Social Evolution and Inclusive Fitness Theory: An Introduction

SOCIAL EVOLUTION AND INCLUSIVE FITNESS THEORY

SOCIAL EVOLUTION AND INCLUSIVE FITNESS THEORY

An Introduction

James A. R. Marshall

Princeton University Press

Princeton and Oxford

Copyright © 2015 by Princeton University Press

Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540

In the United Kingdom: Princeton University Press, 6 Oxford Street, Woodstock, Oxfordshire OX20 1TW

press.princeton.edu

All Rights Reserved

Library of Congress Cataloging-in-Publication Data

Marshall, James A. R., 1976–

Social evolution and inclusive fitness theory : an introduction / James A.R. Marshall.

pages cm

Includes bibliographical references and index.

ISBN 978-0-691-16156-3 (alk. paper)

1. Sociobiology.    2. Social behavior in animals.    3. Behavior evolution. 4. Evolution (Biology)    5. Social evolution.    6. Hamilton, W. D. (William Donald), 1936-2000. I. Title.

GN365.9.M37 2015

304.5–dc23       2014031810

British Library Cataloging-in-Publication Data is available

This book has been composed in Garamond Premr Pro and Avenir LT Std

Printed on acid-free paper. ∞

Typeset by S R Nova Pvt Ltd, Bangalore, India

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

For Ruth, Hannah, and Adam

CONTENTS

List of Figures   xi

List of Tables   xii

Preface   xiii

Acknowlegments   xvii

1 SOCIAL BEHAVIOR AND EVOLUTIONARY THOUGHT   1

1.1 Explanations for Apparent Design   1

1.2 Natural Selection and Social Behavior   3

1.3 Arguments for Group Benefit   7

1.4 Enter Hamilton   11

1.5 Multilevel Selection Theory   13

1.6 The Generality of Inclusive Fitness Theory   14

2 MODELS OF SOCIAL BEHAVIOR   16

2.1 Introduction   16

2.2 The Donation Game   18

2.3 The Nonadditive Donation Game   22

2.4 Other Social Interactions   25

2.5 Public Goods Games   28

2.6 Threshold Public Goods Games   29

2.7 Interactions in Structured Populations   32

2.8 Summary   32

3 THE PRICE EQUATION   34

3.1 A General Description of Selection   34

3.2 Genetic Selection   36

3.3 Illustrative Applications of the Price Equation   39

3.4 Important Caveats   43

3.5 Summary   45

4 INCLUSIVE FITNESS AND HAMILTON’S RULE   46

4.1 Inclusive Fitness Extends Classical Darwinian Fitness   46

4.2 Fitness Effects as Regression on Genes   47

4.3 Deriving Hamilton’s Rule in the Simplest Case   51

4.4 Perceived Limitations of Inclusive Fitness Theory   54

4.5 Summary   58

5 NONADDITIVE INTERACTIONS AND HAMILTON’S RULE   59

5.1 Replicator Dynamics for Interactions between Relatives   59

5.2 Extending Hamilton’s Rule to Deal with Nonadditivity   65

5.3 The Price Equation and Levels of Causal Analysis   69

5.4 Summary   70

6 CONDITIONAL BEHAVIORS AND INCLUSIVE FITNESS   71

6.1 Implicit and Explicit Conditionality   71

6.2 Modeling Conditional Behavior   73

6.3 Claims That Assortment Is More Fundamental Than Relatedness   76

6.4 Summary   77

7 VARIANTS OF HAMILTON’S RULE AND EVOLUTIONARY EXPLANATIONS   78

7.1 Variants of Hamilton’s Rule   78

7.2 Geometric Relatedness Underlies Phenotypic Assortment   83

7.3 Explanations for Greenbeards   86

7.4 Different Viewpoints on Conditional Traits   88

7.5 Summary   89

8 HERITABILITY, MAXIMIZATION, AND EVOLUTIONARY EXPLANATIONS   90

8.1 What Drives Social Evolution?   90

8.2 Selection and Heritability   90

8.3 Do Individuals Act to Maximize Their Inclusive Fitness?   95

8.4 Ultimate Causes and Social Evolution   97

8.5 Summary   103

9 WHAT IS FITNESS?   105

9.1 Introduction   105

9.2 Haldane’s Dilemma   105

9.3 Reproductive Value and Class Structure   107

9.4 Fitness, Fecundity, and Payoffs   109

9.5 Summary   114

10 EVIDENCE, OTHER APPROACHES, AND FURTHER TOPICS   115

10.1 Introduction   115

10.2 Empirical Support for Inclusive Fitness Theory   115

10.3 Some Further Topics in Social Evolution Theory   127

10.4 Other Theoretical Approaches   129

10.5 Conclusion   132

Glossary   135

Notes   139

Bibliography   175

Index   187

LIST OF ILLUSTRATIONS

FIGURES

1.1 Male stag (Cervus elaphus)   4

1.2 Leafcutter ants (Atta colombica)   5

1.3 A long-tailed tit chick (Aegithalos caudatus) is fed an invertebrate by an adult helper   6

1.4 Dictyostelium discoideum forming a stalk and fruiting body   7

1.5 Fitness consequences of siderophore production in Pseudomonas aeruginosa   8

1.6 Pyocins inhibiting growth of Pseudomonas aeruginosa   9

1.7 Aposematism in caterpillars   10

2.1 Visualization of selection pressures in the additive donation game   21

2.2 Visualization of selection pressures in the nonadditive donation game   24

2.3 Visualization of selection pressures in the nonadditive cooperation game   26

2.4 Visualization of selection pressures in the negatively nonadditive group cooperation game   29

2.5 Visualization of selection pressures in the positively nonadditive public goods game   30

2.6 Group payoffs in the threshold public goods game   31

3.1 Lifetime breeding success and antler mass in red deer   38

3.2 Offspring–parent regression of wing length in Drosophila   41

4.1 Fitness effects as simple regression coefficients   50

5.1 Visualization of selection pressures in the positively nonadditive donation game between relatives   61

5.2 Visualization of selection pressures in the negatively nonadditive donation game between relatives   62

5.3 Visualization of selection pressures in the negatively nonadditive donation game with roles between relatives   63

5.4 Visualization of selection pressures in the positively nonadditive donation game with roles between relatives   64

5.5 Regression of fitness change on individual genetic value demonstrating correlated residuals   66

7.1 Four variants of Hamilton’s rule   79

7.2 The geometric view of relatedness   82

9.1 Reproductive value of Australian women   108

9.2 Threshold relatedness with varying scales of competition   112

10.1 A meerkat sentinel (Suricata suricatta) in raised guard   118

10.2 Rates of provisioning by long-tailed tit helpers   119

10.3 A comparative analysis of worker policing theory   122

10.4 Ancestral monogamy predicts the incidence of hymenopteran eusociality   124

10.5 A social-evolution approach to organismality   130

TABLES

2.1 The additive two-player donation game   18

2.2 The nonadditive two-player donation game   23

2.3 Classification of social behaviors   25

2.4 Different evolutionary games and outcomes   27

7.1 Variants of Hamilton’s rule   83

7.2 Possible evolutionary explanations of greenbeards   87

PREFACE

Inclusive fitness theory is probably the most important advance in our understanding of evolution since 1859. Darwin and Wallace’s¹ initial insight, that natural selection could explain adaptation and variation, rested on the assumption that natural selection acts on direct individual reproduction. This assumption remained intact during the mathematical formalization of natural selection theory during the modern synthesis of the 1930s, particularly in the seminal work of R. A. Fisher. Yet if natural selection acts on direct individual reproduction, a fundamental problem arises for the theory: how to explain apparent incidents of self-sacrifice in reproduction by individuals, for the benefit of the reproduction of others? For a while, some researchers assumed the problem could be solved by considering groups as the unit of selection, in which case individuals might sacrifice their own reproduction to benefit that of the group as a whole. However, this naïve group selection was shown to be logically inconsistent. The true resolution of the apparent paradox came in 1963 and 1964 with the publication by William D. Hamilton of the theory of inclusive fitness, according to which individuals should value not only their own reproduction, but also that of genetically related individuals. Hamilton thus extended classical Darwin–Wallace–Fisher fitness to a more inclusive version, able to explain the evolution of self-sacrifice.

The goal of this book is to celebrate the 50th anniversary of the publication of Hamilton’s inclusive fitness theory by providing a relatively brief introduction to, and explanation of, its generality and its correctness. This is necessary, because both the generality and correctness of inclusive fitness theory have been repeatedly criticized since its inception, and the pace and profile of these criticisms has increased in recent years. The situation for a researcher making their first forays into social evolution theory is daunting. The literature is vast and, given the subtleties of the theory and its apparently controversial status, it can seem difficult to know where to start and what to believe. Large literatures usually require effective summarization in books, and several excellent books in social evolution already exist. Yet all these books achieve something slightly different to the aims of this one. Some are conceptually and technically deep, but may be correspondingly difficult to approach. Others provide a more accessible introduction to the relevant theory, but at the expense of some of the technical depth. Still others survey empirical data on social evolution, and interpret it through the lens of theory. This book aims to fall somewhere between these three extremes, providing a brief primer in the basic concepts of social evolution theory, including its major subtleties and its empirical application, while being simultaneously accessible and technically complete.

What will a reader find in this book, and what will they not? In this book I focus on trying to expose the logical core of inclusive fitness theory, in a way that is understandable to the less mathematically confident, while still providing enough of the technical detail for those who are interested to follow up on. By trying to get to the level of causal explanations of the evolution of social traits, I hope to be able to explain why controversies and misunderstandings have arisen. Explaining how divergent views arise is often a good route towards their speedy resolution. Dealing with causation is, of course, a difficult philosophical problem. The tool I have chosen primarily to base the explanations in this book on is quantitative genetics. Quantitative genetics has been used extensively in studying evolution in general, and many believe it provides a useful route to understanding social evolution. Quantitative genetics is an inherently statistical framework, able to deal with complicated phenotypes, with poorly understood underlying genetics, and still to draw meaningful conclusions. Hence there are no complicated population genetical models in this book. Similarly, popular approaches to developing maximization arguments using concepts of evolutionary stability, which require weak selection, do not feature here. Modeling approaches necessarily trade predictive power against generality, and the focus of this book is on showing the generality of inclusive fitness theory. Because of the minimal assumptions it makes, quantitative genetics has very broad applicability, and the conclusions it draws should hold in great generality. For identifying and explaining the general principles of social evolution theory, it thus seems ideal.

This book was written during the 50th anniversary year of Hamilton’s first outline of inclusive fitness theory, and I write these words on the eve of the 50th anniversary year of the first full mathematical presentation of that theory. In the last five decades inclusive fitness has grown from initial obscurity to being generally recognized as the fundamental quantity that natural selection acts on. We should look forward to the even deeper understanding of biological systems that 50 more years of the application of inclusive fitness theory will certainly provide us with.

James A. R. Marshall

Sheffield, December 2013

¹On the centenary of his death it is particularly fitting to note Wallace’s often neglected role in conceiving evolution through natural selection.

ACKNOWLEDGMENTS

Many people have helped to make this book possible, to improve it substantially, or both. First I must thank Tim Clutton-Brock for advice and for putting me in touch with my publisher, Princeton University Press. At the Press I thank the staff, but particularly Alison Kalett for taking a chance on a new and untried author.

In writing this book, and over the years before, I have had useful discussions with a number of people. These include Marco Archetti, Jonathan Birch, Chris Cannings, Tim Clutton-Brock (with particular thanks for a stimulating visit to the Kalihari), Jeff Fletcher, Kevin Foster, Andy Gardner, Bill Hughes, Loeske Kruuk, John McNamara, Samir Okasha, Dave Queller, Corina Tarnita, and Geoff Wild.

Several colleagues were kind enough to provide very useful comments on draft chapters, or parts thereof. These are Marco Archetti, Jonathan Birch, Andrew Bourke, Tim Clutton-Brock, A.W.F. Edwards, Kevin Foster, Steve Frank, Andy Gardner, Melanie Ghoul, Ashleigh Griffin, John McNamara, Samir Okasha, Wenying Shou, and Stu West.

Special thanks are due to Ben Hatchwell, Chris Quickfall, Andy Gardner, and Stu West, who all read the entire manuscript draft and always provided insightful suggestions. Chris Quickfall, in particular, checked and rederived all of my mathematics, and Andy Gardner caught what were hopefully the final few mistakes. Any remaining errors are my own; I have made every effort to avoid these, but should any reader detect what they think is an error, I ask them to contact me.

I should also like to thank the photographers who kindly allowed me to use their images. Photographs are credited where they appear, but here I thank again Artour Anker, Kevin Foster, Melanie Ghoul, Loeske Kruuk, and Chris Tranter. I also thank Tim Clutton-Brock and Amanda Ridley for providing me with the opportunity to take some photographs for the book myself. Finally, Loeske Kruuk was kind enough to provide the source data enabling figure 3.1 to be drawn.

SOCIAL EVOLUTION AND INCLUSIVE FITNESS THEORY

CHAPTER ONE

Social Behavior and Evolutionary Thought

1.1 EXPLANATIONS FOR APPARENT DESIGN

Animals, plants, and other organisms appear to be designed for some purpose. While the ultimate purpose may not always be clear to us, observers of the natural world can readily understand sophisticated devices such as the wing and the eye to be designed for flight and sight, respectively. Until the mid-nineteenth century, natural philosophy explained design in nature as being due to, and evidence for, the existence of a supernatural creator. One of the most famous late examples of this tradition is William Paley’s argument from design [Paley, 1802]; on discovering a pocket watch lying on a heath, the conclusion of any reasonable person is that, due to its apparent complexity and its evident purpose, it must have been designed, and therefore a designer (the watchmaker) must exist. Paley went on to argue that, should the discovered watch have an internal mechanism capable of producing copies of itself, the rational discoverer would still conclude that it had been designed for this purpose, in addition to its purpose of telling the time, and must still therefore have a designer. Similarly, the apparent complexity in construction of animals and plants, and fitness for a purpose which includes reproduction, means they must have been designed, and therefore a designer (God) must exist. Under such a view, of course, an anthropocentric natural theologist might conclude that the animals and plants around us have been designed, by the supernatural creator, with the primary purpose of giving us food to eat, natural resources with which to make things, and so on.

With the work of Charles Darwin and of Alfred Russel Wallace [Darwin and Wallace, 1858, Darwin, 1859], an alternative explanation for the appearance of design arrived and, simultaneously, the question of the ultimate purpose of organisms was answered. The ultimate purpose of organisms was to compete for individual reproduction, and the result of such competition was that natural selection would progressively improve their suitability for this purpose, thereby giving them the appearance of design. If flight would increase the chances of individual reproduction for members of a species, for example, then natural selection acting on heritable variation over many generations could fashion limbs into wings, and then progressively optimize them for the purposes of aerodynamically efficient flight. Design and purpose in nature were both explained, and the explanations did not suggest a supernatural designer.

Darwin and Wallace amassed significant empirical support for the theory of evolution through natural selection, in collections of animals from around the globe,¹ and Darwin also interacted with practitioners of artificial selection, such as pigeon breeders and farmers. Yet the new evolutionary theory was formulated without knowledge of how characteristics, which natural selection was supposed to act on, were inherited by offspring from their parents. In fact, only 8 years after Darwin and Wallace’s papers were read at the Linnean Society in London, Gregor Mendel discovered the particulate nature of inheritance in an abbey in Brno, through his experiments on pea morphology [Mendl, 1866]. Despite being contemporary with and crucially relevant to the theory of natural selection, Mendel’s results were ignored for over 30 years [Bateson, 1909]. Initially thought to be a replacement for Darwinian evolution, the field of genetics was ultimately reconciled with natural selection in a mathematical framework that came to be known as the modern synthetic theory of evolution, or modern synthesis for short [Huxley, 1942]. Primarily the work of three pioneers, Sewall Wright, J.B.S. Haldane, and R. A. Fisher (e.g., [Wright, 1932, Haldane, 1932, Fisher, 1930]), the modern synthesis gave a formal mathematical structure to Darwin and Wallace’s ideas that would enable them to be developed into a predictive theory as never before. Of particular importance, in The Genetical Theory of Natural Selection Fisher mathematically formalized individual reproductive success, which lies at the original heart of natural selection theory [Fisher, 1930]. Thus, with a few exceptions as discussed below, in explaining adaptation the modern synthesis firmly set the focus of natural selection at the level of the individual and their own direct reproduction.

1.2 NATURAL SELECTION AND SOCIAL BEHAVIOR

Although the examples described above of traits designed through natural selection are physical body parts, behaviors also have genetic components, and therefore can be shaped by natural selection. As William D. Hamilton put it very pithily, It is generally accepted that the behaviour characteristic of a species is just as much the product of evolution as the morphology [Hamilton, 1963]. Behaviors that improve the reproductive success expected by an individual often have a negative impact on reproduction of members of the same species; one obvious example is behaviors involved in competition over mates, such as in display and fighting by red deer stags (figure 1.1); by monopolizing access to females, a male improves his own reproductive success at the expense of other males. Natural selection theory as developed by Darwin, Fisher, and others has no problem explaining the evolution of such behaviors; indeed it predicts them. This theory acts according to the reproductive success of individuals, and when the side effects of any trait are to modify the reproductive success of unrelated individuals, these are irrelevant.

Other individual behaviors seem to impact on the reproduction of others in a much more deliberate manner, however. Examples of such social behaviors abound in the natural world. Quite possibly the most well-known examples are among the social insects, considered by Darwin himself [Darwin, 1859]. In these insect species, reproductive division of labor is observed, with one or more castes helping to raise offspring other than their own; this is referred to as eusociality [Crespi and Yanega, 1995]. The simplest pattern is that the daughters of a single reproductive female, the queen, forage for, defend, and raise her offspring. These worker daughters either have suppressed levels of reproduction, as in the honeybee Apis mellifera where workers may both reduce their own levels of reproduction and destroy eggs laid by other workers [Ratnieks and Visscher, 1989], or are completely functionally sterile, as in several species of leafcutter ant for example (figure 1.2). Cooperative breeding is also observed in vertebrates, including many species of birds (e.g., figure 1.3) and mammals, such as meerkats (Suricata suricatta; e.g., [Clutton-Brock et al., 1998]) and naked mole rats (Heterocephalus glaber; [Jarvis, 1981]).² Cooperative breeders exhibit similar behaviors to eusocial species, in that helpers forage for, and guard, the offspring of a single breeding pair, although helpers do not form a distinct caste and may subsequently become reproductives themselves [Crespi and Yanega, 1995]. The presence of helpers has been shown to improve reproductive success by the breeding pair (e.g., [Hatchwell et al., 2004]), yet the helpers necessarily forego their own reproduction while caring for offspring that are not their own (e.g., [Emlen, 1982]).

Figure 1.1: A red deer stag (Cervus elaphus). Stags possess large antlers which impact on performance in fights, dominance rank, and hence access to fertile females. Maintenance of a harem of females, and hence increased reproductive success, negatively impacts on the reproductive success of other males in the population. However, natural selection theory (as developed by Darwin, Fisher, and others) acting on individuals explains the evolution of antlers since the successful male’s net personal reproduction is increased as a result of having them. Photograph by Loeske Kruuk, reproduced from [Kruuk et al., 2014] with the permission of the photographer.

Less frequently appreciated, social behavior is also observed in microbes including amoebae and bacteria [West et al., 2007a]. In social amoebae (Dictyostelium sp.; figure 1.4) normally free-living individuals aggregate at times of ecological stress, with some amoebae sacrificing themselves to form a structure that raises other individuals up in order to facilitate their dispersal to new, potentially richer, locations (figure 1.4) [Raper, 1984]. In the bacterium Pseudomonas aeruginosa, as in many other microorganisms, individuals secrete siderophores, which scavenge iron from insoluble forms in the environment. Siderophore production is individually costly in metabolic terms, resulting in a reduced growth rate, but this is offset by the increase in growth rate that siderophores facilitate when iron is scarce [Griffin et al., 2004, Jiricny et al., 2010] (figure 1.5A). However, siderophores secreted by individual bacteria can also facilitate iron uptake by neighboring individuals, allowing them to benefit from an increased growth rate, even if those neighbors did not contribute to siderophore production themselves [Griffin et al., 2004, Jiricny et al., 2010] (figure 1.5B).

Figure 1.2: Leafcutter ants of the genus Atta have morphologically distinct worker castes [Wilson, 1980], such as this forager (carrying leaf) and minim (sitting on leaf). In eusocial insect colonies, a worker caste or castes are either partially or totally functionally sterile, reducing or foregoing individual reproduction in order to support the reproduction of their mother. Atta colombica workers, although still possessing functioning ovaries, are effectively sterile [Dijkstra et al., 2005, Dijkstra and Boomsma, 2006]. Photograph by Chris Tranter, reproduced with permission of the photographer.

Less munificent examples of social behavior have also been described. Let us take one important example: bacterial production of bacteriocins. Production of colicins by the bacterium Escherichia coli, for example, is fatal for producing cells, as well as killing neighboring cells within a narrow phylogenetic range [Riley and Wertz, 2002]. Thus, bacteriocin production is personally costly (colicin producers pay the ultimate price of death, thereby ceasing personal reproduction), as well as costly to the targets of the behavior. Similarly, Pseudomonas bacteria produce individually costly pyocins that inhibit growth of strains that do not possess corresponding immunity genes [Michel-Briand and Baysse, 2002], as illustrated in figure 1.6.

Figure 1.3: A long-tailed tit chick (Aegithalos caudatus) is fed an invertebrate by an adult helper. Long-tailed tits are facultatively cooperative breeders; individuals that have failed to breed successfully themselves may help to raise the offspring of another breeding pair by feeding their chicks until independence [Hatchwell and Sharp, 2006, Hatchwell et al., 2014]. Photograph by Ben Hatchwell, reproduced with permission of the photographer.

Some purely physical traits can also have positive or negative social effects on conspecifics. One example is aposematism, for example in caterpillars (figure