisol levels among primates. Hormones and there: Migratory performance and orientation. BioScience
Behavior 43:67–82. 57:123–133.
Able, K. P. 1980. Mechanisms of orientation, navigation, and Åkesson, S., and R. Wehner. 2002. Visual navigation in desert
homing. In Animal Migration Orientation and Navigation, ants Cataglyphis fortis: are snapshots coupled to a celestial
edited by S. A. Gauthreaux, Jr. New York: Academic Press. system of reference? Journal of Experimental Biology
Able, K. P. 1982. Field studies of avian nocturnal migratory 205:1971–1978.
orientation I. Interaction of sun, wind, and stars as direc- Alberts, S. C., J. C. Buchan, and J. Altmann. 2006. Sexual
tional cues. Animal Behaviour 30:761–767. selection in wild baboons: from mating opportunities to
Able, K. P. 1993. Orientation cues used by migratory birds: A paternity success. Animal Behaviour 72:1177–1196.
review of cue-conflict experiments. Trends in Ecology and Alcock, J. 2001. Animal Behavior. Sunderland, MA: Sinauer.
Evolution 8:367–371. Alerstam, T. 2003. Bird migration speed. In Avian Migration,
Able, K. P. 1996. The debate over olfactory navigation by edited by P. Berthold, E. Gwinner, and E. Sonnenschein.
homing pigeons. Journal of Experimental Biology Berlin, Germany: Springer-Verlag, pp. 253–267.
199:121–124. Alexander, R. D. 1974. The evolution of social behavior.
Able, K. P., and M. A. Able. 1996. The flexible migratory ori- Annual Review of Ecology and Systematics 5:325–383.
entation system of the Savannah sparrow (Passerculus sand- Alexander, R. D. 1975. Natural selection and specialized cho-
wichensis). Journal of Comparative Neurology 199:3–8. rusing behavior in acoustical insects. In Insects, Science, and
Abramsky, M. V., M. L. Rosenzweig, and A. Subach. 2002. Society, edited by D. Pimental. New York: Academic Press,
The costs of apprehensive foraging. Ecology 83:1330–1340. pp. 35–77.
Abu-Gideiri, Y. B. 1966. The behaviour and neuro-anatomy Alexander, R. D. 1979. Darwinism and Human Affairs. Seattle:
of some developing teleost fishes. Journal of Zoology University of Washington Press.
149:215–241. Allee, W. C. 1951. The Social Life of Animals. Boston: Beacon
Abzhanov, A., W. P. Kuo, C. Hartmann, B. R. Grant, P. R. Press.
Grant, and C. J. Tabin. 2006. The calmodulin pathway and Allen, B. J., and J. S. Levinton. 2007. Costs of bearing a sex-
evolution of elongated beak morphology in Darwin’s ually selected ornamental weapon in a fiddler crab.
finches. Nature 442:563–567. Functional Ecology 21:154–161.
Aceves-Pina, E. O., and W. G. Quinn. 1979. Learning in nor- Allen, J. A. 1988. Frequency-dependent selection by preda-
mal and mutant Drosophila. Science 206:93–96. tors. Philosophical Transactions of the Royal Society of London B
Adams, J., and R. L. Caldwell. 1990. Deceptive communica- 319:485–503.
tion in asymmetric fights of the stomatopod crustacean Altmann, S. A., and J. Altmann. 2003. The transformation of
Gonodactylus bredini. Animal Behaviour 39:706–717. behaviour field studies. Animal Behaviour 65:413–423.
Adkins-Regan, E. 2005. Hormones and Animal Social Behavior. Alvarez, F. 1990. Horns and fighting in male Spanish ibex,
Princeton: Princeton University Press. Capra pyrenaica. Journal of Mammalogy 71:608–616.
Adkins-Regan, E., and C. H. Leung. 2006. Sex steroids mod- Amholt, R. R. H. 2004. Genetic modules and networks for
ulate changes in social and sexual preference during juve- behavior: lessons from Drosophila. BioEssays 26:1299–1306.
nile development in zebra finches. Hormones and Behavior Anderson, J. B., and L. P. Brower. 1996. Freeze-protection of
50:772–778. overwintering monarch butterflies in Mexico: Critical role
Adkins-Regan, E., J.-P. Signoret, and P. Orgeur. 1989. Sexual of the forest as a blanket and an umbrella. Ecological
differentiation of reproductive behavior in pigs: Entomology 21:107–116.
Defeminizing effects of prepubertal estradiol. Hormones and Andersson, M. 1994. Sexual Selection. Princeton, NJ: Princeton
Behavior 23:290–303. University Press.
Adler, K. 1976. Extraocular photoreception in amphibians. Andersson, M., and L. W. Simmons. 2006. Sexual selection
Photochemistry and Photobiology 23:275–298. and mate choice. Trends in Ecology and Evolution 21:296–302.

461
462 References

Andrade, M. C. B. 1996. Sexual selection for male sacrifice in Avens, L., and K. J. Lohmann. 2004. Navigation and seasonal
the Australian redback spider. Science 271:70–72. migratory orientation in juvenile sea turtles. Journal of
Antonov, A., B. G. Stokke, A. Moksnes, and E. Røskaft. 2008. Experimental Biology 207:1771–1778.
Getting rid of the cuckoo Cuculus canorus egg: why do hosts Axelrod, R. 1984. The Evolution of Cooperation. New York: Basic
delay rejection? Behavioral Ecology 19:100–107. Books.
Anway, M. D., C. Leathers, and M. K. Skinner. 2006. Endocrine Ayala, F. J., and C. A. Campbell. 1974. Frequency-dependent
disruptor vinclozolin induced epigenetic transgenerational selection. Annual Review of Ecology and Systematics 5:115–138.
adult-onset disease. Endocrinology 147:5515–5523. Back, S. R., L. A. Beeler, R. L. Schaefer, and N. G. Solomon.
Aoki, S. 1972. Colophina clematis (Homoptera, Pemphigidae), 2002. Testing functional hypotheses for the behavior of res-
an aphid species with “soldiers.” Kontyu 45:276–282. ident pine voles, Microtus pinetorum, toward non-residents.
Aoki, S. 1979. Further observations on Astegopteryx styracicola Ethology 108:1023–1039.
(Homoptera, Pemphigidae), an aphid species with soldiers Backwell, P. R. Y., J. H. Christy, S. R. Telford, M. D. Jennions,
biting man. Kontyu 47:99–104. and N. I. Passmore. 2000. Dishonest signaling in a fiddler
Aoki, S. 1982. Soldiers and altruistic dispersal in aphids. In The crab. Proceedings of the Royal Society of London B 267:719–724.
Biology of Social Insects, edited by M. D. Breed, C. D. Baeza, J. A. 2008. Social monogamy in the shrimp Pontonia
Michener, and H. E. Evans. Boulder, CO: Westview, pp. margarita, a symbiont of Pinctada mazatlanica, off the Pacific
154–158. coast of Panama. Marine Biology 153:387–395.
Arak, A. 1988. Callers and satellites in the natterjack toad: Baker, R. R. 1980. The Mystery of Migration. London: Macdonald.
Evolutionarily stable decision rules. Animal Behaviour
Baker, B. S., B. J. Taylor, and J. C. Hall. 2001. Are complex
36:416–432.
behaviors specified by dedicated regulatory genes?
Archer, J. 1988. The Behavioural Biology of Aggression. Reasoning from Drosophila. Cell 105:13–24.
Cambridge: Cambridge University Press.
Baker, P. J., C. P. J. Robertson, S. M. Funk, and S. Harris.
Armstrong, D. P. 1991. Aggressiveness of breeding territorial
1998. Potential fitness benefits of group living in the red fox,
honeyeaters corresponds to seasonal changes in nectar avail-
Vulpes vulpes. Animal Behaviour 56:1411–1424.
ability. Behavioral Ecology and Sociobiology 29:103–112.
Bakken, G. S., and A. R. Krochmal. 2006. The imaging prop-
Arnold, A. P., and S. M. Breedlove. 1985. Organizational and
erties and sensitivity of the facial pits of pit vipers as deter-
activational effects of sex steroids on brain and behavior: A
mined by optical and heat-transfer analysis. Journal of
reanalysis. Hormones and Behavior 19:469–498.
Experimental Biology 210:2801–2810.
Arnold, E. N. 1988. Caudal autotomy as a defense. In Biology
Balda, R. P. 1980. Recovery of cached seeds by a captive
of the Reptilia, edited by C. Gans and R. B. Huey. New York:
Nucifraga caryocatactes. Zeitschrift für Tierpsychologie
Alan Liss, pp. 235–273.
52:331–346.
Arnold, S. J. 1976. Sexual behavior, sexual interference and
Balda, R. P., and A. C. Kamil. 1989. A comparative study of
sexual defense in the salamanders Ambystoma maculatum,
cache recovery by three corvid species. Animal Behaviour
Ambystoma tigrinum and Plethodon jordani. Zeitschrift für
38:486–495.
Tierpsychologie 42:247–300.
Arnott, G., and R. W. Elwood. 2007. Fighting for shells: How Balda, R. P., and A. C. Kamil. 1998. The ecology and evolu-
private information about resource value changes hermit tion of spatial memory in corvids of the southwestern USA:
crab pre-fight displays and escalated fight behavior. the perplexing pinyon jay. In Animal Cognition in Nature,
Proceedings of the Royal Society of London B 274:3011–3017. edited by R. P. Balda, I. M. Pepperberg, and A. C. Kamil.
Arnqvist, G., and T. Nilsson. 2000. The evolution of San Diego: Academic Press, pp. 29–64.
polyandry: multiple mating and female fitness in insects. Balda, R. P., I. M. Pepperberg, and A. C. Kamil. 1998. Preface.
Animal Behaviour 60:145–164. In Animal Cognition in Nature, edited by R. P. Balda, I. M.
Asdell, S. A. 1946. Patterns of Mammalian Reproduction. Ithaca, Pepperberg and A. C. Kamil. San Diego: Academic Press.
NY: Comstock. Baldaccini, E. N., S. Benvenuti, V. Fiaschi, and F. Papi. 1975.
Asher, G., D. Gatfield, M. Stratmann, H. Reinke, C. Dibner, Pigeon navigation: Effects of wind deflection at the home
F. Kreppel, R. Mostoslavsky, F. W. Alt, and U. Schibler. cage on homing behavior. Journal of Comparative Physiology
2008. SIRT1 regulates circadian clock gene expression 99:177–196.
through per2 deacetylation. Cell 134:317–328. Bandura, A. 1962. Social learning through imitation. In
Aubin-Horth, N., B. H. Letcher, and H. A. Hofmann. 2005a. Nebraska Symposium on Motivation, edited by M. R. Jones.
Interaction of rearing environment and reproductive tactic Lincoln: University of Nebraska Press.
on gene expression profiles in Atlantic salmon. Journal of Baptista, L. F., and L. Petrinovich. 1984. Social interaction,
Heredity 96:261–278. sensitive phases and the song template hypothesis in the
Aubin-Horth, N., C. R. Landry, B. H. Letcher, and H. A. white-crowned sparrow. Animal Behaviour 32:172–181.
Hofmann. 2005b. Alternative life histories shape brain gene Baptista, L. F., and L. Petrinovich. 1986. Song development
expression profiles in males of the same population. in the white-crowned sparrow: Social factors and sex dif-
Proceedings of the Royal Academy of Sciences 272:1655–1662. ferences. Animal Behaviour 34:1359–1371.
 References 463

Barbosa, M., and A. E. Magurran. 2006. Female mating decisions: Bateson, P. 1990. Is imprinting such a special case? Philosophical
maximizing fitness? Journal of Fish Biology 68: 1636–1661. Transactions of the Royal Society of London B 329:125–131.
Barinaga, M. 1994. From fruit flies, rats, mice: Evidence of Bateson, P. 2003. The promise of behavioural biology. Animal
genetic influence. Science 264:1690–1693. Behaviour 65:11–17.
Barkley, C. L., and L. F. Jacobs. 2007. Sex and species differ- Bateson, P. 2005. The return of the whole organism. Journal
ences in spatial memory in food-storing kangaroo rats. of Biosciences 30:31–39.
Animal Behaviour 73:321–329. Bateson, P. P. G., and P. H. Klopfer, eds. 1989. Preface. In
Barlow, G. W. 1968. Ethological units of behavior. In The Perspectives in Ethology, Volume 8. Whither Ethology? New
Central Nervous System and Fish Behavior, edited by D. Ingle. York: Plenum Press.
Chicago: University of Chicago Press.
Bateson, P. P. G., and P. H. Klopfer, eds. 1991. Perspectives in
Barlow, G. W. 1989. Has sociobiology killed ethology or
Ethology, Volume 9. Human Understanding and Animal
revitalized it? Perspectives in Ethology Volume 8, Whither
Awareness. New York: Plenum.
ethology? pp. 1–45.
Batteau, D. W. 1968. The world as a source; the world as a
Barlow, G. W. 1991. Nature-nurture and the debates sur-
sink. In The Neuropsychology of Spatially Oriented Behavior,
rounding ethology and sociobiology. American Zoologist
edited by S. J. Freedman. Homewood, IL: Dorsey, pp.
31:286–296.
197–203.
Barnea, A., and F. Nottebohm. 1994. Seasonal recruitment of
hippocampal neurons in adult free-ranging black-capped Battin, J. 2004. When good animals love bad habitats:
chickadees. Proceedings of the National Academy of Sciences Ecological traps and the conservation of animal populations.
91:11217–11221. Conservation Biology 18:1482–1491.
Bartz, J., and E. Hollander. 2006. The neuroscience of affili- Bauer, E. B., and B. B. Smuts. 2007. Cooperation and com-
ation: Forging links between basic and clinical research on petition during dyadic play in domestic dogs, Canis famil-
neuropeptides and social behavior. Hormones and Behavior iaris. Animal Behaviour 73:489–499.
50:518–528. Baxi, K. N., K. M. Dorries, and H. L. Eisthen. 2006. Is the
Basil, J. A., A. C. Kamil, R. P. Balda, and K. V. Fite. 1996. vomeronasal system really specialized for detecting
Differences in hippocampal volume among food storing pheromones? Trends in Neurosciences 29:1–7.
corvids. Brain, Behavior and Evolution 47:156–164. Beach, F. A. 1974. Effects of gonadal hormones on urinary
Basolo, A. L. 1990. Female preference predates the evolution behavior in dogs. Physiology and Behavior 12:1005–1013.
of the sword in swordtail fish. Science 250:808–810. Beach, F. A. 1976. Sexual attractivity, proceptivity, and recep-
Basolo, A. L. 1995a. A future examination of a pre-existing tivity in female mammals. Hormones and Behavior 7:105–138.
bias favouring a sword in the genus Xiphophorus. Animal Beauchamp, G. 1999. The evolution of communal roosting in
Behaviour 50:365–375. birds: origins and secondary losses. Behavioral Ecology
Basolo, A. L. 1995b. Phylogenetic evidence for the role of a 10:675–687.
pre-existing bias in sexual selection. Proceedings of the Royal Beauchamp, G. 2007. Vigilance in a selfish herd. Animal
Society of London B 259:307–311. Behaviour 73:445–451.
Bass, A. H. 1996. Shaping brain sexuality. American Scientist
Beck, W., and W. Wiltschcko. 1988. Magnetic factors control
84:352–363.
the migratory direction of pied flycatchers (Ficedula
Bass, A. H., D. A. Bodnar, and J. R. McKibben. 1997. From
hypoleuca Pallas). Acta Congress International Ornithology
neurons to behavior: Vocal-acoustic communication in
19:1955–1962.
teleost fish. Biological Bulletin 192:158–160.
Beckerman, A. P., M. Uriarte, and O. J. Schmitz. 1997.
Bateman, A. J. 1948. Intra-sexual selection in Drosophila.
Experimental evidence for a behavior-mediated trophic cas-
Heredity 2:349–368.
cade in a terrestrial food chain. Proceedings of the National
Bateson, M. 2002. Recent advances in our understanding of
Academy of Sciences 94:10735–10738.
risk-sensitive foraging preferences. Proceedings of the
Nutrition Society 61:509–516. Bednarz, J. C. 1988. Cooperative hunting in Harris’ hawks
(Parabuteo unicinctus). Science 239:1525–1527.
Bateson, P. 1976. Specificity and the origins of behavior. In
Advances in the Study of Behavior, edited by J. S. Rosenblatt, Bee, M. A., and H. C. Gerhardt. 2001. Habituation as a mech-
R. A. Hinde, and C. Beer. New York: Academic Press, pp. anism of reduced aggression between neighboring territo-
1–20. rial male bullfrogs (Rana catesbeiana). Journal of Comparative
Bateson, P. 1979. How do sensitive periods arise and what are Psychology 115:68–82.
they for? Animal Behaviour 27:470–486. Beebe, T. 1995. Amphibian breeding and climate change.
Bateson, P. 1982. Preferences for cousins in Japanese quail. Nature 374:219–220.
Nature 295:236–237. Beecher, M. D., I. M. Beecher, and S. Hahn. 1981. Parent-
Bateson, P. 1983. Optimal outbreeding. In Mate Choice, edited offspring recognition in bank swallows (Riparia riparia):
by P. Bateson. Cambridge: Cambridge University Press, pp. Development and acoustic basis. Animal Behaviour
257–277. 29:95–101.
464 References

Beecher, M. D., and E. A. Brenowitz. 2005. Functional aspects Berthold, P. 2001. Bird Migration: A General Survey. Edited by
of song learning in songbirds. Trends in Ecology and Evolution C. M. Perrins. 2nd ed. Oxford Ornithological Series. New
20:143–149. York: Oxford University Press.
Beecher, M. D., J. M. Burt, A. L. O’Loghlen, C. N. Bertram, B. C. R. 1975. Social factors influencing reproduc-
Templeton, and S. E. Campbell. 2007. Bird song learning tion in wild lions. Journal of Zoology 177:463–482.
in an eavesdropping context. Animal Behaviour 73:929–935. Bertram, B. C. R. 1976. Kin selection in lions and evolution.
Beecher, M. D., S. E. Campbell, and J. C. Nordby. 2000. In Growing Points in Ethology, edited by P. P. G. Bateson and
Territory tenure in song sparrows is related to song shar- R. A. Hinde. New York: Cambridge University Press, pp.
ing with neighbours, but not to repertoire size. Animal 281–301.
Behaviour 59:29–37. Bertram, S. M. 2002. Temporally fluctuating selection of sex-
Beecher, M. D., P. Loesche, P. K. Stoddard, and M. B. limited signaling traits in the Texas field cricket, Gryllus tex-
Medvin. 1989. Individual recognition by voice in swallows: ensis. Evolution 56:1831–1839.
signal or perceptual adaptation? In The Comparative Biben, M. 1998. Squirrel monkey playfighting: Making the
Psychology of Audition, edited by R. J. Dooling and S. H. case for a cognitive training function for play. In Animal
Hulse. Mahwah, NJ: Erlbaum. Play: Evolutionary, Comparative, and Ecological Perspectives,
Beehler, M. N., and M. S. Foster. 1988. Hotshots, hotspots, edited by M. Bekoff and J. A. Byers. Cambridge: Cambridge
and female preference in the organization of lek mating sys- University Press, pp. 161–182.
tems. American Naturalist 131:203–219. Bingman, V. P. 1984. Night-sky orientation of migratory pied
Beer, M. F. 1999. Homosynaptic long-term depression: a flycatchers raised in different magnetic fields. Behavioral
mechanism for memory? Proceedings of the National Academy Ecology and Sociobiology 15:77–80.
of Sciences 96:9457–9458. Bingman, V. P., and K. Cheng. 2005. Mechanisms of animal
global navigation: Comparative perspectives and enduring
Bekoff, A. 1992. Neuroethological approaches to the study of
challenges. Ethology, Ecology and Evolution 17:295–318.
motor development in chicks: achievements and challenges.
Journal of Neurobiology 23:1486–1505. Bingman, V., T. Jechura, and M. C. Kahn. 2006. Behavioral
and neural mechanisms of homing and migration in birds.
Bekoff, A., and J. A. Kauer. 1984. Neural control of hatching:
In Animal Spatial Cognition: Comparative, Neural, and
Gate of the pattern generator for leg movements of hatch-
Computational Approaches. [On-line]. , edited by M. F. Brown
ing in post-hatching chicks. Journal of Neuroscience
and R. G. Cook: Available: www.pigeon.psy.tufts.edu/asc/bing-
4:2659–2666.
man/.
Beletsky, L. D., and G. H. Orians. 1987. Territoriality among
Birkhead, T. R. 1996. Mechanisms of sperm competition in
male red-winged blackbirds. II. Removal experiments and
birds. American Scientist 84:254–262.
site dominance. Behavioral Ecology and Sociobiology
Birkhead, T. R., and G. A. Parker. 1997. Sperm competition
20:339–349.
and mating systems. In Behavioural Ecology: An Evolutionary
Beletsky, L. D., and G. H. Orians. 1989. Territoriality among
Approach, edited by J. R. Krebs and N. B. Davies. Oxford:
male red-winged blackbirds. III. Testing hypotheses of ter- Blackwell Science, pp. 121–145.
ritorial dominance. Behavioral Ecology and Sociobiology
Bischof, H.-J. 1994. Sexual imprinting as a two-stage process.
24:333–339.
In Causal Mechanisms of Behavioural Development, edited by
Beling, I. 1929. Über das Zeitgedächtnis der Bienen. Zeitschrift J. A. Hogan and J. J. Bolhuis. Cambridge: Cambridge
für vergleichende Physiologie 9:259–338. University Press, pp. 82–97.
Belovsky, G. E. 1978. Diet optimization in a generalist her- Bischof, H.-J. 2003. Neural mechanisms of sexual imprinting.
bivore: the moose. Theoretical Population Biology 14:105–134. Animal Biology 53:89–112.
Bemis, W. E., and B. Kynard. 1997. Sturgeon rivers: an intro- Bischof, H.-J. 2007. Behavioral and neuronal aspects of devel-
duction to acipenseriform biogeography and life history. opmental sensitive periods. NeuroReport 18:461–465.
Environmental Biology of Fishes 48:167–183. Blackmore, C. J., and R. Heinsohn. 2008. Variable mating
Ben-Shahar, Y., A. Robichon, M. B. Sokolowski, and G. E. strategies and incest avoidance in cooperatively breeding
Robinson. 2002. Influence of gene action across different grey-crowned babblers. Animal Behaviour 75:63–70.
time scales on behavior. Science 296:741–744. Blakemore, R. P., and R. B. Frankel. 1981. Magnetic naviga-
Berec, M., V. Krivan, and L. Berec. 2003. Are great tits (Parus tion in bacteria. Scientific American 245:58–65.
major) really optimal foragers? Canadian Journal of Zoology Bloch Qazi, M. C., J. R. Aprille, and S. M. Lewis. 1998.
81:780–788. Female role in sperm storage in the red flour beetle,
Bereczkei, T., P. Gyuris, and G. E. Weisfeld. 2004. Sexual Tribolium castaneum. Comparative Biochemistry and Physiology
imprinting in human mate choice. Proceedings of the Royal A 120:641–647.
Society of London B 271:1129–1134. Blomqvist, D., O. C. Johasson, U. Unger, M. Larsson, and
Berger, J. 2004. The last mile: How to sustain long-distance L.-A. Flodin. 1997. Male aerial display and reversed sexual size
migration in mammals. Conservation Biology 18:320–331. dimorphism in the dunlin. Animal Behaviour 54:1291–1299.
 References 465

Blumenrath, S. H., and T. Dabelsteen. 2004. Degradation of Bolhuis, J. J. 2008. Chasin’ the trace: the neural substrate of
great tit (Parus major) song before and after foliation: impli- birdsong memory. In The Neuroscience of Birdsong, edited by
cations for vocal communication in a deciduous forest. H. P. Zeigler and P. Marler. Cambridge: Cambridge
Behaviour 141:935–958. University Press, pp. 271–281.
Blumstein, D. T. 2006. The multipredator hypothesis and the Bolhuis, J. J., and H. Eda-Fujiwara. 2003. Bird brains and
evolutionary persistence of antipredator behavior. Ethology
songs: neural mechanisms of birdsong perception and mem-
112:209–217.
ory. Animal Biology 53:129–145.
Blumstein, D. T. 2007. The evolution of alarm communica-
tion in rodents: structure, function, and the puzzle of appar- Bolhuis, J. J., and M. Gahr. 2006. Neural mechanisms of bird-
ently altruistic calling. In Rodent Societies: An Ecological and song memory. Nature Reviews Neuroscience 7:347–357.
Evolutionary Perspective, edited by J. O. Wolff and P. W. Bolhuis, J. J., and R. C. Honey. 1998. Imprinting, learning and
Sherman. University of Chicago Press: Chicago, IL. pp. development: from behaviour to brain and back. Trends in
317–327. Neurosciences 21:306–311.
Blumstein, D. T., and K. B. Armitage. 1997. Does sociality Bollenbacher, W. E., S. L. Smith, W. Goodman, and L. I.
drive the evolution of communicative complexity? A com- Gilbert. 1981. Ecdysone titer during larval-pupal–adult
parative test with ground-dwelling sciurid alarm calls.
development in the tobacco hornworm, Manduca sexta.
American Naturalist 150:179–200.
General and Comparative Endocrinology 44:302–306.
Blumstein, D. T., A. Bitton, and J. DaVeiga. 2006. How does
the presence of predators influence the persistence of Bonadonna, F., S. Caro, P. Jouventin, and G. A. Nevitt. 2006.
antipredator behavior? Journal of Theoretical Biology Evidence that blue petrel, Halobaena caerulea, fledglings can
239:460–468. detect and orient to dimethyl sulfide. Journal of Experimental
Blumstein, D. T., and J. C. Daniel. 2005. The loss of Biology 209:2165–2169.
antipredator behaviour following isolation on islands. Boncoraglio, G., and N. Saino. 2007. Habitat structure and
Proceedings of the Royal Society of London B 272:1663–1668. the evolution of bird song: a meta-analysis of the evidence
Blumstein, D. T., and E. Fernández-Juricic. 2004. The emer- for the acoustic adaptation hypothesis. Functional Ecology
gence of conservation behavior. Conservation Biology 21:134–142.
18:1175–1177.
Bond, A. B. 1983. Visual search and selection of natural stim-
Blumstein, D. T., J. Steinmetz, K. B. Armitage, and J. C.
uli in the pigeon: the attention threshold hypothesis. Journal
Daniel. 1997. Alarm calling in yellow-bellied marmots. 2.
of Experimental Psychology—Animal Behavior Processes
The importance of direct fitness. Animal Behaviour
9:292–306.
53:173–184.
Boag, P. T., and P. R. Grant. 1981. Intense natural selection Bond, A. B., and A. C. Kamil. 1998. Apostatic selection by
in a population of Darwin finches (Geospizinae) in the blue jays produces balanced polymorphisms in virtual prey.
Galapagos. Science 214: 82–85. Nature 395:594–596.
Boal, C. W., and R. W. Mannan. 1999. Comparative breed- Bonduriansky, R. 2001. The evolution of male mate choice in
ing ecology of Cooper’s hawks in urban and exurban areas insects: a synthesis of ideas and evidence. Biological Reviews
of southeastern Arizona. Journal of Wildlife Management 76:305–339.
63:77–84. Bonte, D., S. Van Belle, and J.-P. Maelfait. 2007. Maternal care
Boccia, M. L. 1983. A functional analysis of social grooming and reproductive state-dependent mobility determine natal
patterns through direct comparison with self grooming in
dispersal in a wolf spider. Animal Behaviour 74:63–69.
rhesus monkeys. International Journal of Primatology
4:399–418. Borgia, G. 1986. Satin bowerbird parasites: A test of the bright
Boccia, M. L. 1986. Grooming site preferences as a form of male hypothesis. Behavioral Ecology and Sociobiology
tactile communication and their role in the social relations 19:355–358.
of rhesus monkeys. In Current Perspectives in Primate Social Borgia, G., I. M. Kaatz, and R. Condit. 1987. Flower choice
Dynamics, edited by D. M. Taub and F. A. King. New York: and bower decoration in the satin bowerbird Ptilonorhynchus
Van Nostrand Reinhold, pp. 505–518. violaceus: A test of the hypothesis for the evolution of male
Bodamer, M. D., D. H. Fouts, R. S. Fouts, and M. L. A. display. Animal Behaviour 35:1129–1139.
Jensvold. 1994. Functional analysis of chimpanzee (Pan Borries, C., K. Launhardt, C. Epplen, J. T. Epplen, and P.
troglodytes) private signing. Human Evolution 9:281–296.
Winkler. 1999. DNA analyses support the hypothesis that
Boinski, S., L. Kauffman, E. Ehmke, S. Schet, and A.
infanticide is adaptive in langur monkeys. Proceedings of the
Vreedzaam. 2005. Dispersal patterns among three species
Royal Society of London B 266:901–904.
of squirrel monkeys (Saimiri oerstedii, S. boliviensis and S. sci-
ureus): I. Divergent costs and benefits. Behaviour Bostwick, K. S., and R. O. Prum. 2005. Courting bird sings
142:525–632. with stridulating wing feathers. Science 309:736.
Boles, L. C., and K. J. Lohmann. 2003. True navigation and Bourke, A. F. G., and N. R. Franks. 1995. Social Evolution in
magnetic maps in spiny lobsters. Nature 421:60–63. Ants. Princeton: Princeton University Press.
466 References

Bowmaker, J. K., V. I. Govardovskii, S. A. Shukolykov, L. V. Brockmann, J., A. Grafen, and R. Dawkins. 1979.
Zueva, D. M. Hunt, V. G. Sideleva, and O. G. Smirnova. Evolutionarily stable nesting strategy in a digger wasp.
1994. Visual pigments and the photic environment: the cot- Journal of Theoretical Biology 77:473–496.
toid fish of Lake Baikal. Vision Research 34:591–605. Brodin, A., and K. Lundborg. 2003. Is hippocampal volume
Boyan, G. S., and J. H. Fullard. 1986. Interneurons respond- affected by specialization for food hoarding in birds?
ing to sound in the tobacco budworm moth Heliothis Proceedings of the Royal Society of London B 270:1555–1563.
virescens (Noctuidae): morphological and physiological Broom, M., G. D. Ruxton, and R. M. Kilner. 2008. Host life-
characteristics. Journal of Comparative Physiology A history strategies and the evolution of chick-killing by
158:391–404. brood parasitic offspring. Behavioral Ecology 19:22–34.
Boyan, G. S., and L. A. Miller. 1991. Parallel processing of Brotherton, P. N. M., and M. B. Manser. 1997. Female dis-
afferent input by identified interneurones in the auditory persion and the evolution of monogamy in the dik-dik.
pathway of the noctuid moth Noctua pronumba. Journal of Animal Behaviour 54:1413–1424.
Comparative Physiology A 168:727–738. Brotherton, P. N. M., J. M. Pemberton, P. E. Komers, and G.
Malarky. 1997. Genetic and behavioural evidence of
Boysen, S. T., and G. G. Berntson. 1989. Numerical compe-
monogamy in a mammal, Kirk’s dik-dik (Madoqua kirkii).
tence in a chimpanzee (Pan troglodytes). Journal of
Proceedings of the Royal Society of London B 264:675–681.
Comparative Psychology 103:23–31.
Brotherton, P. N. M., and A. Rhodes. 1996. Monogamy with-
Bradbury, J. W. 1981. The evolution of leks. In Natural
out biparental care in a dwarf antelope. Proceedings of the
Selection and Social Behavior, edited by R. D. Alexander and
Royal Society of London B 263:23–29.
D. W. Tinkle. New York: Chiron, pp. 138–172.
Brower, L. P. 1996. Monarch butterfly orientation: missing
Bradbury, J. W., R. M. Gibson, and M. B. Andersson. 1986. pieces of a magnificent puzzle. Journal of Experimental
Hotspots and the evolution of leks. Animal Behaviour Biology 199:93–103.
34:1694–1709. Brower, L. P., and W. H. Calvert. 1985. Foraging dynamics
Bradbury, J. W., and S. L. Vehrencamp. 1998. Principles of of bird predators on overwintering monarch butterflies.
Animal Communication. Sunderland, MA: Sinauer. Evolution 39:852–868.
Brainard, M. S., and A. J. Doupe. 2000. Auditory feedback in Brower, L. P., D. R. Kust, E. Rendon-Salinas, E. G. Serrano,
learning and maintenance of vocal behaviour. Nature K. R. Kust, J. Mller, C. Fernandez del Rey, and K. Pape.
Reviews 1:31–40. 2004. Catastrophic winter storm mortality of monarch but-
Brantley, R. K., and A. H. Bass. 1994. Alternative male spawn- terflies in Mexico during January 2002. In The Monarch
ing tactics and acoustic signals in the plainfin midshipman Butterfly: Biology and Conservation, edited by K. S.
fish Porichthys notatus Girard (Teleostei, Batrachoididae). Oberhauser and M. J. Solensky. Ithaca: Cornell University
Ethology 96:213–232. Press, p. 155–198.
Braude, S., D. Ciszek, N. E. Berg, and N. Shefferly. 2001. The Brower, L. P., W. N. Ryerson, L. L. Coppinger, and S. C.
ontogeny and distribution of countershading in colonies of Glazier. 1968. Ecological chemistry and palatability spec-
the naked mole-rat (Heterocephalus glaber). Journal of Zoology trum. Science 161:1349–1351.
253:351–357. Brown, C. R. 1986. Cliff swallow colonies as information cen-
tres. Science 234:83–85.
Breed, M. D. 1998. Recognition pheromones of the honey-
Brown, C. R., and M. B. Brown. 1986. Ectoparasitism as a cost
bee. BioScience 48:463–470.
of coloniality in cliff swallows (Hirundo pyrrhonota). Ecology
Breedlove, S. M., and A. P. Arnold. 1983. Hormonal control
67:1206–1218.
of a developing neuromuscular system. II. Sensitive periods
Brown, C. R., and M. B. Brown. 2002. Ectoparasites cause
for the androgen-induced masculinization of the rat spinal
increased bilateral asymmetry of naturally selected traits in
nucleus of the bulbocavernosus. Journal of Neuroscience
a colonial bird. Journal of Evolutionary Biology 15:1067–1075.
3:424–432.
Brown, C. R., and M. B. Brown. 2004. Group size and
Breedlove, S. M., M. R. Rosenweig, and N. V. Watson. 2007. ectoparasitism affect daily survival probability in a colonial
Biological Psychology: An Introduction to Behavioral, Cognitive, bird. Behavioral Ecology and Sociobiology 56:498–511.
and Clinical Neuroscience. 5th ed. Sunderland MA: Sinauer. Brown, C. R., M. B. Brown, and B. Rannala. 1995.
Brenowitz, E. A. 2004. Plasticity of the adult avian song con- Ectoparasites reduce long-term survival of their avian host.
trol system. Annals of the New York Academy of Sciences Proceedings of the Royal Society of London B 262:313–319.
1016:560–585. Brown, F. A., Jr., J. W. Hastings, and J. D. Palmer. 1970. The
Brenowitz, E. A., K. Lent, and E. W. Rubel. 2007. Auditory Biological Clock: Two Views, New York: Academic Press.
feedback and song production do not regulate seasonal Brown, J. L. 1964. The evolution of diversity in avian terri-
growth of song control circuits in adult white-crowned torial systems. Wilson Bulletin 76:160–169.
sparrows. The Journal of Neuroscience 27:6810–6814. Brown, J. L., and E. R. Brown. 1998. Are inbred offspring less
Briffa, M., and L. U. Sneddon. 2007. Physiological constraints fit? Survival in a natural population of Mexican jays.
on contest behaviour. Functional Ecology 21:627–637. Behavioral Ecology 9:60–63.
 References 467

Brown, J. L, E. R. Brown, and S. D. Brown. 1982. Burgess, K. S., J. Singfield, V. Melendez, and P. G. Kevan.
Morphological variation in a population of grey-crowned 2004. Pollination biology of Aristolochia grandiflora
babblers—correlations with variables affecting social behav- (Aristolochiaceae) in Veracruz, Mexico. Annals of the
ior. Behavioral Ecology and Sociobiology 10:281–287. Missouri Botanical Garden 91:346–356.
Brown, J. L., and A. Eklund. 1994. Kin recognition and the Burghardt, G. M. 1998. The evolutionary origins of play revis-
major histocompatibility complex: An integrative review. ited. In Animal Play: Evolutionary, Comparative, and Ecological
American Naturalist 143:435–461. Perspectives, edited by M. Bekoff and J. A. Byers. Cambridge:
Brown, J. S., and B. P. Kotler. 2004. Hazardous duty pay and Cambridge University Press, pp. 1–26.
the foraging cost of predation. Ecology Letters 7:999–1014. Burghardt, G. M. 2005. The Genesis of Animal Play.
Brown, K. S. 1996. Infanticide as a way to get ahead. BioScience Cambridge, MA: MIT press.
46:174–176. Burghardt, G. M., and H. W. Greene. 1988. Predator simu-
Brownell, P. H. 1984. Prey detection by the scorpion. lation and duration of death feigning in neonate hognose
Scientific American 251:86–98. snakes. Animal Behaviour 36:1842–1844.
Brownell, P. H., and J. L. van Hemmen. 2001. Vibration sen- Burkhardt, D., and I. de la Motte. 1983. How stalk-eyed flies
sitivity and a computational theory for prey-localizing view stalk-eyed flies: Observations and measurements of the
behavior in sand scorpions. American Zoologist 41:1229–1240. eyes of Cyrtodiopsis whitei (Diopsidea, Diptera). Journal of
Bruder, C. E. G., A. Piotrowski, C. J. Gijsbers, R. Andersson, Comparative Physiology A 151:407–421.
S. Erickson, T. D. de Ståhl, U. Menzel, J. Sandgren, D. von Burmeister, S. S., E. D. Jarvis, and R. D. Fernald. 2005. Rapid
Tell, A. Poplawski, M. Crowley, C. Crasto, E. C. Partridge, behavioral and genomic responses to social opportunity.
H. Tiwari, D. B. Allison, J. Komorowski, G.-J. B. van PLoS Biology 3:1996–2004.
Ommen, D. I. Boomsma, N. L. Pedersen, J. T. den Burt, J. M., A. L. O’Loghlen, C. N. Templeton, S. E.
Dunnen, K. Wirdefeldt, and J. P. Dumanski. 2008. Campbell, and M. D. Beecher. 2007. Assessing the impor-
Phenotypically concordant and discordant monozygotic tance of social factors in bird song learning: A test using
twins display different DNA copy-number-variation pro- computer-simulated tutors. Ethology 113:917–925.
files. American Journal of Human Genetics 82:763–781.
Butts, K. O., and J. C. Lewis. 1982. The importance of prairie
Brudzynski, S. M. 2005. Principles of rat communication:
dog towns to burrowing owls in Oklahoma. Proceedings of
quantitative parameters of ultrasonic calls in rats. Behavior
the Oklahoma Academy of Science 62:46–52.
Genetics 35:85–92.
Byers, D., R. L. Davis, and J. A. Kiger, Jr. 1981. Defect in
Bruel-Jungerman, E., S. Davis, and S. Laroche. 2007. Brain
cyclic AMP phosphodiesterase due to the dunce mutation of
plasticity mechanisms and memory: A party of four.
learning in Drosophila melanogaster. Nature 289:79–81.
Neuroscientist 13:492–505.
Byers, J. A., and L. Waits. 2006. Good genes sexual selection
Brunton, D. H. 1990. The effects of nesting stage, sex, and
in nature. Proceedings of the National Academy of Sciences
type of predator on parental defense by killdeer (Charadrius
103:16343–16345.
vociferus): Testing models of parental defense. Behavioral
Byers, J. A., and C. B. Walker. 1995. Refining the motor train-
Ecology and Sociobiology 26:181–190.
ing hypothesis for the evolution of play. American Naturalist
Brusca, R. C., and G. J. Brusca.1990. Invertebrates. Sunderland
146:25–40.
MA: Sinauer.
Byers, J. A., A. A. Byers, and S. J. Dunn. 2006. A dry summer
Budzynski, C. A., R. Strasser, and V. P. Bingman. 1998. The
diminishes mate search effort by pronghorn females:
effects of zinc sulfate anosmia on homing pigeons, Columba
Evidence for a significant cost of mate search. Ethology
livia, in a homing and a non-homing experiment. Ethology
112:74–80.
104:111–118.
Buler, J. J., F. R. Moore, and S. Woltmann. 2007. A multi-scale Byers, J. A., J. D. Moodie, and N. Hall. 1994. Pronghorn
examination of stopover habitat use by birds. Ecology females choose vigorous mates. Animal Behaviour 47:33–43.
88:1789–1802. Byrne, J. H. 1987. Cellular analysis of associative learning.
Bullock, T. H., C. D. Hopkins, A. N. Popper, and R. R. Fay Physiological Reviews 67:329–439.
(eds). 2005. Electroreception. New York: Springer Science + Byrom, A. E., and C. J. Krebs. 1999. Natal dispersal of juve-
Business Media, Inc. nile Arctic ground squirrels in the boreal forest. Canadian
Bult, A., and C. B. Lynch. 1996. Multiple selection responses Journal of Zoology 77:1048–1059.
in house mice bidirectionally selected for thermoregulatory Cahalane, V. H. 1961. Mammals of North America. New York:
nest-building behavior: Crosses of replicate lines. Behavior Macmillan.
Genetics 26:439–446. Caldwell, R. L., and H. Dingle. 1976. Stomatopods. Scientific
Bult, A., and C. B.Lynch. 1997. Nesting and fitness: Lifetime American 234:80–89.
reproductive success in house mice bidirectionally selected Calvert, W. H., and L. P. Brower. 1986. The location of the
for thermoregulatory nest-building behavior. Behavior monarch butterfly (Danaus plexippus L.). Journal of the
Genetics 27:231–240. Lepidopteran Society 40:164–187.
468 References

Calvert, W. H., L. E. Hedrick, and L. P. Brower, 1979. Carroll, L. S., and W. K. Potts. 2007. Sexual selection: using
Mortality of the monarch butterfly (Danaus plexippus L.): social ecology to determine fitness differences. In Rodent
Avian predation at five overwintering sites in Mexico. Science Societies: An Ecological and Evolutionary Perspective, edited by
204:847–851. J. O. Wolff and P. W. Sherman. Chicago: The University
Camhi, J. M. 1980. The escape system of the cockroach. of Chicago Press, pp. 57–67.
Scientific American 248:158–172. Carroll, S. P., and L. Moore. 1993. Hummingbirds take their
Camhi, J. M. 1984. Neuroethology. Sunderland, MA: Sinauer. vitamins. Animal Behaviour 46:817–820.
Camhi, J. M. 1988. Escape behavior in the cockroach: Carruth, L. L., R. E. Jones, and D. O. Norris. 2002. Cortisol
Distributed neural processing. Experientia 44:401–408. and Pacific salmon: a new look at the role of stress hor-
Camhi, J. M. , W. Tom, and S. Volman. 1978. The escape mones in olfaction and home-stream migration. Integrative
behavior of the cockroach Periplaneta americana. II. and Comparative Biology 42:574–581.
Detection of natural predators by air displacement. Journal Castellucci, V., and E. R. Kandel. 1974. A quantal analysis of
of Comparative Physiology 128:203–212. the synaptic depression underlying the gill-withdrawal
Candolin, U. 2000a. Increased signalling effort when survival reflex in Aplysia. Proceedings of the National Academy of Sciences
prospects decrease: male-male competition ensures honesty. 71:5004–5008.
Animal Behaviour 60:417–422. Catania, K. C. 2006. Underwater ‘sniffing’ by semi-aquatic
Candolin, U. 2000b. Male-male competition ensures honest mammals. Nature 444:1024–1025.
signaling of male parental ability in the three-spined stick- Catania, K. C., and J. H. Kaas. 1996. The unusual nose and
leback (Gasterosteus aculeatus). Behavioral Ecology and brain of the star-nosed mole. BioScience 46:578–586.
Sociobiology 49:57–61. Catania, K. C., and F. E. Remple. 2004. Tactile foveation in
Caplan, A. L., ed. 1978. The Sociobiology Debate. New York: the star-nosed mole. Brain, Behavior and Evolution 63:1–12.
Harper & Row Publishers. Champagne, F. A., I. C. G. Weaver, J. Diorio, S. Dymov, M.
Caputi, A. A., and R. Budelli. 2006. Peripheral electrosensory Szyf, and M. J. Meaney. 2006. Maternal care associated with
imaging by weakly electric fish. Journal of Comparative methylation of the estrogen receptor–alpha1b promoter and
Physiology A 192:587–600. estrogen receptor alpha expression in the medial preoptic
Caputi, A. A., R. Budelli, K. Grant, and C. C. Bell. 1998. The area of female offspring. Endocrinology 147:2909–2916.
electrical image in weakly electric fish: Physical images of Champalbert, A., and P.-P. Lachaud. 1990. Existence of a sen-
resistive objects in Gnathonemus petersii. Journal of sitive period during ontogenesis of social behavior in a
Experimental Biology 201:2115–2128. primitive ant. Animal Behaviour 39:850–859.
Carballada, R., and P. Esponda. 1992. Role of fluid from sem- Chang, H-S, M. S. Anway, S. S. Rekow, and M. K. Skinner.
inal vesicles and coagulating glands in sperm transport into 2006. Transgenerational epigenetic imprinting of male
the uterus and fertility in rats. Journal of Reproduction and germline by endocrine disruptor exposure during gonadal
Fertility 95:639–648. sex determination. Endocrinology 147:5524–5541.
Carew, T. J. 2000. Behavioral Neurobiology. Sunderland MA: Chapman, T., L. F. Liddle, J. M. Kalb, M. F. Wolfner, and L.
Sinauer. Partridge. 1995. Cost of mating in Drosophila melanogaster
Carew, T. J., H. M. Pinsker, and E. R. Kandel. 1972. Long- females is mediated by male accessory gland products.
term habituation of a defensive withdrawal reflex in Aplysia. Nature 373:241–244.
Science 175:451–454. Chapman, T., D. M. Neubaum, M. F. Wolfner, and L.
Carlier, P., and L. Lefebvre. 1997. Ecological differences in Partridge. 2000. The role of male accessory gland protein
social learning between adjacent, mixing populations of Acp36DE in sperm competition in Drosophila melanogaster.
Zenaida doves. Ethology 103:772–784. Proceedings of the Royal Society of London B 267:1097–1105.
Caro, T. M. 1986a. The functions of stotting: A review of Charnov, E. L. 1976. Optimal foraging: the marginal value
hypotheses. Animal Behaviour 34:649–662. theorem. Theoretical Population Biology 9:129–136.
Caro, T. M. 1986b. The functions of stotting in Thomson’s Charrier, I., and C. B. Sturdy. 2005. Call-based recognition
gazelles: Some tests of the predictions. Animal Behaviour in black-capped chickadees. Behavioural Processes
34:663–684. 70:271–281.
Caro, T. M. 1988. Adaptive significance of play: Are we get- Cheetham, S. A., M. D. Thom, F. Jury, W. E. R. Ollier, R. J.
ting closer? Trends in Ecology and Evolution 88:50–54. Beynon, and J. L. Hurst. 2007. The genetic basis of indi-
Caro, T. M. 1994. Cheetahs of the Serengeti Plains: Group Living vidual-recognition signals in the mouse. Current Biology 17:
in an Asocial Species. Chicago: University of Chicago Press. 1771–1777.
Caro, T. M. 1995. Short-term costs and correlates of play in Cheney, D. L., and R. M. Seyfarth. 1988. Assessment of mean-
cheetahs. Animal Behaviour 49:333–345. ing and detection of unreliable signals by vervet monkeys.
Carpenter, F. L., D. C. Paton, and M. A. Hixon. 1983. Weight Animal Behaviour 36:477–486.
gain and adjustment of feeding territory size in migrant Cheng, M. W., and R. L. Caldwell. 2000. Sex identification
hummingbirds. Proceedings of the National Academy of Sciences and mating in the blue-ringed octopus, Hapalochlaena
80:7259–7263. lunulata. Animal Behaviour 60:27–33.
 References 469

Chittka, L., T. C. Ings, and N. E. Raine. 2004. Chance and Clutton-Brock, T. H., F. E. Guiness, and S. D. Albon. 1982.
adaptation in the evolution of island bumblebee behaviour. Red Deer: The Behaviour and Ecology of Two Sexes. Chicago:
Population Ecology 46:243–251. University of Chicago Press.
Choleris, E., M. Kavaliers, and D. W. Pfaff. 2004. Functional Clutton-Brock, T. H., O. F. Price, and A. D. C. MacColl.
genomics of social recognition. Journal of Neuroendocrinology 1992. Mate retention, harassment, and the evolution of
16:383–389. ungulate leks. Behavioral Ecology 3:234–242.
Christy, J. H., and M. Salmon. 1991. Comparative studies of Clutton-Brock, T. H., S. J. Hodge, G. Spong, A. F. Russell,
reproductive behavior in mantis shrimps and fiddler crabs. N. R. Jordan, N. C. Bennett, L. L. Sharpe, and M. B.
American Zoologist 31:329–337. Manser. 2006. Intrasexual competition and sexual selection
in cooperative mammals. Nature 444:1065–1068.
Clark, C., and M. Mangel. 2000. Dynamic State Variable Models
in Ecology: Methods and Applications. Oxford: Oxford Clutton-Brock, T. H., A. F. Russell, L. L. Sharpe, P. N. M.
Brotherton, G. M. McIlrath, S. White, and E. Z. Cameron.
University Press.
2001. Effects of helpers on juvenile development and sur-
Clark, R. B. 1960. Habituation of the polychaete Nereis to sud-
vival in meerkats. Science 293:2446–2449.
den stimuli. 1. General properties of the habituation
Cockburn, A. 1998. Evolution of helping behavior in cooper-
process. Animal Behaviour 8:82–91.
atively breeding birds. Annual Review of Ecology and
Clark, R. W. 2005. Pursuit-deterrent communication between Systematics 29:141–472.
prey animals and timber rattlesnakes (Crotalus horridus): the
Cockburn, A. 2006. Prevalence of different modes of parental
response of snakes to harassment displays. Behavioral
care in birds. Proceedings of the Royal Society of London B
Ecology and Sociobiology 59:258–261. 273:1375–1383.
Clarke, A. L., B.-R. Sæther, and E. Røskaft. 1997. Sex biases Cockburn, A., R. A. Sims, H. L. Osmond, D. J. Green, M. C.
in avian dispersal: A reappraisal. Oikos 79:429–438. Double, and R. A. Mulder. 2008. Can we measure the ben-
Clarke, B. 1969. The evidence for apostatic selection. Heredity efits of help in cooperatively breeding birds: the case of
24:347–352. superb fairy-wrens Malurus cyaneus? Journal of Animal
Clobert, J., J. O. Wolff, J. D. Nichols, E. Danchin, and A. A. Ecology 77:430–438.
Dhondt. 2001. Introduction. In Dispersal, edited by J. Coleman, S. W., G. L. Patricelli, and G. Borgia. 2004.
Clobert, E. Danhin, A. A. Dhondt, and J. D. Nichols. Variable female preferences drive complex male displays.
Oxford: Oxford University Press, pp. xvii–xxi. Nature 428:742–746.
Clucas, B., M. P. Rowe, D. H. Owings, and P. C. Arrowood. Collett, M., T. S. Collett, S. Bisch, and R. Wehner. 1998.
2008. Snake scent application in ground squirrels, Local and global vectors in desert ant navigation. Nature
Spermophilus spp.: a novel form of antipredator behaviour? 394:269–272.
Animal Behaviour 75:298–307. Collett, T. S., and M. F. Land. 1978. How hoverflies compute
Clutton-Brock, T. H. 1989. Female transfer and inbreeding interception courses. Journal of Comparative Physiology
avoidance in social mammals. Nature 337:70–72. 125:191–204.
Clutton-Brock, T. H. 1991. The Evolution of Parental Care. Compagnone, N. A., and S. H. Mellon. 2000. Neurosteroids:
Princeton, NJ: Princeton University Press. biosynthesis and function of these neuromodulators.
Frontiers in Neuroendocrinology 21:1–56.
Clutton-Brock, T. H. 2002. Breeding together: kin selection
Connor, V. M. 1986. The use of mucous trails by intertidal
and mutualism in cooperative vertebrates. Science
limpets to enhance food resources. Biological Bulletin
296:69–72.
171:548–564.
Clutton-Brock, T. H., and C. Godfray. 1991. Parental invest-
Cook, A. 1981. Huddling and the control of water loss by the
ment. In Behavioural Ecology: An Evolutionary Approach, 3rd
slug Limax pseudoflavus Evans. Animal Behaviour
ed., edited by J. R. Krebs and N. B. Davies. Oxford:
29:289–298.
Blackwell Scientific, pp. 7–29.
Cook, M., and S. Mineka. 1990. Selective associations in the
Clutton-Brock, T. H., and P. H. Harvey. 1979. Comparison
observational conditioning of fear in rhesus monkeys.
and adaptation. Proceedings of the Royal Society of London B Journal of Experimental Psychology 16:372–389.
205:547–565.
Cook, P. A., and N. Wedell. 1999. Non-fertile sperm delay
Clutton-Brock, T. H., and P. H. Harvey. 1984. Comparative female remating. Nature 397:486.
approaches to studying adaptation. In Behavioural Ecology: Cooper, J. B. 1985. Comparative psychology and ethology. In
An Evolutionary Approach, 2nd ed., edited by J. R. Krebs and Topics in the History of Psychology, edited by G. A. Kimble and
N. B. Davies. Sunderland, MA: Sinauer, pp. 7–29. K. Schlesinger. Hillsdale, NJ: Erlbaum.
Clutton-Brock, T. H., and K. Isvaran. 2006. Paternity loss in Cooper, J., A. T. Scholz, R. M. Horral, A. D. Hasler, and D.
contrasting mammalian societies. Biology Letters 2:513–516. M. Madison. 1976. Experimental confirmation of the olfac-
Clutton-Brock, T. H., and A. C. J. Vincent. 1991. Sexual selec- tory hypothesis with homing, artificially imprinted coho
tion and the potential reproductive rates of males and salmon (Oncorhynchus kisutch). Journal of the Fisheries Research
females. Nature 351:58–60. Board of Canada. 28:703–710.
470 References

Cooper, W. E., V. Pérez-Mellado, and L. J. Vitt. 2004. Ease Crespi, B. J. 1992. Eusociality in Australian gall thrips. Nature
and effectiveness of costly autotomy vary with predation 359:724–726.
intensity among lizard populations. Journal of Zoology Crespi, B. J., and J. E. Ragsdale. 2000. A skew model for the
262:243–255. evolution of sociality via manipulation: why it is better to
Cordoba-Aguilar, A. 1999. Male copulatory sensory stimula- be feared than loved. Proceedings of the Royal Society of London
tion induces female ejection of rival sperm in a damselfly. B 267:821–828.
Proceedings of the Royal Society of London B 266:779–784. Crews, D. 1974. Effects of castration and subsequent andro-
Cott, H. B. 1940. Adaptive Colouration in Animals. London: gen replacement therapy on male courtship behavior and
Methuen. environmentally induced ovarian recrudescence in the
Coulon, A., J.-F. Cosson, N. Morellet, J.-M. Angibault, B. lizard Anolis carolinensis. Journal of Comparative Physiology
Cargnelutti, M. Galan, S. Aulagnier, and A. J. M. Hewison. 87:963–969.
2006. Dispersal is not female biased in a resource-defense
Crews, D. 1979a. Neuroendocrinology of lizard reproduction.
mating ungulate, the European roe deer. Proceedings of the
Biology of Reproduction 20:51–73.
Royal Society of London B 273:341–348.
Crews, D. 1979b. The hormonal control of behavior in a
Coulson, J. C. 1966. The influence of the pair bond and age
lizard. Scientific American 241:180–187.
on the breeding biology of the kittiwake gull (Risa tridactyla).
Journal of Animal Biology 35:269–279. Crews, D. 1983. Control of male sexual behavior in the
Canadian red-sided garter snake. In Hormones and Behavior
Cowlishaw, G. 1997. Trade-offs between foraging and preda-
tion risk determine habitat use in a desert baboon popula- in Higher Animals, edited by J. Balthazart, E. Prove, and R.
tion. Animal Behaviour 53:667–686. Gilles. New York: Springer-Verlag, pp. 398–406.
Cox, C. R., and B. J. LeBoeuf. 1977. Female incitation of male Crews, D. 1984. Gamete production, sex hormone secretion,
competition: A mechanism of mate selection. American and mating behavior uncoupled. Hormones and Behavior
Naturalist 111:317–335. 18:22–28.
Crampton, W. G. R., and C. D. Hopkins. 2005. Nesting and Crews, D. 1987. Diversity and evolution of behavioral con-
paternal care in the weakly electric fish Gymnotus trolling mechanisms. In Psychobiology of Reproductive
(Gymnotiformes: Gymnotidae) with descriptions of larval Behavior, edited by D. Crews. Upper Saddle River, NJ:
and adult electric organ discharges of two species. Copeia Prentice Hall, pp. 88–119.
2005:48–60. Crews, D., and T. Groothuis. 2005. Tinbergen’s fourth ques-
Crane, J. 1941. Eastern Pacific expeditions of the New York tion, ontogeny: sexual and individual differentiation. Animal
Zoological Society XXVI: Crabs of the genus Uca from the Biology 55:343–370.
West Coast of Central America. Zoologica 26:145–203. Crews, D., and M. C. Moore. 2005. Historical contributions
Creel, S. 2005. Dominance, aggression, and glucocorticoid of research on reptiles to behavioral neuroendocrinology.
levels in social carnivores. Journal of Mammalogy Hormones and Behavior 48:384–394.
86:255–264. Crews, D., V. Hingorani, and R. J. Nelson. 1988. Role of the
Creel, S., and P. M. Waser. 1994. Inclusive fitness and repro- pineal gland in the control of annual reproductive behav-
ductive strategies in dwarf mongooses. Behavioral Ecology ioral and physiological cycles in the red-sided garter snake
5:339–348. (Thamnophis sirtalis parietalis). Journal of Biological Rhythms
Creel, S. R., and P. M. Waser. 1997. Variation in reproduc- 3:293–302.
tive suppression among dwarf mongooses: Interplay
Crews, D., V. Tranina, F. T. Wetzel, and C. Muller. 1978.
between mechanisms and evolution. In Cooperative Breeding
Hormone control of male reproductive behavior in the lizard
in Mammals, edited by N. G. Solomon and J. A. French.
Anolis carolinensis: Role of testosterone, dihydrotestosterone,
Cambridge: Cambridge University Press.
and estradiol. Endocrinology 103:1814–1821.
Creel, S., N. M. Creel, M. G. L. Mills, and S. L. Monfort.
Crews, D., A. C. Gore, T. S. Hsu, N. L. Dangleben, M.
1997. Rank and reproduction in cooperatively breeding
Spinetta, T. Schaller, and M. D. Skinner Anway, M. L. 2007.
African wild dogs—Behavioral and endocrine correlates.
Behavioral Ecology 8:298–306. Transgenerational epigenetic imprints on mate preference.
Proceedings of the National Academy of Sciences 104:5942–5956.
Creel S., N. Creel, D. E. Wildt, and S. L. Monfort. 1992.
Behavioral and endocrine mechanisms of reproductive sup- Cristol, D. A. 1995. The coat-tail effect in merged flocks of
pression in Serengeti dwarf mongooses. Animal Behaviour dark-eyed juncos: Social status depends on familiarity.
43:231–245. Animal Behaviour 50:151–159.
Creel, S., S. L. Monfort, N. Marusha Creel, D. E. Wildt, and Cristol, D. A., and P. V. Switzer. 1999. Avian prey dropping
P. M. Waser. 1995. Pregnancy, oestrogens and future repro- behavior. II. American crows and walnuts. Behavioral
ductive success in Serengeti dwarf mongooses. Animal Ecology 10:220–226.
Behaviour 50:1132–1135. Crook, J. H. 1964. The evolution of social organisation and
Cremer, S., S. Armitage, and P. Schmid-Hempel. 2007. Social visual communication in the weaver bird (Ploceinae).
immunity. Current Biology 17:R693–R702. Behaviour Supplement 10:1–178.
 References 471

Crook, J. H. 1970. The socio-ecology of primates. In Social Davies, N. B. 1991. Mating systems. In Behavioural Ecology: An
Behavior in Birds and Mammals, edited by J. H. Crook. Evolutionary Approach, edited by J. R. Krebs and N. B.
London: Academic Press, pp. 103–106. Davies. Oxford: Blackwell Scientific, pp. 263–294.
Croze, H. 1970. Searching image in carrion crows. Zeitschrift Davies, N. B. 2000. Cuckoos, Cowbirds and Other Cheats.
für Tierpsychologie Supplement 5:1–86. London: T., and A. D. Poyser.
Cullen, E. 1957a. Adaptations in the kittiwake in the British Davies, N. B., and M. de L. Brooke. 1988. Cuckoo versus reed
Isles. Bird Study 10:147–179. warblers: Adaptations and counteradaptations. Animal
Cullen, E. 1957b. Adaptations in the kittiwake to cliff nest- Behaviour 36:262–284.
ing. Ibis 99:275–302. Davies, N. B., and T. R. Halliday. 1978. Deep croaks and
Cummings, M. E. 2004. Modelling divergence in luminance fighting assessment in toads (Bufo bufo). Nature
and chromatic detection performance across measured 274:683–685.
divergence in surfperch (Embiotocidae) habitats. Vision Davies, N. B., and A. I. Houston. 1984. Territory economics.
Research 44:1127–1145. In Behavioural Ecology: An Evolutionary Approach, 2nd edi-
Cummings, M. E. 2007. Sensory trade-offs predict signal tion, edited by J. R. Krebs and N. B. Davies. Sunderland,
divergence in surfperch. Evolution 61:530–545. MA: Sinauer.
Cummings, M. E., and J. C. Partridge. 2001. Visual pigments Davis, A. K., N. Cope, A. Smith, and M. J. Solensky. 2007.
and optical habitats of surfperch (Embiotocidae) in the Wing color predicts future mating success in male monarch
California kelp forest. Journal of Comparative Physiology A butterflies. Annals of the Entomological Society of America 100:
187:875–889. 339–344.
Curio, E. 1978. The adaptive significance of avian mobbing. Davis, J. M., and J. A. Stamps. 2004. The effect of natal expe-
I. Teleonomic hypotheses and predictions. Ethology rience on habitat preferences. Trends in Ecology and Evolution
48:175–183. 19:411–416.
Curio, E., and K. Regelmann. 1986. Predator harassment Davis, M. S. 1987. Acoustically mediated neighbor recogni-
implies a deadly risk: A reply to Hennessy. Ethology tion in the North American bullfrog, Rana catesbeiana.
72:75–78. Behavioral Ecology and Sociobiology 21:185–190.
Cuthill, I. C., M. Stevens, J. Sheppard, T. Maddocks, C. Davis, R. L., J. Cherry, B. Dauwalder, P.-L. Han, and E.
Alejandro Párraga, and T. S. Troscianko. 2005. Disruptive Skoulakis. 1995. The cyclic AMP system and Drosophila learn-
coloration and background pattern matching. Nature ing. Molecular and Cellular Biochemistry 149–150:271–178.
434:72–74. Dawkins, M. S. 1986. Unravelling Animal Behavior. Essex:
Daanje, A. 1950. On locomotor movements in birds and the Longman.
intention movements derived from them. Behaviour 3:48–98. Dawkins, M. S. 1989. The future of ethology: How many legs
Dacke, M., and M. V. Srinivasan. 2007. Honeybee navigation: are we standing on? In Perspectives in Ethology, edited by P.
distance estimation in the third dimension. Journal of P. G. Bateson and P. H. Klopfer. New York: Plenum Press.
Experimental Biology 210:845–853. pp. 47–54.
Daly, M., and M. Wilson. 1983. Sex, Evolution and Behavior. Dawkins, R. 1976. The Selfish Gene. New York: Oxford
2nd ed. Boston: PWS-Kent. University Press.
Daniels, S. J., and J. R. Walters. 2000. Between-year breed- Dawkins, R. 1980. Good strategy or evolutionarily stable
ing dispersal in red-cockaded woodpeckers: multiple causes strategy? In Sociobiology: Beyond Nature/Nurture, edited by
and estimated cost. Ecology 81:2473–2484. G. Barlow and R. Silverberg. Boulder, CO: Westview, pp.
Darst, C. R., M. E. Cummiings, and D. C. Cannatella. 2006. 331–367.
A mechanism for diversity in warning signals: conspicu- Dawkins, R. 1982. The Extended Phenotype. Oxford:Freeman.
ousness versus toxicity in poison frogs. Proceedings of the Dawkins, R., and J. R. Krebs. 1978. Animal signals:
National Academy of Sciences 103:5852–5857. Information or manipulation? In Behavioural Ecology: An
Darwin, C. 1859. On the Origin of Species by Natural Selection Evolutionary Approach, edited by J. R. Krebs and N. B.
or The Preservation of Favored Races in the Struggle for Life. Davies. Sunderland, MA: Sinauer, pp. 282–309.
London: Murray. Dawson, B. V., and B. M. Foss. 1965. Observational learning
Darwin, C. 1871. The Descent of Man and Selection in Relation in budgerigars. Animal Behaviour 13:470–474.
to Sex. London: Murray. de Belle, J. S., A. J. Hilliker, and M. B. Sokolowski. 1989.
Darwin, C. 1873. Expression of the Emotions in Man and Genetic localization of foraging (for): A major gene for lar-
Animals. London: Murray. val behavior in Drosophila melanogaster. Genetics 123:157–163.
Davies, N. B. 1978. Territorial defense in the speckled wood de Belle, J. S., and M. B. Sokolowski. 1987. Heredity of
butterfly, Pararge aegeria: the resident always wins. Animal rover/sitter: Alternative foraging strategies of Drosophila
Behaviour 26:148–147. melanogaster larvae. Heredity 59:73–83.
Davies, N. B. 1983. Polyandry, cloaca-pecking and sperm de la Motte, I., and D. Burkhardt. 1983. Portrait of an Asian
competition in dunnocks. Nature 302:334–336. stalk-eyed fly. Naturwissenschaften 70:451–461.
472 References

de Vos, G. E. 1979. Adaptedness of arena behavior in black Devillard, S., D. Allaine, J.-M. Gaillard, and D. Pontier. 2004.
grouse (Tetrao tetrix) and other grouse species (Tetraonidae). Does social complexity lead to sex-biased dispersal in polyg-
Behaviour 68:277–314. ynous mammals? Behavioral Ecology 15:83–87.
de Waal, F. B. M. 1991. Rank distance as a central feature of Dewsbury, D. A. 1978. What is (was?) the ‘fixed action pat-
rhesus monkey social organization: A sociometric analysis. tern’? Animal Behaviour 26:310–311.
Animal Behaviour 41:383–395. Dewsbury, D. A. 1982a. Ejaculate cost and male choice.
de Waal, F. B. M. 2001. The Ape and the Sushi Master. New American Naturalist 119:601–610.
York: Basic Books. Dewsbury, D. A. 1982b. Dominance rank, copulatory behav-
de Waal, F. B. M. 2005. A century of getting to know the ior, and differential reproduction. Quarterly Review of
chimpanzee. Nature 437:56–59. Biology 57:135–158.
Dewsbury, D. A. 1984. Comparative Psychology in the Twentieth
Dean, J., D. J. Aneshansley, H. E. Edgerton, and T. Eisner.
Century. Stroudsburg, PA: Hutchinson Ross.
1990. Defensive spray of the bombardier beetle: A biolog-
Dewsbury, D. A. 1988. A test of the role of copulatory plugs
ical pulse jet. Science 248:1219–1221.
in sperm competition in deer mice (Peromyscus maniculatus).
DeCoursey, P. J. 1986. Light-sampling behavior in photoen-
Journal of Mammalogy 69:854–857.
trainment of a rodent circadian rhythm. Journal of
Dewsbury, D. A. 1989. A brief history in the study of animal
Comparative Physiology A 159:161–169.
behavior in North America. Perspectives in Ethology
DeCoursey, P. J. 2004a. The behavioral ecology and evolution Volume 8, Whither ethology? pp. 85–122.
of biological timing systems. In Chronobiology: Biological
Dewsbury, D. A., and D. K. Sawrey. 1984. Male capacity as
Timekeeping, edited by J. C. Dunlap, J. J. Loros, and P. J. related to sperm production, pregnancy initiation, and
DeCoursey. Sunderland, MA: Sinauer, pp. 27–65. sperm competition in deer mice (Peromyscus maniculatus).
DeCoursey, P. J. 2004b. Functional organization of circadian Behavioral Ecology and Sociobiology 16:37–47.
systems in multicellular organisms. In Chronobiology: Dial, B. E., and L. C. Fitzpatrick. 1983. Lizard tail autotomy:
Biological Timekeeping, edited by J. C. Dunlap, J. J. Loros, and Function and energetics of postautotomy tail movement in
P. J. DeCoursey. Sunderland, MA: Sinauer, pp. 145–178. Scincella lateralis. Science 219:391–393.
DeCoursey, P. J., J. K. Walker, and S. A. Smith. 2000. A cir- Díaz, P. C., C. Grüter, and W. M. Farina. 2007. Floral scents
cadian pacemaker in free-living chipmunks: essential for affect the distribution of hive bees around dancers.
survival? Journal of Comparative Physiology A 186:169–180. Behavioral Ecology and Sociobiology 61:1589–1597.
DeCoursey, P. J., J. R. Krulas, G. Mele, and D. C. Holley. Dickinson, J. L., and B. J. Hatchwell. 2004. Fitness conse-
1997. Circadian performance of suprachiasmatic nuclei quences of helping. In Ecology and Evolution of Cooperative
(SCN)-lesioned antelope ground squirrels in a desert Breeding in Birds, edited by W. Koenig and J. Dickinson.
enclosure. Physiology and Behavior 625:1099–1108. Cambridge: Cambridge University Press, pp. 48–66.
Dehn, M. M. 1990. Vigilance for predators: Detection and Dickson, B. J. 2008. Wired for sex: The neurobiology of
dilution effects. Behavioral Ecology and Sociobiology Drosophila mating decisions. Science 322:904–908.
26:337–342. Diesel, R. 1990. Sperm competition and reproductive success
Delaney, K. J., and L. G. Higley. 2006. An insect counter- in the decapod Inachus phalangium (Majiidae): A male ghost
measure impacts plant physiology: midrib vein cutting, spider crab that seals off rival’s sperm. Journal of Zoology
220:213–223.
defoliation, and leaf photosynthesis. Plant Cell and
Environment 29:1245–1258. Dill, P. A. 1977. Development of behavior in alevins of
Atlantic salmon, Salmo salar, and rainbow trout, S. gairdneri.
Delcomyn, F. 1980. Neural basis of rhythmic behavior in ani-
Animal Behaviour 25:116–121.
mals. Science 210:492–498.
Dingemanse, N. J., C. Both, A. J. van Noordwijk, A. L.
Delgado, M. M., and V. Penteriani. 2007. Vocal behaviour and
Rutten, and P. J. Drent. 2003. Natal dispersal and person-
neighbour spatial arrangement during vocal displays in
alities in great tits (Parus major). Proceedings of the Royal
eagle owls. Journal of Zoology 271:3–10. Society of London B 270:741–747.
DeNault, L. K., and D. A. McFarlane. 1995. Reciprocal altru- Dingle, H. 1980. Ecology and evolution of migration. In
ism between male vampire bats, Desmodus rotundus. Animal Animal Migration, Orientation, and Navigation, edited by S.
Behaviour 49:855–856. A. Gauthreaux. New York: Academic Press, pp. 1–101.
Dennis, T. E., M. J. Rayner, and M. M. Walker. 2007. Dingle, H., and R. Caldwell. 1969. The aggressive and terri-
Evidence that pigeons orient to geomagnetic intensity dur- torial behavior of the mantis shrimp Gonodactylus bredini
ing homing. Proceedings of the Royal Society of London B Manning (Crustacea: Stomatopoda). Behaviour 33:115–136.
274:1153–1158. Dingle, H., and V. A. Drake. 2007. What is migration?
Detto, T., P. R. Y. Backwell, J. M. Hemmi, and J. Zeil. 2006. BioScience 57:113–121.
Visually mediated species and neighbour recognition in fid- Dittman, A. H., and T. P. Quinn. 1996. Homing in Pacific
dler crabs (Uca mjoebergi and Uca capricornis). Proceedings of salmon: mechanisms and ecological basis. Journal of
the Royal Society of London B 273:1661–1666. Experimental Biology 199:83–91.
 References 473

Dittrich, G., F. Gilbert, P. Green, P. McGregor, and D. Dupret, D., A. Fabre, M. D. Dobrossy, A. Panatier, J. J.
Grewcock. 1993. Imperfect mimicry—a pigeon’s perspec- Rodríguez, S. Lamarque, V. Lemaire, S. H. R. Oliet, P.-V.
tive. Proceedings of the Royal Society of London B 251:195. Pliazza, and D. N. Abrous. 2007. Spatial learning depends
Dixson, A. F. 1998. Primate Sexuality: Comparative Studies of on both the addition and removal of new hippocampal neu-
Prosimians, Monkeys, Apes, and Human Beings. Oxford: rons. PLoS Biology 5:1683–1694.
Oxford University Press. Dussourd, D. E. 1999. Behavioral sabotage of plant defense:
Dobson, F. S. 1982. Competition for mates and predominant Do vein cuts and trenches reduce insect exposure to exu-
male dispersal in mammals. Animal Behaviour date? Journal of Insect Behavior 12:501–515.
30:1183–1192. Dussutour, A., S. J. Simpson, E. Despland, and N. Colasurdo.
Dobson, F. S., and W. T. Jones. 1985. Multiple causes of dis- 2007. When the group denies individual nutritional wis-
persal. American Naturalist 126:855–858. dom. Animal Behaviour 74:931–939.
Doligez, B., E. Danchin, and J. Clobert. 2002. Public infor- Dyer, A. B., R. Lickliter, and G. Gottlieb. 1989. Maternal peer
mation and breeding habitat selection in a wild bird popu- imprinting in mallard ducklings under experimentally sim-
lation. Science 297:1168–1170. ulated natural social conditions. Developmental Psychobiology
Domenici, P., D. Booth, J. M. Blagburn, and J. P. Bacon. 2008. 22:463–475.
Cockroaches keep predators guessing using preferred Dzieweczynski, T. L., A. C. Eklund, and W. J. Rowland. 2006.
escape trajectories. Current Biology 18:1792–1796. Male 11-ketotestosterone levels change as a result of being
Dong, W., X. Tang, Y. Yu, R. Nilsen, R. Kim, J. Griffith, J. watched in Siamese fighting fish, Betta splendens. General and
Arnold, and H.-B. Schüttler. 2008. Systems biology of the Comparative Endocrinology 147:184–189.
clock in Neurospora crassa. PLoS ONE 3:e3105. Dzieweczynski, T. L., R. L. Earley, T. M. Green, and W. J.
Dorries, K. M., E. Adkins-Regan, and B. P. Halpern. 1995. Rowland. 2005. Audience effect is context dependent in
Olfactory sensitivity to the pheromone androsterone is sex- Siamese fighting fish, Betta splendens. Behavioral Ecology
ually dimorphic in the pig. Physiology and Behavior 16:1025–1030.
57:255–259. Earle, S. A. 1979. The gentle whales. National Geographic
Drake, A., D. Fraser, and D. M. Weary. 2008. Parent-offspring 155:2–17.
resource allocation in domestic pigs. Behavioral Ecology and Eason, P. K. 1992. Optimization of territory shape in hetero-
Sociobiology 62:309–319. geneous habitats—a field-study of the red-capped cardinal
Drent, R., and P. Swierstra. 1977. Goose flocks and food-find- (Paroaria gularis). Journal of Animal Ecology 61:411–425.
ing: field experiments with barnacle geese in winter. Eason, P. K., and P. T. Sherman. 1995. Dominance status,
Wildfowl 28:15–20. mating strategies and copulation success in cooperatively
Drickamer, L. C., P. A. Gowaty, and D. M. Wagner. 2003. polyandrous white-winged trumpeters, Psophia leucoptera
Free mutual mate preferences in house mice affect repro- (Aves: Psophiidae). Animal Behaviour 49:725–736.
ductive success and offspring performance. Animal Eason, P. K., and J. A. Stamps. 1992. The effect of visibility
Behaviour 65:105–114. on territory size and shape. Behavioral Ecology 3:166–172.
Dudai, Y. 1979. Behavioral plasticity in a Drosophila mutant, Eason, P. K., and J. A. Stamps. 1993. An early warning system
dunce DB276. Journal of Comparative Physiology for detecting intruders in a territorial animal. Animal
130:271–275. Behaviour 46:1005–1109.
Dudai, Y., Y.-N. Jan, D. Byers, W. G. Quinn, and S. Benzer. Eason, P. K., and J. A. Stamps. 2001. The effect of visibility
1976. Dunce, a mutant of Drosophila deficient in learning. on space use by territorial red-capped cardinals. Behaviour
Proceedings of the National Academy of Sciences 73:1684–1688. 138:19–30.
Duffield, G. E. 2003. DNA microarray analyses of circadian Eason, P. K., G. A. Cobbs, and K. G. Trinca. 1999. The use
timing: The genomic basis of biological time. Journal of of landmarks to define territorial boundaries. Animal
Neuroendocrinology 15:991–1002. Behaviour 58:85–91.
Duffy, J. E. 1996. Eusociality in a coral-reef shrimp. Nature Ebensperger, L. A., and D. T. Blumstein. 2007. Nonparental
381:512–514. infanticide. In Rodent Societies: An Ecological and Evolutionary
Dugatkin, L. A. 1997. Cooperation Among Animals: An Perspective, edited by J. O. Wolff and P. W. Sherman.
Evolutionary Perspective. New York: Oxford University Chicago: The University of Chicago Press, pp. 267–279.
Press. Eberhard, W. G. 1996. Female Control: Sexual Selection by
Dulac, C. 2005. Sex and the single splice. Cell 121:664–666. Cryptic Female Choice. Princeton, NJ: Princeton University
Dunham, P. J. 1986. Mate guarding in amphipods: A role for Press.
brood patch stimuli. Biological Bulletin 170:526–531. Eckerle, K. P., and C. F. Thompson. 2006. Mate choice in
Dunham, P. J., and A. Hurshman. 1990. Precopulatory mate house wrens: nest cavities trump male characteristics.
guarding in the amphipod, Gammarus lawrencianus: Effects Behaviour 143:253–271.
of social stimulation during the post-copulation interval. Edwards, J. S. 2006. The central nervous control of insect
Animal Behaviour 39:976–979. flight. Journal of Experimental Biology 209:4411–4413.
474 References

Eens, M., and R. Pinxten. 2000. Sex-role reversal in verte- Emlen, S. T. 1975. Migration, orientation, navigation. In
brates: behavioral and endocrinological accounts. Avian Biology, edited by D. Farner. New York: Academic
Behavioural Processes 51:135–147. Press.
Eggert, A.-K., and S. K. Sakaluk. 1994. Sexual cannibalism Emlen, S. T. 1991. Evolution of cooperative breeding in birds
and its relation to male mating success in sagebrush crick- and mammals. In Behavioural Ecology: An Evolutionary
ets, Cyphoderris strepitans (Haglidae: Orthoptera). Animal Approach, 3rd ed., edited by J. R. Krebs and N. B. Davies.
Behaviour 47:1171–1177. Oxford, Blackwell, pp. 301–337.
Eibl-Eibesfeldt, I. 1966. Das Vertridigen der Eiablageplätze Emlen, S. T. 1997. Predicting family dynamics in social ver-
bei der Hood-Meerechse (Amblyrhynchus cristatus venustis- tebrates. In Behavioural Ecology: An Evolutionary Approach,
simus). Zeitschrift für Tierpsychologie 23:627–631. 4th ed., edited by J. R. Krebs and N. B. Davies. Oxford:
Eibl-Eibesfeldt, I. 1975. Ethology: The Biology of Behavior. 2nd Blackwell Science, pp. 228–253.
ed. New York: Holt, Rinehart & Winston. Emlen, S. T., and L. W. Oring. 1977. Ecology, sexual selec-
Eisenberg, J. F., N. A. Muckenhirn, and R. Rudran. 1972. The tion, and the evolution of mating systems. Science
relation between ecology and social structure in primates. 197:215–223.
Science 176:963–874. Emlen, S. T., and S. L. Vehrencamp. 1983. Cooperative
Eisner, T. 1958. The protective role of the spray mechanism breeding strategies among birds. In Perspectives in
of the bombardier beetle, Brachynus ballistarius Lec. Journal Ornithology, edited by A. H. Brush and G. A. Clark, Jr.
of Insect Physiology 2:215–220. Cambridge: Cambridge University Press, pp. 93–133.
Eisner, T., and J. Meinwald. 1995. The chemistry of sexual Emlen, S. T., and P. H. Wrege. 1989. A test of alternative
selection. Proceedings of the National Academy of Sciences hypotheses for helping behavior in white-fronted bee-eaters
92:50–55. of Kenya. Behavioral Ecology and Sociobiology 25:303–319.
Eisner, T., D. F. Wiemer, L. W. Hayes, and J. Meinwald. 1978. Emlen, S. T., and P. H. Wrege. 2004. Division of labour in
Lucibufagins: defensive steroids from the fireflies Photinus parental care behaviour of a sex-role-reversed shorebird, the
ignitus and P. marginellus (Coleoptera: Lampyridae). wattled jacana. Animal Behaviour 68:847–855.
Proceedings of the National Academy of Sciences 75:905–908. Enard, W., M. Przeworski, S. E. Fisher, C. S. L. Lai, V. Wiebe,
Eisner, T., M. A. Goetz, D. E. Hill, S. R. Smedley, and J. T. Kitano, A. P. Monaco, and S. Pääbo. 2002. Molecular
Meinwald. 1997. Firefly ‘femmes fatales’ acquire defensive evolution of FOXP2, a gene involved in speech and lan-
steroids (lucibufagins) from their firefly prey. Proceedings of guage. Nature 418:869–872.
the National Academy of Sciences 94:9723–9728. Endler, J. A. 1978. A predator’s view of animal color patterns.
Ekman, J., J. L. Dickinson, B. J. Hatchwell, and M. Griesser. Evolutionary Biology 11:319–364.
2004. Delayed dispersal. In Ecology and Evolution of Endler, J. A. 1981. An overview of the relationships between
Cooperative Breeding in Birds, edited by W. Koenig and J. mimicry and crypsis. Biological Journal of the Linnean Society
Dickinson. Cambridge: Cambridge University Press, pp. 16:25–31.
35–47. Endler, J. A. 1992. Signals, signal conditions, and the direc-
Ellis, L. 1995. Dominance and reproductive success among tion of evolution. American Naturalist 139:S125–S153.
nonhuman animals: a cross-species comparison. Ethology and Endler, J. A. 2006. Disruptive and cryptic coloration.
Sociobiology 16:257–333. Proceedings of the Royal Society of London B 273:2425–2426.
Emery, N. J., and N. S. Clayton. 2005. Animal cognition. In Engel, J. E., and R. R. Hoy. 2000. A cGMP-dependent kinase
The Behavior of Animals: Mechanisms, Function, and Evolution, gene, foraging, modifies habituation-like response decre-
edited by J. J. Bolhuis and L.-A. Giraldeau. Malden MA: ment of the giant fiber escape circuit in Drosophila. Learning
Blackwell Publishing Ltd. and Memory 7:341–352.
Emlen, S. T. 1967a. Migratory orientation in the indigo Engström-Öst, J., and U. Candolin. 2006. Human-induced
bunting (Passerina cyanea). Part I. Evidence for the use of water turbidity alters selection on sexual displays in stick-
celestial cues. Auk 84:309-342. lebacks. Behavioral Ecology 18:393–398.
Emlen, S. T. 1967b. Migratory orientation in the indigo Epstein, R., C. E. Kirshnit, R. P. Lanza, and L. C. Rubin.
bunting (Passerina cyanea). Part II. Evidence for use of celes- 1984. “Insight” in the pigeon: Antecedents and determi-
tial cues. Auk 84:463-489. nants of intelligent performance. Nature 308:61–62.
Emlen, S. T. 1969. The development of migratory orientation Esch, H. E., S. Zhang, M. V. Srinivasan, and J. Tautz. 2001.
in young Indigo Bunting. Living Bird 8:113–126. Honeybee dances communicate distances measured by optic
Emlen, S. T. 1970. Celestial rotation: Its importance in the flow. Nature 411:581–583.
development of migratory orientation. Science 170:1198–1201. Estep, D. Q., K. Nieuwenhuijsen, K. E. M. Bruce, K. J. De
Emlen, S. T. 1972. The ontogenetic development of orienta- Neef, P. A. Walters, S. C. Baker, and A. K. Slob. 1988.
tion capabilities. NASA Special Publication NASA Inhibition of sexual behavior among subordinate stumptail
SP–262:191–210. macaques, Macaca arctoides. Animal Behaviour 36:854–864.
 References 475

Evans, C. S., L. Evans, and P. Marler. 1993. On the meaning Fisher, A. C., Jr. 1979. The mysteries of animal migration.
of alarm calls: functional reference in an avian vocal system. National Geographic 156:154–193.
Animal Behaviour 46:23–38. Fisher, J., and R. A. Hinde. 1949. The opening of milk bot-
Fagan, R. 1981. Animal Play Behavior. New York: Oxford tles by birds. British Birds 42:347–357.
University Press. Fisher, R. A. 1930. The Genetical Theory of Natural Selection.
Falls, J. B. 1982. Individual recognition by sounds in birds. In Oxford: Oxford University Press.
Acoustic Communication in Birds, Volume 2, edited by D. E. Fitch, W. T. 1997. Vocal tract length and formant frequency
Kroodsma and E. H. Miller. New York: Academic Press, pp. dispersion correlate with body size in rhesus macaques.
237–278. Journal of the Acoustic Society of America 102:1213–1222.
Farrell, W. J., and W. Wilczynski. 2006. Aggressive experience Fitch, W. T., and M. D. Hauser. 2003. Unpacking “honesty”:
alters place preference in green anole lizards, Anolis caroli- vertebrate vocal production and the evolution of acoustic
nensis. Animal Behaviour 71:1155–1164. signals. In Acoustic Communication, edited by A. M.
Feder, M. E., and T. Mitchell-Olds. 2003. Evolutionary and Simmons, R. R. Fay, and A. N. Popper. New York: Springer,
ecological functional genetics. Nature Reviews Genetics pp. 65–137.
4:649–655. FitzGibbon, C. D., and J. H. Fanshawe. 1988. Stotting in
Fedina, T. Y. 2007. Cryptic female choice during sper- Thomson’s gazelles: An honest signal of condition.
matophore transfer in Tribolium castaneum (Coleoptera: Behavioral Ecology and Sociobiology 23:69–74.
Tenebrionidae). Journal of Insect Physiology 53:93–98. Fitzpatrick, M. J., and M. B. Sokolowski. 2004. In search of
Fedina, T. Y., and S. M. Lewis. 2004. Female influence over food: Exploring the evolutionary link between cGMP-
offspring paternity in the red flour beetle Tribolium casta- dependent protein kinase (PKG) and behavior. Integrative
neum. Proceedings of the Royal Society of London B and Comparative Biology 44:28–36.
271:1393–1399. Fitzpatrick, M. J., Y. Ben-Shahar, H. M. Smid, L. E. M. Vet,
Fedina, T. Y., and S. M. Lewis. 2006. Proximal traits and G. E. Robinson, and M. B. Sokolowski. 2005. Candidate
mechanisms for biasing paternity in the red flour beetle genes for behavioral ecology. Trends in Ecology and Evolution
Tribolium castaneum (Coleoptera: Tenebrionidae). Behavioral 20:96–104.
Ecology and Sociobiology 60:844–853. Flamarique, N., G. A. Mueller, C. L. Cheng, and C. R. Figiel.
Feltmate, B. W., and D. D. Williams. 1989. A test of crypsis 2007. Communication using eye roll reflexive signaling.
and predator avoidance in the stonefly Paragnetina media Proceedings of the Royal Society of London B 274:877–882.
(Plecoptera: Perlidae). Animal Behaviour 37:992–999. Foelix, R. F. 1996. The Biology of Spiders, 2nd ed. Oxford:
Feng, A. S., P. M. Narins, C. H. Xu, W. Y. Lin, Z. L. Yu, Q. Oxford University Press.
Qiu, Z. M. Xu, and J. X. Shen. 2006. Ultrasonic commu- Foellmer, M. W., and D. J. Fairbairn. 2003. Spontaneous male
nication in frogs. Nature 440:333–336. death during copulation in an orb-weaving spider.
Fenton, M. B. 1983. Just Bats. Toronto: University of Toronto Proceedings of the Royal Society of London B 270
Press. (Suppl):183–185.
Fenton, M. B. 1992. Bats. New York: Facts on File. Forsman, J. T., J.-T. Seppänen, and M. Mönkkönen. 2002.
Ferkau, C., and K. Fischer. 2006. Costs of reproduction in Positive fitness consequences of interspecific interaction
male Bicyclus anynana and Pieris napi butterflies: effects of with a potential competitor. Proceedings of the Royal Society
mating history and food limitation. Ethology 112:1117–1127. of London B 269:1619–1623
Ferkin, M. H., and I. Zucker. 1991. Seasonal control of odor Fort, K. T., and K. A. Otter. 2004. Territorial breakdown of
preferences of meadow voles (Microtus pennsylvanicus) by black-capped chickadees, Poecile atricapillus, in disturbed
photoperiod and ovarian hormones. Journal of Reproduction habitats. Animal Behaviour 68:407–415.
and Fertility 92:433–441. Foster, M. S. 1977. Odd couples in manakins: A study of social
Féron, C., and P. Gouat. 2007. Paternal care in the mound- organization and cooperative breeding in Chiroxiphia lin-
building mouse reduces inter-litter intervals. Reproduction, earis. American Naturalist 111:845–853.
Fertility, and Development 19:425–429. Fournier, F., and M. Festa-Bianchet. 1995. Social dominance
Ferster, C. B., and B. F. Skinner. 1957. Schedules of in adult female mountain goats. Animal Behaviour
Reinforcement. New York: Appleton-Century-Crofts. 49:1449–1459.
Fields, R. D. 2007. The shark’s electric sense. Scientific Fouts, R. S. 1973. Acquisition and testing of gestural signs in
American 297:74–81. four young chimpanzees. Science 180:978–980.
Fink, L. S., and L. P. Brower. 1981. Birds can overcome the Fouts, R. S. 1974. Language: Origins, definition and chim-
cardenolide defence of monarch butterflies in Mexico. panzees. Journal of Human Evolution 3:475–482.
Nature 291:67–70. Fouts, R., and S. T. Mills. 1997. Next of Kin: What Chimpanzees
Finley, J., D. Ireton, W. M. Schleidt, and T. A. Thompson. Have Taught Me about Who We Are. New York: Morrow.
1983. A new look at the features of mallard courtship dis- Fox, S. F., and M. A. Rostker. 1982. Social cost of tail loss in
plays. Animal Behaviour 31:348–354. Uta stansburiana. Science 218:692–693.
476 References

Francis, D., J. Diorio, D. Liu, and M. J. Meaney. 1999. Gagliardo, A., P. Ioalè, M. Savini, and J. M. Wild. 2006.
Nongenomic transmission across generations of maternal Having the nerve to home: trigeminal magnetoreceptor ver-
and stress responses in the rat. Science 286:1155–1158. sus olfactory mediation of homing in pigeons. Journal of
Franks, N. R., J. W. Hooper, A. Dornhaus, P. J. Aukett, A. L. Experimental Biology 209:2888–2892.
Hayward, and S. M Berghoff. 2007. Reconnaissance and Gahr, M., E. Sonnenschein, and W. Wickler. 1998. Sex dif-
latent learning in ants. Proceedings of the Royal Society of ference in the size of neural song control regions in a duet-
London B 274:1505–1509. ing songbird with similar song repertoire size of males and
Fraser, D., and B. K. Thompson. 1991. Armed sibling rivalry females. Journal of Neuroscience 18:1124–1131.
among suckling piglets. Behavioral Ecology and Sociobiology Galef, B. G., Jr. 1976. Social transmission of acquired behav-
29:9–15. ior: A discussion of tradition and social learning in verte-
Fraser, D., and D. M. Weary. 2005. Applied animal behavior brates. In Advances in the Study of Behavior, edited by J. S.
and animal welfare. In The Behavior of Animals: Mechanisms, Rosenblatt, R. A. Hinde, E. Shaw, and C. Beer. New York:
Function, and Evolution, edited by J. J. Bolhuis and L.-A. Academic Press, pp. 77–100.
Giraldeau. Malden, MA: Blackwell Publishing. pp. 345–366. Galef, B. G., Jr. 1988. Imitation in animals: History, defini-
Frei, U. 1982. Homing pigeons behavior in the irregular mag- tion, and interpretation of data from the psychological lab-
netic field of western Switzerland. In Avian Navigation, oratory. In Social Learning: Psychological and Biological
edited by F. Papi and H. G. Wallraff. Berlin: Springer- Perspectives, edited by T. R. Zentall and B. G. Galef, Jr.
Verlag. pp. 129–139. Hillsdale, NJ: Erlbaum, pp. 3–28.
Frei, U., and G. Wagner. 1976. Die Anfangsorientierung von Galef, B. G., Jr. 1990a. An adaptationist perspective on social
Brieftauben im erdmagnetisch gestörten Gebiet des Mont learning, social feeding, and social foraging in Norway rats.
Jorat. Revue Suisse De Zoologie 83:891–897. In Contemporary Issues in Comparative Psychology, edited by
Freitas, M. B., C. B. C. Passos, R. B. Vasconcelos, and E. C. D. A. Dewsbury. Sunderland, MA: Sinauer, pp. 55–79.
Pinheiro. 2005. Effects of short-term fasting on energy Galef, B. G. 1990b. The question of animal culture. Human
reserves of vampire bats (Desmodus rotundus). Comparative Nature 3:157–178.
Biochemistry and Physiology B 140:59–62.
Galef, B. G., Jr. 1993. Functions of social learning about food:
Freitas, M. B., A. F. Welker, S. F. Millan, and E. C. Pinheiro.
A causal analysis of effects of diet novelty on preference
2003. Metabolic responses induced by fasting in the com-
transmission. Animal Behaviour 46:257–265.
mon vampire bat (Desmodus rotundus). Journal of Comparative
Galef, B. G., Jr., and T. J. Wright. 1995. Groups of naive rats
Physiology B 173:703–707.
learn to select nutritionally adequate foods faster than do
Fretwell, S. 1972. Populations in a Seasonal Environment.
isolated naive rats. Animal Behaviour 49:403–409.
Princeton, NJ: Princeton University Press.
Gallup, G. G., Jr. 1970. Chimpanzees: self-recognition.
Frey-Roos, F., P. A. Brodmann, and H. U. Reyer. 1995.
Science 167:86–87.
Relationships between food resources, foraging patterns,
Gamboa, G. J. 1978. Intraspecific defense: Advantage of social
and reproductive success in the water pipit (Anthus sp. spin-
oletta). Behavioral Ecology 6:287–295. cooperation among paper wasp foundresses. Science
199:1463–1465.
Fromhage, L., and J. M. Schneider. 2006. Emasculation to
plug up females: the significance of pedipalp damage in Garamszegi, L. Z., and M. Eens. 2004. The evolution of hip-
Nephila fenestrata. Behavioral Ecology 17:353–357. pocampus volume and brain size in relation to food hoard-
Fryxell, J. M., J. Greever, and A. R. E. Sinclair. 1988. Why are ing in birds. Ecology Letters 7:1216–1224.
migratory ungulates so abundant? American Naturalist Gardner, A., and S. A. West. 2007. Social evolution: the
131:781–798. decline and fall of genetic kin recognition. Current Biology
Fujiwara, M., P. Sengupta, and S. L. McIntyre. 2002. 17:R810–R812.
Regulation of body size and behavioral state of C. elegans Gardner, K. E., T. D. Seeley, and N. W. Calderone. 2008. Do
by sensory perception and the elg-4 cGMP-dependent pro- honeybees have two discrete dances to advertise food
tein kinase. Neuron 36:1091–1102. sources? Animal Behaviour 75:1291–1300.
Fullard, J. H., J. W. Dawson, and D. S. Jacobs. 2003. Auditory Gardner, R. A., and B. T. Gardner. 1969. Teaching sign lan-
encoding during the last moment of a moth’s life. Journal guage to a chimpanzee. Science 165:664–672.
of Experimental Biology 206:281–294. Gardner, R. A., and B. T. Gardner. 1989. A cross-fostering
Fusani, L. 2008. Testosterone control of male courtship in laboratory. In Teaching Sign Language to Chimpanzees, edited
birds. Hormones and Behavior 54:227–233. by R. A. Gardner, B. T. Gardner, and T. E. Van Cantfort.
Gagern, A., T. Schurg, N. K. Michiels, G. Schult, D. Albany: State University of New York Press, pp. 1–28.
Sprenger, and N. Anthes. 2008. Behavioural response to Garstang, M. 2004. Long-distance, low-frequency elephant
interference competition in a sessile suspension feeder. communication. Journal of Comparative Physiology A
Marine Ecology—Progress Series 353:131–135. 190:791–805.
 References 477

Garthe, S., S. Benvenuti, and W. A. Montevecchi. 2000. Gittleman, J. L. 1989. The comparative approach in ethology:
Pursuit plunging by northern gannets (Sula bassana) feed- Aims and limitations. In Perspectives in Ethology, edited by
ing on capelin (Mallotus villosus). Proceedings of the Royal P. P. G. Bateson and P. H. Klopfer. New York: Plenum, pp.
Society of London B 267:1717–1722. 55–83.
Gazit, I., A. Goldblatt, and J. Terkel. 2005. Formation of an Gobes, S. M. H., and J. J. Bolhuis. 2007. Birdsong memory:
olfactory search image for explosives odours in sniffer dogs. a neural dissociation between song recognition and pro-
Ethology 111:669–680. duction. Current Biology 17:789–793.
Geist, V. 1971. Mountain Sheep. Chicago: University of Goetz, D. W. 2008. Harmonia axyridis ladybug invasion and
Chicago Press. allergy. Allergy and Asthma Proceedings 29:123–129.
Gerhardt, H. C. 2001. Acoustic communication in two groups Goff, M., M. Salmon, and K. J. Lohmann. 1998. Hatchling
of closely related treefrogs. In P. J. B. Slater, J. S. sea turtles use surface waves to establish a magnetic com-
Rosenblatt, C. T. Snowdon, and T. J. Roper, eds. Advances pass direction. Animal Behaviour 55:69–77.
in Behavioral Biology. New York, Academic Press, pp. Goldman, S. A., and F. Nottebohm. 1983. Neuronal produc-
99–167. tion, migration, and differentiation in a vocal control
Gherardi, F. 2006. Fighting behavior in hermit crabs: the com- nucleus in the female canary brain. Proceedings of the National
bined effect of resource-holding potential and resource Academy of Sciences 80:2390–2394.
value in Pagurus longicarpus. Behavioral Ecology and Gonzalez, A., C. Rossini, M. Eisner, and T. Eisner. 1999.
Sociobiology 59:500–510. Sexually transmitted chemical defense in a moth. Proceedings
Gibson, B. M., and A. C. Kamil. 2005. The fine-grained spa- of the National Academy of Sciences 96:5570–5574.
tial abilities of three seed-caching corvids. Learning and
Goodall, J. 1965. Chimpanzees of the Gombe Stream Reserve.
Behavior 33:59–66.
In Primate Behavior, edited by I. DeVore. New York: Holt,
Gibson, J. S., and G. W. Uetz. 2008. Seismic communication Rinehart & Winston, pp. 425–473.
and mate choice in wolf spiders: components of male seis-
Goodson, J. L. 2005. The vertebrate social behavior network:
mic signals and mating success. Animal Behaviour
Evolutionary themes and variation. Hormones and Behavior
75:1253–1262.
48:11–22.
Gilbert, C., S. Blanc, Y. Le Maho, and A. Ancel. 2008. Energy
Goodson, J. L., A. K. Evans, L. Lindberg, and C. D. Allen.
saving processes in huddling emperor penguins: from exper-
2005. Neuro-evolutionary patterning of sociality.
iments to theory. Journal of Experimental Biology 211:1–8.
Proceedings of the Royal Society of London B 272:227–235.
Gilbert, S. F. 1988. Developmental Biology. Sunderland, MA:
Goodwin, N. B., S. Balshine-Earn, and J. D. Reynolds. 1998.
Sinauer.
Evolutionary transitions in parental care in cichlid fishes.
Gill, F. B., and L. L. Wolf. 1975. Economics of feeding terri-
Proceedings of the Royal Society of London B 265:2265–2272.
toriality in the golden-winged sunbird. Ecology 56:33–45.
Gordon, D. M. 1996. The organization of work in social
Gillette, R., M. U. Gillette, D. J. Green, and R. Huang. 1989.
insect colonies. Nature 380:121–124.
The neuromodulatory response: Integrating second mes-
senger pathways. American Zoologist 29:1275–1286. Gordon, D. M., S. Holmes, and S. Nacu. 2008.The short-
term regulation of foraging in harvester ants. Behavioral
Gillis, E. A., and C. J. Krebs. 2000. Survival of dispersing
Ecology 19:217–22.
versus philopatric snowshoe hares: Do dispersers die? Oikos
90:343–346, Gorzula, S. J. 1978. An ecological study of Caiman crocodilus
Giraldeau, L.-A. 1997. The ecology of information use. In inhabiting savanna lagoons in Venezuelan Guayana.
Behavioural Ecology: An Evolutionary Approach, edited by J. Oecologia 35:21–34.
R. Krebs and N. B. Davies. Oxford: Blackwell Science, pp. Goss-Custard, J. D., J. T. Cayford, and S. E. G. Lea. 1998.
42–68. The changing trade-off between food finding and food
Giraldeau, L.-A. 2006. The function of behavior. In The stealing in juvenile oystercatchers. Animal Behaviour
Behavior of Animals: Mechanisms, Function and Evolution, 55:745–760.
edited by J. Bolhuis and L.-A. Giraldeau. Malden, MA: Gottlieb, G. 1965. Imprinting in relation to parental and
Blackwell Publishing. species identification by avian neonates. Journal of
Giraldeau, L.-A., and T. Caraco. 2000. Social Foraging Theory. Comparative and Physiological Psychology 59:345–356.
Princeton, NJ: Princeton University Press. Gottlieb, G. 1968. Prenatal behavior of birds. Quarterly Review
Giraldeau, L.-A., and D. L. Kramer. 1982. The marginal value of Biology 43:148–174.
theorem: A quantitative test using load size variation in a Gottlieb, G. 1978. Development of species identification in
central place forager, the eastern chipmunk, Tamias striatus. ducklings. IV. Change in species-specific perception caused
Animal Behaviour 30:1036–1042. by auditory deprivation. Journal of Comparative and
Giraldeau, L.-A., C. Soos, G. Beauchamp. 1994. A test of the Physiological Psychology 92:375–387.
producer-scrounger foraging game in captive flocks of spice Gottlieb, G. 1985. Development of species identification in
finches, Lonchura puntulata. Behavioral Ecology and ducklings. XI. Embryonic critical period for species-typical
Sociobiology 34:251–256. perception in the hatchling. Animal Behaviour 33:225–233.
478 References

Gotzek, D., and K. G. Ross. 2007. Genetic regulation of Griffin, D. R. 1991. Progress toward a cognitive ethology. In
colony social organization in fire ants: an integrative Cognitive Ethology, the Minds of Other Animals, edited by C.
overview. Quarterly Review of Biology 82:201–226. A. Ristau. Hillsdale, NJ: Erlbaum, pp. 3–17.
Gould, J. L. 1980. The case for magnetic sensitivity in birds Griffin, D. R. 2001. Animal Minds: Beyond Cognition to
and bees (such as it is). American Scientist 68:256–267. Consciousness. 2nd ed. Chicago: Chicago University Press.
Gould, S. J., and R. C. Lewontin. 1979. The spandrels of San Griffiths, M. 1988. The platypus. Scientific American
Marco and the Panglossian paradigm: A critique of the 258:84–91.
adaptationist programme. Proceedings of the Royal Society of Grinnell, J., C. Packer, and A. E. Pusey. 1995. Cooperation
London B 205:581–598. in male lions: kinship, reciprocity, or mutualism. Animal
Gowaty, P. A. 2003. Sexual natures: how feminism changed Behaviour 49:95–105.
evolutionary biology. Signs 28:901–921. Gröning, J., and A. Hochkirch. 2008. Reproductive interfer-
Gowaty, P. A., R. Steinichen, and W. W. Anderson. 2003. ence between animal species. Quarterly Review of Biology
Indiscriminate females and choosy males: within- and 83:257–282.
between-species variation in Drosophila. Evolution Gronquist, M., F. C. Schroeder, H. Ghiradella, D. Hill, E. M.
57:2037–2045. McCoy, J. Meinwald, and T. Eisner. 2006. Shunning the
Grace, J. A., N. Amin, N. C. Singh, and F. E. Theunissen. night to elude the hunter: diurnal fireflies and the ‘femmes
2003. Selectivity for conspecific song in the zebra finch fatales.’ Chemoecology 16:39–43.
auditory forebrain. Journal of Neurophysiology 89:472–487. Groothuis, T. G. G., W. Muller, N. von Engelhardt, C.
Grafen, A. 1984. Natural selection, kin selection, and group Carere, and C. Eising. 2005. Maternal hormones as a tool
selection. In Behavioural Ecology: An Evolutionary Approach, to adjust offspring phenotype in avian species. Neuroscience
edited by J. R. Krebs and N. B. Davies. Sunderland, MA: and Biobehavioral Reviews 29:329–352.
Sinauer, pp. 62–84. Grosberg, R. K., and J. F. Quinn. 1986. The genetic control
and consequences of kin recognition by the larvae of a colo-
Grafen, A. 1990. Biological signals as handicaps. Journal of
nial marine invertebrate. Nature 322:456–458.
Theoretical Biology 144:544–546.
Gross, M. R. 1982. Sneakers, satellites and parentals:
Gray, E. M. 1997. Do female red-winged blackbirds benefit
Polymorphic mating strategies in North American sun-
genetically from seeking extra-pair copulations? Animal
fishes. Zeitschrift für Tierpsychologie 60:1–26.
Behaviour 54:605–623.
Gross, M. R., and R. C. Sargent. 1985. The evolution of male
Greenwood, J. J. D. 1984. The functional basis of frequency-
and female parental care in fishes. American Zoologist
dependent food selection. Biological Journal of the Linnean
25:807–822.
Society 23:177–199.
Gubernick, D. J., and T. Teferi. 2000. Adaptive significance
Greenwood, J. J. D., P. A. Cotton, and D. M. Wilson. 1989.
of male parental care in a monogamous mammal. Proceedings
Frequency-dependent selection on aposematic prey: Some
of the Royal Society of London B 267:147–150.
experiments. Biological Journal of the Linnean Society
Gubernick, D. J., S. L. Wright, and R. E. Brown. 1993. The
36:213–226.
significance of father’s presence for offspring survival in the
Greenwood, P. J. 1980. Mating systems, philopatry and dis-
monogamous California mouse, Peromyscus californicus.
persal in birds and mammals. Animal Behaviour
Animal Behaviour 46:539–546.
28:1140–1162.
Guillette, L. M., K. L. Hollis, and A. Markarian. 2009.
Gregory, P. T., L. A. Isaac, and R. A. Griffiths. 2007. Death Learning in a sedentary insect predator: Antlions
feigning by grass snakes (Natrix natrix) in response to han- (Neuroptera: Myrmeleontidae) anticipate a long wait.
dling by human “predators.” Journal of Comparative Behavioural Processes 80:224–232.
Psychology 121:123–129.
Gunawan, R., and F. J. Doyle III. 2007. Phase sensitivity analy-
Griffin, A. S., and S. A. West. 2002. Kin selection: fact and sis of circadian rhythm entrainment. Journal of Biological
fiction. Trends in Ecology and Evolution 17:15–21. Rhythms 22:180–194.
Griffin, D. R. 1976. The Question of Animal Awareness: Gundersen, G., and H. P., Andreassen. 1998. Causes and con-
Evolutionary Continuity of Mental Experience. New York: sequences of natal dispersal in root voles Microtus oeconomus.
Rockefeller University Press. Animal Behaviour 56:1355–1366.
Griffin, D. R. 1978. Prospects for a cognitive ethology. Gurd, D. B. 2007. Predicting resource partitioning and com-
Behavior and Brain Science 1:527–538. munity organization of filter-feeding dabbling ducks from
Griffin, D. R. 1981. The Question of Animal Awareness. 2nd ed. functional morphology. American Naturalist 169:334–343.
New York: Rockefeller University Press. Gurevitch, J., and L. V. Hedges 1999. Statistical issues in eco-
Griffin, D. R. 1982. Animal Mind—Human Mind. Berlin: logical meta-analyses. Ecology 80:1142–1149.
Springer-Verlag. Gurney, M. E. 1981. Hormonal control of cell form and num-
Griffin, D. R. 1984. Animal Thinking. Cambridge, MA: ber in zebra finch song system. Journal of Neuroscience
Harvard University Press. 1:658–673.
 References 479

Gurney, M. E., and M. Konishi. 1980. Hormone induced sex- Haskell, D. 1994. Experimental evidence that nestling begging
ual differentiation in brain and behavior in zebra finches. behavior incurs a costs due to nest predation. Proceedings of
Science 208:1380–1382. the Royal Society of London B 257:161–164.
Gwinner, E. 1996. Circadian and circannual programmes in Hasler, A. D., and W. J. Wisby. 1951. Discrimination of
avian migration. Journal of Experimental Biology 199:39–48. stream odors by fishes in relation to parent stream behav-
Gwinner, E., and W. Wiltschko. 1978. Endogenously con- ior. American Naturalist 85:223–238.
trolled changes in migratory direction of the garden war- Hasselquist, D. 1998. Polygyny in great reed warblers: A long
bler, Sylvia borin. Journal of Comparative Physiology term study of factors contributing to male fitness. Ecology
125:267–273. 79:2376–2390.
Gwynne, D. T., and W. D. Brown. 1994. Mate feeding, off- Hasselquist, D., S. Bensch, and T. von Schantz. 1996.
spring investment, and sexual differences in katydids Correlation between male song repertoire, extra-pair pater-
(Orthoptera: Tettigoniidae). Behavioral Ecology 5:267–272. nity and offspring survival in the great reed warbler. Nature
Gwynne, D. T., and D. C. F. Rentz. 1983. Beetles on the bot- 381:229–232.
tle: male buprestids mistake stubbies for females Hastings, M. H., A. B. Reddy, and E. S. Maywood. 2003. A
(Coleoptera). Journal of the Australian Entomological Society clockwork web: Circadian timing in brain and periphery, in
22:79–80. health and disease. Nature Reviews Neuroscience 4:649–651.
Gwynne, D. T., and L. W. Simmons. 1990. Experimental Hauber, M. E., and R. M. Kilner. 2007. Coevolution, com-
reversal of courtship roles in an insect. Nature 346:172–174. munication, and host-chick mimicry in parasitic finches:
Hagedorn, M. 1986. The ecology, courtship and mating of who mimics whom? Behavioral Ecology and Sociobiology
gymnotiform electric fish. In Electroreception, edited by T. H. 61:497–503.
Bullock and W. Heiligenberg. New York: John Wiley, pp. Hauber, M. E., and E. A. Lacey. 2005. Bateman’s principle in
497–525. cooperatively breeding vertebrates: the effects of non-
breeding alloparents on variability in female and male
Hagedorn, M. 1995. The electric fish Hypopomus occidentalis
reproductive success. Integrative and Comparative Biology
can rapidly modulate the amplitude and duration of its elec-
45:903–914.
tric organ discharges. Animal Behaviour 49:1409–1413.
Hauser, M. D. 1998. Functional referents and acoustic simi-
Hailman, J. P. 1967. Cliff-nesting adaptations of the
larity: Field playback experiments with rhesus monkeys.
Galapagos swallow-tailed gull. Wilson Bulletin 77:346–362.
Animal Behaviour 55:1647–1658.
Hall, J. C. 1994. The mating of the fly. Science 264:1702–1714.
Hausfater, G., and S. B. Hrdy. 1984. Infanticide: Comparative
Halpern, M., and A. Martinez-Marcos. 2003. Structure and
and Evolutionary Perspectives. New York: Aldine.
function of the vomeronasal system: an update. Progress in
Hausfater, G., H. C. Gerhardt, and G. M. Klump. 1990.
Neurobiology 70:245–318.
Parasites and mate choice in gray treefrogs, Hyla versicolor.
Halpin, Z. T. 1991. Introduction to the symposium: Animal
American Zoologist 30:299–312.
behavior: Past, present, and future. American Zoologist
Hausheer-Zarmakupi, Z., D. P. Wolfer, M.-C. Leisinger-
31:283–285.
Trigona, and H.-P. Lipp. 1996. Selective breeding for
Hamilton, W. D. 1964. The genetical evolution of social extremes in open-field activity of mice entails a differenti-
behaviour. Journal of Theoretical Biology 7:1–52. ation of hippocampal mossy fibers. Behavior Genetics
Hamilton, W. D. 1971. Geometry for the selfish herd. Journal 26:167–176.
of Theoretical Biology 31:295–311. Hayakawa, Y. 2007. Parasperm: morphological and functional
Hamilton, W. D., and M. Zuk. 1982. Heritable true fitness studies on nonfertile sperm. Ichthyological Research
and bright birds: A role for parasites? Science 218:384–387. 54:111–130.
Hare, B., J. Call, and M. Tomassello. 2000. Chimpanzees Hayden, D., A. Jennings, C. Müller, D. Pascoe, R. Bublitz, H.
know what conspecifics do and do not see. Animal Behaviour Webb, T. Breithaupt, L. Watkins, and J. Hardege. 2007.
59:771–785. Sex-specific mediation of foraging in the shore crab,
Hare, B., E. Addessi, J. Call, M. Tomasello, and E. Carcinus maenas. Hormones and Behavior 52:162–168.
Visalberghi. 2003. Do capuchin monkeys, Cebus apella, know Hayes, K. J., and C. Hayes. 1951. The intellectual develop-
what conspecifics do and do not see? Animal Behaviour ment of a home-raised chimpanzee. Proceedings of the
65:131–142. American Philosophical Society 95:105.
Hare, J. F., S. G. Sealy, T. J. Underwood, K. S. Ellison, and Hazelett, D. J., and J. C. Weeks. 2005. Segment-specific mus-
R. L. M. Stewart. 2003. Evidence of self-referent phenotype cle degeneration is triggered directly by a steroid hormone
matching revisited: airing out the armpit effect. Animal during insect metamorphosis. Journal of Neurobiology
Cognition 6:65–68. 62:164–177.
Hart, B. 1980. Neonatal spinal transection in male rats: Hebets, E. A., and D. R. Papaj. 2005. Complex signal func-
Differential effects on penile reflexes and other reflexes. tion: developing a framework of testable hypotheses.
Brain Research 185:423–428. Behavioral Ecology and Sociobiology 57:197–214.
480 References

Hedenström, A., and T. Alerstam. 1997. Optimum fuel loads Heyers, D., M. Manns, H. Luksch, O. Güntürkün, and H.
in migratory birds: Distinguishing between time and Moritsen. 2007. A visual pathway links brain structures
energy utilization. Journal of Theoretical Biology active during magnetic compass orientation in migratory
189:227–234. birds. PLoS ONE 2:e937.
Hedrick, A. V., and S. E. Riechert. 1989. Population variation Hill, G. E. 1990. Female house finches prefer colourful mates:
in the foraging behaviour of a spider: the role of genetics. sexual selection for a condition-dependent trait. Animal
Oecologia 80:533–539. Behaviour 40:563–572.
Hedwig, B., and R. Heinrich. 1997. Identified descending Hill, G. E. 1991. Plumage coloration is a sexually selected
brain neurons control different stridulatory motor patterns indicator of male quality. Nature 350:337–339.
in an acridid grasshopper. Journal of Comparative Physiology Hill, G. E., and R. Montgomerie. 1994. Plumage color sig-
A 180:285–294. nals nutritional condition in the house finch. Proceedings of
Heinrich, B. 1995. An experimental investigation of insight in the Royal Society of London B 258:47–52.
common ravens (Corvus corax). Auk 112:994–1003. Hill, P. S. M. 2001. Vibration and animal communication: A
Heinrich, B., and T. Bugnyar. 2005. Testing problem solving review. American Zoologist 41:1135–1142.
in ravens: string-pulling to reach food. Ethology Hinde, C. A., and R. M. Kilner. 2007. Negotiations within the
111:962–976. family over the supply of parental care. Proceedings of the
Heinroth, O. 1910. Beiträge zur Bilogie, insbiesonder Royal Society of London B 274:53–60.
Psychologie und Ethologie der Anatiden. In Proceedings of Hinde, R. A. 1970. Animal Behavior: A Synthesis of Ethology and
the 5th International Ornithological Congress, Berlin, pp. Comparative Psychology. 2nd ed. New York: McGraw-Hill.
589–702. Hinde, R. A. 1982. Ethology. Oxford: Oxford University Press.
Heinsohn, R., and S. Legge. 1999. The cost of helping. Trends Hirschenhauser, K., and R. F. Oliveira. 2006. Social modula-
in Ecology and Evolution 14:53–57. tion of androgens in male vertebrates: meta-analyses of the
Helbig, A. J. 1991. Inheritance of migratory direction in a bird challenge hypothesis. Animal Behaviour 71:265–277.
species: A cross-breeding experiment with SE- and SW- Hoage, R. J., and L. Goldman (eds.) 1986. Animal Intelligence,
migrating blackcaps (Sylvia atricapilla). Behavioral Ecology and Insights into the Animal Mind. Washington, DC: Smithsonian
Sociobiology 28:9–12. Institution Press.
Helbig, A. J., P. Berthold, and W. Wiltschko. 1989. Migratory Hoar, W. S. 1988. The physiology of smolting salmonids. In:
orientation of blackcaps (Sylvia atricapilla): Population spe- Fish Physiology, vol. XIB, edited by W. S. Hoar and D.
cific shifts of direction during the autumn. Ethology Randall. New York: Academic Press, pp. 275–343.
82:307–315. Höbel, G., and H. C. Gerhardt. 2003. Reproductive charac-
Hennessy, D. F. 1986. On the deadly risk of predator harass- ter displacement in the acoustic communication system of
ment. Ethology 72:72–74. green tree frogs (Hyla cinerea). Evolution 57:894–904.
Herman, L. M., and P. H. Forestell. 1985. Reporting presence Hochner, B., M. Klein, S. Schacher, and E. R. Kandel. 1986.
or absence of named objects by a language-trained dolphin. Additional component in the cellular mechanism of presy-
Neuroscience Biobehavioral Reviews 9:667–681. naptic facilitation contributes to behavioral dishabituation
Herman, L. M., A. A. Pack, and P. Morrel-Samuels. 1993. in Aplysia. Proceedings of the National Academy of Sciences
Representational and conceptual skills of dolphins. In 83:8794–8798.
Language and Communication: Comparative Perspectives, Hodos, W. C., and C. B. G. Campbell. 1969. Scala naturae:
edited by H. L. Roitblat, L. M. Herman, and P. E. Why there is no theory in comparative psychology.
Nachtigall. Hillsdale, NJ: Erlbaum, pp. 403–442. Psychological Review 76:337–350.
Hertenstein, M. J., J. M. Verkamp, A. M. Kerestes, and R. M. Hoffmann, A. A., and Z. Cacoyianni. 1990. Territoriality in
Holmes. 2006a. The communicative functions of touch in Drosophila melanogaster as a conditional strategy. Animal
humans, nonhuman primates, and rats: A review and syn- Behaviour 40:526–537.
thesis of the empirical research. Genetic, Social, and General Hofmann, H. A. 2003. Functional genomics of neural and
Psychology Monographs 132:5–94. behavioral plasticity. Journal of Neurobiology 54:272–282.
Hertenstein, M. J., D. Keltner, B. App, B. Bulleit, and A. Hofmann, H. A. 2006. Gonadotropin-releasing hormone sig-
Jaskolka. 2006b. Touch communicates distinct emotions. naling in behavioral plasticity. Current Opinion in
Emotion 6:528–533. Neurobiology 16:343–350.
Herzog, E. D., and G. Tosini. 2001. The mammalian circa- Hoffmann, K. 1954. Versuche zu der im Richtungsfinden der
dian clockshop. Seminars in Cell and Developmental Biology Vögel enthaltenen Zeitschätzung. Zeitschrift für Tierpsychologie
12:295–303. 11:453–475.
Hews, D. K. 1988. Alarm response in larval western toads, Hoffmann, K. H., K. Dettner, and K.-H. Tomaschko. 2006.
Bufo boreas: Release of larval chemical by a natural predator Chemical signals in insects and other arthropods: from mol-
and its effect on predator capture efficiency. Animal ecular structure to physiological functions. Physiological and
Behaviour 36:125–133. Biochemical Zoology 79:344–356.
 References 481

Hogan, J. A. 2005. Causation: the study of behavioural mech- Holmgren, N. A., and M. Enquist. 1999. Dynamics of mim-
anisms. Animal Biology 55:323–341. icry evolution. Biological Journal of the Linnean Society
Hogan, J. A., and J. J. Bolhuis. 2005. The development of 66:145–158.
behaviour: trends since Tinbergen (1963). Animal Biology Holveck, M-J., and K. Riebel. 2007. Preferred songs predict
55:371–398. preferred males: consistency and repeatability of zebra finch
Hogg, J. T. 1988. Copulatory tactics in relation to sperm com- females across three test contexts. Animal Behaviour
petition in Rocky Mountain bighorn sheep. Behavioral 74:297–309.
Ecology and Sociobiology 22:49–59. Höner, O. P., B. Wachter, M. L. East, W. J. Streich, K.
Hogg, J. T., and S. H. Forbes. 1997. Mating in bighorn sheep: Wilhelm, T. Burke, and H. Hofer. 2007. Female mate-
frequent male reproduction via a high-risk “unconven- choice drives the evolution of male-biased dispersal in a
tional” tactic. Behavioral Ecology and Sociobiology 41:33–48. social mammal. Nature 448:798–802.
Holder, C. F. 1901. A curious means of defense. Scientific Honess, P. E., and C. M. Marin. 2006. Enrichment and
American 2:186–187. aggression in primates. Neuroscience and Biobehavioral
Reviews 30:413–436.
Holekamp, K. E., and P. W. Sherman. 1989. Why male
ground squirrels disperse. American Scientist 77:232–239. Honeycutt, R. L. 1992. Naked mole-rats. American Scientist
80:43–53.
Holland, B., and W. R. Rice. 1998. Chase-away sexual selection:
Honma, A., S. Oku, and T. Nishida. 2006. Adaptive signifi-
antagonistic seduction versus resistance. Evolution 52:1–7.
cance of death feigning posture as a specialized inducible
Holland, R. A., J. L. Kirschvink, T. G. Doak, and M. Wikelski.
defence against gape-limited predators. Proceedings of the
2008. Bats use magnetite to detect the earth’s magnetic field.
Royal Society of London B. 273:1631–1636.
PLoS ONE 3:e1676.
Hopcraft, J. G., A. R. E. Sinclair, and C. Packer. 2005.
Hölldobler, B., and E. O. Wilson. 1990. The Ants. Cambridge:
Planning for success: Serengeti lions seek prey accessibility
Belknap Press.
rather than abundance. Journal of Animal Ecology
Hollis, K. L. 1984. The biological function of Pavlovian 74:559–566.
conditioning: The best defense is a good offense. Journal
Hopkins, C. D. 1974. Electric communication in fish.
of Experimental Psychology–Animal Behavior Processes
American Scientist 62:426–437.
10:413–425.
Hopkins, C. D. 1986a. Behavior of Mormyridae. In
Hollis, K. L. 1990. The role of Pavlovian conditioning in ter- Electroreception, edited by T. H. Bullock and W.
ritorial aggression and reproduction. In Contemporary Issues Heiligenberg. New York: John Wiley, pp. 527–576.
in Comparative Psychology, edited by D. A. Dewsbury.
Hopkins, C. D. 1986b. Temporal structure of nonpropagated
Sunderland, MA: Sinauer, pp. 197–219. electric communication signals. Brain, Behavior and
Hollis, K. L. 1999. The role of learning in the aggressive and Evolution 28:43–59.
reproductive behavior of blue gouramis, Trichogaster tri- Hopkins, C. D. 1999. Design features for electric communi-
chopterus. Environmental Biology of Fishes 54:355–369. cation. Journal of Experimental Biology 202:1217–1228.
Hollis, K. L., M. J. Dumas, P. Singh, and P. Fackelman. 1995. Hopkins, P. M. 1982. Growth and regeneration patterns in the
Pavlovian conditioning of aggressive behavior in blue fiddler crab, Uca pugilator. Biological Bulletin 163:301–319.
gourami fish (Trichogaster trichopterus)—Winners become Horsmann, U., H. G. Heinzel, and G. Wendler. 1983. The
winners and losers stay losers. Journal of Comparative phasic influence of self–generated air current modulations
Psychology 109:123–133. on the locust flight motor. Journal of Comparative Physiology
Hollis, K. L., V. L. Pharr, M. J. Dumas, G. B. Britton, and 150:427–438.
J. Field. 1997. Classical conditioning provides paternity Houck, L. D., and N. L. Reagan. 1990. Male courtship
advantage for territorial male blue gouramis (Trichogaster pheromones increase female receptivity in a plethodontid
trichopterus). Journal of Comparative Psychology 111:219–225. salamander. Animal Behaviour 39:729–734.
Holman, L., and R. R. Snook. 2006. Spermicide, cryptic Houston, A. I. 1995. Parental effort and paternity. Animal
female choice and the evolution of sperm form and func- Behaviour 50:1635–1644.
tion. Journal of Evolutionary Biology 19:1660–1670. Houtman, A. M., and J. B. Falls. 1994. Negative assortative
Holmes, W. G. 1986a. Identification of paternal half-siblings mating in the white-throated sparrow, Zonotrichia albicollis:
by captive Belding’s ground squirrels. Animal Behaviour The role of mate choice and intra-sexual competition.
34:321–327. Animal Behaviour 48:377–383.
Holmes, W. G. 1986b. Kin recognition by phenotype match- How, M. J., J. M. Hemmi, J. Zeil, and R. Peters. 2008. Claw
ing in female Belding’s ground squirrels. Animal Behaviour waving display changes with receiver distance in fiddler
34:38–47. crabs, Uca perplexa. Animal Behaviour 75:1015–1022.
Holmes, W. G., and P. W. Sherman. 1982. The ontogeny of Hsu, Y., R. L. Earley, and L. L. Wolf. 2006. Modulation of
kin recognition in two species of ground squirrels. American aggressive behaviour by fighting experience: mechanisms
Zoologist 22:491–517. and contest outcomes. Biological Reviews 81:33–74.
482 References

Huang, Z. Y., and G. E. Robinson. 1992. Honeybee colony Huxley, J. 1923. Courtship activities in the red-throated diver
integration: Worker-worker interactions mediate hormon- Colymbus stellatus pontopp; together with a discussion on the
ally regulated plasticity in division of labor. Proceedings of the evolution of courtship in birds. Journal of the Linnean Society
National Academy of Sciences 89:11726–11729. of London 2:491–562.
Huber, S. K., and J. Podos. 2006. Beak morphology and song Hylton, R., and R. D. Godard. 2001. Song properties of
features covary in a population of Darwin’s finches (Geospiza Indigo Buntings in open and forested habitats. Wilson
fortis). Biological Journal of the Linnean Society 88:489–498. Bulletin 113:243–245.
Immelmann, K. 1963. Drought adaptations in Australian
Hudson, R., and H. Distel. 1995. On the nature and action of
desert birds. In Proceedings of the 13th International
the nipple-search pheromone: a review. In Chemical Signals
Ornithological Congress, 1962, pp. 649–657.
in Vertebrates, Volume 7, edited by R. L. Doty and D.
Immelmann, K. 1969. Über den Einfluss frühkinlicher
Müller-Schwarze. New York: Plenum, pp. 223–232.
Erfahrungen auf die geschlechtliche Objektfixierung bei
Hughes, K. A., L. Du, F. H. Rodd, and D. N. Reznick. 1999. Estrildiden. Zeitschrift für Tierpsychologie 26:677–691.
Familiarity leads to female mate preference for novel males Immelmann, K. 1972. The influence of early experience upon
in the guppy, Poecilia reticulata. Animal Behaviour the development of social behavior in estrildine finches. In
58:907–916. Proceedings of the 15th International Ornithological Congress,
Hughes, W. O. H., J. Eilenberg, and J. J. Boomsma. 2002. The Hague, 1970, pp. 316–338.
Trade-offs in group living: transmission and disease resis- Immelmann, K. 1980. Introduction to Ethology. New York:
tance in leaf-cutting ants. Proceedings of the Royal Society of Plenum.
London B 269:1811–1819. Immelmann, K., and S. J. Suomi. 1981. Sensitive phases in
Hughes, W. O. H., B. P. Oldroyd, M. Beekman, and F. L. W. development. In Behavioral Development, edited by K.
Ratnieks. 2008. Ancestral monogamy shows kin selection is Immelmann, W. Barlow, L. Petrinovich, and M. Main.
key to the evolution of eusociality. Science 320:1213–1216. Cambridge: Cambridge University Press, pp. 395–431.
Huk, T., and W. Winkel. 2006. Polygyny and its fitness con- Ingram, K. K., P. Oefner, and D. M. Gordon. 2005. Task-spe-
sequences for primary and secondary female pied flycatch- cific expression of the foraging gene in harvester ants.
ers. Proceedings of the Royal Society of London B Molecular Ecology 14:813–818.
273:1681–1688. Inouye, S. T., and H. Kawamura. 1979. Persistence of circa-
dian rhythmicity in mammalian hypothalamic “island” con-
Hunt, G. J., G. V. Amdam, D. Schlipalius, C. Emore, N.
taining the suprachiasmatic nucleus. Proceedings of the
Sardesai, C. E. Williams, O. Rueppell, E. Guzmán-Novoa,
National Academy of Sciences 76:5962–5966.
M. Arechavaleta-Velasco, S. B. C. Chandra, M. K. Fondrk,
Insel, T. R. 2006. From species differences to individual dif-
M. Beye, and R. E. Page, Jr. 2007. Behavioral genomics of
ferences. Molecular Psychiatry 11:424.
honeybee foraging and nest defense. Naturwissenschaften
Ioalè, P., M. Nozzolini, and F. Papi. 1990. Homing pigeons
94:247–267. do extract directional information from olfactory stimuli.
Hunt, G. R. 1996. Manufacture and use of hook-tools by New Behavioral Ecology and Sociobiology 26:301–305.
Caledonian crows. Nature 379:249–251. Isaac, J. L. 2005. Potential causes and life-history conse-
Hunt, G. R., M. C. Corballis, and R. D. Gray. 2006. Design quences of sexual size dimorphism in mammals. Mammal
complexity and strength of laterality are correlated in New Review 35:101–115.
Caledonian crows’ pandanus tool manufacture. Proceedings Iwasaki, M., A. Delago, H. Nishino, and H. Aounuma. 2006.
of the Royal Society of London B 273:1127–1133. Effects of previous experience on the agonistic behaviour of
Hunt, G. R., and R. D. Gray. 2004. The crafting of hook tools male crickets, Gryllus bimaculatus. Zoological Science
by wild New Caledonian crows. Proceedings of the Royal 23:863–872.
Society of London B 271:S88–S90. Jablonski, P. G., and R. S. Wilcox. 1996. Signaling asymme-
Hunter, M. L., and J. R. Krebs. 1979. Geographical variation try in the communication of the water strider Aquarius
remigis in the context of dominance and spacing in the non-
in the song of the great tit Parus major in relation to eco-
mating season. Ethology 102:353–359.
logical factors. Journal of Animal Ecology 48:759–785.
Jackson, R. R., and R. S. Wilcox. 1993a. Spider flexibly
Huntingford, F. A. 1986. Development of behaviour in fish.
chooses aggressive mimicry signals for different prey by trial
In The Behavior of Teleost Fishes, edited by T. J. Pitcher.
and error. Behaviour 127:21–36.
Baltimore, MD: Johns Hopkins University Press, pp.
Jackson, R. R., and R. S. Wilcox. 1993b. Observations in
47–68.
nature of detouring behavior by Portia fimbriata, a web-
Huntingford, F. A., and A. K. Turner. 1987. Animal Conflict. invading aggressive mimic jumping spider from
London: Chapman and Hall. Queensland. Journal of Zoology 230:135–139.
Hupé, G. J., and J. E. Lewis. 2008. Electrocommunication sig- Jacquot, J. J., and N. G. Solomon. 1997. Effects of site famil-
nals in free swimming brown ghost knifefish, Apteronotus iarity on movement patterns of male prairie voles Microtus
leptorhynchus. Journal of Experimental Biology 211:1657–1667. ochrogaster. American Midland Naturalist 138:414–417.
 References 483

Jan, Y., L. Jan, and M. Dennis. 1977. Two mutations of synap- Johnston, R. E. 2003. Chemical communication in rodents:
tic transmission in Drosophila. Proceedings of the Royal Society from pheromones to individual recognition. Journal of
of London B 198:87–108. Mammalogy 84:1141–1162.
Jarvis, E. D., H. Schwabl, S. Ribeiro, and C. V. Mello. 1997. Johnston, T. D. 1988. Developmental explanation and the
Brain gene regulation by territorial singing behavior in ontogeny of birdsong: Nature/nurture redux. Behavioral and
freely ranging songbirds. NeuroReport 8:2073–2077. Brain Sciences 11:617–663.
Jarvis, J. U. M. 1981. Eusociality in a mammal: Cooperative Johnstone, R. A. 1997. The evolution of animal signals. In
breeding in naked mole-rat colonies. Science 212:571–573. Behavioural Ecology, edited by J. R. Krebs and N. B. Davies.
Oxford: Blackwell Science, pp. 155–178.
Jaynes, J. 1969. The historical origins of “ethology” and com-
parative psychology. Animal Behaviour 17:601–606. Jones, C. B., V. Milanov, and R. Hager. 2008. Predictors of
male residence patterns in groups of black howler monkeys.
Jenni, D. A., and B. J. Betts. 1978. Sex differences in nest con-
Journal of Zoology 275:72–78.
struction, incubation, and parental behavior in the polyan-
Jones, T. M., and R. J. Quinnell. 2002. Testing predictions for
drous American jacana (Jacana spinosa). Animal Behaviour
the evolution of lekking in the sandfly, Lutzomyia longipalpis.
26:207–218.
Animal Behaviour 63:605–612.
Jenni, D. A., and G. Collier. 1972. Polyandry in the American
Jones, T. M., R. J. Quinnell, and A. Balmford. 1998. Fisherian
jacana (Jacana spinosa). Auk 89:743–765.
flies: benefits of female choice in a lekking sandfly.
Jennings, H. S. 1906. Behavior of the Lower Organisms. New Proceedings of the Royal Society of London B 265:1651–1657.
York: Columbia University Press. Joron, M., and J. L. B. Mallet. 1998. Diversity in mimicry:
Jennions, M. D. 1993. Female choice in birds and the cost of Paradox or paradigm. Trends in Ecology and Evolution
long tails. Trends in Ecology and Evolution 8:230–232. 13:461–466.
Jennions, M. D., and M. Petrie. 2000. Why do females mate Kacelnik, A., and M. Bateson. 1996. Risky theories—the effect
multiply? A review of the genetic benefits. Biological Reviews of variance on foraging decisions. American Zoologist
75:21–64. 36:402–434.
Jeschke, J. M., and R. Tollrian. 2007. Prey swarming: which Kalmijn, A. J. 1966. Electro-perception in sharks and rays.
predators become confused and why? Animal Behaviour Nature 212:1232–1233.
74:387–393. Kalmijn, A. J. 1971. The electric sense of sharks and rays.
Jivoff, P. 1997. The relative roles of predation and sperm com- Journal of Experimental Biology 55:371–383.
petition on the duration of the post-copulatory association Kamil, A. C., and J. E. Mauldin. 1988. A comparative-eco-
between the sexes in the blue crab, Callinectes sapidus. logical approach to the study of learning. In Evolution and
Behavioral Ecology and Sociobiology 40:175–186. Learning, edited by R. C. Bolles and M. D. Beecher.
Johnsen, S. 2001. Hidden in plain sight: the ecology and phys- Hillsdale, NJ: Erlbaum, pp. 117–133.
iology of organismal transparency. Biological Bulletin Kamil, A. C., and S. I. Yoerg. 1982. Learning and foraging
201:301–318. behavior. In Perspectives on Ethology, edited by P. P. G.
Johnsgard, P. A. 1967. Animal Behavior. William C. Brown: Bateson. New York: Plenum, pp. 325–364.
Dubuque. Kamil, A. C., R. P. Balda, and D. J. Olson. 1994. Performance
of four seed-caching corvid species in the radial-arm maze
Johnson, C. H., J. A. Elliott, and R. Foster. 2003. Entrainment
analog. Journal of Comparative Psychology 108:385–393.
of circadian programs. Chronobiology International
Kamio, M., M. A. Reidenberg, and C. Derby. 2008. To pad-
20:741–774.
dle or not: context dependent courtship display by male blue
Johnson, M. H. 2005. Sensitive periods in functional brain
crabs (Callinectes sapidus). Journal of Experimental Biology
development: problems and prospects. Developmental 211:1234–1248.
Psychobiology 46:287–292.
Kanda, L. L. 2005. Winter energetics of Virginia opossums
Johnson, M. L., and M. S. Gaines. 1990. Evolution of dis- Didelphis virginiana and implications for the species’ north-
persal: Theoretical models and empirical tests using birds ern distributional limit. Ecography 28:731–744.
and mammals. Annual Review of Ecology and Systematics Kanda, L. L., T. K. Fuller, and P. R. Sievert. 2006. Landscape
21:449–480. associations of road-killed Virginia opossums (Didelphis vir-
Johnston, R. E. 1998. Pheromones, the vomeronasal system, giniana) in central Massachusetts. American Midland
and communication—from hormonal responses to individ- Naturalist 156: 128–134.
ual recognition. Annals of the New York Academy of Sciences Kandel, E. R. 1976. Cellular Basis of Behavior: An Introduction
855:333–348. to Behavioral Neurobiology. San Francisco, CA: W. H.
Johnston, R. E. 2000. Chemical communication and Freeman.
pheromones: the types of chemical signals and the role of Kandel, E. R. 1979a. The Behavioral Biology of Aplysia. San
the vomeronasal system. In The Neurobiology of Taste and Francisco: W. H. Freeman.
Smell, edited by T. E. Finger, W. L. Silver, and D. Restrepo. Kandel, E. R. 1979b. Small systems of neurons. Scientific
New York: Wiley-Liss, pp. 101–127. American 241:66–76.
484 References

Kandel, E. R. 2001. The molecular biology of memory stor- Kiepenheuer, J. 1979. Pigeon homing: Deprivation of olfac-
age: A dialogue between genes and synapses. Science tory information does not affect the deflector effect.
294:1030–1038. Behavioral Ecology and Sociobiology 6:11–22.
Kandel, E. R., and J. H. Schwartz. 1982. Molecular biology Kilner, R. M. 1995. When do canary parents respond to
of learning: Modulation of transmitter release. Science nestling signals of need. Proceedings of the Royal Society of
218:433–443. London B 260:343–348.
Kaplan, A., and W. E. Trout III. 1969. The behavior of four Kilner, R. M. 2003. How selfish is a cowbird nestling? Animal
neurological mutants of Drosophila. Genetics 61:399–409. Behaviour 66:569–576.
Kaufmann, J. H. 1983. On definitions and functions of dom- Kilner, R. M., J. R. Madden, and M. E. Hauber. 2004. Brood
inance and territoriality. Biological Reviews 58:1–20. parasitic cowbird nestlings use host young to procure
Kawai, M. 1965. Newly acquired and pre-cultural behavior of resources. Science 305:877–879.
the natural troop of Japanese monkeys on Koshima Islet. Kiltie, R. A. 1988. Countershading: Universally deceptive or
Primates 6:1–30. deceptively universal? Trends in Ecology and Evolution
Kawamura, S. 1959. Sub-culture propagation among Japanese 3:21–23.
macaques. Primates 2:43–60. Kiltie, R. A. 1989. Wildfire and the evolution of dorsal
Keane, B., P. M. Waser, S. R. Creel, N. M. Creel, L. F. Elliott, melanism in fox squirrels, Sciurus niger. Journal of
and D. J. Minchella. 1994. Subordinate reproduction in Mammalogy 70:726–739.
dwarf mongooses. Animal Behaviour 47:65–75. Kimura, K.-I., M. Ote, T. Tazawa, and D. Yamamoto. 2005.
Fruitless specifies sexually dimorphic neural circuitry in the
Keefer, M. L., C. C. Caudill, C.C. Peery, and C.T. Boggs.
Drosophila brain. Nature 438:229–233.
2008. Non-direct homing behaviours by adult Chinook
salmon in a large multi-stock river system. Journal of Fish King, A. P., and M. J. West. 1983. Epigenesis of cowbird
song—A joint endeavor of males and females. Nature
Biology 72:27–44.
305:704–706.
Keeton, W. T., T. S. Larkin, and D. M. Windsor. 1974.
Kirkpatrick, M. 1982. Sexual selection and the evolution of
Normal fluctuations in the earth’s magnetic field influence
female choice. Evolution 36:1–12.
pigeon orientation. Journal of Comparative Physiology
95:95–103. Kirschvink, J. L., S. Padmanabha, C. K. Boyce, and J. Oglesby.
1997. Measurement of the threshold of honeybees to weak,
Keller, F. S. 1941. Light aversion in the white rat. Psychological
extremely low-frequency magnetic fields. Journal of
Record 4:235–250.
Experimental Biology 200:1363–1368.
Keller, L., and K. G. Ross. 1998. Selfish genes: A green beard
Klaassen, M. 1996. Metabolic constraints on long-distance
in the red fire ant. Nature 394:573–575.
migration in birds. Journal of Experimental Biology
Kelley, D. B. 1996. Sexual differentiation in Xenopus laevis. In 199:57–64.
The Biology of Xenopus, edited by R. Tinsley and H. Kobel.
Kleiman, D. G. 1977. Monogamy in mammals. Quarterly
Oxford: Oxford University Press, pp. 143–176.
Review of Biology 52:39–69.
Kelley, D. B., and D. I. Gorlick. 1990. Sexual selection and
Kleiman, D. G., and J. R. Malcolm. 1981. The evolution of
the nervous system. BioScience 40:275–283. male parental investment in mammals. In Parental Care in
Kemp, D. J., and C. Wiklund. 2004. Residency effects in ani- Mammals, edited by D. J. Gubernick and P. H. Klopfer.
mal contests. Proceedings of the Royal Society of London B New York: Plenum, pp. 347–387.
271:1707–1711. Knaden, M., and R. Wehner. 2006. Ant navigation: resetting
Kendrick, K. M., A. P. da Costa, A. E. Leigh, M. R. Hinton, the path integrator. Journal of Experimental Biology
and J. W. Peirce. 2007. Sheep don’t forget a face. Nature 209:26–31.
447:346. Knapp, R., D. K. Hews, C. W. Thompson, L. Ray, and M. C.
Kendrick, K. M., M. A. Haupt, M. R. Hinton, K. D. Broad, Moore. 2003. Environmental and endocrine correlates of
and J. D. Skinner. 2001. Sex differences in the influence of tactic switching by non-territorial male tree lizards
mothers on the sociosexual preferences of their offspring. (Urosaurus ornatus). Hormones and Behavior 43:83–92.
Hormones and Behavior 40:322–338. Knudsen, E. I. 1981. The hearing of the barn owl. Scientific
Kendrick, K. M., M. R. Hinton, K. Atkins, M. A. Haupt, K. American 245:113–125.
D. Broad, and J. D. Skinner. 1998. Mothers determine sex- Knudsen. E. I. 1982. Auditory and visual maps of space in the
ual preferences. Nature 395:229–230. optic tectum of the owl. Journal of Neuroscience 2:1177–1194.
Kent, D. S., and J. A. Simpson. 1992. Eusociality in the bee- Knudsen, E. I. 2002. Instructed learning in the auditory local-
tle Austroplatypus incompertus (Coleoptera: Curculionidae). ization pathway of the barn owl. Nature 417:322–328.
Naturwissenschaften 79:86–87. Knudsen, E. I., and M. Konishi. 1978. A neural map of audi-
Kessel, E. L. 1955. Mating activities of balloon flies. Systematic tory space in the owl. Science 200:795–797.
Zoology 4:97–104. Koenig, W. D., and R. L. Mumme. 1987. Population Ecology of
Kiepenheuer, J. 1978. A repetition of the deflector loft exper- the Cooperatively Breeding Acorn Woodpecker. Princeton, NJ:
iment. Behavioral Ecology and Sociobiology 3:393–395. Princeton University Press.
 References 485

Köhler, W. 1927. The Mentality of Apes. New York: Harcourt Kramer, G. 1949. Über Richtungstendenzen bei der
Brace. nächtichen Zugenruhe gekäfigter Vögel. In Ornithologie als
Kokko, H., R. Brooks, M. D. Jennions, and J. Morley. 2003. biologische Wissenschaft, edited by E. Mayr and E. Schüz.
The evolution of mate choice and mating biases. Proceedings Heidelberg: Carl Winter.
of the Royal Society of London B 270:653–664. Kramer, G. 1950. Weitere Analyse der Faktoren, welche die
Koltermann, R. 1971. 24-std-Periodik in der Langzeiterinnerung Zugaktivität des gekäfigten Vogels orientieren.
an Duft-und Farbsignale ber der Honigbiene. Zeitschrift für Naturwissenschaften 37:377–378.
vergleichende Physiologie 75:49–68. Kramer, G. 1951. Eine neue Methode zur Erforschung der
Komdeur, J. 1992. Importance of habitat saturation and ter- Zugorientierung und die bisher damit erzielten Ergebnisse.
ritory quality for evolution of cooperative breeding in the Proceedings of the 10th International Ornithology Congress:
Seychelles warbler. Nature 358:493–495. 269–280.
Komdeur, J. 1996. Influence of helping and breeding experi- Krause, J. 1993. The effect of “Schreckstoff” on the shoaling
ence on reproductive performance in the Seychelles war- behavior of the minnow: A test of Hamilton’s selfish herd
bler: a translocation experiment. Behavioral Ecology theory. Animal Behaviour 45:1019–1024.
7:326–333. Krause, J., and G. D. Ruxton. 2002. Living in Groups. Oxford:
Komdeur, J., and P. Edelaar. 2001. Male Seychelles warblers Oxford University Press.
use territory budding to maximize lifetime fitness in a sat- Krebs, J. R., and M. I. Avery. 1984. Chick growth and prey
urated environment. Behavioral Ecology 12:706–715. quality in the European bee-eater (Merops apiaster). Oecologia
Komdeur, J., T. Burke, and D. S. Richardson. 2007. Explicit 64:363–368.
experimental evidence for the effectiveness of proximity as Krebs, J. R., and N. B. Davies. 1997. Behavioural Ecology: An
mate-guarding behaviour in reducing extra-pair fertilization Evolutionary Approach. 4th ed. Oxford: Blackwell Science.
in the Seychelles warbler. Molecular Ecology 16:3679–3688. Krebs, J. R., and R. McCleery. 1984. Optimization in behav-
Komdeur, J., A. Huffstadt, W. Prast, G. Castle, R. Mileto, and ioural ecology. In Behavioural Ecology, An Evolutionary
J. Wattel. 1995. Transfer experiments of Seychelles warblers Approach, 2nd ed., edited by J. R. Krebs and N. B. Davies.
to new islands—changes in dispersal and helping behavior. Sunderland, MA: Sinauer.
Animal Behaviour 49:695–708. Krebs, J. R., J. T. Erichsen, M. J. Webber, and E. L. Charnov.
Komers, P. E. 1997. Property rites. Natural History 106:28–31. 1977. Optimal prey selection in the great tit (Parus major).
Animal Behaviour 25:30–38.
Komers, P. E., and P. N. M. Brotherton. 1997. Female space
Krebs, J. R., D. F. Sherry, S. D. Healy, V. H. Perry, and A. L.
use is the best predictor of monogamy in mammals.
Vaccarino. 1989. Hippocampal specialization of food stor-
Proceedings of the Royal Society of London B 264:1261–1270.
ing birds. Proceedings of the National Academy of Sciences
Konishi, M. 1965. The role of auditory feedback in the con-
86:1388–1392.
trol of vocalization in the white-crowned sparrow. Zeitschrift
Krebs, J. R., N. S. Clayton, S. D. Healy, D. A. Cristol, S. N.
für Tierpsychologie 22:770–783.
Patel, and A. R. Jolliffe. 1996. The ecology of the avian
Konishi, M. 1993a. Listening with two ears. Scientific American
brain: Food-storing memory and the hippocampus. Ibis
286:66–73.
138:34–46.
Konishi, M. 1993b. The neuroethology of sound localization
Krohmer, R. W., and D. Crews. 1987. Temperature activation
in the owl. Journal of Comparative Physiology A 173:3–7.
of courtship behavior in the male red-sided garter snake
Konishi, M. 2003. Coding of auditory space. Annual Review (Thamnophis sirtalis parietalis): Role of the anterior hypo-
of Neuroscience 26:31–55. thalamus-preoptic area. Behavioral Neuroscience
Konishi, M. 2004. The role of auditory feedback in birdsong. 101:228–236.
Annals of the New York Academy of Sciences 1016:463–475. Kroodsma, D. E. 2005. The Singing Life of Birds: The Art and
Kowalski, U., R. Wiltschko, and E. Fuller. 1988. Normal fluc- Science of Listening to Birdsong. New York: Houghton
tuations of the geomagnetic field may affect initial orienta- Mifflin.
tion in pigeons. Journal of Comparative Physiology A Kroodsma, D. E., and R. Pickert. 1980. Environmentally
163:593–600. dependent sensitive periods for avian vocal learning. Nature
Krakauer, A. H. 2005. Kin selection and cooperative courtship 288:477–479.
in wild turkeys. Nature 434:69–72. Kroodsma, D. E., P. W. Houlihan, P. A. Fallon, and J. A.
Krama, T., and I. Krams. 2005. Cost of mobbing call to breed- Wells. 1997. Song development by grey catbirds. Animal
ing pied flycatcher, Ficedula hypoleuca. Behavioral Ecology Behaviour 54:457–464.
16:37–40. Kuba, M. J., Byrne, R. A., D. V. Meisel, and J. A. Mather.
Kramer, D. L., and D. M. Weary. 1991. Exploration versus 2006. When do octopuses play? Effects of repeated testing,
exploitation: A field study of time allocation to environ- object type, age, and food deprivation on object play in
mental tracking by foraging chipmunks. Animal Behaviour Octopus vulgaris. Journal of Comparative Psychology
41:443–449. 120:184–190.
486 References

Kummer, H., W. Gotz, and W. Angst. 1974. Triadic differ- Langley, C. M., D. A. Riley, A. B. Bond, and N. Goel. 1996.
entiation: An inhibitory process protecting pair bonds in Visual search for natural grains in pigeons (Columba livia):
baboons. Behaviour 49:62–87. Search images and selective attention. Journal of Experimental
Kunz, T. H., E. B. Arnett, B. M. Cooper, W. P. Erickson, R. Psychology—Animal Behavior Processes 22:139–151.
P. Larkin, T. Mabee, M. L. Morrison, M. D. Strickland, and Laposky, A. D., J. Bass, A. Kohsaka, and F. W. Turek. 2008.
J. M. Szewczak. 2007. Assessing impacts of wind-energy Sleep and circadian rhythms: Key components in the reg-
development on nocturnally active birds and bats: a guid- ulation of energy metabolism. FEBS Letters 582:142–151.
ance document. Journal of Wildlife Management Lashley, K. 1950. In search of the engram. Symposia of the
71:2449–2486. Society for Experimental Biology IV:454–482.
Kupfermann, I., V. Castellucci, H. Pinsker, and E. Kandel. Laundre, J. W., Hernandez, L., and K. B. Altendorf. 2001.
1970. Neuronal correlates of habituation and dishabituation Wolves, elk, and bison: reestablishing the “landscape of
of the gill withdrawal reflex in Aplysia. Science fear” in Yellowstone National Park. Canadian Journal of
167:1743–1745. Zoology 79:1401–1409.
Kuvlesky, W. P., L. A. Brennan, M. L. Morrison, K. K. Laurien-Kehnen, C., and F. Trillmich. 2003. Lactation per-
Boydston, B. M. Ballard, and F. C. Bryant. 2007. Wind formance of guinea pigs (Cavia porcellus) does not respond
energy development and wildlife conservation: challenges to experimental manipulation of pup demands. Behavioral
and opportunities. Journal of Wildlife Management Ecology and Sociobiology 53:145–152.
71:2487–2498. Lawson Handley, L. J., and N. Perrin. 2007. Advances in our
LaBas, N. R., and L. R. Hockham. 2005. An invasion of understanding of mammalian sex-biased dispersal. Molecular
cheats: The evolution of worthless nuptial gifts. Current Ecology 16:1559–1578.
Biology 15:64–67. Le Galliard, J.-F., R. Ferrière, and J. Clobert. 2003. Mother-
Lacey, E. A., and P. W. Sherman. 1991. Social organization of offspring interactions affect natal dispersal in a lizard.
naked mole-rat colonies: Evidence of division of labor. In Proceedings of the Royal Society of London B 270:1163–1169.
The Biology of the Naked Mole-Rat, edited by P. W. Sherman, Le Galliard, J-F., G. Gundersen, H. P., Andreassen, and N. C.
J. U. M. Jarvis, and R. D. Alexander. Princeton, NJ: Stenseth. 2006. Natal dispersal, interactions among siblings,
Princeton University Press, pp. 275–336. and intrasexual competition. Behavioral Ecology 17:733–740.
Lacey, E. A., and P. W. Sherman. 1997. Cooperative breed- Leadbeater, E., and L. Chittka. 2007. Social learning in
ing in naked mole-rats: implications for vertebrate and insects—from miniature brains to consensus building.
invertebrate sociality. In Cooperative Breeding in Mammals, Current Biology 17:R703–R713.
edited by N. G. Solomon and J. A. French. Cambridge: LeBas, N. R., L. R. Hockam, and M. G. Ritchie. 2004. Sexual
Cambridge University Press, pp. 267–301. selection in the gift-giving dance fly, Rhamphomyia sulcata,
Lack, D. 1939. The display of the black cock. British Birds favors small males carrying small gifts. Evolution
32:290–303. 58:1763–1772.
Lack, D. 1943. The Life of the Robin. London: Penguin Books. LeBoeuf, B. J. 1974. Male-male competition and reproductive
Lack, D. 1968. Bird migration and natural selection. Oikos success in elephant seals. American Zoologist 14:163–176.
19:1–9. Lee, J. S. F., and A. H. Bass. 2004. Does exaggerated mor-
Lacy, R. C., and P. W. Sherman. 1983. Kin recognition by phology preclude plasticity to cuckoldry in the midshipman
phenotype matching. American Naturalist 121:489–512. fish (Porichthys notatus)? Naturwissenschaften 91:338–341.
Laidlaw, H. H., Jr., and R. E. Page, Jr. 1984. Polyandry in Lefebvre, L. 1995. Culturally-transmitted feeding behaviour
honey bees (Apis mellifera L.) — Sperm utilization and intra- in primates: Evidence for accelerating learning rates.
colony genetic relationships. Genetics 108:985–997. Primates 36:227–239.
Laland, K. N., and K. Williams. 1997. Shoaling generates Lehrman, D. S. 1953. A critique of Konrad Lorenz’s theory
social learning of foraging information in guppies. Animal of instinctive behavior. Quarterly Review of Biology
Behaviour 53:1149–1159. 28:337–363.
Lande, R. 1981. Models of speciation by sexual selection on Leibrecht, B. C., and H. R. Askew. 1980. Habituation from a
polygenic traits. Proceedings of the National Academy of Sciences comparative perspective. In Comparative Psychology: An
78:3721–3725. Evolutionary Analysis of Animal Behavior, edited by M. R.
Langbauer, W. R., Jr. 2000. Elephant communication. Zoo Denny. New York: John Wiley, pp. 208–229.
Biology 19:425–445. Leinders-Zufall, T., P. Brennan, P. Widmayer, P.
Langkilde, T., R. A. Alford, and L. Schwarzkopf. 2005. No Chandramani, A. Maul-Pavicic, M. Jäger, X. Li, H. Breer,
behavioural compensation for fitness costs of autotomy in F. Zufall, and T. Boehm. 2004. MHC class I peptides as
a lizard. Austral Ecology 30:713–718. chemosensory signals in the vomeronasal organ. Science
Langley, C. M. 1996. Search images: selective attention to spe- 306:1033–1037.
cific visual features of prey. Journal of Experimental Psychology Leitner, S., J. Nicholson, B. Leisler, T. J. DeVoogd, and C. K.
22:152–163. Catchpole. 2002. Song and the song control pathway in the
 References 487

brain can develop independently of exposure to song in the Lima, S. L., and L. M. Dill. 1990. Behavioral decisions made
sedge warbler. Proceedings of the Royal Society of London B under the risk of predation—a review and prospectus.
269:2519–2524. Canadian Journal of Zoology 68:619–640.
Lenneberg, E. H. 1967. Biological Foundations of Language. Lima, S. L., and T. J. Valone. 1986. Influence of predation risk
New York: John Wiley. on diet selection: a simple example in the grey squirrel.
Lent, C. M., M. H. Dickinson, and C. G. Marshall. 1989. Animal Behaviour 34:536–544.
Serotonin and leech feeding behavior: Obligatory neuro- Lima, S. L., T. J. Valone, and T. Caraco. 1985. Foraging-effi-
modulation. American Zoologist 29:1241–1254. ciency—predation-risk trade-off in the grey squirrel. Animal
Leoncini, I., Y. Le Conte, G. Costagliola, E. Plettner, A. L. Behaviour 33:155–165.
Toth, M. Wang, Z. Huang, M. R. Band, J.-M. Bécard, Limongelli, L., S. T. Boysen, and E. Visalberghi. 1995.
D. Crauser, K. N. Slessor, and G. E. Robinson. 2004. Comprehension of cause-effect relations in a tool-using task
Regulation of behavioral maturation by a primer pheromone by chimpanzees (Pan troglodytes). Journal of Comparative
produced by adult worker honey bees. Proceedings of the Psychology 109:18–26.
National Academy of Sciences 101:17559–17564. Linden, M., and A. P. Møller. 1989. Cost of reproduction and
Levi, R., and J. M. Camhi. 1995. Distributing coordinated covariation of life history traits in birds. Trends in Ecology and
motor outputs to several body segments—Escape move- Evolution 4:367–371.
ments in the cockroach. Journal of Comparative Physiology A
Lindström, Å., D. Hasselquist, S. Bensch, and M. Grahn.
177:427–437. 1990. Asymmetric contests over resources for survival and
Levi, R., and J. M. Camhi. 2000. Population vector coding by migration: A field experiment with bluethroats. Animal
the giant interneurons of the cockroach. The Journal of Behaviour 40:453–461.
Neuroscience 20:3822–3829.
Lindström, L., R. V. Alatalo, and J. Mappes. 1997. Imperfect
Levin, L. R., P.-L. Han, P. M. Hwaung, P. G. Feinstein, R. L. Batesian mimicry—The effects of the frequency and the dis-
Davis, and R. R. Reed. 1992. The Drosophila learning and tastefulness of the model. Proceedings of the Royal Society of
memory gene rutabaga encodes a Ca+2/ calmodulin-adeny- London B 264:149–153.
lyl cyclase. Cell 68:479–489.
Linksvayer, T. A., and M. J. Wade. 2005. The evolutionary
Levitan, D. R. 2005. The distribution of male and female origin and elaboration of sociality in the aculeate
reproductive success in a broadcast spawning marine inver- Hymenoptera: maternal effects, sib-social effects, and het-
tebrate. Integrative and Comparative Biology 45:848–855. erochrony. Quarterly Review of Biology 80:317–336.
Lewis, T. L., and D. Maurer. 2005. Multiple sensitive periods
Linn, C. D., Y. Molina, J. Difatta, and T. E. Christenson.
in human visual development: evidence from visually 2007. The adaptive advantage of prolonged mating: a test
deprived children. Developmental Psychobiology 46:163–183. of alternative hypotheses. Animal Behaviour 74:481–485.
Lewis, S. M., C. K. Cratsley, and J. A. Rooney. 2004. Nuptial
Little, A. E. F., T. Murakami, U. G. Mueller, and C. R. Currie.
gifts and sexual selection in Photinus fireflies. Integrative and
2006. Defending against parasites: fungus-growing ants
Comparative Biology 44:234–237.
combine specialized behaviours and microbial symbionts to
Lickliter, R., and G. Gottlieb. 1988. Social specificity: protect their fungus gardens. Biology Letters 2:12–16.
Interaction with own species is necessary to foster species-
Lloyd, J. E. 1975. Aggressive mimicry in Photuris fireflies: sig-
specific maternal preferences in ducklings. Developmental
nal repertoires by femmes fatales. Science 187:452–453.
Psychobiology 21:311–321.
Lloyd, J. E. 1986. Firefly communication and deception: “oh,
Lieberman, D. A. 1993. Learning: Behavior and Cognition. 2nd
what a tangled web.” In Deception: Perspectives on Human and
ed. Pacific Grove, CA: Brooks/Cole Publishing Company.
Nonhuman Deceit, edited by R. W. Mitchell and N. S.
Light, P., M. Salmon, and K. J. Lohmann. 1993. Geomagnetic
Thompson. Albany, NY: SUNY Press, pp. 113–128.
orientation of loggerhead sea turtles: evidence for an incli-
Loeb, J. 1981. The Foundations of Ethology. New York: Springer
nation compass. Journal of Experimental Biology 182:1–10.
Verlag.
Lim, M. M., and L. J. Young. 2006. Neuropeptidergic regu-
lation of affiliative behavior and social bonding in animals. Loeb, J. 1918. Forced Movements, Tropisms, and Animal Conduct.
Hormones and Behavior 50:506–517. Philadelphia: J. B. Lippincott.
Lim, M. M., I. F. Bielsky, and L. J. Young. 2005. Lohmann, K. J. 1991. Magnetic orientation by hatchling
Neuropeptides and the social brain: potential rodent mod- loggerhead sea turtles (Caretta caretta). Journal of
els of autism. International Journal of Developmental Experimental Biology 155:37–49.
Neuroscience 23:235–243. Lohmann, K. J. 1992. How sea turtles navigate. Scientific
Lim, M. M., Z. Wang, D. E. Olazábal, X. Ren, E. F. American 266:100–106.
Terwilliger, and L. J. Young. 2004. Enhanced partner pref- Lohmann, K. J., and C. M. F. Lohmann. 1992. Orientation to
erence in a promiscuous species by manipulating the expres- oceanic waves by green turtle hatchlings. Journal of
sion of a single gene. Nature 429:754–757. Experimental Biology 171:1–13.
488 References

Lohmann, K. J., and C. M. F. Lohmann. 1996a. Detection of Lyons, C., and C. J. Barnard. 2006. A learned response to
magnetic field intensity by sea turtles. Nature 380:59–61. sperm competition in the field cricket, Gryllus bimaculatus
Lohmann, K. J., and C. M. F. Lohmann. 1996b. Orientation (de Geer). Animal Behaviour 72:673–680.
and open-sea navigation in sea turtles. Journal of Lythgoe, J. N., W. R. A. Muntz, J. C. Partridge, J. Shand, and
Experimental Biology 199:73–81. D. McB. Williams. 1994. The ecology of visual pigments
Lohmann, K. J., and C. M. F. Lohmann. 2006. Sea turtles, lob- of snappers (Lutjanidae) on the Great Barrier Reef. Journal
sters, and magnetic maps. Marine Freshwater Behaviour and of Comparative Physiology A 174:255–260.
Physiology 39:49–64. Macdonald, D. W., and D. D. P. Johnson. 2001. Dispersal in
Lohmann, K. J. , C. M. F. Lohmann, and C. S. Endres. 2008. theory and practice: consequences for conservation biology.
The sensory ecology of ocean navigation. Journal of In Dispersal, edited by J. Clobert, E. Danhin, A. A. Dhondt,
Experimental Biology 211:1719–1728. and J. D. Nichols. Oxford: Oxford University Press, pp.
Lohmann, K. J., C. M. F. Lohmann, and N. F. Putman. 2007. 358–372.
Magnetic maps in animals: nature’s GPS. Journal of Macphail, E. M., and J. J. Bolhuis. 2001. The evolution of
Experimental Biology 210:3697–3705. intelligence: adaptive specializations versus general process.
Lohmann, K. J., N. F. Putnam, and C. M. F. Lohmann. 2008. Biological Reviews 76:341–364.
Geomagnetic imprinting: A unifying hypothesis of long-dis- MacRoberts, M. H., and B. R. MacRoberts. 1976. Social orga-
tance natal homing in salmon and sea turtles. Proceedings of nization and behavior of the acorn woodpecker in central
the National Academy of Sciences 105:19096–19101. coastal California. Ornithological Monographs 21:1–115.
Lohmann, K. J., S. D. Cain, S. A. Dodge, and C. M. F. Madden, J. R., and K. Tanner. 2003. Preferences for coloured
Lohmann. 2001. Regional magnetic fields as navigational bower decorations can be explained in a nonsexual context.
markers for sea turtles. Science 294:364–366. Animal Behaviour 65:1077–1083.
Lohmann, K. J., C. M. F. Lohmann, L. M. Ehrhart, D. A. Madsen, T., and R. Shine. 1996. Seasonal migrations of preda-
Bagley, and T. Swing. 2004. Geomagnetic map used in sea- tors and prey—A study of pythons and rats in tropical
turtle navigation. Nature 428:909–910. Australia. Ecology 77:149–156.
Lombardino, A. J., and F. Nottebohm. 2000. Age at deafen- Maeterlinck, M. 1901. The Life of the Bee. New York: Dodd,
ing affects the stability of learned song in adult male zebra Mead.
finches. Journal of Neuroscience 20:5054–5064. Mager, J. N., III, C. Walcott, and W. H. Piper. 2007. Male
London, S. E., and B. A. Schlinger. 2007. Steroidogenic common loons, Gavia immer, communicate body mass and
enzymes along the ventricular proliferative zone in the condition through dominant frequencies of territorial
developing songbird brain. Journal of Comparative Neurology yodels. Animal Behaviour 73:683–690.
502:507–521. Magrath, M. J. L., E. van Lieshout, G. Henk Visser, and J.
Lorenz, K. 1935. Der Kumpan in der Umwelt des Vogels. Komdeur. 2004. Nutritional bias as a new mode of adjust-
Journal für Ornithologie 83:137–213. ing sex allocation. Proceedings of the Royal Society of London B
Lorenz, K. 1952. King Solomon’s Ring. New York: Crowell. 271: S347–S349.
Lorenz, K. 1958. The evolution of behavior. Scientific American Magrath, M. J. L., E. van Lieshout, I. Pen, G. Henk Visser,
199:67–78. and J. Komdeur. 2007. Estimating expenditure on male and
Lorenz, K. 1972. Comparative studies on the behavior of female offspring in a sexually size-dimorphic bird: a com-
Anatinae. In Function and Evolution of Behavior: An Historical parison of different methods. Journal of Animal Ecology
Sample from the Pens of Ethologists, edited by P. H. Klopfer 76:1169–1180.
and J. P. Hailman. Reading, MA: Addison-Wesley, pp. Maguire, E. A., D. G. Gadlian, I. S. Johnsrude, C. D. Good,
231–258. J. Ashburner, R. S. J. Frackowiak, and C. D. Frith. 2000.
Lorenz, K., and N. Tinbergen. 1938. Taxis und Instinkhandlung Navigation-related structural change in the hippocampi of
in der Eirollbewegung der Graugans. Zeitschrift für taxi drivers. Proceedings of the National Academy of Sciences
Tierpsychologie 2:1–29. 97:4398–4403.
Lott, D. F. 1991. Intraspecific Variation in the Social Systems of Maher, C., and D. Lott. 1995. Definitions of territoriality used
Wild Vertebrates. Cambridge: Cambridge University Press. in the study of variation in vertebrate spacing systems.
Low, B. S. 1990. Marriage systems and pathogen stress in Animal Behaviour 49:1581–1597.
human societies. Amerian Zoologist 30:325–339. Maher, C., and D. Lott. 2000. A review of ecological deter-
Lutz, C. K., and M. A. Novak. 2005. Environmental enrich- minants of territoriality within vertebrate species. American
ment for nonhuman primates: theory and application. ILAR Midland Naturalist 143:1–29.
Journal 46:178–191. Malenka, R. C., and M. F. Bear. 2004. LTP and LTD: An
Lynch, C. B. 1980. Response to divergent selection for nest- embarrassment of riches. Neuron 44:5–21.
ing behavior in Mus musculus. Genetics 96:757–765. Mallard, S. T., and C. J. Barnard. 2003. Competition, fluctu-
Lynch, C. B. 1992. Clinal variation in cold adaptation in Mus ating asymmetry and sperm transfer in male gryllid crick-
domesticus: Verification of predictions from laboratory pop- ets (Gryllus bimaculatus and Gyllodes sigillatus). Behavioral
ulations. American Naturalist 139:1219–1236. Ecology and Sociobiology 53:190–197.
 References 489

Mallot, R. W., and J. W. Siddall. 1972. Acquisition of the peo- Martin, R. A. 2007. A review of shark agonistic displays: com-
ple concept in pigeons. Psychological Reports 31:3–13. parison of display features and implications for shark-
Manger, P. R., and J. D. Pettigrew. 1995. Electroreception and human interactions. Marine and Freshwater Behaviour and
the feeding behavior of the platypus (Ornithorhynchus Physiology 40:3–34.
anatinus, Monotremata, Mammalia). Philosophical Martin-Gronert, M. S., and S. E. Ozanne. 2006. Maternal
Transactions of the Royal Society of London B 347:359–381. nutrition during pregnancy and health of the offspring.
Biochemical Society Transactions 34:779–782.
Mank, J. E., D. E. L. Promislow, and J. C. Avise. 2005.
Marx, J. L. 1980. Ape-language controversy flares up. Science
Phylogenetic perspectives in the evolution of parental care
207:1330–1333.
in ray-fined fishes. Evolution 59:1570–1578.
Mason, G., R. Clubb, N. Latham, and S. Vickery. 2007. Why
Manoli, D. S., M. Foss, A. Villelle, B. J. Taylor, J. C. Hall, and
and how should we use environmental enrichment to tackle
B. S. Baker. 2005. Male-specific fruitless specifies the neural stereotypic behaviour? Applied Animal Behaviour Science
substrates of Drosophila courtship. Nature 436:395–400. 102:163–188.
Mappes, J., and R. V. Alatalo. 1997. Batesian mimicry and sig- Massey, A. 1988. Sexual interactions in red-spotted newt pop-
nal accuracy. Evolution 51:2050–2053. ulations. Animal Behaviour 36:205–210.
Marcus, G. F., and S. E. Fisher. 2003. FOXP2 in focus: What Mateo, J. M. 2004. Recognition systems and biological orga-
can genes tell us about speech and language? Trends in nization: the perception component of social recognition.
Cognitive Sciences 7:257–262. Annales Zoologici Fennici 41:729–745.
Marhold, S., W. Wiltschko, and H. Burda. 1997. A magnetic Mateo, J. M., and W. G. Holmes. 2004. Cross-fostering as a
polarity compass for direction finding in a subterranean means to study kin recognition. Animal Behaviour
mammal. Naturwissenschaften 84:421–423. 68:1451–1459.
Markham, J., and W. T. Greenough. 2004. Experience-driven Mateo, J. M., and R. E. Johnston. 2000. Kin recognition and
brain plasticity. Neuron Glia Biology 1:351–363. the ‘armpit effect’: evidence of self-referent phenotype
matching. Proceedings of the Royal Society of London B
Marler, C. A., and M. C. Moore. 1988. Evolutionary costs of
267:695–700.
aggression revealed by testosterone manipulations in free-
Mateo, J. M., and R. E. Johnston. 2003. Kin recognition by
living male lizards. Behavioral Ecology and Sociobiology
self-referent phenotype matching: weighing the evidence.
23:21–26.
Animal Cognition 6:73–76.
Marler, C. A., and M. C. Moore. 1991. Supplementary feed- Mather, J. A., and R. C. Anderson. 1999. Exploration, play,
ing compensates for testosterone-induced costs of aggres- and habituation in octopuses (Octopus dofleini). Journal of
sion in male mountain spiny lizards, Sceloporus jarrovi. Comparative Psychology 113:333–338.
Animal Behaviour 42:209–219. Mathis, A., D. P. Chivers, and R. J. F. Smith. 1996. Cultural
Marler, C. A., G. Walsberg, M. L. White, M. Moore. 1995. transmission of predator recognition in fishes: Intraspecific
Increased energy expenditure due to increased territorial and interspecific learning. Animal Behaviour 51:185–201.
defense in male lizards after phenotypic manipulation. Mattila, H. R., and T. D. Seeley. 2007. Genetic diversity in
Behavioral Ecology and Sociobiology 37:225–231. honey bee colonies enhances productivity and fitness.
Marler, P. 1970. A comparative approach to vocal learning: Science 317:362–364.
Song development in white-crowned sparrows. Journal of Mauck, R. A., and T. C. Grubb, Jr. 1995. Petrel parents shunt
Comparative and Physiological Psychology 71:1–25. all experimentally increased reproductive costs to their off-
Marler, P. 1987. Sensitive periods and the roles of specific and spring. Animal Behaviour 49:999–1008.
general sensory stimulation in birdsong learning. In Maynard Smith, J. 1974. The theory of games and the evolu-
Imprinting and Cortical Plasticity: Comparative Aspects of tion of animal conflicts. Journal of Theoretical Biology
47:209–221.
Sensitive Periods, edited by J. R. Rauschecker and P. Marler.
Maynard Smith, J. 1976. Evolution and the theory of games.
New York: John Wiley, pp. 99–135
American Scientist 64:41–45.
Marler, P. 1996. Social cognition: Are primates smarter than
Maynard Smith, J. 1977. Parental investment: A prospective
birds? In Current Ornithology, Vol. 13, edited by V. J. Nolan
analysis. Animal Behaviour 25:1–9.
and E. D. Ketterson. New York: Plenum, pp. 1–33.
Maynard Smith, J. 1982. Evolution and the Theory of Games.
Marler, P., and M. Tamura. 1964. Culturally transmitted pat- Cambridge: Cambridge University Press.
terns of vocal behavior in sparrows. Science 146:1483–1486. Maynard Smith, J., and G. A. Parker. 1976. The logic of asym-
Martan, J., and B. A. Shepherd. 1976. The role of the copu- metric contests. Animal Behaviour 24:159–175.
latory plug in reproduction of the guinea pig. Journal of Maynard Smith, J., and S. E. Riechert. 1984. A conflicting ten-
Experimental Zoology 196:79–84. dency model of spider agonistic behaviour: hybrid-pure
Martín, J., J. J. Luque-Larena, and P. López. 2006. Collective population line comparisons. Animal Behaviour 32:564–578.
detection in escape responses of temporary groups of Mayr, E. 1983. How to carry out the adaptationist program?
Iberian green frogs. Behavioral Ecology 17:222–226. American Naturalist 121:295–310.
490 References

Mays, H. L. Jr., and G. E. Hill. 2004. Choosing mates: good Mech, L. D. 1970. The Wolf: The Ecology and Behavior of an
genes versus genes that are a good fit. Trends in Ecology and Endangered Species. Minneapolis: University of Minnesota
Evolution 19:554–559. Press.
McBeath, M. K., D. M. Shaffer, and M. K. Kaiser. 1995. How Melletti, M., V. Penteriani, and L. Boitani. 2007. Habitat pref-
baseball outfielders determine where to run to catch fly erences of the secretive forest buffalo (Syncerus caffer nanus)
balls. Science 268:569–573. in Central Africa. Journal of Zoology 271:178–186.
McCann, T. S. 1981. Aggression and sexual activity of male Mellgren, R. L., ed. 1983. Animal Cognition and Behavior.
southern elephant seals. Journal of Zoology 195:295–310. Amsterdam: North Holland.
McCormick, S. D. and D. Bradshaw. 2006. Hormonal control Mello, C., F. Nottebohm, and D. Clayton. 1995. Repeated
of salt and water balance in vertebrates. General and exposure to one song leads to rapid and persistent decline
in an immediate early gene’s response to that song in zebra
Comparative Endocrinology 147:3–8.
finch telencephalon. Journal of Neuroscience 15:6919–6925.
McCormick, S. D., L. P. Hansen, T. P. Quinn, and R. L.
Mello, C., D. S. Vicario, and D. F. Clayton. 1992. Song pre-
Saunders. 1998. Movement, migration, and smolting of
sentation induces gene expression in the songbird forebrain.
Atlantic salmon (Salmo salar). Canadian Journal of Fisheries
Proceedings of the National Academy of Sciences 89:6818–6822.
and Aquatic Sciences 55 (Supplement 1):77–92.
Meltzoff, A. N. 1988. The human infant as Homo imitans. In
McCracken, G. F., and M. K. Gustin. 1991. Nursing behav-
Social Learning: Psychological and Biological Perspectives, edited
ior in Mexican free-tailed bat maternity colonies. Ethology by T. R. Zentall and B. G. Galef, Jr. Hillsdale, NJ: Erlbaum,
89:305–321. pp. 319–341.
McDonald, D. B., and W. K. Potts. 1994. Cooperative display Menzel, R. 2007. Searching for the memory trace in a mini-
and relatedness among males in a lek-mating bird. Science brain, the honeybee. Learning and Memory 8:53–62.
266:1030–1032. Meredith, M. 1998. Vomeronasal, olfactory, hormonal con-
McEwen, B. S. 1976. Interactions between hormones and vergence in the brain: cooperation or coincidence? Annals
nerve tissue. Scientific American 235:48–58. of the New York Academy of Sciences 855:349–361.
McFall-Ngai, M. J. 1990. Crypsis in the pelagic environment. Merilaita, S. 1998. Crypsis through disruptive coloration in an
American Zoologist 30:175–188. isopod. Proceedings of the Royal Society of London B.
McGlothlin, J. W., J. M. Jawor, and E. D. Ketterson. 2007. 265:1059–1064.
Natural variation in a testosterone-mediated trade-off Merilaita, S., and J. Lind. 2005. Background-matching and
between mating effort and parental effort. American disruptive coloration, and the evolution of cryptic col-
Naturalist 170:864–875. oration. Proceedings of the Royal Society of London B
McGowan, K. J., and G. E. Woolfenden. 1989. A sentinel sys- 272:665–670.
tem in the Florida scrub jay. Animal Behaviour Mery, F., and T. J. Kawecki. 2002. Experimental evolution of
37:1000–1006. learning ability in fruit flies. Proceedings of the National
McGuire, B. 1988. The effects of cross-fostering on parental Academy of Sciences 99:14274–14279.
behavior of meadow voles (Microtus pennsylvanicus). Journal Mery, F., and T. J. Kawecki. 2004. An operating cost of learn-
of Mammalogy 69:332–341. ing in Drosophila melanogaster. Animal Behaviour 68:589–598.
McGuire, B., T. Pizzuto, W. E. Bemis, and L. L. Getz. 2006. Messenger, J. B. 2001. Cephalopod chromatophores: neuro-
biology and natural history. Biological Reviews 76:473–528.
General ecology of a rural population of Norway rats
(Rattus norvegicus) based on intensive live trapping. American Metzgar, L. H. 1967. An experimental comparison of screech
owl predation on resident and transient white-footed mice
Midland Naturalist 155:221–236.
(Peromyscus leucopus). Journal of Mammalogy 48:387–391.
McGuire, B., L. L. Getz, J. Hofmann, T. Pizzuto, and B. Frase.
Meyer, A. 1997. The evolution of sexually selected traits in
1993. Natal dispersal in prairie voles (Microtus ochrogaster) in
male swordtail fishes (Xiphophorus: Poecilidae). Heredity
relation to population density, season, and natal social envi-
79:329–337.
ronment. Behavioral Ecology and Sociobiology 32: 293–302.
Meyer, A., W. Salburger, and M. Schartl. 2006. Hybrid ori-
McKaye, K. R. 1977. Competition for breeding sites between
gin of a swordtail species (Teleostei: Xiphophorus clemenciae)
the cichlid fishes of Lake Jiloa, Nicaragua. Ecology driven by sexual selection. Molecular Ecology 15:721–730.
58:291–302.
Meyerriecks, A. J. 1960. Comparative breeding behavior of
McKaye, K. R., and N. M. McKaye. 1977. Communal care four species of North American herons. Publications of the
and kidnapping of young by parental cichlids. Evolution Nuttall Ornithology Club 2:1–158.
31:674–681. Michel, G. F., and A. N. Tyler. 2005. Critical period: a his-
McLaren, A., and D. Michie. 1960. Control of prenatal tory of the transition from questions of when, to what, to
growth in mammals. Nature 187:363–365. how. Developmental Psychobiology 46:158–162.
McLean, I. G., C. Holzer, and B. J. S. Studholme. 1999. Michelson, A., B. B. Andersen, W. H. Kirchner, and M.
Teaching predator-recognition to a naive bird: implications Lindauer. 1989. Honeybees can be recruited by a mechan-
for management. Biological Conservation 87:123–130. ical model of a dancing bee. Naturwissenschaften 76:277–280.
 References 491

Michener, C. D. 1969. Comparative social behavior of bees. Møller, A. P. 1992. Parasites differentially increase the degree
Annual Review of Entomology 14:299–342. of fluctuating asymmetry in secondary sexual characters.
Miklósi, Á., V. Csányi, and R. Gerlai. 1997. Antipredator Journal of Evolutionary Biology 5:691–699.
behavior in paradise fish (Macropodus opercularis) larvae: The Møller, A. P., and A. Pomiankowski. 1993. Fluctuating asym-
role of genetic factors and paternal influence. Behavior metry and sexual selection. Genetica 89:267–279.
Genetics 27:191–200. Møller, A. P., and J. P. Swaddle. 1997. Fluctuating Asymmetry,
Miles, H. L. W. 1990. The cognitive foundations for refer- Developmental Stability and Evolution. Oxford: Oxford
ence in a signing orangutan. In “Language” and Intelligence University Press.
in Monkeys and Apes: Comparative Developmental Perspectives, Møller, A. P., and T. R. Birkhead. 1989. Copulation behavior
edited by S. T. Parker and K. R. Gibson. New York: in mammals: Evidence that sperm competition is wide-
Cambridge University Press, pp. 511–539. spread. Biological Journal of the Linnean Society 38:119–131.
Miller, C. A., and J. D. Sweatt. 2007. Covalent modification Møller, A. P., P. Christie, and E. Lux. 1999. Parasitism, host
of DNA regulates memory formation. Neuron 53:857–869. immune function, and sexual selection. Quarterly Review of
Miller, D. B. 1980. Maternal vocal control of behavioral inhi- Biology 74:3–20.
bition in mallard ducklings (Anas platyrhynchos). Journal of
Moment, G. B. 1962. Reflexive selection: A possible answer
Comparative and Physiological Psychology 94:606–623.
to an old question. Science 136:262–263.
Miller, D. B., and G. Gottlieb. 1978. Maternal vocalizations
Moncomble, A.-S., G. Coureaud, B. Quennedey, D. Langlois,
of mallard ducks (Anas platyrhynchos). Animal Behaviour
G. Perrier, and B. Schaal. 2005. The mammary pheromone
26:1178–1194.
of the rabbit: from where does it come? Animal Behaviour
Miller, R. C. 1922. The significance of the gregarious habit.
69:29–38.
Ecology 3:122–126.
Mönkkönen, M., R. Hardling, J. T. Forsman, and J. Tuomi.
Mineka, S., and M. Cook. 1988. Social learning and the acqui-
1999. Evolution of heterospecific attraction: using other
sition of snake fear in monkeys. In Social Learning:
species as cues in habitat selection. Evolutionary Ecology
Psychological and Biological Perspectives, edited by T. R. Zentall
13:91–104.
and B. G. J. Galef. Hillsdale, NJ: Erlbaum, pp. 51–73.
Moore, F. L., S. K. Boyd, and D. B. Kelley. 2005. Historical
Missoweit, M., and K. P. Sauer. 2007. Not all Panorpa
perspective: Hormonal regulation of behavior in amphib-
(Mecoptera: Panorpidae) scorpionfly mating systems are
ians. Hormones and Behavior 48:373–383.
characterized by resource defence polygyny. Animal
Behaviour 74:1207–1213. Moore, F. L., and S. J. Evans. 1999. Steroid hormones use
non-genomic mechanisms to control brain functions and
Mistlberger, R. E., and B. Rusak. 2005. Biological Rhythms
behaviors: a review of evidence. Brain, Behavior and
and Behavior. In The Behavior of Animals: Mechanisms,
Evolution 54:41–50.
Function, and Evolution, edited by J. J. Bolhuis and L.-A.
Giraldeau. Malden, MA: Blackwell Publishers. pp. 71–96. Moore, F. L., and L. J. Miller. 1984. Stress-induced inhibition
Mock, D. W., H. Drummond, and C. H. Stinson. 1990. Avian of sexual behavior: corticosterone inhibits courtship behav-
siblicide. American Scientist 78:438–449. iors of a male amphibian (Taricha granulosa). Hormones and
Behavior 18:400–410.
Mock, D. W., and G. A. Parker. 1997. The Evolution of Sibling
Rivalry. Oxford: Oxford University Press. Moore, F. R. 1986. Sunrise, skylight polarization, and the early
Mock, D. W., and G. A. Parker. 1998. Siblicide, family con- morning orientation of night-migrating warblers. Condor
flict and the evolutionary limits of selfishness. Animal 88:493–498.
Behaviour 56:1–10. Moore, J., and R. Ali. 1984. Are dispersal and inbreeding
Moehlman, P. D. 1979. Jackal helpers and pup survival. Nature avoidance related? Animal Behaviour 32:94–112.
277:382–383. Moore, M. C. 1991. Application of organization-activation
Molina, A., A. G. Castellano, and J. López-Barneo. 1997. Pore theory to alternative male reproductive strategies: a review.
mutations in Shaker K+ channels distinguish between the Hormones and Behavior 25:154–179.
sites of tetraethylammonium blockade and C-type inacti- Moore, M. C., and C. A. Marler. 1987. Effects of testosterone
vation. Journal of Physiology 499:361–367. manipulations on nonbreeding season territorial aggression
Møller, A. P. 1988. Female choice selects for male sexual tail in free-living male lizards, Sceloporus jarrovi. General and
ornaments in the monogamous swallow. Nature Comparative Endocrinology 65:255–232.
322:640–642. Moore, M. C., D. K. Hews, and R. Knapp. 1998. Evolution
Møller, A. P. 1990. Effects of a haemotaphagus mite on the and hormonal control of alternative male phenotypes.
barn swallow (Hirundo rustica): A test of the Hamilton and American Zoologist 38:133–151.
Zuk hypothesis. Evolution 44:771–784. Moran, G. 1984. Vigilance behavior and alarm calls in a cap-
Møller, A. P. 1991. Parasite load reduces song output in a tive group of meerkats, Suricata suricatta. Zeitschrift für
passerine bird. Animal Behaviour 41:723–730. Tierpsychologie 65:228–240.
492 References

Moreira, P. L., P. López, and J. Martín. 2006. Femoral secre- Müller, M., and R. Wehner. 2007. Wind and sky as compass cues
tions and copulatory plugs convey chemical information in desert ant navigation. Naturwissenschaften 94:589–594.
about male identity and dominance status. Behavioral Munz, T. 2005. The bee battles: Karl von Frisch, Adrian
Ecology and Sociobiology 60:166–174. Wenner and the honey bee dance language controversy.
Morell, V. 1998. A new look at monogamy. Science Journal of the History of Biology 38:535–570.
281:1982–1983. Nagy, M., G. Heckel, C. C. Voigt, and F. Mayer. 2007.
Morena, J., M. Soler, M. Lindén, and A. P. Møller. 1994. The Female-biased dispersal and patrilocal kin groups in a mam-
function of stone carrying in the black wheatear, Oenanthe mal with resource-defense polygyny. Proceedings of the Royal
leucura. Animal Behaviour 47:1297–1309. Society of London B 274:3019–3025.
Morgan, C. L. 1894. An Introduction to Comparative Psychology. Nahallage, C., and M. A. Huffman. 2007. Age-specific func-
New York: Scribner.
tions of stone handling, a solitary-object play behavior, in
Morris, D. 1954. The reproductive behavior of the zebra finch Japanese macaques (Macaca fuscata). American Journal of
Poephila guttata with special reference to pseudofemale
Primatology 69:267–281.
behavior and displacement activities. Behaviour 6:271–322.
Nakano, R., N. Skals, T. Takanashi, A. Surlykke, T. Koike, K.
Morris, D. 1956. The feather postures of birds and the prob-
Yoshida, H. Maruyama, S. Tatsuki, and Y. Ishikawa. 2008.
lem of the origin of social signals. Behaviour 9:75–113.
Moths produce extremely quiet ultrasonic courtship songs
Morris, D. 1958. The reproductive behavior of the ten-spined
by rubbing specialized scales. Proceedings of the National
stickleback (Pygosteus pungitius L.). Behaviour Supplement
Academy of Sciences 105:11812–11817.
6:1–154.
Narendra, A., K. Cheng, D. Sulikowski, and R. Wehner. 2008.
Morris, R. J., and M. D. E. Fellowes. 2002. Learning and natal
Search strategies of ants in landmark-rich environments.
host influence host preference and sex allocation behaviour
in a pupal parasitoid. Behavioral Ecology and Sociobiology Journal of Comparative Physiology A 194:929–938.
51:386–393. Naumann, K., M. L. Winston, K. N. Slessor, G. D. Prestwich,
Morrow, E. H., and G. Arnqvist. 2003. Costly traumatic and F. X. Webster. 1991. Production and transmission of
insemination and a female counter-adaptation in bed bugs. honey bee queen (Apis mellifera L.) mandibular gland
Proceedings of the Royal Society of London B 270:2377–2381. pheromone. Behavioral Ecology and Sociobiology 29:321–332.
Morton, E. S. 1975. Ecological sources of selection in avian Neff, B. D. 2003. Decisions about parental care in response
sounds. American Naturalist 109:17–34. to perceived paternity. Nature 422:716–719.
Mottley, K., and C. Heyes. 2003. Budgerigars (Melopsittacus Neff, B. D., and P. W. Sherman. 2003. Nestling recognition
undulatus) copy virtual demonstrators in a two-action test. via direct cues by parental male bluegill sunfish (Lepomis
Journal of Comparative Psychology 117:363–370. macrochirus). Animal Cognition 6:87–92.
Moum, S. E., and R. L. Baker. 1990. Colour change and sub- Neff, B. D., P. Fu, and M. R. Gross. 2003. Sperm investment
strate selection in larval Ischnura verticalis. Canadian Journal and alternative mating tactics in bluegill sunfish (Lepomis
of Zoology 68:221–224. macrochirus). Behavioral Ecology 14:634–641.
Mouritsen, H., U. Janssen-Bienhold, M. Liedvogel, G. Feenders, Negro, J. J., J. Bustmante, J. Milward, and D. M. Bird. 1996.
J. Stalleicken, P. Dirks, and R. Weiler. 2004. Cryptochromes Captive fledgling American kestrels prefer to play with
and neuronal-activity markers colocalize in the retina of objects resembling natural prey. Animal Behaviour
migratory birds during magnetic orientation. Proceedings of the 52:707–714.
National Academy of Sciences 101:14294–14299.
Neill, S. R. St. J., and J. M. Cullen. 1974. Experiments on
Moyer, K. E. 1976. Psychobiology of Aggression. New York: whether schooling by their prey affects the hunting behav-
Harper and Row.
ior of cephalopods and fish predators. Journal of Zoology
Moynihan, M. 1966. Communication in Callicebus. Journal of 172:549–569.
Zoology 150:77–127.
Nelson, C. M., K. E. Ihle, M. K. Fondrk, and R. E. Page, Jr.
Moynihan, M. 1967. Comparative aspects of communication
2007. The gene vitellogenin has multiple coordinating effects
in New World primates. In Primate Ethology, edited by D.
on social organization. PLoS Biology 5:0673–0677.
Morris. London: Weidenfeld & Nicholson, pp. 236–266.
Nelson, D. A., and P. Marler. 1994. Selection-based learning
Moynihan, M. 1970. Some behavior patterns of platyrrhine
in bird song development. Proceedings of the National
monkeys. II. Saguinus geoffroyi and some other tamarins.
Smithsonian Contributions to Zoology 28:1–77. Academy of Sciences 91:10498–10501.
Mueller, U. G., N. M. Gerardo, D. K. Aanen, D. L. Six, and Nelson, R. J. 2005. An Introduction to Behavioral Endocrinology.
T. R. Schultz. 2005. The evolution of agriculture in insects. Sunderland, MA: Sinauer.
Annual Review of Ecology, Evolution, and Systematics Neumann, D. 1976. Entrainment of a semi-lunar rhythm. In
36:563–595. Biological Rhythms in the Marine Environment, edited by P.
Muller, M. N., and R. W. Wrangham. 2004. Dominance, cor- DeCoursey. Columbia: University of South Carolina Press.
tisol and stress in wild chimpanzees (Pan troglodytes schwe- Nevitt, G. A. 1999a. Foraging by seabirds on an olfactory
infurthii). Behavioral Ecology and Sociobiology 55:332–340. landscape. American Scientist 87:46–53.
 References 493

Nevitt, G. A. 1999b. Olfactory foraging in Antarctic seabirds: O’Brien, E. L., A. E. Burger, and R. D. Dawson. 2005.
A species-specific attraction to krill odors. Marine Ecology Foraging decision rules and prey species preferences of
Progress Series 177:235–241. northwestern crows (Corvus caurinus). Ethology 111:77–87.
Nevitt, G. A., R. R. Veit, and P. Kareiva. 1995. Dimethyl sul- O’Brien, S. J. 1994. A role for molecular genetics in biologi-
fide as a foraging cue for Antarctic procellariiform seabirds. cal conservation. Proceedings of the National Academy of
Nature 376:680–682. Sciences 91:5748–5755.
Nevo, E., G. Heth, and H. Pratt. 1991. Seismic communica- O’Connell-Rodwell, C. E. 2007. Keeping an “ear” to the
tion in a blind subterranean mammal: A major somatosen- ground: Seismic communication in elephants. Physiology
sory mechanism in adaptive evolution underground. 22:287–294.
Proceedings of the National Academy of Sciences 88:1256–1260. O’Connell-Rodwell, C. E., J. D. Wood, C. Kinzley, T. C.
Newman, S. W. 1999. The medial extended amygdala in male Rodwell, J. H. Poole, and S. Puria. 2007. Wild African ele-
reproductive behavior: a node in the mammalian social phants (Loxodonta africana) discriminate between familiar
and unfamiliar conspecific seismic alarm calls. Journal of the
behavior network. Annals of the New York Academy of Sciences
Acoustical Society of America 122: 823–830.
877:242–257.
O’Loghlen, A. L., and S. I. Rothstein. 1993. An extreme exam-
Newton, I. 2007. Weather-related mass-mortality events in
ple of delayed vocal development: song learning in a pop-
migrants. Ibis 149:453–467.
ulation of wild brown-headed cowbirds. Animal Behaviour
Newton, P. N. 1986. Infanticide in an undisturbed population
46:293–304.
of Hanuman langurs, Presbytis entellus. Animal Behaviour
O’Sullivan, C., and C. Yeager. 1989. Communicative context
34:785–789.
and linguistic competence. In Teaching Sign Language in
Noble, G. K. 1939. The role of dominance in the social life Chimpanzees, edited by R. G. Gardner, B. T. Gardner, and
of birds. Auk 56:263–273. T. E. Van Cantfort. Albany: State University of New York
Noë, R., and A. A. Sluijter. 1990. Reproductive tactics of male Press, pp. 269–279.
savanna baboons. Behaviour 113:117–170. Oetting, S., E. Pröve, and H.-J. Bischof. 1995. Sexual
Nonacs, P. 2001. State dependent behavior and the Marginal imprinting as a two-stage process: Mechanisms of infor-
Value Theorem. Behavioural Ecology 12:71–83. mation storage and stabilization. Animal Behaviour
Nordby, J. C., S. E. Campbell, and M. D. Beecher. 2007. 50:393–403.
Selective attrition and individual song repertoire develop- Olberg, R. M., A. H. Worthington, and K. R. Venator. 2000.
ment in song sparrows. Animal Behaviour 74:1413–1418. Prey pursuit and interception in dragonflies. Journal of
Nordeen, E. J., and K. W. Nordeen. 1992. Auditory feedback Comparative Physiology A 186:155–162.
is necessary for the maintenance of stereotyped song in Olson, D. J., A. C. Kamil, R. P. Balda, and P. J. Nims. 1995.
adult zebra finches. Behavioral and Neural Biology 57:58–66. Performance of four seed-caching corvid species in operant
Norris, K. S., and C. H. Lowe. 1964. An analysis of back- test of nonspatial and spatial memory. Journal of Comparative
ground color-matching in amphibians and reptiles. Ecology Psychology 109:173–181.
45:565–580. Olsson, M., T. Madsen, and R. Shine. 1997. Is sperm really
Nottebohm, F. 1981. A brain for all seasons: cyclical anatom- so cheap? Costs of reproduction in male adders, Vipera
berus. Proceedings of the Royal Society of London B
ical changes in song control nuclei of the canary brain.
264:455–459.
Science 214:1368–1370.
Ophir, A. G., S. M. Phelps, A. B. Sorin, and J. O. Wolff. 2008.
Nottebohm, F. 2002. Why are some neurons replaced in the
Social but not genetic monogamy is associated with greater
adult brain? The Journal of Neuroscience 22:624–628.
breeding success in prairie voles. Animal Behaviour
Nottebohm, F. 2005. The neural basis of birdsong. PLoS 75:1143–1154.
Biology 3:759–761.
Orchinik, M., P. Licht, and D. Crews. 1988. Plasma steroid
Nottebohm, F., and J. P. Arnold. 1976. Sexual dimorphism in concentrations change in response to sexual behavior in
vocal control areas of the songbird brain. Science Bufo marinus. Hormones and Behavior 22:338–350.
194:211–213. Orchinik, M., T. F. Murray, and F. L. Moore. 1991. A corti-
Nottebohm, F., T. M. Stokes, and C. M. Leonard. 1976. costeroid receptor in neuronal membranes. Science
Central control of song in the canary (Serinus canarius). 252:1848–1851.
Journal of Comparative Neurology 165:457–468. Ord, T. J., R. A. Peters, B. Clucas, and J. A. Stamps. 2007.
Nunes, S. 2007. Dispersal and philopatry. In Rodent Societies: Lizards speed up displays in noisy motion habitats.
An Ecological and Evolutionary Perspective, edited by J. O. Proceedings of the Royal Society B 274:1057–1062.
Wolff and P. W. Sherman. Chicago: Chicago University Orians, G. H. 1969. On the evolution of mating systems of
Press, pp. 150–162. birds and mammals. American Naturalist 103:589–603.
Nunes, S., and K. E. Holekamp. 1996. Mass and fat influence Orians, G. H., and G. M. Christman. 1968. A comparative study
the timing of natal dispersal in Belding’s ground squirrels. of red-winged, tri-colored and yellow-headed blackbirds.
Journal of Mammalogy 77:807–817. University of California Publications in Zoology 84:1–81.
494 References

Osborne, K. A., A. Robichon, E. Burgess, S. Shaw Butland, Papi, F., L. Fiore, V. Fiaschi, and S. Benvenuti. 1972. Olfaction
R. A., A. Coulthard, H. S. Pereira, R. J. Greenspan, and M. and homing in pigeons. Monitore Zoologico Italiano (N.S.)
B. Sokolowski. 1997. Natural behavior polymorphism due 6:85–95.
to a cGMP-dependent kinase of Drosophila. Science Paranjpe, D. A., and V. K. Sharma. 2005. Evolution of tem-
277:834–836. poral order in living organisms. Journal of Circadian Rhythms
Ostfeld, R. S. 1990. The ecology of territoriality in small 3:7–19.
mammals. Trends in Ecology and Evolution 5:411–415. Parejo, D., J. White, and E. Danchin. 2007. Settlement deci-
Ostreiher, R. 2003. Is mobbing altruistic or selfish behaviour? sions in blue tits: difference in the use of social cues accord-
Animal Behaviour 66:145–149. ing to age and individual success. Naturwissenschaften
Owen, D. 1980. Camouflage and Mimicry. Chicago: University 94:749–757.
of Chicago Press. Parish, A. R., and F. B. M. de Waal. 2000. The other “closest
Owens, D., and M. Owens. 1996. Social dominance and living relative”: How bonobos (Pan paniscus) challenge tra-
reproductive patterns in brown hyaenas, Hyaena brunnea, of ditional assumptions about females, dominance, intra- and
the central Kalahari desert. Animal Behaviour 51:535–551. intersexual interactions, and hominid evolution. Annals of
Owens, I. P. F. 2006. Where is behavioral ecology going? the New York Academy of Sciences 907:97–113.
Trends in Ecology and Evolution 21:356–361. Parker, G. A. 1970. Sperm competition and its evolutionary
Owings, D., and D. Hennessey. 1984. The importance of vari- consequences in the insects. Biological Reviews 45:525–567.
ation in sciurid visual and vocal communication. In Biology Parker, G. A. 1974. Assessment strategy and the evolution of
of Ground-Dwelling Squirrels: Annual Cycles, Behavioral fighting behavior. Journal of Theoretical Biology 47:223–243.
Ecology, and Sociality, edited by J. O. Murie and G. R. Parker, G. A. 1979. Sexual selection and sexual conflict. In
Michener. Lincoln, NE: University of Nebraska Press, pp. Sexual Selection and Reproductive Competition in Insects,
202–247. edited by M. S. Blum and N. A. Blum. London: Academic
Packer, C. 1977. Reciprocal altruism in Papio anubis. Nature Press, pp. 123–166.
265:441–443. Parker, G. A. 1984. Sperm competition and the evolution of
Packer, C. 1986. The ecology of felid sociality. In Ecological mating strategies. In Sperm Competition and the Evolution of
Aspects of Social Evolution, edited by D. J. Rubenstein and R. Animal Mating Systems, edited by R. L. Smith. Orlando, FL:
W. Wrangham. Princeton, NJ: Princeton University Press, Academic Press, pp. 2–60.
pp. 429–451. Parker, G. A. 2006. Sexual conflict over mating and fertiliza-
Packer, C., and A. E. Pusey. 1982. Cooperation and compe- tion: an overview. Philosophical Transactions of the Royal Society
tition within coalitions of male lions: Kin selection or game of London B 361:235–259.
theory? Nature 296:740–742. Parrish, J. K. 1989. Re-examining the selfish herd: Are cen-
Packer, C., and A. E. Pusey. 1987. Intrasexual cooperation and tral fish safer? Animal Behaviour 38:1048–1053.
the sex ratio in African lions. American Naturalist Partan, S. R., and P. Marler. 2005. Issues in the classification
130:636–642. of multimodal communication signals. American Naturalist
Packer, C., D. Scheel, and A. E. Pusey. 1990. Why lions form 166:231–245.
groups: Food is not enough. American Naturalist Patricelli, G., J. A. C. Uy, and G. Borgia. 2003. Multiple male
119:263–281. traits interact: attractive bower decorations facilitate attrac-
Packer, C., D. A. Gilbert, A. E. Pusey, and S. J. O’Brien. 1991. tive behavioural displays in satin bowerbirds. Proceedings of
A molecular genetic analysis of kinship and cooperation in the Royal Society of London B 270:2389–2395.
African lions. Nature 351:562–565. Patterson, F. 1978. The gestures of a gorilla: Sign language
Packer, C., L. Herbst, A. E. Pusey, J. D. Bygott, J. P. Hanby, acquisition in another pongid species. Brain and Language
S. J. Cairns, and M. B. Mulder. 1988. Reproductive success 5:72–97.
in lions. In Reproductive Success, Studies of Individual Variation Patterson, F. L. 1990. Language acquisition by a lowland
in Contrasting Breeding Systems, edited by T. H. Clutton- gorilla: Koko’s first ten years of vocabulary development.
Brock. Chicago: University of Chicago Press, pp. 363–383. Word 41:97–143.
Page, R. E., Jr., and R. A. Metcalf. 1982. Multiple mating, Pavlov, I. 1927. Conditioned Reflexes. Edited by G. V. Anrep.
sperm utilization, and social evolution. American Naturalist London: Oxford University Press.
119:263–281. Payne, R. B. 1962. How the barn owl locates its prey by hear-
Palmer, J. D. 1990. The rhythmic lives of crabs: the same bio- ing. Living Bird 1:151–159.
logical clock that controls circadian rhythms may also drive Payne, K., W. R. Langbauer Jr., and E. M. Thomas. 1986.
crustacean tidal rhythms. BioScience 40:352–358. Infrasonic calls of the Asian elephant (Elephas maximus).
Palmer, J. D. 1995. The Biological Rhythms and Clocks of Behavioral Ecology and Sociobiology 18:297–301.
Intertidal Animals. New York: Oxford University Press. Payne, R. S., and D. Webb. 1971. Orientation by means of
Palmer, J. D. 2000. The clocks controlling the tide-associated long range acoustic signaling in baleen whales. Annals of the
rhythms of intertidal animals. BioEssays 22:32–37. New York Academy of Sciences 188:110–142.
 References 495

Peeke, H. V. S. 1984. Habituation and the maintenance of ter- Perfito, N., R. A. Zann, G. E. Bentley, and M. Hau. 2007.
ritorial boundaries. In Habituation, Sensitization, and Opportunism at work: Habitat predictability affects repro-
Behavior, edited by H. V. S. Peeke and L. Petrinovich. New ductive readiness in free-living zebra finches. Functional
York: Academic Press, pp. 393–421. Ecology 21:291–301.
Penn, D. J., and W. K. Potts. 1998. How do major histocom- Peroulakis, M. E., B. Goldman, and N. G. Forger. 2002.
patibility genes influence odor and mating preferences? Perineal muscles and motoneurons are sexually monomor-
Advances in Immunology 69:411–435. phic in the naked mole-rat (Heterocephalus glaber). Journal
Penn, D. J., and W. K. Potts. 1999. The evolution of mating of Neurobiology 51:33–42.
preferences and the major histocompatibility complex Perrin, N., and J. Goudet. 2001. Inbreeding, kinship, and the
genes. American Naturalist 153:145–164. evolution of natal dispersal. In Dispersal, edited by J.
Penteriani, V., and M. M. Delgado. 2008. Owls may use fae- Clobert, E. Danchin, A. A. Dhondt, and J. D. Nichols.
ces and prey feathers to signal current reproduction. PloS 1 Oxford: Oxford University Press, pp. 123–142.
3:e3014. Perrin, N., and L. Lehmann. 2001. Is sociality driven by the
Pepperberg, I. M. 1987a. Evidence for conceptual quantita- costs of dispersal or the benefits of philopatry? A role for
tive abilities in the African grey parrot: Labeling of cardi- kin discrimination mechanisms. American Naturalist
nal sets. Ethology 75:37–61. 158:471–483.
Pepperberg, I. M. 1987b. Acquisition of same/different con- Peters, R. A., J. M. Hemmi, and J. Zeil. 2007. Signaling
cept by an African grey parrot (Psittacus erithacus): Learning against the wind: modifying motion-signal structure in
with respect to categories of color, shape and material. response to increased noise. Current Biology 17:1231–1234.
Animal Learning and Behavior 15:423–432. Petrie, M. 1992. Peacocks with low mating success are more
Pepperberg, I. M. 1991. A communicative approach to animal likely to suffer predation. Animal Behaviour 44:585–586.
cognition: A study of conceptual abilities of an African grey
Petrie, M. 1994. Improved growth and survival of offspring
parrot. In Cognitive Ethology, the Minds of Other Animals,
of peacocks with more elaborate trains. Nature
edited by C. A. Ristau. Hillsdale, NJ: Erlbaum, pp. 153–186.
371:598–599.
Pepperberg, I. M. 1993. Cognition and communication in an
Petrie, M., and T. Halliday. 1994. Experimental and natural
African grey parrot (Psittacus erithacus). In Language and
change in the peacock’s (Pavo cristatus ) train can affect mat-
Communication, Comparative Perspectives, edited by H. L.
ing success. Behavioral Ecology Sociobiology 35:213–217.
Roitblat, L. M. Herman, and P. E. Nachtigall. Hillsdale, NJ:
Petrie, M., T. Halliday, and C. Sanders. 1991. Peahens prefer
Lawrence Erlbaum, pp. 221–248.
peacocks with elaborate trains. Animal Behaviour
Pepperberg, I. M. 1994. Numerical competence in an African
41:323–331.
grey parrot (Psittacus erithacus). Journal of Comparative
Psychology 108:36–44. Pettigrew, J. D., P. R. Manger, and S. L. B. Fine. 1998. The
sensory world of the platypus. Philosophical Transactions of the
Pepperberg, I. M. 2000. The Alex Studies: Cognitive and
Royal Society of London B 353:1199–1210.
Communicative Abilities of Grey Parrots. Harvard University
Press: Cambridge. Pfennig, D. W., G. J. Gamboa, H. K. Reeve, J. S. Reeve, and
Pepperberg, I. M. 2006. Grey parrot (Psittacus erithacus) I. D. Ferguson. 1983. The mechanism of nestmate dis-
numerical abilities: addition and further experiments on a crimination in social wasps (Polistes, Hymenoptera:
zero-like concept. Journal of Comparative Psychology Vespidae). Behavioral Ecology and Sociobiology 13:299–305.
120:1–22. Phillips, J. B. 1986. Two magnetoreception pathways in a
Pepperberg, I. M., and J. D. Gordon. 2005. Number com- migratory salamander. Science 233:765–766.
prehension by a Grey parrot (Psittacus erithacus), including Phillips, J. B. 1987. Laboratory studies of homing orientation
a zero-like concept. Journal of Comparative Psychology in the eastern red-spotted newt, Notophthalmus viridescens.
119:197–209. Journal of Experimental Biology 131:215–229.
Perdeck, A. C. 1958. Two types of orientation in migrating Phillips, J. B., and S. C. Borland. 1992. Wavelength specific
starlings, Sturnus vulgaris L., and chaffinches, Fringilla effects of light on magnetic compass orientation in the east-
coelebs L., as revealed by displacement experiments. Ardea ern red-spotted newt Notophthalmus viridescens. Ethology,
46:1–37. Ecology and Evolution 4:33–42.
Perdeck, A. C. 1967. Orientation of starlings after displace- Phillips, J. B., and S. C. Borland. 1994. Use of a specialized
ment to Spain. Ardea 51:91–104. magnetoreception system for homing by the eastern red-
Pereira, H. S., and M. B. Sokolowski. 1993. Mutations in the spotted newt Notophthalmus viridescens. Journal of
larval foraging gene affect adult locomotory behavior. Experimental Biology 188:275–291.
Proceedings of the National Academy of Sciences 90:5044–5046. Phillips, J. B., K. Schmidt-Koenig, and R. Muheim. 2006.
Pérez, M., and F. Coro. 1984. Physiological characteristics of True navigation: Sensory bases of gradient maps. In Animal
the tympanic organ in noctuid moths. I. Responses to brief Spatial Cognition: Comparative, Neural, and Computational
acoustic pulses. Journal of Comparative Physiology A Approaches, edited by M. F. Brown and R. G. Cook. [On-
154:441–447. line]. Available: www.pigeon.psy.tufts.edu/asc/phillips/.
496 References

Phoenix, C., R. Goy, A. Gerall, and W. Young. 1959. Poole, J. H., K. Payne, W. R. Langbauer Jr., and C. Moss.
Organizing action of prenatally administered testosterone 1988. The social contexts of some very low frequency calls
proprionate on the tissues mediating mating behavior in the of African elephants. Behavioral Ecology and Sociobiology
female guinea pig. Endocrinology 65:369–382. 22:385–392.
Pietrewicz, A. T., and A. C. Kamil. 1979. Search image for- Porter, R. H., V. J. Tepper, and D. M. White. 1981.
mation in the blue jay (Cyanocitta cristata). Science Experimental influences on the development of huddling
204:1332–1333. preferences and “sibling” recognition in spiny mice.
Pietrewicz, A. T., and A. C. Kamil. 1981. Search images and Developmental Psychobiology 14:375–382.
the detection of cryptic prey: An operant approach. In Portfors, C. V. 2007. Types and functions of ultrasonic vocal-
Foraging Behavior: Ecological, Ethological and Psychological izations in laboratory rats and mice. Journal of the American
Approaches, edited by A. C. Kamil and T. D. Sargent. New Association for Laboratory Animal Science 46:28–34.
York: Garland STPM Press, pp. 311–331. Povinelli, D. J., G. G. Gallup, Jr., T. J. Eddy, D. T.
Pinsker, H., I. Kupfermann, V. Castellucci, and E. R. Kandel. Bierschwale, M. C. Engstron, H. K. Perilloux, and I. B.
1970. Habituation and dishabituation of the gill-withdrawal Toxopeus. 1997. Chimpanzees recognize themselves in mir-
reflex in Aplysia. Science 167:1740–1742. rors. Animal Behaviour 53:1083–1088.
Plassart-Schiess, E., and E. E. Baulieu. 2001. Neurosteroids: Pravosudov, V. V. 2006. On seasonality in food-storing behav-
recent findings. Brain Research Reviews 37:133–140. iour in parids: Do we know the whole story? Animal
Behaviour 71:1455–1460.
Plautz, J. D., M. Kaneko, J. C. Hall, and S. A. Kay. 1997a.
Pravosudov, V. V. 2007. The relationship between environ-
Independent photoreceptive circadian clocks throughout
ment, food caching, spatial memory, and the hippocampus
Drosophila. Science 278:1632–1635.
in chickadees. In Ecology and Behavior of Chickadees and
Plautz, J. D., M. Straume, R. Stanewsky, C. F. Jamieson, C.
Titmice: an Integrated Approach, edited by K. Otter. Oxford:
Brandes, H. B. Dowse, J. C. Hall, and S. A. Kay. 1997b.
Oxford University Press, pp. 25–41.
Quantitative analysis of Drosophila period gene transcription
Pravosudov, V. V., and S. R de Kort. 2006. Is the western
in living animals. Journal of Biological Rhythms 12:204–217.
scrub-jay (Aphelocoma californica) really an underdog among
Plomin, R. , J. C. DeFries, and G. E. McClearn. 1980.
food-caching corvids when it comes to hippocampal volume
Behavioral Genetics—A Primer. San Francisco: W. H.
and food caching propensity? Brain, Behavior and Evolution
Freeman.
67:1–9.
Plomin, R., J. C. DeFries, I. W. Craig, and P. McGuffin. 2003. Premack, D. 1976. Intelligence in Ape and Man. Hillsdale, NJ:
Behavioral genetics. In Behavioral Genetics in the Postgenomic Erlbaum.
Era, edited by R. Plomin, J. C. DeFries, I. W. Craig and P.
Preston, B. T., I. R. Stevenson, J. M. Pemberton, and K.
McGuffin. Washington, DC: American Psychological
Wilson. 2001. Dominant rams lose out by sperm depletion.
Association. Nature 409:681–682.
Plotnik, J. J., F. de Waal, and D. Reiss. 2006. Self-recognition Pribil, S., and J. Picman. 1996. Polygyny in the red-winged
in an Asian elephant. Proceedings of the National Academy of blackbird: Do females prefer monogamy or polygamy?
Sciences 103:17053–17057. Behavioral Ecology and Sociobiology 38:183–190.
Plummer, M. R., and J. M. Camhi. 1981. Discrimination of Pribil, S., and W. A. Searcy. 2001. Experimental confirmation
sensory signal from noise in the escape system of the cock- of the polygyny threshold model for red-winged blackbirds.
roach: The role of wind acceleration. Journal of Comparative Proceedings of the Royal Society of London B 268:1643–1646.
Physiology A 142:347–357. Price, T. D., P. R. Grant, H. L. Gibbs, and P. T. Boag. 1984.
Podos, J., and S. Nowicki. 2004. Beaks, adaptation, and vocal Recurrent patterns of natural selection in a population of
evolution in Darwin’s finches. BioScience 54:501–510. Darwin’s finches. Nature 309:787–789.
Poindron, P., F. Levy, and M. Keller. 2007. Maternal respon- Proctor, H. C. 1991. Courtship in the water mite Neumania
siveness and maternal selectivity in domestic sheep and papillator: Males capitalize on female adaptations for pre-
goats: the two facets of maternal attachment. Developmental dation. Animal Behaviour 42:589–598.
Psychobiology 49:54–70. Proctor, H. C. 1992. Sensory exploitation and the evolution
Pomiankowski, A., M. Denniff, K. Fowler, and T. Chapman. of male mating behavior: A cladistic test. Animal Behaviour
2005. The costs and benefits of high early mating rates in 44:745–752.
male stalk-eyed flies (Cyrtodiopsis dalmanni). Journal of Insect Proctor, M., P. Yeo, and A. Lack. 1996. The Natural History of
Physiology 51:1165–1171. Pollination. Portland: Timber Press.
Pongrácz, P., A. Miklósi, E. Kubinyi, K. Gurobi, J. Topál, and Pruett-Jones, S. 2004. Summary. In Ecology and Evolution of
V. Csányi. 2001. Social learning in dogs: the effect of a Cooperative Breeding in Birds, edited by W. Koenig and J.
human demonstrator on the performance of dogs in a Dickinson. Cambridge: Cambridge University Press, pp.
detour task. Animal Behaviour 62:1109–1117. 228–238.
 References 497

Pruetz, J. D., and P. Bertolani. 2007. Savanna chimpanzees, Rasa, O. A. E. 1986. Coordinated vigilance in dwarf mongoose
Pan troglodytes verus, hunt with tools. Current Biology family groups: The “Watchman’s Song” hypothesis and the
17:1–6. costs of guarding. Ethology 71:340–344.
Pusey, A. E. 1987. Sex-biased dispersal and inbreeding avoid- Rayor, L. S., and G. W. Uetz. 1990. Trade-offs in foraging
ance in birds and mammals. Trends in Ecology and Evolution success and predation risk with spatial position in colonial
2:295–299. spiders. Behavioral Ecology and Sociobiology 27:77–85.
Pusey, A. E., and C. Packer. 1997. The ecology of relation- Redondo, T., and F. Castro. 1992. Signalling of nutritional
ships. In Behavioural Ecology: An Evolutionary Approach, need by magpie nestlings. Ethology 92:193–204.
edited by J. R. Krebs and N. B. Davies. Oxford: Blackwell
Reeder, D. M. 2003. The potential for cryptic female choice
Science, pp. 254–283.
in primates: behavioral, anatomical, and physiological con-
Putz, O., and D. Crews. 2006. Embryonic origin of mate siderations. In Special Topics in Primatology, Volume 3. Sexual
choice in a lizard with temperature-dependent sex deter-
Selection and Reproductive Competition in Primates: New
mination. Developmental Psychobiology 48:29–38.
Perspectives and Directions, edited by C. B. Jones. Norman:
Quinlan, R. J., and J. M. Cherrett. 1977. The role of substrate
American Society of Primatologists, pp. 254–303.
preparation in the symbiosis between the leaf-cutting ant
Reimchen, T. E. 1989. Shell color ontogeny and tubeworm
Acromyrmex octospinosus (Reich) and its food fungus.
mimicry in a marine gastropod Littorina mariae. Biological
Ecological Entomology 2:161–170.
Journal of the Linnean Society 36:97–109.
Quinn, T. P., and A. H. Dittman. 1990. Pacific salmon migra-
tions and homing: Mechanisms and adaptive significance. Reimers, M., F. Schwarzenberger, and S. Preuschoft. 2007.
Trends in Ecology and Evolution 5:174–177. Rehabilitation of research chimpanzees: stress and coping
Quinn, T. P., I. J. Stewart, and C. P. Boatright. 2006. after long-term social isolation. Hormones and Behavior
Experimental evidence of homing to the site of incubation 51:428–435.
by mature sockeye salmon, Oncorhynchus nerka. Animal Reiner, A., D. J. Perkel, C. V. Mello, and E. D. Jarvis. 2004.
Behaviour 72:941–949. Songbirds and the revised avian brain nomenclature. Annals
Quinn, W. G., W. A. Harris, and S. Benzer. 1974. Conditioned of the New York Academy of Sciences 1016:77–108.
behavior in Drosophila melanogaster. Proceedings of the National Reiner, A., D. J. Perkel, L. L. Bruce, A. B. Butler, A. Csillag,
Academy of Sciences 71:708–712. W. Kuenzel, L. Medina, G. Paxinos, T. Shimizu, G.
Rabenoid, K. N. 1990. Campylorhynchus wrens: The ecology Strieddter, M. Wild, G. F. Ball, S. Durand, O. Gütürkün,
of delayed dispersal and cooperation in the Venezuela D. W. Lee, C. V. Mello, A. Powers, S. A. White, G. Hough,
savanna. In Cooperative Breeding in Birds, edited by P. B. L. Kubikova, T. V. Smulders, K. Wada, J. Dugas-Ford, S.
Stacey and W. D. Koenig. Cambridge: Cambridge Husband, K. Yamamoto, J. Yu, S. Siang, and E. D. Jarvis.
University Press, pp. 159–196. 2004. Revised nomenclature for avian telencephalon and
Rado, R., N. Levi, H. Hauser, J. Witcher, N. Adler, N. some related brainstem nuclei. Journal of Comparative
Intrator, Z. Wollberg, and J. Terkel. 1987. Seismic signal- Neurology 479:377–414.
ing as a means of communication in a subterranean mam- Reiss, D., and L. Marino. 2001. Mirror self-recognition in the
mal. Animal Behaviour 35:1249–1266. bottlenose dolphin: A case of cognitive convergence.
Ramirez, J. M., and K. G. Pearson. 1990. Chemical deaf- Proceedings of the National Academy of Sciences 98:5937–5942.
ferentation of the locust flight system by phenotolamine. Remage-Healey, L., and A. H. Bass. 2006. A rapid neuro-
Journal of Comparative Physiology A 167:485–494. modulatory role for steroid hormones in the control of
Ramm, S. A., G. A. Parker, and P. Stockley. 2005. Sperm com- reproductive behavior. Brain Research 1126:27–35.
petition and the evolution of male reproductive anatomy in
Reppert, S. M., and D. R. Weaver. 2002. Coordination of cir-
rodents. Proceedings of the Royal Society of London B
cadian timing in mammals. Nature 418:935–941.
272:949–955.
Rescorla, R. A. 1988a. Behavioral studies of Pavlovian condi-
Randall, J. A., and E. R. Lewis. 1997. Seismic communication
tioning. Annual Review of Neuroscience 11:329–352.
between the burrows of kangaroo rats, Dipodomys spectabilis.
Journal of Comparative Physiology A 181:525–531. Rescorla, R. A. 1988b. Pavlovian conditioning: It’s not what
you think it is. American Psychologist 43:151–160.
Randler, C. 2006. Red squirrels (Sciurus vulgaris) respond to
alarm calls of Eurasian jays (Garrulus glandarius). Ethology Reyer, H.-U. 1980. Flexible helper structure as an ecological
112:411–416. adaptation in the red Kingfisher (Ceryle rudis rudis L.).
Ranson, E., and F. A. Beach. 1985. Effects of testosterone on Behavioral Ecology and Sociobiology 6:219–227.
ontogeny of urinary behavior in male and female dogs. Reyer, H.-U. 1984. Investment and relatedness: A cost/ben-
Hormones and Behavior 19:36–51. efit analysis of breeding and helping in the pied kingfisher
Raouf, S. A., L. C. Smith, M. B. Brown, J. C. Wingfield, and (Ceryle rudis). Animal Behaviour 32:1163–1178.
C. R. Brown. 2006. Glucocorticoid hormone levels increase Reyer, H.-U. 1986. Breeder-helper interactions in the pied
with group size and parasite load in cliff swallows. Animal kingfisher reflect costs and benefits of cooperative breeding.
Behaviour 71:39–48. Behaviour 96:277–303.
498 References

Reyer, H.-U. 1990. Pied kingfishers: ecological causes and Riede, T., and T. Fitch. 1999. Vocal tract length and acoustics
reproductive consequences of cooperative breeding. In of vocalization in the domestic dog (Canis familiaris). Journal
Cooperative Breeding in Birds, edited by P. B. Stacey and W. of Experimental Biology 202:2859–2867.
D. Koenig. Cambridge: Cambridge University Press, pp. Riedl, C. A. L., S. J. Neal, A. Robichon, J. T. Westwood, and
527–558. M. B. Sokolowski. 2005. Drosophila soluble guanylyl cyclase
Reynolds, J. D., N. B. Goodwin, and R. P. Freckleton. 2002. mutants exhibit increased foraging locomotion: behavioral
Evolutionary transitions in parental care and live bearing in and genomic investigations. Behavior Genetics 35:231–244.
vertebrates. Philosophical Transactions of the Royal Society of Riley, J. R., U. Greggers, A. D. Smith, D. R. Reynolds and R.
London B 357:269–281. Menzel. 2005. The flight paths of honeybees recruited by
Ribble, D. O. 1991. The monogamous mating system of the waggle dance. Nature 435:205–207.
Peromyscus californicus as revealed by DNA fingerprinting. Ristau, C.A., ed. 1991. Cognitive Ethology, the Minds of Other
Behavioral Ecology and Sociobiology 29:161–166. Animals. Hillsdale, NJ: Erlbaum.
Richardson, D. S., T. Burke, and J. Komdeur. 2002. Direct Ritz, T., S. Adem, and K. Schulten. 2000. A model for pho-
benefits and the evolution of female-biased cooperative toreceptor-based magnetoreception in birds. Biophysical
breeding in Seychelles warblers. Evolution 56:2313–2321. Journal 78:707–718.
Richardson, H., and N. A. M. Verbeek. 1986. Diet selection Ritz, T., R. Wiltschko, P. J. Hore, C. T. Rodgers, K. Stapput,
and optimization by Northwestern crows feeding on P. Thalau, C. R Timmel, and W. Wiltschko. 2009. Magnetic
Japanese littleneck clams. Ecology 67:1219–1226. compass of birds is based on a molecule with optimal direc-
Rickard, I. J., A. F. Russell, and V. Lummaa. 2007. Producing tional sensitivity. Biophysical Journal 96:3451-3451.
sons reduces lifetime reproductive success of subsequent Ritzmann, R. E., and A. J. Pollack. 1986. Identification of tho-
offspring in pre-industrial Finns. Proceedings of the Royal racic interneurons that mediate giant interneuron-to-
Society of London B 274:2981–2988. motor pathways. Journal of Comparative Physiology A
Ridley, M. 1978. Paternal care. Animal Behaviour 26:904–932. 159:639–654.
Riechert, S. E. 1979. Games spiders play. II. Resource assess- Rivas, J. A., and G. M. Burghardt. 2005. Snake mating sys-
ment strategies. Behavioral Ecology and Sociobiology tems, behavior, and evolution: the revisionary implications
6:121–128. of recent findings. Journal of Comparative Psychology
Riechert, S. E. 1981. The consequences of being territorial: 119:447–454.
Spiders, a case study. American Naturalist 117:871–892. Robbins, R. K. 1981. The “false head” hypothesis: Predation
Riechert, S. E. 1982. Spider interaction strategies: and wing pattern variation of Lycaenid butterflies. American
Communication vs. coercion. In Spider Communication: Naturalist 118:770–775.
Mechanisms and Ecological Significance, edited by P. N. Witt Roberts, T. S. 1907a. A Lapland longspur tragedy. Auk
and J. Rovner. Princeton, NJ: Princeton University Press, 24:369–377.
pp. 281–315. Roberts, T. S. 1907b. Supplemental note to “A Lapland
Riechert, S. E. 1986. Between population variation in spider longspur tragedy.” Auk 24:449–450.
territorial behavior: Hybrid-pure population line compar- Robinson, G. E. 1996. Chemical communication in honey-
isons. In Evolutionary Genetics of Invertebrate Behaviour: bees. Science 271:1824–1825.
Progress and Prospects, edited by M. D. Huettel. New York: Robinson, G. E. 1999. Integrative animal behaviour and
Plenum, pp. 33–42. sociogenomics. Trends in Ecology and Evolution. 14:202–205.
Riechert, S. E. 1993. Investigation of potential gene flow lim- Robinson, G. E. 2002. Sociogenomics takes flight. Science
itation of behavioral adaptation in an aridlands spider. 297:204–205.
Behavioral Ecology and Sociobiology 32:355–364. Robinson, G. E. 2004. Beyond nature and nurture. Science
Riechert, S. E., and R. F. Hall. 2000. Local population suc- 304:397–399.
cess in heterogenous habitats: reciprocal transplant exper- Robinson, G. E., R. D. Fernald, and D. F. Clayton. 2008.
iments completed on a desert spider. Journal of Evolutionary Genes and social behavior. Science 322:896–900.
Biology 13:541–550. Robinson, G. E., C. M. Grozinger, and C. W. Whitfield. 2005.
Riechert, S. E., and Hedrick, A. V. 1990. Levels of predation Sociogenomics: Social life in molecular terms. Nature
and genetically based anti-predatory behavior in the spider, Reviews Genetics 6:257–270.
Agelenopsis aperta. Animal Behaviour 40:679–687. Rodda, G. H. 1984. The orientation and navigation of juve-
Riechert, S. E., and J. Maynard Smith. 1989. Genetic analy- nile alligators: evidence of magnetic sensitivity. Journal of
ses of two behaivoural traits linked to individual fitness in Comparative Physiology A 154:649–658.
the desert spider, Agelenopsis aperta. Animal Behaviour Rodl, T., S. Berger, and L. M. Romero. 2007. Tameness and
37:624–637. stress physiology in a predator-naive island species con-
Riechert, S. E., and T. R. Tracy. 1975. Thermal balance and fronted with novel predation threat. Proceedings of the Royal
prey availability: Bases for a model relating web-site char- Society of London B 274:577–582.
acteristics to spider reproductive success. Ecology Roeder, K. D. 1967. Nerve Cells and Insect Behavior.
56:265–285. Cambridge, MA.: Harvard University Press.
 References 499

Roeder, K. D., and R. S. Payne. 1966. Acoustic orientation of Røskaft, E., T. Järvi, T., M. Bakken, C. Beeh, and R. E.
a moth in flight by means of two sense cells. Symposia of the Reinersten. 1986. The relationship between social status
Society for Experimental Biology 20:251–272. and resting metabolic rate in great tits (Parus major) and
Roeder, K. D., and A. E. Treat. 1957. Ultrasonic reception by pied flycatchers (Ficedula hypoleuca). Animal Behaviour
the tympanic organ of noctuid moths. Journal of 34:838–842.
Experimental Biology 134:127–157. Ross, K. G. 1986. Kin selection and the problem of sperm uti-
Roeder, K. D., and A. E. Treat. 1961. The detection and eva- lization in social insects. Nature 323:799–800.
sion of bats by moths. American Scientist 49:135–148. Rothstein, S. I., and S. K. Robinson. (eds.) 1998. Parasitic Birds
Rodgers, C. T., and P. J. Hore. 2009. Chemical magnetore- and Their Hosts: Studies in Coevolution. Oxford: Oxford
ception in birds: The radical pair mechanism. Proceedings of University Press.
the National Academy of Sciences 106:353-360. Roulin, A. 2002. Why do lactating females nurse alien off-
Romanes, G. J. 1882. Animal Intelligence. New York: D. spring? A review of hypotheses and empirical evidence.
Appleton and Co. Animal Behaviour 63:201–208.
Romanes, G. J. 1884. Mental Evolution in Animals. London: Rousset, F., and D. Roze. 2007. Constraints on the origin and
Keegan, Paul, Trench and Company. maintenance of genetic kin recognition. Evolution
Romanes, G. J. 1889. Mental Evolution in Man. New York: D. 61:2320–2330.
Appleton and Company. Rowell, J.T., S. P. Ellner, and H. K. Reeve. 2006. Why ani-
Romey, W. L. 1995. Position preferences within groups—do mals lie: How dishonesty and belief can coexist in a signal-
whirligigs select positions which balance feeding opportu- ing system. American Naturalist 168:E180–E204.
nities with predator avoidance? Behavioral Ecology and Rowley, I. C. R. 1981. The communal way of life in the splen-
Sociobiology 37:195–200. did wren, Malurus splendens. Zeitschrift für Tierpsychologie
Romey, W. L., A. R. Walston, and P. J. Watt. 2008. Do 3-D 55:228–267.
predators attack the margin of 2-D selfish herds? Behavioral
Rowley, I. C. R., and E. M. Russell. 1990. Splendid fairy-
Ecology 19:74–78.
wrens: Demonstrating the importance of longevity. In
Ronacher, B. 2008. Path integration as the basic navigation Cooperative Breeding in Birds, edited by P. B. Stacey and
mechanism of the desert ant Cataglyphis fortis (Forel 1902)
W. D. Koenig. Cambridge: Cambridge University Press,
(Hymenoptera: Formicidae). Myrmecological News 11:53–62.
pp. 3–30.
Rönn, J., M. Katvala, and G. Arnqvist. 2007. Coevolution
Rozhok, A. 2008. Orientation and Navigation in Vertebrates.
between harmful male genitalia and female resistance in
Berlin: Springer.
seed beetles. Proceedings of the National Academy of Sciences
Rubolini, D., P. Galeotti, F. Pupin, R. Sacchi, P. A. Nardi, and
104:10921–10925.
M. Fasola. 2007. Repeated matings and sperm depletion in
Rood, J. P. 1990. Group size, survival, reproduction, and
the freshwater crayfish Austropotamobius italicus. Freshwater
routes to breeding in dwarf mongooses. Animal Behaviour
Biology 52:1898–1906.
39:566–572.
Rumbaugh, D. M. 1977. The emergence and the state of ape
Rooney, J., and S. M. Lewis. 2002. Fitness advantage from
language research. In Progress in Ape Research, edited by G.
nuptial gifts in female fireflies. Ecological Entomology
H. Bourne. New York: Academic Press, pp. 75–83.
27:373–377.
Rosa-Molinar, E., B. Fritzsch, and S. E. Hendricks. 1996. Rumbaugh, D. M., and E. S. Savage-Rumbaugh. 1994.
Organizational-activational concept revisited: sexual differ- Language in comparative perspective. In Animal Learning
entiation in an Atherinomorph teleost. Hormones and and Cognition, edited by N. J. Macintosh. San Diego, CA:
Behavior 30:563–575. Academic Press, pp. 307–333.
Rose, J. D., and F. L. Moore. 1999. A neurobehavioral model Rumbaugh, D. M., T. V. Gill, and E. C. von Glaserfeld. 1973.
for rapid actions of corticosterone on sensory integration. Reading and sentence completion by a chimpanzee. Science
Steroids 64:92–99. 182:731–733.
Rose, J. D., and F. L. Moore. 2002. Behavioral neuroen- Rüppell, O., T. Pankiw, and R. E. Page, Jr. 2004. Pleiotropy,
docrinology of vasotocin and vasopressin and the sensori- epistasis, and new QTL: The genetic architecture of honey
motor processing hypothesis. Frontiers in Neuroendocrinology bee foraging behavior. Journal of Heredity 96:481–491.
23:317–341. Rüppell, O., S. B. C. Chandra, T. Pankiw, M. K. Fondrk, M.
Rosen, R. A., and D. C. Hales. 1981. Feeding of paddlefish, Beye, G. Hunt, and R. E. Page. 2006. The genetic archi-
Polyodon spathula. Copeia 2:441–455. tecture of sucrose responsiveness in the honeybee (Apis mel-
Rosenheim, J. A. 1993. Comparative and experimental lifera L.). Genetics 172:243–251.
approaches to understanding insect learning. In Insect Russell, A. F., and B. J. Hatchwell. 2001. Experimental evi-
Learning: Ecological and Evolutionary Perspectives, edited by D. dence for kin-biased helping in a cooperatively breeding
R. Papaj and A. C. Lewis. New York: Chapman and Hall, vertebrate. Proceedings of the Royal Society of London B
pp. 273–307. 268:95–99.
500 References

Russell, E. M., and I. C. R. Rowley. 1988. Helper contribu- Sakamoto, T., and S. D. McCormick. 2006. Prolactin and
tions to reproductive success in the splendid fairy-wren growth hormone in fish osmoregulation. General and
Malurus splendens. Behavioral Ecology and Sociobiology Comparative Endocrinology 147:24–30.
22:131–140. Sakurai, T., T. Nakagawa, H. Mitsuno, H. Mori, Y. Endo, S.
Russell, A. F., A. J. Young, G. Spong, N. R. Jordan, and T. H. Tanoue, Y. Yasukochi, K. Touhara, and T. Nishioka. 2004.
Clutton-Brock. 2007. Helpers increase the reproductive Identification and functional characterization of a sex
potential of offspring in cooperative meerkats. Proceedings pheromone receptor in the silkmoth Bombyx mori.
of the Royal Society of London B 274:513–520. Proceedings of the National Academy of Sciences
Rutz, C., L. A. Bluff, A. A. Weir, and A. Kacelnik. 2007. Video 101:16653–16658.
cameras on wild birds. Science Express Brevia, published Sands, J., and S. Creel. 2004. Social dominance, aggression
online October 2007. and faecal glucocorticoid levels in a wild population of
Ruxton, G. D., M. P. Speed, and D. J. Kelly. 2004. What, if wolves, Canis lupus. Animal Behaviour 67:382–396.
anything, is the adaptive function of countershading? Sapolsky, R. M. 1992. Cortisol concentrations and the social
Animal Behaviour 68:445–451. significance of rank instability among wild baboons.
Ryan, M. J. 1988. Energy, calling, and selection. American Psychoneuroendocrinology 17:701–709.
Zoologist 28:885–898. Sapolsky, R. M. 1994. Why Zebras Don’t Get Ulcers. New York:
Freeman and Company.
Ryan, M. J. 1990. Sexual selection, sensory systems, and sen-
sory exploitation. Oxford Survey of Evolutionary Biology Sapolsky, R. M., L. M. Romero, and A.U. Munck. 2000. How
7:157–195. do glucocorticoids influence stress responses? Integrating
permissive, suppressive, stimulatory, and preparative
Ryan, M. J. 2005. The evolution of behaviour, and integrat-
actions. Endocrine Reviews 21:55–89.
ing it towards a complete and correct understanding of
Saranathan, V., D. Hamilton, G. V. N. Powell, D. E.
behavioural biology. Animal Biology 55:419–439.
Kroodsma, and R. O. Prum. 2007. Genetic evidence sup-
Ryan, M. J. 2007. Sensory ecology: see me, hear me. Current
ports song learning in the three-wattled bellbird Procnias tri-
Biology 17:R1019–R1021.
carunculata (Cotingidae). Molecular Ecology 16:3689–3702.
Ryan, B. C., and J. G. Vandenbergh. 2002. Intrauterine
Sato, T., and S. Goshima. 2006. Impacts of male-only fishing
position effects. Neuroscience and Biobehavioral Reviews
and sperm limitation in manipulated populations of an
26:665–678.
unfished crab, Hapalogaster dentata. Marine Ecology Progress
Ryan, M. J., and A. S. Rand. 1993a. Phylogenetic patterns of Series 313:193–204.
behavioral mate recognition systems in the Physalaemus pus- Sato, T., and S. Goshima. 2007. Female choice in response to
tulosus species group (Anura: Leptodactylidae): The role of risk of sperm limitation by the stone crab, Hapalogaster den-
ancestral and derived characters and sensory exploitation. tata. Animal Behaviour 73:331–338.
In Evolutionary Patterns and Processes, edited by D. R. Lees Sauer, E. G. F. 1957. Die Sternorientierung nächtlich ziehen-
and D. Edwards. London: Academic Press, pp. 251–267. der Grasmücken (Sylvia atriciapilla, borin and curruca).
Ryan, M. J., and A. S. Rand. 1993b. Sexual selection and sig- Zeitschrift für Tierpsychologie 14:29–70.
nal evolution: The ghost of biases past. Philosophical Sauer, E. G. F. 1961. Further studies on the stellar orienta-
Transactions of the Royal Society of London B 340:187–195. tion of nocturnally migrating birds. Psychologische Forschung
Ryan, M. J., Tuttle, M. D., and A. S. Rand. 1982. Bat preda- 26:224–244.
tion and sexual advertisement in a neotropical anuran. Sauer, E. G. F., and E. M. Sauer. 1960. Star navigation in noc-
American Naturalist 119:136–139. turnal migrating birds. The 1958 planetarium experiments.
Ryder, T. B., J. G. Blake, and B. A. Loiselle. 2006. A test of Cold Spring Harbor Symposium of Quantitative Biology
the environmental hotspot hypothesis for lek placement in 25:463–473.
three species of manakins (Pipridae) in Ecuador. The Auk Savage-Rumbaugh, E. S. 1986. Ape Language from Conditioned
123:247–258. Response to Symbol. New York: Columbia University Press.
Sack, R. L., R. W. Brandes, A. R. Kendall, and A. Lowy. 2000. Savage-Rumbaugh, E. S., W. M. Fields, and J. P. Taglialatela.
Entrainment of free-running circadian rhythms by mela- 2001. Language, speech, tools and writing: a cultural imper-
tonin in blind people. New England Journal of Medicine ative. Journal of Consciousness Studies 8:273–292.
343:1070–1077. Savage-Rumbaugh, E. S., and R. Lewin. 1994. Kanzi, the Ape
Saether, S. A., P. Fiske, and J. A. Kalas. 2001. Male mate at the Brink of the Human Mind. New York: John Wiley.
choice, sexual conflict and strategic allocation of copulations Savage-Rumbaugh, E. S., D. M. Rumbaugh, and S. Boysen.
in a lekking bird. Proceedings of the Royal Society of London B 1978a. Symbolic communication between two chimpanzees
268:2097–2102. (Pan troglodytes). Science 201:641–644.
Sahara, K., and Y. Takemura. 2003. Application of artificial Savage-Rumbaugh, E. S., D. M. Rumbaugh, and S. Boysen.
insemination technique to eupyrene and/or apyrene sperm 1978b. Linguistically mediated tool use and exchange by
in Bombyx mori. Journal of Experimental Zoology Part A – chimpanzees Pan troglodytes. Behavioral and Brain Sciences
Comparative Experimental Biology 297A:196–200. 1:539–554.
 References 501

Savage-Rumbaugh, E. S., D. M. Rumbaugh, and S. Boysen. Schoech, S. J., R. L. Mumme, and J. C. Wingfield. 1996.
1980. Do apes use language? American Scientist 68:49–61. Prolactin and helping behaviour in the cooperatively
Savage-Rumbaugh, E. S., S. G. Shanker, and T. J. Taylor. breeding Florida scrub-jay, Aphelocoma coerulescens. Animal
1998. Apes, Language, and the Human Mind. New York: Behaviour 52:445–456.
Oxford University Press. Schoener, T. W. 1968. Sizes of feeding territories among birds.
Sazima, I., L. N. Carvalho, F. P. Mendonca, and J. Zuanon. Ecology 49:123–141.
2006. Fallen leaves on the water-bed: diurnal camouflage of Schoener, T. W. 1983. Simple models of optimal feeding-ter-
three night active fish species in an Amazonian streamlet. ritory size: A reconciliation. American Naturalist
Neotropical Ichthyology 4:119–122. 121:608–629.
Schaal, B., G. Coureaud, D. Langlois, C. Ginies, E. Sémon, Schuett, G. W. 1997. Body size and agnostic experience affect
and G. Perrier. 2003. The mammary pheromone of the rab- dominance and mating success in male copperheads. Animal
bit: chemical and behavioural characterisation. Nature
Behaviour 54:213–224.
424:68–72.
Schütz, D., and M. Taborsky. 2005. The influence of sexual
Schaefer, P., G. V. Kondagunta, and R. E. Ritzman. 1994.
selection and ecological constraints on an extreme sexual
Motion analysis of escape movements evoked by tactile
size dimorphism in a cichlid. Animal Behaviour 70:539–549.
stimulation in the cockroach Periplaneta americana. Journal
of Experimental Biology 190:287–294. Schutz, von F. 1965. Sexuelle Prägung bei Anatiden. Zeitschrift
für Tierpsychologie 22:50–103.
Schaller, G. B. 1972. The Serengeti Lion. Chicago: University
of Chicago Press. Schwagmeyer, P., D. W. Mock, and G. A. Parker. 2002.
Schanen, N. C. 2006. Epigenetics of autism spectrum disor- Biparental care in house sparrows: negotiation or sealed
ders. Human Molecular Genetics 15:R138–R150. bid? Behavioral Ecology 13:713–721.
Scheich, H., G. Langner, C. Tidemann, R. B. Coles, and A. Schwind, R. 1983. Zonation of the optical environment and
Guppy. 1986. Electroreception and electrolocation in zonation in the rhabdom structure within the eye of the
platypus. Nature 319:401–402. backswimmer, Notonecta glauca. Cell Tissue Research
Scheiner, R., M. B. Sokolowski, and J. Erber. 2004. Activity 232:53–62.
of cGMP-dependent protein kinase (PKG) affects sucrose Schwind, R. 1991. Polarization vision in water insects and
responsiveness and habituation. Learning and Memory insects living on a moist substrate. Journal of Comparative
11:303–311. Physiology A 169:531–540.
Schleidt, W. 1961a. Reaktionen von Truthühnern auf Scott, J. P., and J. L. Fuller. 1965. Genetics and the Social
fliegende Raubvögel und Veruche zur Analyse ihrer AAM’s. Behavior of the Dog. Chicago: University of Chicago Press.
Zeitschrift für Tierpsychologie 18:534–560. Sear, R., and R. Mace. 2008. Who keeps children alive? A
Schleidt, W. 1961b. Über die Auslösung des Kollern beim review of the effects of kin on child survival. Evolution and
Truthahan (Meleagris galopavo). Zeitschrift für Tierpsychologie Human Behavior 29:1–18.
11:417–435. Searcy, W. A. 1979. Female choice of mates: A general model
Schluter, A., J. Parzefall, and I. Schlupp. 1998. Female pref- for birds and its application to red-winged blackbirds
erence for symmetrical vertical bars in male sailfin mollies. (Agelaius phoeniceus). American Naturalist 114:77–100.
Animal Behaviour 56:147–153.
Searcy, W. A., and S. Nowicki. 2005. The Evolution of Animal
Schmidt, V., H. M. Schaefer, and H. Winkler. 2004. Communication. Princeton, NJ: Princeton University Press.
Conspicuousness, not colour as foraging cue in plant ani-
Sebbel, P., H. Düttmann, and T. Groothius. 1998. Influence
mal signalling. Oikos 106:551–557.
of comfort and social stimuli on a comfort movement and
Schmidt-Koenig, K., and H. J. Schlichte. 1972. Homing in
a display derived from it. Animal Behaviour 55:129–137.
pigeons with impaired vision. Proceedings of the National
Sefc, K. M., K. Mattersdorfer, C. Sturmbauer, and S.
Academy of Sciences 69:2446–2447.
Koblmüller. 2008. High frequency of multiple paternity in
Schneider, D. 1974. The sex-attractant receptor of moths.
broods of a socially monogamous cichlid fish with biparental
Scientific American 231:28–35.
care. Molecular Ecology 17:2531–2543.
Schneider, J. S., M. K. Stone, K. E. Wynne-Edwards, T. H.
Horton, J. Lydon, B. O’Malley, and J. E. Levine. 2003. Selonen, V., and I. K. Hanski. 2006. Habitat exploration and
Progesterone receptors mediate male aggression toward use in dispersing juvenile flying squirrels. Journal of Animal
infants. Proceedings of the National Academy of Sciences Ecology 75:1440–1449.
100:2951–2956. Semm, P., and R. C. Beason. 1990. Responses to small mag-
Schoech, S. J. 1998. Physiology of helping in Florida scrub netic variations by the trigeminal system of the bobolink.
jays. American Scientist 86:70–77. Brain Research Bulletin 25:735–740.
Schoech, S. J., R. Bowman, and S. J. Reynolds. 2004. Food Seney, B., D. Goldman, and N. G. Forger. 2006. Breeding sta-
supplementation and possible mechanisms underlying early tus affects motoneuron number and muscle size in naked
breeding in the Florida scrub-jay (Aphelocoma coerulescens). mole-rats: recruitment of perineal motoneurons? Journal of
Hormones and Behavior 46:565–573. Neurobiology 66:1354–1364.
502 References

Serventy, D. L. 1971. Biology of desert birds. In Avian Biology, Sherman, P. W., H. K. Reeve, and D. W. Pfennig. 1997.
edited by D. S. Farner, J. R. King, and K. C. Parkes. New Recognition systems. In Behavioural Ecology, edited by J. R.
York: Academic Press, pp. 287–339. Krebs and N. B. Davies. Oxford: Blackwell Science, pp.
Seyfarth, R. M., and D. L. Cheney. 2003. Signalers and 69–96.
receivers in animal communication. Annual Review of Sherman, P. W., E. A. Lacey, H. K. Reeve, and L. Keller. 1995.
Psychology 54:145–173. The eusociality continuum. Behavioral Ecology 6:102–108.
Seyfarth, R. M., D. L. Cheney, and P. Marler. 1980. Monkey Sherry, D. F. 2005a. Do ideas about function help in the study
responses to three different alarm calls: evidence of preda- of causation? Animal Biology 55:441–456.
tor classification and semantic communication. Science Sherry, D. F. 2005b. Brain and behavior. In The Behavior of
210:801–803. Animals: Mechanisms, Function, and Evolution, edited by J. J.
Shaffer, D. M., S. M. Krauchunas, M. Eddy, and M. K. Bolhuis and L.-A. Giraldeau. Malden, MA: Blackwell
McBeath. 2004. How dogs navigate to catch frisbees. Publishing.
Psychological Science 15:437–441. Sherry, D. F., A. L. Vaccarino, K. Buckenham, and R. Herz.
Sharma, V. K. 2003. Adaptive significance of circadian clocks. 1989. The hippocampal complex of food storing birds.
Chronobiology International 20:901–919. Brain, Behavior and Evolution 34:308–317.
Sharpe, F. A., and L. M. Dill. 1997. The behavior of Pacific Shettleworth, S. J. 1995. Comparative studies of memory in
herring schools in response to artificial humpback whale food storing birds: From the field to the Skinner box. In
bubbles. Canadian Journal of Zoology 75:725–730. Behavioral Brain Research in Naturalistic and Semi-Naturalistic
Sheldon, B. C. 2002. Relating paternity to paternal care. Settings, NATO ASI Series, edited by E. Alleva, A. Fasolo,
Philosophical Transactions of the Royal Society of London B H. P. Lipp, L. Nadel, and L. Ricceri. Dordrecht: Kluwer,
357:341–350. pp. 159–192.
Shelley, E. L., and D. T. Blumstein. 2005. The evolution of Shettleworth, S. J. 1998. Cognition, Evolution and Behavior.
vocal alarm communication in rodents. Behavioral Ecology Oxford: Oxford University Press.
16:169–177.
Shields, W. M. 1982. Philopatry, Inbreeding, and the Evolution
Shen, J. X., A. S. Feng, Z. M. Xu, Z. L. Yu, V. S. Arch, X. J.
of Sex. Albany: State University of New York Press.
Yu, and P. M. Narins. 2008. Ultrasonic frogs show hypera-
Shields, W. M. 1984. Barn swallow mobbing: Self-defence,
cute phonotaxis to female courtship calls. Nature
collateral kin defence, group defence, or parental care?
453:914–917.
Animal Behaviour 32:132–148.
Sherbrooke, W. C., and J. R. Mason. 2005. Sensory modality
Shields, W. M. 1987. Dispersal and mating systems:
used by coyotes in responding to antipredator compounds
Investigating their causal connections. In Mammalian
in the blood of Texas horned lizards. Southwest Naturalist
Dispersal Patterns, edited by B. D. Chepko-Sade and Z. T.
50:216–222.
Halpin. Chicago: University of Chicago Press, pp. 3–24.
Sherbrooke, W. C., and G. A. Middendorf, III. 2004.
Responses of kit foxes (Vulpes macrotis) to antipredator Shoji, T., Y. Yamamoto, D. Nishikawa, K. Kurihara, and H.
blood-squirting and blood of Texas horned lizards Ueda. 2003. Amino acids in stream water are essential for
(Phrynosoma cornutum). Copeia 3:652–658. salmon homing migration. Fish Physiology and Biochemistry
Sheridan, M., and R. H. Tamarin. 1988. Space use, longevity, 28:249–251.
and reproductive success in meadow voles. Behavioral Shoji, T., M. Ueda, T. Sakamoto, Y. Katsuragi, Y. Zohar, A.
Ecology and Sociobiology 22:85–90. Urano, and K. Yamauchi. 2000. Amino acids dissolved in
Sherman, P. M. 1994. The orb-web: An energetic and behav- stream water as possible home stream odorant for masu
ioural estimator of a spider’s dynamic foraging and repro- salmon. Chemical Senses 25:533–540.
ductive strategies. Animal Behaviour 48:19–34. Shors, T. J., G. Miesegaes, A. Beylin, M. Zhao, T. Rydel, and
Sherman, P. W. 1977. Nepotism and the evolution of alarm E. Gould. 2001. Neurogenesis in the adult is involved in the
calls. Science 197:1246–1253. formation of trace memories. Nature 410:372–376.
Sherman, P. W. 1980a. The limits of ground squirrel nepo- Siegel, R. G., and W. K. Honig. 1970. Pigeon concept for-
tism. In Sociobiology: Beyond Nature/Nurture? edited by G. W. mation: Successive and simultaneous acquisition. Journal of
Barlow and J. Silverberg. Boulder, CO: Westview, pp. the Experimental Analysis of Behavior 13:385–390.
505–544. Sih, A. 2004. Behavioral syndromes: an integrated overview.
Sherman, P. W. 1980b. The meaning of nepotism. American Quarterly Review of Biology 79:241–277.
Naturalist 116:604–606. Sih, A., and B. Christensen. 2001. Optimal diet theory: when
Sherman, P. W. 1985. Alarm calls of Belding’s ground squir- does it work, and when and why does it fail? Animal
rels to aerial predators: Nepotism or self preservation? Behaviour 61:379–390.
Behavioral Ecology and Sociobiology 17:313–323. Silberglied, R. E., J. G. Shepherd, and J. L. Dickinson. 1984.
Sherman, P. W., J. U. M. Jarvis, and S. H. Braude. 1992. Eunuchs: The role of apyrene sperm in Lepidoptera.
Naked mole-rats. Scientific American 267:72–78. American Naturalist 123:255–265.
 References 503

Siljander, E., R. Gries, G. Khaskin, and G. Gries. 2008. Slagsvold, T., B. T. Hansen, L. E. Johannessen, and L. T.
Identification of the airborne aggregation pheromone of the Lifjeld. 2002. Mate choice and imprinting in birds studied
common bed bug, Cimex lectularius. Journal of Chemical by cross-fostering in the wild. Proceedings of the Royal Society
Ecology 34:708–718. of London B 269:1449–1455.
Sillett, T. S., and R. T. Holmes. 2002. Variation in survivor- Slocombe, K. E., and K. Zuberbühler. 2005. Functionally ref-
ship of a migratory songbird throughout its annual cycle. erential communication in a chimpanzee. Current Biology
Journal of Animal Ecology 71:296–308. 15:1779–1784.
Silver, R., J. LeSauter, P. A. Tresco, and M. N. Lehman. 1996. Smale, L., S. Nunes, and K. E. Holekamp. 1997. Sexually
A diffusible coupling signal from the transplanted suprachi- dimorphic dispersal in mammals: Patterns, causes, and con-
asmatic nucleus controlling circadian locomotor rhythms. sequences. Advances in the Study of Behavior 26:181–250.
Nature 382:810–813.
Smedley, S. R., and T. Eisner. 1996. Sodium: A male moth’s
Simmons, L. W. 1988. Male size, mating potential and life-
gift to its offspring. Proceedings of the National Academy of
time reproductive success in the field cricket, Gryllus bimac-
Sciences 93:809–813.
ulatus (De Geer). Animal Behaviour 36:372–379.
Smiseth, P. T., and A. J. Moore. 2004. Behavioural dynamics
Simmons, L. W. 2005. The evolution of polyandry: Sperm
between caring males and females in a beetle with faculta-
competition, sperm selection, and offspring viability. Annual
Review of Ecology and Systematics 36:125–146. tive biparental care. Behavioral Ecology 15:621–628.
Simpson, S. J., R. M. Sibly, K. P. Lee, S. T. Behmer, and D. Smith, M. D., and C. J. Conway. 2007. Use of mammal
Raubenheimer. 2004. Optimal foraging when regulating manure by nesting burrowing owls: a test of four hypothe-
intake of multiple nutrients. Animal Behaviour ses. Animal Behaviour 73:65–73.
68:1299–1311. Smith, T. 2006. Individual olfactory signatures in common
Sinclair, A. R. E. 1983. The function of distance movements marmosets (Callithrix jacchus). American Journal of
in vertebrates. In The Ecology of Animal Movement, edited by Primatology 68:585–604.
I. R. Swingland and P. J. Greenwood. Oxford: Clarendon, Smith, V. A., A. P. King, and M. J. West. 2000. A role of her
pp. 240–258. own: Female cowbirds, Molothrus ater, influence the devel-
Sinervo, B., and C. M. Lively. 1996. The rock-paper-scissors opment and outcome of song learning. Animal Behaviour
game and the evolution of alternative male strategies. 60:599–609.
Nature 380:240–243. Smith, W. J. 1977. The Behavior of Communicating, An
Singer, M. S., and E. A. Bernays. 2003. Understanding Ethological Approach. Cambridge: Harvard University Press.
omnivory needs a behavioral perspective. Ecology Smith, W. J. 1991. Animal communication and the study of
84:2532–2537. cognition. In Cognitive Ethology, the Minds of Other Animals,
Skinner, B. F. 1953. Science and Human Behavior. New York: edited by C. A. Ristau. Hillsdale, NJ: Erlbaum, pp.
Free Press, Macmillan. 209–230.
Skipper, R., and G. Skipper. 1957. Those British-bred pom- Sneddon, L. U., F. A. Huntingford, and A. C. Taylor. 1997.
padours—The story completed. Water Life Aquarium World Weapon size versus body size as a predictor of winning in
12:63–64.
fights between shore crabs, Carcinus maenus (L.). Behavioral
Skow, C. D., and E. M. Jakob. 2006. Jumping spiders attend Ecology and Sociobiology 41:237–242.
to context during learned avoidance of aposematic prey.
Snow, B. K. 1974. Lek behavior and breeding of Guy’s her-
Behavioral Ecology 17:34–40.
mit hummingbird Phaethornis guy. Ibis 116:278–297.
Skutch, A. F. 1935. Helpers at the nest. Auk 52:257–273.
Snow, D. W. 1963. The evolution of manakin displays. In
Skutch, A. F. 1987. Helpers at Birds’ Nests. Iowa City:
Proceedings of the 13th International Ornithological Congress
University of Iowa Press.
1:553–561.
Skutelsky, O. 1996. Predation risk and state-dependent for-
Snyder, W. E., G. M. Clevenger, and S. D. Eigenbrode. 2004.
aging in scorpions: effects of moonlight on foraging in the
scorpion Buthus occitanus. Animal Behaviour 52:49–57. Intraguild predation and successful invasion by introduced
ladybird beetles. Oecologia 140:559–565.
Slabbekoorn, H., and A. den Boer-Visser. 2006. Cities change
the songs of birds. Current Biology 16:2326–2331. Sogabe, A., and Y. Yanagisawa. 2007. The function of daily
Slagsvold, T. 2004. Cross-fostering of pied flycatchers greetings in a monogamous pipefish Corythoichthys
(Ficedula hypoleuca) to heterospecific hosts in the wild: a haematopterus. Journal of Fish Biology 71:585–595.
study of sexual imprinting. Behaviour 141:1079–1102. Sogabe, A., and Y. Yanagisawa. 2008. Maintenance of pair
Slagsvold, T., and J. T. Lifjeld. 1988. Ultimate adjustment of bond during the non-reproductive season in a monogamous
clutch size to parent feeding capacity in a passerine bird. pipefish Corythoichthys haematopterus. Journal of Ethology
Ecology 69:1918–1922. 26:195–199.
Slagsvold, T., and J. T. Lifjeld. 1990. Influence of male and Sokolowski, M. B. 1980. Foraging strategies of Drosophila
female quality on clutch size in tits (Parus spp.). Ecology melanogaster: a chromosomal analysis. Behavior Genetics
71:1258–1266. 10:291–302.
504 References

Sokolowski, M. B., H. S. Pereira, and K. Hughes. 1997. Stallcup, J. A., and G. E. Woolfenden. 1978. Family status and
Evolution of foraging behavior in Drosophila by density- contributions to breeding by Florida scrub jays. Animal
dependent selection. Proceedings of the National Academy of Behaviour 26:1144–1156.
Sciences 94:7373–7377. Stamps, J. A. 1987. Conspecifics as cues to territory quality:
Soler, M., M. Martín-Vivaldi, J. M. Marín, and A. P. Møller. A preference of juvenile lizards (Anolis aeneus) for previously
1999. Weight lifting and health status in the black wheatear. used territories. American Naturalist 129:629–642.
Behavioral Ecology 10:281–286. Stamps, J. A. 1988. Conspecific attraction and aggregation in
Soler, M., J .J. Soler, A. P. Møller, J. Moreno, and M. Lindén. territorial species. American Naturalist 131:329–347.
1996. The functional significance of sexual display: stone Stamps, J. A. 1991. Why evolutionary issues are reviving inter-
carrying in the black wheatear. Animal Behaviour est in proximate behavioral mechanisms. American Zoologist
51:247–254. 31:338–348.
Solomon, N. G. 2003. A reexamination of factors influencing Stamps, J. A. 1994. Territorial behavior—testing the assump-
philopatry in rodents. Journal of Mammalogy 84:1182–1197. tions. Advances in the Study of Behavior 23:173–232.
Solomon, N. G., and B. Keane. 2007. Reproductive strategies in Stamps, J. A. 1995. Motor learning and the value of familiar
female rodents. In Rodent Societies: An Ecological and Evolutionary space. American Naturalist 146:41–58.
Perspective, edited by J. O. Wolff, and P. W. Sherman.
Stamps, J. A. 2001. Habitat selection by dispersers: integrat-
Chicago: The University of Chicago Press, pp. 42–56.
ing proximate and ultimate approaches. In Dispersal, edited
Soma, K. K. 2006. Testosterone and aggression: Berthold,
by J. Clobert, E. Danchin, A. A. Dhondt, and J. D. Nichols.
birds, and beyond. Journal of Neuroendocrinology 18:543–551.
Oxford: Oxford University Press, pp. 230–242.
Sommer, V. 1993. Infanticide among the langurs of Jodhpur:
Stamps, J. A., and J. M. Davis. 2006. Adaptive effects of natal
Testing the sexual selection hypothesis. In Infanticide and
experience on habitat selection by dispersers. Animal
Parental Care, edited by S. Parmigiani and F. vom Saal.
Behaviour 72:1279–1289.
London: Harwood Academic Press, pp. 155–198.
Stamps, J. A., and V. V. Krishnan. 2001. How territorial ani-
Sorenson, M. D., and R. B. Payne. 2002. Molecular genetic
mals compete for divisible space: A learning based model
perspectives on avian brood parasitism. Integrative and
with unequal competitors. American Naturalist 157:154–169.
Comparative Biology 42:388–400.
Stamps, J. A., and R. R. Swaisgood. 2007. Someplace like
Speed, M. P., and G. D. Ruxton. 2007. How bright and how
home: Experience, habitat selection, and conservation biol-
nasty: explaining diversity in warning signal strength.
ogy. Applied Animal Behaviour Science 102:392–409.
Evolution 61:623–635.
Stankowich, T., and R. G. Coss. 2007a. Effects of risk assess-
Speed, M. P., and J. R. G. Turner. 1999. Learning and mem-
ory in mimicry: Do we understand the mimicry spectrum? ment, predator behavior, and habitat on escape behavior in
Biological Journal of the Linnean Society 67:281–312. Columbian black-tailed deer. Behavioral Ecology 18:358–367.
Spencer, H. 1855. Principles of Psychology. New York: D. Stankowich, T., and R. G. Coss. 2007b. The re-emergence of
Appleton and Company. felid camouflage with the decay of predator recognition in
Spencer, K. A., K. L. Buchanan, S. Leitner, A. R. Goldsmith, deer under relaxed selection. Proceedings of the Royal Society
and C. K. Catchpole. 2005. Parasites affect song complex- of London B 274:175–182.
ity and neural development in a songbird. Proceedings of the Stephens, D. W., and J. R. Krebs. 1986. Foraging Theory.
Royal Society of London B 272:2037–2043. Princeton, NJ: Princeton University Press.
Spencer, K. A., S. Harris, P. J. Baker, and I. C. Cuthill. 2007. Stephens, D. W., and E. L. Charnov. 1982. Optimal foraging:
Song development in birds: the role of early experience and Some simple stochastic models. Behavioral Ecology and
its potential effect on rehabilitation success. Animal Welfare Sociobiology 10:251–263.
16:1–13. Stephens, M. L. 1984. Interspecific aggressive behavior of the
Spritzer, M. D., N. G. Solomon, and D. B. Meikle. 2006. polyandrous northern jacana (Jacana spinosa). Auk
Social dominance among male meadow voles is inversely 101:508–518.
related to reproductive success. Ethology 112:1027–1037. Stevens, M. 2007. Predator perception and the interrelation
Srinivasan, M., and S. W. Zhang. 2003. Small brains, smart between different forms of protective coloration. Proceedings
minds: Vision, perception, and “cognition” in honeybees. of the Royal Society of London B 274:1457–1464.
IETE Journal of Research 49:127–134. Stewart, I. J., S. M. Carlson, C. P. Boatright, G. B. Buck, and
Srinivasan, M. V., S. Zhang, M. Altwein, and J. Tautz. 2000. T.P. Quinn. 2004. Site fidelity of spawning sockeye salmon
Honeybee navigation: Nature and calibration of the (Oncorhynchus nerka W.) in the presence and absence of
“odometer.” Science 287:851–853. olfactory cues. Ecology of Freshwater Fish 13:104–110.
Stacey, P. B. 1979. Habitat saturation and communal breeding Stoddard, P., H. Zakon, M. Markham, and M. McAnelly.
in the acorn woodpecker. Animal Behaviour 27:1153–1166. 2006. Regulation and modulation of electric waveforms in
Stacey, P. B., and C. E. Bock. 1978. Social plasticity in the gymnotiform electric fish. Journal of Comparative Physiology
acorn woodpecker. Science 202:1298–1300. A 153:477–487.
 References 505

Stoltz, J. A., and B. D. Neff. 2006. Male size and mating tac- Swaisgood, R. R. 2007. Current status and future directions
tic influence proximity to females during sperm competi- of applied behavioral research for animal welfare and con-
tion in bluegill sunfish. Behavioral Ecology and Sociobiology servation. Applied Animal Behaviour Science 102:139–162.
59:811–818. Swann, J., F. Rahaman, T. Bijak, and J. Fiber. 2001. The main
Storz, J. F., and H. E. Hoekstra. 2007. The study of adapta- olfactory system mediates pheromone-induced fos expres-
tion and speciation in the genomic era. Journal of sion in the extended amygdala and preoptic area of the male
Mammalogy 88:1–4. Syrian hamster. Neuroscience 105:695–706.
Stowe, M. K., J. H. Tumlinson, and R. R. Heath. 1987. Taborsky, M. 1985. Breeder-helper conflict in a cichlid fish
Chemical mimicry: bolas spiders emit components of moth with broodcare helpers: An experimental analysis. Behaviour
prey species sex pheromones. Science 236:964–966. 95:45–75.
Strand, C. R., and P. Deviche. 2007. Hormonal and environ- Taborsky, M. 2006. Ethology into a new era. Ethology 112:1–6.
mental control of song control region growth and new neu- Taglialatela, J. P., S. Savage-Rumbaugh, and L. A. Baker. 2003.
ron addition in adult male house finches, Carpodacus Vocal production by a language-competent Pan paniscus.
mexicanus. Developmental Neurobiology 67:827–837. International Journal of Primatology 24:1–16.
Strandberg, R., and T. Alerstam. 2007. The strategy of fly- Takahashi, D., and M. Kohda. 2004. Courtship in fast water
and-forage migration, illustrated for the osprey (Pandion currents by a male stream goby (Rhinogobius brunneus) com-
haliaetus). Behavioral Ecology and Sociobiology 61:1865–1875. municates the parental quality honestly. Behavioral Ecology
Strassmann, J. E., and D. C. Queller. 2007. Insect societies as and Sociobiology 55:431–438.
divided organisms: the complexities of purpose and cross- Takeda, K. 1961. Classical conditioned response in the honey
purpose. Proceedings of the National Academy of Sciences bee. Journal of Insect Physiology 6:168–179.
104:8619–8628. Tamura, N. 1989. Snake-directed mobbing by the Formosan
Strassmann, J. E., C. R. Hughes, D. C. Queller, S. Turlillazzi, squirrel Callosciurus erythraeus thaiwanensis. Behavioral
R. Cervo, S. K. Davies, and K. F. Goodnight. 1989. Genetic Ecology Sociobiology 24:175–180.
relatedness in primitively eusocial wasps. Nature Tan, E. J., and B. L. Tang. 2006. Looking for food: Molecular
342:268–269. neuroethology of invertebrate feeding behavior. Ethology
Struhsaker, T. T. 1967. Auditory communication among 112:826–832.
vervet monkeys (Cercopithecus aethiops). In Social Tang-Martinez, Z., and T. B. Ryder. 2005. The problem with
Communication among Primates, edited by S. A. Altmann. paradigms: Bateman’s worldview as a case study. Integrative
Chicago: University of Chicago Press, pp. 281–324. and Comparative Biology 45:821–830.
Stutt, A. D., and M. T. Siva-Jothy. 2001. Traumatic insemi- Tanner, D. A., and P. K. Visscher. 2008. Do honey bees aver-
nation and sexual conflict in the bed bug Cimex lectularius. age directions in the waggle dance to determine a flight
Proceedings of the National Academy of Sciences 98:5683–5687. direction? Behavioral Ecology and Sociobiology 62:1891–1898.
Stutt, A. D., and P. Willmer. 1998. Territorial defence in Tanouye, M. A., C. A. Ferrus, and S. C. Fujita. 1981.
speckled wood butterflies: Do the hottest males always win? Abnormal action potentials associated with the Shaker com-
Animal Behaviour 55:1341–1347. plex locus in Drosophila. Proceedings of the National Academy
Sugiyama, Y. 1965. Behavioral development and social struc- of Sciences 78:6548–6552.
ture in two troops of hanuman langurs (Presbytis entellus). Tarsitano, M. S. 2006. Route selection by a jumping spider
Primates 6:213–247. (Portia labiata) during the locomotory phase of a detour.
Sugiyama, Y. 1984. Proximate factors of infanticide among Animal Behaviour 72:1437–1442.
langurs at Dharwar: A reply to Bogess. In Infanticide: Tarsitano, M. S., and A. R., Andrew. 1999. Scanning and route
Comparative and Evolutionary Perspectives, edited by G. selection in the jumping spider Portia labiata. Animal
Hausfater and S. B. Hrdy. New York: Aldine, pp. 311–314. Behaviour 58:255–265.
Summers, C. H., and S. Winberg. 2006. Interactions between Tarsitano, M. S., and R. R. Jackson. 1997. Araneophagic jump-
the neural regulation of stress and aggression. Journal of ing spiders discriminate between detour routes that do and
Experimental Biology 209:4581–4589. do not lead to prey. Animal Behaviour 53:257–266.
Summers, K., and M. E. Clough. 2001. The evolution of col- Taylor, A. H., G. R. Hunt, J. C. Holzhaider, and R. D. Gray.
oration and toxicity in the poison frog family 2007. Spontaneous metatool use by New Caledonian
(Dendrobatidae). Proceedings of the National Academy of crows. Current Biology 17:1504–1507.
Sciences 98:6227–6232. Taylor, P. W., and R. W. Elwood. 2003. The mismeasure of
Surlykke, A. 1984. Hearing in notodontid moths: A tympanic animal contests. Animal Behaviour 65:1195–1202.
organ with a single auditory neurone. Journal of Temeles, E. J. 1989. Effect of prey consumption on foraging
Experimental Biology 113:323–335. activity of northern barriers. Auk 106:353–357.
Sutherland, W. 1998. The importance of behavioural studies Temeles, E. J., I. L. Pan, J. L. Brennan, and J. N. Horwitt.
in conservation biology [Review]. Animal Behaviour 2000. Evidence for ecological causation of sexual dimor-
56:801–809. phism in a hummingbird. Science 289:441–443.
506 References

Tennesen, M. 1999. Testing the depths of life. National Wildlife Tinbergen, N., and W. Kruyt. 1938. Über die Oreintierrung
February/March, volume 2. des Bienenwolfes (Philanthus triangulum Fabr.) III. Die
Tepperman, J. 1980. Metabolic and Endocrine Physiology. 4th ed. Bevorzugung bestimmter Wegmarken. Zeitschrift für ver-
Chicago: Year Book Medical. gleichende Physiologie 25:292–334.
Terborgh, J., and A. W. Goldizen. 1985. On the mating sys- Tinbergen, N., G. J. Broekhuysen, F. Feekes, J. C. W.
tem of the cooperatively breeding saddle-backed tamarin Houghton, H. Kruuk, and E. Szulc. 1962. Egg-shell
(Saguinus fuscicollis). Behavioral Ecology and Sociobiology removal by the black-headed gull, Larus ridibundus L., a
16:293–299. behaviour component of camouflage. Behaviour 19:74–118.
Terrace, H. S. 2005. Metacognition and the evolution of lan- Tobias, M. L., C. Barnard, R. O’Hagan, S. H. Horng, M.
guage. In The Missing Link in Cognition, edited by H. S. Rand, and D. B. Kelley. 2004. Vocal communication
Terrace and J. Metcalfe. Oxford: Oxford University Press, between male Xenopus laevis. Animal Behaviour 67:353–365.
pp. 84–113. Tobias, M. L., S. S. Viswanathan, and D. B. Kelley. 1998.
Terrace, H. S., L. A. Petitto, R. J. Sanders, and T. G. Bever. Rapping, a female receptive call, initiates male-female duets
1979. Can an ape create a sentence? Science 206:891–900. in the South African clawed frog. Proceedings of the National
Terrick, T. D., R. L. Mumme, and G. M. Burghardt. 1995. Academy of Sciences 95:1870–1875.
Aposematic coloration enhances chemosensory recognition Tomaschko, K.-H. 1994. Ecdysteroids from Pycnogonum
of noxious prey in the garter snake Thamnophis radix. Animal litorale (Arthropoda, Pantopoda) act as a chemical defense
Behaviour 49:857–866. against Carcinus maenus (Crustacea, Decapoda). Journal of
Chemical Ecology 20:1445–1455.
Thayer, A. H. 1896. The law which underlies protective col-
oration. Auk 13:124–129. Tomsic D., V. Massoni, H. Maldonado. 1993. Habituation to
a danger stimulus in two semiterrestrial crabs—ontogenic,
Thom C., D. C. Gilley, J. Hooper, and H. E. Esch. 2007. The
ecological and opioid modulation correlates. Journal of
scent of the waggle dance. PLoS Biology 5:e228.
Comparative Physiology A—Neuroethology 173:621–633.
Thompson, C. K., G. E. Bentley, and E. A. Brenowitz. 2007.
Tousson, E., and H. Meissl. 2004. Suprachiasmatic nuclei
Rapid seasonal-like regression of the adult avian song con-
grafts restore circadian rhythm in the paraventricular
trol system. Proceedings National Academy of Sciences
nucleus of the hypothalamus. The Journal of Neuroscience
104:15520–15525.
24:2983–2988.
Thompson, K. V. 1996. Play-partner preferences and the
Trainer, J. M., and D. B. McDonald. 1995. Singing perfor-
function of social play in infant sable antelope, Hippotragus
mance, frequency matching and courtship success of long-
niger. Animal Behaviour 52:1143–1155.
tailed manakins (Chiroxiphia linearis). Behavioral Ecology and
Thorndike, E. L. 1898. Animal intelligence: An experimental
Sociobiology 37:249–254.
study of the associative process in animals. Psychological
Trainer, J. M., D. B. McDonald, and W. A. Learn. 2002. The
Monographs 2:1–109.
development of coordinated singing in cooperatively dis-
Thorndike, E. L. 1911. Animal Intelligence: Experimental playing long-tailed manakins. Behavioral Ecology 13:65–69.
Studies. New York: Macmillan.
Trainor, B. C., and H. A. Hofmann. 2006. Somatostatin reg-
Thornhill, R. 1976. Sexual selection and nuptial feeding ulates aggressive behavior in an African cichlid fish.
behavior in Bittacus apicalis (Insecta: Mecoptera). American Endocrinology 147:5119–5125.
Naturalist 110:529–548.
Travers, S. E., M. D. Smith, J. Bai, S. H. Hulbert, J. E. Leach,
Thornhill, R. 1981. Panorpa (Mecoptera: Panorpidae) scorpi- P. S. Schnable, A. K. Knapp, G. A. Milliken, P. A. Fay, A.
onflies: Systems for understanding resource-defense polyg- Saleh, and K. A. Garrett. 2007. Ecological genomics:
yny. Annual Review of Ecology and Systematics 12:355–386. Making the leap from model systems in the lab to native
Tian, L., B. Xiao, W. Lin, S. Zhang, R. Shu, and Y. Pan. 2007. populations in the field. Frontiers in Ecology and the
Testing for the presence of magnetite in the upper-beak skin Environment 5:19–24.
of homing pigeons. BioMetals 20:197–203. Triefenbach, F. A., and H. H. Zakon. 2008. Changes in sig-
Tinbergen, L. 1960. The natural control of insects in pine naling during agonistic interactions between male weakly
woods I. Factors influencing the intensity of predation by electric knifefish, Apteronotus leptorhynchus. Animal Behaviour
song birds. Archives Néerlandaises de Zoologie 13:265–343. 75: 1263–1272.
Tinbergen, N. 1951. The Study of Instinct. London: Oxford Trivers, R. L. 1971. The evolution of reciprocal altruism.
University Press. Quarterly Review of Biology 46:35–57.
Tinbergen, N. 1952a. The curious behavior of the stickleback. Trivers, R. L. 1972. Parental investment and sexual selection.
Scientific American 187:2–6. In Sexual Selection and the Descent of Man, 1871–1971, edited
Tinbergen, N. 1952b. Derived activities: their causation, bio- by B. Campbell. Chicago: Aldine, pp. 136–179.
logical significance, origin and emancipation during evolu- Trivers, R. L. 1974. Parent-offspring conflict. American
tion. Quarterly Review of Biology 27:1–32. Zoologist 14:249–264.
Tinbergen, N. 1963. On aims and methods of ethology. Trivers, R. L., and H. Hare. 1976. Haplodiploidy and the evo-
Zeitschrift für Tierpsycholgie 20:410–433. lution of the social insects. Science 191:249–263.
 References 507

Trumbo, S. T. 2006. Infanticide, sexual selection and task spe- Venner, S., and J. Casas. 2005. Spider webs designed for rare
cialization in a biparental burying beetle. Animal Behaviour but life-saving catches. Proceedings of the Royal Society of
72:1159–1167. London B 272:1587–1592.
Trumbo, S. T. 2007. Can the “challenge hypothesis” be Verner, J., and M. F. Willson. 1966. The influence of habitats
applied to insects? Hormones and Behavior 51:281–285. on mating systems of North American passerine birds.
Turchin, P., and P. Kareiva. 1989. Aggregation in Aphis vari- Ecology 47:143–147.
ans: An effective strategy for reducing predation risk. Ecology Verwey, J. 1930. Die Paarungsbiologie des Fischreihers.
70:1008–1016. Zoölogische Jahrbucher-Abteilung für Allgemeine Zoölogie und
Turek, F. W., C. Joshu, E. Lin, G. Ivanova, E. McDearmon, Physiologie der Tiere 48:1–120.
A. Laposky, S. Losee-Olson, A. Easton, D. R. Jensen, R. H. Villinger, J., and B. Waldman. 2008. Self-referent MHC type
Eckel, J. S. Takahashi, and J. Bass. 2005. Obesity and meta- matching in frog tadpoles. Proceedings of the Royal Society of
bolic syndrome in circadian clock mutant mice. Science London B 275:1225–1230.
308:1043–1045. Visalberghi, E. and E. Alleva. 1979. Magnetic influences on
Turner, J. R. G. 1977. Butterfly mimicry: The genetical evo- pigeon homing. Biological Bulletin 125:246–256.
lution of an adaptation. Evolutionary Biology10:163–206. Visalberghi, E., and L. Limongelli. 1994. Lack of compre-
Uesugi, K. 1996. The adaptive significance of Batesian mim- hension of cause-effect relations in tool-using capuchin
icry in the swallowtail butterfly, Papio polytes (Insecta, monkeys (Cebus apella). Journal of Comparative Psychology
Papilionidae): Associative learning in a predator. Ethology 108:15–22.
102:762–775. Visalberghi, E., and L. Limongelli. 1996. Acting and under-
Uetz, G. W., and J. A. Roberts. 2002. Multisensory cues and standing: tool use revisited through the minds of capuchin
monkeys. In Reaching into Thought: The Minds of the Great
multimodal communication in spiders: insights from
Apes, edited by A. Russon, K. Bard, and S. Parker.
video/audio playback studies. Brain, Behavior, and Evolution
Cambridge: Cambridge University Press, pp. 57–79.
59:222–230.
Visalberghi, E., and L. Trinca. 1989. Tool use in capuchin
Uetz, G. W., R. Papke, and B. Kilinc. 2002. Influence of feed-
monkeys: distinguishing between performing and under-
ing regime on body size, body condition and a male sec-
standing. Primates 30:511–521.
ondary sexual character in Schizocosa ocreata wolf spiders
Vollrath, F. 1986. Eusociality and extraordinary sex ratios in
(Araneae, Lycosidae): condition-dependence in a visual sig-
the spider Anelosimus eximius (Araneae, Theridiidae).
naling trait. Journal of Arachnology 30:461–469.
Behavioral Ecology and Sociobiology 4:283–287.
Urquhart, F. A. 1987. The Monarch Butterfly: International
von der Emde, G. 1999. Active electrolocation of objects in
Traveler. Chicago: Nelson-Hall.
weakly electric fish. Journal of Experimental Biology
Vahed, K. 1998. The function of nuptial feeding in insects: A
202:1205–1215.
review of empirical studies. Biological Reviews 73:43–78.
von Frisch, K. 1950. Die Sonne als Kompass im Leben der
Vallin, A., S. Jakobsson, J. Lind, and C. Wiklund. 2005. Prey Bienen. Experientia 6:210–221.
survival by predator intimidation: an experimental study of
von Frisch, K. 1967. The Dance Language and Orientation of
peacock butterfly defence against blue tits. Proceedings of the
Bees. Cambridge, MA: Harvard University Press.
Royal Society of London B 272:1203–1207.
von Frisch, K. 1971. Bees: Their Vision, Chemical Senses and
van Praag, H., G. Kempermann, and F. H. Gage. 2000. Neural Language. 2nd ed. Ithaca, NY: Cornell University Press.
consequences of environmental enrichment. Nature Reviews
Voss, R. S. 1979. Male accessory glands and the evolution of
Neuroscience 1:191–198.
copulatory plugs in rodents. Occasional Papers of the Museum
Van Shaik, C. P., and C. H. Janson. 2000. Infanticide by Males of Zoology. Ann Arbor, MI: University of Michigan 689:1–27.
and Its Implications. Cambridge, MA: Cambridge University Waage, J. K. 1979. Dual function of the damselfly penis:
Press. Sperm removal and transfer. Science 203:916–918.
van Staaden, M. J. 1998. Ethology: at 50 and beyond. Trends Waas, J. R. 1991. The risks and benefits of signalling aggres-
in Ecology and Evolution 13:6–8. sive motivation: A study of cave-dwelling little blue pen-
Van Valen, L. 1973. A new evolutionary law. Evolutionary guins. Behavioral Ecology and Sociobiology 29:139–146.
Theory 1:1–30. Wade, J., and A. P. Arnold. 1996. Functional testicular tissue
Vander Wall, S. B. 1982. An experimental analysis of cache does not masculinize development of the zebra finch song
recovery in Clark’s nutcracker. Animal Behaviour 30:84–94. system. Proceedings of the National Academy of Sciences
Vander Wall, S. B., and R. P. Balda. 1977. Coadaptations of 93:5264–5268.
the Clark’s nutcracker and the pinyon pine for efficient seed Wade, J., and A. P. Arnold. 2004. Sexual differentiation of the
harvest and dispersal. Ecological Monographs 47:89–111. zebra finch song system. Annals of the New York Academy of
Velando, A., Beamonte-Barrientos, and R. Torres. 2006. Sciences 1016: 540–559.
Pigment-based skin colour in the blue-footed booby: an Wagner, G. 1976. Das Orientierungverhalten von Brieftauben
honest signal of current condition used by females to adjust im erdmagnetisch gestörten Gebiete des Chasseral. Revue
reproductive investment. Oecologia 149:535–542. Suisse de Zoologie 83:883–890.
508 References

Walcott, C. 1978. Anomalies in the earth’s magnetic field Chronobiology: Biological Timekeeping, edited by J. C. Dunlap,
increase scatter of pigeon vanishing bearings. In Animal J. J. Loros and P. J. DeCoursey. Sunderland, MA: Sinauer.
Migration, Navigation and Homing, edited by K. Schmidt- pp. 325-356.
Koenig and W. T. Keeton. Berlin: Springer-Verlag. pp. Waterman, T. H. 1989. Animal Navigation. New York: W. H.
143–151. Freeman.
Walcott, C. 2005. Multi-modal orientational cues in homing Watson, J. B. 1930. Behaviorism. New York: W.W. Norton &
pigeons. Integrative and Comparative Biology 45:574–581. Company, Inc.
Walcott, C., and R. P. Green. 1974. Orientation of homing Watts , C. R., and A. W. Stokes. 1971. The social order of
pigeons altered by a change in the direction of an applied turkeys. Scientific American 224:112–118.
magnetic field. Science 184:180–182. Watts, D. P. 1987. Effects of mountain gorilla foraging activ-
Waldvogel, J. A., S. Benvenuti, W. T. Keeton, and F. Papi. ities on the productivity of their food species. African
1978. Homing pigeon orientation influenced by deflected Journal of Ecology 25:155–163.
winds at home loft? Journal of Comparative Physiology Watts, D. P. 2006. Conflict resolution in chimpanzees and the
128:297–301. valuable-relationships hypothesis. International Journal of
Primatology 27:1337–1364.
Walker, M. M., and M. E. Bitterman. 1989. Honeybees can
Weatherhead, P. J., and R. J. Robertson. 1979. Offspring qual-
be trained to respond to very small changes in geomagnetic
ity and the polygyny threshold: The “sexy son” hypothesis.
intensity. Journal of Experimental Biology 145:489–494.
American Naturalist 113:201–208.
Walker, M. M., C. E. Diebel, C. V. Haugh, P. M. Pankhurst,
Weatherhead, P. J., and R. J. Roberston. 1981. In defence of
and J. C. Montgomery. 1997. Structure and function of the
the “sexy son” hypothesis. American Naturalist 117:349–356.
vertebrate magnetic sense. Nature 390:371–376.
Weaver, I. C. G., M. J. Meaney, and M. Szyf. 2006. Maternal
Wallraff, H. G. 1980. Olfaction and homing in pigeons: Nerve care effects on the hippocampal transcriptome and anxiety-
section experiments, critique, hypotheses. Journal of mediated behaviors in the offspring that are reversible in
Comparative Physiology 139:209–224. adulthood. Proceedings of the National Academy of Sciences
Wallraff, H. G. 1981. The olfactory component of pigeon 103:3480–3485.
navigation: Steps of analysis. Journal of Comparative Weber, N. A. 1972. The attines: The fungus-culturing ants.
Physiology 143:411–422. American Scientist 60:448–456.
Wallraff, H. G. 2004. Avian olfactory navigation: its empiri- Wedell, N., M. J. G. Gage, and G. A. Parker. 2002. Sperm
cal foundation and conceptual state. Animal Behaviour competition, male prudence and sperm-limited females.
67:189–204. Trends in Ecology and Evolution 17:313–320.
Wallraff, H. G. 2005. Avian Navigation: Pigeon Homing as a Wedell, N., C. Kvarnemo, C. M. Lessells, and T. Tregenza.
Paradigm. Berlin: Springer-Verlag. 2006. Sexual conflict and life histories. Animal Behaviour
Waloff, Z. 1959. Notes on some aspects of the desert locust 71:999–1011.
problem. Report of the FAO Panels of aspects of the strategy Weeks, J. C. 2003. Thinking globally, acting locally: steroid
of the desert locust plague control. FAO Document hormone regulation of the dendritic architecture, synaptic
59–6–4737:23–26. (Cited in Alerstam, T., A. Hedenström, connectivity, and death of an individual neuron. Progress in
and S. Åkesson. 2003. Long-distance migration: evolution Neurobiology 70:421–442.
and determinants. Oikos 103:247–260. p. 249.) Weeks, J. C., and J. W. Truman. 1984. Neural organization
Walter, H. 1979. Eleanora’s Falcon: Adaptations to Prey and of peptide-activated ecdysis behaviors during metamor-
Habitat in a Social Raptor. Chicago: University of Chicago phosis of Manduca sexta. II. Retention of the proleg motor
Press. pattern despite the loss of prolegs at pupation. Journal of
Comparative Physiology A 155:423–433.
Walther, G. R., E. Post, P. Convey, A. Menzel, C. Parmesa,
Weeks, J. C., G. A. Jacobs, and C. I. Miles. 1989. Hormonally
T. J. C. Beebee, J. M Fromentin, O. Hoegh-Guldberg, and
mediated modifications of neuronal structure, synaptic con-
F. Bairlein. 2002. Ecological responses to recent climate
nectivity, and behavior during metamorphosis of the
change. Nature 416:389–395.
tobacco hornworm, Manduca sexta. American Zoologist
Wang, Y., Y. Pan, S. Parsons, M. Walker, and S. Zhang. 2007. 29:1331–1344.
Bats respond to polarity of a magnetic field. Proceedings of
Wehner, R. 1981. Spatial vision in arthropods. In Handbook of
the Royal Society B 274:2901–2905. Sensory Psychology, edited by H. Autrum. Berlin: Springer-
Waser, P. M., and W. T. Jones. 1983. Natal philopatry among Verlag, pp. 287–616.
solitary mammals. Quarterly Review of Biology 58:355–390. Wehner, R., and M. V. Srinivasan. 1981. Searching behavior of
Watanabe, S., J. Sakamoto, and M. Wakita. 1995. Pigeon’s dis- desert ants, genus Cataglyphis (Formicidae: Hymenoptera).
crimination of paintings by Monet and Picasso. Journal of Journal of Comparative Physiology 142:315–338.
the Experimental Analysis of Behavior 63:165–174. Weimerskirch, H., J. Martin, Y. Clerquin, P. Alexandre, and
Waterhouse, J. M., and P. J. DeCoursey. 2004. The relevance S. Jarskova. 2001. Energy saving in flight formation. Nature
of human circadian rhythms for human welfare. In 413:697–698.
 References 509

Weiner, J. 1995. The Beak of the Finch. New York: Vintage Whitlock, J. R., A. J. Heynen, M. G. Schler, and M. F. Bear.
Books. 2006. Learning induces long-term potentiation in the hip-
Weir, A. S., J. Chappell, and A. Kacelnik. 2002. Shaping of pocampus. Science 313:1093–1097.
hooks in New Caledonian crows. Science 297:981. Whiten, A. 1996. When does smart behaviour-reading
Weissburg, M. 1992. Functional analysis of fiddler crab for- become mind-reading? In Theories of Theories of Mind, edited
aging: sex specific mechanics and constraints. Journal of by P. Carruthers and P. K. Smith. Cambridge: Cambridge
Experimental Marine Biology and Ecology 156:105–124. University Press, pp. 277–292.
Weissburg, M. 1993. Sex and the single forager: gender-spe- Whitten, P. L., D. K. Brockman, and R. C. Stavisky. 1998.
cific foraging strategies in the fiddler crab Uca pugnax. Recent advances in noninvasive techniques to monitor hor-
Ecology 74:279–291. mone-behavior interactions. Yearbook of Physical Anthropology
Welsh, D. K., D. E. Logothetis, M. Meister, and S. M. 41:1–23.
Reppert. 1995. Individual neurons dissociated from rat Wich, S. A., and H. de Vries. 2006. Male monkeys remember
suprachiasmatic nucleus express independently phased cir- which group members have given alarm calls. Proceedings of
cadian firing patterns. Neuron 14:697–706. the Royal Society of London B 273:735–740.
Welsh, D. K., S.-H. Yoo, A. C. Liu, J. S. Takahashi, and S. A. Wickler, W. 1968. Mimicry in Plants and Animals. McGraw-
Kay. 2004. Bioluminescence imaging of individual fibrob- Hill, New York. 253 pp.
lasts reveals persistent, independently phased circadian Wickler, W. 1972. Mimicry. 2nd ed. New York: World
rhythms of clock gene expression. Current Biology University Library, McGraw-Hill.
14:2289–2295.
Wiggins, D. A., and R. D. Morris. 1986. Criteria for female
Welty, J. C. 1962. The Life of Birds. 2nd ed. Philadelphia: choice of mates: Courtship feeding and paternal care in the
Saunders. common tern. American Naturalist 128:126–129.
Wenner, A. M. 1964. Sound communication in honeybees. Wilbrecht, L., A. Crionas, and F. Nottebohm. 2002.
Scientific American 210:116–125. Experience affects recruitment of new neurons but not adult
Wenner, A. M. 2002. The elusive honey bee dance “language” neuron number. The Journal of Neuroscience 22:825–831.
hypothesis. Journal of Insect Behavior 15:859–878.
Wilbrecht, L., H. Williams, N. Gangadhar, and F.
Werren, J. H., M. R. Gross, and R. Shine. 1980. Paternity and Nottebohm. 2006. High levels of new neuron addition per-
the evolution of male parental care. Journal of Theoretical sist when the sensitive period for song learning is experi-
Biology 82:619–631. mentally prolonged. The Journal of Neuroscience
West, M. J., and M. J. King. 1988. Female visual displays affect 26:9135–9141.
the development of male song in the cowbird. Nature
Wilcox, R. S., and T. Ruckdeschel. 1982. Food threshold ter-
334:244–246.
ritoriality in a water strider (Gerris remigis). Behavioral
West, M. J., A. P. King, and D. H. Eastzer. 1981. The cow- Ecology and Sociobiology 11:85–90.
bird: Reflections on development from an unlikely source.
Wilcox, R. S., R. R. Jackson, and K. Gentile. 1996. Spiderweb
American Scientist 69:56–66.
smokescreens—Spider trickster uses background noise to
Westneat, D. F., and P. W. Sherman. 1993. Parentage and the
mask stalking movements. Animal Behaviour 51:313–326.
evolution of parental behavior. Behavioral Ecology 4:66–77.
Wiley, R. H. 1974. Evolution of social organization and life
White, D. J., A. P. King, and M. J. West. 2002. Facultative
history patterns among grouse (Aves: Tetraonidae).
development of courtship and communication skills in juve-
Quarterly Review of Biology 49:209–227.
nile male cowbirds, Molothrus ater. Behavioral Ecology
Wilkinson, G. S. 1984. Reciprocal food sharing in the vam-
13:487–496.
pire bat. Nature 308:181–184.
Whitehouse, M. E. A. 1997. Experience influences male-male
Wilkinson, G. S. 1990. Food sharing in vampire bats.
contests in the spider Argyrodes antipodiana. Animal
Scientific American 262:76–82.
Behaviour 53:913–923.
Whiteman, E. A., and I. M. Côté. 2004. Monogamy in marine Williams, C. B. 1965. Insect Migration. London: Collins.
fishes. Biological Reviews 79:351–375. Williams, G. C. 1975. Sex and Evolution. Princeton, NJ:
Whiten, A., D. M. Custance, J.-C. Gomez, P. Teixidor, and Princeton University Press.
K. A. Bard. 1996. Imitative learning of artificial fruit pro- Williams, S. L., K. E. Brakke, and E. S. Savage-Rumbaugh.
cessing in children (Homo sapiens) and chimpanzees (Pan 1997. Comprehension skills of language-competent and
troglodytes). Journal of Comparative Psychology 110:3–14. nonlanguage-competent apes. Language and Communication
Whiten, A., J. Goodall, W. C. McGrew, T Nishida, V. 17:301–317.
Reynolds, Y. Sugiyama, C. E. G. Tutin, R. W. Wrangham, Williams, T. C., and J. M. Williams. 1978. An oceanic mass
and C. Boesch. 1999. Cultures in chimpanzees. Nature migration of land birds. Scientific American 239:166–176.
399:682–685. Willis, P. M., and L. M. Dill. 2007. Mate guarding in male
Whitfield, C. W., A. M. Cziko, and G. E. Robinson. 2003. Dall’s porpoises (Phocoenoides dalli). Ethology 113:587–597.
Gene expression profiles in the brain predict behavior in Wilson, D. M. 1961. The central nervous control of flight in
individual honey bees. Science 302:296–299. a locust. Journal of Experimental Biology 38:471–490.
510 References

Wilson, D. M., and T. Weis-Fogh. 1962. Patterned activity of Wingfield, J. C. 1984. Environmental and endocrine control
coordinated motor units studied in flying locusts. Journal of of reproduction in the song sparrow, Melospiza melodia. I.
Experimental Biology 40:643–674. Temporal organization of the breeding cycle. General and
Wilson, E. O. 1968. Chemical systems. In Animal Comparative Endocrinology 56:406–416.
Communication: Techniques of Study and Results of Research, Wingfield, J. C. 2005. A continuing saga: the role of testos-
edited by T. A. Sebeok. Bloomington: Indiana University terone in aggression. Hormones and Behavior 48:253–255.
Press, pp. 75–102. Wingfield, J. C., and M. C. Moore. 1987. Hormonal, social,
Wilson, E. O. 1971. The Insect Societies. Cambridge, MA: and environmental factors in the reproductive biology of
Belknap/Harvard University Press. free-living male birds. In Psychobiology of Reproductive
Wilson, E. O. 1975. Sociobiology. Cambridge, MA: Belknap Behavior, edited by D. Crews. Upper Saddle River, NJ:
Press of Harvard University Press. Prentice Hall, pp. 148–175.
Wilson, E. O. 1980. Sociobiology, the Abridged Version. Wingfield, J. C., and K. K. Soma. 2002. Spring and autumn
Cambridge, MA: Belknap/Harvard University Press. territoriality in song sparrows: same behavior, different
Wilson, E. O. 2008. One giant leap: How insects achieved mechanisms? Integrative and Comparative Biology 42:11–20.
altruism and colonial life. BioScience 58:17–25. Wingfield, J. C., R. E. Hegner, A. M. Dufty, and G. F. Ball.
1990. The “Challenge Hypothesis”: Theoretical implica-
Wilson, E. O., and B. Hölldobler. 2005. Eusociality: origin
tions for patterns of testosterone secretion, mating systems,
and consequence. Proceedings of the National Academy of
and breeding strategies. American Naturalist 136:829–846.
Sciences 102:13367–13371.
Winston, M. L., and K. N. Slessor. 1992. The essence of roy-
Wilson, P. L., M. C. Towner, and S. L. Vehrencamp. 2000.
alty: Honey bee queen pheromone. American Scientist
Survival and song-type sharing in a sedentary subspecies of
80:374–385.
the song sparrow. Condor 102:355–363.
Wirant, S. C., and B. McGuire. 2004. Urinary behavior of
Wiltschko, R., and W. Wiltschko. 1999. The orientation sys-
female domestic dogs (Canis familiaris): Influence of repro-
tems of birds I. Compass mechanisms. Journal of Ornithology
ductive condition, location and age. Applied Animal
140:1–40.
Behaviour Science 85:335–348.
Wiltschko, R., and W. Wiltschko. 2006. Magnetoreception.
Wirant, S. C., K. Halvorsen, and B. McGuire. 2007.
BioEssays 28:157–168.
Preliminary observations on the urinary behaviour of female
Wiltschcko, R., M. Walker, and W. Wiltschko. 2000. Sun-
Jack Russell Terriers in relation to stage of the oestrous
compass orientation in homing pigeons: compensation for cycle, location, and age. Applied Animal Behaviour Science
different rate of change in azimuth? Journal of Experimental 106:161–166.
Biology 203:889–894.
Wisby, W. J., and A. D. Hasler. 1954. The effect of olfactory
Wiltschko, R., T. Ritz, K. Stapput, P. Thalau, and W. occlusion on migrating silver salmon (O. kisutch). Journal of
Wiltschko. 2005. Two different types of light-dependent the Fisheries Research Board of Canada 11:472–478.
responses in magnetic fields in birds. Current Biology
Wittenberger, J. F. 1978. The evolution of mating systems in
15:1518–1523. grouse. Condor 80:126–137.
Wiltschko, W. 1978. Further analysis of the magnetic com- Wittenberger, J. F. 1979. The evolution of mating systems in
pass of migratory birds. In Animal Migration, Navigation and birds and mammals. In Handbook of Behavioral Neurobiology,
Homing, edited by K. Schmidt-Koenig and W. T. Keeton. edited by P. Marler and J. G. Vandenbergh. New York:
Berlin: Springer-Verlag, pp. 302–310. Plenum Press, pp. 271–349.
Wiltschko, W. 1982. The migratory orientation of Garden Wittenberger, J. F. 1981. Animal Social Behavior. Boston:
Warblers Sylvia borin. In Avian Navigation, edited by F. Papi Duxbury.
and H. G. Wallraff. Berlin: Springer-Verlag, pp. 50–58. Wittig, R. M., C. Crockford, R. M. Seyfarth, and D. L.
Wiltschko, W., and R. Wiltschko. 1972. Magnetic compass of Cheney. 2007. Vocal alliances in Chacma baboons (Papio
European robins. Science 176:62–64. hamadryas ursinus). Behavioral Ecology and Sociobiology
Wiltschko, W., and R. Wiltschko. 1996. Magnetic orientation 61:899–909.
in birds. Journal of Experimental Biology 199:29–38. Wittlinger, M., R. Wehner, and H. Wolf. 2007. The desert ant
Wiltschko, W., and R. Wiltschko. 2005. Magnetic orientation odometer: a stride integrator that accounts for stride length
and magnetoreception in birds and other animals. Journal and walking speed. Journal of Experimental Biology
of Comparative Physiology A 191:675–693. 210:198–207.
Wiltschko, W., and R. Wiltschko. 2007. Magnetoreception in Woelfle, M. A., and C. H. Johnson. 2006. No promoter left
birds: two receptors for two different tasks. Journal of behind: Global circadian gene expression in cyanobacteria.
Ornithology 148:S61–S76. Journal of Biology Rhythms 21:419–431.
Wiltschko, W., R. Wiltschko, and C. Walcott. 1987. Pigeon Wöhr, M., and R. K. W. Schwarting. 2007. Ultrasonic com-
homing: different effects of olfactory deprivation in differ- munication in rats: Can playback of 50-kHz calls induce
ent countries. Behavioral Ecology and Sociobiology 21:333–342. approach behavior? PloS ONE 2:e1365.
 References 511

Wöhr, M., and R. K. W. Schwarting. 2008. Ultrasonic call- Wourms, M. K., and F. E. Wasserman. 1985. Butterfly wing
ing during fear conditioning in the rat: no evidence for an markings are more advantageous during handling than dur-
audience effect. Animal Behaviour 76:749–760. ing the initial strike of an avian predator. Evolution
Wojcieszek, J. M., J. A. Nicholls, and A. W. Goldizen. 2007. 39:845–851.
Stealing behavior and the maintenance of a visual display Wrangham, R. D. 1980. Female choice of least costly males;
in the satin bowerbird. Behavioral Ecology 18:689–695. a possible factor in the evolution of leks. Zeitschrift für
Wolff, J. O. 1992. Parents suppress reproduction and stimu- Tierpsychologie 54:357–367.
late dispersal in opposite-sex juvenile white-footed mice. Wright, J., and I. Cuthill. 1990. Manipulation of sex differ-
Nature 359:409–410. ence in parental care: effect of brood size. Animal Behaviour
Wolff, J. O. 1994. More on juvenile dispersal in mammals. 40:462–471.
Oikos 71:349–352. Wright, J., and A. F. Russell. 2008. How helpers help: disen-
Wolff, J. O., and J. H. Plissner. 1998. Sex biases in avian natal tangling ecological confounds from the benefits of cooper-
dispersal: An extension of the mammalian model. Oikos ative breeding. Journal of Animal Ecology 77:427–429.
83:327–330. Würbel, H. 2001. Ideal homes? Housing effects on rodent
Wolovich, C. K., S. Evans, and J. A. French. 2008. Dads do brain and behaviour. Trends in Neuroscience 24:207–211.
not pay for sex but do buy the milk: food sharing and repro- Wyatt, T. D. 2003. Pheromones and Animal Behaviour:
duction in owl monkeys (Aotus spp.). Animal Behaviour Communication by Smell and Taste. Cambridge: Cambridge
75:1155–1163. University Press.
Wong, B. B. M., U. Candolin, and K. Lindström. 2007. Wyers, E. J., V. S. Peeke, and M. J. Herz. 1973. Behavioral
Environmental deterioration compromises socially enforced habituation in invertebrates. In Habituation, edited by V. S.
signals of male quality in three-spined sticklebacks. Peeke and M. J. Herz. New York: Academic Press, pp. 1–57.
American Naturalist 170:184–189. Yamagata, N., H. Nishino, and M. Mizunami. 2007. Neural
Wong, R. Y., and C. D. Hopkins. 2007. Electrical and behav- pathways for the processing of alarm pheromone in the ant
ioral courtship displays in the mormyrid fish Brienomyrus brain. Journal of Comparative Neurology 505:424–442.
brachyistius. Journal of Experimental Biology 210:2244–2252. Yamamoto, Y., and H. Ueda. 2007. Physiological study on
Wood, D. E. 1995. Modulation of behavior by biogenic imprinting and homing related to olfactory functions in
amines and peptides in the blue crab, Callinectes sapidus. salmon. In North Pacific A: North Pacific Anadromous Fish
Journal of Comparative Physiology A 177:331–333. Commission Technical Report Number 7, pp. 113–114.
Wood, D. E., M. Nishikawa, and C. D. Derby. 1996. Yamazaki, K., E. A. Boyse, V. Mike, H. T. Thaler, B. J.
Proctolinlike immunoreactivity and identified neurosecre- Mathieson, J. Abbott, J. Boyse, Z. A. Zayas, and L. Thomas.
tory cells as putative substrates for modulation of courtship 1976. Control of mating preferences in mice by genes in the
display behavior in the blue crab, Callinectes sapidus. Journal major histocompatability complex. Journal of Experimental
of Comparative Physiology A 368:153–163. Medicine 144:1324–1335.
Woods, W. A. Jr., H. Hendrickson, J. Mason, and S. M. Lewis. Yamazaki, S., R. Numano, M. Abe, A. Hida, R.-I. Takahashi,
2007. Energy and predation costs of firefly courtship sig- M. Ueda, G. D. Block, Y. Sakai, M. Menaker, and H. Tei.
nals. American Naturalist 170:702–708. 2000. Resetting central and peripheral circadian oscillators
Woolfenden, G. E. 1975. Florida scrub jay helpers at the nest. in transgenic rats. Science 288:288–290.
Auk 92:1–15. Yang, D., B. Kynard, Q. Wei, X. Chen, W. Zheng, and H. Du.
Woolfenden, G. E., and J. W. Fitzpatrick. 1978. The inheri- 2006. Distribution and movement of Chinese sturgeon,
tance of territory in group-breeding birds. BioScience Acipenser sinensis, on the spawning ground located below the
28:104–108. Gezhouba Dam during spawning seasons. Journal of Applied
Woolfenden, G. E., and J. W. Fitzpatrick. 1990. Florida scrub Ichthyology 22 (Supplement 1):145–151.
jays: A synopsis after 18 years of study. In Cooperative Yasui, Y. 2001. Female multiple mating as a genetic bet-hedg-
Breeding in Birds, edited by P. B. Stacey and W. D. Koenig. ing strategy when mate choice criteria are unreliable.
Cambridge: Cambridge University Press, pp. 241–266. Ecological Research 16:605–16.
Woolley, S. C., J. T. Sakata, and D. Crews. 2004. Evolutionary Ydenberg, R. C., R. W. Butler, and D. B. Lank. 2007. Effects
insights into the regulation of courtship behavior in male of predator landscapes on the evolutionary ecology of rout-
amphibians and reptiles. Physiology and Behavior 83:347–360. ing, timing, and molt by long-distance migrants. Journal of
Woolley, S. M., and E. W. Rubel. 1997. Bengalese finches Avian Biology 38:523–529.
Lonchura striata domestica depend upon auditory feedback for Ydenberg, R. C., and J. R. Krebs. 1987. The trade-off between
the maintenance of adult song. Journal of Neuroscience territorial defense and foraging in the great tit (Parus major).
17:6380–6390. American Zoologist 27:337–346.
Worden, B. D., and D. R. Papaj. 2005. Flower choice copy- Yeargan, K. V. 1994. Biology of bolas spiders. Annual Review
ing in bumblebees. Biology Letters 1:504–507. of Ecology and Systematics 31:81–99.
512 References

Yoo, S.-H., S. Yamazaki, P. L. Lowrey, K. Shimomura, C. H. Zahavi, A., and A. Zahavi. 1997. The Handicap Principle: A
Ko, E. D. Buhr, S. M. Siepka, H.–K. Hong, W. J. Oh, Missing Piece of Darwin’s Puzzle. Oxford: Oxford University
O. J. Yoo, M. Menaker, and J. S. Takahashi. 2004. Period2: Press.
Luciferase real-time reporting of circadian dynamics reveals Zakon, H. H., D. J. Zwickl, Y. Lu, and D. M. Hillis. 2008.
persistent circadian oscillations in mouse peripheral tissues. Molecular evolution of communication signals in electric
Proceedings of the National Academy of Sciences fish. Journal of Experimental Biology 211:1814–1818.
101:5339–5346. Zedrosser, A., O.-G. Stoen, S. Saebo, and J. E. Swenson. 2007.
Young, L. J., R. Nilsen, K. G. Waymire, G. R. MacGregor, Should I stay or should I go? Natal dispersal in the brown
and T. R. Insel. 1999. Increased affililiative response to bear. Animal Behaviour 74:369–376.
vasopressin in mice expressing the vasopressin receptor Zeh, J. A., and D. W. Zeh. 1996. The evolution of polyandry
from a monogamous vole. Nature 400:766–768. I: Intragenomic conflict and genetic incompatability.
Youthed, G. J., and R. C. Moran. 1969. The lunar day activ- Proceedings of the Royal Society of London B 263:1711–1717.
ity rhythm of myrmeleontid larvae. Journal of Insect Zeh, J. A., and D. W. Zeh. 2001. Reproductive mode and
Physiology 15:1259–1271. the genetic benefits of polyandry. Animal Behaviour
Zach, R. 1978. Selection and dropping of whelks by 61:1051–1063.
Northwestern crows. Behaviour 67:134–148. Zuberbühler, K. 2005. The phylogenetic roots of language:
Zach, R. 1979. Shell-dropping: Decision making and optimal evidence from primate communication and cognition.
foraging in Northwestern crows. Behaviour 67:106–117. Current Directions in Psychological Science 14:126–130.
Zagotta, W. N., S. Germeraad, S. S. Garber, T. Hoshi, and R. Zuberbühler, K. 2006. Language evolution: the origin of
W. Aldrich. 1989. Properties of ShB A-type potassium meaning in primates. Current Biology 16:R123–R125.
channels expressed in Shaker mutant Drosophila by germline Zucca, P., F. Antonelli, and G. Vallortigara. 2005. Detour
transformation. Neuron 3:773–782. behaviour in three species of birds: quails (Coturnix sp.),
Zahavi, A. 1975. Mate selection—A selection for a handicap. herring gulls (Larus cachinnans) and canaries (Serinus
Journal of Theoretical Biology 53:205–214. canaria). Animal Cognition 8:122–128.
Zahavi, A. 1977. Reliability in communication systems and the
evolution of altruism. In Evolutionary Ecology, edited by B.
Stonehouse. London: Macmillan, pp. 253–259.
 Photo Credits

Chapter 1 5.16: Courtesy Irene Pepperberg, Brandeis University,

Figure 1.1: Comstock, Inc.; Figure 1.2: Milton H. Tierney, Psychology.
Jr./Visuals Unlimited; Figure 1.3: S. Muller/Wildlife/Peter
Arnold, Inc. Chapter 6
Figure 6.11: G. Ronald Austing/Photo Researchers, Inc.; Figure
Chapter 2 6.12: Michael Fairchild/Peter Arnold, Inc.; Figure 6.14: Anne
Figure 2.1: American Museum of Natural History Library; Knudsen; Figure 6.21: Heather Cameron, NIMH/NIH; Figure
Figure 2.3a: Culver Pictures; Figure 2.3b: Culver Pictures; Figure 6.27: Stephen Dalton/Photo Researchers, Inc.
2.3c: ©AP/Wide World Photos; Figure 2.8: Corbis–Bettmann;
Figure 2.9: Nina Leen/©Time, Inc. Chapter 7
Figure 7.5: Frank L. Moore, Department of Zoology, University
Chapter 3 of Oregon; Figure 7.11: Courtesy Stephen Goodenough; Figure
Figure 3.1: Bruce Coleman, Inc./Alamy; Figure 3.4: from Lee 7.13: Courtesy Ronald Barfield, Dept. of Biological Sciences,
Drickamer, Animal Behavior, Fifth Edition, 1997, McGraw–Hill Rutgers University; Figure 7.17: Courtesy David Crews,
Companies. Reproduced with the permission of Lee Drickamer Departments of Zoology and Psychology, University of Texas;
and the McGraw–Hill Companies; Figure 3.10a: Photo from: Figure 7.21a: Sharon Wirant; Figure 7.21b: Sharon Wirant;
Ben–Shahar, Y., A. Robichon, M. B. Sokolowski, and G. E. Figure 7.21c: Sharon Wirant.
Robinson. 2002. Influence of gene action across different time
scales on behavior. Science 296 (April 26):741–744; Figure 3.12a: Chapter 8
Photo from: Whitfield, C. W., et al. 2003. Gene expression pro- Figure 8.1: ©NHPA/Photoshot; Figure 8.3a: Scott Camazine/
files in the brain predict behavior in individual honey bees. Photo Researchers, Inc.; Figure 8.3b: Steve Ross/Photo
Science 302 (Oct. 10): 296–299; Figure 3.16a & b: Photo from Researchers, Inc.; Figure 8.3c: Matt Meadows/Peter Arnold, Inc.;
Lim, M. M., Z. Wang, D. E. Olazábal, X. Ren, E. F. Terwilliger, Figure 8.5: Courtesy W. E. Bemis, Department of Ecology and
and L. J. Young. 2004. Enhanced partner preference in a promis- Evolutionary Biology, Cornell University; Figure 8.8: Stephen J.
cuous species by manipulating the expression of a single gene. Krasemann/DRK Photo; Figure 8.9: Mary M. Thacher/Photo
Nature 429:754–757; Figure 3.18a & b: Burmeister, S. S., et al. Researchers; Figure 8.10: Nina Leen/LIFE Magazine, Time Inc.;
2005. Rapid behavioral and genomic responses to social
Figure 8.14: Courtesy Peter H. Klopfer, Dept. of Zoology, Duke
opportunity. PLoS Biology 3(11), 1996–2004; Figure 3.22a:
University; Figure 8.16: ©Arco Images GmbH/Alamy; Figure
Wegner, P./Peter Arnold, Inc.; Figure 3.22b: Courtesy Russell
8.19: ©Robert Shantz/Alamy; Figure 8.21: E. R. Degginger/
Fernald; Figure 3.22c: Image by Aaron1a12 from Wikipedia:
Photo Researchers, Inc.; Figure 8.22a: Arthur Morris/Visuals
http://commons.wikimedia.org/wiki/File:BeeCropped.jpg;
Unlimited.
Figure 3.22d: ©George McCarthy/Corbis Images; Figure 3.22e:
Hans Reinhard/Photo Researchers, Inc.
Chapter 9
Figure 9.5a: Bert L. Dunne/Bruce Coleman, Inc.; Figure 9.8a:
Chapter 4
Figure 4.1: ©Biosphoto/Conchon Laurent/Peter Arnold, Inc.; James L. Amos/Peter Arnold, Inc.; Figure 9.8b: Stephen
Figure 4.2: Steve & Ann Toon/Photolibrary Group Limited; Dalton/© Natural History Photographic Agency; Figure 9.9a:
Figure 4.3: Sumio Harada/Minden Pictures, Inc.; Figure 4.4: Courtesy Steve A. Kay, Jeffrey Plants/Scripps Research Institute;
Jorg & Petra Wegner/Minden Pictures, Inc.; Figure 4.6: Figure 9.10a: Courtesy Steve A. Kay, Jeffrey Plants/Scripps
Courtesy Susan Reichert, Dept. of Zoology, University of Research Institute; Figure 9.14a: Courtesy Rae Silver; Figure
Tennessee; Figure 4.7a: ©Bob Gibbons/Alamy; Figure 4.7b: 9.14b: Courtesy Rae Silver.
Courtesy Susan Reichert, Dept. of Zoology, University of
Tennessee; Figure 4.8: Stan Tekiela/DRK Photo; Figure 4.9: Chapter 10
Tui De Roy/Minden Pictures, Inc.; Figure 4.10: M. & C. Figure 10.5: Courtesy K. Schmidt–Koenig, Department of
Photography/Peter Arnold, Inc.; Figure 4.13: ©Eric and David Zoology, Duke University, Durham, N.C.; Figure 10.6: James L.
Hosking/© Corbis. Amos/Peter Arnold, Inc.; Figure 10.8: Karl H. Maslonski/Photo
Researchers; Figure 10.11a: Johathan Blair/Woodfin Camp &
Chapter 5 Associates; Figure 10.11b: Johathan Blair/Woodfin Camp &
Figure 5.4: Nina Leen/Life Picture Service; Figure 5.6: Frank Associates; Figure 10.15a: Courtesy T. H. Waterman,
Lane Picture Agency; Figure 5.7: Courtesy Masao Kawai, Department of Molecular, Cellular and Developmental Biology,
Museum of Nature and Human Activities, Hyogo, Japan; Figure Yale University; Figure 10.15b: Courtesy T. H. Waterman,
5.8a: Michael Quinton/Minden Pictures, Inc.; Figure 5.8b: Helen Department of Molecular, Cellular and Developmental Biology,
Williams/Photo Researchers, Inc.; Figure 5.8c: Allan D. Yale University; Figure 10.15c: Courtesy T. H. Waterman,
Cruickshank/Photo Researchers, Inc.; Figure 5.11: From “The Department of Molecular, Cellular and Developmental Biology,
Mentality of Apes” by W. Kohler, 1926 ©by Harcourt, Brace & Yale University; Figure 10.16: David Dennis/Tom Stack &
Co., Routledge.; Figure 5.12: Courtesy Alex Weir, Department Associates; Figure 10.19a: Courtesy Charles Walcott, Laboratory
of Zoology, Oxford University, Oxford UK; Figure 5.14: of Ornithology, Cornell University; Figure 10.20: ©Mark
Courtesy Robert R. Jackson, School of Biological Sciences, Conlin/Alamy; Figure 10.21a: Kenneth Lohmann, Department
University of Canterbury, Christchurch, New Zealand; Figure of Biology, University of North Carolina.

513
514 Photo Credits

Chapter 11 Toronto; Figure 14.5: David Woodfall/Natural History

Figure 11.1: Photograph by Boyd Kynard, Department Natural Photographic Agency; Figure 14.8a: Hans Pfletschinger/Peter
Resources Conservation, University of Massachusetts Amherst; Arnold, Inc.; Figure 14.8b: Jonathan Waage, Division of Biology
Figure 11.5a: Photo by Betty McGuire, Department of Ecology & Medicine, Brown University; Figure 14.9: Irene Vandermolen/
and Evolutionary Biology, Cornell University; Figure 11.5b: Photo Researchers, Inc.; Figure 14.10: Lutz Fromhage, School
Photo by Betty McGuire, Department of Ecology and of Biological Sciences, University of Bristol; Figure 14.11a: S. J.
Evolutionary Biology, Cornell University; Figure 11.8: Norbert Arnold, Department of Biology, Oregon State University;
Rosing/Getty Images; Figure 11.9: Charles Ott/Photo Figure 14.11b: S. J. Arnold, Department of Biology, Oregon
Researchers, Inc.; Figure 11.11: Courtesy W. E. Bemis, State University; Figure 14.12: Richard Matthews/ Planet Earth
Department of Ecology and Evolutionary Biology, Cornell Pictures; Figure 14.15: Eric Hosking; Figure 14.16a: Daisuke
University; Figure 11.12b: Gregory Dimijan/Photo Researchers, Takahashi, Department of Zoology, Kyoto University; Figure
Inc.; Figure 11.14: Thomas D. Mangelsen www.mangelsen.com/ 14.17: John Cancalosi/DRK Photo; Figure 14.18:
Images of Nature; Figure 11.15: Chris Newbert/Minden ©franzfoto.com/Alamy; Figure 14.20: Howard Earl Uible/Photo
Pictures, Inc. Researchers, Inc.; Figure 14.23a, b & c: Courtesy Göran
Arnqvist. 2007. Coevolution between harmful male genitalia and
Chapter 12 female resistance in seed beetles. Proceedings of the National
Figure 12.3: Courtesy David Dussourd, Department of Academy of Sciences 104:10921–10925.
Biology, University of Central Arkansas; Figure 12.4: Ross
Hutchins/Photo Researchers, Inc.; Figure 12.6a: Tom Chapter 15
McHugh/Photo Researchers, Inc.; Figure 12.6b: Z. Figure 15.2a: Betty McGuire, Department of Ecology and
Leszczynski/Animals Animals/Earth Scenes; Figure 12.7: Evolutionary Biology, Cornell University; Figure 15.2b: David
Francois Gohier/Ardea London; Figure 12.8: Jim Frazier/ Fraser, Animal Welfare Program, University of British Columbia;
Auscape International Pty. Ltd.; Figure 12.9a: Joe McDonald/ Figure 15.2c: David Fraser, Animal Welfare Program, University
Animals Animals/Earth Scenes; Figure 12.9b: Scientific of British Columbia; Figure 15.5: Aaron Norman; Figure 15.7:
American, May, 1973, R. Igor Gamow and John F. Harris; Tony Heald/Nature Picture Library; Figure 15.8: Eric & David
Figure 12.12: Tom McHugh/Photo Researchers, Inc.; Figure Hosking; Figure 15.9a: George Reszeter/OSF/Photolibrary;
12.13a: Courtesy Barth Falkenberg, School of Biological Figure 15.9b: Justin G. Schuetz; Figure 15.12a: Antonio Baeza,
Sciences, University of Nebraska, Lincoln; Figure 12.13b: Smithsonian Tropical Research Institute; Figure 15.15: Al
Courtesy Al Kamil, School of Biological Sciences, University Lowry/Photo Researchers, Inc.; Figure 15.16: C. Behnke/
of Nebraska, Lincoln; Figure 12.13c: H. J. Vermes/Courtesy Animals Animals/Earth Scenes; Figure 15.17: Konrad Wothe/
Theodore Sargent, Department of Biology, University of Minden Pictures, Inc.
Massachusetts Amherst; Figure 12.15: J. B. & S. Bottomley/
Ardea London; Figure 12.17: Breck Kent/Animals Animals/ Chapter 16
Earth Scenes; Figure 12.18: Z. Leszczynki/Animals Animals/ Figure 16.3: Photograph by Vincenzo Penteriani, Department of
Earth Scenes; Figure 12.20a: David Hosking/Photo Conservation Biology, Estación Biológica de Doñana; Figure
Researchers, Inc.E61; Figure 12.21: Courtesy Paul V. Switzer, 16.5a: Tom Hince; Figure 16.5b: Photograph by Khoi Uong;
Department of Biological Sciences, Eastern Illinois University. Figure 16.6: Stimson Wilcox, Biological Sciences, Binghamton
University; Figure 16.7a: ©tbkmedia.de/Alamy; Figure 16.7b:
Chapter 13 Kate Bemis; Figure 16.8: Pete Oxford/Minden Pictures, Inc.;
Figure 13.1: Courtesy Lincoln P. Brower, Department of Figure 16.9: Kate Bemis; Figure 16.10a: Courtesy Dietrich
Biology, Sweet Briar College; Figure 13.2: Hellio & Van Ingen/ Schneider, Max–Planck Institute, Germany; Figure 16.10b:
NHPA/Photoshot; Figure 13.6a: Lydia M. Mäthger, Marine Courtesy Dietrich Schneider, Max–Planck Institute, Germany;
Biological Laboratory, Woods Hole; Figure 13.6b: Lydia M. Figure 16.11: Kate Bemis; Figure 16.13: Courtesy George W.
Mäthger, Marine Biological Laboratory, Woods Hole; Figure Uetz, Department of Biological Sciences, University of
13.6c: Lydia M. Mäthger, Marine Biological Laboratory, Woods Cincinnati; Figure 16.14: R. M. Meadows/Peter Arnold, Inc.;
Hole; Figure 13.6d: Lydia M. Mäthger, Marine Biological Figure 16.15: Time & Life Pictures/Getty Images, Inc.; Figure
Laboratory, Woods Hole; Figure 13.10: Braude, S., Biology 16.17a: Courtesy Lynne D. Houck, Department of Zoology,
Department, Washington University; Figure 13.11a: Robert and Oregon State University, and S. J. Arnold, Department of
Linda Mitchell; Figure 13.11b: Ivan Sazima, Department Zoology, Oregon State University; Figure 16.19: Roy P.
Zoologia & Museu de História Natural, Universidade Estadual Fontaine/Photo Researchers; Figure 16.27: Mark Moffett/
de Campinas; Figure 13.16: COLOR–PIC, INC./Animals Minden Pictures, Inc.
Animals/Earth Scenes; Figure 13.18: Courtesy T. E. Reimchen,
Department of Biology, University of Victoria, B.C.; Figure Chapter 17
13.20: Leonard Lee Rue/Photo Researchers, Inc.; Figure 13.23: Figure 17.3a: Tui De Roy/Minden Pictures, Inc.; Figure 17.3b:
Courtesy Edmund D. Brodie, Jr., Dept. of Biology, Utah State Alberto Velando; Figure 17.3c: Alberto Velando; Figure 17.4:
University; Figure 13.24: Adrian Vallin, Department of Zoology, Courtesy Joseph Waas, from Waas 1991; Figure 17.6b: Roy
Stockholm University; Figure 13.25: Courtesy Thomas Eisner Caldwell, Department of Integrative Zoology, University of
& Daniel Aneshansley/Cornell University; Figure 13.26: Anup California at Berkeley; Figure 17.7: Courtesy John D. Palmer;
& Manoj Sham/Animals Animals/Earth Scenes. Figure 17.9: Jen and Des Bartlett/Photo Researchers, Inc.; Figure
17.10: Toni Angermayer/Photo Researchers, Inc.; Figure 17.12:
Chapter 14 Phil Devries/Oxford Scientific Films Ltd.; Figure 17.14: Papilio/
Figure 14.1: Piotr Naskrecki/Minden Pictures, Inc.; Figure 14.3: Alamy; Figure 17.16: Molly Cummings, Division of Integrative
Courtesy Darryl Gwynne, Biology Department, University of Biology, University of Texas at Austin; Figure 17.18: Susan
 Photo Credits 515

Kuklin/Photo Researchers, Inc.; Figure 17.21: Courtesy Georgia Earth Scenes; Figure 19.8: terrygray; Figure 19.10: Richard R.
State University & Yerkes/Emory; Figure 17.22: Courtesy Hansen/Photo Researchers; Figure 19.12: Ian Cleghorn/Photo
Georgia State University & Yerkes/Emory; Figure 17.24: Jeffrey Researchers, Inc.; Figure 19.13a: Image from: Grinnell, J., C.
Oonk/Foto Natura/Minden Pictures, Inc. Packer, and A. E. Pusey. Cooperation in male lions: Kinship, rec-
iprocity, or mutualism. Animal Behaviour 49:95–105; Figure
Chapter 18 19.13b: Image from: Grinnell, J., C. Packer, and A. E. Pusey,
Figure 18.1: Hiroya Minakuchi/Minden Pictures, Inc.; Figure 1995. Cooperation in male lions: kinship, reciprocity, or mutu-
18.2: Doug Weschler/Animals Animals/Earth Scenes; Figure alism. Animal Behaviour 49:95–105; Figure 19.15: Leanne T.
18.3: Frank Lane Picture Agency; Figure 18.5: Clem Haagner Nash; Figure 19.16: Steve & Dave Maslowski/Photo
Photo Researchers, Inc.; Figure 18.6: Figure after Fretwell Researchers, Inc.; Figure 19.17a: ©Biosphoto/Fouquet
1972, and Krebs and Davies 1993; Figure 18.9a: ©Brad Franck/Peter Arnold Inc.; Figure 19.17b: Werner Bollmann/
Hamel/Pbase. Photolibrary Group Limited; Figure 19.18: ©Joe McDonald/
©Corbis–Bettmann; Figure 19.20: ©Biosphoto/Dennis Nigel
Chapter 19 J. /Peter Arnold, Inc.; Figure 19.21: M. K. & I. M. Morcombe/
Figure 19.1: Source: de Waal, F. B. M. 2005. A century of get- © Natural History Photographic Agency; Figure 19.22:
ting to know the chimpanzee. Nature 437:56–59, fig. 2; Figure Komdeur, J. 1996. Influence of helping and breeding experience
19.4: Courtesy W. E. Bemis, Department of Ecology and on reproductive performance in the Seychelles warbler: A
Evolutionary Biology, Cornell University; Figure 19.5: translocation experiment. Behavioral Ecology 7:326–333; Figure
Photograph by Bill Romey, Department of Biology, SUNY 19.24: Raymond A. Mendez/Animals Animals/Earth Scenes.
Potsdam; Figure 19.6: Leonard Lee Rue, III/Animals Animals/
This page intentionally left blank
 Permission Credits

Chapter 2 gen receptor aplha- and ß-knockout mice. Proceedings of the

Table 2.1 Fraser, D., and D. M. Weary. 2005. Applied animal National Academy of Sciences USA 100:6192– 6197. Copyright
behavior and animal welfare. In The Behavior of Animals: (2003) National Academy of Sciences, U.S.A. Fig. 3.16c–d
Mechanisms, Function, and Evolution, edited by J. J. Bolhuis and Lim, M. Zuoxin, W., Olazabal, D. E., Ren, X., Terwilliger, E.,
L.-A. Giraldeau. Wiley-Blackwell Publishing. Fig 2.2 et al. 2004. Enhanced partner preference in a promiscuous
Romanes, G. J. 1889. Mental Evolution in Man: Origin of Human species by manipulating the expression of a single gene. Nature
Faculty. Appleton and Lange. Fig. 2.4 Kessel, E. L. 1955. 429:754–757. Copyright 2004. Reprinted by permission from
Systematic Zoology 4. Society of Systematic Zoology. Fig. 2.5 Macmillan Publishers Ltd. Fig. 3.17 Based on Robinson, G. E.
Lorenz, K., and N. Tinbergen. 1938. Zeitschrift für 2004. Beyond nature and nurture. Science 304:397–399. Fig.
Tierpsychologie 2, Paul Parey Scientific Publications. Fig. 2.6 3.18c–f Burmeister, S. S., E. D. Jarvis, and R. D. Fernald. 2005.
Tinbergen, 1989. Courtship ritual of stickleback. The Study of Rapid behavioral and genomic responses to social opportunity.
Instinct. By permission of Oxford University Press. Fig. 2.7 PLoS Biology 3:1996–2004. Fig. 3.19 Mello, C., D. S. Vicario,
Thorndike, E. L. 1911. Animal Intelligence: Experimental Studies. and D. F. Clayton. 1992. Song presentation induces gene
Macmillan Publishing Company. expression in the songbird forebrain. Proceedings of the National
Academy of Sciences 89:6818–6822. Fig. 3.20 Journal of
Chapter 3 Neuroscience 15, Mello, C., and F. Nottebohm, Society for
Table 3.1 Modified from Tan, E. J., and B. L. Tang. 2006. Neuroscience, 1995. Fig. 3.21 Aubin-Horth, N., et al. 2005.
Looking for food: Molecular neuroethology of invertebrate Alternative life histories shape brain gene expression profiles in
feeding behavior. Ethology 112:826–832. Fig. 3.3 Miklósi, A., V. males of the same population. Proceedings of the Royal Academy of
Csányi, and R. Gerlai. 1997. Behavior Genetics 27(3). Plenum Sciences 272: 1655–1662. Figure 3, page 1659. The Royal
Publishing. Fig. 3.5 Lynch, C. B. 1980. Response to divergent Society. Fig. 3.22 From Robinson, G., et al. 2008. Genes and
selection for nesting behavior in Mus musculus. Genetics social behavior. Science. 322:896–900. Figure 1, page 897.
96:757–765. Fig. 3.6 de Belle, J. S., and M. B. Sokolowski. Reprinted with permission from AAAS.
1987. Heredity of rover/sitter: Alternative foraging strategies of
Drosophila melanogaster larvae. Heredity 59:73–83. Fig. 3.7 de Chapter 4
Belle, J. S., and M. B. Sokolowski. 1987. Heredity 59. Blackwell Fig. 4.11 Brockmann J., A. Grafen and R. Dawkins, Journal of
Science Ltd. Fig. 3.8 Fitzpatrick, M. J., Y. Ben-Shahar, H. M.
Theoretical Biology 77, Academic Press Ltd., 1979. Fig. 4.14
Smid, L. E. M. Vet, G. E. Robinson, and M. B. Sokolowski.
Modified from Rosenheim, J. A. 1993. Comparative and
2005. Candidate genes for behavioral ecology. Trends in Ecology
experimental approaches to understanding insect learning. In
and Evolution 20 (2):96–104. Fig. 3.9 Modified from Tan, E. J.,
Insect Learning: Ecological and Evolutionary Perspectives, edited by
and B. L. Tang. 2006. Looking for food: Molecular neuroethol-
D. R. Papaj and A. C. Lewis. New York: Chapman and Hall,
ogy of invertebrate feeding behavior. Ethology 112:826–832.
with kind permission of Springer Science and Business Media.
Wiley-Blackwell Publishing. Fig. 3.10b–c From Ben-Shahar,
Y., A. Robichon, M. B. Sokolowski, and G. E. Robinson. 2002.
Influence of gene action across different time scales on behav- Chapter 5
ior. Science 296:741–744. Reprinted with permission from Fig. 5.1 Reprinted from Clark, R. B. 1960. Habituation of the
AAAS. Fig. 3.11 Modified from Hunt, G. J., G. V. Amdam, D. polychaete Nereis to sudden stimuli. 1. General properties
Schlipalius, C. Emore, N. Sardesai, C. E. Williams, O. of the habituation process. Animal Behaviour 8 (1–2):87.
Rueppell, E. Guzmán-Novoa, M. Arechavaleta-Velasco, S. B. Copyright 1960, by permission of the publisher Academic
C. Chandra, M. K. Fondrk, M. Beye, and R. E. Page, Jr. 2007. Press. Fig. 5.2 Hollis, K. L., V. L. Pharr, M. J. Dumas, G. B.
Behavioral genomics of honeybee foraging and nest defense. Britton, and J. Field. 1997. Classical conditioning provides
Naturwissenschaften 94:247–267, Figure 1. Copyright 2007. paternity advantage for territorial male blue gouramis
With kind permission of Springer Science and Business Media. (Trichogaster trichopterus). Journal of Comparative Psychology
Fig. 3.12 From Whitfield, C. W., A. M. Cziko, and G. E. 111:219–225. Fig. 5.3 Reprinted from Lyons, C., and C. J.
Robinson. 2003. Gene expression profiles in the brain predict Barnard. 2006. A learned response to sperm competition in the
behavior in individual honey bees. Science 302:296–299. field cricket, Grullus bimaculatus (de Geer). Animal Behaviour
Reprinted with permission from AAAS. Fig 3.13 Nelson, C. 72:673–680. With permission from Elsevier. Fig. 5.5 Franks,
M., et al. 2007. The gene vitellogenin has multiple coordinating N. R., J. W. Hooper, A. Dornhaus, P. J. Aukett, A. L. Hayward,
effects on social organization. PLoS Biology 5 (3):0673–0677. and S. M Berghoff. 2007. Reconnaissance and latent learning
Fig. 3.14 Amholt, R. R. H. 2004. Interactions among genetic in ants. Proceedings of the Royal Society of London B 274:1505–
modules. Genetic modules and networks for behavior: Lessons 1509. Fig. 5.9 Balda, R. P., and A. C. Kamil. 1989. A compara-
from Drosophila. BioEssay 26 (12):1299–1306. Reprinted with tive study of cache recovery by three corvid species. Animal
permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Behaviour 38:486–495. Fig. 5.10 Olson, D. J., A. C. Kamil, R.
Sons, Inc. Fig. 3.15 Choleris, E., J. Å. Gustafson, K. S. Korach, P. Balda, and P. J. Nims. Performance of four seed-caching
J. J. Muglia, and S. Ogawa. An estrogen dependent micronet corvid species in operant test of nonspatial and spatial memory.
mediating social recognition. A study with oxytocin-and estro- Journal of Comparative Psychology 109:173–181. Copyright 1995

517
518 Permission Credits

by the American Psychological Association. Adapted with per- and N. G. Forger. 2006. Breeding status affects motoneuron
mission. Fig. 5.13 Shettleworth, S. J. 1998. Cognition, Evolution number and muscle size in naked mole-rats: Recruitment of
and Behavior. Oxford University Press. Fig. 5.15 Reprinted perineal motoneurons. Journal of Neurobiology, 66:1354–1364.
from Tarsitano, M. S., and A. R. Andrew. 1999. Scanning and (2006). Reprinted with permission of Wiley-Liss, Inc., a sub-
route selection in the jumping spider Portia labiata. Animal sidiary of John Wiley & Sons, Inc. Fig. 7.8 Crews, D. 1979.
Behaviour 58:255–265, Copyright 1999, with permission from The hormonal control of behavior in a lizard. Scientific
Elsevier. American 241:180–187. Fig. 7.9 Crews, D. 1979. Biology of
Reproduction 20. Society for the Study of Reproduction. Fig.
Chapter 6 7.10a–b Schneider, J. S., M. K. Stone, K. E. Wynne-Edwards,
Fig. 6.2 Camhi, J. M. 1980. The escape system of the cock- T. H. Horton, J. Lydon, B. O’Malley, and J. E. Levine.
roach. Scientific American 248:158–172. Fig. 6.2 Ritzmann, R. Progesterone receptors mediate male aggression toward
E., and A. J. Pollack. 1986. Identification of thoracic interneu- infants. Proceedings of the National Academy of Sciences 100:2951–
rons that mediate giant interneuron-to motor pathways. 2956. Copyright 2003 National Academy of Sciences, U.S.A.
Journal of Comparative Physiology A 159:639–654. Fig 6.10 Fig. 7.12 Wingfield, J. C. 1984. General and Comparative
Alcock, J. 1975. Animal Behavior. Sinauer Associates. Fig. 6.13 Endocrinology 56. Academic Press, Inc. Fig. 7.15 Orchinik, M.
From Knudsen, E. I., and M. Konishi, 1978. A neural map of P., Licht, and D. Crews. 1988. Plasma steroid concentrations
auditory space in the owl. Science 200:795–797, Copyright change in response to sexual behavior in Bufo marinus.
1978. Reprinted with permission from AAAS. Fig. 6.18 From Hormones and Behavior 22:338–350. Academic Press, Inc. 1988.
Kandel, E. R. 2001. The molecular biology of memory storage: Fig. 7.16 Crews, D., ed. 1987. Psychobiology of Reproductive
A dialogue between genes and synapses. Science 294:1030– Behavior: An Evolutionary Perspective. Prentice-Hall, p. 92.
1038. Reprinted with permission from AAAS. Fig. 6.20 Copyright 1987, Reprinted by permission of Prentice-Hall,
Wilbrecht, L., A. Crionas, and F. Nottebohm. 2002. Inc., Englewood Cliffs, NJ. Fig. 7.18 Dzieweczynski, T. L., R.
Experience affects recruitment of new neurons but not adult L. Earley, T. M. Green, and W. J. Rowland. 2005. Audience
neuron number. The Journal of Neuroscience 22:825–831. Fig. effect is context dependent in Siamese fighting fish, Betta splen-
6.21 Cameron, H. A., and R. D. G. McKay. 2001. Adult neuro- dens. Behavioral Ecology 16:1025–1030, by permission of Oxford
genesis produces a large pool of new granule cells in the den- University Press. Fig. 7.19 Reprinted from Dzieweczynski, T.
tate gyrus. Journal of Comparative Neurology 435:406–417. L., A. C. Eklund, and W. J. Rowland. 2006. Male 11-
Figure 6, page 413. Reprinted with permission of Wiley-Liss, ketotestosterone levels change as a result of being watched in
Inc., a subsidiary of John Wiley & Sons, Inc. Fig. 6.22 Siamese fighting fish, Betta splendens. General and Comparative
Reprinted by permission from Macmillan Publishers Ltd.: Endocrinology 147:184–189. Copyright 2006, with permission
Nature. Shors, T. J., G. Miesegaes, A. Beylin, M. Zhao, T. from Elsevier. Fig. 7.20 Reprinted from Schoech, S. J., R. L.
Rydel, and E. Gould. Neurogenesis in the adult is involved in Mumme, and J. C. Wingfield. 1996. Prolactin and helping
the formation of trace memories. Nature 410:372–376. behaviour in the cooperatively breeding Florida scrub-jay,
Copyright 2001. Fig. 6.23 Dupret, D., A. Fabre, M. D. Apheloma e. coerulesens. Animal Behaviour 52 (3).
Dobrossy, A. Panatier, J. J. Rodríguez, S. Lamarque, V.
Lemaire, S. H. R. Oliet, P.-V. Pliazza, and D. N. Abrous. 2007. Chapter 8
Spatial learning depends on both the addition and removal of Table 8.1 Gottlieb, G. Journal of Comparative and Physiological
new hippocampal neurons. PLoS Biology 5:1683–1694. Fig. 6.24 Psychology, Copyright 1978 by the American Psychological
Maguire, E. A., D. G. Gadlian, I. S. Johnsrude, C. D. Good, J. Association. Adapted with permission. Table 8.2 Gottlieb, G.
Ashburner, R. S. J. Frackowiak, and C. D. Frith. Navigation- 1985. Animal Behavior 33, Academic Press Ltd. Table 8.3
related structural change in the hippocampi of taxi drivers. Rauschecker, J. R., and P. Marler. 1987. Imprinting and Cortical
Proceedings of the National Academy of Sciences USA 97:4398– Plasticity: Comparative Aspects of Sensitive Periods. Reprinted by
4403. Copyright (2000) National Academy of Sciences, U.S.A. permission of John Wiley & Sons, Inc. Fig. 8.2 Journal of
Fig. 6.25 Newman, S. W. 1999. The medial extended amygdala Zoology London 149, Abu-Gideiri, Y. B., Reprinted with the per-
in male reproductive behavior: a node in the mammalian social mission of Cambridge University Press, 1966, Animal
behavior network. Annals of the New York Academy of Sciences Behaviour, Dill, 1977. Fig. 8.2 Animal Behaviour, Abu-Gideiri, Y.
877:742–257. Wiley-Blackwell Publishers. Fig. 6.26 Goodson, B., Academic Press Ltd., 1966., Animal Behaviour, Dill, 1977.
J. L., A. K. Evans, L. Lindberg, and C. D. Allen. 2005. Neuro- Fig. 8.4 Weeks, J. C., and J. W. Truman. 1984. Journal of
evolutionary patterning of sociality. Proceedings of the Royal Comparative Physiology 155. Springer-Verlag. Fig. 8.6 McLaren,
Society of London B 272:227–235. Fig. 6.28a Weis-Fogh, T. A., and D. Michie. 1960. Control of prenatal growth in mam-
1956. Philosophical Transactions of the Royal Society of London mals. Nature 187:363–365. Fig. 8.7 Crews, D., and T.
(Series B). The Royal Society. Fig. 6.28a Horsmann, U., H. G. Groothuis. Tinbergen’s fourth question, ontogeny: Sexual and
Heinzel, and G. Wendler. 1983. The phasic influence of individual differentiation. Animal Biology 55:343–370.
self–generated air current modulations on the locust flight Copyright 2005 by Koninklijke Brill NV Leiden. Fig. 8.11
motor. Journal of Comparative Physiology 150:427–438. Fig. Copyright © 1965 by the American Psychological Association.
6.28b–c Wilson, D. M., and T. Weis-Fogh. 1962. Journal of Adapted with permission. Gottlieb, G. Imprinting in relation to
Experimental Biology 39. Company of Biologists, Ltd. parental and species identification by avian neonates. Journal of
Comparative and Physiological Psychology 59: 345–356. Fig. 8.12
Chapter 7 Bateson, P. 1982. Nature 295. Reprinted by permission from
Fig. 7.1 Bass, A. H. 1996. Shaping brain sexuality. American Macmillan Magazines Ltd. Fig. 8.13 Reprinted from Kendrick,
Scientist 84:352–363. Fig. 7.6 Kelley, D. B., and D. I. Gorlick. et al. 2001. Sex differences in the influence of mothers on the
1990. Sexual selection and the nervous system. BioScience sociosexual preferences of their offspring. Hormones and
40:275–283. Fig. 7.7 Adapted from M. Seney, B., D. Goldman, Behavior 40:322–338. Copyright 2001, with permission from
 Permission Credits 519

Elsevier. Fig. 8.15 Reprinted from Champalbert, A., and J. Elsevier. Fig. 9.14b Silver, R., J. LeSauter, P.A., Tresco, and M.
Lachaud. 1990. Existence of a sensitive period during the onto- N. Lehman. 1996. A diffusible coupling signal from the trans-
genesis of social behavior in a primitive ant. Animal Behaviour planted suprachiasmatic nucleus controlling circadian locomo-
39:850–859. Copyright 1990, by permission of the publisher tor rhythms. Nature 382:810–813. Reprinted by permission of
Academic Press Ltd. 8.17a Adapted from Bolhuis, J. J. 2008. Macmillan Magazines Ltd. Fig. 9.15 Reprinted from Aschoff, J.
Chasin' the trace: The neural substrate of birdsong memory. In 1967. Life Sciences and Space Research. North Holland
The Neuroscience of Birdsong, edited by H. P. Zeigler and P. Publishers; Elsevier Science Publishing Co., Inc., with permis-
Marler. Cambridge University Press. Reprinted with the per- sion from Elsevier Science.
mission of Cambridge University Press. 8.17b From
Nottebohm and Arnold. Sexual dimorphism in vocal control Chapter 10
areas of the songbird brain. Science 194. Copyright 1976. Tables 10.1–10.2 Wiltschko, R., and W. Wiltschko. 2006.
Reprinted with permission from AAAS. Fig. 8.18 From Magnetoreception. BioEssays 28:157–168. Reprinted with per-
Gurney, M. E., and M. Konishi. Hormone-induced sexual dif- mission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons,
ferentiation of brain and behavior in zebra finches. Science 208. Inc. Fig. 10.2 Perdeck, A. C. 1958. Two types of orientation in
Copyright 1980. Reprinted with permission from AAAS. Fig. migrating starlings, Sturnus vulgaris L., and chaffinches,
8.20 Konishi, M. 1965. Zeitschrift für Tierpsychologie 22. Paul Fringilla coelebs L., as revealed by displacement experiments.
Parey Scientific Publications. Fig. 8.22b West, M. J., A. P. King, Ardea 46:1–37. Netherlands Ornithological Union, 1958. Fig.
and D. H. Eastzer. 1981. The cowbird: Reflections on develop- 10.10 Palmer, J. D., A. G. Gilbert, and The Bettmann Archive.
ment from an unlikely source. American Scientist 69:56–66. 1966. Natural History 75. Fig. 10.13 Quarterly Review of Biology
Scientific Research Society, 1981. Fig. 8.23 King, A. P., and M. 62, Able and Bingman, The University of Chicago Press, 1987,
J. West. 1983. Nature 305. Reprinted by permission from modified from Science 170, Emlen, S. T., American Association
Macmillan Magazines Ltd. Fig. 8.24 Reprinted from Reimers, for the Advancement of Science, 1970. Fig. 10.18 Wiltschko,
et al. 2007. Rehabilitation of research chimpanzees: Stress and W., and R. Wiltschko. 2005. Magnetic orientation and magne-
coping after long-term isolation. Hormones and Behavior 51: toreception in birds and other animals. Journal of Comparative
428–435. Copyright 2007, with permission from Elsevier. Physiology A 191:675–693, With kind permission from Springer
Science+Business Media. Fig. 10.18 Data: S. Marhold. 1997. A
Chapter 9 magnetic polarity compass for direction finding in a subter-
Fig. 9.2 This figure was published in The Biological Clock-Two ranean mamma. Naturwissenschaften 84:421–423. With kind
Views by F., A. Brown, J. W. Hastings, and J. D. Palmer, permission from Springer Science+Business Media. Fig 10.19
Copyright Academic Press 1970. Fig. 9.3 Palmer, J. D. 1990. Walcott, C., and R. P. Green. 1974. Science 184. American
The rhythmic lives of crabs: the same biological clock that con- Association for the Advancement of Science. Fig 10.21b
trols circadian rhythms may also drive crustacean tidal rhythms. Lohmann, K. J. 1991. Journal of Experimental Biology 155. The
BioScience 40:352–358. Copyright 1990 American Institute of Company of Biologists Ltd. Fig. 10.22 Data from Beck, W., and
Biological Sciences. Fig. 9.5b Youthed, G. J., and R. C. Moran. W. Wiltschcko, 1988. Magnetic factors control the migratory
1969. The lunar day activity rhythm of myrmeleontid larvae. direction of pied flycatchers (Ficedula hypoleuca Pallas) Acta
Journal of Insect Physiology 15:1259–1271. Reprinted with per- Congress International Ornithology 19:1955–1962. In Lohmann,
mission from Pergamon Press, Inc. Fig. 9.6 Gwinner, E. 1996. K. J., C.M.F. Lohmann, and N. F. Putman, 2007. Magnetic
Circadian and cirannual programmes in avian migration. maps in animals: nature’s GPS. Journal of Experimental Biology,
Journal of Experimental Biology 199:39–48. The Company of 210:3697–3705. Fig. 10.23 From Lohmann, K. J., S. D. Cain, S.
Biologists Ltd. Fig. 9.7 Welsh, D. K., D. E. Logothetis, M. A. Dodge, and C. M. F. Lohmann. 2001. Regional magnetic
Meister, and S. M. Reppert. 1995. Individual neurons dissociat- fields as navigational markers for sea turtles. Science
ed from rat suprachiasmatic nucleus express independently 294:364–366. Reprinted with permission from AAAS. Fig. 10.24
phased circadian firing rhythms. Neuron 14:697–706. Reprinted Gould, J. L. 1980. American Science 68. Sigma Xi Scientific
with permission from Elsevier. Fig. 9.8c Beling, I. 1929. Über Research Society. Fig 10.25 Lohmann, K. J., C. M. F.
das Zeitgedächtnis der Bienen. Zeitschrift für vergleichende Lohmann, L. M. Ehrhart, D. A. Bagley, and T. Swing. 2004.
Physiologie 9:259–338. Fig. 9.9b Plautz, J. D., M. Kaneko, J. C. Geomagnetic map used in sea-turtle navigation. Nature
Hall, and S. A. Kay. 1997. Independent photoreceptive circadi- 428:909–910. Copyright 2004. Reprinted by permission from
an clocks throughout Drosophila. Reprinted with permission Macmillan Publishers Ltd: Nature. Fig 10.26 Ritz, T., S. Adem,
from AAAS. Fig. 9.10 Adapted Reppert, S. M., and D. R. and K. Schulten. 2000. A model for photoreceptor-based mag-
Weaver. 2002. Coordination of circadian timing in mammals. netoreception in birds. Biophysical Journal 78:707–718.
Nature 418:935–941. By permission from Macmillan Reprinted with permission from Elsevier. Fig. 10.27 Stewart, I.
Publishers, Ltd. Fig. 9.11 Inouye, S. T., and H. Kawamura. J., S. M. Carlson, C. P. Boatright, G. B. Buck, and T. P. Quinn.
1979. Persistence of circadian rhythmicity in mammalian hypo- 2004. Site fidelity of spawning sockeye salmon (Oncorhynchus
thalamic “island” containing the suprachiasmatic nucleus. nerka W.) in the presence and absence of olfactory cues. Ecology
Proceedings of the National Academy of Sciences 76:5962–5966. Fig. of Freshwater Fish 13:104–110. Wiley-Blackwell Publishers. Fig
9.12 Based on Breedlove, S. M., M. R. Rosenweig, and N. V. 10.28 Baldaccini, E. N., S. Benvenuti, V. Fiaschi, and F. Papi.
Watson. 2007. Biological Psychology: An Introduction to Behavioral, 1975. Journal of Comparative Physiology 99. Springer-Verlag. Fig
Cognitive, and Clinical Neuroscience. 5th ed. Sunderland MA: 10.29 Ioale, P., M. Nozzolini, and F. Papi. 1990. Behavioral
Sinauer. Fig. 9.12 Based on Reppert, S. M., and D. R. Weaver. Ecology and Sociobiology. Springer-Verlag. Fig. 10.30 Adapted
2002. Coordination of circadian timing in mammals. Nature with permission. Gagliardo, A., P. Ioalè, M. Savini, and J. M.
418:935–941. Fig. 9.13 Reprinted from Herzog, E. D., and G. Wild. 2006. Having the nerve to home: trigeminal magnetore-
Tosini. 2001. The mammalian circadian clockshop. Seminars in ceptor versus olfactory mediation of homing in pigeons. Journal
Cell & Developmental Biology 12:295–303. With permission from of Experimental Biology 209:2888– 2892. Figs. 10.31–10.33 Von
520 Permission Credits

der Emde, G. 1999. Journal of Experimental Biology. The Advancement of Science. Table 13.3 Modifed from Honma, A.,
Company of Biologists Ltd. S. Oku, and T. Nishida. 2006. Adaptive significance of death
feigning posture as a specialized inducible defence against
Chapter 11 gape-limited predators. Proceedings of the Royal Society of London
Table 11.1 Greenwood, P. J. 1980. Mating systems, philopatry, B 273:1631–1636. Table 13.4 Modified from Vallin, A., S.
and dispersal in birds and mammals. Animal Behavior Jakobsson, J. Lind, and C. Wiklund. 2005. Prey survival by
28:1140–1162. Reprinted by permission of the publisher, predator intimidation: an experimental study of peacock butter-
Academic Press London. Table 11.2 Dobson, F. S. 1982. fly defence against blue tits. Proceedings of the Royal Society of
Competition for mates and predominant juvenile male disper- London B 272:1203–1207. Table 13.5 Caro, T. M. 1986a. The
sal in mammals. Animal Behavior 30:1183–1192. Reprinted by functions of stotting: A review of hypotheses. Animal Behaviour
permission of the publisher, Academic Press London. Fig 11.2 34:649–662. Table 13.6 Caro, T. M. 1986b. The functions of
Bonte, D., S. Van Belle, and J.-P. Maelfait. 2007. Maternal care stotting in Thomson’s gazelles: Some tests of the predictions.
and reproductive state-dependent mobility determine natal dis- Animal Behaviour 34:663–684. Reprinted from Animal
persal in a wolf spider. Animal Behaviour 74:63–69. Reprinted Behaviour, Caro, 1986, by permission of the publisher
with permission from Elsevier. Fig 11.3 Höner, O. P., B. Academic Press London. Figs. 13.3–13.4 S. Merilaita, J. Lind.
Wachter, M. L. East, W. J. Streich, K. Wilhelm, T. Burke, and 2005. Background-matching and disruptive coloration, and the
H. Hofer. 2007. Female mate-choice drives the evolution of evolution of cryptic coloration. Proceedings of the Royal Society of
male-biased dispersal in a social mammal. Nature 448:798–802. London B 272:665–670. The Royal Society. Fig. 13.5 Reprinted
Reprinted by permission from Macmillan Publishers Ltd: from Feltmate, B. W., and D. D. Williams. 1989. A test of cryp-
Nature. Fig. 11.4 Devillard, S., D. Allaine, J. M. Gaillard, and sis and predator avoidance in the stonefly Paragnetina media
D. Pontier. 2004. Does social complexity lead to sex-biased dis- (Plecoptera: Perlidae). Animal Behaviour 37:992–999. Figs.
persal in polygynous mammals? Behavioral Ecology 15:83–87, by 13.7– 13.8 Cuthill, I. C., M. Stevens, J. Sheppard, T. Maddocks,
permission of Oxford University Press. Fig. 11.6 Forsman, J. C. Alejandro Párraga, and T. S. Troscianko. 2005. Disruptive
T., J.-T. Seppänen, and M. Mönkkönen. 2002. Positive fitness coloration and background pattern matching. Nature
consequences of interspecific interaction with a potential com- 434:72–74. Copyright 2005. Reprinted by permission from
petitor. Proceedings of the Royal Society of London B Macmillan Publishers Ltd: Nature. Fig. 13.9 Cott, H. B. 1940.
269:1619–1623. The Royal Society. Fig 11.10 Strandberg, R., Adaptive Colouration in Animals. Methuen, London, London.
and T. Alerstam. 2007. The strategy of fly-and-forage migra- Fig. 13.12 Endler, J. A. 1978. Evolutionary Biology 11:346–348,
tion, illustrated for the osprey (Pandion haliaetus). Behavioral 354. With kind permission of Springer Science and Business
Ecology and Sociobiology 61:1865–1875. With kind permission Media. Figs. 13.13–13.14 Kiltie, R. A. 1989. Journal of
from Springer Science and Business Media. Fig 11.13 Mammalogy 70. American Society of Mammalogists. Fig. 13.15
Anderson, J. B., and L. P. Brower. 1996. Ecological Entomology 21 Moment, G. 1962. Reflexive selection: A possible answer to an
(2). Blackwell Scientific Ltd. old question. Science 136:262–263. Reprinted with permission
from American Association for the Advancement of Science.
Chapter 12 Fig. 13.17 Terrick, T. D., R. L. Mumme, and G. M. Burghardt.
Table 12.1 Acknowledgement to the original publication. Zach, 1995. Aposematic coloration enhances chemosensory recogni-
R. 1978, Behaviour. 67, E. J. Brill. Fig. 12.1 Brusca, R. C., and tion of noxious prey in the garter snake. Animal Behaviour 49
G. J. Brusca. 1990. Invertebrates. Sunderland, MA: Sinauer (4):857–866. By permission of the publisher, Academic Press
Associates. Fig. 12.2 Gurd, D. B. 2007. Predicting resource Ltd. Fig. 13.19 Wickler, W. 1968. Mimicry in Plants and
partitioning and community organization of filter-feeding dab- Animals. McGraw-Hill, New York. Fig. 13.21 Burghardt and
bling ducks from functional morphology. American Naturalist Greene. 1988. Predator simulation and duration of death feign-
169:334–343. The University of Chicago Press. Fig. 12.5 ing in neonate hognose snakes. Animal Behaviour
Olberg, R. M., A. H. Worthington, and K. R. Venator. 2000. 36:1842–1844. By permission of the publisher, Academic Press
Prey pursuit and interception in dragonflies. Journal of London. Fig. 13.22 Johnsgard, P. A. 1967. Animal Behavior.
Comparative Physiology A 186:155–162. With kind permission of William C. Brown Publishers. Fig. 13.24 Vallin, A., S.
Springer Science and Business Media. Fig. 12.11 Reprinted Jakobsson, J. Lind, and C. Wiklund. 2005. Prey survival by
from Watt, M., C. S. Evas, and J. M. P. Joss. 1999. Use of elec- predator intimidation: an experimental study of peacock butter-
troreception during foraging by the Australian lungfish. Animal fly defence against blue tits. Proceedings of the Royal Society of
Behaviour 58:1039–1045. Reprinted by permission of the pub- London B 272:1203–1207. The Royal Society. Fig. 13.27
lisher, Academic Press Ltd. Fig. 12.14 Pietrewics, A. T., and A. Reprinted from Hews, D. K. 1988. Alarm response in larval
C. Kamil. 1981. Foraging Behavior: Ecological, Ethological, and western toads, Bufo boreas: Release of larval chemical by a nat-
Psychological Approaches. Garland STPM Press. Fig. 12.19 ural predator and its effect on predator capture efficiency.
Behaviour 68, Zach, R., E. J. Brill Leiden, 1979. Fig. 12.20b Animal Behaviour 36:125–133. By permission of the publisher,
Cowlishaw, G. 1997. Trade-offs between foraging and preda- Academic Press London. Fig. 13.28 Turchin, P., and P. Kareiva.
tion risk determine habitat use in a desert baboon population. 1989. Ecology 70. Ecological Society of America. Fig. 13.29
Animal Behaviour 53 (4):667–686. Reprinted by permission of Neill, S. R. St. J., and J. M. Cullen. 1974. Journal of Zoology 172.
the publisher, Academic Press Ltd. Reprinted with the permission of Cambridge University Press.

Chapter 13 Chapter 14
Table 13.1 Croze, H. 1970. Zeitschrift für Tierpsychologie Table 14.1 Modified from Schütz, D., and M. Taborsky. 2005.
Supplement 5, Paul Parey Scientific Publications. Table 13.2 The influence of sexual selection and ecological constraints on
Dial, B. E., and L. C. Fitzpatrick. 1983. Science 219. Reprinted an extreme sexual size dimorphism in a cichlid. Animal
with permission from American Association for the Behaviour 70:539–549. Table 14.2 Fisher, A. E. 1962. Journal of
 Permission Credits 521

Comparative and Physiological Psychology 55. American a sexually size-dimorphic bird: A comparison of different meth-
Psychological Association. Table 14.3 Modified from Holman, ods. Journal of Animal Ecology 76:1169–1180, British Ecological
L., and R. R. Snook. 2006. Spermicide, cryptic female choice Society, Wiley-Blackwell Publishing. Fig. 15.4b Magrath, M. J.
and the evolution of sperm form and function. Journal of L., E. van Lieshout, G. Henk Visser, and J. Komdeur. 2004.
Evolutionary Biology 19:1660–1670. Table 14.4 Modified from Nutritional bias as a new mode of adjusting sex allocation.
Ebensperger, L. A., and D. T. Blumstein. 2007. Nonparental Proceedings of the Royal Society of London B 271: S347–S349. The
infanticide. In Rodent Societies: An Ecological and Evolutionary Royal Society. Fig. 15.6 Mank, J. E., D. E. L. Promislow, and J.
Perspective, edited by J. O. Wolff, and P. W. Sherman. Chicago: C. Avise. 2005. Phylogenetic perspectives in the evolution of
The University of Chicago Press, 267–279. Fig. 14.2 Allen, B. parental care in ray-fined fishes. Evolution 59:1570–1578,
J., and J. S. Levinton. 2007. Costs of bearing a sexually select- Wiley-Blackwell Publishers. Fig. 15.10 Blackmore, C. J. and R.
ed ornamental weapon in a fiddler crab. Functional Ecology Heinsohn. 2008. Variable mating strategies and incest avoid-
21:154–161, Blackwell Publishing. Fig. 14.4 Rooney, J., and S. ance in cooperatively breeding grey-crowned babblers. Animal
M. Lewis. 2002. Fitness advantage from nuptial gifts in female Behaviour 75:63–70. Reprinted with permission from Elsevier.
fireflies. Ecological Entomology 27:373–377, Wiley-Blackwell Fig. 15.11 Féron, C., and P. Gouat. 2007. Paternal care in the
Publishing. Fig. 14.7 Willis, P. M., and L. M. Dill. 2007. Mate mound-building mouse reduces inter-litter intervals.
guarding in male Dall’s porpoises (Phocoenoides dalli). Ethology Reproduction, Fertility, and Development 19:425–429. CSIRO
113:587–597, Blackwell Publishing. Fig. 14.13 Sato, T., and S. Publishing, http://www.pulish.csiro.au/ nid/45/3365.htm. Fig.
Goshima. 2007. Female choice in response to risk of sperm lim- 15.12b Baeza, J. A. 2008. Social monogamy in the shrimp
itation by the stone crab, Hapalogaster dentata. Animal Behaviour Pontonia margarita, a symbiont of Pinctada mazatlanica, off the
73. Reprinted with permission from Elsevier. Fig. 14.14 Pacific coast of Panama. Marine Biology 153:387–395. With
Eggert, A.-K., and S. K. Sakaluk. 1994. Sexual cannibalism and kind permission from Springer Science+Business Media. Fig.
its relation to male mating success in sagebrush crickets, 15.13 Orians, G. H. 1969. On the evolution of mating systems
Cyphoderris strepitans (Haglidae: Orthoptera). Animal Behaviour in birds and mammals. American Naturalist 103. The University
47(5), Pages 1171–1177. Reprinted with permission of the pub- of Chicago Press. Fig. 15.14 Huk, T., and W. Winkel. 2006.
lisher, Academic Press Ltd. Fig. 14.16b Takahashi, D., and M. Polygyny and its fitness consequences for primary and sec-
Kohda. 2004. Courtship in fast water currents by a male stream ondary female pied flycatchers. Proceedings of the Royal Society of
goby (Rhinogobius brunneus) communicates the parental quality London B 273:1681–1688. The Royal Society. Fig. 15.18 From
honestly. Behavioral Ecology and Sociobiology 55:431–438. Mattila, H. R., and T. D. Seeley. 2007. Genetic diversity in
Reprinted with kind permission from Springer Science and honey bee colonies enhances productivity and fitness. Science
Business Media. Fig. 14.19 Byers, J. A., J. D. Moodie, and N. 317:362–364. Reprinted with permission from AAAS.
Hall. Pronghorn females choose vigorous mates. Animal
Behaviour 47:33–43. Reprinted with permission from Elsevier. Chapter 16
Fig. 14.21 Jones, T. M., R. J. Quinnell, and A. Balmford. 1998. Fig. 16.1 Nakano, R., N. Skals, T. Takanashi, A. Surlykke, T.
Fisherian flies: Benefits of female choice in a lekking sandfly. Koike, K. Yoshida, H. Maruyama, S. Tatsuki, and Y. Ishikawa.
Proceedings of the Royal Society of London B 265:1651–1657. The 2008. Moths produce extremely quiet ultrasonic courtship
Royal Society. Fig. 14.22 Fedina, T. Y. 2007. Cryptic female songs by rubbing specialized scales. Proceedings of the National
choice during spermatophore transfer in Tribolium castaneum Academy of Sciences 105:11812–11817. Copyright 2008 National
(Coleoptera: Tenebrionidae). Journal of Insect Physiology Academy of Sciences, U.S.A. Fig. 16.2 Aidan Martin, R. 2007.
53:93–98. Reprinted with permission from Elsevier. A review of shark agonistic displays: Comparison of display fea-
tures and implications for shark-human interactions. Marine
Chapter 15 and Freshwater Behaviour and Physiology 40:3–34. Taylor &
Table 15.1 Modified from Reynolds, J. D., N. B. Goodwin, and Francis, reprinted by permission of the publisher (Taylor &
R. P. Freckleton. 2002. Evolutionary transitions in parental care Francis Ltd., http://www.tandf.co.uk/journals). Fig. 16.4 How,
and live bearing in vertebrates. Philosophical Transactions of the M. J., J. M. Hemmi, J. Zeil, and R. Peters. 2008. Claw waving
Royal Society of London B 357:269–281. Table 15.1. Cockburn, A. display changes with receiver distance in fiddler crabs, Uca per-
2006. Prevalence of different modes of parental care in birds. plexa. Animal Behaviour 75:1015–1022. Reprinted with permis-
Proceedings of the Royal Society of London B 273:1375–1383. Table sion from Elsevier. Fig. 16.12 Hopkins, C. 1974. American
15.2 Gubernick, D. J., S. L. Wright, and R. E. Brown. 1993. Scientist 62. Sigma Xi Scientific Research Society. Fig. 16.16
The significance of father’s presence for offspring survival in Bradbury, J. W., and S. L. Vehrencamp. 1998. Principles of
the monogamous California mouse, Peromyscus californicus. Animal Communication. Sunderland, MA: Sinauer. Fig. 16.17
Animal Behaviour 46:539–546. Table 15.2 Gubernick, D. J., and Houck and Reagan. 1990. Male courtship pheromones increase
T. Teferi. 2000. Adaptive significance of male parental care in a female receptivity in a plethodontid salamander. Animal
monogamous mammal. Proceedings of the Royal Society of London Behaviour 39:729–734. Reprinted by permission of the publish-
B 267:147–150. Fig. 15.1 Rickard, I. J., A. F. Russell, and V. er, Academic Press Ltd. Fig. 16.18 Bradbury, J. W., and S. L.
Lummaa. 2007. Producing sons reduces lifetime reproductive Vehrencamp. 1998. Principles of Animal Communication.
success of subsequent offspring in pre-industrial Finns. Sunderland MA: Sinauer. Based on Moynihan, M. 1966.
Proceedings of the Royal Society of London B 274:2981–2988. The Communication in Callicebus. Journal of Zoology London 150:77–
Royal Society. Fig. 15.3 Neff, B. D. 2003. Decisions about 127. Fig. 16.20 Marler, P., 1959. Darwin’s Biological Work.
parental care in response to perceived paternity. Nature Reprinted with the permission of Cambridge University Press.
422:716–719. Reprinted by permission from Macmillan Fig. 16.22 Reprinted from von Frisch, K. 1950, 1971. Bees:
Publishers Ltd: Nature. Copyright 2003. 15.4a Magrath, M. J. Their Vision, Chemical Senses, & Language, revised edition.
L., E. van Lieshout, I. Pen, G. Henk Visser, and J. Komdeur. Copyright 1950, 1971, by Cornell University. Used by permis-
2007. Estimating expenditure on male and female offspring in sion of the publisher, Cornell University Press. Figs.
522 Permission Credits

16.23–16.24 Reprinted from von Frisch, K. 1950, 1971. Bees: D. C. Paton, and M. A. Hixon. 1983. Weight gain and adjust-
Their Vision, Chemical Senses, and Language, revised edition. ment of feeding territory size in migrant hummingbirds.
Copyright © 1950, 1971, by Cornell University. Used by per- Proceedings of the National Academy of Sciences 80:7259–7263. Fig.
mission of the publisher, Cornell University Press. Fig. 16.26 Figs. 18.9b–d and 18.10 Eason, P. K., G. A. Cobbs, and K. G.
Riley, J. R., U. Greggers, A. D. Smith, D. R. Reynolds, and R. Trinca. 1999. The use of landmarks to define territorial bound-
Menzel. 2005. The flight paths of honeybees recruited by the aries. Animal Behaviour 58:85–91. Reprinted with permission
waggle dance. Nature 435:205–207. Reprinted with permission from Elsevier. Fig. 18.11 Summers, C. H., and S. Winberg.
from Macmillan Publishers Ltd: Nature. Fig. 16.28 From 2006. Interactions between the neural regulation of stress and
Srinivasan, M. V., S. Zhang., M. Altwein, and J. Tautz. 2000. aggression. Journal of Experimental Biology 209:4581–4589. The
Honeybee navigation: Nature and calibration of the “odome- Company of Biologists Ltd.
ter.” Science 287:851–853. Reprinted with permission from
AAAS. Chapter 19
Table 19.2 Sherman, P. W. 1985. Behavioral Ecology and
Chapter 17 Sociobiology 17. Springer-Verlag. Table 19.3 Reyer, H.-U. 1980.
Fig. 17.1 Redondo, T., and F. Castro. 1992. Signaling of nutri- Behavioral Ecology and Sociobiology 6. Springer-Verlag. Fig. 19.1
tional need by magpie nestlings. Ethology 92:193–204. Wiley- de Waal, F. B. M. 2005. A century of getting to know the chim-
Blackwell Publishers. Fig. 17.5 Cheney D. L., and R. M. panzee. Nature 437:56–59. Reprinted by permission from
Seyfarth. 1988. Assessment of meaning and the detection of Macmillan Publishers Ltd: Nature. Fig. 19.2 From Bednarz, J.
unreliable signals by vervet monkeys. Reprinted from Animal C. 1988. Cooperative Hunting Harris’ Hawks (Parabuteo
Behaviour 36:477–486. Reprinted by permission of the publish- unicinctus). Science 239. Reprinted with permission from
er, Academic Press London. Fig. 17.8 The Study of Instinct. AAAS. Fig. 19.3 Weimerskirch, H., J. Martin, Y. Clerquin, P.
Tinbergen, Oxford Clarendon, 1951. Willkinson, G. S. 1990. Alexandre, and S. Jarskova. 2001. Energy saving in flight for-
Food sharing in vampire bats. Scientific American 262:76–82. mation. Nature 413:697–698. Reprinted by permission from
Fig. 17.11 Orians, G. H., and G. M. Christman. 1968. Zoology Macmillan Publishers Ltd: Nature. Fig. 19.9 From Wilkinson,
84. University of California Press. Fig. 17.13 Hunter, M. L., G. S. 1990. Food sharing in vampire bats. Scientific American
and J. R. Krebs. 1979. Journal of Animal Ecology 48. Blackwell 262 (2):76–82. Scientific American Inc. Fig. 19.11 Sherman, P.
Scientific Publications Ltd. Fig. 17.14b–d Peters, R. A., J. M. W. 1985. Alarm callsof Belding’s ground squirrels to aerial
Hemmi, and J. Zeil. 2007. Signaling against the wind: predators: nepotism or self-preservation? Behavioral Ecology and
Modifying motion-signal structure in response to increased Sociobiology 17. With kind permission from Springer
noise. Current Biology 17:1231–1234. Reprinted with permis- Science+Business Media. Springer-Verlag. Fig. 19.14 Adapted
sion from Elsevier. Fig. 17.15 Slabbekoorn, H., and A. den from Packer, C., L. Herbst, A. E. Pusey, J. D. Bygott, J. P.
Boer-Visser. Cities change the songs of birds. Current Biology Hanby, S. J. Cairns, and M. B. Mulder. 1988. Reproductive suc-
16:2326–2331. Reprinted with permission from Elsevier. Fig. cess in lions. In Reproductive Success: Studies of Individual
17.17 Cummings, M. E. 2007. Sensory trade-offs predict signal Variation in Contrasting Breeding Systems, edited by T. H.
divergence in surfperch. Evolution 61:530–545. Copyright Clutton-Brock. Chicago: University of Chicago Press, pp.
2007, Wiley-Blackwell Publishers. Fig. 17.23 Hauser, M. D. 363–383. Fig. 19.19 Woolfenden, G. E. 1975. Auk 92. The
1998. Functional referents and acoustic similarity: Field play- American Ornithologists Union. Fig. 19.23 Lacey, E. A., and P.
back experiments with rhesus monkeys. Animal Behaviour 55 W. Sherman. 1997. Cooperative breeding in naked mole-rats:
(6):1647–1658. Reprinted by permission of the publisher, implications for vertebrate and invertebrate sociality. In
American Press London. Cooperative Breeding in Mammals, edited by N. G. Solomon and
J. A. French. Cambridge: Cambridge University Press, pp.
Chapter 18 267–301. Fig. 19.26 Hölldobler-Forsyth, Turid. From
Fig. 18.4 Clutton-Brock, T. H., F. E. Guiness, and S. D. Albon, Hölldobler, B. and E. O. Wilson. 1990. The Ants. Cambridge:
1982. Red Deer: The Behavior and Ecology of the Two Sexes. The Belknap Press.
University of Chicago Press. Fig. 18.8 From Carpenter, F. L.,
 Index

Acoustic adaptation hypothesis, Assortative mating, 68 Carnivory, 256–259

392–393 Audience effect, 145–146 Carotenoids, 383
Action potential, 102–104 Auditory map, 112–113 Central pattern generator, 123–124
Activational effects of steroid hor- Autism, 54 Chain of reactions, 16–17
mones, 139–141 Autonomic responses, 388–389 Chase-away model, 329
Adaptation Autotomy, 291–292 Chimpanzees
definition, 60 Avian maternal hormones, 156 language, 397–401
testing hypotheses about, 70–76 Avoidance conditioning rehabilitation after social isolation,
Adoption studies, 33 classical conditioning, 80 180–182
African clawed frogs, 133–134 fruit flies, 35–36 tube task, 94
Aggregations, 423 Chipmunks, 194
Aggression Balloon flies, 13–15 Cichlids
anoles, 135–136 Barn owls, 111–113 dominance relationships, 47–48
game theory and, 406–409 Bats, 109–111 parental care, 340–341
hormones and, 136, 138, 145–146 Bateman, A. J., 306, 308–309 sexual size dimorphism, 310–311
song sparrows, 137–138 Batesian mimicry, 289–290 Circadian time, 187–188
Agriculture, 255 Bees Clamworms, 78–79
Alarm signals, 298–299, 372–373, 402 dances, 374–379 Clark’s nutcrackers, 88–91
Alex the parrot, 96, 99, 401 learning, 85 Classical conditioning
Allee effects, 241 Begging, 382–383 conditioned stimulus, 80
Allele, 30 Behavioral genetics, 27–56 definition, 18, 80
Allogrooming, 365, 372 Behaviorism, 19–20 examples of, 80–82
Alternative male reproductive strategies Biological clock, 185–202 extinction, 80
bluegill sunfish, 312 Bird song Pavlov, I., 80
natterjack toads, 312–313 acoustic adaptation hypothesis and, unconditioned stimulus, 80
plainfin midshipman fish, 127–129 392–393 Cleaner fish, 258
salmon, 51–52 deafening and, 175–176 Clever Hans, 95–96
tree lizards, 141 environment and, 392–393 Coalitions, 437
Altruism, 427–434 own-species bias, 178 Cockroaches, 108
Amfor gene, 39 role of learning, 174–178 Cognition, animal, 23
Anoles sensitive periods and, 176–177 Color
aggression, 135–136 song nuclei and, 172–174 change, 279–280
habitat selection, 241, 242 Bluefooted boobies, 384 functions, 284–285
hormones and aggression, 135–136 Bluegill sunfish polymorphism, 285–287
hormones and courtship, 135–136 alternative male reproductive strate- Coloration matching the visual back-
Antihormone, 137 gies, 312, 337 ground, 277–279
Antipredator behavior, 32, 275–302 certainty of paternity, 337–338 Communication, 355–380
group defenses, 298–302 Bombykol, 363–364 changing views of, 381–382
maintenance of, 302 Bottleneck, 65 channels, 356–366
Ants Bowerbirds, 324–325 definition, 356
agriculture in, 255–256 Breeding dispersal, 234 multimodal, 366–367
behavioral development, 170–171 Brood parasitism, 343–345 Comparative method
castes, 448 Budgerigars, 85–86 definition, 73–74
learning, 84–85 Burrowing owls, 5–9 example in learning studies, 88–91
social behavior, 448–449 Comparative psychology, 17–20
Aposematism, 287–289 Camouflage, 277–284 Compass orientation, 204–206
Artificial selection, 33–35, 59 Canaries, 48–49 Conditional strategy, 410
Associated reproductive pattern, 143 Candidate gene, 38–42 Confusion effect, 301
Associative learning, 80, 82 Cannibalism, 322 Conservation behavior, 24

523
524 Index

Constant reproductive pattern, Directional selection, 62, 63 evolutionarily stable strategies, 68

143–145 Dispersal, 234–240 gene flow, 63
Constraints, in modeling, 265 breeding, 234 genetic drift, 65
Cooperative breeding natal, 234–240 macroevolution, 60
acorn woodpeckers, 439 Displacement activities, 388 microevolution, 60
benefits to helper, 444 Disruptive coloration, 280–282 misconceptions about, 60
benefits to relatives, 439–442 Dissociated reproductive pattern, natural selection, 58–60
costs to helper, 442 143–144 pleiotropy, 65
delayed dispersal and, 443–444 DNA (Deoxyribonucleic acid), 28–31 Evolutionarily stable strategies (ESS)
Florida scrub jays, 146–147, 440, 443 Dogs, 147–148 definition, 68
meerkats, 441–442 Dominance digger wasps, 69–70
pied kingfishers, 442 costs and benefits of, 413–415 hawk-dove game, 408
Seychelles warblers, 443–444 hierarchies, 413 mixed ESS, 68– 69, 608
Cooperative hunting, 424 hormones and, 420–421 pure ESS, 68, 608
Copulatory plugs, 317–318 intrasexual selection and, 309–310, rock-paper-scissors, 70
Correlated traits, 65 312 side-blotched lizards, 70
Cortisol and corticosterone, 131, 181, relationships, 47–48 Evolutionary arms race, 67
420–421 Excitatory postsynaptic potential
Dunce mutation, 35–36
Countershading, 282–283 (EPSP), 106–107
Dwarf mongooses, 3–5
Counting, 95–97 Extinction, in classical conditioning, 80
Courtship Extra-pair fertilizations, 314, 345–346
Ecological traps, 244–245
fruit flies, 30–31 Eyespots, 294–296
Egl-4 gene, 39
stickleback fish, 16–17
Cowbirds Electric fish
False heads, 290
brood parasitism, 343–344 communication, 365–36
Familiarity, 430
song development, 178–179 orientation, 229–230
FAP (Fixed action pattern), 15–16
Crabs Electrolocation, 229–230
Female defense polygyny, 351
blue, 107 Elephant seals, 409
Fighting, 405–422
fiddler, 189, 305–306, 358, 359 Elephants
asymmetric fights, 410–413
hermit, 412 communication, 360–361
experience in, 411
stone, 320 infrasound, 360
game theory, 406–410
Crickets Emancipation of signals, 390
Endocrine gland, 129 resource holding potential, 410
communication, 369
Endocrine system, 129–130 Filial imprinting, 161–165
learning, 82
Entrainment, 187–188 Finches, 121–122, see also Zebra finches
Critical period, 159–161
Environmental enrichment, 157–158 Fireflies, 258, 309
Cross-fostering, 32–33
Crows Epigenetics, 52–54 Fish
carrion, 287 EPSP (Excitatory postsynaptic poten- bluegill sunfish, 312, 337–338
New Caledonian, 93–94 tial), 106–107 cichlid, 47–48
optimal foraging in, 269–270, Erg-1 gene, 48 cleaner, 258
272–273 Escape behavior electric, 365–36
Cryptic female choice, 329–330 cockroaches, 108 paddlefish, 155
Cuckoos, 343–344 moths, 109–111 paradise fish, 32
Currency, in modeling, 264 ESS, see Evolutionarily stable strategies plainfin midshipman, 127–129
Estradiol, 131, 138 salmon, 51–52, 148–150, 224–226
Daily rhythms, 189 Ethology, 13–15 stickleback, 16–17
Dance flies, 13–15 Eusociality swordtail, 73
Darwin, C., 12, 58–60, 306 castes, 448 Fitness
Darwin’s finches, 65, 75 colony organization, 448 definition, 60
Death feigning, 292–293 definition, 445 direct, 60
Deoxyribonucleic acid (DNA), 28–31 evolution of, 446–448 indirect, 60
Desert ants, 205 haplodiploidy and, 446–448 nest building and, 34
Detouring behavior, 94–95 Eusperm, 315–316 Fixed action pattern (FAP), 15–16
Developmental homeostasis, 180–182 Evolution Flehmen, 363
Diet selection model, 265–267 artificial selection, 59 Flies, see also Fruit flies
Dilution effect, 299–300 correlated traits, 65 balloon, 13–15
Direct fitness, 60 definition, 12, 58 sandflies, 328–329, 363
 Index 525

Florida scrub jays hawk-dove, 407–409 History of animal behavior, 11–25

cooperative breeding, 440, 443 Prisoner’s Dilemma, 432–434 Home range, 415
hormones and helping, 146–147 Gene activity, 28–31 Homozygosity, 30
Flying squirrels, 188 Gene flow, 63 Honest signals, 382–387
Foraging (for) gene, 37–42, 45, 50–51 Gene networks, 22–23, 44–45 Honeybees
Foraging, 253–274 Gene regulation, 30–31 candidate genes, 41
aggressive mimicry, 257 Genetic drift, 65 foraging, 38–39
avoiding predation while foraging, Genetic modules, 44–45 juvenile hormone, 43
271–272 Genetic monogamy, 345 learning, 82
carnivory, 256–259 Genetics of behavior, 27–56 microarray analysis, 41–42
definition, 253 Genomics, 22–23, 31 nurse, 39–42
desert ants, 205 Geomagnetic field, 214–215
polyandry, 353–354
detecting prey, 259–264 Global warming and selection, 66
QTLs, 41
fruit flies, 28, 36–39, 40–41 Goal-directed emulation, 85
time sense, 193, 194
harvester ants, 39–40 Good genes
vitellogenin gene, 43–44
herbivory, 254–256 fluctuating asymmetry, 324
honeybees, 38–39 worker, 39–42
models, 327–328
nematodes, 40 Hormone, 129–132
parasite resistance, 323–324
omnivory, 253 peptide, 130
Gould, S. J., 70–71
pursuit, 256–257 Gouramis, 80–82 steroid, 130–132
risk sensitivity, 273 Ground squirrels types of, 130–132
traps, 258–259 Belding’s, 427, 434–436 Hybridization in fruit flies, 37
Free-running period, 187 biological clock, 194 Hyenas
Frequency-dependent selection California, 72–73 natal dispersal, 236–237
definition, 67 natal dispersal, 239–240 spotted, 236–237
frequency-dependent predation, 67 Group living Hypothesis testing, 5–8
frequency-dependent reproduction, benefits, 424–426
67–68 costs, 426–427 Ideal free distribution
polymorphism and, 287 Gulls habitat selection and, 241
Frogs black-headed, 71–72 space use and, 415–416
calls, 133–134, 360, 368, 390–391 herring, 57–58 Imitation, 85
clawed, 133–134 kittiwake, 57–58, 72 Immediate early genes, 47, 48
dendrobatid, 287–288 Inbreeding, 31–33
habituation, 79 Habitat selection, 241–245 Inclusive fitness, 21, 60
túngara, 390–391 anoles, 241, 242 Indirect fitness, 60
ultrasound, 360 conservation biology and, 244–245 Infanticide
Fruit flies conspecific attraction and, 241–242 hormones and, 137
avoidance conditioning, 35–36 heterospecific attraction and, hypotheses to explain, 319–320
biological clocks, 195 242–243
langurs, 319
courtship, 30–31, 308–309 natal experience and, 243–244
mice, 137
CREB gene, 36 phases, 241
dunce, 35–36 Information sharing, 425
Habituation
emergence, 192 Infrasound, 360–361
definition, 78–80
foraging, 28, 36–44 Inhibitory postsynaptic potential
fruit flies, 42–43
habituation, 42–43 (IPSP), 106–107
rhesus monkeys, 402–403
hybridization, 37 Insight, 92–93
Handicap principle, 383–384
learning, 35–36, 42–43 Intellectual continuity, 18–19
Handling time, 265
olfactory learning, 35–36 Hawk-dove game, 407–410 Intention movements, 388
rover, 37–40, 42–43, 50–51 Helpers at the nest, see Cooperative Intruder pressure, 416
rutabaga, 36 breeding IPSP (Inhibitory postsynaptic poten-
Shaker, 104 Herbivory, 254–256 tial), 106–107
sitter, 37–40, 42–43, 50–51 Heritability, 31
Fruitless (fru) gene, 30–31 Heterozygosity, 30 Jacanas, 342–343
Hippocampus Jacobson’s organ, 363–364
Game theory comparative studies of, 91 Japanese quail, 166–167
aggression, 406–409 learning and, 119–120 Juvenile hormone (JH), 43
526 Index

Kin discrimination, 429–432 Major histocompatibility complex Monarch butterflies

Kin selection, 428–429 (MHC), 326 antipredator behavior, 275–276
Knocking in (a gene), 40 Manakins, 438–439 migration, 248–249, 250
Knocking out (a gene), 40–41 Manipulation, 382, 434 Mongooses, dwarf, 3–5
Knockout mice, 44–45, 137 MAP (Modal action pattern), 15–16 Monogamy
Marginal value theorem, 267–268 biparental care, 346–348
Landmark use, 206–207 Mark test, 97 definition, 346
Language Masquerade, 284 genetic, 345
ape language studies, 397–400 Master biological clock, 192, 194 social, 345
definition, 396–397 Mate attraction, 369 Monthly rhythms, 189–190
Langurs, 319 Mate guarding, 313–315 Morgan’s canon, 17–18
Latent learning, 84 Maternal attachment, 169–170
Müllerian mimicry, 289
Learning Maternal behavior
Mutation
Aplysia, 113–115 rats, 27–28, 52–53
definition, 30
classical conditioning, 80–82 taxonomic patterns in vertebrates,
role in learning, 35–36
340–341
comparative studies of, 86–89 role in natural selection, 61–62
Mating systems, 345–354
definition, 78
monogamy, 345, 346–349
epigenetics and, 53–54 Naked mole rats
polyandry, 345, 353–354
habituation, 78–80, 114–115 countershading, 282–283
polygynandry, 345
heritability of, 87 perineal muscles, 134–135
polygyny, 345, 349–353
instrumental conditioning, 82 social behavior, 446
Meadow voles
latent, 84 Natal dispersal
dominance and reproductive success,
olfactory, 35–36
312 conservation biology and, 240
operant conditioning, 82–84
hormones and social preference, costs and benefits, 235–238
reinforcement schedules, 84 132–133 definition, 234
reinforcer, 83 parental behavior, 33, 45–46 sex biases in, 238–240
sensitization, 114–115 Meerkats, 441–442 spiders, 235–236
shaping, 83–84 Memory, 113–115 spotted hyenas, 236–237
social, 85–87 Meta-analysis, 393 squirrels, 239–240
song, 48–50, 118–119, 174–180 Mice Natal philopatry
spatial, 88–91, 119–120 California, 346–347 costs and benefits, 234–235
traditions, 86–87 hormones and infanticidal behavior, definition, 234
Lek polygyny, 352–353 137
Natural selection
Lewontin, R., 70–71 house, 34–35
definition, 12, 58–60
Lexigrams, 399–400 intrauterine effects, 155–156
directional, 63
Lions, 437–438 knockout, 44–45, 137
fitness, 60
Lizards mound-building, 347–348
frequency-dependent, 67–68
anoles, 394 primer pheromones, 364
Jacky dragons, 394 humans and environmental change,
Microarray analysis, 23, 42–43, 45
65–67
leopard gecko, 157 Microevolution, 60
kin selection, 428–429
side-blotched, 70 Migration, 245–251
misconceptions about, 60
tree, 141 benefits, 248–251
nest building and, 34–35
Local enhancement, 85 conservation biology and, 251
Locust flight, 122–126 stabilizing, 53
costs, 246–248
Long-term depression (LTD), 115–116 definition, 245 Negative-assortative mating, 68
Long-term potentiation (LTP), hormones and, 148–150 Nest building, 34–35
115–116 monarch butterflies, 248–249, 250 Neuroethology, 23, 99–125
Lorenz, K., 13–14, 159–160, 161–162, salmon, 148–150 Neurogenesis, 117–120
165 types of, 246 Neuromodulators, 106–107
Lunar day rhythms, 189 Migratory restlessness, 191, 209–210 Neurosteroid, 132
Mobbing, 301–302 Numbers and counting, 95–97
Macroevolution, 60 Modal action pattern (MAP), 15–16 Nuptial gifts, 321–322
Magnetic compass, 214–218 Mole rats, see also Naked mole rats
Magnetite, 223 magnetic compass, 215–216 Observational conditioning, 85
Magnetoreception, 222–224 social behavior, 446 Olfactory learning, 35–36
 Index 527

Olfactory navigation Pigs, 335–336 Resource defense polygyny, 351–352

pigeons, 226–229 Piloerection, 389 Resource holding potential, 410
salmon, 224–226 Piloting, 204 Resting potential, 102
Operant conditioning, 18–20, 82–84 Pinyon jays, 88–91 Ribonucleic acid (RNA), 29–30
Operational sex ratio, 342 Plainfin midshipman fish, 127–129 Risk-sensitive foraging, 273
Opportunistic reproductive pattern, Plant defenses, 254–255 Ritualization of signals, 387–390
143–145 Play, 158–159 Ritualized flight, 390
Optimality modeling, Pleiotropy, 65 RNA (Ribonucleic acid), 29–30
currency, 264 Polarized light orientation, 212–214 RNA interference (RNAi), 43–44
diet selection, 265–267
Pollen QTL, 41 Round dance, 374–377
game theory, 406–410
Polyandry, 353–354 Rover fruit flies, 37–40, 42–43, 50–51
marginal value theorem, 267–268
Polygyny threshold, 349–350 Runaway selection, 328
overview, 75–76, 264–265
Polygyny Rutabaga mutation, 36
rules of thumb, 269–271
female defense, 351
Organizational effects of steroid
hormones, 139–141 lek, 352–353 Salmon
Owls resource defense, 351–352 alternative reproductive histories,
burrowing, 5–9 threshold, 349–350 51–52
hunting, 111–113 Polymorphism behavioral development, 152, 153
Oxytocin, 44–45, 53 color, 285–287 homing, 224–226
Potential reproductive rate, 342 hormones and migration, 148–150
Paddlefish, 155 Prairie voles, 33, 45–46 Scent-marking
Pair bonds, 371–372 Predation, 256–159 chemical communication, 362
Paradise fish, 32 Prisoner’s Dilemma, 432–434 common marmosets, 362
Parasperm, 315–316 Progesterone, 131, 137 dogs, 147–148
Parental care, 333–345 Prolactin, 147 hormones and, 147–148
biparental care and monogamy, Promoter (of gene), 46, 52–53 sex differences in, 147–148
346–348 Pronghorn, 325–326 Schreckstoff, 300
direct, 334 Proteomics, 31 SCN (Suprachiasmatic nucleus), 192,
indicator of male quality, 322–323 Pursuit of prey, 256–257 194
indirect, 334 Scrub jays, 88–91
mode of fertilization, 341–342 QTLs (Quantitative trait loci), 41 Sea turtles
overall patterns, 339–343 Quantitative trait loci (QTLs), 41 magnetic compass, 218
voles, 33, 45–46 magnetic map, 218–219, 220–222
Parental investment, 334 Seabirds, 260–261
Rare male effect in mating, 67
Parrot, Alex, 96, 99, 401 Search image
Rats
Paternity, 337–338
epigenetics and, 52–53 color polymorphism and, 285,
Pavlov, I., 80 286–287
learning, 86
Payoff matrix, 407–408 formation, 263–264
maternal behavior, 27–28, 52–53
Pelicans, 426 Seed caching in birds, 88–91
perineal muscles, 134
Period (per) gene, 195
sexual behavior, 134, 139–141 Selection, see Natural selection
Phenotype matching, 430–432
ultrasonic calls, 361 Selfish herd, 300
Phenotypic variation, 61
Ravens, 92–93 Self-recognition, 97
Pheromones
Receiver-bias mechanisms in communi- Self-shadow concealment, 282–283
aggregation, 373–374
cation, 390–392 Semilunar rhythms, 189–190
definition, 363
Reciprocal altruism, 432–434 Sensitive periods, 159–161
mouse, 364
primer, 363–364 Recognition alleles, 431–432 bird song and, 175–177
releaser, 363 Recombination, 62 examples of, 161–171
silk moth, 363–364 Reductionist approach, 23 Sensory bias, 329
Phylogenies and comparative analysis, Reinforcement schedules, 84 Sensory drive, 395–396
73–74, 288 Reinforcer, 83 Sex allocation, 338–339
Physiological psychology, 20 Relative plasticity hypothesis, 141 Sex-role reversals, 307, 342–343
Pied kingfishers, 442 Repellents Sexual conflict, 330–331
Pigeons defense mechanism, 296–297 Sexual dimorphism, 309–311
concept learning, 96–97 intrasexual selection, 317 Sexual imprinting, 166–169
homing, 206–207, 221 Reproductive suppression, 3–5 Sexual interference, 318–320
528 Index

Sexual selection, 306–331 Spatial learning, Testosterone

intersexual selection, 306, 320–330 corvids, 88–91 aggression and, 138, 419
intrasexual selection, 306, 309–320 mammals, 119–120 sexual behavior and, 136, 140–141,
Sexy son hypothesis, 350–351 Species recognition, 368–369 142
Seychelles warblers, 314, 443–444 Speckled wood butterflies, 412–413 stress and, 420–421
Shaker gene, 104 Sperm competition, 315–317, 318 structure of, 131
Shaping, 83 Sperm heteromorphism, 315–316 Tinbergen, N., 13–14
Sharks, 261, 357 Tobacco hornworms, 152–154
Spiders
Siblicide, 335, 336 Tool use, 92–94
detour behavior, 94–95
Sibling rivalry, 335–336 Traditions, 86–87
fighting experience, 411
Sign language, 397–398 Transcription, of gene, 29–30
funnel web, 63–65
Signals Transformer (tra) gene, 31
gene flow, 63–65
alarm, 298–299 Translation of gene, 29–30
jumping, 94–95, 257
auditory, 358–361 Translocation, 244
multimodal communication, 367 Transparency, 283
body size and, 392
natal dispersal, 235–236 Trial-and-error learning, 18–20
chemical, 361–364
webs, 258–259 Trivers, R. L., 307
cost, 384
wolf, 367 Tube task, 94
definition, 356
electrical, 365 Squirrels Turkeys, 436–437
environment and, 392–395 color polymorphism, 285–286 Twin studies, 33
honest and dishonest, 382–387 fox, 285–286
human effects on, 395 natal dispersal, 239–240 Ultrasound, 360, 361
receiver-bias mechanisms, 390–391 Stabilizing selection, 63
ritualization, 387–390 Star compass, 211–212 Vampire bats, 433–434
seismic, 361 Starlings, 209–210 Vasopressin, 45–47
sensory drive, 395–396 Star-nosed moles, 260 Vasotocin, 132, 142
sensory exploitation, 390–392 Stealth, in pursuit of prey, 257 Vector navigation, 205
tactile, 364–365 Stickleback fish, 16–17 Vervet monkeys, 385–386, 402
unintended receivers, 396 Stimulus enhancement, 85 Vitellogenin gene, 43–44
visual, 356–358 Stimulus filter, 22 Voles
Silk moths, 363–364 Stomatopods, 386, 387 meadow, 33, 45–46, 132–133, 312
Sitter fruit flies, 37–40, 42–43, 50–51 Stotting, 297–298 parental behavior, 33, 45–46
Skinner, B. F., 19–20, 83 Stress, 181, 420–421 prairie, 33, 45–46
Snakes Stridulation, 359 Vomeronasal organ, 363–364
garter, 142, 288–289 Sturgeon, 233–234
hunting, 297 Waggle dance, 374–377
Sun compass, 207–210
rattlesnakes, 258–259, 297 Warning coloration, 287–289
Suprachiasmatic nucleus (SCN), 192,
Snow monkeys, 86–87 Wasps
194
Social behavior network, 120–122 cicada killer, 418–419
Surfperch, 395–396
Social bonding, 44–45 digger, 69–70
Survival of the fittest, 60
Social isolation, 170–171, 180–182 paper, 431
Swordtail fish, 73
Social learning, 84–86 Water mites, 391
Symbol use, 398–401
Social monogamy, 345 Whales
Synapse, 105–106
Social recognition, 44–45 bubble nets, 258–259
Synaptic transmission, 104–107 mate guarding, 314
Sociality, 3–5, 423–450
Sociobiology, 20–22
Song learning, 48–50, 118–119, Temperature compensation, 188–187 Zebra finches
174–180 Territories reproductive pattern, 144–145
Song sparrows, 49–50 boundaries, 418–419 sexual imprinting, 166
Sound localization, 108 definition, 415 song learning, 48–50, 118–119
Sparrows economics of, 416–419 song nuclei, 133, 172–174
song, 137–138, 176 hummingbirds, 417 Zenk, 48–50
white-crowned, 174–176 shape, 417–418 Zugunruhe, 191