Lowe, Joseph John - Walker, Mike - Reconstructing Quaternary Environments-Routledge (2015)

Reconstructing
Quaternary
Environments
This third edition of Reconstructing Quaternary Environments has been completely revised and updated to provide
a new account of the history and scale of environmental changes during the Quaternary. The evidence is extremely
diverse ranging from landforms and sediments to fossil assemblages and geochemistry, and includes new data
from terrestrial, marine and ice-core records. Dating methods are described and evaluated, while the principles
and practices of Quaternary stratigraphy are also discussed. The volume concludes with a new chapter which
considers some of the key questions about the nature, causes and consequences of global climatic and
environmental change over a range of temporal scales. This synthesis builds on the methods and approaches
described earlier in the book to show how a number of exciting ideas that have emerged over the last two decades
are providing new insights into the operation of the global earth–ocean–atmosphere system, and are now central
to many areas of contemporary Quaternary research.
This comprehensive and dynamic textbook is richly illustrated throughout with full-colour figures and
photographs. The book will be of interest to undergraduates, postgraduates and professionals in Earth Science,
Environmental Science, Physical Geography, Geology, Botany, Zoology, Ecology, Archaeology and Anthropology.
John Lowe is former Gordon Manley Professor and now Emeritus Professor of Quaternary Science in Royal
Holloway, University of London, UK.
Mike Walker is Emeritus Professor of Quaternary Science, Trinity Saint David, University of Wales, Lampeter,
UK and Honorary Professor, Aberystwyth University, UK.
For all our colleagues in the INTIMATE project
(INTegration of Ice-core, MArine and TErrestrial records)
2
0
2
-2
0
2
I -2
a
2
-2
0
-2
0 200 400 600 800

A g e (ka)
Reconstructing
Quaternary
Environments
Third Edition
John Lowe and

Mike Walker
Routledge
Taylor & Francis Group
LONDON AND NEW YORK
First published 1984
by Longman Group Ltd
Second edition 1997
by Addison Wesley Longman Ltd
Third edition published 2015
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 1984, 1997, 2015 John Lowe and Mike Walker
The right of John Lowe and Mike Walker to be identified as authors
of this work has been asserted in accordance with sections 77 and 78 of
the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval
system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
Lowe, J. J. (Joseph John)
Reconstructing quaternary environments / John Lowe and
Mike Walker. — Third edition.
pages cm
Includes bibliographical references and index.
1. Geology, Stratigraphic—Quaternary. 2. Paleogeography—
Quaternary. 3. Geomorphology. I. Walker, Mike. II. Title.
QE696.L776 2014
551.7’9—dc23
2013044728
ISBN: 978-0-415-74075-3 (hbk)

ISBN: 978-0-131-27468-6 (pbk)
ISBN: 978-1-315-79749-6 (ebk)
Typeset in Minion and Univers

by Florence Production Ltd, Stoodleigh, Devon, UK
Contents
List of figures and tables xv

Preface to the third edition xxvii
Acknowledgements xxix
Cover image details xxx
1 The Quaternary record 1

1.1 Introduction 1
1.2 Interpreting the Quaternary record 2
1.3 The status of the Quaternary in the geological timescale 2
1.4 The duration of the Quaternary 3
1.5 The development of Quaternary studies 5
1.5.1 Historical developments 5
1.5.2 Recent developments 7
1.6 The framework of the Quaternary 9
1.7 The causes of climatic change 13
1.8 The scope of this book 16
Notes 17
2 Geomorphological evidence 19
2.1 Introduction 19
2.2 Methods 19
2.2.1 Field methods 19
2.2.1.1 Field mapping 19
2.2.1.2 Instrumental levelling 20
2.2.2 Remote sensing 22
2.2.2.1 Aerial photography 22
2.2.2.2 Satellite imagery 22
2.2.2.3 Radar 23
2.2.2.4 Sonar and seismic sensing 24
2.2.2.5 Digital elevation/terrain modelling 24
2.3 Glacial landforms 26
2.3.1 Extent of ice cover 27
2.3.2 Geomorphological evidence and the extent of ice sheets and glaciers during
the last cold stage 30
2.3.2.1 Northern Europe 30
2.3.2.2 Britain and Ireland 33
vi CONTENTS
2.3.2.3 North America 35

2.3.3 Direction of ice movement 39
2.3.3.1 Striations 40
2.3.3.2 Friction cracks 40
2.3.3.3 Ice-moulded (streamlined) bedrock 40
2.3.3.4 Streamlined glacial deposits 42
2.3.4 Reconstruction of former ice masses 43
2.3.4.1 Ice sheet modelling 43
2.3.4.2 Ice caps and glaciers 47
2.3.5 Palaeoclimatic inferences using former glacier elevations 50
2.3.5.1 Cirque floor altitude (CFA) and toe-to-headwall (THAR) methods 50
2.3.5.2 ELA/FLA method 51
2.4 Periglacial landforms 53
2.4.1 Palaeoclimatic inferences based on periglacial evidence 55
2.4.1.1 Rock glaciers 55
2.4.1.2 Pingos and palsas 56
2.4.1.3 Pronival (‘protalus’) ramparts 57
2.5 Sea-level change 58
2.5.1 Relative and ‘absolute’ sea-level changes 59
2.5.2 Eustatic changes in sea level 59
2.5.2.1 Pre-Quaternary eustatic changes 59
2.5.2.2 Quaternary eustatic changes 60
2.5.3 Tectonic influences 67
2.5.4 Glacio- and hydro-isostasy 68
2.5.5 Shoreline sequences in areas affected by glacio-isostasy 69
2.5.6 Palaeoenvironmental significance of sea-level changes 73
2.6 River terraces 73
2.6.1 Origins of river terraces 75
2.6.1.1 Eustatic changes in sea level 76
2.6.1.2 Climatic change 76
2.6.1.3 Glaciation 77
2.6.1.4 Tectonic changes 77
2.6.1.5 Human activity 77
2.6.2 River terraces and palaeoenvironmental reconstruction 78
2.6.3 The terraces of the River Thames 78
2.7 Quaternary landforms in low latitudes 82
2.7.1 Pluvial lakes 82
2.7.2 Dunefields 86
2.7.3 Fluvial landforms 89
2.7.4 Weathering crusts 90
2.8 Conclusions 91
Notes 91
3 Lithological evidence 93
3.1 Introduction 93
3.2 Field and laboratory methods 93
3.2.1 Sediment sections 93
3.2.2 Coring 94
3.2.3 Laboratory methods 96
CONTENTS vii
3.2.3.1 Particle size measurements 96

3.2.3.2 Particle shape 97
3.2.3.3 Surface textures of quartz particles 97
3.2.3.4 Organic carbon content 97
3.2.3.5 Metallic elements 98
3.2.3.6 Heavy minerals 98
3.2.3.7 Clay mineralogy 98
3.2.3.8 Mineral magnetic analysis 98
3.2.3.9 Stable isotope analysis 98
3.3. Glacial sediments 99
3.3.1 Introduction 99
3.3.2 The nature of glacial sediments 99
3.3.2.1 Unstratified and stratified sediments 99
3.3.2.2 Glacigenic facies 100
3.3.3 The classification of tills 102
3.3.3.1 Lodgement, melt-out and ‘flow’ tills 102
3.3.3.2 Deformation tills 102
3.3.3.3 Paraglacial deposits 105
3.3.4 The influence of the thermal regime of glacier ice 107
3.3.5 Analysis of glacigenic sequences 109
3.3.5.1 Particle size and shape analysis 109
3.3.5.2 Lithofacies interpretations 109
3.3.6 Ice-directional indicators 111
3.3.6.1 Erratics 111
3.3.6.2 Till fabrics 113
3.3.6.3 Properties of the till matrix 115
3.4 Periglacial sediments 115
3.4.2 Structures associated with permafrost 116
3.4.3 Palaeoclimatic significance of periglacial structures 118
3.5 Palaeosols 122
3.5.2 The nature of palaeosols 122
3.5.3 Analysis of palaeosols 124
3.5.4 Palaeosols and Quaternary environments 125
3.6 Wind-blown sediments 127
3.6.2 Loess stratigraphy 127
3.6.3 Mid-latitude sand belts (coversands) 130
3.6.4 Low-latitude ‘sand seas’ 131
3.6.5 Wind-blown sediments and palaeoenvironmental reconstructions 131
3.7 Lake-level records from low-latitude regions 132
3.7.2 Pluvial lake sediment sequences 133
3.7.3 Lake-level changes and Quaternary palaeoclimates 135
3.8 Cave sediments and carbonate deposits 140
3.8.2 Detrital sediment in caves 141
3.8.3 Speleothem 143
3.8.4 Speleothem growth and environmental reconstruction 143
3.8.4.1 Speleothem growth and climatic change 143
viii CONTENTS
3.8.4.2 Stable isotope ratios in cave speleothem 145

3.8.4.3 Trace elements in cave speleothem 148
3.8.4.4 Speleothem formation and sea-level variations 149
3.8.4.5 Speleothem formation and tectonic activity 149
3.8.4.6 Speleothem formation and rates of denudation 149
3.8.5 Other carbonate deposits 149
3.9 Lake, mire and bog sediments 151
3.9.2 The nature of lake and bog sediments 152
3.9.3 Palaeoenvironmental evidence from lake sediments 154
3.9.3.1 Lake sediments and landscape changes 155
3.9.3.2 Lake-level variations and climatic changes 159
3.9.3.3 Lake sediments and palaeotemperatures 160
3.9.4 Palaeoenvironmental evidence from mire and bog sediments 161
3.9.4.1 Palaeoprecipitation records from ombrotrophic peats 161
3.9.4.2 Stable isotope records from ombrotrophic peats 163
3.9.4.3 Human impact recorded in ombrotrophic peat 165
3.10 The deep-sea sediment record 165
3.10.1 The nature and origin of ocean sediments 165
3.10.2 Oxygen isotope ratios and the ocean sediment record 166
3.10.2.1 General principles 166
3.10.2.2 Glacial ice storage and the marine oxygen isotope record 167
3.10.2.3 Ice volumes, sea level and the marine oxygen isotope record 169
3.10.2.4 Sea-surface temperatures and the marine oxygen isotope record 170
3.10.3 Limitations of oxygen isotope analysis 170
3.10.3.1 Stratigraphic resolution 170
3.10.3.2 Sediment mixing 171
3.10.3.3 Isotopic equilibrium between test carbonate and ocean water 171
3.10.3.4 Carbonate dissolution and diagenesis 171
3.10.4 Carbon isotopes in marine sediments 171
3.11 Ice-core stratigraphy 172
3.11.1 A brief history of deep-ice coring 172
3.11.2 Ice masses as palaeoenvironmental archives 173
3.11.3 Analysis of ice cores 173
3.11.3.1 Annual ice increments 173
3.11.3.2 Dust content 175
3.11.3.3 Chemical content 175
3.11.3.4 Stable isotope records 175
3.11.3.5 Other trace substances 175
3.11.4 Palaeoenvironmental significance of ice cores 175
3.12 Conclusions 178
Notes 179
4 Biological evidence 181

4.1 Introduction 181
4.1.1 The nature of the Quaternary fossil record 181
4.1.2 The taphonomy of Quaternary fossil assemblages 182
4.1.3 The interpretation of Quaternary fossil assemblages 182
4.2 Pollen analysis 183
CONTENTS ix
4.2.2 The nature of pollen and spores 183

4.2.3 Field and laboratory work 184
4.2.4 Pollen diagrams 185
4.2.5 The interpretation of pollen diagrams 190
4.2.6 Applications of pollen stratigraphy 194
4.2.6.1 Local vegetation reconstructions 194
4.2.6.2 Regional vegetation reconstructions 194
4.2.6.3 Space–time reconstructions 195
4.2.6.4 Human impact on vegetation cover 195
4.2.6.5 Pollen data and climatic reconstructions 197
4.3 Diatom analysis 197
4.3.2 The nature and ecology of diatoms 198
4.3.3 Field and laboratory methods 200
4.3.4 The interpretation of Quaternary diatom records 202
4.3.5 Applications of diatom analysis 202
4.3.5.1 Diatoms as salinity indicators 202
4.3.5.2 Diatoms and pH 203
4.3.5.3 Diatoms and trophic status 205
4.3.5.4 Diatoms and the archaeological record 205
4.3.5.5 Other environmental applications 206
4.4 Plant macrofossil analysis 207
4.4.2 The nature of plant macrofossils 207
4.4.4 Data presentation 208
4.4.5 The interpretation of plant macrofossil data 209
4.4.6 Palaeoenvironmental applications of plant macrofossil studies 212
4.4.6.1 Palaeoclimatic reconstructions 212
4.4.6.2 Forest history 214
4.4.6.3 Charcoal and fire history 214
4.4.6.4 Archaeological records 215
4.5 Fossil insect remains 215
4.5.2 Coleoptera 215
4.5.3 Laboratory methods 216
4.5.4 Coleopteran analysis and Quaternary environments 218
4.5.4.1 Habitat preferences 219
4.5.4.2 Palaeoclimatic inferences based on coleopteran assemblages 221
4.5.4.3 Insect fossils and archaeology 225
4.5.5 Chironomidae 225
4.6 Non-marine Mollusca 228
4.6.2 The nature and distribution of molluscs 229
4.6.4 Taphonomy of non-marine molluscan assemblages 231
4.6.5 Interpretation of non-marine molluscan assemblages: habitat groups
and indices of species diversity 232
4.6.6 Applications of Quaternary non-marine molluscan records 233
4.6.6.1 Biostratigraphic correlation 233
4.6.6.2 Palaeoclimatic reconstructions 234
x CONTENTS
4.6.6.3 Archaeological relevance 234

4.7 Marine Mollusca 235
4.7.2 Analysis of marine molluscan assemblages 235
4.7.3 Marine Mollusca and palaeoclimatic inferences 236
4.7.4 Other applications of fossil marine molluscan records 237
4.8 Ostracod analysis 238
4.8.1 The nature and distribution of ostracods 238
4.8.2 Collection and identification 238
4.8.3 Ostracoda in Quaternary studies 239
4.9 Foraminiferal analysis 241
4.9.1 The nature and distribution of Foraminifera 241
4.9.2 Collection and identification 242
4.9.3 Foraminifera in Quaternary inshore and shelf sediments 242
4.9.3.1 Sea-level change 242
4.9.3.2 Shallow marine water mass and temperature variations 243
4.9.3.3 Other palaeoenvironmental applications 244
4.10 Micropalaeontology of deep-sea sediments 244
4.10.2 Radiolaria 244
4.10.3 Coccolithophores 245
4.10.4 Dinoflagellates (dinocysts) 246
4.10.5 Marine microfossils in ocean sediments 246
4.10.6 Laboratory separation of marine microfossils 248
4.10.7 Marine palaeoclimatology 248
4.10.8 Marine palaeoproductivity and palaeocirculation 253
4.11 Vertebrate remains 254
4.11.2 The structure of teeth and bones 254
4.11.3 Fossilization of bone material 256
4.11.4 Field and laboratory techniques 256
4.11.5 The taphonomy of fossil vertebrate assemblages 257
4.11.5.1 Cave and fissure deposits 257
4.11.5.2 Lacustrine sediments 258
4.11.5.3 Fluvial sediments 258
4.11.6 Quaternary vertebrate records 258
4.11.6.1 Vertebrate biostratigraphy 259
4.11.6.2 Vertebrate biogeography 259
4.11.6.3 Vertebrate fossils and Quaternary environments 260
4.11.6.4 Vertebrate fossils and faunal evolution 262
4.12 Other fossil groups 263
4.12.1 Chrysophytes 263
4.12.2 Cladocera 263
4.12.3 Coral polyps 263
4.12.4 Fungal remains 264
4.12.5 Testate amoebae 264
4.12.6 Biomarkers (ancient biomolecules) 265
4.13 Multi-proxy palaeoecological studies 265
4.14 Conclusions 266
Notes 266
CONTENTS xi
5 Dating methods 267

5.2 Precision and accuracy in Quaternary dating 267
5.3 Radiometric dating techniques 268
5.3.1 The nucleus and radioactivity 268
5.3.2 Radiocarbon dating 270
5.3.2.2 Measurement of 14C activity 271
5.3.2.3 Quality assurance in radiocarbon dating 275
5.3.2.4 Sources of error in radiocarbon dating 275
5.3.2.5 Radiocarbon dating of soils 279
5.3.2.6 Calibration of the radiocarbon timescale 279
5.3.3 Argon-isotope dating 284
5.3.3.1 Potassium–argon dating 284
5.3.3.2 Argon–argon (Ar/Ar) dating 285
5.3.3.3 Problems and limitations of argon-isotope dating 285
5.3.3.4 Some applications of argon-isotope dating 285
5.3.4 Uranium-series (U-series) dating 286
5.3.4.2 Measurement, problems and age range 287
5.3.4.3 Some applications of U-series dating 288
5.3.5 Fission track dating 289
5.3.5.2 Measurement and problems 290
5.3.5.3 Some applications of fission track dating 290
5.3.6 Luminescence dating 291
5.3.6.3 Developments in luminescence dating 292
5.3.6.4 Age ranges and applications of luminescence dating 293
5.3.7 Electron spin resonance (ESR) dating 293
5.3.7.1 General principles and measurement 293
5.3.7.2 Sources of error in ESR dating 294
5.3.7.3 Some applications of ESR dating 294
5.3.8 Cosmogenic radionuclide (CRN) dating 294
5.3.8.3 Some applications of CRN dating 296
5.3.9 Short-lived radioactive isotopes 296
5.3.9.1 Lead-210 297
5.3.9.2 Caesium-137 297
5.3.9.3 Silicon-32 298
5.4 Incremental dating methods 298
5.4.1 Dendrochronology 298
5.4.1.3 Dendrochronological records 300
5.4.1.4 Dendroclimatology 302
5.4.2 Varve chronology 304
5.4.2.1 The nature of varved sediments 304
xii CONTENTS
5.4.2.2 Clastic varves 304

5.4.2.3 Organic (biogenic) varves) 305
5.4.2.4 Chemical varves 306
5.4.2.5 Complex varves 306
5.4.2.6 Sources of error in varve counting 306
5.4.2.7 Applications of varve chronologies 307
5.4.3 Annual layers in glacier ice 310
5.4.3.2 Errors in ice-core chronologies 312
5.4.3.3 Ice-core chronologies 312
5.4.4 Lichenometry 315
5.4.4.2 Sources of error in lichenometric dating 315
5.4.4.3 Some applications of lichenometry 316
5.4.5 Other materials dated by annual increments 316
5.4.5.1 Speleothems 316
5.4.5.2 Sclerochronology 317
5.5 Age-equivalent stratigraphic markers 319
5.5.1 Palaeomagnetism 319
5.5.1.1 Geomagnetic field and remanent magnetism 319
5.5.1.2 Magnetostratigraphy 320
5.5.2 Tephrochronology 325
5.5.2.2 Sources of error in tephrochronology 327
5.5.2.3 Applications of tephrochronology 327
5.5.3 Oxygen isotope chronology 330
5.5.4 Biostratigraphy and molecular clocks 331
5.6 Relative chronology based on processes of chemical alteration 332
5.6.1 Amino-acid geochronology 332
5.6.1.1 Chemistry of proteins 332
5.6.1.2 Amino-acid diagenesis 334
5.6.1.3 Aminostratigraphy and age control 334
5.6.1.4 Problems with amino-acid geochronology 334
5.6.1.5 Recent developments in amino-acid geochronology 336
5.6.1.6 Some applications of amino-acid geochronology 336
5.6.2 Fluorine, uranium and nitrogen content of fossil bones 339
5.6.3 Obsidian hydration dating (OHD) 340
5.6.3.2 Problems with obsidian hydration dating 340
5.6.3.3 Some applications of obsidian hydration dating 340
5.6.4 Weathering characteristics of rock surfaces 340
5.6.4.2 Problems in using surface weathering features as indicators of relative age 341
5.6.3.4 Some applications of surface weathering dating 342
5.6.5 Pedogenesis 342
5.6.5.2 Problems in using pedogenesis as a basis for dating 342
5.6.5.3 Some applications of relative dating based on degree of pedogenesis 343
5.7 Stratigraphic and temporal resolution 343
5.8 Conclusions 344
Notes 345
CONTENTS xiii
6 Approaches to Quaternary stratigraphy and

correlation 347
6.2 Stratigraphic subdivision 347
6.2.1 Principles of Quaternary stratigraphy 347
6.2.2 Stratotypes 349
6.2.3 Elements of Quaternary stratigraphy 349
6.2.3.1 Lithostratigraphy 349
6.2.3.2 Biostratigraphy 353
6.2.3.3 Morphostratigraphy 354
6.2.3.4 Soil stratigraphy 355
6.2.3.5 Oxygen isotope stratigraphy 355
6.2.3.6 Climatostratigraphy 358
6.2.3.7 Chronostratigraphy 361
6.3 Time-stratigraphic correlation 362
6.3.1 Principles of Quaternary correlation 362
6.3.2 Bases for time-stratigraphic correlation 363
6.3.2.1 Palaeomagnetic correlation 363
6.3.2.2 Correlation using tephra layers 364
6.3.2.3 Correlation using palaeosols 364
6.3.2.4 Shoreline correlation 364
6.3.2.5 Correlation on the basis of radiometric dating 365
6.3.2.6 Event stratigraphy and correlation 365
6.3.2.7 Correlation using the marine oxygen isotope record 366
6.3.3 Correlation between continental, marine and ice-core records 366
6.3.3.1 Long-term correlation on Milankovitch timescales 367
6.3.3.2 Correlation on sub-Milankovitch timescales 371
6.3.3.3 Synchronizing records of past environmental change 374
6.4 Conclusions 378
7 Global environmental change during the Quaternary 379

7.2 Environmental simulation models (ESMs) 380
7.2.2 Box models 380
7.2.3 General circulation models (GCMs) 381
7.2.4 Earth system models of intermediate complexity (EMICs) 383
7.2.5 Transient simulations 386
7.2.6 Palaeodata-model comparisons 387
7.2.7 Limitations of ESMs 388
7.2.8 The importance of ESMs in Quaternary research 388
7.3 Climatic change over Milankovitch timescales 389
7.3.2 The Mid-Pleistocene Transition (MPT) 390
7.3.3 The glacial–interglacial cycles of the last 800 ka 394
7.3.4 Overview 397
7.4 Environmental change over sub-orbital (millennial) timescales 399
7.4.2 Ice–ocean–climate interplay in the North Atlantic 400
7.4.3 A bipolar teleconnection 405
xiv CONTENTS
7.4.4 Global teleconnections: linking mechanisms 407

7.4.5 Overview 412
7.5 The Last Termination 413
7.5.2 Definition of the Last Termination 413
7.5.3 Onset of the Last Termination 415
7.5.4 Global teleconnections during the Last Termination 418
7.5.5 Synchronizing records of Lateglacial age 419
7.5.5.1 Introduction 419
7.5.5.2 Lateglacial stratigraphy and chronology 419
7.5.5.3 Lateglacial age models and correlation procedures 420
7.5.5.4 Rapid environmental change during the Lateglacial 421
7.6. Climate and the Holocene 427
7.6.2 Holocene climate trends 427
7.6.3 Holocene climatic events 428
7.6.3.1 The Pleistocene–Holocene transition 428
7.6.3.2 The 8.2 ka event 429
7.6.3.3 The 4.2 ka event 431
7.6.3.4 The 2.8 ka event 433
7.6.3.5 The Little Ice Age 433
7.6.4 Holocene climatic cycles 434
7.6.4.1 Late Holocene solar cycles 434
7.6.4.2 El Niño–Southern Oscillation (ENSO) 435
7.6.4.3 Late Holocene Atlantic and Pacific Oscillations 435
7.6.5 People and climate 436
7.6.5.1 The greenhouse effect 437
7.6.5.2 Early human impact? 438
7.6.5.3 Delayed glaciation? 439
7.6.6 The Anthropocene 439
7.7 Concluding remarks 440
Notes 443
References 445
Index 523
Figures and tables
FIGURES
1.1 The Quaternary relative to the geological timescale 1
1.2 The Cenozoic timescale, as defined in 2014 3
1.3 The Monte San Nicola section in southern Sicily 4
1.4 The maximum glaciation of the Northern Hemisphere during the Quaternary 6
1.5 Climatic trends during the past 3 Ma reflected in a stacked (composite) oxygen isotope record 10
1.6 The MIS record based on a composite of deep-ocean cores and the Quaternary stratigraphy of the
Northern Hemisphere 11
1.7 The δ18O record from the GRIP Greenland ice core showing the Lateglacial event stratigraphy and
the stratigraphic subdivision of the Lateglacial in northwest Europe and the British Isles 12
1.8 The components of the Astronomical Theory of climate change 14
1.9 a) Variations in eccentricity, obliquity and the precessional index over the past 800 ka. b) Normalized
and smoothed variations in the oxygen isotope signal (δ18O) in five deep-sea cores 14
1.10 Variations in temperature, dust flux, CO2 and CH4 from the EPICA Dome C ice-core record,
Antarctica, over the last 800 ka 15
2.1 Geomorphological map of the West Drumochter Hills in the Scottish Highlands 21
2.2 Sonar bathymetry of the north-central English Channel shelf 25
2.3 A digital surface model showing the ice-moulded landscape of part of the Midland Valley of Scotland 26
2.4 Moraine systems in a) Bhutan-Himalaya, b) northwest Scotland and c) Antarctica 28
2.5 a) Features used to identify trimlines and other types of glacial limits in mountainous terrain.
b) Nunataks bordering Liv Glacier. c) High-altitude trimline in the Grimsel Pass, Switzerland 29
2.6 a) Maximal limits of the Late Weichselian ice sheet. b) Recessional stages of the Late Weichselian
ice sheet 30
2.7 a) Cross-section through ice-marginal ridges of Late Weichselian age in Denmark. b) Up-thrusted
and folded Middle Weichselian lacustrine and fluvial sediments, northernmost Jutland 31
2.8 Limits of the Late Devensian British ice sheet 34
2.9 a) The Late Weichselian Laurentide, Cordilleran and Innuitian ice sheet at the LGM.
b) Reconstructed ice lobes in the Great Lakes region at the LGM. c) The pattern of retreat
moraines of the Michigan–Huron lobe after the LGM 36
2.10 Retreat pattern of the Laurentide ice sheet and estimated ages of major arterial flows or ice streams 37
2.11 Wisconsinan terminal moraines dating to the Tahoe series in Pine Creek Canyon, California 39
2.12 Local ice-flow direction based on mapping of lineations of striations in northern Canada 41
2.13 Mega-grooved bedrock surface showing parallel striations and gouge-marks, including crescentic
fractures 42
2.14 A drumlin in the Eden Valley, northwest England, that formed beneath the last British–Irish ice
sheet 42
xvi FIGURES AND TABLES
2.15 Lineation of subglacial landform suites (mostly drumlins) formed by the last Irish ice sheet 43
2.16 Examples of input and output parameters employed in a numerical model simulation of the last
Laurentide ice sheet 44
2.17 Inverse models of the last (Late Devensian) British ice sheet 45
2.18 Model simulations of a) basal ice temperature and b) horizontal basal velocity of the last British Ice
sheet during the LGM 46
2.19 Four numerical simulation models of the ice sheets over North America at the LGM 47
2.20 Reconstruction of the Loch Lomond (Younger Dryas) Stadial icefield in the Pass of Drumochter area,
Scottish Highlands 48
2.21 ASTER images of the Chapman Glacier and adjacent piedmont glaciers, northern Canada 49
2.22 a) The relationship between mean winter snow accumulation and mean temperature during the
ablation season for modern glaciers in Norway. b) Generalized contours for inferred ELAs based
on reconstructed glaciers in Scotland 51
2.23 Mean equilibrium-line altitudes for reconstructed glaciers in eleven districts of the British Isles 52
2.24 Contemporary permafrost distribution in the northern hemisphere 53
2.25 a) Granite tor on the summit of Beinn Mheadhoin, Cairngorm Mountains, Scotland. b) Summit
blockfield on Gros Morne Mountain, western Newfoundland, Canada 54
2.26 a) Active polygonal patterned ground, Mt Edziza, British Columbia, Canada. b) Sorted circles on
the summit of Gros Morne Mountain, western Newfoundland, Canada 55
2.27 Active lobate rock glaciers near Lytton, British Columbia, Canada 56
2.28 Fossil protalus rampart and stratified scree near Cader Idris, Wales 57
2.29 Altitudes of the frontal crests of Loch Lomond (Younger Dryas) Stadial protalus ramparts in the
Scottish Highlands, and reconstructed ELAs of Stadial glaciers 57
2.30 The ‘High Rock Platform’, northern Islay, Scottish Hebrides 58
2.31 Eustatic sea-level variations over the last 9.5 Ma based on marine oxygen isotope measurements 60
2.32 Variations in eustatic sea level over a) the past 1 Ma and b) the last 500 ka 61
2.33 Comparison of seven different eustatic sea-level records from the last glacial cycle 62
2.34 a) Rise in eustatic sea level since the LGM; b) Land elevation changes in Hudson Bay, Canada and
c) in northern Sweden during the Holocene 63
2.35 River bank section in lower Strathearn, eastern Scotland showing evidence for marine regression and
transgression 64
2.36 a) Chronology of isolations from, and connections to, the sea for four littoral basins in northwest
Scotland. b) Relative sea-level reconstructions for selected sites in the UK, based on dated index
points and sea-level tendencies 65
2.37 Relative changes in land or sea level in the UK during the late Holocene 66
2.38 Raised coral reef terraces along the coast of the Huon Peninsula, Papua New Guinea 67
2.39 Schematic diagram of the effect of an ice mass on a land surface 68
2.40 Crustal flexuring in the Pacific margin of Canada reflecting the changing mass of the last ice sheet in
British Columbia 68
2.41 The development of a raised shoreline sequence reflecting isostatic recovery and eustatic rise
following deglaciation 69
2.42 Height–distance diagram of measured altitudinal variations of the surfaces of shoreline fragments in
Fife, eastern Scotland 70
2.43 Isobase and trend surface data reflecting patterns of uplift in a) eastern Canada, b) Scandinavia and
c) Scotland 71
2.44 Sea-level changes in Atlantic Canada 72
2.45 Terraces cut into outwash gravels, Glen Roy, Scotland 74
2.46 a) Schematic representation of the staircase of river terraces preserved in the Middle Rhine area.
b) Areal distribution of terrace gravels in the Lower Rhine 75
2.47 Formation of ‘strath’ and ‘cut-and-fill’ fluvial terraces 75
2.48 Schematic representation of a sequence of river terraces in the valley of the Somme, France 76
FIGURES AND TABLES xvii
2.49 a) Average alluvial sedimentation rate for three catchments in southern Turkey over the past 6 ka.
b) The number of settlements occupied in the region over the same period 78
2.50 Schematic transverse profiles through the best-preserved suite of the terrace staircase of a) the Middle
Thames (c. 35 km west of London) and b) the Lower Thames (c. 15 km east of London) 79
2.51 Schematic representation of the Thames drainage system a) prior to, b) during and c) after the
Anglian glaciation 80
2.52 a) Abandoned shorelines of Pluvial Lake Lahontan near Reno, Nevada, USA. b) Beach deposits and
fan gravels from Pluvial Lake Lahontan 83
2.53 The maximum limits of late Pleistocene pluvial lakes in the American southwest 84
2.54 Location and surface dimensions of modern Lake Chad, Palaeolake Mega-Chad, and the Chad
drainage catchment 85
2.55 a) Hydro-isostatic deformation of the Bonneville Shoreline of Pluvial Lake Bonneville. b) Chronology
of water level changes in Lake Bonneville over the past 30 ka 86
2.56 Principal dune-building episodes during the last 50 ka in the deserts of Africa, Australia, Asia and
North America 87
2.57 Dominant linear dune trends and present-day dust circulation in Australia 88
2.58 Linear trends of crests of sand dunes in Egypt’s Western Desert 89
2.59 Remotely sensed images of the desert area of NW Sudan: a) LANDSAT ETM image of surface sands
with little indication of buried palaeo-drainage features beneath. b) Radarsat-1 image that reveals
the buried palaeo-drainage features below. c) Digital elevation model (DEM) that enhances the
buried river channel system 90
3.1 Monolith tins (50 cm length) in a sediment exposure in the Kopanica Valley, southwest Poland 94
3.2 Examples of coring equipment used to sample Quaternary sediments 95
3.3 X-radiograph images of laminated (varved) lake sediments that accumulated in a former ice-dammed
lake near Loch Lomond, Scotland 96
3.4 Schematic model of supra-, en- and subglacial ‘subenvironments’ within the marginal zone of a
continental ice mass 100
3.5 Schematic model of modes of deposition in proximal zones associated with continental and marine
ice margins 101
3.6 Examples of glacigenic deposits that could be classified as lodgement, melt-out and possible
‘flow’ tills 103
3.7 Highly contorted glacial diamict (till) exposed at West Runton, Norfolk, UK 104
3.8 Range of microstructures and fabrics observed in deformed tills from micromorphological analysis
of thin section samples 105
3.9 Examples of microfabrics in deformed till identified using thin-section analysis on samples obtained
from tills in Ontario, Canada 106
3.10 Scanning electron photomicrographs of vertical sections through the fine matrix of tills 107
3.11 View down the Braldu Valley in the central Karakoram, northern Pakistan 108
3.12 Ternary diagram using a statistical clustering method (eigenvalues) to characterize clast fabrics
obtained from diamictons 110
3.13 Particle size distributions (cumulative frequency curves) showing distinctive curves obtained from
diamictons and sorted sediments 110
3.14 Schematic facies logs of glacigenic sediments in Wales 111
3.15 Idealized lithofacies associations for glacigenic sediment sequences developed under different
thermal regimes 112
3.16 The giant quartzite erratic (‘The Big Rock’) near Okotoks, southwest Alberta, Canada 113
3.17 a) Distribution of some indicator erratics by ice in Britain and northwest Europe. b) The Lennoxtown
boulder train in the Forth Valley, central Scotland 113
3.18 Different methods for representing till fabric data 114
xviii FIGURES AND TABLES
3.19 a) Ice-wedge cast of Anglian age (MIS 12), West Runton, Norfolk, Eastern England. b) Anglian age
sand wedge, Broomfield, Essex, UK 116
3.20 Involutions/cryoturbation structures, formed in Late Weichselian fluvial sediments at the site of
Bosscherheide, Netherlands 117
3.21 Classification of cryoturbation structures according to their form 118
3.22 a) Mean annual temperature in northwest Europe during the 27–20 k14Cyr BP interval. b) Mean annual
temperature in northwest Europe during the 20–13 k14Cyr BP interval 120
3.23 a) Permafrost limits in northwest Europe during the Younger Dryas. b) Maximum mean annual
isotherms in northwest Europe for the coldest part of the Younger Dryas Stadial 121
3.24 Examples of palaeosols 123
3.25 Photomicrographs of pedogenic structures in soil horizons within Weichselian loess deposits in
southern Poland 125
3.26 a) The early Quaternary ‘Wucheng’ sequence of loess units and soil horizons exposed at the
Luochuan site on the Loess Plateau, China. b) The S1 (last interglacial) soil profile exposed at the
Lantian section, near Xi’an in the southern part of the Loess Plateau 128
3.27 The Loess Plateau, major deserts and mountain regions of north-central China 129
3.28 Loess–palaeosol succession at Baoji, in the southern Loess Plateau, showing thirty-two palaeosol
units spanning the last 2 Ma 129
3.29 Coversands and interbedded Usselo soil of Lateglacial (late Allerød) age exposed at a site in the
Netherlands 130
3.30 Magnetostratigraphy, lithology, magnetic susceptibility and grain-size variations of the Lingtai
section on the Chinese Loess Plateau 132
3.31 The water-level record for Lake Lisan over the course of the last 55 ka 133
3.32 The geochemical record from a lake sequence in Tibet that developed between 9 and 4 ka 134
3.33 a) Locations of lakes for which lake-level data are lodged with the PMIP Global Lake Status Data
Base. b) Lake status data for 18 ka. c) Lake status for 6 ka 136
3.34 Lake-status data for 6 ka expressed as anomalies relative to long-term status 137
3.35 Spatial patterns of inferred effective moisture for six time-slices between the Bølling-Allerød period
and the present time 139
3.36 a) Carbonate δ18O records from eight lake sediment sequences from the Indian Monsoon region.
b) A regional moisture index based on the mean of the eight curves shown in a) 140
3.37 Generalized cross-section of a cave with various types of sedimentary infill and associated biological
remains 141
3.38 a) Cross-sectioned surface of a stalagmite from Akçakale Cave, northeast Turkey revealing annual
growth layers spanning the last c. 500 yr. b) Cross-sectioned surface of a small stalagmite from
Rukiessa Cave, southeast Ethiopia, spanning the last century 144
3.39 Compilation of c. 750 TIMS U-series speleothem dates plotted against the latitude of the relevant
site 144
3.40 Composite δ18O curve constructed from twenty-one overlapping speleothem records for the past
185 ka from Soreq Cave, Israel 146
3.41 a) δ18O measurements covering the period 14 ka to the present from Hölloch Cave, Bavarian Alps,
Germany. b) Timberline record from the Eastern Alps, Austria 147
3.42 δ18O record from Hulu Cave, China and the δ8O trace from the GISP2 Greenland ice core 148
3.43 Growth phases of 230Th-dated portions of travertine from sites in semi-arid northeastern Brazil
extending over > 600 ka 150
3.44 Tufa formations around the shore of Mono Lake, California 150
3.45 Middle Pleistocene (Cromerian Complex) organic deposits exposed beneath till at Pakefield,
Suffolk, UK 152
3.46 Some sediment types deposited with increasing depth of water under oligotrophic and eutrophic
conditions 154
3.47 Schematic representation of gradual infill of a small lake 155
FIGURES AND TABLES xix
3.48 Lateglacial and early Holocene environmental changes inferred from lithostratigraphy of a typical
northwest European lake and mire sequence 156
3.49 Variations in the abundance of selected chemical elements in Lateglacial and early Holocene
sediments from two sites in the Isle of Skye, Inner Hebrides, Scotland 157
3.50 Variations in rates of lake sedimentation during the Holocene in sites from the British Isles 158
3.51 The lake sediment sequence at Sevrier-Les-Charretières, Lake Annecy, France 159
3.52 a) A peat profile at Bolton Fell Moss, northern England, showing a lower dark, well-humified peat
overlain by a lighter-coloured, less well-humified peat. b) A peat section from Store Vildmose,
Denmark, showing both lateral and vertical alternations in darker and lighter peat layers 162
3.53 Reconstructed mire-surface wetness changes at Walton, Moss, northern England 163
3.54 Proxy climate reconstructions from Temple Hill Moss, southeast Scotland 164
3.55 Heinrich layers in deep-ocean sediments from the northwestern Labrador Sea 166
3.56 Variations in surface water oxygen isotope ratios during times of glacial maxima and interglacial
high sea-level stands 168
3.57 Schematic representation of oxygen isotope variations for the past 600 ka 168
3.58 Comparison between estimates of sea-level variations for the last 140 ka based on marine isotopic
and coral reef records 169
3.59 Location of some of the principal ice-core drilling stations in Greenland and Antarctica 173
3.60 Annual ice layers exposed in the Quelcayya ice cap, Peru 174
3.61 Seasonal variations in chemistry, dust content and stable oxygen isotope ratios in ice layers in a
section of the NorthGRIP core 174
3.62 Continuous ␦18O profiles through five Greenland ice cores 176
4.1 Subsample of a pollen assemblage typical of last cold stage deposits from the site of St Front, France 184
4.2 Relative pollen diagram showing variations in abundance of principal taxa in a lake sediment
sequence at Ioannina, northwest Greece 186
4.3 Pollen concentration diagram for some of the taxa represented in relative proportions in Figure 4.2 187
4.4 Pollen accumulation rate data for a sequence spanning the last 11 ka at Steel Lake, central
Minnesota, USA 188
4.5 Arboreal pollen taxa and palaeotemperature inferences recorded for last interglacial deposits
at Jammertal, southwest Germany 190
4.6 The changing vegetation cover of the eastern USA from 18 ka to present 195
4.7 Pollen percentage and influx diagram for Abies and Corylus, and charcoal influx record, from a
site in south Switzerland spanning the period 5.1–3.1 cal. BC 196
4.8 Holocene pollen diagram from the site of An Loch Mor, the Aran Islands, western Ireland 197
4.9 Schematic representation of the procedures employed to derive quantitative lake-water pH
reconstructions from fossil diatom records 198
4.10 SEM images of common northwest European diatom frustules recovered from streams and soils
in southeast England 199
4.11 Late Holocene diatom stratigraphy from Beaver Lake in the Nebraska Sand Hills, USA 201
4.12 A marine incursion (positive sea-level tendency) as reflected in a diatom assemblage 202
4.13 Isolation basin studies in southeast Greenland 204
4.14 Diatom-inferred palaeosalinity through a sediment sequence in Oro Lake, Saskatchewan, Canada 205
4.15 Diatom-inferred pH history of Lysevatten Lake, southwestern Sweden, suggesting increased
acidification since the 1960s 206
4.16 Scanning electron photomicrographs of terrestrial macrofossils from a 12 ka peat layer in
North Dakota, USA 207
4.17 Arboreal plant macrofossils in a Lateglacial sediment sequence in Latvia, plotted against the pollen
record of the same taxa 209
4.18 Influx of charcoal fragments and macrofossils of selected plant taxa in a lake sediment sequence in
the North Cascade Range, Washington State, USA 210
4.19 Recruitment pathways and processes by which plant macrofossil remains are delivered to lakes 211
xx FIGURES AND TABLES
4.20 Episodes of climatic deterioration in northern Britain during the past 7.5 ka based on plant
macrofossil evidence for increased bog-surface wetness 213
4.21 Subfossil coleopteran sclerites recovered by flotation from peat deposits overlying a Bronze Age
occupation site near Ballyarnet Lake, Co. Derry, Northern Ireland 216
4.22 Generalized drawings of coleopteran sclerites frequently preserved as Quaternary fossils, showing
a range of diagnostic features used in fossil identification 217
4.23 A coleopteran record from the Lateglacial site of Llanilid, south Wales, UK 218
4.24 Number of obligate beetle species recorded in Holocene sediment sequences from southern and
central England 220
4.25 Present-day European distributions of four coleopteran species found in Lateglacial deposits at the
site of Glanllynnau, north Wales 221
4.26 Schematic representation of the mutual climatic range method of quantitative temperature
reconstructions 223
4.27 Test of the MCR method on assemblages of species found living today at thirty-five localities in
North America 223
4.28 Generalized climatic curves of estimated TMAX variations during the Lateglacial–early Holocene for
different parts of Europe 224
4.29 Head capsules of common chironomid taxa recovered from late Quaternary lake sediment sequences
from northwest Europe 226
4.30 Chironomid-inferred temperature variations for the Lateglacial sequence near the Majola Pass,
Swiss Alps, compared with the NGRIP isotope record 227
4.31 a) Fossil shells of freshwater molluscs (principally gastropods) exposed on an abandoned beach of
Pluvial Lake Lahontan, Pyramid Lake, Nevada, USA. b) Valvata piscinalis, a small gastropod that
inhabits streams, rivers and lakes 228
4.32 Variations in relative abundance of molluscan species from the Lateglacial to mid-Holocene sequence
at Holywell Coombe, southern England 230
4.33 Variations in absolute abundance of terrestrial molluscan species through a Late Quaternary loess
and palaeosol succession at Weinan, China 231
4.34 Variations in species richness, abundance, diversity and habitat types in Holocene tufa sequences
at Courteenhall, near Northampton, UK 232
4.35 Examples of some common marine bivalves of the North Atlantic and their water depth
preferences 235
4.36 Distribution of marine zoogeographical provinces in the Northeast Atlantic 236
4.37 Variations in the dynamics of the winter North Atlantic Oscillation and of annual growth
increments of the bivalve Arctica islandica 237
4.38 Climatically significant ostracods from the Pleistocene of the British Isles 239
4.39 Examples of planktonic foraminiferal species widely employed in Quaternary palaeocean studies 241
4.40 Relative sea-level curves for the last 12 ka for the Western Mediterranean based on benthic
Foraminifera-based transfer functions 243
4.41 a) Examples of siliceous skeletons (tests) of the radiolarian groups Spumellaria and Nassellaria.
b) SEM scans of coccolithophore specimens from the central Adriatic Sea 245
4.42 a) Examples of common Quaternary dinocysts. b) Structural features of some dinoflagellate cysts
(dinocysts) at motile stage 247
4.43 Planktonic foraminiferal provinces in the modern ocean showing the close relationship between
sea-surface temperature gradients and species abundances 249
4.44 Changes in coiling direction of tests of the high-latitude species Neogloboquadrina 250
4.45 Changes in ecological water masses in the Norwegian Sea and the northern North Atlantic over
the past 225 ka 251
4.46 Reconstructions of surface conditions in the Northeast Atlantic during the Last Glacial Maximum
based on marine microfossil records 252
4.47 Mg/Ca calibration results for several species of planktonic Foraminifera 253
FIGURES AND TABLES xxi
4.48 Rapid changes in SST in the Alboran Sea (Mediterranean Sea) over the past 50 ka 253
4.49 a) Fossil right femur of a mammoth of Middle Pleistocene age discovered at the site of West Runton,
Norfolk, UK. b) Occlusal surface and roots of the first lower molar of the ancestral water vole
(Mimomys savini) recovered from Middle Pleistocene deposits at Pakefield, Suffolk, UK.
c) Reconstruction of woolly mammoth (Mammuthus primigenius) based on carcass remains
discovered, in Yakutia, arctic Siberia 255
4.50 Age estimates of Pleistocene records of the spotted hyena (now confined to Africa) in Europe 259
4.51 Reconstructions of the large (Mammuthus colombi) and pygmy (M. exilis) mammoth species that
co-inhabited Santa Rosa Island, California between c. 200 and 11 ka 260
5.1 The ranges of various dating methods discussed in the text 268
5.2 Accuracy and precision with respect to age estimates derived for a sample with a true age of
10 ka BP 269
5.3 Decay curve for radiocarbon 271
5.4 The 5 MV National Electrostatics Corporation Accelerator Mass Spectrometer at the Scottish
Universities Environmental Research Centre, East Kilbride, UK 273
5.5 The exposed sample target wheel (ion source) of the AMS in Figure 5.4 273
5.6 Schematic diagram of the components of a tandem accelerator 274
5.7 Radiocarbon years versus calendar ages for the period 9.0–11.0 ka BP (10.5–12.5 cal. BP) 275
5.8 Radiocarbon calibration dating series for the period 0–50 ka 280
5.9 Age–depth plot of Cariaco Basin sediments after being tuned (matched) to the Hulu Cave record 281
5.10 The influence of the shape of the radiocarbon curve on calibration 281
5.11 The conversion of a Gaussian radiocarbon error range to a non-Gaussian, multi-modal error
distribution in calibrated time 282
5.12 Wiggle-match dating of a peat sequence in the Netherlands 283
5.13 Branching decay of 40K 284
5.14 Chain decay pathways and half-lives of intermediate nuclides during the decay of 238U, 235U and
232
Th to stable lead 286
5.15 Methods of measuring palaeodose (DE) after the natural TL intensity of a sample (N) has been
established 291
5.16 a) Glacial moraine in the Nubra Valley to the north of Leh, Ladakh, northern India. b) Boulders
on the moraine surface dated by 10Be 297
5.17 a) Cross-cut of a tree trunk showing seasonally differentiated growth rings. b) X-ray negatives of
wood surfaces for analysis by a high-resolution microdensitometer, illustrated schematically in c) 299
5.18 Age profile of cross-matched tree-ring series from Finnish Lapland for the period
AD 1200–160 BC 300
5.19 Standardization of ring-width measurements to generate ring-width indices 301
5.20 Ancient bristlecone pines growing in the White Mountains of California 301
5.21 a) Fossil oak trunks excavated from gravel beds of the River Danube valley. b) A section through
one of the trunks of subfossil oak 302
5.22 Part of a continuous 7,519-year pine tree-ring chronology from Finnish Lapland 303
5.23 Simplified model of clastic and organic varve formation 305
5.24 The pattern and rate of ice retreat across Scandinavia based on a combination of clay-varve chronology
and radiocarbon dating 308
5.25 a) Part of a varved sequence from Lake Suigetsu, Japan. b) Photomicrographs of part of the Suigetsu
varve sequence in plain light and polarized light 310
5.26 Line-scan images from the NorthGRIP ice core 311
5.27 The NorthGRIP oxygen isotope record and variations in annual ice-layer thickness plotted against
the GICC05 timescale 313
5.28 Lichenometric growth curves for Rhizocarpon spp. from West Greenland, West Spitsbergen and
Baffin Island 315
xxii FIGURES AND TABLES
5.29 Radiographs of core-slabs cut from a) Pleistocene and b) modern coral reefs (Montastraea) in
Florida Bay, Gulf of Mexico, showing seasonal growth layers. c) Cross-matching of growth
layers of Porites coral from three locations on the central Great Barrier Reef, Australia, using
UV luminescent illumination 316
5.30 A 54-year annual coral stable isotope record from the mid-Holocene 317
5.31 Annual coral δ18O records from various sites in the Pacific and Indian Oceans 318
5.32 Secular changes in magnetic declination and inclination as observed in London, Rome and
Boston 321
5.33 Palaeomagnetic dating of sediments from Lake Päijänne, Finland 322
5.34 The palaeomagnetic timescale of the last 3.5 Ma 323
5.35 Variations in the earth’s magnetic inclination and field strength (palaeointensity) during the last
2.5 Ma 324
5.36 Magnetic susceptibility variations through a complete Quaternary loess sequence at Luochuan on
the Loess Plateau, Central China 325
5.37 a) Yellow and black tephra layers exposed in a soil–peat complex in Iceland. b) The North Atlantic
Ash Zone 2 tephra layer preserved in the NGRIP ice core. c) Fluted distal shards of the Campagnian
Ignimbrite from Campania, Italy. d) Platy distal shards of the Vedde Ash from Katla Volcano,
Iceland 326
5.38 Tephra layers exposed in sections at two sites in Montana, northwest USA 328
5.39 Widespread tephra marker horizons in marine and terrestrial sequences in the central Mediterranean
and eastern Europe 329
5.40 Orbitally tuned chronostratigraphy for a composite (‘stacked’) deep-ocean isotopic record spanning
the last 300 ka 331
5.41 Comparison of orbitally tuned age estimates based on SPECMAP and independently dated U-series
coral ages 331
5.42 Chemistry of amino acids 333
5.43 Amino-acid geochronology of coastal barrier sediments (aeolinites) in the region of the Murray
River mouth, South Australia 337
5.44 An aminostratigraphic framework for the Thames terraces of southeast England 338
5.45 Weathering and soil-forming indices used to establish relative age of landform surfaces 341
5.46 Stratigraphic and temporal resolution in Quaternary sediment sequences 344
6.1 Time-transgression in lithostratigraphic and biostratigraphic boundaries 349

6.2 Lithostratigraphic subdivision of a glacigenic sequence 350
6.3 Sequence of sedimentary units deposited near continental margins during cycles of sea-level fall and
rise 352
6.4 Various types of biozone used in the subdivision and correlation of strata 353
6.5 Schematic diagram showing a series of moraine ridges between which sediments have accumulated
in lake basins following deglaciation 354
6.6 a) Oxygen isotope stratigraphy in a 53 m core obtained from the Indian Ocean east of the
Maldives Platform. b) A stacked record of Core MD9009643 and ODP core 677 from the eastern
equatorial Pacific 357
6.7 The last four terminations in different proxy records 358
6.8 A time–distance diagram showing the onset and end of glaciation at sites at increasing distance from
the ice-dispersal centre 359
6.9 Different ways of defining the onset of an interglacial 360
6.10 Stratigraphic subdivision of a Quaternary sedimentary sequence and the geologic-climatic units that
can be inferred 361
6.11 The Quaternary sequence in the Netherlands matched against the marine oxygen isotope record 368
6.12 Comparison of biogenic silica and diatom records from Lake Baikal with the marine oxygen isotope
signal from ODP site 677 369
FIGURES AND TABLES xxiii
6.13 The sequence of loess–palaeosol units preserved at the Baoji site on the Loess Plateau, China, plotted
against the marine oxygen isotope record 370
6.14 Matching of marine sediments in core ENAM93-21 from the northeast Atlantic Ocean with the
GRIP δ18O ice-core record using Heinrich events (H1–H5) as tie-points 371
6.15 Tentative correlations between records from the Greenland ice sheet, northern North Atlantic,
Arabian Sea and Brazilian Atlantic margin 372
6.16 Comparison between colour reflectance of Cariaco Basin sediments and the GISP2 δ18O record
over the last glacial cycle 373
6.17 Correlation between the δ18O record from Hulu Cave, China, and the δ18O profile from the GISP2
Greenland ice core 374
6.18 Schematic diagram showing the principles and limitations of tuning palaeoenvironmental records
from two cores 375
6.19 Synchronization of four Lateglacial and early mid-Holocene lake records from northern
Germany 376
6.20 Stratigraphic framework for Quaternary marine sequences in the Mediterranean based on sapropel
units, foraminiferal biozones and magnetozones 377
7.1 a) Schematic representation of coupled box model. b) Box model results simulating the
100 ka Milankovitch climatic cycle 381
7.2 Cartesian grid arrangement for a GCM 382
7.3 Comparison of GCM simulations of global temperature change since AD 1950 compared with
instrumentally measured temperature variations 383
7.4 Boundary conditions for the COHMAP simulation for the last 18 ka 384
7.5 Palaeoclimatic model simulations for the last 18 ka 385
7.6 Feedback framework for the Integrated Global System Model of MIT 386
7.7 Deep Pacific records of environmental change during the past 5 Ma 390
7.8 a) Marine oxygen isotope record for the last 2.1 Ma and the spectral signal of the 41 ka and 100 ka
cycles from this record. b) A measure of the strength of correlation between the stacked oxygen
isotope record and Milankovitch forcing over the past 1 Ma 391
7.9 Modelled time-series of ice volume and Northern Hemisphere subarctic surface air temperatures
from 1.5–0.5 Ma, encompassing the Mid-Pleistocene Transition 394
7.10 Records of deuterium (δD), dust flux and aerosols over the last 800 ka obtained from the EPICA
Dome C ice core, East Antarctica, compared with sea-level changes inferred from the LR04
marine isotope record 395
7.11 Isotope time-series data for the past c. 360 ka from three East Antarctic ice-core records 396
7.12 Radiative forcing, temperature profile and the obliquity signal in the EPICA Dome C ice core over
the past 800 ka 396
7.13 Milankovitch insolation record for the past 800 ka a), compared with temperature b), CO2
c) and dust profiles d) from the EPICA Dome C ice core, Antarctica 397
7.14 Temperature anomaly record for the last 800 ka derived from the EPICA Dome C ice-core
deuterium record 398
7.15 The δ18O record from the NorthGRIP ice core over the last 125 ka 400
7.16 Correlation between variations in abundance of N. pachyderma from North Atlantic deep marine
sites DSDP-609 and V23-81, and the δ18O record from the GRIP Summit ice core 401
7.17 Schematic model of the main currents of the global ‘ocean conveyor’ 402
7.18 Cross-section of the Atlantic Ocean basin showing thermal stratification, principal flows and
surface heat exchange 402
7.19 Schematic model of three dominant types of Atlantic circulation that recurred throughout the
last cold stage 403
7.20 Stable isotopic records spanning the last glacial cycle from Antarctica, aligned with the NGRIP
isotopic record from Greenland 405
xxiv FIGURES AND TABLES
7.21 Synchronization of isotopic records from Antarctic and NGRIP ice cores between 10 and 52 ka
using common variations in methane 406
7.22 Schematic representation of Atlantic water mass changes during DO cycles 408
7.23 Variations in concentration of terrestrial dust over the last 80 ka shown against δ18O variations,
a proxy for temperature fluctuations, in the NGRIP ice core 410
7.24 Terrestrial dust flux and sea-level reconstruction for the period 36–64 ka BP based on analysis of
proxy records obtained from core GeoTü-KL11 in the Red Sea 411
7.25 Schematic representation of the role of dust in the earth feedback system 412
7.26 Isotopic records from Antarctica and Greenland for the Last Termination, defined here as the interval
between 20 and 11.7 ka 414
7.27 Stratotype scheme for the Lateglacial–early Holocene in the North Atlantic region 415
7.28 Temperature variations in the Pacific Ocean during the Last Termination shown against ice-core CO2
data, ice-core temperature reconstructions and summer insolation in the Southern Hemisphere 416
7.29 a) Global temperature stack for the Last Termination based on eighty globally distributed records,
compared with temperature and atmospheric CO2 concentration records from Antarctic ice-cores.
b) Mean temperature variations for stacks of proxy records arranged in 30° latitude bands 417
7.30 Hypothesized chain of events during the Last Termination 418
7.31 Quantified palaeotemperature records for the Lateglacial and early Holocene from Scotland and
Norway aligned using tephra isochrons 421
7.32 Lateglacial temperature record from Lake Suigetsu Japan, derived from a pollen-based transfer
function 423
7.33 Location of the Kråkenes and Meerfelder Maar sites that contain high-resolution records of the
GS-1 interval and inferred positions of the Polar Front 424
7.34 a) The end-moraine formed by a cirque glacier that blocked local drainage to create Kråkenes
Lake on the western coast of Norway. b) A varved record of the YD interval which contains the
Vedde Ash layer 425
7.35 Comparison of the timing of the mid-YD transitions reflected in the Kråkenes and Meerfelder
Maar records 426
7.36 Temperature anomalies for the Holocene derived from seventy-three globally distributed
temperature records with its 1σ uncertainty 427
7.37 a) Visual stratigraphy of the NGRIP ice core across the Pleistocene–Holocene boundary.
b) The precise location of the Pleistocene–Holocene boundary 428
7.38 a) The δ18O record showing the position of the Pleistocene–Holocene boundary in the NGRIP
core. b) High-resolution multiparameter record across the Pleistocene–Holocene boundary 429
7.39 Proxy records for the 8.2 ka event 430
7.40 Proxy records for the 4.2 ka event 432
7.41 Solar activity throughout the Holocene 433
7.42 Comparing historical reconstructions of near-global land temperatures (‘CRUTEM4’) with
Callendar (1938) and Callendar (1961), using a reference period of 1880–1935 437
7.43 Atmospheric concentrations of greenhouse gases over the past 2 ka 438
7.44 Concentration of CH4 and CO2 over the last 11 ka, showing departures (increases) from the
natural trace-gas trends 438
7.45 Schematic diagram showing superimposition of global climatic variations at different frequencies 441
7.46 Significant developments in human history plotted against the NGRIP climate record for the last
glacial–interglacial cycle 442
TABLES
3.1 The Wentworth scale of particle size fractions and the equivalent φ (phi) units 97
3.2 Range of air temperature thresholds for the formation of selected periglacial features estimated by different
authorities in the field 119
FIGURES AND TABLES xxv
3.3 Classification of lakes by mode of formation 151

4.1 Estimates of the pollen production of various plant species 191
4.2 Corrosion and oxidation susceptibility of selected pollen and spores 193
5.1 Approximate δ13C values for various materials 277
5.2 The effect of contamination on true age of samples selected for radiocarbon dating, a) by modern
carbon, b) by inert carbon 278
5.3 Reliability of uranium-series dates for terrestrial materials due to deviations from closed system
behaviour and contamination 288
5.4 Summary of characteristics and scope of nuclides used in surface exposure (CRN) dating 295
5.5 GICC05 chronology of climatic events represented in the NGRIP and GRIP ice-core records for
the period c. 47 to c. 8.1 b2k (years before AD 2000) 314
5.6 Age calculations based on D/L ratios of five amino acids measured in fossil terrestrial ostracods
from central and southern Spain 335
6.1 Principal descriptive criteria used in defining lithofacies in glacigenic sequences 351
6.2 Conventional hierarchy of chronostratigraphic and geochronological units 362
Thispageintentionallyleftblank
Preface to the
third edition
Looking again at the Preface to the second edition of the implications for global palaeoceanography and palaeo-
Reconstructing Quaternary Environments, we get a sense of climatology of the deep-ocean marine isotope signal, and
déjà vu. That book took around six years to see the light of the consequent ramifications for subdividing and cor-
day, and we described the gestation period as bordering on relating the Quaternary record. Over the last decade or so,
the elephantine. How therefore should we describe the time it has been the polar ice-core records that have taken us into
that has been spent on producing this third edition? We new areas of Quaternary science, revealing compelling
began work more than ten years ago, fully intending to do evidence of the rapidity and frequency of climate changes
a ‘light touch’ revision of our 1997 text. But we should have on millennial, centennial, decadal and, in some cases,
learned our lesson. We said in 1997 that one of the principal annual timescales. In addition, like the marine isotope
causes of delay in delivering a final manuscript had been records before them, they offer a basis for time-stratigraphic
the enormous volume of Quaternary literature that had subdivision and correlation at regional, hemispherical and
appeared in the previous dozen years or so. Since then, the global scales. When integrated with the marine records, they
acceleration in the rate of publication has been positively afford often startling new insights into the operation of the
exponential, so much so in fact, that to cover the ground global ocean–atmosphere–cryosphere system.
and digest this monumental corpus of material has almost A feature of Quaternary science over the past two
proved to be beyond us. Having just about got away with decades or so has been the emergence of a number of
it this time, our feeling at the moment is that we will be multidisciplinary and interdisciplinary research groups
unlikely to take on a fourth edition! to investigate some of these short- and long-term climatic
In revising the book, we have retained the tried and changes. As we emphasise throughout the book, cutting-
tested format that we used in the earlier editions, in which edge Quaternary research involves scientists from a range
we first review the various forms of evidence (geomor- of different backgrounds, and it is now commonplace to
phological, lithological and biological) that comprise the find research collaborations in which oceanographers
Quaternary record, before moving on to show how we can are involved with ice-core scientists, and isotope geo-
set this evidence in the context of time (dating) and finally chemists work hand-in-hand with palaeoecologists. And,
to consider how we can bring all of these elements together with the rapid development and refinement of high-
within a robust stratigraphic framework. But in Chapter 7, powered computers, a key component of these groupings
we have aimed for something different. In the final chapter is researchers from the numerical modelling community.
of the second edition, we constructed a narrative of events Over the past twenty years or so, it has been our pleasure
for the last glacial–interglacial cycle in the North Atlantic (and privilege) to coordinate two of these multidiscip-
region. Here we take a broader view and examine a series linary research groups: The North Atlantic Seaboard
of themes related to patterns and causes of climate change Programme (1990–5), which was a constituent compo-
at a range of spatial scales, and over a series of time intervals nent of IGCP-253 ‘Termination of the Pleistocene’, and
that become progressively shorter as we approach the INTIMATE (INTegration of Ice-core, MArine, and TErres-
present day. In the 1990s when the second edition of trial Records), initially a Working Group and latterly a
Reconstructing Quaternary Environments was in preparation, Focus Group of the International Union for Quaternary
the Quaternary community was still coming to terms with Research (INQUA). At the outset, the focus of INTIMATE
xxviii PREFACE TO THE THIRD EDITION
was on the North Atlantic province and specifically on Turney, Simon Blockley and Sune Rasmussen who have
events at the end of the Last Cold Stage (c. 8,000–15,000 followed us as the principal officers of INTIMATE, and who
years ago). More recently (2010–14), however, INTIMATE have taken the research group into new and exciting areas.
has become a research group within the European Union Being involved with INTIMATE has been a stimulating
COST Initiative programme, with a pan-global remit and experience for both of us, and has informed much of the
a temporal range extending back to 60,000 years ago. science with which we have been involved over the past two
Many colleagues have been active in both research decades. But it has also been a source of enormous pleasure,
groups and, while some are no longer with us (most notably for an INTIMATE Workshop is not only about science, but
Bernd Becker, Gerard Bond, Russell Coope, Sigfus Johnsen, it is about meeting old friends and making new ones. For
Klaus-Felix Kaiser, Nick Shackleton and Bill Watts), happily all of these reasons, we would like to dedicate this book to
the majority still are and, and in addition, we have brought our INTIMATE colleagues, and to INTIMATE itself – past,
in a steady stream of active, energetic and innovative present and future.
younger scientists along the way. It would be impossible to
mention all of our colleagues individually, but we would John Lowe
like to acknowledge the contributions of Wim Hoek, Chris Mike Walker
Acknowledgements
We are grateful to many friends and colleagues, as well as some new acquaintances, who have provided us with illustrative
material; to our publishers (Pearson International and Routledge) for their patience and for not giving up on us; to our
editorial contacts (Rufus Kurnow, Patrick Bond, Andrew Mould, Pippa Mullins, Sarah Gilkes, Lisa Salonen, and Eliza
Wright) for their constant encouragement and technical advice; to our families for their continued forbearance and support
while yet another book imposed on domestic routines; and most of all, to Jenny Kynaston, for her skilful adaptations of
the often complex diagrams and figures, and for her undying patience. To all of you, very many thanks.
John Lowe
Mike Walker
Cover image details
The Joides Resolution trans-ocean core-drilling vessel. Coring in Lake Suigetsu, Japan. The photograph shows the
The Joides Resolution is a purpose-built ocean drilling floating drilling platform that was used in the summer of
vessel that can recover cores from the sea floor in water 2006 to collect a core sequence from the centre of the lake
depths of up to 5,980 m. It is equipped with drilling in a water depth of approximately 34 m. The coring
apparatus capable of extracting cores from soft and hard employed a hydro-pressure, thin-walled piston sampler
sediment, as well as from rock. The core barrels employed (diameter 7.8 cm) in four parallel bore-holes to obtain a
enable single cores of 9.5 m length and 57 mm diameter continuous composite core (SG06) of over 73 m; this
to be retrieved, and sequential drilling can penetrate up to contains a palaeoenvironmental record extending back to
300 m below the seabed. The vessel is used for a wide range Marine Isotope Stage 6 (c. 150 ka). The upper 46 m of the
of scientific investigations, mainly under contract to the Lake Suigetsu sediment profile are marked by clearly-
International Ocean Drilling Program (IODP), including defined laminations (varves) spanning approximately the
some of the Quaternary palaeoceanographic research last 60 ka, and which provide a basis for calibrating the
described in this book. In this image the vessel is off the Osa radiocarbon timescale (Chapter 5). Photograph courtesy
Peninsula, Costa Rica, taking core samples for a study of of Takeshi Nakagawa, Ritsumeikan University, Kyoto,
earthquake and tsunami mechanisms. Photograph courtesy Japan.
of Arito Sakaguchi, Yamaguchi University, Japan.
1
CHAPTER ONE
The Quaternary
record
1.1 INTRODUCTION were interspersed with warm episodes (interglacials) during

which temperatures in the mid- and high-latitude regions
The Quaternary is the most recent major subdivision of the were occasionally higher than those of the present day.
geological record, and it extends up to, and includes, the During the last interglacial in Britain around 120 ka,1 for
present day. Together with the Neogene and Palaeogene it example, hippopotamuses swam in the River Thames,
forms the Cenozoic, the fourth of the great geological eras while lions and elephants roamed the present site of Trafal-
(Figure 1.1). The Quaternary has long been considered to gar Square in central London! What makes the Quaternary
be synonymous with the ‘Ice Age’ or the ‘Glacial Epoch’, a different, however, is not simply the occurrence of repeated
view that can be traced back to the writings of Sir Edward warm or cold episodes, for fluctuations in global climate
Forbes in 1846. One of the most distinctive features of the are apparent throughout the Cenozoic (Zachos et al., 2001).
Quaternary has certainly been periodic glacier activity Rather, it is a combination of the high amplitude and
during cold periods, with the build-up of major continental frequency of climatic oscillations, coupled with the intensity
ice sheets and the expansion of mountain glaciers in many of the colder periods, that gives the Quaternary its unique
parts of the world. However, these cold or glacial stages character. In some parts of the world, temperatures may
a} O c e a n i c b) Terrestrial
Eons
Eons
Eras
Periods 0 Periods E S S E evidence evidence

0 Quaternary Quaternary Pleistocene ILasI temperat Holocene
Ceng
-zoic
Tertiary 2 ILasI temperate stage

Pliocene 0.2
65 For older
Neogene
UIO Cretaceous 10 0.4

Mesozoic
Quaternary
Jurassic Miocene 0.6 stages see
200 20 F i g u r e 1.5
Triassic 08
23.3
E E 104
Permian 1.0
Oligocene stages
Cenozoic
300 30
Age (Ma)
Carboniferous 1.2 SO
Age (Ma)
Phanerozoic
33.9
Palaeozoic
Devonian cold
Palaeogene
1.4
400 40 temperate
Silurian 1.6
Eocene cycles
Ordovician
1.8
500 50
Cambrian 2 0
570
600 60 E o c eEnoec e n e 2 2
P rote ro zoic
Precambrian
65 2.4
70 2.6
Figure 1.1 The Quaternary relative to the geological timescale. (The oxygen isotope trace (section 3.10) from deep-ocean
sediments a) is after Shackleton et al., 1990).
2 THE QUATERNARY RECORD
have fluctuated by more than 15°C between warm and cold change has produced similar environmental conditions at
episodes, temperature change was frequently rapid different times and, because many records cannot be dated
(measurable, in some cases, over decades), and the last 800 precisely, the process of correlation is frequently beset
ka alone have witnessed ten full glacial–interglacial cycles. with difficulties. Hence, in a single exposure of Quaternary
The deep-ocean sediment record (see section 1.6) suggests sediments, there may be much to perplex the geologist,
that over the course of the full range of Quaternary time, the geomorphologist, the botanist, the zoologist or the
the world may have experienced more than fifty cold or archaeologist, and an explanation of observed geological
glacial stages and a corresponding number of temperate or changes will often require the combined expertise of all of
interglacial periods (Lisiecki & Raymo, 2005). these disciplines. The purpose of this book is to illustrate
The effects of these marked shifts in climate were the very wide range of methods that are currently employed
dramatic. In the mid- and high latitudes, ice sheets and in Quaternary research, and to demonstrate that both a
valley glaciers advanced and retreated, and the areas affected multidisciplinary and an interdisciplinary approach are
by periglacial (cold climate) processes expanded and con- required if a proper understanding of the complexities of
tracted. In low-latitude regions, the desert and savannah the Quaternary environment is to be achieved.
margins shifted through several degrees of latitude as
phases of aridity alternated with episodes of higher
precipitation. Throughout the world, weathering rates and 1.3 THE STATUS OF THE
pedogenic processes varied with changes in temperature QUATERNARY IN THE GEOLOGICAL
and precipitation, river regimes fluctuated markedly, sea
levels rose and fell over a vertical range of c. 150 m, and
TIMESCALE
plant and animal populations were forced to migrate and In the terminology of the geological timescale, intervals
adapt in response to these environmental changes. of time that can be measured in the rock (stratigraphic)
record, are referred to as geochronological units, and the
major episodes (eras) are divided, respectively, into sub-
1.2 INTERPRETING THE eras, periods, epochs, ages and chrons. These are manifest
in the geological record as chronostratigraphic units,
QUATERNARY RECORD which is the term that refers to the sequence of rocks that
The repeated climatic fluctuations that have occurred formed over particular time intervals; here, the hierarchy
throughout this latest chapter of earth history have given of subdivision is erathem, sub-erathem, system, series and
rise to a highly complex record of landforms, sediments, stage. The distinction between geochronology on the one
biological (including human) remains and assemblages hand and chronostratigraphy on the other is important
of human artefacts. From this legacy, it is possible to when the geological record is being considered, and this
reconstruct, often with great clarity and in considerable is discussed further in Chapter 6 (section 6.2.3.7). The
detail, the environmental conditions and associated palaeo- Quaternary has long been considered to be a geochrono-
geography of particular intervals of Quaternary time. logical unit of period rank within the Cenozoic era, and to
There are a number of separate stages in this process of contain two separate epochs: the Pleistocene (originally
palaeoenvironmental reconstruction: first, the establish- meaning ‘most recent’), which ended around 11.7 ka, and
ment of the stratigraphy at each site in order to develop the Holocene (‘entirely recent’), which is the present warm
a geological framework for the investigation; second, interval in which we live (Figure 1.2). These are represented
the analysis of proxy records2 from those stratigraphic in the stratigraphic record as the Quaternary system and
sequences to produce the basic palaeoenvironmental the Pleistocene and Holocene series. In some quarters, there
information; third, the construction of a chronology has been a tendency to regard the Holocene as simply the
of events, which involves the development of a dating latest in a series of warm episodes (interglacials) forming
framework; fourth, the linking of individual sequences part of a long-term climatic cycle (see below), and hence
from different locations by means of correlation; and to consider this as part of the Pleistocene. However, the
finally, the integration of different lines of evidence to Holocene is now widely accepted as a separate unit because
produce an overall palaeoenvironmental synthesis. Each of the importance in the present interglacial of the evolution
one of these stages contains its own set of problems. The of the human environment. Indeed, it is this anthropogenic
terrestrial stratigraphic record is often fragmented; evidence signature that is the hallmark of the Holocene, and which
is absent from many areas, while detailed sequences are only justifies its status as a unit of series/epoch rank within the
locally preserved. Moreover, the cyclical nature of climatic geological timescale (Gibbard & van Kolfschoten, 2005).
THE QUATERNARY RECORD 3
Era them timescale (Gradstein et al., 2005), despite the fact the term
System had been used by earth scientists for more than 150 years.
Period
GSSP
Series Stage Age
Era
Epoch Age Ma However, following a protracted campaign by the Sub-

commission on Quaternary Stratigraphy (SQS) and the
Holocene
International Union of Quaternary Research4 (INQUA),
0.0117 the body that represents Quaternary Science worldwide
Quaternary
Upper (e.g. Gibbard et al., 2005; Bowen & Gibbard, 2007), the
0.126
Ionian Quaternary was, in 2009, reinstated and recognized by
Pleistocene 0.781 the IUGS as a formal unit of system/period rank within
Calabrian
0.781 the geological timescale (Pillans & Gibbard, 2012; Figure
Gelasian 1.2). Note, in passing, that the ‘Tertiary’, which formerly
2.58
Piacenzian
encompassed what are now the Palaeogene and Neogene
Pliocene 3.600 system/periods, is no longer considered by the IUGS to
Zanclean
5.333
be a formal geochronological unit, although the term con-
Messinian tinues to be used informally (as also is the case with
7.246 ‘Precambrian’, for example), and it will be employed in this
Neogene
Tortonian
11.62 way throughout this book. Indeed, it is possible that, in
Serravallian due course, the Tertiary may also be reinstated along with
Miocene 13.82
the Quaternary as a formal period within the geological
Cenozoic
Langhian
15.97 timescale (Knox et al., 2012).
Burdigalian
20.44
Aquatanian
23.03 1.4 THE DURATION OF THE
Chattian
28.1
QUATERNARY
Oligocene
Rupelian Defining the onset of the Quaternary has also been a
33.9
Priabonian contentious issue. For many years, it was considered that
38.0 the Quaternary lasted for approximately one million years
Palaeogene
Bartonian
Eocene 41.3 (a figure derived from extrapolations based on weather-
Lutetian ing profiles), and that it could be differentiated from the
47.8
Ypresian
preceding Tertiary on the basis of evidence for widespread
56.0 glaciation. It is now apparent, however, that many of
Thanetian the high-latitude regions supported glaciers long before
59.2
Paleocene Selandian the onset of the Quaternary. There is evidence, for example,
61.6 of glacial activity during the Late Miocene (10–9 Ma)
Danian
66.0 in Greenland (Lykke-Andersen, 1998) and during the
Middle Miocene (16–15 Ma) in Alaska (Lagoe et al., 1993),
Figure 1.2 The Cenozoic timescale, as defined in 2014. The while in Antarctica the Cenozoic glacial record can be
boundary stratotypes that have been formally ratified by the traced back to 40–35 Ma (Ingólfsson, 2004). These records
International Union of Geological Sciences as global stratotype of early glaciation reflect the fact that although global
section and points (GSSPs: Section 1.4) are shown by the arrow
symbols in the right-hand column. Note that some stages/ages temperatures had oscillated, there had been a long-term
of the Pleistocene series/epoch have yet to be formally ratified high-latitude cooling that began in the Late Miocene
by the IUGS (after Gibbard et al., 2010). (Maslin et al., 1998). In the geological column, therefore,
the Pliocene–Pleistocene (or Neogene–Quaternary) bound-
Although the term ‘Holocene’ was formally adopted ary cannot be drawn simply on the basis of direct terrestrial
by the International Geological Congress in 1885, and glacial evidence. Instead, the boundary has usually been
its lower boundary was ratified (albeit many years later) at located at that point in the stratigraphic record where
11.7 ka by the International Union of Geological Sciences3 there are the first clear indications of climatic cooling,
(Walker et al., 2009), until recently the status of the reflected either in the fossil evidence or in some other
Quaternary (and also of the Pleistocene) as a geological/ climatic proxy.
temporal unit remained unresolved. Indeed, the Quatern- In geology, all formal chronostratigraphic units and
ary was omitted completely from the 2004 geological the boundaries between them are defined on the basis of
type/reference sections or stratotypes, also known as Despite the fact that the Vrica section had been accepted
Global Boundary Stratotype Sections and Points or GSSPs, as the internationally ratified GSSP for the Pliocene–
where significant changes occur in the fossil/climate record Pleistocene boundary, there was a widespread feeling within
(Chapter 6). The need for an objective boundary stratotype the international Quaternary community that the boundary
for the base of the Quaternary (and Pleistocene) was first should be located earlier in the geological record at 2.8–2.4
recognized at the International Geological Congress in Ma (Gibbard et al., 2005; Head et al., 2008a). This is
London in 1948, but it was not until 1982 at the INQUA because it was during that interval that one of the most
Congress in Moscow that the Vrica section in Calabria, significant transitions in the Cenozoic history of the earth
southern Italy was formally proposed as the GSSP for the occurred, notably the initiation of a pattern of glacial–
Pleistocene epoch. There the boundary was placed at the first interglacial cycles that have dominated global climate to the
appearance of the cold-water marine ostracod Cytheropteron present day (Pillans, 2004). As we shall see, this major shift
testudo, and dated on the basis of palaeomagnetic evidence in the earth’s climatic rhythm, centred on c. 2.6 Ma, appears
to c. 1.64 Ma (Aguirre & Pasini, 1985). Subsequent revision to have been driven by variations in the earth’s orbit and
of the palaeomagnetic timescale using ocean-core evidence, axis (section 1.6), and involved a change in pacing from
however, suggests an older age for the Olduvai event (section 23 ka to 41 ka climatic cycles. In particular, the shape of
5.5.1.2), and this has been confirmed by calculations based the climate cycles, as reflected in the marine oxygen isotope
on astronomical parameters which give a date of 1.806 Ma (δ18O) records5 (section 3.10), becomes increasingly asym-
(Lourens et al., 2005). This method, which is often referred metrical (‘saw-toothed’) after this time, suggesting a major
to as astronomical or orbital tuning, employs known change in global climate dynamics (Lisiecki & Raymo,
variations in the earth’s orbit and axis (section 1.7), and is 2007). In deep-ocean cores from the North Atlantic and
discussed more fully in Chapters 5 and 6. North Pacific, the cooling trend that defines the onset of
Figure 1.3 The Monte San Nicola section in southern Sicily, which contains a sequence of uplifted Mediterranean marine sediments
and sapropels6 spanning the Piacenzian, Gelasian and Calabrian. The top of the sapropelic (dark) Nicola bed marks the base of the
Gelasian stage, now the boundary stratotype for the base of the Quaternary system/period and the base of the Pleistocene
series/epoch (photograph by Allan Ashworth, North Dakota State University, USA).
this change in climate rhythms is marked by increased geologist Jules Desnoyers who, writing in 1829, differen-
quantities of ice-rafted detritus accompanied by significant tiated between the strata of ‘Tertiary’ and ‘Quaternary’ age
shifts in the oxygen isotope signal, reflecting both the in the rocks of the Paris basin. The Quaternary was
first major build-up of continental ice masses and major redefined by Henri Reboul in 1833 to include all strata
reorganizations in patterns of oceanic circulation (Kleiven characterized by the remains of flora and fauna whose
et al., 2002; Bartoli et al., 2005). Evidence for the onset of counterparts could still be observed in the living world. The
cooling is also found in a range of continental records, term ‘Pleistocene’ (most recent) was introduced by Charles
particularly from mid- and high-latitude regions, including Lyell some six years later to refer to all rocks and sediments
loess sediments (Ding et al., 2005), pollen data (Kuhlmann in which over 70 per cent of the fossil molluscs could be
et al., 2006) and faunal evidence (Brugal & Croitor, 2007). recognized as living species. Only after the work of Edward
However, low-latitude regions were also affected with, for Forbes in the 1840s did the term ‘Pleistocene’ become
example, increased regional aridity in northwest Africa synonymous with the glacial period.
between 2.8 and 2.4 Ma, and the replacement of closed Quaternary studies represent one of the youngest
canopy forest by grassland savannah (Leroy & Dupont, branches of the geological sciences, with a history that
1994). goes back less than 200 years. Prior to that it was generally
The pan-global events at around 2.6 Ma therefore mark believed that the earth had been created in 4004 BC, a figure
what is perhaps the most fundamental reorganization in the based on genealogical calculations from biblical sources by
earth’s climate system since the cooling trend associated Archbishop James Ussher of Armagh and first published
with the glaciation of Antarctica at around 35 Ma, and in 1658. Since early views on geological and environ-
appears to have been of much greater magnitude than the mental changes were constrained by the Ussher timescale
changes at around 1.8 Ma. This major climatic change had of around 6,000 years, a Catastrophist philosophy held
already been acknowledged by the geological community sway in which the form and character of the earth’s surface
in the definition of a new Pliocene stage, the Gelasian were explained largely through the operation of great
(Figure 1.2), the lower boundary of which is marked by the floods and other cataclysmic events. Around the turn of the
2.6 Ma event. The stratotype for the Gelasian is located in eighteenth century, however, the work of the famous
a section at Monte San Nicola in Sicily (Figure 1.3), and Edinburgh geologists James Hutton and John Playfair
dated by orbital tuning to 2.588 Ma (Rio et al., 1998; began to show that the features of the earth’s surface could
Lourens, 2008). In 2006, the SQS and INQUA formally more reasonably be explained by the operation, over a
requested the International Commission on Stratigraphy protracted timescale, of processes similar to those of the
to accept the proposal that the base of the Quaternary, and present day. This significant departure in geological
also the Pleistocene, be lowered to 2.588 Ma, and that the thinking gave rise to the principle of Uniformitarianism,
GSSP should be that previously defined for the Gelasian first expounded by Hutton, but subsequently popularized
stage. In other words, what had formerly been the upper- by Lyell in his famous dictum ‘the present is the key to the
most stage of the Pliocene would now become the lowest past’. Uniformitarian reasoning, in which present-day
defined stage of the Pleistocene and Quaternary, and this analogues are used as a basis for the interpretation of
was accepted by the IUGS in 2009. This major reclassifi- observed features within the stratigraphic record, is still
cation of the later Cenozoic era met all of the requirements fundamental to many aspects of palaeoenvironmental
of the global Quaternary community, as well as respecting reconstruction (Bell & Walker, 2005).
the historical precedents and established usage for the term The nineteenth century saw a number of significant
‘Quaternary’ (Gibbard et al., 2010). advances in Quaternary studies, many of which stemmed
directly from the introduction and gradual acceptance
1.5 THE DEVELOPMENT OF of the Glacial Theory (Woodward, 2014). Although for
many years there had been speculation that certain Swiss
QUATERNARY STUDIES and Norwegian glaciers had formerly been more exten-
sive, it was not until the 1820s that credence was given to
1.5.1 Historical developments the notion of a glacial epoch. The work of Jens Esmark in
The term ‘Quaternary’ was first used by Giovanni Norway and of Albrecht Bernhardi in Germany, and
Arduino in 1759 to describe the fourth stage or ‘order’ that particularly the investigations of the two engineers Ignaz
he identified in the alluvial sediments of the River Po in Venetz and Jean de Charpentier in Switzerland, produced
northern Italy. It was applied in a wider context to refer to evidence for former glacier activity far beyond the limits of
near-surface, largely unconsolidated deposits by the French present-day glaciers. However, it fell to the Swiss zoologist
Louis Agassiz to expound, in 1837, the first coherent theory

of ‘the great ice period’ involving worldwide climatic
changes. Subsequently, Agassiz visited both Britain and
North America and in both areas demonstrated that
surficial deposits that had previously been interpreted as the
products of marine inundation during the flood (‘diluv-
ium’) could more reasonably be regarded as the results of
extensive glaciation in the relatively recent past.
Although the Glacial Theory did not immediately gain
widespread acceptance, its adherents rapidly refined and
developed the concept. By the 1850s, evidence was begin-
ning to emerge for two glaciations in parts of Britain and
Europe and, as early as 1877, James Geikie was describing
evidence for four separate glaciations in eastern England.
The strata between the glacial deposits (drift) were referred
to as ‘interglacial’, and hence the idea of oscillating warm
(‘interglacial’) and cold (‘glacial’) episodes emerged. By the
end of the nineteenth century, drift sheets of four separate
glaciations (the Nebraskan, Kansan, Illinoian and Wiscon-
sinan), along with deposits of three intervening interglacials
(in descending order of age, the Aftonian, Yarmouthian Figure 1.4 The maximum glaciation of the Northern Hemi-
and Sangamonian) had been identified in North America, sphere during the Quaternary (after Ehlers & Gibbard, 2007).
(Mountain High Maps® copyright © 1993 Digital Wisdom.)
while evidence began to emerge for multiple glaciations in
different parts of Europe. Probably the most influential
work in this respect, however, was that of Albrecht Penck
and Eduard Brückner who, in the early years of the
twentieth century, resolved the river terrace sequences in remarkably close agreement with more recent estimates. In
the valleys of the northern Alps into four separate series, addition to its effects on global sea levels, the results of the
each relating to a glacial episode. The phases of glaciation build-up of ice on the earth’s surface were also noted. A
were named (from oldest to youngest) Günz, Mindel, Riss number of geologists, including John Playfair and Charles
and Würm, after major rivers of southern Germany. In Lyell, had described the raised shoreline sequences in
both Europe and North America, the maximum limits Scandinavia and around the coasts of Scotland, and had
of Quaternary glaciations were first mapped around the inferred that in both regions crustal uplift had occurred.
turn of the twentieth century and have subsequently been The mechanism involved in crustal warping, however,
modified only in detail (Figure 1.4; Hughes et al., 2013), remained unclear. In 1865, the Scottish geologist Thomas
although views on the terminology adopted and on the Jamieson finally made the link between the raised shoreline
number of glacial/interglacial episodes that occurred during evidence and the Glacial Theory when he deduced that
the Quaternary have changed dramatically (see sections 1.3 crustal depression would result from the weight of the ice
and 1.6). sheets and that uplift would follow deglaciation as the
Other effects of glacier expansion and contraction were crust was free to rebound to its pre-glacial state. This was
also recognized at a relatively early stage. The relationship the first clear statement of what are now referred to as
between glaciers and sea level was first considered in a glacio-isostatic effects (section 2.5.4).
systematic manner by Charles MacLaren who, in 1841, During the later years of the nineteenth century,
reasoned that at times of glacier build-up, sea levels would evidence began to emerge for major environmental changes
fall as water was extracted from the ocean basins and in areas beyond those directly affected by glacier ice. In the
locked up in the expanding ice sheets whereas, following semi-arid southwest of the United States, for example,
ice melting, sea levels would rise as water was returned work by Israel Russell and Grove Karl Gilbert in particular
to the oceans. This was the first statement of the Glacio- showed that extensive lakes had existed at some time in the
Eustatic Theory of sea-level change (section 2.5.2). past, and that phases of higher rainfall (pluvial) had
MacLaren suggested that sea levels would fall by 350–400 alternated with more arid (interpluvial) episodes. More-
ft (c. 110–130 m) during a glacial phase, a figure that is in over, a relationship was postulated (although not clearly
articulated) between these climatic oscillations and the 1.5.2 Recent developments
glacials and interglacials at higher latitudes. Similar relict
drainage features in desert and savannah regions in other The last fifty years have seen many important develop-
parts of the world were described by Victorian explorers and ments in Quaternary studies, but five in particular merit
provided further indications of climatic changes in the low attention. The first is the methodological advances that have
latitudes. In the mid-latitude zones, on the other hand, it been made in, and the widespread application of, a range
was gradually recognized that phases of glacier expansion of field and laboratory techniques. Increasingly sophisti-
would be accompanied by an extension of the tundra belt cated methods of sedimentological analysis have offered
where cold-climate (albeit non-glacial) processes predom- new insights into the nature of Quaternary depositional
inated. The term periglacial was first used to describe such environments, while the interpretation of Quaternary
regions by the Polish geomorphologist Walery von Lozinski stratigraphy has been greatly assisted by the development
in 1909. of equipment for coring terrestrial, offshore and deep-
Biological evidence for Quaternary environmental ocean sequences. Analysis of both terrestrial and marine
evidence has been significantly improved by the use of a
change also began to emerge soon after the introduction
range of remote sensing techniques, including airborne
of the Glacial Theory in the middle years of the nineteenth
sensors (e.g. conventional cameras, satellite-mounted
century. The publications of Edward Forbes, in which
imaging systems and radar); ground-based or ship-towed
various geographical components of the British flora and
sonar, radar and seismic systems; and tracer methods for
fauna were related to successive migrations into the British
the analysis of lacustrine and marine processes. Particular
Isles under different climatic conditions, and of Oswald
progress has been made in the mapping, often at very high
Heer wherein ecological changes in Switzerland were
resolution, of sea-bed topography and marine sediment
discussed in the context of Quaternary climatic changes,
architecture through the use of high-resolution sonar
were particularly important milestones. In the later years
and seismic devices. Palaeoecological investigations have
of the nineteenth century, research by the Scandinavian
also benefited from a range of technological advances,
botanists Axel Blytt and Rutger Sernander demonstrated the notably in the extraction, recording and analysis of fossil
wealth of information on climatic and vegetational change assemblages, and in the fields of both light and electron
that could be derived from the stratigraphy and macrofossil microscopy. These various techniques are considered in
content of peat bogs (section 3.9). The scheme of postglacial more detail in Chapters 2–4.
climatic changes constructed by Blytt and Sernander from The second major development has been in the dating
Scandinavian peat bog records was subsequently refined by of Quaternary events. In the nineteenth century, notions
the results of pollen analysis, a technique developed in of time were founded largely on estimates of rates of
Sweden by Lennart von Post and which is still one of the operation of geological and geomorphological processes.
most widely used methods in palaeoecology (section 4.2). Hence, estimated rates of delta construction, cliff retreat,
Systematic investigations of other forms of biological stream dissection, weathering rates and degree of soil
evidence also began during the last century. Important development were all used to assess the duration of
contributions in vertebrate palaeontology included the Quaternary episodes. The first, and for many years the only,
work of Richard Owen, who produced the first compre- quantitative method for estimating the passage of time
hensive volume on British fossil mammals and birds, and was varve chronology developed around the turn of the
of his contemporary, William Buckland, who not only century by the Swedish geologist Gerard de Geer (section
carried out some of the earliest detailed investigations and 5.4.2). A major breakthrough came in the years immediately
analyses of vertebrate assemblages in cave sites, but was one following the Second World War with the discovery, by
of the first British converts to the Glacial Theory. As early Willard Libby, of the technique of radiocarbon dating.
as 1838, James Smith (‘Smith of Jordanhill’) was using fossil Other radiometric methods, notably potassium/argon
shells to demonstrate that the seas around the coast of and uranium-series (U-series) dating, were developed in the
western Scotland had been much colder in the past, thereby 1950s and 1960s, along with the techniques of dendroch-
laying the foundation for subsequent utilization of marine ronology (tree-ring dating) and palaeomagnetism. The
Mollusca as indicators of former marine temperatures 1970s and 1980s saw the refinement of these various
(section 4.7). The seminal works of Alfred Kennard, often methods and a general increase in levels of chronological
in association with Bernard Woodward, in the later part of precision, particularly as a consequence of the introduction
the nineteenth and early years of the twentieth century of mass spectrometry into radiometric dating. In addition,
provided a similar groundwork for the analysis of land and new techniques have been developed, including amino-
freshwater Mollusca (section 4.6). stratigraphy, fission-track dating, electron spin resonance,
luminescence dating, and the use of long-lived cosmogenic A fourth major development over recent decades has
radioisotopes such as 10Be and 36Cl (Walker, 2005). Recent been the coring of polar ice sheets and glaciers. Ice-core
research on ancient DNA involving, in particular, the drilling began on the Greenland ice sheet in the late 1950s,
timing of genetic mutations (‘molecular clock analysis’: and was followed in the 1960s by the drilling of the first
Bromham & Penny, 2003) offers an exciting and potentially deep polar ice core to bedrock at Camp Century, Greenland
valuable addition to the Quaternary dating portfolio. The (Dansgaard et al., 1969). Subsequently, long continuous
principles and applications of the range of dating methods cores have been recovered from other sites in Green-
now available to the Quaternary scientist are discussed in land, from Antarctica, and from other polar ice caps and
Chapter 5. mountain glaciers. The ice layers revealed in the cores
The third important development in Quaternary studies represent annual increments of frozen precipitation, and
during the second half of the twentieth century has been contain a range of proxy indicators (oxygen isotopes,
the stratigraphic investigation of sedimentary sequences on trace gases, chemical compounds, particulate matter) of
the deep-ocean floors. Indeed, it would not be overstating past atmospheric and climatic conditions. Ice-core data not
the case to suggest that the results of research into ocean only provide a temporal framework for Late Quaternary
sediments have revolutionized our view of the Quaternary climatic change (section 3.11), but the upper levels of the
(Imbrie & Imbrie, 1979). In one sense, trying to reconstruct ice cores also record the effects of recent industrial activity.
environmental changes from terrestrial evidence is like The most recent phase of this research has involved the
trying to assemble a jigsaw puzzle and then make sense of drilling to bedrock of cores near the thickest part (3 km)
the picture when more than 90 per cent of the pieces are of the Greenland ice sheet by the European Greenland Ice-
missing. This is because much of the evidence has been core Project (GRIP, NGRIP), the American Greenland Ice
removed by subaerial weathering and erosional processes Sheet Project (GISP2) and the European North Greenland
and, in mid- and high latitudes, by glacial erosion. In parts Eemian Project (NEEM). The data have provided startling
of the deep oceans of the world, however, sediments have evidence not only of the magnitude of climatic change over
been accumulating in a relatively undisturbed manner the last interglacial–glacial cycle (Hammer et al., 1997;
for thousands, or even millions, of years, and therefore North Greenland Ice Core Project Members, 2004), but also
frequently span the entire range of Quaternary time. of the extraordinary rapidity of climate change (Steffensen
Although the investigation of deep-sea sediments et al., 2008). Perhaps even more impressive, given the
actually began in the nineteenth century with the voyage remote location, has been the deep drilling of the Antarctic
of the British government research vessel HMS Challenger ice sheet where the Vostok core has produced a record of
in 1872, it was not until the 1930s that the German climate and atmospheric history for the last four glacial
palaeontologist Wolfgang Schott began the first detailed cycles (Petit et al., 1999), while the European Project for
work on the fossil content of core samples from the ocean. Ice Coring in Antarctica (EPICA) at Concordia Station,
Prior to the Second World War, only short sediment cores Dome C, has generated a palaeoclimatic record extending
(less than 1 m in length) could be raised from the sea back over 800 ka (EPICA Community Members, 2004;
bed. The development of a piston corer by the Swedish Fischer et al., 2010a).
oceanographer Börje Kullenberg heralded the modern The fifth significant advance in Quaternary science,
phase of deep-sea research, for with the Kullenberg corer particularly during the second half of the twentieth century,
and specially equipped research ships, it became possible has been in the development of increasingly sophisticated
to take undisturbed sediment cores of more than 10 m in computer-based models which simulate a range of aspects
length. The changing fossil content of these cores has of Quaternary environments. This type of work began in
provided a remarkable record of changes in ocean water the late 1960s with the development of general circulation
temperatures and, by implication, in global atmospheric models (GCMs), numerical models that were initially
temperatures during the course of the Quaternary (section designed to reconstruct patterns of atmospheric circulation
4.10). Many fossils, however, contain other indices of during the last cold stage, and possible linkages between
environmental change, most notably variations in oxy- terrestrial and atmospheric environments (section 7.2). A
gen isotope content. Pioneered by Cesare Emiliani, oxygen range of increasingly sophisticated models has since been
isotope analysis is now regarded as one of the most powerful developed to explore, in addition to atmospheric circula-
tools in Quaternary stratigraphy and palaeoenvironmental tion, such diverse phenomena as ice sheet behaviour
reconstruction (section 3.10), and continuous isotopic (Siegert & Dowdeswell, 2004), glacio-isostatic effects
records are now available extending back into the Pliocene (Peltier, 2002), oceanographical changes (Weber et al.,
(e.g. Lisiecki & Raymo, 2005). 2007), past vegetation dynamics (Claussen, 2009) and
human migrations (Mithen & Reed, 2002). Some of the for instance, there is only fragmentary evidence for glacier
most impressive results have been achieved, however, activity during the early Quaternary cold stages (Lee et al.,
where scientists from a range of disciplines have collab- 2011), and, indeed, this is also the case for many other
orated to integrate data on Quaternary environmental parts of the world (Ehlers et al., 2011a). It is also apparent
change from a variety of different sources, and to use those that during the last cold stage, Southern Hemisphere ice
data as a basis for both descriptive and predictive modelling contributed less than 3 per cent to the overall increase
of Quaternary environments and environmental change. in global ice volume, prompting the observation that ‘the
Such an approach is typified by the Climate/Long Range growth of ice in the Quaternary was essentially a Northern
Investigation Mapping and Prediction (CLIMAP) group Hemisphere phenomenon’ (Williams et al., 1998). The
(CLIMAP Project Members, 1976, 1981) and by the term ‘glacial’ therefore may have a different connotation
Co-operative Holocene Mapping Project (COHMAP) pro- in the two hemispheres.
gramme (COHMAP Members, 1988; Wright et al., 1993). Because of these difficulties, the terms ‘temperate stage’
These interdisciplinary and multidisciplinary projects are and ‘cold stage’ might be considered more appropriate to
considered in more detail in section 4.10.7. describe the major climatic episodes of the Quaternary.
However, these terms contain their own sets of problems
1.6 THE FRAMEWORK OF THE (defining acceptable thresholds between ‘warm’ and ‘cold’
episodes; quantifying climatic change; conflicting proxy
QUATERNARY records for former climate; etc.) and, as a consequence,
The conventional subdivision of the Quaternary is into are equally arbitrary. Moreover, for historical reasons, it is
glacial and interglacial stages, with further subdivision not always possible to avoid the traditional terminology
into stadial and interstadial episodes. Glacial stages when referring to certain named Quaternary stages. For
have traditionally been regarded as protracted cold phases convenience, therefore, we have opted for the lesser of the
when the major expansions of ice sheets and glaciers took evils and have retained the terms ‘glacial’ and ‘interglacial’
place, whereas stadials have been viewed as shorter cold but, where appropriate, have used these interchangeably
episodes during which local ice advances occurred. Inter- with ‘cold’ and ‘temperate’ stages. This type of categoriza-
glacials are usually recognized as warm intervals when tion, based on inferred climatic characteristics, is known
temperatures at the thermal maximum were as high as, as climatostratigraphy, and is considered further in
or even higher than, those experienced during the Holo- Chapter 6.
cene, and which were characterized in the mid-latitudes Attempts to subdivide the stratigraphic record from
by the development of mixed woodland. Interstadials, by the land areas of the Northern Hemisphere into a coherent
contrast, are traditionally regarded as relatively short-lived scheme of glacial and interglacial stages that has regional
periods of thermal improvement during a glacial phase, or inter-regional application have hitherto proved to be
when temperatures did not reach those of the present day extremely difficult, principally because of the fragmented
and, in lowland mid-latitude regions, the climax vegetation nature of most terrestrial sedimentary sequences. Over the
was boreal woodland. past two decades, therefore, reference has increasingly been
These terms are still widely used in Quaternary science, made to the relatively undisturbed sedimentary sequence
although they clearly lack precision and, as a consequence, in the deep ocean, and particularly to the oxygen isotope
are often difficult to apply. Take, for example, the problem record in the marine microfossils contained within those
of recognizing an interglacial as opposed to an interstadial sediments. As will be shown in Chapter 3, the oxygen
episode on the basis of degree of vegetation development. isotope trace (or ‘signal’) obtained from these microfossils
In northwest Europe, both the interglacials and interstadials reflects the changing isotopic composition of ocean waters
of the Late Quaternary were characterized by a range of over time. Insofar as the marine oxygen isotope balance is
vegetation types (mixed woodland, boreal woodland, open largely controlled by fluctuations in volume of land ice
grassland) depending on latitude, altitude, duration of the (section 3.10), variations in the isotopic signal in fossils
warm stage, etc. In more northerly regions, the ‘vegetational from deep-ocean sediment profiles can be read as a record
signature’ of an interglacial might be boreal woodland; of glacial/interglacial fluctuations (Figure 1.5). Working
further south, this type of forest development would be from the top of the sequence, each isotopic stage has been
more indicative of an interstadial. Hence, the palaeo- assigned a number, with even numbers denoting ‘glacial’
botanical distinction between ‘interglacial’ and ‘interstadial’ (cold) episodes while the ‘interglacial’ (warmer) phases are
becomes blurred by geographical province. Moreover, this denoted by odd numbers. One of the most impressive
terminology may even be misleading. In the British Isles, features of the deep-sea MIS (marine isotope stage) record
is that the isotopic signal is geographically consistent, and between the marine isotope record and terrestrial sequences
can be replicated in cores taken from different parts of the are discussed more fully in Chapter 6.
world’s oceans. Hence, the marine oxygen isotope sequence The most complete system of designated glacial and
provides a climatic signal of global significance. interglacial episodes is that for northern Europe, the British
Twenty-two isotopic stages can be recognized in the Isles and European Russia, with a less detailed formal
past 880 ka or so (Lisiecki & Raymo, 2005), indicating scheme for North America (Figure 1.6). There is general
that around ten or eleven glacials and a similar number of agreement that the Flandrian of the British sequence can
interglacials (or near-interglacials) occurred during that be equated with the Holocene of the European and North
time period, and the total number of isotopic stages American sequences, and that the last cold stage identified
formally identified in the deep-ocean record of the past in Britain (Devensian), northern Europe (Weichselian), the
2.5 Ma now exceeds 100. In the Monte San Nicola type Alpine region of Europe (Würmian) and North America
section (Figure 1.3), for example, the base of the Gelasian (Wisconsinan) can be considered as broad correlatives. The
stage which marks the Pliocene–Quaternary boundary Ipswichian, Eemian, Riss–Würmian and Sangamonian
corresponds to MIS stage 103 (Gibbard et al., 2010). This ‘temperate’ records from each of these regions are also
means that over the course of the Quaternary more than believed to be essentially coeval and are assigned to the last
fifty cold/temperate cycles have occurred, which is many interglacial, despite the fact that identification is often
more temperate and cold stages than have been formally based on quite different types of proxy evidence. In Europe,
recognized and named on the basis of the terrestrial a number of stadials and interstadials are evident in
evidence. Hence, the deep-sea sequence provides an inde- stratigraphic records from areas that lay beyond the limits
pendent, continuous and unique climatostratigraphic of the Weichselian ice sheets.
template against which the often fragmented terrestrial Throughout northern Europe there is a broad measure
sequences can be compared (Figure 1.6). Correlations of agreement over the brief climatic oscillation that
Interglacial earth
Mid-Pleistocene
Revolution
17
G17
17• :17 G7 104
17 0717
>3 Ci •.!
3 77 si 17 17 03 17 101 17• '/.i .
r
5 9 T9 17 17 J 7 17
1 317
J
55 17 75 K17
: 17
A 5
7 13 17 17 17 17
21
7=i J5 17 •1345
17
51 S3 17 61
i-7 5!i
73 17 ii5
104
33
17 41 71
B O nw
.17
r
17
7?, 17 104 104
17
Z CIS
1 E3 70
104
,S
A 5 17
Olduvat 7i
17 Gauss
4 9 17
17
10 17 20 Jaramillo
Z 7
6 9
17
IS 2200 2400 2800 3000
Brunhes 15 MaLuyama 2000 2600
Age (ka)
0 E S I '-00 E00 800 1000 1200 1400 1600 1600
Age (ka)
Glacial earth
Figure 1.5 Climatic trends during the past 3 Ma reflected in a stacked (composite) oxygen isotope record (after Lisiecki & Raymo,
2005). The isotopic trace can be read as a proxy climate record with ‘peaks’ marking warmer (interglacial) intervals and ‘troughs’
colder (glacial) episodes (section 3.10). The astronomically driven millennial-scale climatic oscillations are present throughout, but
their frequency and amplitude changes over time, and especially after the Mid-Pleistocene Revolution (MPR: sections 1.7 and
7.3) at c. 800 ka when climatic extremes increased and Northern Hemisphere ice sheets reached their greatest extents.
MARINE ISOTOPE R E C O R D NORTHERN HEMISPHERE STAGES
QUATERNARY
Composite of cores
MOI stages
WOI stages
NORTHWEST EUROPEAN NORTH AMERICA
Age V19-30 OL>F"-S77 BRITISH ISLES
and ODP-848 EUROPE RUSSIA
LATE
(Ma)
5 fiOôcssn 4 HOLOCENE FLANDRIAN HOLOCENE HOLOCENE
Termination i Weichselian Devensian Valdaian Wisconsinan
5a
0.1 5e EEMIAN I P S W I C H IAN MIKULINIAN SANGAMONIAN
.Tetminfttion fS Wartrie/Drenthe Moscouian
Wolstonian
7s,
0.2 SCHONINGEN OOINTSOVIAN
Saalian
7e
s Dniepnan Pre-lllinoian A
NEOPLEISTOCENE
0,1
0.3 REINSDORF ROMMYAN
13 [3
Tgriningtion f\i Fuhne Pronyan
0.4 •i HOLSTEINIAN HOXNIAN LIKHVINIAN
12 Termination V Elsterian Anglian
MIDDLE
Okian Pre-lllinoian B
0.5 INTERGLACIAL IV
"romerian complex
14 Glacial C Pre-lllinoian G
•5
I N T E R G L A C I A L III MUCHKAPIAN
0.6
1-3 Cromerian D onian Pre-lllinoian D
Glacial B
0.7 I N T E R G L A C I A L II ILYNIAN
IB
Glacial A Pokrovian Pre-lllinoian E
'0
0.8 INTERGLACIAL I PETROPAVLOVIAN
??
0.9 24 Dorst
Pre-lllinoian F
Bavelian
£8 LEERDAM
1.0 23
33 Linge
Krinitsian/
3? BAVEL Krinitsa
1.1
EOPLEISTOCENE
34
1.2 3-3 Menapian Pre-lllinoian G

36
40 Beest onian
1.3 4/
44
1.4 13
WAALIAN
43 Tolucheevkian/
1.5 50 Tolucheevka
52
34
1.6
5'3
Pre-lllinoian H
Eburonian
EARLY
1.8 5S
CO
Pre-lllinoian I
02 S3
1.8 C5-6 PASTONIAN
34
35
1.9 70
Pre-Pastonian/
72 C4c Pre-lllinoian J
74 Baventian
2.0
78
Tiglian
BRAMMERTONIAN/ KHAPROVIAN/
7S C1-3
2.1 30 ANTIAN KHAPRY
32
2.2 34 B Thurnian
33
33
2.3 30
0/ A LUDHAMIAN VERKHODIAN/
2.4 94 VERKHODON
93
93 Praetiglian
2.5 M0 Pre-lllinoian K
•c? Pre-Ludhamian
2 6 104
ici
REUVERIAN C CENTRAL PLIO-I
1C6 W A L T O N IAN NOVORONEZH GENE
Figure 1.6 The MIS record based on a composite of deep-ocean cores (V19-30, ODP-677 and ODP-846) (left) and the Quaternary
stratigraphy of the Northern Hemisphere set against this record (right). The marine isotope signal shows the oxygen isotope stages
back to 2.6 Ma. In the correlation table, temperate (interglacial) stages are shown in upper case, while cold (glacial) stages are in
lower case. Complexes which include both temperate and cold stages are in italics (based on Cohen & Gibbard, 2011).
occurred towards the close of the last cold stage (termed of the Younger Dryas cold episode has been identified
the Devensian Lateglacial in Britain and the Weichselian in eastern North America, and also in Arctic Alaska (Elias,
Lateglacial in northern Europe), for this period can be more 2007b). An alternative stratigraphic scheme for the Late-
precisely dated than older parts of the sequence. However, glacial, based on the oxygen isotope signal in the GRIP
opinions differ over the extent to which the ‘Lateglacial’ can Greenland ice core, has two cold events, Greenland Stadial
be subdivided; in Britain, most scientists accept a twofold 2 (GS-2) and Greenland Stadial 1 (GS-1), the latter broadly
division into a Lateglacial (or Windermere) Interstadial correlating with the Younger Dryas; and an intervening
and a Loch Lomond Stadial, whereas a more complex warmer episode, Greenland Interstadial 1 (GI-1), which
sequence with two interstadial episodes, Bølling and equates with the Lateglacial Interstadial. Two short-lived
Allerød, separated by a brief cold episode (Older Dryas) and colder episodes (GI-1b and GI-1d) are apparent during
followed by the Younger Dryas Stadial has been recog- Greenland Interstadial 1 (Björck et al., 1998; Walker et al.,
nized in records from the European mainland (Figure 1.7). 1999). The Lateglacial record is considered further in
A climatic oscillation that appears to be the correlative section 7.5.5.
G r e e n l a n d ice c o r e
Indicative Age
Northwest
Age British Isles U
C age cal.
1 4
C
GRIP ss08c Europe
(yr) G R I P events yr B P yrBP
5 0 record
1 8
Holocene Holocene Flandrian
11,500 -10,050 -11,500
12,000 Greenland Younger Loch -10,255 -12,000

Stadial 1 Dryas Lomond
[GS-1J Stadial Stadial
12,500 -10,475 -12,500

Weichselian Lateglacial
Devensian Lateglacial
Gl-ia
13,000 Gl-ib -11,000 -13,000

Greenland Interstadial 1
All era d
Interstadial
13,500 GI-1C -11,550 -13,500
[GI-1]
Lateglacial
or
Windermere -12,025 -14,000
14,000 Gl-1d Older Dryas Stadial
Interstadial
Gl-1e
14,500 Belling -12,425 -14,500
Interstadial
15,000 -12,700 -15,000

Greenland
Stadial 2
[GS-2) Dimlington
15,500 Pleniglacial Stadial -13.100 -15,500
-42 - 4 0 -38 - 3 6 %
%«SMOW
Figure 1.7 The δ18O record from the GRIP Greenland ice core showing the Lateglacial event stratigraphy (left) and the strati-
graphic subdivision of the Lateglacial in northwest Europe and the British Isles. The isotopic record is based on the GRIP ss08c
ice-core chronology, and the colder stadial episodes are indicated by dark shading. The radiocarbon timescales (right) are shown
as ‘indicative’ (or ‘average’) radiocarbon (14C) ages and their ‘calibrated’ (cal.) equivalents (partly after Lowe et al., 2001).
Prior to the last interglacial, however, the Quaternary Milutin Milankovitch. The theory is based on the assump-
records are much more difficult to resolve and, although tion that surface temperatures of the earth would vary in
inter-regional correlations have been attempted (Figure response to regular and predictable changes in the earth’s
1.6), these are frequently speculative, and become increas- orbit and axis. Due to planetary gravitational influences, the
ingly so as the age of the depositional record increases. In shape of the earth’s orbit is known to change from almost
the European Alpine area, the sequence of glacial sedi- circular to elliptical and back again (Figure 1.8a), a process
ments and ‘interglacial’ soils now appears to be far more referred to as the eccentricity of the orbit. The timing varies
complicated than in the classical Alpine model, and hence between 95 and 136 ka, although this is generally referred
the standard nomenclature of Günz, Mindel and Riss to as the 100 ka eccentricity cycle. Over a longer time-
glacials can be used only in a very general sense. Similarly scale, however, a 413 ka eccentricity cycle is also apparent.
in North America, the terms ‘Kansan’ and ‘Nebraskan’ In addition, the tilt of the earth’s axis varies from 21°39′ to
glacial periods and ‘Yarmouthian’ and ‘Aftonian’ inter- 24°36′ and back over the space of c. 41 ka (Figure 1.8b).
glacials have been largely abandoned in favour of a series Because the angle of tilt is measured relative to an imaginary
of stages prior to the Illinoian Glacial that are designated line representing the plane of the ecliptic (the plane
simply by letter. In Britain and northern Europe, it is now described by the earth’s elliptical path around the sun), this
generally accepted that several of the established named phenomenon is known as the obliquity of the ecliptic.
stages must contain a number of separate episodes of cold The third variable arises because the gravitational pull
or temperate character (Figure 1.6). Hence, the Cromerian exerted by the sun and the moon causes the earth to wobble
Complex in the Netherlands is believed to encompass on its axis like a top (Figure 1.8c). The consequence of this
four warm (interglacial?) episodes, while in Britain at is that the seasons (or the equinoxes) seem to move around
least five interglacial episodes may be represented in the the sun in a regular fashion, hence the term precession of
‘Cromerian’ of the Middle Pleistocene (Preece & Parfitt, the equinoxes or precession of the solstices. In effect this
2012). In northern Continental Europe, two or maybe means that the season during which the earth is nearest to
three glacial events may have occurred during the classical the sun (perihelion) varies. At present, the Northern
Saalian Glacial (Ehlers et al., 2011b). A further problem Hemisphere winter occurs in perihelion (Figure 1.8c-i)
concerns gaps in the terrestrial stratigraphic records. while the summer occurs at the furthest point on the orbit
Comparison between the Dutch and British Early and (aphelion). In c. 10.5 ka time, the position will be reversed
Middle Pleistocene sequences, for example, suggests that as (Figure 1.8c-iii), while c. 21 ka hence the cycle will be
much as a million years of sedimentary history may be complete. In fact, it now appears that there are two separate
missing from the stratigraphic record in southeastern interlocked cycles, a major one averaging around 23 ka and
England between the Pastonian and Cromerian (Gibbard a minor one at c. 19 ka.
et al., 1991), with other major hiatuses elsewhere (Figure These variables, in combination, exert a profound effect
1.6). Overall, therefore, the individual stages and suggested on global temperatures. The total amount of radiation
correlations between those stages shown in Figure 1.6 must received is determined largely by the eccentricity of the
be regarded as no more than a provisional approximation earth’s orbit, while the other astronomical variables affect
of the Quaternary climatostratigraphic sequence in Europe the way in which that heat energy is distributed at different
and North America. latitudes. In general it seems that solar radiation receipt in
the low- and middle-latitude regions is governed mainly by
precession and eccentricity variations, while in higher
1.7 THE CAUSES OF CLIMATIC latitudes the obliquity cycle predominates. Patterns of
change through time can be calculated from astronomical
CHANGE data (Figure 1.9a), and Milankovitch was therefore able to
It is now apparent that the climatic fluctuations of the obtain estimates for radiation inputs at different latitudes,
past 2.5 Ma or so have followed a series of distinctive and from these to reconstruct long-term temperature
patterns, and hence contemporary explanations of long- changes.
term climatic change have tended to focus on the factors The theory was first published in 1924 and initially
that have given rise to both the regularity and frequency of found favour with many European geologists, for the
climatic fluctuations. The hypothesis that has attracted sequence of warm and cold stages predicted by the radi-
the greatest attention is undoubtedly the Astronomical ation curves appeared to match the record of glacials and
Theory, developed by James Croll a little over 100 years ago interglacials in the classical Alpine region of Penck and
and subsequently elaborated by the Serbian geophysicist Brückner. Increasingly, however, it became apparent
a) Almost elliptical Almost circular

a) 0.05 Eccentricity
North North
0.02
0.0
Obliquity
24.5
Period icily 9$,00Qyr

23.3
bit bi:
North Pole ôrth Pole
22.0
Equator Equator Precession
-0.07
Radiation
Radiation
-0.02
0.04
Period icily 42.DO.0yr ETP

2.7
bit {i;. Mr,-.:-
Periodicity c21.OQ.0yr 0.0
Equator Winter Summer
-2.7
{ii} In c 5,£50yr Summer
0.0 100 200 300 400 500 600 700 800
A g e (ka)
b)
2.2
Winter
[ i ; In c. 10 500/yr
6«0
Equator Wiriter 0.0
2.2
0.0 100 200 300 400 500 600 700 800
Figure 1.8 The components of the Astronomical Theory of A g e (ka)
climate change: a) eccentricity of the orbit; b) obliquity of the
3
0 {%»)
ecliptic; c) precession of the equinoxes.

4
5
1 8
6
0
that the timing and frequency of glacial episodes during the
0.0 100 200 300 400 500 600 700 800
Late Quaternary did not seem to accord with the pattern
A g e (ka)
of climatic changes predicted by the astronomical variables.
This was thrown into sharp relief in the 1940s and 1950s Figure 1.9 a) Variations in eccentricity, obliquity and the
with the development of radiocarbon dating which precessional index over the past 800 ka. The three time series
provided, for the first time, an independent chronology have been normalized and added to form the composite
eccentricity-tilt-precession curve (ETP). The scale for obliquity
for the Late Quaternary glacial sequence. By the mid- is in degrees and for the ETP is in standard deviation units. b)
1950s, the Milankovitch hypothesis as an explanation of Normalized and smoothed variations in the oxygen isotope
climatic change had been almost universally rejected. signal (δ18O) in five deep-sea cores. Note the similarity between
Writing in 1957, for example, Richard Foster Flint discussed this record and the ETP curve above (after Imbrie et al., 1984).
‘geometric variation in elements of the earth’s orbit’ as a
factor in global climate change, but concluded that ‘the
geometric scheme of distribution of insolation heating In the late 1960s and early 1970s, however, work initially
must be considered inadequate in itself to explain the on sea-level changes and subsequently on deep-ocean
Pleistocene climatic changes’. He accepted that ‘geometric sediments reawakened interest in the Milankovitch hypo-
factors are present’ and ‘must have influenced climate to thesis (Imbrie & Imbrie, 1979). The first clue came from
some degree’, but ‘opinions differ as to how closely the the dating of coral reefs around the island of Barbados,
insolation curve matches the Pleistocene record. Presum- which showed high sea-level stands (reflecting warmer
ably the geometric factors are superposed on some other episodes) at around 82 ka, 105 ka and 125 ka, which
cause of climatic fluctuation’ (Flint, 1957, p. 509). coincide closely with the phasing of the precessional
cycle (Mesolella et al., 1969). More significant, however,
Temperature
were oxygen isotope variations found in marine micro- 2
fossils which provided a long-term proxy record of environ- 0

mental and climatic change (section 3.10; Figure 1.9b). 2
xnu isnrj
-2
Spectral analysis7 of ocean-core sequences revealed evidence 0
of cycles of 100 ka, 43 ka, 24 ka and 19 ka in the isotopic 2
I -2
signal, with the longest cycle driving the glacial/interglacial
co 2
a
oscillations of the past 700 ka or so while the others, in 2
-2
combination, modulate or amplify the effects of longer-
'HO
0
term changes (Hays et al., 1976). These data provided the
-2
first unequivocal evidence of the Milankovitch cycles in the
0 200 400 600 800
recent geological record, and were an impressive demon-
A g e (ka)
stration of the potential role of the astronomical variables
in determining patterns of long-term climatic change – Figure 1.10 Variations in temperature, dust flux, CO2 and CH4
hence the title of the seminal paper of Hays et al. (1976), from the EPICA Dome C ice-core record, Antarctica, over the
‘Variations in the earth’s orbit: pacemaker of the Ice Ages’. last 800 ka (based on Masson-Delmotte et al., 2010).
Subsequently, Milankovitch cycles have been detected in a
wide range of proxy records including ice cores (Petit et al., change (or changes) internal to the global climate system
1999), lake sediments (Trauth et al., 2001), loess sequences must have been responsible (see section 7.3).
(Sun et al., 2006b) and pollen records (Torres et al., 2013). The major elements in the climatic equation that serve
Collectively these data confirm the hypothesis that changes to modulate or amplify the effects of the astronomical
in the earth’s orbit and axis, what is often referred to as variables appear to be changes in the disposition of the
orbital forcing, are the primary driving mechanisms in continental land masses, tectonic activity, feedback mech-
Quaternary climatic change (Imbrie et al., 1993). anisms caused by oceanic circulation and changes in the
Although the Astronomical Theory offers a basis for extent of continental ice cover (Denton, 2000). A further
understanding the sequence of major Quaternary climatic influential factor may have been variations in the con-
oscillations, it is now apparent that factors other than stituents of the atmosphere including, for example, aeolian
orbital forcing have influenced the course of global climatic dust (Rea et al., 1998) and particularly atmospheric trace
change. Furthermore, a number of outstanding prob- gases such as CO2 and methane (CH4) (Ruddiman 2003a,
lems relating to the Astronomical Theory still need to be 2006; Figure 1.10). Although the long-term cooling trend
resolved. For example, although the onset of Northern during the late Pliocene has been attributed, in part, to the
Hemisphere glaciation, as reflected in the build-up of influence of obliquity minima between 3.2 and 2.5 Ma
large ice sheets in North America and Europe, dates from (Maslin et al., 1998), other factors may have contributed
around 2.7 Ma (Haug et al., 2005), proxy data from deep- to the cooling and to the onset of continental glaciation.
ocean cores suggest that global climate had cooled, albeit A necessary precursor was the closing of the Isthmus of
in an oscillatory manner, from around 3.6 Ma (Mudelsee Panama which was completed around 2.75 Ma (Schneider
& Raymo, 2005). In addition, the climatic cycles of the & Schmittner, 2006), and which led to an intensifica-
Quaternary have not been constant, but have shifted from tion of North Atlantic circulation and enhanced moisture
a periodicity of around 41 ka prior to 900–800 ka to a flux into high-latitude regions (Bartoli et al., 2005); in
prevailing rhythm of c. 100 ka over the course of the last the western Pacific, the restriction of the Indonesian Seaway
c. 800 ka, a phenomenon often referred to as the Mid- at about the same time (4–3 Ma) could have reduced heat
Pleistocene Revolution (Maslin & Ridgwell, 2005) or the transport to the high northern latitudes (Cane & Molnar,
Middle Pleistocene Transition (MPT). This, in turn, was 2001). Tectonic activity in Tibet may also have contributed
accompanied by an apparent intensification of glaciation, to the long-term cooling trend, as land uplift could alter the
with the growth of Northern Hemisphere ice sheets to wave structure in the airstreams of the upper atmosphere,
volumes very much larger than those attained over the the effects of which would have been to cool the Eurasian
course of the previous 1.7–1.6 Ma. Neither the onset of and American land masses and hence increase their sensi-
Northern Hemisphere glaciation at around 2.7 Ma, nor the tivity to orbitally driven insolation changes (Ruddiman &
shift in climatic phasing around the MPT, can be accounted Kutzbach, 1990). In addition, a more intense monsoonal
for solely by Milankovitch forcing, which suggests that some circulation and increased rainfall following uplift may
have led to accelerated rates of chemical weathering 1.8 THE SCOPE OF THIS BOOK
of silicate minerals. The latter involves CO2 from the
atmosphere, and the products are ultimately deposited on The aim of this book is to describe and evaluate the methods
the ocean floor where they are removed from the global and approaches that are currently employed in the
geochemical carbon cycle. As a consequence, the atmos- reconstruction of Quaternary environments. The work
phere becomes depleted in CO2, a process which may have does not, however, claim to be exhaustive. Indeed, in view
contributed to the long-term Plio-Pleistocene global of the wide range of disciplines involved in Quaternary
cooling trend (Raymo & Ruddiman, 1992). research and, particularly, the ‘information explosion’ that
In addition to these long-term variations in climate has occurred over the past two decades, a comprehen-
during the Quaternary, high-resolution proxy records sive treatment would run far beyond the scope of a single
provide evidence of rapid climatic variations, frequently of volume. Some aspects are, therefore, considered only
large amplitude, which are superimposed on the orbitally briefly, while others (which some will no doubt believe to
driven cycles. These short-lived ‘sub-Milankovitch’ events be important) are omitted altogether. Nevertheless, an
occur over timescales varying from centuries to millennia attempt has been made to present a balanced view of the
and have been found, inter alia, in ice-core records from various methods employed in, and sources of evidence that
Greenland (North Greenland Ice Core Project Members, form the basis for, Quaternary environmental reconstruc-
2004), in Chinese cave speleothems (Wang et al., 2008), tions. Some temporal bias is inevitable, as far more is
in vegetational records from the last glacial stage from Africa known about the later Quaternary than about the earlier
and South America (Hessler et al., 2010) and in marine parts of the period, and therefore the majority of examples
records from the northwest Pacific (Khim et al., 2012). are drawn from the last interglacial and last glacial stages.
Energy transfer in the world’s oceans, driven by salt-density The methods, approaches and principles are, however,
equally applicable to the analysis of Early and Middle
variations (thermohaline circulation), along with chemical
Quaternary environments. In addition, although there is an
changes resulting from biological activity, appear to be
emphasis on evidence from the Northern Hemisphere
major causal factors underlying these events. For example,
mid-latitude regions, particularly from Europe and North
the abrupt climatic changes that have occurred during the
America, it is hoped that readers in other parts of the
Late Quaternary in many low-latitude regions may be
world will find material here that is of interest to them as
related to changes in the nature and rate of North Atlantic
well.
ocean circulation, with fluctuations in sea-surface tempera-
The book falls naturally into three parts. In Chapters
tures (SSTs) influencing the pattern and timing of tropical
2, 3 and 4, the geomorphological, lithological and bio-
monsoons, in turn leading to marked spatial and temporal
logical evidence that forms the basis for environmental
variations in precipitation over tropical Africa. Other causal reconstruction is outlined. Although these are useful
factors of decadal to millennial-scale climatic changes general categories within which to describe particular
include short-term fluctuations in radiatively active atmos- techniques and approaches, they are, to some extent,
pheric trace gases (greenhouse gases), most notably artificial and there are considerable overlaps between them.
CO2, CH4, N2O; variations in solar output (as reflected, for Hence, in Chapter 2, where the emphasis is on geomor-
example, in sunspot cycles) and in the intensity of the solar phology (i.e. surface architecture), certain aspects of the
wind (the stream of protons and electrons emitted by the stratigraphy of river terraces and raised shoreline sequences
sun); and volcanic eruptions during which both particulate need to be considered also, while in Chapter 3, where
matter and sulphur volatiles were injected into the sedimentological evidence is being discussed, reference is
atmosphere, and which might lead to short-term tempera- frequently made to landform evidence as, for example, in
ture reductions on the earth’s surface through the screen- the analysis of sand dunes formed in loess and coversand
ing out of incoming radiation. All of these, singly or in deposits. In all three chapters, field and laboratory tech-
combination, could, through a complex series of feed- niques are introduced in order to give an indication of the
back loops, serve to modulate or amplify the effects of procedures that are involved in generating the basic data.
climate change resulting from orbital forcing or from Chapters 5 and 6 make up the second part of the book.
oceanographical changes. We return to the factors driving The various dating methods that are currently employed
Quaternary climate change, on both long- and short-term in Quaternary science are described and evaluated in
timescales, in Chapter 7, while further discussion can be Chapter 5, while the principles of stratigraphy and correla-
found in Bradley (1999), Bell & Walker (2005) and Cronin tion which enable the researcher to construct meaningful
(2009). spatial and temporal sequences from often fragmentary
evidence are outlined in Chapter 6. The final part, Chapter grains, isotopic records, glacial sediments, tree rings or animal
7, illustrates how insights into the timing, rate and impacts bones.
of past climate change can be gained by synthesizing 3 The International Union of Geological Sciences (IUGS) is the
evidence obtained using the various methods and body that oversees and arbitrates on all matters relating to
geological nomenclature. It operates through a series of
approaches outlined in Chapters 1–6. The aim here is to
Commissions (the International Commission on Stratigraphy
highlight recent advances in understanding of the processes [ICS], for example, deals with issues relating to stratigraphy and
that drive climatic change over different timescales correlation), and is responsible for the production of the
(Milankovitch to decadal), and of their environmental international geological timescale and associated stratigraphic
consequences. charts (e.g. Gradstein et al., 2012). Within the ICS are a number
of subcommissions, each responsible for a particular interval of
geological time, such as the Subcommission on Quaternary
Stratigraphy (SQS).
NOTES 4 INQUA is the International Union for Quaternary Research,
1 A number of conventions are currently employed for established in 1928 to encourage and facilitate the research of
abbreviating geological ages. In this book we use the shorthand Quaternary scientists in all disciplines (http://www.inqua.org/).
form ‘ka’ to refer to ages in thousands of years (e.g. 3 ka = 3,000 It holds a major international congress every four years.
years ago). Similarly, ‘Ma’ refers to ages in millions of years. 5 Isotopes are atoms of an element that are chemically similar,
When radiocarbon dating (14C dating) is used to determine age, but have different atomic weights (section 5.3.1). Oxygen, for
however, the dates are expressed in years BP (before present), example, consists principally of two isotopes, the ‘heavier’ 18O
the reference or baseline year being AD 1950 (e.g. 3 ka BP). As isotope and the ‘lighter’ 16O. 18O/16O ratios (termed δ18O) in
we shall see in Chapter 5, radiocarbon dates do not equate the atmosphere, seas, groundwater and ice are often controlled
precisely to calendar years, but can be ‘calibrated’ to these, in by oceanographic or climatic conditions.
which case they are expressed as ‘cal. BP’ (section 5.3.2.6). In 6 Sapropels are dark-coloured marine sediments rich in organic
ice-core chronologies, the reference year is AD 2000 (section matter that develop during periods of reduced oxygen avail-
5.4.3.3). Although other age estimates (based on both radio- ability in bottom waters. For further details, see sections 5.5.4
metric and incremental methods: Chapter 5) are often expressed and 6.3.3.3.
in years BP, the baseline year is the actual year in which the 7 Spectral analysis is a statistical technique which aims to identify
measurement was made. cycles in time-series data. Cycles may be characterized in two
2 The term ‘proxy’ or ‘proxy record’ is used to refer to any line ways: the period of a cycle is the length of time between
of evidence that provides an indirect measure of former climates consecutive repeats, while the frequency is the number of cycles
or environments. It can include materials as diverse as pollen (i.e. repeats) that occur per unit of time (Green, 1995).
2
CHAPTER TWO
Geomorphological
evidence
2.1 INTRODUCTION dence (Chapter 3) and that, wherever possible, the two
should be used in conjunction in the reconstruction of
The marked oscillations in global climate that occurred Quaternary environments.
during the Quaternary led to major changes in modes
and rates of operation of geomorphological processes.
Undoubtedly the most spectacular manifestations of 2.2 METHODS
climatic change were the great continental ice sheets
whose passage resulted in widespread modification of the 2.2.1 Field methods
land surface of mid- and high-latitude regions. The growth
2.2.1.1 Field mapping
and decay of the ice sheets were accompanied by the
expansion and contraction of areas affected by periglacial The production of a map illustrating the type and dis-
activity, there were major changes in the regimes of many tribution of the principal landforms is often the first
river systems, and the nature and effectiveness of geomor- stage in the investigation of the Quaternary history of an
phological processes were strongly influenced by changes area, and a wide range of automated survey, mapping and
in the distribution and type of vegetation cover. In low- imaging tools, as well as user-friendly software for their
latitude regions, phases of aridity alternated with periods manipulation, is now available (Pike, 2000). Because these
of wetter climatic conditions leading to migration of enable large tracts of the earth’s surface to be mapped and
desert, savannah and rainforest margins, lake levels rose and analysed very rapidly, they are often now preferred to the
fell, and fluvial and colluvial processes varied both spatially more time-consuming process of ground survey using
and temporally. On the global scale, sea level during the manually operated instruments. The latter have not yet been
glacial phases was frequently more than 100 m lower entirely superseded, however, because they can resolve
but rose to positions above present levels during some of features too fine in scale to be detected by remote methods
the interglacial stages. (e.g. Hubbard & Glasser, 2005), and they also provide data
Landforms that developed under a previous climatic for the ‘ground-truthing’ of remotely captured images
regime have often survived, albeit sometimes in a much (Lane & Chandler, 2003).
modified form, as ‘relict’ or ‘fossil’ features. Analysis of Morphological mapping, using either remote or manual
these can often provide valuable information about the techniques, is the recording and depiction of individual
nature of the climatic regime under which they evolved, slope elements in the landscape, and the nature of the
as well as about other environmental parameters such as junctions between them (Griffiths, 2002). A simple hand-
glacial and fluvial processes, slope stability, groundwater held instrument such as an Abney Level or a clinometer can
movement and tectonic activity. However, the use of geo- be used for slope measurement. Typical morphological
morphological evidence for this purpose requires a proper maps are produced at scales of 1:10,000 or larger, on which
understanding of the relationships between geomorpho- even the most subtle changes in the shape of the land
logical processes and landforms. Moreover, it must be can be recorded (Griffiths et al., 1995). This approach
emphasized that there is frequently a close relationship has been widely employed in land survey, but because it
between geomorphological evidence and lithological evi- is not specifically concerned with landscape evolution,
20 GEOMORPHOLOGICAL EVIDENCE
it has found less favour with Quaternary scientists. more detailed investigation is being conducted, the altitudes
Geomorphological mapping, on the other hand, is one of and surface gradients of landforms must be obtained by
the most important techniques in Quaternary research, instrumental measurements in the field.
for the maps produced contain information not only on The comparison of altitudes of landforms or strati-
morphology but also on the genesis and, in some cases, graphic horizons, especially from widely separated localities,
on the age of the landforms. This type of mapping can be requires a common geodetic datum,1 a plane or point of
carried out at different scales ranging from very detailed precisely fixed altitude to which all subsequent measure-
maps of small areas (typically 1:10,000) to maps at the ments can be referred (e.g. North American Datum
national scale (e.g. IGS Quaternary Map of the British NAD83; the European Datum of 1950, with its origin at
Isles 1977, scale 1:625,000; see also the regional compila- Potsdam, Germany; and the UK’s Ordnance Datum (OD),
tions in Ehlers & Gibbard, 2003, and Ehlers et al., 2011a). measured at Newlyn, Cornwall). A frequently employed
Geomorphological mapping is essentially interpretative regional datum has been sea level, but as this varies both
and therefore requires both an appreciation of the com- spatially and temporally, altitudinal data are more reliable
plexity of landform assemblages and a detailed know- when related to precisely surveyed surface altitudes of fixed
ledge of their genesis. It also needs an eye for detail, some geodetic coordinates; in the UK, these are known as
training in field mapping and surveying techniques, and Ordnance Survey Bench Marks and are marked on the
a knowledge of aerial photographs and satellite images, ground surface by a symbol cast in metal.
as the mapping of landforms can be greatly aided by In many parts of the world, however, only low-quality
comparison with high-quality optical images of the field maps and limited geodetic information are available,
area. Geomorphological mapping, particularly at scales but the surface altitudes and spatial relationships of land-
of 1:10,000 or greater, has been most effectively employed forms in these regions can be estimated using a range of
in the analysis of glacial landscapes, including those result- instruments. These include aneroid barometers, hand-
ing from the passage of the last ice sheets and, in particular, held (e.g. Abney) levels, surveyor’s levels, theodolites and
from more recent phases of glacier activity (Figure 2.1). electronic distance measures (EDMs), the more sophisti-
cated versions of which are commonly termed ‘total
stations’ (Bannister et al., 1998). Aneroid barometers are
2.2.1.2 Instrumental levelling
affected by variations in barometric pressure, while the
In reconstructing the Quaternary history of an area, it is other instruments, although providing positional data of
often essential to determine the precise altitude of, and increasing degrees of precision, also have operational and
differences in altitude between, particular landforms and technical limitations (Ghilani & Wolf, 2008). For example,
landform assemblages. The same applies equally to litho- hand-held levels are useful for rapid surveys, but tend to
logical units (Chapter 3). Altitudinal data can aid in the produce variable results due principally to operator errors.
interpretation of landforms, and may also enable land- In most geomorphological fieldwork in the mid-latitude
forms of different age to be identified. For example, only regions, therefore, particularly where precise altitudinal data
fragments of former river terraces may be preserved in a are required, surveyor’s levels, theodolites and EDMs are
particular area, and it may be impossible to identify and most frequently used and traverses opened and closed at
correlate fragments of similar age, and to establish a national survey benchmarks in order to provide an estimate
chronology of terrace development on the basis of field of accuracy. However, in those parts of the world where
mapping alone. By obtaining precise altitudinal meas- no such benchmarks are available, such as Arctic Canada,
urements on each terrace fragment, however, formerly Greenland, and parts of Africa and Antarctica, the local
continuous features can be reconstructed, gradients can marine tidal level or some prominent landmark may be
be measured and altitudinal relationships between indi- selected to provide an arbitrary temporary datum. While
vidual terraces can be established (section 2.6). This, in turn, these may provide a reasonable basis for determining the
may enable the relative order of age of the features within relative heights of features within the survey area, the
a terrace sequence to be deduced. Similar principles can ‘absolute’ geodetic position of the temporary datum cannot
be applied in the investigation of abandoned shoreline be established, and this limits the potential for altitudinal
remnants (section 2.5). Where only a general impression correlation with features in other areas.
of altitude is required, and where the mapping is being An important innovation in geomorphological studies,
carried out at a relatively small scale, it may be sufficient especially in remote locations, is the Global Positioning
to obtain the altitudinal data from spot heights and System (GPS). This is a method of triangulation based on
contours on the relevant base maps. However, where a automated computation of distances between points on the
GEOMORPHOLOGICAL EVIDENCE 21
Mountain-top detritus
Terraces
ETerraces
3
Drift limit or trimline
Rock-slope failure
Clftls
Meltwater channel
SoM uctiori lobes
Ice-moulded bedrocfc, ZQQO Contours a\
showing direction 50 metre intervals
Frost shattered bedrock ZQQO Mountain summits
in metres
Scree
0 ZQQOm
Snore line
Figure 2.1 Geomorphological map of the West Drumochter Hills in the Scottish Highlands (from Benn & Ballantyne, 2005). The
nature and distribution of glacial landforms enable the areal extent and thickness of the former ice mass to be determined (copyright
© 2005 John Wiley & Sons, Ltd).
earth’s surface and a number of earth-orbiting satellites. It features subsequently obscured by afforestation, the
was developed initially for military purposes by the USA in products of catastrophic events such as landslips or
the 1970s, and civilian access to the network was granted tsunamis, or urbanization; and (d) allow the monitoring
for the first time in the early 1990s. A range of affordable of changing landscapes and landform assemblages, or
hand-held receivers is now available which can achieve changes in other features, such as glacier and ice sheet
remarkable precision. At the lower end of the price range, margins (Khan et al., 2012).
‘recreational’ GPS receivers enable the position and altitude Disadvantages of aerial photographs include distortions
of the ground at any point to be determined to within a due to camera tilt or variations in camera altitude, loss of
few metres, depending on the receiver model, the nature detail due to cloud cover or shadow effects, poor tonal
of the local terrain, the number of satellites that the receiver contrasts so that, for example, unconsolidated surface sedi-
is able to track and the positioning correction factors that ments sometimes cannot be distinguished from bedrock,
can be applied (Spencer et al., 2003a). More sophisticated and difficulty in the detection of small-scale landforms.
‘differential receivers’2 (Trimble Navigation, 1989) are Field mapping, therefore, remains essential, and even where
able to pinpoint locations to within a few centimetres. mapping is based on large-scale, good-quality aerial
photographs, the results must be viewed as no more than
a provisional map of the Quaternary geomorphology of a
2.2.2 Remote sensing region until the interpretations can be checked in the field.
‘Remote sensing’ refers to the acquisition of images or scans For areas where good topographic maps are unavailable,
of the earth’s surface and subsurface by instruments that aerial photographs or satellite images provide the only
can detect elements of the electromagnetic spectrum realistic basis for the mapping of landforms, and enlarge-
reflected or emitted by different surface materials. A range ments can be made specifically for this purpose.
of devices can be mounted on aircraft, satellites and boats
to capture signals of selected wavelength in order to ‘sense’
2.2.2.2 Satellite imagery
and record spatial and temporal variations in the earth’s
surface albedo, roughness, lithology, sediment architecture Satellite imagery offers great advantages over aerial
or other physical attributes. Remote sensing includes con- photography as distortions are minimized, the process
ventional photography using visible and non-visible (e.g. is much more rapid, and repetitive images of large parts
infra-red) light spectra, multispectral (multi-wavelength) of the earth’s surface can be obtained. In terms of image
scanning systems, radar sensing, sonar signals (echo- generation, conventional photographs are produced by
sounding) and, more recently, laser technology (Lillesand the simultaneous recording on film of all features viewed
et al., 2004; Jensen, 2007). through the lens of the camera. Satellites, on the other hand,
carry a wide range of sensors and filters that receive and
process images in various light-wave bands, and hence are
2.2.2.1 Aerial photography
considerably more versatile.
Since the First World War, aerial photographic recon- The development of non-military satellite earth moni-
naissance has increased both in frequency of use and in toring began in the late 1960s when the National Aero-
degree of sophistication. Good-quality aerial photographs nautics and Space Administration (NASA), with the
are now available even for some of the most inaccessible cooperation of the US Department of the Interior, initiated
parts of the world, allowing at least preliminary maps to be a programme to place in orbit a series of Earth Resource
made of landforms and landform assemblages (Poole et al., Technology Satellites (ERTS), the first of which, ERTS-1,
2002). A system of grid corrections using ground control was launched in 1972. Just before the launch of the second
points can be used to transfer details from photographs to ERTS satellite in January 1975, it was renamed Landsat, and
maps where scales differ, or where the photographs contain all subsequent satellites in the series have carried the Landsat
serious distortions (Baily et al., 2002). Aerial photographs designation. In January 1983, NASA transferred Landsat to
are especially useful in the mapping of landforms in that the National Oceanic and Atmospheric Administration
they (a) direct attention to areas where landforms are most (NOAA), but in October 1985, the whole operation was
evident or abundant, thereby avoiding wasted ground commercialized. Space Imaging, Inc. of Colorado, USA,
reconnaissance; (b) reveal larger-scale landform patterns obtained exclusive sales rights to all Landsat images,
that may go undetected in ground mapping, such as the although the U.S. Geological Survey (USGS) retained
shorelines of lakes in semi-arid regions or streamlined primary responsibility as the government’s long-term
patterning in glacial bedforms; (c) record morphological archive of Landsat data. In 2001, Space Imaging relin-
quished their commercial rights and Landsat data are now generates precipitation, evaporation and water cycle data;
jointly owned by the USGS and NASA (http://landsat. and Aura (EOS CH-1: 2004) studies ozone, air quality and
usgs.gov/). Seven Landsats have been launched since 1972, climatic variables.
current data being supplied mostly by Landsat-5 (1984) and Several satellites carry instruments that focus on
Landsat-7 (1999). particular earth surface or atmospheric properties. These
All seven Landsats have been equipped with two types include the TOPEX/Poseidon satellite (1995), a French–
of sensor: a set of return beam vidicon (RBV) cameras USA collaboration for monitoring the oceans, and its
and a multispectral scanning system (MSS). While RBVs successor Jason-1 (2001) which has provided measure-
do not contain film, images are received and stored on a ments of the surface altitude of the world’s oceans to an
photosensitive surface in each camera. This is then scanned accuracy of 3.5 cm; GEOSAT (Geodetic Satellite: 1985),
by an internal electron beam to produce a video signal, a US Navy observation satellite which can determine sea-
which is similar to that in a conventional television camera surface altitude with a precision of c. 5 cm; and NASA’s
(Lillesand et al., 2004). In MSS systems, a scanner pro- ICESat (Ice, Cloud, and Land Elevation Satellite: 2003),
duces a set of corresponding digital images of terrain in which measures ice sheet, cloud and aerosol properties, as
four different wavelength bands of the electromagnetic well as surface topography and vegetation.
spectrum. Digital images comprise a grid of cells (picture The potential of satellite imagery in Quaternary research
elements or pixels), each of which is assigned a value that is clearly considerable, not only for detecting and mapping
corresponds to the intensity of electromagnetic radiation ancient landforms, but also for monitoring modern earth
measured by the sensor from a portion of the terrain in the surface processes. Satellite images have been employed to
sensor’s instantaneous field of view (IFOV). The satellites map, inter alia, desert landforms (Al-juaidi et al., 2003),
orbit at a height of c. 700 km above the earth’s surface, and glacial moraines and streamlined glacial bedforms (Stokes
the MSS sequentially scans sectors of the surface measuring et al., 2005), ancient lake shorelines (Leblanc et al., 2006)
185 km by 170 km, recording radiance from an IFOV of and a range of periglacial landforms (Walsh et al., 2003).
approximately 80 m by 80 m. A second sensor, the thematic Examples of modern geomorphic processes that can be
mapper (TM), was added to Landsats launched after 1982. observed using satellite imagery include the evolution of
This records electromagnetic reflectance of sunlight from alluvial channels (Gupta et al., 2002), mountain mega-fan
the earth’s surface in seven spectral bands ranging from the development (Leier et al., 2005), glacier outburst floods
visible to infra-red, and at a higher spatial resolution (IFOV (Worni et al., 2012) and formation of thermokarst
of 30 m by 30 m). Landsat-7 carries an enhanced thematic (permafrost) features (Grosse et al., 2005). Satellite imagery
mapper that provides images with a spatial resolution of has also been used to measure the changing positions of the
15 m, greatly improving the definition of scanned surface margins of glaciers and ice caps (Berthier et al., 2010), and
features. also to monitor variations in sea-ice cover (Laxon et al.,
A number of other satellites specifically designed for 2013).
terrain monitoring have been placed in orbit during the
past thirty years. These include the French SPOT (Système
2.2.2.3 Radar
Pour l’Observation de la Terre) constellation of five satel-
lites, the first of which was launched in 1986 and the most Radio detection and ranging (radar) is based on the
recent (SPOT-5) in 2002. SPOT-5 carries high-resolution emission of pulsed signals from a transmitter, usually in the
visible (HRV) multi-spectral linear array (MLA) sensors microwave and higher radio frequencies, and the recording
which sequentially scan surface areas of 60 km by 60 km, of the ‘echoes’ of these signals as they bounce back from
from which stereo-pair images can be generated enabling the ground surface. The returning ‘back-scattered’ signals
accurate topographic mapping at a resolution of 10 m. are affected by ground surface roughness, by the orientation
Three satellites have been launched by the European of upstanding features, and by the density and electrical
Space Agency (ESA), the latest of which, Envisat (2002), properties of ground materials. Back-scattered radiation is
carries an array of nine instruments which capture a wide recorded in pixels and, as a general rule, the rougher the
range of information from land, water and ice surfaces, as terrain, the more back-scattered energy is returned, and the
well as from the atmosphere. Perhaps the most advanced brighter the resulting images. In dry sediments or cold ice,
earth observation programme, however, is NASA’s Earth boundaries between stratigraphic units or ice layers can
Observing System (EOS), which comprises three main often be detected.
artificial satellites: Terra (EOS AM-1: 1999) monitors Airborne radar equipment (‘echo sounders’) have been
environment and climate; Aqua (EOS PM-1: 2002), developed that automatically transform received signals
into images (imaging radar), usually referred to as side- (Piro & Campana, 2012), but it also has applications in
looking airborne radar (SLAR) because the radar antenna sedimentology (Neal, 2004), in investigations of periglacial
fixed below the aircraft is pointed to the side. As in satellite landforms, such as pingos (Ross et al., 2005) and rock
scanners, the pulsed signals scan the terrain and the received glaciers (Degenhardt, 2009), and in the measurement of soil
signals are subsequently converted into electrical impulses and debris mantle thickness (Sucre et al., 2011).
that are digitized or transformed into a photographic
image. In practice, however, SLAR systems are limited by
2.2.2.4 Sonar and seismic sensing
resolution problems to relatively short-range and low-
altitude operations (Lillesand et al., 2004). They have now Several techniques have been developed that are based on
been largely replaced by more sophisticated systems known the gravitational, magnetic or electrical properties of the
as synthetic aperture radar (SAR), which combines the earth. Transmission of acoustic or sonic waves through
echo signals through the radar instrument’s aperture as it the ground is affected by the density and other charac-
moves along the flight track, thereby significantly enhancing teristics of different materials, and these form the basis
signal resolution. SAR systems have been mounted on for seismic surveys that have been extensively used in
aircraft and on many earth-observing satellites, for example geophysical exploration. The sonic or seismic waves are
on Seasat and the ERS satellites (see above). A new phase created by controlled, minor explosions or induced
in SAR began in 2007 with the launch of the TerraSAR-X vibrations at the ground surface, the returning reflections
satellite, a joint project between the German Aerospace being captured by specialized receivers (‘geophones’ on
Centre and EADS Astrium, an aerospace subsidiary of the land, ‘hydrophones’ in water). Various devices are em-
European Aeronautic Defence and Space Company ployed to emit short, compressional seismic waves, one
(EADS). This earth observation satellite uses an X-band of the most common being a spark chamber, or ‘Sparker’.
SAR (the X band is a part of the microwave sector of the In studies of submarine geomorphology, side-scan sonar
electromagnetic spectrum) to provide very high-quality rapidly surveys and images large sectors (swathes) of the sea
topographic images for both commercial and scientific floor to establish bathymetry (swathe bathymetry) and
applications. bottom sediment architecture (acoustic profiling). Side-
As with other remote sensing systems, the great scan is a device that transmits acoustic signals in a wide arc
advantages of radar imaging are that micro- and radiowaves or fan towards the seabed, and the apparatus can be towed
can penetrate thick clouds and are unaffected by adverse behind a boat or attached to a ship’s hull. The vessel surveys
weather conditions, while the reflections received are fre- the sea floor along parallel tracks, and the continuous
quently more sensitive to ground surface textural variations sonic reflections are integrated to provide 2-D and 3-D
than is visible light. Radar has proved to be a useful and bathymetric images (Figure 2.2). Definition depends upon
versatile technique for geomorphological mapping and the sonar frequencies used, with higher frequencies
terrain evaluation, for example in the analysis of mass delivering better resolution, but possessing a more limited
movement phenomena (Catani et al., 2004), palaeodrainage range.
features (Robinson et al., 2006) and the movement of rock High-resolution acoustic profiling is used in studies of
glaciers (Rignot et al., 2002). It also has a range of glacio- seabed processes and stratigraphy (Sacchetti et al., 2013),
logical (Gao & Liu, 2001; Rémy & Parouty, 2009) and geo- in the mapping of submerged landforms (Dowdeswell et al.,
archaeological applications (Wiseman & El-Baz, 2007). 2010) and in investigations of sedimentary sequences
A specialized aspect of radar imaging is ground beneath polar ice shelves (Johnson & Smith, 1997). Other
penetrating radar (GPR), a geophysical investigative applications include investigations of lake-bed sediments
procedure that employs radar signals to image subsurface (Eyles et al., 2003), calculation of suspended sediment
phenomena. It offers a non-destructive and rapid method loads (Thorne & Hanes, 2002) and offshore palaeoseis-
for analysing subsurface objects and structures in a variety micity (Hutri & Kotilainen, 2007). Further details on
of materials, including rock, sediments, soil, ice and geophysical exploration can be found in Kearey et al.
water. Penetration is limited by a number of factors, (2002).
including the electrical conductivity of the medium through
which the signals are transmitted, and the maximum depth
2.2.2.5 Digital elevation/terrain modelling
of penetration (c. 15 m) is only achievable in dry, sandy
media. GPR has been widely employed in archaeology, for Remote sensing methods enable accurate topographic
example to detect and record shallow burials (Cheetham, models of the earth’s surface, known as digital elevation
2005) and to map subsurface archaeological structures models (DEM) or digital terrain models (DTM), to be
1.5'W 1 W C5 W
palaeo-Arun
valley
50.5 N
palaeo-Solent Bench
valley
Solent
Scarp confluence
Platform
M5
N
Headcut
Scour
W a t e r d e p t h (m)
Sinuous channel 20 km
Inner c h a n n e l s Seoul- c 20 4C so so 100
SON
Figure 2.2 Sonar bathymetry of the north-central English Channel shelf showing (in blue) the deeply incised channel of a major
river which was cut following the confluence of the Thames, Rhine, Seine and other major European rivers at a time when sea
level was >100 m lower than today (from Gupta et al., 2007). Onshore topography is shown as black and white relief (Isle of Wight
and southern shores of England) (image provided by Sanjeev Gupta, Imperial College, London, UK. Reprinted by permission from
Macmillan Publishers Ltd: Nature, Gupta et al., copyright 2007).
constructed. DEMs/DTMs are created from regularly obvious on maps or generated by ground-based surveying.
spaced grids of elevation data which are referenced to a The azimuth of light incidence can be rotated through
common geographic coordinate system. These can then be 360° in DTMs to vary the length and direction of shadows
used to create 2-D and 3-D topographic representations cast by surface features; this can greatly enhance the 3-D
of the area of ground surface that they cover. DEMs show appearance of individual landforms and also reveal pat-
all the features captured by the survey process, including terns in the landscape that may be less obvious under high
woodland and buildings for example, whereas the super- incident light or overcast conditions (Figure 2.3). The data
ficial features are filtered out in a DTM to show the ground can also be linked into GIS software such as ARC/INFO,
surface only (‘bald-earth’ elevation mapping). A number which enables other data to be overlain on the terrain
of different types of remotely sensed data can be used for surface.
terrain modelling, but one of the more important is A number of companies offer access to DEM/DTM
interferometric synthetic aperture radar (IFSAR), which data-sets. InterMap Technologies, for example, provides
can generate DEMs with a resolution of less than 10 m. NEXTMap, a DEM program that covers most of Europe
DEMs have revolutionized the study of landforms as and the USA, and selected areas such as some Pacific
the images often show ground surface detail that is not islands. It is now widely used to generate base surfaces for
Forth
Valley
t
ul
Loch
Fa
Somond
ry
da
un
Bo
nd
la
gh
Hi
0 5 10 k m
Figure 2.3 A digital surface model (DSM) based on NEXTMap showing the ice-moulded landscape of part of the Midland Valley
of Scotland to the northeast of Glasgow. The Highland Boundary Fault marks the southern margin of the Grampian Highlands.
Ice-moulded bedrock and drumlin orientations show that ice flowed from the Highland valleys along the Forth Valley towards the
North Sea (arrows indicate inferred ice-flow directions) (DSM model generated by Chaoyuan Chen, Royal Holloway, University of
London, UK using interferometric synthetic aperture radar [IFSAR] data obtained by Internet Technologies, Inc. and provided courtesy
of the British Geological Survey).
geomorphological mapping. The spatial resolution and

accuracy of DEMs and DTMs depends on a number of 2.3 GLACIAL LANDFORMS
factors, including the type of sensing device employed and Ever since the acceptance of the ‘Glacial Theory’ by the
the algorithms used to convert the raw data into eleva- geological community in the middle years of the nine-
tion data. Remotely sensed images have to be corrected teenth century, it has been recognized that landforms
(a process known as orthorectification) to remove dis- of glacial erosion and deposition are important palaeo-
tortions caused by topographic slope, lens distortion environmental indicators. When mapped systematic-
(parallax), camera tilt and other influences. This process ally, they reveal a great deal about the extent, thickness
can improve the spatial resolution of the resulting images and behaviour of former ice masses, the direction of ice
to 1 m or less. movement at both local and regional scales, and the nature
A recent development in digital elevation/terrain model- and pattern of glacier retreat. In certain circumstances, the
ling is light detection and ranging (LIDAR) technology. evidence may be used to reconstruct the configuration of
This operates on the same basis as radar, except that it former ice sheet or glacier surfaces, and to allow former ice
employs laser beams rather than radiowaves for detection. volumes to be estimated. Comparisons with present-day
LIDAR supports airborne laser swath mapping (ALSM) glaciers, particularly those where a close relationship has
and laser altimetry which can scan ground surfaces at a been established between glacier behaviour and climatic
much higher density of beam points than radar, and hence parameters, enable inferences to be made about former
can generate elevational data and images with, potentially, climatic regimes. Glacial geomorphological evidence is
a centimetre-scale resolution. LIDAR mapping is therefore also key to the development of computer models that
rapidly becoming the new generation DEM. Further details simulate the geometry and behaviour of former ice sheets
can be found in Hatzopoulos (2008). and glaciers. Global ice volume co-varies with changes in
sea level, ocean circulation and hemispherical climate The land lying beyond or above the area directly affected
(Chapter 7), and an understanding of these complex by glacier ice will have been affected by periglacial activity,
interrelationships is an important element of contemporary with the shattering of exposed rocks by freeze–thaw
environmental science. The first stage in this process, processes, the development of gelifluction features (lobes,
however, is the production of an accurate map of the terraces, etc.) and the formation of structures associated
extent of Quaternary ice masses for different times in the with the action of ground ice (wedges, patterned ground,
past. etc.). The distribution of these periglacial features can,
in certain cases, provide further evidence of the former
extent of glacier ice. In upland areas, for example, the
2.3.1 Extent of ice cover boundary (or, more commonly, the zone of transition)
Establishing the extent of former ice sheets and glaciers has between glacially scoured and frost-shattered bedrock is
long been regarded as one of the most challenging objectives referred to as the trimline, and indicates the approximate
of Quaternary research. In North America, the systematic positions of the former ice margins (Figure 2.5). Careful
field mapping of the outer limit of Pleistocene glaciation measurement of the trimline altitude on different mountain
began soon after 1860. The maximal extent of ice cover was peaks enables the upper limit of glaciers and ice sheets to
based largely on the evidence of conspicuous ‘end moraines’ be established at the regional scale (Ballantyne et al., 2008).
or on the limits of glacigenic deposits (‘Drift Border’), and The lateral extent of the area affected by frost action can
by 1878, a map had been produced of the southern margin be used in a similar manner to delimit formerly glacierized
of the glaciated area between Cape Cod and North Dakota. areas.
Similar investigations of glacial drift cover and end A number of difficulties arise in mapping the former
moraines were underway in Europe, Asia and parts of the extent of Pleistocene glaciers on the basis of geomor-
Southern Hemisphere, so that by 1894 James Geikie was phological evidence. First, at the height of the last glaciation
able to compile maps of the worldwide distribution of in both Europe and North America, ice masses submerged
glaciers during what he referred to as the ‘Great Ice Age’. many of the upland areas. Hence, although geomorpho-
The principal types of geomorphological evidence used logical evidence marking the lateral extent of the ice sheets
in the reconstruction of former ice-marginal positions and glaciers is widespread, the vertical extent of the ice
are lateral, terminal, end and retreat moraines, outwash mass is much more difficult to establish from field evi-
spreads and sandar,3 ice contact features such as kame dence alone. In order to obtain estimates of ice thickness,
terraces, and valley-side or down-valley limits of stagna- therefore, recourse may have to be made to models of
tion moraine, boulder spreads (boulder limits) or drift former ice sheets (section 2.3.4). Second, successive ice
limits (Bennett & Glasser, 2009; Benn & Evans, 2010). sheets covered broadly the same areas, except towards the
Lateral and terminal moraines (and, in some cases, mar- outer margins (Figure 1.4 and section 2.3.2), so that geo-
ginal meltwater channels) mark the maximal positions of morphological evidence from earlier glacial episodes has
glacier margins, whereas within those limits, linear moraine usually been destroyed by later ice advances. Consequently,
ridges will reflect subsequent recessional stages as the ice in many areas of the mid-latitudes that were affected by
becomes temporarily stabilized during deglaciation (Figure Pleistocene glaciers, the majority of the landforms that
2.4), while widespread moundy topography (‘dump’ or have been preserved date only from the later stages of
hummocky moraine) may result from glacier stagnation in the last glaciation. Third, many glacial landforms have
situ (section 3.3). Kame terraces, which reflect glaciofluvial been considerably modified by periglacial and paraglacial
deposition along a decaying ice margin, may also preserve activity (section 3.3.3.3) both during and after regional
a record of ice-marginal positions during glacier wastage deglaciation, and this often poses problems in field map-
(Bitinas et al., 2004). The types of deposits and landform ping and interpretation. Fourth, some glacial landforms
assemblages produced during deglaciation will be deter- may resemble ice-marginal features, but may not, in fact,
mined by a range of often interconnected factors, including be so. Glaciofluvial landforms (e.g. kames and eskers) often
manner and rate of glacier retreat, debris content of the display linear trends and have, on occasion, been inter-
ice, position of entrainment of debris within the ice, and preted erroneously as evidence for former ice limits. In the
topographical influences (Glasser and Hambrey, 2001; majority of cases, the lineations displayed by such features
Benn & Lukas, 2006). In general, however, the overall reflect local patterns of ice disintegration rather than the
distribution of a range of glacial landforms will broadly retreating margins of active ice masses. Hence, a proper
define the extent of the formerly glaciated area, with both understanding of the nature and origin of glacial landforms
lateral (ice-marginal) and vertical (valley-side) limits. is necessary if this type of evidence is to be used to determine
a)
b) c)
Figure 2.4 Moraine systems. a) NASA satellite image of the glacier icefield in Bhutan-Himalaya. Note the glaciers are receding,
leaving a series of arcuate (‘horse-shoe’) moraines and lakes near their termini. b) Marginal lateral moraines (dotted box) and
‘hummocky’ or ‘dump’ moraines of Loch Lomond Stadial age (c. 12.9–11.7 ka) in Coire na Phris, northwest Scotland (from Lukas
& Benn, 2006; photograph by Sven Lukas, Queen Mary University of London, UK. © Royal Scottish Geographic Society, reprinted
by permission of Taylor & Francis Ltd, http://www.tandfonline.com on behalf of The Royal Society Geographic Society). c) Moraine-
mound comprising diamicton and sandy gravel melting out of glacier ice at Alpha Glacier on James Ross Island, Antarctica – a
possible modern analogue for ‘hummocky moraine’ (photograph by Neil Glasser, Aberystwyth University).
the former extent of Pleistocene glacier ice. Fifth, mapping Sixth, the outer margins of large areas of the last great ice
of the upper limits of glaciers and ice sheets on the basis sheets lay beyond the present coastline on the continental
of trimline evidence may be problematical, for high-level shelves, and accurate determination of the ice limits has
trimlines may not always reflect former periglacial weather- only recently become possible through the application of
ing limits, but rather the position of a thermal boundary remote sensing techniques (Bradwell et al., 2008a). Finally,
separating wet-based ice at pressure-melting point from it should be noted that the outer margins of drift sheets
cold-based ice on summit plateaux (Fabel et al., 2012). often have no distinctive geomorphological expression,
a)
a)
(Jagged) Mountain summit blockfield
Tor-like s u m m i t s Frost shattered
Solifluction bedrock
terraces. Solifluction lobes
Solifluction
sheet
Trimline
(Smooth)
Solifluction- Ice m o u l d e d
apron bedrock
Till
(Smooth)
Debris-strewn
slope
Perched
boulders
Fossil scree
Roches
ck
dro
moutonnees
Trimline
(Smooth)
Glacial pavements
Be
G u l l i e d till Hummocky moraine
b) c)
Figure 2.5 a) Features used to identify trimlines and other types of glacial limits in mountainous terrain (from Thorp, 1981).
b) Nunataks bordering Liv Glacier in the 3,000 m high Queen Maud Mountains, a major outlet glacier from the East Antarctic Ice
Sheet to the Ross Ice Shelf. Flow is from right to left (photograph © Mike Hambrey, Aberystwyth University, from
http://www.glaciers-online.net). c) High-altitude trimline (dashed line) marking the maximum late Würm ice surface at c. 2,670 m
in the Grimsel Pass, Switzerland, this is marked by the upper limit of ice-polished bedrock above which the rock surfaces show
evidence of prolonged frost shattering (photograph by Sven Lukas, Queen Mary University of London, UK).
and end moraines in particular are absent. In some cases, Where end moraines are found within areas formerly
the evidence has been destroyed, either by meltwater activity covered by glacier ice, further problems of interpretation
during deglaciation, or by postglacial erosion or subaerial are encountered. There has been considerable debate over
weathering. In other situations, the glaciers either did not whether such moraines are, strictly speaking, recessional,
carry sufficient debris, or did not maintain a steady-state4 in that they formed during the still-stand of the ice margin
position for the length of time required for the construction during a phase of overall glacier retreat, or whether
of an end moraine. they have been produced by a renewed episode of glacier
expansion, and therefore reflect a glacier readvance. Where MKt-

prominent end moraines occur, the latter interpretation has
usually been adopted, as it has been considered unlikely that
large constructional forms would have been produced
during a still-stand of the ice margin. The geomorphological
evidence, however, is frequently equivocal, and it must be
emphasized that conclusive proof of a glacier withdrawal
and a subsequent readvance can only be obtained from
stratigraphic evidence (where organic sediments are found
interbedded with two glacigenic units, for example) or, in
some instances, from other geomorphological features,
such as indicators of changing directions of ice flow between
successive glacial episodes (Clark, 1997). An additional MKt-
complication in the interpretation of end-moraine evidence
is that it is frequently very difficult to distinguish between
those landforms that reflect a glacier readvance induced by
a deterioration in climate, and constructional forms that
have been produced by a glacier surge5 resulting from
short-lived instability within the former glacier system
l
rg
which may, or may not, have been climatically determined m
bu
(Lingle & Fatland, 2003). Despite these difficulties, however, ta

H
Bergen*^
there is now a broad consensus about the positions of the tH a m b u r g l

tH a m b u r g l Hamburgl tH a m b u r g l
maximal limits and general patterns of decay of the last ice MKt- Middle
.''.'iVinn Hamburgl
t Swedish
Hamburgl
sheets that occupied North America and northern Eurasia, tH a m b u r g l N
Riga
although in some areas local details remain contentious
(Ehlers et al., 2011a; Hughes et al., 2013). Hamburgl AFrtmkfurtc
HtHaammbbuur g
r gl l
Bremen tH a m b u r g l BQlOQOyQ'EdflOVO
0 200km
tH a m b u r g l
H
am
Hb
Poznan
am
ur
bgul
gr
2.3.2 Geomorphological evidence and the

extent of ice sheets and glaciers during Figure 2.6 a) Maximal limits of the Late Weichselian ice
the last cold stage sheet in MIS 2 (after Svendsen et al., 2004). b) Recessional
stages of the Late Weichselian ice sheet (after Böse, 2005).
2.3.2.1 Northern Europe The southern ice-marginal zone on the north German
In northern Europe, conspicuous ‘end moraines’ mark the plain is characterized by three prominent moraine belts
maximal limits (Figure 2.6a) and initial recessional stages named (from south to north) the Brandenburg, Frankfurt
of the last (Weichselian) Fennoscandian ice sheet (Böse, and Pomeranian Moraines. These are not single structures,
2005). These extend over considerable distances from however, for each comprises a narrow band of subparallel,
northern Denmark through Germany into Poland, and sometimes branching and occasionally linked ridges. In
then northwards through the Baltic states and into between lie flatter till plains with minor moraine mounds
European Russia (Figure 2.6b). In places they are discrete and kames, incised through which are large abandoned
narrow ridges, while elsewhere they comprise broad tracts fluvial channels (Urströmtaler) which were the principal
of hummocky ground a kilometre or more wide; sometimes meltwater conduits along the receding ice margin (Böse,
the ridges fade into flat, almost featureless ground and 2005). The geomorphological evidence points to episodic
cannot be traced continuously (Rinterknecht et al., 2005). retreat of the last ice sheet from its most southerly position
Moreover, the term ‘moraine’ that is widely applied to these marked by the Brandenburg Moraine (Figure 2.6b).
features may not always be appropriate, since some Radiocarbon dating of organic material beneath the
segments comprise glaciofluvial or outwash sediments, moraine suggests a maximal age for the feature of c. 21 ka
14
while others have been formed by the bulldozing action of C BP (c. 24 ka cal. BP: Chapter 5), which is very similar to
the ice margin, which has thrust and deformed superficial the age estimate (23.2 ka) inferred for the Brandenburg
deposits or pre-Quaternary strata (Figure 2.7). Moraines from an annually laminated lake sediment record
a) b)
Figure 2.7 a) Cross-section through ice-marginal ridges (push moraines) of Late Weichselian age in the Limfjord Area, Jutland,
Denmark, exposing a series of glaciotectonally displaced slabs of Eocene diatomite (white cliff faces) with volcanic ash layers
(black) overlain by Quaternary diamictons (ice-push from north-east [right to left], cliff height 25 m). b) Up-thrusted and folded Middle
Weichselian lacustrine and fluvial sediments exposed on the North Sea coast of northernmost Jutland; deformation is from north
to south (left to right) (photographs by Michael Houmark-Nielsen, University of Copenhagen, Denmark).
in the Eifel district of Germany (Zolitschka et al., 2000). The regional correlation. In the eastern Baltic lowlands, for
lake sediment data suggest maximal ages for the Frankfurt example, the CRN evidence suggests that the last ice sheet
and Pomeranian stages of 19.7 ka and 16.5 ka, respectively. advanced after c. 25 ka, reached its maximal extent by
Cosmogenic radionuclide dating (CRN dating: Chapter 5) c. 21 ka and began to retreat c. 19 ka, a chronology that is
of boulders on these moraines implies ice retreat northward broadly comparable with that obtained for the north
from the Brandenburg Moraine around 19 ka, with retreat German Plain (Rinterknecht et al., 2006). In Poland the
from the Frankfurt and Pomeranian Moraines c. 18 ka and Lezno phase is CRN-dated to around 24 ka, the Poznan
c. 16 ka, respectively (Heine et al., 2009). These ages are phase to c. 19 ka and the Pomeranian phase to 17–16 ka BP
broadly confirmed by more recent CRN of Pomeranian (Marks, 2012). Further east in Belarus, the maximal
stage deposits which show a tightly clustered distribution extent of the last ice sheet is indicated by the Orsha
of ages between 17.2 and 16.4 ka (Rinterknecht et al., Moraine, with ice retreat from that position dated at 17.7
2012). ka (Rinterknecht et al., 2007). However, subsequent ice
Correlations between the north German moraines and wastage of this south-eastern flank of the Scandinavian ice
those in other parts of Europe have proved problematical. sheet may have been largely through surface thinning
In Denmark the maximum extent of Weichselian ice rather than marginal retreat (Bitinas, 2012). In the far
is marked by the Mid Jutland (Figure 2.6b) or Main north, the situation is less clear, for it is only recently that
Stationary Line: this is not, however, correlated with the geomorphological evidence from around the Barents Shelf
Brandenburg, but with the Frankfurt Moraine, while the has become available to scientists outside the former USSR,
more northerly East Jutland Line is considered to be a while interpretation of marine records from the Barents
continuation of the Pomeranian Moraine (Lundqvist, 1986; Sea is far from straightforward (Svendsen et al., 2004).
Houmark-Nielsen, 1989). In Poland, the Frankfurt Moraine It does appear, however, that during the Late Weichselian,
has been correlated with the Poznan Moraine and the an extensive and dynamic, possibly multi-domed ice sheet
Pomeranian with the Pomorze Moraine, because more formed over the Svalbard-Barents Sea (Ingólfsson &
southerly ridges (the Leszno Moraine in Poland and Landvik, 2013). This extended to Svalbard in the far north
the Belogoye Moraine in Russia) have been judged to be and to the edge of the Barents Shelf, and was confluent with
extensions of the Brandenburg Moraine. Attempts have also the northern margins of the Fennoscandian ice sheet
been made to link prominent linear ridges in the eastern (Figure 2.6a).
Baltic region to the Brandenburg–Frankfurt–Pomeranian More is known about the limits of the last Fenno-
model, although this frequently rests on visual matching of scandian ice sheet along its western margin due largely to
linear trends and often equivocal stratigraphic evidence. several decades of exploration of submarine deposits for
Where CRN dates are available, however, these offer a oil and gas. A combination of remote-sensing and marine
basis for construction of glacial chronologies and inter- core data has revealed submarine ridges close to the edge
of the continental shelf with large fans of glacigenic sedi- of the northern Baltic consist of three prominent moraine
ment along the continental slope (Sejrup et al., 2005). belts commonly over 60 m in height, with the two outer-
These were produced by major ice streams from the most moraines extending almost continuously for
mountains of Norway that reached the shelf margin on almost 600 km from the coast of southwestern Finland to
at least three occasions during the Weichselian, the last North Karelia in eastern Finland (Rainio et al., 1995).
being between 28 and 22 ka. Off the northwest coast of All were once considered to be true end moraines, but
Norway, the limit of the last ice advance is marked by the the Salpausselkäs consist largely of deltas and smaller till
Egga Moraines, the oldest of which is dated to > 25 ka ridges which appear to have formed in the Baltic Ice Lake,
(Vorren & Plassen, 2008). Further south, data from the extensive water body that was impounded to the south
exploration activities in the northern North Sea have also of the Fennoscandian ice sheet during deglaciation (Björck,
helped to settle a long dispute about whether or not the Late 1995). The composition of the Salpausselkä Moraines in
Weichselian Fennoscandian ice sheet was confluent with eastern Finland was largely determined by the nature of
the Late Devensian British ice sheet (section 2.3.2.2), the the ice-marginal environment; where the ice terminated
evidence now suggesting coalescence between the two ice in water, massive glaciofluvial sediment sequences accum-
masses in the central North Sea between 29 and 25 ka ulated, whereas discontinuous and narrow moraine ridges
(Sejrup et al., 2009; Figure 2.6a). formed where the ice-front stabilized on dry land. The form
Following the Pomeranian stage, episodic retreat of the of the Salpausselkä Moraines indicates a lobate ice margin,
last Fennoscandian ice sheet is marked by a series of sub- with the outermost of the Salpausselkä ridges possibly
parallel moraines across northern Denmark and south- marking the most advanced position during the Younger
west Sweden. From south to north these are the Halland Dryas (Rainio et al., 1995). CRN dating indicates that
(18–16 ka), Göteburg (15.4–14.5 ka), Berghem (14.4–14.2 retreat from the outermost moraine occurred at c. 12.5 ka,
ka), Trollhätten (14.2 ka) and Levene Moraines (13.4 ka) which is consistent with an age of 12.3 ka based on varve
(Lundqvist & Wohlfarth, 2001). The Trollhätten Moraine chronology (Rinterknecht et al., 2004). However, it appears
is correlated with the prominent Hvaler Moraine in the that the Salpausselkäs may not be synchronous features
outer Oslofjord of southern Norway (Andersen et al., 1995) throughout southern Finland, but could represent re-
and possibly with the Little Fiskebank Moraine of Denmark advances by different ice lobes at slightly different times.
(Nesje & Sejrup, 1988). It is not clear, however, whether the The same situation obtains in Norway, for new dating
linear moraine systems of Germany, Poland and the eastern evidence indicates that the Younger Dryas maximum
Baltic are the temporal equivalents of those in Denmark occurred significantly later in western Norway than else-
and southern Sweden. Indeed, it seems likely that retreat where in Scandinavia, whereas small glaciers in the north
of the southern ice margin was a time-transgressive pro- expanded to their maximal positions very early during the
cess, which was interrupted by both local and regional Younger Dryas (Bondevik & Mangerud, 2002). These
readvances (Lagerlund & Houmark-Nielsen, 2008). contrasts reflect differing glaciological responses to topo-
The best-known and most prominent moraines of the graphic and climatic factors at different locations around
Fennoscandian ice sheet formed as a result of a readvance the ice sheet margin and suggest that, on the continental
during the Younger Dryas Stadial (c. 12.9–11.7 ka). These scale at least, former ice limits as shown by geomorpho-
have been mapped as almost continuous belts across logical evidence are more likely to have been metachronous
southern Norway (Ra Moraines), Sweden (Middle Swedish than synchronous. This has implications for ice sheet
Moraines) and Finland (Salpausselkä Moraines). Along or models derived from such data (section 2.3.4).
close to the west coast of Norway, distinctive terminal During the Holocene, the last ice sheet over Scandinavia
moraines dated to this period can be traced within a narrow retreated from its Younger Dryas limits, although episodic
belt stretching from the Russian border in the north to the readvances formed moraines some distance beyond
Swedish border in the south, a distance of some 2,500 km younger (‘Neoglacial’) maxima. These included the Erdalen
(Andersen et al., 1995). They include the Herdla Moraines Event between 10.1 and 9.7 ka (Dahl et al., 2002), and
of the Bergen area and the Tromsø–Lyngen Moraines the Finse Event c. 8.3 ka (Nesje et al., 2000). Between c. 8
of northwest Norway. Morphologically and genetically, and 4 ka, however, most glaciers in Norway had melted
however, the Younger Dryas terminal moraines are very completely due to a combination of high summer tem-
different. In the west of Norway, they frequently comprise peratures and reduced precipitation (Nesje et al., 2008).
single or paired subparallel ridges, while those that termin- But throughout the high mountain areas of Norway and
ated in the sea are occasionally associated with prominent Sweden, there is abundant evidence for renewed glacier
ice-front deltas. By contrast, the Salpausselkä Moraines activity after that time. This consists of ‘fresh’ and relatively
unvegetated terminal and lateral moraines down-valley falsified by CRN dating (Phillips et al., 2008b). Indeed,
from present-day active glaciers (Nesje et al., 2001). In some evidence from offshore (Figure 2.8) shows that ice covered
areas, as many as seven readvances occurred during the the whole of the Scottish mainland at the Late Devensian
late Holocene or ‘Neoglacial’ period (Matthews & maximum (Chiverrell & Thomas, 2010).
Dresser, 2008), the youngest of which occurred during As in other areas of Europe (section 2.3.2.1), CRN
the ‘Little Ice Age’, the period of widespread cooling that dating is one of a number of important recent develop-
lasted from the fourteenth to the eighteenth centuries ments that have provided new insights into the dimensions
(section 7.6.3.5). and behaviour of the last British–Irish ice sheet (Ballantyne,
2010). A second is digital elevation mapping based on
high-quality satellite imagery by the BRITICE project,6
2.3.2.2 Britain and Ireland
and which provides high-resolution data on glacial
In Britain, the maximum expansion of the last ice sheet landform assemblages that can be viewed at both regional
occurred towards the end of the Devensian cold stage, and national scales (Clark et al., 2004; Evans et al., 2005).
during the Dimlington Stadial (Rose, 1985). The limits Landform patterns that reflect ice-marginal or subglacial
(Figure 2.8) were initially inferred from geomorphological contexts can be readily distinguished, providing a first
evidence, with products of the Late Devensian ice sheet approximation of the dimensions and flow patterns of the
(‘Newer Drift’) in England and Wales being differen- last British–Irish ice sheet at its maximum extent (Figure
tiated from those of earlier glaciations (‘Older Drift’) on the 2.8). Third, considerable advances have been made in the
basis of contrasts in degree of dissection and relative mapping of submarine landforms and deposits on the
preservation of ‘constructional’ glacial topography (Bowen continental shelf (Bradwell et al., 2007) and in the North
et al., 2002). In Ireland, the ‘South Irish End Moraine’ has Sea (Graham et al., 2007). Fourth, the recognition of
long been considered to delimit the maximum extent of trimlines in some upland regions has enabled the vertical
Late Midlandian (cf. Late Devensian) ice. In many places, dimension of the ice sheet to be established and hence
however, the geomorphological evidence is equivocal, regional ice-surface gradients of the last ice sheet to be
and it is clear from more recent stratigraphic records and delimited (McCarroll & Ballantyne, 2000). CRN dating
from other data that the geomorphic features that have provides the essential chronological control on these
traditionally been regarded as marking the maximal limits landform features (Ballantyne et al., 2008).
of the ‘Newer Drift’ may not, in fact, do so. In southwest Bathymetric data from the seabed around northern
Wales, for example, it has long been known that the ice Britain has revealed remarkable geomorphological evi-
maximum lay to the south of the mapped position of dence in the form of moraines and tunnel valleys7 of the
the ‘South Wales End Moraine’ (Bowen, 1981), while in dimensions and behaviour of the former British–Irish
Ireland, it is now apparent that the Late Midlandian ice and Fennoscandian ice sheets (Bradwell et al., 2008a). At
sheet extended well beyond the ‘South Irish End Moraine’ its maximum, a grounded ice sheet flowed southeast to
(Ó Cofaigh & Evans, 2007), with a major lobe of ice northwest across the northern North Sea basin and
spreading southwards down the Irish Sea basin to reach the terminated at the continental shelf edge, the northwestern
Scilly Isles where it is marked by a well-defined terminal part of the ice sheet being most extensive between c. 27 and
moraine (Hiemstra et al., 2006). Indeed, most of southern 25 ka (Everest et al., 2013). The zone of confluence with the
Ireland was glaciated at the Last Glacial Maximum (LGM), much larger Fennoscandian ice sheet was probably across
and hence the morainic belts which have traditionally been the northern Orkney Islands. The period of maximum
interpreted as marking the last glacial limits, such as the confluent glaciation (c. 29–25 ka) was followed by the
South Irish End Moraine, are now considered as recessional opening of a marine embayment in the northern North
features during ice sheet retreat (Ó Cofaigh et al., 2012a). Sea, the decoupling of the two ice sheets and stabilization
In the Vale of York, it remains uncertain whether the at new margins marked by a second distinct set of moraines
Escrick and York Moraines mark the maximal extent of the (24–18 ka). The lobate, over-printed morphology of these
Late Devensian ice sheet (as has long been considered to moraines, particularly on the mid-continental shelf west of
be the case) or whether the ice extended c. 50 km further Orkney and Shetland, indicates that this reorganization
south into a large ice-dammed lake (Glacial Lake Humber), of the British ice sheet was extremely dynamic and punc-
possibly accompanied by a glacier surge (Evans et al., tuated by a series of readvances. To the west of the British
2005). In Scotland, the long-held view that areas of Isles, numerous large moraines along the continental shelf
northeast Scotland (‘moraineless Buchan’) remained free record an extensive pattern of retreat stretching from
from glacier ice during the Late Devensian has now been southwest Ireland to the Shetland Isles (Clark et al., 2012).
Figure 2.8
Limits of the Late Devensian
British ice sheet and principal
geomorphological lines of
evidence (generalized) on
which they are based (from
Clark et al., 2004. Copyright ©
2008, John Wiley and Sons).
Tunnel valleys
Moraine ridges
Me:.:;\\::J Channels
Lateral meltwater channels
Loch Lomond readvance limit
Loch Lomond re-advance limit {inferred)
Shelf-edge fans
Esker
Lilhological limit relevant to erratic dispersal
Inferred erratic path
Erratic Limit
Drumlin
Limit or glaciogenio deposits
Trimline
Indicator erratic source area
Moraines
Glaciolacusmne deposits
Ice-dammed lake at its lower stand
Ice-dammed lakes
Geochronological data from the Irish Sea basin indicates the Cheshire–Shropshire lowlands, have long been regarded
that the ice maximum in the Celtic Sea to the southeast of as marking readvances of the ice sheet margin. In other
Ireland was reached around 25.3–24.5 ka (Chiverrell et al., areas, a combination of glacial geomorphology and strati-
2013). graphy provide firm evidence for readvances, as in the
Elsewhere, readvances of the British–Irish ice sheet Solway Lowlands of Cumbria (Livingstone et al., 2010)
during deglaciation from the LGM are reflected in both and on the Isle of Man where the stratigraphic succession
geomorphological and stratigraphic evidence. Prominent associated with the major constructional form of the
moraine complexes up-glacier from the maximal ice limits, Bride Moraine indicates a readvance episode between 22
such as the Wrexham–Ellesmere–Whitchurch Moraine of and 18 ka (Thomas et al., 2004).
In the Irish Sea Basin, the interdigitation of glaciomarine glacier activity now appears to reflect a short-lived re-
muds and glacial diamictons also provides compelling advance of the downwasting residual mass of the Late
evidence for ice-margin readvances, particularly where Devensian ice sheet (Bradwell et al., 2008b). Some Scottish
marine fossils contained within the glaciomarine sedi- glaciers seem to have reached their maximum extent during
ments have been radiocarbon dated. In Dundalk Bay on the the early part of the Stadial (Golledge, 2010), while others
northeast coast of Ireland, for example, moraines record attained maximal positions towards the end of the cold
two readvances of the ice sheet into the northern Irish phase (MacLeod et al., 2010). At the onset of the Holocene,
Sea basin, the Clogher Head Readvance (c. 18.3–17.3 ka) glacier ice wasted rapidly and has not reappeared since in
and the Killard Point Readvance (c. 17.1–15.2 ka) (McCabe the British Isles.
et al., 2007). In western Ireland, CRN dating of moraines
to c. 15.6 ka suggests that these also formed during the
2.3.2.3 North America
latter glacial event (Clark et al., 2009a). In western Scotland,
glacial erosional and depositional landform assemblages There were three major ice sheets in North America during
revealed in high-resolution remote sensing datasets indi- the Wisconsinan: the Laurentide, which covered the
cate renewed ice sheet thickening in the Firth of Clyde Canadian Shield and northern USA as far south as the Great
during the Killard Point Stadial (Finlayson et al., 2010). Lakes; the Cordilleran, in the mountains of Alaska, Yukon
Further north, the prominent Wester Ross Moraine, and British Columbia; and the Innuitian over Ellesmere and
a distinctive boulder-strewn linear ridge which extends the Queen Elizabeth Islands (Figure 2.9a). Smaller moun-
across the peninsulas of Wester Ross, has been CRN-dated tain glacier complexes also developed over the Rockies in
to c. 16.3 ka (Everest et al., 2006) and, more recently, to the western USA (Hughes et al., 2013).
c. 14–13.5 ka (Ballantyne et al., 2009), the latter age estimate The Laurentide ice sheet formed from three ice centres
being very similar to those obtained on major readvance located in Quebec–Labrador, Keewatin and Foxe Basin,
moraines elsewhere in northwest Scotland (Bradwell et al., near Baffin Island (Dyke et al., 2002), and reached its
2008b). maximum towards the end of the Wisconsinan (28–25 ka)
The most convincing evidence for a renewed period of when it was more extensive than in any previous glacial
glacial activity following retreat from the Late Devensian stage (Andrews & Dyke, 2007). For more than half its
maximum can be found in the uplands of Scotland, length, the outer limit of the ice sheet between the Atlantic
northern England, Wales and southern Ireland, where Ocean and the Rocky Mountains is marked by prominent
terminal and lateral moraines, spreads of hummocky terminal moraines, although there are large areas where no
recessional moraine (Figure 2.4b), well-developed trim- geomorphological distinction can be made between Late
lines on mountain sides (Figure 2.5), intricate meltwater Wisconsinan and older deposits, and where mapping of the
channel systems and the distribution of periglacial fea- last ice limit rests largely on stratigraphic evidence. Within
tures define the limits of the Loch Lomond Readvance. the Late Wisconsinan limits of the Laurentide ice sheet,
This is broadly equivalent to the Younger Dryas readvance end moraines marking possible stillstands or readvances
of Scandinavia, and is dated to between 12.9 and 11.7 ka of the ice margin occur (Figure 2.10), and in many areas
(Golledge, 2010). Confirmation of a readvance of ice is the deglacial chronology of the last ice sheet can be
provided by radiocarbon dates of >11.0 ka BP on marine reconstructed (Dyke, 2004). Unlike western Europe, CRN
shells which subsequently became incorporated into dating has not been as widely used on moraines around
terminal deposits as ice invaded estuarine localities (Sissons, much of the former Laurentide ice sheet margin, but it
1979), and on organic debris buried beneath till (MacLeod has been employed to date landforms and exposed
et al., 2011). bedrock surfaces in the western mountains and Alaska
The freshness and relatively unweathered nature of the (e.g. Owen et al., 2003; Refsnider et al., 2008; Badding
landforms have enabled the extent of the readvance to et al., 2013).
be mapped in great detail (Benn & Ballantyne, 2005; In the Great Lakes region, where different ice lobes
Ballantyne, 2007a). This has not only allowed the ice limits formed during ice wastage from 15 ka onwards, there
to be established accurately, but has also made possible the are large numbers of end moraines, some of which can be
reconstruction of individual cirque and valley glaciers traced over considerable distances and which appear to
(section 2.3.4.2). In many upland areas of Britain, glaciers reflect broadly synchronous regional readvances of the ice
formed anew during the Loch Lomond Stadial, but in margin in both the western and eastern parts of the
Scotland, where a substantial ice sheet existed across the Great Lakes basin. These include the Port Huron Advance
west-central highlands (Figure 2.8), the episode of renewed in the Erie–Ontario region (c. 14.5 ka), the Greatlakean
a) lb) ON
Arctic O c e a n MN Ml
ND
Superior
Cordilleran Iceland
Cordilleran
De
Sheet
IIS MN
Jam
y n
sM
Baffin
Michigan
Ba ree
Baffin SD
es
oi n
G
ON
Ml
es
Labrador
i Sea lA
ke
ie
-Er e
La
Cordilleran NE ron Eri
Ice Laurentide Hu
Sheet Ice S h e e t IN OH
IL
MN
Mountain KS
glacier
complexes
0 100 200 miles KY
0 150 300 miles N
Lak
C) 89- W
Phases:
eM
C = Crown Point M 28.5
W - Woodstock 28.5
ichi
L = Livingston
42 P = Putnam
gan
S = Shelbyville 28.5
SE
M = Marengo
C Chicago
17.6
21.8
El p
23.0
0 25
2525 km
kmkm
0 25 km
Figure 2.9 a) The Late Weichselian Laurentide, Cordilleran and Innuitian ice sheet (IIS) at the LGM (after Andrews & Dyke, 2007).
b) Reconstructed ice lobes in the Great Lakes region at the LGM. c) The pattern of retreat moraines of the Michigan–Huron lobe
after the LGM (b) and c) are from Curry et al., 2011).
Figure 2.10
Retreat pattern of the Laurentide ice 180° 18 c a l k a
sheet and estimated ages of major Arctic Ocean 16 c a l k a
arterial flows or ice streams. The ages 14 c a l k a
are in calibrated radiocarbon years BP 13 c a l k a
(after Stokes & Tarasov, 2010). 35
37 12 c a l k a
11 c a l k a
19 10 c a l k a
160" 36
18
1 9 cal ka
8 2?
11 Baffin 20
8 "3 Bay
2
S 10 21
3
8
10 20 23
5
24 Atlantic
140° 6
Ocean
Hudson
14 Bay 16
17 AO-
15 27 34
33 29
23' 26
10
31
30 32
124°W 0 km 1000
112 W 100" 80-
Advance in the Michigan, Huron and Green Bay areas (Hooyer & Iverson, 2002). Surging appears to have been
(c. 13.5–13.2 ka), and the Marquette Advance into areas less widespread after c. 11.5 ka, by which time the ice sheet
presently occupied by Lake Superior and northern margin had retreated northwards onto the more permeable
Michigan culminating around 11.7 ka (Larson, 2011). This substrates of the Canadian Shield. Further factors
last-named readvance has been traced along a broad promoting surging may have been the degradation of
front from the prairies (Cree Lake Moraine) to Ontario permafrost prior to ice advance, which reduced the shear
(Dog Lake and Hartman Moraines), and along the strength of subglacial materials (Cutler et al., 2000), and the
northern shore of the St Lawrence. There, the prominent termination of many ice lobes in proglacial lakes that
Narcisse (12.8–12.2 ka) and Mars Batiscan Moraines developed around the southern margin of the ice sheet
(12.2–11.5 ka) have been equated with the early and later (Cutler et al., 2001). Although extensive, therefore, land-
stages of the Younger Dryas cooling of northwest Europe form evidence for ice-marginal positions in the Great Lakes
(Occhietti, 2007). region is often difficult to interpret, and recourse must be
In many parts of the Great Lakes region, however, the made to lithostratigraphic and biostratigraphic evidence in
moraine pattern is complex and younger moraines are order to reconstruct the deglacial sequence of events.
frequently found overlapping or cross-cutting older land- To the west and southwest of the Great Lakes, a range
forms. These reflect rapid and irregular changes in the of geomorphological features, including end moraines,
position of the ice margin, driven largely by instabilities outwash fans, meltwater channels and hummocky glacial
within the ice sheet, and which led to repeated glacier topography, has been used to infer former ice-marginal
surges. These may have been caused by deforming sedi- positions, while readvances are indicated by truncating and
ments at the glacier bed, as high pore-water pressures in the overlapping patterns of end moraines and disintegration
low-permeability tills of the Great Lakes region created features, cross-cutting and overriding relationships of melt-
conditions conducive to surging of individual ice lobes water channels and glacial spillways, and superposed till
units (Fullerton & Clayton, 1986). The landform pattern residual ice over Hudson Bay, has again been attributed to
is complicated by ice advances from different directions that repeated surging of the ice margin (Clark, 1994). Subse-
reflect shifting ice divides in the Laurentide ice sheet (Figure quently, two episodes of catastrophic collapse of the ice
2.9b). In the Dakota–Minnesota–Iowa region, for example, sheet (at c. 9 ka and 7.5 ka) led to rapid deglaciation. The
moraines of the earliest glacial advance (>20 ka) from the decaying ice sheet split into three residual masses, those
Hudson Bay/Labrador area are truncated by the Bemis in Keewatin and Labrador Ungava having virtually dis-
Moraine (c. 16.25 ka) formed by ice moving south from appeared by c. 7 ka, leaving only the Barnes Ice cap on Baffin
Keewatin (Lowell et al., 1999). Subsequent fluctuations of Island as the last vestige of the Laurentide ice sheet.
the ice margin, possibly reflecting repeated surging, In the mountains of western North America, the
produced a series of cross-cutting moraines (e.g. the Late Wisconsinan Cordilleran ice sheet covered British
Altamont and St Croix Moraines) that extend throughout Columbia, the southern Yukon Territory, parts of Alaska,
northern Minnesota and northwest Wisconsin. In much of and northern areas of Washington, Idaho and Montana.
Minnesota, the former ice margins are not marked by In its major source areas, the high mountain ranges of
single-crested moraine ridges, but more commonly are the Canadian Cordillera, the ice sheet began to develop
reflected in broad complexes of hummocky topography c. 33–28 ka, and achieved its maximum extent c. 18 ka. Ice
(Jennings, 2004; Jennings & Johnson, 2011). These include sheet decay was rapid and was characterized by complex
the Alexandria Moraine Complex near the centre of the frontal retreat at the periphery, accompanied locally by brief
state, which may have been formed by the interaction readvances, and by downwasting and stagnation (Clague
between four different ice lobes. In southern Wisconsin and & James, 2002). Further south, valley and piedmont glaciers
Illinois, the sequence is even more complicated with the left abundant evidence in the form of lateral and terminal
Lake Michigan ice lobe forming numerous distinctive moraines, outwash terraces and trimlines which mark the
curvilinear moraines with gently hummocky topography, maximal extent of Wisconsinan ice, and also recessional
the morainic pattern delimiting five distinctive sublobes stages and readvances during deglaciation. In the southern
(Figure 2.9c). The maximum southwesterly extent of ice, and central cordillera, for example, nested moraines of
marked by the extensive Shelbyville Moraine, occurred Wisconsinan age show that ice descended from the eastern
around 23 ka, after which ice retreated in an irregular flanks of the mountains into the Great Basin of the western
fashion, marked by numerous recessional moraines (Curry USA, sourced from up to forty individual ranges, plateaux
et al., 2011). or massifs (Osborn & Bevis, 2001). In California, the CRN-
On the Canadian Prairies, the distribution of hum- dated Tahoe Moraines (Figure 2.11) show that many valley
mocky drift and patterns of meltwater channels delimit glaciers in the Sierra Nevada reached their maximum extent
the extent of Late Wisconsinan ice, although geomorpho- in the Early and Middle Wisconsinan, while later ice
logical evidence is absent in many areas and the limits of advances, which terminated further up-valley, are marked
the last ice sheet can only be established on stratigraphic by the Tioga Moraines (Sampson & Smith, 2006). Although
grounds. Large expanses of hummocky moraine suggest an Early Wisconsinan glacial advance may have occurred
widespread stagnation following the Late Wisconsinan further north in the Rocky Mountains of Idaho, Colorado
glacial maximum (Eyles et al., 1999), while till lineations and Wyoming, there are few surviving moraines that date
and subglacial bedform assemblages revealed by remotely to this period, in contrast to the widely developed Pinedale
sensed data provide evidence of ice streaming and major Moraines of the Late Wisconsinan (Pierce, 2003). Moraines
changes in ice-flow configuration (Ó Cofaigh et al., 2009; resulting from glacier readvances that may equate with the
Ross et al., 2009). On the eastern flank of the Laurentide Younger Dryas of Europe include those of the Crowfoot
ice sheet, the extensive lateral moraines of the Saglek Advance in Glacier National Park, Montana (MacLeod et
system, CRN-dated to c. 13.4 ka (Clark et al., 2003), define al., 2006), the Titcomb Lakes Moraine in the Wind River
the upper limits of the Wisconsinan ice in northern Mountains, Wyoming (Gosse et al., 1995), the multiple
Labrador. Numerous end moraine sequences, dated to moraines of the Sumas Drift in the Fraser Lowland of
between 10 and 9 ka (the Cockburn Event) and marking Washington and British Columbia (Kovanen & Easter-
recessional stages of the last ice sheet, have been mapped brook, 2002), and the Finlay Moraines of northern British
throughout the Canadian Arctic and sub-Arctic. These Columbia (Lakeman et al., 2008).
include moraines of the Cochrane Advances of the eastern Some of the most prominent moraine sequences
Canadian Arctic, the Hudson Bay region and the James formed during glacier readvances in the Neoglacial
Bay lowlands. This reactivation of the Laurentide ice sheet, period of the mid- and late-Holocene. In the southern
which resulted in advances and retreats of up to 75 km from Coast Mountains of British Columbia, the earliest of these
Figure 2.11
Wisconsinan terminal moraines dating
to the Tahoe series in Pine Creek
Canyon on the eastern flank of the
Sierra Nevada, near Bishop, California
(photograph by Ron Wolf,
EyeOnNature.com).
date to c. 8.6–8.0 ka (Osborn et al., 2007), while readvances than in the Laurentide ice sheet to the south, with the
from c. 7.5 ka have been reported in the Canadian Rockies absence of deglacial landforms in some areas prior to 11 ka
(Luckman et al., 1993) and from c. 9.5–8.4 ka further BP suggesting that the majority of the retreat post-dates the
south in Washington state (Thomas et al., 2000). In Alaska, Younger Dryas (England et al., 2006). By contrast, retreat
the earliest Neoglacial advances appear to have occurred of the Greenland ice sheet from the West Greenland
much later, around 4 ka (Barclay et al., 2009). Subsequent shelf bordering Baffin Bay (Figure 2.9a) began much earlier
readvances, many marked by well-developed moraines, (c. 14.9 ka), and there is evidence for an extensive readvance
include the Tiedemann–Peyto Advance of the Canadian of outlet glaciers during the Younger Dryas Stadial
Rockies which are dated to c. 3.5–1.9 ka (Clague et al., (Ó Cofaigh et al., 2012b).
2009). Radiocarbon dates and lichen-dated moraines Glacial landforms therefore continue to act as a
suggest that a widespread advance occurred in many regions cornerstone in establishing the extent of former glacier
of the Pacific North American cordillera between AD 400 ice. They are of greatest value in the investigation of recent
and 700 (Reyes et al., 2006), but the most widespread late (i.e. Lateglacial and Neoglacial) patterns of glacier activity,
Holocene advances appear to have occurred during the especially in highland regions where the features are
Little Ice Age, with a number of glaciers achieving their often well preserved and relatively easily mapped, and
maximal Holocene extent at that time (Luckman, 2000; from which the vertical and lateral extent of ice can often
Barclay et al., 2009). In the Coast Mountains of British be reconstructed. For earlier glacial episodes, morainic
Columbia, ten separate moraine-building episodes have landforms in particular can still be used to delimit the
been identified from around AD 1450, with glacier re- glacierized area, and as a basis for mapping of readvances,
advances occurring, on average, every 65 years (Larocque although it is clear that a proper appreciation of glacial and
& Smith, 2003). deglacial sequences rests as much on stratigraphic verifi-
The third of the great ice masses of North America, the cation as on geomorphological evidence. This is considered
Innuitian ice sheet, reached its maximum after c. 22 ka, at further in Chapter 3. Meanwhile, one further line of
which time ice caps that had formed over the mountains evidence is required before ice sheets and glaciers can be
of eastern Ellesmere Island were confluent with the western reconstructed, namely the former directions of ice move-
margins of the Greenland ice sheet, the configuration of the ment, and it is to this aspect of glacial landforms that we
Ellesmere Island and Greenland ice being based on moraine now turn our attention.
and meltwater channel patterns and on erratic distributions
(England, 1999). To the south, the ice sheet was confluent
with the northern sector of the Laurentide ice sheet (Figure
2.3.3 Direction of ice movement
2.9a). Retreat of the Innuitian ice sheet began along In many formerly glaciated areas, a ‘grain’ or streamlined
the southwestern margins around 13 ka, considerably later sculpture is evident in the landscape (Figure 2.3) reflecting
the former direction of ice movement. At a small scale, glaciological conditions have changed over time, or when
bedrock protuberances are scratched (striated), fractured, ice has readvanced into a region. If the later ice advance
polished and grooved, and at larger scales, whalebacks, has a different direction of flow and the striations resulting
roches moutonnées 8 and glaciated valleys (troughs) are from the initial ice advance have not been completely
fashioned by overriding ice. This preferred alignment erased, a second set of striations will become superimposed
of erosional forms in a glaciated landscape is often best upon the earlier ones (Figure 2.12). In some instances, it
seen where ice has emphasized local bedrock contrasts, may be possible to distinguish between sets of cross-
particularly when exploiting the trend of geological striations of different age; for example, in Snowdonia in
weaknesses such as relatively incompetent strata and joint North Wales, striae produced by the last ice sheet can be
and fault lines (Bennett & Glasser, 2009; Benn & Evans, differentiated from those superimposed upon them by ice
2010). Certain glacial depositional landforms, such as of the later Loch Lomond (Younger Dryas) Stadial (Sharp
drumlins and fluted moraines (section 2.3.3.4), may also et al., 1989). In most areas, however, such distinctions are
be aligned in the direction of glacier flow. Careful field difficult to make. Overall, therefore, it would seem that
mapping and remote sensing of landforms that are ice- striations are best used in conjunction with other lines of
directed, therefore, allows the dominant patterns of ice evidence in the reconstruction of regional patterns of ice
movement to be reconstructed. movement.
2.3.3.1 Striations 2.3.3.2 Friction cracks

Striations (or striae) form where stones entrained A range of bedrock fractures or arcuate ‘friction cracks’
within the basal layers of the ice are dragged across bedrock results from stones in basal ice being forced against
surfaces, the size of the indentation being determined underlying bedrock (Glasser & Bennett, 2004). The best
by the load and relative hardness of the stone and the known are ‘crescentic gouges’ and ‘crescentic fractures’.
substrate across which it is dragged. The plotting of striation Crescentic gouges are believed to form concave down-ice
trends is a relatively straightforward field exercise and, and crescentic fractures concave up-ice, the direction of
given the availability of exposed striated bedrock, regional concavity reflecting the former direction of ice movement
directions of ice movement can often be ascertained (Figure 2.13). Consistent patterns do not always emerge,
fairly rapidly. Where the evidence is abundant and has however, and the use of friction cracks in isolation as ice-
been mapped over a sufficiently large area (e.g. Catto, directional indicators now seems a doubtful procedure,
1998), the dominant directions of ice movement become although friction-crack data often lend useful support to
apparent, and the results may reveal local deflections of the reconstruction of regional ice-flow patterns and past
ice flow caused by topographic obstructions and inter- glacier activity based on other lines of evidence (Kelly et al.,
ference between competing ice streams (Figure 2.12). 2004).
In practice, however, the interpretation of striation data is
seldom straightforward. Not all ‘scratch marks’ on bedrock
2.3.3.3 Ice-moulded (streamlined) bedrock
surfaces in formerly glaciated regions have resulted from
the passage of ice. Many may simply reflect lines of An irregular bedrock surface presents numerous obstacles
weaknesses in the rock, accentuated perhaps by subaerial to the passage of ice. This leads to compression of the ice
weathering, while others may have resulted from fluvial, and erosion of the upstream (stoss) sides of the obstruc-
glaciofluvial, snowcreep or avalanche activity. Where tion, whereas plucked and shattered craggy surfaces tend
striations are of glacial origin, they may indicate basal ice to characterize the downstream (lee) sides. A series of bed-
movements determined by local bedrock irregularities. rock ridges running at right angles to the direction of basal
They can frequently be seen, for example, to follow the ice flow will, after prolonged glaciation, be smoothed only
curvature of the face of bedrock protuberances, and on on the up-ice side to form stoss-and-lee landforms, and if
the lee sides (with respect to ice movement) of rock a consistent pattern is evident in the landscape, this can
obstacles are often oblique to the dominant or regional be used to infer ice-flow directions. Where a bedrock ridge
direction of glacier flow. In some areas, diverging sets of runs parallel to the direction of ice flow, the bedrock
striae can reflect more than one direction of ice movement, will become smoothed on the up-ice side and also along
while on certain rock outcrops striations with significantly the flanks to produce the landforms known as roches
different trends may be found crossing or superimposed moutonnées or ‘whalebacks’. The process of smoothing
upon one another. Crossing striations can arise where operates at a range of scales, from micro-features etched on
64' N 64'N
94 w
94 w
2
1 2
Cross 2
Say, 2
2 2
2
2
2
2 Ch 2
22 es
ter
2 fie
22 ld
inl
et
2 2
22 2 2
2
.Chesterfield Inlet
2
2
2
22 2
2
2
2 Hudson
Rankin Inlet
Bay
2
2 22
2
2 2
2 .Chesterfield
.Chesterfield Inlet Inlet
sw SE
Fe
2
rg
•old' 1
us
SSE
on
Ri
Calving bay/
ve
•aid'
r
Whale ESE esker pulling

Cove D km 25
94 w
94 w
62<N 62<N
Figure 2.12 Local ice-flow direction based on mapping of lineations of striations in the Keewatin district of Ninavut, northern
Canada. Crossing striation patterns are indicative of changing ice dynamics, with the older sets reflecting ice advance, while younger
striations result from ice-divide migration and diversion of flow towards melting ice margins in Hudson Bay (after McMartin &
Henderson, 2004).
polished bedrock surfaces, visible only by close observa- glacially moulded terrain (Figure 2.13) and, where strongly
tion, through ‘mega-grooved’ terrain composed of linear, developed, may indicate the courses of palaeo-ice streams,
parallel bedrock undulations of a few metres in vertical the former arterial pathways of fast-flowing ice moving
amplitude (Figure 2.13), to whole mountain sides worn by from source areas towards the ice sheet margins. Ice streams
the passage of ice (Bradwell et al., 2008c). Where localized may also be reflected in patterns of glacial troughs, for
outcrops of particularly resistant bedrock occur, such as while over-deepening of pre-existing valleys commonly
volcanic plugs, these serve to protect the bedrock on the occurs during ice-sheet build-up, troughs continue to
lee side and glacial erosion results in the development constitute the principal avenues of flow within the resulting
of a crag-and-tail feature, the orientation of the often ice sheet. By plotting the trends of glacial troughs, therefore,
drift-veneered ‘tail’ indicating the former direction of ice the major routes taken not only by valley glaciers but
movement. The overall directional ‘grain’ imposed on the also ice streams beneath the former ice sheets can be estab-
landscape is commonly described as glacially sculpted or lished, and centres of ice accumulation and dispersal can
Figure 2.13
Mega-grooved bedrock surface
showing parallel striations and gouge-
marks, including crescentic fractures,
located in the vicinity of the present-
day terminus of the Findelengletscher,
near Zermatt and the Matterhorn,
Switzerland (photograph by John
Lowe).
be identified. The mapping and analysis of ice-moulded outline, usually less than 60 m in height, and are composed
landforms, in conjunction with other ice-directional mainly of till, although some contain stratified material and
indicators such as striations and friction cracks, can there- have a bedrock core (Figure 2.14). They frequently occur
fore provide important insights into former directions in ‘fields’ or ‘swarms’ in lowland areas where there was
of ice flow and positions of ice divides, although the fact little obstruction to the passage of ice, or in piedmont zones
that some large-scale ice-moulded features can survive where flow was radiative or dispersive. Most possess a
more than one phase of glaciation can lead to problems of prominent stoss end with a trailing distal slope. It is gen-
interpretation where the direction of ice movement has erally agreed that the direction of the drumlin long axis
changed between successive glacial episodes. reflects local direction of ice movement with the stoss end
pointing up-glacier, and hence they have been particularly
widely used as ice-directional indicators (e.g. Greenwood
2.3.3.4 Streamlined glacial deposits
& Clark, 2009). Detailed field mapping of drumlins reveals
Subglacial debris deposited beneath moving ice is fre- the dominant local ice-flow paths that prevailed over a
quently streamlined in the direction of ice movement
(Figure 2.3). Streamlined glacial deposits are often found
in glaciated landscapes and range in size from small-scale
fluted moraine with heights of a metre or less (Glasser &
Hambrey, 2001) to larger flutes and drumlins (see below)
sometimes comprising drift accumulations some tens of
metres in thickness (Clark et al., 2009b). These depositional
landforms invariably record the basal movements of the last
ice mass to have affected an area, for any subsequent glacier
with a different direction of movement would have erased
or at least substantially modified such features. Almost all
of the detailed work that has been published on drumlins
and fluted moraines relates to forms that have developed
in drift of the last glaciation.
Drumlins are among the most intensively studied of all
Figure 2.14 A drumlin in the Eden Valley, northwest England,
glacial landforms, with at least 1,300 contributions in the that formed beneath the last British–Irish ice sheet; ice flowed
literature and more than 400 scientific papers since 1980 from right to left across the photograph (photograph by Mike
alone (Clark et al., 2009a). They are low hills with an oval Walker).
enable comparisons to be made between drumlins in

different areas. These studies frequently equate elongation
ratios with relative speed of ice flow during drumlin
formation, a notion that is supported by remotely sensed
images of bedforms beneath the present-day polar ice
sheets (King et al., 2009a). Although most measurements
of drumlin shapes have been made from maps and aerial
photographs, detailed field mapping is essential, for
subtleties in form are not always expressed on topographic
maps, or may not always be clearly identifiable on aerial
photographs.
2.3.4 Reconstruction of former ice masses
2.3.4.1 Ice-sheet modelling

Glacial landforms provide empirical evidence for deter-
mining the margins, upper limits and, to an extent, the
internal architecture of former ice sheets. For example, ice-
directional indicators such as drumlins, fluted moraines,
drift lineations and striations, in association with evidence
from glacial erratics and till provenance studies (section
0 km 100
3.3), enable the principal flow lines of the ice to be estab-
lished, from which major ice-dispersal centres, ice domes
and ice divides can be inferred. Such reconstructions have
Figure 2.15 Lineation of subglacial landform suites (mostly improved substantially in recent years with the develop-
drumlins) formed by the last (Late Midlandian) Irish ice sheet. ment of remote sensing techniques, for these enable many
These enable regional flow-lines (arrows) and principal ice more features and much larger areas to be surveyed than
divides (thick black lines) to be inferred (after Greenwood &
Clark, 2009).
is possible by mapping on the ground. The flow paths
of the last ice sheet in Ireland (Figure 2.15), for example,
have been reconstructed by mapping more than 39,000
particular region and, when the evidence is viewed collec- individual landforms from satellite images (Greenwood &
tively, macro-scale patterns of ice movement frequently Clark, 2009). Large data-sets such as these provide import-
become apparent (Figure 2.15). More than any other line ant baseline information for modelling former glaciers
of evidence, perhaps, maps based on drumlins and other and ice sheets.
streamlined deposits have provided the most striking Models of glacier and ice sheet behaviour are based on
images of the major arterial flow systems of the last great established physical parameters that govern ice flow rate
ice sheets. However, drumlins remain enigmatic features, and surface form. As glaciers increase in thickness, the
for while it is generally agreed that they form in the sub- ice deforms and flows, the rate and direction of which
glacial environment as a response to fluctuating stress and is determined, inter alia, by ice volume, glacier length/
strain conditions within a deforming sediment layer thickness ratio, inclination of underlying topography and
trapped between a rigid bed and overlying mobile ice ice temperature (Ng et al., 2010). Ice will flow from high
(Hart, 1997), precisely how they develop continues to be surface points (e.g. ice domes) to lower areas (e.g. ice
the subject of considerable debate (Benn & Evans, 2010). margins), but if the ice is sufficiently thick, the basal layers
In addition to the regional ice flow trends displayed can be forced to flow uphill, for example over saddles
by their long axes, the overall shape and distribution between adjacent valleys or up-gradient into ice-free valleys.
of drumlins can provide information on former glacier The rate of build-up of the ice, and hence of migration of
dynamics, such as basal ice pressure, rate and type of ice ice margins, will depend on the mass balance of the ice, the
flow, and basal shear stress variations (Mooers, 1989). ratio between the total amount of snow added to the glacier
Numerous shape indices have been developed using axial system and the amount lost through melting and other
and outline ratios (Briner, 2007; Hess & Briner, 2009) that processes (Paterson, 1994; Knight, 1999).
Simulations of the growth and behaviour of glacial ice sheet might be independent from (Figure 2.17b) or
masses can be developed from empirically tested mass confluent with (Figure 2.17a) the Scandinavian ice sheet.
balance and ice-flow equations (Carr & Coleman, 2007), But simply using ice-marginal evidence to constrain the
and for individual cirque or valley glaciers, a simple Excel models may not be sufficient to generate simulations that
spreadsheet model is available (Benn & Hulton, 2010). explain all aspects of the geological evidence for former ice
Numerical simulations of ice sheets are more compli- masses. For example, as we have seen above, the last British
cated, however, since these have multiple ice centres with ice sheet appears to have been confluent with Scandinavian
complex flow paths, and they take much longer to achieve ice at least during the early stages of the last cold stage, and
equilibrium states (Siegert, 2001). Nevertheless, steady-state hence the eastern ice margin cannot be clearly delimited.
models of the last great ice sheets have been developed In addition, geomorphological and geophysical evidence
(Marshall et al., 2002) that integrate the physical behaviour suggest that the ice sheet had a relatively low profile, low
of ice, the underlying topography and assumed climatic summit elevation, and extensive elongated lobes at the ice
variables (e.g. Figure 2.16). In addition, dynamic numerical margin (Figure 2.17). In order to explain these charac-
models that are calibrated against geomorphological and teristics, Boulton and Hagdorn (2006) developed a dynamic
geophysical data (relative sea level, marine limits, present- thermo-mechanically coupled ice sheet model, driven by
day rates of uplift, etc.), enable the inception and growth proxy climate, which simulated the behaviour of the British
of ice sheets, and their behaviour over time, to be simulated ice sheet throughout the last glacial cycle. This showed
(Stokes et al., 2012). In some instances, models can simulate that the ice sheet may indeed have been confluent with
ice sheet behaviour over very long timescales, such as the the Scandinavian ice sheet during parts of its history, and
growth and collapse of the West Antarctic ice sheet over that unforced periodic and asynchronous oscillations
several million years (Pollard & DeConto, 2009). In these could have occurred in different areas of the ice marginal
cases, outputs of the models can be tested against recon- zones creating the distinctive ice lobes. These are indica-
structions based on empirical data (e.g. Stokes & Tarasov, tive of streaming within the ice which drew down ice
2010), and compared with the behaviour of modern ice from the ice sheet interior (Figure 2.18). The interior ice
sheets. on upland surfaces thinned as a result and was largely
Where ice sheet models are prescribed to yield outputs cold-based or immobile, while the ice streams were
that approximate empirically based (field-based) recon- warm-based and flowed at rates of up to 500–1,000 m yr–1
structions, they are referred to as inverse models. In the compared with a surface velocity of 10–50 m yr–1 in
models shown in Figure 2.17, the southern and western interstream zones. The simulations suggest that 60–84 per
margins of the Late Devensian ice sheet (cf. Figure 2.8) were cent of the ice flux was delivered to the margins via the ice
predetermined by the geomorphological evidence for the streams.
positions of the ice margins, and the input parameters were In northern Europe, ice-sheet numerical modelling was
varied to influence simulation outputs. For example, by used to reconstruct a history of the Eurasian ice sheet
altering basal shear stress values, the resulting simulated ice during the last glacial cycle that again was compatible with
INPUTS
initial B e d Topography h B (A, 9, y
BASIC MODEL OUTPUTS
Subgrid Topographic Attributes: Ice T h i c k n e s s H (A.Q, i)

Glaciologic &
h ™ , (*VOI " W ( X © ) , roughness
Geodynamic
B e d Topography h (A,@,t)
Model
B
Initial Ice T h i c k n e s s H (A,S, trf

Ice A r e a A((). V o l u m e V(t)
Integration from
G e o t h e r m a l Heat Flux 0 (A.O)
G
f tot„,at2-5
0
Ice Velocity v{,\,0,z,t)
y e a r time s t e p s
Deform able B e d Fraction • (A,0)
Ice Temperature T(A.0,z,t)
Monthly A i r Temperature a n d
Precipitation, T (A,Q,t), P(A,Q,t)
A
Figure 2.16 Examples of input and output parameters employed in a numerical model simulation of the last Laurentide ice sheet
(from Marshall et al., 2002).
a) Scandinavian b)
ice
00
Assumed zone
0
12
of c o n f l u e n c e
80
00
16
00
18
1700
0m
20
0
100
750,
750,
W i c k low 750,
ice cap
W e l s h ice c a p
0 100km 0 100km
Figure 2.17 Inverse models of the last (Late Devensian) British ice sheet. For a) the basal shear stress is assumed to be 100
kPa9 throughout, while in b) the basal shear stress is 70 kPa on the modern land and 30 kPa outside the modern coastline (from
Boulton & Hagdorn, 2006).
a range of geological datasets, including evidence for ice- the Gulf of Bothnia, where the surface altitude was around
sheet limits, sedimentary records of palaeo-ice streams 2,000 m (Arnold & Sharp, 2002).
and uplift information relating to ice-sheet configuration Models such as these not only provide insights into
and the overall pattern of deglaciation. The result was a the dynamic nature of ice sheets, but also the scale of their
quantitative assessment of the behaviour of the ice sheet, environmental impacts. As ice sheets grow in size, they
its mass balance and climate, and predictions of glacio- influence local wind patterns, changing the supply routes
logical outputs, including sediments, icebergs and melt- for snow, and ultimately leading to migration of ice divides.
water (Siegert & Dowdeswell, 2004). In North America, a In the case of the Laurentide ice sheet, for example, the ice
modelling exercise of the Laurentide ice sheet by Marshall domes were sufficiently large to prevent snow-bearing
et al. (2002) generated 190 different simulations for a range winds from reaching the original source areas. Moisture-
of combinations of input variables (Figure 2.16), and run bearing air masses would have been forced to rise over
over a model period of 17 ka, which is the approximate time the expanding ice on lower ground, depositing more snow
taken for the last ice sheet to reach its greatest extent at the than previously, before reaching the mountains. A positive
LGM. Of these, thirty-three simulations closely resemble feedback mechanism accentuates this situation, creating a
the geomorphologically determined LGM limits (Figure ‘snow shadow’ (equivalent to rain shadow) effect that
2.19), when the ice volume is estimated to have been increasingly isolates the original snow gathering grounds.
between 28.5 and 38.9 × 1015 m3 (Marshall et al., 2002). Both Ice-sheet models also provide insights into past sea-level
empirically based and simulated models confirm the changes resulting from ice-sheet melting. For example, a
locations of maximum ice thickness at the LGM. For the model simulation of the Greenland ice sheet, forced by data
Laurentide ice sheet, this was over Hudson Bay, with a from an independent climatic model, and constrained
surface altitude in excess of 3.5 km (Figure 2.19), while the by empirical evidence from Greenland ice cores, indi-
Fennoscandian ice sheet was thickest in the western part of cates that Greenland ice melt contributed at least 0.6 m to
348° 350° 352° 354° 3 5 6 ' 3 5 8 ' 0' 2° 346° 350° 352° 354° 356° 358° 0° 2"
b)
61' 61
61° 61
60° 60'
60° 60
59° 59°
59° 59"
58° 58°
58° 58¬
57° 57°
57" 57"
56' 56°
66° 86°
65' 55°
55° 55
54' 54'
54° 54¬
53' 53°
53° 53
52' 52'
52° 52°
51' 51'
51° 5V
50° 50°
60° 50°
s t a n d a r d B f i t a i n . c o n y 2 . ela£ -20.00ka standard Britain. cony2. ela2 -20.00ka
348° 350* 352° 354' 356' 358' 0' 346" 350° 352° 354° 356° 359° 0°
-50 -40-35-30-25 -20-15-10 -5 -1 0 1 0 10 20 501001502002505007501000

homologous basal ice temperature horizontal basal velocity
Figure 2.18 a) Simulated basal ice temperature and b) horizontal basal velocity of an ice sheet generated by proxy climate inputs
over a model-equivalent period of 17 ka, considered equivalent to LGM conditions. See text for further explanation (from Boulton
& Hagdorn, 2006).
sea-level rise during the Last Interglacial (Stone et al., imposed streamlined landforms suggest migrations of the
2013). main ice divide by up to 500 km (McMartin & Henderson,
Numerical reconstructions of ice sheets may also help 2004).
explain many of the observed geomorphological charac- In addition to reconstructing the Quaternary ice sheets,
teristics of glaciated landscapes. Of particular signifi- models have been developed that simulate the behaviour
cance are ice divides, those parts of a former ice mass of the present Antarctic and Greenland ice sheets (Payne,
where flow was radiative or dispersive. Erosion is believed 1999; Denton & Hughes, 2002). Climatic and glaciological
to be minimal in such areas, and hence ancient land surface data from contemporary glaciated regions provide modern
features, such as tors, may be preserved despite the fact analogues for the last great ice sheets at the LGM. Such
that they were covered by considerable thicknesses of ice models of the polar ice sheets can help clarify the loca-
(Phillips et al., 2006; Goodfellow, 2007). However, non- tion of ice streams (Bamber & Rivera, 2007), the processes
eroded pre-glacial features tend to be preserved only in leading to the development of subglacial lakes (Siegert,
those areas where ice divides remained fixed throughout 2005) and the transient positions of grounding lines (where
a glacial cycle. Where ice divide migration occurred, as the ice margin remains in contact with the substrate
the ice sheet models predict for parts of the Laurentide ice beneath, rather than floating), an important consideration
sheet, no areas seem to have escaped the effects of erosion. for mass balance calculations (e.g. Conway et al., 1999). The
In Keewatin, for example, crossing striations and super- models also provide a basis for estimating future glacio-
a! I.-;
Ice shveet surface (m)
3950
3300
7750
IBM
IBM
1133
:no
-
a! left shfral surface (m) a!

3B5C
330C
2750
m>c
• 850
'100
550
"
Figure 2.19 Four numerical simulation models of the ice sheets over North America at the LGM based on different combinations
of input variables. In each case, the ice margin is fixed, but different model parameters generate different ice-sheet configurations
reflected, for example, in the number and locations of the main ice domes (after Marshall et al., 2002).
logical changes, particularly under scenarios of human- 2.3.4.2 Ice caps and glaciers
induced global warming (Huybrechts, 2006). Modelling
results suggest that the ice sheets are more sensitive to More detailed reconstructions are possible for smaller
warming than previously thought (Alley et al., 2005), ice caps and for glaciers that were restricted to cirque and
inducing fears of accelerated sea-level rise (Overpeck et al., valley situations. These include many of the Loch Lomond
2006). The modelling results can be compared with data Readvance glaciers of Britain (Benn & Ballantyne, 2005),
from satellite measurements. In the period 1992–2002, for the Younger Dryas and early Holocene glaciers and ice caps
example, these show that the Greenland ice sheet thinned of Scandinavia (Bakke et al., 2005) and the Neoglacial
at its margins, the West Antarctic ice sheet also lost mass, glaciers that developed in the mountains of both the
while the East Antarctic ice mass gained a small amount Northern and Southern Hemispheres (Klok & Oerlemans,
(Zwally et al., 2005). Subsequent measurements between 2004; Thompson Davis et al., 2009). The shape of the
2002 and 2005 by the Gravity Recovery and Climate former glacier margin can be traced by joining those points
Experiment (GRACE) satellite monitoring system have or areas where clear ice-marginal evidence (e.g. terminal
confirmed these trends, the data suggesting an overall ice or lateral moraines, trimlines or valley side drift-limits) is
loss equivalent to a sea-level rise of 0.47 mm ± 0.1 mm a–1 preserved (Figure 2.4b). However, while glaciers often
(Ramillien et al., 2006). These and other data suggest that leave abundant depositional evidence in the lower ablation
ice sheets are more dynamic than previously believed, zones, there is little in the higher accumulation zones, and
and that, valuable as they are, numerical models do not at if trimline evidence is unavailable, then extrapolation
present capture this dynamism adequately (Bamber et al., between scattered ice-marginal indicators becomes
2007). It is important that improvements continue to be necessary in order to delimit the former glacier. When the
made in glaciological modelling, for ice sheets play crucial glacier outline has been reconstructed, ice-surface contours
roles in the global environmental system, as we shall see can be inferred by analogy with typical contour patterns on
again in Chapter 7. present-day glaciers (Figure 2.20). These are commonly
Figure 2.20
Reconstruction of the Loch Lomond
(Younger Dryas) Stadial icefield in the
Pass of Drumochter area, Scottish
Highlands, based on the type and
distribution of the glacial
geomorphological features shown in
Figure 2.1 (from Benn & Ballantyne,
2005).
Forms glacier rnargn
Fame* ice-dammed bfce
Contours m SO metre
Intervals on reconstructed
Contour
Contours at
50 metre rtorvata
Mountain sumrrils
In metres
n metres n metres
normal to valley walls near the median altitude of a valley consolidated granular snow (firn) recedes on surviving
glacier, and become progressively more convex towards the a full summer season’s melt (Figure 2.21). In fact, the two
glacier terminus and more concave towards the upper are not quite the same10 but for the purposes of the present
reaches (glacier source). In practice, contour drawing is discussion, the equilibrium line altitude (ELA) and firn
constrained by features indicating direction of ice move- line altitude (FLA) can be taken to be synonymous. Once
ment, such as striae and fluted moraines, for contours tend the ELA/FLA has been established for individual glaciers,
to be normal to direction of ice movement. the regional firn line (or snowline) can then be recon-
Once the ice-surface contours have been drawn, the structed either by averaging the altitudes from individual
altitude of the equilibrium line can be estimated. The equi- glaciers, or by means of trend-surface analysis.
librium line is the line on a glacier separating the accumu- A range of approaches has been employed to estimate
lation area, the area where the glacier gains in mass, from the altitude of the ELA or FLA on reconstructed glaciers.
the ablation area where a net loss of mass occurs. A term These include the balance ratio (BR), accumulation area
that is often used synonymously with equilibrium line is ratio (AAR), maximum elevation of lateral moraines
firn line, which is the altitude on a glacier surface to which (MELM), toe-to-headwall altitude ratio (THAR), toe-
Figure 2.21 ASTER images (north to right) of the Chapman Glacier (left) and adjacent piedmont glaciers, Nunavut Territory,
Ellesmere Island, northern Canada. During the summer months, the positions of glacier firn lines (FLAs) are shown by marked
contrast between debris-laden ice in the lower ablation zones (darker ice surfaces) and the snow-covered accumulation zones in
the upper areas of the glaciers. The FLAs approximate the positions of the equilibrium-line altitudes (ELAs) of the glaciers. Note
how the lines of debris (medial moraines) reflect flow patterns within the ice (NASA image dated 29 July 2000 from NASA web
site, http://www.jpl.nasa.gov/news/news/2009–059, reproduced with permission of NASA/GSFC/METI/ERSDAC/Japan Space
Systems, and U.S./Japan ASTER Science Team).
to-summit altitude (TSA), cirque floor altitude (CFA) and 2005). A typical AAR value of between 0.60 and 0.65 is
area-altitude balance ratio (AABR). Some incorporate an normally assumed when estimating the ELAs of former
aspect of glacier mass balance assessment (BR, AABR, glaciers in mid- to high-latitude regions, that is, the accum-
AAR, MELM), while others approximate the elevation of ulation area accounted for 60–65 per cent of the total
an attribute of the glacier catchment, because the size of the glacier area. ELAs can be computed from maps or from
glacier itself is not known (THAR, TSA, CFA; Benn & photographs where the altitudinal distribution of the
Lemkuhl, 2000). The latter are less precise and are referred former glacier surface can be measured, and this approach
to as ‘glacier elevation indices’. has been employed to establish the ELA of Late Pleistocene
The most widely employed and easily accomplished is and Holocene glaciers in many parts of the world (e.g.
the calculation of an AAR, which is the ratio between Porter, 2000). On present-day glaciers, the colour contrast
the accumulation area and the total area of the glacier. The in late summer between the ablation and accumulation
AARs of modern steady-state glaciers in the mid- to high surfaces (Figure 2.21) enables the AARs and ELAs to be
latitudes can range between 0.5 and 0.8, but most com- monitored by remote sensing (Bamber & Rivera, 2007), and
monly lie between 0.55 and 0.65; by contrast, the AARs of the links to contemporary climatic parameters provide
glaciers in the humid tropics tend towards the upper part the necessary modern analogue for palaeoclimatic
of that range, whereas debris-covered glaciers in the reconstruction (section 2.3.5.2).
Himalayas can have AARs as low as 0.2–0.4 (Benn et al., Although the AAR method provides a useful first
approximation, it is limited because it takes into account balance of the glaciers that occupied the cirques is not
only the plan dimensions and not the three-dimensional known. However, they offer first-order estimates of the
geometry of the (inferred) glacier. Sissons (1974) intro- average ELA reduction in formerly glaciated regions and
duced a procedure for taking glacier hypsometry into hence are still widely employed (Munroe & Mickelson,
account, based on the area-weighted mean altitude 2002; Lachniet & Vazquez-Selem, 2005). The precise
distribution of the reconstructed glacier surface, using the relationship between the ELA of a small cirque glacier and
following formula: the altitude of the cirque is difficult to determine, though
a close correspondence is generally assumed. Since the

n
Aihi regional snowline approximates the ELA of cirque glaciers
x= i =0
(the exact relationship can be measured for any particular

n
Ai
i =0 area), the latter can be estimated for times in the past from
measurements of the average altitude of cirque floors. The
where x = the altitude of the firn line in metres; Ai = the
alternative index, THAR, is based on the premise that
area of the glacier surface at contour interval i in km2;
ELAs are located a certain fraction (normally 0.4) of the
hi = the altitude of the mid-point of contour interval i;
vertical distance between the headwall and the palaeo-
and n = the number of contour intervals. This formula
has since been widely employed to calculate equilibrium glacier ‘toe’ or inferred terminus. Where the average annual
ELAs of reconstructed Loch Lomond (Younger Dryas) temperature is known for the altitude of the present
Stadial glaciers in many parts of the British Isles (section snowline, the temperature reduction required to lower the
2.3.5). However, the method assumes a linear correspond- snowline to the altitude of cirque floors presently devoid
ence between the accumulation and ablation gradients, of glaciers can be calculated from regional temperature
whereas studies of modern glaciers have shown that the lapse rates. This approach suggests that, at the LGM,
ablation gradient usually tends to be steeper than the regional ELAs and mean temperatures, respectively, were
accumulation gradient (Oerlemans, 2001). The AABR depressed by 900 ± 135 m and 5.4 ±0.8° C in tropical alpine
method overcomes this problem by allowing variable mass regions (Porter, 2000), 1,100 m and c. 8–9°C in Northern
balance gradients to be used in the computation of the ELA Baikal, Russia (Osipov, 2004) and 1,100 m and c. 7–8°C in
(Rea, 2009). Ballantyne (2007a) used the AABR method to Tasmania (Mackintosh et al., 2006).
calculate the ELAs of ten reconstructed Loch Lomond There are, however, a number of problems in the use
Stadial glaciers on the Island of Harris, Scottish Hebrides of cirque floor altitudes as a basis for palaeoclimatic
(location 3 in Figure 2.23b) and found that the Sissons’s reconstruction. First, the snowline associated with cirque
(1974) equation tended to over-estimate the ELA by and valley glaciers is normally lower than the average
c. 25–30 m. altitude on exposed summits and slopes, since wind-drifting
and low insolation protect accumulated snow and ice
within cirque basins. Second, the method can only provide
2.3.5 Palaeoclimatic inferences using a means of estimating the snowline elevation, and hence
former glacier elevations palaeotemperatures, for the time when the glaciers were
Expansion of the ice masses in mid-latitude regions resulted confined to cirques or individual valleys and it is not
from a combination of reduction in temperatures and suitable for calculating temperatures during periods of
increased snowfall which, in turn, caused widespread more extensive ice cover. Third, not all cirques in an area
lowering of regional ELAs. If regional ELAs can be estimated were occupied by ice at the same time, and since some
for times in the past, then by studying the relationships variation in the altitude of cirque glaciers can be expected,
between present-day climatic parameters and ELAs of contemporaneity of cirque glacier development may be
contemporary ice masses, former temperature regimes difficult to establish. Finally, it is an implicit assumption in
and, in some cases, seasonal or annual precipitation values the lapse-rate calculations that the same annual accumu-
can be inferred. lation at the ELA of a modern glacier in a mountain region
occurred at the ELA of Late Pleistocene glaciers, which
may have been several hundred metres lower. However,
2.3.5.1 Cirque floor altitude (CFA) and toe-to-
empirical data show that precipitation in mountain regions
headwall (THAR) methods
typically decreases with decreasing altitude, and hence the
As mentioned above, the CFA and THAR methods provide precipitation on the lower-altitude Pleistocene glaciers
relatively imprecise estimates of ELAs, because the mass would also have been lower.
2.3.5.2 ELA/FLA method reasonable limits, the appropriate value for the other
variable can be calculated. Similar relationships have been
An alternative, and possibly more accurate, approach calculated for modern glaciers elsewhere, which demon-
for deriving temperature estimates from reconstructed strates that glaciers worldwide are restricted to a relatively
glaciers involves the use of present or recent glacier mass narrow range of combined mean accumulation and mean
balance/climate relationships as an analogue for past summer temperature conditions (Ohmura et al., 1992).
glacier/climate conditions. The ELA on a glacier is deter- For reconstructed Quaternary glaciers, therefore, reference
mined by both seasonal precipitation and temperature to such data allows summer temperature values to be
regimes within the glacier catchment. The mean summer estimated where the altitude of the former equilibrium
ablation season temperature (t) on modern glaciers is line and associated accumulation data are known. On the
closely related to accumulation at the equilibrium line Isle of Mull in western Scotland, for example (locality 6
(Figure 2.22a) which, in turn, approximates average in Figure 2.23b), geomorphological mapping of former
accumulation (A) over the whole glacier (Sutherland, Loch Lomond Stadial glaciers and subsequent application
1984). For ten Norwegian glaciers, Ballantyne (1989) of the AABR indicated a former regional ELA of 250 m
found that this relationship corresponds to the regression (Ballantyne, 2002b). A combination of modern meteoro-
equation logical data and the application the equation above
A = 0.915e0.339t (r2 = 0.989, P = 0.0001) suggested a mean temperature for July at the ELA of
2.5 ± 0.5°C (5.7 ± 0.5°C at sea level) and annual precipi-
The ELA is also linearly related to temperature, which tation values of between c. 2,700 and 3,800 mm (Figure
Sutherland (1984) calculated to be 0.58°C per 100 m. 2.22). In some instances, it may be possible to use inde-
Hence, a rise or fall in ELA can be directly related to a pendent proxy data to infer either former temperature or
change in temperature, or precipitation, or some com- precipitation values. For example, on reconstructed Loch
bination of the two. If either former precipitation values or Lomond Stadial glaciers on the island of Arran in western
former temperatures are known or can be estimated within Scotland (locality 7 in Figure 2.23b), mean July–August
uoiteiidpsjd ienuue ueay\j
3.6 5400 b)
a)
3.4 5100 400 500
700
ma
3.2 4800
Accumulation (A) at the E L A
3.0 4500
(metres water-equivalent)
2.8 4200
331
2.6 .3 3900
E0
24 5 3600
91
0. 700
v i e v i e ie (dVW)
2.2 Best estimate 3300

=
A
i
2.0 3000
1000
1.8 2700 EES Use
1.6 2400 500
(ww) (ww)
1.4 2100
1.2 1800
400
1 0 1.5 2.0 2.5 3.0 3.5 4.0 CS
M e a n a b l a t i o n s e a s o n t e m p e r a t u r e (t) at t h e E L A
300
4.5 5.0 5 5 6.0 6 5 7.0 X

M e a n J u l y t e m p e r a t u r e at s e a l e v e l
Figure 2.22 a) The relationship between mean winter snow accumulation (defined as October–April period) and mean
temperature during the ablation season (May–September) for modern glaciers in Norway. This curve has been used to estimate
the temperature and precipitation values for former glaciers in Scotland. b) Generalized contours (m) for inferred ELAs based on
reconstructed Loch Lomond Readvance (Younger Dryas) glaciers in Scotland (from Sissons, 1980).
temperatures at the regional climate ELA estimated from conditions. This is almost impossible to establish for
fossil chironomid (non-biting midge) records (section individual ‘palaeoglaciers’ but if, as seems likely, lags have
4.5.5) in southeast Scotland lay between 5.7 ±0.1°C and occurred between recent climatic change and glacier
4.1 ± 0.2°C (Ballantyne, 2007b). Empirical relationships response, cirque and valley glaciers must frequently have
between temperature and precipitation at modern glacier been out of synchrony with climate during the cold stages
ELAs indicates equivalent mean annual precipitation at of the Quaternary. Finally, some reconstructed glaciers
the ELA on Arran lay between 2,002 ± 409 mm and have been found to have anomalously low long-profile
2,615 ± 449 mm. gradients due largely, it appears, to the effects of deforma-
There are a number of potential difficulties in the tion of subglacial sediment. Where this has occurred, it may
application of this method, however. First, precipitation/ undermine the assumption of a linear relationship between
temperature ELA graphs have to be employed that are accumulation and ablation in the calculation of former
applicable to the area under investigation, and in the case ELAs (Ballantyne, 1989).
of the British Isles, for example, it has been assumed that Despite these limitations, this approach offers the
the curves calculated for Norwegian glaciers provide the potential for deriving quantitative palaeoclimatic data
closest analogues for the last glaciers in northern Britain. from, in the first instance, glacial geomorphological
This may not necessarily be the case. Second, the method evidence. It is encouraging that in areas such as the British
depends on the establishment of a sound statistical rela- Isles, ELA-based reconstructions exhibit consistent regional
tionship between ELA/FLA and regional climatic data, but glacier-climate trends. From the time that Sissons (1980)
glaciers in many mountain glaciers have been shrinking in first demonstrated a strong SW–NE gradient across Scot-
recent decades and may not be in phase with prevailing land for Loch Lomond Stadial ELAs (Figure 2.22b), subse-
climatic conditions (Solomina et al., 2008), while not all quent research has tended to reinforce this pattern (Benn,
glaciers in a particular area respond to climatic shifts in 1997). For example, ELA reconstructions from sites
exactly the same way (Pelto & Hedlund, 2001). Third, the extending along the western seaboard of the British Isles
approach rests on the assumption that the ‘palaeoglaciers’ show a clear latitudinal gradient (Figure 2.23a), equivalent
reconstructed from the geomorphological evidence were to a northward temperature decrease of 0.42°C/100 km
themselves in equilibrium with the prevailing climatic (Ballantyne, 2007b). This is remarkably close to the present
D i s t a n c e (D) in k i l o m e t r e s s o u t h o f O r k n e y b) L o c a t i o n of S a m p l e s i t e s
a)
800 700 600 500 400 300 200 100 0 1 - Orkney
700 2 - S W Lewis
11
3 - Harris
E L A 2 = 138 + 0.695 D
600 4 - Skye
10' (n=11; r = 0.963)
2 5 -Rhum
E L A 2 (BR = 1.8) in metres
500 9 6 - Mull
8
7 - Arran
400 8 - S.W. Scotland
7
5 9 - L a k e District
300
6 4
7
200
3 10 - S n o w d o n i a
E L A 2 = 4 4 6 3 . 5 - 7 6 . 8 1 4 Latitude 1
100
0 1 = 1 1 : ^ = 0.963) 11 - S . W . Ireland
51 52 53 54 55 56 57 58 59
Latitude fN)
Figure 2.23 Mean equilibrium-line altitudes for reconstructed Loch Lomond (Younger Dryas) Stadial glaciers in eleven districts
of the British Isles shown on the right (from Ballantyne, 2007a).
summer temperature gradient (0.45°C/100 km) between however, it has been used in a much broader sense to refer
the Outer Hebrides and Snowdonia, which lends credence to non-glacial processes and features of cold climates,
to the method and to its underlying assumptions. irrespective of age or proximity to glacier ice (André, 2009).
In addition to providing evidence of past tempera- Periglacial processes and features are generally associated
ture and precipitation, former ELAs offer insights into with zones of continuous or discontinuous permafrost
other aspects of Late Quaternary palaeoclimate. In western (Figure 2.24), that is, ground that is permanently or
North America, for example, where precipitation patterns seasonally frozen. Today continuous permafrost exists
during the LGM varied due to the influence of continental where mean annual air temperature (MAAT) is about or
ice sheets, there has been a degree of uncertainty as to below –6 to –8°C (Smith & Riseborough, 2002). Those areas
the moisture sources for many mountain glaciers. In the that are characterized by cold-climate processes where
Sangre de Cristo mountains of southern Colorado, how- frost action predominates constitute the periglacial domain
ever, the observed ELA pattern is very similar to modern (French, 2007) and are found in both high-altitude and
precipitation patterns, suggesting that southeasterly derived high-latitude regions of the world. They currently cover c.
precipitation had a significant influence on the mass 25 per cent of the earth’s land surface (Anisimov & Nelson,
balances of LGM glaciers (Refsnider et al., 2009). Likewise, 1996), but tracts of permafrost also lie beneath the Arctic
in the Brooks Range of Alaska, the regional west-to-east Ocean and on the northern continental shelves of North
gradient in reconstructed ELAs on LGM glaciers is com- America and Eurasia. The occurrence of fossil or relict
parable to that of contemporary glaciers, suggesting the periglacial landforms and deposits throughout the tem-
same primary moisture source in the North Pacific as perate mid-latitude regions suggests that perhaps a further
today. However, a reduction in ELAs in the northeastern fifth of the earth’s land surface was affected by cold-climate
part of the range may point to a secondary source of processes on occasions during the Quaternary (French,
moisture in the Beaufort Sea (Balascio et al., 2005). In the 2008).
European Alps, analysis of ELAs during the Younger Dryas
period shows the greatest depression relative to modern in
the west and northwest. This points to a more zonal type
of atmospheric circulation than that which characterizes the
region today, an inference that accords with data from other
proxies, as well as with atmospheric circulation models
(Kerschner et al., 2000). In western Norway, variations in
glacier ELAs during the Holocene suggest a close corre-
spondence between glacier size and storm-track variability
over the North Atlantic, which regulates winter precipi-
tation levels (Bakke et al., 2008).
Spatial patterns and altitudinal trends exhibited by past
ELAs therefore provide a basis for palaeoclimatic recon-
structions. Inferred palaeotemperatures often show a broad
measure of agreement with estimates from other proxy
data (section 4.5 and Chapter 7), as well as with contem-
porary climatic records. They may also offer a basis for
distinguishing between those changes in mass balance of
modern glaciers that are due to natural climate variability,
and those that reflect more pervasive, anthropogenically
forced climatic warming (Owen et al., 2009).
F r o s t Index S c a l e
0.0 0.5 0.6 0.67 1.0

2.4 PERIGLACIAL LANDFORMS
The term ‘periglacial’ was first used by the Polish geologist Figure 2.24 Contemporary permafrost distribution in the
Walery von Lozinski in the early years of the twentieth northern hemisphere based on modern climatic data and
computed ‘frost index’ values (FIVs) which define the southern
century to describe both the climate and characteristic limits of permafrost zones, as follows: continuous permafrost
cold-climate features (landforms and sediments) found in – FIV of 0.67; extensive – FIV of 0.60; sporadic – FIV of 0.50
areas adjacent to the Pleistocene ice sheets. Since then, (from Anisimov & Nelson, 1996).
The periglacial domain is characterized by an extremely bedrock by frost action and the transport of frost-riven
active geomorphological environment in which processes material by gelifluction produce altiplanation or cryo-
operating on the ground surface include: planation terraces (Thorn & Hall, 2002). Footslopes, by
contrast, tend to develop smooth, low-angled profiles, with
(a) frost-shattering of bedrock and of particles within the accumulation of gelifluction sheets and terraces
unconsolidated sediments; (Kneisel, 2010). The latter are typically lobate in form and
(b) the growth of ground ice, leading to upheaval of the evolve through the combination of creep induced by frost
ground surface, and the lateral displacement of surface heave and the downslope movement of saturated surficial
materials; debris. Modern lobes are most active under high pore-water
(c) accelerated wind erosion and transport in environ- pressures in spring, when snow melts over frozen substrates
ments where vegetation cover is sporadic and un- (Harris et al., 2008), inducing down-slope movements of
consolidated materials are exposed over wide areas; up to 63 mm a–1 (Ridefelt et al., 2009).
(d) thermal erosion by fluvial activity; On valley floors, on hillside benches, and on plateau
(e) accelerated downslope movement of materials surfaces, a number of features occur in geometric patterns,
(gelifluction) where near-surface thawing results in a referred to generally as patterned ground (Ballantyne &
saturated surface layer overlying a still-frozen substrate, Harris, 1994). This term covers landform assemblages that
resulting in mass flow on slopes with angles as low have been produced by a variety of processes. Perhaps the
as 2°. best known is that of the ice-wedge polygon (Figure 2.26a),
which can develop either through ground cracking at very
Some of these processes are unique to periglacial environ- low ground temperatures (Christiansen, 2005), or by
ments, most notably those associated with the growth of sorting processes, where coarser materials are selectively
ground ice, while others, including fluvial, aeolian and separated from finer particles. More common, however,
gelifluction processes, are particularly effective in high- are sorted nets and circles (Figure 2.26b). Sorted stripes
latitude and high-altitude regions of the world. Collectively, tend to form where sorting processes operate on gentle to
they give rise to a suite of landforms that is highly distinc- moderate slopes up to about 25°. Patterned ground may
tive and that is characteristic of a periglacial landscape also be found, however, on mountain summits, on geli-
(Ballantyne & Harris, 1994; French, 2007). fluction terraces and on altiplanation terraces. Regular
On upper slopes, exposed bedrock is highly frac- depressions reflect the original locations of ground ice
tured, angular and craggy in appearance as a result of frost lenses (palsas), which accumulate in substrates where
action, with upstanding masses of bedrock (tors: Figure groundwater is preferentially diverted under high pore-
2.25a) and frost-weathered mountain-top debris (block- water pressure during ground freezing. Following thaw,
fields: Figure 2.25b). A characteristic step-like, hillslope subsidence occurs and, in present-day arctic environments,
profile typically evolves in which a process of excavation of the resulting hollows are normally filled with water (thaw
a) b)
Figure 2.25 a) Granite tor on the summit of Beinn Mheadhoin, Cairngorm Mountains, Scotland. Also included is Colin Ballantyne,
University of St Andrews, UK, who provided the photograph. b) Summit blockfield on Gros Morne Mountain, western Newfoundland,
Canada (photograph by Mike Walker).
a) b)
Figure 2.26 a) Active polygonal patterned ground, Mt Edziza, British Columbia, Canada (photograph by Neville Alley, University
of Adelaide, Australia). b) Sorted circles on the summit of Gros Morne Mountain, western Newfoundland, Canada (photograph by
Mike Walker).
lakes). The landscape is commonly described as thermo- Some landforms, however, may form the basis for more
karst (Luoto & Seppälä, 2003), from a visual analogy detailed palaeoclimatic reconstructions, and these are
with karst regions, which are often marked by numerous considered in the next section.
sink-holes. Thermokarst is normally associated with the
degradation of continuously frozen ground. Finally, there
2.4.1 Palaeoclimatic inferences based on
are the polar deserts, which occupy the driest parts of the
periglacial domain, such as in the McMurdo Dry Valleys
periglacial landforms
of Antarctica today, where extensive sand sheets and dune Certain periglacial landforms are unique to present-day
systems occur (Speirs et al., 2008). arctic and alpine environments, and aspects of the pre-
Many of these features of the periglacial landscape have vailing climate under which they have evolved can
been recognized in relict form in Eurasia and North sometimes be quantified. Where comparable relict features
America, either through field mapping or through the can be identified, therefore, they can be used as a basis for
use of remote sensing techniques (Grosse et al., 2005), and estimating climatic parameters for earlier times during the
reflect the extensive areas of continuous and discontinuous Quaternary. Examples of periglacial landforms that have
permafrost that developed during the cold stages of the been employed in this way are rock glaciers, pingos and
Quaternary. Patterned ground forms, for example, can palsas, and protalus ramparts. The potential of periglacial
frequently be identified on aerial photographs, for the geo- deposits for palaeoclimatic reconstruction is discussed in
metrical patterns tend to be emphasized by differences in section 3.4.3.
crop growth resulting from drainage variations. They are
particularly clear on flat alluvial plains or on the coastal
2.4.1.1 Rock glaciers
terraces of northwest Europe (Svensson, 2005), and on the
Late Wisconsinan drift plains of North America (Lusch These are active tongue-shaped or lobate accumulations
et al., 2009). The tracts of former polar deserts are marked of rock debris (Figure 2.27) that move slowly downslope
by the extensive dune fields and ‘sand belts’ that stretch as a result of deformation of interstitial ice or frozen
across the northern European plains, the northern belt of sediment, with rates of movement ranging from 1 cm a–1
the Great Plains of North America, and large parts of to greater than 130 cm a–1 (Giardino & Vitek, 1988). Two
northern China and Mongolia (Yang et al., 2004). Relict main types can be distinguished on the basis of their
desert landforms often enable wind patterns at their time characteristic length/breadth ratios (Hamilton & Whalley,
of formation to be inferred. Patterned ground and polar 1995). Rock glaciers sensu stricto (also termed morainic
desert features therefore provide good evidence for the rock glaciers), are formed through the burial and subse-
former existence of periglacial conditions, although in quent incorporation of a core of glacier ice under a thick
most cases they only allow the most generalized of climatic cover of morainic debris. These are frequently found
inferences to be made (Humlum & Christiansen, 2008). spreading down-valley from a cirque glacier, but might also
2.4.1.2 Pingos and palsas

Pingos are dome-shaped mounds or hills that occur in
permafrost regions as a result of the uplift of frozen ground
by the growth of a large mass of ground ice in the sub-
stratum, and are prominent periglacial landforms in vast
regions of the Arctic and Subarctic (Grosse & Jones, 2011).
Two broad types of pingos have been recognized: the
‘closed-system’ (hydrostatic) or Mackenzie Delta type that
form by groundwater expulsion during permafrost aggrada-
tion; and the ‘open-system’ (hydraulic) or East Greenland
type, in which the ice core is fed by subsurface groundwater
seepage from unfrozen parts of discontinuous permafrost.
Smaller cryogenic mounds also occur in the discontinuous
and sporadic permafrost zones, the most common of which
Figure 2.27 Active lobate rock glaciers near Lytton, British are palsas and mineral palsas, the former developing in peat
Columbia, Canada (photograph by Neville Alley, University of while the latter form in mineral soils. Mineral palsas are also
Adelaide, Australia).
known as lithalsas (Pissart, 2002). Melting of the ice core
within a pingo or palsa leads to ground collapse, forming
be initiated by avalanches which concentrate large amounts a central depression or crater, with a characteristic ring-
of rock debris (Humlum et al., 2007). Protalus lobes, shaped rampart around the depression (Mollard, 2000).
sometimes referred to as ‘protalus rock glaciers’, develop Relict pingo ramparts have been recognized in many
through the deformation of thick accumulations of talus mid-latitude regions where periglacial conditions once
in ice-rich permafrost regions, and are nearly always found obtained and, by using prevailing climatic conditions in
at the foot of steep cliffs, where they form step-like or lobate areas where pingos are currently forming as an analogue,
extensions of the lower parts of talus slopes (Harrison et al., MAATs at times of former pingo growth may be inferred
2008). Protalus lobes and rock glaciers are characteristic (Ballantyne & Harris, 1994; French, 2007). For example,
features of the discontinuous permafrost zone in mountain relict pingo scars and other periglacial indicators suggest
regions, and their formation appears to be governed by a MAATs at sea level of around or below –8°C in Fenno-
threshold MAAT of –2° (Brazier et al., 1998). Hence, in a scandia, northern Britain and Ireland during the Younger
mountain region where relict protalus rock glaciers are Dryas Stadial, while in the Netherlands, upland Belgium,
found at lower elevations than modern ones, the degree northern Germany and Poland, MAATs are estimated to
of cooling required to generate the relict features can be be in the range –8 to –1°C (Isarin, 1997). The presence of
estimated from regional temperature lapse rates. This lithalsas in upland Belgium, Wales and Ireland adds support
line of reasoning suggests a reduction of mean annual to these reconstructions, as the features are found today in
temperature by 2–4°C in the Swiss Alps during the Late- northern Quebec where MAATs are in the range –4 to –6°C
glacial period (Frauenfelder et al., 2001) and by 8–9°C in (Pissart, 2001). Modelling of palsa distribution in northern
northwest Greece during the LGM (Hughes et al., 2003), Europe suggests the optimum areas for their development
while shorter episodes of climate cooling during the are those with a MAAT of –3 to –5°C and with low levels
Neoglacial period have also been inferred from relict rock of precipitation (< 450 mm) (Luoto et al., 2004).
glaciers in the Front Range mountains of Colorado There are two important limitations to the use of
(Refsnider & Brugger, 2007). However, this type of evidence this approach, however. First, while both pingos and
often needs to be supported by independent proxy records, palsas occur in areas of discontinuous permafrost (MAAT
as climate is not the only variable that directly affects rock ⭐ –3°C), pingos are also features of the continuous perma-
glacier formation and dynamics (Harrison et al., 2008; frost zone (⭐ –7°C), while palsas are found in regions of
Janke & Frauenfelder, 2008). Moreover, attribution of sporadic permafrost (⭐ –1°C). Hence, appropriate con-
debris accumulations in upland areas to relict rock glaciers temporary analogues for relict features may be difficult to
is often contentious, as many such features can be explained determine (Gurney, 2003). This problem may be com-
without recourse to forward movement controlled or pounded by the difficulty of classifying relict ramparted
facilitated by incorporated ice as it deforms and melts out depressions in former periglacial regions as either pingos
(Jarman et al., 2013). or palsas. Second, the remains of these ground-ice features
display a variety of forms, from isolated ridges to clustered where snow survives the ablation season but where
collectives of ramparts, the often complex structures of accumulation is insufficient to promote the development
the latter suggesting not only polygenetic origins (Mollard, of glacier ice, the snowbed being partly nourished by direct
2000), but that factors other than climate may have influ- precipitation input, but also often by redistribution of
enced their development. snow from surrounding mountain and plateau areas (Brook
& Williams, 2012). Their survival is governed partly by local
temperature regime, but the primary control on their
2.4.1.3 Pronival (‘protalus’) ramparts development appears to be precipitation, since the
A pronival (protalus) rampart, also termed a nivation accumulation of too much snow would result in rapid snow
ridge, is a lobe, ridge or ramp formed of debris that patch growth and the transition to glacier ice (Ballantyne
accumulates along the downslope margins of a perennial & Benn, 1994). Fossil protalus ramparts (Figure 2.28),
snow patch. The term ‘pronival’ is now generally preferred therefore, mark the positions of former snow patches that
to ‘protalus’, because the presence of a snow patch is accumulated under colder temperatures than today, but in
essential to ridge formation. Originally it was assumed conditions marginal for glaciation. In the uplands of the
that the ridge was largely composed of frost-riven bedrock British Isles, for example, such features appear to be mostly
from above the snow patch, and that build-up of debris of Loch Lomond/Younger Dryas Stadial in age, and in
occurred mainly over the distal slope of the ridge. However, Scotland their altitudinal range is very similar to that of
analysis of both active and fossil features has shown that reconstructed glacier ELAs (section 2.3.4.2), with protalus
rounded as well as angular clasts occur within ramparts, ramparts increasing in altitude from west to east (Ballantyne
suggesting a variety of transport processes (Shakesby, 1997), & Kirkbride, 1986). This appears to reflect heavier snow-
while debris also accumulates on both proximal slopes and fall in western Scotland by comparison with the mountains
rampart crests (Hedding et al., 2007). Perennial snow in the east, a pattern that is also apparent in the ELA
patches develop in sheltered situations on mountainsides reconstructions (Figure 2.29).
Figure 2.28 (left) Fossil protalus rampart (middle distance)

and stratified scree (foreground: see section 3.4.1) near Cader
Idris, Wales (photograph by Mike Walker).
Figure 2.29 (below) Altitudes in metres of the frontal crests

of Loch Lomond (Younger Dryas) Stadial protalus ramparts in
the Scottish Highlands, and reconstructed ELAs of Stadial
glaciers at their maximal extent (see also Figure 2.22) (after
Ballantyne & Kirkbride, 1986).
Foinaven 500, >Ben Hope 475
470
Baosbheinn
An TeaHach 3 1 5
50
Baosbheinn
470
Creag Meagaidh
730 725 3 5 :
9-0 SO 3
750 700
35? s - o
mo
420
47
47
0
0 470
780 Rampart altitudes (m)
274
500-- Finn line altitudes (m)
Ben Bowie
0 SO km 237
The distribution of fossil periglacial landforms, there- 2.5 SEA-LEVEL CHANGE

fore, not only provides evidence for the former extent of
the periglacial domain, but in certain cases their occurrence There is abundant geomorphological evidence in many
can be used to derive palaeoclimatic data for times during parts of the world for fluctuations in sea level during the
the Quaternary. In addition to the three examples discussed course of the Quaternary (Murray-Wallace & Woodroffe,
above, other periglacial landforms used as a basis for 2014). This includes former coastal landforms now standing
palaeoclimatic reconstruction include blockfields (Nelson above present sea level, such as rock platforms sometimes
et al., 2007) and small cryogenic mounds known as earth with prominent backing cliffs (Figure 2.30), ‘raised beaches’
hummocks or thúfur (Grab, 2005). Care must be exer- consisting of estuarine or littoral sand and gravel deposits,
cised in the use of this type of evidence, however, for the deltas, spits, shingle ridges, stacks, caves and coral reefs.
occurrence of fossil periglacial landforms indicates only that Evidence for sea-level change can also be found offshore
certain critical temperature thresholds were transgressed. in the form of submerged landforms, including caves,
It is possible, and in many cases likely, that tempera- platforms, beaches, reefs and river valleys. Mapping and
tures were lower than the threshold values, and hence the altitudinal measurement of such features enables the
derived palaeotemperatures must be regarded as maximal positions of former coastlines to be established and the
values only. The influences of non-climatic variables on the vertical range of sea-level variations to be determined.
development and distribution of periglacial landforms In many situations, however, a complete picture of sea-level
poses further difficulties (Harris et al., 2009). These and change can only be obtained using both geomorphological
other problems associated with the use of fossil periglacial and lithostratigraphic data. Not only do marine deposits
evidence in palaeoenvironmental reconstruction are form a more continuous record of sea-level change than the
considered further in section 3.4. fragmentary geomorphological evidence, many littoral
Figure 2.30 The ‘High Rock Platform’, northern Islay, Scottish Hebrides: this raised marine platform that is cut into quartzite
dominates the picture. Its surface altitude varies between 32 and 35 m above present mean tide level and it extends for between
400 and 600 m from seaward edge to backing cliff. It is locally covered by marine gravels, indicating that it was covered by a
higher marine level after it was formed. It has been etched into by a lower and younger platform of Younger Dryas age (the ‘Main
Lateglacial Platform’) that in this locality is between c. 3 and 5 m above present mean tide level. It has formed the backing cliff
and caves seen in the foreground (a figure in the foreground provides scale) while isolated stacks can be seen closer to the present
shore. A gravel ridge that covers the lower platform (lower right of photo) was formed during the Holocene. This sequence of
shorelines reflects the complex interplay between glacio-isostatic rebound and eustatic sea-level variations during the waning stages
of the last British ice sheet (photograph from http://www.islay.org.uk/tag/mala-bholsa/).
sedimentary sequences contain fossils that provide an movements. The term ‘isostasy’ refers to the state of balance
additional source of information for the reconstruction of that exists within the earth’s crust so that a depression of
sea-level histories. In addition, sedimentological and the crust by the addition of a load (sediment, lava, ice,
biological evidence offers a basis for the dating and cor- water, etc.) in one locality will be compensated for by a rise
relation of sea-level changes, whereas geomorphological in the crust elsewhere (Teixell et al., 2009). The state of
features (platforms, beaches, etc.) are usually more difficult isostatic equilibrium is considered to be maintained by
to date precisely. By combining landform evidence with viscous flow of the mantle, although the precise nature
sedimentary and fossil records from boreholes or exposures, of the processes involved is still a matter of debate (Thorson,
therefore, the sea-level history of a particular locality can 2000). In order to understand Quaternary sea-level
be established, often in considerable detail. In this section, variations, therefore, it is first necessary to establish how
therefore, the geomorphological evidence and the the separate effects of isostatic and eustatic changes have
lithostratigraphic evidence for changes in sea level are affected a region. In more tectonically stable areas (such as
considered together. The biological evidence for sea-level the islands of Bermuda and the Bahamas, for example),
change is discussed in Chapter 4. where the eustatic effect appears to have been the major
factor influencing sea levels, it may be possible to recon-
struct a sequence of what have been termed ‘absolute’ as
2.5.1 Relative and ‘absolute’ sea-level
opposed to relative sea-level changes. Absolute sea levels
changes cannot easily be discerned in areas where crustal move-
The level of the sea relative to the land can vary through ments have occurred, however, as it is often difficult to
the vertical change of either the sea or the land surface or discriminate between the isostatic and eustatic compo-
both of these. Where subsidence of the land takes place nents jointly responsible for movements of the sea
at a time of stable ocean levels, there will be a local rise in surface relative to land. One way to resolve this problem is
sea level; conversely, land uplift will lead to the elevation to compile records of sea-level change for tectonically
of littoral features, and an apparent fall in sea level. Where ‘stable’ coastlines, for these (in theory) should reflect
changes in sea level take place either through land or eustatic variations only, and these data should then enable
through sea-level movements, they are referred to as relative the magnitude of land movement to be calculated for
sea-level changes, that is, a change in the position of the sea tectonically active regions. However, as discussed in the
relative to the land. Such changes are essentially local in following sections, there are important methodological
effect. Worldwide sea-level changes, on the other hand, limitations to this approach. An alternative strategy is to use
that result from fluctuations in the volume of water in the an independent measure of sea-level change, such as the
ocean basins, are termed eustatic. At one time it was oxygen isotope record from marine microfossils in deep-
assumed that the extent of any eustatic change would be ocean cores (Figure 2.31). This type of evidence is
uniform worldwide, but it is now acknowledged that local considered in the following section and is discussed in more
gravitational effects can influence the rates and magnitudes detail in section 3.10.
of sea-level rise or fall in different coastal regions (Milne &
Mitrovica, 2008). In effect, this means that while eustatic
changes may be broadly similar at the local or regional scale,
2.5.2 Eustatic changes in sea level
this cannot be assumed when comparing records at the
2.5.2.1 Pre-Quaternary eustatic changes
continental or global scale (section 2.5.2). Moreover, even
where the amplitude of eustatic change has been the same Major changes in global sea level have occurred through-
in two coastal regions, this need not result in the same out the Phanerozoic.11 Most reconstructions suggest much
geomorphological response. For example, a eustatic rise of higher stands of the sea during pre-Quaternary periods
50 m over a 1,000 year period will result in a relative rise (e.g. Figure 2.31), although estimates of the magnitude and
in sea level of 50 m in areas where the land surface is stable, timing of high sea levels vary considerably (Miller et al.,
but a 50 m fall in relative sea level (RSL) along coastlines 2005; Müller et al., 2008). The discrepancies arise because
being uplifted at a rate of 10 cm per year. long-term sea-level trends are difficult to establish, as they
Some land movements are long term and result from reflect, inter alia, continental-scale and localized tectonic
tectonic activity associated with the migration of the effects; changes in mass distribution and shape of the earth;
great lithospheric plates across the surface of the globe. changes in the volume and mass of the hydrosphere
Others may be of shorter duration and are generally more through the addition of juvenile water;12 and temporal
localized in their effects; these are known as isostatic variations in the rate of rotation or in the axial (angular)
2.5
50
3.0
0
3.5
Sea Level (m)
OblS
4.0
-50-
4.5
-100-
5.5
Quaternary Pliocene Miocene
0 1 2 3 4 5 6 7 8 9
A g e (Ma)
Figure 2.31 Eustatic sea-level variations over the last 9.5 Ma based on marine oxygen isotope measurements (modified from
Miller et al., 2005). Note the distinct cooling trend reflected in the oxygen isotope record that begins c. 2.6 Ma.
momentum of the earth (Johnston & Lambeck, 2002; and by the lack of fixed stable reference points from which
Gross et al., 2005). In addition, temperature and chemical to determine the height of eustatic sea level for specific times
variations within the water column can lead to ‘steric’ in the past (Moucha et al., 2008).
(density) changes that affect ocean water volume and thus
sea level (Chambers, 2006). Fluctuations in volume and
2.5.2.2 Quaternary eustatic changes
mass of sea water will also influence the hydro-isostatic load
(section 2.5.4) on the underlying seabed, a process that may More is known about sea-level variations during the late
explain the emergence of coral atolls from the Pacific Pliocene and the Quaternary. The Pliocene was generally
Ocean during the late Holocene, when eustatic sea-level rise warmer than the Quaternary by approximately 3°C (annual
was reversed by hydro-isostatic depression (Dickinson, average) and, with less polar ice cover, eustatic sea level
2004). In addition, long-term changes in sea level can has been estimated to have been between 10 and 25 m
result from changes in the configuration of ocean basins higher than present (Haywood & Valdes, 2004). Global sea-
caused either by sediment infill and the consequent dis- levels show progressive lowering towards the end of the
placement of ocean water, or by the lateral movement of Pliocene, and this pattern appears to have continued
lithospheric plates and associated vertical displacements throughout the Quaternary (Figure 2.31). Superimposed on
of crust and mantle in the vicinity of subduction zones. this long-term trend, however, is a cyclical pattern of
Accelerated rates of sea-floor spreading lead to an increase shorter-term oscillations in eustatic sea level related to the
in volume of the mid-ocean ridges, resulting in a sea-level expansion and contraction of the great ice sheets. During
rise; conversely, slower rates of spreading reduce both glacial episodes, ice sheets grow by storing water abstracted
ridge volume and sea level. Reconstructing changes in from the oceans (section 3.10), while melting of ice during
ocean bathymetry over long geological timescales is warmer episodes returns water to the oceans. Sea-level
therefore a complicated exercise which is compounded changes controlled by the growth and contraction of the
further by chronological uncertainties (Müller et al., 2008), ice sheets are termed glacio-eustatic. During the Late
Quaternary, sea levels were lowered glacio-eustatically by (Lourens & Hilgen, 1997), but more sophisticated models
as much as 130 m at glacial maxima, but returned to levels are now available that can isolate the ice-volume com-
similar to, or higher than, those of the present during ponent from other influences affecting marine isotope
interglacials. ratios (Lea et al., 2002; Bintanja et al., 2005).
This link between changes in global ice volume and sea Chronologies of Quaternary sea-level change can also
level is reflected in oxygen isotope profiles from deep- be obtained from the direct dating of shoreline features
ocean cores where the isotopic ‘signal’ can be read as a proxy and deposits, most notably in those areas considered to be
sea-level record, and a chronology of long-term sea-level tectonically stable. Of particular importance are coral reefs
change can be established (Shackleton, 1987). While it is found, for example, around Caribbean or Pacific islands
now recognized that oxygen isotope ratios in marine and atolls, for some reef-forming corals occur only within
microfossils are not governed solely by ice-ocean volumetric the intertidal zone, and hence submerged or raised inactive
changes (section 3.10), the fragmented nature of empirical reefs provide direct evidence for changes in sea level.
evidence from coastal localities means that the marine Reef carbonate can be dated by U-series and amino-acid
oxygen isotope curve remains the only basis for establish- racemization methods (sections 5.3.4 and 5.6.1) to provide
ing a long-term continuous eustatic record (Figure 2.32). chronologies of sea-level change, for instance around the
Initially, sea-level reconstructions based on the oxygen islands of Bermuda (Peltier & Fairbanks, 2006) and
isotope approach used simple inversions of the isotopic data Mururoa (Camoin et al., 2001). In Tunisia interglacial
a)
20
0
Ice volume (m sea level)
relative to present
-20
-40
-60
-80
-100
-120
-140
0 200 400 600 800 1000
A g e (ka)
b)
50
(m relative to present)
0
Global sea level
-50
-100
-150
0 100 200 300 400 500
A g e (ka)
Figure 2.32 a) Variations in eustatic sea level (with 1σ confidence limits shown in grey shading) over the past 1 Ma based on
ice-volume changes reflected in marine oxygen isotope records. b) The sea-level record for the last 500 ka compared with
independent isotopic measurements from the Red Sea basin (white circles) and from coral reef data from New Guinea and Barbados
(black circles) (from Bintanja et al., 2005).
highstands of sea level have also been dated by U-series, that are too frequent to be explained solely by reference to
although in this case using marine Mollusca associated ice volume changes, while Chappell (2002) concluded that
with the shorelines (Jedoui et al., 2003). Luminescence temporary reversals of the general downward trend of sea
dating (section 5.3.6) has also been used to establish level during the last glacial are connected with ‘Heinrich
chronologies of sea-level change, for example on interglacial events’, extremely cold climatic pulses recorded in North
shorelines on the northeast coast of Brazil (Barreto et al., Atlantic marine records (section 3.10.1). It has also been
2002), and on Holocene marine sequences in southeast suggested that the separate influences of the polar ice sheets
India (Thomas, 2009). Another approach employs evidence can be detected in sea-level cycles, with the Antarctic ice
from submerged caves that contain cave speleothems sheet dominating during cooling phases, but the Greenland
(section 3.8.4.4). Since these only develop under subaerial ice sheet becoming the more important driver during
conditions, speleothem evidence provides constraints warming events (Siddall et al., 2010b).
on maximum heights of sea level, while the termination of At the Last Glacial Maximum (LGM) around 21 ka
speleothem formation reflects the time when caves were (Mix et al., 2001) when the Laurentide, Fennoscandian,
submerged beneath the sea. Dating of speleothem carbon- British–Irish and Antarctic ice sheets were at their greatest
ate, again using U-series, has been employed to reconstruct extent, global sea levels may have been 130 and 135 m below
sea-level histories in many coastal limestone regions, such those of the present (Yokoyama et al., 2000). The sea-level
as along the coastlines of Italy (Antonioli et al., 2004) and rise that accompanied the wastage of these ice sheets is
Croatia (Surić et al., 2009). Other features that enable a reflected in records from far field locations, coastal regions
chronology of sea-level change to be established include
carbonate aeolianite beach ridges on some coastal fringes
as, for example, in western Australia (Hearty & O’Leary, MIS 1 M ! S 2 MIS 3 MIS 5 MIS 5
2007). North Atlantic &
In some coastal localities, there is evidence for several Equatorial Pacific
glacial–interglacial sea-level cycles. In Bermuda and the

Papua New Guinea & Barbados
Bahamas, interglacial highstand positions are represented
by coral limestone while low sea levels during cooler
periods are marked by palaeosols (section 3.5). The record
extends over 1.2 Ma and includes evidence for at least Mass balance global ice model
seven interglacial highstands between 2.5 and >20 m above
present sea level, and dated to MOI stages 11–5e (Hearty
& Kindler, 1995; Kindler & Hearty, 2000). In the Mediter-
Red S e a
ranean region, ‘staircases’ of raised marine terraces are
widespread, and while some of these reflect episodic
tectonic uplift (section 2.5.3), a number mark eustatic sea-
level highstands. The oldest pre-date MIS 11, while up to Red S e a
three highstand positions have been recognized for the last

interglacial stage (MIS 5e) alone (Zazo et al., 2003). East Equatorial Pacific
Some of the most detailed records of eustatic sea-level
change are available for the last glacial–interglacial cycle (the
last c. 120 ka), for not only is the evidence better preserved,
SPECMAP
but shoreline sequences are often less disturbed than is the
case for earlier periods. Reconstructions are based on a
variety of approaches, but frequently use dated coral reef
records combined with marine oxygen isotope data 0 20 40 60 80 100 120
(Lambeck & Chappell, 2001; Siddall et al., 2003). Records A g e (ka)
from different regions are remarkably similar (Figure 2.33),
and reflect changes in global ice volume as the principal Figure 2.33 Comparison of seven different eustatic sea-level
controlling factor in overall sea-level trends. Superimposed records from the last glacial cycle. The interval between
horizontal lines represents 20 m vertical depth, and each curve
on this pattern, however, are more subtle influences. ranges between modern sea level (zero datum) at the top and
Thompson and Goldstein (2005), for example, found c. –120 to –130 m at the bottom (adapted from Siddall et al.,
evidence for persistent short-term oscillations in sea level 2010b).
a) b)
0 0 300
R i c h m o n d Gulf
Barbados H u d s o n Bay, C a n a d a
250
Tahiti
Relative sea level (m)

Huon P e n i n s u l a
Christchurch 200
S u n d a Shelf
Ice-volume-equivalent sea level (m)
Bonaparte Gulf
150
-50 18.5
100
Ice volume (10 km )

3
MWP-1b
50
6
0
c) 10 8 6 4 2 0
MWP-1a
200
Angermanland
-100 S2E Northern S w e d e n -
150
100
50
-150 57.5
25 20 15 10 5 0
A g e (ka) 10 8 6 4 2 O
Figure 2.34 a) Rise in eustatic sea level since the LGM based on evidence from several independent studies; the mean sea-
level curve has been interpolated between the data points. b) Land elevation changes in Hudson Bay, Canada and c) in northern
Sweden during the Holocene (modified from Lambeck & Chappell, 2001). Vertical and horizontal bars represent errors of altitudinal
and age estimates respectively; MWP – meltwater pulse. For further explanation, see text.
that were unaffected by glacier activity (Fleming et al., sediments, while a slight rise can result in the submergence
1998). Six such records are shown in Figure 2.34a, and show of terrestrial or freshwater deposits. As a consequence, in
a rapid rate of sea-level rise, with two episodes of further estuarine contexts there is often a stacked sequence of
acceleration at around 14 and 11.5 ka that are assumed marine, estuarine, freshwater and terrestrial sediments,
to reflect increased rates of global ice melt (Alley et al., from which a detailed history of sea-level change can be
2005). Meltwater pulse (MWP) 1a (14 ka), the earlier and reconstructed. Figure 2.35, for example, shows a terrestrial
most pronounced of the two, was originally considered peat interbedded between two layers of estuarine marine
to be the signature of rapid retreat of the Northern sediment. The boundary between the lower marine unit
Hemisphere ice sheets, although contributions from and the base of the peat marks the emergence of mud-
Southern Hemisphere ice cannot now be excluded (Peltier, flats above local high-tide level, the surface of which was
2005). The less pronounced MWP 1b (11.5 ka) has proved subsequently colonized by land plants whose remains make
to be more enigmatic, for while it seems to be clearly up the peat. If dated, this horizon would provide a sea-level
represented in some records, for example from the Sea of index point, from which not only the age and altitude of
Japan (Tanabe et al., 2009), it is absent or very poorly RSL could be established, but also the tendency of relative
defined in others, such as in the Indian Ocean (Camoin et sea-level change could be deduced (Shennan et al., 2006a).
al., 2004) and around Tahiti (Bard et al., 2010). The ‘tendency’ of an index point refers to the trend of RSL,
Eustatic sea-level rise during the Holocene is most in other words whether there is an increase or decrease in
clearly reflected in estuarine sediments and in stratigraphic marine influence at that locality. This can be inferred
records from coastal lakes. Sedimentation in estuaries is not only from changes in the nature of the sediments
usually rapid and the deposits tend to be protected from (from terrestrial to marine or vice versa), but also from
erosional processes. Relatively shallow waters are often pollen, diatom or other microfossil evidence (sections 4.3.5
found in estuaries, so that even minor falls in sea level can and 4.9.3). In this particular case, the transition from
lead to the exposure of marine and brackish water marine to terrestrial deposits indicates a decrease in marine
M a r i n e silt
Positive marine tendency
C transgression'}
Peat b e d
(terrestrial)
N e g a t i v e marine tendency
('regression')
Marine s a n d
Figure 2.35 River bank section in lower Strathearn, eastern Scotland, showing Holocene marine sands overlain by a peat layer
which in turn is overlain by up to 6 m of marine silts and clay. All the deposits lie above present sea level. Radiocarbon dating of
the bottom and top of the peat bed indicates a relative sea-level fall (marine regression) at c. 9.6 ka cal. BP and a relative sea-level
rise (marine transgression) at c. 7.5 ka cal. BP (Cullingford et al., 1980; photo by John Lowe).
influence and hence a negative sea-level tendency. In Figure In northwest Scotland, for example, analysis of sedimentary
2.35, the boundary between the top of the peat and the base sequences from a number of littoral basins shows that
of the overlying marine deposits reflects the re-submergence these had once been flooded by the sea (Figure 2.36a). In
of the estuary, and provides a sea-level index point for an each basin, sea-level index points and tendencies suggest
increase in marine influence, in other words a positive sea- that, relative to the land, sea level fell by about 14 m
level tendency. In the older literature, such evidence was between 12 and 10 ka as a result of isostatic recovery
often interpreted in terms of marine regressions and following deglaciation. Subsequently, sea level rose by
transgressions (falls and rises in sea level), although these 2–3 m relative to the land so that between 9 and 4 ka
terms are now less widely used as they imply absolute sea waters reoccupied some of the basins, while all had re-
changes in sea level, whereas only relative sea-level emerged from the sea by c. 3 ka (Figure 2.36a). In parts of
tendencies can really be inferred from such evidence. southern England, by contrast, sea level appears to have
Lake basins that lie near the coast may also preserve sea- risen relative to the land by c. 30 m or more between 10
level index points where, for example, such basins have been and 4 ka (Figure 2.36b). The difference in sea-level tenden-
raised above the sea and subsequently begin to accumulate cies between the English and Scottish records reflects the
freshwater sediment. These are referred to as isolation glacio-isostatic effects of the last ice sheet, which depressed
basins. In some instances, a rise in RSL may lead to the the northern parts of the country, while southern areas lay
re-submergence of these basins, in which case brackish or beyond its impact. There is still a small, residual reflex
marine deposits accumulate over freshwater sediments. movement today (Figure 2.37), for northern Britain has
a) Figure 2.36 a) Chronology of isolations from, and

20 connections to, the sea for four littoral basins in northwest
18 Rumach Meadbonach Scotland: Loch nan Eala upper basin; Loch nan Eala lower
basin; Rumach Lochdar; and Rumach Meadhonach. The
16
curve shows the sequence of relative sea-level change
14 (from Shennan et al., 1994). b) Relative sea-level
Metres (OD)
12 reconstructions for selected sites in the UK, based on

10 RiJmscb Locbdar dated index points and sea-level tendencies. The horizontal
8 and vertical bars represent age and altitude errors,
Upper basin respectively. The x-axes show time (ka cal. BP) and y-axes
6 Lower basin give height (m) relative to modern sea level. Sites closest
4 to former thick ice masses show initial relative falls in sea
prasem MHWST
2 level following deglaciation. Sites distant from ice centres
13 12 1 27 7 'i7 7 31011
31011
0 a show a marked relative rise from low glacial to modern
12 1112 11
1 0 9 8 7 6 5 4 3 2 1 0 levels (based on Shennan et al., 2006a).
kaBP
'3 i: i4 4 4 44 7 3 13 1!
13 12 11 10 9 8 7 6 5 4 3 2 4 1 0 14 2
A g e (ka)
1-13 Index point number Marine episode in basin

Marine episode in basin
b)
(19) Ardyne 10 (3) Wick
40 0
15
2
-1C
3
20
7 ft A -20
• 5 13
o c'i 10 14 -30
11
12 15 15
IS
0 5000 10000 15000
-20 10 17 15
0 10000 20000 15
IB
15
22
54 21 15 24 (17) Forth Valley
15
22 40
28 15
2 Li 25
30 21
31 3? ft;: 20
37
:;3 33
(52) S W England, Devon 39 15
•
0
15
35 3.; 15
0 •' •3 41
15
• '•I -2C
'•ft
15 0 5000 10000 15000
-10 44 15
43 47
48
50 (40) Fenland (Tidal range corrected)
20 52 5.1 40
b'J,
0
-30
0 5000 10000 15000
-10
sea level index point
sea level index point from basal peat
20
limiting index point 0 5000 10000 15000
been rising during the late Holocene by rates up to 2.0 mm continental ice sheets to small residual remnants, and
yr–1, while southern England has been subsiding by rates broadly stabilized between 3 and 2 ka (Intergovernmental
of up to 1.2 mm yr–1 (Shennan & Horton, 2002). It must Panel on Climate Change, 2007). Indeed, sea levels may
be emphasized, however, that unlike the curves in Figures have fallen slightly during the late Holocene as ocean water
2.31–2.33, those in Figure 2.36b are relative sea-level curves, was transferred from equatorial to high-latitude regions to
as they are derived from localities where crustal warping has fill space in the ocean basins caused by the subsidence
occurred. They therefore provide an indication of local of glacial forebulges (section 2.5.4). This process has been
trends in sea level only. reversed in recent centuries, however, as both tide gauge
The rate of Holocene eustatic sea-level rise slowed records and empirical evidence from, for example, salt
markedly after c. 5 ka, reflecting contraction of the great marsh contexts (Gehrels et al. 2004, 2006), show evidence
< 00 eustatic sea-level data from ‘stable’ coastlines, but it is now

acknowledged that few parts of the world have been
0.6, 0.4 unaffected by some form of tectonic activity and hence
,0.7 0. 0
0.5, can be considered as ‘stable’. In southern Australia, for
11. 0.6,
o.a
example, which occupies a passive enclave within a craton
11.0 1.1 0.7,
(the stable core of a lithospheric plate) and is located far
11. 1.4
0.5' 1.0 from the isostatic impacts of Quaternary ice sheets,
0.5
1.0 1.6
1. 1 2, 0.0
interglacial shorelines have, nevertheless, been tectonic-
1,5 ally deformed (Murray-Wallace, 2002). Similarly in the
•<2.0 0.7,
,0.1 Caribbean region, again often considered to be relatively
1 0 1.1 £2
-0.5
• 0.9
stable, a c. 80 ka raised shoreline that can be traced from
,-0.2/
0 5 1-0, Barbados, through the Bahamas, Florida and Bermuda,
0.5 J . 7. to the Atlantic coast of the USA, has clearly been disrupted,
'-OlQ :-o.2>
0.0 11. I-0,8 (-0 7] partly by differential hydro-isostatic effects (section 2.5.3.),
,-0. 2 -0.6 -o.e:
and partly by glacio-isostatic influences of the last North
-0.3 American ice sheet (Potter & Lambeck, 2003). A clearer
-o.s (-0.6]
-0.4 -0.9 [0.4] differentiation of the eustatic component in long-term
-0.6
-0.9, Quaternary glacial–interglacial sea-level oscillations will
-0.5. -0.7. therefore require the improved definition of palaeo sea-level
-0.7
-0. 6 indicators, more secure chronologies, and the development
-1.0- vO-8
-1.2 -04 of geophysical models that integrate eustatic and tectonic
-0.5
-11 -0.5 data (Mastronuzzi et al., 2005).
A second major difficulty is posed by geoidal eustasy.
Figure 2.37 Relative changes in land or sea level in the UK The earth is not spherical, but is flattened at the poles and
during the late Holocene (mm yr–1). Positive values (filled bulging at the equator, and it has generally been assumed
circles) indicate relative land elevation or sea-level fall; open that the free ocean surface or geoid, which is the averaged
circles indicate land subsidence or sea-level rise. Figures in
elevational surface that is everywhere perpendicular to the
parentheses are model-adjusted (from Shennan & Horton,
2002). direction of gravity, is smooth and regular and parallels the
shape of the earth. However, satellite measurements have
of rising sea levels from the early 1800s onwards. Indeed, shown that this is not the case: the real geoid is highly
data from the east coast of the USA suggest a rise of around irregular, even over the ocean surface, and varies by as much
1.7 mm yr–1 during the twentieth century (Engelhart et al., as 200 m between elevational highs and lows (http://
2011). Modelling studies confirm the sea-level acceleration earthobservatory.nasa.gov/Features/GRACE/; Tapley et al.,
from the nineteenth to the twentieth century and point to 2004). Furthermore, its shape is constantly changing. Some
anthropogenic forcing as a partial cause (Gregory et al., of the geoidal irregularity is caused by gravitational
2006), although discrepancies between model simulations deformities resulting from the earth’s rotation or from the
and observed sea-level rise mean that further work is spatial complexity of its internal density structure. But
required to reconcile observations from tide-gauge and there are also glacio-isostatic effects caused by the transfer
proxy data with the modelling studies (Gehrels, 2010). This of mass from the oceans to the higher latitudes during
is an important research area because concern over possible build-up of ice sheets, and the reversal of this process
future sea-level rise caused by anthropogenically forced during phases of ice melt. Furthermore, oscillations in sea
climate change is leading to greater demands for better level lead to variations in hydro-isostatic pressure on the
baseline data from which to predict future sea-level beds of the ocean basins. The result is that there is unlikely
scenarios (Church et al., 2008; Woodroffe & Murray- to be any part of the geoid that remains unaffected (Milne
Wallace, 2012). & Mitrovica, 2008). Not only does the geoidal ocean surface
Despite the advances that have been made over recent intersect different land masses simultaneously at different
years in understanding long-term eustatic changes, quanti- altitudes, but the overall pattern of gravitational deform-
fying the eustatic element in Quaternary sea-level records ation of the geoid changes over time. While this may not
on the basis of empirical evidence remains extremely affect the construction of sea-level curves for individual
problematical. Attempts have repeatedly been made to localities, it does mean that realistic eustatic sea-level curves
separate the isostatic from the eustatic component by using can only be constructed for individual regions, and that
Last interglacial platform
Platform at b a s e of M I S - 5
Terrace sequence of last glacial stage
H o l o c e n e platform {age 7 ka)
Figure 2.38 Raised coral reef terraces extend for over 80 km along the coast of the Huon Peninsula, Papua New Guinea, reflecting
tectonic uplift rates of up to c. 5.0 m ka-1. In the part of the sequence shown here, the Holocene (MIS-1) platform is now at 18 m
and the last interglacial (MIS-5e) terrace at 350 m above present sea level, indicating a mean uplift rate of c. 2.8 m ka-1 (from Esat
et al., 1999; photograph by Tezer Esat, Australian National University, Canberra; reprinted with permission from AAAS).
those curves remain area-specific. The corollary, of course, number of sites around the western margin of the Pacific
is that a single, globally valid eustatic sea-level curve based Rim, similar numbers of Holocene terraces with typical
on shoreline evidence, so long regarded as the holy grail of maximum uplift rates of 4 m ka–1 have been observed,
sea-surface studies, is effectively unattainable (PALSEA, although much higher rates of uplift (15 m ka–1) are evi-
2010). dent along the east coast of Taiwan (Ota & Yamaguchi,
2004).
While raised shorelines provide evidence of long-term
2.5.3 Tectonic influences activity and recurrence intervals of major faults and
Shoreline displacement and deformation resulting from earthquakes in tectonically active regions (e.g. Gardner
long-term earth movements is apparent in many parts of et al., 2008), they may also be indicative of rates of tectonic
the world and provides important evidence for the rates activity in less active areas. For example, beach ridges of Last
of operation of tectonic processes. In New Guinea, for Interglacial age (MIS 5e, c. 125 ka) confirm the long-term
example, the spectacular flights of raised coral terraces stability of the Carmel coastal plain of Israel, for they are
that extend for almost 80 km along the emergent coastline all no higher than 9 m above present sea level, implying
of the Huon Peninsula (Figure 2.38) have been uplifted at a maximum possible uplift of 48 mm ka–1 (Galili et al.,
rates of 0.5–4 m ka–1 over a period of >300 ka, the oldest 2007). By contrast, on the coast of northeastern Brazil, there
and highest terrace now lying at c. 400 m above present sea is clear evidence of down-faulting of the MIS 7c marine
level (Esat et al., 1999). Similar ‘staircases’ of raised marine terrace and uplift of the MIS 5c terrace, the latter being
terraces occur near other active plate boundaries, including locally c. 12 m higher than features of similar age some 1,000
New Zealand (Claessens et al., 2009), eastern Patagonia km to the south (Barreto et al., 2002). In this case, the
(Pedoja et al., 2011) and the Aegean (Vött, 2007). At a shoreline evidence suggests that this passive plate margin
region, formerly considered to be relatively stable, is far Saxony in Germany, westwards through the Nether-
more tectonically active than hitherto believed (Barreto lands to the Dogger Bank area of the southern North Sea
et al., 2002). basin (Vink et al., 2007), while RSL evidence from the
Gulf of Mexico shows that although this area is far removed
from the maximal limits of glaciation, the region is still
2.5.4 Glacio- and hydro-isostasy responding glacio-isostatically, by means of forebulge
Crustal deformation resulting from the expansion and collapse, to the melting of the Laurentide ice sheet
contraction of the great ice sheets has been, and continues (Törnqvist et al., 2004). In other areas, ice sheet forebulge
to be, a major cause of relative sea-level change during collapse is superimposed on regional isostatic influences,
the Quaternary (Milne et al., 2009). An increase in ice for example in west Greenland, where the RSL history of
mass leads to increased loading on, and depression in, the the Late Holocene reflects a combination of local and
underlying crust, while the reverse occurs during ice sheet non-Greenland RSL processes, notably the reloading of the
wastage. This process of crustal warping is referred to as earth’s crust caused by Neoglacial expansion of the
glacio-isostasy. The oceans also have a loading effect on Greenland ice sheet, and continued subsidence associated
the crust, which also varies as the ice sheets expand and with the collapse of the Laurentide ice sheet forebulge
contract, and the deformation of the crust that results is (Long et al., 2009). The vertical extent of crustal displace-
known as hydro-isostasy. ment associated with forebulge development is often
The consequences of glacial loading will vary with the considerable. In western Canada, for example, evidence
rigidity of the crust, although it is clear that the earth’s crust from the coast of British Columbia indicates land uplift
does not behave as a solid block when under stress, but from the forebulge of the Cordilleran ice sheet of the order
rather in a malleable, visco-elastic manner, and this enables of 85 m, which exposed parts of the currently submerged
it to respond flexibly to differential ice loads (Plag et al., coastal shelf close to Queen Charlotte Island (Figure 2.40).
1998). In general, maximum loading and crustal depression
occur near the centre of an ice sheet and there is a gradual a) a) a) a)
a) a)
rise in crustal surface towards the ice sheet margins (Figure a) a)
2.39). However, crustal depression at one point must be a) Queen a)
Queen
Queen
compensated for elsewhere, one result being an upward Queen
bulging of the crust (forebulge) beyond the margins of the a) a)

ice mass (Figure 2.39). The areal extent of former glacial a) a)
forebulges can be inferred from RSL index points, regional
a) a)
variations in which display complex patterns of differen-
tial crustal movement in areas that lay beyond (in some a) a)
instances well beyond) the margins of the last ice sheets. a) a)a)
Queen Queen
Queen
In western Europe, RSL data suggest that the zone of a)
a)
maximum forebulge subsidence extends from Lower a)a)
a)
a)
a)
C e n t r e of i c e m a s s a n d of a)
isostatic d e p r e s s i o n a)
Ice s u r f a c e a)
Ice Forebulge a)
A Ice Ice
a)a)
Ice a)
Ice m a r g i n
Isostatically d e p r e s s e d land s u r f a c e Figure 2.40 Crustal flexuring in the Pacific margin of Canada
reflecting the changing mass of the last ice sheet in British
Columbia at a) 13.2–12.7 ka, b) 11.7–11.2 ka and c) 10.2–9.7
Figure 2.39 Schematic diagram of the effect of an ice mass ka. Areas in shades of blue/purple are depressed, while those
on a land surface. In general terms, the amount of isostatic in green, yellow and pink are elevated, relative to the present.
depression of the land surface A–B increases proportionately The latter represent a transient glacial forebulge situated
with ice load, so that greater depression occurs towards the beyond the perimeter of the ice mass (from Hetherington &
centre of an ice mass than around its margin (cf. h1 and h2). Barrie, 2004, reprinted with permission from Elsevier).
In a now classic work on the Canadian Arctic, Andrews (Lewis et al., 2012). In smaller ocean basins, the interaction
(1970) proposed three developmental stages in the process between deglacial sea-level rise and local flexuring of the
of isostatic recovery after ice-sheet melt. The rapid crustal crust can result in continental levering. In the Mediter-
adjustment that occurs at a site between initial ice sheet ranean region, for example, subsidence caused by increased
contraction and subsequent deglaciation is referred to as weight of water resulting from sea-level rise is pronounced
restrained rebound. Following ice wastage, more gradual in the central, deeper parts of the basin, leading to uplift
postglacial rebound takes place, which continues up to and hence emergence of strandlines around the periphery
the present day. The relatively small amount of uplift still (Stocchi & Spada, 2007). A combination of hydro-isostatic
required to re-establish crustal equilibrium is termed effects and crustal warping from sedimentary loading may
residual rebound. Isostatic recovery can therefore be also occur close to large deltas, such as in the Gulf of
seen as a process that accelerates rapidly at first, but Mexico (Simms et al., 2007).
which then slows down gradually as the pre-glacial state of
crustal equilibrium is approached. In many areas, isostatic
2.5.5 Shoreline sequences in areas
uplift is not yet complete (Figure 2.37). Parts of Norway and
the Gulf of Bothnia, for example, are presently rising at a
affected by glacio-isostasy
rate of c. 1.0 mm yr–1 (Fjeldskaar et al., 2000), while in North During glacial episodes, eustatic sea levels were low, but
America, GPS measurements show the most rapid current because the crust was also depressed beneath the weight of
rate of land uplift to be c. 10 mm yr–1 in the Hudson Bay
area, where ice cover was thickest at the LGM. The com- C e n t r e of isostatic d e p r e s s i o n a n d recovery
puted uplift rates decline with distance from Hudson Bay
and, further south, change to subsidence by c. 1–2 mm yr–1
Ice
in the region of the southern Great Lakes (Sella et al., 2007).
The magnitude and regional impacts of hydro-isostatic
TA ML1 SL-1
effects are more difficult to estimate. The problem is that
both the sea surface and seabed have been subject to
differential flexuring, the former by geoidal variations, and a)
the latter by glacio-isostatic effects which, as we have seen,
extended far beyond the ice sheet margins. Moreover,
Ice
spatial variations in the rheological behaviour of the crust ML2
ML2
and in ocean density mean calculation of past hydro- TB
ML2 SL-2
isostatic effects is not straightforward (Lambeck et al.,
2003; Mitrovica, 2003) and, moreover, it is particularly
a)
difficult to separate the impacts of the hydro-isostatic
component from those of other flexural stresses. A partic-
%
ular problem arises from the process of ocean siphoning Ice 5 SL
-3
(Mitrovica & Peltier, 1991). As the ice sheets melted and TC
forebulges collapsed, the continental shelves were tilted ML3 SL-3
towards the oceans, partly by continental isostatic recov-

ery and partly by the increased mass of the oceans from ca))
meltwater input. Sea water could therefore encroach across SL
SL -3 SL
previously glaciated and up-tilted shelf areas, progressively Ice
-3 -3
drawing (siphoning) water away from the open oceans. TD
ML4 SL^4
Some regions may therefore have experienced rapid rise in
sea level during the early Holocene, followed by a relative
fall in sea level as a consequence of the siphoning effect R e l a t i v e isostatic uplift
(section 2.5.2). Indeed, marked Holocene shorelines a)
traceable throughout the central Pacific between New
Zealand and Japan are considered to reflect this process Figure 2.41 The development of a raised shoreline sequence
reflecting isostatic recovery and eustatic rise following
(Dickinson, 2000), while variations in the timing and
deglaciation. ML – marine limit; SL – sea level; RS – raised
elevation of the Holocene high sea-level stand around the shoreline. The effects of the forebulge (Figure 2.39) have been
coast of Australia are attributed largely to ocean siphoning omitted from this diagram. See text for further explanation.
ice, shorelines may have formed in some localities close to that the RSL in the area is rising and therefore later
the margins of the ice sheets. This is indicated by the fact shorelines will progressively truncate the older and more
that in many areas, raised shorelines terminate inland in steeply inclined features (Figure 2.41d). In general,
glacial outwash and related deposits. Following deglaci- therefore, the oldest and most steeply tilted shorelines will
ation, both land uplift and sea-level rise occurred and form at the greatest distance from the ice centre, younger
hence sequences of raised shorelines have developed that shorelines will be less steeply inclined and more extensively
reflect the interplay of isostatic and eustatic factors. developed, and older features will have been destroyed or
Figure 2.41 shows the type of shoreline sequence that partly destroyed during the formation of younger shorelines
might be found in an area undergoing isostatic recovery. or, in certain cases, will either be buried beneath later
As the ice front recedes from terminus TA to TB, uplift sediments or will be found below present sea level. The
occurs so that the shoreline formed when the sea level vertical interval between individual shorelines at particular
stood at SL-1 is raised above the new sea level, SL-2. localities shows the amount of isostatic uplift that has
Because isostatic recovery decreases with distance from the occurred between the times of shoreline formation (‘x’ in
centre of an ice sheet, the shoreline RS-1 will be tilted away Figure 2.41d).
from the ice-sheet centre. During glacier retreat from TB Raised shorelines can be both depositional and erosional
to TC, the shoreline that developed while the sea stood forms, and the extent to which a clear geomorphological
at SL-2 is raised and tilted (RS-2), but it will be less steeply feature develops depends on a range of factors including
inclined than RS-1, which has now been even further the length of time that sea level remained constant relative
deformed. However, because the rate of isostatic recovery to the land, and on the operation of local glacial, fluvial
at that time was accelerating due to rapid ice wastage, the and marine processes. Continuous shoreline features will
marine limit (ML2) of RS-2 has been raised to a higher have evolved in some areas, while in others the geomorpho-
altitude than the marine limit ML1 of shoreline RS-1. logical expression of relative sea-level change may be more
Subsequently, a third raised shoreline (RS-3) develops, sporadic. However, postglacial subaerial and marine activity
but by this time isostatic recovery has slowed down, so that may have destroyed or extensively modified much of the
the marine limit ML3 is found at a lower altitude than either evidence, even in those localities where coastal landforms
ML1 or ML2. Moreover, a combination of decreased uplift were originally well developed. Consequently, only shore-
and an increase in the rate of eustatic sea-level rise means line fragments remain in most areas, and careful mapping
WNW ESE
Feet St. A n d r e w s Boarhills Kingsbarns Fife N e s s
120
Metres above present sea level

110
35
100
90 30
Kb
SO
25
70
20
60
heights on raised shorelines

50
• heights o n fluvio glacial t e r r a c e s
15
18 16 14 12 10 8 6 4 2
Kilometres
Figure 2.42 Height–distance diagram of measured altitudinal variations of the surfaces of shoreline fragments in Fife, eastern
Scotland (modified after Cullingford & Smith, 1966). For further explanation see text.
and instrumental levelling of each shoreline remnant is

a)
necessary before individual shorelines can be reconstructed
and inferences made about former sea levels (Smith et al.,
2006), a procedure that has become easier to conduct in
more remote regions by using airborne LIDAR data and
digital terrain modelling (Kovanen & Slaymaker, 2004). In 0
some sheltered localities, however, because of the complex 28
interplay between land uplift and eustatic sea-level rise, 180
some shorelines may become buried by younger marine -160
sediment. Further isostatic recovery may raise these ?3?3
0
0
above present sea level to produce raised buried shorelines 'J

80
(Gray, 1984). In order to reconstruct the complete history 60 0 600 km
30
of isostatic and eustatic changes in such areas, therefore,
a)
subsurface stratigraphic information obtained from bore-
holes needs to be integrated with the results of detailed
mapping of visible shoreline fragments.
Raised shoreline data are usually presented in the form
of a height–distance diagram (Figure 2.42), which is a plot
of all the individual data points in a vertical plane running
parallel to a line towards the assumed ice centre. Shoreline
fragments are resolved into a series of inferred shorelines,
and the gradients of the features can then be calculated,
usually by means of regression analysis. Where prominent
0
shorelines of the same age are found in different areas,
60600
isobases can be constructed for these shorelines. Isobases 0
join points of equal altitude (or uplift) on shorelines of the 60600
0
60 0
same age. The pattern of isobases, reconstructed either
manually or using trend surface analysis, gives a three-
dimensional image of the deformation of the land surface 60 00
by the weight of glacier ice (Figure 2.43). It is important to 6
appreciate, however, that while the isobase patterns may 0 400 km
indicate in a very general way those areas that experienced C)
maximum glacio-isostatic depression, the values of the
10
isobases themselves bear no relation to the amount of 2
depression (or subsequent rebound) that has actually 4
occurred. Hence, isobase maps such as Figure 2.43 show 6
only uplift relative to present sea level, and not absolute uplift 8
following deglaciation. Moreover, large-scale isobase maps 10
tend to be generalized, and fail to represent the complex
deformational effects that arise when there was more than 12
one ice accumulation centre, or the consequences of local
crustal flexuring that only become apparent when there is
sufficient evidence to reveal isobase trends at the regional
scale (e.g. Figure 2.43).
Figure 2.43 a) Isobase map showing uplift (in metres) in

eastern Canada since 7.5 ka (Hillaire-Marcel & Ochietti, 1980).
b) Isobases (in metres) showing absolute uplift of Scandinavia
during the Holocene (Mörner, 1980). c) Quadratic trend surface
showing isobases (in metres) on the Main Postglacial Shoreline 0 100 km
(c. 6.5 ka) in Scotland (Firth et al., 1993).
a) 76
77 a)
60N
76 ,36
77 35
34
33
56 N
32 31
3c
23
52N
26, 27
76
73 7? 76 7:
76 2-1
21 23
0 7 73 55
76
li 34
19 22
48N 53 55 '•5
76 17
76 13
61 57 60 5976 23
116C IS 76•
76
13 76 76
76 44 ••
•
7 76 76 76 43 92
44N 76
5 43 •1; 31
1C 76 76 •- •
il 76 3
=3 2
72 W 68 64 60 76
56 52
b)
52
52 NN 12 ka
52 N 52 N
30
0
48 N 48 N
0-
0
80
0
12 40
12
40
44 N
40
44 N -so •so
9
72 W 68 64 60 56 52 48 W 72 W 68 64 60 56 52 48 W
11 ka 10 ka
52 N 52 N
0
0 16 160
16 40
•0
48 N 48 N
0'
20 0 0
1 12 12 -40
40
44 N 44 N
-30
44 N
72 W 68 64 80 56 52 48 W 72 W 68 64 60 56 52 48 W
Figure 2.44 Sea-level changes in Atlantic Canada. a) Location of sites from which relative sea-level data have been obtained. b)
Isobases (in m) for shoreline uplift for 13, 12, 11 and 10 ka. c) Palaeogeographic map for 12 ka based on isobase data: green and
yellow are land areas, grey is shallow littoral zone, blue is sea and white is glacial ice cover (from Shaw et al., 2002).
Recent years have seen the development of more 2.3.4); on the ages of raised shorelines where dating by
sophisticated numerical models of glacio-isostatic rebound direct means (e.g. radiometric dating) is not possible; on
and associated sea-level fluctuations. Such models provide the palaeogeography of the changing coastline (e.g. Figure
information on, inter alia, the earth’s response to surface 2.44); and on the extent to which sea-level measurements
loading; on ice-sheet dimensions and behaviour (section from different sites can be combined to form representative
sea-level curves for a given region (Shennan et al., 2006b; Sundaland; and Australia and New Guinea were connected
Smith et al., 2012). The models combine glaciological by the Sahul Shelf (Bird et al., 2005). It was across these
evidence, empirical data relating to sea-level change, and land bridges that the peopling of the Americas and Australia
geophysical parameters relating to the rheological13 took place (Davidson, 2013). By contrast, rising sea levels
properties of the upper mantle (Peltier, 2002; Shennan submerged former land-links, a process that may have
et al., 2006a). In northern Ireland, for example, RSL encouraged the spread of early farmers into Europe during
evidence combined with glacial isostatic adjustment model- the Neolithic after previously occupied fertile lowlands
ling have been used to constrain the evolution of the last were lost to the sea (Turney & Brown, 2007). The recon-
Irish ice sheet, the ‘best-fit’ ice model pointing to the struction of former coastlines can therefore provide an
development of a spatially extensive ice sheet of around important aid to understanding human history, not least
700 m thickness over much of northern and central Ireland because of the resource potential of littoral zones for
at the LGM, with very rapid deglaciation after 21 ka (Brooks prehistoric communities (Smith et al., 2010). Finally, there
et al., 2007a). In the high Arctic to the north of Norway, are indirect effects of the crustal stresses caused by glacio-
outputs from a fully integrated ice-mass and glacio-isostatic and hydro-isostatic change, for these may have caused
model of the Late Weichselian ice sheet suggest that initial increased eruptive activity in major volcanic centres, such
ice growth over Scandinavia and adjacent island archi- as the Campanian region in Italy (D’Argenio et al., 2004),
pelagos caused uplift of the central Barents Sea, and and perhaps more widely in the Mediterranean area
subsequent growth of ice across the entire Barents Shelf (McGuire et al., 1997). These crustal stresses may have also
(Howell et al., 2000), a reconstruction that could not have influenced (and continue to influence) the rates of release
been made from empirical evidence alone. Elsewhere, of fluids and gas, such as methane, from sea-bed reposi-
geophysical modelling and empirical data on sea-level tories, and thus affect the greenhouse gas content of the
variations have been used to reconstruct patterns of crustal atmosphere (Boles et al., 2001).
tilting and uplift at the north Cascadia subduction zone
of western Canada during the retreat of the Cordilleran ice
sheet (James et al., 2000), while empirical evidence of RSL
2.6 RIVER TERRACES
and ice extent were employed to calibrate a glaciological River valleys throughout the world contain abundant
model of the behaviour of the Greenland ice sheet between evidence for changes in fluvial activity. Short-term adjust-
the LGM and the present day (Simpson et al., 2009). ments in river regime and channel course lead to the
Isostatic rebound models, combined with empirical data of erosion of floodplain deposits, especially during peak
sea-level change, therefore provide a basis for assessing the discharge (flood) events. If the river permanently changes
rheological behaviour of the earth’s crust, and also for course, former channels may be marked by distinctive
testing glaciological reconstructions. erosional scars or remnant meander loops, and abandoned
flood-plain levels by inactive sand or gravel (braid) bars
and overbank, finer-grained deposits (Blum and Törnqvist,
2.5.6 Palaeoenvironmental significance of
2000). The dominant factor governing these hydrological
sea-level changes changes is precipitation, although the impact of rainfall
Understanding the causes and effects of sea-level change is variations on fluvial systems is modulated by other com-
of fundamental importance in the analysis of Quaternary ponents of the landscape, notably soil and vegetation cover,
environments, for not only do they enable sea-level histories which are themselves influenced by changes in climate
to be reconstructed, but as we have seen in the foregoing (Törnqvist, 2007). To a large extent, therefore, relict fluvial
sections, they provide additional insights into glacier and landforms reflect the impacts of regional climatic variations,
ice sheet behaviour, into deglacial chronologies and into the although during the Late Quaternary, the increasing
tectonic history of littoral zones. Changing levels of land influence of people on hydrological processes adds a further
and sea impact on other aspects of Quaternary research, dimension to the interpretation of river histories (Hoff-
however. For example, lowered sea levels in the past have mann et al., 2010).
provided migration corridors and access routes to new In many river valleys, long-term changes in river regime
territories for plants and animals, while shorter sea passages are reflected in river terraces preserved on valley sides or
may have encouraged movements of humans by boat. alongside the present river channel on the floodplain (Figure
Land bridges formed at times of lower sea level, such as that 2.45). Sometimes these exist as single features, but on
across the Bering Strait (Dixon, 2001); the Indonesian occasions they are arranged in vertical succession forming
islands were joined together to form the now submerged a flight or ‘staircase’, and, in major river valleys such as the
Figure 2.45 Terraces cut into outwash gravels, Glen Roy, Scotland (photograph by John Lowe).
Rhine (Figure 2.46), Somme or Danube, these represent lateral migration of the river channel, while episodes of
episodes of alluviation and incision over a considerable fluvial downcutting are usually interspersed with aggrada-
part of the Quaternary period (Bridgland & Westaway, tional phases, a process referred to as ‘cut and fill’ (Figure
2008a). Fluvial terraces may be erosional, with bedrock 2.47).
being planated to form a low-gradient strath which is often Because river terraces are most frequently developed on
covered by a thin veneer of alluvium, or they may represent unconsolidated alluvial sediments, they are easily destroyed
the upper levels of aggradation (alluvial sediment accum- by subsequent fluvial action, and hence a previous flood-
ulation) on the valley floor before subsequent downcutting, plain surface will usually only be preserved in the form of
that is, the surfaces of former floodplains (Figure 2.46). River individual, sometimes isolated, terrace fragments. Instru-
terraces occur in all geomorphological and climatic environmental levelling of the terrace remnants and collation of
ments and may be preserved as either ‘paired’ or ‘unpaired’ the data by means of height–distance diagrams enables
features. Where there has been an episode of rapid incision down-valley gradients of former floodplain levels to be
by the river, then ‘paired terraces’ may form on both sides reconstructed. It has usually been assumed that the highest
of the valley. During times of slow downcutting or land (and generally the most fragmented) forms in a terrace
surface stability, however, lateral migration of the stream series represent the oldest river levels, and lower terraces
channel leads to erosion of floodplain gravels on the outer reflect successively younger stages. In broad outline, this
edges of meanders. If prolonged, this may lead to com- relationship seems to hold, but it is now apparent that the
plete removal of former floodplain gravels from one flank sequences in many river valleys is more complicated than
of the valley, leaving a single or unpaired terrace on the this, and stratigraphic evidence is usually required if a
other (protected) side. Terrace formation and preservation complete fluvial history of a river valley is to be recon-
therefore reflects the rate and magnitude of incision and structed. Older terrace surfaces can be buried beneath
a) a) .Kr^Feta, WT Zt3
Push r-Kjrj.rc
.Kr
WT Zt3 WT Zt3
WT Zt3
^Fe
WTZt3
WT Zt3
.Kr
m a.s.l.
ta,
^Fe
.Kr^Feta,
kliddta Ttrraoas
M T 6Zt3
WT
ta,
WT Zt3
Pleistocene
300 .Kr^Feta,
WT Zt3
.Kr^Feta,
Terraces
Pliocene Terraces Upper Terraces HT1 Zt3
WT
Lower
.Kr^Feta, Upper TGTOOM
250 Tertiary gravel beds ,
Feta
UT '.
.Kr^
Pre-OuBtemBr*
UT 2/3
.Kr^Feta,
.Kr^Feta,
.Kr^Feta,
LIT 4
200 Middle Terraces
1
a)
a) HT
LMT
Lower Terraces
MMT
150
a)
LT
a)a) BEDROCK •J.
a) 2
LMT
HT1
LT
a)a)a)
a)
MMT
100
MMT
a)
a)a)
MMT
a)a) 1 2
MMT
a)
MMT
ve
MMT
gra
a)a
53
a)a))
HT1
a)a)a)
a) HT1
a) a)a) a)a)
0
0 iO »m » »km
Figure 2.46 a) Schematic representation of the staircase of river terraces preserved in the Middle Rhine area, which spans the
Pliocene to Late Quaternary. b) Areal distribution of terrace gravels in the Lower Rhine, reflecting long-term lateral migration of
the river’s course (after Boenigk & Frechen, 2006).
C u t - a n d fill-history
Fill ( a g g r a d a t i o n a l ) Incision
Strath terrace terrace Strath terrace
Fill ( a g g r a d a t i o n a l ) Aggradation
terraces Strath
Cut (degradational) terrace
terraces
Strath
terrace
Bedrock
Time
Figure 2.47 Formation of ‘strath’ and ‘cut-and-fill’ fluvial terraces. Strath terraces are formed by lateral fluvial erosion of bedrock,
while cut-and fill terraces are composed of fluvial deposits (from Merritt, 2007).
younger alluvial fills or other deposits, and might only be struction of palaeoenvironments associated with alluvial
discovered where exposed by erosion or through systematic deposition and also in the dating and correlation of indi-
borehole survey, the former terrace surfaces being indicated vidual terrace remnants (Bridgland, 2000; Gibbard & Lewin,
by buried soils, weathering profiles, peat layers, fossiliferous 2009). Hence, geomorphological evidence for past fluvial
beds or archaeological features. In the Somme valley of activity represents only one aspect of these wider palaeo-
northern France, for example, a 700 ka terrace sequence has hydrological investigations.
been buried by aeolian sediments (loess) with intercalated
palaeosols (Figure 2.48). Where exposures are not available,
therefore, and river terraces are recognized purely on the
2.6.1 Origins of river terraces
basis of observable geomorphological characteristics, a
greatly over-simplified history of fluvial activity could be River incision into a valley floor leading to the aband-
inferred. In the study of river terrace sequences, there- onment of floodplain levels, or to renewed aggradation
fore, recourse is increasingly being made to sedimentary, along the river’s course, may result from a number of
biological and archaeological evidence, both in the recon- factors. These include the following.
80 Relative height a b o v e M o d e r n valley bedrock
•3
70
High 7
level
- 1- 9169k6ak a
E
60 gravels •3 •3
•3 •4
Grace-Au to route r•4 •rj 1
•4
50 9 9 7 +/- 1 4 1 k a
1?
Grace 5
1 105+/- 196 ka •4
Saveuse •4
40 1 105+/- 196 k
Renancourt ft
5 •4
Freville 6
30 •4
400 +/-100 k a Garenne 7
2
•4
20 2 9 6 +/- 5 3 k a
Epinette •4 •4
• River S o m m e
Argceuves 2B
10 12 4 ka
2 0 0 +/- 5 7 k a Montieres
•4
9 5 +/•- 4 k a Etouvie
0 •4•4
M o d e r n valley gravels
9.0 ka
Figure 2.48 Schematic representation of a sequence of river terraces in the valley of the Somme, France, which extends over
the last c. 1 Ma. The terraces can be dated by a variety of means and assigned to oxygen isotope stages (numbers in circles).
The terrace surfaces have subsequently become buried beneath a thick accumulation of windblown silt (loess) and intercalated
soils (from Antoine et al., 2000).
2.6.1.1 Eustatic changes in sea level mental variables, such as river discharge and sediment
yield from the catchment, or tectonic influences (Hancock
Changes in sea level (base level) have long been considered & Anderson, 2002).
to be a major variable governing river terrace development.
Eustatic sea levels appear to have fluctuated through a
2.6.1.2 Climatic change
range of perhaps 150 m during the glacial and interglacial
cycles of the Quaternary (section 2.5.2). Traditional models The influence of climatic change on river valley evolution
envisage a situation in which active downcutting and river has long been recognized, and is still regarded as a major
incision accompanied falling sea levels (i.e. during cold influence on river terrace development (Starkel, 2003;
stages), while increased aggradation was associated with Bridgland & Westaway, 2008a). The relationship between
high sea-level stands, particularly in the lower reaches of river behaviour and climate depends on the nature of
river valleys inundated by the rising sea. Reconstructions climatic change and on the effects of such changes on
of the long profiles of former floodplains should therefore discharge and sediment load. In semi-arid and arid regions,
show convergent grading to these interglacial sea-level river incision occurred primarily during pluvial episodes
positions. Terraces related to sea level in this way have been (section 2.7.1), when discharge was more constant and
referred to as thalassostatic terraces (‘thalassa’ is Greek sediment yield was reduced by an expansion in regional
for sea). However, this paradigm is not well supported by vegetation cover. Phases of aggradation, on the other hand,
empirical evidence for past river behaviour, as reflected in were more characteristic of arid periods, when discharge
the geomorphological and stratigraphical archives pre- was reduced and sediment yield from more sparsely
served in major river valleys. Rather, aggradation appears vegetated catchments was enhanced. In the northwest
to have dominated in the lower reaches of fluvial valleys Himalayas, for example, phases of fluvial aggradation
during cold stages, while alternation between phases of have been correlated with intensification of the Indian
incision and aggradation has clearly been a driving force summer monsoon, while downcutting coincided with
behind river terrace formation in river reaches well- weaker monsoonal activity (Brookhagen et al., 2006). In
upstream of any marine influence (Bridgland & Westaway, temperate regions, increased aggradation occurred during
2008a). Moreover, spatial and temporal variations in colder periods when rivers were more heavily debris-
geomorphic activity within a river system make it extremely laden. Present arctic nival discharge regimes are marked by
difficult to isolate those changes in fluvial activity induced a major flood event during the spring resulting from the
by base-level changes from those caused by other environ- rapid melt of winter snow (Syvitski, 2002). Hence, areas
affected by periglacial conditions in the past would have are best seen in the highly active orogenic areas of, for
experienced seasonal flooding due to snowmelt, and example, southeast Asia and New Zealand, where rapid
increases in sediment yield due to a combination of a uplift has produced flights of widely separated terrace
relatively sparse vegetation cover and ground disturbance levels in many river valleys (Tsai et al., 2007; Claessens
by periglacial processes (Vandenberghe & Woo, 2002). et al., 2009). River long profiles frequently show abrupt
During warmer periods, sediment yield would have been changes in gradient that are thought to reflect variations in
reduced and discharge variations less marked. Terraces uplift rates (‘knickpoints’), while more subtle deforma-
related primarily to climatic changes have been referred tion reflecting epeirogenic influences14 can sometimes be
to as ‘climatic terraces’, but clearly some geographical detected (Brocklehurst, 2010). River terraces that have
variation in the relationship of terrace formation to climatic clearly been warped by tectonic activity enable rates and
factors is to be expected. Moreover, it must be stressed that patterns of uplift to be determined, as well as magnitudes
climate is only one of a number of variables affecting of folding and faulting (Yang et al., 2011). River valleys
discharge, sediment load and sedimentation (Vanden- affected by glacio-isostasy would also have experienced
berghe, 2003). gradient changes as a result of differential land uplift as, for
example, in the Rhine–Meuse system in northwest Europe
2.6.1.3 Glaciation and rivers draining the Sudeten Mountains in southwest
In areas formerly covered by glacier ice, deglaciation was Poland (Busschers et al., 2007). Even in small alpine glacier
typically associated with the accumulation of glaciofluvial systems, ice loading can influence river terrace gradients
sediments (mainly outwash gravels), followed by alternating (Pazzaglia & Brandon, 2001). In some large river valleys,
periods of incision and aggradation or channel stability, however, only part of the basin may have been affected by
which resulted in the formation of terrace sequences in uplift or tectonic down-warping, so that the original river
river valleys (Evans & Twigg, 2002). These are usually des- long profiles traced by contemporaneous terrace frag-
cribed as outwash terraces, and are frequently preserved ments have become deformed or differentially displaced.
as complex terrace systems comprising several surface This complicates the interpretation of river terrace records
levels along the margins of upland river valleys (Figure in valleys such as the Rhine and Danube, where the lower
2.45). In some areas, catastrophic release of water from reaches of the rivers have been affected by downwarping,
glacier-dammed lakes (jökulhlaups) may have been a while uplift has occurred in the higher parts of the drainage
further factor in the formation of outwash terraces (Alho systems (Ruszkiczay-Rüdiger et al., 2005; Nivière et al.,
et al., 2005). Throughout the mid-latitudes, moreover, 2008). A compilation of river terrace sequences in a number
many major drainage basins may not have been directly of major river catchments has revealed not only a wide
affected by glacier ice, but were sufficiently close to the variation in the extent of tectonic deformation, but also
ice margins to be indirectly influenced by glaciofluvial that tectonic uplift has affected some areas previously con-
processes. Consequently, the geomorphology of these sidered to have been tectonically stable (Westaway et al.,
proglacial drainage basins may have been affected not only 2009). The possibility of crustal warping should therefore
by an increase in river discharge, but also by higher always be considered when interpreting and comparing
sediment loads as debris was released from melting ice height–distance diagrams of river terrace sequences.
margins. During cold stages, therefore, considerable aggrad-
ation might be expected in the upper reaches of such rivers,
and terraces would remain as evidence of these aggraded 2.6.1.5 Human activity
surfaces following river incision during the subsequent It is now apparent that in many river basins, hydrological
warmer phase (Bridgland & Westaway, 2008a, 2008b). In regimes, sediment yield and consequent changes in flood-
some river catchments affected by glaciation in their upper plain levels have been affected by anthropogenic activity.
reaches, there are complex terrace suites that reflect the Throughout the Mediterranean region, for example, early
impact of several episodes of enhanced bedload supply Holocene palaeoenvironmental records show increases in
during successive glacial cycles (Boenigk & Frechen, 2006;
alluvial sedimentation immediately after the expansion
Tyrácek & Havlícek, 2009).
of agriculture and settlement (Figure 2.49), while archaeo-
logical evidence points to a further acceleration in sediment
2.6.1.4 Tectonic changes yields during the Hellenic and Roman periods (Hooke,
In general terms, tectonic uplift leads to rejuvenation of 2006). Forest clearance and cultivation practices remove the
rivers and therefore to accelerated incision. These effects natural vegetation cover, which reduces the water-retentive
4000 3000 2000 1000 BCWD 1000 2 0 0 0 2.6.2 River terraces and palaeo-
6.86 environmental reconstruction
6
River terraces and their underlying sediments have long
5
a) A v e r a g e 467 attracted the interest of those concerned with Quaternary
floodplain 4
environments. Not only do they provide evidence of former
aggradation 3
(m/1000 yrs]
river regimes, which can offer insights into aspects of both
2 environmental change and human history (Mishra et al.,
1 0.40 2007), but they often contain sediments with floral and
0.16
faunal remains (Schreve et al., 2007). Fossils are preserved,
60 L o w l a n d sites for example, in abandoned meander scrolls, in oxbow
U p l a n d sites
50 lakes or in overbank and ‘backswamp’ deposits, and from
b) N u m b e r
of sites in 40
these the nature of the environment under which the
the J e b e l 30
terraces evolved can often be established. In some subsiding
al-Aqra basins, fluvial deposits appear to have accumulated contin-
20
uously over long timescales, such as the stacked alluvial
10
sediments of the River Danube in the Pannonian Basin in
Hungary, which contain a record of glacial–interglacial
Middle Islamic
Late Roman
Early Bronze
Middle Bronze
Iron Age
Early Islamic
Late Islamic
Roman
Hellenistic
cycles spanning the whole of the Quaternary (Nádor et al.,

2003). In the majority of river catchments, however, the
palaeoenvironmental record is much more fragmentary,
due to the intermittent formation and erosion of terraces,
and piecing together the fluvial historical record requires
Figure 2.49 a) Average alluvial sedimentation rate calculated the combined efforts of geomorphologists, sedimentolo-
for three catchments in southern Turkey over the past 6 ka. b) gists, palaeoecologists, archaeologists and geochronologists.
The number of settlements occupied in the region over the Such reconstructions include important reference horizons
same period; lowland sites are those on valley bottoms, and for linking sedimentary sequences in different areas,
upland sites were located on hill-slopes or hilltops (from Casana,
2008). while also providing a time-stratigraphic framework
for Quaternary environmental changes (Bridgland, 2010).
In addition, both short- and long-term climate signals
can be recognized in river terrace records. For example,
capacity of the soil and leads to increased runoff. Over- the Dansgaard–Oeschger oscillations (section 3.11.4) and
grazing and over-cultivation also result in substantially centennial events like the Little Ice Age (section 7.6.3.5 )
increased erosion and therefore higher river sediment have been detected in fluvial sequences (Macklin & Lewin,
loads. In the Po valley in northern Italy, it has been 2008), while in many areas there are indications that
estimated that during the Roman period, at least 60 per cent downcutting and terrace formation occurred close to
glacial–interglacial transitions (Pan et al., 2003). On longer
of the area was deforested and converted to cultivation,
timescales, links have been inferred between terrace
commencing in the second century BC and continuing
formation and Milankovitch cycles. In Turkey, for instance,
for about 400 years. During this period, topsoils were
a 41 ka obliquity signal is reflected in river terrace sequences
removed from valley sides and the field systems on the Po
(Maddy et al., 2005), while in the Thames basin of south-
plain became buried beneath extensive spreads of alluvial east England, a distinct aggradational terrace formed in each
deposits, leading eventually to abandonment of the area of the last five 100 ka eccentricity cycles (see below).
(Marchetti, 2002). Similar episodes of alluviation associated Quaternary fluvial records may therefore provide important
with prehistoric woodland clearance, and with prehistoric baseline data for predictive modelling of the relationships
and historic farming activity, have been recorded in Britain between long-term river behaviour and climate change
and Ireland (Foulds & Macklin, 2006) and throughout (Macklin & Woodward, 2009).
much of mainland Europe (Notebaert & Verstraeten, 2010).
In many areas of the world, therefore, Late Holocene
alluviation and associated terrace development is as likely 2.6.3 The terraces of the River Thames
to have been the product of human activity as of natural One of the most intensively studied river terrace sequences
processes (Syvitski et al., 2005). is that of the River Thames in southern England. The
Thames basin contains a long record of fluvial activity spatial relationships of the Thames terrace sequence (e.g.
and river terrace genesis that reflects the influence of a Bridgland et al., 2013). This new work has demonstrated
number of the factors discussed above. Research on the that the terrace sequence and associated sediment
terraces of the River Thames has a long history extending stratigraphy of the Thames Valley are considerably more
back well over a hundred years (Gibbard, 1994; Bridgland, complicated than previously envisaged. However, many of
1994). Work during the first half of the twentieth century the terrace names introduced by Wooldridge and Linton
focused primarily on the geomorphology of the Thames have been retained, as these refer to features that repre-
Valley, and on the discrimination of different terraces sent important stratigraphic marker horizons through-
(Wooldridge & Linton, 1955). Over the past two decades, out the Thames Valley, although the terrace fragments are
a more comprehensive approach involving not only geo- now named after the sedimentary units of which they
morphological mapping, but also lithostratigraphic and are predominantly composed (e.g. Boyn Hill Gravel;
biostratigraphic investigations of the terrace gravels and Lynch Hill Gravel; see Figure 2.50) rather than after the
their included organic deposits and associated archae- particular terrace levels (i.e. Boyn Hill Terrace; Lynch Hill
ology, has provided new insights into the temporal and Terrace).
a) T e r r a c e s t a i r c a s e of the M i d d l e T h a m e s ,
Nettlebed w e s t of L o n d o n , S E E n g l a n d
200 -2.5 Ma interglacial deposits
c. 81 Satwell
c S6 32
Gerrards Cross
150
Stoke Row
Altitude O.D. (m)
IS W i n t e r Hill U p p e r (deltaic)
c. 82
Stoke Row 12
100 W eW
s tel as nt lda n d Kempton Park
c. 66 4 64
Red Crag 12 L y n c h Hill 4
Formation Westland a
B e a c o n sfi e l d Westland Alluvium 1
50 (Little H e a t h / Green
Lane End c. 54 22 Hi R Thames
B o y n Hill
deposits; marine)
W i n t e r Hill L o w e r 12 10 Taplow Shepperton 2
0-
P r e - D i v e r s ion Post Diversion
b)
60 T e r r a c e s t a i r c a s e of t h e p o s t - A n g l i a n L o w e r T h a m e s ,
Dartford e a s t of L o n d o n , S E E n g l a n d
Heath S w a n s c o m b e B o y n Hill /
gravel deposits Aveley /
Orsett Heath
West Thurrock
40 Gravel
deposits Taplow I
11 10 early
L y n c h Hill I Mucking
Gravel Kempton Park /
Altitude O.D. (m)
C o r b e t s Tey
12 Gravel E a s t Tilbury M a r s h e s River
6
20 8 early Gravel Thames
late early
10. 4
Honchurch 9
till late Alluvium
7
12 1
0 Purfleet / G r a y s
deposits 8 late Buried
Trafalgar S q u a r e 6 channel
deposits mm
Interglacial d e p o s i t s Shepperton
-20 (upstream)
Gravel
Cold-climate gravels Glacial deposits 5e 2
Figure 2.50 Schematic transverse profiles through the best-preserved suite of the terrace staircase of a) the Middle Thames
(c. 35 km west of London) and b) the Lower Thames (c. 15 km east of London). Note that the higher and older terraces of the
upper Thames are not included, and the sequence has been simplified. The circled numbers refer to the marine isotope stage
(MIS) to which each terrace fragment has been assigned (from Bridgland & Westaway, 2008b).
Despite the fact that much has been learned about the the Devensian stage. In fact, much less time is actually
history of the Thames from these recent studies, difficulties represented in the sedimentary record at any one locality,
continue to be encountered in the interpretation of the for there are numerous non-sequences and minor uncon-
record. One reason for this is that the period of time formities within sequences that arise from the complex
represented by preserved deposits may be relatively short, nature of the depositional processes. Downcutting between
with evidence from the Devensian part of the sequence aggradations must account for a substantial part of the
suggesting accumulation within a timespan of no more than missing record, but the consequence is that many terraces
35–30 ka, that is, around 30 per cent of the time range of are locally poorly defined or fragmented and hence
a) Colchester a)
Anglian
Ice
Sheet
St. A l b a n s St. A l b a n s
Harlow CI acton
Moor Mill
d Chelmsford
tfor Chelmsford Lake
Wa
Southend
Watford LONDON
Slough LONDON
Southend Slough
Dartford
W
Wa
Dartford
at
Wtaf o
W
fo
Sl
at
rd
then
ug
rd
n
t frod
he
So ul
fo rd
fo
ut
S
tfo
Su lgoh
ou
rd
So
oguh
at
Wa
rd
th
en
W
Sou
Slohg
a)
0 10 k m
Other gravel terraces
Lower Thames Gravel
Low-level East Essex

Gravel
Till
High-level East E s s e x
Gravel
Kesgrave Group Gravels
Figure 2.51 a) Schematic representation of the drainage system prior to the advance of the Anglian ice, c. 450 ka. b) Diverted
drainage pattern and limits of the Anglian ice in eastern England during MIS 12. c) Distribution of gravel terraces in the lower
Thames Valley and adjacent parts of southeast England that post-date the Anglian glaciation (after Bridgland, 2006, and Bridgland
& Shreve, 2009).
correlation between individual terrace remnants and their basin were far removed from the coast, and are therefore
contained sediments is seldom straightforward. highly unlikely to be thalassostatic. The influence of sea-
A further complication is that the River Thames has level changes on the formation of the Thames terrace
changed its course over the course of the Quaternary, system has therefore been over-estimated (Bridgland,
principally as a result of ice sheets advancing into the 2000).
drainage basin from the north. Buried valleys, spreads of This view has been further reinforced by evidence which
fluvial gravels, and the distribution and gradient of certain shows that the Thames terraces, particularly in the lower
terraces show that the river originally flowed out to sea on reaches, form a well-ordered ‘staircase’, with the oldest
the present-day Norfolk coast some 100 km to the north terraces highest and younger lowest. Given that eustatic sea
of its present outlet. On at least two occasions prior to levels were close to (or slightly higher than) the present level
MIS 12 (before c. 450 ka), however, southward-moving ice during interglacials, and were lower (>100 m) during
sheets deflected the river into a new course that turned glacial periods, terraces related to former sea levels should
northeastwards immediately downstream from Reading, be confined within a narrow altitudinal range, and should
through the Vale of St Albans and across the present-day also show evidence of repeated reworking of sediments.
coastline near Colchester in Essex (Figure 2.51a and b). The That this is evidently not the case suggests the operation of
river finally gained its present course after the Anglian another mechanism, namely crustal uplift, induced by
glaciation (MIS 12). The consequence of these changes in repeated loading and unloading of the crust by glacier ice
river direction is that terraces of different age have diverging over successive glacial–interglacial cycles (Maddy, 1997;
surface gradients, which adds a further complication to Westaway, 2002). Glacial advance would have led to the
the correlation of terrace remnants. In addition, glacial removal of surface mass by erosion and transfer to the
encroachment into the Thames basin affected sediment adjacent North Sea basin, and that would have led to a
input to the river, with higher sediment yields in areas compensatory net inflow of the more mobile lower crust
adjacent to the ice margins, while glacio-isostatic loading after the ice load is removed. The overall result is incision
depressed the crust beneath the ice, and compensatory of the River Thames to ever lower valley floor levels in
forebulge effects (section 2.5.4) raised ground level beyond response to this background of regional uplift (Bridgland,
the ice margins. It has been estimated, for example, that 2006).
glacio-isostatic adjustments have raised the terraces of An equally important development in understanding
Anglian age in the Middle reaches of the Thames some the evolution of the Thames terrace sequence, particularly
55 m higher than corresponding terraces in the Upper in the lower reaches of the drainage system where post-
Thames, as a result of glacio-isostatic adjustments (Bridg- Anglian diversion terraces are preserved (Figure 2.51c),
land & Schreve, 2009). has been the emergence of evidence for climatic forcing
It was implicit in early interpretations of the Thames of terrace formation. A combination of biostratigraphy,
terrace system that the sequence, particularly in its lower notably mammalian and molluscan records, artefact
reaches, reflected primarily the influence of sea-level assemblages and amino-acid geochronology (Bridgland
changes, and hence terrace remnants standing above sea & Schreve, 2009; Penkman et al., 2013), has enabled the
level were generally regarded as thalassostatic features. It is interglacial deposits associated with individual terrace
now apparent, however, that the majority of the terrace fragments to be assigned to different warm stages, which
deposits aggraded under cold conditions, when sea levels can be correlated with marine oxygen isotope stages (Figure
were well below those of the present. In the upper parts of 2.50). The evidence demonstrates that a separate aggrad-
the drainage system, most terrace forms relate to increased ational terrace formed in each of the last five 100 ka
aggradation during deglaciation, when meltwaters were Milankovitch cycles, and that downcutting took place close
feeding large quantities of debris into the Thames system to warming transitions (Bridgland, 2006, 2010). During
from the northwest. Moreover, these cold-stage terraces the preceding cold phase, sediment transfer to the river
that lie above present sea level would have been some channel from poorly vegetated slopes led to aggradation
distance from the sea at the time of their formation. At the on the braided river floodplain. As the climate warmed
LGM, for example, the Thames flowed over exposed parts and vegetation developed, sediment supply diminished
of the English Channel to join the Rhine, Seine and other and river flow was confined to single channels, where
major European rivers, forming a large fluvial system that erosion could be focused and incision accelerated, while
eventually drained southwards to join the sea near the edge interglacial deposits accumulated in pools and areas of
of a wide continental shelf (Figure 2.2; Gupta et al., 2007). impeded drainage. Renewed onset of cold conditions initi-
Hence, terraces formed at that time in the lower Thames ated a further aggradational episode, which buried and
preserved remnants of these interglacial sediments. A much phases of expansion and contraction during the
longer record of terrace deposits is preserved in the pre- Quaternary, while levels of present-day lakes in tropical and
diversion section of the Middle Thames, although here subtropical regions also rose and fell. Abandoned lake
correlations with the MIS sequence are more speculative. shorelines (Figure 2.52a) and associated littoral deposits
After more than a century of research, a coherent (Figure 2.52b) are often found around large structural
account of the evolution of the complex sedimentary basins and surface depressions in which drainage is pre-
archive preserved within the Thames terrace sequence is dominantly internal (endoreic). Lakes that show evidence
now beginning to emerge. The current view is that the of expansion and contraction unrelated to worldwide
river terrace staircase is generated by long-term crustal changes in base level have been termed pluvial lakes, as their
uplift, while the internal stratigraphical detail of individual high-water stages have been attributed to wetter climatic
terraces is determined primarily by long-term climatic phases, termed ‘pluvials’ or pluvial episodes. Low water
variations, with the 100 ka eccentricity cycle proving to levels or periods of complete desiccation are, in turn,
be an important driving mechanism. The Thames terrace assumed to reflect ‘interpluvials’ or interpluvial episodes.
sequence therefore provides a highly complex record Because of the close correspondence that appears to exist
of Quaternary environmental change, and the multi- in closed-basin catchments between the status of a lake
disciplinary approach that has been employed to recon- (i.e. its volume, depth, surface area, etc.) and the balance
struct the history of this river system constitutes a template between seasonal precipitation and evaporation, temporal
for the analysis of fluvial sequences in other areas of the variations in lake surface level provide a proxy record of
world. continental climatic change, particularly during the Late
Quaternary.
2.7 QUATERNARY LANDFORMS Geomorphological evidence for the existence of pluvial
lakes and for oscillations in water level includes former cliff-
IN LOW LATITUDES lines, shorelines, beaches, bars, spits and deltas, as well as
There is abundant geomorphological and lithological abandoned watercourses that acted as overflows at times
evidence to show that major climatic changes have affected of high lake-level stands (Jewell, 2007). Frequently associ-
the tropical, subtropical and warm temperate regions of the ated with these landforms are freshwater evaporates, such
world during the Quaternary. The periodic expansions as marls and tufas (sections 3.7 and 3.8). The landforms and
and contractions of the great ice sheets in the high and mid- their associated deposits are not only distinctive in the field
latitudes were accompanied by shifts in the climatic zones (Figure 2.52), but often appear as prominent features on
of the low latitudes, leading to marked spatial and temporal aerial photographs and satellite images (e.g. Maxwell et al.,
variations in regional rainfall values, seasonal distribution 2010). Strandline evidence in particular enables the former
of rainfall, annual temperatures, and wind directions extent of Quaternary pluvial lakes, some of which occupied
and strengths. The effects were most pronounced in the considerable areas, to be established. In the southwest
desert and savannah margins of, for example, the Sahara United States, for example, where one of the greatest
(LeBlanc et al., 2006), northwest India (Juyal et al., 2006), concentrations of pluvial lake features in the world is to be
China (Zhang et al., 2004b) and Australia (DeVogel et al., found (Figure 2.53), Lake Bonneville, a precursor of the
2004), although fluctuations between humid and more Great Salt Lake, had a maximum surface area of 51,300 km2
arid periods also affected some mid-latitude regions, (the present-day Great Salt Lake is c. 4,000 km2), a volume
such as parts of western USA (Sack, 2009) and southern of c. 9,500 km3 and a depth of 370 m, making it the largest
Eurasia (Dogan, 2010). In many of these regions, fossil known pluvial lake in North America. The second largest,
landforms and deposits are preserved that can be used Lake Lahontan in Nevada, covered an area of 23,000 km2,
to infer climatic changes during the Late Quaternary. Of more than fourteen times larger than the existing lake
particular importance are lacustrine features that provide (Adams, 2010). In North Africa, Lake Chad (Figure 2.54)
evidence for wetter conditions at times in the past, and had a surface area of c. 350,000 km2 at its maximum during
sand-dune complexes (or ‘sand seas’) indicating former the Late Quaternary (the present lake fluctuates between
episodes of increased aridity. 10,000 and 25,000 km2), although its depth may have
been no more than 50 m (Schuster et al., 2005). Other large
Late Quaternary pluvial lake systems include Palaeolake
2.7.1 Pluvial lakes Makgadikgadi in the Kalahari Desert (Burrough et al.,
In many arid and semi-arid regions of the world, saline 2009), Palaeolake Suguta in the African Rift Valley of
lakes and ephemeral water bodies (playa lakes) experienced northern Kenya (Garcin et al., 2009) and Lake Mega-
a)
b)
Figure 2.52
a) Abandoned shorelines of Pluvial
Lake Lahontan near the northern end
of Pyramid Lake, near Reno, Nevada,
USA. b) Beach deposits (lower) and
fan gravels (upper) from Pluvial Lake
Lahontan (Mike Walker for scale;
photographs by Allan Ashworth, North
Dakota State University, USA).
Frome in the Lake Eyre Basin, which contains Australia’s factors contributing to lake-level fluctuations (Chepalyga,
largest endoreic drainage network (Cohen et al., 2011). 2007).
The most extensive Late Quaternary palaeolake so far Abandoned lake shorelines in many low-latitude
recorded, however, is the Caspian and Aral Sea systems of regions of the world therefore provide clear evidence of
central Eurasia which reached maximum dimensions both the former extent of palaeolakes and of fluctuations
around 16–15 ka (Chepalyga, 2007). At that time, the lake in lake water level, and where these primarily reflect changes
covered an area in excess of 1.5 million km2, stood more in atmospheric moisture balance, they can provide a basis
than 75 m above the level of the present Caspian Sea, for reconstructing past changes in precipitation regimes.
and extended some 3,000 km up the Volga River valley. The However, as is evident in the case of the Caspian–Aral
lake was different from other pluvial lakes, however, as it Sea palaeolake, lake basins may be affected in complex
was fed by waters from the Yenisei, Ob and Pechora rivers ways by climate change, especially those with catchments
which had been diverted southwards by the Eurasian ice periodically fed by meltwaters from glaciers or permafrost.
sheet (Mangerud et al., 2004), while spring flooding and Moreover, as enclosed lakes expand, their surface levels may
permafrost melt during warmer periods were additional reach new overspill cols, with the result that lake volume
Canada
USA"
4 54 5
Pluvial
Lationtan
Lake
•EH" B o nBnoenvni lelvei l l
Lationtan
USA
Bonnevill
Gu
lf o
Ice s h e e t s a n d g l a c i e r s
fC
ali
Pluvial lakes
fo
m
ia
Slate boundary
International boundary
0 500 km
Figure 2.53 The maximum limits of late Pleistocene pluvial lakes in the American
southwest (from Orme, 2008).
is no longer solely controlled by precipitation/evaporation shorelines appear as horizontal features and can readily be
ratios, which further complicates the links between lake correlated across the lake basin, in certain cases the
volume and rainfall regime. An additional difficulty may differential effects of hydro-isostatic loading and unloading
be the modification of lake drainage systems by humans, can lead to shorelines of a similar age being found at
notably during the Holocene, as appears to have been the different altitudes. In the Great Basin of Utah, for example,
case around the Aral Sea (Boomer et al., 2009). One final the weight of the waters of Lake Bonneville caused isostatic
problem arises from the use of geomorphological evidence depression of the crust, especially towards the middle of the
alone to infer past lake volume changes and their palaeolake where it was deepest (Godsey et al., 2005; Sack, 2009).
climatic significance. Although the majority of former lake Following the disappearance of the lake around 10 ka,
c a t c h m e n t limits
Ahaggar N
p r e s e n t - d a y rivers
m o d e r n L a k e C h a d ( 2 8 5 m)
L a k e M e g a - C h a d ( - 3 2 5 m)
Tibesti
20°N
Air'
Ennedi
Modem
Lake .
Chad
Darfur
10°N
1J0o°s N
r
ga
ag
Ah
over 1000m
r
ga
a 600- 1000m
aou
ag
dam 400 - 600m

Ah
A 200 - 400m
10°E 0 200 k m 20°E
Below 200m
Figure 2.54 Location and surface dimensions of modern Lake Chad, Palaeolake Mega-Chad, and the Chad
drainage catchment (from Schuster et al., 2005).
isostatic recovery led to warping of the shorelines so that be possible to construct a record (a hydrograph) of pluvial
the main Bonneville Shoreline is now c. 74 m above the lake-level fluctuations through time. In Lake Bonneville, for
former lake margins (Figure 2.55a). example, dated strandline evidence indicates at least six
The reconstruction of histories of lake-level variations, major water-level fluctuations with amplitudes of up to
therefore, requires not only detailed mapping of the pre- 50 m between 21 and 10 ka (Figure 2.55b), and it has been
served palaeolake features, but also a reliable chronology suggested that these may reflect the millennial-scale climatic
of shoreline formation, as well as a knowledge of the wider fluctuations that affected the North Atlantic region during
palaeoenvironmental context in which the lake system has that time interval (Oviatt, 1997; Orme, 2008). More detailed
evolved (Bohacs et al., 2003). In these circumstances, it may palaeoclimatic reconstructions based on data from closed-
a)
300 200 100 km 0 km 100 200
Lakeside Mountains
n
SSW NNE
1 6 2166 2 6
ine
h orel 1620
lle s
elevatio
crj
evi
dBonn 1600
adjusted
de 74 m
oun 5
Escalante
Reb a) 1580 Cache
Desert Valley
Original Bonneville shoreline 1560 1552m
ta) 1 5 5 0 Bonneville
adjusted elevation (m)
£ Provo,
o 1450
=•
,Stansbury
1350
3
Gilbert
1250
5 10 15 20 25
r a d i o c a r b o n a g e (ka)
Figure 2.55 a) Post-drainage hydro-isostatic deformation of the Bonneville Shoreline that formed during the maximal extent of
Pluvial Lake Bonneville (from Sack, 2009). b) Chronology of water level changes in Lake Bonneville over the past 30 ka. Note the
catastrophic drop in lake level of around 100 m from the Bonneville to the Provo Shorelines. This was caused by a major flood
event and is reflected in a distinctive marker bed in deep-water sediments (after Oviatt, 1997).
basin lakes can be made where the geomorphological Amazonia and the Gulf of Mexico coast plain (Latrubesse
evidence is supported by stratigraphic records, and these & Nelson, 2001; Otvos, 2004), and on the pampas of
aspects of pluvial lake sequences are discussed further in Argentina (Tripaldi & Forman, 2007). The geomorpho-
section 3.7. logical evidence in all of these regions indicates phases of
significantly increased aridity or ‘megadroughts’ in the
past, and particularly around the time of the Last Glacial
2.7.2 Dunefields Maximum (LGM) (Hanson et al., 2010).
Increased aridity at times in the past can be demonstrated In some areas it may be difficult to make a clear
by the existence of desert landforms, especially sand dunes, distinction between active and fossil dunes, particularly
in areas where such features are no longer forming. Even where sand has been remobilized and secondary dune
when heavily vegetated, dunefields often stand out clearly patterns have become superimposed on an older primary
on aerial photographs and satellite images (Brookes, 2003), set. Most fossil dunes have been distinguished on the basis
so that mapping is possible of inactive dune systems beyond of features indicative of a period of stability under humid
the margins of the present desert regions. Such evidence climatic conditions. These include evidence of gullying
shows, for example, that the Sahara once extended south and truncation by fluvial activity, and features produced by
into the Sahel to latitude 10–12°N over a distance of some pedogenesis and chemical weathering, including chemical
5,000 km, while fossilized dune systems occur throughout alteration of clay minerals, decalcification and staining by
southwest Africa, northwest India and around the margins iron oxide. In hyper-arid regions, these usually take the
of the Australian arid zone (Figure 2.56). Stabilized dunes form of calcrete or gypsum concentrations (section 2.7.4),
have also been found in present-day humid forest zones of sometimes associated with illuviated clay coatings that
NORTH AMERICA
Great Southwestern
AFRICA AUSTRALIA ASIA Great Plains
Basin Deserts
North Eastern Namibia

Central & Eastern
Strzelecki Desert
Namib Sand S e a
Soldier Mountain
Western Zambia
Simpson Desert
Sonoran Desert
Kalahari Desert
Kalahari Desert
Kalahari Desert
Arabian Desert
Wahiba Sands
Sahara Desert
Sahara Desert
Mojave Desert
Mojave Desert
Mojave Desert
Saskatchewan
Sinai & Negev
Gran Desierto
Namib Desert
Kelso Dunes
Thar Desert
AFRICA
Mauritania
Minnesota
Dale Lake
Nebraska
Wyoming
Colorado
Deserts
Nevada
Oregon
Oman
Texas
Libya
UAE
0
10
15
20
25
30
35
40
45
50
Figure 2.56 Principal dune-building episodes (shaded) during the last 50 ka in the deserts of Africa, Australia, Asia and North
America (from Tchakerian, 2009).
have formed around pedogenically altered sand grains truncating older dunes or being buried by later sand-dune
(Fitzsimmons et al., 2009). Former dune surfaces may also expansion (Drake & Bristow, 2006). Such geomorpho-
be indicated by archaeological evidence, for example logical relationships provide convincing evidence for the
concentrations of human or other animal bones, indicating alternation of arid and pluvial phases. In the Erg of Bilma
a sufficient degree of sand stabilization and moisture in eastern Nigeria, for example, dune formation in the
availability to permit humans to occupy previously hyper- period 23–19 ka was followed by a more humid episode
arid regions (Nicoll, 2001). Vegetated dunes may indicate that culminated during the early Holocene (c. 10–9 ka) in
relict features, although as some dunes may be preferential the development of an extensive pluvial lake system. Drier
sites for plant growth in arid environments, vegetation conditions around 6 ka led to a second phase of dune
cover may simply reflect better water-retaining properties construction and, following disappearance of the lakes, a
and deep-rooting opportunities offered in such areas. third episode of dune expansion began after c. 3 ka. The first
Chronologies of dune formation can be obtained using two generations of dunes show evidence of weathering, but
luminescence dating of sand grains (Lancaster, 2008; section the third is still active at the present day (Völkel & Grunert,
5.3.6.) and, in less extreme environments, by radiocarbon 1990). Dunefields may also occur in conjunction with
dating of organic carbon residues in palaeosols (Goble et fluvial landforms and deposits, with erosion of dunes by
al., 2004). In addition, CRN dating (section 5.3.8) enables streams during humid periods and the avulsion of stream
ages to be obtained on eroded rocks and on desert pavement channels by dunes during more arid episodes. In the eastern
surfaces (Fujioka et al., 2009). MacDonnell Ranges of central Australia, for example
In some areas, fossil dunefields have been found on the (Figure 2.57), fluvial erosion prior to the LGM created
beds of former pluvial lakes, with shorelines either desert surfaces of different ages on which successive
MacDonnell
Lake Eyre
'Ranges
Basin
Basin
Miirrav-Dariina
Basin
Basin
Basin
N
0 500km
Arid z o n e boundary Longitudinal dune systems
Lunette z o n e s Dust paths
Figure 2.57 Dominant linear dune trends and present-day dust circulation in Australia (from Hesse & McTainsh, 2003).
generations of dunes developed, with an earlier generation been identified, the oldest emplaced during the LGM
(75–65 ka) preceding the main fluvial phase and a later set (25–15 ka), a second set during the Younger Dryas (13–10
of dunes forming on abandoned river floodplains during ka) and the youngest after c. 5 ka. The trends of the dune
the LGM and Holocene. crests suggest that the pre-Holocene features reflect easterly,
Fossil dunefields also provide valuable data on past northerly and northwesterly winds, resulting from the
wind directions and strengths, thereby permitting former intensification of the seasonal high-pressure air cell over
synoptic atmospheric circulation patterns to be inferred. In North Africa (Lancaster et al., 2002).
the example from central Australia referred to above, Dune patterns also provide valuable input data for, as
contrasts in alignment and distribution of the two dune sets well as independent tests of, numerical simulations of past
suggests that the circular Australian wind system shifted atmospheric circulation. For example, the mapped pattern
southwards by some 160 km (1.5°) after the LGM (Hollands of parabolic dune crests and other aeolian indicators in
et al., 2006). In the Western Desert of Egypt, north- and northern Europe, dating to between 20 and 15 ka, accord
northeast-trending dune crests cross-cut older west- and closely with model-based simulations of the prevailing
northwest-trending ones (Figure 2.58). The latter are of wind pattern for that period. Both the empirical evidence
early Holocene age and were formed by ‘palaeo-westerly’ and modelling outputs suggest winds that were generally
winds that steered moist Atlantic/Mediterranean air masses more westerly-to-northwesterly and significantly stronger
across Egypt towards the Red Sea. The younger dunes, than those of the southwesterly air stream that dominates
orientated along a north–south axis, developed later the region today, the former being attributed to a stronger
than 4–3 ka, at about the time the north African savannah circulation regime induced by the northern ice sheets and
boundary stabilized at its present position (Brookes, 2003). a colder North Atlantic Ocean at the end of the last cold
In Mauritania, three generations of dune formation have stage (Renssen et al., 2007a). Finally, there is evidence in
some palaeo-dune records of cyclical variations in aeolian if correct inferences are to be made about the nature and
deposition, for instance a possible precessional-forced cycle timing of periods of dune formation, and their links to
of dune accretion in the Thar Desert in India over the last climatic change. A further problem arises from the activities
200 ka (Singhvi et al., 2010) and a 2.6 ka cycle in dust of human groups on the desert and savannah margins. It
delivery to the Gulf of Carpenteria in northern Australia is clear that, over recent millennia, people have played an
over the last 40 ka (De Deckker, 2001). increasingly important role in the expansion of the desert
Although fossil dunefields are therefore potentially margins, and human-induced desertification has to be
valuable indicators of the former extent of arid environ- considered when palaeoenvironmental inferences are being
ments, again care must be exercised in their use as made on the basis of dunefield evidence, particularly for the
palaeoclimatic indicators. Desert conditions result from late Holocene (Nicoll, 2004).
a variety of climatic factors, the most important being
temperature, precipitation and wind. A change in any of
these variables may be sufficient to alter the balance between
2.7.3 Fluvial landforms
adequate groundwater retention for plant growth and Relict fluvial features are found throughout the tropics and
excessive evaporation leading to drought. Hence, a change subtropics, and are clearly revealed in remote sensing
towards an episode of dune construction may result from surveys (Ori et al., 2009). For example, Landsat and SLAR
higher temperatures, lower precipitation levels, increased imagery has identified integrated fluvial channel networks
wind strength, or a combination of all three. Similarly, it around oases in the eastern Sahara (Robinson et al., 2006),
cannot be assumed that higher precipitation levels con- in the Lake Chad basin of the southwestern Sahara (Leblanc
stitute the only environmental factor leading to sand-dune et al., 2006), and in the arid region of Rajasthan in north-
stability, as this is partly dependent on a balance between west India (Rathore et al., 2010). In parts of interior
sediment availability, supply and mobility (Lancaster, Australia, complex systems of palaeochannels and palaeo-
2008). There is also evidence to suggest that stabilization floodplains, many of which contrast markedly with the
of dunes following a major climatic shift can be time- form of contemporary alluvial channels, appear to have
transgressive over large areas (Werner et al., 2011). For developed under climatic regimes very different from those
transitional periods, therefore, additional lithological, of the present day (Nanson et al., 2008). Fluvial response
biological and chronological evidence may be required to climatic fluctuations is also reflected in river sediment
sand
plain
hamad
a
sand
plain
0 km 10 0 km 10
Figure 2.58 Linear trends of crests of sand dunes in Egypt’s Western Desert (from Brookes, 2003).
records, for example on the Deccan Plateau in India where 255crcr

,
ae-'O'o -
2$°w00 ze-artr
alternations between episodes of alluviation during wetter
periods and fluvial incision during drier phases are
20 600
attributed to centennial or millennial variations in the
20 600
strength of the Indian monsoon (Kale, 2007).
The identification of palaeo-drainage channels in semi-
arid areas is not always straightforward, however, because
modern rivers in these areas are prone to seasonal flooding
and frequent channel shifts, making it difficult to dis-
20 600
20 600
tinguish between degraded modern and much older, 0 10
a fkOl
permanently abandoned, features (Tooth, 2000). Moreover,
palaeo-channels often tend to become obscured by shifting
aeolian and alluvially deposited sands. To some extent, the
20 600
20 600
latter problem can be overcome by the use of synthetic
aperture radar (SAR), which can penetrate surface sands
to reveal buried fluvial channel networks (Figure 2.59). SAR
also enables groundwater aquifers to be detected beneath
sand sheets and dune systems, and it has also revealed
archaeological sites in close proximity to palaeo-drainage
20 600
20 600
networks in what are currently hyper-arid and inhospitable
regions (Kindermann et al., 2006).
In reconstructing palaeoenvironments of the arid
and semi-arid zone from relict fluvial landforms, however,
20 600
20 600
a range of non-climatic factors must be taken into account

(Tooth, 2000). Fluvial processes are governed by a range
of environmental variables that include, in addition to
climate, geology, relief, soil, vegetation cover and the influ-
ence of people. All of these, in combination, will affect the
20 600
20 600
rate at which fluvial activity proceeds as well as the geo-

morphological response, and again climate may not always
be the dominant factor. For example, in highly arid regions, K-Crrjr HîWT » 2C(r
-
increased sediment load in rivers may result from increased

rainfall, whereas in the semi-arid zone, lower rainfall levels Figure 2.59 a) LANDSAT ETM image of part of the desert
area of northwest Sudan, showing the dominance of surface
may lead to higher sediment yields as a result of reduced sands, with little indication of the palaeo-drainage forms buried
vegetation cover. Moreover, where large rivers drain into beneath. b) Radarsat-1 image of same area which uses syn-
arid areas, changes in flow regime may not necessarily thetic aperture radar (SAR) to penetrate dry sand deposits to
reflect the ambient climatic conditions in the desert, but reveal buried palaeo-drainage features below. c) Digital
elevation model (DEM) of the same area based on 90 m spatial
may have been caused by changes in precipitation in the
resolution Shuttle Radar Topography Mission (SRTM) data to
region beyond the desert margins (Lancaster, 2000). enhance the buried river channel system (from Ghoneim &
El-Baz, 2007, reprinted with permission from Elsevier).
2.7.4 Weathering crusts

Weathering crusts, or duricrusts, are resistant ground concentrations develop as hardpans and, when exposed,
surface mantles or cappings commonly found as protec- form cemented (indurated) layers which are more durable
tive layers at the surface of eroded bedrock or sediments in than adjacent sedimentary horizons (Nanson et al., 2005).
low latitudes (Dixon & McLaren, 2009). They originate There is a wide range of weathering crusts of different
through the concentration of chemical elements in soils, chemical constituents but, in general, they can be divided
sediments or permeable rocks, either by accumulation of into those that originate as weathering layers in humid
residual compounds following the removal of more mobile tropical environments, and those that develop under
elements by solution or translocation, or by the reprecipi- arid or semi-arid conditions. Laterites, for example, evolve
tation of soluble compounds from groundwaters. These through the accumulation of residual hydroxides of
aluminium and iron in humid tropical soils which, when 2.8 CONCLUSIONS
exposed, form cemented surface layers. Laterite or ferricrete
(iron-rich) crusts are typically found as caprocks in desert Geomorphological evidence provides a valuable starting
and savannah regions where their presence on plateau point in the investigation of Quaternary environments.
surfaces reflects inversion of relief (Ollier & Sheth, 2008). By using modern landforms as analogues, aspects of former
Their occurrence also indicates a major change in climate, glacial, periglacial, fluvial, marine and aeolian environ-
for they are believed to form where both temperature and ments can be inferred. In many cases, only relatively
precipitation levels are high, although the precise climatic general conclusions are possible, but in certain instances,
parameters governing the formation of laterites are diffi- specific climatic parameters can be deduced. Throughout
cult to establish. Moreover, as deep-weathering profiles this chapter, however, it has been stressed that geomor-
and ferricretes are usually polygenetic in origin and of phology is but one of the lines of evidence used in the
considerable antiquity, they tend to be of relatively limited reconstruction of Quaternary environments and that,
value in inferring former climatic conditions during the wherever possible, landform records should be employed
course of the Quaternary (Widdowson, 2007). in conjunction with other forms of evidence. A second
Weathering crusts that formed under more arid condi- major data source lies immediately beneath the earth’s
tions include calcrete (sometimes termed caliche), which surface, and it is to the stratigraphic record that we now turn
is composed of cemented calcareous horizons and often our attention.
forms the hard rim of exposed escarpments in deserts and
savannah regions (Alonso-Zarza & Wright, 2010), gypcrete
or gypsum cement (Aref, 2003) and silicious crusts known NOTES
as silcrete (Nash et al., 1994). These weathering crusts and 1 For survey purposes, small sectors of the earth’s surface can be
related structures in soils have been found in many low- considered flat, and no account need be taken of surface
latitude regions and have been interpreted as reflecting curvature. For larger areas, however, the earth’s curvature needs
Quaternary climatic change; they are often degraded, to be accounted for and appropriate corrections applied. This
however, which suggests that some may be relict forms from is the study of geodesy, and the determination and application
previous arid phases (Hirmas & Allen, 2007). Again, the of fixed positional reference points is known as geodetic survey.
climatic conditions governing their formation are diffi- 2 A differential global positioning system (DGPS) provides more
cult to quantify, partly because they are highly variable in accurate measurements of ground position than the less
composition, frequently intergrading between different expensive ‘recreational’ GPS instruments. It involves the use
of at least two receivers, one of which is left stationery (the
duricrust types (Ringrose et al., 2009). Other factors also
‘reference station’) at a point which has been accurately surveyed
contribute to their formation, especially in the case of
for position and altitude; the second (‘roving’) receiver can then
pedogenic carbonate, which appears to be largely bio- measure the location and altitude of other points relative to the
logically produced (Alonso-Zarza & Wright, 2010). The position of the reference station using more accurate compu-
ages of different phases of carbonate production can be tations of coordinate error (see http://www.trimble.com/
determined by uranium-series dating (section 5.3.4), and gps_tutorial/dgps or Spencer et al., 2003a).
when the results are compared with independent palaeo- 3 ‘Sandar’ is an Icelandic term to describe the outwash plains that
climatic estimates, climatic control over carbonate pro- form in lowland areas in front of valley glaciers and ice sheets.
duction seems to be implied. For example, data from Spain The singular is sandur.
spanning the last 120 ka suggest that pedogenic carbonate 4 In glaciology, the term ‘steady-state’ refers to the condition
has formed predominantly under warmer and more arid whereby a glacier or ice sheet maintains equilibrium over time
climatic conditions (Candy & Black, 2009), while in Chinese as a result of a balance between the major controlling variables
loess sequences, carbonate nodules in soil horizons appear such as seasonal climatic variations, albedo, rate of flow, etc.
5 Glacier surges are short-lived, often substantial, advances of an
to have developed during colder and more arid phases
ice sheet or glacier over timescales ranging from a few years to
(Rowe & Maher, 2000). Where a chronology of carbonate
several centuries (Benn & Evans, 2010).
production can be established, therefore, calcrete–climate 6 BRITICE is a GIS-based compilation of glacial and associated
relationships may prove to be useful as generalized palaeo- landforms in and adjacent to the British Isles (see http://www.
climatic indicators. In many cases, however, establishing the shef.ac.uk/geography/staff/clarkchris/britice.html).
age of weathering crusts is seldom straightforward for, as 7 Tunnel valleys are large over-deepened channels cut into
with laterites, not only do many have polygenetic origins, bedrock or sediment near the margins of continental ice sheets,
but they may have been forming since early or even pre- and they often terminate abruptly at major moraines (Benn &
Quaternary times. Evans, 2010).
8 A roche moutonnée is a form of ice-moulded and polished and therefore the firn line and the equilibrium line coincide
bedrock, elongated in the direction of ice flow, with a distinct (Paterson, 1994).
shallow gradient on the up-glacier (stoss) surface and a steeper 11 The Phanerozoic is the period of time during which sediments
irregular surface (caused by ice plucking of rock fragments) on have accumulated that contain remains of plants and animals.
the down-glacier (lee) surface. The term was coined by the It refers to the geological systems from the Cambrian to the
Alpine geologist Horace-Bénédict de Saussure, who saw a Quaternary and covers approximately 540 Ma of the geological
resemblance between a land surface with well-developed roches timescale.
moutonnées and the curled ornamentation of wigs worn by 12 Juvenile water originates deep in the earth, possibly from
French gentry in the eighteenth century. processes associated with degasification of mantle material
9 The Pascal (Pa) is the SI unit for pressure, and is the measure during metamorphism and magmatism. Juvenile water reaching
of force per unit area. In glaciology, pressure of ice is usually the crust mixes with water derived from precipitation (meteoric
measured in kilopascals (kPa), in other words 1 kPa = 1,000 Pa. water). Higher levels of carbon dioxide, helium and oxygen in
10 In some glaciers, particularly in the High Arctic, there may be groundwater are indicative of the presence of juvenile water.
a complex zone of transition between accumulation and ablation 13 Rheology is the branch of Newtonian mechanics that deals with
areas. This is usually referred to as the superimposed ice zone, the deformation and flow of materials that are neither solid nor
in which ice will have formed from the refreezing of meltwater completely liquid, such as semi-molten (ductile) rock or glacier
runoff or avalanche material from up-glacier. The lower ice. The term deformation refers to the process by which a
boundary of the superimposed ice zone is the equilibrium line substance changes shape without breaching continuity.
(up-glacier from which there is a gain in mass), whereas the firn 14 Epeirogenic movements of the earth’s crust and surface are those
line, that is, the line dividing fresh snow from ice, forms the that involve gentle uplift or subsidence over very large areas,
upper boundary. On most maritime temperate glaciers, where with little folding or other localized deformation of rock
snowfall is heavy, there is frequently no superimposed ice zone, formations.
3
CHAPTER THREE
Lithological
evidence
3.1 INTRODUCTION 3.2 FIELD AND LABORATORY

Although geomorphological evidence can provide useful METHODS
insights into former climatic regimes and environmental
conditions, a more detailed impression of events during the
3.2.1 Sediment sections
Quaternary can often be gained from the sedimentary Wherever possible, lithological investigations should be
record. Not only can valuable data on Quaternary envir- carried out on open sections so that variations in strati-
onments be obtained from the sediments themselves, by graphy, both vertically and horizontally, can be carefully
relating observations on present depositional environ- recorded. Before commencing fieldwork, the section should
ments to features preserved in the recent stratigraphic be cleaned of slumped material, and a ‘fresh’ face revealed
record, but since many deposits are fossiliferous, inferences by cutting back into the exposure. On large sections, steps
based on lithological changes can often be supported can be cut on slumped material or directly into the face to
directly by those based on fossil evidence. Furthermore, provide temporary working surfaces, but a ladder may be
because sedimentary sequences frequently reflect sediment necessary to reach the less accessible parts. Careful draw-
accumulation over an extended time period, some appreci- ing of the exposure is the first stage in analysis, and this
ation can be gained of both spatial and temporal aspects of should be supported wherever possible by a photographic
environmental change. Finally, while geomorphological record. It may be useful to grid the face with measuring
evidence is restricted largely (although not wholly) to the tapes as this will provide an accurate scale for section
terrestrial environment, sedimentary data can be obtained drawing. Detailed notes should be taken on all aspects of
from beneath the waters of present-day lakes, from the exposed stratigraphy, using Munsell colour charts to obtain
world’s ice caps and, perhaps most important of all, from a relatively precise description of colour changes between
the deep-ocean floors where lengthy sequences of virtually and within lithostratigraphic units. Where a mixture of
undisturbed deposits are preserved. organic and inorganic sediments is under examination,
Quaternary sediments are generally unconsolidated and descriptive schemes can be used to classify the different
are of two principal types: inorganic (clastic) deposits, sediment types (e.g. Schnurrenberger et al., 2003). Instru-
consisting of mineral particles (termed clasts) ranging in mental levelling from a benchmark will enable the various
size from large boulders to very fine clays; and biogenic stratigraphic features to be related to a common datum,
sediments, consisting of the remains of plants and animals. for purposes of altitudinal comparison both within
Biogenic sediments can, in turn, be divided into an organic and between sites. A detailed guide to the description of
component of humus and the decayed remains of plants Quaternary stratigraphic field sections is provided by
and animals, and an inorganic component of such elements Jones et al. (1999).
as mollusc shells and diatom frustules. In this chapter, we The type of sampling framework employed will depend
are concerned primarily with inorganic sediments such as on the nature and purpose of the investigation. For certain
tills, aeolian and cave sediments, and fossil soils, although types of study, for example the analysis of soil or loess
some of the properties of biogenic sediments are also profiles (both discussed in this chapter), a sequence of
considered. A full discussion of the fossil record contained samples may be required from a representative vertical
in Quaternary sediments can be found in Chapter 4. section of the exposure. Depending on the aims of the
94 LITHOLOGICAL EVIDENCE
investigation, these might vary from relatively small samples section. These can be selected subjectively, but random
(perhaps no more than 1 cm3) to larger bulk samples sampling of a face that has been gridded will provide a more
weighing several kilograms, both taken at a set vertical objective sampling framework. The same approach can be
interval. In other circumstances, it might be necessary to applied where measurement of sedimentary properties
remove a complete sequence of samples, either in the form needs to be undertaken in the field, such as the recording
of monoliths (perhaps 20–30 cm3) cut from the section of the orientation and dip of pebbles or other clasts (termed
with a spade, or in metal boxes (monolith tins) which can ‘sediment fabrics’: see section 3.3.6.2).
be hammered into the face (Figure 3.1). In each of these
cases, sampling horizons should be carefully related to a
measuring tape attached to the free face at the side of the
3.2.2 Coring
sampling line; the trowels, spades, knives and spatulas Although section work is preferable to coring, relatively
should be cleaned between the extraction of individual few natural sections are to be found, and there are many
samples; and care must be taken over the packaging, sealing situations where it is impossible to excavate exposures
and labelling of the sediment samples. Detailed notes because of, inter alia, problems of time, expense, sediment
should be made throughout the sampling process. In other thickness or the likelihood of waterlogging. Hence, recourse
types of investigation, it may be necessary to take a number must be made to coring. A range of coring equipment is
of samples from points scattered across the face of the now available, with different operating mechanisms and
different levels of success in core recovery. Hand-operated
corers of either the side-sampling (e.g. ‘Russian’) or piston
(e.g. ‘Livingstone’) type are widely used for sampling peat
and lake sediments, although motor-driven equipment
(e.g. ‘rotary’ or ‘percussion’ corers) is required for more
cohesive sediment such as gravels and tills (Figure 3.2).
Specialized motor-driven or hydraulically operated corers
are also used to sample sediments in deep lakes or on the
seabed, and are also employed to obtain cores from glaciers
and ice sheets.
Numerous problems are encountered in coring
operations, including the facts that (1) cores may be dis-
torted during recovery or extrusion from the sampling
chamber, a problem that becomes particularly acute when
cores are being taken from poorly consolidated sediments,
such as certain types of peat and lake muds; (2) unless
overlapping cores are taken, stratigraphic units that are
either thin or of limited lateral extent may be missed during
sampling; and (3) it is often difficult to transport heavy
coring equipment to remote areas where other logistical
problems may also be encountered, such as a lack of water
for hydraulic coring operations. Despite such difficulties,
however, the development of increasingly sophisticated
coring machinery in recent years has enabled the successful
drilling of marine sediments, of the polar ice sheets and of
deep lake sequences, all of which have revolutionized our
understanding of the Quaternary environmental record
(section 1.5).
Figure 3.1 Monolith tins (50 cm length) in a sediment expo- Since most Quaternary sedimentary sequences, when
sure in the Kopanica Valley, southwest Poland. The monolith traced laterally, vary in both thickness and complexity,
tins were used to collect contiguous sediment samples for the sedimentary history of a site is often difficult to ascer-
tephrochronological analysis (section 5.5.2); overlapping sam-
tain when only a single core is taken. It is common prac-
ples between adjacent tins ensured the complete sequence
was sampled (photograph by Alison MacLeod, Royal Holloway, tice, therefore, to obtain several borehole records, usually
University of London, UK). arranged in transects, so that a two-dimensional, or even
LITHOLOGICAL EVIDENCE 95
a) b) c)
Figure 3.2 Examples of coring equipment used to sample Quaternary sediments. a) The hand-operated ‘Russian’ sampler with
internal chamber of 1 m length and 5 cm internal diameter, employed mainly to sample soft, Late Quaternary lake sediment and
peat. These can be manufactured to suit different purposes, for example b) a wide-diameter (10 cm) version is suitable for plant
macrofossil studies, as much larger samples can be recovered. c) An Ecologia Joy-4 rotary drilling machine recovering cores from
an 18.5 m thick sequence of resistant bedded sands, silts and laminated clays at the Middle Pleistocene site of Marks Tey, Essex,
UK, where sediments spanning MIS 12 to MIS 10 are preserved. Overlapping cores of 3 m length and 10 cm diameter were obtained
from two adjacent boreholes to ensure that the complete sequence was sampled (photographs by Alison MacLeod and Adrian
Palmer, Royal Holloway, University of London, UK).
three-dimensional, schematic model of lithological varia- provides a detailed and precise record of the lithological
bility can be constructed. The number of boreholes em- variations represented in the cores before destructive
ployed in such an exercise will depend upon the area to be sampling procedures are applied. Enhanced imaging using,
surveyed, the complexity of the lithological sequence, the for example, X-radiography, often reveals important
time available for fieldwork and other logistical consid- lithological variations that are not clear to the naked
erations, such as costs. Remote sensing techniques, particu- eye or shown by standard photography (Figure 3.3). A
larly seismic and sonar sounding methods, may also be used number of scanning methods are also available that can
in conjunction with coring (section 2.2.2). measure down-core variations in abundance of chemical
Once recovered, the surfaces of the cores should be elements in sediment, such as X-ray fluorescence (XRF),
cleaned and then analysed using a range of rapid, non- with a potential scanning resolution of c. 100 μm (Rothwell,
destructive, down-core scanning methods that are now 2006). Figure 3.3 shows how relative variations in the ratio
becoming routine. The first step is photo-imaging, which of Fe (iron) to Ca (calcium), measured using an ITRX core
scanner, reflect alternations between silt and clay layers 3.2.3.1 Particle size measurements
(varved deposits: section 5.4.2) in a former lake basin, and
which enables lithological boundaries between layers to be Particle size distribution is a most important diagnostic
clearly identified. The images and comprehensive suites of property of a body of sediment, for even very subtle
analytical data can be visualized and integrated using variations in average grain size or in the range of sizes
software developed specifically for the analysis of core of clasts may reflect important changes in sedimentary
material (marine, terrestrial and polar ice cores), such as environment. The particle size distribution of coarser
CoreWall (http://www.corewall.org/). grades of sediment (sand size and above: Table 3.1) can be
established by sieve analysis, but sedimentation methods
are required for finer materials. Different grades of sedi-
3.2.3 Laboratory methods
ment suspended in a liquid will settle under gravity at
Laboratory analysis of Quaternary sediments is an inte- different rates and these can be established either by using
gral part of palaeoenvironmental investigations. Both the a hydrometer to record changes in the density of the
physical and chemical properties of sediments can provide suspension over time, or by extracting subsamples from
valuable data on the nature of former depositional environ- the suspension using a pipette and then measuring the
ments, and are often useful indices of climatic and other changing concentrations of suspended matter over time by
environmental changes. A very wide range of laboratory successive weighings (Gale & Hoare, 2011). More sophisti-
methods is now available, and a detailed account is beyond cated approaches to particle size analysis include the use
the scope of this book. In this section, however, we provide of X-radiography (Andrews & Principato, 2002), pulse
a brief introduction to the methods that are most com- counters (such as the Coulter counter) and laser diffraction
monly used in the description and analysis of Quaternary (Blott et al., 2004). In pulse counters, sediments are
sediments. For further details the reader is referred to Gale suspended in an electrolyte and passed through an
& Hoare (2011), Evans & Benn (2004), and to the Technical electrode-flanked aperture; voltage pulses proportional to
Guides produced by the Quaternary Research Association the volumetric size of the particles are counted and the
of Great Britain (e.g. Jones et al., 1999; http://www.qra. particle size distribution established (Molinaroli et al.,
org.uk). 2000). Laser diffraction uses the forward scattering of a laser
Glacial Lake Blane: Varve Analysis
ISOGWOOO 'MIT RAX

K HMJklfifUfTH CORF IkUGf 4
0 0 150 150 150 150 150 150 150 150 150 150 150 150 150
150 150 150 150
150 150
150 150
150 150
150
150 150 150

150 150 150 150 150 150
150 150
150 150 150 150
150 150
150 150 150 150
150 150 150 150 150 150
150 150 150150 150
150 150
150 150 150
150 150
150
150 150
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 ITRAX XRF:
X-Rattiographs; L i g h p C l a y ; Dark^Sill F e F/ Ce a/ C a
Figure 3.3 X-radiograph images of laminated (varved) lake sediments that accumulated in a former ice-dammed lake near
Loch Lomond, Scotland. On the right is a graph showing variations in Fe/Ca as measured by an ITRAX core scanner (see text)
based at the National Oceanography Centre, Southampton, UK (images provided by Alison MacLeod, Royal Holloway, University
of London, UK).
Table 3.1 The Wentworth scale of particle size the division of pebbles into classes ranging from angular
fractions and the equivalent φ (phi) units. Phi units to rounded, and the direct measurement of particles
are obtained by conversion from the millimetre scale, themselves. Shape measurements are then based on the axial
where phi is –log2 of the diameter in millimetres. ratios of the particles and include a variety of indices of
The phi scale has the advantage of using integer elongation, roundness, flatness and sphericity (to derive
numbers only, making the statistical description of classes of blades, rods, spheres and discs, for example).
sediments more straightforward. These methods are described in Gale & Hoare (2011) and
Name mm scale φ unit
Evans & Benn (2004). Automated methods of particle
form analysis (e.g. image analysis) are discussed in Blott &
Boulder > 256 > –8
Pye (2008).
Cobble 256–64 –8 to –6 In recent years, micro-scale features and structures of
Pebble 64–4 –6 to –2 sediments have been increasingly widely used as diagnostic
Granule 4–2 –2 to –1 properties of different depositional environments. Micro-
Very coarse sand 2–1 –1 to 0 morphological analysis of samples is typically undertaken
Coarse sand 1–0.5 0–1 either at low magnification (10–100×), using thin sections
Medium sand 0.5–0.25 1–2 of sediment and petrographic microscopes, or at higher
magnification (100× and greater) using a scanning electron
Fine sand 0.25–0.125 2–3
microscope1 (Carr, 2004). Micromorphology was initially
Very fine sand 0.125–0.0625 3–4 most widely used in Quaternary soil science (section 3.5.3),
Coarse silt 0.0625–0.0312 4–5 but is now being employed in other sedimentary contexts,
Medium silt 0.0312–0.0156 5–6 most notably in the analysis of glacigenic deposits (e.g. van
Fine silt 0.0156–0.0078 6–7 der Meer et al., 2003; Kilfeather et al., 2010).
Very fine silt 0.0078–0.0039 7–8
Coarse clay 0.0039–0.00195 8–9 3.2.3.3 Surface textures of quartz particles
Medium clay 0.00195–0.00098 9–10
Different sedimentary environments (e.g. glacial, marine,
aeolian) give rise to particular textural features on the
beam by particles passing through it, and back-calculation surfaces of quartz and sand grains, and these can be analysed
of the sizes of these particles from the properties of the using an electron microscope. Characteristic textural
diffraction pattern produced (Hoey, 2004). Comparison of features on quartz grains that can be detected by scan-
the results obtained using these different methods suggests ning electron microscopy (SEM) include fracture patterns,
that measurement of particle settling velocities by sedigraph scratches, grooves, chatter marks, solution pits and cleavage
is the most reliable for fine-grained deposits, as other flakes (Fuller & Murray, 2002). Moreover, it is often
approaches over-estimate the significance of platy-shaped possible to identify superimposed features and hence more
particles (McCave et al., 2006). An alternative approach to than one palaeoenvironment of modification can some-
the derivation of grain-size data is to make measurements times be inferred. A recent innovation is the use of atomic
from digital images, which can be scanned and analysed force microscopy (AFM), a technique that enables 3-D
using image analysis software (Sime & Ferguson, 2003). representations of microscopic surface textures to be
Particle-size data are usually presented in the form of determined, and these can be quantified to derive empirical
sigmoidal curves on probability graph paper (section measures and indices of surface texture that corroborate the
3.3.5.1; Figure 3.13) or in ternary diagrams (triangular qualitative data derived from SEM analysis (Konopinski
graphs: section 3.3.5.1; Figure 3.12) or histograms. et al., 2012).
3.2.3.2 Particle shape 3.2.3.4 Organic carbon content

Particle shape has been used to distinguish between The determination of organic carbon content is of con-
sediments that have accumulated in different depositional siderable importance in palaeolimnology where it provides
environments, and has been employed particularly effec- an index of biological productivity in former lake basins.
tively in the analysis of glacial, glaciofluvial and fluvial It is also useful for establishing the amount of organic
sediments. Techniques include the purely visual assessment material that is likely to be required for the radiocarbon
of particle shape, the use of prepared charts as a basis for dating of a sample (section 5.3.2). The most widely used
method is loss on ignition, in which the amount of carbon 3.2.3.7 Clay mineralogy
in a sample is indicated by the weight loss following com-
bustion in a furnace, and which can be used for estimating The clay mineralogy of a sediment can provide information
both organic carbon and carbonate in sediments (Heiri on both the origins of the material and on any chemical
et al., 2001). More accurate results can usually be obtained changes that have resulted from the effects of different
using standard titration methods or colorimetry techniques weathering processes since deposition. Examples of the
(Meyers & Teranes, 2001). use of clay mineral analysis in Quaternary studies include
the provenancing of glacigenic sediments (Hillebrand
et al., 2009), and the verification of trimlines (section 2.3.1)
3.2.3.5 Metallic elements in mountain regions where differences in lengths of time
The variations in proportions of metallic ions, such as over which surfaces above and below the former ice limit
calcium, potassium, sodium or magnesium, in Late Qua- have been exposed are reflected in contrasts in clay fraction
ternary lake sediments are now regarded as important mineralogy (McCarroll & Ballantyne, 2000). The most
indicators of the changing erosional history of lake widely applied method in clay mineral analysis is X-ray
catchments (see section 3.9.3.1). The concentration of such diffraction (XRD), which involves the rotation of a sample
elements in a sediment sample can be determined using in a stream of directed electrons. The clay minerals present
either a flame photometer or an atomic absorption spectro- (e.g. illite, chlorite, montmorillonite) are identified by
photometer (AAS). The former operates on the principle observing and comparing the spacing and intensity of
that a metallic salt drawn into a non-luminous flame peaks on diffractometer traces: this approach enables clay
ionizes and emits light of a characteristic wavelength, mineral ratios and inferred environmental changes to be
while the AAS measures the concentration of an element reconstructed at high stratigraphic and temporal resolution
by its capacity to absorb light of its characteristic resonance (e.g. Fagel & Boës, 2008).
while in an atomic state. In both cases, the light emissions
are recorded photoelectrically. More rapid methods, which 3.2.3.8 Mineral magnetic analysis
measure the proportions of metallic ions present in minute Because sediments vary enormously in the quantity, size
quantities of sediment, use mass spectrometers to analyse and type of magnetic minerals that they contain, they
atomic emissions with a high degree of precision. The can often be readily characterized on the basis of their
inductively coupled plasma atomic emission spectrometry magnetic properties. A range of magnetic signals can be
(ICP-AES) method, for example, involves the creation of measured rapidly, both in the field and in the laboratory,
an aerosol (or plasma) of ionized gas from a homogenized and these enable the concentration and types of magnetic
sediment sample, which is then analysed for elemental minerals present in sediments to be determined (Walden
content using mass (ICP-MS), atomic emission (ICP-AES) et al., 1999; Evans & Heller, 2003). Variations in the mag-
or atomic fluorescence (ICP-AFS) spectrometry. Applica- netic properties of sediments can then be used to establish
tions of ICP-MS in archaeology are described by Pollard & changes in rates of sedimentation and to correlate sedi-
Heron (1996) and Ciliberto & Spoto (2000). ment units (Maher & Hallam, 2005). In lake sediment
records, environmental changes around the catchments
3.2.3.6 Heavy minerals can be inferred on the basis of mineral magnetic properties
(e.g. Oldfield et al., 2003). The application of magnetic
Heavy-mineral assemblages often reflect the derivation or measurements in the analysis of sediment sequences is
provenance of Quaternary deposits. They have been used discussed in more detail in section 5.5.1. Wider applications
in a number of Quaternary studies including the investi- of mineral magnetic analysis in the field of Quaternary
gation of weathering profiles, the differentiation of tills climatic and environmental reconstruction can be found
(section 3.3.6.3), and the analysis of aeolian and marine in Maher & Thompson (1999).
deposits (e.g. Eitel et al., 2001; Schüttenhelm & Laban,
2005). Heavy minerals are those with a specific gravity
(SG) greater than 2.85, and are usually separated from the 3.2.3.9 Stable isotope analysis
lighter mineral fraction in a sample by settling in a heavy Several common elements (e.g. oxygen, carbon and hydro-
liquid such as bromoform (SG 2.89). The heavy mineral gen) exist in nature in different isotopic states, reflecting
assemblage is then dried and mounted on a slide, and the variations in the number of neutrons in the nuclei (see
percentage of individual types can be determined using a section 5.3). This results in slight, but very important,
petrological microscope. differences in their physico-chemical behaviour, and a
selective separation between molecules of different atomic most secure palaeoenvironmental inferences are likely to
mass (see e.g. sections 3.8.4.2 and 3.10.2) often occurs as a be drawn from situations where geomorphological and
result of crystallization, evaporation, precipitation, osmosis, sedimentological evidence are employed together. Good
metabolism, etc. Changes in the ratios of isotopes of accounts of Quaternary glacial sediments can be found in
different atomic mass can subsequently become locked into Evans & Benn (2004), Bennett & Glasser (2009) and Benn
fossils or precipitates, and hence into the sedimentary & Evans (2010).
record. For example, where lake water levels fall during
drought conditions, the ‘environmental stress’ that results
may be reflected in a shift in isotopic ratios, and this will
3.3.2 The nature of glacial sediments
be registered in the tissues and skeletons of biota inhabiting
3.3.2.1 Unstratified and stratified sediments
the lake as well as in chemical precipitates. The isotopic
composition of meteoric water may also change due to The nomenclature of glacial deposits can be very confus-
climatic variations, and these may also be reflected in ing reflecting, at least in part, the bewildering variety and
isotopic ratios in lake sediments, ice sheets and cave complexity of sediment sequences in glacial environ-
speleothem. Variations in isotope ratios in sediments are ments (Benn & Evans, 2010). The superficial sediments
measured using mass spectrometers, and the results not that blanket the landscape in many parts of Europe were
only provide insights into the environmental conditions originally believed to have been derived from a great
under which the sediments accumulated, but also constitute flood and were termed diluvium, although Sir Charles
a proxy record of regional and, in some cases, global Lyell referred to them as drift because, along with many of
climatic change (Leng, 2004). Stable isotope data from lake his contemporaries in the early years of the nineteenth
sediments, speleothems, deep-sea sediments and ice cores century, he believed that the deposits had been derived
are described in more detail later in this chapter. primarily from the melting of icebergs that had drifted in
during a marine inundation. Curiously, the latter term
has survived in the literature to the present day and is
3.3 GLACIAL SEDIMENTS still used to refer to deposits formed by, or in association
with, glacier ice or by ice-melt. Glacial ‘drift’ has long
3.3.1 Introduction been divided into stratified and unstratified sediment,
Glacial sediments of Quaternary age cover large areas of and although more recent work has tended to blur the dis-
the earth’s surface, particularly in the mid-latitude regions. tinction between these two categories (section 3.3.3), it
In Europe, for example, glacially derived diamicton2 remains a useful first step in classification of glacial deposits.
forms an intermittent blanket over at least one-third of the Unstratified drift is usually referred to as till or (inappro-
land area, while in North America, such deposits are spread priately) ‘boulder clay’. The former term was first used
over half the continent. It has already been shown (section by Archibald Geikie in 1863 to describe coarse stony soil
2.3.3) that the moulding of these deposits into characteristic commonly found on the glacigenic deposits of northern
glacial landforms can be used to establish former glacier Britain. It is preferable to, and now more widely used
extent and direction of ice movement, and such data form than, the term ‘boulder clay’ since many non-glacigenic dia-
the basis for glacier modelling and climatic reconstruction. mictons contain boulders and clay whereas till frequently
However, equally important palaeoenvironmental evidence does not. The term ‘till’ is also more satisfactory than
can be derived from an analysis of the deposits themselves, ‘ground moraine’, which has often been employed to
for the distinctive properties of many glacial sediments describe glacigenic deposits. ‘Moraine’ is a geomorpho-
allow inferences to be made about former glacier types, the logical term and hence should not be used for lithological
mode of sediment deposition, ice-flow directions and classification. Diamictons that have been transported and
the sources of sediment supply. An understanding of these deposited by or from glacier ice with little or no sorting
properties can also have economic benefits as, for example, by water should be termed till (Menzies et al., 2006).
in tracing the sources of ore bodies or modelling ground- Stratified drift is characterized by sorting of material
water flow in glacigenic deposits. Indeed, in view of the by the action of glacial meltwater. These deposits show
widespread nature of glacial deposits by contrast with many affinities with fluvial sediments and are therefore
well-defined landforms, such as moraines and drumlins, referred to as glaciofluvial (or glacifluvial). The term covers
it could be argued that lithological evidence has a more a range of sedimentary environments, including ice-contact
important role to play in the reconstruction of Quaternary deposits (sediments formed adjacent to, or in contact with,
environments although, as in many other instances, the glacier ice), proglacial or outwash deposits (sediments
accumulating close to the frontal margin of an ice sheet or the ice is confined within a narrow valley, or spreads out
terminus of a glacier) and glaciolacustrine deposits into a piedmont lobe), the concentration and type of clasts
(sediments accumulating in lakes or located on, within, or contained within the ice, and the zone or part of the ice in
adjacent to, glacier ice, or fed directly by glacial meltwaters). which the sediments were initially entrained (Figure 3.4).
The sediments found in these various contexts range from Sediments may accumulate in a subglacial (at the base of
coarse gravels and sands formed in braided streams on land the ice), englacial (within the ice body) or supraglacial (on
to laminated clays and silts formed in lakes. Where the the ice surface) position. The supraglacial environment is
latter are seasonally frozen, varves (section 5.4.2) may especially complex with a mixture of glacially derived and
develop. In the marine environment, distinctive, mostly subaerially derived sediments often forming debris flows
stratified, glacimarine deposits accumulate around the (section 3.3.3). Short-term oscillations of the ice margin
margins of grounded tide-water glaciers or floating ice may lead to the superimposition of sediments from each
shelves, while ice-rafted debris may be deposited by icebergs of these contexts, as well as the incorporation of materials
and can be carried some considerable distance from the ice (including glaciofluvial sediments) deposited either prior
margins (Ó Cofaigh, 2007). to glacial advance or during ice wastage. As a consequence,
glacigenic sequences frequently reveal intercalations of
tills, flow materials and stratified sediments with complex
3.3.2.2 Glacigenic facies
vertical and spatial facies variations (Evans & Benn, 2004).
The nature and composition of glacigenic sediment depend The term ‘facies’ (from the Latin word meaning ‘external
on a number of factors, including the thickness of the ice, form’ or ‘appearance’) refers to a body of sediment that is
the rate of melting, the topographic context (e.g. whether characterized by a combination of lithological, physical or
supraglacial subenvironment
(l-A)
supraglacial
deltaic facies supraglacial
supraglacial (l-A-4) fluvial f a c i e s supraglacial
crevasse facies supraglacial
supraglacial (l-A-3) melting-ice
(l-A-2) lacustrine
fluvial f a c i e s facies (l-A-2)
facies (l-A-5)
(l-A-3)
supraglacial
melting-ice
facies (l-A-1)
terminoglacial
englacial .
terrestrial englacial
melting-ice m e I (water-tunnel
facies (ll-A-3) crevasse
facies (l-B-1) facies (l-B-3)
facies (l-B-2)
englacial subenvironment
(l-B)
terminoglacial meltwater-tunnel subglacial meltwater-tunnel

subglacial
subglacial facies (l-C-1) melting-ice facies (l-C-1)
subenvironment lacustrine
lacustrine facies (l-C-3)
(I l-A) facies (l-C-2)
facies (l-C-2)
subglacial subenvironment
(l-C)
Figure 3.4 Schematic model of supra-, en- and subglacial ‘subenvironments’ within the marginal zone of a continental ice mass.
Characteristic sediment facies are associated with each ‘subenvironment’, and classified according to the dominant mode(s) of
sedimentation, with the following main facies types recognized: in the supraglacial subenvironment (I-A), melting-ice (I-A-1), crevasse
(I-A-2), fluvial (I-A-3), deltaic (I-A-4) and lacustrine (I-A-5) facies; in the englacial subenvironment (I-B), melting-ice (I-B-1), crevasse
(I-B-2) and meltwater-tunnel (I-B-3) facies; in the subglacial subenvironment (I-C), meltwater-tunnel (I-C-1), lacustrine (I-C-2) and
melting-ice (I-C-3) facies (after Brodzikowski & Van Loon, 1987 reprinted with permission from Elsevier).
a) continental ice-marginal environment(ll)
extraglacial ice-marginal fan f a c i e s

aeolian facies stream way (ll-B-1)
(ll-C-1)
outwash
plain
mass flows
loess coversands drift s a n d s

(ll-C-1-c) (ll-C-1-b) (ll-C-1-a) near-termmal terminoglacial lacustrine
fluvial d e p o s i t s ( l l - B - 1 - c ) facies (ll-A-1)
extraglacial near-terminus terminoglacial

b) subenvironment (ll-C) s u b e n v i r o n m e n t (ll-A)
subenvironment (ll-B)
terminoglacial
terrestrial f a c i e s
proglacial near-terminus near-terminus near-terminus
(ll-A-3)
fluvial d e p o s i t s lacustrine facies deltaic facies fan facies
(ll-B-1-c) (ll-B-3) (ll-B-2) (ll-B-1)
t e r m i n o g l a c i a l fluvial f a c i e s
(ll-A-2)
z o n e of i c e - m a r g i n a l
streamways
m a r i n e i c e - m a r g i n a l e n v i o n m e n t (II)
floating dead-ice
b) ice shelf
blocks (icebergs)
undercurrents
extraglacial
full-marine facies (ll-E-2-d)
(ll-F-2)
g l a c i o m a r i n e turbidities
(ll-F-2-a)
glaciomarine
tunnel-mouth
near-terminus near-terminus
deposits (ll-D-2-b)
marine deltaic marine melting ice
extraglacial marine
facies(ll-E-l) facies (ll-E-2)
dropstones (ll-F-1-a) terminoglacial
marine mass-flow
deposits (ll-D-1-b)
extraglacial marine near-terminus marine terminoglacial marine

s u b e n v i r o n m e n t (ll-F) subenvironment (ll-E) s u b e n v i r o n m e n t (ll-D)
Figure 3.5 Schematic model of modes of deposition in proximal zones associated with continental a) b) and marine c) ice margins.
Characteristic facies in the continental proximal ice-marginal zone include: in the terminoglacial subenvironment (II-A), lacustrine
(II-A-1), fluvial (II-A-2) and terrestrial (II-A-3) facies; in the near-terminus subenvironment (II-B), fan (II-B-1), deltaic (II-B-2) and lacustrine
(II-B-3) facies; and in the extraglacial subenvironment (II-C), aeolian facies (II-C). In the marine ice-marginal zone, a range of facies
typifies each of the terminological (II-D), near-terminus (II-E) and extraglacial (II-F) marine subenvironments (after Brodzikowski &
Van Loon, 1987, reprinted with permission from Elsevier).
biological properties, and which can be distinguished from till were recognized: lodgement till, melt-out till and flow
adjacent bodies of sediment by a well-defined geometry and till. Lodgement till (Figure 3.6a) is deposited at the ice base
structure. Facies, or combinations of facies, reflect certain and accumulates on the subglacial floor either through
types of sedimentary process or a particular sedimentary pressure against bedrock protuberances or against patches
environment. Facies that are defined solely on the basis of of stagnant ice underneath the moving glacier body. Melt-
their physical properties are termed lithofacies; by contrast, out till (Figure 3.6b) consists of englacial debris released
genetic facies are defined using descriptive terms that imply from melting ice either above the glacier sole or at the
a mode of formation, e.g. ‘varved deposit’ or ‘debris flow’. glacier surface. In the former situation, it is confined
The analysis of glacigenic facies in contemporary ice- beneath the overlying ice body and the glacier bed, while
marginal environments reveals diagnostic features that can at the surface it is trapped beneath the overburden of
be employed in the interpretation of older Quaternary surface debris. Hence, melt-out till can be either subglacial
sequences (the ‘analogue’ approach). A number of classifi- or supraglacial in origin (Boulton, 1980), although the
catory schemes of glacigenic sediment sequences have now latter may be indistinguishable from the overlying surface
been developed, but all are simplifications of the complex debris. Indeed, there have been few studies of tills in
range of facies associations that form within and adjacent modern environments where till genesis by melt-out can
to glacial ice (Benn & Evans, 2010). Figure 3.5 illustrates be clearly demonstrated (Benn & Evans, 2010). In dry arid
the range of ‘subenvironments’ encountered in ice-marginal regions, such as parts of Antarctica, ice may vaporize
areas, and the lithofacies sequences that characterize them. directly without passing through the liquid phase, a process
For each subenvironment, a number of distinctive facies that is a form of freeze-drying. This is known as sublimation
types and associations can be identified, depending on the and the material that is released from the ice is sublimation
local style of ice melting, as well as on whether the sediments till (Marchant et al., 2002). Material that has been termed
accumulate against, above or beneath active or inactive ice. flow till (Figure 3.6c) is thought to be released as a water-
In the continental ice-marginal environment (Figure saturated fluid mass from the downwasting glacier surface,
3.5a,b), for example, the type of sediment deposited is and to be augmented by material from subaerial sources,
determined by distance from the ice margin and position such as avalanches and rockfalls. This debris tends to be
of the material within the ice mass. Both of these factors highly unstable resulting in flowage, a process that is
influence hydrological conditions as well as the physical and especially common near the snouts of receding or stag-
mechanical properties of the accumulating sediment which, nating glaciers (Menzies & Zaniewski, 2003). However,
in turn, will determine the rate of sedimentation and the because processes and products of debris flow can be
extent of sediment deformation3 (section 3.3.3.2). A identical in both glacial and non-glacial environments,
characteristic range of facies can also be found in marine many glacial sedimentologists now prefer to use the terms
ice-marginal zones (Figure 3.5c). ‘debris flow deposit’ or ‘mass flow diamicton’ (Figure 3.6c)
in preference to ‘flow till’, although the prefix ‘glacigenic’
may be added where direct associations with glacier ice can
3.3.3 The classification of tills be demonstrated (Benn & Evans, 2010). Flowage of
successive generations of material leads to interbedding of
3.3.3.1 Lodgement, melt-out and ‘flow’ tills
flow deposits, not only with outwash and lacustrine
Prior to the 1970s, till classification and genesis were sediments deposited on the ice surface, but also with
based mainly on the analysis of glacigenic sequences that spreads of outwash beyond the glacier terminus. As a
accumulated during the last, or earlier, glacial stages in consequence, complex stratigraphic sequences of till, debris
Europe and North America. It was not until the later 1960s flow, outwash and lacustrine sediments can result from a
and the early 1970s that detailed observations of the active single phase of ice wastage.
processes operating today in ice-marginal zones began to
appear regularly in the literature. A number of important
3.3.3.2 Deformation tills
new approaches to the study of glacigenic sequences
emerged in publications from this period. One example was In recent years, a considerable amount of attention has
the classification of tills and of the factors influencing their been focused on the subglacial environment, where the
formation introduced by Geoffrey Boulton (e.g. Boulton & recognition of the importance of coupling between the
Eyles, 1979). glacier and its bed has resulted in a paradigm shift in
On the basis of observations of deposits associated with glacial geological studies. This emphasizes the importance
contemporary glaciers in Spitsbergen, three main types of of subglacial deformation as opposed simply to subglacial
a) b)
c)
Figure 3.6 Examples of glacigenic deposits that could be

classified as: a) lodgement till (subglacial traction till sensu
Evans et al., 2006) from Isle of Skye, Scotland (reprinted with
permission from Elsevier); b) melt-out till, showing water-lain
sediments interbedded with diamicton, from Alberta, Canada;
and c) possible ‘flow till’, showing partial alignment of large
clasts suggesting mass flow, and hence now more typically
classified as a mass flow deposit (photographs by Dave Evans,
Durham University, UK).
deposition and sliding (Evans et al., 2006). The principal result of sliding or slumping under saturated conditions,
product of this process is what has become known as while it may be difficult to distinguish between intra-
deformation till (Figure 3.7), a rock or sediment that has formational deformation structures (those generated
been disaggregated completely or largely homogenized by during the initial settling of the till) and those resulting from
shearing in the deforming layer at the base of the ice (Benn post-depositional deformation, by the readvance of ice
& Evans, 2010). It now appears that many subglacial tills over older, unconsolidated glacigenic sediments (Phillips
are not, as previously thought, built up gradually in layers et al., 2008a).
by the release of material from the base of the ice An important innovation in the analysis of glaci-
(sometimes referred to as a ‘plastering-on’ process), but genic deposits has been the use of micromorphology,
rather that a body of soft sediment, transported en masse which has revealed a remarkable array of microstructures
at the base of the ice, is continuously subjected to shear within subglacially deformed deposits (Menzies et al.,
stresses and remoulding by deformation. The concept of a 2006). These include various forms of foliation, rotational
‘deforming bed’ is based on observations of modern glaciers structures, shear lines and zones, water escape structures
which indicate that most of the movement of glaciers is and a range of microfabrics (Figures 3.8, 3.9). Scanning
achieved by continuous deformation of the subglacial electron micrographs also reveal these microstructures
debris (Boulton et al., 2001). The confining, compressional (Figure 3.10). The characteristic micromorphological fea-
stresses generate a range of deformational features, tures that can be observed in deformed tills are partly
including, for example, isoclinal recumbent folds, shears, dependent on grain size. At the microscopic level, sediment
boudins (Figure 3.7), hook folds, diapirs, tension fractures, dominated by particles < 20 μm in size is termed plasma
sediment wedges, brecciated beds and conjugate fracture (formerly termed ‘groundmass’ or sediment ‘matrix’),
patterns (Roberts & Hart, 2005). However, deformational while larger individual mineral or organic particles are
structures, mainly folds, can also form in flow tills, as a termed skeleton grains. During deformation, micro-
structures and fabrics that form in plasma can be dis-

tinguished from those that form where plasma and skeletal
components are combined (S-matrix structures). These can
be divided into those characteristic of ductile or brittle
deformation, and those associated with high pore-water
content. Analysis of the type and distribution of these
features can reveal the processes of till formation (Menzies
et al., 2006). One important advantage of the thin-section
micromorphological approach is that it can be employed
to analyse small samples obtained from boreholes drilled
into unexposed sediments, including submarine strata, to
discriminate between subglacial till and other sediments
(Carr et al., 2006).
Macro- and micro-deformational features are com-
monly encountered in subglacial tills, so much so that
there is growing support for recognizing deformational
stress as probably the most important mechanism affecting
till formation during the Pleistocene (van der Meer et al.,
2003). While this view has been contested (e.g. Piotrowski
et al., 2004), some geologists argue that since all basal or
lodgement tills show evidence of subglacial deformation,
they should be described in tectonic rather than sedi-
mentary terms (e.g. Menzies et al., 2006; Evans et al., 2006).
Hence, many researchers now use the term ‘glacitectonite’
or ‘tectomicts’ to refer to subglacially deformed material
in general, while some maintain a distinction between
glacitectonite, materials that have undergone subglacial
shear but retain some of the structural characteristics of the
parent material, and subglacial traction till, sediment
deposited by basal melting and subsequently disaggregated
and perhaps homogenized by shearing (Evans et al., 2006).
Some have gone further, with Menzies et al. (2006) sug-
gesting that all subglacial tills are effectively deforming
glacier beds and hence the terms ‘lodgement till’ and ‘melt-
out till’ are effectively redundant. Future work on the
subglacial environment and its deposits will show whether
or not this is indeed the case.
The presence of a deforming or mobile bed at the base
of an ice sheet has important palaeoenvironmental implica-
Figure 3.7 Highly contorted glacial diamicton (till) exposed at
tions. It has generally been considered, for example, that
West Runton, Norfolk, UK (from Roberts & Hart, 2005, reprinted climate change has been the primary driver of glacier
with permission from Elsevier). a) Folded layers of till and chalk behaviour, with a direct relationship between temperature
wrapped around large ‘pods’ of rafted chalk. b) A layer of chalk and precipitation on the one hand, and glacier advance and
that has been compressed and strained under high subglacial retreat on the other. However, if the majority of glacier
pressures, leading to formation of boudins, which are bead- or
lozenge-shaped moulded remnants of the original layer; boudins movement takes place in the basal sediment, then changes
are common features along faults and compressional stresses, in the deforming layer, such as drainage of the sediment and
where a competent bed (in this case a chalk layer) is moulded changes in sediment texture, may have a more significant
and stretched by lateral stress, while less competent adjacent impact on glacier behaviour than climate (Hart & Rose,
material (in this case the grey till) migrates into any resulting
voids. c) Recumbent fold structures displayed by light-grey silt
2001). Similarly, deforming bed processes may help to
layers within deformed subglacial till (photographs by Dave explain a continuing problem in glaciology, namely dis-
Roberts, Durham University, UK). continuous fast ice flow in surging valley glaciers. Indeed,
Figure 3.8 Microfabrics and microstructures within the

Range of microstructures and Plasma and S-matrix of glacial sediments
fabrics observed in deformed tills
from micromorphological analysis
of thin section samples (from Plasma Skeleton grains
Menzies et al., 2006, reprinted with
permission from Elsevier).
Plasmic microfabric S-matrix
Ductile Brittle Polyphase Porewater

(Ductile/Brittle) Influenced
SKELSEPIC or Induced
PLASMIC FABRIC
VARIETIES OF STRAIN CAPS f-AULTr-n [KlMAINJj UUI.TIPIE OIAMICTON CUTANS'

MA5EPIC PLASMIC A SHADOWS DOMAIN 5 . AHÎi. i ANS.
-AWHIL:
IAI ritinvn: FOLD STRUCTURE DISCRETE "COMST- WATER ESCAPE

PLASMIC FABRIC SMEAR LINES STRUCTURE STRUCTURES
miNlSFPlC LAYeRlHq .1 SHEAR ZONES SILL 4 D K E SILT C A P S

?LASMlC FA&BlC FOLIATION STRUCTURES
UNISTRrAL M-::KIN:.- REVER-St FAULT T I L E D UNITS "POLYGONAL "

PLASMIC FABRIC 5TRUCTURES O F LAMINATED STRUCTURES
C L A Y S * SILT
IN3EPIC "ROTATIONAL" KINK BANDS SILT i C L A Y

PLASMIC FABRIC STRUCTURE COATINGS
"BANDED" PLASMA SECONDARY CRUSHED GRAINS

I 01 IATION
KINKING CRENULATlON
PLASMIC FABRIC FOLIATION
deformation of subglacial till is now thought to be one of concentrated under high pressure at the base of modern ice
the primary mechanisms maintaining the high rates of ice streams, saturates basal sediment reducing its friction and
movement associated with ice streams in Antarctica, and increasing its plasticity and mobility; it is therefore difficult
may also be responsible for the generation of glacier surges to distinguish between the mechanical effects of sediment
(Dowdeswell et al., 2004; Winsborrow et al., 2010). Episodes deformation and lubrication as agents of basal sliding
of fast flow over a mobile glacier bed may also explain the (Marshall, 2005).
massive discharges of the Laurentide ice sheet into the
North Atlantic ocean (Heinrich events: section 3.10.1).
3.3.3.3 Paraglacial deposits
While simulation modelling of palaeo-ice sheets (e.g.
Breemer et al., 2002) and modern observational studies The retreat or withdrawal of glacier ice exposes land-
of Antarctic fast-flow (e.g. Smith et al., 2007) appear to scapes that are inherently unstable and which are liable
support these ideas, there is still uncertainty as to whether to modification, erosion and sediment release at rates
soft-bed deformation is more important than basal sliding greatly exceeding background denudation rates. Such accel-
in the promotion of fast-flowing ice. Water, which is erated geomorphological activity has been described as
Figure 3.9
Examples of microfabrics in deformed
till identified using thin-section analysis
on samples obtained from tills in
Ontario, Canada (from Menzies et al.,
2006, reprinted with permission from
Elsevier). The images on the right are
annotated versions of those
on the left, with characteristic
microstructures outlined and labelled
(images provided by John Menzies,
Brock University, Ontario, Canada).
paraglacial. The term was introduced by Ryder (1971) example, slope failure, debris flow and fluvial reworking
but subsequently defined by Church & Ryder (1972, of materials. All of these are characterized by high rates of
p. 3059) to refer to ‘nonglacial processes that are directly sediment delivery from slopes into fluvial and aeolian
conditioned by glaciation’ and both to ‘proglacial processes, systems, and while they are not unique to the paraglacial
and to those occurring around and within the margins of landscape, they are especially effective during the
a former glacier that are the direct result of the former adjustment of newly deglaciated landscapes to non-glacial
presence of ice’. The essence of the concept is that of conditions (Figure 3.11). The temporal operation of these
glacially conditioned sediment availability (Ballantyne, processes (the paraglacial period) will vary because different
2002a). Recently deglaciated terrain is often initially in elements of the paraglacial landscape will relax at different
an unstable or metastable state, and is therefore vulner- rates: steep, debris-mantled hillslopes may achieve stability
able to modification by subaerial processes including, for within a few centuries of ice retreat, whereas large fluvial
systems may still be reworking glacigenic sediment more

than 10 ka after deglaciation (Ballantyne, 2002a). It is also
possible for external factors (e.g. climate change and sea-
level change) to induce sediment release millennia after
termination of the initial (deglacial) period of paraglacial
activity (Ballantyne, 2002b). It is important, however, to
note the key distinction between paraglacial and periglacial
processes, with the latter being characteristic of all cold,
non-glacial environments, irrespective of whether glacier
ice is or was present in the catchment (Knight & Harrison,
2009).
In the last two decades increasing attention has focused
on paraglacial geomorphology and, in particular, on
paraglacial facies represented in Quaternary stratigraphic
sequences. In many parts of the British Isles, for example,
landscapes that were previously interpreted in terms of
more conventional glacial or periglacial models have been
re-evaluated in the context of the paraglacial concept. In
coastal areas of Wales, exposures of diamictons hitherto
considered to be of glacial or periglacial derivation are now
considered to reflect widespread redistribution of glacial
sediments during and after deglaciation (e.g. McCarroll &
Rijsdijk, 2003). Remobilization of valley-side tills and other
glacigenic deposits by debris flow and erosional processes
is increasingly being recognized and reflected in the
mapping of sediment–landform associations in other
regions as well, for example in studies of recently deglaciated
terrain in Patagonia (Glasser et al., 2009), the Himalayas
(Iturrizaga, 2008) and the Alps (Curry et al., 2006). The
identification of paraglacial facies within Quaternary
sequences is not always straightforward, however, and in
some cases it may be impossible to distinguish paraglacial
diamictons from, for example, flow or deformation tills.
Paraglacial sediments therefore constitute a source of
continuing ambiguity in the Quaternary stratigraphic
record (Ballantyne, 2002a).
3.3.4 The influence of the thermal regime

of glacier ice
A fundamental factor in determining the type of till that
Figure 3.10 Scanning electron photomicrographs of vertical will be deposited at any one locality is the position at
sections through the fine matrix of tills. a) Slate-rich till, showing
anisotropy with elongated clasts dipping towards the right, which debris is transported within the ice and this, in turn,
indicating ice movement direction towards the left. b) Shears is governed largely by the thermal regime of the glacier
in silty, deformation till indicating movement of ice towards the (Benn & Evans, 2010). The thermal regime is determined
right (SEM Images provided by Lewis Owen, University of by ice thickness, mass balance and, above all, climate, and
Cincinnati, USA). can be used to define four boundary conditions at the
glacier sole. These are:
A. a zone of net basal melting where more heat is provided
to the glacier sole than can be conducted through the
glacier;
Figure 3.11 View down the Braldu Valley in the central Karakoram, northern Pakistan (photograph by Lewis Owen, University
of Cincinnati, USA). In the foreground is a large moraine ridge, while in the background is a series of massive fans, created by
paraglacial mass movement of glacigenic sediments that took place soon after the ice retreated from the valley slopes (see Seong
et al., 2009).
B. a zone in which a balance exists between melting and through shearing action. Subsequent deposition may
freezing where the heat provided at the glacier sole is therefore be in the form of mass flow deposits. In zone D,
approximately equal to the amount that can be the bed is frozen, no basal sliding occurs and glacier
conducted through the glacier per unit time; movement is entirely a result of internal shearing. Again,
C. a zone of net basal freezing but where sufficient mass flow diamictons may be the dominant depositional
meltwater may still be present to raise the temperature types. Zones A and B tend to be associated with ‘warm-
and maintain parts of the sole at the melting point; based’ or temperate glaciers, while zones C and D are
D. a zone in which the amount of heat provided at the sole found principally in ‘cold-based’ or polar glaciers. It is
is insufficient to prevent freezing throughout. important to appreciate that within a single ice sheet, some
or all of these zones may be present, and that during the
In zone A, the glacier slips over its bed and material course of an ice sheet cycle (ice growth, ice maximum,
entrained in the basal ice layers will subsequently be ice wastage), the various boundary conditions will vary in
deposited where frictional retardation against the bed both space and time (e.g. Boulton et al., 2001). Changing
is high. Similar processes operate in zone B, although dynamic and thermal regimes have also been noted in
lodgement (and deformation) of material will tend to be valley glaciers (‘polythermal glaciers’) leading to contrasting
greater with lower amounts of meltwater present. In zone styles of sedimentation (Glasser & Hambrey, 2001).
C, plucking of subglacial material occurs as the glacier The importance of this concept for the interpretation
slides over its bed, little lodgement/deformation till is of glacigenic sediments and sequences is that it explains why
deposited and material tends to be carried up into the ice in some areas subglacial tills tend to be the dominant
depositional type, while in others supraglacially derived of the nature of the contacts between them, any lateral
debris is more common. It also explains why, in regions variations both within and between beds, and any evidence
where cold-based conditions occurred in ice-marginal for superimposition of sediments as well as the degree of
zones, stacked multiple-depositional sequences including deformation or of other post-depositional modifications.
a range of interbedded glacigenic facies (lodgement/ The analysis of facies relationships and facies variations
deformation till, mass flow diamictons and glaciofluvial can provide greater insights into the nature of the former
sediments) can be the product of a single wasting ice depositional environment than the more traditional anal-
mass (McCarroll & Rijsdijk, 2003; Lukas et al., 2005). This ysis of individual sedimentary units (Thomas & Chiverrell,
is a significant departure from earlier thinking where, 2007). Some of these aspects of glacial sediment analysis are
for many years, multiple-till sequences were commonly now considered in more detail.
interpreted as reflecting two or more glacial advances.
Equally important has been the recognition that glacigenic
3.3.5.1 Particle size and shape analysis
sediments are spatially complex as a consequence of vari-
ations in basal thermal regime. Research on the beds of Individual till units often have characteristic grain-size
modern glaciers has shown that subglacial conditions vary variations and particle shapes. Particle size distributions are
between cold-based and warm-based ice over short dis- a function of a number of factors, including rock and
tances leading, in some areas, to a highly complex mosaic mineral types of which the clasts are composed, transport-
of glacigenic deposits (Evans et al., 2006). The local topo- ational processes, transport distance and mode of deposi-
graphic context will also influence the basal thermal regime, tion. In lodgement/deformation tills, for example, the
and thus the type and complexity of glacigenic sequences particles tend to be closely packed and little winnowing
that develop in ice-marginal zones. This is particularly the by drainage takes place, so that if the sediment has a high
case where proglacial and subglacial lakes develop, for not clay content, this will tend to be preserved. Mass flow
only does this result in interdigitation of outwash sediments diamictons, on the other hand, may lose much of the finer
with diamictons, but it can also induce instability in the ice matrix through rapid drainage which removes clay and silt
margin, leading to glacitectonic disturbance of the sedi- particles, a process known as illuviation. If the depositional
mentary facies (e.g. Knight, 2012). Complex sedimentary process is passive, however, much of the fine matrix can be
facies are also characteristic of ice-marginal zones that preserved, whereas active slides followed by free water
extend into the sea, reflecting local variations in basal escape may reduce the amount of fine particles during
thermal regime, especially where the ice front oscillates settling. Laboratory measurement of particle size distribu-
across the grounding line (Ó Cofaigh, 2013). tions, in the form of ternary plots (Figure 3.12) or
cumulative frequency curves (Figure 3.13) can sometimes
discriminate till units of different genesis, or isolate tills from
3.3.5 Analysis of glacigenic sequences other diamictons, and can even reveal subtle differences in
In view of the complexities of till genesis, the analysis of lithology between tills of similar origin (Curry et al., 2009).
glacigenic sequences requires the application of a range of An analysis of particle shape and surface textures, along with
field and laboratory techniques in order to distinguish signs of micro-wear, may also be important for establishing
between different types of glacigenic sediment and to the mode of formation of tills. There will be a tendency for
establish their field relationships (Hubbard & Glasser, particles to be more strongly affected by abrasion, faceting
2005). The aim should be to gain an understanding of and crushing in lodgement/deformation than in supra-
both the internal structure and constituents of each sedi- glacial tills, for example, and so this evidence, in combina-
mentary unit, as well as the three-dimensional geometry of tion with particle size data, may help to differentiate
the glacigenic sequence under investigation. A detailed between tills of different genesis (e.g. Evans et al., 2012).
description should be obtained of each lithological unit,
including changes in grain size, shapes of particles, and the
3.3.5.2 Lithofacies interpretations
presence or absence of bedding and deformational struc-
tures such as shears and folds (Chapter 6, Table 6.1). Most interpretations of glacigenic sequences are based on
Evidence of flowage of material, for example current bed- the synthesis of a range of lithological data, involving
ding or ripple structures in stratified sediments or clast detailed vertical logs of sediment exposures, and the analysis
fabrics (section 3.3.6.2) in unstratified deposits, should also of clast lithology, fabric (section 3.3.6.2) and grain-size
be recorded. It is particularly important to note the facies variations from as many beds as possible. These are plotted
relationships between lithological units, including details to provide an overview of the three-dimensional geometry
a) 1211109 8 7 6 5 4 3 2 1 0 -1 - 2 - 3 4 - 6 - 6
100
90
80
0.6 0.2 70
Percentage
SO
50
40
0.6 0.4 30
20
S FS 5F 5 10
C lC s tr e r
u sl ut e
SF5 0
0 6 b) 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 0 -1 - 2 - 3 - 4 - 5 - 6
0.4
100
F 4 .F6 90
0 2
80
SF6.
F5 EB3
SF6.
70
Percentage
00.6
2 BF5 SF1 \0.8
BF3 60
SF4 F3
BF4 50
ES E
F6.
Bf2 40
30
20
Girdle Cluster 10
0
F Filey B r i g g s fabric O k s t i n d b r e e n fluted m o r a i n e s
c) 1211 10 9 8 7 6 5 4 3 2 1 0 -1 -2 -3 - 4 - 5 - 6
BF B a r m s t o n fabrics U n d e f o r m e d l o d g e m e n t till
100
SF S k i p s e a fabrics D e f o r m e d l o d g e m e n t till 30
S e d i m e n t gravity flows 80
Percentage 70
00
Figure 3.12 Ternary diagram using a statistical clustering 50
method (eigenvalues) to characterize clast fabrics obtained 40
from different types of diamicton in eastern England (from 30
Evans et al., 1995). 20
10
0
d) 1211 10 9 8 7 6 5 4 3 2 1 0 -1 -2 -3 - 4 - 5 - 6
100
of lithofacies represented at a site, and this forms the basis ao
for the correlation of beds and reconstructions of their so
70
Percentage
genesis (e.g. Figure 3.14). The number of logs and associated 60

field measurements depends on the local stratigraphic 50
complexity. The more sedimentary characteristics that can 40
30
be established for each bed, the more confidently can the 20
mode of deposition be inferred. Examples of the sort of 10
0
characteristics that can be logged, and how the combined 1211 10 9 8 7 6 5 4 3 2 1 0 -1 -2 - 3 -4 -5 - S
8)
information may help to differentiate between particular 10C
types of till and between tills and other diamictons can be SO
80
found in Pawley et al. (2004) and Thomas & Chiverrell 70
Percentage
(2007). 80
50
Despite the meticulous and painstaking nature of this 40
type of work, the evidence often remains frustratingly 30
equivocal. This can be exemplified by the differences of 20
10
view that currently exist over the interpretation of the 0
0.002 0_06 2 mm
sedimentary record of ice wastage in the Irish Sea basin in
CLAY
HNh M b LI CHS FSANDMh M E D CRS "i\r.SAND
MED.SAND
western Britain. Some have interpreted the stratigraphic SILT SAND SAND
SAND
evidence as indicating deposition within a glaciomarine Particle size (diameter! BS/mm & 0
Figure 3.13 Particle size distributions (cumulative frequency curves) on the mm and φ (phi) scales showing the distinctive curves
obtained from diamictons and sorted sediments a) and b) and the way in which the technique may distinguish between different
diamictons c), d) and e). Note the more consistent frequency curves obtained from diamicton d) compared with those of c)
and e). (a and b from McCarroll & Harris, 1992; c, d and e from Evans et al., 1995).
m
RR•
m
27 515m
26
25 RR
59Sm
24
23
22 320 m North
105m
21
lb
20 228m
5
19 160m
16
R
17
'0
16 n
IS n
54 m 3 26 05 m
m 10
14
13 0
12
VA WR
446 m
11 7
10
m
9
8 6
7
a
s
s
3
2
1
0
A3CDEFGH
A - C l a y B-Silt C - F i n e S a n d O - Medium Sand E - Coarse Sand F - P e b b l e G - Cobble H - Boulder
Figure 3.14 Schematic facies logs of glacigenic sediments in Wales, including rose diagrams of preferred orientations of long
axes of clasts and clast roundness data (after McCarroll & Harris, 1992).
environment, while others have concluded that the coastal direction of ice movement across a formerly glaciated area.
deposits around the Irish Sea were derived largely from However, some characteristics of glacial sediments can
wasting land-based glaciers (McCarroll et al., 2001). There also yield valuable ice-directional information. The most
is no single set of sedimentary features that can conclusively widely used are indicator erratics, till fabrics and certain
resolve such differences of opinion; interpretation depends properties of the till matrix.
on an evaluation of all of the lithological (and geo-
morphological) evidence that is available, and the match-
3.3.6.1 Erratics
ing of this information with what are considered to be
best-analogue facies sequences from contemporary glacial The far-travelled particles found within, or on the surface
contexts. Much depends, therefore, upon having an of, a body of glacial sediment are known as erratics. The
adequate knowledge of the links between modern glacier, term is derived from the phrase terrain erratique, and was
glaciomarine and glaciofluvial processes, and of the used initially by the French geologist Horace-Bénédict de
lithofacies variations they produce. Our understanding of Saussure in the late eighteenth century to describe areas
Pleistocene glacigenic sequences is therefore an iterative and where material of foreign origin overlay local bedrock.
ongoing process, with constant refinement of the facies The shortened term is now generally applied to a particle
models upon which environmental reconstructions are of any size that is not indigenous to the area in which it is
based, and increasing recognition of the regional variation currently found. The most valuable types of erratic are those
in glacigenic facies associations (Figure 3.15; Hambrey & for which the sources are known, which are resistant to
Glasser, 2012) . erosion, and which have distinctive appearances, unique
mineral assemblages or unique fossil contents, thereby
allowing unequivocal identification. These are usually
3.3.6 Ice-directional indicators termed indicator erratics and range in size from finely
It has already been shown how certain landforms, such as comminuted fragments to large blocks weighing several
drumlins and roches moutonnées, can be used to infer the hundred tonnes. The famous Okotoks erratic (‘The Big
TEMPERATE - TERRESTRIAL-
A B C D E
Alpine Alpine Icelandic Alpine POLYTHERMAl.
(ice-proximal recession) (tee-distal recession) (ice-prowmal advance) (glaciolaeusirine) TERRESTRIAL
(.proglacial)
(.proglacial)
drjiian:
supra basal
glacial i i
t.M
aupra-
basal ij BCial
till debr.s
•••r :•• ••
chaPEic,
•jlacio.ecE-
J.1T Vf!.
•Jwnptd 'luviai lurbrdilea gravel lag
oni$«d
lOa-rYiargin 'luviai with drop-
stones,
•gravel
• lag ::T
pv-i
iuhAqurlOui
gravel lag gravel lag
•
gravel laglag
gravel
•ndicalrje
suogiaciai
icale
till
flEi&O-
riuvial
glaCiu+nic waalhered
1m 1m
gHCi&'
dabfl».flcw
ram-out bedrock
indicative
sjba:n; i>I
scale
gravellag
gravel lag from lo*
Ml rronllnd beorock
todrMk iCrtorgs
Glasser 4 Hambrey i?0C 1
Eyles A LazareH, 2007 Eyles £ Leaorok, 2007 Evans 2003 Eyles A Lezorok. 2O07
c 5 5 g c 3 sg C 3 s g C C3 C C C S C
clay
Sill
sand
I-V-'LVH
F G H I J
POLYTHERMAL COLD POLYTHERMAL POLYTHERMAL COLD GLACIOMARINE
TERRESTRIAL TERRESTRIAL GLAClOMARINE GLAOOMARINE Antarctic (>ce she. r to
Himalayan Dry valleys E lies me re Island Greenland (mid-fjord) continental shelf:)
(terminal moraine) i ce-coniact j^roni (grounding-tine fan sea 'loor
10 tfeltaic.em&fgerit) sea floor
ia ;u air i" e
s^dimente;
diaiofnactout
pobibly
(Opsel nanable
flftifdyditiom
iiiind i. r arid M l L"
pobibly rain-ouE
deltaic
diamicton
.MM .in Mine (erne roam |
(fom iceberg .raNing
drape* •cebergs and aeuhan
Eidal • r. i-n |j!S
t
and
cyclop*li Hop J si • •. a'
^hdotaicMiad •Ad cyr.lnp 11 n i drtlal glaD0rriar.n(.
glscioflurtal (bwt) grounding EOnfr
Band, basal plume aelihng of
ice and snow fines, rewarKmg
hetero- •<i.t!.i:;ii. by i ran i on c-urrents
•<i.t!.i:;ii
(e.j. tidal pumping)
geneous ••..:.v.r.h
ice-fronEal ran, ica-
deposition prommal pro* gittioffit/int;
1.0*1 i M Eurbtdily da •• L-M. wjlh drop.
<
• i if re n and
1
lidal rhythmiEes; J slooes near
•<i.t!.i:;ii t re;; s(rjn r -, grounding zone
•<i.t!.i:;ii
flows.
n-ii.r.i:,!::
old
flows. subgler>al
subglacial
IN HI
toeberg lain-cul
Hambrey A. Filzermona 7010 Slc*art Ifi91 Dowd rjs*eil al i?i , 1999 DomackereJ.. 1&9B
Slc*art
Slc*art Ifi91 Ifi91
C C C C Cc 3 C 9 c s s S C S • : C S. S C
Lithofacies
'1 3 3 4 5 5 7 8
Bandylmridy massive htfflroggiwui ilratrfwd dwm*don tKniHetfcotoble missive sandy massme lo weakly massi»e <o waaKly
grovelvrttti liNim :::nri «ndYgravel ¥*thoVotHlun* grtvdl (angular I bouldergravel straliliedtardyparal alrabfitd aftnd wuh
tadrotf relti urueluret with rumor tand wdh nceasionBI rara cro»-t4dd>ng
anssbecdinn or climbing ripples
9 10 11 12 13 14 15
wen stratified mud laminated day/silt, siliceous mud foliated and

anssbecdinn lam of bedrock
Wnd willl DWugh (**t&tliy| Sand in CoupwjC* with and D M I V : : „anssbecdinn m of m of
Wnd willl DWugh Wnd willlWnd willl DWugh
DWugh
Figure 3.15 Idealized lithofacies associations for glacigenic sediment sequences that developed under different thermal regimes
in terrestrial and marine settings. The depositional mechanisms inferred are also shown. The grain-size indicators (from left) are:
c – clay; s – silt; s – sand; g – gravel (from Hambrey & Glasser, 2012, reprinted with permission from Elsevier).
Figure 3.16 The giant quartzite erratic (‘The Big Rock’) near Okotoks, southwest Alberta, Canada (photograph by Dave Evans,
Durham University, UK).
Rock’: Figure 3.16) c.18 km southwest of Calgary in western B

Canada, for example, is the world’s largest known glacial
erratic, with a mass of c. 18,000 tonnes. It is part of the
‘Foothills Erratics Train’ that stretches more than 580 km
from the Athabasca Valley in Jasper National Park along B
the eastern edge of the Rocky Mountain Foothills to the B B F
B F
southern border of Montana, USA, and was carried into
A
place during coalescence between Cordilleran glaciers and E
2B D
the Laurentide ice sheet during the Late Wisconsinan A E
B
C B
c
glaciation (Jackson et al., 1997). At the other end of the F
scale, distinctive indicator minerals can also be used to infer B

E
F F
till origin or provenance (e.g. Klassen, 2001; Roy et al., 2007).
Erratic distributions can provide information on both local ©AMPS in M/wi/res™ 2009
and regional ice-flow directions (Figure 3.17), and they have E s s e x i t e outcrop
also been used to reconstruct patterns of ice-sheet
deglaciation, where changing flow pathways and migrations Glasgow Firth of
Forth
of ice divides can be inferred from erratic distributions (e.g.
Ehrmann et al., 2011). In this respect they are also an Edinburgh
important data source for the development of ice-sheet b)
models (section 2.3.4).

One problem with using erratics as ice-directional Figure 3.17 a) Distribution of some indicator erratics by ice
indicators, however, is that their presence in a glacial in Britain and northwest Europe: A – Galway granite; B –
Rannoch granite; C – Ailsa Craig riebeckite–eurite; D – Criffell
deposit may not always reflect primary derivation (i.e. an granite; E – Oslo rhomb porphyry; F – Dala porphyries. b) The
erratic could have been removed during a previous glacial Lennoxtown boulder train in the Forth Valley lowlands of
episode and reincorporated into a younger till) and this can central Scotland (based on diagrams in Sissons, 1967 and West,
lead to erroneous interpretations of former patterns of ice 1977).
movement. Hence, although erratics can frequently provide
useful ice-directional information, they are perhaps best glacial geological studies that stones (clasts) within a till
used in conjunction with other independent sources of often displayed a preferred orientation, although it was
evidence. somewhat later that a quantitative relationship was
established between stone orientations in till and patterns
of ice movement. Over the last few decades till fabric
3.3.6.2 Till fabrics
analysis, involving the measurement of orientation and
The arrangement of particles in a till is termed the till fabric. dip of particles within a till matrix, has become one of
It was observed at a very early stage in the development of the most widely used techniques for reconstructing former
ice-flow directions (e.g. Kjaer et al., 2003). More recently,

a) b)
fabric data have been used to infer depositional processes
and deformational mechanisms (Larsen & Piotrowski,
2003; Hart et al., 2009). The technique rests on the
assumption that, within the constantly deforming layer at
the base of the ice, stones (clasts) will become orientated
to adopt the line of minimal resistance to flow, that is, with
their long axes parallel to the flow direction. In theory,
therefore, subsequent deposition of the subglacial debris in
the form of lodgement/deformation till should preserve a
record of the former direction of ice movement, and this c) c)
can be established by measuring the orientation of pebbles
in lodgement/deformation till exposures using a compass.
Measurement of the dip by means of a clinometer or a
similar instrument may also provide useful ice-directional
information, as a tendency has been observed for pebbles
in a lodgement/deformation till to dip up-glacier.
Till fabric data are often presented in the form of
stereonets or polar graphs. Where two-dimensional data
only have been obtained, a rose diagram is constructed
showing the number or proportion of pebbles in different Ec) c)
azimuthal classes. However, because each measured stone
is represented by two opposite azimuthal values (e.g. 30°
and 210°), the rose diagram actually consists of two reflected
halves or mirror images. The data can be shown by a line
through the middle of each sector (Figure 3.18a), by the
linking of such lines to form the typical rose diagram
(Figure 3.18b) or by the shading of each azimuthal class
to the extent of the line marking the outer limit of each
class (Figure 3.18c). Where the dip of the pebbles has been
recorded, orientation measurements are taken in the down- Figure 3.18 Different methods for representing till fabric
dip direction, and therefore each pebble is represented by data. For explanation see text.
a single dip and orientation value. In this case, the diagram
will show a full 360° distribution (Figure 3.18d). Orienta-
tion and dip can be plotted together in the form of a differentiating between different types of sediment facies,
scattergram with the radius divided into degrees show- for facies with contrasting depositional and deformational
ing the angle of dip, and the circumference divided into histories may have distinctive ranges of eigenvalues. If this
degrees showing the orientation of the pebble (Figure is so, then it may be possible to employ the fabric charac-
3.18e). This is one of the most commonly used methods teristics (as reflected in eigenvalues) of sediments of known
for depicting till fabric data, and in some cases the visual origin to assist in the interpretation of facies whose mode
effect is enhanced by contouring the diagram as shown in of genesis is not known (Benn, 1994; Hart, 1994). Not all
Figure 3.18f. authorities agree on the validity of this approach, however,
While stereonets are useful in that they provide a and the use of fabric data in the discrimination of glacigenic
good visual impression of fabric data, a quantitative basis facies and in the genesis of tills continues to be debated
for comparison between fabric data is often required. (Benn & Ringrose, 2001; Carr & Rose, 2003).
Perhaps the most widely used approach has been the eigen- In using fabric analysis as a basis for determining
vector method. Eigenvalues reduce large datasets to simple regional ice-flow direction, it is necessary to be certain of
descriptive statistics. These reflect the strength and orienta- the genesis of the till under investigation. Mass flow tills,
tion of directional properties of a sediment, and thus allow for example, which were generated under saturated condi-
the ready comparison of fabric data from several localities tions and flowed down a steep ice surface before settling,
(Benn, 2004). This method has also been used as a basis for tend to show a very strong preferred orientation through-
out the sediment body. However, fabrics from such tills Scherer, 2005). In addition to providing insights into
invariably indicate only very localized conditions of flow- former patterns of ice movement, the analysis of indicator
age with neighbouring till units often exhibiting markedly minerals in tills also has applications in economic geology,
different flow directions. Other tills may show little pre- for ore bodies can be located using evidence of glacially
ferred orientation where clast concentration was low and transported debris (Lehtonen et al., 2005; Sarala &
melting from within the ice was slow; in these cases, the Peuraniemi, 2007).
englacial fabric may be retained (Johnson et al., 1995), A recent development in studies of till matrices has
but as this is easily lost during the final melting phase, these been the analysis of cosmogenic nuclides. As we shall see
too can be unreliable indicators of regional ice movement. in Chapter 5 (section 5.3.8), cosmogenic nuclides that
The key, therefore, is to employ only tills that derive from are formed at the earth’s surface through interactions
the subglacial environment. However, fabrics in lodgement/ between incoming cosmic rays and certain terrestrial
deformation tills can also display considerable spatial minerals can be used as a basis for dating. Here, however,
variability, even within a single exposure with the lack of variations in cosmogenic nuclide content of tills and
consistent pattern in relation to ice-flow direction perhaps exposed ground surfaces can be used to determine whether
reflecting stress field variability within the deforming glacier these surfaces were formerly covered by warm-based as
bed (Carr & Rose, 2003). Fabrics may also be affected by opposed to cold-based ice. In Baffin Island, for example,
the morphology of the glacier bed, especially where basal the concentrations of two cosmogenic nuclides (26Al and
10
ice is forced around prominent bedrock protuberances Be) were higher in those areas where independent studies
(Catto, 1990), and they may also show marked spatial suggested cold-based ice predominated (Staiger et al.,
variability where the tills were affected either by changes 2006). This may reflect the fact that surfaces that were not
in the direction of regional ice movement during subject to intense erosion (i.e. were covered by cold-based
deposition (e.g. Hicock et al., 1996), or by post-depositional ice) have retained higher cosmogenic nuclide levels that
modification of the sediment (e.g. Hambrey et al., 2001). were inherited from preglacial times. Greater effective
Till fabric analysis is, therefore, a technique that must be erosion takes place under warm-based ice, which removes
undertaken with care, and is perhaps best employed in former regolith, exposing younger surfaces with a much
association with other indicators of former ice movement. lower cosmogenic nuclide content. This contrast is also
reflected in the cosmogenic isotope content of tills associ-
ated with cold- or warm-based ice. A broadly similar find-
3.3.6.3 Properties of the till matrix
ing emerged from the analysis of 10Be content of regolith
Certain physical and chemical properties of the till matrix and tills in northern Sweden (Ebert et al., 2012). Given the
can be used to make inferences about regional ice move- importance of distinguishing between cold-based and
ment, and also to differentiate between tills from different warm-based ice for understanding the behaviour of former
source areas. This approach rests on the assumption that ice masses, this new technique appears to offer a potentially
tills will inherit certain textural and chemical characteristics useful independent method for reconstructing former
from bedrock over which the glacier has travelled. In glacier thermal regimes.
certain cases, where a particularly distinctive lithological
type lies up-glacier, it may be possible to detect the influence
of that rock outcrop on the till matrix. In a sense, therefore,
3.4 PERIGLACIAL SEDIMENTS
certain properties of tills can be used in the same way as
indicator erratics. A widely used technique has been the
3.4.1 Introduction
analysis of clay mineral ratios in the till matrix as, for In the ‘periglacial domain’ (section 2.4), freeze–thaw
example, in the establishment of ice-transport pathways activity causes fracture of the country rock and the
for sediments lying beneath the McMurdo Ice Shelf in accumulation of coarse, angular debris. This material moves
Antarctica, where it has been employed in conjunction with downslope through the combined processes of flowage
heavy mineral data (Georgetti et al., 2009). Other methods (gelifluction) and creep induced by the growth and melt
that have been used for ‘fingerprinting’ till matrices are of interstitial ice and a landscape of low-angled slopes with
geochemical analysis of mineral composition (e.g. Kubala- smooth profiles results. The deposits are known by a variety
Kukus et al., 2013), mineral magnetic properties such as of names, including ‘head’, ‘coombe rock’, ‘tjaele gravel’
magnetic susceptibility (Ojala et al., 2011a), major and and, where coarse stratification has developed, ‘stratified
trace element ratios and isotope analysis (Farmer et al., screes’ or ‘grèzes litées’ (e.g. Figure 2.28). Sediments that
2006) and micropalaeontological content (Sjunneskog & have been affected by periglacial action during the
Quaternary are widespread throughout the mid- and high- They can be up to 3 m in width and 10 m in depth,
latitude regions of the world. Frequently, their presence but mature wedges are typically 1–1.5 m in width and 4 m
reflects not only the breakdown of bedrock by cold-climate in depth. They typically form in unconsolidated sedi-
processes, but also the reworking and redistribution of pre- ments, but on occasions have been recorded penetrating
existing drift deposits. As described above (section 3.3.3.3), underlying bedrock (Owen et al., 1998). If they develop
where such processes operate in proximity to glacier ice or some time after the accumulation of the sediments in
in areas of recently deglaciated terrain where rates of which they form, they are referred to as epigenetic ice
sediment delivery are high, the term ‘paraglacial’ is now wedges. Occasionally, however, cracks form in sediments
used in preference to periglacial to describe the geomorphic that are still accumulating, and the wedges extend upwards
processes that operate in such environments and the to keep pace with sediment aggradation; these are referred
depositional sequences that are produced (Ballantyne, to as syngenetic (or synsedimentary) ice wedges and can
2002a, 2002b). grow to exceptional depths. Upon melting, the ice is
Periglacial sediments can be recognized by a number of replaced by material falling into the cracks from above and
distinctive characteristics, including the occurrence of from the sides. In this way a cast or pseudomorph of the
predominantly angular material within the sediment matrix original form of the ice wedge is preserved (Figure 3.19a).
as a result of frost-riving, the vertical alignment of many Ice wedges typically form as part of a network of thermal
stones reflecting the upward movement of particles with the contraction cracks which appear as interconnected poly-
expansion and melting of ice lenses, the presence of gons on the ground surface, although in sections they are
structures produced by ground cracking as well as flow of often found as single features.
saturated sediment (see below) and, where the deposits have Active ice wedges in present-day cold regions can
been moved by gelifluction, the preferred alignment of the develop in a variety of sediments and soil, but they occur
larger particles downslope. Clast fabric analysis in particular only in the zone of continuous permafrost (Ballantyne &
may be used to distinguish between, for example, geliflucted Harris, 1994). They therefore provide unambiguous evi-
and undisturbed glacial drift, for in periglacial deposits dence of the former existence of perennially frozen ground.
fabrics taken over a wide area should exhibit a consistent In the discontinuous permafrost zone, most ice wedges
preferred orientation parallel with the local slope. Further appear to be inactive. In arid and semi-arid periglacial
details of the range of sediments and associated processes regions and in some localities that are free-draining, frost
of the periglacial zone can be found in Ballantyne & Harris fissures that develop from thermal contraction of the
(1994), French (2007) and Martini et al., (2011). ground are frequently filled with wind-blown sediment and
are termed sand wedges (Figure 3.19b). Occasionally,
3.4.2 Structures associated with
c) c)
permafrost
Although the presence of frost-shattered bedrock and
extensive spreads of gelifluction deposits is indicative of a
former periglacial climatic regime within a particular area,
it normally provides only the most generalized of inform-
ation about former environmental conditions. However,
where sediments show evidence of ground-ice activity,
more precise palaeoclimatic inferences are possible. This
section deals with two types of structure that result from
the deep penetration of ground ice: ice-wedge casts and
involutions.
Ice wedges are considered to be diagnostic structures of
permafrost. When temperature falls rapidly, thermal
contraction cracks open in the permafrost surface, and
where these occur repeatedly at the same location, water Figure 3.19 a) Ice-wedge cast or pseudomorph of Anglian
seeps in and freezes, leading to the formation of a vertical age (MIS 12) penetrating sediments of the Cromerian complex,
West Runton, Norfolk, Eastern England (photograph by Mike
wedge of ice. The incremental accumulation of ice in veins Walker). b) Anglian age sand wedge in profile, from Broomfield,
and fissures along the axes of the contraction cracks Essex, UK (photograph by Peter Allen, Royal Holloway,
leads to the growth of wedges (Mackay & Burns, 2002). University of London, UK).
though, they are filled with other materials, including permafrost table, such as a stone pavement or cryoturbated
soil, so that in eastern Europe the more general term of horizon.
‘ground wedge’ is used, while the term ‘soil wedge’ refers In many areas where periglacial conditions prevailed,
to those features that have a high soil content. Ancient/relict unconsolidated sediments in open sections frequently
sand wedges of 2 m or more in depth probably indicate display contortions in bedding, the interpenetration of
the former presence of permafrost (Murton et al., 2000; one layer by another, and pockets that resemble load
Vandenberghe et al., 2004), whereas narrow sand wedges structures.4 Such cryostructures are frequently termed
(frost cracks) are more ambiguous as they form not only involutions, cryoturbations or festoons (Figure 3.20) and
in the active layer above and within continuous permafrost, are considered to reflect differential pressures induced by
but also in seasonally frozen ground in non-permafrost freezing within the active layer above the permafrost table
areas. (French & Shur, 2010). They sometimes appear to be
Sand-wedge pseudomorphs can sometimes be dis- irregular in spacing and heterogeneous in form, but a close
tinguished from ice-wedge casts by the characteristics of inspection often reveals a degree of order. Vandenberghe
the sediment infill, but a clear differentiation between the (1988) has recognized six different types of involution on
two is not always possible in the field, and the two forms the basis of their morphology (Figure 3.21): (1) isolated
may be found in close association in both present-day folds of small amplitude (i.e. depth of structure) but large
and former permafrost regions (Liu & Lai, 2013). Care wavelength; (2) regular, symmetrical and well-developed
also needs to be taken to distinguish ice-wedge casts from forms with amplitudes of 0.6–2 m; (3) smaller-scale
other deformational structures, such as water escape fea- versions of type 2; (4) solitary features of teardrop (4a) or
tures which, at first sight, may often appear very similar. diapiric (4b) form; (5) sediment injected upwards into
Characteristic features of wedges that can assist identifica- polygonal cracks; and (6) irregular deformation structures.
tion include: (1) depth/width ratios of between 3:1 and 6:1, Three main modes of formation have been proposed for
which conform with the dimensions of modern wedges involution structures. Periglacial loading involves the
(Ballantyne & Harris, 1994); (2) slump structures and density inversion of sediments during the thaw of frozen
stratification within the cast-fill sediment, which is usually ground. Melting may produce pockets or layers of mobile
concave downwards (‘sag’ structures); (3) the presence of saturated sediment that become liquefied and injected into
large joints and normal faults within the wedge structure overlying deposits to be replaced by denser, less fluid
and of associated micro-joint patterns in adjacent sedi- material which sinks. This process probably accounts for
ments, produced by freezing during the development most of the features classified as types 2, 3 and 4 in Figure
of patterned ground; and (4) if the wedge cast has not 3.21. The second process involves the movement of material
been truncated by erosion, the top of the cast may merge under cryohydrostatic pressure, as the ‘freezing front’
with other evidence indicating the position of the former descends from the surface each autumn towards the upper
Figure 3.20
Involutions/cryoturbation structures,
formed in Late Weichselian fluvial
sediments during the Younger Dryas
Stadial, exposed at the site of
Bosscherheide, Netherlands (see
Bohncke et al., 1993) (photograph by
Jef Vandenberghe, Free University,
Amsterdam, Netherlands).
is often the close field relationship between cryoturbations

Type 1
and other undoubted periglacial phenomena, such as ice
wedges, stone pavements and upturned stones, and thus
T y p e 2 a) where the depositional context is clearly periglacial, the
involutions can be interpreted more confidently as frozen
ground phenomena.
b) 3.4.3 Palaeoclimatic significance of

periglacial structures
T y p e 3 a) m As ice wedges and, in some cases, involutions are associ-

ated with permafrost, the former extent of permanently
frozen ground can be deduced from the distribution of
ice-wedge casts and involution structures. Thus, if the
Type 4 a) b) conditions under which permafrost is generated can be
ascertained, then the fossil forms will enable climatic
reconstructions to be made (Table 3.2). However, perma-
frost covers very large parts of the Northern Hemisphere,
extending southwards into Alaska, Canada, Siberia and
Type 5 a) b) c) other parts of Asia (Figure 2.24). It occupies some 1.59
million km2 (more than 15 per cent of the country) of China
today (Ran et al., 2012) and some 25 per cent of the
southern circumpolar region (Bockheim, 1995). Clearly,
Type 6 with such an areal extent, there can be no single definition
of what constitutes a ‘permafrost climate’. It is generally
accepted, however, that permafrost will only occur where
Figure 3.21 Classification of cryoturbation structures accord- the mean annual air temperature (MAAT) drops below
ing to their form (symmetry, amplitude, wavelength and pattern 0°C, although an upper MAAT limit for permafrost
of occurrence) (based on Vandenberghe, 1988, after Ballantyne development of –2°C, especially for mountain regions,
& Harris, 1994). For further detail see text.
has also been suggested. The development of a thick
permafrost layer requires centuries of sustained freezing.
In North America, the present southern limit of continuous
surface of the permafrost table at the base of the active layer. permafrost occurs between a MAAT of –6 and –8°C (Smith
This leads to a build-up of high pore-water pressures in the & Riseborough, 2002), while the southern limit of dis-
material trapped within the unfrozen part of the active layer, continuous permafrost in North America coincides with the
and subsequent deformation of the liquefied sediments. The –1°C isotherm. These relationships are not always con-
third mechanism is differential frost heave, which results sistent, however, for isolated but locally continuous areas
from the differential rate of freezing in sediments of varying of permafrost occur south of both of these isotherms.
composition. Frost will penetrate coarser sediment more Altitudinal limits for permafrost have been identified in
quickly than finer deposits, and the water in the latter has some mountain regions, such as Norway, where continuous
a lower freezing point. This results in differential pressures permafrost occurs at MAATs that fall below –6°C and dis-
as the ground freezes, leading to mass displacement continuous permafrost below –1.5°C. In the Rocky Moun-
(heaving) of sediment (Ballantyne & Harris, 1994). tains of North America, discontinuous permafrost occurs
As is the case in the interpretation of ice-wedge casts, where MAATs fall below –1°C.
however, care has to be taken before attributing a periglacial More important in the context of palaeoclimatic
origin to involutions, since very similar structures can reconstructions, however, are the specific climatic condi-
also be produced under non-freezing conditions where, for tions required for the generation of ice-wedge casts and
example, high pore-water pressures build up and escape cryoturbations (Ballantyne & Harris, 1994). Where the ice
routes are created in overlying layers by tectonic effects, or wedges have developed in fine sediment, a MAAT of less
where density variations lead to the injection of liquefied than –4°C is probably required, and continuous permafrost
sediments into overlying sediment bodies. However, there is likely to be present (Table 3.2). If the feature is developed
Table 3.2 Range of air temperature thresholds for the formation of selected periglacial features estimated by
different authorities in the field (after Matsuoka, 2011). MAAT – mean annual air temperature; MATCM – mean air
temperature of the coldest month; a – value for fine (clay-silt) sediment; b – value for coarse (sand/gravel)
sediment.
Warm limits by MAAT (°C) Warm limits by MATCM (°C)
Péwé, Washburn, Karte, Ballantyne Huijzer Karte, Van
1966 1980 1983 and Harris, and Isarin, 1983 Huissteden,
1994 1997 et al. 2003
Earth hummocks +3
Small involutions,
amplitude < 0.6 m –1
Large involutions,
amplitude ≥ 0.6 m –8a, –4b
Sorted patterned ground,
diameter < 1 m +4
Sorted patterned ground,
diameter ≥ 1m 0 –4 –2 to 0
Rock glaciers 0 0 to +2 –2 to –1
Organic palsas 0 –3 to 0
Thermokarst depressions –2 to 0 –1 –8 to –6
Ice wedge polygons –8 to –6 –5 –8 to –4 –6a, –3b –8a, –4b –20 –15
Soil wedge polygons –4 to 0 –1a, +1b –1 –8
Open-system pingos –2 –1 –5 to –4 –4
Closed-system pingos –6 –5 –6
in sand and gravel, a MAAT of –8°C is implied, and almost The distribution of ice and sand wedges and cryo-
certainly continuous permafrost (Matsuoka, 2011). Effec- turbations, in conjunction with other periglacial indicators
tively, this means that mean annual air temperatures (e.g. landforms), along with data from fossil Coleoptera
between –4 and –8°C produce continuous permafrost in (section 4.5.4.2), forms the basis for reconstructions of
loess regions, no permafrost in sandy-gravelly regions the periglacial environment of western Europe during the
and discontinuous permafrost in regions with both coarse- last cold stage (Huijzer & Vandenberghe, 1998). The evi-
grained and fine-grained soils (Renssen & Vandenberghe, dence suggests that, in the period from c. 27–20 k14 C yr BP
2003). Sand wedges are thought to develop under a similar (prior to and during the Last Glacial Maximum: LGM),
climatic regime, although more severe conditions may be mean annual temperatures were below –8°C in the
required, with a MAAT of as low as –12 to –20°C (Karte, Netherlands, Belgium, England, Germany and Poland, and
1983). However, sand wedges are known to form under that the –4°C isotherm of mean annual air temperature was
warmer, and wetter, conditions where sand supply is abun- situated near the French–Belgian border (Figure 3.22a). The
dant; indeed, small primary sand wedges may be associated data therefore point to a transition from continuous
with seasonally frozen ground where MAATs are very permafrost across the former countries to discontinuous
much higher (Murton et al., 2000). Soil wedges can also perma-frost (northern France) during the maximum cold
occur in seasonally frozen ground, with a MAAT of +1°C of the last cold stage. The southern boundary of permafrost
(fine sediments) to –1°C (sand and gravel). Cryoturbation lay across southern France. Mean annual temperatures
structures, on the other hand, reflect not only MAAT, but remained below –8°C over much of the region immediately
also the depth of the active layer, with large-scale structures following the LGM (prior to c. 18 ka). During that interval,
(amplitude 0.6 m) indicating permafrost conditions and a the southern boundary of the continuous permafrost zone
MAAT of –4°C to –8˚C (Huijzer & Isarin, 1997; French & was marginal to the decaying Weichselian ice sheet, and a
Shur, 2010). zone of discontinuous permafrost extended from northern
France towards northern Germany. After 20 k14 C yr BP, sheets in Poland and northern Germany (Renssen &
however, mean annual temperatures in Britain and Belgium Vandenberghe, 2003). Precipitation reconstructions suggest
rose by up to 10°C, and the permafrost zone shifted rapidly an initial period of relatively high precipitation (around
northwards (Figure 3.22b), so that by 17–15 k14 C yr BP, only 30 ka), but an increasing trend to aridity thereafter (Huijzer
a narrow zone of permafrost existed near the retreating ice & Vandenberghe, 1998).
a) shoreline
V3
-8 -8
-8
limit o f i c e s h e e t .-8
-4
,6
.-4
-S -2 .-g
V
. 3 V
.-8 V -1 to 0.5
.-8 V -4 ,8. -8
-V
8 to -6 -10 to-6,5 .8
^4 B
2 •A V -10 to -4 -4
-10to-0.S -4 V continuous
V .3 S P8 -8 permafrost V V
.4 V .-4
V .4 periglacial e v i d e n c e
-4 -8
•-4 3 coleoptera evidence
-A -4
-4 km
-1 discontinuous
-4 0 100 2 0 0 300 400
permafrost
b)
shoreline
^8
-4
-8 continuous
-8 permafrost r8
limit of i c e s h e e t
,8
,8 ,8 - 9 VV 1>5 ,4
r-8. -4 •4
-4
-B ,-8 .-4
-8 .-4 .-4 *4 ..8
.-1 K-1
-1 -4
-1 d i s c o n t i n u o u s ,-8
-1 v-1 n e r m a f m s t -4
tc .-4
- 6 . 5 to 0 . 5 V c4
,V
4 c4 deeply, seasonally periglacial evidence
frozen ground
sporadic coleoptera evidence
-4
permafrost
,-4 km
0 100 200 300 400
seasonally frozen ground
Figure 3.22 a) Mean annual temperature in northwest Europe during the 27–20 k14 C yr BP interval based on a combination of
periglacial and coleopteran evidence (after Huijzer & Vandenberghe, 1998). b) Mean annual temperature in northwest Europe during
the 20–13 k14 C yr BP interval based on a combination of periglacial and coleopteran evidence (after Huijzer & Vandenberghe, 1998).
a) b)
50 N
continuous
-15
53
permafrost
-15 -15
-15
discontinuous
-8
permafrost -16
50 N
-15 -14 -13

-14
50 N
-12 -4
-13 -11
-1 -13
-13
0 200 400
Land Sea ice Km
6°W 0" 6°E 12°E 0° 10°E 20°E
Figure 3.23 a) Permafrost limits in northwest Europe during the Younger Dryas based on periglacial evidence (after Renssen &
Vandenberghe, 2003). b) Maximum mean annual isotherms in northwest Europe for the coldest part of the Younger Dryas Stadial
based on periglacial evidence (after Isarin et al., 1998).
During the Younger Dryas cold phase (c. 13–11.7 k14 C snow cover and hence the relative abundance of ice wedges
yr BP), relict periglacial phenomena suggest that continuous and related structures (Renssen et al., 2002; Renssen &
permafrost existed north of 54°N in Fennoscandia, the Vandenberghe, 2003).
northern parts of the British Isles, and Ireland during Fossil periglacial phenomena, therefore, constitute a
the coldest part of the phase (Figure 3.23a). Mean annual potentially valuable data source for the reconstruction
air temperatures at sea level near or below –8°C and mean of Quaternary environments. However, the use of this
temperatures of the coldest month well below –20°C evidence is not without its difficulties (Murton et al., 2000).
(Figure 3.23b) characterized this zone. Discontinuous Present-day Arctic areas are not necessarily good analogues
permafrost was present between 54°N and 50°N (i.e. in for the periglacial regimes that prevailed in the mid-
central and southern England and Ireland, the Nether- latitudes during former cold stages for, while the physics
lands, upland Belgium, northern Germany and Poland). of the freeze–thaw process are essentially the same, mid-
In this zone, mean annual air temperatures were between latitude environments would have experienced differences
–8 and –1°C, with mean temperatures of the coldest month in insolation budgets, atmospheric circulation, vegeta-
above –20°C (Isarin et al., 1997). These reconstructions tion cover and levels of ground instability (French, 2007).
suggest winter temperatures were depressed by as much as In particular, there would have been considerably more
20–30°C compared with today, with summer temperatures diurnal freeze–thaw cycles than in the present High Arctic
around 4–5°C below present values. The marked decrease where the seasonal periodicity of daylight and darkness
in annual temperature range generated a Younger Dryas favours longer, more severe cycles and deeper ground
temperature regime in western Europe that may well be freezing. Certainly, mid-latitude areas did not experience
comparable to that of western Alaska today (Isarin & the extremes of freezing that now characterize, for example,
Renssen, 1999). A combination of the periglacial (and interior Siberia and Canada and, in view of the proximity
other proxy) evidence and atmospheric model simulation of large glacier masses, probably had a climatic regime
experiments suggests that during both the Younger Dryas which differed in a number of respects from that of the
and also early cold phases, the southern margin of perma- present-day high latitudes, particularly in terms of precipi-
frost in western Europe was controlled largely by the tation, wind direction and wind intensity (French, 2007).
latitude of the winter sea-ice margin in the North Atlantic Difficulties in palaeoclimatic reconstruction may also arise
Ocean. Cold winds blowing off the ice pack (situated from the fact that geomorphological processes in periglacial
around 52°N during the Younger Dryas) would account for regions are frequently determined as much by local site
(1) the very low winter temperatures; (2) the low levels factors as by prevailing climatic conditions, giving rise to
of precipitation; and (3) the resulting absence of a thick a complex variation in cryostatic facies (French & Shur,
2010). Hence, local variations in vegetation or snow cover, (section 5.6.5) and, insofar as they form time-parallel
for example, can mean that mean annual ground tempera- marker horizons, they can be employed in the subdivision
ture (MAGT) is more likely to be the determining factor and correlation of Quaternary successions (section 6.2.3.4).
in ice-wedge formation than MAAT (Smith & Riseborough, In this respect, they have proved to be especially valuable
2002). Indeed, modelling of ice-wedge networks suggests as markers in loess sequences (section 3.6), where the
that the spacing, wedge width and fracture frequency in interbedded palaeosols constitute key reference horizons for
wedges may be more sensitive to infrequent climatic events correlation at the local and regional scales (section 6.3.2.3).
and initial ground conditions, rather than prevailing
climatic parameters such as MAAT (Plug & Twerner,
2008). Further interpretational problems may arise from the
3.5.2 The nature of palaeosols
complex stages of wedge growth and degradation that have Soils are formed by chemical, physical and biological
been observed in both relict Pleistocene wedges in mid- processes operating in combination at the earth’s surface.
latitude sites (Kolstrup, 1993) and in High Arctic contexts Where the surface remains stable, near-surface layers will
(Murton et al., 2000). Finally, the dating of periglacial become progressively altered by soil-forming processes,
structures is frequently problematic, and such features in some instances resulting in well-differentiated horizons
have only been dated either on the basis of their stratigraphy that are often diagnostic of particular bioclimatic zones
or on radiocarbon dating of sub-adjacent or supra-adjacent (‘zonal’ soils). Soils therefore evolve over time, and reflect
organic material. This difficulty has, to some extent, been the influences of prevailing environmental conditions.
overcome, at least in the dating of wedges, by measuring Where soils are buried beneath younger sediments, they
the luminescent properties of sand grains (section 5.3.6) may no longer be affected by soil-forming processes,
in the sedimentary infill (Porter et al., 2001; Kolstrup, and become relict features or palaeosols. Buried palaeosols
2004). These various limitations notwithstanding, the may also be found in archaeological contexts where they
close agreement between palaeoclimatic inferences based may have been buried by either natural (e.g. alluvial or
on periglacial phenomena and those based upon other colluvial) processes, or as a consequence of anthropogenic
independent lines of evidence (see above) suggests that relict activity. Frequently, palaeosols are often found beneath
periglacial features can, when used judiciously, provide monuments, such as wall lines or field boundaries, some-
important insights into Quaternary climatic conditions, times in the form of a buried turf layer (Figure 3.24a).
particularly during cold stages; however, they are perhaps The distinction between a soil and a palaeosol, how-
best employed as a supplementary rather than as a primary ever, is not always straightforward. Most soils are, in fact,
data source. polygenetic or ‘polycyclic’, for they can evolve over such
long periods that environmental conditions may change
significantly during the period of soil formation, and thus
3.5 PALAEOSOLS soils often contain ‘relict features’, such as red colouration
(Figure 3.24b) or cryoturbation structures (Figure 3.24c)
3.5.1 Introduction inherited from a previous or subsequent climatic regime
A ‘palaeosol’ is a soil that has developed on a land surface (Kemp et al., 1993). In arid to semi-arid environments, for
of the past, and which is preserved as a fossil soil usually example, long-term changes in regional climate may lead
(but not always) by burial beneath younger sediments to alternation between episodes of clay enrichment under
(Retallack, 2001). Palaeosols frequently formed under more humid conditions and episodes of carbonate accum-
environmental conditions that differ markedly from those ulation and surface salt deposition under arid conditions
currently prevailing at a site, and therefore by relating the (Birkeland, 1999). Where relict features are particularly well
fossil horizons to those of present-day soils, deductions developed in soils, therefore, these are referred to as ‘relict
can be made about environmental conditions that obtained palaeosols’.
at the time of their formation. In particular, inferences When buried by sediment, a soil may be modified
can often be made about two of the principal soil-forming physically (e.g. by root penetration) or chemically (e.g. by
factors, namely climate and vegetation. Additional palaeo- solute percolation) through the action of subsequent soil-
environmental information may be derived from the fossil forming processes at the new ground surface. A ‘welded
content of palaeosols, for acid soils frequently contain soil’ may develop if the younger soil profile is superimposed
pollen and other microfossils, while molluscan remains may upon, or merges with, the older one, and is typical of a
be found in calcareous soils. Palaeosols have also been used protracted period of pedogenesis (Olson & Nettleton,
as a basis for the relative dating of landforms and sediments 1998). Welded soils are common in loess, alluvial and
a) a)
a)
Figure 3.24 Examples of palaeosols. a) A buried turf layer (dark horizon) beneath chalk upcast from a
Neolithic ditch, Windmill Hill, Salisbury Plain, Southern England. b) Valley Farm Soil, Stebbing, Essex, eastern
England, showing clay enrichment and ‘rubification’ under warm interglacial conditions (see text). c) Barham
Arctic soil, Great Blakenham, Suffolk, eastern England showing cryoturbation structures and sand wedges
that have been developed by ice that deposited the overlying till (photographs by Mike Walker).
colluvial sequences (Figure 3.26b). In these ‘pedocom- Quaternary environments than deeply buried soils. Equally
plexes’, it often becomes difficult to distinguish the separate difficult are exhumed palaeosols, which are those soils
effects of the different phases of soil formation. A further that were formerly deeply buried but have subsequently
problem arises in the case of accretionary or cumula- become exposed by erosion (e.g. Whiteman & Kemp,
tive soils. These form where the deposition of colluvial, 1990). Exhumed palaeosols will be affected by contem-
alluvial or aeolian sediment over a soil is so slow that soil porary pedogenic processes and they may therefore grade
formation can keep pace with aggradation at the surface laterally into buried palaeosols and welded soils. Exhumed
(Hall & Anderson, 2000). The soil profile develops upwards palaeosols have not been widely recognized, however,
and often generates very thick organic-rich horizons for in only a few areas, such as the loess landscape of the
(A-horizons) as a result. It is possible, of course, that envir- American Midwest, can they be shown to have emerged
onmental conditions may change during the accretionary from beneath an eroding overburden and thus the exhumed
process, so that lower horizons contain relict features, nature of the palaeosol can be clearly demonstrated. In the
and, where exceptional rates of accumulation are experi- absence of such stratigraphic evidence, it is often impossible
enced, the lowest horizons may become isolated from soil- to differentiate between relict and exhumed palaeosols,
forming processes. Clearly, the distinction between active and thus the value of exhumed palaeosols in Quaternary
soil and palaeosol units may be difficult to make in these research tends to be limited.
circumstances. Buried soils may also be affected by diagenesis5 as a
Distinguishing between relict, welded and accretionary result of burial and other post-formational influences.
soils is far from straightforward, and hence these types of These include disturbance by periglacial processes,
pedocomplexes are more problematic for the student of compaction due to the weight of any overburden (especially
where glaciers advance over the sediments in which the underlying geological structures, and the areal or lateral
palaeosols are developed: Figure 3.24c), or the effects of continuity of soil are further characteristics that aid in the
changing groundwater levels and conditions. All of these recognition of buried palaeosols.
can lead to changes in the physical and chemical properties The soil profile forms the upper part of the weathering
of soils, such as microfabric structures and iron oxide, profile, although some confusion has arisen, particularly
carbonate and organic matter content (Kemp et al., 1994). in the American literature where the two terms have some-
Where these changes have been particularly marked, it may times been used interchangeably. In certain cases, however,
be difficult to distinguish true palaeosols from ‘pseudosoils’, the weathering and soil profiles may be indistinguishable
which are distinctive, coloured horizons in sediment where, for example, erosion has removed the A and B
sequences, caused by the mobilization and subsequent horizons leaving only the weathered subsurface materials
accumulation of iron, manganese and other elements exposed in sections. In many palaeosols, the organic content
during diagenesis. True palaeosols can be distinguished, of the A horizon is not retained after burial (it may be lost
however, using criteria applied in the classification of by decomposition or as a result of erosion), although the
modern soils, such as degree and nature of soil horizonation mineral part of the A horizon may still be present and may
(see Krasilnikov & García-Calderón, 2006). be recognized by a clay content that is markedly different
Where a palaeosol is exposed across a site or region, it from the underlying B horizon. Generally, however, it is the
may be possible to establish the extent to which the nature B horizon which is of greatest importance in the identifi-
of the profile varies laterally as well as vertically, to form a cation of buried palaeosols. Some important diagnostic
palaeocatena. A soil catena (from the Latin word mean- features of the B horizon include colour, texture variations
ing ‘chain’), or ‘toposequence’, describes the gradational (e.g. clay-enriched or depleted horizons), weathered
changes in soil profiles as a result of variations in surface minerals, and enrichment (e.g. soils in semi-arid areas) or
gradient and the topographic position of the profile. In depletion (acid soils) in carbonate content. These properties
other words, it is a chain of related soil types (or facets) can be used either singly or, more preferably, in combin-
down a slope (e.g. Applegarth & Dahms, 2001). Relatively ation, to demonstrate evidence of pedogenesis. A careful
few examples of palaeocatenas have been published, but analysis of the properties of the B horizon may also establish
in a now classic study, Valentine and Dalrymple (1975) the type of environmental conditions under which the soil
demonstrated their value in distinguishing true soil hori- evolved. Features that should be carefully recorded in the
zons from diagenetic or weathering phenomena, for the analysis of palaeosols, therefore, include colour changes
former should show lateral variations in development (using Munsell colour charts), particle size distributions,
which bear a logical relationship to original topography clay mineral composition, organic matter content, evidence
(e.g. variations in clay content and mineral or salt con- of soil macrostructures (e.g. peds, pans and nodules) and
centrations) whereas the latter will not. Not surprisingly, variations in calcium carbonate content (Kemp et al.,
therefore, greater attention is now being paid to the spatial 2004).
variations in the nature and composition of modern soils Over the last two decades, increasing emphasis has
in order to develop models of catenas that can assist in the been placed on soil micromorphology as both a descriptive
identification and interpretation of palaeocatenas (Walk- and a diagnostic tool in palaeopedology (e.g. Stoops, 2003;
ington, 2010). Kühn et al., 2006). Soil micromorphology is the term used
to describe the distinctive arrangement of particles and
voids making up a soil fabric, which can be established by
3.5.3 Analysis of palaeosols an examination of soil thin sections under a microscope
The recognition of buried palaeosols is not always easy, (Figure 3.25). It is now widely regarded as one of the most
for a wide range of weathered materials will have been reliable methods for detecting evidence of pedogenesis,
buried in Quaternary landscapes, which may or may not elucidating the pedosedimentary history of complex sedi-
be soils. The most important property of a soil, and mentary sequences (such as loess–palaeosol sequences),
which distinguishes it from other sediments, is that it and inferring changes in environmental conditions (Kemp
has developed distinctive, vertically differentiated layers et al., 2003; Mroczek, 2013). Soil micromorphology has
or horizons in response to variations in physical, chemical also been employed in archaeological research where it has
and biological weathering, and the subsequent move- been used to address a range of archaeological questions
ment of weathering products up and down the profile at both site and landscape scale (Davidson & Simpson,
(Krasilnikov & García-Calderón, 2006). The often abrupt 2001; French, 2003). These include issues relating to site
nature of the horizon boundaries, the truncation of formation processes, agricultural practices, economic
activity and ritual practice. Soil micromorphological soils in order to identify both the types and concentra-
analysis can also reveal evidence of weathering alteration tions of minerals present (Walden et al., 1999). The method
of minerals, packing or orientation of clay particles, con- offers a rapid way of analysing complex soil-stratigraphic
centration and arrangements of voids, presence and type sequences, for magnetic minerals produced or enriched
of calcite crystal growths, animal excrements, clay coatings, through pedogenic processes generate diagnostic mineral
rootlet pseudomorphs and a range of other features (Figure magnetic signals. Further details can be found in section 3.6.
3.25). Careful analysis of such evidence helps to decipher
the history of soil formation and the changes in environ-
3.5.4 Palaeosols and Quaternary
mental conditions that may have occurred during pedo-
genesis.
environments
A second technique that is being increasingly widely The interpretation of Quaternary environments on the
used in the investigation of palaeosols is mineral magnetic basis of palaeosol evidence rests, as in other fields of
analysis (section 5.5.1). This involves the measurement Quaternary investigation, on the uniformitarian principle
of the magnetic properties of the mineral constituents of of inferring past conditions from the observed relationships
a b
LuJLul 2000 |jm
1000 um LuJLul
ws
f
la
2000 Mm 1000 (jm
Figure 3.25 Photomicrographs of pedogenic structures in soil horizons within Weichselian loess deposits in southern Poland.
a) Clay particles infilling voids. b) Irregular Mn–Fe nodules (black) and disturbed clay infillings (orange). c) Irregular Fe nodule.
d) Ferrous staining and lenticular pedogenic microstructures. e) MN-FE nodules with chemically depleted coatings. f) irregular large
MN-FE nodule. (from Mroczek, 2013, reprinted with permission from Elsevier).
between present-day soils and environments. Although example, relatively stable phases with translocation of clay
this approach seems to work reasonably well with most and calcium carbonate, a process that is characteristic of
types of biological evidence (Chapter 4), it is, perhaps, temperate soil climate, and intervening unstable phases with
less satisfactory in geomorphological and pedological the physical disruption and mixing of soil material, typically
contexts where the dangers of equifinality (different as a consequence of cold-climate processes (Rose et al.,
processes leading to the production of similar forms) are 2000).
always present. Soils with very similar physical and chemical In Europe, the textural B horizon of the last (Eemian)
characteristics can develop through a variety of genetic interglacial soil constitutes an important stratigraphic
pathways, and these may be impossible to differentiate in marker (Haesaerts & Mestdagh, 2000), and these inter-
the fossil soil profile. Thus, although buried palaeosols glacial palaeosols are especially well preserved in the loess
may be similar in morphology and in other characteristics sequences of central and eastern Europe. Comparisons
to soils forming at the present day, they may not necessarily between Eemian palaeosols and Holocene soils in Austria
be analogous in terms of palaeoenvironment and regional and the central Russian Plain showed that the former are
conditions. A further difficulty with the use of modern soils better developed in terms of a range of soil-forming pro-
as analogues for Quaternary pedogenesis is that it is seldom cesses (weathering of primary minerals, clay transforma-
possible to separate the influences of environmental vari- tion, etc.), reflecting not only a longer period of soil
ables (e.g. climatic regime) from those of other soil-forming formation, but also a warmer and possibly moister climate
factors (e.g. parent material). Moreover, some soils may not during the Eemian (Sedov et al., 2013). This accords with
be in equilibrium with present-day environmental condi- other proxy records for higher temperatures in many parts
tions. Although fossil soils have tended to be difficult to of Europe during the last interglacial (e.g. Coope, 2000).
date, particularly by means of radiocarbon or uranium- In central Asia, by contrast, data from a loess–palaeosol
series methods (section 5.3), new approaches using optical sequence in Tajikistan showed that the interglacial periods
dating (Mauz & Felix-Henningsen, 2005; Ahr et al., 2013), represented by the B horizons of the buried palaeosols of
cosmogenic isotopes (Graham et al., 2001) and mineral late, middle and early Pleistocene age were climatically
magnetism (Schellenberger et al., 2003) have proved similar to the Holocene (Bronger et al., 1998). Warm inter-
successful means of dating of palaeosols. In particular, a vals of different duration and intensity have also been
combination of dating methods, for example radiocarbon inferred on the basis of palaeosols. In southern Spain,
and luminescence (Clarke et al., 2003) and mineral mag- for example, strongly and weakly developed palaeosols
netism and luminescence (Dearing et al., 2001), are now have been correlated with interglacials and interstadials
providing increasingly robust chronologies for many respectively (Günster et al., 2001), while in the Basin of
pedosedimentary sequences. Persepolis in southern Iran, well-developed palaeosols have
A number of inferences have been made about former been correlated with major soil-forming episodes during
environmental conditions on the basis of palaeosol evi- MIS 5, while immature, weakly developed palaeosol
dence. For example, in southeast England, the widely horizons correspond with interstadials of MIS 3 (Kehl
developed Early and Middle Pleistocene palaeosol known et al., 2005). Younger palaeosols of Lateglacial age (typically
as the Valley Farm Soil is clay enriched and distinctly late Allerød and Younger Dryas) are commonly found
reddish in colour (‘rubified’), characteristics (Figure 3.24b) in aeolian sequences in western and north-central Europe.
that are normally associated with a temperate or even The Usselo soil, for example, has been found at sites in the
Mediterranean environment. Hence it has been inferred Netherlands (see Figure 3.29), Germany and Poland, and
that the soil formed under climatic conditions that were at widely dated to around 10,950 14C yr BP (Hoek, 1997). In
least as warm as, or possibly warmer than, those prevailing Germany, the Finow Soil also began to form during the
in the region at the present time (Kemp, 1987). This late Allerød (Kaiser et al., 2009). These palaeosols have
contrasts with the Barham Soil, a Middle Pleistocene provided important palaeoecological data (e.g. both are rich
palaeosol which is frequently superimposed on the Valley in charcoal and reflect widespread burning), and they are
Farm Soil unit, but which displays both macromor- also important pedostratigraphical marker horizons across
phological characteristics (incorporated aeolian sediments large areas of Europe.
and ground-ice structures) and micromorphological Finally, palaeosols are often encountered in archaeo-
features characteristic of severe arctic conditions (Rose logical contexts, frequently marking former occupation
et al., 1985; Figure 3.24c). A key element in the palaeo- horizons, for example in cave sequences (Mallol et al.,
environmental reconstruction based on these and other 2010). Careful analysis of pedogenetic features can establish
palaeosols is the soil microstratigraphy which shows, for whether a horizon was occupied several times, how much
disturbance has affected the archaeological layers, and what 3.6.2 Loess stratigraphy
types of activity were conducted, for example micro-
charcoal pieces preserved within soil horizons indicating the Loess deposits cover about 10 per cent of the earth’s land
use of fire (Holliday, 2004). It can also reveal environmental surface, blanketing hilltops, plateau surfaces and valley
factors that may have affected the archaeological record, floors alike. Extensive spreads of loessic sediments occur
such as changes in the position of the local water table or on the ‘Loess Plateau’ of North China (see below), in
disturbance by animals (Tsatskin & Ronen, 1999). central Asia and southern Siberia (Chlachula, 2003), in
Palaeosols, therefore, are of considerable value in central Europe, where loess extends as a discontinuous
Quaternary research. They constitute important strati- belt from northern France to the Ukraine (Haase et al.,
graphic markers, they provide a basis for correlation, they 2007), in the mid-continent United States (Bettis et al.,
can be used as a means of relative dating and, above all, 2003), and in the Chaco-Pampean plains and adjacent
they can provide palaeoenvironmental information that mountain regions of southern South America (Zárate,
is additional to, and often independent of, that derived 2003). The loess of North China frequently exceeds 100 m
from other sources such as fossil assemblages. Indeed, in in thickness, and near the city of Lanzhou reaches 300 m
some regions, they may provide the only source of (Derbyshire et al., 1995). In central Asia, loess accumula-
palaeoenvironmental information for periods during which tions of about 90 m are common, but locally can exceed
little or no sediment has accumulated. In some regions, 200 m, whereas in Europe and North America, thicknesses
buried soil sequences represent a substantial proportion of about 30 m are more usual. More restricted spreads of
of the Quaternary period (Hall & Anderson, 2000). While loess occur in many other parts of the world, including the
it is true that there is still much to be learned about modern Mediterranean region, the Middle East, India and Pakistan,
soils that is crucial to the interpretation of palaeosols, it is the western USA, New Zealand and Alaska (Derbyshire,
equally the case that much can be learned by pedologists 2003).
from the Quaternary archive of palaeosols, about the ways Loess has a high carbonate content, sometimes exceed-
in which modern soils have evolved, and about the rates and ing 40 per cent by weight, and it frequently possesses a
effects of weathering and related pedogenic processes. distinctive range of sedimentary properties, including heavy
mineral and clay mineral suites, sediment magnetic vari-
ations and micromorphological features, all of which can
3.6 WIND-BLOWN SEDIMENTS be used to characterize and correlate loess sequences.
The fine-grained and friable nature of loess deposits gives
3.6.1 Introduction rise to unstable landscapes into which rivers are rapidly
Large areas of wind-blown sediment (sand seas) are incised, and this creates numerous exposures for detailed
currently forming in arid and hyper-arid regions of the study (Figure 3.26a). In section, the sediments usually dis-
world, while blankets of sandy material (grain size 64 μm play little visual evidence of stratification, although careful
to 2 mm) or coversand were deposited in many mid- examination often reveals faint bedding. Microscopic and
latitude temperate regions during the last cold stage. SEM examination of the particle fabric (micromorphology)
Elsewhere, silt particles (2–64 μm) form the bulk of wind- can distinguish differences in texture that are important for
blown sediment, material which is referred to as loess. palaeoenvironmental reconstruction (Kemp et al., 2001;
Spreads of loess are found throughout the world, and Josephs, 2010). These analyses have shown that there are
these often accumulate to great thicknesses. Analysis of two types of loess, one that is typical of areas such as China
these loess sequences has shown that not only are they that have experienced arid conditions during glacial periods
frequently of considerable antiquity (some extend back and humid conditions in interglacials, and a second type
into the Pliocene), but they also contain evidence of cyclical that has accumulated in areas with persistently high
climatic change, with phases of aeolian sedimentation humidities, such as western Europe. The most distinctive
(reflecting cold stages) interspersed with episodes of soil features in loess sequences, however, are palaeosols, and
formation during warm stages (section 3.5). Wind-blown these often show up in section as darker units in exposed
sediments, both loess and coversands, are also valuable profiles (Figure 3.26b). Differences in pedogenic history of
indicators of former wind directions. In addition, dust of these buried soil units revealed, for example, by grain size
terrestrial origin can be identified in marine and ice cores, variations, micromorphological characteristics and the
and these archives enable both the former transport paths magnetic susceptibility record, provide a basis for
and the relative flux of air-transported dust to be correlation between individual loess–palaeosol sequences
reconstructed at a global scale. (Xiong et al., 2001).
a) b)
Figure 3.26 a) The early Quaternary ‘Wucheng’ sequence of loess units and soil horizons (darker bands) exposed at the Luochuan
site on the Loess Plateau, China; prominent in the photograph are loess unit 15 (L15) and the Wucheng group of soil horizons
(WS–1, comprising S15–S23) dating to between 1.2 and 1.6 Ma. b) The S1 (last interglacial) soil profile (dark red layer) exposed
at the Lantian section, near Xi’an in the southern part of the Loess Plateau, where three separate soil horizons exposed in other
areas converge to form a welded soil (section 3.5.2) due to unusually low rates of dust accumulation (photographs by Thomas
Stevens, Royal Holloway, University of London, UK).
The largest area of loess deposits in the world is found Tajikistan, for example, the loess–palaeosol record extends
on the Chinese Loess Plateau. This covers an area of around back to c. 1.77 Ma, and the climate cycles correspond
440,000 km2 between 33–40˚N and 98–115˚E (Figure 3.27). closely with both the Chinese loess and deep-sea oxygen
The sedimentary record contains a sequence of interbedded isotope records (Ding et al., 2002a). In the middle and
loess and palaeosol units which developed during successive lower Danube basin of southeastern Europe, loess deposits
cold (glacial) and warm (interglacial) episodes, with the extend to the base of the Pleistocene. The period from
earliest, the ‘red clays’, dated to c. 7 Ma (Ding et al., 1999). around 1 Ma onwards (MIS 27) was characterized by
The loess appears to have three principal sources: silt- alternating loess deposition and pedogenesis during glacial
sized material derived from adjacent deserts particularly to and interglacial periods, respectively. Loess accumulation
the north and west of the Loess Plateau, glacially derived rates increased after ~700 ka reflecting a shift to a more
material from the surrounding mountains, and recycled steppic, and therefore relatively more arid, environment
older loess that has been fluvially transported and deposited compared with the lower parts of the profiles (Fitzsimmons
in river valleys within the Loess Plateau region (Kohfeld & et al., 2012). In South America, the beginning of loessic
Harrison, 2003). The loess–palaeosol sequence contains sedimentation dates back to the Late Miocene (c.10 Ma),
up to 37 soil units that have formed during the last 2.5 Ma although the Plio-Pleistocene record is mostly composed
(Figure 3.28). Some of these are very distinctive and wide- of loessic sediments modified by pedogenesis, which
spread, such as the S1 (last interglacial) and S5 (MIS 13, 14, produced welded palaeosols (Zárate, 2003). The longest
15) palaeosols. However, a number have proved to be continuous loess–palaeosol sequence (La Carreras, north-
complex polygenetic units, consisting of several soil west Argentina) contains thirty-one intercalated palaeosols
horizons separated by thin ‘second order loess beds’ with and has a minimum age of 1.15 Ma for the onset of loess
fainter ‘second order pedogenic features’ in some of the accumulation (Schellenberger et al., 2003). In western
more distinctive soil units (Kemp et al., 1995). These Europe and North America, by contrast, the loess sequences
variations appear to reflect climatic change, the well- are generally younger, and the deposits are usually thinner
developed palaeosols probably indicating relatively long and span a much more limited age range. In southeast
periods of pedogenesis during interglacial stages, while the England, which lies at the western extremity of the central
less prominent soil units reflect much shorter episodes of European belt of wind-blown sediments, loess-like deposits,
soil formation during interstadial periods (section 3.5.4). locally termed ‘brickearths’, date largely from the last cold
Numerous palaeosols, comparable in age with those stage (MIS 2), but there are also localized patches of older
of China, have also been discovered in central Asia. In (mainly MIS 6 and 12) loess, which are the dissected
inn= 110"
Rocky Desert
Rocky Desert
Desert
4C Beijing
ng
ua
Qi
ng
zh
ha
ua
Chai D a m i-X
ijia
oy
iza Vinchua'n
Basin ng
ish
ca
Desert :
Yulin
al
ijaiyuan
jJ
ishijiazhuang
Lishi
Dingbain
oyuan
oyua
Jingbian
ua ng Wucheng
zhng
Qtnghat ijiaijia
ishish zh ua
jJalca
Xming
jJalca
Lak& jJalcaoyuan
jJa
ua
Xifeng
oy
Qinghai-Xizang Plateau Lanzhou
lca
ca
Luochuan
oy
al
jJ
ua
35° We 35-
ina
n
n Weinan
Baoji Zhengzhou
Qin Xi'arv
gha
i-X Qin
iza Q
g hi na g
ng i -hXa i - X
Pla iza iza Hua-Bei Plains
0 300 km tea n n
u
100° 110°
Figure 3.27 The Loess Plateau, major deserts and mountain regions of north-central China (from Ding et al., 1994).
So
•Si
S2-
11
•Si S 1
•Si 2
S
•Si
-II
4 s 3
S5-1
•Si S5-111
•Si
•Si
s, B/M
•Si S
9-ll J
•Si
» 1 J •S11
•Si metres
375 14
S
^13 0
?1& S,
ias 5
•Si 10
;
S-7
•Si 21
$2j •Si
S
20
•Si s
•Si
24
^26
0
•Si 30
S .=
2
$27
30 SRI
40
S
•Si
M/G
S32 0 10 2 0 3 0 4 0 5 0 metres SO
R e d clay
Alluvia
Palaeosol
Loess
Figure 3.28 Loess–palaeosol succession at Baoji, in the southern Loess Plateau, north-central China (for location see Figure
3.27) showing thirty-two palaeosol units spanning the last 2 Ma. The palaeomagnetic timescale (see Figure 5.34) is shown on the
right. B/M – Brunhes–Matuyama boundary c. 0.78 Ma; J – Jaramillo event, 1.07–0.99 Ma; O – Olduvai event; 1.956–1.79 Ma; M/G
– Matuyama–Gauss boundary, c. 2.6 Ma (from Ding et al., 1994).
remnants of originally more extensive deposits (Antoine

et al., 2003). In North America, complex successions of
loess, sand units and at least four cycles of pedogenesis,
dating approximately from the last 500–400 ka, have been
reported from the central Great Plains (Feng et al., 1994),
but most loess sequences in the USA only span the period
from the last interglacial (Bettis et al., 2003), the most
widespread being the Peoria Loess, dated to c. 25–14 ka,
with the highest dust accumulation rates between 18 and
14 ka (Roberts et al., 2003). In places, the loess is up to
48 m in thickness, making it probably the thickest Last
Glacial Maximum loess accumulation in the world (Maat
& Johnson, 1996).
3.6.3 Mid-latitude sand belts (coversands)

Coversands dating from the last cold stage are found in
many parts of Europe and North America, their presence
reflecting both availability of material and increased wind
strength at that time. Much of the sediment was derived
from the greatly expanded periglacial regions, where a
combination of sparse vegetation and seasonally dry ground
meant that large areas of unconsolidated and friable sedi-
ment were left exposed to wind action. Particularly vulner-
able were the unvegetated surfaces of glacigenic and out-
wash sediments, which were rapidly stripped of their finer
components (Wright, 2001). Considerable quantities of Figure 3.29 Coversands and interbedded Usselo soil of
Lateglacial (late Allerød) age (section 3.5.4) exposed at a site
sand were also removed from the continental shelves, in the Netherlands, with periglacial geomorphologist Kees
which had been exposed by a eustatic fall in sea level of Kasse of the Free University, Amsterdam (photograph by Mike
over 100 m, while unvegetated floodplains of large rivers Walker).
provided a further source of material. A more vigorous
atmospheric circulation around the ice sheets, coupled
with strong katabatic winds blowing off the glaciers and are driven by the wind to form interdigitating layers
ice sheets, resulted in the stripping of these surfaces and (Ballantyne & Harris, 1994). Subsequent melting of the
the deposition of the coversand and loess belts of the snow destroys any bedding and leads to deformation and
northern mid-latitude regions. In addition, aeolian sedi- cracking of the sediment layers, collectively termed
ments themselves are liable to severe erosion, and consti- denivation features.
tute a secondary source of minerogenic dust (Derbyshire The European coversands typically comprise a number
et al., 2000). of distinctive sand units, which are interbedded with finer
The coversands form a featureless surface, occasionally sand and loamy layers (Figure 3.29). In northwest Europe,
fashioned into undulating dunes, over large parts of the these typically comprise an Older Coversand (mostly
northern Great Plains of the USA and Canada (Forman fluvially reworked aeolian sands) and dated to 25–22 ka, an
et al., 2008). In lowland Europe, they stretch from northern Older Coversand II/Younger Coversand I (sandy to silty
France, the Netherlands and Belgium, eastwards through sand sheets) dated to 17–13 ka, and finally a Younger
Denmark, northern Germany and Poland into Russia. This Coversand II (cross-bedded dune sands), with an age of
mantle of yellow to grey sand is often several metres in 13–11.5 ka (Frechen et al., 2001; Kasse et al., 2007). This
thickness, and individual sand units are relatively homo- complex sequence suggests that, towards the end of the last
geneous, with little or no stratification. The sediments are cold stage, northwest Europe experienced abrupt but short-
frequently deformed by folding and tension cracks, reflect- lived changes in climate from polar desert to relatively
ing niveo-aeolian deposition, where both sand and snow humid and more temperate conditions.
3.6.4 Low-latitude ‘sand seas’ Pleniglacial periods (MIS 4 and 3), and a dominant
northwesterly wind during the Late Pleniglacial (MIS 2:
In addition to the extensive spreads of wind-blown Bokhorst et al., 2011). Proxy records from northwest
sediment that form the present-day ‘sand seas’ of arid and Europe also indicate a predominantly westerly to north-
hyper-arid low-latitude areas of the world, aeolian sands westerly wind in winter during the Late Pleniglacial.
have been found beneath the present-day soils and vege- Inferred easterly winds in more northern areas appear to
tation of many savannah and forest regions, which indicates reflect katabatic winds from the Scandinavian ice sheet.
that these areas have experienced more arid conditions Modelling results suggest that the prevailing winds were
in the recent past. In South America, for example, some stronger than the predominantly southwesterly winds over
25 per cent of the continent is covered with palaeo-aeolian western Europe at the present day, caused by an eastward
features which reveal a complex history related to atmos- shift of the Icelandic Low pressure system, and an enhanced
pheric circulation regime change during the Late Quatern- pressure gradient over northwest and central Europe during
ary (Tripaldi & Forman, 2007). Similarly in northwest the last cold stage (Renssen et al., 2007a).
China, there are enormous spreads of aeolian material, In eastern Asia, a range of palaeoclimatic data has
much of which has been fashioned into dunes (section been obtained from the analysis of wind-blown sediments
2.7.2), and which provides evidence of the former devel- and related deposits. On the Chinese Loess Plateau,
opment of extensive sand seas during the last cold stage glacial–interglacial cycles are reflected in particle-size,
(Wang et al., 2004). The same is true of the great Sinai- geochemical and soil-micromorphological evidence
Negev erg (sand sea) of Egypt and Israel which comprises from loess–palaeosol sequences, arid (glacial) phases being
aeolian material transported from the Nile delta region characterized by increased dust deposition and semi-arid
during the last cold stage (Muhs et al., 2013). In Australia, episodes (interglacial/interstadials) by episodes of soil
the marine record for dust flux during the LGM is at least formation (Kemp & Derbyshire, 1998). The loess–palaeosol
three times greater than that for the Holocene, driven by sequence has also provided a proxy record for the Asian
both weakened monsoon rains and drier westerly monsoon which may have been initiated as long ago as the
circulation (Hesse & McTainsh, 2003). Many present-day late Miocene (Porter, 2001). Different properties of the sedi-
arid or hyper-arid regions contain evidence of earlier, ments have been used to reconstruct changes in monsoon
wetter phases. Interbedded within, or buried beneath, the intensity and circulation. For example, grain-size varia-
surficial sand seas and dunes can often be found palaeo- tions reflected in mineral magnetic signals demonstrate
drainage systems (Figure 2.59) that testify to the former changes in both the strength and position of the Asian
existence of wetter conditions (e.g. Robinson et al., 2000; monsoon cell during the Late Quaternary. Horizons char-
Craddock et al., 2010). In the Sahara, for example, satellite acterized by ultra-fine grains (high magnetic susceptibility)
imagery reveals the presence of a dense palaeo-river are typical of episodes of reduced aeolian deposition and
network and this, combined with archaeological and fossil pedogenesis, while periods of increased influx of wind-
evidence, indicates a markedly more humid phase during blown dust are indicated by lower magnetic susceptibility
the early Holocene (Drake et al., 2010). measurements. The latter represent periods when the winter
monsoon (associated with dust transport from Siberia) was
dominant, while evidence of pedogenic development
3.6.5 Wind-blown sediments and
reflects a reduction in the strength of northerly winds and
palaeoenvironmental reconstructions an increase in intensity of the summer monsoon from the
Former wind directions can be reconstructed on the basis south (Liu & Ding, 1998). Hence the magnetic susceptibility
of grain-size or mineralogical variations of wind-blown record through a loess–palaeosol sequence can be used as
sediments. For example, grain-size analysis of the Peoria a proxy index of the intensity of the summer monsoon
Loess on the Great Plains of the USA shows that the deposit (Figure 3.30). Changes in the strength of the Asian mon-
becomes progressively finer eastwards towards the soon on the Chinese Loess Plateau have also been deduced
Mississippi River (Mason, 2001). Maximum loess thick- from variations in the concentration of a range of chem-
ness also declines towards the east. The evidence suggests ical and isotopic indicators. For example, in the Huanxian
predominantly northwesterly loess-transporting winds profile in northern China, variations in 87Sr/86Sr and Zr/Rb
across the Great Plains region during the last glacial period ratios, and mean grain size, were interpreted as reflecting
(Muhs & Bettis, 2000). In central and eastern Europe, a weakening in the East Asian monsoon between 40 and 30
spatial variations in loess grain-size distributions suggest ka, but an increase in monsoon strength from 30–10 ka
a prevailing westerly wind during the Early and Middle (Yang et al., 2005). Finally, magnetic parameters have been
source in Patagonia and the Altiplano of Bolivia (Delmonte

Magnetic Mean
Strata Polarity et al., 2010). The Antarctic dust record extends back over
Susceptibility (SI) (pm)
Age
0 150 300 25 15 5
800 ka and shows a significant correlation with temperature,
Malan (Ma)
Loessf 0 with a ~25-fold increase in dust concentration during
glacial periods (Lambert et al., 2008). Dust concentrations
Brunhes
have also been measured in several ice cores from

Lishi Loess
40
Greenland where again there is a strong relation with
climate, the ice from the last glacial period containing dust
0.78
concentrations 10–100 times higher than the Holocene;
0.99 moreover, abrupt changes in climate tend to coincide with
J 1.07 80
Lishi Loess
sudden changes in dust concentration (Steffensen et al.,

2008). Continental dust has been found in deep-marine
Wucheng Loess
cores using geochemical and magnetic methods and has

120
O 1 77 been used, for example, to correlate marine and ice-core
(w) mdaa
1 95
O records (Pugh et al., 2009). Increased aeolian flux to the
oceans may also have climatic consequences, for increased
160 concentrations of Fe-rich dust may be responsible for
Lishi Loess
2.60
phytoplankton blooms in ocean waters, and these may draw
down atmospheric CO2. This scenario has been suggested
361 200 for the Southern Oceans (Watson et al., 2000), while
enhanced dust flux from continental North America could
Red Clay
4.19
Lishi Loess
have had significant effects on the productivity of the

240
North Atlantic Ocean during the Last Glacial period (Bettis
523 et al., 2003).
588 Quaternary wind-blown sediments are therefore valu-
able data sources for palaeoenvironmental reconstruction.
240
2B0
7 04 At the local and regional scales, they provide evidence
of sediment source and transport, and hence of palaeo-
Loess Reddish PaJeosal Bed Rock
wind directions. At the global scale, these data can be used
Loess to reconstruct former atmospheric pressure systems, and
also to infer those times when atmospheric dust transport
Figure 3.30 Magnetostratigraphy, lithology, magnetic sus- was greater than that of the present time. The loess–
ceptibility and grain-size variations of the Lingtai section palaeosol sequences are important archives of palaeo-
(35°04’N, 107°39’E) on the Chinese Loess Plateau (after An,
2000).
environmental data, and constitute some of the longest and
most detailed records of Quaternary climatic change. They
also provide one of the principal sources of evidence for the
used, in association with other pedostratigraphic charac- correlation of terrestrial and marine sequences, a topic that
teristics to obtain quantitative estimates of dust deposition is discussed in greater detail in section 6.3.3.
on the Loess Plateau (Kohfeld & Harrison, 2003). From
these, palaeoprecipitation estimates have been obtained
which again enable the varying strength of the East Asian 3.7 LAKE-LEVEL RECORDS FROM
monsoon to be inferred (Liu & Ding, 1998). Variations in
grain-size parameters also enable individual loess–palaeosol
LOW-LATITUDE REGIONS
sequences to be correlated and resolved into a single
‘stacked’ climatic record which, in turn, can be correlated
3.7.1 Introduction
with the oxygen isotope record of the deep-ocean sequence The geomorphological evidence for environmental
(Ding et al., 2002b). changes in low-latitude regions during the Quaternary
Dust transport fluxes to the polar ice sheets can also be was reviewed in section 2.7. However, much longer and
reconstructed using geochemical and isotopic tracers. For more detailed lake-level histories can be reconstructed
example, isotopic tracers in dust records from six different using sedimentological data. The records can be broadly
ice cores drilled on the east Antarctic Plateau suggest a divided into those from humid, low-latitude regions, and
Elevation (m) -200 -200
-300 -300
-400 -400
-700 -700
0 10,000 20,000 30,000 40,000 50.000
A g e (ka)
Figure 3.31 The water-level record for Lake Lisan over the course of the last 55 ka. The lake-level fluctuations reflect the
hydrological conditions in the large watershed of the lake which, in turn, is influenced by the hydro-climatic conditions in the central
Levant region. The highest lake levels occurred during MIS 2, while lower lake levels are recorded during MIS 3 and especially
during the Holocene (MIS 1). The highest stand was reached at c. 26 ka during the coldest stage of the Last Glacial cycle. Note
that lake-level heights are shown in metres below mean sea level (mbsl) (after Torfstein et al., 2013).
those more typical of arid and semi-arid areas. In some low- California, for example, the earliest sediments may be as old
latitude lakes, long sediment records are preserved, some as Late Pliocene, with the most detailed part of the record
of which span the whole of the Quaternary and extend back (the last 150 ka) revealing a series of climatically related
into the Pliocene (Torres et al., 2013). Most attention has lake-level fluctuations that appear to be driven by orbital
focused, however, on the ‘pluvial lake’ sequences (section forcing (Smith, 2009). Further north in the Bonneville
2.7.1) of the arid and semi-arid regions, for not only do Basin of Utah, a long sequence of deep-lake cycles is also
these records provide some of the most dramatic evidence recorded, the oldest of which appears to correlate with
of Late Quaternary environmental change, but the data MIS 16 (c. 650 ka) of the deep-ocean record (Oviatt et al.,
from these lakes are used in the construction of palaeo- 1999). In many cases, however, the borehole records and
climatic models at both the continental and global scales the majority of surface exposures span much shorter time
(Qin et al., 1998; Kohfeld & Harrison, 2000). intervals and hence a continuous sequence of lake-level
change can usually only be constructed for the Holocene
and last cold stage. In the Jordan Valley of Israel, for
3.7.2 Pluvial lake sediment sequences instance, the sedimentary record for ancient Lake Kinneret,
Cores from saline lakes or from salt-encrusted pans of the Last Glacial Dead Sea (Figure 3.31), extends back to
former playa lakes often reveal complex sequences of lake around 70 ka (Torfstein et al., 2013).
sediments (silts, marls, clays and organic muds), inter- The sedimentary sequence in closed basin (endoreic)
bedded with units of alluvium, colluvium, aeolian sedi- lakes is closely related to changes in water balance which,
ments and occasional soil horizons. Unconformities are in turn, reflect changes in regional climatic regime. If the
also common. In some places, for example in parts of links between lithological variations and water balance
Egypt and northern Kenya, the upper parts of the sequences can be quantified, it is possible to reconstruct the former
are exposed and the intercalated lacustrine and terrestrial climatic conditions under which the lake sediments accum-
sediments can be examined in section (Brookes, 1993). ulated. The hydrological budget of a lake is controlled by
Elsewhere in east Africa and further north in the Jordan the balance between the inflows and outflows from the
Valley of Israel, sections through older lake sediments system as follows:
provide ‘snapshots’ of episodes of lake history (Hay &
Kyser, 2001; Stein, 2001). The sedimentary sequences often dV
= P + Si +Gi E So Go
reveal episodes of lake expansion (reflecting higher rainfall dt
regimes) interrupted by periods of contraction during
drier conditions (Figure 2.55), and some records extend where dV is change in lake volume; dt, the time interval;
back throughout the Quaternary. In the Searles Lake basin, P, precipitation on to the lake surface; Si , surface inflow
from rivers and/or overland runoff; Gi , ground water lake, secondary gypsum deposition occurs on the exposed
inflow; E, evaporation from the lake surface; So, surface sediment surface and clay pellets also form. Prolonged
outflow; Go, ground water outflow (Jones et al., 2007b). desiccation will result eventually in soil formation. A
Changes in lake area or volume could, therefore, reflect the re-expansion of the lake will reverse these trends, the
influence of any of these variables. However, not all causes precise nature of the sedimentary sequence depending
of lake-level changes are necessarily related to climate. For upon the rate and magnitude of the water-level changes.
example, overspill from adjacent lakes, influx of glacial Other indicators of the fluctuating hydrologic balance of
meltwaters during cold stages, changes in basin configura- lakes include changes in magnetic properties, with reduced
tion due to tectonic activity, or the creation of dams by magnetic values being noted during episodes of high
avalanche debris, talus cones or lava flows can all affect lake productivity during shallow water periods (Zic et al., 2002),
volume and lead to fluctuations in the lake-water level. and in total inorganic carbon and δ18O, both of which
Where sedimentary (and indeed geomorphological) decrease with increasing lake size (Benson et al., 2003). The
evidence can be obtained from a number of lake basins links between the hydrological balance and chemistry
within a particular region, and where these show a con- (e.g. stable isotope ratios) of a lake body can be quantified,
sistent trend in water balance changes for a given time using modified versions of the water balance equation
period, a regional climatic signal may reasonably be presented above, and these are now widely used as palaeo-
inferred, and hence the lake sediment record can be used climatic proxies (Jones et al., 2007b). The sedimentary
as a proxy for climate change. and geochemical responses to lake-level fluctuations
As lakes change in volume, the sedimentary and geo- will vary considerably between individual lake basins,
chemical environment also changes. In salt lakes, for however, and each sedimentary sequence must therefore
example, salinity declines as water volumes increase, and be interpreted in the context of local sediment chemistry,
hence clastic sediments of relatively low salinity tend to detrital input and hydrogeology. Detailed analysis of
accumulate when lakes are expanded. By contrast, when the geochemical variations in particular can enable water-
lakes decrease in size, salinity levels rise, until primary level variations to be reconstructed which, in turn, may
carbonates are deposited first, followed by gypsum and provide insights into former regional climatic controls,
then by halite. Algal limestone growths (stromatolites) such as changes in the strength of monsoon circulation
may also develop at this stage. With seasonal drying of the (Figure 3.32).
% Dolomite % Carbonate ( P°Db Bu ik carb ( ( P D) B

6 1 3 p D B
%Organiccarbon
0 5 10 15 0 40 80 -2 0 2 4 6 40 0 20 4 0
4000
5000
6000
cal. y r B P
7000
8000
9000
-8 -10 -12 -14 25 20 15 10 -8 -10 -12 -14 -10 -14 -18 -22
° 1 8
CW t , (PDB) C/N oi90 (PDB) (PDB (PDB)
6^0
a f o r g e n c ( u s l a l i o n
(PDB0fa
Figure 3.32 The geochemical record from a lake sequence in Tibet that developed between 9 and 4 ka. Lake geochemistry appears
to have been particularly sensitive to water-level fluctuations driven by precipitation changes associated with variability in monsoon
strength. The grey-shaded horizontal bars, for example, represent intervals of reduced carbonate content and the spread of the
aquatic plant Potamogeton, reflecting lower water levels and a weakened monsoon (from Morrill et al., 2006).
Quantifying the linkages between water balance and considerable ‘residence time’ in soils around the catchment
climate on the one hand, and between water chemistry and (section 5.3.2.5). Again, this will contribute to an ‘ageing’
water budgets on the other, rests heavily on reference to effect in radiocarbon dates obtained from lake sediments.
modern analogues. Where present-day equilibrium condi- In addition, the effective age range of radiocarbon dating
tions can be established for a lake basin, modelling tech- extends back only to c. 45 ka (section 5.3.2) and hence the
niques can be used to simulate the conditions that give technique is not applicable to the dating of long lake
rise to changes in either water budget or water chemistry. sediment sequences. Other techniques have therefore been
For Lake Malawi, for example, Kumambala & Ervine (2010) employed to establish chronologies of lake-level change,
developed a water balance model based on a combination including U-series dating of carbonates (Ma et al., 2004)
of lake outflow records and precipitation/evaporation and stromatolites (Lisker et al., 2009); amino-acid dating
estimates from climate stations along the lake shore, of molluscs and other organic materials (Kaufman, 2003a);
and combined this with climate change modelling using magnetic stratigraphy and tephrochronology (Benson et al.,
the IPCC emissions scenarios to assess the likely future 2003); luminescence dating of lake sediments and inter-
behaviour of the lake. Further north in Ethiopia, Kebede bedded aeolian or fluvial deposits (English et al., 2001); and
et al. (2006) generated a hydrological model for Lake Tana varve chronology (Prasad et al., 2009). All of these methods
based on similar series of parameters, which showed that are described in Chapter 5.
in this case, the lake was less sensitive to recent rainfall
variations than other large lakes in Tropical Africa. Using
3.7.3 Lake-level changes and Quaternary
a different approach, Benson & Paillet (2002) developed
a hydrologic-isotopic balance model for palaeolakes for
palaeoclimates
application to palaeolake δ18O records based on a combina- Because of the potential value of lake-level changes in re-
tion of recent and historical datasets. Applications indicate constructing Quaternary climates, a Global Lake Status
that when lake volume increases, δ18O decreases and vice Data Base of palaeo-lake records has been compiled (http://
versa. Hence the δ18O record can be used to infer the direc- www.ncdc.noaa.gov/paleo/lakelevel.html), which classi-
tion of climate change, that is, whether climate is becoming fies the records for particular time intervals on the basis
drier or wetter (Benson et al., 2002). of lake-level status (low, intermediate and high). When
Other approaches to the study of the relationships last updated (2011), the database contained modern lake-
between lake-level variations and climatic change include status data for 599 lakes, and estimates of the status of
the use of biological indicators of lake salinity, particularly 413 lakes at 6 ka and of 103 lakes at 18 ka (Figure 3.33). The
diatom and ostracod assemblages (sections 4.3 and 4.8), or number of lakes in the record therefore decreases with
the chemical analysis of sediments or fossils. The latter age, and comparatively few lakes date from pre-Holocene
approaches include the analysis of the carbon, nitrogen and times. There are also significant regional gaps, with very
hydrogen isotopes of lake organic matter (Abbott et al., few records from South America, central and southwest
2000), the carbonate mineralogy of the sediments (Nelson Africa, or for large parts of Eurasia and the Far East. The
et al., 2005), and the oxygen isotope content of calcite, classification into ‘low’, ‘intermediate’ and ‘high’ status is
sediments, molluscs and ostracods (Benson et al., 2003). also rather generalized, while some of the lake sequences
Complementary information can be obtained from the may not have been dated accurately. Despite these limita-
study of fluvial sediments in low latitudes, since evidence tions, however, the database constitutes an important
of enhanced fluvial activity may reflect periods of higher archive of palaeoclimatic information, and the palaeo-lake
precipitation, and might provide independent support for reconstructions also provide valuable benchmarks for
lake sediment evidence of ‘pluvial’ episodes (e.g. Blanchet evaluating numerical climate-model outputs (e.g. Qin et al.,
et al., 2013). 1998).
Prior to the 1980s, most chronologies of lake-level Street & Grove (1979) found that tropical lake records
changes were based on radiocarbon dating of carbonate from Tropical Africa revealed evidence of widespread
materials, such as algal limestones, calcite-cemented sands aridity at around the Last Glacial Maximum (LGM), but
and mollusc remains. These are not ideal media for radio- generally achieved their highest levels during the early
carbon dating, however, for they are easily contaminated Holocene, after which conditions became increasingly arid
by older carbon residues (Licciardi, 2001). The problem is up to the present. They attributed this pattern to increased
exacerbated by the mobility of CaCO3 in arid and semi-arid sea-surface evaporation following the global rise in eustatic
environments, and by the fact that organic carbon that has sea level that resulted from the melting of the last ice
been washed into lake sediments will often have spent a sheets. In the southwest USA, the lake records showed a very
0 -120 -60 0 60 120 0
a) O k a
60
-60
0 -120 -60 0 60 120 0
b)6ka
60
-60
0 -120 -60 0 60 120 0
c) 18 ka
60
high
intermediate
low
-60
Figure 3.33 a) Locations of lakes for which lake-level data are lodged with the PMIP Global Lake Status Data Base (updated
2011). b) Lake status data for 6 ka. c) Lake status for 18 ka (source of data: http://pmip2.lsce.ipsl.fr/synth/lakestatus.shtml).
different pattern, with high levels from LGM times until the to an increase in aridity in this region. Higher lake levels
early Holocene, followed by arid conditions from c. 11–10 in North Africa, Arabia and India reflected enhancement
ka (Smith & Street-Perrott, 1983). The timing of lake-level of the Afro-Asian monsoon circulation. In some parts of
variations in Australia appears to have been intermediate Africa, precipitation may have been around 15 per cent
between those of Africa and the USA, with lake levels gen- higher than at present (the ‘African Humid Period’, AHP;
erally falling between c. 30 and 18 ka, rising steadily through Bergner et al., 2003). By 6 ka, a marked insolation anomaly
the first half of the Holocene, and subsequently falling again over northern mid-latitudes led to further aridification
after achieving maximum levels at c. 8 ka (Harrison & of the American Southwest, and also of parts of Eurasia,
Dodson, 1993). while North Africa remained humid under a strong Afro-
This general pattern of lake-level changes was broadly Asian monsoon (Figure 3.34). By 3 ka, declining seasonal
confirmed, and further developed, in a geostatistical analysis insolation in the northern hemisphere was accompanied
of the Global Lake Status Data Base by Viau & Gajewski by an increase in precipitation in the American Southwest
(2001). They examined the data over the last 12 ka, and a rise in lake levels, while a concomitant weakening
and generated lake-status patterns (grids) at 3 ka intervals. of the Afro-Asian monsoon led to aridification in North
The results suggest the following scenarios. By c. 12 ka, the Africa, Arabia and parts of Asia. Viau & Gajewski (2001)
northern ice sheets were in rapid retreat and most of also compared the 6 ka lake-level status grids with simu-
North America and Europe was wet, perhaps due to the lations from four atmospheric circulation models (section
northward migration of the jet stream. At the same time, 7.2.3) to underline their usefulness in validating broad-scale
high lake levels in Africa probably reflected higher land–sea climate model outputs (e.g. Coe & Harrison, 2002).
temperature contrasts and an enhanced African monsoon. Conversely, because modelling methods can experiment
By 9 ka, the Eurasian ice sheet had almost disappeared and with different climatic inputs, they can also inform the
the Laurentide ice sheet was significantly reduced in size. debate concerning the climatic parameters that lead to
The northward migration of the jet stream was accom- marked water balance differences, and hence to lake-level
panied by the storm tracks that had previously delivered response (e.g. Li & Morrill, 2010).
precipitation to the lakes in the American Southwest and Not all lakes are completely enclosed, however, or
this, in combination with higher levels of solar radiation respond solely to climate changes in terms of the local
(due to Milankovitch forcing) in the early Holocene, led evaporation–precipitation balance: a number of factors
M u c h wetter
Wetter
No change
Drier
M u c h drier
Figure 3.34 Lake-status data for 6 ka expressed as anomalies relative to long-term status. The site records are averaged for
grid squares to enable comparisons with outputs from numerical modelling experiments (after Sawada et al., 2004).
affect lake basins, such as tectonic or geomorphological (8–4 ka), and a decreasing moisture trend during the late
processes that alter the dimensions of a drainage basin or Holocene. Across this region, the pattern of Holocene
the height of the overflow (section 2.7.1). In south-central effective-moisture development appears to have been
Africa, for example, the Makgadikgadi Salt pans of northern determined largely by westerly (as opposed to southerly
Botswana are the desiccated remains of a former major monsoonal) airstreams, with a strong influence of North
inland lake system with well-preserved fossil shorelines, Atlantic sea-surface temperatures (Chen et al., 2008).
which have long been considered to be the result of Late Isotope variations in lake sediments (section 3.7.2)
Quaternary climate forcing, but which now seem more provide an additional index of past climatic variations that
likely to reflect tectonically initiated changes in regional can be set alongside, or used in combination with, climatic
drainage pattern as the underlying control over lake inferences obtained from lake-level changes. For example,
evolution (Moore et al., 2012). Furthermore, some lake Zhang et al. (2011) compiled carbonate δ18O records from
basins may respond to climate changes more quickly than eight lakes in the present-day Indian monsoon region,
others, reflecting perhaps the size of the basin, drainage each spanning the Holocene. The records are remarkably
sources, or other factors. In certain cases, this may lead to consistent in the long-term trends they portray (Figure
a delay in, or an asynchronous lake response to climate 3.36), indicating that climate changes during the Holocene
forcing. In an analysis of lake status records from the north were broadly synchronous across the monsoon region,
of the Tibetan Plateau, for example, Wünnemann (2007) with the wettest conditions during the early Holocene
found that while centennial- to millennial-scale changes in (11–7 ka) and progressive aridification since 7 ka. This
lake status during the last glacial stage were a direct response confirms the general pattern of Holocene climate change
to synchronous feedbacks of variations in the Asian inferred from other lake records (e.g. Figure 3.35). In the
monsoon, since the onset of the Asian summer monsoon Mediterranean region, a comparison of δ18O variations in
13 ka ago, many desert lakes have responded asyn- biogenic and endogenic carbonates recovered from lake
chronously with temporal and spatial shifts of some 100 sediment sequences showed that conditions were arid
years between short-term climate-induced variations in around much of the Mediterranean at times of cooling
moisture availability. in the North Atlantic, such as during Heinrich events
As with many other forms of proxy climate data, (section 3.10.1) and the Younger Dryas. In addition, the
therefore, lake-level records are perhaps best used alongside records suggest that changes in local rainfall were more
independent lines of sedimentological and biological important than temperature in driving the isotopic signal
evidence. Hence, in Figure 3.34, for example, the inferred in carbonate lake deposits in the eastern Mediterranean
drier conditions in the north-central USA are consistent region, but not in the west, reflecting a strong west–east
with independent pollen-based and numerical palaeo- contrast across the region during the Holocene (Roberts et
climatic simulations which indicate a band of positive al., 2008).
temperature anomalies across the continent at that time Much of the foregoing discussion has been on long-term
(Sawada et al., 2004). In central Asia, lake-level records were (millennial-scale) climatic changes, but it now appears
one of a number of environmental proxies that provided that lakes have also responded sensitively to shorter-term
the basis for a time-slice reconstruction of effective moisture climatic oscillations over decadal and centennial timescales.
changes across the monsoonal-affected areas over the past Although these typically involve very small-scale shifts in
13 ka (Figure 3.35). Although there is a considerable degree the pattern of pressure cells, particularly around the inter-
of complexity in the records, particularly during the tropical convergence zone, they can lead to marked changes
Bølling-Allerød and Younger Dryas periods, there is a in regional precipitation patterns, and may perhaps be
broader degree of consistency during the Holocene, with linked to short-term cycles of solar activity. For example,
wetter conditions reflecting a strong summer monsoon and peaks in the ~11-year sunspot cycle (section 7.6.4.1)
humid climate during the early and middle Holocene, but have been accompanied during the twentieth century by
an increasingly drier climate indicating a weakened summer water-level maxima in Lake Victoria, reflecting positive
monsoon during the late Holocene (Herzschuh, 2006). In rainfall anomalies ~1 year before solar maxima. Similar
a similar multi-proxy study in arid central Asia (the region patterns occurred in at least five other East African lakes
extending from the Caspian Sea eastwards across the suggesting that these sunspot–rainfall relationships were
northern Tibetan Plateau and into northeast China), a broadly regional in scale (Stager et al., 2007). Centennial-
different spatial and temporal pattern emerged from scale events in monsoonal climate have also been recog-
the lake records, with a dry early Holocene (pre-8 ka), nized in Asia (Wang et al., 1999) and it is these, rather than
maximum moisture conditions during the mid-Holocene Milankovitch-scale orbital forcing, that may be responsible
BBlling/Allered: 13.3 cal. k a (11,500 1 4

C-vears BP) Younger Drvas: 12.7 cal. ka (10,700 1 4
C-vears BP)
E80' E80' E80' 130° BP E80' irfjfr tar
Kazakhstan Kazakhstan
E80' Mongolia E80' ir» Mongolia

E80'
Tlbalan Plaleau
China China
Nil)- N30"
India India
NM: E80'
E80'
E80' LL2L E80' I EM' cc- E80'
Early Holocene: 10.9 cal. k a (9,500 1 4

C-years BP) Early m i d - H o l o c e n e : 6.8 cal. k a (6,000 1 4
C-years BP)
E80' IE M r •or E80' E80' E80' E80' E80'
.10=
Kazakhstan
Mongolia L-o-
H I Mongolia
E80' E80'
Tlbelan Plateau Tibetan Plalaau

China China
31 iridia
E80'
K.30-
India
E80'
'-an- J«L_ t30 5
E M' IM" !20"
E80'
Late m i d - H o l o c e n e : 3.8 c a l . k a (3.5 1 4

C-years BP) Present-day
ISM ~7TT m i jy E80' 190" '103' E80'
E80'
V
Kazakhstan
E80' Mongolia E80' Mongol la E80'

E80'
Tlbatan Plateau Tibetan Plateau
China Mongol la
E80' E80'
India India
NzHT.
E80' N3td
E80' E80' E80' 'EB0"

trial .
moderate moderate
wet dry
wet dry
Figure 3.35 Spatial patterns of inferred effective moisture for six time-slices between the Bølling-Allerød period and the present
time, based on lake-level variations and other palaeoclimatic records from Central Asia (after Herzschuh, 2006).
a) H e t o n g c h a n a Nur Donggi Cona

Hamaertai Lake Koucha Co
1.0
Xingyun Lake Selin Co
Qilu Lake
0.8 Qinghai Lake
0.6
0.4
0 2
0.0
Moisture index p
0 1 2 3 4 5 6 7 8 9 10 11 12
A g e ( C a l ka B P )
1.0
bj
Wet
0.8
0.6 Moderately wet
0.4
M o d e r a t e l y dry
0.2
Dry
0.0
0 1 2 3 4 5 6 7 8 9 10 11 12
A g e ( C a l ka B P )
Figure 3.36 a) Carbonate δ18O records from eight lake sediment sequences from the Indian monsoon region. Each isotope series
has been normalized to reduce them to a common scale (shown in probability units: p). b) A regional moisture index based on the
mean of the eight curves shown in a) (after Zhang et al., 2011).
for a weakening of monsoonal circulation over central and 3.8 CAVE SEDIMENTS AND
northern Australia during the late Holocene, the subsequent
trend to increasing aridity being reflected, inter alia, in
CARBONATE DEPOSITS
palaeo-lake records (Wyrwoll & Miller, 2001; Hesse et al.,
2004).
3.8.1 Introduction
Fluctuations in water level in closed-basin lakes, com- Caves form natural sediment traps in which the deposits
bined with isotopic studies and other proxy climate indica- are largely protected from the effects of subaerial weather-
tors, therefore, constitute a valuable means of investigating agencies and erosion (Gunn, 2003). In karstic regions
ing Late Quaternary climatic change. Most ‘pluvial’ lakes especially, the sediments that have survived in caves fre-
appear to respond rapidly to climate variations and are quently cover a much longer time interval than those on
often more direct indicators of water balance than vege- the neighbouring land surface. Three main types of material
tation or soils. Moreover, many pluvial lake deposits are contribute to cave sediment sequences: clastic detritus,
found in areas where biological and other indicators of organic detritus and precipitated carbonates, the relative
former environmental conditions are scarce. Not only proportions of each depending upon rock type, size of
do they provide direct evidence of former precipitation fissure, groundwater regime, topographical or geological
regimes, but they also offer a basis for modelling climatic context and geographical location (section 4.11.5.1). Clastic
change and for reconstructing atmospheric circulation detritus includes rock rubble, cave earth and water-lain
patterns in the low-latitude regions of the world. sediments. Organic detritus may consist of skeletal parts of
animals that occupied the caves and those of their prey, palaeoclimatic histories (Richards & Dorale, 2003); and (4)
while many caves acted as occupation sites for humans and offers a basis for correlation between terrestrial, ice-core
so cave successions may contain a rich legacy of organic, and ocean-core proxy climate records (Bar-Matthews
artefactual and cultural material of anthropogenic origin. et al., 2003). Reference will also be made to other types
In limestone regions, reprecipitated carbonates, collectively of carbonate material that are found outside the cave
known as speleothem, constitute a third important com- environment (e.g. travertine and tufa), since the processes
ponent of the sedimentary fill. Precipitation of carbonate of formation, and the analytical techniques employed in
from flowing or dripping water in caves can generate a their study, are often similar to those associated with
variety of forms of dripstone, the best known of which are speleothem (section 3.8.5).
stalactites and stalagmites, or flowstone. In many caves,
dripstone and flowstone production appears to have been
periodic or even cyclic, resulting in inter-stratification of
3.8.2 Detrital sediment in caves
speleothem and detrital layers, and artefacts, bones and A major factor governing the processes of sedimentation
clastic material can often become wholly embedded (and in caves is the shape of the cave itself. A distinction is often
very well preserved) within the precipitated carbonate made between exogene caves, which are shallow niches in
(Figure 3.37; Ford, 2006). the hillside, referred to by archaeologists and anthro-
In this section, the palaeoenvironmental potential of pologists as rock shelters, and endogene caves, which
detrital and carbonate deposits in caves will be examined. penetrate deep into the ground as chambers or passages
Of particular importance is speleothem, since this mater- (Figure 3.37). In the interior passages of endogene caves,
ial (1) provides the longest and most detailed palaeo- relatively equable conditions prevail, for the air is protected
environmental records from cave contexts (Fairchild & from the temperature extremes that occur at the ground
Baker, 2012); (2) contains stable isotopic signatures that surface. Daily and seasonal changes in weather and climate
form a basis for high-resolution palaeoclimatic recon- therefore rarely penetrate, at least not directly, and therefore
structions (Lauritzen & Lundberg, 1999); (3) can be dated only the major and long-term climatic changes can affect
by U-series dating (section 5.3.4) to provide a chronology the mode of sedimentation. Unless endogene passageways
both for the cave sedimentary sequences and for the derived are near the ground surface, in which case exotic material
Accumulation of bat
remains beneath roosting
area and accumulation of
small mammals beneath
owl roosting/nesting area
/ in the cave roof Water transported mud from
Breccias f u r t h e r i n s i d e the c a v e
formation under
a pitfall t r a p Waterlain Speleothem formation
silts stalactites, stalagmites & flow stone
Den Bo
nes
accumulation, /se
dim
ent
of bones s
T r a n s p o r t of: Bedrock Gravel

Sediments Loam 'Guano' R o o f fall; a c c u m u l a t i o n o f b o n e s o f
Bones C a v e earth Speleothem c a v e bear, d i e d d u r i n g h i b e r n a t i o n
Bones/sediments in t h e d e e p e r r e g i o n s o f t h e c a v e
F o r m e r l o w e r c h a m b e r of t h e c a v e that a c t e d a s a n a t u r a l t r a p / a c c u m u l a t i o n r o o m of a m a l g a m a t e d d e b r i s f l o w s
Figure 3.37 Generalized cross-section of a cave with various types of sedimentary infill and associated biological remains (modified
from http://historyofgeology.fieldofscience.com/2010_10_01_archive.html).
may be introduced into the system through clefts in the it is now widely recognized that cave sediment sequences
cave roof, the majority of sediments are derived from the constitute a potentially valuable archive of environmental
bedrock within which the fissures have formed, the prin- and climatic change (Woodward & Goldberg, 2001; White,
cipal exception being water-lain sediments from conduits 2007), while important environmental and archaeological
connected to the surface. evidence can be obtained from both macro- and micro-
In exogene caverns, and near the entrance to endo- morphological analyses of sediments in caves (Mallol et al.,
gene caves, sedimentation is more directly influenced by 2010). In a wider context, sedimentological analysis can
prevailing weather conditions outside. Moreover, as the provide a basis for correlating cave stratigraphies with
deposits are derived both from within the cave (autoch- the marine oxygen isotope record (Zhou et al., 2001).
thonous component) and from the surrounding area However, cave sediment sequences are normally complex
(allochthonous component), the stratigraphy in these and characterized by numerous hiatuses in sedimentation.
situations is often complex. As a rule, exogene caves will Furthermore, the facies variations found in cave sediments
tend to have a higher allochthonous component than are often not as well understood as those associated with
endogene caves. The entrances of exogene caves in formerly surface processes. The reconstruction of environmental
glaciated terrain may contain till or other glacigenic sedi- conditions from clastic cave sediments is therefore not
ments, while coastal caves may contain beach gravels or (in without its problems.
the inner reaches) finer marine sediments. Wind-blown A wide range of organic materials can be found in cave
material is also commonly encountered in exogene cave sediments, incorporated within or interstratified with clastic
sediments as well as colluvial and soliflucted sediment. material (e.g. Hearty et al., 2004). The most obvious organic
Autochthonous sediments consist principally of rock components are usually the skeletal remains of natural cave
rubble and cave earth. Angular fragments of rock are com- dwellers and their prey, and the analysis of such fossil
mon deposits in many caves and have been weathered assemblages can provide information on the nature and
from the roofs and walls to form thermoclastic scree. In the ecology of both predators and prey (Parmalee, 2005). Many
outer parts of caves, some rock fragments will have been caves contain rich snail assemblages, and in karstic regions,
derived from insolation weathering (expansion and con- these may often be of considerable antiquity (Goodfriend
traction at the rock surface due to temperature changes), & Mitterer, 1993). Pollen is often found in clastic cave
while in limestone areas, solutional weakening by percol- sediments (Carrión et al., 1999) and sometimes in speleo-
ating groundwaters will result in pieces of rock breaking thems (McGarry & Caseldine, 2004). Plant macroremains
off from the walls and roofs of a cave. Cave earth is com- may also be preserved in cave sediment sequences (Hansen,
posed of much finer materials (sand size and less) and may 2001), such materials having been transported into caves
have a variety of origins. Near the cave entrance, it is often either by natural processes, or by animals and people.
largely allochthonous, being composed of wind-blown or These include the decayed parts of plants carried in for
water-lain sand or silt, or even inwashed colluvial sediment. bedding, litter or food, plant macrofossils from animal
In the deeper parts of limestone caves, however, cave earths excreta and carcasses, and charcoal deposits from former
are formed either from the acid-insoluble residues left by hearths. In addition, pollen and other microfossils (e.g.
the solutional breakdown of the country rock, or from the phytoliths6), derived from local vegetation may be carried
secondary weathering of angular rock fragments that have in by animals and humans (Wallis, 2001). The fossil content
accumulated on the cave floor. Cave earths are frequently (biostratigraphy) of caves is discussed in Chapter 4.
red or brown in colour, due partly to the presence of Many cave sequences are depositional palimpsests with
oxides of iron and aluminium, but in some cases reflecting successive cultural horizons superimposed (Auban et al.,
the influence of phosphate derived from fossilized faecal 2001), and these may provide valuable insights into historic
matter. and prehistoric human activity, such as early agriculture
Until recently, relatively few detailed studies had been (Peña-Chocarro et al., 2005). Caves have also frequently
made of clastic sediments in caves, at least for the purposes been used in prehistory for stabling of animals, especially
of reconstructing Quaternary environments, since most in upland areas that provided rich pasture in the summer
interest has been focused on the fossil, archaeological or months. Evidence of the use of caves for this purpose can
speleothem content of cave sequences. Indeed, many caves be found in cave sediment layers that are rich in phosphates,
were excavated in the late nineteenth and early twentieth coprolites (animal excreta), charcoal and concentrations of
centuries primarily to recover archaeological remains, calcium oxalate crystals that are derived from coprolites
with scant regard for the stratigraphical context of this (Canti, 1998). These and other cultural horizons can readily
material. The situation has changed in recent decades, and be identified by soil micromorphological and chemical
analysis of the cave sediment sequence, and provide mentary record capable of analysis at a very high resolu-
evidence of, inter alia, transhumance in mountain regions tion (Fairchild & Baker, 2012). In some sites the growth
during the Neolithic and Bronze Ages (Karali et al., 2005). banding in speleothem may be annual in nature, while
Human coprolites have also been recovered from cave speleothem can also be dated using the U-series dating
sediments, sometimes in association with artefacts, and method (section 5.3.4) and, in some instances, by
provide a basis both for DNA analysis and also for dating means of radiocarbon (Yadava & Ramesh, 2005). Once a
the cultural contexts (Jenkins et al., 2012). chronology of speleothem growth has been established,
palaeoenvironmental inferences can be made on the basis
of (1) the rate and abundance of speleothem formation and
3.8.3 Speleothem (2) temporal variations in the isotope ratios contained
Speleothem is a secondary mineral deposit formed in caves within the carbonate structure. These aspects will now be
in karst regions as dripstones or flowstones. Dripstones discussed in more detail.
are deposits of calcium carbonate (although some may be
composed of varying quantities of aragonite, gypsum or
3.8.4 Speleothem growth and
halite) formed by water dripping from the ceilings or walls
of a cave, or from the overhanging edge of a rock shelter.
environmental reconstruction
The most common features that develop in this way are Precise dating of speleothem has demonstrated that the
stalactites and stalagmites. Flowstones are deposits of development of carbonate growth bands in cave systems
calcium carbonate, gypsum or other mineral matter that may be cyclic or intermittent. This largely reflects envir-
have accumulated on the walls or floors of caves in places onmental controls on cave hydrology, and so an analysis
where water trickles or flows over the rock. Upon reaching of the growth frequency of speleothem within cave systems
the cave floor, water may percolate into the interstices of can be used to reconstruct a number of different aspects
clastic sediments, cementing them into a coherent, often of former environmental conditions, as the following
very hard porous rock known as cave breccia. In some cases, examples demonstrate.
a complete cover of precipitated calcium carbonate blankets
the floor of the cave, where it may become interbedded with
3.8.4.1 Speleothem growth and climatic
the screes, breccias and cave earths (Figure 3.37). Such a
change
flowstone cover has been referred to as a stalagmite floor.
Other precipitated calcium carbonate deposits occasionally Carbonate precipitation is strongly influenced by prevail-
found in caves in karst regions include travertine, a light, ing climatic conditions. It is either strongly reduced or
compact and generally concretionary substance, extremely arrested during cold episodes and increases to a maximum
porous or cellular varieties of which are known as cal- during warm intervals. In permafrost environments,
careous tufa, calcareous sinter or spring deposit. Compact groundwater percolation is at a minimum, while CO2 con-
banded varieties are sometimes referred to as ‘cave marble’ centrations in these groundwaters are also reduced because
or ‘cave onyx’. Many of these carbonate features are also of the restricted biogenic activity associated with skeletal
common in subaerial environments in karstic regions, periglacial soils. There may, however, be some limited
around streams and springs (section 3.8.5). or sporadic speleothem growth, depending upon local
The precipitation of calcium carbonate is caused by conditions. In Norway, for example, caves that were
degassing of CO2. Water that drips or flows into the cave overlain by ice were flooded by meltwater, which resulted
chamber has usually originated on the surface where it has in phreatic7 conditions and the deposition of clastic
acquired a high concentration of CO2 from biogenic sediments rather than speleothem carbonate (Lauritzen,
production in soils. Due to ventilation effects, the air in 1993), whereas in Austria there are indications that
the cave chamber has a lower partial pressure of CO2 than speleothems were forming in some alpine caves when these
the incoming water and so automatic degassing of CO2 were covered by glacier ice (Spötl & Magnini, 2007). During
from the water takes place, leading to supersaturation and interglacial conditions, by contrast, higher precipita-
then precipitation of carbonate. With a continuous supply tion levels result in greater water penetration, biological
of water over long periods, large speleothem structures can productivity increases, vadose (see note 7) conditions
build up by successive growth layers of carbonate (Figure prevail in cave systems, and maximum speleothem forma-
3.38). It is this characteristic that makes speleothems tion occurs. In mid- and high-latitude regions, therefore,
so valuable for the reconstruction of Quaternary palaeo- speleothem formation reflects the sequence of climatic
environments, for they often preserve a continuous sedi- conditions experienced during the course of a glacial–
30
30 a) 2005 AD 70
60
50
40
30
20
Latitude
1 2 3 4 5 6
10
-10
-20
244 ± 10 -30
-40
-50
Micro-milling
transect 0 50 100 150 200
A g e (ka)
353 ± 30
Figure 3.39 Compilation of c. 750 TIMS U-series speleothem

dates plotted against the latitude of the relevant site. MIS
stages are also shown. Note the virtual absence of dates from
512 ± 18
the mid- and high-latitude regions of the Northern Hemisphere
during MIS 2 (after McDermott, 2004).
interglacial cycle. This pattern is clearly reflected in Figure

3.39, which shows a compilation of approximately 750
U-series age (yr)
TIMS U-series dates (section 5.3.4.2) on speleothems that
a) Micro-milling transect
have been studied over the course of the last decade, plotted
2004 AD
against the latitude of the relevant cave site. The absence
1984 A D of speleothem deposition in the mid- and high latitudes
of the Northern Hemisphere during MIS 2 is striking.
By contrast, speleothem deposition appears to have been
essentially continuous throughout the glacial period at
40mm
lower latitudes in the Northern Hemisphere (McDermott,

2004).
The correlation between interglacial intervals and
1935 AD periods of enhanced speleothem growth is evident at many
1905 AD
sites around the world. In the Austrian Alps, for example,
speleothems grew from around 127 ka, with peak inter-
glacial conditions (high δ18O values and growth rates)
Figure 3.38 a) Cross-sectioned surface of a stalagmite from Akçakale Cave, northeast Turkey revealing annual growth layers
spanning the last c. 500 years, the chronology confirmed by U-series dating. Small samples were extracted (micro-milled) at 0.1–1.0
mm sampling intervals along transects (light shading in the figure) where series of contiguous growth layers could be traced (Jex
et al., 2010). Stable isotope measurements obtained from these samples, cross-checked with modern instrumental records, provide
a record of annual variations in local precipitation and have been successfully forward-modelled using GCM data for this region
coupled with a karst hydrology model (Jex et al., 2011). b) Cross-sectioned surface of a small stalagmite from Rukiessa Cave,
southeast Ethiopia, spanning the last century. Seasonal layers are indicated by colour variations that reflect impurities within the
drip waters in the cave. Samples 1–3 mm in thickness were drilled along a continuous transect (light shading) that enabled seasonal
climatic variations and land-use changes to be reconstructed (Baker et al., 2007; Blyth et al., 2007) (photographs by Andy Baker,
Catherine Jex and Asfawossen Asrat, University of New South Wales, Sydney, Australia, reprinted with permission from Elsevier).
between 127 and 124 ka. Speleothem growth terminated at ture in the vicinity of the site. Calcite deposition occurs by
c. 114 ka, which is in in line with widespread ice rafting in degassing from carbon-saturated dripwaters that enter the
the North Atlantic and the onset of full stadial conditions cave from the surface and the speleothem carbonate tends
in MIS 5d (Meyer et al., 2008). Further north at Rana, north- to form in isotopic equilibrium8 with the water from which
ern Norway, stalagmite deposition has been dated to between it is precipitated. The isotopic composition of percolating
c. 123 and 73 ka, in other words spanning much of MIS 5. water will be registered in the isotopic ratios locked into the
The most rapid growth phase (~46 μm yr–1) occurred carbonate crystal or in fluid inclusions trapped within it.
between 123 and 119.5 ka, but had declined to ~0.7 μm yr–1 By measuring the ratio of heavy (18O) to light (16O) oxygen
by c. 107 ka. Between c.107 and 73 ka, growth is barely notice- isotopes (measured as a deviation, δ, from a standard:
able. The transition between rapid and slow growth rate section 3.10.2) at regular intervals along the axis of speleo-
between c. 119 and 107 ka reflects the termination of inter- them growth bands, a record of isotope variations over time
glacial climate in the region (Linge et al., 2001). In southern can be obtained. Such data have shown that the oxygen
France, most of the major growth phases in stalagmite in the isotope ratios in cave speleothem often vary in a cyclic
Grotte de Clamouse correspond with warmer intervals fashion (Figure 3.40) and that they contain a climatic
corresponding to MIS 5 and 7 (Plagnes et al., 2002). signal (see below). Dating of these records by the uranium-
Similar trends are apparent in evidence from the Southern series method allows direct correlation both with ice-core
Hemisphere. In Australia, for example, data from Tasmania and marine oxygen isotope profiles, and also with other
indicate peak last interglacial conditions between c. 129 and records of global climatic change (McDermott, 2004).
122 ka, which also coincides with a time of prolific coral Not all speleothem is suitable for this type of research,
growth in western Australia (Zhao et al., 2001), while in New however, for a number of conditions must be satisfied
Zealand, U-series dating of almost 150 speleothems shows before the isotopic ratios can be regarded as providing
that calcite deposition has taken place somewhere in the reliable palaeoclimatic data. These include: (1) the carbon-
country with little interruption for more than 500 ka, and ate precipitate must be in isotopic equilibrium throughout
that most growth also occurred during interglacial and the period represented; (2) the oxygen isotope content of
interstadial episodes (Williams et al., 2010). the water from which the carbonate was precipitated must
Speleothem records may also reveal periods of more be known; (3) no diagenetic alteration should have
arid conditions. In caves in middle- to low-latitude regions, occurred; and (4) the deposit must be sufficiently free of
where variations in groundwater are strongly influenced detrital contamination for satisfactory dating of the
by regional changes in precipitation regime, speleothem sequence. Given the complexity of cave sediment sequences,
growth is often episodic. A change to drier climatic con- it is not always easy to satisfy all four of these conditions.
ditions will lead to a reduction in the amount of water Perhaps the most difficult is (2), for isotope variations in
percolating from the surface, while soil biogenic activity will cave carbonate can be the result of the combined influences
also be reduced. In many caves in southern Australia, for of temperature, the fractionation9 which occurs when
example, speleothem deposition over the past 500 ka has carbonate precipitates from dripwater (a decrease in cave
been intermittent and concentrated during cooler and temperature will lead to higher 18O concentration in the
wetter stadials and interstadials of the past four glacial cycles carbonate precipitate), and changes in the isotopic ratios
(Aycliffe et al., 1998; Desmarchelier et al., 2000). For the of dripwater before entry to the cave system. The extent of
Holocene, speleothem records indicate that, following an fractionation during carbonate precipitation can be
early phase of more effective precipitation, with peak estimated by analysing variations in other isotopes, such as
moisture at 7–6 ka, an arid climate similar to present carbon and hydrogen, since these are unaffected during
was established by 5 ka, reflecting the onset of an ENSO- crystal formation and they therefore enable the pre-
like climate (section 7.6.4.2) across arid Australia (Quigley fractionation ratios of oxygen isotopes in dripwater to be
et al., 2010). determined (Lauritzen, 1993).
As temperature is a major factor influencing temporal
variations in δ18O in cave speleothem, and numerous
3.8.4.2 Stable isotope ratios in cave
studies have been published in which quantitative estimates
speleothem
of past temperature have been derived from the oxygen iso-
In deep caves, where there is no direct contact with the tope signal in speleothem calcite (e.g. Mangini et al., 2005;
external atmosphere and there is little or no air circulation, Affek et al., 2008). However, the temperature-dependence
the temperature remains more or less constant throughout of δ18O in rainfall is temporally and spatially variable,
the year and is similar to the mean annual air tempera- and is frequently site-dependent (Fricke & O’Neill, 1999).
-9
•8
-7
ft O (%o PDB)
-e
-5
te
-3
-2
150 150 100 50 0
Age (ka)
Figure 3.40 Composite δ18O curve constructed from twenty-one overlapping speleothem records for the past 185 ka from Soreq
Cave, Israel. The black circles at the top show the positions of samples dated by TIMS U-series (section 5.3.4.2) (after Ayalon
et al., 2002).
In addition, on centennial to millennial timescales, a range widely recognized 8.2 ka event (section 7.6.3.2), match very
of factors other than temperature may lead to temporal closely variations in the timberline in the eastern Austrian
variations in δ18O. These include changes in the δ18O of the Alps (Figure 3.41b). In addition to reconstructing tempera-
ocean surface due to fluctuations in continental ice volume ture trends, variations in oxygen isotope ratios can provide
(section 3.10); temporal variations in temperature differ- valuable chronological control on the timing of major
ence between the sea-surface water vapour source area and shifts in δ18O ratios in precipitation, which can then be
the cave site of interest; long-term changes in moisture interpreted in terms of changes in atmospheric circulation
sources or storm tracks; changes in the proportion of patterns, changes in the δ18O of ocean vapour sources,
precipitation that has been derived from non-oceanic or first order climate changes, such as the Dansgaard–
sources, in other words obtained from rivers, lakes and Oeschger oscillations (section 3.11) of the last glacial cycle
other continental sources; and also the seasonally varying (McDermott, 2004). For example, a δ18O record from
isotopic composition of precipitation (McDermott et al., Hulu Cave, China, shows a remarkable resemblance to
1999; Fairchild & McMillan, 2007). the oxygen isotope record from the GISP2 Greenland
These and other interpretational difficulties have led, ice core, suggesting that East Asian monsoon intensity
in more recent years, to a move away from quantitative changed in concert with Greenland temperatures between
temperature estimates to more qualitative climatic recon- 75 and 11 ka (Figure 3.42). This record links North Atlantic
structions, in which δ18O shifts are interpreted in terms of climate with the meridional transport of heat and moisture
warming and cooling trends (e.g. Onac et al., 2002; Williams from the warmest part of the ocean where the summer East
et al., 2004a). Figure 3.41a shows a δ18O record from a Asian monsoon originates (Wang et al., 2001).
stalagmite in Hölloch Cave in the Bavarian Alps, Germany. In addition to δ18O values, many recent studies have
The sequence spans the period from around 14 ka to the included the analysis of carbon isotope ratios (δ13C) in
present and shows the climatic oscillation of the Lateglacial, speleothem carbonate. The δ13C record in the speleothem
with the marked cooling in the Younger Dryas cold interval is a reflection of the nature of the overlying vegetation
(c. 13–11.5 ka) and the abrupt warming at the beginning cover, albeit modified by fractionation and other pro-
of the Holocene (Wurth et al., 2004). The marked cesses and, in some instances, this can be interpreted as a
oscillations in climate during the Holocene, including the palaeoclimatic record. In semi-arid and arid regions,
b)
Goschenen 2
Gaschenen 1
Little Ice Age

Rot moos 1
Rotmoos 2
Schams
Lobben
Mi sox
modern potential
Palu
Timberline [Am];
natural position
deviation from
Warm + 100
-100
Cold -zoo
a) 1
deviation from mean value
0
6 0 [ M » PDB]
-1
, (
8,200yr BP
-2
Younger Dry as
Su bat I antic
Sub boreal
Pre boreal
Allerad
Atlantic
Belling
Boreal
Pleniglacial
8,200yr BP 8,200yr BP
14,000 12,000 10,000 8,000 6,000 4,000 2,000 0
A g e (cal, yr BP)
Figure 3.41 a) δ18O measurements covering the period 14 ka to the present from Hölloch Cave, Bavarian Alps, Germany. Note
the significant isotopic shift (to cooler conditions) between c. 13 and 11.5 ka (Younger Dryas) and the fluctuating nature of the
Holocene climatic record. b) Timberline record from the Eastern Alps, Austria. Note the correlation between decline in timberline
elevation (implying cooler climates) and the shifts to heavier values in the isotopic curve (after Wurth et al., 2004).
significant shifts in δ13C per mil values in speleothems may record is often less straightforward, although a correlation
be reflective of climate-driven changes in vegetation (i.e. has been suggested between δ13C and changes in soil
between C3 and C4 plant types ), with ‘lighter’ δ13C per mil CO2 production, with higher δ13C values reflecting reduced
values being associated with C3 plants and ‘heavier’ values soil CO2 as, for example, during the Younger Dryas cold
with C4 vegetation (Bar-Matthews et al., 1997). In eastern phase (Támas et al., 2005). Another potentially fruitful
China, for example, Zhand et al. (2004) have interpreted area of research concerns hydrogen ratios (δD) in fluid
δ13C shifts in a 6,000-year Holocene stalagmite as a climatic inclusions that were trapped within the cave speleothem at
proxy record, the lighter δ13C values reflecting the the time of their formation. The δ18O ratios of such
dominance of C3 plants under relatively warm or rainy inclusions may be ambiguous, because of the possibility of
conditions, while heavier δ13C values are associated with C4 post-depositional changes between oxygen isotopes and the
plants that thrive under cooler, arid conditions. In parent carbon material. Significantly, there is no post-
temperate regions, however, the interpretation of the δ13C depositional hydrogen isotope exchange. The correlation
10000 20000 30000 40000 50000 60000 70000
-10
12 13
11 13 16 17
12 14 21
4 5 8 10 11 480
10 15 20
0 ( V VPDB)
Insolation (W/m )
5
2
9
-8 4 19
9 18
4 Das
2
4
1 3
-6 MSD MSL
6
PD
440
•A-
YD H1 H2 H3 H4 H5 H6
Warmer
0 ( V VSMOW)
19 20
-36¬ 3 12 14
5 8 1617 21
4 10 11 15
2 4 7 13 18
9
-40-
Colder
B a a
l a
6
10000 20000 30000 40000 50000 60000 70000
A g e (ka)
Figure 3.42 δ18O record from Hulu Cave, China (upper) and the δ18O trace from the GISP2 Greenland ice core. The Younger
Dryas and Heinrich events (episodes of ice rafting reflecting cooler conditions in the North Atlantic: section 3.10) are shown by
the vertical bars (after Wang et al., 2001).
between D/H ratios and δ18O is well known, and thus (Fairchild, 2009). For example, at Soreq Cave, Israel, a
where δD and δ18O are measured on the same fluid marked decline in concentrations of strontium, barium
inclusions, D/H can be used as a proxy for δ18O. Because of and uranium, and in 87Sr/86Sr and 234U/238U ratios, that
the known temperature relationship between meteoric are recorded in cave speleothem, began around 17 ka and
water (rainwater) and hydrogen isotopes, there is a basis for reached a minimum around 9 ka. This trend coincides
deriving a quantified temperature record from the δ18O trace with a climatic shift from a more arid to a wetter climatic
in the speleothem (Genty et al., 2003). Using this approach regime. During the initial dry phase, the higher trace
on speleothems from three caves in Israel, McGarry et al. element concentrations are interpreted as reflecting an
(2004) obtained a temperature range for the last interglacial increase in the contribution of salts derived from exogenic
(c. 130–120 ka) of 17–22°C and temperatures of the order sources (sea spray and aeolian dust) along with reduced
of 10°C below modern values (18°C) for the Last Glacial weathering of local bedrock, while the subsequent lower
Maximum. Further developments along these lines would concentrations are indicative of a reduction in exogenous
not only enable unambiguous estimates to be derived from material and enhanced weathering of the local dolomitic
speleothem stable isotope records, but would also allow rocks (Ayalon et al., 1999). In the Grotte de Clamouse,
realistic estimates to be made of the uncertainties associated southern France, McMillan et al. (2005) showed that trace
with reconstructed climatic parameters. element Mg/Ca and Sr/Ca ratios in speleothem carbonate
are enhanced during seasonally dry periods, and this led
them to infer a multi-decadal period of aridity between
3.8.4.3 Trace elements in cave speleothem c. 1.25 and 1.15 ka (AD 750–850). Interestingly, a compar-
In addition to stable isotopes, palaeoclimatic and palaeo- able arid phase based on lake-level data (section 3.7) has
environmental records can be derived from trace elements been suggested for this time interval in Ethiopia, Niger and
(e.g. Mg, Sr and Ba) contained within cave speleothem Mexico (Street-Perrott et al., 2000).
3.8.4.4 Speleothem formation and sea-level the island after MIS 5a. The overall differential displace-
variations ment between the north and south of the east coast is
~1.5 m and, assuming that tilting has been continuous
Submarine cave systems in coastal karst areas often con- over the past 85 ka, this suggests an average rate of tilting
tain speleothem structures, indicating that sea level of 0.02 mm yr–1 (Fornós et al., 2002).
must have been below the level of the caves at the time
the speleothem formed. Dating the onset and cessation
of speleothem growth throughout the cave systems has 3.8.4.6 Speleothem formation and rates of
enabled the amplitude, rate and timing of global sea level denudation
to be reconstructed. In some instances, the age of low Dating of speleothem in mountain regions suggests that
sea-level stands can be determined by radiocarbon dating the oldest material is frequently found in the highest
organic inclusions in the speleothem (Smart, 2003), but caves, while the onset of speleothem formation is pro-
in the majority of cases, long-term records of sea-level gressively later in caves at lower altitudes. This may reflect
change are based on U-series dating of speleothem car- long-term lowering of the water table through valley
bonate from the currently submerged caves. This approach incision, which results in the higher caves being perched
has been used to develop a record of sea-level change above the vadose zone so that carbonate precipitation will
during MIS 5a on the Adriatic coast of Croatia (Surić then cease (Atkinson & Rowe, 1992). As the water table falls,
et al., 2009), while a 215 ka history of sea-level oscillations so new caves are formed or come within the vadose zone,
in the Mediterranean region has been constructed from and speleothem formation is initiated. Dating speleothems
the dating of marine and continental layers in speleothems in different cave levels therefore enables rates of denudation
from the presently submerged Argentarola Cave in Italy to be inferred and may provide valuable insights into rates
(Antonioli et al., 2004). These data are valuable in that they and processes of bedrock incision (Fei et al., 2004). In
provide independent tests of the models of global ice northwest Scotland, for example, U-series dates on speleo-
volume based on oxygen isotope ratios from marine thems, in combination with geomorphological and other
microfossils (section 3.10), since the changes in sea level data from cave surveys, suggests that, over the course of the
inferred from the speleothem evidence should match past 300 ka, the mean rate of valley deepening lies between
the ice-volume changes deduced from the marine isotope 47 and 68 m per glacial–interglacial cycle. This indicates a
records. glacial erosion rate of about 2 mm a–1, a figure that is in
agreement with modern process measurements (Hebdon
et al., 1997).
3.8.4.5 Speleothem formation and tectonic
activity
3.8.5 Other carbonate deposits
Speleothem evidence also provides useful information on
local and regional tectonic histories. In the Bahamian Other forms of carbonate accumulate in or around sur-
archipelago, for example, while speleothem growth is face streams and springs in karstic regions. Some, such as
abundant in many submerged caves, it is less well developed travertine (carbonate coatings on bedrock), are produced
in caves above sea level, being confined to a very narrow in much the same manner as flowstone in caves (Pentecost,
zone between c. +1 and +7 m. The restricted elevations 2005). Groundwater saturated in CaCO3 undergoes CO2
of emerged caves, and the fact that speleothem older than degassing where it emerges into the open and carbonate is
100 ka is absent within them, led Carew & Mylroie (1995) precipitated at the surface. In arid and semi-arid regions,
to conclude that the Bahamian banks have been tectonically the rate of production of CO2-charged water depends upon
stable throughout the Late Quaternary. By implication, local soil and microbial activity, and so the formation of
therefore, it should be possible to derive a Late Quatern- travertine may reflect times of increased biological produc-
ary eustatic sea-level curve from Bahamian shoreline tivity associated, perhaps, with wetter climatic conditions.
evidence (section 2.5.2). On the east coast of Mallorca, In some semi-arid regions, isolated patches of travertine
caves partly filled with brackish water contain phreatic over- may mark the levels at which higher, perched water tables
growths on speleothems which mark sea-level positions developed during wetter periods. Dating of such features
during the Late Quaternary. The altitudinal positions of in the Grand Canyon, Arizona, for example, suggests that
phreatic speleothems that formed during high sea-stands wetter episodes occurred at c. 338, 171, 71 and 15 ka, dates
of MIS 5a, 5c and 5e are found at increasing elevations that correspond with those obtained for high lake-level
northwards and reflect significant southward tilting of stands in the region (Szabo, 1990). Travertine precipitation
forms and can merge imperceptibly with other carbonate
Pacui Junction
Jose Gregorio
deposits, such as travertine, calcareous cements and strom-
Salgadinho
Salgadinho
Salgadinho
Salgadinho
atolites (see below). Tufas commonly mark the margins of
high lake stands in arid and semi-arid regions (Figure
3.44), and the chronology of lake-level variations in such
areas is frequently based upon the U-series dating of tufa
11 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 22 deposits. Oxygen isotope ratios in freshwater tufa also
A g e (ka) appear to be a sensitive indicator of past climatic changes
Brejoes La go a (Andrews & Brasier, 2005). Tufas are more difficult to date
Da Gruta Branca by U-series than most speleothems because of problems
with detrital contamination (Garnett et al., 2004).
300 400 500 600 700 300 900 1000 A third category of carbonate growths is stromatolites,
A g e (ka)
which are carbonate encrustations around algae, bacteria
Figure 3.43 Growth phases of 230Th-dated portions of or other organisms. These can take a wide variety of forms,
travertine from sites in semi-arid northeastern Brazil extending depending upon the range of organisms present, the
over more than 600 ka. Note that the four phases of growth chemical composition of the water and the nature of the
between 22 and 11 ka reflect periods of increased rainfall (after climatic regime. Fossil stromatolites constitute the most
Auler et al., 2004).
ancient records of life on earth, with some remains dating
to over 3.5 ba. In the contemporary littoral environment,
from hyper-alkaline springs in Northern Oman indicates they commonly accumulate to form reefs at the edges of
wet conditions prior to 19 ka, a period of extreme aridity lakes or in the intertidal zone where they are often found
between 16.3 and 13 ka, a pluvial episode between 12.5 and in association with travertines (Valero-Garcés et al., 2001).
6.5 ka, and a late Holocene phase of hyper-aridity which As algal forms, they are reliant upon sunlight in the upper
continues to the present day (Clark & Fontes, 1990). Again, water layer, and they can therefore be used as proxies
this is very similar to the climatic sequence obtained from for water level changes in lakes; their aragonite mineralogy
lake-level records. In semi-arid northeastern Brazil, dating also means that they can be dated by U-series. As such,
of travertines provides a record of pluvial activity extending they constitute excellent proxies for reconstructing Late
back 900 ka (Figure 3.43), with the present dry conditions Quaternary lake-level sequences (Lisker et al., 2009).
in the region being established at the onset of the present Finally, carbonate deposition is also an important
interglacial (Auler et al., 2004). pedogenic process, particularly in arid and semi-arid
Tufa is a calcareous crust that develops around lake regions where pedogenic carbonate is deposited as nodules,
margins, stream edges or springs. It can take a variety of root casings, pebble coatings and irregular concretions
(Candy & Black, 2009). Precipitation is largely of inorganic
carbonate, but biomineralization of carbonate in soils by
soil fungi and other organisms also occurs. The analysis of
pedogenic carbonate may provide evidence of climatic
change, ground surface stability and substrate age, for the
amount, type and depth of carbonate production in soils
is climatically determined. In the western United States,
pedogenic carbonate on alluvial gravels has been dated
by U-series to provide a chronology for fluvial terrace
sequences (Sharp et al., 2003), while in western Europe,
analysis of stable carbon and oxygen isotope ratios
preserved in pedogenic needle fibre calcite (a particular
form of pedogenic carbonate) at sites in France, Spain and
the UK revealed a good correlation between mean monthly
climatic parameters (including air temperature, frost-
Figure 3.44 Tufa formations around the shore of Mono Lake, free days, humidity and precipitation), suggesting that
California. Many of these tufas occur on shorelines that are well
above the level of the present-day lake and reflect highstands
stable isotope compositions preserved in carbonate in soil
of formerly more extensive pluvial lakes (photograph by Mike horizons may be a potential tool for palaeoenvironmental
Walker). reconstructions (Millière et al., 2011).
3.9 LAKE, MIRE AND BOG the many hundreds of small lakes that typically form in
glaciated regions. They include kettle lakes (which develop
SEDIMENTS in hollows created by the melting of buried ice) and lakes
dammed behind or between glacial landforms (e.g. terminal
3.9.1 Introduction moraines or drumlins) as a result of the blocking of
Preserved within sediments that have accumulated in lakes, drainage outlets. In some respects the term is ambiguous,
mires and peat bogs is a diverse and often detailed record however, as it can also be used to refer to lakes that form
of environmental change. Given sufficient time, all lakes on, or within, glacier ice (section 3.3) or to ice-marginal
become infilled with sediment to form mires and bogs. lakes where glaciers advance to block a drainage outlet.
Thus lake, mire and bog sediments are genetically related Other types of lake listed in Table 3.3 tend to be more
and frequently grade into one another. For this reason, the ephemeral or to contain sediment records that span only
sediments are considered under a single heading. Lake, mire short periods of time.
and bog deposits are important in a number of respects. In the glaciated regions of Europe and North America,
First, the contained fossil flora and fauna provide evidence present-day lakes are no older than c. 15 ka, as they formed
of both local and regional ecological changes. Second, the following the retreat of the last ice sheets, and hence the
character of the sediments offers clues about former sediments in mires and bogs in these regions are usually
environmental conditions. This is particularly true in the
case of lake sediments where variations in the physical
and chemical properties reflect developments in the lake Table 3.3 Classification of lakes by mode of formation
ecosystem, and also changes in the rates at which pro- according to Håkansson & Jansson, 1983.
cesses operated around the lake catchment. In both cases,
Type of lake Mode of formation
the observed variations may be interpreted in terms of
Tectonic lakes a. Epeirogenic movements
environmental change. Third, fossil lacustrine sediments
(e.g. Caspian Sea);
and associated shoreline features often reveal a record of
b. tilting, folding or warping
fluctuations in lake levels in response to climatic changes
(e.g. East African rift system)
during the later part of the Quaternary. This topic is
discussed in sections 2.7.1 and 3.7 in relation to lakes in arid Volcanic lakes a. Maars, calderas and crater lakes
(e.g. Crater Lake, Oregon);
and semi-arid regions, where prolonged episodes of
drought frequently result in desiccation of the lakes. Here, b. by damming of drainage by lava
or volcanic debris (e.g. Lake Kivu,
however, we focus on lake sediments accumulating in the
Sea of Galilee)
humid regions of the world, where drought is less of a
hazard and sediment supply tends to be more continuous. Landslide lakes Rockslides, mudflows and screes
Lakes can have a variety of origins. In a classic Glacial lakes Large variety of lakes formed in
contribution published more than thirty years ago, glacigenic sediments by ice-melting
(‘kettle’ lakes) or through impeded
Håkansson and Jansson (1983) recognized eleven main
drainage behind or between glacigenic
types of lake (Table 3.3) of which three – tectonic, volcanic
landforms (e.g. moraine, drumlins)
and ‘glacial’ – are of most interest to the Quaternary
Solution lakes Ground solution of limestone, other
scientist, since it is within these that the most detailed
calcareous rocks, gypsum, rock salt
sedimentary records have been found. Tectonic lakes form
Fluvial lakes Plunge pools, delta lakes, meander
in areas of subsidence caused by folding or faulting, and
lakes (oxbows, levées)
include some of the largest lakes in the world, such as Lakes
Baikal (Siberia), Titicaca (Bolivia), Tanganyika and Victoria Aeolian lakes Deflation basins; lakes dammed by
wind-blown sediment
(East Africa) and Biwa (Japan), and the Black, Dead and
Caspian ‘Seas’. Some of these contain sediment sequences Shoreline lakes Damming by material transported by
longshore drift; tombolas, spit-lakes
that span the whole of Quaternary time. Volcanic lakes
(or maars) occur in calderas and craters, the length of the Organic lakes Blocking by vegetation, beaver dams
sediment record depending upon the time at which volcanic (‘phytogenic’ lakes); coral lakes
activity ceased. In Europe, a number of the maars in the Anthropogenic Dams and excavation fills (e.g. Lake
French Massif Central, in central Italy and in Greece contain lakes Mead, Arizona)
sediments that extend over several glacial–interglacial cycles Meteorite lakes Water accumulating in meteorite
(Reille et al., 2000; Tzedakis et al., 2003). ‘Glacial lakes’ are craters
considerably younger. In those areas that lay beyond the ice that are commonly found in lake and bog sediments are
sheets, however, sediments continued to accumulate in lake discussed in Chapter 4.
basins throughout the last cold stage, and these sites often In this section we are concerned with those lithological
contain records extending back into the last interglacial and characteristics of lake, mire and bog deposits that can
beyond. Coring of such open-water sites is technically be used as a basis for palaeoenvironmental reconstruction.
difficult and relatively expensive, and it is only in the The emphasis is placed on sediments that have accumulated
last two decades, with the development of sophisticated during the Late Quaternary (particularly during the
hydraulic coring equipment and the formation of inter- Lateglacial and Holocene periods) in North America and
national research teams to provide the necessary logistical northwest Europe, for these are often relatively accessible
and financial support, that long records have begun to and can be sampled either in section or with hand-operated
be obtained from these deep lake basins (e.g. Brauer & corers. As a consequence, they are frequently known in
Negendank, 2004; Nakagawa et al., 2012). much greater detail than older deposits.
The importance of these long lake sequences is that they
provide a continuous record of changes in ecosystems, lake
3.9.2 The nature of lake and bog
sediment processes and climate, sometimes over several
glacial–interglacial cycles. Hence they offer a means of
sediments
correlating terrestrial records with those obtained from Lake sediments are both allochthonous and autochthonous
deep-ocean sequences and from ice cores (section 6.3.3). in origin, being derived partly from organic production
In addition to these long lake records, fragmentary remains within the lake ecosystem, and partly from the inwash
of lake and peat deposits dating to before the Last Glacial of both organic and inorganic material from around the
Maximum are also found in many localities. In higher lake catchment (Cohen, 2003). If the lake is rich in mineral
latitudes these are often intercalated between tills and other nutrients, organic productivity will be high and the con-
terrestrial sediments (Figure 3.45) and usually reflect limnic ditions are described as eutrophic. The typical deposit will
accumulation during interglacials and interstadials or, in be a green-brown organic-rich sediment known as nekron
areas dominated by coarse-grained sediments that are mud or by the Swedish name gyttja. In deeper waters, this
free-draining, during intermittent periods of higher will be extremely fine in texture and will consist of
water tables when temporary lakes or mires developed. comminuted and largely unrecognizable plant material,
Although they are useful in stratigraphic subdivision, the but will grade into detritus gyttja with recognizable plant
palaeoenvironmental significance of these deposits usually macrofossils (fruits, seeds, leaves, etc.) in shallow waters.
lies more in the fossils they contain than in the nature of Where the lake substrate is calcareous, lime may be
the sediments themselves. The faunal and floral remains precipitated from the water by aquatic plants (e.g. by some
of the pondweeds, such as Potamogeton, and algae such as
Chara) and other organisms, and a fine cream-white clay-
rich sediment (marl) will accumulate. In general, sediments
deposited under eutrophic conditions are predominantly
autochthonous. Where the lake is poor in nutrients and
organic productivity is low (oligotrophic), allochthonous
sediments will often predominate. If the inwashed materials
are low in organic content, clastic sediments (sands, silts
and clays) will dominate, the finer grades of sediment
being encountered in deeper waters. Where organic pro-
ductivity is low, but the inwashed materials are dominated
by humic substances from, for example, peats around the
lake catchment, the lake waters are typically brown in
colour due to the dissolved humic acids. In some lakes,
Figure 3.45 Middle Pleistocene (Cromerian Complex) organic a dark-brown gel-mud composed largely of colloidal
deposits (the ‘Pakefield Rootlet bed’) exposed beneath till at precipitates will accumulate. Such conditions are often
Pakefield, Suffolk, UK (photograph by Mike Walker). The described as dystrophic and the deposit is known by the
deposits contain a diverse range of plant and animal fossils, and
date to around 700 ka. Flint artefacts recovered from the site
Swedish term dy. Finally, where the lake waters support a
represent some of the earliest evidence for human presence rich diatom flora, the sediments are sometimes charac-
in northern Europe (Parfitt et al., 2005). terized by a white siliceous mud composed almost entirely
of diatom frustules (section 4.3), which is termed diatomite. these conditions. Heavy metals are reactive with sulphides
These deposits may form under either eutrophic or oligo- and form compounds that are typically black or blue-
trophic conditions depending on the ecological affinities of black in colour. These are usually very obvious in lake sedi-
the diatom species. ments, as they contrast markedly with the lighter-coloured
The nature of lake sediments is also partly related to lake oxidized clastic sediments or the brown colour of oxidized
size and the temperature and density variations experi- organic sediments. The former also often emit a pun-
enced by lake waters. Lakes can be classified according to gent sulphurous odour and have a distinctive chem-
whether they are chemically or thermally stratified, a ical composition. Changes in lake stratification and the
distinction that may be important for the interpretation former degree of oxygenation of lake waters can there-
of lake sediment sequences as well as their fossil content. fore be inferred from an analysis of sediment chemistry. In
The density structure of lake water is controlled by tem- diamictic lakes which experience marked stratification,
perature, by solids in the water column and by dissolved seasonal turnover may cause sudden and large-scale ‘die-
compounds (Boehrer & Schultze, 2008). In freshwater, off’ of aquatic organisms, such as algae (notably diatoms),
the temperature of maximum density is about 4°C at the and in some cases this can result in the formation of
surface and declines to about 3.4°C at a depth of 500 m due annual laminations (organic varves: Figure 5.23) in the
to increased pressure. In lakes where summer water tem- sediments accumulating on the lake floor. A seasonal
peratures exceed 4°C, the surface waters become less dense alternation between oxygenated and anoxic bottom waters
than bottom waters and a thermal stratification evolves. may also lead to seasonal production of calcareous deposits,
The warm surface waters (epilimnion) are separated from iron oxides or sulphur-rich sediments and hence chemical
deeper, cooler water (hypolimnion) by a marked thermal varves.
gradient, termed the thermocline. Stratification may also Over time, lakes silt up, plants encroach from the
occur in winter due to freezing at the surface. Most lakes marginal zones, and areas of open water are progressively
in temperate regions therefore experience vertical mixing eliminated. The succession from open water to mire and
(breakdown of thermal stratification) twice a year (dia- bog is known as a hydrosere and the sediments gradually
mictic lakes), during what are called the spring and autumn change in character from muds to peats. Three broad
‘overturns’. In high-altitude and high-latitude lakes, how- categories of peat can be identified and each is characteristic
ever, where warming does not exceed 4°C, mixing may of a particular stage in the hydroseral succession. These are:
occur only once per year (monomictic lakes). Similarly, (1) limnic peats which form beneath the regional water
lakes in low-latitude regions may not cool to below 4°C, and table and which are composed partly of transported plant
are also usually monomictic. Some lakes (meromictic) debris and partly of decayed vegetation formerly growing
remain stable throughout the year, and do not overturn, in situ; (2) telmatic peats which form in the swamp zone
usually because of a strong density gradient as a result of between high and low water levels and which are largely
light freshwater overlying denser water at depth, the latter autochthonous in origin; and (3) terrestrial peats which
containing high concentrations of solids or dissolved salts. accumulate at, or above, the high water mark and which
In very general terms, large (>10 km2) and deep (>10 m) are entirely autochthonous in derivation. Each of these peat
lakes tend to be stratified, whereas shallow or small lakes types will be composed of the remains of particular peat-
tend to be unstratified. forming plants, depending on the stage in the hydrosere and
The importance of density/thermal stratification lies the trophic status of the lake water (Figure 3.46). The rate
in its effects on oxygen levels in the water column, on at which hydroseral succession progresses, and lakes
the preservation of organic detritus and on the influences become infilled, depends upon a number of factors, includ-
these have on lithological composition. The epilimnion is ing the size of the lake, the size of the catchment, the rate
usually well oxygenated, while the hypolimnion is anoxic of sediment supply and productivity within the lake. Many
(anaerobic), and this leads to important differences in smaller lake basins in northwest Europe and North America
lake chemistry, lake biota and the types of sediment have become completely infilled since the end of the last
that accumulate. Organic remains, for example, are more glacial stage, and are currently covered by peat deposits,
likely to survive degradation under anoxic conditions. whereas a number of larger lakes are still accumulating
On the other hand, anoxic conditions may limit the variety limnic sediments, except perhaps in shallow marginal areas
and abundance of organisms inhabiting the lake, for they where telmatic and terrestrial deposits have formed.
can lead to the production of hydrogen sulphide (H2S) The terminology applied to peat-forming environments
which, in high quantities, is toxic for most organisms. can be confusing. In Europe and Russia, all waterlogged
Exceptions include reducing bacteria that can thrive under areas where peat develops as a result of reduced vegetal
Figure 3.46
Motinia peat
Birch/Pine wood peat

a) Oligotrophic Some sediment types deposited
Sphagnum-Eriophorum
Calluna
with increasing depth of water under
Motinia peat
Sphagnum peat
oligotrophic a) and eutrophic b)
peat
Terrestrial conditions (after Birks & Birks, 1980).
Sphagnum-
(mires)
Enopborum
peat
peat ' . ,
High water level
Motinia peat
Aquatic
Telmatic
.
Sedgei peat
(swamps)
Low water level
t
Fine detritus
Coarse
detritus
Diatomite
Fine
Limnic
(lakes)
Fen wood peat

b) Eutrophic Sedge peat
Terrestrial
Sedge
peat;
(mires)
seat
High water level

peat;
k i Moss
Telmatic
(swamps) High water level
Reed
peat
Fine detritus
detritus
Coarse
Marl
Limnic
Gyttja
(lakes)
decay under anaerobic (anoxic) conditions are usually bogs form the final stage of topogenous hydroseral suc-
termed mires. In North America, however, the term ‘peat- cessions (Figure 3.47) whereas blanket bogs are typical of
land’ is more generally applied to such areas (Charman, upland areas and develop as a continuous cover over the
2004). Mires can be divided into those in which the high landscape where rainfall is high.
water table that induces peat formation is a consequence
of groundwater conditions, either where drainage is im-
3.9.3 Palaeoenvironmental evidence from
peded (soligenous mires) or where water accumulates in
enclosed basins (topogenous mires), and those where the
lake sediments
water table is maintained by high atmospheric moisture Like caves, lakes form natural sediment traps, and analysis
levels (ombrogenous mires). Topogenous mires are, of of the sedimentary record in lake basins enables infer-
course, part of the hydroseral sequence and are usually ences to be made about environmental changes around the
referred to as fens if eutrophic and valley bogs if oligo- lake catchment. These include changes in the composition
trophic. Soligenous mires are almost always oligotrophic. of the vegetation cover, which may reflect both natural
Ombrogenous mires (usually termed bogs) can be sub- (climatically induced) or anthropogenic processes, and
divided into raised bogs and blanket bogs. Raised bogs changes in the rates of operation of geomorphological
develop mainly in lowland areas where the peat-forming processes on catchment slopes (solifluction, gelifluction,
plants, principally Sphagnum mosses, produce a domed etc.). In addition, both the levels of lake waters and the
surface above the level of the surrounding ground. Raised pattern of sedimentation within the lake will be influenced
content (see Figure 3.48). The whole sequence often rests

1
Lake upon glacial gravels or sands (deposited during the last
cold stage) and commonly lies beneath Holocene peats
Limnic and organic muds. Palaeobotanical evidence suggests
that the two minerogenic horizons accumulated during
2 periods of reduced vegetation cover, the former during
the pioneer phase immediately following local deglaci-
Telmatic
ation, while the latter represents the cold phase of the
Limnic Loch Lomond or Younger Dryas Stadial when a periglacial
regime prevailed. Minerogenic material was therefore
,3
transferred from the catchment to the lakes, especially
Fen
during the Stadial phase, when surrounding slopes were
Telmatic affected by freeze–thaw activity, gelifluction and overland
Limnic flow. The organic-rich sediments, however, contain a fossil
record reflecting a vegetation of shrub or open woodland
Terrestrial
that developed under more stable conditions, as indicated,
4
in turn, by a substantial reduction in the inwash of minero-
genic material.
Telmatic
Other approaches to the investigation of lake deposits
Limnic include analyses of the chemistry and magnetic properties
of lake sediments. In Britain, pioneering research on Late-
glacial sediments by John Mackereth, Winifred Penning-
5
ton and colleagues (e.g. Pennington et al., 1972) demon-
Raised strated how changes in the chemical composition of lake
bog sediments could most readily be explained if the sediments
were regarded as sequences of soils derived from the
catchment. Increased soil erosion results in the transfer of
large amounts of relatively unweathered material into lake
Figure 3.47 Schematic representation of gradual infill of a
small lake with lake muds in deep water, fen peat in shallow basins, and the mineral fraction of the sediments that
water, and Sphagnum peat formed in a raised or domed bog accumulates under such conditions is therefore charac-
(modified from Foss, 1987). terized by higher proportions of metal elements, most
notably sodium (Na), potassium (K) and magnesium (Mg)
and, in certain cases, calcium (Ca), iron (Fe) and manganese
by changes in precipitation, while in calcareous lakes the (Mn), derived from exposed sediment or bedrock. During
isotopic record of the sediments will be determined, at least periods of reduced erosive activity and soil maturation
in part, by former temperature regimes. Hence it may be under a vegetation cover, the mineral material transported
possible to make inferences about regional climatic change into the lakes would be leached of its content of potassium,
from the analysis of lake sediment records. These aspects magnesium and sodium in particular, and during such
of palaeolimnology (the study of ancient lake sediments) ‘stable’ phases, the lake sediments record lower concen-
are considered in the following section. trations of those elements. Periods of reduced erosion also
coincide with higher values for organic carbon, which
result partly from increased aquatic productivity within
3.9.3.1 Lake sediments and landscape
the lake basins, and partly from the increase in organic
changes
matter washed in from the catchment. Where acid soils and
It has long been customary for those working on the pollen peats develop in the catchment, this may result in high
content of lake sediments to draw parallels between inferred iodine/carbon ratios or a high Fe and Mn content in lake
vegetation changes and variations in sediment stratigraphy. sediments.
In northwest Europe, for example, Lateglacial lake deposits Chemical data from lake sediments can therefore be
typically consist of a threefold sequence of organic-rich used to augment pollen-stratigraphic and other palaeo-
lake muds (often gyttja or clay–gyttja) which overlie and environmental data in the reconstruction of lake catch-
are underlain by mineral sediments with a very low organic ment histories. Figure 3.49 shows chemical profiles from
Lithe-stratigraphy I 4 * y g p
9
r Local developments Regional developments
Enhanced growth of
ombrogenous peat in
response to increased climatic
wetness Rise in water table
submerges trees
5,000 Wetter conditions in mid-
Holocene and major climatic
Colonization of mires by shift to a more oceanic regime
alder and birch
Falling water levels in early-
mid Holocene refecting drier
Telmatic peat accumulates conditions. Spread of forests
as hydrosere progresses onto mire surfaces
beyond the open water stage
Organic sedimentation from

internal productivity and Climatic amelioration in early
inwash of soluble humic Holocene. Beginning of soil
10.000 acids and plant debris stability and soil maturation
M in erogenic inwash following Periglacial regime of the

break-up of vegetation cover Younger Dryas Stadial.
and frost disturbance of soils Cryoturbation and gelifluction
Organic productivity in lake widespread. Glaciers in
basin low highland areas of NW Europe
11.000
More equable climatic regime

Organic sediment of Lateglacial Interstadial.
accumulation from internal Vegetation colonization of
productivity and inwash of slopes, soil stability and
plant debris and soluble maturation. Higher temperatures
humic acids and development of lake
13.000 ecosystems
M in erogenic inwash into lake Climatic amelioration, but

prior to closure of the continued inwash of minerogenic
vegetation cover. Internal material prior to stabilization
productivity low of slopes by vegetation
Glacial outwash and other

Deglaciation
sediments
Figure 3.48 Lateglacial and early Holocene environmental changes inferred from lithostratigraphy of a typical northwest
European lake and mire sequence.
two lake sequences on the Isle of Skye, western Scotland, Stadial (13–11.5 ka) led to the break-up of the interstadial
which date from the Lateglacial and early Holocene periods vegetation cover, the destruction of soils by freeze–thaw and
(c. 15–10 ka). The lowermost sediments accumulated in frost heave, and the renewed inwash of mineral material
the basins following the wastage of the last ice sheet from into the basins. These environmental changes are reflected
Skye and contain relatively high quantities of Na, K, Ca in the reduced organic carbon content of the lake sediments
and Mg, and a low organic carbon content. These data and the significant increase in base element content. The
reflect the inwash into the lakes of unweathered, finely com- abrupt climatic amelioration at the beginning of the
minuted clastic material derived from freshly exposed Holocene is marked by a sharp upturn in the organic
glacigenic sediments. Subsequently, an increase in carbon carbon curve and a marked decline in the curves for Na,
content reflects the gradual stabilization of slopes around K, Ca and Mg. In general terms, therefore, the chemical
the catchments and maturation of soils as the vegetation record from these lake sediments can be read as a proxy for
cover expanded in response to climatic amelioration. regional climatic change (Walker & Lowe, 1990).
The reduction of mineral inwash is marked by the decline Anthropogenic effects on the landscape during the
in base elements in the sediment profile. A return to cold Flandrian can also be inferred from the chemical record
conditions during the Loch Lomond (Younger Dryas) of lake sediments. In his pioneering work, Mackereth
Figure 3.49 a) Loch Ashik
)
)
)
)
m
Variations in the abundance of selected chemical
m
m
(c
)
(c
(c
m
(c
m
elements in Lateglacial and early Holocene
(c
th
(c
th
th
th
p
p
p
th
p
th
De
De
De
sediments from two sites in the Isle of Skye, Inner
De
p
De
De
Hebrides, Scotland: a) Loch Ashik; b) Druim Loch a'jO
Flandrian
(from Walker & Lowe, 1990).
MO
Bio
Lomond
Sladial
Loch
520
r>:>(!
Laleglacial Interstadial
MO
MO
E-50
MO
O 10 20 30 0 10 20 0 10 20 0 10 0 10
% foss-on-ignilion mg/g
b} Dniim Locti
)
m
)
m
m
m
)
)
m
(c
(c
(c
(c
(c
(c
h
h
h
pt
pt
pt
h
h
pt
pt
De
De
pt
De
De
De
De
MO
Lomond Flandrian
!)*::
MO
Sladial
Loch
950
960
Lategfacial Interstadial
S7C
MO
MO
1000
0 10 20 0
0 10 20 30 0 10 20 0 10 0 10
% foss-on-ignilion mgyg
(1965) noted the increase in concentration of base elements Aberdeen in northeast Scotland, the onset of pastoral
in several lake cores in northern England following the activity after c. 5.39 ka, as indicated by the pollen record,
decrease in woodland cover recorded in pollen records at resulted in a threefold increase in sediment accumulation,
c. 6 ka. This, he suggested, was a reflection of increased soil while during the period of modern agricultural practices
erosion around the lake catchments following forest clear- (radiocarbon dated to 370 ± 250 years), about 25 per cent
ance by Neolithic people. Accelerated soil erosion arising of the sediment deposited during the past 12 ka was
from human activity during the mid- and late Holocene is accumulated (Figure 3.50b). Overall, lakes with no evidence
also reflected in the significantly increased sediment influx for accelerated sedimentation, or where it occurs very late,
into lake basins in many parts of the world (Bell & Walker, are all found in the west or north of the British Isles (Figure
2005; Boardman & Poesen, 2006). Analysis of rates of 3.50a). Periods of increased sedimentation have been dated
sediment accumulation in fifty lakes in Britain and Ireland to 5.9, 4.9 and 3 ka (Figure 3.50c). Similar increased rates
(Figure 3.50) has shown that accelerated sedimentation of erosion during the late Holocene have been noted at sites
rates typically occur in those basins where there is pollen in Europe. A record from Dallun So in Denmark, for
analytical evidence for human activity (Edwards & example, shows that the lake was relatively insensitive to
Whittington, 2001). For example, at Braeroddach Loch near catchment disturbance during the Neolithic and Early
a, b) g cm" a" 2 1
1 0 0.01 0 0 2 0.03 004

2
1 3
100
43 43
5 43 370 Cal BP
10 15
1 43 200 300 BP
1 i? 2100 Cal BP
1 11 20
1 12 19 21 /090 BP
18
Depth (cm)
22
25 43 300
26 24
27 3405 Cal BP
36
37 39 43 400 3110 BP
43
38 41«
42 3329 Organic carbon
32 30
43 Mineral sediment
31 500
43
n 43 5 5390 Cal BP
4600 BP
35 600
34 10645 Cal BP
9600 BP
b) 0
1000
fa
2<i
2000 43
35
44
C ka BP
3000 .3
13 53
4000 4 5 19
, 4
2: 30 38 32 27
5000 23 0 2<J
10 4- 44 •2
3S
6000
24
7000
Figure 3.50 Lake sedimentation in the British Isles. a) Locations of studied lakes showing those with (bold) and those without
accelerated late Holocene sediment accumulation. b) Record of sediment deposition at Braeroddach Loch, Scotland. c) Dates
at which accelerated sedimentation occurs at sites in (a); note that the timescale is in 14C yr BP in (c) and in both calibrated and
14
C yrs in (a) after Edwards & Whittington, 2001, b) after Edwards & Rowntree, 1980; from Bell & Walker, 2005).
Bronze Age periods (c. 5.9–3.0 ka), but there were dramatic increased lake productivity but, interestingly, had com-
impacts on the lake ecosystem by environmental changes paratively less impact on diatom assemblages and carbon-
and resultant erosion associated with a major phase of storage pathways in the lake ecosystem than the initial
deforestation at the transition from the Late Bronze Age to Iroquois disturbance (Ekdahl et al., 2004).
the Pre-Roman Iron Age around 2.5 ka (Bradshaw et al., Changes in lake sediment chemistry are also associated
2005). Lakes in North America generally show few signs of with changes in the magnetic properties of the sediments,
increased sediment input resulting from human impact and these can often be diagnostic of the types of processes
until after the European contact. At Crawford Lake, operating around the catchment. The quantity, size and
Ontario, Canada, however, there is evidence for increased mineralogy of ferrimagnetic particles in particular can be
nutrient input caused by Iroquoian horticultural activity established using magnetic susceptibility measurements.
from AD 1268–1486, and this elevated lake productivity In some lake sequences, magnetic parameters have been
caused bottom water anoxia and changed diatom com- used to reconstruct vegetation patterns around the catch-
munity structure within the space of a few years. A second ment and to relate these to patterns of climate change
phase of cultural eutrophication starting in AD 1867 (e.g. Geiss et al., 2003). In other lake contexts, magnetic
initiated by Canadian agricultural disturbance also data reflect increased erosion rates caused by human
disturbance, and offer insights into the type of land use silting up of the lake, blockage of outflowing streams
practised in the catchments. For example, at Gormire Lake by vegetation growth or by landslips, or the influence of
in northeast England, mineral magnetic properties human activity on lake catchments (see above). However,
combined with other proxy records (pollen analysis, regionally synchronous changes in lake level can be assumed
organic biogeochemical analysis) formed the basis for the to be climatically driven, in other words they reflect changes
development of a catchment history which revealed major in precipitation regime (Magny, 2007). A range of sedi-
phases of deforestation and erosion during the Late Iron mentological characteristics can be employed to recon-
Age/ Romano-British and Medieval/post-Medieval periods. struct histories of lake-level change (Figure 3.51). These
Before and between these periods, the lake catchment include grain-size variations (coarser deposits corres-
appears to have been mainly forest covered and erosion pond to near-shore areas); lithological properties (littoral
more limited and/or sporadic (Oldfield et al., 2003). sediments are often characterized by larger quantities of
organic matter from near-shore vegetation, while fine lake
marl is deposited in deeper water); sediment hiatuses,
3.9.3.2 Lake-level variations and climatic
which mark either erosion or non-deposition during
changes
phases of lake lowering; and the macroscopic components
Changes in water level in lake basins in temperate regions of lake sediments (molluscan abundance increases in near-
can be caused by a range of local factors, including the shore sediments; ostracods are more common in deeper
West East
C1 25m 50m
C1
C2
0
B
8780 C3
NAP 20% B i 260 BP
C1
late PB 9710
+ 60 RP
B
1
Bel. late P B
YD
bedrock Betula 3 6 % 11060

Junipews 16% ± 1 1 0 BP
At.
metres
iake-marl
oncolltes 11650
± 150 BP
sand
2 erosion surface Bel
Firbas zones
B : Boreal
PB : Preboreal 12360
Y D : Younger Dryas t 60 B P
A L : Allerad
Bel : Boiling Betula 4 2 %
Junipetvs 18%
Datum (0) = 443 69 metres
3
Figure 3.51 The lake sediment sequence at Sevrier-Les-Charretières, Lake Annecy, France. Deeper water phases are indicated
by the marl units, and shallow water phases by sand layers and oncolites (carbonate concretions). Erosion surfaces in the marginal
core mark episodes of very low lake level (after Magny, 2001, reprinted with permission from Elsevier).
water). Plant macroscopic remains also provide evidence consistent with the development of an anticyclone over the
of contrasting shallow and deeper water communities decaying Scandinavian ice sheet with enhanced south-
(Magny, 2001). westerly flow along the west coast and easterlies south of
During the Lateglacial period in Europe, significant the ice. After c. 9 ka, falling lake levels suggest drier condi-
fluctuations in lake level have been recorded and have tions than at present over much of northern Europe,
been linked to regional changes in climate. In the Nether- although wetter conditions are indicated by higher lake
lands, for example, higher water levels during the early and levels in the far north, along the west coast, in eastern
middle interstadial (Bølling–early Allerød: c. 14–13 ka), Finland and in western Russia. Differences in lake status
probably related to melting of permafrost under a warmer during the mid–late Holocene are consistent with streng-
climatic regime, were followed by fluctuating lake levels thening of the blocking anticyclone over the Baltic Sea in
in the later interstadial. During the Younger Dryas cold summer, resulting in a more meridional circulation than
stage (12.9–11.7 ka), lake levels fell markedly as climate that which obtains today (Yu & Harrison, 1995). Fluctu-
became drier (Bos et al., 2006). In Switzerland, lake-level ating lake status in central and southern Europe during the
records for the Lateglacial and early Holocene show a Holocene has been attributed to alternate southwards/
close correlation with the climatic sequence reflected in the northwards displacement of the Atlantic westerly jet stream.
Greenland ice-core records (section 3.11). Lower lake levels A southward movement of the jet could have led to
are recorded during the warmer episodes of the Interstadial enhanced cyclonicity, with the resulting increase in
(GI-1e, GI-1c and GI-1a: Figure 1.7) during the middle precipitation being reflected in higher lake levels (Harrison
period of the Stadial (GS-1), during the GS-1/Holocene et al., 1993; Magny et al., 2003).
transition, and immediately after the 11.2 ka event. Higher
lake levels correspond with the colder phases of the
3.9.3.3 Lake sediments and
Interstadial, that is, with GI-1d and 1b, early and late
palaeotemperatures
GS-1 and with the 11.2 ka event (Magny, 2001).
Even during the Holocene, however, regional climate In calcareous lake sediments, a direct relationship has been
changes appear to have been of sufficient amplitude to affect found between stable isotope variations in lake sediments
lake levels in temperate Europe. For example, a compilation and palaeotemperatures, notably the relation between water
of Holocene lake-level records obtained from twenty-six temperature and δ18O of lacustrine carbonates during
lake basins located in the French pre-Alps and Swiss Plateau equilibrium precipitation. For example, at Lake Torreberga
show common tendencies, with evidence for as many as in Sweden, the shift in δ18O at the end of GS-1 was con-
fifteen distinct lake-level highstands; these generally accord sidered to represent an increase in mean annual air
with episodes of increased climatic instability inferred from temperatures of c. 5.3°C (Hammarlund et al., 1999), while
Greenland ice-core records and with North Atlantic ice- in Hawes Water, northern England, a δ18O record from car-
rafting events (Magny et al., 2003). The data suggest that bonate sediments provided evidence for climatic warming
the Greenland ice sheet, the North Atlantic Ocean and at c. 12.5 ka, and four submillennial climatic events during
north-central Europe were responding to common climatic the course of the Lateglacial Interstadial (GI-1) prior to the
forcing agencies over the course of the Holocene, with onset of the Loch Lomond/Younger Dryas Stadial (GS-1),
variations in solar activity (section 7.6.4.1) perhaps playing that could be correlated with the climatic signal in the
a major role (Magny, 2004). In similar vein, changes in level Greenland ice-core record (Jones et al., 2002). Although
of a number of Swiss lakes during the last 3.5 ka correlate interpretation of the δ18O record in terms of tempera-
closely with fluctuations in extent of Swiss Alpine glaciers, ture is not always straightforward (Leng & Marshall, 2004),
suggesting regulation of both by winter cooling and and these data are perhaps best employed in conjunction
summer moisture, a correspondence that has again been with other palaeoclimatic and palaeoenvironmental indi-
attributed to variations in solar activity (Holzhauser et al., cators (e.g. Hammarlund et al., 2002, 2003), stable isotope
2005). ratios in lake sediments do, nevertheless, offer a potentially
Lake-level records also enable inferences to be made valuable independent method of inferring palaeo-
about changes in atmospheric circulation patterns. temperature changes and trends.
In northern Europe, for example, lakes during the early In addition to palaeotemperature reconstructions,
Holocene show conditions similar to or drier than present stable isotopes in lake sediments offer a basis for a
across southern Britain, southern Scandinavia and into number of other palaeoclimatic inferences. These include
the eastern Baltic, while wetter conditions obtained along seasonal changes in precipitation and temperature, decadal
the west coast and into central Europe. This pattern is and millennial variations in precipitation regime, and
linkages between lake sediment records and atmospheric climatic regimes and a range of palaeoclimatic data can be
circulation. Details of these and other aspects of climate- obtained from the careful analysis of peat stratigraphy. Peat
related isotopic records in lake sequences can be found in chronologies are usually based on radiocarbon dating,
Leng (2004) and Leng et al. (2013). although increasingly tephra horizons (section 5.5.2) are
being used as a basis for correlation and dating (Langdon
& Barber, 2004, 2005).
3.9.4 Palaeoenvironmental evidence from
In the early years of the twentieth century, data from
mire and bog sediments Scandinavian peat bogs enabled the Scandinavian botanists
In topogenous mires and bogs that develop over former Blytt and Sernander to identify what they considered to
lake deposits, the different types of peat may also provide be clearly defined climatic episodes (Pre-Boreal, Boreal,
indications of former climatic conditions. The rate at which Atlantic, Sub-Atlantic, Sub-Boreal) in the plant macrofossil
hydroseral succession has proceeded cannot usually be record of Holocene peat bog sequences in northern Europe.
interpreted simply in terms of climatic change, as local site What became known as the Blytt–Sernander climatic
factors will often exert a degree of control over the direction scheme was subsequently related to regional pollen zones
and rate of the succession (Charman, 2004), but in certain (section 4.2.4), and was enthusiastically adopted by
circumstances, it may be possible to make inferences about palaeoecologists and particularly by archaeologists as the
variations in the height of former water tables and hence basis for climatic subdivision of the Holocene. For some
about former precipitation levels. Good examples would time now, however, it has been apparent that the relation-
be where there is evidence for disturbance of the hydrosere ships between peat stratigraphy, pollen assemblage zones
(e.g. a transition from bog peat to reedswamp peat) reflect- and climatic change are much more complicated than
ing a major change in local environmental conditions, was envisaged in the 1920s and, as a consequence, the
such as a rise in silty water, possibly resulting from a scheme has now been largely abandoned, although,
change to a wetter climatic regime; or the occurrence of tree curiously, the terminology still appears even in current
roots in a raised mire sequence, which may indicate inva- literature (e.g. Roos-Barraclough et al., 2004). Early work
sion of the bog surface by trees during a drier climatic on peat stratigraphy also attempted to identify horizons in
period. peat sequences where there had been rejuvenation of peat
growth as reflected, for example, by a change from dark,
well-humified peats to light-coloured, less humified
3.9.4.1 Palaeoprecipitation records from
Sphagnum peats (Figure 3.52a). These became known as
ombrotrophic peats
‘recurrence surfaces’ and were believed to mark a change
Ombrotrophic peats can provide valuable palaeoclimatic to wetter mire surface conditions brought about by
information. A number of factors determine the rate and an increase in precipitation. Perhaps the most famous
type of peat formed, including temperature (which affects (or infamous) of these was the grenzhorizont (boundary
metabolic and decomposition rates), local topography and horizon) of northern Germany, first described by Weber
groundwater pH. The most important control, however, is more than 100 years ago, and radiocarbon-dated to c. 500
the height of the local water table. Above the water table, BC, the transition from the Bronze to the Iron Age on the
decomposition can take place relatively rapidly because archaeological timescale. As with the Blytt–Sernander
of increased microbial activity under aerobic conditions. scheme, however, questions have been raised about the
The height of the water table fluctuates throughout the year validity of recurrence surfaces as regional palaeoprecipi-
due to seasonal variations in regional drainage conditions tation indicators. Radiocarbon dating has shown that not
but, when examined over a timescale of decades, it will tend only do ages of inferred ‘recurrence surfaces’ from nearby
to have a minimum level, below which little decomposition sites differ by several hundred years, but that there may
can take place. This lower permanently waterlogged zone be considerable within-site variations in both age and
is termed the catotelm. Above this, a shallow near-surface stratigraphic position of individual ‘recurrence surfaces’
zone (the acrotelm) is regularly exposed for at least a brief (Figure 3.52b). This partly reflects changing rates of peat
period each year, and perhaps experiences prolonged growth on different parts of a mire surface, but it might also
dryness during the summer. The thickness of the acrotelm indicate mixing of carbon residues between peat horizons
will be determined by a combination of water supply and during oxidation and decomposition of the peat-forming
summer drying, especially in ombrotrophic peats that are materials. Where this has occurred, it can lead to major
totally dependent upon precipitation for their water supply. discrepancies between radiocarbon dates from the same
Ombrotrophic peats therefore contain a record of past stratigraphic horizon. A second problem relates to the
Figure 3.52 a) A peat profile at Bolton Fell Moss, northern

England, showing a lower dark, well-humified peat overlain by
a lighter-coloured, less well-humified peat. The boundary
between the two peat units has often been described as a
‘recurrence surface’, reflecting a regional change from drier
to wetter conditions (photograph by Mike Walker). b) A peat
section from Store Vildmose, Denmark, showing both lateral
and vertical alternations in darker and lighter peat layers. The
boundaries between units characterized by different degrees
of humification are marked by white pins in the peat face
(photograph by Keith Barber, Southampton University, UK).
extent to which recurrence surfaces develop in response to to changes in the strength and location of the westerlies,
regional climatic (i.e. precipitation) changes, or whether linked to large-scale North Atlantic Ocean and atmospheric
they may also form as a result of site-specific factors, such circulation (Charman et al., 2009).
as local changes in mire hydrology, particularly where A second approach to the reconstruction of mire surface
human activity has been a factor (Evans & Warburton, wetness records involves the use of plant macrofossils
2007). (Mauquoy et al., 2010). This technique exploits the strong
More recent approaches to climatic reconstruction response of peatland plants, especially of bryophytes
based on ombrotrophic peat sequences have focused on (notably Sphagnum) to changes in surface wetness. Some
down-profile variations in indicators of surface wetness species (e.g. Sphagnum sect. Cuspidata) are associated with
change, which provide a more realistic portrayal of trends wetter conditions on the mire surface, whereas others such
in past climate than individual features of the peat strati- as Sphagnum sect. Acutofolia are characteristic of drier
graphy (Charman, 2004). A widely used technique that has habitats. Hence, down-core changes in plant macrofossil
been applied to both blanket peats and raised mire peats is assemblages will provide evidence of surface wetness
peat humification (Chambers et al., 2011). This employs changes and, by implication, of changes in effective
colorimetric methods, and the speed and relative ease of precipitation (Barber et al., 2000, 2003). Figure 3.53 shows
analysis means that high temporal resolution measurements a palaeoprecipitation record from Walton Moss, northern
are possible. Although most humification studies have England, derived from plant macrofossil records. The
been undertaken in Britain and Ireland, where proximity sequence extends from the early Holocene to the present
to the North Atlantic Ocean means that the peat bogs tend and shows significant wet shifts at 7.8 ka, 5.3 ka, 4.41–3.99
to be more sensitive to changes in effective precipita- ka, 3.5 ka, 3.17–2.86 ka, 2.32–2.04 ka, 1.75 ka, 1.45 ka,
tion (e.g. Blackford & Chambers, 1995; Charman et al., 0.3 ka and 0.1 ka. Time-series analysis suggests that these
2009), mire surface wetness records have been generated climatic changes are not stochastic, but there are underlying
from other parts of Europe (Roos-Barraclough et al., periodicities of c. 600 and c. 100 years between wet-shifts
2004; Andersson & Schoning, 2010), from New Zealand (Hughes et al., 2000).
(McGlone & Wilmshurst, 1999) and from Alaska (Payne A third technique that has been applied to ombrogenous
& Blackford, 2008). These records largely reflect changes peat sequences in order to derive a record of mire surface
in precipitation which, in western Europe, are due largely wetness is that of testate amoebae analysis. Testate amoebae
c a l . yr B P
0 2000 4000 6000 8000 10000

0 •
50¬
100¬
• C A a x i s 1 (SD units)
Eigen value 0 82
150¬
200¬
250¬
300¬
350-
Figure 3.53 Reconstructed mire-surface wetness changes at Walton Moss, northern England. Increased wetness (and, by
implication, higher levels of precipitation) is reflected in the higher values of detrended correspondence analysis (DCA) shown on
the y axis. DCA is a standard multivariate statistical technique which is used here to analyse the variation in plant macrofossil
components through the peat profile and to relate them to particular controlling factors (such as surface wetness) (after Hughes
et al., 2000, reprinted with permission of SAGE).
(rhizopods) are a group of the protozoa that are abundant reliable palaeoclimatic interpretation and strengthening
in almost all peatland environments and certain species the correlations with other climatic reconstructions
are reflective of surface wetness changes (Charman et al., (Mauquoy et al., 2004; Langdon & Barber, 2005; Nichols
2000). Again, therefore, down-core variation in fossil et al., 2006). A multi-proxy ombrogenous peat record for
testate amoebae can be read as a record of surface wetness the past 7.5 ka from Temple Hill Moss, southeast Scotland
changes (Booth, 2002). Testate amoebae can also provide is shown in Figure 3.54. The record not only contains a
other insights into peatland environments, for example number of the wet-shifts identified at Walton Moss (Figure
in characterizing hydroseral transitions from open water to 3.53), but it also displays the same millenial-scale periodicity
fen, and from pioneer raised mire to ombrotrophic bog of 1,100 years suggesting that both sites are responding to
(Elliott et al., 2012). Other proxies used to reconstruct a regional climatic forcing factor (Langdon et al., 2003).
variations in bog surface wetness include the ratios of
abundances of n-alkanes (components of plant lipids) in
3.9.4.2 Stable isotope records from
selected plant macrofossils (Nichols et al., 2006), and fungal
ombrotrophic peats
spore records (see Chambers et al., 2012, for an overview).
Increasingly, however, these various methods are In ombrotrophic bogs, which are entirely rain-fed, the
being used in combination, for such a multi-proxy isotopic composition of rainwater (reflected in D/H and
18
approach enables changes in each proxy reconstruction to O/16O ratios) will be reflected in plant tissue and, in
be validated against another, thereby producing a more theory, isotopic variations through the peat sequence can
-18
-16
Residuals
a)
Increasing dryness
-14
Testate Amoebae
Residuals
-12
(absorbance
-10
(absorbance
-8
-6
at 540
-4
at 540
nm)nm)
-2
2 0.4
Residuals (absorbance at 540 nm)

0 2
Residuals (absorbance at 540 nm)

b) 0.0
Residual
Residual
-0.2
-0.4
400 -0.6
G l e n Garry Lairg 'A'

350
Tephra Tephra
300
c)
s (absorbance at 540 nm)
Increasing wetness
250
200
DCA
150
100
50
0 1000 2000 3000 4000 5000 6000 7000

cal. y r B P
Figure 3.54 Proxy climate reconstructions from Temple Hill Moss, southeast Scotland. a) Variations in water table depth based
on testate amoebae analysis. b) Peat humification record. c) Detrended correspondence (see Fig. 3.53) based on plant macrofossil
data. The peaks in the curves represent drier intervals and the troughs wetter episodes. The vertical shaded bars highlight periods
of wetter/cooler climate (after Langdon et al., 2003).
be linked to past changes in temperature. If so, then the woodland clearance (Martínez-Cortizas et al., 2005), while
stable isotope record can be read in terms of past precipi- historic and prehistoric mining and metallurgy is reflected
tation and temperature variations. In practice, however, the by the presence in peat sequences of heavy metals such as
situation is much more complex, and there is considerable lead and mercury (Martínez-Cortizas et al., 1997; Farmer
uncertainty about how isotopic ratios from peat should be et al., 2009).
derived and interpreted, principally because of strong inter-
species variation between different plants in the ways in 3.10 THE DEEP-SEA SEDIMENT
which water is moved to the growing tissues (Charman,
2004).
RECORD
To a degree, some of these problems have been resolved
3.10.1 The nature and origin of ocean
by the recent development of more sophisticated analytical
techniques (e.g. continuous flow isotope ratio mass spectro- sediments
metry), which have enabled oxygen isotopic values to be On the deep-ocean floors, sediments have been accum-
measured on only a few mg of cellulose (Loader et al., 2007). ulating in a relatively undisturbed manner for thousands,
This means that genus-specific analysis can now be carried or even millions, of years. They consist partly of terrigenous
out, thereby avoiding complications arising from the deposits, that is, detrital material derived from erosion of
different chemical signatures in different plants. Using the the land masses surrounding the ocean basins, and partly
established relationships between modern temperature and of biogenic sediments composed largely of accumula-
the δ18O precipitation signal, Daley et al. (2009, 2010) tions of the calcareous and siliceous skeletal remains of
developed a Holocene climate record for Nova Scotia micro-organisms that formerly lived in the ocean waters.
and northern England, although in the latter case the Terrigenous detritus (ranging in size from fine sand to
reconstructions show larger temperature anomalies than clay) arrives on the ocean floor by a number of different
have been found in other archives. This may reflect a range pathways, but the principal transporting agencies are
of factors, including the influences of evaporation/peatland turbidity currents, bottom currents, wind and ice. In the
surface wetness, and changing δ18O of the precipitation mid- and high latitudes, both coarse and fine terrigenous
source. More promising appears to be the compound- detritus appears to have been delivered to the ocean
specific analysis of hydrogen isotope ratios obtained from floors mainly during glacial periods, reflecting in particular
Sphagnum leaves; the reliability of this approach has been the ice-rafting of glacially eroded debris (see below) and,
tested on modern peat surfaces and through comparisons to a lesser extent, the transport of aeolian sediments
with bog-surface wetness trends over the last 3 ka, obtained from the greatly expanded periglacial regions (section 3.4).
from other proxy indicators (Nichols et al., 2010). The Indeed, wind-blown sediment may have constituted a
results suggest that the interpretation of stable isotope major proportion of the fine detrital input in low latitudes
ratios in ombrotrophic peat is less problematic when during glacial times. Sea-level lowering of over 100 m
restricted to selected compounds and plant species. would also have resulted in the discharge of large quantities
of terrestrial debris from the major rivers as they flowed
across the continental shelves, and this material would
3.9.4.3 Human impact recorded in
subsequently have been spread down the continental
ombrotrophic peat
slopes and across the abyssal plains in gravity-controlled
The effects of prehistoric and historic human activity sediment flow.
are often preserved in the geochemical record of peat In many ocean sediment sequences, therefore, a broad
sequences. Arable farming, for example, may be reflected correlation can be detected between the deposition of
in an increase in the immobile elements silicon (Si) and terrigenous material and former glacial episodes, and this
titanium (Ti) as these are readily removed from bare is reflected most clearly in the large volumes of ice-rafted
ground by wind action and subsequently deposited on debris (IRD) that are found in deep-ocean sediments. In
peat surfaces (Hölzer & Hölzer, 1998). The intensity and the North Atlantic, for example, IRD deposition appears
timing of farming activity can be corroborated by pollen- to have been by far the most important mechanism for
analytical and other evidence (Lomas-Clarke & Barber, supplying terrigenous sediment to the ocean floor and it
2006). The occurrence of these and other chemical elements has been estimated that IRD may make up as much as
(K, Ca, Rb, Sr, etc.) in peat profiles provide additional 40 per cent of the total amount of sediment deposited in
evidence of human impacts, such as an increase in atmos- Quaternary cold stages (Robinson et al., 1995). Cycles
pheric dust flux from exposed soils following anthropogenic of IRD deposition in the North Atlantic during the course
Figure 3.55
0
Heinrich layers in deep-ocean
4.0 ka sediments. The two core segments
(right) were recovered from the
10.4 ka northwestern Labrador Sea between
Baffin Island and Greenland (lat.
61°30’N; long. 58°26’W).
He in rich Sedimentological analyses (left) show
13 2 k a
Layer 1 that the Heinrich layers are
characterized by significantly higher
I DC
levels of carbonate and mass
b
magnetic susceptibility (MSS), and in
14.7 ka the cores they are reflected in a colour
change from grey to brown (‘a’ and ‘b’
Depth cm
mark the end of Heinrich events 1 and

2, respectively). The two Heinrich
layers shown here are radiocarbon-
19.5 ka
dated to 14.7–14.2 ka BP (c. 18.0–17.4
k cal. BP) and 21.5–19.5 ka BP (c.
200 25.6–23.3 k cal. BP), and are coeval
with Heinrich events 1 and 2 in the
Heinrich eastern North Atlantic
Layer 2 (sedimentological data from Andrews
& Tedesco, 1992; photographs by
b John Andrews, University of Colorado,
21.5 ka
USA).
25.5 ka
300
0 25 50 300 150 0
Total carbonate (%) MMS (x 1u- m /kg)
B 3
Dolomite (%)
of the last cold stage are reflected in distinctive layers of 3.10.2 Oxygen isotope ratios and the
glacially derived material in the ocean sediments around
ocean sediment record
45°N (Figure 3.55). These have been termed Heinrich
layers and reflect episodic deposition of carbonate-rich IRD
3.10.2.1 General principles
from icebergs drifting eastwards across the Atlantic from
the margins of the Laurentide ice sheet (Heinrich, 1988; An extremely important source of palaeoenvironmental
Hemming, 2004). The Heinrich events appear to have had information is the chemical and isotopic content of the
a major effect on the global, or at least the Northern Hemi- marine organisms contained within ocean sediments.
sphere, climate system and their significance in the context Variations in the ratios between, for example, aluminium,
of the history of the North Atlantic during the course of the barium, calcium and cadmium, and between the iso-
last cold stage is discussed in section 7.4. topes of carbon, oxygen and uranium, reflect the com-
In the deeper oceans, the sediments tend to be finer bined influences of circulation, nutrient supply and water
grained and are often dominated by biogenic material temperature, and therefore provide a basis for the recon-
consisting of the accumulations of the carbonaceous and struction of oceanographical changes. Some of these
siliceous remains of micro-organisms that formerly lived lines of evidence are discussed in section 4.10. However,
in the ocean waters. Such sediments are known as marine it is the application of oxygen isotope analysis to deep-
oozes (section 4.10) and are frequently characteristic of ocean sediments that has had by far the greatest impact
interglacial or warmer episodes. They contain recogniz- in Quaternary science. The method not only provides one
able fossil remains, and these provide a record of ocean of the principal indices of global environmental change
circulation, ocean water temperature and, by implication, during the Quaternary, but also serves as a basis for global
atmospheric temperatures throughout the Quaternary. stratigraphic subdivision and correlation (Chapter 6).
The use of fossils in the reconstruction of changing patterns Moreover, it was the marine oxygen isotope record that
of oceanic circulation is considered in Chapter 4. first convincingly demonstrated the influence of the
Milankovitch radiation cycles on the earth’s climatic reflect the ratios prevailing in the oceans at the time of their
history (Imbrie & Imbrie, 1979). Oxygen isotope analysis secretion. Analyses have been carried out on the remains
has already been touched upon in the discussion of cave of a range of marine micro-organisms, but by far the most
speleothems (section 3.8.4.2) and lake carbonates (section widely used fossils are the tests of the planktonic (near-
3.7.3), but because deep-ocean oxygen isotope stratigraphy surface) and benthic (deep-water) Foraminifera (section
has such far-reaching implications for Quaternary science, 4.10). Data from a large number of ocean cores show that,
the principles and applications of the method are discussed over time, the isotopic balance of the world’s oceans has
here in more detail. An overview and historical review of changed. Several factors are involved in this process, but
oxygen isotopes in marine Foraminifera is provided by the two principal elements are the isotopic composition
Pearson (2012). of the ocean waters at the time of secretion and water
Oxygen can exist in three isotopic forms (16O, 17O and temperatures. The former largely reflects changes in land-
18
O) but only two (16O and 18O) are of importance in ice volume, and it is important to appreciate why this is so
oxygen isotope analysis of marine deposits. 18O/16O ratios before moving on to consider other influences on the
in the natural environment vary between about 1:495 and marine isotope signal.
1:515, with an average of approximately 1:500. This means
that only about 0.2 per cent of oxygen in natural circulation
3.10.2.2 Glacial ice storage and the marine
is 18O. Ratios of oxygen isotopes are measured not in
oxygen isotope record
absolute terms but as relative deviations (δ18O per mil) from
a laboratory standard value. The standards normally Changes in δ18O values in marine microfossils are a result
employed are PDB (belemnite shell: see section 5.3.2) for of a natural fractionation of oxygen isotopes as water
the analysis of carbonates and standard mean ocean evaporates from the sea surface. During evaporation, the
water (SMOW) for the analysis of water, ice and snow. The lighter H216O molecule is drawn into the atmosphere in
latter is used in the isotopic analysis of glacier ice cores preference to the heavier H218O molecule. This process is
(section 3.11). PDB is +0.2 per mil in relation to SMOW. temperature dependent and so it is particularly marked at
The standard materials are monitored and distributed to higher latitudes where colder air masses are increasingly less
laboratories by the International Atomic Energy Agency able to absorb the heavier isotope. Thus the moisture-
in Vienna.10 Mass spectrometric analyses are carried out bearing winds that nourish the polar glaciers contain
on CO2 gas prepared from the fossil material and on the relatively higher quantities of the lighter 16O, and this, in
standard reference material. Oxygen isotope ratios are turn, is reflected in the isotopic composition of glacier ice.
then expressed as positive or negative values relative to During the cold phases of the Quaternary, with markedly
the standard (δ = 0), thus: expanded ice masses in both Northern and Southern
Hemispheres, large quantities of H216O were trapped in the
18
O/16Osample – 18O/16Ostandard ice sheets leaving the oceans relatively enriched in H218O
␦ 18O = 1000 × 18
O/16Ostandard (i.e. isotopically more positive, or heavier). Conversely, the
melting of the ice masses during interglacial periods
liberated large volumes of water enriched in H216O back
A δ18O value of –3 per mil indicates that the sample is into the oceans (Figure 3.56). The result, therefore, is a
0.3 per cent or 3.0 per mil deficient in 18O relative to the cyclical down-core oscillation in the marine oxygen isotope
standard. A δ18O value of –10 per mil is even more deficient signal, reflecting alternating glacial (expanded land ice
in δ18O and is therefore described as isotopically lighter volume) and interglacial (reduced land ice volume)
than a value of –3 per mil. It is important to understand conditions (Figure 3.57). Analysis of 18O/16O ratios in
this because reference is commonly made in the literature Foraminifera has revealed that the overall glacial–
to isotopically ‘light’ and isotopically ‘heavy’ segments of interglacial variation in isotopic composition of ocean
oxygen isotope records (see below). waters was relatively small, the δ18O record from planktonic
The variation in the isotopic composition of ocean species suggesting that ocean waters at the Last Glacial
waters over time can be reconstructed from the δ18O values Maximum (LGM) were typically about 1–1.5 per mil
of carbonate shells and skeletons preserved in deep-sea more positive than at the present day (Cortijo et al.,
sediments. Many marine organisms secrete (or build) car- 2000). Isotopic data obtained from benthic species have
bonate structures and oxygen is abstracted from seawaters much heavier δ18O values than surface species, typically in
for this purpose. Thus, the oxygen isotope ratios in fossil the range +2.5 to +5.0 per mil (Shackleton et al., 1990;
carbonates buried in sediments on the ocean floors should Dwyer et al., 1993), reflecting the higher δ18O content of
common to all marine isotope profiles, and hence the

a) G l a c i a l
clearly defined terminations form marker horizons that can
Glacial maximum
be used as a basis for inter-core correlation (section 6.2.3.5).
expanded ice sheets
6 0 -30t»
, e The diagram also shows that, for much of the last 600 ka,
global land ice cover was far greater and, by implication,
the climate was cooler than at present. Hence, the current
MSL
relatively equable climate is atypical when viewed in the
ca 120m LSL
context of the isotope records for the last 600 ka (section
Continental
o 1 8
0 ocean 7.3.3).
land mass
-+1.5%o
0
i
b) I n t e r g l a c i a l I
Ice s h e e t s " * 100

reduced in size II
MSL
200
Continental 5 1 8
0 ocean in
land mass ~0.0%o
Terminations
Age (ka)
300
IV
Figure 3.56 Variations in surface water oxygen isotope ratios
during times of glacial maxima and interglacial high sea-level
stands (minimal ice cover).
400
V
deep-ocean water. Overall, however, the glacial–interglacial
oscillations reflected in isotope profiles from benthic species
match very closely those obtained from planktonic species,
which suggests that, in spite of the influence of other 500 VI
factors (see below), both isotopic signals contain a
significant ice-storage component.
Oxygen isotope profiles from marine deposits can
therefore be interpreted principally as a record of global
palaeoglaciation. In Figure 3.57, the marine isotope signal 600-
VII
has been drawn to show patterns of global ice accumula- I n c r e a s i n g ice v o l u m e
shaded
tion and wastage over the course of the last 600 ka (the scale Temperature
is the reverse of that normally depicted, with isotopically Increasing Decreasing

heavier values to the right). The isotopic trace is asym-
metrical, reflecting a gradual increase in ice volume during Figure 3.57 Schematic representation of oxygen isotope
variations for the past 600 ka. Isotopically light values to the
the course of a cold stage, followed by a sudden decline left, and isotopically heavier values to the right. The isotopic
indicating rapid ice wastage. These latter points on the signal is interpreted as a proxy for extent of global palaeo-
isotope curve are referred to as terminations (I–VII in glaciation. I–VII are termination events (section 6.2.3.5). The
Figure 3.57). Small amplitude variations in isotope values vertical line represents the isotope ratio that corresponds to the
limited ice cover typical of the late Holocene. Only three short
between successive terminations suggest that the gradual
periods during the preceding 600 ka appear to have experienced
build-up of land ice was interrupted by short-lived epi- comparable climatic conditions (after Broecker & Denton,
sodes of glacier retreat. The overall shape of the curve is 1990).
Oxygen isotope traces through cores of deep-ocean suggests an LGM–Holocene change attributable to ice
sediment reveal records of glacial–interglacial changes volume of 1.1–1.4 per mil (Mix et al., 2001). This is roughly
spanning, in many instances, the whole of the Quaternary. equivalent to an excess LGM ice volume (i.e. more ice than
There is, moreover, a remarkable similarity between the at present) of 48–59 × 106 km3 (modern ice volume being
isotopic profiles from different parts of the world’s oceans, estimated at 32 × 106 km3). Other reconstructions of the
in terms of both the number and amplitude of oxygen ice volume component in the benthic isotopic record of
isotopic cycles. This suggests that the oceans as a whole have the last glacial–present interglacial transition suggest a
been responding in a coherent fashion to a common forcing slightly lower isotopic shift. Shackleton (2000), for example,
mechanism (the Milankovitch radiation cycles), and hence obtained an estimate of ~1.0 per mil based on a comparison
variations in the isotopic signals are broadly synchronous of the benthic marine δ18O record and the δ18O signal from
worldwide. The common peaks and troughs in the curves the Vostok (Antarctica) ice core (section 3.11), while Schrag
can therefore be designated as isotopic stages, and these et al. (1996) derived a value of 0.8–1.0 per mil from direct
form the basis for stratigraphic subdivision of the individual measurements of the isotopic composition of pore waters
profiles and for correlation at the global scale. These aspects in deep-sea sediment cores. These estimates, if correct,
of oxygen isotope stratigraphy are considered further in would imply somewhat lower global ice volumes at the
sections 6.2.3.5 and 6.3.3. LGM.
In terms of sea-level changes, it has been suggested
that a sea-level change of c. 10 m will be represented by a
3.10.2.3 Ice volumes, sea level and the
c. 0.1 per mil shift in the oxygen isotope signal (Shackleton
marine oxygen isotope record
and Opdyke, 1973). Based on the inferred ice-volume
Given that there is a close relationship between the glacial effect in the benthic δ18O record noted above, this would
storage of oxygen isotopes and the marine oxygen isotope imply a sea-level lowering at the LGM of –110 to –140 m,
record, it should, in theory, be possible not only to interpret an estimate broadly confirmed by other δ18O-based
the magnitude of the δ18O shift between glacial and inter- reconstructions (Waelbroeck et al., 2002). This figure is also
glacial events in terms of fluctuating global ice volumes, consistent with independent evidence for global sea-level
but also to derive a quantitative estimate of global ice change. For example, sea-level estimates from raised coral
volume changes. Similarly, insofar as ice volumes are terraces in New Guinea point to a sea-level lowering of
related to sea level, the marine δ18O record could be around 125–130 m at the LGM (Figure 3.58). A similar
employed as a proxy for global sea-level change. The plank- estimate (–125 ± 5 m) has been obtained by combining
tonic δ18O record is perhaps less suitable for these purposes evidence from a range of low-latitude sites with data
because of temperature and other influences, hence ice from ice modelling (Fleming et al., 1998), while faunal
volume and sea-level changes are usually inferred from and sedimentary records from the northwest Australian
benthic foraminiferal oxygen isotope signals. continental margin suggest sea level at the LGM at –135
Benthic Foraminifera inhabit the deep oceans of the to –130 m (Yokoyama et al., 2000). A sea-level lowering of
world, which are generally more stable and uniform in 125 ± 5 m corresponds to a volume of grounded ice in
terms of temperature and salinity than are ocean surface excess of the present volume (see above) of 46–49 × 106 km3,
waters. As a consequence, it has generally been considered
that temperature change in the deep oceans of the world
has been minimal, and thus the deep-ocean isotopic signal 50
Metres below present
Sea level estimated

is more readily interpreted as reflecting glacial storage vari- from New Guinea
0
ations, as opposed to the temperature-influenced frac- terraces
tionation effects of carbonate precipitation. However, -50

it is now recognized that most benthic foraminiferal
-100 Sea level estimated
δ18O curves, which typically range from 1.5–1.9 per mil, are using Plan ktonic and
also influenced to some degree by deep-sea temperature Benthic O data
1b
-150
changes of as much as 4.5°C. This is equivalent to a δ18O 0 20 40 60 80 100 120 140
range in carbonate of c. 1.1 per mil, and clearly represents A g e (ka)
a significant proportion of the overall glacial–interglacial
Figure 3.58 Comparison between estimates of sea-level
isotopic shift. In general, however, the likely tempera-
variations for the last 140 ka based on marine isotopic variations
ture component is significantly less and, when other local and those derived from New Guinea coral reef records (based
effects are taken into account, benthic foraminiferal δ18O on Waelbroeck et al., 2002).
while the lower sea-level value (–130 to –135 m) suggests of numerous empirically derived palaeotemperature rela-
an ice volume excess at the LGM of 52 × 106 km3. tionships, which may produce results that differ by several
Again, both these estimates are consistent with the degrees, even when applied to the same dataset. Planktonic
figures for reconstructed ice volumes described above. foraminiferal data can be affected by a range of factors,
Where the ice-volume component can be extracted from including local changes in ambient temperature, fluctu-
the benthic marine isotope record, therefore, this con- ations in the regional balance between evaporation and
stitutes a useful proxy not only for global ice-volume precipitation, and variations in salinity arising, for example,
variations, but also for glacio-eustatic sea-level change. from river inflow or glacial meltwater discharge. Indeed, it
For example, a long-term record of isotopically derived sea- has been suggested that as much as one third to one half
level change, which links ice volume and sea level, suggests of the total variability in planktonic δ18O may reflect local
that the growth and decay of the Antarctic ice sheet from environmental changes (Mix et al., 2001). Studies of living
~33.5 Ma caused variations of 50–60 m in sea level and that planktonic Foraminifera indicate that the δ18O signal is
this amplitude of sea-level fluctuation also characterized also affected by seasonal temperature variations and by the
the Early and Middle Quaternary, while from c. 780 ka effects of vertical migration of foraminiferal species through
onwards, a much larger range, of the order of ~100 m, is the upper part of the water column. They also show that
evident in the eustatic sea-level record (Miller et al., 2011). the δ18O of living foraminiferal species is much lighter than
that predicted by the equilibrium equations derived for
calcite (Waelbroeck et al., 2005).
3.10.2.4 Sea-surface temperatures and the
These various limitations mean that the reconstruction
marine oxygen isotope record
of sea-surface temperatures using the δ18O record in plank-
In the early years of research on marine oxygen isotopes, tonic Foraminifera is far from straightforward and, where
the dominant influence on δ18O was thought to be water this approach is employed, it requires a careful analysis of
temperature. This was because a second fractionation the constraints that might influence the temperature
process, that which occurs when carbonate is precipitated reconstructions. This includes obtaining a realistic estimate
slowly in seawater is, at least partly, temperature dependent. of the likely ice-volume effect and, in northern oceans, of
The pioneering work of the American scientist Harold the effects of changing sea-surface salinity. An example of
Urey (published in 1947) suggested that former sea-water how this might be achieved is described by Meland et al.
temperatures could be established by measuring the degree (2005) who reconstructed sea-surface salinity changes
of isotopic fractionation that had occurred in marine for the Nordic Seas at the LGM and then employed these
carbonate fossils. Subsequently, Samuel Epstein and others data in the interpretation of the δ18O record to con-
developed an empirical equation for this relationship that struct a map of summer sea-surface temperature changes
gave a value of c. 0.23 per mil per 1°C. In order to use this (see section 4.10.7 and Figure 4.46).
equation to obtain palaeotemperatures, however, it was
necessary to know the second component, namely the
3.10.3 Limitations of oxygen isotope
former isotopic composition of seawater (see above), and
as this could not be established directly, it had to be
analysis
estimated. Cesare Emiliani, who pioneered the strati- There are a number of limitations affecting the inter-
graphic application of oxygen isotope analysis in the 1950s, pretation of oxygen isotope data. These include the
estimated that only about 0.4 per mil of the average 1.7 per following.
mil δ18O variation between glacial and interglacial stages
reflected changes in the isotopic composition of sea-
3.10.3.1 Stratigraphic resolution
water, and hence the more important factor was water
temperature variation. Using Epstein et al.’s equation, he Sedimentation rates vary markedly throughout the oceans.
interpreted the fluctuations in isotopic content of plank- In deeper waters, where the terrigenous influx is often
tonic Foraminifera in cores from the Caribbean and Equa- minimal and biogenic production is dominant, the rate
torial Atlantic as reflecting palaeotemperature changes of of sediment accumulation is usually very low. Although
6°C between the last glacial and the postglacial periods. long records, sometimes spanning the whole of the Qua-
In subsequent years, however, research has shown ternary, can occasionally be obtained from these deposits,
that the relationship between δ18O values in planktonic an average sample for oxygen isotope analysis may span
Foraminifera is more complicated than was initially a time interval of several thousand years. Much higher
envisaged. This is reflected in the publication in recent years temporal resolution can be obtained in sequences that
have accumulated more rapidly, such as those on or close before reaching the ocean floor. Some foraminiferal species
to the continental shelves (Ebbesen & Hald, 2004) or from are more susceptible to dissolution than others, so that
deep-water deposits dominated by IRD (Bond et al., during settling there is a selective removal, usually of the
2001b). However, these sedimentary records are often species that lived closer to the surface. Between 3 and 5 km
very much shorter, and are usually restricted to the last depth, CaCO3 solution equals CaCO3 supply, the level
glacial–interglacial cycle. defined as the carbonate compensation depth (CCD).
Below that depth, only the most robust microfossils arrive
3.10.3.2 Sediment mixing on the seabed. Even where complete dissolution does not
take place, the water depth plus depth of sediment burial
Where mixing has occurred, as a result of bottom-dwelling may be sufficient to lead to diagenesis in the form of car-
burrowing organisms or through the action of turbidity bonate recrystallization (Carter et al., 2000). Stratigraphic
currents, the clarity of the oxygen isotope record will tend studies of foraminiferal assemblages are, therefore, usually
to be blurred. Benthic organisms typically affect sediments restricted to sediments that have accumulated in waters
to a depth of 20 cm as they feed and burrow, although in that are shallower than the CCD, and where evidence of
some cases the depth of bioturbation may be much deeper. diagenesis is absent or negligible.
Bioturbation is especially problematical where sedimen- In spite of these potential limitations, however, oxygen
tation rates are slow, and may lead to attenuation of isotope analysis of deep-sea sediments is a technique of
millenial-scale events registered in the sediment record great importance in Quaternary research. Indeed, in many
(Anderson, 2001). Reworking caused by the scouring of respects it has revolutionized Quaternary science. Although
ocean bottom currents can also result in depositional perhaps not as secure an indicator of palaeotemperatures
hiatuses, while distortions of the sediment and contained as was at one time believed, oxygen isotopes do, neverthe-
microfossil record may occur during sampling. less, provide a unique record of glacial–interglacial cycles,
of changing global ice volumes, and of glacio-isostatic
3.10.3.3 Isotopic equilibrium between test oscillations of sea level. Moreover, the recognition of com-
carbonate and ocean water parable isotope stages from different areas of the world’s
Because of fractionation effects, some species of benthic oceans provides a means of correlating environmental
Foraminifera do not secrete carbonate that is in isotopic changes on a global scale (Chapter 6).
equilibrium with the ocean water that they inhabit. Hence,
when benthic tests are being measured, more import- 3.10.4 Carbon isotopes in marine
ance is attached to species such as Uvigerina senticosa sediments
and Globocassidulina subglobosa that are known to deposit
carbonate in isotopic equilibrium with deep-ocean waters. The analysis of carbon isotopes in marine sediments can
A number of planktonic species also calcify at differ- also provide valuable data on oceanographic changes during
ent depths during their life cycle (e.g. some species of the Quaternary. As with oxygen, carbon has two naturally
Globorotalia), and this can lead to differences in isotopic occurring stable isotopes (13C and 12C) that are fraction-
ratios between adults and juveniles of the same species. ated during a range of natural processes (see section 5.3.2).
Some knowledge of the species dependence of such effects δ13C profiles in marine sediment sequences show cyclic
is therefore required and appropriate corrections can then variations, very similar to those in oxygen isotope traces,
be applied. Certain species are regarded as particularly and it appears, therefore, that these also reflect important
good indicators of the original isotopic ratios of surface environmental changes. They can be used, inter alia, to
waters (e.g. Globigerinoides sacculifer and Globigerina reconstruct changes in ocean circulation patterns, marine
bulloides), while others (e.g. Globorotalia menardii and productivity, air–sea gas exchange and biosphere carbon
Neogloboquadrina pachyderma) are selected because they storage (Oliver et al., 2009). Similar problems to those of
are thought to provide reliable records of isotopic ratios in oxygen isotope analysis are experienced in the analysis
deeper waters. of carbon isotopes, including, for example, variations in
carbon isotope content between microfossils caused by
fractionation as a result of different vital effects, and
3.10.3.4 Carbonate dissolution and
regional variations due to localized circulation changes.
diagenesis
Carbon isotope data in planktonic Foraminifera, however,
After death, carbonate microfossils sink within the water provide information on former productivity changes in the
column, and many will become dissolved or disaggregated upper layers of the oceans and on the flux of 12C in surface
waters. This evidence also offers insights into factors from the GRIP site, reached bedrock at a depth of 3,053 m
affecting atmospheric variations in CO2, for example during in 1993. In the last ten years, drilling has been completed
the last deglaciation (Tschumi et al., 2010), and there- on two new Greenland ice cores: NorthGRIP on the summit
fore forms a complementary source of information to of the ice sheet to the north of the GRIP site where the ice
the ice cores (section 3.11). Benthic Foraminifera record is 3,085 m thick, and at NEEM in the far north of Greenland
deep-water circulation changes in the oceans. Vertical near the original Camp Century core where the ice is over
circulation brings oxygenated waters into the deeper parts 2,500 m in thickness (North Greenland Ice Core Project
of the ocean, a process referred to as ventilation and, at Members, 2004; NEEM Community Members, 2013). The
times of reduced vertical mixing, oxygen levels fall, oldest records from Greenland are from the last (Eemian)
productivity is reduced and this will be reflected in the interglacial, with the NEEM core recording the onset of
δ13C signatures obtained from the fossil records. For interglacial conditions around 130 ka.
example, in the North Atlantic a close correspondence has In Antarctica, where rates of snow accumulation are
been observed between IRD events and the δ13C record, lower than in Greenland, much longer ice-core records
with lighter δ13C values suggesting a ventilation mini- have been obtained. The earliest core was obtained from
mum associated with IRD events (Boyle, 2000). In the Byrd Station in the interior of West Antarctica (1968),
Mediterranean, by contrast, fluctuations in δ13C values in and was followed by successful drilling at Dome C, the
benthic Foraminifera reflect a more vigorous thermohaline Concordia Research Station on the Antarctic Plateau
circulation during the cold stadials of Dansgaard–Oeschger (1979), and Vostok Station at the centre of the East Antarc-
oscillations of the last cold stage (Cacho et al., 2000; Moreno tic ice sheet (1985: Figure 3.59b). During the 1990s, deep
et al., 2005). ice cores were recovered from Taylor Dome on the East
Antarctic ice sheet (1994) and Siple Dome in West Antarc-
tica (1999), these sequences extending back to at least the
3.11 ICE-CORE STRATIGRAPHY last interglacial. Further work at the Vostok site resulted
(in 1998) in the deepest ice core drilled to date (3,623 m),
3.11.1 A brief history of deep-ice coring with a continuous record extending back over 400 ka.
Since the 1960s, groups such as CRREL (United States More recently, the European Project for Ice Coring
Army Cold Regions Research Engineering Laboratory), in Antarctica (EPICA), has recovered a core (known as
the Centre for Ice and Climate in Copenhagen, and the EDC) from Dome C near to Concordia Station with a
British Antarctic Survey have developed specialist coring logged depth of 3,260 m and spanning eight complete
equipment that can recover undisturbed cores from deep glacial–interglacial cycles; the basal ice has an age of around
within the world’s ice sheets. Analysis of deep cores from 800 ka, almost doubling the length of time covered by the
Greenland and Antarctica, and of shallower ones obtained Vostok record. Within the last ten years, further cores
from the smaller ice caps and glaciers, has revealed a record have been obtained from the Antarctic ice sheet. These
of annual increments of snow and ice accumulation include the Kohnen Station core (2005) in Dronning
extending back to the last interglacial and, in some cases, Maud Land (part of the EPICA programme and known
into deeper Quaternary time. as EDML), which provides a very high-resolution record
The first deep-ice core was recovered from Camp of the last glacial–interglacial cycle and represents the
Century in northwest Greenland in 1966, and was followed first direct southern counterpart of the Greenland records;
by cores from Dye 3 (1981) and Renland (1988), also in and the Talos Dome core (2007) obtained by the inter-
Greenland (Figure 3.59a), and from Devon Island (1976) national TALDICE consortium from the edge of the East
in northern Canada. Analyses of these cores not only Antarctic plateau, which extends back over the last two
yielded valuable stratigraphical and palaeoenvironmental glacial–interglacial cycles to around 250 ka (Barbante et al.,
information, but also led to technical improvements in 2010; Buiron et al., 2011).
drilling, core recovery and analytical techniques, as well as In addition to these polar drilling programmes, ice
providing new insights into the behaviour of the great ice cores have also been obtained from glaciers and ice caps
sheets. In the 1990s, two new deep cores were obtained in high-altitude regions in the mid-latitudes. These include
from the summit of the Greenland Ice Sheet. The Greenland cores from the Tibetan Plateau, from Africa (Mt Kiliman-
Ice-core Project (GRIP), coordinated through the Euro- jaro), and from the Andes (Thompson, 2000; Thompson
pean Science Foundation, reached bedrock at 3,029 m in et al., 2013). The records from these ice cores are much
1992, while the North American Greenland Ice Sheet shorter than those from the polar ice sheets and typically
Project (GISP), which was drilled only 30 km or so away contain only a Holocene or late last cold stage sequence.
32°W
a) 48'W b) land ice
81"N 81°N
30'W 0* 30°E
C a m p Century
60" 69°N
Weddell Dronning Maud 60"
NEEM Sea
75'N
75°N NGRIP
D o m e Fuji
Berkener
Island
Summit
GISP2 she 69°N

Scoresby
South
Siple Pole
Renland 6Sund
9°N 90" 90°
Station
69°N Vostok
69°N
Byrd
Dominon Range
Law
Siple D o m e Dome
32°W Dome C
Taylor D o m e
Dye-3
63°N 120' 120'

63°N Talos D o m e
48"W
Figure 3.59 Location of some of the principal ice-core drilling stations in a) Greenland and b) Antarctica.
The data from these various investigations at both high various exotic materials reflecting the vigour of atmos-
and low latitudes have revolutionized our understanding pheric circulation, wind direction and aerosol flux. Other
of the patterns and rates of past global climatic change, and sources of palaeoenvironmental data preserved in annual
of the linkages between the ocean–atmosphere–terrestrial increments of ice sheets and glaciers include trace gases
systems (Chapter 7). (e.g. carbon dioxide and methane) that become trapped in
minute air bubbles within the ice crystals, and which
provide evidence of both short- and long-term changes in
3.11.2 Ice masses as palaeoenvironmental
atmospheric gas composition, and stable isotopes (particu-
archives larly isotopes of oxygen), which not only act as a proxy for
Glacier ice accumulates in a sequence of annual layers, climate change but also offer a basis for correlation between
and analysis of cores from the high-latitude ice sheets has marine and terrestrial records (section 6.3.3). Finally,
shown that these contain a wealth of palaeoclimatic natural and artificial radioactive isotopes contained within
evidence. The annual increments of ice reflect the balance the ice layers provide an independent means of dating ice
between accumulation and ablation over the course of a cores. Ice cores, therefore, contain a range of evidence that
year, and hence the variation in thickness of the annual can be used to reconstruct past climatic conditions and are
layers may provide useful information on, for example, widely regarded as one of the most important archives of
amount of winter snowfall, degree of melting (determined Late Quaternary palaeoenvironmental data (see below).
by summer temperature regimes), and so on. More detailed
palaeoclimatic information can be obtained, however, from
aerosol particles and other exotic material that has settled
3.11.3 Analysis of ice cores
on the glacier and which subsequently becomes incorp-
3.11.3.1 Annual ice increments
orated into the ice layers. These include dust particles from
volcanic or desert sources, a variety of trace substances, Near the surface of an ice sheet, the annual increments of
and microbial or other biological materials (e.g. pollen ice comprise a darker and a lighter component. The winter
grains and fungal spores), the relative abundance of these layers are lighter in colour whereas those of the summer
Figure 3.60 Annual ice layers exposed in the Quelcayya ice

cap, Peru (photograph by Lonnie G. Thompson, Ohio State
University, USA).
melt season are darker (Figure 3.60), the lower rate of

accumulation and partial melting in the summer leading
to a higher concentration of impurities and hence a darker
colour by comparison with the winter layers. In deeper ice,
the annual layers are less obvious because they become
thinner and distorted through pressure from the accumu-
lating overburden, and as a result of flow deformation. They
can be detected, however, using light transmission, X-ray
methods or digital scanners (Figure 5.26), or on the basis
of changes in physical or chemical properties of the annual
ice increments (Figure 3.61). This ‘multi-parameter
approach’ to ice layer counting is discussed further in
section 5.4.3.
Once an incremental record has been established, each
layer of ice can be assigned an age in ice-accumulation
years (number of annual layers below present surface). For
the youngest part of the record, these are equivalent to
calendar years, but with increasing depth, the clarity of the
seasonal increments diminishes, so that it becomes more
difficult to distinguish between individual layers. The depth
in the ice at which ice-accumulation years cease to provide
reliable ages will vary, but in Greenland ice and in some
Antarctic cores, ice-accumulation years have been estab-
lished throughout the Holocene and into the last cold
ECM(ueq. <H*1)
10°
10°
10° 0
(«qdd) [ Q H]
Z
Z
50
20
[Ca *] (ppbw)
100
10
50
(Mqdd) [JHNI
2
-30
6 O (per mill)
0
35
i e
1393.5 1393.7 1393.9 1394.1 1394.3 1394.5 1394.7

GRIP depth (m)
Figure 3.61 Seasonal variations in chemistry, dust content and stable oxygen isotope ratios in ice layers in a section of the
NorthGRIP core. The annual layers (shown by the vertical grey bands) are identified as matching pairs of spring and summer
indicators: spring is characterized by high dust content leading to peaks in Ca2 and dips in the H2O2 curve, while summer is
characterized by high NH4 and corresponding minima in the ECM curve. Note that the ECM and H2O2 curves are plotted on reversed
scales. The annual layer identifications are supported by high-resolution δ18O data (after Rasmussen et al., 2006; copyright 2006
by the American Geophysical Association).
stage. In general, however, dating of the older section of in the CFA conductivity and sulphate profiles, as well as in
ice cores relies on other methods, and again these are con- ECM and DEP measurements (Svensson et al., 2012).
sidered in section 5.4.3.
3.11.3.4 Stable isotope records
3.11.3.2 Dust content
The oxygen isotopic composition of ice is measured by mass
Wind-blown dust particles are an important strati- spectrometry and is expressed as deviations (δ18O) per mil
graphic component of ice cores. Studies of the variation from a standard, which is SMOW (section 3.10; but see note
of dust concentration with depth have shown that dust 10). During evaporation from the ocean surface, atmos-
concentration in the ice varies seasonally, and hence the pheric water becomes depleted in the heavier 18O isotope
measurement of annual dust cycles along an ice core can by, on average, about 10 per mil; however, there is seasonal
help in the identification of annual layers of snow deposi- variability in this value, so that the range of δ18O between
tion. The concentration of dust particles, which typically summer and winter precipitation over the ice sheets is
have a grain size of 0.1–2 μm, can be measured using a commonly about 15 per mil. Thus seasonal changes can be
Coulter (or Elzone) particle counter (section 3.2.3.1), detected in ice cores by precise measurements of oxygen
but more rapidly and effectively by laser light scattering isotope variations. Although the amplitude of the δ18O
(LLS) along the length of the core (Ram et al., 2012). Dust signal decreases with depth due to diffusion effects,
content can also be determined using electrical conductivity significant variations are still detectable over at least the
measurements (ECM) or measurements of the dielectrical last 125 ka (Figure 7.15), and these predominantly reflect
properties (DEP) of ice. These methods involve the trans- changes in regional and/or global climate (section 3.11.4).
mission of an electrical current through the ice, the strength Hydrogen isotopes behave in much the same way as oxygen
being directly proportional to the balance of acids and bases isotopes, the ratio of normal hydrogen to the heavier
present. The current increases with higher concentrations deuterium in atmospheric water (and hence in snow
of strong acids, especially sulphuric and nitric acids, but a accumulation) being determined by saturation vapour
decrease occurs where the acids are neutralized by, for pressure and molecular diffusivity in air. Hence in ice-core
example, alkaline dust from aeolian sources (Taylor et al., records, long-term variations in δD tend to match the
1997). Dust records constitute valuable inputs to climatic trends in the δ18O signal (Sime et al., 2009).
models, and hence are important in palaeoclimatic recon-
structions (sections 7.4 and 7.5).
3.11.3.5 Other trace substances
Seasonal variations in anions and, to a lesser extent, in some
3.11.3.3 Chemical content
cations have been detected in many ice-core sequences,
Although the chemical content of ice can be determined and the changing concentrations of, for example, Ca2, NH4
using ECM and DEP analyses, most chemical analysis and H2O2, enable seasonal ice increments to be deter-
now involves continuous flow analysis (CFA) (Rasmussen mined, and hence are important for the construction of
et al., 2005). This technique is carried out in specially ice-core chronologies (section 5.4.3). Trace gases, such
constructed laboratories on the ice sheet and involves as CO2 and CH4, provide a record of changing atmos-
continuous melting of the inner parts of a newly drilled pheric concentrations, reflecting both natural and anthro-
ice core. Air bubbles are removed from the meltwater pogenic influences; other trace substances, such as soot
and the sample stream is split into a number of analytical particles and heavy metal elements, are evidence of atmos-
systems. Measurements are then made on a range of pheric pollution over both industrial and pre-industrial
parameters, including the amount of insoluble dust and size timescales (see section 3.11.4), while ash layers are evidence
distribution of the particles; the concentration of a range of past volcanic eruptions and form important time-
of chemical ions (NH4, Ca2, Na, etc.); and the conductivity stratigraphic markers in ice cores (Mortensen et al., 2005).
of the meltwater (which is closely related to the ion con-
tent). As the impurities come from different sources,
3.11.4 Palaeoenvironmental significance
they will reveal different information about the climate
system. For example, Ca2 (which is derived mainly from
of ice cores
dust) has an annual cycle, and occurs in the highest Down-core variations in stable isotope content, and in the
concentrations in spring. Impurities in the ice also provide abundance of other trace elements in polar ice, provide
evidence of volcanic activity, and these too are detectable a powerful tool for the reconstruction of the pattern and
amplitude of Late Quaternary climatic change. Moreover, short-lived, each lasting for no more than 500–2,000 years,
because an independent chronology can be established and appear to reflect the operation of a range of feedback
for each ice core, a temporal framework is available for mechanisms involving, inter alia, ice sheet fluctuations,
these continuous stratigraphic records. Where similar oceanographic changes and atmospheric circulation vari-
isotopic signals have been obtained from different cores ations (section 7.4).
within a single ice sheet, this provides compelling evidence In the Greenland NorthGRIP core where the DO cycles
for a regional/global climatic signal. Moreover, as the frac- are particularly strongly registered (Figure 7.15), the
tionation of isotopes is temperature dependent (section isotopic value of the ice at the base of the core differs from
3.10), the stable isotope variations provide a proxy record the present by 3 per mil. If attributed solely to temperature,
for air temperature change. Figure 3.62 shows continuous this would suggest that Eemian temperatures around
δ18O profiles from five drill sites on the Greenland ice 122 ka were at least 5°C warmer than the present day
sheet. A distinctive feature of each of the records is the (North Greenland Project Members, 2004). This inference
sequence of high-frequency oscillations during the course has recently been confirmed by data from the base of the
of the last 120 ka, with shifts in isotopic values of up to most recently drilled core in Greenland, at NEEM (Figure
8 per mil, reflecting amplitudes in air temperature change 3.59a), which indicate that temperatures during the Eemian
of between 5 and 16°C. These ‘Dansgaard–Oeschger peaked at 8 ± 4°C (NEEM Community Members, 2013).
(DO) events’, as they have become known, were relatively In NGRIP, the subsequent slow decline in δ18O values to
C a m p Century Dye-3 GRIP GISP2 Renland
1600 305
ss
1600
1 Gl
1150 1
1800 1 1 1
1800 1
2 1800
2
4 310 2
2
2
1200 1850 2 2000 4
a 4 2000 5
4 8
4
55
12
Depth (m)
12
14 B 8 • 2200
1900 2200 8 8 14
1250 315
19 12¬
12 12
12
14 2400
14 2400 14
21
1950
1300
19;
19 19 19
23
2600 2600 320 21
21 23 23
2000 19; 23 23
1350 2800 2800
Silty ice Silty ice
-45 -40 -35 -30 -25 -35 -30 -25 -45 -40 -35 -30 -45 -40 -35 -30 -30 -25
o i s
O [%o]
Figure 3.62 Continuous δ18O profiles through five Greenland ice cores. Some of the warm Greenland Interstadials (GI) of the
Dansgaard–Oeschger (DO) cycles are defined by reference to the GRIP δ18O record and are numbered to the right of each profile;
these were used to guide correlation between the isotope traces. Some of the cold Greenland Stadials (GS) are also numbered
to the left of the GRIP record (after Johnsen et al., 2001).
115 ka indicates that the last interglacial did not end with 2004), and also that the Antarctic Peninsula may have
a rapid climatic change, but rather with a long and gradual warmed faster than the globally averaged rate of warming
decline in temperature. Within this overall trend, there is over the last century (Barrett et al., 2009).
evidence for abrupt shifts in climate. Before full glacial Interpreting temperature records obtained from ice
conditions are reached, the record reveals an abrupt cool- cores is not without its problems, however (Jouzel et al.,
ing, with the first δ18O decrease at ~119 ka, followed by 1997). This is particularly the case in older records, where
relatively stable depleted δ18O (Greenland Stadial 26) and the deeper parts of the ice sheet may have flowed some
followed, in turn, by a sharp increase at ~115 ka, at the onset considerable distance, and could therefore have originated
of the first interstadial of the DO cycles (Greenland in a different source region from the ice in the upper
Interstadial 25). What the record shows, therefore, is that layers. If there is a significant isotopic gradient between the
the DO cycles, which constitute such a distinctive feature source area for the ice and the present drill site, this could
of the Greenland ice-core record during the course of the be reflected in down-core isotopic variability that is
last cold stage, were registering in the isotopic signal at a independent of temperature. Corrections may have to be
time when the major ice caps were beginning to form. This made for this effect, depending upon the regional isotopic
is puzzling because one explanation for the DO events is variability of the ice. One advantage of the GRIP, GISP2 and
that they were triggered, at least in part, by melting ice or NorthGRIP records, however, is that the drill sites were all
other freshwater input to the North Atlantic (section 7.4). close to the summit of the Greenland Ice Sheet (as also is
An alternative method for obtaining past temperature NEEM) and hence, providing that the ice divide has not
variations from ice cores is to measure the temperature of migrated significantly, flow within the ice sheet at these
layers in the ice directly by lowering specially designed locations should have been minimal. A second problem
thermistors11 down the borehole, an approach termed concerns the relationship between stable isotope ratios
borehole thermometry. As air diffuses through the surface and temperature. Evidence suggests a direct link between
layer of the ice sheet, the ice acquires a temperature attuned stable isotope ratios in precipitating snow and temperature
to contemporary climatic conditions. Because ice is a poor of water vapour at the time of precipitation, assuming
conductor, it retains this temperature after becoming unchanged temperature and humidity in the moisture
buried beneath later layers. Each successive layer of ice sources. Empirical evidence from the Greenland ice sheet
therefore retains a ‘memory’ of the ambient climatic con- shows a clearly defined relationship between δ18O and
ditions at the time of formation (Johnsen et al., 1995). This mean annual surface air temperature, but it is uncertain
can be changed by other factors, such as geothermal heat whether this relationship obtained in the past. One way
flux, the local ice-flow pattern and the rate of ice accum- around this difficulty is to model the borehole tempera-
ulation, but these can be corrected for from a knowledge ture profile using parameters derived from instrumental
of modern surface temperature variations and the age– records of recent temperatures, and the well-dated near-
depth profile of the ice sheet (Dahl-Jensen et al., 1998). surface δ18O profile. The modelled temperature profile
Borehole thermometry generates palaeotemperature recon- can be fitted to the measured temperature profile and these
structions that are broadly in line with those derived from data can then be employed to calibrate the deep-core δ18O
isotopic measurements, and can clearly resolve distinct signal to temperature (Johnsen et al., 2001). Using this
climatic events such as the Last Glacial Maximum (LGM), approach, it has been estimated that Greenland tempera-
Holocene Climatic Optimum, Medieval Warm Period and tures were around 20°C colder than present during the
Little Ice Age (section 7.6), although increasing statistical LGM, and that the DO events were marked by extremely
uncertainties down-core restrict the application of this abrupt temperature shifts of around 15°C. We consider the
method to the last c. 50 ka (Dahl-Jensen et al., 1998). GCM Dansgaard–Oeschger cycles further in Chapter 7.
simulation modelling (section 7.2) confirms that borehole The Greenland ice-core records, therefore, provide a
palaeothermometry not only enables coherent temperature highly detailed record of climate change in the North
estimates to be inferred, but that these may constitute a Atlantic region extending back to the last interglacial. In
correction for isotopic palaeotemperature reconstructions Antarctica, however, ice cores spanning a much longer time
that may be biased by seasonality effects (Werner et al., period have been recovered, the oldest ice so far investigated
2000). Borehole palaeothermometry can also provide other being at the base of the EPICA core and dated to 960 ± 20
palaeoglaciological insights. For example, it has revealed ka (EPICA Community Members, 2004). The climatic
unusually high rates of geothermal heat flux at the base of record therefore extends back over eight glacial–interglacial
the West Antarctic ice sheet, which may explain the cycles. Figure 7.10 shows the inferred temperature profile
locations of some fast-flowing ice streams (Engelhardt, (δD) from the EPICA core, plotted against carbon dioxide
(CO2), methane (CH4), dust flux and aerosols. The highest contaminants from historic and prehistoric metallurgical
CO2 and CH4 mixing ratios are found during the inter- industries (Krachler et al., 2009).
glacials and the lowest during the glacial maxima. The In addition to the above, ice-core data have contributed
100 ka periodicity reflected in these records provides to an understanding of many other aspects of environ-
striking confirmation of the Milankovitch hypothesis of mental change. For example, the dust content of ice pro-
long-term climate change described in section 1.7. How- vides evidence of changes in atmospheric aerosol load-
ever, the remarkable correlation between the CO2 and CH4 ings and of the former extent of arid or poorly vegetated
and the δD profiles in Figure 7.10 suggests that trace gases landscapes. Thus, data obtained from two ice cores in the
are also important as amplifiers of the initial orbital forcing north-central Andes of Peru suggest that during the last
(section 7.3). This record also reflects the dynamics of the glacial stage the atmosphere may have contained 200 times
oceanic and continental biospheric carbon reservoirs in as much dust as today, while a major phase of enhanced
relation to climatic changes. dust deposition around 4.5–4.2 ka marks an extensive
The CH4 signal in both Antarctic and Greenland ice- drought event lasting for 300 years, and which may have
core records is of particular significance. Because the impacted on both climate and culture throughout the
atmosphere is well mixed on timescales of a few years, world (Thompson, 2000; Walker et al., 2012). Variations
variations in the concentrations of methane trapped in in sea-salt concentration in polar ice cores has been linked
bubbles in the ice can be used as global time markers. By to changes in sea-ice formation (Rankin et al., 2004), while
matching the methane records in Greenland and Antarctic ammonium and sulphate in the Antarctic ice have been
cores, northern and southern records can be correlated. used as an indicator of changes in biogenic productivity
This is extremely important to climate scientists and involving algae and bacteria in the Southern Ocean
modellers who are trying to understand the forcing (Kaufmann et al., 2010). Finally, acidity levels (acidity
mechanisms underlying millennial-scale climate change profiles) in ice cores, which can be established by ECM and
and, once again, it is a topic that we discuss further in CFA methods (see section 3.11.3), provide a record of
Chapter 7 (section 7.3). changes in sulphuric acid content, reflecting variations in
Over shorter timescales, the gas content in the upper volcanic aerosols. Peaks in the acidity profiles constitute a
layers of polar ice sheets constitutes an important record proxy record of former volcanic eruptions, and constitute
of recent human activity. This shows, in particular, an in- key marker horizons for the correlation of ice-core records
creasing anthropogenic influence over the last 200 years or (Parrenin et al., 2012).
so leading to present-day levels of atmospheric trace gas
concentration (CO2, CH4, N2O) that are unprecedented
over the past 800 ka (Figure 7.10), or possibly over the past
3.12 CONCLUSIONS
20 Ma (Prentice et al., 2001)! We consider these anthro- Almost all Quaternary sediments contain within their
pogenic records in more detail in section 7.6 in the context matrix important clues about their mode of deposition, and
of recent greenhouse warming. But the polar ice sheets may often about the climatic regime under which the sediments
also provide evidence of earlier increases in atmospheric accumulated. This information can be extracted from the
trace gases that predate the Industrial Revolution; for stratigraphic record by the application of various physical
example, variations in CH4 concentrations over the last and chemical methods, such as those described at the
two millennia have been linked to changes in population beginning of this chapter, and by analogy with sedimento-
and land use during the decline of the Roman Empire, logical processes that can be observed at the present day.
the Han dynasty in China and population expansion in Technological developments in field and laboratory
Europe during the Medieval period (Sapart et al., 2012). methods, perhaps best exemplified by the extraction and
As well as changes in atmospheric trace gas, the ice analysis of cores from the world’s ice sheets, have resulted
sheets contain a record of recent pollution histories; in remarkable advances in our understanding of the
these show soot, lead and other particulates all rising operation of the global climate system over timescales
sharply during the industrial era (Weiss et al., 1999; ranging from decades to thousands of millennia. Equally
Osterberg et al., 2008). Indeed, increased black carbon impressive are the data that have been obtained from
concentrations over the past 150 years have even been loess–palaeosol sequences, where the application of a range
detected in an ice core from Mt Everest (Kaspari et al., of analytical techniques has provided a high-resolution
2011). Trace substance records, such as for lead (Pb) environmental record, in some cases spanning the whole
and arsenic (As) in ice cores also provide a longer-term of the Quaternary. It must be emphasized, however, that
perspective on human industrial activity, including the sedimentary evidence can seldom be properly evaluated
in isolation. For example, in the analysis of glacigenic in which they are found. They may, therefore, provide important
deposits, periglacial sediments or pluvial lake sequences, the clues to components of past vegetation.
stratigraphical record should, wherever possible, be 7 Water-filled passages in bedrock, which are below the ground-
integrated with the geomorphological evidence to produce water table, are termed phreatic passages. Water flow is very slow
and little, if any, downward erosion occurs. Immediately above
a synthesis of landscape or climate change. The same
the long-term water table, passages are only filled with water on
applies equally to the fossil record, the third category of a seasonal basis, or they may contain streams flowing down to
evidence used in the reconstruction of Quaternary the water table. This is termed the vadose zone, in which
environments, and which forms the subject matter of the significant down-cutting can take place.
following chapter. 8 Biological or chemical compounds may contain isotopes of
various elements that can be exchanged during reactions or that
will be subject to fractionation (see note 9) until a stable ratio
NOTES is attained, known as isotopic equilibrium.
1 An electron microscope consistsofofaacathode ray tube
cathode-ray tubethrough
through 9 Fractionation is the selective separation of chemical elements
which a beam of electrons is passed. The electrons are con- or isotopes during natural physical, chemical or biochemical
centrated on, and pass through, the specimen to produce a processes, such as during evaporation, condensation, trans-
magnified image on a photographic plate. After development, piration and metabolism.
the electron photomicrograph shows the structure of the object 10 In the 1990s, a problem arose because the supply of PDB
in terms of its electron density. (Pee Dee belemnite) standard material became exhausted and
2 The term ‘diamicton’ refers to non-sorted terrigenous sediments because SMOW (standard mean ocean water) does not have a
and rocks containing a wide range of particle sizes, regardless unique definition; in other words, there is no globally accepted
of genesis. standard. This means that because laboratories do not use
3 Deformation is the alteration of the primary (original) bedding the same reference material to establish their isotopic ratio
or attitude of lithological units or their components through scales, they are reporting different values for the same material.
stress forces (e.g. compression, extension, or shear caused by In order to eliminate confusion, therefore, the Commission of
traction) resulting in folding, faulting or alteration of internal Atomic Weights and Isotopic Abundances of the International
structures and fabric (see also Chapter 2, note 13). Union of Pure and Applied Chemistry has recommended that
4 Load structures often form where coarse material (e.g. sand) is the use of SMOW and PDB be discontinued, and that isotopic
deposited over a hydroplastic or fluid mud layer. Under the abundances of oxygen- (and also hydrogen- and carbon-)
weight of the sand, the mud layer becomes distorted and bent bearing materials should be reported relative to the reference
downwards. In some cases, the sand layer sinks and forms water VSMOW (Vienna standard mean ocean water) and VPDB
lobes; in others, the mud layer becomes pushed up to form (Vienna Pee Dee belemnite). These are defined by adopting a
of a cathode
tongues or ‘flame’ structures injected into the ray tube through
overlying material. δ18O value of –2.2% and a δ13C value of +1.95 per mil for NBS
5 Diagenesis refers to the alteration of minerals and sediments by (National Bureau of Standards) carbonate relative to VPDB.
the influences of oxidation and reduction, hydrolysis, solution, SLAP (standard light Antarctic precipitation) isotopic abun-
biological changes (e.g. induced by anaerobic bacteria), dance scales should be normalized so that values for 2H (or D,
compaction, cementation, recrystallization, and the alteration for deuterium) and δ18O are –427.5 per mil and –55.5 per mil,
of the lattice structure of clays by expulsion of water and ion respectively, relative to SMOW (International Atomic Energy
exchange. Agency, 2009).
6 Most plants secrete opaline silica bodies and these assume the 11 A thermistor is a sensor that records temperature variations
shape of the cell in which they are deposited. These forms are automatically through electrical resistance measurements which
known as phytoliths, and many are characteristic of the plants are temperature dependent.
4
CHAPTER FOUR
Biological
evidence
4.1 INTRODUCTION Occasionally spectacular fossils, such as skeletons of large

vertebrates (see Figure 4.49) and tree trunks or stumps (see
Biological evidence, in the form of plant and animal Figure 5.21) are preserved. The range of biota that can be
remains, has always been a cornerstone in the reconstruc- identified in Quaternary fossil material is extensive, and
tion of Quaternary environments. The analysis of fossil includes pollen, diatoms, plant macrofossils, insects,
evidence employs uniformitarian principles, namely that
molluscs, ostracods, foraminifers and vertebrates, all of
knowledge of the factors that influence the abundance
which are discussed in this chapter. Other organisms for
and distribution of contemporary organisms enables in-
which more limited palaeoecological data are available,
ferences to be made about the dominant environmental
including Cladocera, fungal spores and certain types of
controls on plant and animal populations in the past.
marine plankton, are considered more briefly.
Applying this approach to the interpretation of Quatern-
Quaternary fossils fall broadly into three size cat-
ary fossil assemblages, therefore, the majority of which have
egories. The largest, termed macrofossils, ranges from
living counterparts, it should be possible to reconstruct
whole skeletons of large vertebrates to component parts
former environmental conditions with a reasonable degree
(e.g. nuts or seeds) or fragments (e.g. pieces of leaf or
of confidence. The use of modern ecological information
wood) derived from plants or animals; these can be identi-
in this way is an essential element of palaeoecology, the
fied by eye or with low-power magnification (up to c. ×40).
study of the interrelationships of organisms in the past,
Microfossils are those organic components that are
both with their physical environment and with other plants
generally less than 1 mm in size and hence require the use
and animals (Battarbee, 2000; Birks & Birks, 1980, 2006).
Variations in the type and diversity of plant and animal of more powerful microscopes for identification. They
remains preserved within sedimentary sequences are also include the skeletal remains of tiny organisms (e.g. diatoms,
used to subdivide the geological record, a field of study Foraminifera, etc.) and any small components derived
known as biostratigraphy. This chapter is concerned almost from larger organisms, such as pollen grains, stomatal
entirely with the palaeoecological aspects of biological cells, epidermal tissue, skeletal platelets or delicate scales.
evidence; the principles and practices of biostratigraphy are A third category comprises the biomolecular products of
considered more fully in Chapter 6. the degradation of organic tissue, such as DNA molecules,
pigments, proteins and cellulose, which can account for a
significant fraction of soil humus or of the fine organic
4.1.1 The nature of the Quaternary fossil muds that accumulate in some mires and lakes (Hofreiter,
record 2008). Sophisticated analytical methods are required
When sediments are disaggregated and closely examined, to detect and identify these submicroscopic residues.
the organic components preserved within them can be Representatives of all three categories may be found
surprisingly abundant and diverse. Even sediments that together within the same organic unit, collectively providing
appear on first sight to be ‘barren’ may contain a rich array an often detailed record of the biological communities
of tiny, but recognizable, plant and animal remains. These from which they were derived, as well as insights into the
might include seeds, fruits, leaves, pieces of wood, char- nature of the sedimentary processes that operated during
coal, insect remains, molluscs, fish scales, teeth and so on. their burial.
182 BIOLOGICAL EVIDENCE
4.1.2 The taphonomy of Quaternary fossil section 4.2). Second, for valid inferences to be made using
the uniformitarian approach, a number of assumptions
assemblages
about contemporary plant and animal populations must
The study of the processes that lead to the formation, hold true. It must be assumed, for example, that the spatial
deposition and burial of a fossil assemblage is referred to distributions and population densities of present-day plant
as taphonomy, and understanding the taphonomy of and animal taxa are fully known, that these reflect clear
an assemblage is an essential prerequisite for palaeo- preferences for specific environmental conditions or niches,
environmental interpretation. Fossils are best preserved and that an equilibrium has been reached between a taxon’s
where deterioration resulting from the activities of micro- distribution and the dominant environmental factors that
organisms and the operation of chemical agencies is influence it. Third, similar assumptions have to be made
minimal, and hence anaerobic environments are more when interpreting the fossil record in terms of what is
likely to lead to survival rather than those where oxidation known about the ecology and distribution of modern biota
has occurred (Briggs, 2003). Some organisms may be found (the modern analogue approach). It is generally accepted,
in the position of growth (in situ fossils), such as tree for example, that the taxa represented in fossil assemblages
stumps buried by the rapid accumulation of peat, volcanic were themselves in equilibrium with the environmental
ash or wind-blown sand. The majority of fossils, however, conditions that prevailed at the time they lived, that plants
appear to have been transported from their growth or life and animals have coexisted in the past in communities that
position by, for example, wind (e.g. pollen and spores), were similar to those than can be observed today (i.e. they
water (a great range of fossils) and animals (e.g. prey of closely match the modern analogues), and that the eco-
carnivorous taxa). These processes can introduce bias logical affinities of plants and animals have not changed
through the selective transport of fossils, or by differential through time. The extent to which these assumptions hold
degradation which selectively removes the less robust true varies both with the type of fossil evidence (pollen
specimens. Following initial (primary) deposition, some grains, animal bones, shells, etc.) and with the nature of
fossils may subsequently become disturbed and remobilized each fossil assemblage (how well preserved, how diverse,
by processes of erosion, bioturbation, solutional collapse etc.). It is rare, however, for all of the above conditions to
or some other agency. As a result, they may be redeposited be satisfied. Indeed, given the frequency and rapidity of
in a new locality and even in an entirely different sedi- Quaternary climate change (Chapter 7), the question arises
mentary context as, for example, when rivers or tidal waves as to whether organisms ever achieve a state of complete
erode terrestrial peats, removing material which may then equilibrium with their environment.
settle in a new repository, such as in a flood plain or In this chapter, the different types of biological evidence
estuarine basin. Fossils remobilized in this way are that are routinely employed in the analysis of Quaternary
commonly referred to as secondary, reworked or derived environments are described and evaluated in the light
fossils. Clearly, it is important to distinguish such secondary of these qualifications. Although the different forms of
components from primary fossils in palaeoenvironmental evidence are discussed separately, organisms are of course
reconstructions using this form of evidence. integral components of ecosystems, and are dependent
upon, as well as in competition with, other organisms for
4.1.3 The interpretation of Quaternary survival (Begon et al., 2006). When attempting to recon-
struct past environments, therefore, it is important to
fossil assemblages
appreciate the complex interrelationships between different
In addition to the taphonomical problems introduced organisms, and hence the wider palaeobiological context
above, other difficulties arise in the interpretation of the of fossil assemblages. This is difficult to achieve in a single
fossil record. First, fossil remains have to be identified proxy study (i.e. one that focuses on only one fossil type).
to a sufficiently low taxonomic level for reasonably well- Increasingly, therefore, Quaternary scientists are collab-
constrained inferences to be made about former environ- orating in multi-proxy palaeobiological studies, where
ments. Whereas some types of fossil (e.g. beetles, diatoms) environmental reconstructions are based on the integra-
can be identified to species level and may show unam- tion of evidence from a range of fossil types (e.g. Birks
biguous ecological preferences, others, such as certain types and Wright, 2000; Kucera et al., 2005a). Moreover, while
of fossil pollen and spores, may only be identifiable to the it is clear that palaeoecologists must have an understand-
genus or family level. This can pose problems where several ing of the principles of modern ecology in order to draw
different species within a genus, or genera within a family, reasonable inferences from fossil assemblages, knowledge
have contrasting or very broad ecological affinities (see of the Quaternary fossil record is equally vital for sound
BIOLOGICAL EVIDENCE 183
interpretation of the origins and spatial distributions of 4.2.2 The nature of pollen and spores
the modern biota (MacDonald et al., 2008). The present
may indeed provide a key to the past, but the reverse Pollen grains (Figure 4.1) are reproductive propagules
is equally true: contemporary ecological theory is best formed in the anthers of the seed-producing plants
served by a secure foundation in Quaternary palaeoecology (angiosperms or gymnosperms). They contain the male
(Willis et al., 2007). Moreover, while the modern analogue gamete of the plant and aim to reach the stigma of the
approach is undoubtedly a valuable vehicle for making female part of the flower where fertilization can take
inferences about the past based on present-day climatic and place. Spores, which are commonly included in pollen
environmental parameters, it may be that contemporary diagrams, represent the sporophyte stage of lower plants
biota are attuned to increasingly unusual or atypical climatic (cryptogams), such as ferns (Pteridophyta) and mosses
circumstances, especially with respect to increasing levels (Bryophyta). The sporophyte is dispersed to suitable
of atmospheric CO2 and the climatic consequences that habitats where the second stage in plant generation, the
follow from this (section 7.6.5.1). In addition, many species gametophyte, can grow. Pollen grains and spores, parti-
distributions have been markedly influenced by human cularly those that are spread by wind (see below), are
activity, especially during the late Holocene (Jackson & frequently dispersed in very large numbers in order to
Williams, 2004). We return to these matters towards the maximize the opportunities for successful pollination
end of this chapter. or gametophyte growth, and many accumulate on the
ground surface or in water bodies. Some will subsequently
become incorporated and fossilized in sediments, and it is
4.2 POLLEN ANALYSIS the extraction, identification and counting of these pre-
served fossil grains that form the basis of the technique of
4.2.1 Introduction pollen analysis.
Most pollen grains and spores are extremely small, few
Of all the palaeobiological methods currently employed exceeding 80–100 μm in diameter, with the majority falling
in the reconstruction of Quaternary environments, un- in the size range 25–35 μm. A typical pollen grain consists
doubtedly the most widely adopted and arguably the most of three elements. The central portion, the living part,
versatile is pollen analysis (or, more correctly, pollen which consists of one or two generative cells and one
stratigraphy1). The technique is also referred to as palyn- vegetative cell, is surrounded by a covering of cellulose
ology. First developed in the 1920s, the method enables the known as the intine. None of these parts survives in the
history of late Quaternary vegetation change to be recon- fossil form. The outer layer or exine, however, consists of
structed at the local, regional, continental and, potentially, a remarkably resistant, waxy coat of material called
global scales (e.g. Williams et al., 2004b; Gajewski, 2008). sporopollenin, a chemically complex substance consist-
It can also be used as a basis for correlating Quaternary ing of biopolymers of carotenoids, phenolics and fatty
stratigraphic units (Magri, 2010), including continental– acid derivatives. The prime function of this outer wall is to
marine correlations (Roucoux et al., 2005), for palaeo- protect the inner, reproductive elements from desiccation
climatic reconstructions (Seppä & Bennett, 2003), and for and microbial attack. It achieves this so efficiently that
evaluating human impacts on late Quaternary vegetation pollen grains continue to be well preserved in sediments
and landscape (Mackay et al., 2003). Pollen records are also when almost all other organic constituents have been
increasingly being used in the formulation of conservation reduced to structureless and unrecognizable components.
policies because of the important historical context that The exine is characterized by a variety of morphological and
they provide (Froyd & Willis, 2008). An extensive literature structural features which, along with the number and
on the principles and applications of pollen analysis is distribution of germinal apertures, and the overall size
now available, and it is not possible to do full justice to and shape of the grain, form the basis for pollen and spore
the method within the space available here. The aim, identification (Figure 4.1).
therefore, is to provide a general introduction to pollen Pollen grains and spores are disseminated by a variety
analysis as a palaeoenvironmental technique, and to outline of means. Spores are usually dispersed by wind, but pollen
the principal strengths and weaknesses of the method. grains can also be spread by water, insects, birds and
More detailed accounts can be found in Birks & Birks animals (including humans). Those plants that liberate
(1980), Faegri & Iversen (1989) and Moore et al. (1991), wind-borne pollen are termed anemophilous, and generally
while useful summaries are provided by Bennett & Willis produce far greater numbers of grains than entomophilous
(2001) and Chambers (2002). taxa, which rely on insects or other zoological vectors for
4
,10
,10
• 9
10.
,10 ,10
,10
J6
,10
I5 10
.1 ,10 ,10
6 _
,10
7
,10
5 1 2 ,10
Figure 4.1 Subsample of a pollen assemblage typical of last cold stage deposits from the site of St Front, France. The arboreal
component is dominated by Pinus (1), while Picea (2) and Betula (3) are also recorded in smaller numbers. Pollen of Cedrus (4)
are atypical of this assemblage, and probably indicate long-distance transport. A wide variety of non-arboreal pollen is represented,
including, for example, Helianthemum (5), Plantago (6), Ephedra (7), Calluna (8) and several genera within the families
Caryophyllaceae (9), Chenopodiaceae (10), Poaceae/Gramineae (11) and Liliaceae (12). The long axis of the Helianthemum grain
(5) is c. 45 μm (photo-montage courtesy of Maurice Reille and Valerie Andrieu-Ponel, Aix-Marseille University, France).
transfer. Wind dispersal is facilitated by the small size, analytical work. They are also found in soils (Davidson
smooth surface features and low specific gravity of the et al., 1999), cave earths (Navarro et al., 2000), cave calcite
grains, while in the gymnosperms such as pine (Pinus) or deposits (McGarry & Caseldine, 2004), ocean floor sedi-
spruce (Picea), air bladders or sacs have evolved enabling ments (Roucoux et al., 2005), glacial ice cores (Liu et al.,
pollen of these taxa to stay airborne for very long periods 1998) and rodent middens in desert regions (Maldonado
and also to travel considerable distances. The entomo- et al., 2005). The degree of preservation of pollen and
philous grains possess a hardy, armoured surface which spores depends upon a range of factors, the most import-
often has prominent spines and a coat of sticky material that ant of which are the chemical composition and grain size
causes them to adhere to each other or to other surfaces, of the host material, the extent to which anaerobic con-
such as the bodies of insects or animals. Generally, these ditions have persisted since deposition, and exine thickness
grains tend either to be large (>60 μm) or very small (<15 and structure. Pollen grains tend to be poorly preserved
μm) and are usually less well represented in the fossil in loosely compacted materials, such as certain types of
record than wind-dispersed types. Pollen production and peats or soils which allow aerobic or microbial attack, and
dispersal is complex, and is considered in more detail in they can also be damaged or destroyed by desiccation or
section 4.2.5. by mechanical abrasion in, for example, coarse-grained
sediments on river flood plains, in estuaries, or close to
points of stream inflow into lakes.
4.2.3 Field and laboratory work Samples containing fossil pollen and spores can be
Pollen grains and spores are frequently well preserved in extracted from sections exposed in river banks, cliffs, road
lake and pond sediments and in peats, and it is these cuttings or building excavations, or by digging pits.
deposits that have been most widely investigated in pollen Alternatively, they can be obtained by coring (Last & Smol,
2001). Samples must be sealed air-tight and are best kept pollen types; this is termed the pollen assemblage or pollen
in a refrigerated store (at c. 2–3°C) to prevent dessication spectrum. Analysis of a series of horizons (or spectra) may
and/or microbial attack. This also protects them from reveal changes in pollen content which may, in turn, be
contamination by pollen circulating in the atmosphere, interpreted as indicating vegetational changes in the area
especially during the pollen and spore production season. adjacent to the site. Pollen counts are displayed as pollen
In the laboratory, following sediment dispersal, sieving diagrams of which there are three principal types: those
and/or chemical flotation (density separation), samples based on variations in the relative proportion of each
are chemically treated in a variety of ways to remove as taxon represented (percentage or relative pollen diagrams:
much of the sediment matrix as possible. Organic residues Figure 4.2); those based on variations throughout the
can be concentrated using oxidation, acetolysis and sequence in the actual concentrations of each pollen and
filtration methods, or by using thermochemical or ultra- spore type (concentration pollen diagrams: Figure 4.3);
sonic techniques (Kim et al., 2003). Minerogenic sediments and those recording changes in the rate of accumulation
can be removed by digestion in hydrofluoric acid or of pollen and spores over specified stratigraphic intervals
separated from an organic matrix by sieving through fine (influx or pollen accumulation rate (PAR) diagrams:
filters, by differential centrifugation or by flotation using a Figure 4.4). The last mentioned are also sometimes referred
‘heavy liquid’ such as sodium polytungstate. Carbonates to as ‘absolute pollen diagrams’.
and calcareous sediments can be treated with hydrochloric In percentage (or relative) pollen diagrams, the pro-
acid. The residues containing the pollen and spores may be portion of each taxon represented in each spectrum is
stained with an organic dye such as safranin, which expressed as a percentage of the selected pollen sum (see
enhances the surface detail of some grains (although this section 4.2.3). Sometimes the data will be calculated as
stage is often omitted as some dyes limit the quality of a percentage of the total pollen and spores counted,
photomicroscopy), and then mounted on glass slides in although this is not preferred by most analysts because not
a suitable medium such as glycerine jelly or silicon oil. all plant types have an equal chance of being represented.
Counting is performed at magnifications of ×100 to ×1,000 In the case of a lake sediment sequence, for example,
depending on the resolution required for identification plants that were growing in or at the margins of the lake
purposes. By traversing the slide in a systematic way, a count (e.g. pondweeds and water lilies) have more chance of
can be made of all of the identifiable pollen and spores until being represented than those growing further away. Hence
a predetermined number (the pollen sum) has been pollen derived from aquatic plants are usually excluded
reached. This should be high enough to account for most from the pollen sum. The same applies to spores, because
of the variability in the spectrum. Pollen sums of 300–500 they are formed in a different way and have a different
grains are usually employed, although there may still be function from pollen grains. In the majority of instances,
significant statistical errors in counts of fewer than 1,000 therefore, the data are calculated on the basis of a ‘total
(Weng et al., 2006). Identifications, which can generally land pollen sum’, that is, the total number of pollen
be made to family level, commonly to genus level but less derived from trees, shrubs and terrestrial herbaceous plants
frequently to species level, are based on distinctive exine (Figure 4.2). Variations in the frequencies of aquatic and
characteristics using pollen keys, photographs and labora- spore taxa are then expressed relative to the land pollen
tory reference collections of modern pollen samples. sum, and are also included on the diagram. In most
Details of the laboratory procedures and examples of instances, however, it is variations in the representation of
classification keys and photographs can be found in Moore the ‘land’ taxa that are the chief concern of the pollen
et al. (1991), Reille (1995) and Colinvaux et al. (1999), and analyst. Indeed, where researchers have been especially
in pollen atlases (Hooghiemstra & van Geel, 1998). Because interested in forest history, they have often chosen to use
pollen counting can be time-consuming and subjective, an arboreal pollen sum as a basis of calculation. In this
attempts have been made to automate the process of pollen approach, all pollen and spore types are counted until a
classification (Zhang et al., 2004a), though these have yet specified number of tree taxa have been identified, and each
to generate reliable procedures for routine use. taxon is expressed as a percentage of the tree pollen sum.
This style of presentation is still occasionally used in
archaeological investigations and in studies of Holocene
4.2.4 Pollen diagrams landscape change, where the focus is on the changing
Where samples have been taken from a stratified body composition of forest communities (Booth et al., 2004).
of sediment, such as a lake or peat sequence, the pollen Where sediments of the cold stages of the Quaternary are
content of a single horizon will reveal a mixture of being investigated, much lower frequencies of tree pollen
uso
du
s
u seci
s
uosu
us
dou
uso d us
ou
iu
us
cd
uo
du
cricduoduuso
cei
i
e
id
e
c
d
sded
ec
d
a
s
ucsu
u
la
l
l e
recr la
us
l
u
tu sQu ideuci
la
u
e
la
tu
u
rc
e la u s
Q
B l tue a
u lat
cu d
tu a
tu la
rc
u
tu
Q ruceursc ous
QQ
Beuercu udsecd
B tu a du
ue ula
er la
Beetuldeci
B tul
B cus
er us
e
B et a
Qu tu cid uou
Be s de ecid
Betula
B ercua cid
Berc la iduo
B tula de us
Beercu ciduo
B et la ous
Q etu s d uou
B s
Quetuus d
B tula
B et ula
Beetu
Be a
Qu
Q et
Beetul
Be
B
Quetu dec
B et ula
Be
Qu etuula
Qu erc
Qu
3
4 1-284-1
5
S I-284-2
7
I-284-3
e
s I-284-4
10
11
I-284-5
12
13
Depthhmm
14 I-284-6
15 [-284-7
I-2B4-8
16 -284-9
17 1-284-10
18
19
20 1-284-11
21
22
23 1-284-12
0 0 0 0 0 20 40 0 0 20 0 0000 0 0 0 00 000 0000 4040 80 0 0 00 0 0 20 0 0 0 20 40 60

%
Figure 4.2 Relative pollen diagram showing variations in abundance of principal taxa in a lake sediment sequence (Core I-284) at Ioannina, northwest Greece. The
data are based on a pollen sum of all terrestrial trees, shrubs and herbs. AP– total arboreal pollen; NAP – total non-arboreal pollen. The sequence spans approximately
the last 20 ka (modified from Lawson et al., 2004).
are usually recorded, and a ‘land pollen sum’ is considered a site will automatically lead to a suppression of the
more appropriate. In long sedimentary sequences spanning percentages of other taxa represented in the pollen diagram.
several glacial–interglacial cycles, simplified indices such as In the upper part of Figure 4.2, for example, where marked
the ratio of arboreal to non-arboreal (AP/NAP) pollen may reductions in percentages of deciduous Quercus (oak) and
be employed to emphasize the changes in forest cover Ostrya (Hop-hornbeam) coincide with an increase in
associated with successive warm and cold stages (Tzedakis Gramineae (grass) pollen, the question arises as to whether
et al., 2004). the rise in grass pollen percentages reflects a real ecological
One problem arising from the representation of change, that is, an expansion of open grassland habitats at
pollen data in percentage form is that the curves for the expense of trees, or whether it represents a decline in
individual taxa are, of necessity, interdependent. In other tree pollen influx (e.g. from a reduction in flowering) and
words, an increase in the influx of one type of pollen to hence is a statistical artefact of that process.
us
tul
usl be
s
tu us
ou
s be pin
du
lu s us s
ulu s a rs
tu lu l n u
pitu Cul
ci
be etu m tu t
be ar e
de
m us b m th m m m s h m uCs b be
h m
pt th th m in nus th eppth th h m pth h m u
n pt s
us
i n u
e p
r pi p D e p t e t rp e rp
i i n
rc
D ep ep a
C ar De e
D D ep D De p rp
Ca D Ca
ue
D D C D Ca
Q
3
1-284-1
4
S I-284-2
1
I-284-3
8
9 I-264-4
1U
11
I-234-5
1?
13
14
I I-2B4-6 I
I-284-7
15
1-284-10
16 I-2 84-9
17
1-284-10
18
19
1-284-11
20
21
22
23 1-284-12
0 0 0 5 0 0 0 0 0 0 0 0 5 0 0 0 0 5
x 10^grains cnrP
Figure 4.3 Pollen concentration diagram for some of the taxa represented in relative proportions in Figure 4.2 (from Lawson
et al., 2004).
One way of solving this problem is by the use of pollen at different levels in a profile can be depicted diagram-
concentration or pollen influx diagrams described above, matically (Figure 4.3).
as these are not based on percentage changes, but on Although this approach will overcome the difficulty of
changes in the total number of pollen grains per unit statistical interdependence of taxon curves, and is there-
volume of sediment. A number of techniques are available fore potentially of considerable value in palaeoenviron-
for estimating pollen concentration, but the one most mental reconstruction, it too has problems of applica-
commonly employed involves the addition of a known tion. Variations in rates of deposition within a body of
quantity of exotic pollen grains (sometimes referred to as sediment will affect pollen accumulation, and hence in a
a ‘spike’) to the fossil sample during laboratory prepara- single profile, concentrations will vary between different
tion. In northwest Europe, tablets containing known types of sediment. In Figure 4.3, for example, pollen
amounts of Eucalyptus pollen are commonly used for this concentrations are generally low in the lower part of the
purpose, since the plant is exotic to the region and the diagram, increase markedly at c. 16 m, where a change in
pollen can therefore be easily distinguished from fossil sediment type reflects the transition from the last cold stage
grains. The spike is added to measured volumes of sedi- to the Holocene interglacial, and decline abruptly at c. 9 m,
ment and hence becomes mixed with the fossil pollen on where there is a further change to shallow-water type
the slides prepared from the residues. Since the quantity sedimentation. The extent to which the fluctuations in
of exotic material added to the sample is known, then the pollen concentrations reflect variations in pollen influx or
observed ratio of exotic to fossil pollen enables the total in the rate of sediment supply is difficult to determine from
number of fossil pollen per unit volume of sediment these data alone. Supposing, for example, that both pollen
to be calculated, and thus changes in pollen concentrations and host sediment (clay, mud) are delivered to a lake basin
us us us
ul ul ul
eluts sbuelut s us eluts mus
bu l ut us mlus ul b
m m est tuubse m t ul h u et stu thtul
nbu
m b
s pus
e
i n
thmbe t t m m b nbue pm be
mDtheuspmb th m pth
h psi e psi
m th m u
s
pt pth h m ainru th th m inCrupainr th mentphush m h pth pin
r
ain u Dt hues
De De ept CarCp Dep ep CarCp a ep aeDrppiept epth erppinDep De ept Deep Car AP rp erppi
C n
2200 D D D CD D D CDa D D AP Ca D Ca
2200
0
22120000 ST-8
MMS SO O
2200
2200
1000 ST-7
2300
2350
200O
2400 ST-6
3000 2450 [sT-6
25C0
•WOO
2200
S00O 2eco
265 a
6000 ST-4
12
7 02 00 0
2200
700O
2800
2350
BOOO
2600
ST-3
2 " 3
900O
I000O 3000 ST-2

1100O
;05<J
20 40 80 180 40 2040 20 4 02 8
200 22024000 20 20 20 40 80 20 20 100 200 3 0 0 tOO 200 20 300 600
K100 KlOOO *1000
Figure 4.4 Pollen accumulation rate (PAR) data for a sequence spanning the last 11 ka at Steel Lake, central Minnesota, USA
(from Wright et al., 2004).
at constant rates, then the concentration of pollen will during the preceding glacial episode (Figure 4.5). For many
remain uniform. But pollen concentrations would increase years, it was the convention in published pollen diagrams
if either the supply of pollen increased or the rate of to show all pollen taxa counted, even where only single
sediment decreased, and vice versa. grains were recorded. More recently, however, it has
In order to resolve this uncertainty, therefore, the rate become the practice to publish diagrams showing only
of sediment accumulation must also be determined, so that the principal pollen taxa, as these fit more easily onto the
‘absolute’ changes in concentration per unit measure of printed page, and they focus attention on the key elements
time (pollen influx) can be established. This requires of the pollen record. The complete pollen counts may
an independent chronology, which is usually provided by appear online as supplementary information, or are usually
means of radiometric dating (Chapter 5). A series of age available on request from the author.
estimates from the profile enables the rate of sediment Diagrams are normally divided into a series of pollen
accumulation between dated horizons to be determined, zones, which are distinct stratigraphic intervals that are
and the pollen concentration diagrams can then be characterized by broadly similar pollen composition.
converted into pollen influx diagrams. These express the Each zone is classified according to the dominant taxa, and
data in the form of number of pollen grains per square highlights what are considered to be the key changes in
centimetre of surface of sediment per year or for some other vegetation cover reflected in the pollen record over the
unit of time. Again, however, care needs to be exercised time interval represented. Until recently, pollen diagrams
when interpreting pollen influx data. In Figure 4.4, for were traditionally zoned subjectively, but it is now more
example, the record appears to indicate an increase in the usual to employ multivariate statistical methods to avoid
influx of arboreal taxa, notably Pinus (two species) and ‘operator bias’. These include, for example, agglomera-
Betula (birch), after c. 3.4 ka. As is the case with many dated tive ordination methods, such as principal components
pollen records, however, the age scale has been derived from analysis, correspondence analysis and cluster analysis,
statistical curve-fitting through a series of radiocarbon which classify data according to degree of similarity as well
dates, and is based on the assumption that the rate of as ‘dissimilarity’ (hierarchical partitioning techniques),
sedimentation between the dated horizons was constant, and which enable a series of pollen spectra to be subdivided
something that can seldom be established unequivocally. into discrete groups (Dale & Dale, 2002a). As Moore et al.
Furthermore, in many palaeoecological studies, the (1991) have noted, however, all zonation systems are merely
constructed chronology is frequently based on a relatively aids to interpretation and even an ‘objective’ system of
low number of dates, all of which have statistical zonation could be misleading if accepted without critical
uncertainties (Chapter 5). In the case of calibrated radio- knowledge and appraisal of the ecological affinities of the
carbon dates, for example, the statistical errors are taxa under consideration. Although some have questioned
frequently centennial in scale (Telford et al., 2004). Before whether pollen zones need to be defined at all, there is a
interpreting pollen influx data, therefore, it is important general consensus amongst palaeoecologists that in order
to be aware of possible limitations in the age estimates and to summarize, and ultimately to make sense of, the large
in the age–depth model upon which the influx calculations body of complex multivariate data that is contained within
are based. a pollen diagram, some form of ordination scheme is
Pollen diagrams, whether percentage or ‘absolute’, are essential (Bennett, 1996).
constructed in different ways, depending partly on tradi- Most approaches to pollen zonation continue to follow
tion and partly on objective. Many diagrams show principles first introduced by Cushing (1967), in that the
arboreal pollen on the left, and move through shrub, pollen diagram is first divided into local pollen assemblage
herb and aquatic taxa, with spores completing the record zones, usually on the basis of the principal terrestrial taxa.
on the right-hand side. In other instances, the taxa may Regional pollen assemblage zones can then be established
be arranged in alphabetical order or by plant families on the basis of observed common features between pollen
within the broad divisions of tree, shrub and herbaceous diagrams obtained from neighbouring sites. These regional
plants. In records from temperate regions, particularly pollen zones therefore typify the suite of changes in
from interglacial sequences, it is common to plot tree vegetation that characterize given intervals, and examples
genera in the order in which they first appear, the advan- include those for New England, USA (Shuman et al., 2004),
tage being that this may reflect the order of immigration the Netherlands (Hoek, 1997) and North Island, New
of trees following their elimination from the landscape Zealand (Wilmshurst et al., 2004).
T e m p e r a t u r e (*C) P o l l e n g r a i n s (%)
-20 0 20 50 100 0 20 20 40 20 20 1U°S££ 20 40 60 10 20 40 S 20 20
12.0
Melisey 1
12.5
1
2
13.0 3
4
Eemian interglacial
13.5 5
14.0
7
8
14.5 9
10
11
15.0
Corylus
Tilia
Lithology
Taxus
Depth (m)
Satix
Annua]
Trees
Winter
Summer
Herbs
Samples
Stratigraphy
Hectera
Betula
litmus
Pinus
Buxus
Abies
Pivea
Corylus
Carpinus
Quercus
Fraxinus
Artemisia
er
Corylus
Juniperus
Gramineae
u f A
I
Silt m u d Fine-detritus mud
Figure 4.5 Arboreal pollen taxa recorded for last (Eemian) interglacial deposits at Jammertal, southwest Germany and
palaeotemperature inferences based on the mutual present-day climate tolerances of the taxa. The grey horizontal bars, numbered
to left, mark the positions of recurring short-lived cold events (from Müller et al., 2005).
4.2.5 The interpretation of pollen which are self-pollinating and which liberate very few
pollen grains into the atmosphere, while in cleistogamous
diagrams
plants (e.g. Viola), the flowers never open and thus pollen
The interpretation of a pollen diagram is undoubtedly is very rarely released. Within each of these plant types,
the most difficult part of pollen analysis, for it requires however, there is considerable variation. Flowers of the lime
a knowledge of pollen production and dispersal, pollen (Tilia cordata) and ling (Calluna vulgaris), for example, are
source and deposition, pollen preservation and the rela- both insect-pollinated yet they usually liberate large
tionship between fossil pollen and former plant com- quantities of pollen. On the other hand, beech (Fagus
munities. Only when these have been carefully evaluated sylvatica) and oak (Quercus petraea) are both wind-
can inferences be made about former vegetation cover pollinated, yet are often relatively low pollen producers.
and, by implication, about former climates and environ- Table 4.1 gives some indication of the general variability in
ments. pollen production of some taxa that are commonly
First, it is important to appreciate the fact that not all encountered in pollen diagrams from Europe. Although
plants produce the same quantities of pollen. It has already there are marked differences in the absolute quantities of
been noted that entomophilous species usually produce pollen produced by these plants in different areas or
much less pollen than anemophilous plants and will vegetation associations, the rank order between them is
therefore be under-represented by comparison in modern often similar (Moore et al., 1991). Indeed, some analysts
surface samples and in the fossil record. Less well repre- have attempted to derive correction factors (or R-values)
sented are the autogamous plants such as wheat (Triticum) for tree and shrub taxa based on such rankings (e.g. Sugita,
Table 4.1 Estimates of the pollen production of various plant species. The index of relative pollen production in
the final column is based upon estimates of the pollen production of an individual plant over a period of fifty years
and is expressed relative to the estimated production of Fagus (beech) over the same period, which is used as a
standard (from Erdtman, 1969).
Species Number of pollen Number of pollen Number of pollen Index of relative
grains per anther grains per flower grains per catkin pollen production
(cf. Fagus = 1.0)
Trifolium pratense 220
Acer platinoides 1,000 8,000
Malus sylvestris 1,400–6,250
Calluna vulgaris 2,000 tetrads
Fraxinus excelsior 12,500
Secale cereale 19,000 57,000
Rumex acetosa 30,000 180,000
Juniperus communis 400,000
Pinus sylvestris 160,000 15.8
Picea abies 600,000 13.4
Betula pubescens 6,000,000
Alnus glutinosa 4,500,000 17.7
Quercus robur 1,250,000
Fagus sylvatica 1.0
Quercus petraea 1.6
Carpinus betulus 7.7
Betula pendula 13.6
Corylus avellana 13.7
Tilia cordata 13.7
1994; Odgaard, 1999), based on the premise that taxa with transport mechanisms by which the pollen are transferred
high R-values are likely to be over-represented in the pollen from source to eventual point of deposition. Pioneering
record, while those with low values will be under- investigations by Tauber (1965) suggested that, in a forested
represented. Attempts to apply scaling or weighting ratios region, airborne pollen arrives at a bog or lake surface by
to correct for statistical biases caused by differential pollen one of three pathways: either through the trunk space,
production have proved problematic, however, as other through the forest canopy, or from raindrop impact.
factors (considered below) affect the mix of pollen delivered Factors such as wind speed through the trunk space and
to sedimentary archives. Alternative approaches to canopy, the density of woodland cover, thickness of
calibrating fossil pollen spectra include modern analogue foliage, time of pollination of the trees, and the size, shape
studies, which test for characteristic pollen suites associated and proximity of a bog or lake surface to pollen source,
with particular vegetation zones or niches (Whitmore et al., will all play a part in determining the composition of the
2005), and simulation models of pollen dispersal and pollen deposited. Since Tauber’s innovative work, num-
deposition (Sugita, 2007). erous studies have investigated the complexities of pollen
Second, it is necessary to know something about the dispersal and recruitment and their influence on pollen
source of fossil pollen in a body of sediment. It is import- deposition in different sedimentary contexts (e.g. Nielsen
ant to establish whether plants were growing on the & Odgaard, 2004). Contemporary pollen deposition can be
bog surface or within the lake basin, around the margins monitored by analysing soil surface samples (Elenga et al.,
of the site, in the immediate vicinity or some distance 2000) or moss polsters collected from bog surfaces
away. Moreover, it is necessary to know something of the (Wilmshurst & McGlone, 2005b), or by the use of pollen
traps designed to capture pollen from the atmosphere, presence of arboreal pollen in contemporary spectra from
lake surface or sediment–water interface (Gosling et al., the Greenland tundra (Rousseau et al., 2006) and the
2003). These approaches provide valuable quantitative high alpine heathlands of Fennoscandia (Hicks et al., 2001).
data on pollen compositional variation at the local, regional Far-travelled arboreal pollen has even been found in
and continental scales (see section 4.2.6), and offer a basis high Arctic ice cores (Bourgeois et al., 2000) and in small
for calibrations to correct for over- or under-representation lakes in East Antarctica (Bera, 2004). These and other
of plant taxa in the pollen record (Sugita, 2007). studies indicate that upper air currents often contain a
The mix of airborne pollen deposited at any one locality ‘background’ component of long-distance pollen that
will also depend upon the surface area of a site in relation can potentially contribute to local or regional pollen
to surrounding vegetation. Pollen influx to a small lake assemblages. This will tend to be insignificant (perhaps
basin surrounded by a dense tree cover will be dominated undetectable) where the local vegetation cover generates a
by pollen from the immediate vicinity, whereas a large high pollen influx, but in poorly vegetated polar or subpolar
open lake or bog surface will receive a higher proportion regions where local pollen influx is low, this far-travelled
of airborne pollen derived from a larger regional catchment component may register more strongly and may be a
or from even further afield (Giesecke & Fontana, 2008). feature of, for example, pollen assemblages in Quaternary
Analysis of sediment samples from the former site will, cold stage deposits in mid- and high-latitude regions.
therefore, provide evidence of local vegetation composition, A third factor to be considered when pollen diagrams
while data from the latter will be more likely to reflect the are being interpreted concerns the nature of pollen
regional vegetation cover. According to Sugita (1994), the deposition. Differential settling velocities of pollen in
dominant source area reflected in pollen assemblages lakes and ponds, coupled with the disturbance of sediment
accumulating in small woodland hollows of 2 m radius lies on the lake floor, either by currents or by burrowing
within 50–100 m of the hollow, whereas the equivalent organisms, can lead to complications in the fossil record.
sources lie 300–400 m and 600–800 m from the edges of Equally misleading can be the occurrence of redeposited
sites with radii of 50 and 250 m, respectively. Transport by or secondary pollen that has been washed into the lake by
inflowing streams is also an important process in pollen stream flow, overland flow, solifluction or collapse of the
recruitment in lake sediments (Brown et al., 2007a), and basin edge sediments, and the subsequent redistribution
in some lakes in temperate regions, up to 90 per cent of the of material across the lake floor (Campbell, 1999). These
pollen accumulating on the lake bottom may be derived grains will clearly be of a different age from those arriving
from inflowing streams and groundwater (e.g. Bonny, at the lake surface from the atmospheric pollen rain, and
1980). Data from Africa show that river-transported pollen although they can often be distinguished from the primary
is especially high around deltas (DeBusk, 1997), while in pollen by signs of exine deterioration (see below), they are
China, pollen transport by streams flowing from mountain potential sources of confusion in the interpretation of the
areas into lowland lakes has generated pollen spectra that biostratigraphic record.
are markedly different from those obtained from plant In general, there are fewer uncertainties where mire
communities in the lake hinterlands (Chen et al., 2006). and bog sites are used in preference to lakes, although
There will also be a local pollen input from aquatic plants here too complications may still arise. Studies have shown
growing in the lake or from mire plants growing on a bog that when pollen arrives on the surface of a bog, there
surface or at the edges of a lake. Finally, there may also may be a tendency for both lateral and vertical mixing to
be a secondary component from pollen which has been occur, with the larger grains remaining on the surface
deposited around the catchment and which has subse- while smaller grains may migrate downwards into the peat
quently been remobilized and incorporated into lake (Clymo & Mackay, 1987). However, these movements are
sediments at a later date (see below). believed to be relatively insignificant when set against the
Studies of contemporary pollen dispersal in relation to timescales usually involved in peat accumulation. More
species composition of the vegetation around the samp- problematical is the behaviour of pollen and spores in
ling site suggest that most wind-borne pollen is deposited soils. A major difficulty with soil pollen analysis is that
within a few kilometres of its source, and only a very small the processes of leaching and capillary action will have the
proportion of the pollen grains liberated into the atmos- effect of moving fossil grains up and down the profiles
phere is likely to travel very far (e.g. Seppä & Bennett, 2003). (Davidson et al., 1999). Mixing by earthworms and other
In certain circumstances, however, and under particular soil organisms further exacerbates the problem, although
synoptic conditions, some tree pollen grains may travel discrete pollen assemblages have been detected where
over considerable distances as reflected, for example, in the earthworm populations have burrowed to progressively
shallower depths as soil profiles have developed (Miura, tion) or bacterial attack, and the latter by mechanical
1987). The analysis of pollen assemblages in cave sediment abrasion, rupture or compaction during transportation,
is also considered problematic, partly because of the settling and burial. Four categories of deterioration in
possibility of downward migration of pollen in coarse- fossil pollen grains are commonly observed: corrosion,
grained deposits (Bennett & Willis, 2001), and also because where the exine is pitted or etched; broken, where grains
cave animals may disturb cave-floor deposits by trampling, are ruptured or split, or where pieces have completely
or, in the case of bats, introducing pollen foraged from broken away; crumpled, where grains are folded, twisted
vegetation far removed from the cave (Hunt, 2005). or collapsed; and degraded, where the structural elements
Fourth, many fossil pollen grains show signs of deteri- are fused together presenting a ‘solid’ or ‘fossilized’ (waxy)
oration resulting from physical, chemical and biological appearance to the grain. Fossil pollen grains frequently
attack on the exine. Experimental work has shown that display more than one of these deterioration characteristics
pollen and spores vary in their susceptibility to such (Tweddle & Edwards, 2010).
processes as oxidation and corrosion (Table 4.2). Some The type of deterioration is largely a function of the
grains, for example spores of the clubmosses (Lycopodium) nature of the depositional environment (Twiddle &
and certain ferns (Polypodium), show remarkable resist- Bunting, 2010). Corroded grains usually reflect oxidation
ance to deterioration, while others, such as the more in poorly compacted peats or soils subject to periodic
delicate grains of nettle (Urtica) and poplar (Populus), drying, while degraded grains are frequently indicative of
may be destroyed altogether. As a consequence, some secondary deposition, the exine surfaces having under-
pollen types tend to be under-represented in the fossil gone structural modification through reworking. As such,
record while others may be over-represented. Damage to deteriorated pollen records may provide useful corrob-
pollen may be chemical or physical in nature, the former orative information on local environmental conditions,
caused by groundwater solutions, exposure to air (oxida- and may also reveal the occurrence of redeposited pollen
(Wilmshurst and McGlone, 2005; Tweddle & Edwards,
2010). Pollen in soils and loess deposits are sometimes so
Table 4.2 Corrosion and oxidation susceptibility of poorly preserved that identification becomes difficult or
selected pollen and spores (from Havinga, 1964). even impossible (Tomescu, 2000), but even in undisturbed,
a) Sequence of increasing corrosion susceptibility waterlogged fen deposits some degree of pollen degradation
of selected pollen and spores may be encountered (Jones et al., 2007a). Given that high
Lycopodium LOW levels of degradation might have led to the preferential
removal of important plant types from the record, analysis
Conifers
of the amount and type of pollen degradation should
Tilia
perhaps be more widely employed in fossil pollen analysis.
Corylus Finally, there is the vexed question of how far it is
Alnus, Betula possible to relate pollen assemblages to plant communities,
Quercus and how far we are justified in making inferences about
Fagus HIGH former climatic and environmental conditions on the basis
b) Sequence of increasing oxidation susceptibility of pollen analytical data. It is now generally accepted that
of selected pollen and spores many former plant communities, especially those dom-
Lycopodium clavatum LOW inated by herbaceous taxa, which were characteristic of
large areas of the mid-latitude regions of the Northern
Polypodium vulgare
Hemisphere during the cold phases of the Quaternary,
Pinus sylvestris
have no analogues in the modern flora (West, 2000).
Tilia spp. Research in both western Europe and North America has
Alnus glutinosa, Corylus avellana shown that at times of climatic stress, each plant type
Betula sp. responds individually, so that even though successive warm
Carpinus betulus or cold stages were characterized by broadly similar climatic
Populus spp.; Quercus spp.; Ulmus spp. conditions, quite different plant associations developed
(Williams et al., 2004b). In some instances, it seems, plant
Fagus sylvatica, Fraxinus excelsior
communities may never have been in equilibrium with
Acer pseudo-platanus
prevailing climatic conditions. Pollen records from the
Salix spp. HIGH Colombian Andes, for example, indicate that low-latitude
montane ecosystems were much more dynamic in the years, a considerable international effort has been directed
recent geological past than previously realized, and here it towards the development of regional pollen databases.
seems that the composition of the vegetation changed with Under the aegis of approved nationally funded data centres
such relative rapidity and frequency that stable plant (such as NOAA in the USA), or international collaborative
communities rarely, if ever, existed (Hooghiemstra & van initiatives (e.g. the European Pollen Database), pollen
der Hammen, 2004). records are held in standardized formats in electronic
These difficulties are further compounded by the archives for use by the global scientific community. The
limitations imposed on palaeoecological inferences by data can be remotely accessed and interrogated in a variety
the taxonomic imprecision of pollen identification. Under of ways. Individual site records can be downloaded, or
normal microscopy, it is occasionally possible to identify pollen data can be synthesized to enable reconstructions of
pollen grains to the species level; distinctions can usually vegetation change to be made at a range of spatial scales (see
be made, for example, between species of plantain below).
(Plantago), saxifrage (Saxifraga) or clubmoss (Lycopodium).
More frequently, identifications are made only to the
4.2.6.1 Local vegetation reconstructions
generic level, for example in identifying pollen of tree birch
(Betula), willow (Salix) or mugwort (Artemisia), while in Tracing the course of local vegetation developments has
other cases it is often difficult to subdivide beyond the been, and still remains, a central theme of pollen analysis.
family level. Grass (Poaceae) and sedge (Cyperaceae), for Pollen records from peat and lake cores enable inferences
example, are rarely taken to the generic or specific level. to be made about the history of a particular peat bog or lake
However, considerable progress has been made in recent ecosystem (Woolfenden, 2003), while in sites where pollen
years in the quality of instrumentation (e.g. SEM analysis diagrams have been obtained from a number of different
and optics of higher resolution in standard microscopy) profiles (‘three-dimensional pollen analysis’), changes in
while detailed taxonomic studies have provided clearer the local vegetation cover can be mapped through time (Bos
diagnostic criteria for identifications to lower taxonomic et al., 2006). In coastal regions, pollen analysis can help
levels (Reille, 1995; Punt et al., 2009). Despite these elucidate the history of sea-level change, as positive and
technical and methodological advances, however, pollen negative sea-level tendencies (section 2.5.2.2) will be
diagrams remain a data bank at a variety of taxonomic reflected in local pollen records by changes between
levels, and this obviously imposes major constraints upon saltmarsh and terrestrial or freshwater plant communities
the reconstruction of former plant communities, particu- (Roe & van de Plassche, 2005). Understanding the nature
larly as some plant families and genera include species with of local vegetation changes is a prerequisite to establishing
markedly contrasting ecological affinities. the scale and pattern of regional changes in vegetation, as
The reader may be forgiven for thinking that these well as the impact of humans on vegetation communities
difficulties in the interpretation of pollen data render the (section 4.2.6.4).
technique of dubious value to the analysis of Quaternary
environments. That this is clearly not the case is demon-
4.2.6.2 Regional vegetation reconstructions
strated by the remarkable degree of consistency in the
large number of pollen-based research publications that For many years, pollen data have constituted one of the
have appeared in the Quaternary literature in recent years. principal lines of evidence for reconstructing vegetational
Some of the many important applications of pollen analysis history at the regional and extra-regional scales. Not only
are considered briefly in the following section. does the technique enable large-scale vegetation patterns
to be established, it allows the history of both individual
species and entire vegetational assemblages to be traced
4.2.6 Applications of pollen
through time (Webb et al., 2003a). Regional biotic catas-
stratigraphy trophes may also be reflected in pollen records including,
In the ninety or so years since the technique was developed, for example, the Tsuga (hemlock) decline at c. 4.9 ka in
fossil and modern pollen data have been published from a eastern North America (Bhiry and Filion, 1996) and the
large number of sites around the world. While the density Ulmus (elm) decline at around 5.8 ka in western Europe
of records is highest in Europe and North America, (Parker et al., 2002), both of which have been attributed to
numbers from other parts of the globe are increasing disease, although in the case of the elm decline, human
rapidly. This enormous body of data is difficult for indi- activity may have been a contributory factor (Peglar &
vidual researchers to store and to collate and so, in recent Birks, 1993).
4.2.6.3 Space–time reconstructions 18 ka 12 ka
Where a network of securely dated pollen records exists,

maps can be compiled showing changes in vegetation
pattern through time. These may be based on isopollen
maps (sometimes referred to as ‘isochrone maps’), which
provide ‘snapshots’ of regional vegetation cover for selected
time periods (Birks, 1989). Pollen percentages of selected
taxa are obtained from all sites within a region for a
particular time interval; these are then plotted on maps
which are contoured to reflect spatial variations in each
taxon’s abundance. Data can be displayed for consecutive Ska 6ka
periods showing the changing distributions of selected
taxa over time (Latalowa & Van der Knaap, 2006) and
palaeovegetation maps can be generated to show the overall
change in vegetation composition and distribution through
time (Figure 4.6).
Isopollen and palaeovegetation maps for the Lateglacial
and Holocene periods have been constructed for both
Europe (Huntley, 1992) and North America (Williams,
2002). Such maps provide a basis for testing hypotheses
relating, for example, to the responses of individual taxa or 3ka[
vegetation as a whole to global climatic changes or other
environmental influences; to areas where temperate trees
survived (plant refugia) during glacial stage conditions
(Magri, 2010); and to patterns and rates of plant migration
during climatic transitions (Klotz et al., 2004). Isopollen
maps also have applications in studies of faunal history,
soil–vegetation relationships, human impacts on vegetation
and the estimation of former biomass (Adams & Faure, Modern
1998). The last-named is of particular importance in the
development of global climate and earth system models, Southeast
Ice M i x e d Forest
and particularly to the role that vegetation plays in carbon Forest
storage and the global carbon cycle (see Chapter 7). In Tundra D e c i d u o u s Forest No A n a l o g
comparisons between palaeovegetation reconstructions
Forest Tundra A s p e n Parkland No Data
and general circulation models, however, the pollen data
are usually simplified using the concept of plant functional Boreal Forest Prairie
types (PFT)2 or plant biomes,3 in order to represent the
dynamics of vegetation change over millennia at the Figure 4.6 The changing vegetation cover of the eastern
continental or global scale (Williams et al., 2004b; Edwards USA from 18 ka to present. These maps, lodged on the NOAA
et al., 2005). web site, are based on 11,700 fossil pollen samples and 1,744
modern pollen samples. Twenty-one pollen taxa were included
in the analyses: Alnus, Fraxinus, Populus, Tilia, Fagus, Betula,
4.2.6.4 Human impact on vegetation cover Ulmus, Abies, Celtis, Corylus, Tsuga, Carya, Ostrya/Carpinus,
Quercus, Pinus, Prairie, Cyperaceae, Picea, Liquidambar,
Human impact has been the dominant factor affecting the Platanus and Juglans (based on original work by Overpeck &
vegetation cover of Europe since the beginning of the Webb, 1992, republished with permission of Geological Society
of America; for additional information see the NOAA web site
Neolithic period (around 7 ka in northwest Europe), and at http://www.ncdc.noaa.gov/paleo/vegmap.html).
the effects of widespread woodland clearance, farming
practices and the introduction of new plant species into
regions have left clear imprints in pollen records, especially
in sites closely associated with human activities. In southern
Switzerland, for example, pollen records spanning the last Age is followed by a further reduction in tree/shrub pollen
7 ka (Figure 4.7) show marked changes in woodland cover and by the appearance of cereal pollen, reflecting the
resulting from burning, with peaks of charcoal repeat- establishment of a mixed farming economy from the
edly associated with reductions in pollen of fir (Abies), Medieval period onwards (Molloy & O’Connell, 2004).
ivy (Hedera), lime (Tilia), elm (Ulmus) and beech (Fagus), Pollen diagrams with clear anthropogenic indicators can
but with increases in hazel (Corylus), willow (Salix) and also provide insights into land management techniques
elderberry (Sambucus). In many northwest European pollen (Berglund et al., 2008), prehistoric and historic farming
records, clearance of woodland to facilitate pastoral practices (Veski et al., 2005b) and activities around ancient
activities or cultivation is reflected not only by reduced settlement sites, such as those of the initial colonization
frequencies of arboreal pollen, but also by increases in of the Faroe Islands and New Zealand (Wallin, 1996; Arge
pollen of ruderal taxa that are characteristic of arable et al., 2005). Pollen records can also reflect the environ-
or pastoral land, or by the appearance of pollen grains of mental impacts of early mining operations, especially when
cereals. These pollen signals are commonly referred to as combined with geochemical evidence from the same
‘anthropogenic indicators’ (Behre, 1986), and an example sequence (Breitenlechner et al., 2010), and of military
of the way in which they register in pollen records is shown invasion and occupation (Manning et al., 2007). Along with
in Figure 4.8. This diagram, from the Aran Islands in other indicators of human activity, such as increased
western Ireland, demonstrates the degree to which people sediment accumulation in lakes, the occurrence of charcoal
modified the vegetation cover during the mid-Holocene. into peat and lake sediments, and changes in soil charac-
Trees and shrubs begin to disappear from the record at the teristics, pollen-stratigraphic records provide important
end of the Neolithic period, and are replaced by herbaceous insights into the evolution of the ‘cultural landscape’ in
plants associated with pastoral activity. A short-lived different regions of the world, particularly during the mid-
episode of woodland regeneration during the late Iron and late-Holocene period (Hellman et al., 2009).
)
c m )
) mh m)) ( ) m )
m t )
(c
m ) (c (cm )
(c h e(cp h (c (c
m
h (cm th th m
t h ptD pt t h ep ep (c
ep De De th ep pt D D th
D ep D D
e
ep
D D 14c
33210000
820 3200 14c

3300
840 3400 EH
14a
3500
860 3600
13
3700
3800
880 3900
4000 12
4100
900
4200
4300
920 4400 11
4500
4600
940 10
4700
4800
960 4900
9
5000
5100
8
980 5200
0 20 40 400 800 20 40 0 1000 10 20 30
x10 x10
Figure 4.7 Pollen percentage (%) and influx (I) diagram for Abies (fir) and Corylus (hazel), and charcoal influx record, from a site
in south Switzerland spanning the period 5.1–3.1 cal. BC (from Tinner et al., 1999).
climatic affinities (Zagwijn, 1994). However, problems
historical periods
relating to the taphonomy of pollen assemblages (see
(pasture)
Wetland
section 4.1.2) and also the differential migrational response
(arable)
Corylus
Shrubs
Herbs
Herbs
Ferns
Trees
Samples
Norman to Modern Cultural/

of plants, especially trees, to climatic change, can lead
Tall
Depth
to erroneous palaeoclimatic inferences. More recent

(cm)
approaches to climatic reconstructions using pollen data

2250 have involved the development of pollen response
2300 functions, which measure, in a quantitative way, the
2350 dependence of a range of plant taxa or of broad-scale
2400 vegetation patterns on climate (Seppä & Bennett, 2003).
2450 Relationships are determined between spatial variations in
2500 modern vegetation and contemporary climatic parameters,
Medieval
2550 usually mean annual or summer temperatures. These form
2600 the basis for a pollen–climate training set which enables the
inter-dependence of these variables to be tested statistically.
Iron Age
2650
2700
Late Selection of an optimal mathematical relationship provides
a statistical transfer function4 that is used to infer past
Mid & Early
2750
IranAge
2800
climatic conditions from analogous pollen assemblages
2850
(calibration). So long as the training set encompasses the
Late B-A
2900
full range of plants represented in the fossil record, then
2950
variations of climate through time can be inferred (Figure
Mid-BA
3000
4.9). However, the training set must accurately reflect
the vegetation composition of the study area. Moreover,
Early BA
3050
many of the plants commonly recorded in pollen spectra
3100
have wide climatic tolerances, and those with the narrow
Neolithic
3150
climatic preferences that are most valuable in palaeo-
3200
climatic reconstructions are frequently present in relatively
3250
low abundance. In addition, transfer functions assume
3300
Mesolithic
equilibria between biota and the environment, and may

3350
produce aberrant results at times of climate–environment
3400
disequilibria (Birks & Seppä, 2004; Finsinger et al., 2007).
3450
These limitations notwithstanding, however, pollen
Varved
response functions are undoubtedly a powerful tool for
20 40 60 80 100
sediment climatic reconstruction, and are being increasingly widely
used in integrated modelling of past climates at a range of
Figure 4.8 Holocene pollen diagram from the site of An Loch spatial and temporal scales (Seppä et al., 2004; St. Jacques
Mor, the Aran Islands, western Ireland. Apart from Corylus,
which was a dominant shrub in the early Holocene, other taxa et al., 2008).
are grouped into plant types and their varying proportions are
related to archaeological (cultural) and historical periods (from
Molloy & O’Connell, 2004). 4.3 DIATOM ANALYSIS
4.3.1 Introduction
Although several members of the algal kingdom have
4.2.6.5 Pollen data and climatic
been studied by Quaternary palaeoecologists, including
reconstructions the green algae, blue-green algae and chrysophytes (Arm-
Pollen data have long been used to reconstruct Quatern- strong & Brasier, 2005), it is diatoms that have attracted
ary climates. Indeed, some of the earliest systematic most attention, since their classification and ecological
attempts to derive climatic parameters from fossil evidence preferences are much better understood. Diatoms have
employed pollen records (Faegri & Iversen, 1989). The been studied for over two centuries and the analysis of
initial approaches relied largely on indicator species in the diatom content of Quaternary sediments actually
pollen diagrams, that is, individual plants with recognized predates pollen analysis (Round et al., 1990). By the end
Contemporary environment
(climate, water, chemistry)
Environmental
Transfer reconstruction
function
Modern biology
Calibration dataset
a
sim
a
im nsuim tisa a a a a ima
is s iis sim sim im
sim utiss
Fossil biology ut snumt a tis tis t iss
i n i u u a a a
u tis es min
m ithme tissim min tissim in
a a
im im iss
im
sim i n u
m i s s
is s a t is
m in h
es aens minu es ut
u uat it
ut simin
m tcishsnimana
es
inim
in isms
es
u in ism
ith hinth in uts
es
m ust
th m
th es
m inhe hine
ith
uA
an nAacn anithe ani nithe
s nit
ni es
m nmit
n
s
a ith ith
ensa
itchh neas n i th hes min
a an hnanit
n n cith
hn chna
n
A
hn
h h chn a n
hn
a a aA
hn hn
Ac Ac hn hn
A Ac Ac Ac Ac
Ac Ac
Ac hn Ac
Ac
A
0
4 05.0 5.4 5 a
Lake-water pH
Depth (cm)
10
10
30 20 20 20
20
Bio stratigraphy
Figure 4.9 Schematic representation of the procedures employed to derive quantitative lake-water pH reconstructions from fossil
diatom records (from Birks & Seppä, 2004, but based on an unpublished diagram by Steve Juggins, Newcastle University, UK).
Although this schematic relates to diatoms, the procedure can be adapted to quantify the relationships between other biological
proxies and environmental variables, for example pollen assemblages and climate.
of the nineteenth century, a considerable amount of 4.3.2 The nature and ecology of
work had been undertaken on diatom remains, and while diatoms
most of these studies had little stratigraphic or palaeo-
ecological value, they laid the foundations for later research Diatoms are microscopic, unicellular members of the
on diatom taxonomy. Diatoms are ubiquitous, being Bacillariophyta of the algal kingdom (Figure 4.10). They
found in virtually every body of water (Jones, 2007), while secrete a siliceous shell or structure, known as a frustule
in the oceans, they generate most of the organic matter which can range in length from 5 μm to c. 2 mm, depending
that supports marine life, accounting for something on the species (Battarbee et al., 2001). The frustule is often
like one-fifth of all photosynthesis on earth (Armbrust, compared to a pill-box, as it consists of two overlapping
2009). Quaternary diatom remains have proved extremely valves or thecae, the larger one (epitheca) fitting over the
useful as indicators of local habitat changes, particularly in smaller one (hypotheca) in a box fashion to enclose the
lake sediments, but also in both shallow and deep marine living (protoplasmic) mass. The valves are linked together
deposits. The analysis of diatom floras has provided new by connecting girdle bands (copulae). The wall of the
insights into a wide range of palaeoenvironmental issues, frustule may be a single layer of silica or it may be more
such as the reconstruction of lake-level changes, water complex, consisting of a double silica wall separated by
chemistry variations, sea-level changes and the disturbance vertical silica slats. Frustules are commonly circular (centric)
of lake ecosystems by human activities (Smol & Stoermer, or elliptical to rod-like (pennate) in shape and are perforated
2010). by intricate patterns of tiny apertures (punctae or areolae).
Figure 4.10 SEM images of common northwest European diatom frustules recovered from streams and soils in southeast
England. 1 – Diatoma sp. (valve view). 2 – Cocconeis placentula (valve view). 3 – Melosira varians (girdle view). 4 – Encyonema
ventricosum (valve view). 5 – Unknown naviculoid (valve view). 6 – Melosira varians (girdle view). 7 – Surirella sp. (valve view).
8 – Cocconeis placentula (valve view). 9 – Achnanthes sp. (valve view). 10 – Planothidium frequentissimum (valve view). 11 –
Hantzschia amphioxys (fragment) (valve view). 12 – Pinnularia borealis (valve view) (images by Kirstie Scott, University College
London, UK).
The arrangement of the perforations is one of the most to stones) and free-floating (planktonic) forms, and while
important diagnostic characteristics of diatoms, although all species require light and are therefore limited to the
other structural details, as well as in tiny reticulations, canals photic zone (usually less than 200 m water depth), they
and ribs, are important for classification below the generic occupy a large number of ecological niches. In the sea, they
level, so that careful scrutiny of the valves under a high- are found in lagoons, shelf seas and deep oceans; they are
powered microscope is necessary for species identifications. common in the intertidal zone in estuaries and salt marshes;
The frustules are composed of amorphous hydrated silica, and they are often abundant in ponds, lakes and rivers.
similar to opal, which enhances their preservation potential Certain species even live on wetted rocks, in the soil or
in different sedimentary environments. attached to trees. The freshwater, soil or epiphytic niches
Diatoms are found in a wide range of aqueous to are dominated by pennate diatoms, while the centric forms
subaqueous environments, including soils (Clarke, 2003), tend to be more common as free-floating plankton in
and make up about 80 per cent of the world’s primary ocean waters, especially in the subpolar and temperate
producers. They exist in bottom-dwelling (benthic), latitudes. Marine benthic habitats, however, are charac-
attached (epiphytic – attached to plants; epilithic – attached terized by pennate forms.
The distribution of diatom species is determined by a As with pollen grains, identifications are based upon type
number of variables, including water acidity and salinity, collections, keys and photographs in diatom manuals and
oxygen availability, nutrient content and water tempera- catalogues (e.g. Round et al., 1990). Nevertheless, diatoms
ture. Freshwater diatoms are controlled largely by salinity, are often difficult to classify and recent inter-comparison
pH and trophic status (Battarbee et al., 2001), while sea- exercises have revealed discrepancies in taxonomic
surface temperatures, oceanic frontal contrasts and nutrient conventions between different countries, a problem that is
up-welling influence the distribution of many marine taxa being addressed by the development of international
(Burckle, 1998). Changes in any of these parameters can taxonomic quality control standards (e.g. Stoermer, 2001).
have a major effect on the structure and composition of the Online guides to diatom identification are available, for
diatom community. The autecology of modern diatoms has example via the US Geological Survey at http://western
been extensively studied (Smol & Stoermer, 2010) and a diatoms.colorado.edu/, while attempts are being made to
range of quantitative methods has been developed to relate automate the process, such as the Automatic Diatom
modern diatom assemblages to contemporary habitat and Identification And Classification (ADIAC) project (Bayer
environmental conditions (see below). & Juggins, 2002).
Three principal methods have been developed for the
determination of diatom concentrations in sediment
4.3.3 Field and laboratory methods samples. The aliquot method uses a measured volume
Diatom valves, like pollen grains, are best preserved in fine- of sample suspension which is pipetted on to a circular
grained sediments since they can be easily damaged or coverslip and the total number of frustules per unit volume
destroyed in coarse-grained deposits. Samples for analysis is estimated. The evaporation tray method uses a meas-
can be obtained from vertical exposures in shallow water ured volume of suspension which is added to a coverslip
marine or estuarine deposits, though more frequently they and from which the water is allowed to evaporate at room
are extracted from sediment cores obtained from lakes, temperature. The diatoms are then counted and since
shelf seas or the deep ocean floor (e.g. Koç, 2007; Barron the original volume is known, the concentration can be
& Bukry, 2007). Diatoms are not susceptible to oxidation measured. The use of microsphere markers is the third
or microbial degradation, but cores for diatom analysis method. These are tiny spherules made of glass or plastic
are usually sealed air-tight to prevent drying out of the which are added in known numbers to an aliquot of sample
sediment, which can lead to fracturing of the valves. suspension, and concentrations of diatoms can be calcu-
Diatom frustules may be separated from the sediment lated from the ratio of diatoms to spherules. This third
matrix by a variety of laboratory procedures (Serieyssol method is similar in principle to the estimation of pollen
et al., 2011). Organic matter is removed by oxidation, concentrations using exotic pollen markers, and sometimes
the most common methods being by digestion in H2O2 yields better results (Wolfe, 1997). Measures of diatom
or in a mixture of potassium dichromate and sulphuric concentration underpin estimates of diatom productivity,
acid, while carbonates and certain other salts can be and are important for quantifying silica storage and flux in
dissolved by heating gently in dilute hydrochloric acid. the global silica cycle (Bradtmiller et al., 2006).
Much more difficult is the removal of minerogenic matter, Diatom counts based on samples selected from a
since diatoms are soluble in some acids, and especially stratified sediment sequence are normally presented in
in hydrofluoric acid (HF) which is commonly used in the the form of a percentage diagram using bar histograms
preparation of pollen samples. This means that diatom and (e.g. Figure 4.11). These can be subdivided in a similar way
pollen counts cannot be carried out on the same slides since to pollen diagrams (see above), either subjectively on the
the samples have to be prepared separately. Coarse mineral basis of visual inspection of the data, or by statistical
particles (> 500 μm) can be removed by sieving or by gentle ordination techniques that test for structure and spatial
swilling in a beaker, but finer particles require either some variability within the data (Webster & Oliver, 2000). Other
form of flotation using heavy liquids, or the less efficient methods of data presentation that have been employed
differential centrifugation method. The residues are by diatom analysts include the tabular format, where
mounted on slides and counted under a microscope using counts are listed in percentages, and the composite or
phase-contrast illumination and magnification of up to ratio diagram, whereby groups of diatoms with particular
×1,000. Diatoms are often more abundant than pollen in environmental affinities are added together to show the
sediment samples, and a count of 500 or 1,000 valves may ratio variations between the different groups (e.g. Figure
form a statistically significant total (the ‘diatom sum’). 4.12).
Planktic species Tycho plankton Benthic species
p)
)
.b
p)
p
)
p) .b
bp
e (ca ) l .b
al
.b al
(c
l.
A.bg p)
p))
Ag e .bp )ca
)
al (c)
p)
bp
pe)
ca)
gee .b )
(c )
.bbp
p)
(c ebp
al .bp
p)
b( p
AAg cal l .bp

al .b
p
((c p)
Ag cal g.bep(
b
.b
.b
l.
Ag caall.
.b
e Aagl .
.
Ag .bp
gael .
(c l
e (ca
a
e cal
al
al
al
al
e caAl
Ag (c
(c
(Ac
(c
(c
(c
Ag e (
Ag e
e
Ag (
e
(
(c
e
Ag
e
e
Ag
(
e
Ag
e
Ag
Ag
Ag
Ag
Ag
0 B-8
250
500 B-7
750
1000 B-7
B-7
1250
1500 B-4
1750¬
2000
2250
2500
B-3
2750
3000
3250
3500 B-2
3750
4000 B-1
6 20 010 6 20 40 60 80 0 20 40 60 80 0 20 40 60 O~*20 0 20 0 20 0 20 40 0 20010 6~7ooTo 0 20 0 20
Tycho- Diatom-inferred Mean

Planktics plankton Bent hies depth grainsize
0 B-7
250
500 B-7
750
1000 B-7
B-7
1250
1500 B-4
1750
2000
2250
2500
B-3
2750
3000
3250
3500 B-2
3750
4000 B-1
0 20 40 60 80 100 0 20 40 0 20 40 60 80 100 0 2 4 6 40 120 200 290
Depth (m) B-7
Figure 4.11 Late Holocene diatom stratigraphy from Beaver Lake in the Nebraska Sand Hills, USA. Variations in percentages of
planktic, benthic and tychoplanktic species enable variations in lake-water depth to be inferred. Tychoplankton (or pseudo-plankton)
are siliceous forms, particularly algae, that become enmeshed in vegetation mats near lake shores (from Schmieder et al., 2011).
4.3.4 The interpretation of Quaternary Moreover, diatom frustule size varies markedly between
species. These various factors complicate the interpretation
diatom records
of diatom assemblages and the estimation of diatom
The difficulties that arise in the interpretation of diatom productivity. Species-specific correction factors can be
assemblages are, in many ways, similar to those discussed applied to the data in order to reduce some of the bias
above for fossil pollen. Diatom valves are light and easily (Korhola, 2007), but an element of uncertainty will always
transported, and thus in estuarine sediments, for example, remain.
there is frequently an admixture of marine, brackish and Despite these problems, diatom analysis has proved to
freshwater forms, while lake muds may contain diatoms be a particularly valuable technique for environmental
derived not only from the lake ecosystem itself, but also reconstructions, particularly in studies of eutrophication,
from inflowing streams and catchment soils. Freshwater acidification and salinity changes in lakes (Battarbee et al.,
diatoms often occur in marine sediments, having been 2001; Smol & Stoermer, 2010). These and other applications
blown in by the wind, while in the deep oceans, diatom of diatom analysis are discussed in the following section.
remains have been found that have been transported many In addition, diatoms in deep-ocean sediments constitute
hundreds of kilometres from their source (Marshall & a valuable source of palaeoecological data, an aspect
Chalmers, 1997). Selective destruction of diatoms is another considered later in this chapter (section 4.10).
potential error source, with complete or partial dissolution
of the frustules under pressure at depth in the oceans, while
in brackish and freshwater contexts, the less robust and
4.3.5 Applications of diatom analysis
weakly silicified forms will tend to dissolve where conditions
4.3.5.1 Diatoms as salinity indicators
are very alkaline. In these environments the diatom death
assemblage will be biased in favour of the stronger and more A major control on diatom distributions is salinity, and
heavily silicified forms. Grazing by herbivores may also hence individual diatom species can be classified on the
affect the composition of fossil assemblages. Reworked basis of their salinity preferences. The halobian system
diatom frustules can sometimes be detected where valves of classification, introduced by Robert in 1927 for applica-
are broken or partially dissolved, or demonstrate signs of tion to marine and estuarine assemblages, and subse-
mechanical abrasion, although secondary diatoms within quently modified by Friedrich Hustedt in the 1950s, has
a body of sediment may not always be easily recognizable. four main groupings:
Sediment Marine Brackish Freshwater Highly Tolerant

sequence Poly halo bous species Mesohalobous species Oligohalobous species Euryhaline
Marine
silts &
Saltwater
clay
transitional
unit
Organic
clay Brackish
transitional
unit
Freshwater
Peat
0 25 SO 75 0 25 50 75 0 25 50 75 0 25 50
Percentage
Figure 4.12 A marine incursion (positive sea-level tendency) as reflected in a diatom assemblage (from Carter, 1992).
(a) polyhalobous diatoms: those that thrive in salt Closed-basin lakes

concentrations of >30 per cent; These lakes are frequently sensitive to changes in effective
(b) mesohalobous diatoms: those that thrive in salt moisture, especially where this affects the salt content of
concentrations of 0.2–30 per cent; the lake water (‘salt lakes’), and hence past variations in
(c) oligohalobous diatoms: those that generally require salt salinity can be used as a basis for reconstructing climatic
concentrations of <0.2 per cent; change (Fritz, 2008). Fluctuations in lake volume and salt
(d) halophobous diatoms: those that cannot tolerate even concentrations in lake waters will, in turn, be reflected in
slightly salty water. changing compositions in diatom assemblages, because
of the wide range in salt tolerance between obligate fresh-
Other terms commonly employed today are ‘halophilous’ water and hypersaline species (Figure 4.14). The relation-
for those diatoms that are particularly attracted to salt- ship between diatom assemblage variations and salinity
rich waters and ‘indifferent’ or ‘euryhaline’, for those concentrations in modern lakes can be measured and
seemingly unaffected by variations in salt content, and forms the basis for diatom training sets or calibration
hence not restricted to any of the above categories. The sets. These are now available for a number of regions
sensitivity of diatoms to salinity changes is exemplified throughout the world, and enable past variations in lake
in two particular areas of Quaternary research: sea-level salinity to be quantified (Davies et al., 2002a). For example,
change, and the environmental record preserved in closed- Laird et al. (2007) interpreted an abrupt increase in lake
basin lakes. salinity in Oro Lake, western Canada at around 9.6 ka as
marking the onset of a more arid climate that continued
Sea-level variations throughout much of the Holocene, albeit with several short
Diatoms have long been employed as indicators of chang- episodes of reduced salinity that suggest shifts to wetter
climatic conditions (Figure 4.14). Lake waters, however, are
ing sea level, the earliest accounts being from Scandinavia
chemically complex and the ratios of the principal salts
in the 1920s, where evidence of marine or brackish
(chlorides, carbonates and sulphates) vary in response
water conditions in lakes now well above sea level formed
not only to hydrological changes, but also to a range of
an important line of evidence in the reconstruction of the
environmental variables that affect dilution and mineral
extent of glacio-isostatic recovery following the wastage of
supply (Fritz et al., 2010a). Understanding the response of
the last ice sheet (Tuovinen et al., 2008; section 2.5.4).
diatom communities to such changes is essential if valid
Subsequently, the analysis of diatom assemblages from
inferences about palaeosalinity variations are to be made.
coastal localities has become, as with pollen data (see
Recent approaches to this problem have involved multi-
above), a standard technique for identifying positive and
variate statistical analysis of modern diatom assemblages to
negative sea-level tendencies (section 2.5.2.2; Figures 2.35 gauge which chemical variables have the greatest influence
and 2.36) in littoral sediment sequences. Positive tenden- on the structure and composition of diatom assemblages
cies are characterized by the replacement of halophobous (Ryves et al., 2003).
and oligohalobous diatoms first by mesohalobous and
subsequently by polyhalobous forms (Figure 4.12), whereas
the reverse is the case where negative sea-level tendencies 4.3.5.2 Diatoms and pH
have occurred (Figure 4.13). Hence changes in diatom The distribution and abundance of many diatom species
assemblages can reflect precisely the highest point of marine also vary with water pH or with a wide range of environ-
influence, and can therefore provide important strati- mental factors that co-vary with pH, such as alkalinity.
graphic markers for the reconstruction of shoreline dis- Following pioneering work by Friedrich Hustedt in the
placement curves and patterns of isostatic rebound. The 1930s, diatoms have traditionally been divided into the
evidence illustrated in Figure 4.13, for example, provides following groups according to their pH preferences:
important indices of the timing and relative rate of sea-
level fall in southeast Greenland as a result of land uplift (a) alkalibiontic: occur in waters of pH values > 7;
following the retreat of the Greenland Ice Sheet (Long (b) alkaliphilous: occur at pH values of about 7, but with
et al., 2008). Similar investigations have provided valuable widest distribution at pH > 7;
sea-level index points relating to the wastage of the last (c) circum-neutral: occur equally above and below a pH
ice sheets in the British Isles (Selby & Smith, 2007), Scandi- of 7;
navia (Lohne et al., 2007) and North America (Hutchinson (d) acidophilous: occur at pH values of about 7, but with
et al., 2004). widest distribution at pH < 7;
a) b)
65 40-
ali
Scoresby
ss
Disko Sund
k a
Ammassalik
ali m
Bugt
ssAm
Send re
ma
Stromfjord
Am
a)
Ammassalik
0 Kilometres 10
_Wan ortaWt
Qaqortoq<
Kap Farvet 37 30W 37" 10
65 30N
a)
1 2 3 4 6 S 7
23
22
Altitude (metres)
.7
21 •
5 water
20
54
19
3 18
,2
17
1
16
15
14
Area
N m 10
of sill 13
QT23 Gyttja Sand Silt Sample core
s) ) ) ) s) es
)
s)
di) QT23 tre es es es dii) QT11 tre s)
me etr etr etr me tre etr tre
( m e(
m
e(
m
de
(
(m
e
e(
m (me
de e( de tud de
titu itu
d itud itud titu itu i itu
Al Alt Alt Alt Al Alt Alt
741 348
Alt
742-
350
743-
8 5 2 3 ± S7
O S 54-9411 744 352
cal. BP, 745- 7594153
SUERC-9425) (8538-8324 354
746
cal. BP.
747- SUERC-9424) 356
748
358
749-
750- 360
751
50 100 SO 100 50 1C0 50 10 0 5 0 10 0 50 100
Figure 4.13 Isolation basin studies in southeast Greenland. Transects of sediment cores were obtained from six small basins
on islands in Ammassalik Fjord (circle in b). c) shows an example of the sediment infill recorded from a transect of boreholes for
one of the basins, and d) the changes in diatom content in two of the basins. The change from polyhalobous to oligohalobous
assemblages marks the time when the basin became isolated from the sea due to land uplift after wastage of the Greenland Ice
Sheet, the timing of basin isolation being dependant on the height of each basin’s sill (from Long et al., 2008). For further explanation
see text.
(e) acidobiontic: occur at pH values of less than 7, with for some lowland lakes, deforestation and the spread of
optimum distribution at pH of 5.5 or under. farming in the mid- to late Holocene has led to marked
changes in nutrient balance (Hall & Smol, 2010), whilst
Fossil diatom assemblages that can be classified in this the recent increase in sewage effluents, phosphorus-
way provide a basis for inferring changes in lake water rich detergents and intensive agricultural methods has
pH in the past (Battarbee et al., 2010). The links between resulted in a rapid nutrient enrichment of many lakes
modern diatom assemblage variations and pH are tested throughout the world (Battarbee & Bennion, 2011), as
and quantified using transfer functions based on a range well as impact-ing on coastal waters (Clarke et al., 2006a).
of multivariate statistical methods (Holden et al., 2008) Such cultural eutrophication is marked in diatom records
and these provide a basis for inferring past variations in pH from lake sediments (Ekdahl et al., 2004), particularly in
from fossil diatom records (Simpson et al., 2005). This lakes that were naturally oligotrophic, by increases
approach has enabled palaeolimnologists to demonstrate in overall diatom productivity, in the relative abundance
the acidification of many lakes during the late Holocene of planktonic taxa, or in diatom-inferred phosphorus
(Figure 4.15), partly as a consequence of increased levels concentrations.
of acid deposition caused by atmospheric pollution
(Stoddard et al., 1999), but possibly linked also to climate
4.3.5.4 Diatoms and the archaeological
change (Wolfe, 2002).
record
Diatom assemblages are increasingly widely used as
4.3.5.3 Diatoms and trophic status
indicators of past human activity (Juggins & Cameron,
Diatom communities are also very sensitive to changes in 2010). Archaeological applications include the identifi-
nutrients and are therefore good indicators of trophic cation of prehistoric agricultural practices (Horrocks et al.,
status (section 3.9.2). Many lakes have experienced a 2002), the detection and mapping of traces of ancient
trophic stability through most of the Holocene. However, dwellings (Bathurst et al., 2010), and the provenancing
Fresh Sgbsaline Hyposalin e

ql\
Hypersaline
; 0.5 gl ' 0.5-3 gfl ' 3-20 20-50 gl\
e ns s
ed en
e rc c ed s
i nt t er en
in ed
s
v en
s
c
s
a v ed
er
en is
si
at
s
si
c
er
si
en
ta t
en um s
t
in
en
n
oc e hum th en
in
na
um
is
in
nse is
v v
ne eo me s um
si ns
ise ns
ta
o m
in
th
p ta
m sis
na
e
s
en is
u sith
s
nuem ise
ios
ne m is
a
co is um sis
s
en m
oc neo ne hu
si
is
si
ne hu is
si
oc is n oisthnu nsi
in
ira p
toh sm
m s
s
um ns
nn
si
th no
hu n
is thu ns
s
hu si
en
um us
en
is ot ns
en
C on hum nsi
C s ne hum s
p
C ne eo mene
ot me
u
n
si
i
ira
th oethn
ot me
coc ethn
ot en
) ros
C one ncecoo sis
s ot
a
C co eo me
en ir
ne ne e
m
ne onne um
p
e
ed
eo em
co is n theuis
os
noe mo
os
hu
BP tau
c so en
ne ei e
ne hu
rc
ra
isC thnue
oc s n hu
te
ur
h
si
co on n
r
n
in
oc ne th
ta
ot
is ot
ot
v
ot
al os au
os
ne eois
ta
ot
o
ne
s
ne ne
C coCn ne
ur
t
na
ne
d
co nne
st
C cc ei
ne
eu
in
c
t
co n
p
is
o
a
co is
is
e( eud
oc ecicso
s
oc eC
is
st
a
o n
co ics
do
ir
oc is
P
ne
is
i
ne
os
c
i
ne
is
C co
oc oe
ne
C con
ur
do
ne
Ag Ps
ono
co
ta
co
u
co
os
cC
co
i
u
co
c
oc
se
ne
ud
oc
oc
se
oc
oc
se
oc
oc
oc
oc
C
co
P
C
C
oc
P
C
P
C
C
C
oc
0
C
0
1000
200 0 C3
3000
4000
500 0 C2
6000
E3
300 0 CI
accc
10000 80
80
11000
80 80 80 20 30
20 60 20 20 4 0 20 40 20 '•0 20 20 80 20 80 20
20 •10
SO ! 00 20 4 0 20 20 20 0 1 1 '0 00
Relative a b u n d a n c e Diatom-inferre d salinity (gtL)
Figure 4.14 Diatom-inferred palaeosalinity through a sediment sequence in Oro Lake, Saskatchewan, Canada, based on
variations in abundance of dominant diatom species (from Laird et al., 2007).
g
ag
nag g
agg iaag ag
a n g
a n a a
an a in
ta egtiza ata a gi
a atat
n gi r at t a r a u t t a i n t a eerar ta
i e a e ki e n r a z a m
et
z
om er m atz ra er uet er glolom era
r ku gl m glo tkeulle glom ome lom k om lalag lom
a lla lla
l o o
lla cllla lla
l g ll a l
g tell g
ar aYr e clote
g g
e e c lo
te
o t el
la
l o te lCotye ote ella
c y yc c l t o t el
la lote
y c l o
lla lote a
te cyycclo tell
o e ar
Y Y C y c y
cy
c l y
c C c
cy
c l o
cy
c l C cy
c c
cy
c l Y
198S 0
3
1970 4
5
5
1950
7
8
9
1900
10
11
1850 •
12
13
14
15
16
17
18
19
20
20 4D 60 20 40 20 20 20 20 20 40% 5.0 6.0 7.0
Percent of total sum of diatom valves
Figure 4.15 Diatom-inferred pH history of Lysevatten Lake, southwestern Sweden, suggesting increased acidification since the
1960s. Note how the alkaliphilous diatom species on the left are progressively replaced by diatoms which are associated with
more acid conditions (centre and right). The curve for inferred pH change is shown on the right (from Renberg et al., 2009).
of manufactured materials such as pottery and bricks long-term changes in a number of other environmental
(Kligmann & Calderari, 2012). Diatom stratigraphy is also contexts. They can provide evidence of variations in, inter
frequently employed in multi-proxy investigations of alia, lake-water depth (Wolin & Stone, 2010), the extent and
archaeological sites, especially those associated with duration of ice cover over lakes (Thompson et al., 2005) and
waterfronts (e.g. Veski et al., 2005a) or lake dwellings, sea surfaces (Gersonde & Zielinski, 2000), wind trajectories
such as crannogs (O’Brien et al., 2005). As outlined above, and strength (Romero et al., 2003) and rates of ocean
diatoms also reveal the impacts of humans on local upwelling (Anderson et al., 2009). They have also been used
ecosystems and water bodies, from the introduction of to detect prehistoric tsunami activity (Dawson & Smith,
agriculture in the early to mid-Holocene (Bradshaw et al., 2000) and to reconstruct high-resolution climate histories
2005) through to the toxic effects of metal pollution in the (Fritz, 2008). Finally, while most of this section has focused
industrial era (Cattaneo et al., 2004). on diatom evidence from the late Quaternary, diatom
analysis has also been applied to earlier Quaternary records,
for example, in the investigations of long lake sediment
4.3.5.5 Other environmental applications
sequences such as that in Lake Baikal in central Asia
Because diatoms respond sensitively to a range of variables, (Mackay, 2007), and diatom-rich oozes on the deep ocean
they are proving to be useful indicators for evaluating floor (Romero & Schmieder, 2006).
4.4 PLANT MACROFOSSIL vascular plants (plants with specialized tissues for trans-
porting water), such as fruits, seeds, stamens, buds, scales,
ANALYSIS rhizomes, roots and bark, and non-vascular plants, mainly
the leaves of bryophytes (mosses) and liverworts, but
4.4.1 Introduction also oospores of algae, the latter being especially abundant
The study of fossil plant remains is one of the earliest in calcareous lakes (Figure 4.16). Carbonized plant macro-
branches of Quaternary research, the Quaternary floras of fossils, usually in the form of wood or seeds, are also
the British Isles being investigated from as long ago as the commonly encountered, especially in archaeological
1840s. Similar studies were being undertaken in Denmark contexts (van der Veen, 2007).
and Germany and the results of these nineteenth-century Plant macrofossils are found in a variety of depositional
investigations were synthesized in Clement Reid’s remark- environments, but most commonly in lacustrine and
able volume The Origin of the British Flora published in 1899 fluviatile sediments, especially fine alluvium, and in acid
(Birks, 2008). This book contained the first clear state- peats. Occasionally, rich assemblages of plant remains are
ment of Quaternary vegetation changes in western Europe, recovered from soils or sediments on archaeological sites
and appeared almost twenty years before the develop- where the fossil remains may include those of cultivated
ment of pollen analysis as a palaeoenvironmental technique. plants, weeds associated with cultivation, and uncultivated
The analysis of plant macrofossils can provide valuable species collected for food (e.g. Kubiak-Martens, 1999).
complementary information to microfossil data, but it can Fossil dung or animal middens may also preserve plant
also provide an independent approach to the reconstruction remains in good condition, and in some regions these have
of environmental conditions. provided valuable records of vegetational and environ-
mental change (van Geel et al., 2008). However, it is in acid
peat deposits that the remains are often best preserved, as
4.4.2 The nature of plant macrofossils the fossils have been protected from oxidation, and may
Plant macrofossils range in size from minute fragments of even be found in growth position.
plant tissue or tiny seeds, to pieces of wood, and even to The preservation of fossil plant material is very variable.
whole trees that can be measured in cubic metres (Birks, Wood may survive in recognizable condition for many
2001, 2007). They include recognizable remains of both thousands of years, either in waterlogged sites or in very dry
perigynium
m
perigynium niu per
erigy igyn
p ium
perigynium
2 mm
1 mm
Carex perigynium
ci.C.rostrata
Ca
rex
Lanx
pcei.
Cri.g laricina d
roys
ntriau
needle tam
glauca^
balsam^ perigynium
Abies Carex
glauca^ ci.C.rostrata
glauca^
perigynium
Picea glauca needle
Picea
glauca^
glauca^
Figure 4.16 Scanning electron photomicrographs of terrestrial macrofossils from a 12 ka peat layer in North Dakota, USA (from
Fisher et al., 2008, reprinted by permission of SAGE; image provided by Timothy Fisher, University of Toledo, USA and Catherine
Yansa, Michigan State University, USA).
soils in arid environments, but in other situations decom- glycerine or other preserving fluids. The more delicate
position can be extremely rapid. Seeds and fruits will structures such as translucent leaves and seeds are best kept
survive in most deposits, their resistance to decay reflecting mounted on microscope slides. Wood, except in the case
adaptation to periods of dormancy. Some of the very small of some very obvious species (e.g. birch), must be either
seeds, however, such as those produced by orchids and by macerated or prepared in thin section for identification.
certain members of the heather family (Ericaceae), are Most plant macrofossil remains can be examined on a
rarely preserved. The leaves of deciduous trees, with their white plate under binocular scanners or low-powered
delicate structure, are particularly vulnerable to mechanical stereo-microscopes. On occasions, however, high-powered
breakdown and decomposition and hence are usually microscopy is required, and in recent years, the electron
preserved as very small fragments. Perfect specimens can microscope (Figure 4.16) has been employed for the
survive, however, especially in annually laminated fine- differentiation of closely similar taxa (Carlquist, 2001). As
grained sediments (Staff et al., 2011) or in deposits in with pollen analysis, identifications are based on a reference
sinkholes with water deep enough to be anoxic (Steadman collection of seeds, fruits, leaves, wood and so on from the
et al., 2007). Needles from coniferous trees are often present flora, on atlases of macroscopic plant remains and,
abundant as macrofossils, occurring in a variety of deposi- in certain cases, on keys of particular plant families (e.g.
tional situations (Tobolski & Ammann, 2000), while leaves Hather, 2000; Velichkevich & Zastawniak, 2008).
of dwarf shrubs from tundra environments such as willow
(Salix), dwarf birch (Betula nana) and bilberry (Vaccinium)
have also been found in a good state of preservation in
4.4.4 Data presentation
lake sediments (Rundgren & Björck, 2003). Of the lower The results of plant macrofossil analysis can be presented
plants, mosses preserve very well in the macrofossil in a number of different ways. At many sites, particularly
form, but lichen and liverwort remains are seldom found, where archaeological investigations are being carried out
though impressions of the latter occur in tufa and travertine or where a single stratum is being investigated, a simple
deposits (Ali et al., 2003). In the majority of cases, therefore, species list is compiled of all taxa discovered. Where several
Quaternary macrofossil analysis is concerned primarily levels are being examined, the presence or absence of
with the study of wood, seeds, fruits and mosses, augmented particular plant remains may be indicated by simple dot
by information provided by a limited number of easily symbols (Figure 4.17) or the results may be shown in
identifiable plant remains such as conifer needles and tabular format. Alternatively, the data may be expressed as
certain leaves. estimates of abundance using such descriptive terms as rare,
occasional or frequent, and thus an impression can be
gained of changes in frequency of taxa through time.
4.4.3 Field and laboratory work More commonly, however, plant macrofossil data are
The larger plant macrofossils can be collected in the field presented in quantitative form, where the plant remains
from exposed sections or from sediment cores, but in most from different levels in a profile are expressed as a per-
cases extraction takes place in the laboratory. There is a centage of the total number of macrofossils identified
variety of techniques for the removal of the fossil remains from each sample, and these are plotted on a vertical time
from the sediment matrix, but the majority involve sequence. A major problem with this type of diagram,
disaggregation of the material with either dilute hydro- however, is that certain macrofossil types (such as seeds
chloric acid or sodium/potassium hydroxide, followed by of aquatic plants in limnic sediments or stems and leaves
sieving (Birks, 2007; Mauquoy et al., 2010). Use of stronger of sedge or moss in peat) will be over-represented and
chemicals, such as nitric acid or hydrogen peroxide tends thus the curves for other taxa will be suppressed. Equally,
to be avoided as these can degrade delicate plant tissues. it may be difficult to arrive at a satisfactory macrofossil
Finer lake sediments and ombrotrophic peat can usually sum, particularly when a range of different types of plant
be broken down simply with the aid of a jet of water and macrofossil material is present. An alternative strategy,
washing through sieves (90–250 μm in mesh size), in order therefore, is to construct a concentration diagram showing
to capture tiny seeds (Beaudoin, 2007). During disaggrega- the occurrence of total numbers of plant macrofossils
tion, some fruits, seeds or leaves will rise to the surface of per unit volume of sediment at different levels in the
the liquid and these can be picked off with a fine paintbrush, profile, and this may be converted to a macrofossil influx
while macrofossils can be removed in a similar fashion from diagram showing accumulation per year if a secure
the mesh of the sieves. Some fossils, such as fruit stones, dating framework can be established (Figure 4.18). As with
can be kept dry, but others need to be stored in alcohol, pollen concentration diagrams, however, fluctuations in
,mc cm
cm t,hc hm,
m
h, etph ethp,tc
th hm, m
cm
pt cm m
, c m
Deepthth,c,m m
m
Depthcm m
DeDeepptehp,ctthm,c
pt h,ccm
eDp
De th, hm,c
c
t,hcm
p,tcm
D eppt hm,c
h, m
,c
De ,cm
h Dcm
De ,cm
DeDpethp,cm
Depth th,c,m
h,
cm
De cm
pt ,Decpm
D Dp ehp,c
p tc
De ethp,tc
,
tph
Dpeth
h,
h
De pth
h
NGRIP
DeptD
DeeDpt
Dpe
pt
pt
Dp
De
De
events
De
D
D
1140- 11400
Holocene
11600
11800
12000
KE3
12200
GS-1
12400
1240
12600
12800
1290 13000 Gl-1a
13200 Gl-1b
1340 13400
13600 Gl-1c
1390'
13800
14401 14000 Gl-1c
1490 14200
Gl-1e
14400
EE 14600
50 1000
2000 6000 5000 15000 5 10
Figure 4.17 Arboreal plant macrofossils (shown as dots) in a Lateglacial sediment sequence in Latvia, plotted against the pollen
record of the same taxa (from Veski et al., 2012).
sedimentation rates and an insufficiently sensitive timescale identified to the species level, and therefore many of the
can pose interpretative problems and, as yet, comparatively problems arising from taxonomic imprecision which often
few influx diagrams have been constructed for plant prove so frustrating in pollen analysis are not encountered
macrofossil data. Another approach to data analysis is to to the same degree in the study of plant macrofossils.
use a square grid graticule for the quantitative estimation Moreover, there are some plants, such as the rushes
of macrofossil remains examined under a low-power (Juncaceae) and poplar (Populus), whose pollen seldom
microscope, and the resulting data can be analysed statistic- survive in the fossil form, and which are only occasionally
ally using, for example, weighted averages, and then represented in the pollen record. The former presence of
presented in tabular form (e.g. Mauquoy et al., 2010). these types in a vegetation community may, however, be
revealed by their macrofossil remains. The same applies
equally to those plants that are low pollen producers.
4.4.5 The interpretation of plant
Unfortunately, however, the occurrence of plant macro-
macrofossil data fossils tends to be sporadic. Many deposits, while rich in
The majority of plant macrofossils found in a body of fossil pollen, are entirely devoid of recognizable plant
sediment are derived locally (autochthonous) and provide remains (and vice versa). Other sediments may contain
data on the composition of former plant communities plant macrofossils, but a large quantity of material is often
growing in and around the site of deposition. Important needed in order to produce relatively few fragments of
information can often be gained, therefore, on local fossil vegetative matter. Moreover, although identification
hydroseral developments and also on changes in the trophic is frequently possible to the species level, the abundance of
status of lake waters and associated fens. Interpretation of diagnostic detail in different fossil remains is very variable,
the record is aided by the fact that, unlike pollen grains, a and while there are seeds and fruits that are relatively easy
very large number of plant macrofossil remains can be to recognize, the identification of small fragments of achene
m
cm
cm
cm
cm m
,c
m
m
h, ,c
c m
th
pt p,tchm
De thm
,c
,c
h,
m
h,
h,
ptm
pctm
cm
pctm
ecpm
h, ,mc
De cm
De m
,tch
ph,c
cm
c
pt pht,hc
cm
he ,
DeDpeth
he,
h,
he,
h,
thp
D
,
th
Dpet
Dpe
h,
pDt
pDt
pt
pDt
pt
DeDpet
h,
De
p
De
pt
De
De
De
De
pt
De
De
0
De
1000 LH2
2000
3000
LH1
4000
5000
6000
MH
7000
8000
9000
EH
10000
11000
0 6 120 6 12 0 2 0.0 0.2 0 0 0.20.0 o.c 0.60.0 0 1 0.0 1.00.0 0.2
Charcoal fragments and needle equivalents cm- y r2 1
Figure 4.18 Influx of charcoal fragments (CHAR), total plant macrofossils (MFAR) and macrofossils of selected plant taxa in a
lake sediment sequence in the North Cascade Range, Washington State, USA (from Prichard et al., 2009).
or epidermis, for example, may require a great deal of work. dispersal such as fruits of birch (Betula) or seeds of
For these reasons, plant macrofossils may be worth studying sycamore (Acer). Hence, these records will usually be
only if they are abundant, well preserved and easily dominated by the remains of the peat-forming plants such
extracted from the sediments in which they occur, or where as Sphagnum mosses and the cotton sedge (Eriophorum
they make up most, if not all, of the sediment body, such vaginatum), accompanied by species such as the bog myrtle
as ombrotrophic peats. In some cases, however, it may be (Myrica gale) and common heather (Calluna vulgaris)
worth examining the macrofossil content of a site where which are often found growing on bog surfaces (Hughes et
relatively few remains are preserved, either because it might al., 2000). In lake sediments, however, plant macrofossil
help to solve a particular ecological problem, encountered assemblages are more diverse, for although locally derived
in pollen analysis, or because the macrofossil remains are fossils (particularly those of aquatic plants) will tend to
required for radiocarbon dating (section 5.3.2). predominate, exotic elements from outside the lacustrine
As with pollen analysis, a proper understanding of the ecosystem may also be present and will have been brought
origins of a plant macrofossil assemblage is required before into the lake by wind, stream or animal transportation
palaeoecological inferences can be attempted. In fen or bog (Figure 4.19). The proportion of autochthonous to alloch-
sites, macroremains tend to be almost entirely of local thonous fossil material will therefore be determined by a
origin (Figure 4.19), apart from a few with very good wind range of taphonomic factors affecting the production and
dispersal of seeds, fruits, leaves, etc., the mode of trans- there have been marked fluctuations in lake-water level and
portation to and within the lake, and sedimentary processes intermittent erosion of the shoreline. Unlike secondary
operating during and after deposition. pollen, however, it is not always easy to detect reworked
Vegetational productivity varies considerably between plant macrofossil material by signs of physical or chemical
species, and even between individuals of the same species, deterioration, and often a meticulous evaluation of the
depending upon reproductive strategies (in the case of assemblage is necessary in order to isolate such exotic
fruits and seeds) and vegetational response to environ- components.
mental conditions. Woodland shrubs and herbs tend to There are, finally, a range of broader environmental
produce relatively few seeds, while certain trees and annual factors that may affect the formation of plant macrofossil
weeds, especially those found growing on muddy surfaces assemblages. These include lake bathymetry and catchment
around lakes with fluctuating water levels, may have a
higher rate of seed production. Other factors will also
a) Temperate lake
come into play. Many seeds and fruits will not find their transport by
way into the fossil record because they are taken for food terrestrial or surface water
marsh plants run-off and streams
by birds and animals, while others may be subject to attack
-wind
by fungal parasites. Different plants have different strategies
float float
for combating predation, which affects the seed survival
ratio (Nathan & Casagrandi, 2004), and hence some taxa aquatic
plants sink lake
sink
have a much higher probability of being represented in
death assemblages than others. The potential dispersal loss by:
range of seeds also varies widely between taxa due primarily, incorporated germination
into sediment predation
but not exclusively, to seed size: small-seeded species death and decay
usually produce more seeds per plant that persist longer in
seed stores than is the case with larger-seeded species b) Arctic or Alpine lake
(Dieffenbacher-Krall, 2007). Dispersal efficiency depends
direct wind transport by
upon the mode of transportation. The most effective agent transport by
deposition surface water
snow meltwater run-off and
for transporting seeds to lake surfaces is wind, especially and streams
•it. solifluction
where propagules or other plant parts such as leaves are
launched from the canopies of tall trees. Transport by .snow float
streams and surface runoff is also important, especially deposition from mm
during periods of high rainfall (Hölzel & Otte, 2004). Seeds melting snowbed sink
that are not waterlogged remain afloat and tend to be
deposition
carried into shallow water by wind and wave action; indeed, of floating
incorporated
studies of the seed content of surface sediments in modern into sediment material
on shore
lakes show low numbers in the deeper areas and a much
greater concentrations close to the shores (Zhao et al.,
2006). Seeds of some species, however, become rapidly c) Bog/fen
local deposition
waterlogged and sink, and are then moved along the lake wind- from plants on
bed by currents and turbulence until they settle in the fer the surface
coarser marginal sediments or are trapped near the lake bog
stream
shore by submerged plants.
Although the foregoing discussion has been concerned stream^ peat
transport
largely with lake sites, many of the points apply equally filtered
to plant macroremains found in riverine sediments. by fen
In fluvial deposits, however, a further complicating factor
is the occurrence of plant macrofossils from different
time periods that have been incorporated into a single
assemblage. Hence reworked or secondary macrofossils are
Figure 4.19 Recruitment pathways and processes by which
frequently encountered in river terrace sequences. Mixed plant macrofossil remains are delivered to lakes in a) a
assemblages of plant macrofossil material are also occasion- temperate environment, b) an arctic or alpine environment, and
ally encountered in lake sediments where, for example, c) to a bog or fen site.
size, the type and density of vegetation surrounding a site palaeoenvironmental reconstructions can be refined.
(dispersal through a woodland stand, for example, is They also aid the Quaternary palaeoecologist in a num-
less efficient than wind transport over open ground), and ber of other ways, for example by indicating the degree
both annual and longer-term climatic variations that can of bias associated with assemblage types, the sedimentary
modulate seed production. In addition, there are geo- niches (e.g. type of lake) where bias is best constrained,
graphically determined influences, such as the role of and the optimal locations within a site for the collection of
melting ice and snow in the transportation of plant debris plant macrofossil samples (Dieffenbacher-Krall, 2007).
in arctic landscapes (West, 2000), the activities of animals One problem, however, is that some associations of plant
in the creation of plant macrofossil assemblages in arid and species appear to have no modern analogues. Jackson &
semi-arid terrains (McCarthy & Head, 2001), and the Williams (2004) attribute these non-analogue vegetation
ecological dynamics in areas of frequent burning, either by types to unique climatic regimes and possibly to lower
natural forest fires or initiated by humans, where charred atmospheric CO2 levels in the past, while in Europe they
macrofossils predominate in the palaeobotanical record may also reflect the long-term impacts of humans (van der
(section 4.4.6.3). Knaap et al., 2011). Future research may clarify the nature
Understanding the complex web of taphonomic and origins of these non-analogue vegetation types and
processes leading to the development of plant macrofossil their implications for contemporary plant ecology, evolu-
assemblages has been aided in recent years by the use of tionary biology and palaeoecology (Jackson & Williams,
modern analogue studies. Some of these are based on 2004).
observation of contemporary ecological processes, which These various difficulties notwithstanding, plant macro-
inform interpretations of palaeoecological records. For fossil analysis remains a valuable tool in palaeoecology.
example, Stearta et al. (2009) studied the present-day decay While pollen analysis is perhaps the more widely used
rates of the leaves of five Australian tree species (Acacia, technique in vegetational reconstructions, plant macrofossil
Atherosperma, Eucalyptus, Lomatia and Nothofagus) in analysis is an important adjunct to palynological studies
order to establish which species might be over- or under- and, when used in conjunction, the two methods offer a
represented in the fossil record. Cellot et al. (1998) exam- more secure basis for palaeoecological inference than either
ined the drift of aquatic macrophyte propagules along technique used in isolation.
the River Rhône during flood conditions, and were able
to demonstrate the importance of floods in propagule
4.4.6 Palaeoenvironmental applications
dispersal, while Dilcher et al. (2009) compared the charac-
teristics of leaf litter on the floor of a Florida swamp
of plant macrofossil studies
woodland with those of leaves in the adjacent woodland
4.4.6.1 Palaeoclimatic reconstructions
canopy, from which the litter was derived.
Other modern analogue studies have focused on the Climatic inferences can be drawn from plant macrofossil
relationships between the mix of plant remains that initially remains in a number of different ways. First, the indicator
accumulate on land surfaces, and those that survive decay species and transfer function approaches used to derive
and predation to become preserved in sedimentary pollen–climate relationships (section 4.2.6.5) can also be
sequences, in order to determine how well observed death applied to plant macrofossil assemblages. The former was
assemblages reflect the local or regional vegetation cover used by Isarin & Bohncke (1999) to infer mean July
from which they were derived. For example, Goman (2001) temperatures of 12–13°C (c. 5°C below those of the present)
studied the seed and vegetal remains dispersed across in western Europe during the Younger Dryas cold period,
modern tidal marshes in San Francisco Bay to develop a while the latter, when applied to deposits of Eemian age
calibration dataset of seed assemblages for the last 4,000 from a site in Germany by Kühl et al. (2002), indicated that
years, while Kalis et al. (2006) linked fossil plant assem- summer temperatures were similar to those at the present
blages recovered from a fen in the Vosges Mountains, day. The mutual climatic range method, developed for
France, with typical phyto-sociological communities cur- deriving quantitative climatic estimates from fossil beetle
rently occupying the region, and Davidson et al. (2005) assemblages (section 4.5.4.2), can also be applied to fossil
compared documentary evidence of changes in the com- plant remains (Sinka & Atkinson, 1999). The precision with
position of submerged vegetation in a small lake with which former climatic conditions can be reconstructed
plant macrofossil assemblages in the lake sediments that using these methods depends, however, on the number of
accumulated over the same period. Modern analogue taxa in a fossil assemblage for which modern climatic
studies such as these are an important means by which parameters can be quantified, and on the range of climatic
conditions that each indicator species can tolerate (Pross considered to be linked to variations in local hydrological
et al., 2000). conditions (bog surface wetness) which are, in turn,
A different approach to palaeoclimate reconstruction influenced by regional climatic conditions (section 3.9.4.1).
involves the analysis of stomata (guard cells) found on the The composition and state of preservation of plant
surface (epidermis) of plants, as both stomatal density macrofossils in ombrotrophic peats provide a basis for
(number of stomata per unit area of leaf) and stomatal measuring these relationships (Chambers et al., 2012) and
index (percentage of epidermal cells that are stomata) vary are reflected, for example, in changes in the proportions
inversely with atmospheric CO2 concentration (Beerling & of different species of the bog moss Sphagnum, since each
Royer, 2002; Hetherington & Woodward, 2003). Measure- species has a different tolerance to bog-surface wetness,
ment of these indices on fossil leaves enables past variations or in the relative importance of macrofossils of plants
in atmospheric CO2 concentration to be inferred, and these requiring drier bog conditions, such as Erica (e.g. Langdon
data complement those derived from other methods, such et al., 2003). These changes can be quantified using
as the analysis of gas bubbles in polar ice cores (section weighted averaging methods, and the results often show a
3.11.3). The results not only contribute to global models correlation with independent temperature reconstructions
of environmental change (Chapter 7), but provide (Barber & Langdon, 2007). Moreover, a number of peat
additional insights into the possible role of atmospheric macrofossil studies reveal cyclic variations in bog-surface
CO2 variations as a driver of climatic changes during the wetness during the Holocene (Figure 4.20), which may
Late Quaternary (Mcelwain et al., 2002). reflect solar variations (e.g. Mauquoy et al., 2004; section
A third line of evidence for past climate change 7.6.4.1).
comes from the analysis of plant remains in bog peat. Finally, plant remains concentrated and preserved
The stratigraphy of ombrotrophic bogs has long been through the actions of animals can provide valuable climatic
Scotland
Temple Hill Moss
Langdonetal £ 0 03
Kentra Moss
Ellis a n d Tall is 2000
NW Scotland"
Baker e t a l . 1999
Moine Mhor
Barber e t a l 1999
Talla Moss, Scottish Borders
Chamber s et al. 1997
Scottish Borders and Cumbria
Sloneman 1993
Northern England
Northern England
Hendon e t a l 2001
May Moss
Chiuenell 2001
Walton Moss
Hughes et al 201)0
Coom Riga Moss
Charman e t a l . 1998
Coom Rigg Moss

Felecia Moss
Mauquoy a n d Barber 19gg
Bolton Fell Moss

Barber 1961. Barber et al. 1994
0 1000 2000 3000 4000 5000 6000 7000 8000

Years BP
Figure 4.20 Episodes of climatic deterioration in northern Britain during the past 7.5 ka based on plant macrofossil evidence for
increased bog-surface wetness. Thin bars represent wet shifts and thicker bars wet phases (from Langdon et al., 2003). See also
Figures 3.53 and 3.54.
information in places where there are few alternative lines 4.4.6.3 Charcoal and fire history
of evidence. For example, remains of the aquatic plants
Najas flexilis and Typha latifolia found among a rich array Some regions, such as the western mountains of North
of other plant macrofossils in a beaver dam in central and South America, experience regular wildfires as part of
Alaska indicate that temperatures were significantly warmer a natural ecological cycle, and records of historical fire
during the early Holocene than today (Robinson et al., activity are reflected in fossil charcoal records. Analysis of
2007), while in the desert regions of the USA, plant macro- the charcoal content of lake sediments in the North Cascade
fossil remains preserved in packrat middens (nests) provide Mountains in Washington State, USA, for example, suggests
a record of changes in vegetation and climate, notably significant changes in fire frequency throughout the last
precipitation, in some instances extending back over 40 ka 10.5 ka, with climate probably the key driver (Prichard
(Jackson et al., 2005). et al., 2009). Charcoal studies from the Argentinian Andes
also implicate climate as the principal forcing factor in
fire frequency, with increases in fire incidence coinciding
4.4.6.2 Forest history with increased ENSO strength (section 7.6.4.2) and greater
A long-standing debate in Quaternary palaeoecology climatic variability (Whitlock et al., 2006). The relationship
concerns the altitudinal and latitudinal positions of former between fire and climate may not always be straightforward,
treelines. During full glacial conditions, temperate trees however, because changes in forest composition also
are believed to have receded to ‘refugia’ in lower-latitude modulate fire dynamics by altering the mix of species
warmer areas, but the location and nature of many of susceptible or resistant to fire (Prichard et al., 2009). In
these refugia remain to be established. Determining forest addition, local factors may over-ride regional climatic
limits during warmer episodes has also proved problematic. controls, while the intensity of fire damage also varies at
The main difficulty has been the reliance on pollen- the local and regional scales (Gavin et al., 2007). A further
stratigraphic evidence for detecting past treelines, because complication in the interpretation of charcoal records is
the proportion of trees can be over-estimated in pollen that charcoal fragments can be transported over consider-
assemblages obtained from areas with few or no trees able distances, due to the powerful convection currents
(Eide et al., 2006), but also under-estimated in pollen generated by intense forest fires (Pisaric, 2002). Hence,
records obtained from wooded areas (Opgenoorth et al., although charcoal stratigraphy in lake sediments provides
2010). Macrofossil tree remains, on the other hand, are a basis for reconstructing fire histories, care is required in
unequivocal evidence for tree presence, and are providing the interpretation of the data.
important new insights into former forest refugia. It has In many areas of the world, fire histories reflect human
long been assumed, for example, that the key refugia in activity as opposed to climate. Throughout North America,
Europe during the Last Glacial Maximum lay in areas such for example, there is abundant ethnohistorical and charcoal
as the Balkans, Italy and Iberia, but plant macrofossil data evidence for burning, with fire being used as part of hunt-
now suggest that coniferous and some deciduous trees ing practices to drive animals, to improve pasture and
were growing much further north and east than previously human mobility, and to increase species diversity of food
envisaged (Willis & van Andel, 2004). Macrofossil evidence plants (Boyd, 1999). Anthropogenic burning inferred
also shows that both black spruce (Picea mariana) and white from charcoal records has also been described, inter alia,
spruce (P. glauca) were able to regenerate in an unglaciated from the British Isles (Barton et al., 1995), Madagascar
enclave of Yukon Territory, Canada, after the onset of (Burney et al., 2004) and Australia (Wroe & Field, 2006).
harsh glacial conditions between 26 and 24 ka when most The principal difficulty with attributing fire to human
of the territory was arctic tundra (Zazula et al., 2006). agency, of course, is that in many areas, with long,
Similarly, in the Himalayas, near the highest known seasonally dry climates, such as the American Midwest, it
treelines in the world, macrofossil evidence indicates that is difficult to distinguish between natural and anthropo-
trees survived in small refugia throughout the last cold stage. genic burning from the charcoal record alone (Bell &
These small protected enclaves of persistent tree growth Walker, 2005). This is exemplified by evidence from
through harsh glacial conditions are referred to as cryptic Australia where fire incidence after the arrival of humans
refugia or glacial microrefugia (Opgenoorth et al., 2010), from c. 60 ka onwards has been interpreted on the one hand
and woody macrofossil evidence from these localities is as indicating burning as part of a hunting strategy which
providing new insights into the history of trees and how led to the eventual extinction of the Australian megafauna
they responded to the major climatic changes of the Late (Miller et al., 2005), while on the other it has been suggested
Quaternary. that the fire regime in Australia has been predominantly
forced by climate, with a strong similarity in temporal Many different orders of fossil insects have been observed
variation to the Dansgaard–Oeschger cycles of the last cold in Quaternary deposits, including bugs (Hemiptera–
stage (Mooney et al., 2011; section 3.11.4). Homoptera), two-winged flies (Diptera), caddis flies
(Trichoptera), stink bugs (Pentatomidae), seed bugs
(Lygaeidae), leaf hoppers (Cicadelidae), bees, ichneumons
4.4.6.4 Archaeological records
(Hymenoptera), dragonflies (Odonata), non-biting midges
Humans have altered natural vegetation in a variety of ways, (Chironomidae), water striders (Gerridae), shore bugs
through woodland clearance for pastoral and agricultural (Saldidae), water boatmen (Corixidae), backswimmers
activities, exploitation of aquatic resources, land irriga- (Notonectidae) and beetles (Coleoptera) (Elias, 2010),
tion and reclamation, domestication of grain and crop while members of the ant family (Formicidae) have also
species, and transfer of selected plants to new territories been investigated. Although not technically insects, the
as ‘exotic’ newcomers; evidence for all of these changes can mites (Arachnida), especially oribatid mites, have been
be discerned in the plant macrofossil record (Jacomet, considered together with insect assemblages, and are
2007). Plant macrofossil data augment inferences drawn particularly important components of the soil fauna of
from pollen records about anthropogenic activity, and the arctic and alpine ecosystems (Elias, 1994).
two are collectively referred to as archaeobotanical It is the beetles, however, that are the best-known of this
evidence. Macrofossils may provide important dietary array of fossils (Figure 4.21). This is because in a body
information; for example, remains of wild fruits found of sediment they are often visible to the naked eye and
in sediments from Neolithic and Bronze Age dwellings at commonly display brilliant and often iridescent colouring
Lake Neuchatel, Switzerland, reveal that the occupants of blues and greens. As a consequence, their presence in
consumed Prunus spinosa (sloe), Cornus sanguinea (dog- Quaternary sediments has long been a source of fascination
wood), Malus sylvestris (apple), Rubus species (raspberry/ to laymen, naturalists and entomologists alike. More
dewberry/blackberry), Fragaria vesca (wild strawberry), recently, however, and largely through the pioneering work
Rosa sp. (hip), Quercus sp. (acorn), Corylus avellana of Russell Coope (see section 4.5.4), fossil Coleoptera
(hazelnut) and Fagus sylvatica (beechnut) (Karg & Märkle, have proved to be a powerful tool for the investigation of
2002). Plant macrofossil remains also provide evidence Quaternary environments, and particularly of past climatic
of the earliest farming activity, with charred cereal grains conditions. Although fossils of the other insect groups
indicating cultivation in the Euphrates Valley, Syria have occasionally proved useful in palaeoecological studies,
around 13 ka (Hillman et al., 2001), while in the New such as caddis flies and oribatid mites (Greenwood et al.,
World, peanut, squash and cotton appear to have been 2003; Hodgson & Convey, 2005), they tend to be less
grown as early as 10 ka in the Peruvian Andes (Dillehay well preserved due to their delicate structures, some of the
et al., 2007). Other archaeobotanical evidence reveals the groups are difficult to identify beyond generic level, and
nature of plants used in the dyeing of fabrics (Hall, 1996), their present-day distributions are less well known than
the development of plant-based fabric industries, such as those of the Coleoptera. As a consequence, the remainder
flax and hemp (Viklund, 2011), and the types of plants of this section is devoted mainly to Coleoptera, although
adopted for sacrificial ceremonies (Megaloudi, 2005). In Chironomidae are also considered as they too have proved
addition, evidence of routine domestic practices such as the to be useful palaeoclimatic indicators.
preparation of animal fodder, and the plant types and
parts preferred for bedding, tempering agents, insulation
and roofing, may also be preserved in the archaeobotanical
4.5.2 Coleoptera
record. Coleopteran remains are usually the most diverse group
of insect fossils in Quaternary deposits and often they
are the most abundant. The chitinous exoskeletons of
4.5 FOSSIL INSECT REMAINS which they are composed are highly robust and contain
sufficient structural detail to permit a large number of
4.5.1 Introduction beetle fossils to be identified to species level (Figure 4.22).
Fossil insects are often abundant in a wide range of They have been collected and studied by entomologists
Quaternary deposits. Typically these include sediments in many parts of the world and there is now a consider-
that accumulated in ponds or near lake margins, in able body of knowledge in the form of atlases and
backwaters of rivers, in peats or indeed in any depositional monographs that summarize their distribution and
environment conducive to the preservation of plant debris. ecological associations. The order Coleoptera is one of the
4.5.3 Laboratory methods

In the laboratory, insect fragments can be removed by hand
from the sediment matrix where, for example, they occur
on bedding planes in clays or felted peats. More frequently,
however, flotation techniques are employed (Elias, 2007a).
The most commonly used method involves disaggregation
of the sediment in water or in a sodium carbonate solution,
followed by sieving (300 μm), and the residues remaining
on the sieves are then mixed with kerosene (paraffin),
which adheres to insect sclerites but not to plant detritus.
When water is subsequently added, the insect remains
float to the surface, along with some plant macrofossils.
The floating fraction is decanted, washed gently in detergent
and dehydrated in 95 per cent ethanol before being sorted
under a low-power microscope. The insect fossils range in
size from less than 1 mm to several centimetres, and close
examination may be necessary at this stage in order to
ensure that very small specimens are not overlooked and
that the subsequent collection is not heavily biased in
favour of the more conspicuous species.
The fossil remains are then glued onto cards or stored
Figure 4.21 Subfossil coleopteran sclerites (see Figure 4.22)
in dilute (c. 20 per cent) alcohol and examined under a
recovered by flotation from peat deposits overlying a Bronze
Age occupation site near Ballyarnet Lake, Co. Derry, Northern microscope. Electron microscopy may be necessary where
Ireland. Elytra (wing-cases) dominate in this view with an examination of very fine structural detail is required for
occasional pronotum (p). 1. Pterostichus nigrita – a predatory identification of some species. Since the entire insect fossil
ground beetle (10–15 mm) widespread in marshes and is rarely recovered from the sediment body (commonly
water margins. 2. Cercyon atomarius – a small (1.5–4.5 mm)
dung beetle. 3. Helophorus spp. – a water scavenger genus,
identification is made on an elytron [wing cover] or thorax,
with most species occupying emergent vegetation in standing or even a small fragment of an elytron), keys to identifi-
or slow-moving water. 4. Trechus rubens – a small (5–6.5 mm) cation are of limited value and identifications require
subterranean ground beetle, under stones and leaf piles on careful comparisons with modern specimens. A compre-
riverbanks and lake shores. 5. Anobium punctatum – a beetle hensive comparative collection of modern beetles is there-
that bores into and feeds upon wood (3–4.5 mm). 6. Hydraena
britteni elytrum (6e) and pronotum (6p) – a water beetle fore essential for palaeoecological research. Some parts of
common in shallow lake margins (photograph by Nicki the fossils possess few diagnostic features, but in many cases
Whitehouse, Plymouth University, UK). the heads, thoraces, elytra and genitalia (particularly in the
male specimens) display a wealth of characters, which
enable specific determinations to be made (Figure 4.22).
Data from fossil insect analysis are usually presented in
largest in the animal kingdom, accounting for 25 per cent the form of a species abundance list showing numbers
of all known species of organisms. They form the most of individuals occurring within a particular sample. The
important insect order, with more than 300,000 known results tend to be presented in tabular form because of the
species, 30,000 of which occur in North America while size of the overall dataset, although subsets of particular
within Britain alone there are over 3,800 named species. interest, such as stenothermic species, may be displayed
They occupy a very wide range of habitats, having colonized graphically (Figure 4.23). Occasionally, further information
almost every terrestrial and freshwater niche, some even is provided in species lists on the specific parts of insects
being found within the intertidal zone. Many of these that have been recovered, for example heads and elytra. The
species are stenotopic, which means that they show a numbers listed are the minimum numbers of individuals
marked preference for particular environments (e.g. those that are represented by the recorded skeletal parts. Thus a
adapted to narrow temperature ranges or specific habitats collection of three elytra, one head and two thoraces of the
or substrates), and it is this characteristic above all others species Olophrum fuscum (Grav.) obtained from one
that makes the Coleoptera such valuable palaeoecological stratigraphic horizon would indicate a minimum number
indicators. of two individuals of that species.
a) HEAD
PROMOTUmI
ELVTfa:
ChvUlACT
ER;
b) CHAJlACTEf!. rAcrosculpruie
'Dectn
'Dectn
/Shape of Mflsh
/Sue -J" MW1
f o w or Sndfcw Meshes
character:
F\nclufecl
F\nclufecl cc Smooth
cc Smooth ChvUlACTER;
ChvUlACTER;
fi&n t Angla-.lnjlwa l Sino, ChvUlACTER;
ChvUlACTER;
VAflttBlK:
'Dectn
/Shape
/Angle
Jtaferai ifnpisssori
* Shakw n Deep
CHARACTER: La1erg-f>asal Fovea

VAJHASLES; ChvUlACTER;
Hind Anata &
'Dectn Latefoflaaal Punch™
'Shape VARA
I BLES:
JWm or Wttwxri Stnae Shade of Ang*j
F al Outaae Margin Ppjiticm Of PuncMe
F\nclufecl cc Smooth $i» arm $hace of Pundufe
fieKJtion K? LatefQ-Saial fovea:
c) CHARACTER: Shoulder Toolh

VARIABLES:
Size a' foo h_
Shape of footh CHARACTER: Setaceous Punctures

Angle of shoulder
VAU A3_fcS
Strtallon past shoulder?
Number of Punctures
Posittcn on Erytron
Size of Puncture
Shape of Puncture
D e e p vs. Shadow F\inctures
CHARACTER; Striahons
VARIABLES:
CHARACTER: MfcfOsculpture
Deptn
VARIABLES:
| Broken vs. Continuous
Icurved vs. Straight Shape of Mesh
Convexity ot Intervals Size of Mesh
belween Sirioe Deep a Shallow Meshe*
Iriaescent Potterra
Dull vs. Shiny Surface
CHARACTER: ElylrOl Apex
•/ARABLE S
Shape (Truncate l a Eiongote)
Angle of Apex
Do Striae Reach Apen?
Position of Setaceous
PuriClures orf Aoex
Figure 4.22 Generalized drawings of coleopteran sclerites frequently preserved as Quaternary fossils, showing a range of
diagnostic features used in fossil identification: a) dorsal surface, b) pronotum and c) elytron of a ground beetle (Carabidae) (from
Elias, 1994).
^Thermophiles; Cold-adapted species

cm cm cm
hs s s
cmm cm cm
e pt cm pctmh p th cm c m
s h s hs m
cdm cm s thdse cmcm de s
th cm pth s cm
s h c t mt c
s
l e s t h lpe m s l e p e pt ths cm c m dep dsecp hs
h h p pe mc tmh hms pt m
cm aecm ptpecpmt dcem m s p m p mde mhs dme tmh d p s
cpmlcem dcem pctmh m sm le plcetm
thc pcm
h m
s S
sd d
s hes slaecmdhpsthchdse c esptc s c s acm s
p lec sepct pslec dsepc s s l es e c hps s
m m
d esp sdce s c m
h
l S
h t p te t dt h h S th pthd pmth lpeth pthpatmh ppth ethd hs s e h
h
pt plteh pplte ep pmth epd pelpe leep ept pth pm pdt tah lepat ptle th c
de adme adme Se admdSea epdleaemd m pd d de d
e
Sea dpel dSea mdpe d edSe daem pelp p t
pdlee depS mdpSe adme dep ths
l l l le em e ae le e eS m d e e ae e p
e
pl plSe pSle mp ple mapmmpS mSpa ple ple pl
e
pSla mpl Spl mp pl mpl pSlae le
d pml le S l plS le e
m m m a m a S a a m m m m m m m p aSma mp ampam mp le d
Sa Sa Sa S Sa S S S Sa Sa Sa Sa a
S Sa Sa Sa a
Sa Sa m S Sa S S S p
Sa m
Sa FU-5
128
122
117
112
106
102
99
94 FI M
88
82
78
72
63
61
56
47
40 FU-3
35
28
20
14
9
4 FU-2
0
20 40 20 20 20 20 20 20
Figure 4.23 A coleopteran record from the Lateglacial (c. 14.7–11.5 ka) site of Llanilid, south Wales, UK. The warmest part of
the sequence (early Interstadial) is represented by faunal unit FU-2, cooler conditions are recorded in FU-3 (later Interstadial) and
marked by the arrival of cold-adapted species, while the cold Younger Dryas/Loch Lomond Stadial is characterized by an increased
number of cold-adapted species. Note how these disappear in the early Holocene (FU-5) to be replaced by thermophiles (from
Walker et al., 2003).
A useful aid for the study of palaeoentomology, entom- of deposits, but they appear to combine both evolutionary
ology and ecology is the Bugs-CEP web site: http://www. and physiological stability with acute sensitivity to climatic
ncdc.noaa.gov/paleo/insect.html or http://www.bugscep. change (Elias, 2010). In addition, they frequently display a
com. This provides a comprehensive set of search and marked preference for very restricted environmental niches
reporting functions, including pictorial images of species, or conditions.
information on their habitat and distributions, guidelines Our present understanding of the ways in which fossil
for data collection and storage, and tools for climatic and beetles can be used to reconstruct Quaternary environ-
environmental reconstruction (Buckland & Buckland, ments stems, in large measure, from the work of Russell
2006). Coope (Ashworth et al., 1997). He noted, in particular,
that the skeletal elements of the fossil insects showed a
remarkably close similarity in shape and fine anatom-
4.5.4 Coleopteran analysis and Quaternary
ical detail to the corresponding parts of living species,
environments and that this remained the case irrespective of the age
Coleoptera exhibit a number of characteristics which make and context of the Quaternary assemblages (Coope,
them one of the most valuable components of the terres- 1967). Subsequent research has confirmed his view that the
trial biota for the reconstruction of Quaternary environ- vast majority of beetle species have remained morpho-
ments. Not only are they relatively abundant in a wide range logically identical, possibly since Miocene times; moreover,
there is only one confirmed record of species extinction isolated for long periods (Previsic et al., 2009; Marra &
during the Pleistocene (Elias, 2010; Marra & Leschen, Leschen, 2011). In such areas, therefore, the habitat
2011). This apparent long-term evolutionary stasis preferences of some beetle species might possibly have
(anatomical stability) initially seemed counter-intuitive, for changed over the course of the Quaternary. It is also worth
it was assumed that the cyclical climatic changes that noting that the current global databank of Quaternary
characterize the Quaternary would have resulted in a high fossil beetle records is biased towards northern temperate
species turnover. Coope’s research suggested quite the and Palearctic regions and to records from the last 15 ka.
opposite: repeated environmental disruption appears to The database may therefore reflect a greater emphasis on
have engendered stasis, at least as far as beetles were species that were forced to migrate over long distances as
concerned (Coope, 2004). a consequence of repeated expansion and demise of the
Coope’s second key observation related to the mech- great ice sheets. In other areas, by contrast, past changes in
anism by which stasis in beetle populations was main- the geographical and temporal ranges of beetle species
tained, for he noted that the ecological preferences of appear to have been much more conservative (Abellán
most coleopteran species also appear to have remained et al., 2011).
unchanged during the Quaternary. The available evidence
suggested that, in the great majority of cases, species of
4.5.4.1 Habitat preferences
beetles are found in similar associations in both fossil
and living assemblages. In Britain, for example, the warm- Any assemblage of fossil insect remains will contain
adapted insect assemblages of one interglacial period are species from a variety of local habitats. Botanical factors,
essentially similar to those of others, even though the soil type, microclimatic environment, hydrological condi-
interglacials may differ in age by hundreds of thousands tions and chemical variations will all restrict the distribu-
of years (Coope, 2010). The beetle assemblages of various tion of insects at the local scale (Elias, 1994, 2010). To a
cold stages, in particular, have a large number of species in palaeoecologist who is interested in the environmental
common. Moreover, independent palaeobotanical and history of a particular site, therefore, it is important that
geological evidence indicates that most fossil beetle species the range of habitats represented by the fossil assemblage
were associated with environments that are similar to those is identified and, as far as is possible, quantified. Some
that they occupy today. It does seem, therefore, that in the beetle species are substrate dependent, such as Bembidion
majority of cases, physiological stability of coleopteran obscurellum, which lives on dry, sandy soils; Dyschirius
species was accompanied by morphological and ecological globosus, which requires moderately humid soils with clay,
constancy throughout the greater part of the last two sand or peat and a sparse vegetation cover; and Bembidion
million years, and exceptions to this rule are comparatively schueppeli, which is restricted to river banks. A large number
rare (Coope, 1977). The key point is that beetles, being of beetles are associated with aquatic habitats. Thus actively
highly mobile, did not need to evolve to cope with marked flowing water is indicated by Esolus, Limnius volckmari
climatic or ecosystem change: they simply moved to main- and Ochthebius pedicularius while Potamonectes depressus
tain their association with migrating climatic and vegeta- (elegans) and Halyplus obliquus live in clear ponds with
tional zones, thereby remaining within their ecological sandy or silty bottoms. Other beetles indicate the pres-
comfort zones and avoiding pressures to adapt. ence of particular plants or other animals upon which
Although the evidence for morphological stasis appears they depend for food. Most of the staphynilid beetles,
strong, however, recent developments in molecular re- for example, are predators, living on small arthropods
search are beginning to reveal that more speciation of and worms in leaf litter. A profusion of dung beetles in an
beetles may have taken place during the Quaternary than assemblage would indicate the local presence of mammals;
has been recognized hitherto. For example, on the Iberian indeed, some beetle species are selectively associated
Peninsula, eighteen to nineteen new species appear to with the dung or decaying carcasses of particular animals,
have emerged during the Quaternary (Ribera & Vogler, for example mammoths (Allen et al., 2009). A number
2004), while in North America, tiger beetles have continued of phytophagous Coleoptera feed only on reeds, such as
to diversify over the course of the last 5 Ma, and at an several species of the Donacia genus which live on Carex,
increasing rate during the Quaternary (Barraclough & Scirpus, Sparganium and tall marsh grasses. Hydnobius
Vogler, 2002). However, most studies suggest that beetle punctatus feeds on fungal hyphae, Hypera postica is
speciation during the Quaternary was confined to low- dependent upon various leguminous plants, such as
latitude areas not directly affected by glaciation, or to Medicago, Melilotus and Trifolium, and Simplocaria semi-
mountain micro-refugia, where beetle faunas became striata feeds exclusively on moss.
Beetle assemblages can therefore provide valuable beetles have been used to augment palynological evidence
information on a diverse range of contemporaneous for the presence of trees and to establish the changing
habitats or co-dependent biota, and may provide environ- composition of woodland over time (Figure 4.24; section
mental insights that are difficult to obtain from other 4.5.4.3). Fossil insect assemblages preserved in packrat
lines of evidence. For example, large numbers of the middens in the deserts of the southwest USA contain a
genus Bledius were found among the earliest colonizers mixture of temperate and desert species not found together
of the surface of the lower cover sands (deposited during in any part of North America today (Elias, 2007b). This may
the last cold stage) in the Netherlands in deposits contain- indicate a late Pleistocene climatic regime for which there
ing remains of the blue-green alga Gleotrichia (Van Geel is no modern analogue.
et al., 1989), an important nitrogen fixer. The algae may In addition to these macro-scale environmental re-
have provided an important food source for these beetles constructions, mosaics of former micro-habitats may
at a time when few other organisms had succeeded in also be inferred from fossil beetle assemblages. For example,
colonizing the area. Large numbers of Bledius recovered Olsson & Lemdahl (2009) used fossil beetle records from
from early Lateglacial sands in a site in eastern England southern Sweden to amplify the history of woodland
may therefore indicate that nitrogen-fixing algae played changes during the Holocene, with different species
an important role in the colonization process at that site groups providing evidence of variation in woodland density
also (Walker et al., 1993). Some beetle species are obligate and diversity, of the amount of deadwood on the ground,
dwellers in tree bark or leaf litter, and fossil remains of such of the impacts of grazing and fire, and of the eventual
9500-6000 cal BC 6000^1000 cal BC 4000-200

0 cal BC
12
10
6
"Oak / beech
Pine
4 Willow
Birch
Elm
2 Ash
"Hazel
Lime
0 Alder
Etton •
Silbury •
Sweet Track •
Archaeological sites
Atlas Wharf'
Thome Moors II
Lea Marsdon B
Bole Ings C 25-24

Bole Ings C 29-26
Westwood Ho 26 & 29
Hatfield HAT 3
Bole Ings C 17-23

Thorne Waste
Goole blkw/pine
World's End
Holywell (s 12,13)
Holywell (s 14)
Mingles Ditch (0)
Rossington (19-20)
West Heath Spa (WS1)
Tyrham Hall (Oak)

West Heath Spa (WS2)
Tyrham Hall Quarry (pine)
Hatfield Lind B{18&15)

Rossington (18-17)
West Bromich
Croft Neolithic
Goole blkw/oak
Etton i-k
Hatfield Lind A
Goldcliff
Langford
Shustoke
Rowlands
Baker
Stileway
Runneymede
Figure 4.24 Number of obligate beetle species (key indicates tree affinities) recorded in Holocene sediment sequences from
southern and central England. The site names are listed along the base; archaeological sites are indicated by closed circles; the
remainder are considered ‘palaeoecological’ (not disturbed by archaeological activities). For further explanation see text (from
Whitehouse & Smith, 2010).
replacement of woodland by heathland at c. 2.8 ka. 4.5.4.2 Palaeoclimatic inferences based on

Similarly, Davis et al. (2007) were able to decipher small- coleopteran assemblages
scale variations in the topography and channel patterning
of lowland river flood plains from the characteristic fossil One of the most important factors that has governed the
beetle assemblages that occupy different flood-plain micro- distribution of most insect species during the Quaternary
habitats, while Kuzmina et al. (2011) used evidence of has been climate, particularly thermal conditions (Coope,
micro-habitat preferences of modern beetles in present-day 1990, 2004). Distribution maps of modern beetles show
arctic tundra regions to reconstruct climatic gradients for that the geographical range of many species corresponds
the former periglacial zone of the exposed continental with well-defined climatic zones (Figure 4.25) and especially
shelf around the Bering Sea during the last cold stage. with summer temperature thresholds. Those insect species
These various examples show how a knowledge of the aut- whose distributions are narrowly restricted are termed
ecology of beetles can provide valuable additional insights stenotherms, while those that can tolerate a broader range
into Quaternary palaeoenvironments at a range of spatial of climatic conditions are termed eurytherms. The former
and temporal scales. are much more important in palaeoclimatic research,
(a) Asaphidion cyanicorne (b) Bembidion ibericum
(c) Helophorvs sibericus (d) Boreaphilus

henningianus
Figure 4.25 Present-day European distributions of four coleopteran species found in Lateglacial deposits (c. 14.7–11.5 ka) at
the site of Glanllynnau, north Wales (from Coope & Brophy, 1972).
since they enable more precise inferences to be made about method, in which the modern distributions of species
former temperature regimes. The acute sensitivity of (indicator species) represented in a fossil assemblage are
beetles to temperature variations is reflected in the growing plotted, and the zone of overlap of the ranges is identified
body of evidence that reveals how beetles have adjusted (e.g. Coope, 1959). Modern climatic statistics obtained
their modern geographical ranges in response to global from meteorological stations that lie within the zone of
climate change within the last century or so (e.g. Parmesan, overlap, such as mean annual temperature or the annual
1999; Hill et al., 2011). temperature range, can then be used to provide a quantita-
As with most other lines of fossil evidence, however, there tive estimate of the regional macroclimate that prevailed at
are problems in utilizing fossil beetles as climatic proxies. It the time the fossil taxa coexisted. In practice the method
can never be established for certain that an insect species works best when a large number of stenothermic species
has colonized the entire climatic range to which it is suited, are represented in an assemblage, and this often enabled
nor that past distributions were entirely in equilibrium relatively precise palaeoclimatic estimates to be inferred.
with the prevailing climatic conditions. On the other hand, The main difficulty with this approach, however, is that a
Coope (1977) has argued that the enormous scale of the species might not occupy its full potential geographical
changes in geographical distribution of species in response range. Furthermore, some taxa may temporarily coexist
to climatic shifts during the Quaternary, and the rapidity during a transitional phase of adaptation to new climatic
of many of these changes in species range, indicate that conditions, and the resulting mix of fossils may therefore
Coleoptera must have been able to colonize new available be largely an ephemeral one, with no modern analogue
habitats extremely quickly. In the majority of cases where (a non-analogue assemblage). This problem may be
the range limit of a coleopteran species coincides with a particularly acute in the interpretation of fossil insect
climatic boundary, this relationship has, therefore, long been assemblages derived from sediments that accumulated
used to derive quantitative palaeotemperature estimates. For during an episode of abrupt climatic change, since insects
the British Isles, Coope (1987) classified fossil coleopteran appear to have responded much more rapidly than other
species ranges into eight categories: biota to changing climatic conditions. One of the challenges
of Quaternary palaeoecology, therefore, is to distinguish
(a) southern European species; such temporary associations from those representing more
(b) southern species whose normal ranges just fail to reach stable climatic episodes, and, while meeting this challenge,
Britain; to note that non-analogue faunas do not necessarily imply
(c) southern species whose normal ranges are south of non-analogue environments.
central Britain; In an attempt to avoid the errors that may arise from
(d) widespread species whose normal ranges are north of the use of the indicator species approach, the mutual
central Britain; climatic range (MCR) method was developed to obtain
(e) boreal and montane species whose normal ranges more representative palaeotemperature estimates from
extend down into the upper part of the coniferous beetle records (Atkinson et al., 1987). This is an extension
forest belt; of the range overlap method, but it employs the ranges
(f) boreal and montane species whose normal ranges are of all of the taxa included. Modern distribution maps
above the tree line; are first obtained for as many as possible of the species in
(g) eastern Asiatic species, some of which also range into the fossil assemblage, and the climatic range of each beetle
North America; type is then established using contemporary meteoro-
(h) cosmopolitan species with very wide geographical logical data. The two most important variables governing
ranges. beetle distributions appear to be the temperature of the
warmest month (TMAX) and the temperature range between
Hence, an interglacial assemblage might therefore be the warmest and coldest months (TRANGE), the latter pro-
dominated by fossils from categories (a), (b) and (c) with viding an index of seasonality. By knowing the distribution
representatives of group (h), while cold-stage faunas in terms of TMAX and TRANGE, the geographical range of
are more likely to have a high representation of categories each species may be plotted in ‘climate-space’, and for each
(d), (e) and (f) with some elements of (h). The problem, species a ‘climatic envelope’ is thus produced (Figure 4.26).
however, is how to make a quantitative estimate of pre- For any fossil assemblage, therefore, the mutual climatic
vailing macroclimate from such data. range can be determined from a computer-generated plot
Early attempts to derive macroclimatic reconstructions of the climatic parameters relating to each beetle in the
from fossil assemblages employed the range overlap assemblage. From these plots, the values of TMAX, TRANGE
and TMIN (temperature of the coldest month) can be a)

obtained, and these constitute the ‘best estimates’ of the 25
mutual climatic conditions within which the particular
Reconstructed temperature
mix of fossils formerly coexisted. The method is most
successful (i.e. produces the narrowest range estimates) 2C
where an assemblage contains a large number of species,
and where a number of these are obvious stenotherms. 15
The method can be tested by deriving MCR values for
modern beetle assemblages and measuring their statist-
10
ical relationship with modern climate data (Figure 4.27).
The results usually show a stronger linear relationship for
summer than for winter temperatures, mainly because •
many beetle species can survive a wide range of winter
temperatures (Elias, 1997). The advantages of the MCR
0
method over the indicator species approach are that it 0 5 10 15 20 25
avoids subjective interpretations and possible bias, as well Observed mean July temperature
as over-generalization from the use of geographical over- a)
lays. Moreover, geographical range limits are often too 0

broad, and cannot take into account such factors as altitude,
oceanicity, microclimatic variations and so on. The MCR
approach ignores geographical location, and focuses Reconstructed temperature E
entirely on climatic parameters governing species dis-
tributions. Hence a complex geographical distribution may 10
be reduced to a narrow climatic range, reflecting the fact
that the often diverse geographical locations in which a
15
species occurs may, in fact, have common characteristics
when plotted in climate–space. Also, and most importantly,
it does not really matter if the species does not occupy its 2C
25
Thermal envelopes for hypothetical species A, B and C
25
T
Grange T
Grange T
Grange
3t
T
Grange
T
Grange T
Grange
30 25 20 •15 10 •5 0
Observed mean January temperature
T
Grange
Mutual Climatic Range
for species A, B a n d C
Figure 4.27 Test of the MCR method on assemblages of
species found living today at thirty-five localities in North
100% overlap America. The reconstructed mean July temperatures (TMAX) are
Partial overlap shown on the vertical axis of graph a); the reconstructed mean
No overlap January temperatures (TMIN) on the vertical axis of graph b).
Vertical bars represent the mutual climatic range of the beetles
found living at the localities studied. The horizontal axes of the
two graphs show the observed TMAX and TMIN values for the
thirty-five sites. The slopes and positions of the gradient lines
are linear regressions of predicted against observed values for
the test sites (from Elias, 1997).
T
Grange
Figure 4.26 Schematic representation of the mutual climatic

range (MCR) method of quantitative temperature recon-
structions (courtesy of Adrian Walkling).
full (potential) geographical range, so long as it reaches coincident with the Younger Dryas interval, again reflecting
potential climatic boundaries in a sufficient number of the prevailing influence on climate of the last ice sheets
places. (Elias, 2007a). The two exceptions are the Maritime Region
Climatic reconstructions based on beetle MCR data of Canada, where beetle MCR temperature evidence from
from the British Isles (Figure 4.28 main curve) bear a a number of Lateglacial sites shows a significant decline in
striking resemblance to those derived from Greenland temperature (c. 5°C) around 13.0–12.5 k BP, coincident
ice-core records (section 3.11.4), most notably in respect with the onset of the Younger Dryas in Europe (Miller &
of the abrupt increases (by around 7°C) in mean July Elias, 2000), and arctic Alaska, where again beetle data
temperatures at around 14.7 and 11.5 ka where the rate appear to indicate a Younger Dryas equivalent cooling
of warming may have been as rapid as 1°C per decade (Elias, 2000).
(Atkinson et al., 1987), and the episode of markedly lower The MCR method is not without its problems, however.
temperatures between c. 12.9 and 11.5 ka, in the Younger One drawback is the incomplete knowledge of the present
Dryas period or Greenland Stadial 1 (GS-1: section 1.6). geographical ranges of many species, especially in Asia.
So close is the match between the British and Greenland As a result, the palaeoclimatic significance of records of the
records, that it seems likely that both areas were climatically many cold-adapted Asian species that migrated to Europe
phase-locked at times during the Lateglacial period (Walker and Alaska during cold intervals is difficult to quantify
et al., 2012). This does not appear to have been the case for precisely (Zinovjev, 2006; Alfimov and Berman, 2009).
other parts of Europe, however, for MCR reconstructions A different constraint affects the use of the MCR method
from Poland, Sweden, Norway and France (Figure 4.28) in geographically isolated regions, such as Australia:
show significant differences in regional temperature although conditions appear to have been much colder
trends (Ponel et al., 2005). These largely reflect the delayed in the past than is experienced in any part of Australia
melting of the Scandinavian ice sheet compared with its today, these lower temperatures are not reflected in the
British counterpart, with the former maintaining cooler fossil beetle record, because no inward migration of cold-
climatic conditions around its periphery for longer. This adapted species was possible (Porch, 2010). Other problems
is clearly indicated by the beetle MCR data which reveal with the MCR method include the possibility that some
marked thermal gradients between Britain and Scandinavia modern species are prevented from attaining their full
throughout the Lateglacial period (Coope et al., 1998). climatic range because of physical barriers to migration,
In contrast to Europe, fossil beetle MCR data from many while the range of climate states that exists today may not
parts of North America indicate a relatively smooth include the extreme conditions or variants that some species
transition from glacial to interglacial climates from around could potentially tolerate (Porch, 2010). Finally, the MCR
17 ka onwards, and few sites record a climatic reversal method has been challenged on statistical grounds, mainly
because beetle data are unlikely to have normal (Gaussian)
distributions in climate–space, but are dispersed in more
20
complex ways throughout their ranges (Bray et al., 2006).
NF These and other difficulties notwithstanding, there is no
•
SI doubt that the MCR method has proved to be a valuable
15- CP tool in palaeoclimatology, and remains one of the key
SW;
T C
techniques for quantifying the magnitude and rate of past

WN
max
temperature change.
A recent development in the use of fossil beetles as
10 palaeoclimatic indicators involves the analysis of the
hydrogen isotope chemistry of the chitin in fossil beetle
carapaces (Gröcke et al., 2006). Data from modern beetles
suggest that this varies predictably with changes in mean
13 12 11 10 precipitation and annual temperature. The methodology is
Radiocarbon ka BP still at an early stage, however, and requires further experi-
mentation to establish the degree to which H-isotope varies
Figure 4.28 Generalized climatic curves of estimated TMAX both within and between beetle specimens in a systematic
variations during the Lateglacial–early Holocene for the British
and predictable way, and the extent to which these vari-
Isles (BI), central Poland (CP), southern Sweden (SW), western
Norway (WN) and northern France (NF) (from Coope & ations correlate with climatic parameters (Gröcke et al.,
Lemdahl, 1995). 2011).
4.5.4.3 Insect fossils and archaeology macro-environment, may contain microclimatic niches
that are very similar, and in some cases identical. Moreover,
Coleopteran assemblages, in common with a number of it would be expected that, in those insects that are mark-
other proxy biological indicators, clearly record the effects edly synanthropic, such as the grain weevil (Sitophilus
of human impact on the landscape of Europe during granaricus), slight physiological changes may have occurred
the mid-Holocene (Elias, 2010). Obligate woodland beetle to adapt to these artificial habitats. Although in most
species not only record natural changes in woodland cases the general archaeological interpretation will not be
composition throughout the Holocene (Figure 4.24) but affected, care clearly needs to be exercised in the palaeo-
also reveal the degree to which woodland in large parts ecological inferences that are drawn from entomological
of Europe was reduced in density or cleared completely records from archaeological sites based upon the known
to meet increasing agricultural or pastoral demands ecological affinities of modern insect species.
(Whitehouse & Smith, 2010). This in turn led to the Recent technological developments may help to resolve
elimination of a number of beetles dependent on natural some of these problems, and may also shed new light on
woodland cover and their replacement by species associ- past human activities. An important advance has been
ated with farming practices and human settlements in the extraction of DNA preserved in fossil beetle material
(Whitehouse, 2006). The latter include pests of grain as, for example, from exoskeletal fragments of Sitophilus
production (Smith & Kenward, 2011), dung and parasitic granaricus discovered in deposits of Roman and Medieval
beetles associated with domesticated or working animals age (King et al., 2009b). Although DNA in beetle remains
(Buckland et al., 2009), and beetles that are associated with is susceptible to contamination and requires relatively
human occupation, and are found in house floors, waste large specimens for analysis (Reiss, 2006), this approach
pits, roofing structures and fills, tanneries, cess or refuse may provide important new insights into the ways in which
pits, stables and textile stores (Kenward & Carrott, 2006; beetles have developed adaptational traits in response to
Kenward et al., 2012). Human activity is also reflected in human influences. A second potentially important devel-
the occurrence of synanthropic species (those closely opment for archaeology is the analysis of isotopic frac-
associated ecologically with humans), especially parasitic tionation within subfossil beetle chitin, as this enables
species or predators on human corpses (Panagiotakopulu inferences to be made about past farming practices, such
et al., 2012), while other bio-indicator species are indicative as transhumance, or changes in the types of herds or crops
of more specialized human practices, such as ancient managed by humans during different archaeological
mining and metal-working (Mighall et al., 2002). periods (King, 2012).
A number of factors serve to complicate the study of
archaeoentomology, however. Fossil beetle assemblages
from sites of human occupation often consist of a mix 4.5.5 Chironomidae
of fossils derived from several different habitats, or are Chironomidae are non-biting midges, a family of the
dominated by species that are not restricted to very specific Diptera (true flies), and their species abundance and
niches (Kenward & Carrott, 2006). Also, while most composition are related to such factors as pH, salinity,
assemblages from occupation sites contain a large pro- water depth, temperature, oxygen concentration and
portion of insect remains that originated from near the trophic status (e.g. Porinchu & MacDonald, 2003). This
point of deposition, they often also contain an element of makes them potentially valuable for Quaternary palaeo-
‘background’ fauna composed of insects that have flown environmental reconstruction, although it is their use as
to the site, or that have been derived from the regurgitated palaeotemperature indicators that has attracted the greatest
pellets or faeces of birds and other animals. Because of interest (Brooks, 2006).
the complexities of insect death assemblages in archaeo- Chironomidae produce larvae in the bottom of almost
logical contexts, therefore, large samples (over fifty species) all freshwater habitats, and these eventually develop into
should be analysed, in order that the full species divers- mature forms that consist of a robust, strongly sclerotized
ity is adequately represented and that the background head capsule made of chitin (Figure 4.29) and a body that
component can be isolated. A further difficulty arises, resembles a maggot. The final, adult stage is mosquito-like
however, where people have created a range of wholly but lacks a proboscis and a biting habit. It is the head
artificial environments, for ecological changes may have capsules of the larval stage that are often abundant and well
occurred in certain insect species as they adapt to these preserved in freshwater sediments. Most genera possess
new biotopes (Anderson, 2000). Many natural and artifi- head capsules with diagnostic forms, structures or surface
cial environments, while being very different in terms of markings that enable them to be identified, using both type
material and keys (e.g. Brooks et al., 2007b). Chironomid they can be picked individually using forceps under low-
head capsules can be separated from most sediment power magnification and mounted on slides in a suitable
matrices by procedures that include deflocculation in KOH medium, commonly Euparal or Hydromatrix (Brooks et al.,
(or, for calcareous deposits, HCl), the use of a sonic bath, 2007b).
and removal of fine sediment by sieving (Lang et al., 2003). A number of studies have demonstrated a close
Kerosene flotation, used to extract beetle remains, also relationship between summer air and lake surface water
works well with chironomid head capsules (Rolland & temperatures and the distribution of chironomid species in
Larocque, 2007). After concentrating the head capsules, arctic, subarctic, temperate and tropical zones of both
Figure 4.29 Head capsules of common chironomid taxa recovered from late Quaternary lake sediment sequences from
northwest Europe: a) Glyptotendipes severini-type (tribe Chironomini); b) Phaenopsectra flavipes-type (tribe Chironomini); c)
Cricotopus obnixus-type (Orthocladiinae); d) Heterotrissocladius grimshawi-type (Orthocladiinae); e) Prodiamesa olivacea-type
(Prodiamesinae); f) Guttipelopia (Tanypodinae); g) Stempellina (Tanytarsini); h) Micropsectra radialis-type (Tanytarsini) (images
provided by Steve Brooks, Natural History Museum, London).
hemispheres (Eggermont & Heiri, 2012). Calibration inferred and instrumental temperature measurements,
datasets have been generated that measure the strength which suggests that chironomids could provide useful
of the relationship between predicted (i.e. chironomid- proxy climate data over the course of recent centuries
inferred, mainly using weighted-average regression (Larocque & Hall, 2003).
methods) and measured summer surface water tempera- Interest in the palaeoclimatic potential of chironomid
tures. Research was initially concentrated in Europe (e.g. stratigraphy has grown considerably over the past ten to
Heiri et al., 2011) and North America (e.g. Porinchu fifteen years, principally because this approach offers two
et al., 2009), but has subsequently been extended to Africa, important advantages over other methods of Quaternary
Australia, New Zealand and South America (Verschuren & palaeotemperature reconstruction. First, many hundreds
Eggermont, 2006; Vandergoes et al., 2008). Some studies of head capsules can be obtained from as little as 1 cm3 of
have also shown a close relationship between chironomid- sediment. This should enable a higher-resolution record
Greenland Ice Core Chronology 2005, yr b2k

NGRIP core 10500 11000 11500 12000 12500 13000 13500 EH2D
Period Pre bo real 11653 Younger Dryas 12846 Allercd
G i - Gl- Gl-i
Isotope event Preboreal GS-1 Gl-1c
la ib id -
35
37
5 1 S 0 %o S M O W
39
-41
Reconstructed mean
oscillation
oscillation
-43
Chironomids
Maloja Riegel 16-
core
Aegeisee
Gerz ensee
14
July air temperature, °C
Reconstructed mean
10
Figure 4.30 Chironomid-inferred temperature variations for the Lateglacial sequence from a former lake near the Majola Pass
in the Swiss Alps, compared with the NGRIP oxygen isotope record and climatic event sequence (from Iliyashuk et al., 2009).
than is possible using other proxies, such as fossil Coleop- to be inferred (Ruiz et al., 2006). Elsewhere, they have
tera, where far larger samples of sediment are required. provided evidence of changes in lake productivity or degree
This is particularly well illustrated in chironomid-inferred of eutrophication (Brodersen & Quinlan, 2006), of flow
temperature reconstructions for the Lateglacial period, regime within fluvial palaeochannels (Howard et al., 2010),
which enable detailed comparisons to be made with and of palaeosalinity variations and hence changes in sea
Greenland ice-core records (Figure 4.30). Second, the level (Heinrichs & Walker, 2006).
method enables summer water-surface temperatures of
small lakes to be inferred, providing valuable microclimatic
information which can be compared with independently
4.6 NON-MARINE MOLLUSCA
derived ambient air or ground temperature estimates.
In some Lateglacial records, for example, the chironomid
4.6.1 Introduction
data suggest a steady increase in temperatures over intervals Mollusc shells are some of the most common fossil remains
for which pollen and other proxies suggest the opposite. in terrestrial Quaternary sediments (Figure 4.31) and
This is thought to reflect different forcing factors within consequently they have a long history of investigation.
the overall climatic environment, with the midges respond- As in other branches of palaeontology, much of the early
ing exclusively to summer season conditions, while land work in the eighteenth and nineteenth centuries was con-
plants were more sensitive to changes in seasonality or cerned with taxonomy and, by comparison, little consid-
to variations in annual precipitation (Lotter et al., 2012). eration was given to the palaeoecology of molluscan
Chironomids therefore have much to offer in palaeo- assemblages. In the late nineteenth and twentieth centuries,
climatic reconstructions, a potential that may be further however, workers in Britain such as Alfred Kennard,
enhanced by the discovery that δ18O abundance in head Bernard Woodward and Frederick Harmer began to utilize
capsules may also be temperature-dependant (e.g. Wooller molluscs as palaeoclimatic indicators, and also as a means
et al., 2004). of dating geological events (Keen, 2001). The increasing
While the palaeoclimatic applications of chironomid use of pollen analysis as a palaeoenvironmental technique
analysis have attracted most attention, the technique is led to a gradual decline in interest in molluscan studies in
proving to be useful in other areas of Quaternary research. the period before and after the Second World War, and the
In archaeological investigations, for example, chironomid modern phase of molluscan investigations did not begin
assemblages have enabled changes in lake levels resulting until the 1950s with the first quantitative analyses of shell-
from the construction of dwellings or drainage channels bearing deposits (Miller & Tevesz, 2001). This approach,
Figure 4.31 a) Fossil shells of freshwater molluscs (principally gastropods) exposed on an abandoned beach of Pluvial Lake
Lahontan (section 2.7.1) on the eastern side of Pyramid Lake, Nevada, USA (photograph by Allan Ashworth, North Dakota State
University, USA). b) Valvata piscinalis, a small gastropod (up to 5 mm height and width) that inhabits streams, rivers and lakes,
preferring running water and tolerant of low calcium levels. Such small specimens can be readily identified, when magnified, by
examination of morphological and ornamental details, some of which are illustrated (photograph by Jenni Sherriff, Royal Holloway,
University of London, UK).
developed initially for the investigation of non-marine large calcareous, external, coiled shell, through the squids,
Mollusca, has since been applied to the study of marine which have only a thin horny vestige of a shell embedded
shells (section 4.7) and, over the last thirty years or so, has in the mantle, to the octopuses which have no shell. Some
produced a considerable body of data on local environ- gastropods breathe by gills (prosobranchs) and are mainly
mental and climatic changes during the Quaternary. More aquatic, while others such as snails and slugs breathe by a
recently, however, attention has also turned to other rudimentary lung (pulmonates) and, although they can live
palaeoenvironmental aspects of molluscan studies, such as in water, are primarily terrestrial in habitat. The crystalline
shell morphology and colour-banding on the shells of form of calcium carbonate in the shells of most molluscs
certain species, stable isotopic composition of shell organic is pure aragonite, although the internal shells of certain slugs
matter and shell carbonate, and processes of amino-acid are composed of calcite. In both cases, the shells are subject
diagenesis (Ellis et al., 1996; Rousseau, 2001). to little change in either their crystalline or their chemical
Mollusca possess a number of advantages over other composition following the death of the organism, fre-
terrestrial and freshwater fossil groups that have been used quently retaining residues of the organic compounds bound
in the reconstruction of Quaternary environments. First, within the carbonate matrix, and are therefore sometimes
specimens can nearly always be identified to species level referred to as subfossil rather than fossil.
and therefore more secure palaeoenvironmental conclu- Land and freshwater Mollusca are extremely useful
sions can be drawn. Second, fossil shells are often present palaeoecological tools because of their wide distribution and
in oxidized sediments, such as slope-washes, loess or preservation in a variety of deposits. They show a marked
tufa, in which other fossil remains, such as pollen or preference for habitats that contain sufficient lime for
Coleoptera, are either absent or poorly preserved. Third, building their shells, although they are found not only in
many specimens are large enough to be identified in the chalk and limestone regions, but also in calcareous drifts,
field (Cameron, 2003; Sturm et al., 2006) and hence a in colluvial deposits, in cave earths, in loess, on coastal
good idea of the general palaeoecological context of each dunes and beaches, and in estuarine muds. Molluscs do
mollusc can generally be gained. This may help to occur in non-calcareous environments, for some species are
determine where samples should be taken for other fossil strongly calcifuge and are found only in non-calcareous
groups. Fourth, much is known about their present-day regions (e.g. Zonitoides excavatus, which lives in acid litter
ecology and geographical distributions (Dillon, 2000; of birch or alder). In such cases, however, the number of
Barker, 2001) and, since molluscs are frequently sensitive species is more restricted, shells are often thinner and less
to changes in the physical or chemical environment well preserved, and weathering and leaching in acid
(Horsák, 2005), they provide useful insights into past environments are more likely to lead to shell dissolution.
changes in both the local and regional environment. In general, the richer the base status of the locality, whether
it be land or freshwater, the richer the fauna, and molluscan
remains would be expected to be discovered in a wide range
4.6.2 The nature and distribution of
of river, marsh, lake, woodland and open-land sediments
molluscs (Miller & Tevesz, 2001; Sturm et al., 2006). Terrestrial
Molluscs are invertebrates in which the soft parts of the molluscs are also found in a variety of archaeological
body are generally enclosed within a hard shell or exo- deposits including soils, ditch, pit and well sediments,
skeleton composed of calcium carbonate bound by ploughwash and other colluvial deposits, and occupation
proteins. They have a wide range of morphological charac- horizons and building debris (Allen et al., 2009).
teristics, many of which are shared between marine and
continental species. The two principal classes as far as the
Quaternary palaeoecologist is concerned, however, are
4.6.3 Field and laboratory work
the Gastropoda (snails) or univalves, which usually possess Although molluscan remains can be collected from open
a single spiral or conical shell (although in the case of sections in the field by hand, they are best extracted under
the slugs the shell is reduced to an internal remnant), laboratory conditions because hand-picking of individual
and the Bivalvia (mussels and clams), in which the animal shells will bias the resulting samples towards the larger
possesses two hinged valves. Other classes include the species (Kidwell, 2002). Bulk samples from sections or
Polyplacophora (commonly known as chitons, sea cradles from cores are air dried and immersed in water, a small
or ‘coat-of-mail’ shells), Scaphopoda (tusk-shells) and the quantity of a dispersive agent such as H2O2 or NaOH being
Cephalopoda. The last-named are the most highly organ- added if there is organic material present. The froth, which
ized of the Mollusca and range from Nautilus, which has a includes the snails, is then decanted through a 0.5 mm
sieve. The process is repeated several times until no more exceed 200 (cf. the 3,800 or more species of Coleoptera).
snails are present in the froth and the residual slurry is then Nevertheless, identifications can be complicated by the
poured into a second 0.5 mm sieve. Both sieves are dried fact that many mollusc species vary markedly in mor-
and the residues passed through another set of sieves phology and in markings from juvenile to adult stage,
(1 mm, 710 μm, 2,411 μm) for ease of sorting. Molluscan while both colouring and fine sculpture can be affected by
remains can be either removed by hand or with the aid of local site condition. Some fossils may also be damaged
a moistened brush under a low-powered binocular by mechanical abrasion, some can have their surface
microscope or scanner. Identifications are always based markings masked by carbonate overgrowths, while others,
on type collections of modern shell material aided by the particularly the fragile specimens, may be highly com-
numerous molluscan reference works that are now available minuted. In certain cases, specialized techniques allow
(e.g. Kerney, 1999; Perez et al., 2008). In many respects, even small fragments to be identified. For example, the
identification of molluscan remains is not as difficult as marine genera Mytilus, Modiolus and Pinna have shells
in other branches of palaeontology. This is particularly which possess a characteristic crystal structure that can
true of land and freshwater molluscs, for which not be recognized with a high-powered microscope. Similarly,
only is there an extensive taxonomic literature but also, differences in shell microsculpture are often diagnostic
in Britain at least, the total number of species (both living of land Mollusca, so that specific determination of
and extinct) in late Quaternary assemblages does not fragmentary remains can be made using a light microscope,
PtStduiii* + Lymn&ea t/utrcalula
Carychiwn tridantatum
Spprmocfejj iamettaia
CattNnma calurnotto
Monacha canus&na
Oxychilus cefSamrs
& AegoprnettB sop )
Monacha Ca/itmna
Oxychilus aUianus
Lguna cyhndracea
Poten assemblage zones

PomaUas nkxjans
Vttrligo evjgtrUtOf
Pupifo muscorwm
Discus rotundstus
Vertigo subsinafg
Discus rucktrftluS
Radiocarbon dates (yr BP)
Vertigo pygmaes
L&ostvls angiica
Vertigo nJptmtw
VcKtgo gamsif
VeftigopusiJto
Total numbe* of shells
Hetrx Bspersa
ACKttlJ} fusco
HehceNo tiala
Vertigo asyen
AbXJ.l MO'ito
Swamp species
Milan spp
VaNonia
(catholic sop)
Mollusc zone
Terrestrial B'
Terrestrial A'
Tnchra
(mainly
100 153
153
d
f
Main section
100 153
153
153
3814*00 I 153
153 d
3814*00 I 100 153
153
153
153
153 Caotfus
d flvflivarts
153
153
Caotfus
cd
7550 1 SO '
153
Trench 3
6 6 3 0 ± 1 20 153
153
153 C a otfus
Coryfus
153
153 Caotfus
dh .^rl?l'l'.l.'?.J
2T3
153
9230 ± 75
153 Batata
Caotfus
9630 ± 75
53-
153
153
d Caotfus
Trench 6
153
0520 ± 90
153
153
Caoi„tfu, ••s| ,
: M
153
153 communis
10150± 110 153 - Gramineae
153 d
153
1 1 3 7 0 ± 1 50 |Cul 153
153
153
GOYB 153
153
153
Trench A Section
BS
1 2 1 5 0 * 110 | 153
153
153 Bd Caotfus
153
153
•Mr::) • 400 153 G ieae
153
0 20 40 0 20 0 0 0 0 0 d d
20 4 0 60 0 20 40 0 0 0 0 0 0 0 0 0 0 0 0 0 d 20 40 60 0 0 0d 20 40 0d 0 0 0 0
• Single shell
Figure 4.32 Variations in relative abundance of molluscan species from the Lateglacial to mid-Holocene sequence at Holywell
Coombe, southern England. Several taxa have been combined to produce Terrestrial Group A (‘catholic’ species, of wide
environmental tolerance, mainly open ground), and Terrestrial Group B (more narrow in environmental tolerance, mainly deciduous
woodland). Note how Group B taxa expand in molluscan zone b, which coincides with woodland expansion (Corylus avellana) as
reflected in the pollen record (right) (from Preece & Bridgland, 1999).
augmented by the use of scanning electron photo- 4.6.4 Taphonomy of non-marine

micrographs (Karney et al., 2011).
molluscan assemblages
The results of molluscan analyses can be presented in
a variety of ways. Early studies tended to employ species The interpretation of molluscan assemblages can be
lists, perhaps using symbols to depict the frequency of complicated by a number of taphonomic problems that
occurrence (+ = presence; * = common; 0 = abundant, etc.). should be taken into account before palaeoenvironmental
Nowadays the data are more usually depicted graphically, reconstructions are attempted. Species bias in fossil snail
either in histogram form for single samples or, where a assemblages has long been recognized, for example by
sequence of sediments is being investigated, on a diagram marked differences between the characteristic assemblages
which has the vertical axis for depth in the sequence and found in modern river flood plains compared with those
the horizontal axis for the number of species plotted. The associated with older, abandoned flood plains (river
results can be presented in terms of relative abundance terraces); these differences are likely to reflect the combined
(Figure 4.32) or absolute numbers (Figure 4.33), in both effects of selective transportation, burial and preservation
cases the histogram bars showing the depth below ground (Briggs et al., 1990). Species with fragile shells are generally
datum or a timescale, and being drawn proportional to the under-represented in subfossil assemblages, for example in
thickness of the sampled horizons. Once constructed, the shell middens, primarily due to the action of predatory
diagram can be divided into molluscan assemblage zones animals such as birds (Carter, 1990). Although abundant
(Figure 4.32) which allow further palaeoecological or and well-preserved molluscan remains are often found in
biostratigraphical inferences to be made. These zones will caves where an equable microclimate (temperature and
initially be of local significance only, but may be extended humidity) is maintained and greater protection from
(as in pollen analysis) to form a zonation scheme that has erosion and weathering is afforded, the taphonomy of cave
regional applicability, implying an orderly immigration molluscan assemblages is still not well understood. The
sequence of molluscan species within the region (Meyrick assemblages are frequently diverse, with material being
& Preece, 2001; Meyrick & Karrow, 2007). derived from the litter and vegetation that accumulate
pi eo oli
pi ailleao oli
PPu la a aoelio i
l
up e i
pi llae ei o
l
lla e i
pP iallae li
Pu la a alie
pi a a eoli l
Pp ilal li
ae oli
Pu ullap eooli
Pu ilulap oalei
Pu a a i
Pu illa oli
Pu uilplai aeoal
Pu iullpa eo
l
Pu pill aaaeo
Pu pillial oli
Pu plliall li
Pu la a li
i o
i
pPi aaae
i
pi aeo
ol
ol
Pp a
Pup e
ae
l ug
Pu lla
ll
l
lla
p
Sl
Depth
l
l
AMS C
Slug
pi
1 4
pi
and TL Age (m)
Pu
Pu
4000
(year BP)
0 400 0 JOOO JOOO 4000 4000 50 0 4000 150 i 0 e 5 S 0 2 0 2 0 20 0 4000 4000
SO 1400
o.
M- i 1
8640 + 190-1
10270 ± 3 8 0 -
*11740±3 2 8 0 - 2 -
1 7 1 3 0 ± 2 6 0-
2
3'
1 7 5 7 0 ± 1 80 M- i
"19750+ 1440 .4
21990 ± 320
25150 + 380
M- i
5 3
28360 ± 380
3 1 3 8 0 ± 3 1 30
L1.3
6
1 2 3 TL age
Figure 4.33 Variations in absolute abundance (number of individuals per 15 kg of sediment) of terrestrial molluscan species through
a Late Quaternary loess and palaeosol succession at Weinan, China. Note that species abundance and diversity are dependent
on weathering intensity (pedogenic alteration). 1: Holocene soil; 2: Loess; 3: Weakly-weathered soil (from Wu et al., 2002).
close to the cave mouth, from the action of scavenger while the transfer and subsequent spread of ‘alien’ species
animals that inhabited the cave, from washing in of by humans when colonizing new areas is a problem that
specimens either from surface soils though fissures, or particularly affects the interpretation of Holocene records
from rock faces, and from the dropping of shells into cave (Erthal et al., 2011).
fissures by birds (Hunt, 1993).
These various taphonomic problems have led, in recent
4.6.5 Interpretation of non-marine
years, to the increasing application of quantitative
taphonomic studies that test the degree of compatibility,
molluscan assemblages: habitat groups
or ‘fidelity’, between life, death and fossil molluscan and indices of species diversity
assemblages. Results have shown a relatively close corre- Terrestrial Mollusca can be classified in a number of ways
spondence between life and death assemblages in a range that allow inferences to be made about former local habitats
of sedimentary environments, for example in calcareous and environmental change. One approach is to divide
soils (Yanes et al., 2011) and in stream and lake deposits modern species into groups according to their common
(Cummins, 1994; Tietze & De Francesco, 2012), although habitat preferences. The most widely recognized sub-
in estuarine sediments, death assemblages may contain large divisions include a ‘catholic’ group (species tolerant of a
amounts of reworked material (De Francesco & Hassan, wide range of habitats), an aquatic group (species associated
2008). Much greater differences are to be found, however, with standing or running water, such as ditches, pools and
between life and death assemblages on the one hand, and streams), marsh or ‘swamp’ species, open ground species
ancient (subfossil) assemblages on the other. The principal (intolerant of shade), and shade-tolerant species, i.e. those
reasons for this appear to be post-burial taphonomic normally associated with woodland. The last group can
processes, including weathering, fragmentation of shell be further subdivided according to their preferences
material, chemical decomposition and carbonate secretions, for light, medium or heavy shade. Figures 4.32 and 4.34
a) Top
Be rger-Parker Local
Depth Ecological summary
Species richness Abundance mollusc
(cm) diversity index diagram
zones
50-
Middle
Holocene CH5
100
CH4
150
CH3
200
Early CH2
Holocene cm
b) Bottom
CH2
150
CH1
Early
Holocene
0 20 40 0 200 400 600 0 2 4 6 a o 20 40 60 80 100
Number of Individuals 1/d
species 100g- 1
Open ground Shade demanding Catholic Marsh Aquatic
Figure 4.34 Variations in species richness, abundance, diversity and habitat types in Holocene tufa sequences at Courteenhall,
near Northampton, UK (from Meyrick & Preece, 2001). For explanation see text.
(right-hand column) show how this type of classification thereby to reconstruct the history of local sea-level change
enables temporal trends in the dominant groups in a between MIS 4 and MIS 2 (Carboni et al., 2010).
molluscan record to be identified, and corresponding
local and regional environmental changes to be inferred.
4.6.6 Applications of Quaternary non-
The habitat data summarized in Figure 4.34, for example,
show a decline in open-ground taxa in biozones CH1 to
marine molluscan records
CH3 during the early Holocene, and a gradual increase in
4.6.6.1 Biostratigraphic correlation
shade-tolerant species through CH3 to CH5, culminating
in the mid-Holocene with a predominance of species It has long been recognized that molluscan assemblages
tolerant of heavy woodland cover (Meyrick & Preece, offer a basis for relative dating and correlation, because
2001). the species composition of the assemblages changes over
Non-marine molluscan populations have also been time, while certain species are confined to discrete intervals,
classified into groups on the basis of the overall diversity and hence constitute range fossils (see section 6.2.3.2). The
of the fossil assemblage, and various indices are used to development of regional molluscan ‘biozonation’ schemes,
reflect this diversity. Some of these are shown in Figure 4.34. for example, has proved particularly useful for correlat-
They include species richness (a count of the number of ing loess units and sequences, such as the Peoria Loess
different species represented in a sample) and abundance in North America, within which molluscan remains are
(the number of individuals recovered from a unit weight particularly abundant and well preserved (Rousseau, 2001).
or volume of sample: in Figure 4.34, the number per In the British Isles, some interglacial deposits of different
100 g of sample). Species diversity is also an important age can be discriminated by virtue of diagnostic molluscan
indicator, for it shows the relative abundance of each assemblages (Keen, 2001; Preece, 2001). In part, this reflects
species represented in the assemblage. Two such indices that climatic influences, for certain thermophilous taxa, such
are commonly used are the Berger–Parker index (BPI) and as the gastropod Belgrandia marginata (Michaud) and the
the Shannon diversity index (SDI). The former is a measure freshwater mussel Unio crassus (Philipsson), could only
of species dominance (d), and reflects the proportional survive in Britain during periods slightly warmer than the
abundance of the most common species in the assemblage;
present interglacial (Candy et al., 2010). A second important
the inverse of this measure (1/d) gives an indication of
control was the periodic isolation of the British Isles as a
relative diversity. Hence an increase in 1/d indicates an
result of sea-level rise: the mix of molluscan species able
increase in diversity and reduction in dominance by a few
to migrate from the continent during transitions between
species. In Figure 4.34, for example, species richness
cold glacial and warm stages depended on the time it took
generally increased throughout the Holocene, the greatest
for rising sea levels to flood the English Channel and form
abundance and diversity of species occurring in zone CH3,
an effective barrier to migration (Meijer & Preece, 1995).
during the transition from open ground to woodland. The
SDI also measures the degree to which faunal assemblages Insofar as some interglacials can be recognized by different
are dominated by a small number of species, and shows the combinations of non-marine molluscan species, this
degree of evenness or equitability of species in an therefore provides a basis for correlation (Bridgland et al.,
assemblage on a scale of 0 (dominated by one species) to 2004; Schreve & Candy, 2010). Interglacial deposits on the
1 (even representation of several species). The SDI has been European continent, however, tend to show less marked
widely used in modern ecological studies to reveal how differences in molluscan species because there were fewer
molluscan species richness and diversity vary systematically effective barriers to migration, although there is one notable
with environmental variables, for example with altitude exception. A distinctive assemblage found in the Somme
(Aubry et al., 2005) or regional rainfall patterns (Tattersfield and Seine Valleys of northern France, and that dates to the
et al., 2001). Application of this index to molluscan Middle Pleistocene, is characterized by the presence of
assemblages in loess sequences has revealed how their shells of a small extinct land snail of the genus Retinella
diversity co-varies with changes in rates of loess accumu- (Lyrodiscus) and by the remains of several species found well
lation, presumably because an increase in dust flux reduces beyond their modern ranges. This marker bio-assemblage,
the range of mollusc species able to survive (Rousseau, locally termed the ‘Lyrodiscus biome’, is thought to correlate
2001; Rossignol et al., 2004). In a sediment record from with the Hoxnian Interglacial molluscan assemblages of the
the coastal plain of Tuscany, Italy, the SDI was used to British Isles, and provides a unique basis for linking British
discriminate between fully marine, lagoonal and alluvial and western European sequences (Limondin-Lozouet &
faunal assemblages (molluscs and Foraminifera), and Antoine, 2006).
4.6.6.2 Palaeoclimatic reconstructions by diet (as opposed to climate) and in some areas reflect
variations in the proportions of C3, C4 and CAM plants
Many palaeoclimatic reconstructions from molluscan in the local vegetation. Oxygen isotope ratios, on the other
records have been based on the indicator species approach. hand, appear to be governed by climate, and hence offer a
Candy et al. (2010) employed this methodology to identify basis for palaeoclimatic reconstruction. For example,
interglacial episodes in Britain that were warmer than regional variations in the δ18O/δ16O ratio of modern
present, while Wu et al. (2002) also used variations in the land snails closely accords with regional isotopic trends in
abundance of selected molluscan indicator species (both meteoric water (Kehrwald et al., 2010), and hence variations
warm- and cold-adapted) to reconstruct the climatic history in δ18O in fossil shells provide insights into long-term
of part of the Loess Plateau, China, during the MIS 12 to changes in precipitation source and/or amount. Oxygen
MIS 10 interval. In southern Siberia, Horsák et al. (2010) isotope data from mollusc shells has been used to
identified eight ‘index species’ for extreme cold (Columella reconstruct, inter alia, shifts in the northern margin of the
columella, Pupilla alpicola, P. loessica, Vallonia tenuilabris, African summer monsoon in Sudan during the Holocene
Vertigo genesii, V. parcedentata and V. pseudosubstriata), and (Abell & Hoelzmann, 2000), an episode of increased
concluded that these indicator molluscs would indicate ‘full humidity and cooler temperatures in New Mexico, USA,
glacial’ conditions in other parts of Eurasia during the at the end of the last cold stage (Balakrishnan et al., 2005),
Pleistocene. and the history of drought phases in central Mexico
Other palaeoclimatic reconstructions have been based between 40 and 8 ka (Stevens et al., 2012). In some instances
on the statistical analysis of the range of species recorded it has proved possible to obtain palaeoclimatic data from
in molluscan assemblages. For example, a transfer function single mollusc shells. For example, high-resolution meas-
derived from a modern calibration or training dataset urement of δ18O variations along the growth axis of a land
suggests that contemporary non-marine molluscan assem- snail shell from Ethiopia revealed a cyclic periodicity that
blages reflect two main climatic influences, a temperature may reflect seasonal changes in rainfall, temperature and
gradient and a moisture gradient, and that these can be evaporation (Leng et al., 1998). If these inferences are
determined from characteristic groupings of species correct, it shows that palaeoclimatic records at annual or
(Rousseau et al., 1993). Using this approach, temperature even seasonal resolution can be generated from freshwater
and moisture variations in Burgundy, eastern France, molluscs, as has also proved to be the case with marine
during the Holocene have been reconstructed from fossil mollusc shells (see Figure 4.37).
molluscan assemblages, the data showing a two-step
warming at the beginning of the Holocene, which is similar
to that inferred from oceanic evidence (Rousseau et al., 4.6.6.3 Archaeological relevance
1994). The mutual climatic range method (section 4.5.4.2) Both land and freshwater snails have also proved valuable
has also been applied to fossil molluscan records, although in archaeological investigations, allowing inferences to
the resulting temperature estimates often tend to have be made about the environmental contexts of former
relatively large uncertainties (Moine et al., 2002). This may occupation sites (Preece et al., 2006), about human-induced
be due, in part, to the fact that climate is only one of a vegetation changes (see Figures 4.32 and 4.34) and about
number of variables that affects molluscan distribution, agricultural activities and associated erosional effects
and that the climatic influence may be difficult to isolate (Davies, 2008). In this respect, non-marine Mollusca have
from other factors, such as vegetation cover or soil chem- proved to be particularly useful in chalkland areas where
istry, which may be more important especially at the local the archaeological evidence is often abundant, but where
scale (Horsák, 2011). A further problem is that modern there is little scope for pollen analysis due to the general
molluscan distributions may not provide reliable analogues absence of peats or limnic sediments (Davies & Griffiths,
for fossil assemblages because of the degradation of their 2005). Mollusc shells are also important indicators of
natural habitats over recent millennia by human activity, foraging practices, ornamental craftwork and trading
most notably through woodland clearance, drainage operations: although most usually associated with marine
alteration and farming practices (Kiss et al., 2004, and see species (see section 4.7.4), land snails have also featured
below). in these prehistoric activities (Stiner, 1999). Over the course
In recent years, stable isotope ratios in molluscan shells of the late Quaternary, and especially during the late Holo-
have increasingly been used to infer past environmental cene, the impacts of humans on non-marine molluscan
and climatic conditions. Studies of modern shells have populations has been considerable. These include ‘alien’
shown that carbon isotope ratios are largely determined introductions, either deliberate or accidental, of species
not endemic in areas to which humans have migrated, and although those from the intertidal zone may be fragmented
which have often led to ecological destabilization and the by wave action.
favouring of synanthropic species, a process that has
extended to even the remotest of islands (Ó Foighil et al.,
4.7.2 Analysis of marine molluscan
2011). In addition, progressive alteration and loss of natural
habitats has led to a general reduction in species diversity,
assemblages
with many now in danger of extinction (Lydeard et al., 2004; Approaches to the analysis of marine molluscan assem-
Cuttelod et al., 2011). blages are essentially the same as those employed in the
study of terrestrial and freshwater forms. Marine molluscs
are affected by a number of environmental factors, such as
4.7 MARINE MOLLUSCA substrate, food supply, temperature, salinity, oxygen level,
nutrient availability, ‘depth’ (which affects, in combination,
4.7.1 Introduction
The shells of marine Mollusca have been found in a range
of deposits in coastal areas. They often occur in beach
gravels, sands and estuarine clays now lying some distance
above or below sea level, having been isostatically raised
following uplift of areas formerly glaciated or submerged
by rising sea levels, and dating of these marine fossils
provides a chronology of sea-level change and deglaci-
ation (Dyke et al., 1996; Pedoja et al., 2011; section 2.5).
Molluscan assemblages have been recovered from boreholes
both onshore and from the seabed, while shell remains,
often highly fragmented, are found at localities inland,
having been stripped from a former seabed and transported
to their present position by glacier ice (England et al.,
2000). Although perhaps less widely used in Quaternary
studies than their freshwater counterparts, marine Mollusca
are nonetheless an important additional source of palaeo-
environmental information. Indeed, the Pleistocene epoch
was originally defined on the basis of the composition of
marine molluscan faunas.
Marine molluscs occupy a great range of ecological
niches from pools and rock outcrops in the intertidal
zone to the deeper waters off the edge of the continental
shelf, although they are seldom found at depths greater
than 1 km. The mode of life of marine molluscs varies
considerably, but many taxa are infaunal, burrowing with
a muscular foot into soft sediments (e.g. Turritella, Mya,
Figure 4.35 Examples of some common marine bivalves of
Macoma, Arctica, Nucula) or boring into bedrock (Hiatella, the North Atlantic and their water depth preferences. 1. Spisula
Zirfaea). Others are epifaunal, attaching themselves by elliptica – open water, on various substrates, to 100 m depth.
threads to surfaces or other organisms (Mytilus, Modiolus) 2. Callista chione – continental shelf to 200 m depth. 3. Tapes
or by cementing one valve to the surface (Ostrea edulis). decussatus – lower shore and shallow sublittoral. 4. Mya aren-
Only a few bivalves are free-swimming (e.g. Chlamys) aria – inter-tidal mudflats. 5. Mya truncata – widely distributed
on sandy substrates to 70 m depth. 6. Arctica islandica –
while some species (mainly pectinids, or scallops) are un- widespread on muddy sand substrates from very shallow tidal
attached and are recumbent on the sea bed (Saxena, 2005). to (exceptionally) 200 m depth; noted for its longevity. 7.
The gastropods, scaphopods and bivalves are principally Aequipecten opercularis – typically found on hard surfaces in
benthic and sedentary in life and, upon death of the depths of 20–45 m in shallow subtidal areas, but can extend
to 180 m depth. 8. Macoma balthica – open sea in inter-tidal
individual, often become fossilized in situ to form autoch- zone, and to 100 m depth in the brackish Baltic Sea
thonous death assemblages. Marine Mollusca are often (photographs by David Roberts and James Scourse, School of
well preserved in Quaternary sediments (Figure 4.35), Ocean Sciences, Bangor University, UK).
food supply, habitat, shelter, light, temperature and degree 4.7.3 Marine Mollusca and palaeoclimatic
of turbulence), competition, predation and life strategy. The
inferences
main controls over their distribution and abundance,
however, are current flux and water temperature (Hiscock There are four main approaches to deriving palaeoclimatic
et al., 2004). As a result, marine fauna can be grouped into information (changes in sea-surface or water-column
distinct zoogeographical provinces, the boundaries of temperature) from marine molluscan assemblages. The
which are determined by marked temperature gradients first is based upon evidence of past migrations of the
that are, in turn, influenced by major ocean currents and principal zoogeographical provinces referred to above
gyres (Spalding et al., 2007). The latter not only regulate (Figure 4.36), and has been characteristic of attempts to
water temperature, but also affect nutrient supply as well use molluscs as palaeoclimatic indicators, particularly
as the dispersal of larval stages. Within the Northeast in the North Atlantic (e.g. Peacock, 1993). Other oceanic
Atlantic and adjacent seas, for example, modern and regions that show clearly defined province migrations during
fossil molluscan assemblages can be categorized on the basis the late Quaternary include the Pacific Shelf adjacent to
of their thermal and ecological affinities as Lusitanian North America (Roy et al., 1995), and the Southern Atlantic
(warm, southern species),5 Boreal (temperate), Subarctic off the coast of Argentina (Aguirre et al., 2011).
or Arctic on the basis of a strong south–north tempera- The second approach employs indicator species.
ture gradient and the impacts of dominant surface water Mangerud (1977), for example, used the variation in occur-
currents (Figure 4.36). These groupings provide the basis rence through a sequence of marine deposits at Ågotnes,
for palaeoenvironmental reconstruction, although there Norway, of three species (Modiolus modiolus, Littorina
are no universally accepted definitions of the provinces littorea and Mytilus edulis) whose present-day distribution
in which they occur, or of the geographical locations of and ecology are well known, to plot the palaeo-positions
the province
the province boundaries
boundaries. Furthermore, the taphonomy of of the North Atlantic Polar Front6 off the west coast of
fossil marine molluscan assemblages is generally not well Norway during the Lateglacial period. A similar approach
known, although relationships between life, death and was adopted by Ingólfsson et al. (1997) to reconstruct
fossil assemblages are increasingly being investigated by water mass variations close to Iceland during the Late-
marine biologists (Walker & Goldstein, 1999; Powell et al., glacial period, in which Macoma calcarea and Chlamys
2011). islandica were identified as key indicators of boreal–arctic
a) b)
Arctic
v2m
an
ni
ia n
itan
ta
Lus
si
Lu
Boreal
NAC-
Lusitanian v2m
Figure 4.36 Distribution of marine zoogeographical provinces in the Northeast Atlantic: a) present time; b) during the Eemian
(Last) Interglacial. NAC – North Atlantic Current; EGC – East Greenland Current (based on Funder et al., 2002 – province boundaries;
and Knudsen et al., 2001 – surface current flows).
conditions and Portlandia arctica and Buccinum groen- can be matched to develop a growth chronology (sclero-
landicum as arctic indicators. In a similar study, five chronology) that may span several centuries (Figure 4.37).
‘guide fossils’ (Mytilus edulis, Modiolus modiolus, Arctica The isotopic profile obtained from such shells provides
islandica, Littorina littorea and Zirphaea crispata) were an annually resolved palaeoclimatic record over the same
selected by Salvigsen et al. (1992) to infer variations in interval (Scourse et al., 2006). Using this approach, Schöne
the temperature of the sea close to Svalbard during the & Feibig (2009) compared seasonal and inter-annual
Holocene. Molluscan indicator species were also employed temperature patterns in the North Sea for Allerød Inter-
in a multi-disciplinary reconstruction which suggested stadial and Medieval and modern times, while Butler et al.
that summer temperatures were 4–5°C above present (2013) were able to develop a c. 1,350-year proxy marine
over Arctic regions during the last interglacial (CAPE-Last climate record for the North Icelandic shelf using growth
Interglacial Project Members, 2006). In addition, the increments in Arctica islandica.
projected northward migration of marine Mollusca in the
North Atlantic as a consequence of current and expected
4.7.4 Other applications of fossil marine
ocean warming has involved the use of key indicator species
(Lima et al., 2007).
molluscan records
The third approach involves the analysis of elemental Marine molluscs also contribute to other areas of
or isotope ratio variations within marine molluscan shells Quaternary research. For example, they provide a basis
which are considered to reflect, inter alia, changes in sea- for biostratigraphic correlation, as exemplified by the
water temperature. The most frequently employed indices work of Garilli (2011) who employed selected range
are Mg/Ca, 18O/16O, 13C/12C and Sr ratios. Examples of the fossils and a distinctive group of thermophilous molluscan
types of palaeoclimatic inference that have been based on taxa (the ‘Senegalese fauna’) to differentiate glacial from
this approach include the effects of a freshwater outburst interglacial assemblages in long Quaternary sequences
from Glacial Lake Agassiz into the Champlain Sea (eastern from the Mediterranean region. Marine molluscan assem-
Ontario) at the start of the Younger Dryas episode (Brand blages also provide palaeosalinity data, which underpin
& McCarthy, 2005), and variations in sea-surface tem- long-term sea-level histories for a number of tectonically
perature (SST) values and also in ENSO strength (section active basins including, for example, the Wanganui Basin,
7.6.4.2) in the Pacific Ocean adjacent to Peru during the New Zealand (Hughes & Kennedy, 2009), the eastern
Holocene (Carré et al., 2005). Mediterranean (Lécuyer et al., 2012) and the Black Sea, the
A final approach involves the detailed analysis of the last-named being a basin with a particularly complex
structure and/or isotopic composition of shells of certain history of marine incursion since the Last Glacial Maximum
marine molluscan species that secrete growth layers (Major et al., 2006). Fossil marine molluscs also have
on an annual or seasonal basis (Wanamaker et al., 2011). archaeological importance, as they have been a key part
As is described in section 5.4.5.2, annual growth layers of of hominin diet since the Lower Palaeolithic, and shell
specimens of long-lived species, such as Arctica islandica, middens provide evidence of the importance of marine
shellfish in prehistoric foraging and other subsistence strat-
Reconstruction Validation Calibration period egies (Colonese et al., 2011). Many sites of former human
3
occupation are now submerged beneath the sea, but can be
2 identified through characteristic midden assemblages, and
are therefore key elements in the reconstructions of long-
WNAO index
1
term migration and habitation patterns (Bailey & Flem-
0 ming, 2008). The analysis of seasonal climatic variations
1
reflected in annually layered shells (section 4.7.3) can also
have archaeological significance, for example, in the
E 1800 1850 1900 1950 2000 provision of an environmental context for the demise of the
Time (yrA.D.) Norse colonies in Iceland and Greenland (Patterson et al.,
3
2010). Finally, marine molluscan assemblages are proving
Figure 4.37 Variations in the dynamics of the winter North particularly valuable in the development of conservation
Atlantic Oscillation (dotted curve: section 7.6.4.3) and of annual strategies, notably in the identification of reserves for the
growth increments of the bivalve Arctica islandica obtained
protection of marine biota (Gladstone, 2002) and for
from the central North Sea and the Norwegian Shelf: all values
shown as deviations from long-term means (from Schöne the modelling of biotic responses to future global climate
et al., 2003). change (Belanger et al., 2012).
4.8 OSTRACOD ANALYSIS parasitism and natural associations. However, it is often

difficult to cite any one controlling factor as being univer-
Ostracoda (‘mussel shrimps’) are small, laterally com- sally dominant, for while many feel that, in the case of
pressed, bivalved crustaceans whose time range extends marine ostracods in particular, water temperature is the
over 450 million years from the Ordovician to the present most important, others would argue that salinity is more
(Rodriguez-Lazaro & Ruiz-Muñoz, 2012). During that fundamental, while in the freshwater situation the nature
period they have expanded from exclusively marine habi- of the substrate may be the overriding influence (Holmes,
tats to colonize most aquatic environments, encompassing 2002; Lord et al., 2012). Nevertheless, where autecological
a wide range of salinity and temperature conditions, studies have been able to establish, with some degree of
including ephemeral lakes and ponds. An estimated 20,000 confidence, the principal parameters that govern ostracod
species exist today, though many have never been formally distributions, those species may be of value in palaeo-
described. Some have restricted ecological preferences environmental reconstruction (Holmes & Chivas, 2002).
and are therefore useful palaeoenvironmental indicators.
Their fossil record is well documented, the first fossil
ostracod having been described as long ago as 1746. Most 4.8.2 Collection and identification
stratigraphical research has been devoted to investiga- Fossil ostracods are often collected, along with Foraminifera
tion of marine ostracods, where the majority of species and molluscs, from lacustrine and marine sediments. The
are to be found, although there is an expanding databank deposits are usually disaggregated in water (although
on the stratigraphical occurrence and ecological preferences occasionally hydrogen peroxide may be required), sieved
of brackish and freshwater species (Forester et al., 2005; and then dried. The ostracod carapaces and valves can
Martens et al., 2008). Recently, the analysis of isotopic ratios be picked out by hand, using a very fine brush. The use of
and trace element contents of ostracod carapaces has a low-powered binocular scanner of ×40 or ×60 magnifi-
provided an additional basis for inferring past environ- cation allows the majority of determinable remains to be
ments, and especially for differentiating salinity conditions collected. Individual ostracods are then mounted on a
(Holmes & Chivas, 2002; Horne et al., 2012). slide and examined under a high-powered microscope. The
carapaces possess a considerable range of morphological
4.8.1 The nature and distribution of features that aid identification, including extensive orna-
mentation of frills and spines, and internal details such as
ostracods
muscle scars, pore canals and duplicature.7 A distinction
The majority of ostracods are generally between 0.2 and is made between the valve (hard parts) and the body with
2.0 mm in adult length, but exceptionally up to 30 mm. its appendages (soft parts). Many zoologists distinguish
They consist of an outer shell or carapace, which contains several ostracod species on the basis of the number of
the soft body parts of the living organism. The carapace bristles on their appendages, whereas palaeontologists are
is usually ovate, kidney-shaped or bean-shaped (Figure forced to use other (less satisfactory) criteria. Ostracods
4.38) and consists of two chitinous or calcitic bivalve-like are usually studied under reflected light, but transmitted
valves that hinge above the dorsal region of the body. light may be necessary to see the internal features. Identifi-
The biological classification of recent ostracods rests very cations are based on modern type collections, stereoscan
largely on the characteristics of the soft parts, but as these photographs and descriptions in micropalaeontological
features are very rarely preserved in the fossil form, the manuals, and the data are expressed either as species lists
taxonomic classification of extinct species has to be based or in diagrammatic form showing the change in frequency
on the nature of the carapace, which fossilizes relatively of occurrence through time. Further details on collection,
easily. preparation, taxonomy and study can be found in Holmes
The majority of marine ostracods are bottom-dwelling & Chivas (2002), Cohen et al. (2007) and Namiotko et al.
forms, and only a small number occupy the planktonic (2011). In the case of specimens selected for elemental or
realm. Moreover, pelagic species usually possess weakly isotopic chemical analysis (section 4.8.3), more stringent
calcified shells and are therefore relatively rare in fossil preparation procedures are required in order to remove
assemblages. The distribution patterns of living ostracod minute organic or clay particles that may have adhered to
communities are governed by a wide range of factors carapace surfaces. These include soaking in ethanol or
(Mesquita-Joanes et al., 2012), including physical param- methanol, or heating in H2O2, oxygen plasma or a vacuum,
eters such as water temperature, salinity and nature of although of these, multiple methanol ultrasonic cleaning
the substrate, and biological factors such as food chains, tends to yield optimal results (Jin et al., 2006).
A B C
D E F
1 H I
1 K L
Figure 4.38 Climatically significant ostracods from the Pleistocene of the British Isles. A–H are cold-adapted taxa (A–E marine;
F–H freshwater) and I–L are warm-adapted (I, freshwater; J–L brackish/marine). A. Baffinicythere howei, flv (female left valve)
910 μm. B. Roundstonia globulifera, mlv (male) 450 μm. C. Cluthia cluthae, flv 360 μm. D. Hemicytherura clathrata, frv (right valve)
490 μm. E. Semicytherura affinis, mlv 520 μm. F. Leucocythere batesi, flv 700 μm. G. Limnocythere falcata, frv 660 μm.
H. Amplocypris tonnensis, flv 1,870 μm. I. Ilyocypris salebrosa, frv 800 μm. J. Callistocythere curryi , flv 470 μm. K. Leptocythere
cribrosa, flv 520 μm. L. Semicytherura arcachonensis, flv 560 μm. The measurements are valve lengths and arrows point in anterior
direction (SEM images provided by David Horne, Queen Mary University of London, UK).
in ostracod analysis. At the generic level, the poor state of

4.8.3 Ostracoda in Quaternary studies taxonomy often inhibits the comparison of fossil and recent
Certain rapidly evolving ostracod lineages are useful forms. Also there is evidence to suggest that several ostra-
markers in marine biostratigraphic sequences, especially cod species that are now benthic in character developed
where Foraminifera are absent (Griffiths, 2001). However, from shallow-water ancestors. Fortunately, it appears
as they lack planktonic larvae, many shallow- and warm- that migration in the opposite sense, in other words from
water species cannot cross physical barriers and are there- deep and cold to shallow and warm water, seems unlikely
fore restricted to particular geographical areas. Moreover, to have occurred. Further, there are indications that the
some of the problems already considered in the inter- dominant elements in certain benthic fossil assemblages
pretation of terrestrial fossil assemblages are also found may be due less to environmental factors than to selective
preservation of the more thick-shelled species. Where these In general, however, the abundance and distribution of
difficulties can be overcome, however, marine ostracods can most present-day ostracod species (both marine and non-
be extremely useful for reconstructing a range of palaeo- marine) co-vary with both temperature and salinity, and it
environmental conditions, including salinity, temperature, is not always clear which of these is the dominant con-
hydrodynamic balance, substrate characteristics, sea-level trolling variable. As a consequence, multivariate statistical
variations, oxygenation levels and nutrient availability methods and modern training sets tend to be used to
(Frenzel & Boomer, 2005). The most frequently inferred determine the association between the species composition
parameters, however, are palaeotemperature, palaeo- of modern ostracod communities and selected environ-
salinity and (by association) palaeobathymetry, which mental factors, the resulting transfer function informing
co-vary between ocean masses and currents. They are also interpretations of the fossil assemblage data (Viehberg &
the principal controlling factors in the distribution and Mesquita-Joanes, 2012). This approach has been used,
abundance of non-marine ostracods in freshwater and for instance, to reconstruct variations in lake level and
saline lakes (Mesquita et al., 2005). salinity in regions susceptible to seasonal or longer-term
Some palaeoclimatic reconstructions obtained from evaporation cycles (Kemp et al., 2012a). Multivariate
ostracod assemblages have used the indicator species statistical methods are also increasingly being employed
approach. For example, two modern Lusitanian (Mediter- in studies of sea-level variations based on ostracod assem-
ranean) ostracod species, Aurila arborescens and Callisto- blages. For example, Reeves et al. (2007) used cluster
cythere badia, which are found in the seas adjacent to analysis to identify characteristic faunal assemblages
northern Denmark only during the mid-Holocene, suggest associated with marine, marginal marine, tidal, estuarine
that sea-surface temperatures during the coldest month and non-marine sediment facies in the Gulf of Carpentaria,
at that time were more than 5–6°C above present and then employed the results to develop a sea-level history
values (Vork & Thomsen, 1996). Similarly, Bennike et al. for the shallow shelf that formed a land bridge between
(2010) concluded that the optimum warmth of the seas Australia and New Guinea at times of low sea level. A similar
around west Greenland during the Holocene was achieved approach, based on the taxonomic composition of ostracod
between 7 and 6.5 ka, as the thermophilous (non-Arctic) assemblages, was adopted by Boomer et al. (2010) to recon-
ostracod species Ilyocypris bradyi is confined to records struct the complex history of the Black Sea, which over the
dating from that period. On a longer timescale, glacial course of the late Quaternary has been periodically joined
and interglacial episodes in a 130 ka sediment sequence to, and isolated from, the Mediterranean Sea.
in south-central Illinois could be identified using the An alternative and increasingly important palaeo-
ostracod species Limnocythere friabilis as a cold indicator environmental approach is the analysis of variations in
and Candona caudata and Heterocypris punctata as warm the chemical composition of fossil ostracod carapaces.
indicators (Curry & Baker, 2000). Although this approach Recent studies have shown that the ratios of chemical
has been less extensively used with non-marine ostracods, elements (e.g. Mg, Sr, Ba, measured proportionately to the
because fewer species appear to be controlled exclusively abundance of Ca) and of isotopes of oxygen, carbon and
by temperature, some recent research suggests that a strontium in modern ostracod carapaces vary predictably
number of non-marine ostracod species do show common with environmental gradients (Holmes & Chivas, 2002;
distributional patterns with respect to modern climatic Horne et al., 2012). The chemical composition of ostracod
variables, and where this is the case their mutual tem- carapaces has been used to reconstruct the history and
perature range can be calculated enabling an estimation to pattern of lake-level variations during the Late Quaternary
be made of former temperatures (Horne et al., 2012). This in North America (Bright et al., 2006), to determine the
approach constitutes a variant of the MCR technique strength of the East Asian monsoon system (Zhai et al.,
developed for the analysis of fossil beetle assemblages 2011), to identify sources of water delivered to lake systems
(section 4.5.4.2). in southern Spain (Anadón & Gabàs, 2009), and to establish
Some ostracod species, such as Cyprideis torosa, rates of dissolved load influx and weathering in the
appear to be particularly sensitive to salinity variations catchment of Lake Constance in central Europe (Kober
and have therefore been used as salinity indicator species et al., 2007)
(Pint et al., 2012). Other ostracods are able to survive in the Fossil ostracod assemblages have also been employed
inter-tidal zone, where salinity, light, degree of oxygenation in other areas of Quaternary research, for example in the
and other properties vary markedly over a diurnal cycle, and study of high-energy tsunami deposits (Ruiz et al., 2010),
these constitute useful indicator taxa for the reconstruc- and in the reconstruction of pollution histories in both
tion of past variations in sea level (Yasuhara et al., 2003). marine and freshwater environments (Yasuhara et al.,
2003). They can also provide useful archaeological inform- by various cements) components. The tests may be single
ation on the nature of sites occupied or exploited by chambered, but more frequently consist of a number of
humans (Holmes et al., 2010; Lord et al., 2011). Finally, chambers separated by septae. Connections between the
they are an important component of current micro- chambers, through which cytoplasmic material can move,
palaeontological work that is providing new insights into are formed by small holes in the septae known as foramina,
deep-sea biodiversity and ecosystems, and the degree to from which the group derives its name. In many common
which these vary over glacial–interglacial cycles (e.g. species, the chambers are added in a spiral pattern,
Yasuhara et al., 2012). producing a coiled shell, while others develop far more
complicated structures (Figure 4.39). They are classified on
the basis of a number of characteristics, such as the
rhizopodia (cytoplasmic extensions used in locomotion and
4.9 FORAMINIFERAL ANALYSIS feeding), degree and form of coiling, number of chambers,
Foraminifera are protists (or prokaryotes – single-celled number and pattern of apertures, and surface ornamen-
organisms8) that possess a hard calcareous shell often dis- tation. Not all of these, however, are preserved in fossil
tinctively coiled to resemble that of a gastropod or cephal- forms.
opod. They were first described and illustrated in the Foraminifera range in size from less than 0.40 mm
sixteenth century, but were not studied systematically (the planktonic forms) to some of the benthic species
until the latter part of the nineteenth century following which may measure up to 10 cm in width (so-called larger
the remarkable voyage of HMS Challenger which began Foraminifera). They are tolerant of a range of salinity and
in 1872. The discovery during that expedition of living temperature, being found in saltmarshes, shallow brackish
Foraminifera in deep-sea waters, and of fossil remains in water in estuaries, on the continental shelf and in the
sediments that were dredged from the sea floor, revolu- waters of the deep oceans of the world. Most Foraminifera
tionized marine micropalaeontology. Since then, Foramini- are marine and benthic, although a few genera are pelagic,
fera have become staple tools in Quaternary stratigraphy, while a very small number of species (thecamoebids) are
palaeoceanography and palaeoclimatic reconstruction. adapted to freshwater environments. The marine plank-
tonic and benthic forms have proved to be particularly
useful in global correlation and climatic reconstruction and
4.9.1 The nature and distribution of
these will be considered in the next section of this chapter.
Foraminifera The present discussion will be concerned principally with
Foraminifera consist of a soft body (protoplasm) enclosed foraminiferal remains in shelf seas and inshore waters.
within a shell or test secreted by the organism, which is Further details on the nature, distribution and ecology of
variously composed of organic matter, minerals (calcite or Foraminifera can be found in Haslett (2002), Sen Gupta
aragonite) or agglutinated (foreign particles held together (2002) and Murray (2006).
1 2 3 4 5
Figure 4.39 Examples of planktonic foraminiferal species widely employed in Quaternary palaeoceanographical studies. The use
of sinistral (1) and dextral (2) forms of Neogloboquadrina pachyderma as palaeoclimatic proxies in high-latitude surface waters is
referred to in section 4.10.7. Globigerinoides ruber (3) and G. sacculifer (4) are abundant today in subtropical waters (Figure 4.43)
and hence employed as proxy indicators of warm SSTs. Globigerina bulloides (5) can tolerate a wide range of temperatures but
is most abundant in cool upwelling ocean waters. (SEM images provided by Alessandra Asioli, CNR Institute of Geoscience and
Geo-resources, Padua, Italy.)
4.9.2 Collection and identification the water column, and (to a lesser degree) water tem-
perature. But because these conditions can change quickly
Foraminifera can be extracted from sediments obtained in shallow marine environments, most benthic forams
from surface samples or from cores. The matrix is usually possess wide environmental tolerances (Van der Zwaan
disaggregated using either water or Calgon solution, but et al., 1999), and are therefore less valuable as palaeo-
boiling in a dilute solution of hydrogen peroxide may environmental indicators than planktonic species (see
prove necessary for strongly cemented sediment matrices. section 4.10). Palaeoenvironmental inferences are also
The samples are then washed through sieves and the constrained by taphonomic factors such as reworking and
residues dried; the retained Foraminifera can be picked out redeposition, the resulting mixed assemblages being
by hand with the aid of a binocular microscope. Where large indicated by wear, poor preservation and unusual popula-
numbers of sand grains are present (e.g. in some shelf tion structure. Further problems arise from the pyritization
sediments), the foraminiferal remains can be concentrated or permineralization of some tests, which makes them
using a heavy liquid such as ethylene bromide/absolute heavier leading to segregation during transport (Berkeley
alcohol solutions, the tests being ‘floated’ from the sand et al., 2007), and differential dissolution, which particularly
using detergent (Lehmann & Röttger, 1997). Removal of affects calcareous tests (Murray & Alve, 1999), although
surface contaminants, which is essential where elemental agglutinated forams are more susceptible to dissolution
or isotope ratio measurements (e.g. Mg/Ca) are to be than either calcareous or siliceous forms.
obtained from foraminiferal tests, is usually achieved by These problems notwithstanding, benthic foraminiferal
dilute acid leaching, although care needs to be exercised in studies have provided valuable data on inland and shelf-
the application of this procedure as it may lead to partial sea sequences, some examples of which are summarized
dissolution of sample carbonate, which can affect the below.
isotopic ratios (Barker et al., 2003).
The smaller Foraminifera are examined under a high-
powered microscope using reflected light. Occasionally 4.9.3.1 Sea-level change
staining (e.g. with malachite green or a similar food dye) A number of studies have shown a strong statistical
is required to bring out the surface structures more clearly. relationship between the altitudinal distribution of modern
Larger Foraminifera are often studied in thin section where benthic foraminiferal species and tide level, while cali-
wall and growth plan may be better seen under transmitted brations between benthic foram assemblages and tide-
light. As with ostracods, identifications are based on type gauge records indicate that some species track changes in
collections, descriptions and stereoscan photographs in tidal reach (Edwards, 2007). Transfer functions that
foraminiferal manuals, which can be explored through quantify these relationships have provided a basis for
online resources, such as the World Modern Foraminifera the reconstruction of Holocene changes in sea level on,
Database (http://www.marinespecies.org/foraminifera/), for example, the Great Barrier Reef complex, Australia
the Marine Species Identification Portal (http://species- (Horton et al., 2007), mangrove swamps on various
identification.org) or eForams (http://www.eforams.org/). Pacific shores (Woodroffe et al., 2005), saltmarshes in
The data can be presented simply as species lists or, more North America (Kemp et al., 2009) and inter-tidal mud-
commonly, as ‘range charts’ expressed in percentage form flats and inter-dune slacks in the British Isles (Edwards &
or as abundance per unit volume of sediment (e.g. Lloyd Horton, 2006). There is a limit to the precision with which
et al., 2012). Further information on the collection and mean tidal altitude can be inferred using this approach,
study of Foraminifera can be found in Green (2001) and however, due partly to taxonomic issues, but also to the
Schönfeld (2012). fact that most inter-tidal species seem able to tolerate a
range of salinity and other conditions, so that small-scale
changes in sea level may be difficult to detect (Edwards,
4.9.3 Foraminifera in Quaternary inshore
2007; Woodroffe, 2009). Where significant changes in
and shelf sediments sea level are sustained over longer periods, however, the
Foraminiferal remains in sediments of most inshore corresponding changes in benthic foraminiferal assem-
waters and the shelf seas are dominated by benthic forms, blages are more evident. In the Western Mediterranean, for
in contrast with the deep-sea sediments in which planktonic example, benthic foram records from shallow-water shelf
Foraminifera are more abundant. These bottom-water sequences reflect a rise in sea level over the last 12 ka that
assemblages are influenced by a number of factors, which matches reconstructions based on independent lines of
include light, oxygenation, organic and nutrient flux within evidence (Figure 4.40).
0 ERSL W A . P L S {this study)
ERSL core 342-1

ERSL core 367-1
2000
ERSL core 401-1
4000
Global ERSL
Age cal BP
Tahiti corals
6000
Global SL change model
ERLS Mediterr. Sea

8000 Tyrrhenian Sea
French coast
Israeli coast
10000 Tyrrhenian coast
12000-
-70 60 -50 -40 -30 -20 -10 0 10 20

Relative sea level (m)
Figure 4.40 Relative sea-level (ERSL) curves for the last 12 ka for the Western Mediterranean based on benthic Foraminifera-
based transfer functions. Data from three shelves (core 342-1: Alboran Platform; core 367-1: Oran Bight; and core 401-1: Mallorcan
Shelf) show broadly similar trends to independent reconstructions of sea-level change over the same period (from Milker et al.,
2011).
influences of the West Greenland and East Greenland

4.9.3.2 Shallow marine water mass and Currents during the Roman Warm Period, the Medieval
temperature variations Climate Anomaly and the Little Ice Age. Eberwein &
Most palaeotemperature reconstructions based on fora- Mackensen (2008) also employed benthic foraminiferal
miniferal evidence are derived from planktonic or deep indicator species as one of a number of tools to recon-
marine benthic species, and particularly on the chemical struct changes in palaeoproductivity and water mass in an
composition of their tests (see section 4.10). Nevertheless, upwelling region of the Atlantic off the coast of northwest
shallow-water benthic species can also provide important Africa. In the eastern Pacific, benthic foraminiferal assem-
palaeoclimatic insights, because shelf and estuarine assemblages from the Santa Barbara Basin, California, show
blages are sensitive to changes in water mass, the positions distinctive faunal ‘switches’ during the Late Quaternary,
of dominant surface currents, or migrations of important which Cannariato et al. (1999) attributed to the influence
sea-surface boundaries such as the North Atlantic Polar of Dansgaard–Oeschger climatic oscillations (section
Front, and indicator species can be used to reconstruct these 3.11.4). Other studies of shallow benthic foraminiferal
past changes in marine conditions. For example, Perner assemblages have employed a transfer function approach,
et al. (2011) identified the benthic foram species Cassidulina for example in the reconstruction of water mass tem-
reniforme as an indicator of an Atlantic water mass, and perature and salinity variations over the North Icelandic
Elphidium excavatum f. clavata, Cuneata arctica and Spiro- Shelf during the Holocene (Knudsen et al., 2012). Finally,
plectammina biformis as indicators of an Arctic water modern benthic foraminiferal assemblages have been used
mass, in their reconstruction of centennial-scale climate to establish the rate and magnitude of the warming of
variations off West Greenland over the past 3.6 ka. They shallow seas in response to the recent increase in global
attributed changes in water mass to variations in the temperatures (Saher et al., 2012).
4.9.3.3 Other palaeoenvironmental and microfloral assemblages themselves can also provide
applications valuable palaeoenvironmental data, since they commonly
retain a record, at least in part, of former ocean-water
Apart from the palaeoclimatic and sea-level applications temperatures that will, in turn, be a reflection of former
described above, the wider palaeoenvironmental poten- climatic conditions. From an analysis of the biostrati-
tial of shallow marine benthic Foraminifera has been graphy of the deep-sea sediments, therefore, it is possible
relatively under-exploited until quite recently (Schönfeld, to make reasoned inferences about climatic regimes
2012). For example, some modern benthic species have that prevailed over large areas of the world’s oceans at dif-
been found to be diagnostic of glacimarine margins; ferent times during the Quaternary period. In addition,
hence equivalent fossil assemblages can be used to locate marine microfossil records have been used to reconstruct
the positions of former ice margins and grounding lines, changes in oceanic circulation (palaeoceanography) and
and these have contributed to the mapping of retreat stages in nutrient supply (palaeoproductivity), as well as in
of the Laurentide Ice Sheet (Jennings et al., 1996), the the dissolved oxygen content of ocean waters and
Antarctic ice sheet (Pudsey et al., 2006) and fjord glaciers bathymetry.
in Spitsbergen (Korsun & Hald, 1998). Other assemblages Some of the marine organisms found in the deposits of
appear to be indicative of seasonal or permanent sea ice, the deep-ocean floors that are employed in palaeoclimatic
and provide evidence of sea-ice cover in the past (Scott research have been discussed above in relation to fresh- or
et al., 2008). Many benthic species are now known to be brackish-water situations, or to relatively shallow shelf
sensitive to changes in water chemistry, and are proving to seas. These include foraminifers, diatoms and ostracods.
be valuable monitors of past and current pollution levels Of the remaining organisms, the most valuable in terms
and water quality (Bouchet et al., 2012), while others of their application to palaeoclimatic research have proved
provide evidence of submarine methane seepage (Panieri to be Radiolaria, coccolithophores and dinoflagellates
et al., 2012). Shallow marine benthic forams also have (Armstrong & Brasier, 2005; Hillaire-Marcel & De Vernal,
important archaeological applications, for example for 2007).
provenancing of pottery (Santacreu & Vicens, 2012),
discriminating between natural and cultural shell deposits
(Rosendahl et al., 2007) and deciphering the detailed coastal 4.10.2 Radiolaria
context of ancient settlements now isolated from the sea Radiolaria are marine, amoebic protozoans that secrete
(Bernasconi et al., 2010). elaborate skeletons composed largely of amorphous
Shallow water benthonic Foraminifera, therefore, can (opaline) silica, which is extracted from seawater in the
be used as indicators of both local and regional environ- same way that Foraminifera extract calcium carbonate,
mental conditions although, as with ostracod assemblages a group with which they have close affinities in terms
described earlier, they are best employed in conjunction of ecological and environmental requirements (Haslett,
with other fossil remains, particularly molluscs, ostracods, 2004). The skeleton, which consists of a complex net-
diatoms and pollen. But it is the planktonic and deeper- work of elements, is contained within the living proto-
water benthic Foraminifera that have made the most plasm and thus the hard parts forming the fossil do
significant contributions to Quaternary environmental not dissolve in seawater until the creature dies. The single-
reconstruction, as explained in the following section. celled radiolarians are commonly circular in shape and
average between 100 and 2,000 μm in diameter, though
their skeletal tests vary markedly in form, measuring
4.10 MICROPALAEONTOLOGY OF c. 30–300 μm in length (Figure 4.41a). There are about
400–500 relatively common species and these are found
DEEP-SEA SEDIMENTS in all ocean waters from the tropics to the subpolar
seas, in surface waters down to depths of over 4 km.
4.10.1 Introduction Most of the taxa have specific, sometimes narrow,
It has already been shown (section 3.10) how the ratios bathymetric preferences, though some can occur in a very
of oxygen isotopes in marine microfossils can provide a wide range of water depths (eurybathyal) (Baumgartner
record of the expansion and contraction of ice sheets et al., 2006).
during the Quaternary, and the applications of this The composition of modern radiolarian communities
technique in Quaternary stratigraphy will be considered appears to be influenced by water depth, water masses,
further in Chapter 6. However, the marine microfaunal hydrographic boundaries (e.g. surface frontal systems) and
Figure 4.41 a) Examples of siliceous skeletons (tests) of the radiolarian groups Spumellaria (S) and Nassellaria (N), important
biomarkers commonly preserved in Late Quaternary deep marine sediment; their tests range in size from c. 30–300 μm. 1. Callimitra
carolotae (N). 2. Euchitonia elegans (N). 3. Lamprocyclas maritalis (S). 4. Mitrocalpis araneafera (N). 5. Nephrospyris knutheieri (N).
6. Rhizoplegma boreale (S), recently renamed Cleveiplegma (from Dumitrica, 2013; SEM images provided by Kjell Rasmus Bjørklund,
Natural History Museum, University of Oslo, Norway). b) SEM scans of coccolithophore specimens from the central Adriatic Sea.
1. Coccosphere of Emiliania huxleyi TYPE A comprising an outer (single) layer of calcitic platelets (coccoliths) which enclose the
living organism. This form of the species is typical of nutrient-rich environments and is characterized by rapid growth during bloom
conditions (scale bar: 1 μm). 2. Emiliania huxleyi TYPE A with multiple layers of coccoliths, a larger form that is typical of nutrient-
poor environments and which grows more slowly (scale bar 1 μm). 3. Complete coccosphere of Calcidiscus leptoporus (ssp.
quadriperforatus), an important carbon storage species that thrives in tropical and temperate latitudes (scale bar: 10 μm). 4. Close-
up of 3 showing coccolith detail (scale bar: 1 μm) (images provided by Luka Supraha, Uppsala University, Sweden).
nutrient availability (Abelmann & Gowing, 1997), and 4.10.3 Coccolithophores

hence past variations in these parameters can be inferred
Coccolithophores are the most common members of a
from fossil assemblages. Examples include the reconstruc- group of unicellular autotrophic marine algae known as
tion of variations in the strength of the East Asian monsoon calcareous nannoplankton. They are generally spherical
over the last glacial cycle (Ikehara & Itaki, 2007), and of the or oval in shape, and are mostly less than 100 μm in
upwelling currents in the Atlantic close to northwest Africa diameter (Figure 4.41b). The living organism is covered by
over the past 40 ka (Haslett & Smart, 2006). Radiolarian a layer of organic scales upon which small calcite platelets,
stratigraphy has also been used to estimate past variations commonly ranging in size from 1–25 μm and called
in SSTs, for example in the East China Sea during the past coccoliths, are secreted (Flores & Sierro, 2007). These may
10.5 ka (Chang et al., 2008a), to reconstruct stratifica- envelop the cell completely to form a hollow sphere
tion and current patterns in the southwest Pacific over the (coccosphere) which eventually disintegrates and falls to
past 600 ka (Lüer et al., 2008), and to assess oceanographic the ocean bed. The individual button-shaped coccoliths
responses to orbital glacial–interglacial cycles (Itaki et al., are usually all that remain of the former living creature.
2007). In addition, Quaternary radiolarian records often Like other marine flora, coccolithophores are autotrophic,
contain distinct biostratigraphic marker horizons (e.g. possessing chloroplasts that are used to photosynthesize
range boundaries of certain taxa), and these provide a food. They possess whip-like threads (flagella) to generate
basis for correlation between palaeoceanographical records motion, and a few are known to ingest bacteria and small
(Haslett, 2004). algae. They are therefore difficult to classify, possessing
characteristics of both plants and animals. Although a few propelled by the spiralling motion of two flagella. The cell
species are adapted to either fresh or brackish water, the wall of many dinoflagellates is protected by armoured
majority of present-day coccolithophores are marine, plates, called theca, the distinctive structure and arrange-
but are much less common at salinities greater than 37 per ment of which provides a basis for classification (Figure
mil or less than 32 per mil. Being photosynthetic, they 4.42a). Their life cycles are complex and frequently involve
are mostly confined to the photic zone, predominantly a non-motile stage during which a cyst, composed of
within the upper 80 m of the water column, and are rarely sporopollenin, is formed; it is this resistant dinocyst that
encountered below 200 m depth (Baumann et al., 2005). is commonly preserved in fossil form (Figure 4.42b).
Coccolithophores are found in very large numbers in Dinoflagellates are one of the most abundant types of
present-day ocean surface waters where they rival diatoms marine plankton, occasionally generating ‘algal blooms’ that
as the most abundant phytoplankton (Flores et al., 2012). colour the sea surface red or brown (so-called ‘red tides’),
Their abundance and species distributions are governed by caused by the pigments that their living cells contain.
a combination of light, salinity and temperature, but it was Further details can be found in Haq & Boersma (1998) and
their temperature requirements that initially attracted Dale & Dale (2002b).
ocean scientists, the down-core variations of key coccolith Comparatively little was known about the potential
indicator species forming the basis of some of the early of dinoflagellates for Quaternary stratigraphy and palaeo-
research on Quaternary palaeoceanography (McIntyre & environmental reconstruction until the early 1990s (Harland,
Ruddiman, 1972; McIntyre et al., 1972). This approach 1988). Over the last twenty years, however, dinoflagellate
continues to be used today, providing evidence, for records have been obtained from many of the world’s
example, of shifts in the position of the Inter-Tropical oceans and a range of palaeoenvironmental parameters
Convergence Zone (ITCZ) during the late Quaternary have been inferred, including sea-surface temperatures
(Mertens et al., 2009), the effects of Dansgaard–Oeschger (Esper & Zonneveld, 2007), palaeosalinity (Mudie et al.,
cycles (section 3.11.4) and Heinrich Events (section 3.10.1) 2001), ocean palaeoproductivity (Höll et al., 1999), ocean
in the Atlantic and Mediterranean basins over the past stratification (Marret & Scourse, 2002) and the limits
50 ka (Colmenero-Hidalgo et al., 2004), and precession of sea-ice extent (Harland et al., 1999). Past variations
orbital forcing effects on the Arabian Sea over the last in the strength and location of major oceanographic
200 ka (Rogalla & Andruleit, 2005). Recent research has currents have also been reconstructed, for example periodic
also focused on the link between climate change and incursions of Atlantic waters into the Eurasian Arctic
coccolith productivity, particularly over glacial–interglacial (Matthiessen et al., 2001) and latitudinal migrations of
cycles, with marked increases in coccolith abundance the Antarctic Circumpolar Current (Verleye & Louwye,
apparent during glacial periods and peak abundances 2010). A number of statistical methods have been employed
reached during Terminations (section 6.2.3.5) (Flores et al., in these studies, again including transfer functions. The
2012). Coccolith productivity also appears to have latter approach, however, tends to generate SST estimates
fluctuated in concert with the precession orbital cycle (e.g. that are higher than those derived from other proxies,
Ivanova et al., 2012). Over shorter timescales, however, such as forams, diatoms and radiolarians, and this has been
the factors affecting coccolith productivity are more attributed either to selective degradation of the more
complex, and appear to reflect the positions and strengths delicate dinocyst types (Esper & Zonneveld, 2007) or to
of dominant marine currents (Giraudeau et al., 2010). It inherent statistical bias (Telford, 2006). Nevertheless,
now seems that the body size and morphology of some dinoflagellate records not only add to the Quaternary
coccoliths, such as Emiliania huxleyi (Figure 4.41b), may scientist’s proxy toolkit, but also have applications in
be controlled by temperature, for an increase in size appears non-marine contexts. For example, seasonal dinoflagellate
to correspond with successive colder episodes; if so, this blooms in large freshwater lakes have formed annually
would be a useful additional palaeoclimatic indicator laminated sediment, allowing past environmental changes
(Flores et al., 2010). to be reconstructed at a high temporal resolution (Chu
et al., 2008).
4.10.4 Dinoflagellates (dinocysts)
4.10.5 Marine microfossils in ocean
Dinoflagellates are microscopic, usually unicellular, protists
measuring up to 2 mm in diameter, though the majority
sediments
are considerably smaller. They are members of the algae Planktonic Foraminifera are the major contributors to
(division Dinoflagellata), and are free-swimming, being deep-sea sediments and, along with coccoliths, account
a)
1 2 2
e
5 6 7 E
9 10 11 12
b)
apex or anterior
apical
horn plates
flagellar sutures
epitheca - pore(s)
cingulum-
hypotheca - ^flagella
-ântapical
sulcus' 1
horns
dorsal
antapex or view
posterior
ventral view dorsal view
Figure 4.42 a) Examples of common Quaternary dinocysts. 1. Protoperidinium stellatum (Black Sea Holocene sediments). 2.
Echinidinium transparantum (Black Sea Holocene sediments). 3. Peridinium ponticum (Black Sea Holocene sediments). 4. Cyst
of Polykrikos schwartzii (Black Sea Holocene sediments). 5. Operculodinium centrocarpum (North Icelandic Shelf Recent
sediments). 6. Bitectatodinium tepikiense (Celtic Sea Recent sediments). 7. Tuberculodinium vancampoae (Gulf of Guinea Recent
sediments). 8. Quinquecuspis concreta (Celtic Sea Recent sediments). 9. Spiniferites cruciformis (Black Sea Holocene sediments).
10. Spiniferites ramosus (Celtic Sea Recent sediments). 11. Lingulodinium machaerophorum (Black Sea Holocene sediments).
12. Trinovantedinium applanatum (Celtic Sea Recent sediments). Scale bar represents 10 μm (images provided by Fabienne Marret,
University of Liverpool, UK). b) Structural features of some dinoflagellate cysts (dinocysts) at motile stage.
for more than 80 per cent of modern carbonate deposition estimated that perhaps as few as 10 per cent of all Radiolaria
in seas and oceans (Hüneke & Mulder, 2011). Most of the find their way into the fossil record. Similar low values have
tests now being deposited are from planktonic species of been suggested for diatoms. Both Radiolaria and diatoms
Globigerina and it has been estimated that about 30 per cent are prone to exhumation and reburial in younger sediments
of the present ocean floor (over 60 million km2) is covered and this poses further problems of interpretation for the
by the grey mud known as Globigerina ooze. These oozes, marine biostratigrapher.
forming at depths up to 5 km in ocean waters between 50°N
and 50°S, are fed by a continuous rain-out from the water
4.10.6 Laboratory separation of marine
column of mixed organic detritus, termed ‘marine snow’.
Coccolith oozes form principally in the tropical and
microfossils
subtropical regions, where the remains may average up to Faunal and floral remains are extracted from deep-ocean
30 per cent by weight of the sediments. In arctic regions, cores in the laboratory by disaggregation of the sediment
by comparison, the values may be as low as 1 per cent. By using a water jet to flush sediment through a series of sieves
contrast with Foraminifera, however, coccolith remains of varying mesh size. This procedure alone may be sufficient
settle much more slowly, and are therefore more susceptible to recover most specimens from soft, unconsolidated
to carbonate dissolution. Although some coccoliths may sediment, but for more resistant materials, immersion in
settle out more rapidly if they are contained within the dilute hydrogen peroxide (Foraminifera, Radiolaria), nitric
faecal pellets of planktonic grazers, it has been estimated acid or hydrochloric acid (Radiolaria), or, in the case of
that less than 25 per cent of all coccolith species are actually coccoliths, sodium hexametaphosphate (Calgon), may be
preserved in the fossils of ocean sediments, though precise required. The larger fossils can be hand-picked from the
census studies of coccolithophore production and survival meshes of sieves, while for the smaller remains, particu-
are rare (Robert, 2009). Below 3–5 km, the calcium car- larly coccoliths and dinoflagellate cysts, it is necessary to
bonate compensation depth, nearly all CaCO3 enters into concentrate the microfossils into a liquid which can then
solution, and thus only the most resistant calcareous fossils be mounted on a microscope slide (‘smear slides’). Diatoms
will be found. The sediments there will be dominated by and some other small siliceous microfossils (less than
siliceous remains, predominantly of Radiolaria. 100 μm) require a more lengthy laboratory preparation
Radiolaria accumulate in abundance in equatorial procedure involving several chemical immersion steps and
sediments where productivity is high in the water column centrifugation. High-powered microscopy (up to ×1,600)
above. However, as the productivity of calcareous organ- may be necessary for ultra-detailed study using trans-
isms is also high, the radiolarian remains are often masked mitted, reflected and polarized light and, as in other micro-
by foraminiferal and coccolith fragments. Only in areas palaeontological work, increasing use is being made of the
such as the tropical northern Pacific and Indian Ocean, electron microscope. Identifications may be made difficult
where large areas of the sea floor lie below the carbonate by the solution of diagnostic parts, by the mechanical wear
compensation depth, are radiolarian oozes encountered. of the skeletal remains, and by the tendency, especially in
Diatomaceous oozes are also found in the abyssal the case of carbonate fossils, for calcite overgrowth and
depths of the Pacific and Indian Oceans and in parts of the recrystallization to obscure the morphology of the surface
Atlantic (Sverdrup & Kudela, 2011). Some of these oozes features. Further detail on extraction procedures can be
are laminated and appear to have formed along oceanic found in De Vernal et al. (2010) and on identification in
temperature ‘fronts’ where upwelling leads to large con- Armstrong & Brasier (2005).
centrations of diatoms near the surface, and periodic die-
off promotes sudden accumulation (diatom mats) on the
sea floor (Shimada et al., 2008). Siliceous oozes are most
4.10.7 Marine palaeoclimatology
common in the high-latitude areas of the north Pacific and The distributions of Foraminifera, Radiolaria, coccolitho-
around Antarctica, locally containing a high abundance phores, dinoflagellates and marine diatoms are partly
of siliceous dinoflagellate cysts (e.g. Marret & Zonneveld, determined by nutrient requirements. The planktonic
2003). In these regions, calcareous fossils are rare and both forms are all found in abundance in zones of upwelling,
radiolarian and diatom remains are abundant. As with for example, or pronounced vertical mixing, where
carbonates, however, silica is soluble in seawater, dissolu- mineral nutrients are readily available. For this reason,
tion being especially rapid in the upper levels of the water large numbers of these micro-organisms are frequently
column. Only those radiolarian species with a solid opaline encountered just seaward of the continental slope. In most
skeleton reach the sea floor and, overall, it has been cases, however, the fundamental determinant is water
Mean annual sea-surface temperature °C
21
24
18
6
27
3
15
12
0
60
0
Sphaeoidinelia dehiscens
;
30"
-60°
-30 •
Globigerinoides ruber (pink)
120°
Pulleniatina obliquiloculala
Globorotalia menardii + tumida
0
Globigerinoides sacculifer
TROPICAL
Globigerinoides '.MODUS
25
150°
Globorolalia crassaformis
50
Globigerineiia siphonifera
180°
Globigerinoides ruber (white)
75
Globoturbotalita lenella
100
Average abundance %
Globigerinoides conglobatus
•150°
Neogloboquadrina dutertrei
2007, reprinted with permission from Elsevier).

Globigerineiia calida
Beela digitata
120"
Globoturbotalita albescens
SUBTROPICAL
Globorotalia truncatulinoides
Globigerina falconensis
Orbulina universa
•90°
Globorotalia hirsute
Globorotalia infiata
-60"
Globorotalia scitula
Globigerinila glutinata
-30°
Globigerina bulloides
TRANSITIONAL
0
Neogloboquadrina incompta
30
Turborotalita quinqueloba
SUBPOLAR
60°
90'
Neogloboquadrina pachyderms
POLA R
BIOLOGICAL EVIDENCE
9
e
2
0
21
27
15
12
18
24
249
temperature gradients and species abundances. The species abundance plots (top) are averaged at 1°C intervals (from Kucera,
Figure 4.43 Planktonic foraminiferal provinces in the modern ocean showing the close relationship between sea-surface
temperature, and detailed ecological studies have shown is a left-coiling species, whereas right-coiling species, e.g.
that many species are associated with water masses that N. incompta, favour warmer waters (Figure 4.44). Both the
possess distinctive thermal characteristics. From the present abundance of N. pachyderma and the ratio of left- to right-
distribution of marine plankton, and allowing for current coiling Neogloboquadrina species are still widely employed
circulation and Coriolis effects, it is possible to recognize in palaeoceanography as key palaeotemperature indicators
distinct equatorial, tropical, subarctic and arctic provinces (see Kucera, 2007).
(Figure 4.43). Hence, the analysis of marine microfossil Although this work provided valuable insights into
assemblages can provide a unique source of information on Quaternary climatic changes, the foraminiferal assemblage
ocean palaeotemperatures and, by implication, on former evidence was not always easy to interpret. It has already been
climatic conditions. shown that only a small proportion of the planktonic
The initial approach to Quaternary temperature ocean fauna and flora actually reaches the sea floor to
investigations using data from deep-ocean cores was based enter the fossil record due largely to dissolution, a process
simply on the presence or absence of certain key species that favours the preservation of cold-adapted species, since
in fossil assemblages. Early work demonstrated that the they tend to have more robust skeletal parts. The death
abundance of the planktonic foraminiferal species Globo- assemblage in a body of ocean sediment, therefore, rarely
rotalia menardii could be used to infer climatic change, reflects accurately the former living assemblage in the water
an idea developed by David Ericson and his colleagues column above. Indicator species associated with these
(e.g. Ericson & Wollin, 1968) to construct a series of biased death assemblages are therefore not always a reliable
palaeotemperature curves based on the abundance of index of palaeotemperature change. Moreover, although
G. menardii in sediments from the floors of the Caribbean temperatures are generally believed to be the major
and subtropical Atlantic. High percentages of G. menardii determinant in the distribution of planktonic fauna and
were interpreted as indicating warmer, possibly inter- flora, other factors need to be considered such as salinity
glacial periods, while reduced frequencies reflected cold, variations, seasonal temperature fluctuations and food
glacial periods. In a series of influential papers during supply, and the prime factor may not be the same for all
the 1970s, McIntyre et al. (1972), McIntyre & Ruddiman species in a given assemblage.
(1972) and Kellogg (1976) used selected planktonic faunal In an attempt to overcome this problem, Imbrie and
indicators, particularly the markedly polar foraminifer Kipp (1971) developed a transfer function approach to
Neogloboquadrina pachyderma, along with the absence of derive palaeotemperature estimates from microfossil
coccolith remains at certain levels in cores from the North data which led to one of the important milestones in
Atlantic, to record the migration of the North Atlantic Polar palaeoceanography and palaeoclimatology: the establish-
Front since the last interglacial. This important indicator ment of the CLIMAP research programme (see especially
Cline & Hays, 1976). This generated a considerable
Right coiling specimens
100 body of data on the Quaternary history of the world’s

90 W. pachyderma oceans and atmosphere, and laid the foundations for
90 90 high-resolution palaeoceanographic reconstructions that
90
TO N. incompta form such an important component of contemporary
GO Quaternary research (see Chapter 7). Within the North
50 Atlantic, for example, where most of the late Quaternary
dominantly sinistral
40
(left-handed coilinal palaeoceanographic investigations were initially concen-
30 trated, Ruddiman & McIntyre (1976) and Kellogg (1976)
dominantly dextral
20 (right-handed ooiling) were able to reconstruct time–space variations in surface-
^0
water mass boundaries for the last 225 ka based on the
0
occurrence of distinctive foraminiferal and coccolith
0 12 14
2 A 6
E 10 16 18 20
assemblages in the deep-ocean cores (Figure 4.45). These
Annual Average SST, °C
water masses appear to have migrated across more than
Figure 4.44 Changes in coiling direction of tests of the high- 20° of latitude (some 2,000 km) at rates of up to 200 m per
latitude species Neogloboquadrina. The proportion of right- year. From these data it would appear that glacial surface
coiling specimens increases markedly in surface waters with water temperatures in the area were up to 12.5°C lower
mean temperature of between 6 and 10°C, reflecting the
replacement of N. pachyderma, which produces mainly sinistral
in winter and 13.0°C lower in summer than at present, and
tests, by the dextral-coiling N. incompta (modified from Kucera, within the last 600 ka alone, at least eleven separate
2007). prolonged southward movements of polar water occurred.
Figure 4.45 Changes in ecological water masses in the Norwegian Sea and the northern North Atlantic over the past 225 ka;
the reconstruction is based on analysis of foraminiferal assemblages in a transect of marine sediment cores obtained from the
sea bed (after McIntyre & Ruddiman, 1972).
The CLIMAP Project (1981) produced maps showing transfer function models have provided a basis for more
inferred sea-surface temperatures (SSTs) for the world’s sophisticated palaeoceanographical and palaeoclimatic
oceans at the Last Glacial Maximum (LGM, c. 18 k 14C yr reconstructions (Kucera, 2007). A significant methodo-
BP) compared with the present. Not only did these provide logical advance has been the development of SIMMAX,
the first quantified model of the state of the oceans during a modern analogue technique that uses similarity indices.
full glacial conditions, but they also highlighted those CLIMAP employed a limited number of indicator spe-
parts of the oceans that experienced the largest tempera- cies and hence could not accommodate the diversity of
ture differences between the LGM and the present day, fossil assemblages commonly encountered in Quaternary
with the major change in the North Atlantic perhaps oceanographical records. As a result, too many fossil
being the most distinctive feature. Such reconstructions samples deviated appreciably from any known modern
provide essential boundary conditions9 for the develop- assemblage (‘non-analogue samples’), and hence could
ment of global circulation models (GCMs) which aim to not be calibrated satisfactorily. SIMMAX is based on a more
simulate past climatic changes at a global scale (section 7.2). comprehensive species list with closer palaeoecological
For example, models of the global climate at 18 k 14C yr BP constraints (Pflaumann et al., 1996). It is periodically up-
developed by the COHMAP group employed the CLIMAP dated and now underpins most palaeoclimatic reconstruc-
global SST reconstructions as boundary conditions for the tions based on marine microfossil assemblages (e.g. Meland
ocean surface (COHMAP Members, 1988). et al., 2005; Kandiano et al., 2012). SIMMAX results
CLIMAP and COHMAP were landmark collaborative are often significantly different from those generated by
programmes that provided new insights into the way the CLIMAP: for example, the latter suggests that the Nordic
global ocean-climate system operates. Since then, however, Seas were permanently frozen during the LGM (Figure
a marked increase in the number of published fossil marine 4.46a), whereas seasonal ice-free conditions extending
records, a more detailed knowledge of the distribution much further north are implied by SIMMAX data (Figure
and ecology of marine organisms, and refinements to 4.46b). A further difference is the location of the strongest
80
a)
N N
b)
80
80 N N
80 80
N N
80 80 N
N 80
80 N
N 80 80
N N
80 80
N N
80 80 N
N
80N
80 0 N 8
8 00NN 0
N8 N8 80N 0 N8
0N 80N N 8 0 80N 80N 80 N 8
80N 80
-2 1-1 20 13
0 0 641 752 863 74 85 96
4 2 53 7 108 -20 -11 021 3 24 350 461 572 683 74 8 5 9 6 710 8
SST (°C ) SST (*C)
80 80
N N N N
80
di
80
di
80 N 80 N
N 80 N 80
80 N 80 N
N 80 N 80
80 N80N N
N 80 80
N N
80
N8 N80 80N N80
0N 80N N 8 0 8 0N 80N N 8 0
80N 80 80N 80
33.0 334 33.3 34.2 34.6 35.0 35.4 35.8 -2 -1 0 1 2 30 41 52 63 74 85 96 7 10

8
S(%.) T PC)
Figure 4.46 Reconstructions of surface conditions in the Northeast Atlantic during the Last Glacial Maximum (LGM) based on
marine microfossil records. a) Summer SSTs based on CLIMAP (1981). b) Summer SSTs based on SIMMAX. c) and d) More detailed
reconstructions for LGM salinities and SSTs based on a larger number of palaeo-data sites and calibrated using SIMMAX (from
Meland et al., 2005).
salinity and temperature gradients, which SIMMAX data Mg/Ca ratio in planktonic foram tests appears to be closely
place to the north, not south, of Iceland (Figure 4.46, b–d). related to temperature at the time of carbonate secretion,
Equally detailed palaeoceanographic reconstructions using an association confirmed by studies of living organisms
SIMMAX have been generated for other marine sectors, recovered from sediment traps and by laboratory culture
such as the Mediterranean Sea (e.g. Hayes et al., 2005), the experiments (Anand et al., 2003; Lea et al., 1999), while
Pacific Ocean (Chen et al., 2005), and the Indian Ocean and many species show remarkably similar temperature res-
Australian margin (Barrows & Juggins, 2005). ponses (Figure 4.47). In addition, some species are restricted
Marine palaeotemperature estimates have also been to narrow temperature ranges that accord with their habitat
obtained from the chemical composition of planktonic requirements, and this enables temperature differences
foram tests (Katz et al., 2010), the most widely used and gradients through the water column to be recon-
being Mg/Ca palaeothermometry (Barker et al., 2005). The structed. Finally, the measurement of Mg/Ca ratios is a
6 20¬ -34
(0.09T
1
G truncatulinoides
0.38*exp(0.09T
obliquiloculata 18 -36
Gsacculifer
sacculifer 12 14
*expT
3 6 10
0.38(0.09
5
5 G sacculifer inflata
with sac 16 :4 67
2 11 13 -38
G. ruberruber
(white) 9
0.38*exp
14
G. ruberruber
(pink)
-40
) T
P. hirsuta
obliquiloculata 12
T(0.09
4 G. hirsuta
inflata
YD 42
G. hirsuta 10 H5
*exp
HI2 H3
(0.09
H2 H4
N. dutertrei
Mg/C
0.38
G. hirsuta
crassiformis 0 10 20 30 40 50
0.38*exp
3
G. hirsuta
conglobatus cal. ka B P
Figure 4.48 Rapid changes in SST in the Alboran Sea

2 (Mediterranean Sea) over the past 50 ka, inferred from
variations in Uk37 abundance, and correlation with Greenland ice-
core events; note that marked declines in temperature in the
Mg/Ca = 0.38*exp(0.09T) Alboran Sea coincide with Heinrich Events and the Younger
1
Dryas cold stage (from Cacho et al., 1999).
10 15 20 25 30
Temperature (°C )
Figure 4.47 Mg/Ca calibration results for several species of initiative (Multiproxy Approach for the Reconstruction of
planktonic Foraminifera. Temperatures shown are the the Glacial Ocean), which has synthesized and quality-
isotopically derived calcification temperatures; the equation tested all the available proxy palaeotemperature records in
defines the correspondence between temperature and
calcification (r = 0.93) (based on Anand et al., 2003). order to reconstruct the global ocean surface temperature
pattern during the LGM (Kucera et al., 2005b; MARGO
Project Members, 2009). MARGO has provided a snapshot
relatively rapid process and requires only small samples,
of ocean circulation at the time of maximum ice extent, and
enabling temperature estimates to be generated at a high
it has computed ocean surface temperature gradients
temporal resolution (James & Austin, 2008).
that differ markedly from those in other climate model
Another approach to marine palaeoclimatology in-
simulations. Similarly, the EPILOG project (Environment
volves the measurement of long-chain carbon compounds
Processes of the Ice Age: Land, Ocean, Glaciers) has
(C37–39 ketones, or ‘alkenones’) preserved in microfossil
integrated surface temperature data for the LGM in the
remains, one of a number of trace organic derivatives (bio-
Southern Ocean, a sector that appears to have played a
markers) now employed in Quaternary palaeoenviron-
major role in past global climate change (Gersonde et al.,
mental research (see section 4.12.6). The compound most
2005). Collaborative programmes such as these represent
commonly measured, the alkenone unsaturated Uk37 index,
the cutting edge in current palaeoceanographical research,
is biosynthesized by certain algae, including coccolitho-
providing important new insights into the ocean’s role in
phores (Rosell-Melé & McClymont, 2007). This process is
global climate behaviour (Chapter 7).
temperature-sensitive, and hence down-core variations
in Uk37 abundance enable temperature fluctuations to be
4.10.8 Marine palaeoproductivity and
quantified. The method has generated highly resolved
palaeotemperature records for the last glacial cycle that
palaeocirculation
correlate closely with Greenland ice-core records (Figure Just as the gyres, thermal stratification, major currents and
4.48); moreover, it appears to have application throughout climate fronts (water mass boundaries) of the contem-
the entire Quaternary period (Cacho et al., 1999; porary oceans are determined by the prevailing circulation
McClymont & Rosell-Melé, 2005). pattern, so too are the distribution of nutrients and the rate
A range of palaeoclimate proxies is therefore now of exchange between the ocean and the atmosphere on
available to the Quaternary palaeoceanographer and, as the one hand, and between shallow and deep ocean water
is the case with other aspects of Quaternary research, on the other. Changes in any of these parameters will be
multi-proxy datasets are being compared in order to test reflected in chemical changes in the water column and in
the degree of compatibility between different indices, ocean sediments. In addition to oxygen isotope signals
and hence to generate more secure palaeoceanograph- (section 3.10), marine microfossils contain a record of
ical reconstructions. A good example is the MARGO former trace element variations in ocean water. This is a
powerful new tool in palaeoceanography, enabling past although the fossil remains indicate that extinct animals
variations in circulation and productivity to be recon- such as the mammoth (Mammuthus primigenius), woolly
structed, and related to changes in atmospheric gas content rhinoceros (Coelodonta antiquitatis) and straight-tusked
(Sigman & Boyle, 2000). A number of chemical tracers elephant (Palaeoloxodon antiquus) were to be found in
(often chemical ratios) are now widely employed (Elder- many areas of the Northern Hemisphere during the
field, 2006). Some provide evidence of water mass and Quaternary (Stuart, 2005), and also that creatures such as
productivity changes, including Mg/Ca, Cd/Ca, Ba/Ca, the hippopotamus (Hippopotamus amphibius) and musk-
Mn/Ca and Zn/Ca ratios, while others, such as 14C content ox (Ovibus moschatus) formerly occupied ranges that are
and 231Pa/230Th ratios, provide information on the pattern very different from those of their present-day counterparts,
and rates of circulation (Lynch-Stieglitz et al., 2007). Some albeit at different times, until comparatively recently much
tracers, such as isotopes of neodymium (Nd), are useful for less was known about the smaller vertebrates and their
detecting the mixing of water masses or significant influx Quaternary distributions. In the last few decades, however,
of freshwater into the marine realm (Scrivner et al., 2004). considerable progress has been made in understanding
An important indicator of past variations in nutrient supply the spatial and temporal variations in both large and small
is phosphorus (P), a vital nutrient for all living organisms. animal populations during the Quaternary, and their
Past variations in P concentration cannot, however, be relationships to environmental changes.
measured directly, but are inferred from Cd/Ca ratios. A range of vertebrate remains is found in Quaternary
Cd is preferentially taken up by planktonic forams during sediments. Occasionally hair, muscle and horn-sheaths
skeletal building, and because Cd/Ca ratios in foram tests are preserved, and, in exceptional locations such as the
are proportional to the amount of P in the ocean, they can permafrost of arctic Siberia, tar pits and peat bogs, mum-
be used as a proxy for nutrient (P) levels (Elderfield & mified carcasses have been found (Spencer et al., 2003b;
Rickaby, 2000). Other marine water mass properties that Fisher et al., 2012). Other evidence of the former presence
can be inferred from chemical tracers in palaeoplankton of animals includes nests and middens of rats, burrows,
include alkalinity, pH, degree of oxygenation and nutri- hyena dens, coprolites (droppings: section 3.8), bird pellets,
ent uptake (Henderson, 2002). Chemical tracers can be diagnostic teeth marks by predators on other bones, ‘trace
measured in deep-water benthic species and hence provide fossils’ (prints and tracks) and skincasts. Overall, however,
information on ocean circulation at depth, allowing it is teeth and bones (and, occasionally, antlers) that make
inferences to be made about intermediate and deep-ocean up the majority of fossil vertebrate remains (Figure 4.49),
properties and behaviour (Flower et al., 2000; Rickaby and and these form the principal focus of the following
Elderfield, 2005). By analysing chemical tracer information discussion.
for particular time-periods therefore, ocean density
stratification and overall circulation can be inferred
(Rogerson et al., 2011), information that is key to
4.11.2 The structure of teeth and bones
understanding the role of oceans in the global climate Teeth are structurally complex but in most mammals
system (Rahmstorf, 2002). consist of three distinct substances of differing hardness:
the hard brittle outer casing (enamel), the softer dentine
of which the greater part of the tooth is composed, and
4.11 VERTEBRATE REMAINS cement which covers the dentine of the roots and occa-
sionally the valleys and folds of the main tooth body. In
4.11.1 Introduction the fossil, the enamel provides the most durable element
Fossil animal bones and teeth, particularly those of large except where burning has affected the original dental
vertebrates, have long been an attraction for amateur material, in which case the dentine of the tooth roots may
collectors and, as a result, museums are full of the skeletal prove to be the most resistant. Teeth are of consider-
parts of Pleistocene mammals. Many of these remains able importance in palaeoenvironmental work for not
were removed from exposures in cliffs and in river valleys only do they provide data on the age (years of life) of the
towards the end of the last century by well-meaning animal, but they also give an indication of dietary prefer-
Victorian enthusiasts who, unfortunately, often paid scant ences (i.e. herbivore or carnivore), while isotopic analysis
regard to the stratigraphic context within which they lay, can provide additional information on diet, habitat and
or indeed to the less spectacular but equally important climatic conditions (Kohn & McKay, 2012). In more recent
collecting of smaller animal remains which together formed sediments, teeth tend to be outnumbered by remains of
the total assemblage of the stratum (Schreve, 2007). Thus, bone, but in older Quaternary deposits, teeth are occasion-
a)
b) c)
Figure 4.49 a) Nigel Larkin cleaning the fossil right femur (length 145 cm) of a mammoth of Middle Pleistocene age discovered
at the site of West Runton, Norfolk, UK (photograph by Nigel Larkin, Norfolk Museums & Archaeology Service, Norfolk, UK; from
Larkin, 2010, reprinted with permission from Elsevier). b) Occlusal surface (top) and roots of the first lower molar of the ancestral
water vole (Mimomys savini) recovered from sediments of early Middle Pleistocene age at Pakefield, Suffolk, UK. Variations in
dentition enable fossil vole teeth to be assigned to species and the relative age of temperate stages to be inferred (section 5.5.4).
Scale bar represents 1.0 mm (photograph by Harry Taylor & Simon Parfitt, Natural History Museum, London, UK; from Maul &
Parfitt, 2010, reprinted with permission from Elsevier). c) Reconstruction of woolly mammoth (Mammuthus primigenius) based
on soft tissue, skin, hair, parts of the intestines and delineation of toes recovered from one of the best preserved mammoth carcasses
ever discovered, in Yakutia, arctic Siberia (reconstruction and photograph by Remie Bakker of Mammal Works & Dick Mol of the
Natural History Museum, Rotterdam, Netherlands; from van Geel et al., 2008, reprinted with permission of Elsevier; Lister & Bahn,
2009).
ally more strongly represented than elements of the post- and water. Mineral salts in solution in the surrounding
cranial skeleton (Dayan et al., 2002). sediment, particularly calcium and iron, will be deposited
Fresh animal bone consists of both organic material in the vacant pore spaces and the bone may eventually
and inorganic material in the approximate ratio by weight become completely permineralized, and more resistant to
of 30:70. The organic fraction is contained within the further decay (Turner-Walker, 2008). This is one reason
shafts of long bones (e.g. femurs, tibias and vertebrae) and why mammalian remains tend to be well preserved in cave
comprises cell tissue (fats, etc.) and a fibrous protein called deposits in limestone regions. However, Quaternary fossil
collagen. The collagen is very resistant to decay and may bones are often mechanically weak, because loss of struc-
survive for thousands of years following the death of tural collagen happens quickly, while the hardening pro-
the animal, while the remaining organic matter under- cess of mineralization is usually much slower; particular
goes autolysis10 after death and is rapidly decomposed. care is needed, therefore, during excavation and sample
Surrounding the collagen fibre is bone mineral material, transfer, especially with large bones (Larkin, 2010; Figure
the principal component of which is a phosphate of cal- 4.49a). In acid soils and peats, which are depleted in
cium, hydroxyapatite (Ca10OH(PO4)6). The structure and bases, both the organic and mineral fractions decompose
composition of animal bones are of considerable interest quickly and the bone will disappear completely, leaving no
to the palaeoenvironmentalist as they affect the way in trace of its former existence. Thus while prehistoric burials
which fossilization takes place, and the chemical structure on chalklands in areas such as southern England have
of bones in particular provides a means of dating the fossil often yielded well-preserved bones, those in adjacent
material (see section 5.6.2). regions where porous, sandy soils are found contain few
bone remains.
In waterlogged contexts, a completely different set of
4.11.3 Fossilization of bone material reactions occurs. In deep lakes in limestone regions where
Quaternary vertebrate remains have been recovered from bases are abundant, bones are not only well preserved but
a wide range of deposits. These include cave and fissure are often extremely hard. In some cases, even the organic
sediments, lacustrine and marine deposits, fluvial sediments elements have been converted into a stable wax-like
(especially river terraces), peat bogs, soils and a variety of substance, composed of fatty acids, known as adipocere
situations associated with human activities such as middens, (Ubelaker & Zarenko, 2011). At the other extreme, in peat
cesspits and burial chambers. At some sites, whole skeletons bogs or in oligotrophic lakes, the anaerobic nature of the
have been found, but more frequently the fossil assemblages depositional environment often results in the organic
consist of disarticulated skeletons and a mixture of bones portions of the bones being preserved, while attack by
of varying sizes and in differing states of preservation. humic acids leads to complete destruction of the mineral
Animal bones are perhaps more vulnerable to physical fraction. Skeletal remains will, therefore, be found in a soft
and chemical changes than any other biological remains or pulpy state in advanced stages of decalcification (Adams
encountered in Quaternary deposits. They are highly et al., 2007).
susceptible to ‘weathering’ by a variety of processes, if Finally, there are the effects of burial on bone that are
exposed to oxidation, temperature extremes or chemical purely physical. The seasonal drying of clay soils, for
exchange with ground solutions. In some instances, the example, will result in the fissuring and eventual destruction
extent of weathering has been used to estimate duration of of even the strongest bones. Bones may be similarly
exposure of bone assemblages, but this is no longer con- shattered by frost-heaving and by the action of ground ice.
sidered to be a reliable approach, because bones degenerate Soil creep, solifluction and mechanical abrasion in river
differentially, even within the same site and sedimentary gravels will have similar damaging effects and will result in
layer (Nielsen-Marsh & Hedges, 2000), and more sophisti- the progressive fragmentation of bone remains.
cated methodologies are needed to establish the diagenetic
history of bones and teeth (Szostek, 2009).
As soon as a bone becomes incorporated into a body of
4.11.4 Field and laboratory techniques
sediment, it begins to undergo chemical changes that vary Before any excavation of bone assemblages can take place,
in nature and degree with the chemistry of the surrounding legal and ethical requirements must be considered, not only
matrix. In most deposits where air is freely circulating, in the interests of conservation, but also because of the
the mineral parts of bone will tend to be more resistant to possibility that human remains form part of the assemblage.
decay, while the organic substances will break down rapidly It should not be assumed that all the bones are ancient,
into simple compounds such as ammonia, carbon dioxide while some sites have particular religious or cultural
importance and may be protected by legislation (Marquez- cation purposes. In spite of these difficulties, however,
Grant & Fibiger, 2011). Appropriate permissions must positive identifications of Quaternary vertebrate remains
be obtained in advance, and excavation protocols strictly are steadily increasing and it is now proving possible to
followed. construct fairly detailed faunal lists for the major stages of
Because bones can be found in such a variety of con- the Quaternary (Kurtén, 2009; Currant & Jacobi, 2011). In
ditions, particular care must be exercised in the excavation addition, the palaeobiogeography and evolutionary trends
of bone-bearing deposits. Mapping and surveying aug- in many groups of animals can now be reconstructed in
mented by field description, sketches and photographs detail, and are frequently accompanied (and enhanced) by
should precede the removal of bone fragments from the DNA analysis (section 4.11.6.4).
sediment matrix. In some cases, it may be possible to
remove the larger bones by hand. These can be left to dry
4.11.5 The taphonomy of fossil vertebrate
out and then cleaned with a brush or by gentle agitation in
water. Large bones that are cohesive can be protected
assemblages
by encasing in plaster of Paris and hessian, prior to removal The first stage in the interpretation of fossil bones and
(Larkin, 2010). Many bone remains, however, even if teeth is to establish how a particular grouping of vertebrate
heavily mineralized, are quite brittle and it may be necessary remains came to be associated together. The various factors
to treat these with a penetrative epoxy resin hardener that can influence the reconstruction of the living com-
before removal from the matrix can be attempted; this munity that is represented in the fossil assemblage are
process is best preceded by anatomical examination of reviewed by O’Connor (2005). Three different depositional
subsamples of the bones to establish the degree of decay environments will serve to demonstrate the complexities of
(Jans et al., 2002). If the bones are wet, resins or emulsions fossil vertebrate assemblages.
of lower viscosity may be needed in order for the
strengthening material to penetrate deeply into the bone
4.11.5.1 Cave and fissure deposits
fibres (Smith, 2003). Bones that are so treated, however,
cannot be used for subsequent chemical analysis or for Some of the richest vertebrate assemblages in the world
radiocarbon dating. The presence of very small bones or are those found in cave sediments, particularly in limestone
teeth of micro-vertebrates (e.g. rodents) can usually only regions, yet the ecological history of cave faunas is fre-
be detected by sieving the matrix following the removal of quently very difficult to interpret because of the multiple
the larger faunal remains by hand, but all sample material origins of the fossil material (Jass & George, 2010). Some
needs to be examined for small bone fragments to reduce bones, for example, may have been washed into the caves
sampling bias (Stahl, 1996). The smaller specimens (< 2 or fissures by streamflow and are therefore allochthonous
mm in size) are identified under a binocular microscope to the site (Figure 3.37). Caves were often occupied during
(Maul & Parfitt, 2010). the Quaternary by carnivores including cave bear (Ursus),
Identification of bone remains is usually carried out in wolf (Canis lupus), red fox (Vulpes vulpes), sabre-toothed
the laboratory and often proceeds in two stages. As most cats (e.g. Smilodon) and spotted hyena (Crocuta crocuta),
identifications are based upon fragmentary evidence, and hence many cave assemblages will be biased towards
the first step is to identify the bone of which the fragment the prey of these animals. Small vertebrate remains could,
is a part. This is usually achieved by comparing the fragment for example, have been derived almost entirely from
with fresh skeletal material from a range of animals of birds of prey, especially from pellets dropped by owls
different sizes. The second, and more difficult, stage is to nesting in the cave roof, and could include either woodland
track down the animal from which the unknown bone was or open-country rodents depending on the species involved
derived. Here a reference collection of type material is (Marín-Arroyo et al., 2009). Many of the large vertebrate
essential, although the development of a type collection for bones will have been dragged into the cave by predators
the Quaternary vertebrates involves many more difficulties so that the resulting assemblage will give some indication
(and considerably more expense) than are encountered in of the original large vertebrate fauna of the vicinity. How-
the construction of a reference collection for Quaternary ever, the cave assemblages will inevitably be biased towards
pollen grains or coleopteran remains. The Mammalia, for the predators themselves as many would have eventually
example, include a proportion of taxa that are now extinct, died in the caves and thus contributed their bones to
while evolution and speciation during the Quaternary the assemblage. This is particularly the case where caves
pose additional complications. Moreover, few museums acted as natural pitfall traps with animals having fallen
possess a reference collection that is suitable for identifi- in through holes in the cave roof (Kos, 2003), or where the
configuration of a cave opening allows animals to enter are, however, more difficult to interpret. Some may be
the cave but from which they were unable to escape. In the remains of waterside creatures such as voles and rats,
both instances, the resulting bone assemblages will be while others may have been delivered to stream banks
partly biased towards scavenging animals, such as the or the water directly by carnivore droppings or by predator
hyena, which would have been attracted to the cave by dead birds (Louchart et al., 2009). Bird remains, however, are
and dying animals. Early humans may also have been seldom preserved in lacustrine or fluvial deposits due to the
attracted to such easy prey, and it is frequently difficult to extreme fragility and buoyancy of their bones; as a
distinguish their activities from those of other scavengers consequence the Quaternary history of the avian fauna is
(Kuhn et al., 2010). The difficulties of interpretation of still poorly understood.
vertebrate assemblages in cave sites are further exacerbated A further difficulty in the interpretation of animal
by the often complex stratigraphy of cave sediments (Figure remains from fluvial deposits is that the assemblages are
3.37; section 3.8). frequently biased as a result of hydraulic sorting. Fluvial
transportation, particularly during floods, leads to selective
fragmentation or removal of bones according to size and
4.11.5.2 Lacustrine sediments
state of fragility, a process that tends to be more common
Lake sediments often contain whole or partial skeletons of in open channels where stream power is high and bed
mammals, amphibians and fish. Remains of large mammals material coarser (channel-lag assemblages) than in over-
such as elk (Alces alces), reindeer (Rangifer tarandus) and bank or meander beds (channel-fill assemblages) (Behrens-
mammoths (e.g. Mammuthus primigenius) found in lake meyer, 1988), and careful analysis of both the bones
deposits probably represent animals that died either by and their host sediment is needed in order to gauge the
drowning after breaking through thin ice, or after having possible effects of hydraulic sorting (Price & Webb, 2006).
become trapped in the soft mud on the lake floors in their An additional complication is that vertebrate remains of
efforts to drink, wallow or feed (Lister, 2009). Often the different ages may be incorporated into the sediments
adjacent sediments will have been disturbed by the struggles as the river banks are eroded. Because similar (but not
of the animal to become free. Many human occupation necessarily identical) animal populations existed during
sites, such as the early Mesolithic hunting settlement at Star successive warm and cold stages of the Quaternary, the
Carr in Yorkshire, were by lakes and rivers and therefore a likelihood of erroneous ecological interpretations from
proportion of the remains of animals that were hunted also mixed bone assemblages is very real, and care needs to be
found their way into the lake (Milner et al., 2011). Fish and taken to establish the degree to which the mix of skeletal
amphibians clearly reflect the former presence of these remains is contemporaneous (Lewis et al., 2006). Some-
animals in the lake waters, but again humans may have been times it may be possible to recognize bones of different ages
responsible for the concentration of faunal remains in the on the basis of varying degrees of physical deterioration.
littoral sediments. Alternatively, relative ages of bones may be established by
chemical means, for example by the use of amino-acid
ratios (section 5.6.1) or of the fluorine, uranium or nitrogen
4.11.5.3 Fluvial sediments
content of bones (section 5.6.2), although in some cases
The origins of vertebrate remains in river sediments can direct dating using the U-series method may be possible
be almost as diverse as those found in cave deposits, (section 5.3.4).
and the assemblages may include the bones of a wide
variety of species, particularly if deposited by rivers with
large catchments (Rogers et al., 2007). Large vertebrate
4.11.6 Quaternary vertebrate records
remains become incorporated in riverine deposits in sim- The record of Quaternary vertebrates in both space and
ilar ways to those outlined above for lacustrine contexts, time is extremely patchy, for fossil bone assemblages,
though allochthonous material is common because corpses particularly those of large mammals, tend to accumulate
can float downstream. From an analysis of the assem- episodically. In river terrace deposits, for example, they are
blage, it may be possible to gain some impression of rela- often found in channel fills that were quickly abandoned
tive population densities, of the lifespan of particular because rivers change course (Bridgland & Schreve,
taxa, and of the distance of their habitats from the site 2009), while in caves, intermittent occupancy by predators
of deposition (Faith et al., 2009). Fish and amphibian or humans, flooding of cave passages and collapse of roofs
remains will be locally derived and will tend to be over- and walls means that cave bone assemblages are com-
represented in the assemblage. The small animal remains monly discontinuous. In addition, vertebrate species have
experienced marked and recurrent shifts in territorial range 4.11.6.2 Vertebrate biogeography
in response to Quaternary environmental and climatic
changes, and many have been characterized by a degree of Correlation of vertebrate stratigraphic records also
biogeographical, and in some cases evolutionary, change. enables the geographical ranges of individual species to be
In this section we consider these different aspects of the reconstructed, and the ways in which these have changed
Quaternary vertebrate record, but we begin with vertebrate in response to climatic and other environmental influ-
biostratigraphy, as this provides the foundation on which ences. These range shifts were especially marked in the
the other studies depend. high latitudes with the successive advances and retreats of
the great continental ice sheets, and the corresponding
latitudinal shifts in the permafrost zone. Although much
4.11.6.1 Vertebrate biostratigraphy remains to be learnt about the consequent changes in
Analyses of Quaternary vertebrate assemblages have faunal population dynamics (Lyons, 2005), it is clear that
enabled Quaternary stratigraphic ranges (appearance and many large mammals, particularly carnivores, had more
extinction dates) to be established for many vertebrate extensive distributions in the past than they do today
species in Europe (Kurtén, 2009; Currant & Jacobi, 2011), (Hofreiter & Stewart, 2009). The cave lion (Panthera
North America (Webb et al., 2003b), Africa (Werdelin spelaea), for example, was widespread in northern Eurasia
& Lewis, 2005) and South America (Prevosti et al., 2009). and the Yukon–Alaska region during the late Pleistocene,
The stratigraphic ranges of individual vertebrates and of but became extinct at the end of the last cold stage, allowing
vertebrate assemblages have proved to be of particular the modern lion (Panthera leo) to expand into some of its
value in distinguishing between interglacial deposits of former territory during the Holocene (Stuart & Lister,
different ages in fluvial records (Schreve, 2007). In the lower 2011). The European spotted hyena (Crocuta crocuta) was
Thames Valley, for example, diagnostic vertebrate assem- common throughout Europe (Figure 4.50) until the end
blages in terrace sequences enabled a detailed biostrati- of the last cold stage, when it contracted in range and is
graphy for the interglacials of the Middle Pleistocene to be now confined to Africa (Varela et al., 2010). It appears that
established (Figure 5.44), and this could be used to correlate larger-bodied vertebrates were able to expand their ranges
these fluvial records with those from central Germany more readily and over greater distances than smaller-
(Bridgland et al., 2004). Although the majority of such bodied species, perhaps because the former had better
studies have tended to focus on large mammals, first dispersal abilities and were less inhibited by physical barriers
appearance and extinction evidence has also been obtained (Lyons et al., 2010).
for a number of small mammal species, from which dis- The Quaternary fossil vertebrate record shows how
tinctive faunal assemblages or ‘provinces’ (also termed many mammals adapted to changing habitats during the
‘guilds’) have been identified, and which are diagnostic of Quaternary, some to exploit new niches that arose as global
particular Quaternary intervals. For example, Cuenca- climate cooled, while others became more flexible in their
Bescós et al. (2010) have shown that the stratigraphic
record of the last 1.5 Ma in Spain is characterized by
seven distinct provinces defined by changes in the fossil • Last interglacial
assemblages of small mammals, while Markova (2005) a c. 42 ka
30 ha
subdivided the stratigraphic record of the last 1 Ma in
21 ka |
Russia using fossil rodent remains. In other cases, a com-
bination of small and large vertebrate data have been used
as biostratigraphic indicators as, for example, in the
differentiation between MIS 9 and MIS 11 interglacial
stages in fluvial deposits from Essex, eastern England
(Roe et al., 2009). In addition to being diagnostic fossils of
particular time intervals, some vertebrates have been
employed as biomarkers in stratigraphic sequences, most
notably the evolutionary transition from Arvicola to
Mimonys in the vole lineage, which forms an important
marker horizon within Cromerian (Middle Pleistocene) Figure 4.50 Age estimates of Pleistocene records of the
deposits in western Europe, and which is discussed further spotted hyena (now confined to Africa) in Europe (from Varela
in section 5.5.4. et al., 2010).
requirements and were able to exploit a range of different appears to have occurred spontaneously when species
habitats (Lister, 2004). Also apparent from the fossil record competition and population dynamics have been altered.
is the gradual decline, from the early Pleistocene onwards, But there are examples where the island rule has not applied
in the importance of solitary carnivore hunters. By the consistently, as in the vertebrate record from Santa Rosa
middle Pleistocene, these had become increasingly out- Island, California, where a pigmy mammoth, Mammuthus
numbered by larger, more social predators such as Panthera exilis, seems to have coexisted with the much larger M.
leo, Crocuta crocuta and Canis lupus, perhaps reflecting the colombi (Figure 4.51) between c. 200 and 11 ka (Agenbroad,
expansion of grasslands that supported large herbivores 2012). Much clearly remains to be learned about morpho-
(Croitor & Brigal, 2010). logical changes in both current and historical isolated
One feature of the Quaternary vertebrate record is animal populations, and the Quaternary vertebrate record
evidence for changes in animal morphology, particularly is an important source of evidence in these lines of enquiry
in isolated populations. It has long been known that when into faunal biogeography.
animals become isolated for long periods of time they can
change size: some small mammals grow progressively larger
4.11.6.3 Vertebrate fossils and Quaternary
(gigantism) while others become smaller (dwarfism). The
environments
Quaternary fossil vertebrate record contains many instances
of such changes, including the skeletal remains of ‘pygmy’ It has long been assumed that there is a close link between
hippopotami and elephants (Figure 4.51) as well as giant Quaternary vertebrates and prevailing climatic regime
rodents (Millien et al., 2006). In the case of predator–prey (Blois & Hadly, 2009). In low-latitude regions, vertebrate
dynamics, it has been suggested that there is a reduction in records tend not to reflect major thermal changes between
the size of large predators if the size or number of prey glacial and interglacial episodes, but rather longer-term
declines, whereas small animals become larger when trends related mainly to precipitation. In central Queens-
isolated from their natural predators (Raia & Meiri, 2006). land, Australia, for example, contrasting mammalian
Of particular interest are islands, where a marked change assemblages indicate that tropical rainforests were present
in body mass, reduced species diversity and accelerated throughout the early and middle Pleistocene (until c. 280
evolution in animal communities is such a recurrent ka), after which there was a switch to the arid regime
phenomenon that it has come to be known as ‘the island that characterizes the region today (Hocknull et al., 2007).
rule’ (van der Geer et al., 2010). Indeed, this rule may also In the Turkana Basin, east Africa, vertebrate records indi-
apply to hominids, for on the island of Flores, Indonesia, cate that a savannah regime persisted in that region for
the skeletal remains of a dwarfed hominid species (Homo almost the whole of the Quaternary, whereas rainforest
floresiensis) have been found in association with bones of and deciduous woodland had characterized the earlier
other animals showing signs of either dwarfism or Pliocene period (Fernández & Vrba, 2006). In the Pampas
gigantism, and are believed to be the result of isolation region of South America, the modern wet subtropical
sometime between 90 and 17 ka (Van den Bergh et al., fauna was established as recently as 1.5–1 ka, prior to
2009). In these and other instances, evolutionary change which conditions were drier, especially during the Last
Glacial Maximum when the area was largely a cold desert
(Tonni et al., 1999).
In higher latitudes, however, vertebrate faunal assem-
blages are more reflective of contrasting thermal condi-
tions of interglacial and glacial episodes (Lister, 2004),
although the relationships between faunal assemblages
and prevailing climate have not been quantified to the same
degree as for other biological proxies discussed earlier in
this chapter. Traditionally, the indicator species approach
has been used to derive palaeoclimatic information from
fossil vertebrate records. During the last interglacial,
for example, hippopotamus (Hippopotamus amphibius), the
Figure 4.51 Reconstructions, based on skeletal remains, of pond tortoise (Emys orbicularis) and the lesser white-
the large (Mammuthus colombi) and pygmy (M. exilis) mam-
toothed shrew (Crocidura cf. suaveolens) were all found in
moth species that co-inhabited Santa Rosa Island, California
between c. 200 and 11 ka (from Agenbroad, 2012, reprinted southern Britain. The hippopotamus is now confined to
with permission of Elsevier). tropical Africa, the pond tortoise is now found only in the
Mediterranean and in southeast Europe with its northern change, some in scattered cryptic refugia. As we saw earlier
breeding range apparently limited by the 18°C July iso- (section 4.4.6.2), these are refugia that are situated at
therm, while the lesser white-toothed shrew is also essen- different latitudes or longitudes than would normally be
tially southern European in its distribution. These data expected, and often resemble climatic islands in which
are, therefore, strongly suggestive of warmer summers conditions differ, often favourably, from the surround-
and milder winters in the British Isles during the Last ing areas. The manner by which animal communities
(Ipswichian) Interglacial by comparison with the present were subsequently reassembled depended on complex
day, a hypothesis supported by both palynological and population interactions and perhaps an element of seren-
coleopteran evidence (Stuart, 1979). dipity (Stewart et al., 2010). As a result of this continuous
More recent studies have attempted to derive quanti- mixing and renewal of faunal associations during successive
fied climatic envelopes for groups of vertebrate species glacial–interglacial cycles, non-analogue species combina-
with common thermal preferences. For example, Fernández tions may have been the rule rather than exception, the
(2006) applied modern bioclimatic limits for rodent degree of divergence between ancient and modern faunal
species to fossil rodent assemblages in order to reconstruct groups seemingly increasing with the age of the assemblage
temperature variations in continental Europe over the last (Stewart, 2009).
interglacial–glacial cycle. Polly & Eronen (2011) used an A further constraint on the use of modern bioclimatic
approach similar to the MCR method described above data as a basis for inferring past environmental conditions
(section 4.5.4.2) to construct ecological niche models for is the degree to which humans have affected animal
groups of vertebrate animals typical of cold (Alopex lagopus, populations and distributions during the late Quaternary,
Lemmus lemmus, Ovibos moschatus and Rangifer tarandus), either through hunting or domestication, the latter being
warm (Crocuta crocuta, Panthera leo and Hippopotamus accompanied by accelerated genetic modification (Bar-Oz
amphibius) and temperate climates (Arvicola terrestris, & Nadel, 2013). Humans have been implicated in the mass
Cervus elaphus and Sus scrofa). Application of these models extinction of large mammals between c. 50 and 7 ka, when
to British fossil vertebrate records generated maximum more than 178 of the world’s largest mammals disappeared.
probability temperatures that accorded with indepen- In North America, for example, some thirty-five genera
dent estimates based on other proxies, and particularly on (more than seventy species) became extinct between 13 and
beetle MCR data. However, a straightforward relation- 11 ka (Faith & Surovell, 2009). Many small mammals also
ship between animal distribution and climatic parameters disappeared at around the same time (Blois et al., 2010),
cannot always be assumed for, as pointed out by Frederick but it is the fate of the larger animals that has attracted most
Zeuner in 1959, the primary adaptation of many vertebrates attention. The peopling of the Americas appears to have
during the Quaternary may well have been to vegetation occurred shortly before this extinction event (Davidson,
and only secondarily to climate. In the fifty years since 2013), but the demise of the mammals also coincided with
Zeuner made that observation, a number of other problems the climatic warming at the Pleistocene–Holocene tran-
have emerged. For example, from an analysis of the fossil sition, and the debate continues as to which of these factors
mammal record of the United States, Graham et al. (1996) was primarily responsible for the disappearance of large
concluded that climatic change could not by itself explain numbers of the megafauna from the American landscape
the changing distributions of mammals during the late (Haynes, 2009). In northern Eurasia, megafaunal extinc-
Quaternary, and that other factors, such as habitat change, tions seem to have begun earlier, at around 50 ka, close to
species interactions and stochastic events must also have the time of the arrival of anatomically modern humans,
been involved. A further difficulty is that it now seems who brought superior hunting tools and skills, but again
that many fossil vertebrate assemblages have no modern this was also a time of rapid environmental change, with
analogues, which clearly creates a problem in using modern the major climate shifts associated with the Dansgaard–
bioclimatic data to infer past climatic conditions (Hofreiter Oeschger cycles (section 3.11.4). As in North America, there
& Stewart, 2009). This problem has been thrown sharply appears to have been a further wave of extinctions around
into focus by recent work on animal refugia during the the time of rapid warming at the onset of the Lateglacial at
Quaternary. The traditional view has been that in Eurasia c. 14.7 ka (Stuart & Lister, 2012). In the Australia–New
and North America, temperate-adapted vertebrates were Guinea region, about sixty megafaunal species disappeared
confined to shared southern refugia during glacial periods during the Middle to Late Pleistocene, which some have
and cold-adapted fauna to northern refugia during interlinked to the arrival of humans, perhaps from c. 60 ka
glacials. The vertebrate fossil record shows, however, that onwards, although here too a climatic influence cannot be
species responded individually to rapid environmental excluded (Field et al., 2008). That said, however, it is clear
from the fossil record that extinctions on this scale did not been an assumption, seemingly initially confirmed by
occur earlier in the Quaternary at similar times of marked DNA analysis (Lindqvist et al., 2010), that the polar bear
climate change and, moreover, that the selective dis- (Ursus maritimus) is a relatively recent evolution from the
appearance of large mammals is unusual, both of which brown bear (U. arctos), though more recent DNA results
would tend to favour human agency and lend support to suggest that the polar bear is a much older and distinct
the ‘overkill hypothesis’ which attributes the mass extinc- bear lineage that evolved during the Middle Pleistocene
tions to the indiscriminate slaughter of animal herds (Koch (c. 600 ka: Hailer et al., 2012), or maybe even earlier (Miller
& Barnosky, 2006). The consensus view, however, is that a et al., 2013).
combination of these two factors may have been respon- DNA analysis is also proving to be important in
sible, with abrupt climatic changes reducing and scattering vertebrate phylogeography, i.e. the study of how the
the large mammal populations, before the development modern distributions of different animals evolved, which
of human hunting technology tipped the balance towards has provided new insights into animal history and
extinction (Haynes, 2013). Other theories of the cause of behaviour, including the identification and locations of
extinctions involving, for example, transmission of patho- key animal refugia and colonization routes (Hewitt, 2011).
gens and spread of disease by humans or animals (e.g. dogs) In North America, for example, DNA analysis of fossils
travelling with them (Fiedel, 2005), or the environmental of the woodland caribou (Rangifer tarandus caribou)
effects (e.g. wildfires) of a cometary impact (Firestone et al., indicates that this animal withdrew during the last glacial
2007), have proved more controversial and have not gained stage to three geographically discrete refugia: in the Rocky
widespread support (Boslough et al., 2012). Mountains, east of the Mississippi and the Appalachian
Mountains (Klütsch et al., 2012). The present phylo-
geography of this animal comprises three genetic lineages
4.11.6.4 Vertebrate fossils and faunal
that reveal a pattern of expansion from these refugia after
evolution
the wastage of the Late Wisconsinan ice sheet. Miller et al.’s
The frequent climatic and environmental changes that (2013) study of polar bears (see above), suggests that these
characterized the Quaternary appear to have resulted in probably came into contact and interbred with brown
accelerated rates of morphological adaptation, and to a bears during interglacial stages at times when their
degree of evolution, certainly when compared with pre- territories overlapped; hence, although polar bears are now
ceding geological periods. This is reflected in calculations exclusively Arctic animals, they must have been sufficiently
of ‘turnover rate’ (the sum of first and last appearances adaptable throughout the Quaternary to cope with a long
of species during a geological interval, divided by the total series of climatic fluctuations. In Europe, DNA analyses of
number of species recorded for that interval), and estimates fossil remains of the collared lemming (Dicrostonyx
for the Quaternary are very much higher than for com- torquatus) indicate that this animal was driven to extinction
parable time spans during the Neogene. Remarkably, nearly on numerous occasions as a result of climatic fluctuations
all the mammal species present in Europe at the start of the and, since this small mammal is an important prey species,
Quaternary have now disappeared, while none of the this must have impacted on the wider steppe–tundra animal
present-day mammal species were present at around 2.5 Ma community (Brace et al., 2012). This process of episodic
(Lister, 2004). DNA analysis has certainly shown that extinction and regeneration led to a significant loss of
genetic diversity changed more often, and in unpredictable genetic diversity in collared lemmings, a trend that is also
ways, during the Quaternary, much more so than had noted in a number of other vertebrate species during the
been anticipated from previous studies of bone assemblages late Quaternary (Hofreiter & Barnes, 2010; Barnosky et al.,
(Hofreiter & Stewart, 2009). 2011).
In considering adaptation of animals to environment, These various studies attest to the dynamic nature of
a distinction needs to be made between short-term changes phenotypic and phylogenetic interchange during the
in animal traits (e.g. body shape, skin colour and markings), Quaternary, and it seems clear that Quaternary climate
which is termed phenotypic plasticity (a phenotype is the oscillations had a marked effect on the gene pools and
sum of all the traits that make up the physical appearance phenotypic adaptability of vertebrate animals (Stewart,
of an individual organism), and genetic modification, 2009). As noted in Chapter 1, Milankovitch-driven climatic
which is a much longer process that results in irreversible oscillations became more pronounced after the mid-
changes that lead to speciation. Data from Quaternary Pleistocene Revolution, in turn forcing the biota to respond.
mammals provide evidence of the rate and manner in This may have militated against specialization in species,
which these processes take place. For example, there has leading instead to the development of animal guilds that
were more adaptable, had larger territorial ranges and were been less well studied than other lake biota (Eggermont &
able to disperse quickly. This evolutionary trend, a product Martens, 2011). They mainly occupy the littoral zones of
of serial environmental upheaval, has been termed ‘orbitally lakes and ponds, though a number of species are planktonic
forced species range dynamics’ (Dynesius & Jansson, 2000). and occur in open water, but skeletal fragments of both
Over shorter timescales, the question arises as to how forms accumulate in the deeper parts of lakes. They have
animals coped with the rapid climatic and environmental been most widely used as indicators of past changes in lake
changes that occurred during, for example, the last glacial level, because cladoceran assemblages and the ratio of
stage and where temperature fluctuations of several degrees planktonic to littoral cladoceran species vary with water
Celsius occurred in a matter of decades (section 3.1.4). The depth (Nevalainen et al., 2011). Their potential as palaeo-
possible phylogenic effects that may have resulted from climatic indicators is less clear, however. For example,
these changes are only now beginning to attract the Kattel et al. (2008) developed a Cladocera-based tem-
attention of geneticists and evolutionary biologists (e.g. Hof perature transfer function for small lakes in Scotland,
et al., 2011). but found that inferring a relationship between cladoceran
assemblages and climate was complicated by other environ-
mental factors. A similar problem was encountered in
4.12 OTHER FOSSIL GROUPS Canadian Arctic lakes, where the response of Cladocera to
recent global warming appears more muted compared
4.12.1 Chrysophytes with that of diatoms (Sweetman et al., 2008). These and
Chrysophytes are a group of planktonic (‘golden’) algae other studies seem to suggest that the climatic signal in fossil
that produce resting cysts (stomatocysts or statospores) cladoceran assemblages may be masked by the influence of
during their life cycle, while a subset are also covered by additional ecological variables, including trophic state,
siliceous scales (‘scaled chrysophytes’). As siliceous water depth and habitat availability (Eggermont & Martens,
organisms, chrysophyte cysts and scales are commonly 2011).
found in association with diatoms, and are frequently
analysed together. Chrysophyte remains are deposited in
great abundance and preserve well in sediments, while the
4.12.3 Coral polyps
majority can be identified to species level. Since most are Coral reefs are formed by a number of interlocking
sensitive to a range of environmental variables, including elements, including autochthonous and allochthonous
pH, lake chemistry, nutrient status and metal concen- remains of animals and plants, clastic sediment and
tration, they have been widely employed in palaeolimnology chemical alteration processes (Montaggioni & Braithwaite,
(Zeeb & Smol, 2001), but they can also be useful palaeo- 2009). The main ‘builders’, however, are coral polyp
climatic indicators. In Alpine lakes, for example, variations colonies that are normally long-lived, spanning hundreds
in stomatocyst assemblages have been found to co-vary to thousands of years. Together with the other elements
with the time of spring thaw (Kamenik & Schmidt, 2005), mentioned above, they produce massive calcareous
while in lakes in southwest Greenland, they show a strong structures close to mean sea level. Growth rates of about
relationship with longitudinal temperature gradients (Pla 10–30 mm yr–1 are common. Gradual changes in sea
& Anderson, 2005). They also appear to be influenced level have enabled reefs to build vertically, maintaining
by seasonal changes in lake temperature (Pla-Rabes & their shallow-water habitat over time, and thus provide
Catalan, 2011), and constitute one of the few biological important records of sea-level variations (section 2.5).
proxies that provide a basis for reconstructing former As the reef grows, the growth rates and chemical com-
winter temperatures (De Jong & Kamenik, 2011). In position of coral are affected by the physical and chemical
addition they are proving to be a key proxy for monitor- condition of the surrounding seawater as well as by
ing the response of lake systems to recent global warming biological factors, and a record of variations in these
(e.g. Moos et al., 2009). influences is therefore contained within the coral structure.
The most important parameters governing rate of reef
growth appear to be water temperature and water depth.
4.12.2 Cladocera Since these vary seasonally, many species of coral produce
Cladocera are small invertebrate crustaceans, commonly seasonal growth bands with clear density differences that
called ‘water fleas’, the chitinous exoskeletons of which are are visible by X-radiography or fluorescence techniques
often abundant in lake sediments. Although they are an (Lough & Cooper, 2011). The rapid growth and annual
important component of freshwater ecosystems, they have banding of coral provide a basis for palaeoenvironmental
reconstruction at high temporal resolution, potentially as been used in studies of past agricultural activity, as in
detailed as records based upon tree rings (section 5.4.1). Norse Greenland during the early Middle Ages (Gauthier
Moreover, a number of chemical indices can be used in et al., 2010), and in the Pyrenees Mountains where fluc-
conjunction with ‘coral stratigraphy’ as a basis for palaeo- tuations in grazing pressure over the last two millennia
environmental reconstructions. The most widely employed were inferred from coprophilous fungal assemblages
are stable isotopic composition (mainly oxygen and carbon) (Cugny et al., 2010). Others are associated with the dung
and trace element composition, such as barium/calcium of wild animals. For example, ascospores of three types of
(Ba/Ca) and strontium/calcium (Sr/Ca) ratios. While a coprophilous fungi, Sordaria, Sporormiella and Podospora,
number of environmental variables can affect these ratios, are often found together in Quaternary records, where
the principal controlling factor appears to be variations in they are indicative of the dung of large herbivores, such as
SST, which can be measured on seasonal to millennial mammoth (Van Geel et al., 2007). Their importance
timescales (Corrège, 2006). A global compilation of coral in palaeoenvironmental reconstruction is reflected in the
δ18O records shows temporal trends dominated by El Niño work of Gill et al. (2009), who found a marked decline in
Southern Oscillation (ENSO) variability (section 7.6.4.2), influx of Sporormiella spores in sites in North America
in terms of both long-term freshening of the oceans and around 14.8–11.3 ka, and which provides further evidence
decadal- to seasonal-scale SST fluctuations (Grottoli & of the late Pleistocene megafaunal collapse (section
Eakin, 2007). Coral stratigraphic records also provide the 4.11.6.3). Finally, the unique insight that fungal spores can
long-term historical context for assessing the scale of provide into Quaternary palaeoecology is demonstrated
ecological problems affecting modern reefs, such as coral by the contents of the lower intestine of a preserved
‘bleaching’ (Baker et al., 2008a). mammoth carcass from Siberia, which contained fruit-
bodies of coprophilous fungi that require at least one week
of exposure to air for germination, suggesting that the last
4.12.4 Fungal remains meal of this particular mammoth must have included some
Fungal remains, especially hyphae and fruiting structures dung (Van Geel et al., 2008)!
(which resemble spores), are very common in Quaternary
deposits, and are often encountered during routine pollen
analysis. They are especially common in lake sediments
4.12.5 Testate amoebae
and peats. The fossil components are difficult to classify, As we saw in Chapter 3 (section 3.9.4.1), testate amoebae
even to genus level, although systematic recording and are a large and diverse group of Protozoa, classified together
classification of the most common types found in late with Foraminifera in the taxonomic superclass Rhizo-
Quaternary deposits have led to the identification of key poda. Unlike forams, however, the amoebae are found
‘marker’ fossils (e.g. Van Geel & Aptroot, 2006). Fungal predominantly in lakes and ponds, rivers, mire deposits
remains appear to offer considerable potential for Quatern- and wet soils, though a few inhabit estuaries. Some amoebae
ary palaeoenvironmental reconstruction. For example, the have thin shells, or tests, that they build through secre-
identification in a sediment profile of saprophytic fungi tion or the agglutination of organic or mineral particles
known to invade peats and organic lake sediments during to protect the cytoplasm (section 4.9), and it is the shells
phases of desiccation could provide evidence of previous of these testate amoebae (occasionally referred to as
drier episodes (Yeloff et al., 2007). Episodes of catchment ‘testaceans’ or ‘thecamoebians’, or imprecisely as ‘Rhizo-
erosion may be reflected in lake sediment records by high pods’) that are preserved in sediments (Charman et al.,
numbers of spores of fungi associated with disturbed 2000). They are morphologically distinctive, permitting
ground (Kramer et al., 2010). Phases of natural or human- identification to species level.
induced burning can be inferred from the presence of Testate amoebae have proved to be valuable palaeo-
pyrophilous fungi, species that appear after plant ecological indicators in a range of palaeoenvironmental
communities have been destroyed by fire (Riera et al., contexts (Charman, 2001). They are particularly sensitive
2006). Parasitic fungi associated with crop plants, for to variations in water table depth, a relationship that has
example those that cause rust disease, have been used to again been quantified using transfer functions. Applying
reconstruct past farming practices of indigenous peoples these to data from ombrotrophic peat bogs, for example,
and European settlers in Canada (McAndrews & Turton, shows that the western Great Lakes region in North
2010). The most widely employed, however, have been America experienced severe drought between 1,000 and 700
coprophilous (dung-loving) fungi, some of which are years ago (Booth et al., 2006). Testate amoebae have also
indicative of the dung of pastoral animals. These have been employed in the interpretation of archaeological
records. In bog sequences adjacent to archaeological sites sediment sequences allows the former extent of sea-ice
in County Tipperary, Ireland, for example, records of cover to be reconstructed (Stein & Fahl, 2013).
testate amoebae indicate episodes of low water table level Palaeoenvironmental biomarkers have also been found
that accord with independent archaeological evidence in terrestrial and freshwater sediments. A peat sequence
for increased anthropogenic activity (Gearey & Caseldine, in China, for example, showed that lipids derived from
2006). Other applications include the reconstruction of sea- subaerial plants could be distinguished from those of
level variations (Charman et al., 2010) and of palaeo- waterlogged plants, the ratio between the two providing
limnological properties of Arctic lakes, such as pH, oxygen an index of effective precipitation over the past 16 ka
levels and degree of eutrophication (Patterson & Kumar, (Zhou et al., 2010). Romero-Viana et al. (2012) have
2002). applied a similar approach in order to reconstruct past
rainfall anomalies from lake sediment records, using a
measure termed the DiTe index which is derived from the
4.12.6 Biomarkers (ancient biomolecules) ratio of two molecular compounds. Analysis of the
The decay of organic tissue leads ultimately to chemical concentration of lipids derived from Archaea (single-celled
breakdown and the release of organic molecules, some microbes that are abundant in soils and peat) provides a
of which have chemical structures that are species-specific basis for estimating past changes in methane production
and are frequently preserved in sedimentary deposits. and cycling (Pancost et al., 2011). Some biomarkers also
Because they provide useful palaeoecological and palaeo- have archaeological applications, for example organic
environmental information, they can be considered as molecules derived from human and animal faeces, which
chemical fossils, though they are more commonly referred can indicate the onset of human settlement (D’Anjou et al.,
to as chemical biomarkers (Eglinton & Calvin, 1967). An 2012).
example is vestigial DNA, which is ubiquitous in the
environment. In addition to its role in phylogeographical 4.13 MULTI-PROXY PALAEO-
and evolutionary studies (section 4.11.6.4), DNA analysis
is important in taxonomy, for it enables taxa to be identified
ECOLOGICAL STUDIES
to subspecies level, and can thus be a more powerful and For many years, Quaternary palaeoecological research
reliable classificatory mechanism than one based on the projects generally involved the analysis of only one or two
morphological characteristics of organisms alone (Hofreiter lines of biological evidence. This is a reflection both of the
et al., 2012). In some deposits lacking visible fossil material, time-consuming nature of palaeoecological research, and
a diverse range of DNA molecules may have been preserved, of the particular specialisms of most palaeoecologists.
reflecting the mix of plants and animals that once inhabited Increasingly, however, research in Quaternary science
the vicinity of the site, including some that are rarely has followed a multi-proxy, as opposed to a single-proxy,
preserved as fossils (Willerslev et al., 2007). Further details approach, principally because more sophisticated environ-
can be found in Willerslev & Cooper (2005) and Hofreiter mental reconstructions are possible when several different
et al. (2012). proxies provide converging and mutually supporting
A number of other organic molecular proxies have data. A much wider range of biological evidence is now
applications in Quaternary palaeoenvironmental research, being investigated, and a great deal more is known about
particularly in palaeoceanography (Rosell-Melé & McCly- the ecological affinities of modern biota and, in particular,
mont, 2007). The majority are derived from long-chain about their environmental controls. In an increasing num-
alkenones and lipids, such as the alkenone-based U37 k
index ber of cases, therefore, it is now possible to model envir-
that co-varies with sea-surface temperature (section 4.10.7). onmental change using these different lines of evidence,
Another marine palaeotemperature biomarker that is but the models are, of course, much more powerful when
widely employed is the lipid-based TEX86 index (TetraEther they combine evidence from more than one climatic
indeX), a group of tetraether compounds consisting or environmental proxy. As a consequence, Quaternary
of eighty-six carbon atoms, the composition of which research is now characterized by teams of scientists, each
co-varies with mean annual temperature of surface ocean perhaps with a different expertise, cooperating in collab-
water (Kim et al., 2008). A further recently discovered orative research programmes which are not only multi-
biomarker that indicates the presence of sea ice is the IP25 disciplinary but frequently interdisciplinary in nature.
index (Ice Proxy with 25 carbon atoms), a lipid-based Perhaps the first modern example of such collabora-
compound derived from diatoms during the Spring (Belt tion was the CLIMAP project in which marine micro-
& Müller, 2013). Detection of this compound in marine palaeontologists, marine geochemists and atmospheric
scientists came together to develop models of ocean– be correlated in both the spatial and temporal dimen-
atmosphere–biosphere–cryosphere interactions for the sions. These form the subject matter of the following two
Last Glacial Maximum (section 4.10). Multi-proxy investi- chapters.
gations such as CLIMAP are essential if the multi-faceted
nature of the Quaternary record is to be properly inter-
preted and reasonable inferences drawn about patterns,
NOTES
processes and causes of environmental change. The validity 1 Used literally, the term ‘pollen analysis’ refers to the description
of the inferences that we make depends, however, on and classification of pollen grains, whereas environmental
establishing the past distribution of biota (the palaeo- reconstruction is based upon interpretation of pollen
biogeography of the period under investigation) as assemblages and their stratigraphical variations. However, the
term is so widely used that it is retained here, though strictly the
accurately as possible, and also on improving still further
term ‘pollen stratigraphy’ is perhaps more appropriate.
our understanding of the palaeoecology of former organ- 2 Plant function types (PFTs) is a system employed in vegetational
isms whose remains make up the fossil record. Just as modelling by climatologists and ecologists to classify plants
the interpretation of modern distributions of individual according to their physical, phylogenic (evolutionary relation-
species requires a knowledge of their mutual associations, ships) and phenological (life cycle) characteristics.
so must be the case for palaeoecosystems. This, however, 3 Plant biomes are communities of plants occupying a major
is a two-way process: studying the past is just as important geographical area and usually related to climate that are
to contemporary ecology as modern ecology is to under- characterized by similarity of vegetation structure rather than
standing past conditions, a point that we have been stressing by similarity of species composition.
throughout this chapter. 4 Transfer functions are essentially variants on multiple linear
regression models. In palaeoecological studies, they have been
employed to establish quantitative relationships between
biological data and environmental variables. If it is assumed that
4.14 CONCLUSIONS an assemblage of organisms is related to the environment by
Palaeobiological evidence, in the form of fossil fauna and some complex function (the transfer function), and if these
relationships can be determined for modern situations, then
flora, is probably the most effective and direct means we
multivariate numerical analysis should allow that function to be
have at our disposal for reconstructing past environmental applied to the fossil assemblages, thereby enabling former
conditions. The analysis of all forms of biological evidence, environmental parameters to be reconstructed quantitatively
however, is time-consuming, often costly and requires a (Birks & Birks, 1980).
very high level of expertise. These factors must be weighed 5 Lusitanian, a term used to denote warm-adapted fauna and flora
against the type of information required, the level of in or adjacent to the Northeast Atlantic, is named after the
sophistication in the data that are being sought and the province (Lusitania) established in Portugal by the Romans.
importance of the research topic to which they are being 6 The North Atlantic Polar Front is a prominent hydrographic/
applied. No single technique can provide all of the evidence oceanographic boundary, also termed the Subarctic Conver-
that we need in order to understand fully the nature of gence, which separates warm water of high salinity flowing
northwards from cold low-salinity water flowing from the
Quaternary environmental changes. Each data source
Arctic. Weather patterns are usually ‘trained’ along the southern
outlined in this chapter offers a slightly different perspec- flank of this boundary, which is informally termed the
tive, and the point has been made repeatedly that the most atmospheric ‘Polar Front’ in the literature.
fruitful lines of enquiry are frequently those in which 7 The term ‘duplicature’ in ostracod anatomy refers to calcified
several techniques are employed in conjunction, or where inner lamellae that extend along the free margin of the two valves
biological evidence is supported by geomorphological, that encase the soft parts of the animal.
sedimentological or geochemical data. In these circum- 8 Prokaryotes differ from all other organisms principally because
stances, the tools are available to enable Quaternary the DNA within their cell is loosely organized and not bounded
scientists to attempt reconstructions not only of environ- by a membrane into a nucleus. They lack chromosomes.
ments that existed at specific times in the past, but also 9 Boundary conditions are the assumed or measured surface
conditions used to constrain global climate models, and include
of the history of environmental changes and of specific
such factors as sea-surface temperatures, albedo values,
responses of the biota over time. Before these steps can be incoming solar radiation and atmospheric transparency. For
taken, however, two further aspects of Quaternary research further explanation see section 7.2.
need to be examined, namely the establishment of a 10 Autolysis is the self-‘ingestion’ of organic matter caused by
timescale for environmental change and the means whereby organic chemical reactions triggered by the enzymes of which
Quaternary sequences at widely separated localities can it is partly composed.
5
CHAPTER FIVE
Dating
methods
5.1 INTRODUCTION 5.2 PRECISION AND ACCURACY IN

There are three broad categories of Quaternary dating
QUATERNARY DATING
methods. First, there are techniques that provide age Many of the dating techniques currently employed in
estimates, in other words techniques that enable the age of Quaternary research can be applied only to restricted spans
fossils, sediments or rocks to be established directly in of Quaternary time (Figure 5.1), and each method has its
years. These include radiometric methods, which are based own distinctive set of problems which lead to uncertainties
on the radioactive decay of certain unstable chemical in interpretation. In evaluating any age determination,
elements or related phenomena (such as damage to crystal particularly those obtained by radiometric means, it is
lattices caused by radiation), and incremental methods, important to make a distinction between precision and
which involve measurements of regular accumulations of accuracy. The former refers to the statistical uncertainty
sediment or biological materials through time. Second, that is attached to any physical or chemical measure-
there are techniques that establish age-equivalence based ment, while the latter relates to the degree of cor-
on contemporaneous horizons in separate and often respondence between true age and that obtained by the
quite different sedimentary sequences. Certain distinctive dating process (Figure 5.2). In other words, it refers to
stratigraphic marker horizons are regionally and, in some the extent of bias in an age determination. In consider-
cases, globally synchronous, and where these can be traced ing accuracy and precision, it is useful to think of the
laterally between sediment profiles, they represent common analogy of a watch. A precise watch that tells the time to
time-planes in the stratigraphic records. If the age of the the nearest second may actually be inaccurate by 10 min-
markers can be established in one locality by the application utes; conversely, an imprecise watch with no second
of any of the age-estimate methods (above), then equivalent hand may still be accurate and tell exactly the correct time.
horizons within other sequences can be indirectly dated Both accuracy and precision determine the reliability
by correlation. Third, there are relative age methods, tech- of dates, but establishing whether or not a date can be
niques that establish the relative order of age of fossils or regarded as reliable also requires a knowledge of other
stratigraphic units. The relative antiquity of geological factors such as contexts of deposition, taphonomy of
materials is most obvious where superposition1 can be fossil material and post-depositional diagenetic processes.
established but, under certain circumstances, the relative For example, precise measurements can be obtained
age of Quaternary landforms and sedimentary units can be from fossils containing contaminants, but if the con-
determined from the degree of degradation or alteration taminants are not recognized as such and corrected for,
resulting from the operation of chemical processes through then the inferred age will be inaccurate and therefore
time. Further details of these various approaches can be invalid.
found in Walker (2005).
268 DATING METHODS
3
3 5
0.38*exp(0.09T
a) Marin e oxyge n 3
isotop e stratotyp e A5
Ja'arril' o Ja'arril' o
5 Ja'arril' o
Palaeomaanetism Brunnas Ja'arril' o 2000 2200 2400 2600 2800 3000
Ag e (Ka)
0 200 400 600 300 1000 1200 1400 1600 1BC-0
Ag e <ka)
Araon-Araon
Potassium-Argon
Thorium/Uranium
Pi'otactiniu m/Uranium
Helium/Uranium
Fission track dating
Luminescence dating
Electron spin resonance
Losmogenic nuclides
Antarctic ice cores
Tephrochronology
Amino-acid ratios
Rock weathering
Chemical weathering ol bone
Obsidian hydration
fapeleothems
Pedogenesis
0.38*exp(0.09T
Thousand s of years befor e A. D. 2000 (ka b2k)

40 40 60 80 100 120
b) Greenlan d oxyge n -31
isotop e stratotyp e 36
-31 Horocen e
Eemisf T
36
Annual ice-layers (Greenland)
Dendrochronology
Varve chronology
Radiocarbon dating
Corals/molluscs
Lead-210, Caesium-137,
Silicon-32
Applicable range Potential extension of range Intermittent application
Figure 5.1 The ranges of various dating methods discussed in the text, divided into a) those that apply to the longer Quaternary
timescale and b) those generally restricted to the Holocene and last glacial cycle. Broken lines with short dashes indicate where
extensions to dating ranges may be possible with further methodological improvements; broken lines with larger dashes indicate
time ranges in which a method can presently only be applied intermittently.
5.3 RADIOMETRIC DATING methods has now been developed, some of which are
capable of dating the oldest rocks on earth (Macdougall,
TECHNIQUES 2008), but in this section only those techniques that are
Radiometric dating methods are based on the radioactive directly applicable to the Quaternary timescale are discussed.
properties of certain unstable isotopes which undergo
spontaneous changes in atomic organization in order to
achieve a more stable atomic form. A number of radio-
5.3.1 The nucleus and radioactivity
active elements, such as uranium, occur naturally and The nucleus of an atom contains positively charged par-
are commonly found in rocks, sediments and fossils. ticles called protons and particles with no electrical
Radioactive decay (atomic transformation) is time- charge known as neutrons. These are densely packed in the
dependent, and if the rate of decay is known, the age of the nucleus so that although the nucleus occupies only about
host rocks or fossils can be established. Rates of radioactive 10–14 of the volume of an atom, it contains nearly all of
transformations vary markedly: some elements decay in the mass. The other major type of particle contributing
days or even seconds, whereas others transform gradually to the structure of an atom is the electron. These are tiny
over millions of years. A range of radiometric dating particles of negative charge and negligible mass that spin
DATING METHODS 269
c Chemical elements are classified according to atomic

number (Z), which is the number of protons contained
in the nucleus. Hydrogen has an atomic number of 1,
oxygen 8 and uranium 92. The atomic mass number (A)
of an element is the number of protons plus neutrons;
b
that of hydrogen is 1 and of oxygen is 16. It is convention
a to give the numerical value of A as a superscript and Z
as a subscript on the left-hand side of the symbol for
a chemical element, for example, 238 92U (uranium-238).
The atomic mass number of elements can vary, since the
number of neutrons in the nucleus is not always constant.
As we saw in Chapter 1, elements with the same number
12 11 10 9 ka B P
of protons, but a different number of neutrons (e.g. 16O and
18
"True" A g e O; 12C and 14C) are known as isotopes. They have the same
1.i
chemical properties, since the number of electrons remains
a Precise but inaccurate
2ii constant for each element, but isotopes differ in mass.
b Accurate but imprecise
c A c c u r a t e a n d precise
Each isotope of an element is called a nuclide. The particles
Mean
that constitute the nucleus are bound together in a way that
Figure 5.2 Accuracy and precision with respect to age
is not fully understood, but if a nucleus contains too many
estimates derived for a sample with a true age of 10 ka BP. or too few neutrons, it becomes unstable and the repulsive
forces between the similarly charged particles overcome the
binding forces keeping them together. This results in
around the nucleus, and they vary in number for different spontaneous emission of particles or energy, which is the
chemical elements. At one time, it was believed that basis of radioactivity. Isotopes involved in such radioactive
electrons orbited the nucleus in shells (or orbitals), and processes are known as radioactive nuclides.
that in each of these orbits they had a certain energy level. Three types of energy emission occur during radioactive
It now appears, however, that this is too simplified a view, decay. Alpha (α) particles consist of two protons plus two
for modern physics has shown that it is not possible to neutrons and are the positively charged nuclei of helium
determine both the location and velocity of subatomic atoms. They collide with surrounding atoms and acquire
particles. Electrons are now considered to exist in volumes electrons to form helium gas. Nuclides that emit alpha
of space around the nucleus, even though their precise particles lose mass and positive charge. By this process, the
location cannot be established. These volumes are known atomic number changes, and thus one chemical element
as atomic orbitals, and the build-up of electrons within can be formed by the ‘decomposition’ of others. Beta (β)
these orbitals enables scientists to account for many of the particles are negatively charged electrons, and their
properties of elements, and forms the basis of modern emission does not alter mass, but changes atomic number.
chemistry. Gamma rays (γ) are powerful forms of radiation that occur
As atoms gain or lose electrons, they acquire a net during radioactive decay. They are not important in the
electrical charge, in which case they are referred to as ions. calculation of decay constants but do contribute to the
The charge can be positive or negative; a positively charged build-up of thermoluminescence properties in minerals
ion is known as a cation and a negatively charged ion as (section 5.3.6). Moreover, the cosmic rays that constantly
an anion. The degree of electrical charge is determined by bombard the earth’s upper atmosphere (section 5.3.2)
the number of protons minus the number of electrons consist largely of gamma rays.
and is termed the valence. For example, a chemical element The atom that undergoes atomic transformation is
with six protons and six electrons has a net electrical charge termed the parent nuclide (or ‘mother nuclide’) and the
of zero. If it loses two electrons, it develops a positive product is the daughter nuclide. This single-stage trans-
charge (valence = 2+) and becomes a cation; conversely, if formation is known as simple decay. Many radioactive
it gains two electrons it now has a negative charge (valence transformations, such as uranium-series (section 5.3.4),
= 2–) and becomes an anion. The process whereby electrons involve more complex pathways where the transformation
are removed (usually) or added (occasionally) to atoms is of the nuclide with the highest atomic number to a stable
known as ionization, and is an important element in nuclide involves the production of a number of inter-
radiation (see below). mediate unstable nuclides. This is known as chain decay
270 DATING METHODS
(see Figure 5.14). Intermediate nuclides involved in such measurements of the progressive disappearance of nuclides
chains are both the product of previous transformations during disintegration, while others (‘accumulation clocks’)
and the parents in subsequent radioactive decay, and such measure the increasing quantity of a particular nuclide
nuclides are termed supported. Unsupported decay through time. The principal radiometric methods employed
involves the transformation of a parent nuclide that is not, by Quaternary scientists are discussed in the following
in itself, the product of decay, or is separated from earlier sections.
nuclides in the chain as a result of physical, biogenic or
sedimentary processes (section 5.3.4.1).
Radioactive decay processes are governed by atomic
5.3.2 Radiocarbon dating
constants. The number of transformations per unit time is
proportional to the number of atoms present, and for each
decay scheme there is a decay constant (λ) which represents This was one of the earliest radiometric methods to be
the probability that an atom will decay in a given period of developed and, despite the fact that it is only applicable to
time. The transformation of an individual atom occurs a small proportion of Quaternary time (Figure 5.1), radio-
spontaneously and unpredictably, but where a large number carbon dating has perhaps been the most widely used of all
of atoms of a particular nuclide are considered, there is a the radiometric techniques. A useful overview can be found
predictable time rate at which overall disintegration in Bowman (1990), while the journal Radiocarbon, along
proceeds. The law of radioactive decay is given by: with its website (http://www.radiocarbon.org) and associ-
ated links, is a valuable source of information on recent
␦N
= ␭N developments in, and applications of, the method.
␦t The basic principles were formulated during the late
where N is the number of atoms, t is a time constant and 1940s by the American scientist Willard Libby who syn-
λ is the decay constant for that nuclide. For all nuclides, thesized evidence from radiochemistry and nuclear physics
the decay is exponential (see Figure 5.3), and is best to determine the effects of high-energy cosmic radiation
considered in terms of the half-life (t0.5). This is the period (the cosmic-ray flux) on the atmosphere (Libby, 1955).
of time required to reduce a given quantity of a parent Free neutrons resulting from nuclear reactions in the upper
nuclide to one half. For example, if 1 g of a parent nuclide atmosphere collide with other atoms and molecules, and
is left to decay, after t0.5 only 0.5 g of that parent will one effect is the displacement of protons from nitrogen
remain. It will then take the same period of time to reduce atoms to produce carbon atoms:
that 0.5 g to 0.25 g, and to reduce the 0.25 g to 0.125 g, and
7 neutrons 8 neutrons
so on. The relation between the half-life and the decay N+ neutron C+ proton
constant is given as: 7 protons 6 protons
log e 2 0.693 The carbon nucleus produced by this reaction, 14C, is a

Half-life (t 05. ) = =
␭ ␭ radioactive isotope of carbon which eventually decays to
form the stable element 14N:
The application of the principle of radioactivity to geo-
logical dating requires that certain fundamental condi- 8 neutrons 7 neutrons
C ——— N+
tions be met. If an event (such as the cooling of a magma, 6 protons 7 protons
the formation of salt precipitates, the death of an animal
and the burial of its bones, etc.) is associated with the Decay is by beta (β) transformation, i.e. the emission of β–
incorporation of a radioactive nuclide, then assuming particles.
14
that (a) none of the daughter nuclides are present in the C atoms are rapidly oxidized to carbon dioxide and,
initial stages, and that (b) no parent or daughter nuclides along with other molecules of carbon dioxide (12CO2),
are added to or lost from the materials to be dated (i.e. the become mixed throughout the atmosphere and absorbed
radioactive process has proceeded within a closed system), by the oceans and by living organisms. Thus 14C, which
then an estimate of the age of that event can be obtained, is continually being produced in the upper atmosphere,
providing that the ratio between parent and daughter becomes stored in various global reservoirs, the atmos-
nuclides can be established, and that the decay rate is phere, the biosphere and the hydrosphere. All living
known. All estimates of time derived by radioactive decay matter absorbs carbon dioxide during tissue building in
are termed radiometric clocks; some methods are based on a ratio that is broadly in equilibrium with atmospheric
DATING METHODS 271
carbon dioxide. As long as the organism is alive, carbon residual 14C activity in a sample: (1) radiometric dating
used to build new tissues will be in isotopic equilibrium or ‘beta decay’, which involves the detection and count-
with (i.e. will exist in similar isotopic ratios to) those in ing of β emissions from 14C atoms over a period of time
the contemporaneous atmosphere. Upon death, 14C within in order to determine the rate of emissions and hence
the organic tissues will continue to decay, but no replace- the activity of the sample; and (2) accelerator mass
ment takes place. Hence, if the rate of decay of 14C is spectrometry, which uses particle accelerators as mass
known, date of death can be calculated from the measured spectrometers to determine the isotope ratio of 14C relative
residual 14C activity. to that of the stable isotopes of carbon (13C and 12C), and
The activity of 14C in the atmosphere is approximately age is then determined by comparing this ratio with that
15 dpm/g (15 disintegrations per minute per gram), and of a standard of known 14C content. The development and
this activity is halved every 5,700 years or so (Figure 5.3). refinement of these very different approaches to 14C
The half-life of 14C was originally calculated at 5,568 ± 30 measurement are reviewed by Povinec et al. (2009).
years, but this was subsequently more accurately deter-
mined as 5,730 ± 40 years. However, because a large Radiometric dating
number of 14C dates were published prior to the measure- Two methods are employed in radiometric radiocarbon
ment of the new half-life, it has been convention to base laboratories to detect emissions of beta particles: gas
radiocarbon dates on the former half-life value, and 5,570 proportional counting and liquid scintillation counting. In
± 30 years remains the internationally agreed fixed constant the former, a suitable gas (usually carbon dioxide, ethylene
for all radiocarbon measurements. This avoids confusion or methane) is prepared from the carbon in the sample and
as dates calculated using the same half-life, irrespective of collected in a chamber, down the centre of which runs
value, are directly comparable. Similarly, all radiocarbon age a charged wire. This detects, and counts, pulses of current
estimates are expressed in years before present (BP), the that flow through the gas when it is ionized by radioactive
reference year being AD 1950. decay. The current is proportional to the energy of the β
particle (electron), and hence it is possible to discriminate
between decays from different radioactive elements in a
16
proportional counter. In liquid scintillation counting,
14
samples are first combusted to CO2, reacted with molten
*expT(0.09T
12 lithium metal to give lithium carbide, mixed with water

10 to release acetylene and finally polymerized to benzene. A
‘scintillator’ is then added, usually a phosphoric substance
(0.09
a
that emits pulses of light (photons) in response to radio-
0.38
0.38*exp
6
active disintegrations, and these can be counted by photo-
4 electrical means.
2 From the decay curve (Figure 5.3), it can be seen that
ii)i)
ii) 0
material approximately of 10 ka age will have an activity of
only 4 dpm/g, and older samples correspondingly lower
0 4 8 12 16 -20\ 2 4 : 2 8 32 36 40 44 48
i) +VS dpm g - ' j ; ii) -Vi dpm g-
1 values. The limit of practical counting using conventional
(3500 years)•' i (5200 years) methods is eight half-lives (about c.45 ka), for beyond
Age (ka) that age the curve becomes so flat and insensitive that it is
difficult to separate samples of different activities with any
Figure 5.3 Decay curve for radiocarbon. See sections 5.3.2.1
and 5.3.2.2 for further explanation. statistical certainty. However, greater ages have been
measured by the technique of isotopic enrichment where
the amount of 14C present in a sample is enhanced so that
the frequency of decay can be more accurately determined
5.3.2.2 Measurement of 14C activity
by gas or liquid scintillation counters. This method, which
In order to detect 14C activity in organic materials, takes advantage of isotopic fractionation (Chapter 3, note
extremely sensitive equipment is required. This is because 9) uses either thermal diffusion or photo-dissociation by
the natural occurrence of 14C is so small that for every means of a laser beam. With the former, finite ages in excess
one million million atoms of 12C in a living organism, there of 60 ka have been obtained (Grootes, 1978; Woillard &
is only one atom of 14C. Moreover, 14C is a low-energy β Mook, 1982). However, the technique has not been widely
particle emitter. Two approaches are used to measure the applied, for it requires relatively large samples of material
272 DATING METHODS
and long counting times and, by comparison with standard probability that the true age of the sample lies between 1,900
dating procedures, it is time-consuming and hence more and 2,100 years BP, or, for a 95 per cent probability (2σ),
costly. Another approach to extending the radiocarbon the age is within the range 1,800–2,200 years BP. Even with
timescale is to use large-volume, high-precision counters two standard deviations on the ‘date’, there is still a one in
in which very old samples (up to ten half-lives) can be twenty chance that the true age lies outside the range
measured, and finite ages in the range 50–60 ka have been 1,800–2,200 years BP.
obtained in this way (Behre & van der Plicht, 1992). Again, It is important to remember, however, that age is not
however, substantial quantities of sample material and the quantity that is being measured, but activity of the
extended counting times are needed. sample which, on the basis of a number of assumptions,
In the calculation of radiocarbon dates obtained by is interpreted as indicating ‘age’. The plus-or-minus refers
conventional methods, laboratories compare sample activ- to the uncertainty associated with determining activity, and
ities with a modern reference standard. The internationally this is why on older dates the plus value of the standard
accepted reference standard for all 14C dating is the modern deviation is often quoted as being larger than the minus
activity level of NBS oxalic acid held by the American value. This can be seen in Figure 5.3 by considering an
Bureau of Standards. Because of the relative scarcity of this activity of 1 dpm/g ± 0.5 dpm/g which, translated into ‘age’,
material, however, some laboratories now use a secondary gives a significantly higher plus than minus value. Because
standard, such as Australian National University (ANU) of the asymptotic decay curve, technically there is always a
sucrose, or a paper cellulose sample provided by the difference between the plus and minus value, but this is
International Atomic Energy Authority (IAEA). The reason regarded as insignificant for younger material.
why such standards are necessary is that there have been
variations in 14C production rates through time, and Accelerator mass spectrometry (AMS dating)
modern levels are artificially high (see below). Compara- Mass spectrometers are widely used in physics to detect
bility between laboratories and between samples of different atoms of specific elements based on differences in atomic
ages therefore requires reference to a standard. The time weights (section 3.10). Charged particles moving in a
in years (R) since the death of an organism can be calculated magnetic field will be deflected from a straight path by a
from the following equation: factor that is in proportion to atomic weight; the lighter
the particle, the greater will be the amount of deflection.
1 A Normal mass spectrometers cannot discriminate, however,
R= log e 0
␭ A between 14C and other elements of similar weight (e.g. 14N),
but if the particles are subjected to large voltage differences
where λ is the decay constant of 14C, A0 is the 14C activity so that they travel at very high speeds, even the very small
of the modern reference standard, and A is the measured number of 14C atoms in a sample can be detected. This is
14
C activity of the sample of unknown age. In arriving at the principle of accelerator mass spectrometry. The most
a measure of R, account has to be taken of a number commonly used system is a tandem accelerator (Figure 5.4),
of factors that affect the determination of the activity of a so called because there are two stages of acceleration (see
sample. These include sample volume or gas pressure, below). Samples are converted to graphite (although a
dilution ratio (especially for small samples), atmospheric number of laboratories use a CO2 source) and mounted on
pressure (which affects background radiation during a metal disc (Figure 5.5). Caesium ions are then fired at this
measurement), and loss of sample during counting (which ‘target’ and the negatively ionized carbon atoms (C–)
results in samples having to be reweighed at the end of produced are accelerated towards the positive terminal
counting and a correction applied). Because these cannot through the two focusing devices and injection magnet
always be accurately determined, however, some uncer- (Figure 5.6). Nitrogen does not form negative ions, and
tainty is always associated with the calculated age of a hence almost all of the 14N, which tends to mask the 14C
sample. One source of uncertainty that can be quantified signal in a conventional mass spectrometer, is eliminated
is the probable effects of the randomness of radioactive before it can reach the detector. During passage through
decay on the counting statistics. As a consequence, radio- the ‘stripper’, four electrons are lost from the C– ions and
carbon dates along with other radiometric age deter- they emerge with a triple positive charge (C3+). Other
minations are always reported as mean determinations molecules are also lost at this stage. Repulsion from the
with a plus or minus value of one standard deviation (1σ) positive terminal leads to a second acceleration of the
about the mean. A radiocarbon date of 2,000 ± 100 years carbon ions through focusing magnets where deflection
should be interpreted as indicating that there is a 68 per cent occurs according to mass, and the concentration of 14C and
DATING METHODS 273
Figure 5.4
The 5 MV National Electrostatics
Corporation Accelerator Mass
Spectrometer (AMS) at the Scottish
Universities Environmental Research
Centre, East Kilbride, UK. The
accelerator itself is on the left of the
photograph, while the target wheel or
ion source (Figure 5.5) is in the
screened area to the right (photograph
by Sheng Xu, Scottish Universities
Environmental Research Centre, East
Kilbride, UK).
Figure 5.5
The exposed sample target wheel (ion
source) of the AMS in Figure 5.4. The
wheel holds 134 graphite samples, but
not all of these will be of fossil
material as standards of known age
are interspersed at regular intervals
(photograph by Sheng Xu, Scottish
Universities Environmental Research
Centre, East Kilbride, UK).
of the stable carbon isotopes 13C and 12C can therefore be important to stress, however, that it is not the absolute
measured. number of 14C atoms that is being measured: the abundance
In order to obtain an age for the sample material, the of 14C atoms is so small that it would be extremely difficult
14
C/12C ratio measured for the sample is compared with to measure total amounts. Rather, it is the ratio between
those for the targets in the same set which are made up from 14
C and the stable isotopes that form the basis for deter-
a material of known 14C activity (usually oxalic acid – see mining age, and thus this approach is often referred to as
above). This gives a sample/modern ratio from which, isotope ratio mass spectrometry. The plus-or-minus value
after corrections, a radiocarbon age can be calculated. It is that accompanies all AMS dates reflects, as in conventional
274 DATING METHODS
2 M V tandem accelerator
focusing C 3 +
C" injection
magnet magnet
focusing -stripper
device focusing
devices
power
supply
caesium
ion dectector MM
ion C" ions
MM
ion
ion
MM
Faraday collectors
sample
Figure 5.6 Schematic diagram of the components of a tandem accelerator. For further explanation see text (after Bowman, 1990).
dating, statistical uncertainties associated with the precise The potential of AMS is therefore clearly considerable,
measurement of the 14C decay curve, difficulties in quan- and its versatility is shown by a range of applications which
tifying the natural background 14C, as well as random and include the dating of small plant macrofossils (Hillman
systematic errors that inevitably occur during the measure- et al., 2001), fossil pollen grains (Vandergroes and Prior,
ment process. These can arise, for example, from contam- 2003), charcoal fragments (Meyrick, 2003), fossil insect
ination by modern carbon of samples, targets or the ion remains (Walker et al., 2001) and microfauna from marine
beam itself. cores (Flower et al., 2009), as well as more unusual materials
such as fragments of cloth from the Shroud of Turin
Evaluation of counting techniques (Damon et al., 1989), pieces of old parchment (Donahue
There are two principal advantages of AMS dating over et al., 2002), carbon residues (from smelting) in iron
gas proportional or liquid scintillation counting. First, artefacts (Hüls et al., 2004) and in calcium oxalate skins
very small samples of material can be dated, with most AMS covering ancient rock engravings (Smith et al., 2009),
laboratories routinely counting samples containing 1 mg organic-rich coatings (from food) on ancient pottery (Stott
of organic carbon or less. This compares with a typical et al., 2001) and lime mortar from old buildings (Heine-
sample size of 5–10 g of organic carbon required by most meier et al., 2010).
radiometric dating laboratories. The second advantage of AMS dating does have certain disadvantages, how-
AMS is one of time. The actual determination can take ever. The capital cost of establishing an AMS facility is an
only a matter of hours whereas typical liquid scintillation order of magnitude greater than that involved in develop-
or gas proportional counters may be occupied for days. ing a conventional radiocarbon laboratory, and running
Hence an AMS laboratory could perform more than 1,000 costs are also commensurately higher. Hence AMS dates
analyses per year, far more than most decay counters can are still more costly than radiometric radiocarbon age
measure to a comparable precision. It was initially envisaged determinations. Initially, analytical precision was also a
that AMS would enable the radiocarbon timescale to be problem, with AMS laboratories unable to match the levels
significantly extended, and while it still has not equalled of precision obtained by radiometric dating where the
the isotope-enriched beta counting dates in excess of 70 ka error estimate is typically around 1 per cent; in other
(see above), it is possible that the dating of large samples, words, ± 50 years at 5.5 ka and 120 years at 12.0 ka. How-
possibly back to 80 ka, might become feasible in due course ever, over the last decade or so, counting statistics have
(Hedges, 2001). Indeed, recent developments involving rapidly improved, and many AMS facilities are now more
more stringent pretreatments (e.g. the ‘ABOX pretreat- or less comparable with conventional radiometric labora-
ment’, and filtered gelatin preparations) and innovations tories in terms of analytical precision. Neither of these can
in target preparation for AMS dating (‘stepped combus- compare with the high-precision radiometric laboratories
tion’) to remove contaminants have resulted in finite ages (such as Belfast, Groningen, Heidelberg, Pretoria and
for charcoal samples in the range 50–60 ka BP (Bird et al., Seattle) which can produce dates with a standard deviation
2003; Higham et al., 2004). of less than 20 years, and sometimes as low as ± 12 years,
DATING METHODS 275
although in order to achieve these levels of precision, large 11,000

amounts of sample material (up to 20 g carbon) and mark-
edly longer counting times are required.
5.3.2.3 Quality assurance in radiocarbon P

dating
P
yr Bp
Because different practices have been adopted in different 10.000
laboratories, the international radiocarbon community
has agreed to participate in regular inter-laboratory com-
P
parison programmes designed to ensure reliability of, and
consistency between, results from radiocarbon laboratories.
The most recent of these, VIRI (the fifth international
radiocarbon intercomparison) which was completed in
2008, involved the preparation and distribution of typical
9,000
sample materials (peat, bone, shell, etc.) to more than 12,500 12,000 11,500 11,000 10,500
sixty laboratories worldwide. In this exercise, many more cal. BP
AMS than radiometric laboratories participated but, on
average, the results showed little evidence of differences Figure 5.7 Radiocarbon years (y axis) versus calendar ages
for the period 9.0–11.0 ka BP (10.5–12.5 cal. BP). Note the
between laboratory types (Scott et al., 2010a, 2010b). These marked ‘radiocarbon plateaux’ at c. 10.4, 10.0 and 9.6 ka BP.
intercomparison programmes are an essential part of the Radiocarbon dates that fall on these plateaux will have calendar
quality assurance process in radiocarbon dating, for they age ranges of up to several hundred years (after Goslar et al.,
demonstrate to the user community that laboratories 2000).
worldwide are delivering results that are reliable and in
accordance with good practice. each plateau is considered to reflect an episode of reduced
Δ14C (Figure 5.7). On a longer timescale, comparisons
between AMS determinations of 14C and uranium isotopes
5.3.2.4 Sources of error in radiocarbon on carbonate materials such as coral and cave speleothem
dating (section 5.3.4) indicate that radiocarbon dates under-
estimate true age by as much as 3.5 ka at 20 ka BP, with even
Temporal variations in 14C production greater discrepancies beyond 25 ka BP (Hughen et al.,
A fundamental assumption in radiocarbon dating is that 2004a).
atmospheric 14C concentrations (expressed as the decay- The divergence between radiocarbon age determin-
corrected deviation from the standard pre-industrial atmos- ations and those obtained by other dating methods means
pheric 14C concentration, and indicated by Δ14C) have not that it is necessary to make a clear distinction between
varied significantly over time. We now know, however, that ‘radiocarbon years’ and ‘calendar years’ when discussing
this is not the case. Over forty years ago, scientists working Late Quaternary chronologies. The two timescales can be
on tree-ring chronologies found a clear discrepancy reconciled, however, by comparing radiocarbon dates with
between ages of wood obtained by dendrochronological those obtained from the same samples of material using
dating (section 5.4.1) and those based on radiocarbon, independent dating methods. In this way it is possible to
with the dendrochronological dates being older than the calibrate the radiocarbon timescale against calendar years
radiocarbon ages (Renfrew, 1973). These data showed, for (see section 5.3.2.6).
example, that wood with a radiocarbon age of c. 5.0 ka BP The causes of the long-term atmospheric variations in
corresponds approximately to 5.8 ka BP on the dendro- Δ14C remain to be established, but several factors seem
chronological timescale. In addition, comparisons between to be involved. Of particular significance are variations in
dendrochronological and radiocarbon dating series indi- the strength of the earth’s geomagnetic field and changes
cated that atmospheric 14C activity has fluctuated markedly in the intensity of solar activity (Stuiver et al., 1991).
throughout the Holocene, apparently in a quasi-periodic A reduction in magnetic field intensity could result in a
manner (Sonnett & Finney, 1990). The record of long-term weakening of the earth’s geomagnetic shield and the
14 resultant increase in the number of cosmic rays entering the
C variations shows pronounced ‘plateaux’ of constant
radiocarbon age, for example at 10.4, 10.0 and 9.6 ka BP, atmosphere would lead to an increase in 14C production.
276 DATING METHODS
Similarly, a reduction in the strength of the solar wind (the 1010 is 14C. In nature, however, a fractionation of this ratio
stream of charged particles emitted by the sun) would also commonly occurs. Photosynthesis, for example, results in
enable more cosmic rays to enter the atmosphere, which an enrichment of 12C relative to the other isotopes in most
again would result in an increase in 14C production (van plant tissues, whereas ocean waters preferentially absorb
Geel et al., 2003). A further factor may have been changes 14
C. These effects are small, but can significantly affect
in the carbon distribution in the ocean–atmosphere system, radiocarbon dates where measurement to less than ± 1 per
especially in the oceans, which constitute one of the major cent error is required. In addition, fractionation can also
global carbon reservoirs (Beck et al., 2001). During the last occur in the laboratory during the conversion of sample
glaciation, for example, lower atmospheric CO2 levels were carbon to the gas or liquid form. Most radiocarbon lab-
accompanied by increased atmospheric radiocarbon oratories today make corrections for the probable effect of
concentrations, reflecting greater storage of CO2 in deep fractionation based on thermodynamic laws which show
ocean basins. The end of the last cold stage was marked that the heavier isotope 14C is twice as enriched as 13C. The
by a rapid increase in atmospheric CO2 as this old, 14C- latter can be measured in a small subsample of the material
depleted CO2 was released back into the atmosphere (Rose to be dated. Originally, the 13C/12C ratio was compared with
et al., 2010). One consequence of this process is the plateaux a standard from carbonate of a marine fossil from the Pee
in radiocarbon ages at the onset of the Holocene described Dee Formation in South Carolina, USA (PDB belemnite),
above. Over longer timescales, palaeomagnetic records but this material has now been exhausted and has been
suggest a close relationship between changes in the earth’s replaced by a new limestone standard using marble, and
geomagnetic field and 14C production, with fewer radio- referred to as Vienna PDB or VPDB (Chapter 3, note 10).
carbon nuclides being produced during periods of high Values are published as deviations from this standard:
magnetic field intensity and more during low geomagnetic
13
field intensity phases (Muscheler et al., 2005). Data from C12Csample – 13C12Cstandard
␦ 13C‰ =
sediment cores in the Iceland Sea, for example, show that 13 12
C C standard
over the time interval from 27–54 ka, three intervals of
highly increased atmospheric Δ14C have been recorded Most terrestrial samples have a negative δ13C value
(c. 34 ka cal. BP; c. 38 ka cal. BP; c. 41 ka cal. BP) and these compared with the PDB standard. Different photosyn-
are coincident with low values of magnetic field intensity thetic pathways exist in plants so that very different levels
(Voelker et al., 2000). The palaeoenvironmental significance of fractionation occur (Table 5.1). However, it is practice
of long-term changes in 14C is considered further in section to ‘normalize’ 14C activities during calculations of radio-
5.3.2.6. carbon age, by treating each sample as if an average
In addition to natural variations, atmospheric 14C enrichment had taken place. The normal value is taken to
content has recently been affected by human activity. Over be –25 per mil, which is the mean isotopic composition
the past 200 years, 14C levels have been progressively diluted of wood. If the δ13C value of a sample is found to be
as a result of the combustion of fossil fuels, which has –25 per mil, then no adjustment is made. With a value
liberated large quantities of ‘inert’ 12C into the atmosphere of –30 per mil, however, a 5 per mil depletion in the
13
(Solomon et al., 2007). In the last forty-five years or so, C/12C ratio is implied, which in turn indicates a prob-
however, this industrial effect has to some extent been able 10 per mil depletion in the 14C/12C ratio. Thus, the
offset by greatly increased production of 14C resulting from measured 14C activity would be increased by 10 per mil,
the detonation of thermonuclear devices. The combined which is equivalent to about 83 years.
effects of industrial activity and atomic explosions mean
that modern organic samples are unsuitable as reference Circulation of marine carbon
samples for radiocarbon activity. A value of 0.95 times Because 14C is transferred from the atmosphere to the
the measured activity of the NBS standard is regarded as oceans only across the ocean surface, and because the
equivalent to the natural 14C activity of AD 1890 wood (pre- mixing rate of surface and deep waters is very slow, 14C
industrial effect), and this is corrected to AD 1950 which in deep ocean waters decays without replenishment. Sea-
is, as noted above, the reference year for all ages quoted in waters therefore have an apparent age (known as the
radiocarbon years BP. reservoir age). The ageing effect in ocean surface waters
ranges from c. 30 years or less in parts of the northern
Isotopic fractionation Indian Ocean (Dutta et al., 2001), to around 400 years in
Of the three naturally occurring isotopes of carbon about the North Atlantic (Bard et al., 1991) and over 700 years
98.9 per cent is 12C, 1.1 per cent is 13C and only l part in in parts of the northwestern Pacific (Kuzmin et al., 2001).
DATING METHODS 277
Table 5.1 Approximate δ13C values for various A further problem is that the apparent age (R(t) + δR)
materials. The ranges on these data are typically ± 2 or of present-day marine molluscs in a region may not
3 per mil, but substantially more variability is possible. represent an appropriate correction factor for earlier
With each per mil deviation from –25 per mil periods of the Quaternary. This is partly because of spatial
representing c. 16 years, these data clearly illustrate and temporal changes in patterns of ocean water move-
the need for fractionation corrections to be applied to ment, but may also reflect oceanographical responses to
measured 14C age results (from Bowman, 1990). variations in atmospheric 14C. Data from the high-latitude
Material δ13C value
North Atlantic, for example, indicate that during Green-
(‰) land Stadial 1 (GS-1)/Younger Dryas, the marine reservoir
age increased to 590 (±130) radiocarbon years (Cao et al.,
Wood, peat and many C3 plants –26
2007), while even greater values (1,000 ± 250 radiocarbon
Bone collagen* –19
years) have been suggested for the same time interval for
Freshwater plants (very variable) –16 the Norwegian Sea (Björck et al., 2003). These age estimates
Arid zone plants (C4 plants) –13 are significantly higher than those obtained from modern
Marine plants –15 marine molluscs from the North Atlantic region. Even
Atmospheric CO2 –8 during the late Holocene (post 4.5 ka), there is evidence
Marine carbonates –0 from the Icelandic Sea that marine reservoir ages varied
from tens to hundreds of radiocarbon years as a result of
* For direct or indirect C3 consumers.
changes in water mass distribution (Eiríksson et al., 2004).
Collectively, these data suggest that finding the appropriate
correction factor for marine radiocarbon dates may be more
In the deep oceans, however, longer residence time problematical than has hitherto been envisaged.
means that seawater may have an apparent age in excess of
2,000 years. Hence, radiocarbon dates on Foraminifera Contamination
from deep-ocean cores have to be corrected for the age of Contamination can occur in organic sediments because
seawater, with different correction factors being applied younger or older carbon has been added to the sample
to planktonic and benthic species. Similarly, corrections material. The former can arise from root penetration
have to be made to dates on marine molluscan fossils through a profile, infiltration by younger humic acids
and also to other samples with a marine component, such through older peat or soil horizons, and the downward
as human bone from coastal regions. At present, the global movement of younger sediments through bioturbation. The
average marine reservoir age of surface waters, R(t), deter- possible effects of contamination by younger carbon are
mined from measurements on modern marine organ- shown in Table 5.2a. Because of the high activity of modern
isms, is c. 400 radiocarbon years, but regional values deviate carbon in comparison to fossil material, relatively small
from this as a function of climate and ocean circulation amounts of contaminant can result in major errors in
systems (Ascough et al., 2005). These local deviations from radiocarbon dates.
R(t) are expressed as δR values and have been collated in Contamination by older carbon can also take a number
an online global database (http://www.calib.org; see also of forms. Inwash of older inorganic carbon residues
http://radiocarbon.1deo.columbia.edu/research/resage.htm). (graphite, coal, chalk, etc.) into lake basins leads to a
In general, polar waters tend to exhibit greater reservoir ages dilution of the local 14C/12C ratio and hence an ageing factor
(δR = +400 to +800 radiocarbon years) than equatorial (mineral carbon error) will affect radiocarbon dates from
waters (δR = c. 0 14C years), but there is considerable local such sediments. In formerly glaciated regions, such older
and/or regional spatial variation in the data. For example, carbon may have been brought into the area by the passage
empirical values for modern Mediterranean surface waters of glacier ice (Walker et al., 2001). In late Holocene sedi-
vary between 280 and 665 radiocarbon years, and are, ments, inwash of soils or sediments can arise as a con-
exceptionally, in excess of 1,000 radiocarbon years (Reimer sequence of anthropogenically induced erosion around
& Reimer, 2001). It is difficult, therefore, on present lake catchments (e.g. Oldfield et al., 2003). Sub-aquatic
evidence, to specify a δR that would be representative of the photosynthesis, water uptake in carbonate-rich ground-
Mediterranean region as a whole, and this appears to be the waters, and carbonate secretion by freshwater or offshore
case for other marine sectors as well. An online marine organisms may also be affected by diluted 14C levels. In these
reservoir correction database is now available and can be instances, the resulting age error is termed hard-water
accessed at the website http://www.calib.qub.ac.uk/marine. error, and may add up to 1,200 years to the apparent age
278 DATING METHODS
Table 5.2 The effect of contamination on true age of samples selected for radiocarbon dating, a) by modern
carbon, b) by inert carbon (after Harkness, 1975).
a) Measured age as a result of weight % contamination of sample by modern carbon
True age (years) 1% contamination 5% contamination 10% contamination
600 540 160 Modern
1,000 910 545 160
5,000 4,870 4,230 3,630
10,000 9,730 8,710 7,620
Infinitely old 36,600 24,000 18,400
b) Effect of contamination by inert carbon on true 14C age

Contamination (%) Years older than
true age
5 400
10 850
20 1,800
30 2,650
40 4,100
of limnic material. One way around this problem is to date Tibetan Plateau, reservoir ages of 2,010 and 3,260 radio-
inwashed terrestrial materials (such as leaves and fruits) as carbon years have been measured in samples of bulk organic
these are unlikely to have been influenced by hard-water matter (Wu et al., 2010).
contamination if the plants have been growing above the Problems of contamination are often encountered
level of the carbonate-rich water table. where shells are used for dating, since carbon exchange
Contamination of lake sediments may also arise from takes place more readily in carbonate structures. Contam-
the inwashing of older organic carbon detritus, and this ination may be by older or younger carbon, depending
redeposited or allochthonous carbon will also produce on the dissolved contaminants introduced, and can result
an ageing effect in dated samples. The extent to which from the gradual accumulation of particulates or solu-
contamination by older carbon can affect radiocarbon tional carbon in the interstices of the carbonate matrix,
dates is shown in Table 5.2b. In general, samples that are or from the recrystallization of the carbonate matrix and
initially rich in organic carbon will only be significantly an exchange of sample carbonate with contaminant carbon.
affected where the amount of contaminant is high. Many Exchange usually affects the outer part of a shell more than
organic sediments such as lake gyttjas, however, typically the inner layers, and it has therefore become practice to
consist of only 4–5 per cent (and sometimes as low as 1 per remove the external portion prior to dating. Procedures
cent) organic carbon, and therefore relatively small vary between laboratories, but up to 25 per cent of the outer
amounts of inert carbon incorporated into the sediments part is typically leached and discarded. The remainder is
can introduce significant errors into radiocarbon dates. then treated to produce an ‘outer’ and an ‘inner’ fraction
A further problem in the dating of limnic sediments which are dated separately, the inner date being preferred
arises because the 14C/12C ratio in lake waters may be lower where a noticeable difference between the two ages
than that in the atmosphere (the reservoir effect). This is occurs. The development of AMS dating has resolved
partly a consequence of the fact that the exchange rate at some of the difficulties that have arisen with the dating
the lake surface is relatively slow and hence lake waters of shell, however, for age determinations can now be
may have a somewhat lower activity than the atmosphere, made on very small samples of inner shell material.
but it may also reflect seepage into lakes of groundwater Moreover, radiocarbon dates on Mollusca can now be
containing dissolved carbonates. Some lakes in central and obtained from hitherto unpromising contexts, such as
eastern Europe show variations in reservoir age of almost marine boreholes, where only small amounts of shell
1,000 years (Geyh et al., 1998), while in two lakes on the carbonate are available.
DATING METHODS 279
In the case of bone material, carbon exchange after death tion. However, buried soils are important components of
is always likely to have occurred, and hence although in the Quaternary stratigraphic record (section 3.5) and a
theory both elements of bone (calcium hydroxyapatite and considerable amount of effort has been directed towards
collagen) can be dated, it is usually only the proteinaceous obtaining coherent radiocarbon dates from such contexts
fraction which is used for 14C measurements, although (Scharpenseel & Becker-Heidemann, 1992). Meaningful
careful sample selection and biochemical purification is dates have been obtained from soil profiles by dating
needed in order to obtain reliable dates (van der Plicht and comparing different organic components, such as soil
et al., 2004; Higham et al., 2006). With the development of organic matter, the humin (acid-insoluble) fraction and
AMS dating, however, it has proved possible to obtain age charcoal (Pessenda et al., 2001), by modelling the measured
14
determinations on individual amino acids (section 5.6.1) C content of soil organic matter (Wang et al., 1996), or
within the collagen to check for consistency in dates and by verifying radiocarbon dates from soils against results
hence for the presence of contaminants (McCullagh et al., obtained from other dating methods (Dalsgaard &
2010). Odgaard, 2001).
Biogeochemistry of lake sediments 5.3.2.6 Calibration of the radiocarbon

AMS dating has enabled radiocarbon ages to be obtained timescale
on different biogeochemical components of lake sediments
(humic acid fractions, lipids, chlorite treatment residues, Bases for calibration
HF/HCl treatment residues, etc.). Where this type of dating Long-term variations in atmospheric Δ14C which were first
has been carried out, significant age differences have been detected over forty years ago through comparisons between
found within a single sediment sample and between samples radiocarbon and tree-ring chronologies have resulted in
from contemporaneous horizons (Lowe et al., 1988). More- an increasing divergence back in time between radiocarbon
over, marked discrepancies have also been recorded in 14C and calendrical ages (section 5.3.2.4). Over the course of
activity between macrofossil cellulose and the sediments the past thirty years, and particularly following the advent
from which the plant macrofossils have been obtained, with of AMS and the radiocarbon dating of small samples of
macrofossils typically providing younger radiocarbon ages material, the international radiocarbon community has
than the sediments within which they were contained devoted a considerable amount of effort to the develop-
(Turney et al., 2000; Walker et al., 2003). Clearly, more ment of calibration programmes that enable radiocarbon
needs to be learned about the biogeochemistry of lake years to be converted to ‘actual’ or ‘calendar’ years (Kromer,
sediments, if reliable age estimates are to be obtained from 2009). A major focus has been dendrochronological
these media by means of radiocarbon dating. records, as these enable a direct comparison to be made
between the ages of wood as determined by counting
annual tree rings (section 5.4.1) and radiocarbon dates
5.3.2.5 Radiocarbon dating of soils
obtained from individual wood increments. High-precision
One of the most difficult materials to date by radiocarbon radiocarbon measurements on bristlecone pine from the
is soil. All soils contain both organic and inorganic carbon, western United States, on Irish oak and, for the earliest
but soils are dynamic systems and receive organic matter part of the sequence, on German oak, have generated a
over long time periods. Any radiocarbon date on a soil continuous dendrochronological calibration curve extend-
will therefore be heavily influenced by the mean residence ing back over 10,000 years (Stuiver et al., 1998). By linking
time of the various organic fractions in the soil. The this European oak-based record to tree-ring series from
continual addition of organic matter throughout its devel- subfossil pines, the tree-ring calibration record has been
opment means that the measured radiocarbon age of a soil extended to 12,594 cal. (calibrated) years BP (Hua et al.,
is normally younger than its true age (Pessenda et al., 2009). Beyond the limit of continuous dendrochrono-
2001). When a soil is buried, addition of organic matter logical series, calibrations have been based on matched
ceases and a radiocarbon age will reflect both the mean uranium-series (U-series: section 5.3.4) and radiocarbon
residence time and the time that has elapsed since burial. dates on fossil corals (Fairbanks et al., 2005), radiocarbon-
The date at which pedogenesis commenced, which is of dated organic material from laminated (i.e. annually
primary interest to the stratigrapher, will be almost im- accumulating) marine sediments, and synchronized palaeo-
possible to establish. Further complications are caused by climatic signatures in ice cores and U-series dated cave
the constant recycling that takes place within the soil speleothems (Hughen et al., 2006; Weninger & Jöris, 2008).
profile, notably by humic acid filtration and root penetra- A number of radiocarbon calibration models have been
280 DATING METHODS
generated for the period 0–50 ka (Figure 5.8), and some of foraminiferal record from the Cariaco Basin in Venezuela,
these are considered below. where parts of the sediment sequence are also laminated
(section 5.4.2.7). The radiometric ages provide the tie-
IntCal: Probably the most widely employed calibration points for matching (or tuning) the different records
model has been developed by an international Working (e.g. Figure 5.9). The temporal trends in the calibration
Group (IntCal) of specialists in radiocarbon dating and datasets were statistically analysed and averaged using a
in other areas of Quaternary geochronology. Their remit ‘random walk’ procedure (Buck & Blackwell, 2004). This
is to produce the most coherent calibration model, and to generates and tests various iterations within the data, and
make this available to the wider user community. The selects the most coherent results that are then statistically
latest iteration at the time of writing is IntCal13 which is smoothed. The calibration curve confirms a progressive
based on radiocarbon-dated tree-ring samples back to divergence between radiocarbon and calendar time (Figure
13.9 ka BP, and then extended to 50 ka BP using 14C 5.8), but also shows fluctuations (‘radiocarbon wiggles’)
and independently dated samples from both terrestrial which reflect both long-term variations in atmospheric 14C
and marine contexts (Reimer et al., 2013). These include production and statistical noise in the radiocarbon dataset.
14
C-dated plant macrofossils from the varved sequence in
Lake Suigetsu, Japan (section 5.4.2.7), 14C measurements Calibration procedures and effects: Calibrating radio-
from U-series-dated speleothem samples from the Bahamas carbon to calendar time, which takes account of the
and China, paired 14C and U-series dates from pristine statistical errors in both the original radiocarbon measure-
corals and planktonic foraminifera from a number of ment and in the relevant part of the calibration curve, can
localities in the Atlantic and Pacific Oceans, and a 14C-dated be achieved by the use of online programs such as CALIB
Tree Ring
Lake Suigetsu
50000 Bahamas Speleothem
Cariaco Basin
PS2644
Lake Lisan
Papua New Guinea
40000
30000
yr Bp
1:1 Line
20000
10000
0 10000 20000 30000 40000 50000

cal. yr B P
Figure 5.8 Radiocarbon calibration dating series for the period 0–50 ka; some of the IntCal datasets are shown by the light green
(dendro) and grey (Cariaco) records (after Fairbanks et al., 2005).
DATING METHODS 281
2800
Cariaco-Hulu Age/Depth Model Radiocarbon-calendar
30 Cariaco-GISP2 AqeyDepth Model calibration curve
Mod el
2700
Hulu Cave ° T h Dates
2 3
25
Caria co-G ISP2 AqeyDepth
2600
20
2500-
yr Bp
RD1
15 2400-
RD1
10 2400-
5 2200
10000 20000 30000 40000 50000 60000 70000

cal. yr BP 2100
2200 2300 2400 2500 2600 2700 2800 2900
Figure 5.9 Age–depth plot of Cariaco Basin sediments after cal. yr B P
being tuned (matched) to the Hulu Cave record (Wang et al.,
2001). The close convergence between the two datasets Figure 5.10 The influence of the shape of the radiocarbon
suggests that both are reflecting past variations in Δ14C (after curve on calibration. Radiocarbon date RD1 intercepts a steep
Hughen et al., 2006). part of the curve, while RD2 intercepts a flatter part of the
curve. The consequence for calibrating radiocarbon dates are
discussed in the text.
7.0 (IntCal: http://calib.qub.ac.uk/calib/; Stuiver et al.,
2013) and OxCal (Oxford Radiocarbon Accelerator Unit:
http://c14.arch.ox.ac.uk/). It is important to distinguish linear; calibration is essential if the true order and temporal
between calibration datasets (the raw data), such as precision of a series of radiocarbon dates is to be
IntCal13, and the programs developed to interrogate determined.
the datasets (i.e. to perform the calibration calculations)
such as CALIB and OxCal, as they are often confused. Once Other calibration models: In addition to IntCal, a number
the user has selected the calibration model and conver- of other curves for calibration are now available. These
sion software of preference, the radiocarbon data can be include the Fairbanks (2005) calibration model (http://
input and the output generated fairly quickly, although radiocarbon.ldeo.columbia.edu/research/radiocarbon.htm),
interpretation of the results is frequently far from straight- which is based on high-precision paired U-series and
forward. In Figure 5.10, for example, radiocarbon date radiocarbon dates on pristine coral samples, and the CalPal
RD1 intersects a steep part of the calibration curve, and (Cologne Radiocarbon Calibration and Palaeoclimate
a calibration with a small error range is obtained; indeed, Research Package) calibration which synchronizes the
in this instance, calibration leads to an apparent reduc- palaeoclimatic signature of several radiocarbon dating
tion in age uncertainty. The opposite is the case with radio- series with the uranium-series dated Hulu cave speleothem
carbon date RD2, where the age estimate intersects a sequence and the Greenland ice-core record (Weninger &
part of the calibration curve that is flatter and also uneven. Jöris, 2008; http://www.calpal.de/). The differences (some-
Here, calibration leads to an increased uncertainty in times marked) between the various datasets partly reflect
age. Furthermore, due to the vagaries of the calibration the limited analytical precision with which samples of low
curve, the chronological order of samples measured in residual 14C activity can be measured, but they also arise
radiocarbon time can, in some instances, be reversed after from both systematic and stochastic errors that affect each
calibration (see Bowman, 1990). An additional difficulty is dataset. For example, marine records may be affected by
that whereas radiocarbon dates have Gaussian (normally reservoir errors that are not yet sufficiently constrained
distributed) errors, the calibrated equivalents are statistically (section 5.3.2.4); varved records may be inaccurate because
more complex and may be multi-modal, which makes of missing or indistinct annual layers (section 5.4.2.6);
interpretation of age even more complicated (Figure 5.11). speleothems often contain old carbon residues from soil
Site chronologies should no longer be expressed in radio- (dead carbon fraction: DCF) which results in aberrantly
carbon time, therefore, since radiocarbon time is not older radiocarbon dates; other carbonate materials, such
282 DATING METHODS
precision radiocarbon-dated tree-ring series are available,

R_Date (3000,30}
3200 95.4% probability
for example, the curve derived by plotting dendro-age
1375BC (8.2%) 1340BC against radiocarbon age should show undulations that
3100 1320BC (87 2%) 1129BC correspond to those in the calibration curve; in other
3000 words, the ‘wiggles’ should match (Figure 5.12). This
yr Bp
approach has also been applied to the dating of peat

2900
sequences, although here only the radiocarbon age of
2800 individual horizons is known, and the independent (i.e.
non-radiocarbon-based) timescale has to be derived using
2700 assumptions about the rate of peat accumulation (Blaauw
2600
et al., 2003, 2004). The wiggle-matching approach is best
used over a short section of the calibration curve where, in
1400 1300 1200 1100 1000
cal. yr B P
some cases, it may prove possible to match the dated
section to within a few years on the calendar timescale
Figure 5.11 The conversion of a Gaussian radiocarbon error (Hogg et al., 2003).
range (1σ: at left-hand edge) to a non-Gaussian, multi-modal A more sophisticated way of achieving optimal matches
error distribution in calibrated time (lower curve), using IntCal04. between a radiocarbon dataset and a calibration curve
The 1σ error range of c. 200 radiocarbon years is spread over
nearly 300 cal. years but there is no single dominant mode. The involves the use of statistical methods, particularly
probability distribution of the radiocarbon error is proportionally Bayesian-based statistical modelling. Bayesian probability
distributed according to the pattern of intercept with the differs from classical statistics in that prior knowledge
calibration curve. Hence, those parts of the Gaussian curve and assumptions are expressed explicitly in the calcula-
close to the mean value (transparent shading) are allocated a
tion of probability, allowing such prior information to be
commensurately higher probability intercept value than those
nearer to the tails of the radiocarbon error range (figure obtained incorporated into the radiocarbon calibration process.
from the University of Oxford Radiocarbon Web-Info site at This approach can be most usefully applied in the gen-
http://www.c14.arch.ox.ac.uk). eration of age–depth models for sedimentary sequences,
especially where there is a degree of scatter in the original
radiocarbon dates and inversions in age–depth relationships
as aragonite and mollusc shells, are generally unreliable (Walker et al., 2003; Blockley et al., 2004). In such cases,
media for radiocarbon dating due to fractionation and it is difficult to judge the coherency of the dataset and the
other chemical effects; while freshwater lake sediments degree to which individual values represent statistical
may include in-washed secondary organic materials. Each ‘outliers’ merely by visualizing the plotted data, particularly
dataset, therefore, has limitations and, while the IntCal because the probability densities of the calibrated dates are
calibration is perhaps the most widely used, there is, as yet, not normally distributed (Figure 5.11). Bayesian-based
no consensus on which calibration curve is the most age-modelling approaches have therefore been developed
reliable. to deal with this problem and are now widely employed
in Quaternary stratigraphic and archaeological studies
Age modelling based on calibrated radiocarbon data (e.g. Wohlfarth et al., 2006; Dee et al., 2013); the relevant
So far we have addressed the calibration of individual software is available online through the OxCal web site (see
radiocarbon dates and the resulting output; a more above). The models use a Bayesian form of Markov Chain
complicated task, however, is the calibration of a whole Monte Carlo analysis and a prior rule that age must increase
series of radiocarbon dates from a single section and the with sediment depth (Bronk Ramsey, 2008). The method
construction of calibrated age–depth models. Two main tests sequentially through all possible combinations for the
approaches have evolved, both of which attempt to find highest probability match between the new site chronology
the optimal match of a series of radiocarbon dates to and the radiocarbon calibration model. The probability
a calibration model where the stratigraphic order of the distributions of individual radiocarbon age measure-
dated samples is known: visual matching to the radiocarbon ments are treated independently in the analysis, so that
calibration curve, and statistical matching. The first, termed the outcome is not based on knowledge of the other radio-
radiocarbon wiggle-matching, takes advantage of the short- carbon dates in the set, or on a pre-judgement of the order
term fluctuations in the radiocarbon calibration curve, in which the dates should plot; the data only need to satisfy
and which should also be evident in a dated sequence, the prior rule. The appeal of this approach is that it provides
providing there are sufficient data-points. Where high- an objective test that is independent of stratigraphic
DATING METHODS 283
a) 4000 b) 4000
middle
subset
3500 3500
ISPDept
ISPDept
lower
suDset
Aqey
Aqey
Caria co-G
Caria co-G
3000 3000
upper
subset
2500 2500
;>nnrt 2000
2500 2000 1500 • M l 500 o 2500 2000 1500 1000 500 0
Calendar age (BC) Calendar age (BC)
Figure 5.12 Wiggle-match dating of a peat sequence in the Netherlands. a) The individual radiocarbon ages (shown with error
bars) have been wiggle-matched to the radiocarbon calibration curve based on the assumption of a constant rate of accumulation
for the peat, with the younger part of the series showing the closest pattern match. b) The most probable calendar-year conversions
for all dated horizons in the peat profile are shown, if the constant growth rate assumption is relaxed (after Blaauw et al., 2004).
assumptions, it employs the full radiocarbon inventory from the fossil kauri (Agathis australis) trees preserved in
available for each sequence (no data are excluded until peats in North Island, New Zealand (section 5.4.1.3), as
analysis is complete) and it makes any data selection explicit these have been growing since at least MIS 5 (D’Costa
(Blockley et al., 2007a). et al., 2009), and radiocarbon dates back to 60 ka have
been obtained from wood samples (Hogg et al., 2006). An
Extending and refining radiocarbon calibration alternative approach to the provision of an extended
models radiocarbon calibration is to use annually laminated
The most secure bases for radiocarbon calibration are (varved) sediments in long lake sequences (section 5.4.2.7).
radiocarbon-dated tree-ring chronologies that extend back The most promising site in this regard is currently Lake
continuously from the present day. Further refinement of Suigetsu, Japan (Nakagawa et al., 2012), where a continuous
radiocarbon calibration would therefore be best achieved varved sequence extends back over at least the last 70 ka,
by extending the dendrochronological calibration beyond and from which a comprehensive radiocarbon inventory
the current limit of 12,556 cal. BP (Reimer et al., 2013). has been obtained that spans the whole of the radiocarbon
A 1,500-year extension to the present record may be pro- time range (Bronk Ramsey et al., 2012).
vided by the radiocarbon wiggle-matched Huon Pine
chronology from Tasmania and the Stuttgart-Hohenheim Palaeoenvironmental applications
chronologies from Europe (section 5.4.1.3), as this has As noted above (section 5.3.2.4), Δ14C in the atmosphere
enabled a continuous record of atmospheric 14C variations is modulated, inter alia, by geomagnetic and solar
to be established back to 14 ka cal. BP (Hua et al., 2009). influences, and hence the long-term deviations in Δ14C
In addition, linking of currently ‘floating’ tree-ring that are reflected in the radiocarbon calibration curve
chronologies (i.e. tree-ring series that are currently not provide a record of variations in the earth’s magnetic field
anchored to a calendrical timescale) to the present con- and in solar activity over the past 50 ka (Laj et al., 2002;
tinuous central European tree-ring record could extend the Muscheler et al., 2005). The level of detail that can be
radiocarbon calibration back to 14.2 ka cal. BP (section obtained from the 14C data is remarkable, as exemplified
5.4.1.3). Although older dendrochronological records have by Miyahara et al.’s (2007) study of the modulation of
been obtained from a number of sites around the world, the eleven-year sunspot cycle between AD 1413 and 1745.
the best prospects for deriving an even longer dendro- Such precise reconstructions of past geophysical behav-
chronology-based radiocarbon calibration appear to be iour enable temporal fluctuations in the global carbon
284 DATING METHODS
cycle to be quantified, and this has applications in global radioactive process. This converts a proton into a neutron
carbon modelling (Naegler & Levin, 2006), and in the (Figure 5.13) to reduce the atomic number by one and
evaluation of other influences on global climate change, yield argon, a gas. Only one of these branches, the 40K/40Ar
such as fluctuations in ocean circulation and ventilation pathway, is useful for dating, for 40Ca is so ubiquitous in
(Muscheler et al., 2004). nature that it is not possible to separate 40Ca atoms
produced by the decay of 40K from those already present
in rocks at the time of formation.
5.3.3 Argon-isotope dating 40
K/40Ar transformation is an accumulation radio-
metric clock. Once a volcanic rock has cooled, 40Ar from
5.3.3.1 Potassium–argon dating
the decay of 40K becomes trapped within the lattice, and
Two approaches are employed in argon isotope dating. its abundance increases with time. Hence measurement
Potassium–argon dating is a technique that allows the age of the amount of 40Ar enables an estimate to be made of
of volcanic rocks to be established using the radioactive the time that has elapsed since the formation of the rock.
isotope of potassium, 40K. This is a radioactive nuclide It has to be assumed, however, that no original 40Ar is
that undergoes branching decay, leading to one of two present and that the system remains closed to both 40Ar
daughter nuclides depending on the type of transformation and 40K after crystallization; hence dates will only be valid
(Figure 5.13). Most 40K decays by β emission to produce for host rocks from which argon gas cannot escape.
40
Ca, each particle emitted from the nucleus resulting in Fortunately, many mineral lattices retain argon, and it
the conversion of a neutron to a proton. The atomic is only if rocks become melted, recrystallized or heated to
number is therefore increased by one, resulting in an a critical temperature that substantial loss of argon will
element of different chemical properties, but with a virtually occur. In the case of a volcanic rock that has been reheated
unchanged atomic mass. Electron capture (from the sur- or metamorphosed, therefore, the method will deter-
rounding atomic orbitals) by the nucleus is the alternative mine the age at which the final phase of modification
ceased and not the age of initial formation. It also has
B- to be assumed that no non-radiogenic 40Ar is present (i.e.
a) argon from the atmosphere). In practice, some atmospheric
40
Ar will be included in rocks and minerals, but this can
be corrected for by comparing the 36Ar/40Ar ratio in the host
2l/ n rock with the known atmospheric ratio for the two gases.
20n
Potassium concentrations are measured by means of atomic
19* P 20 P
absorption spectrophotometry or flame photometry, while
the abundance ratios of the three argon isotopes (36Ar, 38Ar
(|T emission)
and 40Ar) are determined by gas isotope mass spectrometry
(Lanphere, 2000).
Potassium -40 Calcium-40 Radiometric dating by the 40K/40Ar method is largely
restricted to volcanic and metamorphic rocks, since
b) sedimentary rocks do not retain argon. Yet not all volcanic
and metamorphic rocks are suitable, for sufficient potas-
sium must be present to make dating possible. The nature
22n
of the mineral lattice is also an important factor, for not all
21j f n>
minerals retain argon over long time periods, particularly
sf^lOPy 18P
when under stress. Orthoclase and microcline, for example,
do not retain argon well. Biotite, on the other hand, is
(Electron capture}
one of the most suitable minerals, for not only does it retain
argon, but it is also rich in potassium. The long half-life of
40
Potassium -40 Argon - 4 0 K (1,250 million years) means that it can be used to date
some of the oldest rocks on earth. In the Quaternary,
Figure 5.13 Branching decay of 40K. a) Conversion of atoms however, the 40K/40Ar method is only appropriate for dating
of 40K to 40Ca through the emission of a β particle from the
volcanic rocks that are older than c. 100 ka. Although dates
nucleus. b) Conversion of 40K to 40Ar through electron capture
by the nucleus from one of the atomic orbitals. The latter are younger than this have been obtained, the large standard
shown in schematic form. errors (typically ± 100 per cent) mean that these have little
DATING METHODS 285
practical value. However, recent methodological developing a volcanic event; in other words, there has been no
ments have enabled meaningful ages to be obtained from loss of argon following crystallization. This could occur
volcanic rocks as young as 30 ka (Chernyshev et al., 2006), if the mineral sample has been weathered or if a further
suggesting that the method may now have a much wider episode of heating has occurred; in both cases, an under-
application in Late Quaternary geochronology. estimate of age will result. Careful petrographic examin-
ation may reveal evidence of weathering, while the presence
of secondary minerals (e.g. xenocrysts) may indicate that
5.3.3.2 Argon–argon (Ar/Ar) dating
recrystallization has occurred. If older mineral material
One difficulty with 40K/40Ar dating is that measurements has become incorporated into rocks, a relatively common
have to be made on separate aliquots (equal proportions phenomenon in volcanic materials, this may lead to an
of a sample) and this can lead to erroneous results over-estimate of age (Richards & Smart, 1991). Over-
where the sample material is heterogeneous (e.g. in basalts). estimate of age can also be caused by the presence of excess
The development of argon–argon dating circumvented argon in the sample; in other words, previously accum-
this problem, for here measurements are made on a single ulated gas that has failed to escape while the rock was
sample of material. In this method, the 40Ar content is molten.
measured directly, but the 40K concentration is determined A second key assumption is that all the 40Ar in the
indirectly by using the known proportions between the sample has been derived from the decay of 40K. The problem
potassium and argon isotopes. Sample grains are placed in here is that 40Ar is also a constituent of the atmosphere and
a nuclear reactor and irradiated. This converts a portion hence a proportion of this atmospheric 40Ar (as opposed
of the stable 39K isotopes to another form of argon, 39Ar. to radiogenic 40Ar) will be present in minerals. This can be
The 39Ar abundance is proportional to that of 39K which is, corrected for, however, by using another argon isotope,
in turn, proportional to 40K. Hence a single mass spectro- 36
Ar, which accompanies atmospheric 40Ar, and which will
metric analysis can be employed to determine the 40Ar/40K only be present in the sample as a result of atmospheric
ratio (McDougall & Harrison, 1999). The advantage of contamination (Kelley, 2002a). Measurement of 36Ar in
this approach is that dating can be carried out using very the sample will therefore enable a correction to be made
small samples, typically less than 10 g of material (e.g. for the presence of atmospheric 40Ar, using the known
tephra). Indeed, recent refinements to the technique enable ratio of 40Ar/36Ar in the contemporary environment. A
single mineral crystals of the order of 1 mg or less to be further difficulty occurs in argon–argon dating, how-
targeted. This method, which employs a high-powered ever, because 36Ar is produced from calcium when the
laser to drive off the argon for measurement using a super- sample is exposed in a nuclear reactor. In this case,
sensitive mass spectrometer, is known as single laser correction for the presence of atmospheric argon involves
crystal fusion (SLCF) 40Ar/39Ar dating, and has led to a the measurement of the non-naturally occurring argon
significant improvement in both accuracy and precision isotope, 37Ar, which is also produced from calcium (Aitken,
of argon–argon age estimates (Wintle, 1996). Although this 1990). A good overview of the principles and problems of
technique has been most widely employed in the dating of both Ar–Ar and K–Ar dating can be found in Kelley
older volcanic materials, such as the 2.1 Ma old Huckleberry (2002b).
Ridge Tuff in Yellowstone, northwest USA (Ellis et al.,
2012), it has been used to date more recent volcanic events,
5.3.3.4 Some applications of argon-isotope
such as the Laacher See eruption of 12.9 ka in the Eifel
dating
region of Germany (Van den Bogaard, 1995) and the Toba
eruption in Indonesia around 73 ka (Mark et al., 2013). The principal contributions of argon-isotope dating in
Over shorter timescales, the dating of the eruption of Quaternary research include the dating of early hominid
Vesuvius in AD 79 (Renne et al., 1997) demonstrates that sites in Africa (McDougall et al., 2005) and the provision
the 40Ar/39Ar method can provide a practical geochron- of an outline timeframe for early hominid migration from
ometer into the late Holocene. Africa to eastern Europe and the Far East (Garcia et al.,
2010; Sémah et al., 2000). The method has also been used
in the development of glacial chronologies by dating tephras
5.3.3.3 Problems and limitations of argon-
interbedded with glacial deposits, for example in the Yukon
isotope dating
(Westgate et al., 2001) and Patagonia (Singer et al., 2004),
Two fundamental assumptions underlie argon-isotope and it provides a chronology for the palaeomagnetic
dating. The first is that the system remains closed follow- timescale (section 5.5.1).
286 DATING METHODS
5.3.4 Uranium-series (U-series) dating can occur in natural systems for a variety of reasons. These
include the loss of radon (Ra) by gaseous diffusion through
a porous rock matrix, and the separation of different
chemical elements during weathering, transport and
238
Uranium, 235uranium and 232thorium all decay to stable deposition in the hydrosphere. Where the U-decay is
lead isotopes through complex decay series of intermediate interrupted, and some decay products selectively removed,
nuclides with widely differing half-lives (Figure 5.14). The the uranium-series disequilibrium dating method can be
helium (He) gas formed by a particle emission may become applied. Overviews can be found in Smart (1991a) and
trapped within host rocks, or may slowly diffuse out, Latham (2001), while more in-depth accounts are provided
ultimately to be liberated into the atmosphere. In theory, by Ivanovich & Harmon (1995), Bourdon et al. (2003) and
the age of a rock or mineral can be obtained from the van Calsteren & Thomas (2006).
amount of stable lead produced, or from the amount of Disequilibrium methods are based on the following
He liberated, but these measures are usually restricted to geochemical principles. Uranium (U) and weathering
the dating of much older rocks (but see below). Within the products containing U are highly soluble, whereas other
more limited timescale of the Quaternary, only those products of the uranium decay chain, such as thorium
intermediate nuclides with relatively short half-lives can (230Th) and protactinium (231Pa) are readily absorbed
usually be employed. However, nuclides with half-lives or precipitated. Thus 230Th and 231Pa are co-precipitated
of only a few years or less are impractical for dating and with other salts to accumulate on the floors of lakes and
even the intermediate nuclides with half-lives of hun- on the seabed, while U remains in solution. A selective
dreds or thousands of years cannot be used for dating of separation (fractionation) of these decay products there-
materials where radioactive disintegration has proceeded fore occurs. Accumulating sediments will contain quant-
in an undisturbed system. This is because in most rocks, ities of 230Th and 231Pa but will be deficient in U, whilst
an equilibrium state has been achieved in which nuclides organisms that secrete carbonate direct from ocean waters
formed by decay are disintegrating at rates similar to their (such as molluscs and corals) will build a carbonate shell
rate of production by the parent nuclide. If the decay chain or skeleton that contains U but very little 230Th or 231Pa.
remains unbroken, parent and daughter isotopes remain The same principles apply to carbonate precipitates
in radioactive equilibrium. However, if the decay chain is such as speleothems and travertines, where fractionation
broken, then the system will be in disequilibrium until results in the separation of uranium from decay products,
equilibrium is restored through subsequent radioactive and the age of precipitation can be measured from the
decay. Disequilibrium between the longer-lived isotopes renewed accumulation of radioactive decay products in
U-238 Series U-235 Serie s Tb-232 Serie s

U-234 Th-226
U-233 250,0001 U-23S Pa-231 Th-232 2y
4.5x10 v 9
a 713x10 s
34,300y 1.4i10 y
1u
it Pa-234 a Th-230
| l ft aa . aa Th-227 Ac-223
1 18 75,200 y 25.6 h 4 6h
Th-234 rr. Th-231 ^0-22 7 Ra-223 Ra-224
24d a 25.6b 22 v a a 5.7 V 3.64d
Ra 22S
Ra 223
1S22v
11.Id
it
a Rn-220
Rn-222
Po-214 Po-210 Rn-219 Po-212
3 8d [ 140d Bi-212
a
I' a<t Bi-211 1h
Po-218m Bi-214 a Po-216
BI-S10 a 2 16m
Po-215
a 19.7m 19.7m Pb-207 Pb-206
aa a a (Slaljls : (stable )
Pb-214 Pb-210 a TI-2CB
Pb-206
27m a 22.2 y Pb-211 TI-307 Pb-212 3.1m
(stable )
36.1m 4.79m a 10.6h
summary : summary : MBBBM
8 alpha s 7 alpha s 6 alpha s
33S(J alpha ('Radiu m lead'] 235[J ™pti ['Actiniu m lead' ) 2J!Th 20SPI) Thoriu m lead' ;
3 betas 4 betas 4 betas
Figure 5.14 Chain decay pathways and half-lives of intermediate nuclides during the decay of 238U, 235U and 232Th to stable lead.
The elements are arranged vertically according to atomic number. Loss of an α particle results in a decrease in atomic number,
whereas emission of a β particle leads to an increase. Some of the very short-lived nuclides within the decay chain have been
omitted; d – days; h – hours; m – minutes.
DATING METHODS 287
the speleothem calcite. Age of lake or ocean floor sediments difficult to satisfy, and a number of materials show evidence
can be estimated by measuring the rate of decay of 230Th of departure from closed system behaviour (Table 5.3).
or 231Pa down a sediment profile, while the age of carbonate Dating of such material necessitates the use of correction
fossils, speleothems, teeth and bone can be derived from factors which, in turn, requires a detailed knowledge of the
measurement of the accumulation of decay products of geochemistry that gives rise to the isotope disequilibrium
U within the carbonate matrix. The former has been (see below). Open-system behaviour may be reflected in
referred to as the daughter excess (DE) type of U dis- reversals in isotopic ages in stratigraphic sequence or the
equilibrium series dating method, that is, the daughter samples themselves may show evidence of open-system
nuclides are initially present in excess of concentration activity. For example, petrographic study of calcite may
at secular equilibrium before decaying over time, while the indicate recrystallization and similar evidence may be
latter is termed the daughter deficient (DD) group of detected in bone (Scholz et al., 2004). A fourth assumption
methods, in which the daughter nuclide is initially absent is that the sample has not been contaminated at the time
but increases with time until equilibrium is achieved. of formation by detrital materials that already contain
DE methods have been most widely used in the dating daughter nuclides. Hence in the measurement of the
of ocean sediments using the decay of 230Th or 231Pa, while 234
U/230Th ratio in speleothem calcite, for example, it is
the DD methods are principally based on measurement of assumed that the 230Th content was zero at the time of
230
Th/234U ratios in a variety of materials including tufa, crystal formation, for any residual 230Th in the sample
speleothem, shell, bone and phosphates, all of which are would result in an age that was too old. Again, empirical
initially deficient in thorium. evidence has shown that this assumption does not always
hold, and that dated materials may contain varying
amounts of detrital 230Th. The presence of detrital 230Th can
5.3.4.2 Measurement, problems and age
be detected by measuring the amounts of 232Th in a sample.
range
This long-lived isotope occurs in water as a trace impurity
Conventional measurement of U-series ages is by means and, where present, is indicative that contamination has
of alpha spectrometry following chemical extraction of occurred. The 232Th/230Th ratio may then be used as a basis
thorium and uranium from the sample material. Since the for the correction of detrital contamination. One way to
late 1980s, however, thermal ionization mass spectrometry identify the effects of detrital contamination and to correct
(TIMS) has been employed to determine U-isotope ratios. for them is by the isochron technique (Dorale et al., 2007).
This approach enables individual atoms to be counted This involves measuring the activity ratios of different
directly as opposed to the monitoring of alpha particles minerals, or of whole rock fragments, obtained from the
emitted during radioactive decay. It is therefore more rapid dated sample. If straight line plots are obtained, this
and, as count rates are not restricted by the half-life of the indicates that the different mineral phases were formed
isotope, there is the potential for extending the age range simultaneously, and the best estimate of age is that ratio at
of the technique. In addition, the method offers a greater which the isochron lines intersect. A large scatter in the
level of analytical precision enabling, for example, dataset indicates that detrital contamination is a major
meaningful ages to be obtained from Lateglacial and problem.
Holocene materials (Camoin et al., 2012). Over the past The age range of the uranium-series disequilibrium
decade, further advances have been made in thermal method varies with the nuclides employed. The practical
ionization techniques, and the advent of alternative mass dating range using alpha-particle spectrometric methods
spectrometric methods, especially multi-collector induc- is around five half-lives, and hence 230Th/234U has been
tively coupled-plasma mass spectrometry (MC-ICPMS) employed to date samples in the range 5–350 ka, while
231
(Goldstein & Stirling, 2003) and laser-ablation ICPMS Pa/235U and 231Pa/230Th have upper limits of around 200
(Eggins et al., 2005), have continued to improve the quality ka and 250 ka, respectively. Recent developments in mass
of U-series studies. spectrometric counting (TIMS, ICPMS), however, have
A number of assumptions underlie the U-series dis- reduced the lower limits of the 230Th/234U method to less
equilibrium method. It must be assumed that the decay than 100 years, while increasing the upper age limit to
coefficients have been accurately determined and that the c. 500 ka. Materials in the range 100 ka to 10 Ma may, in
activity ratio of daughter to parent nuclide can be measured theory, be dated using 4He/U ratios, while the 234U/238U
to a high level of precision. These two requirements can method has a potential age range of up to 1.5 Ma. Although
largely be met. A third assumption, namely that there has the very long half-life of 238U (4.5 × 109 yr) means that
been no loss or gain of nuclides since deposition, is more U-Pb (lead) methods had previously been largely applied
288 DATING METHODS
Table 5.3 Reliability of uranium-series dates for terrestrial materials due to deviations from closed system
behaviour and contamination by 230Th and 234U from detritus (after Smart, 1991a).
Reliability Material Closed system? Contaminated?
Reliable Unaltered coral Closed Clean
Clean speleothem Clean
Volcanic rocks –
Dirty speleothem Contaminated
Possibly reliable Ferruginous concretns Possibly closed Contaminated
Tufa Contaminated
Mollusc shells Contaminated
Phosphates Contaminated
Generally unreliable Altered coral* Open Clean
Bone ?
Evaporites Contaminated
Caliche Contaminated
Stromatolites Contaminated
Peat and wood ?
* Diagenetically altered.
only to the dating of old rocks, under certain favour- analysis has been used, as in radiocarbon, in the develop-
able circumstances (e.g. radiogenic speleothems with ment of speleothem chronologies (Millard, 2006). The
very low common Pb) and using recently developed MC- dating of cave speleothem and associated material forms
IPCMS analytical methodologies, it has proved possible to the chronological basis for a range of environmental
generate excellent age resolution on much older materials applications (Richards & Dorale, 2003); these include
(Woodhead et al., 2006). Examples include the dating of climatic reconstructions (Fleitmann et al., 2004), cave
an Early Pleistocene Italian cave stalagmite (Bajo et al., archaeology (Pike & Pettitt, 2003), the depositional history
2012), and flowstones of around 1 Ma in caves from South within caves (Hearty et al., 2004), sea-level change (Surić
Africa (Pickering et al., 2013). Further refinements in this et al., 2009), and in an unusual application, the dating of
area could eventually lead to a combination of U-Th and cave art (Pike et al., 2012). Further applications and details
U-Pb methods covering the whole of the Quaternary time of recent developments in speleothem dating can be found
range and beyond (Latham, 2001). in Drysdale et al. (2012).
Corals
5.3.4.3 Some applications of U-series dating
Corals are generally regarded as one of the best media
Speleothems for U-series dating for, after death, coral skeletons act as
230
Th/234U dating has been widely applied to the dating closed systems until the coral is dissolved or changes to
of cave calcite precipitates – stalagmites, stalactites, flow- calcite. Moreover, they contain sufficient uranium (typically
stones, etc., the palaeoclimatic significance of which has 2–3 ppm) for the application of both the 230Th/234U and
already been considered (section 3.8). 234U is precipitated 231
Pa/235U methods. Since these are independent decay
from karst waters during the formation of speleothem chains, they can provide an internal check on calculated
carbonate, and almost all of the 230Th subsequently found ages. U-series dating of raised coral reef complexes has been
appears to be authigenic, that is, it has originated from widely employed in studies of Late Quaternary environ-
decay of 234U that forms part of the speleothem chem- mental change, with high stands of sea level during the
istry. Because not all speleothems can be layer-counted, recent interglacials commonly recorded in raised reef
age–depth models have been produced using U-series complexes (Thompson et al., 2011) although corals of
determinations spread along the length of the speleothem Middle Pleistocene age have also been dated using this
(Scholz et al., 2012), and in some instances, Bayesian method (Stirling & Andersen, 2009). In addition, the
DATING METHODS 289
technique has been used to determine the duration of therefore, the reappearance of Th and Pa in bone apatite
interglacial periods as reflected in periods of high sea level can be determined using either the 230Th/234U or 231Pa/235U
(Edwards et al., 2003). As noted above (section 5.3.2.6), ratios, but discrepancies between 230Th/234U and 231Pa/235U
U-series dating of fossil corals provides a basis for calibrat- dates on the same sample, and between 14C dates and U-
ing the radiocarbon timescale (Fairbanks et al., 2005), and series dates on bone, suggests that problems of leaching
can also be used to calibrate orbitally tuned timescales based (resulting in U depletion) and open-system behaviour
on the marine oxygen isotope record (Thompson & (resulting in U enrichment) remain. However, because
Goldstein, 2006; section 5.5.3). both human and animal bone is such an important and
relatively common fossil material (Pike & Pettitt, 2003),
Carbonate deposits considerable efforts have been made to resolve the problem
These include such diverse materials as travertines, calcretes, of open-system behaviour in buried bone (Pike et al., 2002;
lake marls, stromatolites, phosphates and evaporites. Grün et al., 2010). These have largely involved the devel-
Travertines and calcareous tufa can be difficult materials opment of uptake models to try to describe how and when
to date because of their complex growth mechanisms and the U was absorbed into bone (e.g. Sambridge et al., 2012),
contamination by detrital and organic particles (Garnett but these may need to be calibrated using other dating
et al., 2004), but reliable U-series age determinations can techniques, such as electron spin resonance (section 5.3.7).
be obtained providing they are based on primary calcite Further progress is still required before reliable U-series
(e.g. micrite and spar), and iron (Fe) and aluminium (Al) dates can be obtained from bones and teeth.
levels are low (Mallick & Frank, 2002). Dating of pedo-
genic calcareous nodules in alluvial terrace sequences is Peat and other applications
possible, but only if multiple ages are generated from a Peat, along with other organic material, takes up U from
single site and are shown to be stratigraphically consistent groundwaters and can become relatively enriched in the
(Candy et al., 2005). U-series dating of lake sediments element. Since peat has a high adsorption capacity, any
usually requires significant corrections to account for percolating groundwater will transfer its U content to the
detrital inputs of Th and U, but the technique has enabled upper peat surface layer, thus protecting the older layers
coherent long chronologies and sediment accumulation from further acquisition of U. As with bone, the incorp-
histories to be established (Torfstein et al., 2009). oration of U-bearing inorganic detrital material into the
peat deposits has proved to be a problem. However, some
encouraging results have been obtained on peat (Walker
Molluscs et al., 1992; Vaughan et al., 2004), and refinement of the
Thus far, U-series dating of fossil molluscs has been method offers the prospect of dating interglacial peats and
relatively unsuccessful, due partly to the fact that they con- other organic deposits that currently lie beyond the range
tain initially only very small amounts of uranium (e.g. one- of radiocarbon (Frechen et al., 2007).
fiftieth of that contained in corals), and also to the fact U-series has also been employed in the dating of young
that they do not function as geochemically closed systems, volcanic rocks (Miallier et al., 2004) and of wood samples
for diagenetic uptake of uranium is common following from fluvial sediments (Allard et al., 2012).
death of the organism (McLaren & Rowe, 1996). Hence,
uranium-series dates on molluscs have tended to be
regarded as unreliable. On the other hand, coherent age 5.3.5 Fission track dating
estimates on molluscs have been obtained in some instances
(Jedoui et al., 2003), and particularly from contexts where 5.3.5.1 General principles
these can be compared with the results from other dating This method, which dates uranium-bearing crystals, is
methods, such as radiocarbon (Magnani et al., 2007). As based on the spontaneous fission of 238U: that is, the
many shorelines are not characterized by coral (a more nucleus (of atomic number 92) divides to form elements
suitable dating medium), but often contain shells, attempts of medium atomic number from about 30–65 (e.g. barium
are likely to continue to find a means of obtaining reliable 56). An important consequence of spontaneous fission is
U-series dates from fossil molluscs. that the energy released leads to high-speed collisions
between fission fragments and neighbouring atoms. In
Bone rocks containing uranium, fission fragments cause damage
Following the death of an animal, U from groundwater trails or tracks in the wake of their movement through
enters and is trapped within the bone apatite. In theory, the host crystal lattice. The ‘damage’ induced is a result of
290 DATING METHODS
ionization of the atoms that come into contact with fission fading is reflected in a reduced number of fission tracks and,
fragments. The positive charge acquired by adjacent atoms if not detected, will lead to an under-estimate of age. Not
leads to mutual repulsion and therefore disorder in the all materials show the same tendency to anneal, how-
crystal lattice. The tracks can be retained for millions of ever; in some minerals, such as zircon and titanite, fission
years and their number is a function of both U content and tracks are relatively stable and fading is less of a problem.
time. Fission track (FT) dating is mostly applicable to Apatite, by contrast, is much less stable, while in glasses
volcanic extrusive rocks such as basalts, tephras and tuffs, stability decreases with decreasing silica content. A number
where the fission track clock is zeroed by the heating event, of experimentally established correction procedures can be
although it has also been applied to volcanic glasses such applied to counter this problem. These involve the step-wise
as obsidian and pumice. Zircon is the most widely used heating of paired samples containing spontaneous and
mineral because of its naturally high U content and also its induced tracks until the ratio of the spontaneous track
widespread occurrence in volcanic rocks. Details of fission density in the natural sample to the induced track density
track dating can be found in Wagner & van den Haute in the irradiated sample reaches a plateau level, this plateau
(1992) and van den Haute & Corte (1998). value providing a corrected age for the sample (plateau
A variant of FT dating is alpha-recoil track (ART) correction procedure). A variant on this method is the
dating. This method is also based on the analysis of damage isothermal plateau fission track technique (ITPFT) in
trails etched in crystals by natural radioactivity, but ARTs which paired samples are subjected to a single heat
are formed by recoil of the nucleus that emits alpha particles treatment until a pre-determined temperature is achieved
during the decay of U and Th, as well as from other and sustained for a specified period (Westgate, 1989).
daughter nuclei. Although relatively new and not yet as Since the density of fission tracks depends on U content
widely applied as FT dating, tests of the concentrations of and age, the method cannot be applied where the density
ARTs in dark mica from the Quaternary volcanics of the of tracks is too low (typically in samples of less than 100
Eifel region, Germany, suggest that this technique may have ka) or, at the other extreme, so high that counting becomes
considerable potential for dating Quaternary events (Gögen impossible. Ironically, rocks with a rich U content are
& Wagner, 2000). not suitable because of the very high density of tracks, and
in these instances the selection of minerals with a low
U content, such as sphene, may form a more suitable
5.3.5.2 Measurement and problems
counting medium.
The age of a mineral or glass can be obtained by counting
‘spontaneous’ 238U fission tracks under a microscope after
5.3.5.3 Some applications of fission track
the surface has been polished and treated with a chemical
dating
etchant to enlarge the fission tracks. In order to use these
data to infer age, it is necessary to know the original U Materials that have been dated by this method range from
content of the sample and while this cannot be determined glass manufactured in the nineteenth century to some of
by direct measurement, it can be inferred indirectly by the oldest rocks on earth (Wagner, 1998). In the context
measuring the amount of 235U. After the ‘spontaneous’ of the Quaternary, the most significant contribution of
fission tracks have been counted, therefore, the sample is the technique has been in the dating of tephras and tuffs
irradiated in a nuclear reactor with thermal neutrons to in the age range 50 ka (the upper limit of radiocarbon) to
induce fission in atoms of the less abundant U isotope 235U; 500 ka. Where these have been found interbedded with
this produces a new set of fission tracks which can also be glacial deposits, fission track dating can provide a timescale
etched and counted. The ‘induced’ fission track count for glacial events, as has been the case, for instance, in
provides a measure of 235U abundance, from which the western Argentina (Espizua et al., 2002), the west-central
concentration of 238U in the sample can be obtained from United States (Colgan, 1999) and Alaska (Westgate et al.,
the known 235U/238U ratio in volcanic rocks (Hurford, 2001). Widespread tephras form key marker horizons
1991). (section 5.5.2) and fission track dating may enable ages
The method, however, is not without its problems. to be assigned to these. Examples include the Rockland
Fission tracks can ‘heal’ or be erased (a process known Tephra (c. 400 ka), a widespread pyroclastic layer and
as fading or annealing) through the heating of the host hence a key stratigraphic marker horizon in western North
materials or through spontaneous diffusion of ions. The America (Meyer et al., 1991), and the much older but
result is that the tracks become narrower and shorter over equally significant Bishop Tuff (c. 700 ka) also from
time until they eventually disappear completely. Partial the western United States (Izett & Naeser, 1976). Other
DATING METHODS 291
applications include the dating of the Early Pleistocene a)

hominid colonization of southeast Asia (O’Sullivan et al.,
2001), of Middle Pleistocene (800–900 ka) stone tool Natural
radiation Laboratory irradiation
assemblages on the island of Flores in Indonesia (Morwood
irradiation
et al., 1998) and of Middle Pleistocene glacial–interglacial [N)
sequences in Italy (Marcolini et al., 2003). The technique
Laboratory
has also provided chronologies for uplift and erosion in the
Himalayas (Burbank et al., 2003) and for the Pleistocene
drainage history of the European Alps (Bernet et al., 2004).
•Natura l intensity
In addition, the method has also been used in the tracing
of trade routes in the Andes based on fission track dates
-100 -50 0 50 100 150 200 250 300
obtained from obsidian artefacts (Dorighel et al., 1998).
Laboratory dose (Gy)
b)
5.3.6 Luminescence dating

N.
irradiation
Any material that contains U, Th or potassium (K) (all
Laboratory
sediments and volcanic rocks contain all three), or lies in
close proximity to other materials containing these radio-
active substances, is subject to continuous bombard- i
ment by α, β and γ particles. This leads to ionization in the
host materials and the ‘trapping’ of metastable electrons
within minerals. These electrons can be freed by heating,
and under controlled conditions, a characteristic emission 0 50 100 150
of light occurs which is proportional to the number of Laboratory dose (Gy)
electrons trapped within the crystal lattice of mineral grains,
principally quartz and feldspar. This is termed thermo- Figure 5.15 Methods of measuring palaeodose (DE) after the
natural TL intensity of a sample (N) has been established. a) In
luminescence and is the basis of thermoluminescence the additive method, different aliquots of the sample are
(TL) dating. The electrons can also be released from traps irradiated in the laboratory at increasing intensities of irradiation
by shining a beam of light onto the sample, and again the and the corresponding TL properties measured. b) In the
luminescent signal is a reflection of the number of elec- regenerative method, aliquots of the sample are initially
bleached to low radiative intensities and then progressively
trons trapped within the crystal lattice. This is optically
irradiated by known amounts, with corresponding TL values
stimulated luminescence and hence the technique is noted. Method b can be applied using multiple or single aliquots
referred to as optically stimulated luminescence (OSL) (single aliquot regeneration, or SAR) of a sample. Each
dating, although this is now often abbreviated to optical approach establishes the mathematical correspondence
dating (Aitken, 1998). between DE, N and TL. Radiation intensity is measured in units
of gray (Gy) (based on Grün, 2001 and Aitken, 1998).
All electron traps can be emptied by exposure to
heat, but only some are sensitive to light. The luminescent
clock is reset by heating, for example by firing of pottery
or, in the case of sediments, by exposure to heat from luminescence dating can be found in Aitken (1998), Lian
a campfire or forest fire, or by exposure to sunlight. This & Huntley (2001), Duller (2004) and Wintle (2008). Case
process is often referred to as bleaching. In both cases, studies are described in two recent symposium volumes on
the majority of the electron traps are effectively emptied luminescence and ESR dating (Grün, 2010; Grün &
but then progressively fill following burial of the pottery or Preusser, 2012).
sediment. The ‘natural luminescence signal’ therefore
provides an indication of the time that has elapsed since
firing, in pottery, or, in the case of sediment, since burial 5.3.6.2 Measurement and problems
and removal from sunlight. In volcanic rocks, the zero- TL measurements are carried out on a sample of mineral
ing of the electron clock will be a direct consequence material, usually a separated quartz or feldspar fraction.
of both heat and mineral formation. Further details on This is heated to temperatures in excess of 500°C and as
292 DATING METHODS
light (photons) is emitted from the luminescence centres, Once the dose rate has been established, then TL/OSL
these photons are converted to electric pulses using a age can be calculated from the following:
sensitive light detector. The light emission (TL intensity)
is then plotted against heating temperature to produce Equivalent dose
TL/OSL age (yr) =
a glow curve, in which the peaks are reflective of the Annual dosee
thermal lifetimes of the various electron trap populations
within the sample. The most important are the traps with Possible sources of error in luminescence dating include
long thermal lifetimes (‘deep traps’) because the electrons systematic errors in the calibration of laboratory radiation
within them will remain there over relatively long periods sources, light contamination during field sampling, and the
of time. Once a ‘natural signal’ has been obtained, this is accurate determination of the dose rate discussed above. A
compared with the ‘artificial signals’ obtained from a number of potential difficulties surround the zeroing of the
portion of the sample to which known doses of radiation luminescence clock because, particularly in TL dating,
have been administered from a calibrated laboratory there may be a residual TL component which, if not
radioisotope source (Figure 5.15). This enables the equiva- corrected for, will result in an aberrant age. Reworking of
lent dose (DE) to be determined, which is a measure of sediments, intermittent exposure to sunlight, and grain-to-
the amount of radiation that would be needed to generate grain variability in the extent of bleaching are all problems
a TL signal equal to that which the sample has acquired encountered in the dating of sediments. Inadequate
subsequent to the most recent firing (zeroing) event or bleaching and reworking of older sediments may be
exposure to sunlight. The DE is sometimes referred to as the particularly problematical in the dating of glacial sediments
palaeodose or palaeodose equivalent. (Lukas et al., 2007). Dating errors can also arise as a result
Many of the principles of TL dating underlie OSL of the leakage of electrons from thermally stable electron
dating, but the major difference is that the trapped electrons traps, a problem that affects both TL and OSL dating of
are released by light rather than by heat. In the earlier feldspars (but not quartz) and which is usually referred to
studies, a beam of green light from an argon ion laser was as anomalous fading (Huntley & Lamothe, 2001). This
employed, but other light sources have subsequently been results from disequilibrium in the uranium decay chain and
used, including filtered halogen lamps and high-powered variations in past water content of sediments, the latter in
light-emitting diodes (LEDs). A further development has particular being very difficult to assess.
been the employment of infrared stimulated luminescence
(IRSL), although this approach can only be applied to
5.3.6.3 Developments in luminescence dating
feldspar grains as electron traps in quartz are insensitive
to infra-red stimulation. For quartz grains, high-powered The increasing use of OSL in the dating of sediments has
blue-green LEDs are now routinely used. stimulated a number of methodological developments. An
In order to arrive at an estimate of age using either important advance was the single aliquot regeneration
TL or OSL, the environmental dose rate (or annual dose method (SAR) which uses repeated measurements on a
rate) has to be determined. This is a measure of the single sample rather than multiple aliquots from the same
radiation dose per unit of time absorbed by particular sample, as is the case in TL, to determine DE (Wintle &
minerals since the zeroing of the luminescence clock either Murray, 2000). This enables much smaller samples of
by firing or by exposure to sunlight. The dose rate is material to be dated, and it also avoids the problems
calculated from an analysis of the radioactive elements in of differing luminescence sensitivity resulting from grain-
both the sample and its surroundings, the so-called internal to-grain variations within a sample (section 5.3.6.2). A
dose rate and external dose rate. However, the deter- second important development has been the ability to
mination of the total radiation dose that the sample has obtain OSL measurements from single grains of quartz or
received since burial is not straightforward for, in addition feldspar. Again, this makes it possible to determine whether
to radiation from the surrounding sediments, the sample all of the grains in a sample of sediment possess the same
will also have been affected by cosmic rays from deep apparent age (Duller, 2004). More recently, studies have
space, and these need to be corrected for. In addition, water been undertaken into a new signal from quartz, termed the
and organic matter, where present in a body of sediment, thermally transferred optically stimulated luminescence
absorb radiation at a different rate from mineral sediment, (TT-OSL). This appears to have the potential to date
and this has to be accounted for in dose rate calculations. sediments over much longer periods than OSL, but further
Overall, this can usually be measured with a precision of comparisons with independent chronological methods are
around 5 per cent (Grün, 2001). needed (Duller & Wintle, 2012). These and other techno-
DATING METHODS 293
logical advances are reviewed in Wintle (2008, 2010). dating of beach sands (Roberts & Plater, 2007), glacial
Although OSL is a relatively recent development, it has deposits (Puthusserry et al., 2008), lake deposits (Wilkins
effectively replaced TL in the dating of sediments, partly et al., 2012), colluvial deposits (Fuchs et al., 2004) and
because of the technical innovations described above, but fluvial sediments (Popov et al., 2012). The technique has
also because residual signals are usually much smaller for proved to be especially valuable in the dating of sediment
OSL than for TL dating, which enables younger sample sequences and landforms in desert regions, where there is
materials to be dated (Lian & Huntley, 2012). frequently no alternative chronometric method (Singhvi &
Porat, 2008; Singhvi et al., 2010).
5.3.6.4 Age ranges and applications of
luminescence dating 5.3.7 Electron spin resonance (ESR)
The lower age range of luminescence dating reflects the
dating
sensitivity of the sample and the efficiency of the zeroing
5.3.7.1 General principles and measurement
mechanism, while at the upper end of the dating range,
saturation (the point at which the electron traps become Electron spin resonance (ESR) dating was initially used in
completely filled) and the long-term stability of the TL/OSL Quaternary research to determine the age of speleothem
signal are the principal constraints. At the younger end of calcite, but it has since been employed to date a variety
the dating range, TL ages of around 100 years have been of materials, most notably tooth enamel, but also coral,
obtained from fired pottery, while OSL dates obtained molluscs, quartz-bearing rocks and sediments, and burnt
on quartz grains from dune sands in Denmark that accum- flint. The basic principles of the technique are similar to
ulated during the last few centuries are entirely consistent those of optical dating in that it involves the direct meas-
with independent dates from historical sources (Clem- urement of radiation-induced paramagnetic electrons
mensen & Murray, 2006). At present, the upper practical trapped in crystal defects in a body of rock or other material
limit for quartz appears to be around 150–200 ka, although (indeed, the technique is sometimes referred to as electron
luminescent ages of around 450 ka have been obtained on paramagnetic resonance). These ‘free’ electrons are gen-
quartz from fluvial sediments in the Thames Valley which erated by α, β and γ radiation from natural radioelements
were consistent with an independently assigned age (MIS (e.g. U, Th and K) and have accumulated in the crystal
12) for the deposits (Pawley et al., 2010). Optical dates on lattices of minerals over time. Exposure to high-frequency
feldspar may have a similar upper age limit to quartz. electromagnetic radiation in a strong magnetic field ‘excites’
Here, problems of anomalous fading are a major limitation, the electrons and their resonance can be detected as the
but recent studies using infra-red IRSL and the application magnetic field is changed. This is because when the trapped
of carefully selected correction factors suggest that this electrons (or ‘spins’) are in resonance, electromagnetic
range might, in due course, be extended (Kars et al., 2008). power is absorbed in proportion to the number of elec-
With the TT-OSL method, there is the potential for dating trons present; the greater the number of electrons, the
sediments back to one million years (Duller & Wintle, greater the absorption. Hence the latter is a reflection of age
2012). This has been recently confirmed by the dating of and, as in luminescence dating, is known as the palaeodose
quartz from cave sediments in South Africa, where TT-OSL or equivalent dose. Again, this is a measure of the amount
dates of around 1 Ma were supported by independent age of radiation that would be needed to generate an ESR
estimates based on U-series and palaeomagnetic evidence signal equal to that which the sample has acquired
(Pickering et al., 2013). subsequent to the most recent zeroing event.
The range of luminescence dating applications is In order to obtain an age for the sample, an estimate has
considerable, and the technique is now used almost as also to be made of the dose rate (or annual dose), which
widely as radiocarbon in the dating of Late Quaternary is a measure of radiation dose per unit of time since the
events. It has been applied to the dating of a variety of zeroing of the electron clock. This includes both internal
archaeological contexts, including stratified cave sequences and external radiation, plus cosmic ray radiation. The age
with interbedded artefacts (Barton et al., 2009), burnt of a sample can then be obtained by dividing the value for
flints from Palaeolithic contexts (Richter, 2007), ancient equivalent dose by the dose rate. It is important to note,
hearths (Sun et al., 2012), historic iron smelters (Godfrey- however, that in ESR dating, the electron traps are not
Smith & Casey, 2003), pottery (Zink et al., 2012) and emptied, as in the case of luminescence dating, and this
calcite overgrowths on ancient rockwall cave paintings means that replicate measurements can be made on a
(Watanabe et al., 2003a). Other applications include the single sample of material. Overviews of ESR dating can be
294 DATING METHODS
found in Rink (1997), Blackwell (2001a) and Schellmann that coherent ESR ages can be obtained from younger
et al. (2008). coral material that was formed only a few hundred years
or even a few decades ago (Radtke et al., 2003). Further
applications of the ESR method can be found in Grün
5.3.7.2 Sources of error in ESR dating
(2010) and Grün & Preusser (2012).
Almost all of the error sources discussed above in relation
to luminescence dating apply equally to ESR dating. These
5.3.8 Cosmogenic radionuclide (CRN)
include the non-zeroing of the electron clock, leakage
from electron traps and problems relating to sample con-
dating
text, such as reworking of materials or the effects of ground-
water movement. The last-named poses particular
difficulties in estimating the annual dose, for while ideally Cosmogenic nuclides form when high-energy cosmic rays
both the dated sample and surrounding deposit should entering the earth’s atmosphere collide with nuclei, trig-
form a geochemically closed system with regard to the gering a cascade of high-energy neutrons that impact on
relevant isotopes, in practice erosion or deposition of rela- the earth’s surface. The collisions between these neutrons
tively radioisotope-rich sediments surrounding the sample and target nuclei within certain minerals cause these
can often result in major changes in the radiation flux nuclei to fragment (a process known as spallation) resulting
over time (Smart, 1991b). A further complication is that in the creation of new radioactive nuclides. These accum-
many materials conventionally dated by ESR show post- ulate in the surface layers of rocks and boulders and their
depositional uranium uptake, the extent of which cannot abundance is directly related to the time that the surfaces
normally be established, and recourse has to be made have been exposed to cosmic ray activity. By measuring
to models to attempt to simulate the process (Grün, 2001). the concentration of cosmogenic nuclides in rock surface
As a consequence, although ESR dates may be accurate samples, therefore, an estimate can be obtained on the time
(in other words, they provide a reasonable approximation of exposure of that rock surface. This is the basis of
of the true age of the sample: section 5.2), they tend to be cosmogenic radionuclide (CRN) dating, also known as
of low analytical precision. Precision will also be affected terrestrial in situ cosmogenic nuclide (TCN) dating. Details
by the strength of the ESR signal: a weak signal, for example, of the method and its applications can be found in Gosse
could result in a precision of no better than ± 100 per cent. & Phillips (2001), Ivy-Ochs & Kober (2008), Dunai (2010)
Overall, therefore, ESR dating is a technique that seldom and Balco (2011).
generates ages with errors of < 10 per cent (Rink, 1997). In addition to surface exposure dating, however,
CRN dating can be used to estimate weathering and erosion
rates and therefore to develop timescales for long-term
5.3.7.3 Some applications of ESR dating
landscape evolution, and also to determine ages for buried
ESR dating has a greater time range than many of the depositional sequences. The first approach involves model-
techniques discussed in this book, extending from a few ling measured surface concentrations of cosmogenic
thousand years to about 2 Ma in the case of tooth enamel. nuclides in an eroding rock outcrop, and has been used to
Most of the applications on other materials lie in the establish site-specific denudation rates in different geo-
time range between 20 and 700 ka. Examples include the morphological contexts (Cockburn & Summerfield, 2004).
dating of glacial sediments in western China up to 710 ka Cosmogenic burial dating uses a pair of nuclides (usually
10
of age (Zhou et al., 2002); of teeth from a site of Hoxnian Be and 26Al) that are produced in a rock or mineral,
interglacial age (MIS 11) in eastern England (Grün & but with different decay constants. On an exposed surface,
Schwarcz, 2000); of the first human settlements of the the concentrations of the two nuclides conform to the
Loire Basin, western France, around 1.1 Ma (Voinchet et production ratio. Once buried, however, nuclide produc-
al., 2010); of molluscs associated with Palaeolithic artefacts tion ceases and the inventories of both nuclides decline due
in Egypt of MIS 2 and MIS 4 age (Blackwell et al., 2012); to radioactive decay. As they decay at different rates, the
and of quartz phenocrysts from the Mt Toba super- ratio between the two nuclides gradually diverges from the
eruption where the ESR ages (mean value of 81 ± 17 ka) surface production ratio. Hence by measuring this ratio in
accord with the previously derived argon isotope ages of a body of sediment, the time when these nuclides were first
74 ± 4 ka (Wild et al., 1999). Dating of corals from the shielded from the effects of cosmic rays (i.e. buried beneath
Netherlands Antilles has shown not only that the method a surface overburden) can be determined (Dehnert &
is appropriate for the dating of older Pleistocene corals, but Schlüchter, 2008).
DATING METHODS 295
5.3.8.2 Measurement and problems usually determined on a particular mineral such as olivine
or garnet; 14C, 10Be and 26Al are measured on pure quartz;
For surface exposure dating, samples 1–2 cm in thick- whereas 36Cl analyses are generally undertaken on whole
ness are chiselled off exposed rock surfaces or off boulders rock samples.
lying on the rock surfaces or on deposits such as moraines. There are two principal sources of error in CRN dating,
Sampling sites should be horizontal or near-horizontal the first relating to factors affecting the CRN signal in the
surfaces where there are no obvious signs of weathering or dated rock surface, and the second to the determination
erosion in order to ensure that the layers of sampled rock of CRN production rates. In terms of the first, it is assumed
contain representative nuclide concentrations. that there has been no erosion or weathering of the surface
A number of cosmogenic nuclides are now commonly since the time exposure. On bedrock surfaces, this can be
used as a basis for dating (Table 5.4). Beryllium-10 (10Be), checked by replicate measurements, but it is a problem
carbon-14 (14C), aluminium-26 (26Al) and chlorine-36 in other contexts, for example in the dating of boulders
(36Cl) are radioactive and are measured using AMS, while on moraines where erosion and exhumation of fresh
the ‘noble gases’ helium-3 (3He) and neon-21 (21Ne) are boulders can cause an under-estimate of age (Putkonen &
stable and can be measured by mass spectrometry. The Swanson, 2003). Under-estimates of age may also result if
different half-lives of the various nuclides mean that they the surface has been shielded from cosmic rays by snow
are appropriate to different time ranges. For the dating cover or by soil and/or sediment (Benson et al., 2004).
of surfaces less than 5 ka in age, 3He and 36Cl are most It must also be assumed that the surfaces have acted as
generally employed, although technical advances in extrac- closed systems since exposure and that there has been
tion and measurement suggest that in situ produced 14C no loss of nuclides or contamination by others. Similarly,
may become more widely used in the future (Naysmith it has to be assumed that there is no inherited signal from
et al., 2004). For longer time periods, 10Be, 21Ne, 26Al and earlier exposure events; in other words, the ‘exposure
36
Cl are more applicable. Cosmogenic 3He and 21Ne are accumulation clock’ has been effectively zeroed. This may
Table 5.4 Summary of characteristics and scope of nuclides used in surface exposure (CRN) dating
(from Ivy-Ochs & Kober, 2008). Production rates (atoms per grams per year) are from Gosse & Phillips
(2001).
Nuclide Half-life Other Target Production Advantages Disadvantages
isotopes elements rate
10 9
Be 1.51 Ma Be O, Si 5 Quartz resistant; Low production rate;
10
ubiquitous Be interference in AMS;
generally restricted to
quartz
26 27
Al 716 ka Al Si 31 High production rate; Restricted to quartz (low Al);
quartz resistant accurate measure of 27Al
ubiquitous required
36 35
Cl 301 ka Cl Ca, K, 10, granite Low detection limit; Complex production;
37 35 36
Cl Cl 20, limestone low AMS background; S interference in AMS;
any rock type; accurate measure of total Cl
silicates and carbonates required
14 12
C 5.73 ka C O 16 Useful for short time- Short half-life;
13
C scales; quartz resistant; atmospheric 14C
ubiquitous contamination
3 4
He Stable He Many 120 High production rate; Diffuses out of quartz or
useful for long time- volcanic groundmass;
scales; pyroxene; olivine corrections required;
pre-exposure possible
21 20
Ne Stable Ne Mg, Si 20 Useful for long time- High air background possible;
22
Ne scales (> 50 ka); corrections required;
quartz, olivine, pyroxene pre-exposure possible
296 DATING METHODS
be a particular problem in the dating of glacial and fluvial 5.3.8.3 Some applications of CRN dating
sediments and careful field sampling is needed to resolve
the difficulty. CRN dating has evolved rapidly over the past two
Obtaining a reliable estimate of CRN production rates decades, and the applications of the technique have been
is also problematic as these are known to vary with alti- extremely varied (Dunai, 2010). Some of the earliest uses
tude, latitude, depth below ground surface and degree were in glacial geology to determine surface exposure
of shielding from cosmic rays (Stone, 2000; Gosse & ages of moraines and glaciated surfaces (Figure 5.16), and
Phillips, 2001). Changes in past cosmic ray flux also have this approach has since been extensively applied in the
to be taken into account (Florinski et al., 2004). In order development of glacial and deglacial chronologies (Kelly
to address these issues, CRN production rate calibration et al., 2004; Balco et al., 2013), and in developing con-
datasets have been developed, including the globally straints on the lateral and vertical dimensions of ice
averaged dataset of Balco et al. (2008) which can be accessed sheets (Ballantyne et al., 2008; Fabel et al., 2012). Other
applications include the dating of landslide and rock
online (http://hess.ess.washington.edu). This enables sci-
avalanche deposits (Ivy-Ochs et al., 2009), of alluvial fans
entists to calculate exposure ages and erosion rates and to
(Machette et al., 2008), of pluvial lake shorelines (Kurth
compare previously published exposure ages or erosion
et al., 2010), of granite tor forms (Gunnell et al., 2013)
rate measurements on a common basis. However, because
and of river terrace sequences (Rixhon et al., 2011). The
of the marked geographical variations in CRN produc-
cosmogenic burial approach has been less widely used than
tion, regional calibration sets are now being generated
surface exposure dating, but has been successfully applied
for different areas of the world, for example for western
in a number of depositional contexts including cave
Norway (Goehring et al., 2012), Greenland (Briner et al.,
(Anthony & Granger, 2007) and fluvial sequences (Dehnert
2012) and the Baffin Bay region of Arctic Canada (Young
et al., 2011).
et al., 2013). A further complication is that some cosmo-
Cosmogenic radionuclides are also found in glacier ice.
genic isotopes may occur naturally in rocks. For instance,
36 These are not produced at the earth’s surface but by
Cl may be produced in rocks as part of the decay series
spallation processes in the atmosphere (as in the case of
of uranium and thorium isotopes, while cosmogenic 14
C), and they become incorporated into glacier ice via
nuclides can occur in rocks not as a result of spallation
precipitation. They can be used in the dating of older ice
reactions, but by direct adsorption from the atmosphere.
(greater than 100 ka); for instance, a 36Cl age of c. 760 ka
In both instances, those nuclides have to be identified and
was obtained from near the base of the Guilya ice cap
their presence corrected for.
on the Qinghai-Tibetan Plateau (Thompson et al., 1997),
All of the above mean that there are, at present,
but they are more widely used in the correlation of ice-
considerable uncertainties in CRN ages, which are typically
core records. An example is the clearly defined peak in
in the range 10–20 per cent. When these are added to the 10
Be concentration at c. 41 ka which has been used to link
analytical errors arising from laboratory measurement Greenland and Antarctic ice cores (Raisbeck et al., 2007).
(7 per cent or less for AMS; ± 3 per cent for mass spectro- Because cosmogenic radionuclide production in the
metry), this means that CRN dating is the least precise of atmosphere is influenced by changes in the earth’s geo-
all the radiometric methods considered so far. However, magnetic field, past variations in the geomagnetic field can
in view of the ubiquity of rock surfaces (e.g. by compar- also be reconstructed from the record of cosmogenic
ison with ancient organic deposits), the potential of the radionuclides preserved in glacier ice (Wagner et al., 2000).
technique is clearly considerable and hence sustained In addition, since cosmogenic radionuclide production
efforts are being made to improve both the accuracy and is also modulated by solar activity (the ‘solar wind’),
precision of CRN dating. A major initiative is the CRONUS fluctuations in solar activity can be inferred from the record
(Cosmic-Ray Produced Nuclide Systematics)-Earth Project, of cosmogenic radionuclides, such as 10Be in polar ice
an international consortium of scientists from Europe (Vonmoos et al., 2006).
and North America, whose goals are not only to develop
and refine the calibrations for CRN production rates
referred to above, but also to evaluate the sources of
5.3.9 Short-lived radioactive isotopes
uncertainty in CRN production with a view to reducing Radioisotopes with much shorter half-lives that have been
current levels of uncertainty in measurements to around 5 used in dating Late Quaternary events include lead-210
per cent. If this could be achieved, it would bring CRN (210Pb: half-life 22.26 yr) and caesium-137 (137Cs: half-life
dating more into line with other radiometric techniques 30 yr). These have, in the main, been applied to the dating
currently employed in Quaternary dating. of lake sediments that have accumulated over the past two
DATING METHODS 297
Figure 5.16 a) Glacial moraine near Leh in the Indus Valley, Ladakh, northern India. b) 10Be ages on the boulders from the moraine
surface indicate formation during the Leh glacial stage (Owen et al., 2006) which dates to 317±57 ka, that is, in MIS 9/10 (Dortch
et al., 2013; photographs by Mike Walker).
210
or three centuries, although other depositional contexts Pb has been widely used in the dating of recent lake
have also been dated. A third radioisotope, silicon-32 (32Si), sediments, and especially in studies of human impacts
has a longer half-life (178 ± 10 yr) and is applicable over on lake catchments and ecosystems (Oldfield et al., 2003).
a timescale of 30–1,000 years. Dating of recent sediments It has also been used to date recent peat accumulations
using these various isotopes is reviewed by Appleby (2008). (Hendon & Charman, 2004), salt-marsh sediments (Kemp
et al., 2012b) and marine deposits (Zillén et al., 2012), while
210
Pb dates have also been obtained from ice cores (Eichler
5.3.9.1 Lead-210 et al., 2011).
Lead-210 dating uses the radiogenic isotope 210Pb, which
forms in the atmosphere from terrestrial radon-226 (226Ra).
5.3.9.2 Caesium-137
The ‘unsupported’ 210Pb is subsequently removed from the
137
atmosphere by precipitation to accumulate in lacustrine Cs is an artificially generated radioactive nuclide that
and marine sediments, and in soils, peats and glacier ice, has only been produced in significant quantities as a result
where it subsequently decays to stable 206Pb. Measurement of thermonuclear weapons testing which began in 1945.
of the remaining ‘unsupported’ 210Pb as a function of depth The first pronounced atmospheric increase in the Northern
in a body of sediment provides an estimate of the time that Hemisphere was detected in 1954 and a clear maximum
has elapsed since the lead was deposited (up to a maximum in 1963, after which atmospheric concentrations declined
of c. 150 years, i.e. 5–7 half-lives), and hence enables significantly with successive nuclear test-ban treaties.
the rate of sediment accumulation to be established. The The 1963 maximum, however, shows up clearly in many
main problem encountered with this method is that most recent lacustrine sequences, and forms a distinctive time-
sediments contain small amounts of 210Pb derived from the stratigraphic marker horizon. In Europe, the 137Cs peak due
decay of 226Ra, and this ‘supported’ 210Pb must be deter- to the Chernobyl fallout (1986) is more strongly marked
mined and subtracted from the ‘unsupported’ 210Pb pro- than the weapons peak, but the latter is global whereas the
duced in the atmosphere (Appleby & Oldfield, 1992). Chernobyl signal is regional. Applications of 137Cs include
Other difficulties arise from the fact that the fluxes of 210Pb the dating of pollution histories of lake sediments (Rember
and its carriers are site-specific and may vary through time, et al., 1993) and the determination of rates of erosion on
while bioturbation and reworking of sediments provide archaeological sites (Davidson et al., 1998). 137Cs has also
further constraints on the technique. Particular problems been used as an independent check on 210Pb chronologies
are encountered in the dating of peat sequences as a result (Abril, 2003).
of erosion and sediment loss in the upper parts of the profile In addition to 137Cs, other artificially generated iso-
and the downwash of 210Pb in the lower levels (Oldfield topes, including plutonium-239 (239Pu: half-life 2.4 ×
et al., 1995). 104 yr), plutonium-240 (240Pu: half-life 6.5 × 103 yr) and
298 DATING METHODS
americium-241 (241Am: half-life 432.7 yr), also show 1963 5.4.1 Dendrochronology
maxima and can be used in the dating of recent peat and
lake sediment sequences (Böllhofer et al., 2004). 5.4.1.1 General principles
In most softwood (coniferous) trees, new water- and
5.3.9.3 Silicon-32 food-conducting cells (tracheids) are added to the outer
Cosmic-ray produced 32Si is removed from the atmosphere perimeter of the trunk each growth season following an
by precipitation and accumulates in lake and marine inactive period in winter. The new cells that begin to grow
in the spring tend to be larger and more thin-walled than
sediments, and also in snow. It can therefore be used in the
those produced in late summer, as a result of heavier
dating not only of sediment sequences, but also of glacier
demands on water supply early in the growth season.
ice. The principal difficulty in 32Si dating relates to the
Later in the year, the cells become gradually smaller and
detection of 32Si because of its very low natural specific
develop thicker walls. There is normally, therefore, a dis-
activity compared with stable Si. Hence, in the dating of
tinct ‘line’ between successive annual increments of wood
sediments, stringent radiochemical purification procedures
growth (Figure 5.17), and counting of these lines (tree
are required. AMS can be used for samples of rain, snow
rings) allows the age of the tree to be established. In hard-
and ice, but for limnic and marine sediments where
wood (deciduous) trees, growth trends are more variable,
biogenic silica (diatoms and radiolaria) forms the dating
however. They can be divided into ring-porous types where
medium, the 32Si/Si ratio is below the detection limit for the spring vessels are normally distinctly larger than those
AMS, and hence beta radiation radiometry has to be of the summer wood (e.g. oak, ash, elm) and diffuse-
employed (Morgenstern et al., 2000). porous types in which the pores are more uniform in size
Applications of 32Si include the dating of glacier ice (e.g. beech, birch, alder, lime). The result is a considerable
(Morgenstern et al., 2000) and the dating of recent difference in the nature of tree rings between species, and
lacustrine and marine sequences (Nijampurkar et al., 1998; as some do not display clearly defined annual bands, not
Suckow et al., 2001). The applicable timescale of 32Si dating all trees are suitable for dendrochronology. The most
(30–1,000 years) means that it bridges the time gap between widely used are oak (Quercus) and pine (Pinus), although
the present and the younger end of the 14C range (Fifield tree-ring chronologies have also been obtained from
& Morgenstern, 2009), and it could therefore be especially Sequoia and Douglas fir (Pseudotsuga). The principles and
valuable for archaeologists in the dating of historical events applications of dendrochronology are described by Baillie
over the course of the last millennium (Morgenstern et al., (1995), Schweingruber (1996), Fritts (2001) and Speer
2001). (2010), while the journal Dendrochronologia provides a
regular update on recent developments in the technique.
5.4 INCREMENTAL DATING 5.4.1.2 Measurement and problems

METHODS
Subfossil or dead trees may be obtained from a range of
Incremental dating methods are those based on regular contexts: they may be buried in peats, in riverine deposits
additions of material to organic tissue or to sedimentary or on archaeological sites, or they may form part of a
sequences. Those which have been most widely used are standing or former building. In all of these cases, the wood
dendrochronology (tree-ring dating) and varve chron- can be cut to enable a complete cross-section to be exam-
ology, which is based on annual accumulations of sedi- ined, allowing comparisons of ring width to be made in
ments in lakes or in the sea. Annual layers in glacier ice also several radial directions. Living trees can be sampled with
form a basis for dating. Measurements of variations in a metal increment corer which extracts small-diameter
lichen size (lichenometry) can be used to determine the age cylinders of wood from the tree trunk. In the laboratory,
of exposure of rock surfaces, while annual increments of samples are dried, polished and mounted prior to exam-
calcite can also be detected in speleothems. Annual banding ination, or else cut with a sharp blade while still damp.
in corals and molluscs enable short-term chronologies Counting and measuring can be carried out by visual
(sclerochronologies) to be established in marine environ- inspection of the rings under normal magnification, or
ments. on a moving stage under a binocular microscope. Other
approaches employ electronic measuring equipment
linked to both micro- and main-frame computers, or
DATING METHODS 299
rings
responses in the trees. The most important determining
a)
factor for many trees is climate. Under conditions of
transverse^
section
""medullary stress, growth is retarded, and a narrow tree-ring will
rays
result; conversely, under more favourable conditions,
^radial growth rates are increased, and wider annual rings are
section-
produced. As a consequence, climatic variations over very
short and precisely dated timescales can often be inferred,
heartwood or a field of study known as dendroclimatology. In temperate
duramen
areas, such as northwest Europe, there is usually sufficient
moisture available all year round, and hence the primary
sapwood limiting factor to tree growth is temperature (accumulated
cambiun
tangential
summer degrees). In semi-arid regions, such as parts of the
b} section southwest USA, moisture availability is more important.
Climatic changes over time are reflected in a ring-width
pattern within which distinctive rings (representing
particularly good or bad growth years), or characteristic
groups of rings, form markers and these, in turn, form a
basis for cross-matching or cross-dating between wood
of overlapping ages. Tree-ring records can therefore be
extended back in time and a master chronology developed
(Figure 5.18). Where wood samples are recovered from peat
bogs, for example, or from old buildings, counting of the
tree-ring sequences enables a floating chronology to be
established. This can be matched to the master chronology
by cross-dating and thus wood of hitherto unknown date
can be assigned a precise calendar age.
b}
Trees tend to grow more vigorously in youth than in
old age, and as a result there is usually a reduction in
ring-width with age. This may cause difficulties in cross-
dating for the climatically induced ring-width variations
can be masked by ring-width variations due to age. A
further problem is that the width of tree rings tends to
vary with the height of the trunk, and there is no way of
knowing precisely where on the trunk a piece of subfossil
wood originated. Each tree-ring series is therefore
standardized, by transforming the measured ring-width
values to ring-width indices, for example by fitting a
regression line to the measured ring-width values to pro-
Figure 5.17 a) Cross-cut of a tree trunk showing seasonally vide an indication of the general decline in tree-ring
differentiated growth rings, reflecting variations in wood cell width with age (see Figure 5.19). Further difficulties arise
density. b) X-ray negatives of wood surfaces can be magnified from the lack of variations in ring width, from intermittent
and analysed using a high-resolution microdensitometer: the absence of rings, or from the presence of false rings.
optical density of the X-ray negatives is inversely proportional
to wood density, as illustrated schematically in c). Where trees are growing in situations where there is
little or no climatic variation or where minimal climatic
stress is exerted on the trees, ring widths may show little
variation through time. A tree-ring series of this type
X-ray densitometry which detects annual variations in is termed a complacent series, and is of little value in
wood density by degree of penetration of X-ray beams. dendrochronology as the distinctive markers that form
Tree-ring width is seldom uniform, for tree growth is a basis for cross-dating are absent. Dendrochronologists
influenced by a range of environmental factors, varia- therefore prefer to select trees from stressed situations,
tions within which will produce different physiological where some climatic factor has been critical to growth.
300 DATING METHODS
1.0¬ Correlation between samples (RESAR)
0.5¬ R = 0.40
0.0-
PIT3125
RIF3185
KOM6707
KOM6712.
KOM6810
KOM6820
KOM6820
KOM6750.
KOM6740:
KOM672 4
KOM6827
KOM6530
KOM6131
KOM6814
KOM6717
KOM5975
KOM6143
KUL3006.
NAK4912i
PIT6568
LU03122«
LU05624.
PIT4963
RIE5640
KOM6706
NAK6Q47<
VAL5031.
VAL5022 -
KUMb54b i
VAL6087
-1200 -1000 -800 -600 -400 -200 0 200
Year scale
Figure 5.18 Age profile of cross-matched tree-ring series from Finnish Lapland for the period AD 1200–160 BC, showing part of
the record with many overlapping tree-ring series (after Eronen et al., 2002). The complete record of which this forms a part, and
which extends over the past 7,519 years, is shown in Figure 5.22.
Wood production will be reduced at times of stress, and this 5.4.1.3 Dendrochronological records
is shown by a sensitive series of rings. However, if the stress Dendrochronological records are now available from
is too extreme, there is always the possibility that during many parts of the world (Briffa, 2000), and form the basis
certain years a tree may fail to manufacture new cells, for dating environmental change, as well as providing a
or may only produce new material on restricted parts of chronology for climatic and environmental reconstructions
the trunk. The result is missing or partial rings. On the derived from dendroclimatological data (section 5.4.1.4).
other hand, where the spring growth period is interrupted They are also key to the calibration of the radiocarbon
or curtailed by late frosts, it is possible for more than one timescale (section 5.3.2.6). Moreover, as tree-ring sequences
growth band to develop in a particular year (false rings contain a record of long-term atmospheric 14C variations,
or intra-annual growth bands). These may not be easy to they offer insights into the forcing mechanisms (solar
identify in individual tree-ring series, and can only be variability, ocean circulation variations, geomagnetic
corrected for by replication of a number of records through changes, etc.) that influence atmospheric 14C variability over
systematic cross-dating. both long and short timescales (Hua et al., 2009).
DATING METHODS 301
cm
2.0- a) Measured ring width values
1.0-
2.0 b) Ring width indices
1.C
Figure 5.20 Ancient bristlecone pines growing in the White
Mountains of California (photograph by Mike Walker).
n
450 500 500 600 650 700

Age AO longer chronologies may be developed, as floating chron-
ologies from subfossil remains of Quercus from the central
Figure 5.19 Standardization of ring-width measurements to USA have been radiocarbon dated up to 13.8 ka cal. BP
generate ring-width indices. a) A linear regression line has been (Stambaugh & Guyette, 2009), while in New York State,
fitted to the measured ring widths and this provides an
indication of the overall decrease in ring width with age. The
floating chronologies from Picea have been radiocarbon
value for each year is then divided by the value for that year dated back to c. 14.5 ka cal. BP (Griggs & Kromer, 2008).
obtained from the regression curve. b) The derived ring-width
indices (after Baillie, 1982). Western Europe
In western Europe, there are no trees with an age range
comparable to the bristlecone pine, and hence long chron-
North America ologies have only been established by the laborious cross-
In the arid mountain environments of the southwest dating of large numbers of subfossil wood samples whose
United States are the oldest living trees in the world. The ring patterns often span no more than 100–200 years. The
bristlecone pine (Pinus longaeva, formerly Pinus aristata) most commonly used species are the oaks (Quercus robur
thrives on dry and rocky sites up to 4,000 m and in the and Quercus petraea) and the Scots pine (Pinus sylvestris),
White Mountains of California, in the rain shadow of the obtained from fen and raised bog peats, or from river
Sierra Nevada, large numbers of twisted and stunted gravels. In the British Isles, the longest chronologies so far
bristlecone pines are found (Figure 5.20), with the famous developed have been on oak, with a record from Ireland
‘Methuselah Tree’ being more than 4,700 years old. These extending back 7,429 years (Brown & Baillie, 1992). In
trees are growing under particularly stressed conditions, central Europe, a continuous oak chronology has been
and the limited growth period (perhaps only one or two constructed from the remains of oak found in the riverine
months each year) produces very narrow rings and a ring deposits of southern and eastern Germany (Figure 5.21),
width that is highly sensitive to climatic change (Johnson, the oldest oak in the German Hohenheim master chron-
1999). By cross-dating between living and dead wood ology being dated at 10,429 years (Friedrich et al., 2004).
and then between subfossil samples, a continuous master By linking the sequence based on oak to a chronology based
chronology has been developed which extends back to on pine, the European dendro-record has now been
8681 years BP (Ferguson & Graybill, 1983). This has since extended back to 12,594 years; in other words, beyond the
been extended to 6828 BC (8778 BP) and may eventually onset of the Holocene and into the Younger Dryas period
link with an older 3,000-year floating chronology (Harlan (Hua et al., 2009). For the Lateglacial, two floating
& Robertson, 2006). Other relatively long tree-ring chron- chronologies (the Hohenheim series: 1,116 rings; the Swiss
ologies have been established in North America on species Dättnau series: 1,606 rings) have been synchronized into a
of Sequoia, Douglas fir (Pseudotsuga), and pine, the longest combined chronology, the youngest section of which
of which, based on foxtail pine (Pinus balfouriana), extends forward into the Younger Dryas (Kromer, 2009;
presently extends back over 3,000 years. In due course, even Kaiser et al., 2012). If these floating records can, in due
302 DATING METHODS
Figure 5.21 a) Fossil oak trunks excavated from gravel beds of the River Danube valley, with dendrochronologist Michael Friedrich
of the University of Hohenheim, Germany. b) A section through one of the trunks of subfossil oak, with the late Bernd Becker,
one of the pioneers of European dendrochronology (photographs by Bernd Kromer, University of Heidelberg, Germany).
course, be linked to the Hohenheim master chronology, this back 3,722 years (Boswijk et al., 2006). More importantly,
will extend the continuous European dendro-curve back perhaps, the remains of subfossil Kauri have been found
beyond 14,000 years. For earlier periods, there are floating buried in peats that pre-date the Last Glacial Maximum,
pine and larch chronologies from northern Italy, the oldest and a preliminary group of chronologies has been devel-
of which dates to around 17,500 years (Kromer, 2009). Long oped that combine to produce a floating chronology
tree-ring records based on pine have also been constructed spanning 10,719 years of MIS 3 (Palmer et al., 2006). This
for northern Scandinavia, with a 7,400-year chronology tree, therefore, not only has potential for the construction
from Swedish Lapland (Grudd et al., 2002) and a 7,500-year of long and continuous Holocene dendrochronological
chronology from Finnish Lapland (Figure 5.22). In both series but it may, as discussed above (section 5.3.2.6), also
North America and Europe, there are shorter chronologies offer a basis for a dendrochronological calibration of the
spanning the last 1.0–1.5 ka (Watson & Luckman, 2001; radiocarbon timescale well beyond that provided by the
Gunnarson & Linderholm, 2002). Floating chronologies Northern Hemisphere tree-ring series.
have been used, inter alia, to date prehistoric trackways
(Hillam et al., 1990), ancient boat timbers (Nayling &
5.4.1.4 Dendroclimatology
McGrail, 2004) and prehistoric tombs (Panyushkina et al.,
2007). Dendroclimatology is the science of reconstructing past
climatic conditions and histories from tree rings. The
Australia and New Zealand importance of dendroclimatological records is that palaeo-
Although long continuous tree-ring chronologies compar- climatic information can be precisely dated, and inferences
able with those from North America and Europe have yet can even be made about seasonal variations (Linderholm
to be developed from the Southern Hemisphere, some et al., 2010). Further, as a proxy record of climate, it enables
trees offer the potential for the construction of long the record to be extended beyond those based on
dendrochronological records. These include the Huon instrumental measurements (Hughes et al., 2011).
Pine (Lagarostrobus franklinii) from Tasmania from which Palaeoclimatic reconstructions based on tree rings rest
a 4,136-year annual tree-ring chronology has been recon- on assumed or demonstrable relationships between ring
structed, and the Silver Pine (Lagarostrobus colensoi) from width, or some other ring characteristic, and climatic
the South Island of New Zealand which has yielded a parameters. These relationships are often complicated by
cross-matched chronology extending back to 323 BC (Cook lag effects between climatic inputs and tree response, for
et al., 2006). Of particular significance in terms of its trees have the ability to store food reserves and water that
dendrochronological potential, however, is the Kauri may then be used in adverse years. Once the linkages
(Agathis australis), a conifer which grows on the North between climatic and tree-growth parameters have been
Island of New Zealand. This long-lived tree (over 500 established, however, inferences can be made about past
years) has produced a continuous chronology that extends climate for the timespan of the dendrochronological record.
DATING METHODS 303
1.5
1.0
0.5
0.0
-5500 -5000 -4500 -4000 -3500 -3000
1.5¬
10¬
0.5¬
0.0-
-3000 -2500 -2000 BEUj -1000 -500
15 -I
1.0 -
05 -
0.0 -
500 0 soo 1000 1SOO 2000
yr BC/AD
Figure 5.22 Part of a continuous 7,519-year pine tree-ring chronology from Finnish Lapland. The average ring width remains
remarkably constant at about 0.6 mm yr–1 throughout the whole length of the chronology. The thin line indicates annual ring-width
variability (in mm) and the bold line the 50-year running mean (after Eronen et al., 2002).
Dendroclimatology has become an integral tool in the particularly during the summer, they offer the prospect for
study of environmental change, particularly over the course finer resolution of climatic relationships than ring widths
of the last millennium, allowing inferences to be made which integrate the effects of conditions over several
about temporal variations in summer temperature (Helama seasons. Studies from North America (Wilson & Luckman,
et al., 2002), precipitation regimes (Watson and Luckman, 2003), Europe (Büntgen et al., 2006) and China (Wang
2001), summer drought patterns (Zhang et al., 2004b), et al., 2010) have shown that late-wood width and maxi-
extreme arid events (Stahle et al., 2007), linkages between mum density appear to be closely related to climate, while
streamflow and climate (Cleaveland, 2000), and relation- dendrochronological records from Tasmania suggest a link
ships between climatic change and late Holocene volcanism between maximum density and regional streamflow at the
(Gervais & MacDonald, 2001). Moreover, because tree-ring end of the growing season (Allen et al., 2012). Although
sequences provide very high-resolution records of climate much still remains to be learned about the precise nature
change, they constitute one of the key sources of inform- of the climatic signal in wood density variations, density
ation for the reconstruction of short-term climatic events, measurements are becoming increasingly widely employed
such as the Pacific Decadal Oscillation (PDO) or the North alongside conventional ring-width values in dendroclimatic
Atlantic Oscillation (NAO: section 7.6.4.3). research (Briffa, 2000; Mannes et al., 2007).
An alternative approach to dendroclimatology involves A further climatic index derived from tree-ring data is
variations in wood density and, in some instances, this may based on stable isotope ratios of carbon, hydrogen and
provide clearer palaeoclimatic information than simple oxygen in wood cellulose, a field of research known as
ring-width measurements. As wood densities respond isotope dendroclimatology. The carbon isotope composi-
primarily to conditions during restricted growth periods, tion in cellulose of trees is determined by the ratio between
304 DATING METHODS
the CO2 concentrations in the cell fluid and that in the but needs to be combined with experimental studies of
atmosphere. Warmer and drier conditions generally lead isotopic pathways and exchanges during cellulose synthesis
to more restricted CO2 availability and to an enrichment in living trees (Barbour, 2007).
in the heavier isotope of carbon, 13C, relative to 12C. The
D/H and 18O/16O ratios in tree rings are largely determined
by the isotopic signature of the source water available for
5.4.2 Varve chronology
tree growth. If this is mainly meteoric water (rainwater),
5.4.2.1 The nature of varved sediments
then the isotopic ratios in the tree rings reflect the isotopic
effects in the regional precipitation, and in mid- and Rhythmical sediment layers in lacustrine deposits are
high-latitude regions, these are closely related to surface common in the geological record and are generally referred
air temperatures. Hence δ13C, δD and δ18O values in tree- to as rhythmites. Where these reflect an annual cycle of
ring series can be employed as proxy climatic indicators accumulation, they are termed varves, from the Swedish
(Tang et al., 2000). Examples include the reconstruction varv for lap, turn or revolution. They usually consist of
of precipitation patterns over the last millennium in the couplets of alternating sediment colour, grain size or fabric,
Canadian Rockies (Watson & Luckman, 2001), the resulting from seasonal differences in depositional pro-
detection of a mid-Holocene warm and dry phase in cesses. Clastic varves are minerogenic and reflect variations
northwestern Russia (Boettger et al., 2003), and the in rate or type of sediment influx, the best known being
development of a record of temperature and sunshine for glacigenic varves which accumulate in lakes close to a
northern Sweden over the past 100 years (Loader et al., glacier margin. Organic varves form as a result of seasonal
2013). The majority of these studies have been carried out changes in the deposition of organic materials, for example
on coniferous trees growing in remote high-latitude/high- alternations between microfossil-rich summer accum-
altitude areas, but recent results from west Wales, where a ulation formed during plankton ‘blooms’, and lower-
160-year palaeoclimate signal has been obtained from productivity winter layers (Figure 5.23). Chemical varves
sessile oak (Quercus petraea) and which correlates closely develop where the seasonal contrasts in the water column
with the central England temperature record, suggest that (e.g. in oxygen content or temperature) lead to the pre-
it may be possible to derive robust palaeoclimate records cipitation of different chemical compounds. Some varves
from trees in more temperate regions, and particularly from may be the product of all three of these seasonal influences.
the well-replicated oak chronologies of western and central Where varved sequences can be dated by independent
Europe described above (Young et al., 2012). means (e.g. by radiocarbon), then the events and intervals
Despite these encouraging results, there are a number that they represent can be estimated in calendar time,
problems in isotope dendroclimatology that remain to be potentially at an annual resolution. Even if their calendar
resolved, particularly in relation to the nature of the age is unknown, they nevertheless provide a ‘floating
fractionation and equilibrium processes that determine chronology’ (section 5.4.1.2) for the interval of time over
isotopic ratios in tree rings. For example, it appears that the which the varved sequence has accumulated.
stable carbon isotope signal in tree-ring cellulose from
northern pines reflects a combination of stomatal conduc-
5.4.2.2 Clastic varves
tance (controlled by air humidity and soil moisture status)
and photosynthetic rate, the latter controlled primarily by Large quantities of sediment are deposited in proglacial
sunshine and air temperature. As the balance of these lakes and shallow seas from rapid ice melt during the
controls varies spatially, and perhaps also locally, during the spring and early summer. The coarser particles settle
Holocene, without some independent estimate of either quickly, but the silts and clays remain in suspension
stomatal conductance or photosynthetic rate, the δ13C throughout the summer season as the water column is
value alone cannot provide a clear and consistent climatic agitated by surface winds and currents. During winter,
signal. One solution is to adopt a multi-proxy approach however, when the lake is frozen, the smaller particles
in which stable isotope data are analysed alongside other settle out to form a fine silt- or clay-dominated layer that
potential climatic proxies (ring width, wood density, contrasts with the coarser summer layer, thereby forming
etc.). By combining the proxies, the strength of climate glacigenic varves (Figure 5.23). Clastic varves can also form
correlations is increased and the range of extractable in non-glacial lakes that receive a mixed-grade sediment
climatic parameters extended. This type of study provides supply and freeze over in winter.
a powerful new means of extracting climatic information The potential of glacigenic varves as a basis for dating
from long tree-ring chronologies (McCarroll et al., 2003), was first recognized by the Swedish geologist Gerard De
DATING METHODS 305
Clasti c Varv e Organi c Varv e
Winter
Winter
Fall
Summer
Summer
Sprin g
Calcite crystals
Sand /coar s
Sand/coarse silt
Centric diatoms
; Fine silt/clay
Pennate diatoms
Crtrysophyte casts
Figure 5.23 Simplified model of clastic and organic varve formation, showing variations in sedimentological and biological
characteristics of annual layers (after Sturm & Lotter, 1995).
Geer who, as early as 1884, was investigating the exposed the clastic couplets may not necessarily represent annual
varve sequences in the Stockholm area (Petterson, 1996). incremental accumulations. In order to establish whether
He discovered that as the last ice sheet wasted northwards such sequences are true varves, reference may need to be
across southern Sweden, it left behind a complex of made to independent age estimates from the same sedi-
moraines and proglacial lakes. In many places, overlaps mentary sequence, or the use of stratigraphic markers of
of varve ‘histories’ had developed, the uppermost varves in known age (e.g. Lekens et al., 2005). Another approach is
one lake sequence being of the same age as those in the to employ some of the techniques described below (section
lower part of a neighbouring series. In a now classic 5.4.2.6) to compare the fine internal detail of clastic varve
publication, De Geer (1912) first presented measure- series with modern analogues (Hambley & Lamoureux,
ments of individual varves and curves of relative thickness 2006).
for each site investigated. Comparisons between sites
enabled correlations to be established using similar
5.4.2.3 Organic (biogenic) varves
principles to those now employed in the matching of ring-
width series in dendrochronology. By methodically extend- Many lakes do not contain laminated sediments because of
ing his master chronology northwards and southwards reworking of deposits by currents or by bottom-dwelling
across Sweden, De Geer compiled a long varve sequence organisms (holomictic lakes). In deeper waters, however,
which was believed to extend back continuously to the there are fewer bottom-dwelling fauna, and vertical circu-
beginnings of deglaciation. This seminal work laid the lation (caused by thermal lake stratification and its seasonal
foundations for all subsequent work on ice recession in the breakdown: see section 3.9.2) does not extend to the bottom
Baltic region (section 5.4.2.7). of the water column. In these meromictic lakes, fine
Finely graded laminated sequences have also been found laminations (organic/biogenic varves) may form. Laminae
in other ice-marginal locations, and especially in shallow that develop during the summer typically exhibit higher
marine settings (Ó Cofaigh & Dowdeswell, 2001). However, organic content, reflecting increased biological produc-
it is now apparent that the pulsed supply of sediment tivity within the lake and reduced allochthonous influx
that is responsible for the formation of the laminations from the catchment, while the reverse is the case in laminae
may not always be a seasonal phenomenon, and hence formed during the winter. Fossil assemblages (pollen,
306 DATING METHODS
diatoms, etc.) may be preserved in the summer layers 5.4.2.5 Complex varves
(Baier et al., 2004; Nakagawa et al., 2005) and, in some cases,
increments laid down in spring, summer and autumn can Some varve series have complex origins, reflecting the
be identified on the basis of these assemblages (St. Jacques various physical, chemical and biological factors that
et al., 2009). However, in many lake systems, organic varves affect productivity and deposition in lacustrine and marine
are thin and comprise only decomposed organic detritus, systems. The varves that form around the Antarctic ice
margin (section 5.4.2.3), for example, are characterized
in which case seasonal layers can only be established by
not only by contrasts in diatom content, but by seasonal
molecular and isotopic analytical methods (Fuhrmann
grain-size (clastic) alternations. Moreover, some varve
et al., 2004).
sequences that have developed over long time intervals
Similar processes lead to the formation of organic
contain different types of varve, reflecting changes in local
varves in the marine realm. Most common are diatom-
and regional environmental conditions. In Lac d’Annecy,
rich varves that develop where biological productivity is
France, for example, the varves that developed after the
high, notably where upwelling occurs (Pike & Stickley,
end of the last glacial stage are clastic, indicating low
2007). In some marine basins, such as the Gulf of California,
initial lake productivity, but these gradually change into
productivity is so high that diatom frustules form ‘mats’
chemical varves as the lake waters warmed and bio-
on the seabed (Kemp, 2003). However, as in lakes, the
logical activity increased (Brauer & Casanova, 2001).
number of fossils reaching the sea floor is likely to be low.
Similarly, in the long lacustrine record of Lake Lisan,
In the Cariaco Basin off Venezuela, for example, estimates
Jordan, that spans MIS 4 to MIS 2, there are marked vari-
suggest that only 1–2 per cent of the organic carbon gen- ations between aragonite-dominated and clastic-dominated
erated in the water column accumulates on the seabed, varved sequences that reflect cyclical changes in lake water
with perhaps less than half of that subsequently preserved depth (Schramm et al., 2000).
in the sediment record (Thunell et al., 2000). Diatom-rich
varved sediments are found throughout the world’s
oceans, and even in areas adjacent to the Antarctic ice sheet, 5.4.2.6 Sources of error in varve counting
where diatom oozes develop in response to seasonal changes Problems are encountered in varve counting that are
in nutrient availability and ocean stratification caused by similar in many respects to those in dendrochronology.
seasonal fluctuations of the sea-ice margin (Stickley et al., Local site factors can lead to incorrect estimates of age,
2005). and also cause difficulties in the correlation (cross-dating)
between individual varve sequences. For example, adverse
5.4.2.4 Chemical varves weather conditions in particular years can result in
a reduced input of sediment into a proglacial lake, or to
Water bodies vary in the degree to which dissolved reduced biomass or chemical precipitation in summer,
chemicals precipitate out to form crystalline compounds so that an individual varve fails to develop, or is too thin
(carbonates, oxides, hydroxides) which may combine to be recognized. Seasonal increments of sediment may also
with other particulate matter raining out of the water be absent because of the intermittent erosion of bottom
column to the lake or sea floor. The chemical processes may layers. Alternatively, more than one set of laminae (sub-
be seasonally affected, for example through summer–winter laminations) can develop within an annual increment
temperature contrast or degree of oxygenation, resulting in of sediment reflecting, for example, episodic sedimenta-
chemical varves that can be recognized on the basis of tion from intermittent local wind-driven currents, or two
colour or texture. In temperate lakes, sediments deposited or more periods of snowmelt during the course of a year
in summer are frequently much darker than winter (Lamoureux et al., 2001). Further complications can arise
accumulations, caused by the enrichment in iron sulphide as a result of subaqueous slumping and/or the action of
(FeS) or Fe reduction in deoxygenated water (Hongve, turbidity currents (Hambley & Lamoureux, 2006), and as
2003). Water enriched in dissolved calcium carbonate may a result of flocculation of sediment which can affect the
precipitate carbonatic compounds in the warmest months, formation of glaciolacustrine varves (Hodder, 2009).
through either inorganic or biologically enhanced processes, Chronological errors may also arise through inconsistencies
to form carbonate varves (Wittkop et al., 2009). In semi- in varve counting, either by a single operator or between
arid and arid regions, evaporitic varves may form in high- operators, although where attempts have been made to
salinity lakes where water levels are reduced during summer quantify error estimates, these may amount to no more
droughts (Zolitschka, 2007). than 1–3 per cent (Ojala et al., 2011b).
DATING METHODS 307
Overcoming these difficulties requires careful assess- section 5.4.2.2) and Matti Sauramo in Finland, glacio-
ment of each of the potential error sources. For example, lacustrine varves have continued to attract considerable
structural disturbance of sedimentary sequences may be interest as a means of dating the wastage of the last ice sheet.
identified by a survey of sediment architecture using 3-D The starting point for De Geer’s timescale was a key marker
seismic reflectors that can provide high-resolution images varve in a section in the Indalsälven Valley of north-central
of internal sedimentary structures (Vardy et al., 2010). Sweden, which he considered to reflect the sudden input
Ultra-thin varves and fine-scale sub-laminations can be of large quantities of meltwater following the bipartition of
detected using a range of imaging methods, including the residual Scandinavian ice mass. By reference to this
X-ray radiography (Ojala & Alenius, 2005) and thin-section varve year ‘zero’, which was taken to mark the change from
micromorphology (Palmer et al., 2010). Variations in glacial to postglacial conditions, older varves were given
chemical composition can be established by, for example, negative numbers and younger varves positive numbers.
ITRAX (high-resolution continuous microfluorescence- Subsequently, De Geer’s master chronology has been
X) or CAT-scan (computerized axial tomographic) analyses repeatedly revised as more data have become available,
(Guyard et al., 2007). Also, the use of replicated records while the connection of the upper part of the varve sequence
from different parts of a basin, and the employment of to the present (AD 1978) has enabled a calendrical age of
different operators in the varve counting of each core 9,238 years to be assigned to the ‘zero year’. The Swedish
sequence provide a basis for identifying and quantifying Timescale (STS) now extends back for 13,300 varve
errors in the resulting chronology (Ojala & Tiljander, years and is based on more than 1,000 inter-connected
2003). Finally, correlation between varve sequences is varve-thickness records. The youngest varves in the
increasingly being based on other parameters, including the sequence are still being deposited today in the estuary of
palaeomagnetic properties of the sediments and pollen- the Ångerman River in northeastern Sweden. However,
stratigraphic records, coupled with calibrated radiocarbon independent evidence, including AMS dating of plant
dates on plant macrofossils from the varved sequences macrofossils extracted from the varved clays, indicates
(Lundqvist & Wohlfarth, 2001; Nakagawa et al., 2005). that several hundred years are missing from the varve
chronology (Wohlfarth & Possnert, 2000). This means
that, at present, the STS is effectively a ‘floating varve
5.4.2.7 Applications of varve chronologies
chronology’. Considerable efforts are now being made to
Patterns of regional deglaciation identify and quantify the extent of the time gaps resulting
The development of a regional varve chronology span- from the missing varves, for example by wiggle-matching,
ning a long period of deglaciation has, hitherto, been an calibration and/or synchronization of the AMS 14C-dated
approach largely confined to Scandinavia. In the British intervals (Wohlfarth & Possnert, 2000); by correlating
Isles and throughout most of western Europe, relatively few the STS with independently dated records such as the
varved sequences have been recorded and analysed; Greenland ice cores (Andrén et al., 1999); and by link-
consequently, there has been little interest in the use of ing individual varve sequences using, inter alia, pollen
varves as a basis for dating ice recession, although in Britain, stratigraphy, palaeomagnetic secular variations (section
glaciolacustrine varves have recently been used as a climate 5.5.1.2) and lead-pollution chronologies (Stanton et al.,
proxy and to correlate climatically inferred events with 2010). These difficulties notwithstanding, the STS remains
those in polar ice cores (Palmer et al., 2012). In North a valuable chronological tool, particularly at the regional
America too, there are relatively few studies that have scale, and has enabled the deglacial history in parts of
involved varved sediments as deglacial chronometers. southern Sweden to be reconstructed in considerable
However, the early work of Ernst Antevs who, between the detail. It also provides an overall temporal framework
1930s and 1950s, developed preliminary varve chronologies for the deglaciation of the Baltic region as a whole (Figure
for Lake Agassiz and for the Great Lakes region has been 5.24).
revisited in recent years, and his original New England
Varve Chronology (NEVC) has been consolidated and Duration of Quaternary time intervals
recalibrated to produce a continuous (5,659-year) varve Varve chronology offers a basis for estimating the dura-
sequence spanning the retreat of the Laurentide ice sheet tion of Quaternary time intervals. For example, analysis
(18.2–12.5 ka BP), and which again has been linked to the of the annually laminated sequence in Lago Grande
Greenland ice-core records (Ridge et al., 2012). di Monticchio, Italy, indicates that the last (Eemian)
In Scandinavia, however, following on from the interglacial spanned an interval of 17.7 ± 0.20 ka (Brauer
pioneering studies of Gerard De Geer in Sweden (see et al., 2007). Varved sequences have been used to date the
308 DATING METHODS
5000
^-1200
^-12000
r-11000
10OO0
~^00 0
0
27
-1
0
27
-1
0
27
0
0
0 0
27
27
-1
27 7
-1 70
-1 -12
-1
2 0
-1 27
-1 0
27 0
-1 127 0
- 27
-1
0
27
-1
0
27
-1
0
27
-10
27 0 70
- 2 7o - 1 2
1
—5fo 70
-1 7 0 12
2 - 0
-1 - 1 7
2
27 -1
0
-1
2 7-1 27 0
0
0 - 1 0- 1 2 27
27 2 7 -1 70
-1 -1 70 0 2
27 - 1
-1 0 0
27 27 0
0- 1 -1
27 27
2°°°JSooo X -1
;
0 12400,
-1
2
• 12600
7
0-
-12700
1
2
7
0
©MAP S IN MINUTES™ 2009
Figure 5.24 The pattern and rate of ice retreat across Scandinavia based on a combination of clay-varve chronology and radiocarbon
dating. The dates from northern Sweden and along the southwest Baltic coast are based on clay-varve chronology and should be
corrected by +365 years (Cato, 1985). Other dates are based on radiocarbon (after Lundqvist, 1986, with permission from Elsevier).
While subsequent radiocarbon dating may have revised the ages of some of the retreat stages, the overall pattern of ice wastage
remains essentially unchanged.
duration of earlier European interglacial periods (Bittman consistent with the hypothesis that this was a result of a
& Müller, 1996), as well as more recent, shorter time pathogenic attack (Peglar, 1993). Varve records have also
periods, such as the Younger Dryas Stadial (Litt et al., been employed to determine the duration of former ice-
2001). Non-climatic periods can also be dated by varve dammed lakes, and different phases of their development
chronology. At Diss Mere in southeast England, the Ulmus (Palmer et al., 2010).
(elm) decline, a feature common to most western European
mid-Holocene pollen records and dated to c. 5.8 ka cal. BP, Dating Late Quaternary events
is preserved in a series of annually laminated sediments. Where annually laminated sediment sequences, which
The record shows that Ulmus pollen values fell by 73 per are effectively floating chronologies (section 5.4.2.1), can
cent over a six-year interval, such a rapid decline being be linked to calendar time by, for example, calibrated
DATING METHODS 309
radiocarbon dates, the varve sequence can provide the provided evidence for a higher dust flux in the mid-
basis for a high-resolution timescale of climatic and envir- Holocene compared with the Little Ice Age and more
onmental changes. For example, a 14,000-year compo- recent ‘dust bowl events’ (Dean, 1997). Dust and other
site carbon isotope record from Lake Holzmaar in the pollutants in varved sediments also provide records of
Eifel region of western Germany, which shows evidence of human activity. In China, Chu et al. (2009) found evidence
significant ecosystem reorganizations at around 14.2, 9.6, for intensive dust storms during the historic period (AD
5.5 and 2.7 ka, each reflecting major changes in climatic 1050–1330, 1590–1690 and over the past 200 years) which
regime, has been dated by varve chronology, the mean time they linked to expanding populations and associated
resolution for the isotope sequence being 14 years (Lücke farming activity, while Renberg et al. (2002) detected
et al., 2003). Varve chronology has also been used to increases in lead pollutants in a Swedish varve record, with
provide a similar high-resolution timescale for a palyno- markedly higher levels in Roman and Medieval (AD 1000,
logical record from a lake in the Hartz Mountains of 1200 and 1530) times, and especially in the period post
Germany, which shows land-use changes from Bronze World War II. An increase in metal pollutants over the past
Age to Roman times (Zolitschka et al., 2003). In the marine 150 years has also been observed in varved sequences in
Cariaco Basin, a seasonally resolved titanium influx record Arctic lakes (Sun et al., 2006a).
from the varved sequence shows that the collapse of the
Mayan civilization in the Terminal Classic Period occurred Palaeoseismicity
during an extended regional dry episode, punctuated by Deformed laminated deposits caused by seismic events
more intense multi-year droughts centred at approximately (seismites) provide evidence of past earthquake and
AD 810, 860 and 910 (Haug et al., 2003). associated activity. In the Dead Sea graben, there is a record
in varved sediments of major earthquakes occurring in
Climatic cycles and patterns 31 BC, and AD 749, 1033, 1212, 1837 and 1927 (Migowski
Statistical analysis of varved sequences suggests that a et al., 2004). Seismic events are also reflected in the Swedish
history of cyclical climatic activity and patterns may be varve record, with evidence for increased tectonic activity
imprinted in varve records. Holocene varves in the Santa in the Stockholm area just after the onset of ice-sheet
Barbara basin, California, for example, show evidence of a wastage (Tröften, 2000).
c. 1,000-year cycle, similar to that recorded in the Greenland
ice cores, suggesting a teleconnection (similar response to Calibration of the radiocarbon timescale
a common stimulus) between the Pacific and Polar realms The IntCal13 calibration described in section 5.3.2.6
(Nederbragt & Thurow, 2005). Some varved sequences incorporates marine-varve data from the Cariaco Basin,
appear to reflect the influence of more than one climatic Venezuela, where lighter layers, rich in the remains of
rhythm, such as Turkish lake deposits which appear to have diatoms, are formed during the winter–spring upwelling
responded to variations in both the Indian monsoon and season, and these alternate with darker laminae formed
the North Atlantic winter climate regime (Jones et al., largely of terrestrial material inwashed into the basin during
2006). Other varve records contain evidence of decadal- the late summer–autumn rainy season. The varve chron-
scale climate cycles, such as the North Atlantic Oscillation ology, which spans the time period 15–10 ka, is floating but
(NOA) (Hubeny et al., 2006; section 7.6.4.3) and the El it can be anchored to the calendrical timescale by cross-
Niño Southern Oscillation (ENSO) (Boës & Fagel, 2008; correlating radiocarbon dates obtained on planktonic
section 7.6.4.2). In addition, solar and lunar periodicities Foraminifera with the 14C record in the pine tree-ring
may be recorded in some varved sequences (Berger & von chronology (Hughen et al., 2004b). A more extended
Rad, 2002). timescale for the Cariaco varve sequence, spanning the
interval 55–15 ka, has been cross-matched with the annual-
Atmospheric dust and pollutants layer counted chronology of the GISP2 ice core (Hughen
Dust in the atmosphere can be an important factor affect- et al., 2004b).
ing climate, and varved sediments are key records for The longest continuous varve record in a terrestrial
quantifying past variations in atmospheric dust flux (Maher site is that from Lake Suigetsu, Japan (Figure 5.25) which
et al., 2010). In northeast China, Schettler et al. (2006) extends back over 70 ka (Nakagawa et al., 2012). The
analysed the annual concentrations of atmospheric dust varves consist of lighter-coloured, diatom-rich, spring–
in a varve record for the past 220 years and related these summer units alternating with darker, diatom-poor,
to seasonal variations in the East Asian monsoon regime, autumn–winter horizons. Radiocarbon dating of organic
while in Minnesota, USA, a varved sediment sequence materials from the laminated sediment sequence provides
310 DATING METHODS
b
plain light polarized light
1 yesr
l year
1 year
1 year
1 year
1 year
1 year
1 year
Figure 5.25 a) Part of a varved sequence from Lake Suigetsu, Japan. The white curves reflect measured grey-scale variations
which are proportional to sediment colour (lighter to the right). In the left-hand image, the ages of the varves range from 49,951
± 168 vyr (varve years) (top) to 50,088 ± 171 vyr (base), and in the right-hand image from 51,497 ± 201 vyr to 51,627 ± 204 vyr.
b) Photomicrographs of part of the Suigetsu varve sequence in plain light (left) and polarized light (right). The varves are composed
primarily of amorphous biogenic material and diatoms (the latter appear black in polarized light), along with chemical precipitates,
principally siderite (yellowish in polarized light) and recycled iron (Fe). While biogenic material makes up most of the sediment
generated in a year, there is a marked seasonal component in the varve record, with distinct diatom-rich layers forming in spring
and siderite layers in the late autumn. The latter are the most frequent seasonal layers in the Lake Suigetsu core, and varve counting
relies mainly on these horizons. This part of the sequence dates to c. 26.8 k vyr (photographs and micrographs by Takeshi Nakagawa,
Ritsumeikan University, Kyoto, Japan).
a comprehensive calibration based on terrestrial samples 5.4.3 Annual layers in glacier ice
for the period 52.8–11.2 ka and constitutes a unique
source of information on the connection between global 5.4.3.1 General principles
atmospheric and marine radiocarbon levels (Bronk Ramsey
et al., 2012). This remarkable record may, in due course, Annual additions of snow can be observed as clearly defined
enable even further improvements to be made in radio- layers in the upper parts of many glacier masses and these
carbon calibration (section 5.3.2.6). provide a high-resolution chronology for environmental
DATING METHODS 311
change (section 3.11). Annual layering has been observed (Figure 5.26). Identification and subsequent counting of
both in polar ice sheets (Taylor et al., 2004; Rasmussen annual layers in ice cores can be based on a single parameter
et al., 2006) and in smaller ice masses, such as those found that is known to exhibit annual cycles, but where parallel
on the summits of high mountains in the tropics (Vimeux data series with sufficient resolution are available from
et al., 2009). In addition to changes in the visual properties the same segment of an ice core, then it is preferable to use
of the ice, seasonal layering can also be detected by all of the data in the development of a chronology. This
measurements of such physical and/or chemical parameters multi-parameter approach is now routinely employed in
as stable isotope ratios (δ18O, δD), electrical conductivity the construction of ice-core timescales (Meese et al., 1997;
of the ice, dust content, microparticle content and chemical Rasmussen et al., 2006).
element composition. More recent approaches using digital In deeper ice, however, the annual layers are more
scanners, powerful computers and large storage media closely spaced and they become increasingly deformed
(Svensson et al., 2005) have enabled the visual stratigraphy and diffuse, hence annual variations are more difficult
of ice cores to be determined with extraordinary clarity to distinguish. Older ice, therefore, cannot be dated by
a) a
a a a a a a a a a
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Ligh t sourc e Lens
0.0
Camera 0.0
l u core Camera
0.0
0.0
-
Ligh t sourc e Lens Mirro r Dark field
0.0
0.0
0.0 -
0.0
0.0
-
0.0
Figure 5.26 Line-scan images from the NorthGRIP ice core. a) A camera moves along the upper side of an ice-core section while
an indirect light source is moving below the ice. The light enters the ice at an angle of 45° so that the camera detects only light
that is scattered in the ice. Transparent ice thus appears black in the record, whereas the cloudy bands appear white. b) The images
show alternating darker and lighter layers, the latter containing relatively high amounts of impurities, in particular micrometre-sized
dust particles from dry areas in eastern Asia. The visual stratigraphy is essentially a seasonal signal. However, other impurities
also show on the line-scan images; for example, the bright layer at 1.33 m in core b is the Vedde Ash (after Svensson et al., 2005;
images by Anders Svensson, Niels Bohr Institute, University of Copenhagen, Denmark).
312 DATING METHODS
straightforward incremental means, and recourse has to be as inconsistencies in the counting of annual (or, in some
made to other techniques. For example, age estimates cases, subannual) horizons. While replicate counts by at
can be derived from theoretical ice-flow models based on least two different operators are the norm, some degree
a knowledge of ice dynamics. This method is often com- of error will be inevitable. All ice-core chronologies there-
bined with the use of marker horizons, such as tephra and fore contain an element of uncertainty. For example, in the
sulphate from volcanic eruptions, cosmogenic nuclides younger part of the Greenland NGRIP ice core (14.8–7.9
(e.g. 10Be, 14C) and geomagnetic events (section 5.5.1.2) to ka BP), the maximum counting error (MCE) (broadly
provide ‘pinning points’ for the ice-core chronology equivalent to a 2σ error in a Gaussian distribution) ranges
(Parrenin et al., 2007). Alternatively, the isotopically defined from 2 per cent in the Holocene to 3 per cent in the Late-
events in the ice cores can be correlated with independently glacial. The base of the Holocene at 11,700 has an MCE
dated events (based on radiocarbon or U-series dating) of 99 years, and the onset of the Lateglacial at 14,692 an
in marine cores or in speleothem records, and the ice- MCE of 190 years (Rasmussen et al., 2006). In the deeper
core timescale can then be ‘calibrated’ using these age part of the core where the chronology extends to c. 60 ka,
measurements (Shackleton et al., 2004). Other methods the MCE varies between 4 per cent in the warm interstadial
include the tuning (matching) of proxy climate records periods to 7 per cent in the cold stadials, although the
obtained from ice cores, such as atmospheric trace gases overall MCE is c. 5 per cent or 950 years (Andersen et al.,
(e.g. methane) to variations in the earth’s orbital parameters 2006; Svensson et al., 2008).
(section 1.7) which have an independent chronology In Antarctica, while annual layer counting has been
based on astronomical calculations (Ruddiman & Raymo, possible at some sites (Taylor et al., 2004), in general low
2003). A difficulty with this approach, however, is that there rates of snow accumulation mean that layer counting is not
will be an age difference between the trapped gas and feasible, especially in deeper ice, and age–depth profiles
the surrounding ice matrix. This is because the gas itself have largely been based on ice-flow modelling constrained
is trapped several tens of metres below the ice-sheet surface by marker horizons and by orbital tuning of trace gas
where it is surrounded by ice that was deposited as surface records. The longest continuous chronology dated in this
snow, possibly hundreds or even thousands of years earlier, way is the EDC3 timescale from the EPICA Dome C site in
and firm densification models are required in order to East Antarctica (EPICA Community Members, 2004; Wolff
evaluate this ice age–gas age difference (Loulerge et al., et al., 2010). The 2σ error on the timescale is estimated to
2007). More recently, an integrated method for dating of be of the order of 100 years at 6–2 ka, increasing to 400 years
ice cores has been developed which combines the modelling at 14 ka, 1.5 ka at 40 ka, 3 ka at 100 ka and 6 ka at 130 ka,
and empirical approaches and develops a ‘best compromise’ after which it remains stable to the base of the record
between model-based dating scenarios and chronological (Parrenin et al., 2007).
information from data (direct measurements; age markers,
etc.). This probabilistic technique employs a Bayesian
5.4.3.3 Ice-core chronologies
framework (section 5.3.2.6) in order to estimate confidence
intervals for the new dating scenario (Lemieux-Dudon In the Greenland ice sheet, where snow accumulation
et al., 2010). was relatively rapid, chronologies span the last glacial–
interglacial cycle. They include the GRIP (Greenland
Ice Core Project) core, which is ice-layer counted back
5.4.3.2 Errors in ice-core chronologies
to 14.5 ka (Johnsen et al., 2001), and GISP2 (Greenland
Errors in ice-core chronologies arise from two sources: Ice-Sheet Project 2), where multiparameter continuous
imperfections in the nature of the record and human (or counting has been possible down to a depth of 2,800 m and
technical) errors that occur during sampling, recording and where the ice is dated at c. 110 ka (Meese et al., 1997). These
interpreting the ice-core sequence. The former includes loss chronologies agree within 750 years back to c. 40 ka, but
of material from the ice surface through wind scour which diverge by up to several thousand years beyond that time.
will leave gaps in the ice-core stratigraphy, and deform- The most highly resolved chronology from Greenland is
ation within the ice which causes folding and possible the Greenland Ice Core Chronology (GICC05) extend-
loss of stratigraphic continuity (Alley et al., 1997). It also ing back to 60 ka (Figure 5.27; Table 5.5). GICC05 is a
includes the ice age–gas age difference described above. composite stratigraphic timescale and is based on multi-
Technical or human errors include incomplete core recov- parameter counting using δ18O and δD variations, electrical
ery, which means that interpolations of age may have to be conductivity measurements (ECM) and continuous flow
made to bridge the missing parts of the sequence, as well analysis (CFA) of water-soluble ions in three separate
DATING METHODS 313
Greenland ice cores: Dye-3, GRIP and NGRIP (Rasmussen ice thickness, heat flow, ice melt and integral ice-flow
et al., 2006). Ages are expressed in years before AD 2000, dynamics (Johnsen et al., 2001; Dahl-Jensen et al., 2003).
which is abbreviated to b2k; hence the Holocene boundary As we saw in section 3.11.1, the oldest ice so far investi-
in the core referred to above is dated at 11,700 b2k, i.e. gated in Greenland is in the 2,540 m long NEEM core from
before AD 2000. As radiocarbon ages are measured in years the northern part of the ice sheet which contains the onset
before 1950, this means that ice-core ages on the GICC05 of the Eemian interglacial (NEEM Community Members,
timescale are 50 years older than radiocarbon ages. The 2013). By linking the CH4 and δ18O profiles from the
7.9–0 ka section of GICC05 is based on annual layer bottom section of the core with other records from
counting in the Dye-3, GRIP and NGRIP ice cores, the Greenland (NGRIP) and Antarctica (EPICA), the Eemian
14.8–7.9 interval on ECM and CFA in the GRIP and NGRIP sequence in the NEEM ice core can be placed on the EDML
cores, while the chronology for the 60–14.8 ka section is timescale (see below).
based on the counting of annual layers identified by visual A number of chronologies have been generated from
stratigraphy and in ECM and CFA data in the NGRIP core Antarctic drilling programmes, including the Vostok
only (Svensson et al., 2008). Station and Dome Fuji records, dated by ice-flow modelling
The NorthGRIP ice-core record extends to bedrock at to c. 420 and 340 ka respectively (Petit et al., 1999; Watanabe
3,085 m depth below the ice surface and dates to c. 123 ka et al., 2003b). The oldest record obtained so far is the EPICA
BP, the later part of the last (Eemian) interglacial (North core, which has reached ice estimated to date to 960 ± 20
Greenland Ice Core Project Members, 2004). Dating by ka (EPICA Community Members, 2004). The EDC3
layer counting may, in due course, be possible to a depth timescale from that core is based on snow accumulation
of 2,700 m (c. 90 ka BP), but below this the annual layers rates and a mechanical ice-flow model, and by reference to
are extremely thin, indistinct and deformed (Svensson et a set of age markers along the core. These include volcanic
al., 2005). The age of the lower parts of the ice-core record, events and independently dated cosmogenic records;
therefore, will still have to be estimated using ice-flow geomagnetic events; and dated CH4 profiles in other ice-
models based on observed physical relationships between core records (e.g. GICC05). Construction of the timescale
NorthGRIP depth (m)

1400 1600 1700 1800 1900 2000 2100 2200 2300 2400
GI-1
GI-1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
-36
YD
NorthGRIP 6 O i e
-40
(m)
-0,06
-44
Annual layer thickness -0.04
0.02
0,00
10.000 20,000 30,000 40,000 50,000 60 000
G I C C 0 5 age (yr)
Figure 5.27 The NorthGRIP (NGRIP) oxygen isotope record and variations in annual ice-layer thickness plotted against the GICC05
timescale. GI indicates Greenland Interstadials; GI-1 – Bølling-Allerød Interstadial; YD – Younger Dryas Stadial (after Svensson
et al., 2008).
314 DATING METHODS
Table 5.5 GICC05 chronology of climatic events represented in the NGRIP and GRIP ice-core records for the
period c. 47 to c. 8.1 b2k (years before AD 2000). GS – Greenland Stadial; GI – Greenland Interstadial (from
Blockley et al., 2012).
Event Depth in GICC05 Maximum Uncertainty Depth in
NGRIP age counting of definition GRIP core
core (m) (years b2k) error (2␴) (years) (m)
in years
End of 8.2 ka BP event 1,219.47 8,140 45 +50/–10 1,325.11
Volcanic peak, 8.2 ka BP event 1,228.67 8,236 47 1 1,334.04
Start of 8.2 ka BP event 1,234.78 8,300 49 +10/–40 1,340.5
End of 9.3 ka BP event 1,322.88 9,240 68 +30/–10 1,432.43
Start of 9.3 ka BP event 1,331.65 9,350 70 +10/–20 1,442.10
Start of Holocene 1,492.45 11,703 99 4 1,624.27
Start of GS-1 1,526.52 12,896 138 4 1,662.41
Start of GI-1a 1,534.5 13,099 143 30 1,671.7
Start of GI-1b 1,542.1 13,311 149 +30/–10 1,680.5
Start of GI-1c 1,570.5 13,954 165 +30/–10 1,713.7
Start of GI-1d 1,574.8 14,075 169 10 1,718.5
Start of GI-1e 1,604.64 14,692 186 4 1,753.39
Start of GS-2a Feature not consistent between records; no date recommended
Start of GS-2b 1,745.31 20,900 482 100 1,899.7
Start of GS-2c 1,783.62 22,900 573 40 1,940.28
Start of GI-2 1,793.19 23,340 596 60 1,950.46
Start of GS-3 1,861.69 27,540 822 20 2,018.09
Start of GI-3 1,869.12 27,780 832 20 2,025.39
Start of GS-4 1,882.62 28,600 887 20 2,037.7
Start of GI-4 1,891.57 28,900 898 20 2,046.05
Start of GS-5 1,938.95 32,000 1,103 60 2,087.33
Start of GI-5 1,951.66 32,500 1,132 20 2,098.82
Start of GS-6 1,964.3 33,360 1,191 20 2,109.73
Start of GI-6 1,974.56 33,740 1,212 20 2,118.58
Start of GS-7 1,990.28 34,740 1,286 20 2,131.77
Start of GI-7 2,009.45 35,480 1,321 20 2,148.69
Start of GS-8 2,026.66 36,580 1,397 40 2,162.56
Start of GI-8 2,070.03 38,220 1,449 20 2,200.32
Start of GS-9 2,094.64 39,900 1,569 20 2,219.69
Start of GI-9 2,099.62 40,160 1,580 20 2,223.6
Start of GS-10 2,109.71 40,800 1,615 20 2,231.55
Start of GI-10 2,124.03 41,460 1,633 20 2,243.27
Start of GS-11 2,134.99 42,240 1,682 20 2,251.94
Start of GI-11 2,157.49 43,340 1,736 20 2,270.77
Start of GS-12 2,170.0 44,280 1,780 60 2,280.48
Start of GI-12 2,222.3 46,860 1,912 20 2,324.22
DATING METHODS 315
also involves tuning of atmospheric trace gas records in the pioneered and developed by Roland Beschel a decade or so
EPICA core to orbital parameters (Parrenin et al., 2007). later, and rests on the principle that there is a direct
More recently, a multi-site, multi-proxy timescale, the relationship between lichen size and age. Where a surface
Antarctic Ice Core Chronology 2012 (AICC2012), has been has been recently exposed to lichen colonization, providing
developed for four Antarctic ice cores: Vostok, EPICA (a) that the growth patterns of the lichens are known and
Dome C (EDC), EPICA Dronning Maud Land (EDML) (b) that no major time lapse has occurred between surface
and Talos Dome (TALDICE). The section from 120–0 ka exposure and lichen colonization, an estimate of the age of
includes data from the Greenland NGRIP GICC05 the substrate can be made. In studies of glacial retreat, for
timescale and has been constructed using the Bayesian example, lichen size (usually maximum diameter of the
tool DATICE (Lemieux-Dudon et al., 2010) that combines largest lichen) is first established for morainic surfaces of
both glaciological inputs and a wide range of relative and known age (dated by radiometric methods or by historical
absolute gas and ice-stratigraphic markers (Veres et al., evidence such as old photographs) and a lichen growth
2012). The record from 800–120 ka also employs DATICE, curve can then be constructed based on these ‘fixed points’
and is based on a combination of modelling (snow (Figure 5.28). Surfaces of unknown age can then be dated
accumulations rates, snow densification into ice, and ice by relating lichen diameters on those surfaces to the growth-
flow), and observational data (absolute ages of key reference rate curve and deriving a calendar age. In this way, a
horizons, stratigraphic links between the different cores, detailed deglacial chronology can be established for an
and orbitally derived ages). The new chronology shows only area. The biological basis of lichenometry is discussed by
small differences, well within the uncertainty range, when Loso & Doak (2006) who confirm its reliability as a dating
compared with the previous Antarctic ice-core reference age method.
scale, EDC3, described above (Bazin et al., 2012). Some lichen species (e.g. Rhizocarpon geographicum)
will continue to grow for several thousand years and
therefore, in theory, lichenometry is a technique that may
5.4.4 Lichenometry be applicable to most of the Holocene. In practice, however,
the dating limit is around 4,500 years in extremely cold and
dry continental regions, such as west Greenland, whereas
Lichens are complex organisms consisting of algae and in the majority of cases the age range for lichenometry as
fungi living together symbiotically. The algae provide a dating technique is 500 years or less (Matthews, 1992).
carbohydrates via photosynthesis, while the fungi pro-
vide the protective environment in which the algal cells
5.4.4.2 Sources of error in lichenometric
can function. Although lichens had been used as a basis
dating
for dating by Knut Faegri in the 1930s, the technique was
Not all lichens are suitable for lichenometrical purposes,
150-
for only those that show a gradual and progressive rate
of growth can be employed. Moreover, lichen growth is
. affected by local environmental conditions that vary with
. le A.u
Moraine daleA.u
A ua
.d both latitude and altitude, for example air temperature,
150- aal iene
e dor
orain M day length and snow cover. A lichen growth curve must
M
therefore be constructed for specific lichens and will only
be applicable to particular geographical areas (Figure 5.28).
50-
Moreover, recent climate change will probably result in
changed lichen growth curves over time (Hansen, 2010).
However, in remote regions, such as the mountains of
0 -
1950 1900 1850 1800 1750 Norway or Greenland, no documentary or other evidence
Moraine daleA.u. may be available for establishing fixed points on a lichen
growth curve (Winkler et al., 2003). Further, it is assumed
Figure 5.28 Lichenometric growth curves for Rhizocarpon that in the study of glacier recession, for example, there is
spp. from West Greenland, West Spitsbergen and Baffin Island. no significant delay in lichen colonization of exposed
The ages of unknown surfaces (moraines, etc.) can be obtained
by measuring average lichen size on those surfaces (y axis)
surfaces following ice retreat. This, however, can never be
and reading off the appropriate age on the x axis (after Forman proved and must always remain a source of uncertainty.
et al., 2007). Conversely, lichens are found on actively forming medial
316 DATING METHODS
moraines and therefore in some areas lichen growth must substrates in proglacial areas (Matthews, 1992); and the
have preceded ice wastage. Finally, problems have been dating of talus accumulations (McCarroll et al., 1998),
encountered in sampling and measurement of lichens in the glacier outburst floods (Winchester & Harrison, 2000),
field and very thorough preliminary investigations are fluvial deposits (Gob et al., 2003), debris-flow activity
required to establish reproducibility of results within any (Innes, 2006), and historic and late prehistoric earthquakes
one region. These and other constraints on lichenometric (Bull, 2003). The technique has also been used to date
dating are reviewed by Hansen (2008). structures built of lichen-covered rock, and may have other
applications in archaeology (Benedict, 2009).
5.4.4.3 Some applications of lichenometry
5.4.5 Other materials dated by annual
Lichenometry has been most widely employed in the dating
of Late Holocene moraines, particularly those of the Little
increments
Ice Age period, for example in Scandinavia (Matthews,
5.4.5.1 Speleothems
2005), Alaska (Wiles et al., 2010), the North American
Cordillera (O’Neal & Schoenenberger, 2003) and New As we saw above (section 5.3.4.3), speleothems, which
Zealand (Winkler, 2004). However, lichenometry has a often contain valuable palaeoclimatic evidence, can be
range of other applications. These include the provision of dated by U-series. However, in certain circumstances, it
a timescale for plant colonization of newly exposed may be possible to identify individual growth bands in the
a b 1974
lc 1775
5 cm
177.1
1773
11959-70
1772
1771
1770
\TM
1768
[1963-64
1768
1767
1766
1765
11957-58 1764
I 10 cm 1783
1762
1768
176fJ
175fa
I7r f
1757
Broo k Islan d
.1756
1755
nr,i
1753
1752
1751
1750
1941-42
nlavannat i Islan d Pandor a Reel
Figure 5.29 Radiographs of core-slabs cut from a) Pleistocene and b) modern coral reefs (Montastraea) in Florida Bay, Gulf of
Mexico, showing seasonal growth layers (from Gischler et al., 2009; images supplied by Eberhard Gischler, University of Frankfurt,
Germany). c) Cross-matching of growth layers of Porites coral from three locations on the central Great Barrier Reef, Australia,
using images enhanced by UV luminescent illumination, an approach that enabled a 373-year chronology of ENSO variations
to be reconstructed (from Hendy et al., 2003, reprinted with permission of SAGE; image supplied by Erica Hendy, Bristol
University, UK).
DATING METHODS 317
speleothem (Figure 3.38) that represent annual accumula- physical and chemical processes of skeletal construction.
tions of calcite, and counting these laminations forms The construction of a timescale based on these growth
the basis for a timescale (Baker et al., 2008b). The floating increments is referred to as sclerochronology, and the
chronology that results can be anchored to calendar years most widely employed media thus far have been corals and
either by counting back from the present day or by using molluscs (Gröcke & Gilliken, 2008; Oschmann, 2009).
other dating techniques (U-series; 210Pb) to calibrate the
sequence (Paulsen et al., 2003). Speleothem laminae can Corals
be identified visually using UV and visible light, or they Corals grow rapidly through successive increments of
can be detected by fluorescence (or luminescence), the aragonite that can accumulate at up to several centi-
signal reflecting seasonal variations in concentrations of metres per year. Some species contain clear evidence of
organic matter derived from the overlying soil by percol- annual banding (Figure 5.29) which can be detected by
ating groundwaters (McGarry & Baker, 2000). Annual X-radiography or UV luminescence (Hendy et al., 2003).
banding in speleothems has been used to provide a The annual bands are characterized by variations in
timescale for recent precipitation records (Fleitmann chemical properties, such as stable isotope and trace-
et al., 2004), to calibrate climatic reconstructions over the element ratios, which reflect seasonal variations in
course of the last millennium (McMillan et al., 2005) and environmental conditions (Grottoli & Eakin, 2007), and
to provide an independent chronology for Dansgaard– from which past sea-surface and salinity changes can be
Oeschger events (Spötl & Mangini, 2002). reconstructed with an annual, monthly or even weekly
temporal resolution (Figure 5.30). Although coral growth
bands can be cross-matched in the same manner as tree-
5.4.5.2 Sclerochronology
ring sequences, there are, as yet, few records extending back
Many organisms in the aquatic realm display clear more than c. 400 years. Typical errors associated with
annual growth bands reflecting seasonal controls over the coral-chronology are of the order of 1–2 years per century.
a) W/W
-6¬
Nomi nal year Nomi nal year
-5¬
-4¬ yea
-1¬
yea
0.5¬
0-}
0.5¬ yea
1- I l I
0I14 0 4 12 6 14 8 1610 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54
a) Nominal year
•0.6 -r •0.6
Nomi nal year Nomi nal year
0.4¬ Nominal yea

year •0.6
Nomi nal yearNomi nal year
0.2¬ •0.6
0 - •0.6
0.2-. •0.6
0.4- •0.6
•0.6
Nominal year
•0.4 ¬ •0.6
0.2¬ •0.6
0¬ •0.6
0.2¬ •0.6
0.4¬ •0.6
•0.6
0.6- •0.6
20 10 12 614 816 10
4 4 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54
Nominal year
Figure 5.30 A 54-year annual coral stable isotope record from the mid-Holocene. a) The δ18O record (above) and deviations in
δ18O (Δδ18O) after removal of the sea-surface temperature component W/W – warm/wet; C/D – cold/dry. b) Annual variability in
winter (December–February) sea-surface temperatures and in summer sea-surface salinity (SSS) (after Sun et al., 2005).
318 DATING METHODS
1. Panama
2. Galapagos
3 Tarawa
year nal year

4. Vanuatu
Nomi nal Nomi

Nomi nal year
5. New Caledonia
Nomi nal year
6 Great Barrier Reef
Nomi nal year

7. Philippines
Nomi nal year
3. Seychelles
9 Red S e a
1600 1700 1800 1900 2000

Year
Figure 5.31 Annual coral δ18O records from various sites in the Pacific and Indian Oceans. Horizontal lines represent the mean
δ18O value for each site. Black triangles mark approximate times of abrupt shifts in δ18O, heralding warmer/wetter conditions related
to ENSO variations (after Gagan et al., 2000).
DATING METHODS 319
As global sea levels have fluctuated, so too has the tidal 5.5 AGE-EQUIVALENT
range within which corals live, and hence fossil corals of
varying ages occur at different altitudes, while many are
STRATIGRAPHIC MARKERS
submerged below present sea level. Coral records have In many Quaternary deposits, distinctive marker horizons
been used to reconstruct past sea-surface temperature are found that are broadly synchronous and form time
changes (McCullough & Esat, 2000), ENSO events (Figure planes across different sedimentary sequences. The horizons
5.31; section 7.6.4.2), variations in coastal run-off (Hendy themselves cannot be used in the first instance to date
et al., 2003) and Holocene precipitation changes (Deng Quaternary successions, for other methods are required to
et al., 2009). In addition, samples of fossil corals have been establish their age. However, once dated by radiometric
used to develop floating chronologies for older (interglacial) or incremental methods at one locality, they allow age
reef sequences (Tudhope et al., 2001). estimates to be extended to other sequences where the
marker horizon is present. As such, they form an indirect
Molluscs means of dating. Moreover, in view of their often
Like corals, some marine molluscs also develop annual widespread distribution, they also form a basis for strati-
growth bands in their shells. These range from semi-diurnal graphic subdivision and time-stratigraphic correlation
to annual and reflect the influence of a range of environ- (Chapter 6).
mental factors, including water temperature, salinity and, Four methods of dating using age-equivalent strati-
in particular, nutrient and food availability (Goodwin et al., graphic markers are considered here: palaeomagnetism,
2001). Variations in growth banding may provide valuable which is based on the changes in the earth’s magnetic field
palaeoenvironmental data, as well as the basis for a time- preserved in rocks and sediments; tephrochronology, the
scale for geochemical proxies such as δ18O and δ13C from use of volcanic ash layers as a means of dating; oxygen
which variations in, for example, bottom-water tempera- isotope chronostratigraphy, which employs globally
ture can be inferred (Weidman et al., 1994). The shells of synchronous changes in the oxygen isotope signal in
some freshwater molluscan species may also contain deep-ocean sediments; and biostratigraphy, in which
evidence of former climatic/environmental conditions, dating is based on palaeoecological or evolutionary events
such as changes in air temperature (Schöne et al., 2004). recorded in stratigraphic sequences. Other marker
Composite sclerochronologies have also been developed, in horizons that are more widely employed in stratigraphic
which living molluscs are linked to subfossil species using subdivision and correlation than in dating, such as palaeo-
distinct ‘marker’ bands or sequences of bands, following the sols and shorelines, are discussed further in section
principles of cross-dating employed in dendrochronology 6.3.2.
(section 5.4.1.2). Many of these chronologies are relatively
short (< 100 years) because of the limited lifespan of most
Mollusca (Schöne, 2003), but some are longer-lived and
5.5.1 Palaeomagnetism
hence longer chronologies may be possible. One example
5.5.1.1 Geomagnetic field and remanent
is Arctica islandica, a relatively long-lived (over 100 years)
magnetism
species from the North Atlantic, from which a contin-
uous chronology back to AD 1843 has been obtained for The earth’s geomagnetic field is dipolar in that it possesses
northwest Scotland (Stott et al., 2010). An even older two poles that we refer to as north and south. Over time,
Arctica record has been obtained from the North Sea where the magnetic field changes in both field strength and
a cross-matched floating chronology spanning the period direction. There are two components to these changes:
from c. AD 1000–1400 has been constructed, and which major dipole changes where the dynamo currents that
includes a 267-year continuous series from the longest- generate the main dipole field become reversed so that
living Arctic islandica specimen yet recorded from the area effectively the magnetic north pole becomes the magnetic
(Scourse et al., 2006). Further development of these sclero- south pole and vice versa (polarity changes), and secular
chronological records could have major implications, not variations, which operate largely through a process known
only for palaeoceanographical reconstructions, but also for as ‘dipole wobble’ whereby the axis of the dipole field
our understanding of spatial and temporal variations in the (magnetic north) precesses around the earth’s axis of
marine radiocarbon reservoir (section 5.3.2.4). In addition, rotation (true north). The major polarity changes are meas-
sclerochronology of shell middens provides an opportunity urable over timescales of thousands or millions of years,
for the examination of climate and environmental change whereas, secular variations occur over centuries or less
in relation to human activity (Andrus, 2011). during which field direction changes at a rate of 1° every
320 DATING METHODS
few decades, while field strength changes by several per cent angle of magnetic dip. Magnetic intensity can be measured
per century (Sternberg, 2001). by the amount of torque required to prevent the needle
A record of both of these magnetic changes can be found returning to the angle of magnetic dip after it has been
in rocks and sediments containing magnetic minerals, rotated through 90°.
for these are magnetized during formation and hence
individual crystals or particles will reveal a natural
5.5.1.2 Magnetostratigraphy
remanent magnetism (NRM) which is a reflection of the
geomagnetic field at the time of rock or sediment formation The study of variations in magnetic properties through a
(McElwhinney & McFadden, 2000). Volcanic rocks will sequence of rocks or sediments is termed magneto-
acquire a remanence through heating, and this is referred stratigraphy. Geomagnetic field variation can be detected
to as thermoremanent magnetization (TRM). Archaeo- in rocks or sediments that contain even small amounts of
logical materials that have been reheated, such as pottery magnetic minerals, and the identification of major polarity
and hearths, will also contain a TRM signal, and again this changes in volcanic rocks, in ice cores, and in deep-sea
will reveal the direction of the earth’s magnetic field at sediment records, or variations in inclination or declina-
the time of firing. Sedimentary rocks and unconsolidated tion (secular changes) in different depositional sequences,
sediments accumulating on the sea floor or in lakes also provides a basis for relative dating and correlation (Roberts
contain evidence of former geomagnetic fields, for a record et al., 2013). In addition, however, mineral magnetic
is preserved in the alignment of ferromagnetic sedimentary ‘potential’, that is, the concentration and magnetic
particles as they settle in water or in water-saturated susceptibility of magnetic minerals, also varies within
sediments. This is referred to as detrital remanent mag- sediments, and under certain circumstances this too can
netism (DRM). In some sediments, however, NRM can also provide a basis for correlation.
be acquired by chemical action, where the crystallization
of ferromagnetic oxides results in a chemical remanent Secular variations
magnetism (CRM). This process may occur later than, and Secular variations, which are due to both dipole and
under a different magnetic field from, that of DRM in the non-dipole field activity, are variable in both space and
same sediment unit. In this way, a secondary magnetization time, and are reflected in changes in declination, inclination
is introduced into both volcanic and sedimentary rocks, and field intensity. Past variations in these three parameters
which serves to complicate the study of palaeomagnetic are referred to as palaeosecular variations (PSV). Direct
variations. Thompson and Oldfield (1986) provide further measurements of values of field declination, inclination
details of the magnetic properties of minerals, rocks and and intensity have been collected over the past 400 years2
sediments, while various aspects of environmental for cities such as London, Rome and Boston (Figure 5.32).
magnetism are discussed in Walden et al. (1999), Evans & These suggest complex radial variations in field activity
Heller (2003) and Maher (2007). around the globe and may hold the key to understanding
There are three components in the earth’s magnetic the causes of major excursions and reversals (Jackson et al.,
field. Declination is the angle between magnetic north and 2000). In London, for example, declination has varied
geographic (true) north, while inclination refers to the angle from 11°E in AD 1570 to 24°W in AD 1820. Since then,
of dip of the magnetic field. Hence, a freely suspended declination has decreased to the present values of 5°W, and
needle at the surface of the earth will align with the pre- it continues to decrease at a rate of 9 minutes each year
vailing magnetic field (declination), and the amount of (Thompson & Oldfield, 1986). Magnetic inclination has
dip exhibited by the needle relative to the horizontal is a varied from a maximum of over 74° in AD 1700 and is now
measure of the inclination. The inclination value varies near to 66°. Figure 5.32 shows that secular magnetic
from 0° at the magnetic equator to 90° at the magnetic variations in London and Rome have been similar whereas
poles. The third component, intensity, refers to the strength in Boston, some 5,000 km from Europe, the pattern has
of the geomagnetic field. At the present day, the field been very different. These curves, based on observational
strength at the geomagnetic poles is twice that at the evidence, can be extended into the historic period and
geomagnetic equator. The strength of the field can be beyond by using archaeomagnetic measurements from
estimated in the following way. Suppose a magnetic needle features dated by documentary sources, pottery typology,
is fixed to a horizontal axle, so that the axle passes through dendrochronology or other methods (Batt, 1997). Within
the centre of gravity of the needle and is orientated along the historic period, data from a number of sites in western
magnetic east–west. If the needle (aligned north–south) is Europe and the eastern Mediterranean suggest that the
allowed to swing freely, it would eventually stabilize at the region experienced ‘archaeomagnetic jerks’ (significant
DATING METHODS 321
-30° -20° -10° 0° -10°

IOOO
V-
1510
' Rome
-30° 60'
1800
'1600
Declinatio n
1700
, London
1900
70° 70"
Boston
1800 •
f 1800
1540
1600
1800
-30° -20° -10° -10 -10*
Declination
Figure 5.32 Secular changes in magnetic declination and inclination as observed in London, Rome and Boston. The solid curves
begin at the time of the first declination measurements at each locality. The earlier inclination changes (dashed lines) are based
on archaeomagnetic data (after Thompson & Oldfield, 1986).
short-term departures in geomagnetic field direction and profiles from the mid- and high latitudes of the South-
intensity) on at least two occasions, most notably at c. AD ern Hemisphere, and reveals consistent millennial-scale
200 and 1400 (Gallet et al., 2003). variability (Lisé-Pronovost et al., 2013). The derivation of
Long-term records of secular geomagnetic variations an accurate record of long-term secular variations is not
can also be obtained from lake sediment sequences, and always straightforward, however, partly because of the
these can be used as regional chronostratigraphic tools difficulties in obtaining an accurate independent time-
(Barletta et al., 2008). High-resolution magnetostrati- scale for lake sediment profiles, and partly because secular
graphic sequences, which can be dated by independent variation patterns rarely have amplitudes of more than
methods such as radiocarbon and varve chronology (Snow- 20° (compared with 180° for polarity changes) and
ball & Sandgren, 2002), constitute type profiles against hence the palaeomagnetic signal may not always be easily
which other secular magnetic records can be matched distinguishable from background ‘noise’ (Snowball et al.,
(Figure 5.33). These master curves, some of which extend 2007).
back into the early Holocene, may be applicable to
sediments found up to 2,000 km from the type site. Core Field reversals and the palaeomagnetic timescale
matching can be achieved using the distinctive inflections From time to time, the geomagnetic field reverses so that
or turning points in the magnetic profiles. Secular geo- the geomagnetic poles change relative positions through
magnetic variations recorded in lake sediments have been 180°. These polarity reversals can be detected in the
employed in the correction of errors in Holocene varve geological record and are of fundamental importance in
sequences in west central Sweden (Stanton et al., 2010) palaeomagnetic studies. The present-day magnetic field is
while, on a longer timescale, a 50 ka palaeomagnetic record regarded as possessing normal polarity, and the opposite
of secular variation has been obtained from a lake sequence is referred to as reversed polarity. Periods of long-term fixed
in southern Argentina. This compares with other magnetic polarity (105–107 years) are known as polarity epochs.
322 DATING METHODS
Pohjajarvi
Declination of master curve
0 hood 2000 3000yr
50
Declin atio
Declin atio
Declin atio
Declin atio
n
n
n
n
100-
150.:
200
(cm)
Inclination of master curve
Figure 5.33 Palaeomagnetic dating of sediments from Lake Päijänne, Finland, based on a correlation of the declination (left) and
inclination (right) record with the dated PSV master curves from Lake Pohjajärvi. Open boxes indicate matching declination features
and solid boxes show matching inclination features (after Saarinen, 1999).
These are interrupted by a large number of polarity reversals timescale to be established for the Quaternary and parts
of shorter duration (104–105 years), which are termed of the pre-Quaternary sequence (Figure 5.34). Three
polarity events, and also by polarity excursions, in which polarity epoch boundaries are shown on Figure 5.34: the
the geomagnetic pole changes direction through 45° or Brunhes–Matuyama, which is K–Ar dated at c. 0.73 Ma, the
more for a short period only (100–1 ka). Polarity epochs Matuyama–Gauss at c. 2.47 Ma and the Gauss–Gilbert at
and polarity events are experienced globally and can be used c. 3.41 Ma. Important polarity events include the Jaramillo
as a basis for worldwide correlations. ‘normal’ event, which has a K–Ar age of between 0.90 and
Where polarity epochs and events are found in vol- 0.97 Ma, and the Olduvai ‘normal’ event between 1.67 and
canic rocks, they can be dated by the argon isotope dating 1.87 Ma. Argon isotope dating is, however, not sufficiently
method (section 5.3.3), thereby enabling a palaeomagnetic precise to date some of the relatively short-lived polarity
DATING METHODS 323
A g e (k-Ar), Astronomically-tuned
events, and their positions on the palaeomagnetic timescale
Ma A g e (Ma) have therefore been established by extrapolation based on
the ages of epoch boundaries. As a consequence, the dating
B l a k e 0.12
of polarity events tends to be less secure.
Mo unt ai
An alternative approach to dating the palaeomagnetic
timescale involves the use of the ocean sediment record. As
E m p e r o r 0.42
sediments accumulate on the deep-ocean floors, individual
particles adopt the direction of the earth’s magnetic field,
and hence a continuous record of geomagnetic changes is
preserved within the sediment sequence. Astronomical
0,73 0,78 tuning of the oxygen isotope signal obtained from the
microfossil record within these sediments (sections 5.5.3
0,90 0.99 and 6.3.3) provides the basis for a timescale for the
0,97 Jaramillo geomagnetic changes that is independent of that based on
1 07
argon isotope dating of volcanic rocks (Bassinot et al.,
1,15
1994). Ages of the polarity epochs and principal polarity
C o b b Mountain 1.19
1,17 events that have been obtained using this method are
shown on the right of Figure 5.34. In general, the dates
tend to be older than those based on K–Ar. Hence, the
Brunhes–Matuyama boundary is dated at c. 0.78 Ma and
the Matuyama–Gauss boundary at c. 2.6 ka BP, while the
Mou ntai
1,55 G i l s a 1.6S
ages of the Jaramillo and Olduvai events are 0.99–1.07 Ma
Mou ntai
1,67 1.79
and 1.79–1.95 Ma BP, respectively. The resolution and
length of the polarity timescale in the ocean sediment
Olduvai
record is, however, dependent on rates of sedimentation.
1,87 1,95 For example, in parts of the oceans experiencing com-
paratively slow sedimentation rates (e.g. 0.1–1.2 m per 1,000
2.01
2,04
years), a core may contain a complete record of Pleistocene
2.12 R e u n i o n 2.13 polarity changes. Conversely, in some areas of the oceans,
2,14
such as the North Atlantic, where sedimentation rates are
rapid (e.g. 0.25–0.5 m per 1,000 years), coring may fail to
reach the first major geomagnetic boundary, the Brunhes–
Matuyama transition.
• 2,48 2.6 The causes of these geomagnetic changes are not
fully understood, but some have pointed to a possible link
between magnetic variations and astronomical cycles.
For example, Yamazaki & Oda (2002) have detected a
100 ka periodicity in a 2.25 Ma palaeomagnetic inclination
and intensity record from the equatorial Pacific (Figure
Mou ntai
2.92 5.35), and have suggested that the earth’s magnetic field
Kaena
3.01 may be modulated by orbital eccentricity. A relationship
3,05
Mammoth has also been proposed between geomagnetic polarity
3,15 reversals/excursions and glacial–interglacial cycles, with
the suggestion that the growth of polar ice sheets may
affect the differential rotation between the earth’s field
• 3,40 and core, and that increased ice volumes are accompanied
by a weakened geomagnetic field (Westaway, 2009). Other
Figure 5.34 The palaeomagnetic timescale of the last 3.5 Ma. possible causes of geomagnetic reversals are discussed by
Shaded areas indicate periods of normal polarity; unshaded Gubbins (2008).
areas show episodes of reversed polarity. K–Ar ages are shown
on the left; astronomically tuned ages are on the right (after The use of palaeomagnetic stratigraphy as a means of
Cande & Kent, 1995; Funnell, 1995). correlating between individual deep-sea cores, and also
324 DATING METHODS
30¬
Mou ntai
Mou ntai
30-
Mou ntai Mou ntai

Mou ntai Mou ntai
i 2¬
Mou ntai
Mou ntai Mou ntai

1¬
0
0 500 1000 1500 2000
A g e (ka)
Figure 5.35 Variations in the earth’s magnetic inclination and field strength (palaeointensity) during the last 2.5 Ma (from Yamazaki
& Oda, 2002).
between the marine and terrestrial records, is discussed 3.5 Ma (Sun et al., 2006b). The magnetic susceptibility
further in Chapter 6. signal, which is usually higher in soils than in loessic
sediments, reflects the production or concentration of
Magnetic susceptibility, isothermal remanent
magnetic minerals with pedogenesis, and therefore forms
magnetism and coercivity
a basis for correlation between widely separated profiles
A number of magnetic characteristics of sediments do not
(Figure 5.36).
depend on variations in the earth’s magnetic field, but
rather reflect the nature and origins of magnetic minerals Isothermal remanent magnetization (IRM): This is the
in the sediments (Walden et al., 1999). While these pro- magnetic moment activated in and retained by a sample
perties cannot in themselves be used as a basis for a placed in a magnetic field at room temperature. With a
chronology, they offer a potential basis for correlation. This gradual increase in the strength of the field, IRM will
is because changes in mineral magnetic potential are often increase non-linearly until saturation isothermal remanent
a reflection of environmental change, and as these will be magnetization (SIRM) is reached. This is the level at which
broadly synchronous, the magnetic signal in sediments may a further increase in the magnetic field will not result in
offer a means of correlation between depositional contexts any increase in IRM in the sample. SIRM measurements
and across a range of timescales. have been used in a range of environmental studies, for
example in reconstructions of fire and pollution histories
Mineral magnetic susceptibility (MS): This measures the
(Gedye et al., 2000; Rosen & Dumayne-Peaty, 2001), and
degree to which a material can be magnetized, that
again these properties have been employed in the
is, its ‘magnetizability’. Variations in natural magnetic
correlation of Holocene lake sediment records (Snowball
assemblages, which will be reflected in susceptibility meas-
& Thompson, 1992).
urements, have been used to make inferences about a
number of environmental processes (e.g. as an indicator of Coercivity of IRM: This is the reversed field strength
sediment flux and erosion in lake catchments). Applications required to reduce the remanent magnetism to zero after
include the correlation of Holocene lake sediment cores saturation. Low coercivity values appear to be characteristic
(Evans & Slaymaker, 2004), the relative age correlation of large-grained magnetite, while high coercivities tend
of alluvial fans (Harvey et al., 2003) and the correlation of to be associated with fine-grained haematite. In practice,
magnetic susceptibility variations in cave sediments with coercivity curves may differentiate between assemblages of
the MIS record to date Neanderthal skeletal remains soil and sediment types, and offer a further means whereby
(Ellwood et al., 2004). On a longer timescale, magnetic correlation between sediment sequences, and between
susceptibility measurements have been obtained from components of those sequences, can be effected (Andrews
loess–palaeosol sequences from China extending back over et al., 2003).
DATING METHODS 325
these horizons constitute distinctive isochronous marker
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
horizons that provide a basis for inter-site correlation.
Magneti
Many tephra layers contain a unique geochemical finger-
Magneti Unit print relating to the source of eruption, and it is often
Magneti SO 0 possible to date the tephra layers using radiometric or
Magneti L1
Magneti other techniques (see below). Tephras therefore provide a
S1 10
basis for dating (tephrochronology) and also for correlation
SSuusscceepptti ibbi il li itt
L2
Magneti
Susceptibilit
S2 (tephrostratigraphy). Good overviews can be found in
L5 20
L5 Alloway et al. (2007a) and Lowe (2011).
L5
L5 30 In areas close to a volcanic eruption, tephras may be
L5
SS
visible as distinctive light-coloured horizons in stratigraphic
L6
4C sections and in cores of peats or lake sediments. Such ash-
S6 fall deposits are usually referred to as proximal tephras, in
S7 50 780
contrast to distal tephras that can be deposited hundreds
Susceptibilit
S3
Magneti
L9 or even thousands of kilometres downwind of the point of

60
eruption (Figure 5.37a). For example, the Y5 tephra from
%12 S90
S
%12 Jarimilb the Campanian Ignimbrite eruption near Naples (c. 39 ka)
70
%12 1070
L15 reached southwest Russia, more than 2,500 km away
S14
80 (Figure 5.37c; Pyle et al., 2006), and it is likely that non-
L15
visible distal ash layers (microscopic glass shards known as
Susceptibilit
90
WS1 microtephras or cryptotephras) travelled even further
WL1 afield. Meticulous laboratory procedures involving, for
WS2 100
example, density and magnetic separation, are required to
SSuusscceeppttiibbiilliitt
WL2
Magneti
110 1770 extract these from minerogenic lacustrine and marine

WS3 Olduvai
1950 sediments where the volcanic glass shards are often present
WL3 120 in very low concentrations (Turney & Lowe, 2001). The
WS4 tephras can be identified on the basis of their physical
WL4 130 characteristics (colour, grain-size distribution, lithic
Red | 2580 content, refractive indices, etc.), but for distal tephras in
140
Clay particular, the chemical composition of the glass shards has
0 100 200
S t r a t i gMr a gp n
r eti Magnetic proved to be a more effective basis for identifying and
Susceptibility distinguishing between particular tephras. This geochemical
(SI)
fingerprinting not only enables a distinctive signature to be
obtained for each tephra, but it also provides a basis for
Figure 5.36 Magnetic susceptibility variations through a linking tephras directly to centres of eruption. The
complete Quaternary loess sequence at Luochuan on the
Loess Plateau, Central China. The sequence is divided into geochemical composition is usually determined by electron
pedogenetically altered loess (S – soil, or palaeosol) and probe micro-analysis (EPMA), and usually about ten major
unaltered loess (L) units. The strong magnetic signals in the soil elements (expressed as oxides) are selected for analysis
units reflect higher concentrations of iron oxides (after Balsam (Davies et al., 2002b). More sophisticated geochemical
et al., 2004).
techniques involving, for example, mass spectrometric
analysis of rare earth or trace elements and isotopic
composition are now being employed to distinguish
5.5.2 Tephrochronology tephras, particularly where low numbers of shards have
been recovered (Pearce et al., 2004). Considerable efforts
5.5.2.1 General principles are also being made to standardize the analytical protocols
Following a volcanic eruption, ash or tephra is often used in different laboratories, the reporting procedures and
spread rapidly over a relatively wide area and forms a thin the ways in which tephrochronological data are presented.
cover over peat surfaces, lake-floor sediments, estuarine The age of a tephra can be established by radiocarbon
sediments, river terraces, etc. Thin ash layers have also been dating of associated organic material such as plant
found in deep-sea sediments. As the deposition of the macrofossils (Plunkett et al., 2004) or, in older deposits,
tephra layers is effectively instantaneous in geological time, by argon isotope dating (Mark et al., 2013), fission track
326 DATING METHODS
a) b)
d) d)
2 5 82 5 8
258 2 5 82 5 8
Figure 5.37 a) Yellow and black tephra layers exposed in a soil–peat complex in Iceland (photograph by Christopher Bronk Ramsey,
Oxford University, UK). b) The North Atlantic Ash Zone 2 tephra layer (dark grey band) preserved at 2,359 m depth in the NGRIP
ice core, and with an age of 55.4 ka BP (photograph by Inger Seierstad, Niels Bohr Institute, University of Copenhagen, Denmark).
c) Fluted distal shards of the Campanian Ignimbrite from Campania, Italy, with an age of c. 39 ka (SEM by Suzanne MacLachlan,
National Oceanography Centre, Southampton, UK). d) Platy distal shards of the Vedde Ash from Katla Volcano, Iceland. In the
Greenland GICC05 ice-core chronology, this tephra is dated to 12,171 with a maximum counting error (MCE) of 114 years (photograph
by Jan Mangerud, Bergen University, Norway).
DATING METHODS 327
dating (Marcolini et al., 2003), or ESR and luminescence heterogeneous chemical composition, further complicating
dating (Toyoda et al., 2006) of some of the primary mineral geochemical classification (Shane et al., 2008). Finally, not
constituents. Where tephras are found in ice cores (Figure all geochemical data have been obtained using the same
5.37b), the date of the ash fall can be obtained by annual analytical procedures and protocols, leading to inconsist-
layer counting (Davies et al., 2008), and calendar ages for ency and a lack of comparability in the published data
tephras can also be obtained from varved sediment (Pollard et al., 2006).
sequences (Blockley et al., 2007b). Other means whereby
tephra can be dated include stratigraphical position in
5.5.2.3 Applications of tephrochronology
relation to other tephra layers, palaeomagnetic correlations
and relationships to oxygen isotope stage boundaries in Tephrochronology is now widely and routinely employed
deep-ocean sediments (Wastegård & Rasmussen, 2001). to correlate and date Quaternary sequences. A large number
of tephras have been discovered since the mid-1990s, and
with the associated technical advances, particularly in the
5.5.2.2 Sources of error in tephrochronology
detection and geochemical fingerprinting of non-visible
Two major difficulties constrain the effectiveness of tephra distal tephras (section 5.5.2.1), the global tephra database
as a chronological and correlative tool: the first relates has expanded enormously (Froese et al., 2008). The result
to the distribution of tephra and its incorporation into is that regional tephrostratigraphic frameworks have been
sedimentary sequences, while the second involves ongoing developed for many parts of the world, particularly for the
analytical problems. In terms of distribution, it is clear Late Quaternary (Haflidason et al., 2000; Shane, 2005;
that tephrochronology can, at best, only form a basis for Lane et al., 2010). In each of the regional schemes, there are
regionally applicable schemes of dating and correlation, for normally distinctive marker tephras that form the thickest
individual ash layers are restricted spatially by factors such and most widely distributed of the layers, and that can be
as the magnitude and type of volcanic eruption, the strength detected even in the most distal locations. These have
of the prevailing wind, and the direction of the prevail- usually resulted from cataclysmic, caldera-forming events
ing wind at the time of eruption. Many volcanic plumes (‘super-eruptions’: Mason et al., 2004). Examples include
appear to have been relatively narrow, and hence the the Lava Creek B tephra from the Yellowstone Caldera that
spatial distribution of tephra fall-out will be restricted. Post- was spread across much of the western USA at c. 640 ka
depositional processes will also influence tephra presence (Alloway et al., 2007a) and the Y5 tephra from Campania
in the sedimentary record. Tephras falling over the high- (see below). The largest known eruption of the last two
latitude oceans may be prevented from reaching the seabed million years, the Toba Caldera in Sumatra (c. 74 ka),
by surface ice cover; moreover, these tephras may sub- deposited tephra over India, Malaysia, much of the Indian
sequently be transported by ice-rafting and deposited Ocean, and the Arabian and South China Seas (Williams
only when the ice melts. In this case, older tephras would et al., 2009). Tephras from eruptions of lower magnitude,
be incorporated into younger stratigraphic horizons (Bond such as Glacier Peak and Mt St Helens (c. 7.5 ka) in the
et al., 2001a). In peatland areas, a combination of protracted northwest USA, tend to have more limited geographical
snow cover, wind scour and redeposition by meltwater on distributions (Figure 5.38). Some volcanoes have been
the bog surface may result in only a fragmentary tephra active over long periods during the Quaternary, leaving
record being preserved in peat sequences (Bergman et al., multiple tephra layers that can help address a number of
2004). chronological issues.
In terms of analytical problems, correct geochemical The rate of dissemination of tephra in the atmosphere
sourcing to a particular eruptive event may not always be and the potential extent of its distribution was graphic-
possible because of a lack of data on volcanic histories and ally illustrated by the eruption of Eyjafjallajökull in
diagnostic chemistries of eruptive events. Moreover, some southeastern Iceland in April 2010. Following the onset
tephras (especially basaltic layers), once deposited in of the eruption, satellite imagery showed the rapid trans-
sediments, may have a poor chemical stability. This could portation of ash by westerly winds over western Europe,
result in major alterations to the tephra geochemistry and eventually expanding to large swathes of the North Atlantic
pose problems for the correct geochemical typing and Ocean and the eastern seaboard of Canada. This event
provenancing of the tephra (Pollard et al., 2003). Further caused unprecedented disruption to air traffic over Europe,
difficulties arise because successive tephras from the same a region generally considered to be free from hazards
volcanic source can have very similar chemical signatures, associated with volcanic eruptions, and it constitutes a
while individual tephras can be generated by magmas of remarkable modern analogue for Late Quaternary Icelandic
328 DATING METHODS
b) Reworked Glacier Peak G b)
Glacier Peak G
Glacier Peak G
k G
Pea
cier
Gla
M L St. Helens J
Mt. St. H e l e n s J
56 •125* -120° -115* -110'

0 50 100150km 5
Saskatchewan
Alberta
0 50 100 mi 3 (Irvine), 67
-120
British Columbia 1 2
EJtliillHrir'iriTn 56 56 Frater Lake

13 .21 SO
33 52 24
IS
Glacier 42
Peak! 56 56 26 (Marias Pass;
44 5S 30
37
4/
36 56
54'I 45 is
(Trinity 39 Montana
mine) 5S- 29
-120• Porter ff 978)
49 57 13 G distribution
48 43 McFeelu Rd 25'
55 22
Washington 60 Zanoar Jet ,15 2B
7 27
32
31 a 23
33
45 Oregon 17
9 McCallfenj 65
10
,11 Poison Lafce bog' 61
6S
E3
Tephra localities •H Van Wyck bog 53
E
B (Chiwawa) 64
Porter (1978)
12
G (Manyberries) B distribution
Two or more beds
Not reported or ambiguous
Distribution areas; Idaho Wyoming
B G
Figure 5.38 Tephra layers exposed in sections at two sites in Montana, northwest USA: a) Maris Pass and b) Frater Lake. The
tephras are from the Glacier Peak G and Mt St Helens J volcanic eruptions, the former dating to c. 11,600 14C yr BP and the latter
to a few decades earlier (from Kuehn et al., 2009, reprinted with permission of Elsevier; photographs by Steve Kuehn, Concord
University, West Virginia, USA). c) Mapped distributions (‘footprints’) of the Glacier Peak G and B eruptions, the ages of which
are too close to discriminate between.
DATING METHODS 329
eruptions and their potential effects throughout the North Pacific (Newnham et al., 2003), Japan to the Pacific
Atlantic region and adjacent land areas (Davies et al., (Machida, 1999) and Latin American to the Caribbean
2010). As in other parts of the world, some Icelandic (Jordan et al., 2006). Ash layers in polar ice sheets enable
volcanoes have been active over long periods of the Qua- correlations to be established between ice-core records
ternary, leaving multiple tephra layers that can be used as and those from both the marine and terrestrial realms
a basis for dating and correlation. Some examples are (Davies et al., 2008). Examples from Antarctica, and parts
considered in the following section. of the Southern Ocean and adjacent lands, are described by
Narcisi et al. (2005, 2010), and from Greenland and the
Land–marine-ice correlations North Atlantic region by Haflidason et al. (2000) and
Tephra isochrons constitute one of the most important Wastegård & Rasmussen (2005). Also in the North Atlantic
bases for correlating terrestrial records. For example, the province, tephra isochrones underpin an event stratigraphy
Y5 tephra mentioned above is one of several clear tephra (section 6.3.2.6) for the Last Termination which links ice-
isochrons that link central Mediterranean and eastern core, marine and terrestrial records (Lowe et al., 2008b).
European stratigraphical records (Figure 5.39), while the
Icelandic Vedde Ash (Figures 5.26 and 5.37d) provides a Marine reservoir errors
basis for correlation of Lateglacial sequences across large Radiocarbon marine reservoir errors (section 5.3.2.4)
areas of northwest Europe (Blockley et al., 2007c; section can be estimated for times in the past by comparing
7.5.5). Tephra layers can also be used to link terrestrial and radiocarbon dates on marine fossils with those from
marine sequences, for example New Zealand to the western contemporaneous terrestrial materials (preferably plant
S o u r c e area Distal archives
South Tyrrhenian Ionian Russia

Volcano Event Monticchio Adriatic S e a Balkans
Gregorio Sea Sea Plain
0 CM92-42 MM0-S17 KET8218
Pomici di
Aveltino
Somma-Vesuvio TM-3C
Aveltino
10
C a m p i Flegrei N . Y e l l o w Tuff S21 TM-8 C2 200 395 C2
Etna B i a n c a villa TM-11 Et-1 490 Y-1
P o m i c i di 20
Somma-Vesuvio TM-13 595
Base
C a m p i Flegrei VRa 30 S19 TM-15 C7 Y3 Y3

Somma-Vesuvio? Codola S18 TM-16b C10 400 C10
C a m p i Flegrei
Campanian
Ignimbrite
40 S18 TM-18 C13 450 C13 Y5 Y5 m
50
T. V e r d e
Ischia S16 TM-24 C18 Y7
Epomeo
60
Susceptibilit
70
Susceptibilit
80
i
100
Campi Flegrei? Unknown S11 TM-24 C27 710 X5
Campi Flearei? Unknown S10 TM-25 C31 X6
110
Figure 5.39 Widespread tephra marker horizons in marine and terrestrial sequences in the central Mediterranean and eastern
Europe (after Sulpizio et al., 2008).
330 DATING METHODS
macrofossils), and one way of establishing contemporaneity activity, but they form time-stratigraphic frameworks
is by tephrochronology. For example, by using tephra within which different environmental processes can be
horizons as time-stratigraphic markers, Eiríksson et al. compared and evaluated.
(2004) found that the magnitude of the marine reservoir
error in the Iceland Sea varied by several hundred years Volcanism and human evolution
over the last 4,500 years, probably as a result of periodic Volcanism has also been implicated as an influence in
incursions of Arctic water masses. A similar approach human evolution. For example, in the East African Rift
using tephra isochrones has been employed to determine Valley, volcanic activity may have been part of a range of
the extent of marine reservoir offsets in the Mediterranean tectonic processes that affected behaviour or even speci-
and Black Seas (Kwiecien et al., 2008). ation within early human groups (King & Bailey, 2006).
Similarly, it has been suggested that volcanic winters
Volcanic histories initiated by major volcanic eruptions, such as Toba, may
Eruption histories for individual volcanic centres are have caused ‘population bottlenecks’, that is, marked
often incomplete because of destruction of older beds by reductions in human population levels which catalysed
younger explosions, and the limited registration of minor genetic drift and local adaptations in behaviour (Ambrose,
volcanic eruptions (Bacon & Lanphere, 2006). Further- 1998), although this proposal has proved to be conten-
more, older volcanic materials may be buried by younger tious (Petraglia et al., 2007). In Europe, the Campanian
deposits and hence are inaccessible. Distal records of tephra Ignimbrite tephra fallout at c. 39 ka may have fostered the
deposition can therefore provide evidence of volcanic expansion of Homo sapiens at the expense of Neanderthals
activity that is missing from the proximal sequences (Wulf who were less adaptable to cope with the disaster (Fedele
et al., 2004). Well-dated eruption histories are particu- et al., 2002), although in this case also other evidence
larly important for establishing the recurrence patterns of suggests that the volcanic effects on human populations
volcanic activity (Marra et al., 2009). These form the basis may have been less far-reaching (Lowe et al., 2012). In all
for the development of hazard prediction models (Turner of these instances, however, tephrochronology provides a
et al., 2008), and for evaluating the effects of, for example, basis for evaluating the competing hypotheses by providing
glacial unloading (Jellinek et al., 2004) and sea-level change a secure chronological framework within which to examine
(D’Argenio et al., 2004) on long-term volcanic behaviour. the different lines of evidence.
Volcanism and climate

Possible links between volcanic activity and climate
5.5.3 Oxygen isotope chronology
remain controversial, with some arguing that there is little It has already been seen (section 3.10.2) that the oxy-
evidence for significant climatic impacts from recent gen isotope trace in deep-ocean sediments represents a
eruptions (Shindell et al., 2003), while others point to proxy record of long-term climatic change. Moreover, as
coincidences between volcanic activity and recurrent the isotopic signal is geographically consistent and can
climatic phenomena such as ENSO (Adams et al., 2003; be replicated in cores taken from different areas of the
section 7.6.4.2). Over longer timescales, Bay et al. (2004) world’s oceans, inflections in the isotopic profiles are
have noted a possible causal connection between increased essentially time-parallel events, and constitute age-equiva-
volcanic activity (as reflected in ash layers in polar ice) lent marker horizons. As we explained in Chapter 1, these
and millennial-scale climate change, while Prueher & Rea form the basis for a globally applicable scheme of oxygen
(1998) have suggested that a tenfold increase in the number isotope stages (section 1.6) and, moreover, it is possible to
and thicknesses of volcanic ash layers across the North derive ages for the major isotopically defined horizons
Pacific implicates volcanic activity in the rapid onset of using the technique of orbital tuning; in other words,
global cooling at 2.67 Ma. There is also the possibility that using the astronomical forcing frequencies of the 23 ka
climatic change could, by contrast, have affected volcanic (precession), 41 ka (obliquity) and 100 ka (eccentricity)
activity, perhaps through increased stress on the lithosphere Milankovitch variables to time the cycles reflected in the
by ice sheet loading and unloading, by associated fluc- marine oxygen isotope records. This approach was pion-
tuations in global sea level, and perhaps also by changes in eered by Imbrie et al. (1984) in their production of a
the earth’s axial spin rate, all modulated by Milankovitch standard template for oxygen isotope records known as
harmonics (Rampino et al., 1979). Tephrostratigraphical the SPECMAP timescale. A high-resolution chronology
records are important in the examination of these issues, was based on the amalgamation of several isotopic records
as they not only provide a basis for quantifying past volcanic (‘stacked’ records), and the composite curve which spans
DATING METHODS 331
tibilit
Susceptibilit
3.0
sceptibilit
3.0
3.0 3.0
3.0
3.0
3.03.0
e p.0
3.03.0
3.0
3.03.0
3.0
3.03.0
3.0
u s c3
3.0
3.0
3.0
3.0
3.0
S3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
S u3.0
3.0
3.0
3.0
1
l i tu s c e p t i b i l i t
1
S u s c e p t i b iS
-1
0 50 100 150 200 250 300
T i m e (ka B P )
Figure 5.40 Orbitally tuned chronostratigraphy for a composite (‘stacked’) deep-ocean isotopic record spanning the last 300 ka.
The numbered vertical lines indicate distinctive features in the record. In the hierarchical numbering scheme, each number before
the decimal point refers to the marine isotope stage (1, 2, 3, 4, 5 . . .) while the number following the decimal refers to substages
within each of the major isotopic stages (see section 6.2.3.5) (after Martinson et al., 1987).
250 the last 300 ka was then smoothed, filtered and tuned to
the known cycles of the astronomical variables (Figure
7.5
5.40). This now classic orbitally tuned chronology has
70 7 3
200 since been confirmed by independent dating of fossil
7.1
corals using U-series (Figure 5.41). Since then, a num-
t usceptibilit
ber of tuned stacked records have been constructed (e.g.

6.5
150 6.0 Martinson et al. 1987; Karner et al., 2002), the longest
5 51 6.41
6 4 and most comprehensive of which is the LR04 stack of
S u s c e p t i b i l iS
52 Lisiecki & Raymo (2005). This consists of fifty-seven

100 51 5.4
globally distributed benthic δ18O records, and forms a con-
4.23 5.33
5.31 tinuous orbitally tuned sequence extending back into the
3.1 early Pliocene (Figures 1.5 and 5.1). The stacked record
50 7 21
2 2 3.31 shows a significant coherency with insolation as reflected
3.3 in the obliquity band throughout the entire 5.3 Ma record,
0
and in the precession band throughout more than half of
0 SO 100 150 200 250
that time interval. As such, it constitutes the most detailed
C o r a l a g e (ka) oxygen isotope chronostratigraphy developed to date, and
forms a ‘type sequence’ against which all palaeoceano-
Figure 5.41 Comparison of orbitally tuned age estimates graphic measurements can be compared.
based on SPECMAP and independently dated U-series coral
ages. The numbered data points correspond to the SPECMAP
marine isotopic events as defined by Martinson et al. (1987): 5.5.4 Biostratigraphy and molecular
5.51 – Last Interglacial; 7 – Penultimate Interglacial (after
Thompson & Goldstein, 2006). clocks
As noted in Chapter 4, within a restricted geographical area,
vegetational changes will have been broadly synchronous,
and hence the boundaries between regional pollen
assemblage zones will be effectively time-parallel. In
332 DATING METHODS
essence, therefore, these form marker horizons in lake or to occur at predictable rates, then molecular clock analysis,
peat profiles and can be used as a basis for correlation and which is based on the timings of genetic mutations, may
dating. They have been of particular value in the dating enable dates to be assigned to events or to stratigraphic
of Holocene sequences. In Britain, for example, regional sequences (Bromham & Penny, 2003; Kumar, 2005). This
pollen assemblage zones have been established largely on approach has already been applied to the dating of river
the basis of the establishment and expansion of key drainage and other geological processes (Craw et al., 2007,
indicator taxa, notably particular woodland trees. Radio- 2008).
carbon dating enables the events to be set within a time-
stratigraphic framework, and includes the arrival of Corylus
avellana, dated to 10.5–10.0 ka, and the expansion of 5.6 RELATIVE CHRONOLOGY
Alnus around 8.0 ka. A key marker horizon is the Ulmus BASED ON PROCESSES OF
decline, an event that is considered to be broadly syn-
chronous and has been widely dated in both Britain and in
CHEMICAL ALTERATION
western Europe to c. 5.8 ka (Parker et al., 2002). Such data Fossils, sediments and rocks are affected by a number of
form the basis for isopollen maps (Chapter 4) showing the chemical reactions that are partly time-dependent. Upon
patterns of expansion of arboreal taxa and threshold ages the death of an organism, tissues are broken down by
for the arrival of certain tree species in particular localities a variety of chemical processes to produce compounds of
(van der Knaap et al., 2005). Planktonic assemblage zones a more simple chemical structure; the surfaces of fossils
can provide similar biochronological frameworks for the or minerals may be altered by the effects of hydration or
marine realm (Capotondi et al., 1999), while sapropels the accumulation of precipitates of certain chemicals in
(Chapter 1, note 6) may also form isochronous marker groundwaters, while weathering and pedogenic processes
horizons that offer a basis for correlation of some marine will gradually effect changes on rock and sediment surfaces.
deposits (Meyers, 2006). In all of these cases, the degree of alteration brought about
Although not common, evolutionary changes do by different chemical reactions increases with time, and this
occur within the Quaternary (section 4.11.6.3) but one that therefore offers a basis for relative dating. Some of the more
is of particular significance for dating the Middle Pleisto- widely used techniques are considered in the following
cene in Europe involves the lineage of the large water section.
vole, Arvicola terrestris cantiana. This species, with un-
rooted molars, is thought to have evolved from an earlier
species of vole, Mimonys savini (with rooted molars) during
5.6.1 Amino-acid geochronology
the ‘Cromerian’, that part of the Middle Pleistocene record Living bone consists of approximately 23 per cent colla-
which may contain at least five temperate episodes (Preece gen (protein-bearing) fibrils bound within a phosphatic–
& Parfitt, 2012). The Mimonys–Arvicola boundary, which calcitic matrix. Proteins can survive in bones and shells
occurs after the Brunhes–Matuyama geomagnetic boun- for extremely long periods,3 but undergo a number of
dary (~MIS 19) but earlier than the Elsterian–Anglian molecular changes. Protein residues were discovered in
glaciation (MIS 12), forms an important marker horizon fossil bones and shells in the 1950s, and since then the
within Cromerian deposits at sites in western Europe (van study of protein transformation in the geological record
Koenigswald & van Kolfschoten, 1996). It has proved has developed rapidly. Some of the chemical changes in
extremely valuable in dating and correlating sites for a time proteins that occur after the death of organisms are time-
period for which few other dating methods are available dependent, and thus the characteristics of certain protein
(e.g. Stuart & Lister, 2001; Preece et al., 2009). residues from the Quaternary record can provide the basis
Recent developments in the recovery and sequencing for a relative chronology. Useful overviews of amino-acid
of mitochondrial DNA from fossil remains, however, geochemistry and its applications in geochronology can be
indicate that evolutionary changes may be a more common found in Goodfriend et al. (2000), Wehmiller & Miller
feature of the Quaternary fossil record than hitherto (2000), Blackwell (2001b) and Penkman (2010).
considered. Evidence from a number of different fossil
residues, including reptiles (Douglas et al., 2006), mammals
5.6.1.1 Chemistry of proteins
(Lister, 2004) and birds (Lambert et al., 2002), suggests that
faunal turnover and speciation may have been both Proteins are large and complex molecules and are basic
common and significantly increased during the Quaternary ingredients of all living organisms. They are composed of
(Barnes et al., 2007). If evolutionary turnover can be shown amino acids, which have the generalized chemical formula
DATING METHODS 333
shown in Figure 5.42a. The ‘R’ linkage differs for each commonly found in proteins can exist in two molecular
amino acid, ranging from a simple hydrogen atom in forms (isomeric forms). The chemical and biochemical
glycine, to a methyl group (CH3) in alanine, and to highly properties of the two forms of amino acid are similar, but
complex chemical structures in other amino acids. Around they rotate plane-polarized light in opposite directions.
eighty amino acids occur in nature, of which about twenty Effectively, they constitute two non-superimposable mirror
are commonly found in proteins. Proteins form through images (Figure 5.42d), rather like left and right hands, and
the combination of several amino acids into peptide chains, these optical isomers are referred to as L-amino acids and
each amino acid linked to the next by a peptide bond D-amino acids (after the Latin laeva – left-handed, and
following the loss of a water molecule (Figure 5.42b and c). dextra – right-handed). The carbon atom at the centre of
The patterns formed in this way are specific to each protein the isomers (the chiral carbon atom) forms the point of
type, and the various amino-acid arrangements create asymmetry and allows the development of the two optical
thousands of different proteins, including enzymes and isomers. The biological significance of this distinction
antibodies. Some of the larger proteins may contain up to between the L- and D-configuration in amino acids is that
3,000 amino-acid residues. A list of the more common only L-isomers occur in living (active) proteins. D-isomers
amino acids includes alanine, glycine, isoleucine, leucine can occur in a free state as components of non-protein
and aspartic acid. structures, and in fossil organic materials as a result of the
With the exception of glycine, all amino acids break-down of proteins (see below).
a) H"
H"
H"
H H.
R H" SI
H" o H
Side chain group C C 0
N. c N G
C
H H
9 3
H H" H" OH
Amine H H
group H G Carboxyl water O O water
group H
o
R.
H O
i
St
c N H" carboxyi terminus
amino terminus C c.
d)
(N terminus)
N 0 ) (C terminus)
H 0 H H" H
D-A!anine L-Alanine
COOH iCOOH
E R,
R.
3
H H o H H o
G G
CH 3
CH 3
C N C c N c
H NH 2 NH, H N c c N C c N
H in
H M
R
ffl 3 M
2
R,
Figure 5.42 Chemistry of amino acids. a) Generalized formula for amino acids. b) Schematic representation of chemical
combination of amino acids through loss of water molecule to form a peptide bond. c) Combination of carboxyl and amino
groups between amino bonds to form peptide chains. d) Isomers (mirror-image molecular arrangements) of alanine (from Petsko
& Ringe, 2004).
334 DATING METHODS
5.6.1.2 Amino-acid diagenesis and organic residues in cave speleothems that were origin-
ally derived from surface soils (Blyth et al., 2008). AARs
Chemical alteration in proteinaceous residues following the have also been obtained from organic materials incorp-
death of an organism results in the disruption of peptide orated into carbonate rocks in coastal areas, often referred
chains to release free amino acids. Where proteins are to as ‘whole-rock aminostratigraphy’ (Hearty & Kaufman,
exposed to the atmosphere or to biological processes, very 2000; Hearty, 2003). More problematic media include
rapid degradation will take place, but if they are protected wood, where difficulties arise from such non-carbonaceous
by skeletal hard parts so that a ‘closed’ system prevails, then materials not being buffered against changes in environ-
much slower chemical alterations occur. Some reaction mental pH (Sykes, 1991), and bone, where open-system
times are in the range of 50 ka to a few million years, behaviour results in post-burial loss of proteins through
whereas other amino-acid reactions operate over a timescale leaching or contamination by younger amino acids from
of only a few thousand years. the surrounding sediment matrix (Collins et al., 2000;
Several indices have been employed to measure the Fernández et al., 2009).
amount and rate of degradation of amino acids in fossil
samples, including the ratio between bound- to free-amino
acids, with respect to either the total amino compounds in 5.6.1.3 Aminostratigraphy and age control
the sample, or for selected amino acids. From the point of Amino-acid racemization (or epimerization) does not
view of Quaternary geochronology, however, the most progress at the same rate in all species, nor in fossils of
important diagenetic change that takes place in subfossil the same species recovered from different sites (see below).
material is the inversion (or intraconversion) of L- to By themselves, therefore, AARs cannot provide numerical
D-configured molecules in the constituent amino-acid age estimates. The rate of the degradation process needs to
residues, a reaction that proceeds until an equilibrium be calibrated for individual sites and species, and for
state has been achieved. In the case of amino acids that form different types of host material, using independent dating
stereoisomers, this conversion is called racemization. methods such as radiocarbon, U-series or luminescence
However, some common amino acids (e.g. aspartic acid) measurements (Owen et al., 2007; Kosnik et al., 2008).
are diastereomers: their D- and L-molecules are not mirror Calibrated AAR age models (AAR chronology) show that
images of each other. In this instance, the conversion D/L conversion values can be traced back over considerable
process is known as epimerization. D/L ratios start at zero periods, for well-preserved amino acids have been recovered
at the time of death of an organism and increase to an from ostracods (Ortiz et al., 2004) and marine molluscs in
equilibrium ratio, which for the process of epimerization deposits older than 1 Ma (Table 5.6). Aminostratigraphy
in some molluscan species is a value of 1.30 ± 0.05. Fossil appears to be potentially applicable to the entire timespan
ratios are measured proportional to position on the of the Quaternary (and beyond?), for it has been employed
degradation continuum from 0–1.30 (e.g. a ratio of 0.65 to date, inter alia, some of the earliest Pleistocene marine
indicates that the sample is half of the way along the formations in the British Isles (Bowen, 1991), Plio-
conversion continuum, which is reported as a value of Pleistocene raised marine sediments in New Zealand
0.5). Amino-acid ratios (AARs) are normally measured (Bowen et al., 1998) and Pliocene glacial sediments in
by means of ion-exchange chromatography or gas–liquid Antarctica (Colhoun et al., 2009).
chromatography coupled to mass spectrometers, and
kinetic models are employed to determine the trend of
5.6.1.4 Problems with amino-acid
protein diagenesis.
Although in theory amino-acid determinations can
geochronology
be made on any protein-bearing materials, such as hair, The relationship between D/L ratio and time is complicated
teeth, Foraminifera, shells, bones and tusks, in practice by a number of factors. All reactions are temperature-
suitable samples are restricted to those with ‘tight’ skeletal dependent, with epimerization proceeding more rapidly
carbonate matrices (Penkman et al., 2008). The most at higher ambient temperatures. Hence, meaningful com-
widely used media have been marine molluscan shells parisons of amino-acid ratios can only be made between
which form an essentially closed system that is affected sites that have experienced a similar climatic, and par-
by a minimum of external factors. Other materials have ticularly thermal, history. Amino-acid composition, relative
been used with some degree of success, however, includ- abundances of amino acids, and rates of various amino-
ing freshwater molluscs, land snails and ostracods (Ellis acid reactions are all genus- and, in some cases, species-
et al., 1996; Kaufman, 2003a), ostrich eggshells (see below), dependent, and hence precise identifications of shell
Table 5.6 Age calculations based on D/L ratios of five amino acids measured in fossil terrestrial ostracods recovered from alluvial and travertine
sequences in central and southern Spain. All error ranges are 1σ (from Ortiz et al., 2004).
Level in sequence D-alloisoleucine/ D/L leucine D/L aspartic acid D/L phenyl-alanine D/L glutamic acid Age (ka)
above datum (m) L-isoleucine
206.5 0.997 ± 0.021 0.710 ± 0.030 0.780 ± 0.066 0.805 ± 0.007 0.703 ± 0.083 1,012 ± 169
223.8 1.045 ± 0.035 0.735 ± 0.007 0.820 ± 0.014 0.780 ± 0.000 0.725 ± 0.092 1,008 ± 125
228.1 0.819 ± 0.000 0.828 ± 0.000 0.711 ± 0.000 0.678 ± 0.000 0.594 ± 0.000 863 ± 173
253.0 0.725 ± 0.046 0.570 ± 0.061 0.700 ± 0.013 0.502 ± 0.035 0.534 ± 0.042 743 ± 88
269.3 0.736 ± 0.055 0.544 ± 0.055 0.683 ± 0.040 0.519 ± 0.094 0.579 ± 0.100 736 ± 115
281.2 0.877 ± 0.046 0.603 ± 0.031 0.687 ± 0.004 0.603 ± 0.029 0.599 ± 0.007 737 ± 68
303.4 0.675 ± 0.021 0.490 ± 0.014 0.605 ± 0.021 0.435 ± 0.007 0.445 ± 0.007 517 ± 56
314.0 0.520 ± 0.071 0.373 ± 0.051 0.534 ± 0.006 0.436 ± 0.099 0.376 ± 0.037 407 ± 58
323.3 0.529 ± 0.106 0.397 ± 0.098 0.522 ± 0.016 0.429 ± 0.085 0.397 ± 0.031 409 ± 95
327.1 0.492 ± 0.011 0.397 ± 0.031 0.492 ± 0.004 0.430 ± 0.040 0.342 ± 0.015 389 ± 62
329.8 0.395 ± 0.037 0.336 ± 0.001 0.481 ± 0.054 0.320 ± 0.000 0.40 ± 0.000 339 ± 68
352.5 0.332 ± 0.001 0.279 ± 0.081 0.441 ± 0.019 0.346 ± 0.051 0.325 ± 0.005 279 ± 77
336 DATING METHODS
samples, for example, are necessary before analyses are to aragonite. As the former is more stable, inter-crystalline
undertaken (Miller & Clarke, 2007). A further difficulty is proteins should be better preserved in this biomineral,
that proteins in different structural layers of a shell may and hence are likely to provide better temporal resolution
often differ in terms of their amino-acid composition and than those in aragonitic shells. This has been confirmed by
rate of epimerization (Penkman et al., 2008). In theory, recent analyses of calcitic opercula4 from the freshwater
therefore, comparable portions of different shells should mollusc Bithynia from sites in southeastern England
be used for analysis, but this principle is difficult to apply and the Netherlands, the results of which have formed the
where samples are taken from fragmented molluscan basis for a coherent aminostratigraphic framework for
assemblages. Degradation of proteins due to micro- the whole of the Quaternary. In addition, samples from
biological activity and varying degrees of oxygenation, the Eocene demonstrate the persistence of closed system
particularly during the early stages of diagenesis, may also protein within the calcitic opercula extending back over 30
result in significant differences in amino-acid ratios in Ma (Penkman et al., 2013).
molluscs from sites of the same age and with the same
thermal histories (Takano et al., 2005). Furthermore, some
5.6.1.6 Some applications of amino-acid
AAR diageneses may be characterized by non-linear and
geochronology
reversible reactions (Clarke & Murray-Wallace, 2006).
Finally, AARs are naturally variable, and it will frequently Correlation and dating of Quaternary sequences
be the case that a range of values will need to be obtained One of the most widely used applications of this technique
in order to generate an index that adequately reflects the has been in the correlation and dating of stratigraphic
AAR signature of any one stratigraphic context. In other sequences containing molluscs. This involves the develop-
words, relative age estimates must be sufficiently large to ment of aminozones, each of which is characterized by a
capture this natural variability (McCarroll, 2002). These distinctive range of D/L ratios. Stratigraphic units can be
difficulties are not insurmountable, however, and a number correlated on the basis of comparable D/L values, providing
have been resolved by some recent developments in AAR they were obtained from the same species and molecular
analysis, which are discussed in the following section. compounds. This approach is applicable only to regions
that are likely to have experienced the same thermal
histories during the Quaternary, however. Order of super-
5.6.1.5 Recent developments in amino-acid
position of the stratigraphic units provides an independent
geochronology
test of the integrity of the aminozones, which is important
Since the mid-1990s, there have been a number of method- in view of potential constraints on aminostratigraphy
ological advances in amino-acid geochronology. These outlined above, and the sequence can be calibrated to real
involve improvements in sample pretreatment and new time using radiometric or other methods. For example,
analytical methods (Sykes et al., 1995; Kaufman & Manley, Ortiz et al. (2004) used palaeomagnetic dating to test the
1998), including the technique of sample bleaching to AAR chronology for central and southern Spain shown in
remove the leachable, open-system matrix of shell protein. Table 5.6. Amino-acid geochronology has been employed
This leaves a component, known as the intra-crystalline to date eustatic changes and interglacial marine sequences
fraction, that exhibits closed system behaviour, and within in, for example, the Bahamas, western Australia, the eastern
which protein degradation is dependent on only two United States and South Africa (Hearty & Kaufman, 2000;
variables, time and temperature (Penkman et al., 2008). Parham et al., 2007; Carr et al., 2010). Coastal aeolianite
In this fraction, it has proved possible to isolate five amino sequences, in some cases extending back over 500 ka (Figure
acid D/L pairs (aspartic acid, glutamic acid, serine, alanine 5.43), have been dated by this method, but it has also been
and valine). As these racemize at different rates, this applied to the dating and correlation of much younger
provides a cross-check on the geochemical integrity of deposits (Hearty et al., 2004).
the sample (see also Table 5.6), and also enables differ- The extension of the amino-acid technique to non-
ent aspects of diagenesis to be monitored within a closed marine Mollusca has provided a basis for the dating and
system. Moreover, as the overall extent of inter-crystalline correlation of terrestrial deposits, such as loess–palaeosol
protein decomposition is being measured, the method sequences (Oches & McCoy, 2001) and fluvial terrace
offers a significantly improved temporal resolution over sequences (Westaway, 2009). In the lower Thames Valley
earlier analyses based on the racemization of a single amino of southeast England, for example, D/L ratios from opercula
acid (Penkman et al., 2007). A further development has of the freshwater gastropod Bithynia provided a chrono-
been the use of the calcitic component of shell as opposed logical context for the terrace sequence extending back to
DATING METHODS 337
O
138°E 140°E N
ADELAIDE b!
O
AR
b!
G AG R
O
SOUTH AUSTRALIA
AR
N
Guff Murray Bridge
KA
KAN
N OG
St Vincent.
Pinnaroo
K AA R
NG
140°
Alexandria
Cape
KA
Jan* Coonalpyn
KAN
GAR M e n ingle
K A NOISLAN
GAR
CspsVmoighby
O
KANGAROO 361S
Coorong Lagoon Coorong Coastal Rain
ISLAND
KA
Bordertown
NG
Susceptibilit
KA
AN
RG
KA
OAR
SOUTHERN
KANGARO
NG
K OCEAN
ANGARO KA b!
O
AR
NG
Kingston S E
OK
KAAR O
KA KA
A
NG
NG
NG NG
A
K AA R OA R O
AR
Robe
R KO
Relict c o a s t a l barriers
NG
O
AN
AR
GA
MacDonnel
O
RO
Mount
Gambler
AUSTRALIA
38TS
Port M a c D o n n e l l
0 Gamble
d)
1.0 y = 0.00091x+0.189
r = 0 . 9 8 8 9 r2 = 0 . 9 7 8 1
n = 8
0.8
CWT MM
Susceptibilit
06 KB PS PMcll
RS
RR
0.4 SB
02
PB
0
0 100 200 300 400 500 600 700
J T L A g e (ka)
Figure 5.43 Amino-acid geochronology of coastal barrier sediments (aeolinites) in the region of the Murray River mouth, South
Australia. a) Location of the aeolinite ridges. b) Cliff section in one of the barrier ridges showing cross-bedded aeolianite deposits
(dune limestone or calcarenite) of MIS 5a and 5c age overlain (at surface) by Holocene coastal dunes. c) Backbarrier shell-rich
lagoonal facies of the last interglacial age, dominated by the cockle Donax deltoides, exposed on Hindmarsh Island, Murray River
mouth. d) Extent of valine racemization (total hydrolysable amino acids) in ‘whole-rock’ sediment samples plotted against the square
root of TL age. Letters indicate names of beach and spit sections from which the samples were obtained (after Murray-Wallace
et al., 2010, reprinted with permission of Elsevier; photographs by Colin Murray-Wallace, University of Wollongong, Australia).
338 DATING METHODS
MIS stage 11 (Figure 5.44a), while data from a number of

m OH
sites in southern and eastern England enabled a temporal a)
40
framework based on aminostratigraphy to be reconstructed
for faunal (Figure 5.44b) and archaeological (Figure 5.44c) 40
hj U
records that extends back into the Middle Pleistocene, that 40

is, beyond MIS 12 (Penkman et al., 2013). Elsewhere in
40
Europe, terrace sequences of the River Danube in eastern
40
Europe (Oches & McCoy, 1995) and the River Dnieper
in the Ukraine (Oches et al., 2000) have also been differ-
Susceptibilit
entiated on the basis of aminostratigraphy. Similarly, in 40
the Somme Valley of northern France, the chronology

of a well-developed staircase of fluvial terraces is based 40
on a combination of aminostratigraphy, radiometric and

palaeomagnetic dating, enabling the sequence to be cor- 40 DD 01 04 0*
D L FAAAIa
related with the marine oxygen isotope record (Bridgland
b)
et al., 2004). 1 D
Dating of ratite eggshell in archaeological contexts

Unlike bone or mollusc shell, ratite eggshell (ostrich, emu 0I
and the extinct ratite, Genyornis) has a compact structure
which protects the protein within it from changes in soil Susceptibilit
D f.
chemistry. Indeed, experimental work has shown that Mimomys
ostrich eggshell more closely approximates a closed system

for amino acids than any other biomineral yet evaluated Arvicola
cJ
(Miller et al., 1992), and that in tropical and subtropical
regions, isoleucine epimerization in egg shell should allow Corbicuta
dating back to 200 ka (Brooks et al., 1990). cZ
AARs on ostrich shell, calibrated to a timescale using Hippopotamus

radiocarbon, have been used to develop an aminostrati-
CD
graphic context for human occupation at the key site of 0.0 D.J 0.4 o.s Q.B 1.0
Border Cave in South Africa (Miller et al., 1999a). The DA FAAAIa

b)
reconstructed chronology showed that anatomically D/L FAAAIa
modern human remains were present in levels more than 00
0.2 0.4 0.6 OB
100 ka old, thereby lending support to an African origin Mesolithic

for Homo sapiens. Other examples of the use of AARs in Upper Palaeolithic
Levallois
eggshell include an analysis of the likely role of humans
0.2
in the extinction of Genyornis newtoni (a flightless, ostrich- Palaeolithi
sized bird) in Australia, around 50 ka (Miller et al., 1999b),
possible human-induced ecosystem collapse and mega-
,1.!
faunal extinction, also in Australia (Miller et al., 2005),
Susceptibilit
and Late Holocene faunal extinctions in Madagascar

Clactonian
(Clarke et al., 2006b).
OLfl
Core-and-tiake
industries
OLfl
Figure 5.44 An aminostratigraphic framework for the Thames terraces of southeast England (section 2.6.3): a) the terrace deposits;
b) the faunal record; c) the archaeological record. The numbers in a) refer to MIS stages; in b) and c) THAA refers to the total
hydrolysable amino-acid fraction of the sample; FAA refers to the free amino acids analysed; Ala is the amino acid alanine (after
Penkman et al., 2013).
DATING METHODS 339
Dating of molluscs in archaeological contexts et al., 2006). Screening is particularly important for estab-
One of the most widespread indicators of human lishing the integrity of archaeological sequences and their
exploitation of marine resources and occupation of coastal contained fossils (Ellis et al., 1996).
environments are shell middens. Dating of these deposits
using conventional techniques such as radiocarbon can be
5.6.2 Fluorine, uranium and nitrogen
problematic, however, as the shell middens accumulated
content of fossil bones
relatively rapidly, and there is often a mixing and inversion
of materials (Stein & Deo, 2003). Middens have been Hydroxyapatite, the principal mineral constituent of
successfully dated by AAR, however (Bateman et al., 2008), bones and teeth (section 4.11), absorbs fluorine from
and recent work involving the isolation and extraction of groundwater progressively over time. The rate at which the
amino acids from the inter-crystalline fraction of shell fluorine content increases varies from locality to locality,
offers the potential for even more secure dating of middens but bones that have been buried for the same length of
(Demarchi et al., 2011). Not only is establishing a reliable time in a particular deposit will have approximately the
chronology for these molluscan accumulations important same fluorine content. As the fluorine fixed in the bone is
for understanding human activity in the coastal zone, but not readily removed, a specimen that is washed out of an
it may also be significant in studies of human evolution and older deposit will show a much higher fluorine content
dispersal. than bones that are contemporaneous with the bed, while
bones from a younger stratum will have accumulated
Palaeothermometry
substantially less fluorine. Hence the technique is useful in
Since racemization rates are temperature-dependent,
the relative dating of bones in a deposit, and it also enables
where the age of a fossil protein can be established, amino-
intrusive (i.e. younger or older) material in a fossil assem-
acid ratios can be used to infer the ambient temperature
blage to be identified (Demetsopolous et al., 1983). The
conditions that affected the proteins since the onset of
fluorine to phosphate ratio is established by X-ray diffrac-
diagenesis (McCoy, 1987). On this basis, Miller et al. (1997)
tion techniques, or by the use of an ion-selective electrode
found that AARs from emu eggshells indicated low-latitude
(Schurr & Gregory, 2002). The latter approach has been
cooling in the southern hemisphere during the last glacial
employed in the dating of mammal bones in a range of
stage, while Oches et al. (1996) inferred depressed north–
archaeological contexts (e.g. Johnsson, 1997; McFarlane &
south temperature gradients in the Mississippi Valley
Blake, 2005).
during the past 25 ka on the basis of AARs from fossil
The analysis of uranium incorporated into fossil bones
gastropod species, the southern part of the valley having
from groundwater has been employed in a similar way to
cooled by up to 13°C at the Last Glacial Maximum. Using
the fluorine method, the former having the advantage that
AARs from fossil ostracods, Kaufman (2003b) suggested
the counting of uranium emissions does not involve the
that temperatures in the Bonneville Basin, Utah were about
destruction of the bone material, but it has the disadvantage
10°C cooler during the last glacial stage by comparison with
that bone material does not always conform to the
the late Holocene, and concluded that reduced evaporation
assumptions of a closed system (Millard & Hedges, 1996).
rather than increased precipitation could have accounted
The relative ages of fossil bones in assemblages can also be
for an expanded ancient Lake Bonneville (section 2.7.1).
established by analysing the nitrogen content of fossil bone,
While these temperature estimates broadly concur with
for as proteinaceous materials disappear from bone
those from independent lines of evidence, other studies
collagen, so too will nitrogen. Hence decreasing nitrogen
suggest that local environmental and preservation factors
content will reflect increasing age.
can cause problems for the application of amino-acid
A further means of dating using fluorine content of
palaeothermometry (Reichert et al., 2011), and the
bone involves the establishment of fluorine profiles through
technique, at present, remains largely experimental.
sections of bone using a nuclear microprobe. In older
Screening of fossils for sample integrity bones, the profile is relatively flat, whereas in younger
Because amino-acid racemization studies are relatively samples, the concentration of fluorine falls off steeply with
inexpensive, they provide a cost-effective basis for screen- increasing distance into the bone. The shape of the profile
ing fossils prior to other analyses, such as radiometric is therefore age-dependent and offers the basis for dating.
dating (Kosnik & Kaufman, 2008) or the extraction of The technique can be used to establish relative age in
DNA (Collins et al., 2009). Unusual variations in amino- mixed bone assemblages, providing useful data on the
acid diagenetic behaviour may reveal atypical site condi- taphonomy of animal bone assemblages (Johnsson, 1997;
tions, or the contamination of the assemblage (O’Donnell Tankersley et al., 1998). It may also be a useful screening
340 DATING METHODS
process in the selection of samples for radiocarbon dating, serious are problems relating to the determination of the
and may provide approximate ages for very recent hydration rate which, as noted above, is influenced by a
materials, that is, over the timespan of the last few hun- number of different factors. Calibration of obsidian
dred years. However, uranium profiling suffers from the hydration rates to an absolute timescale is even more
same problems of non-uniform alteration and low levels problematical. These and other limitations have led some
of accuracy and reliability as other chemical methods to question the validity of the method (Ridings, 1996;
considered in this section; moreover, the distribution of Anovitz et al., 1999). Others, however, have come to the
fluorine in archaeological samples appears to be strongly defence of OHD, pointing out that although the temporal
influenced by local environmental factors (Gaschen et al., resolution of the method may not be as good as was once
2008). Indeed, it has been suggested that fluorine (or anticipated, as a relative dating technique it can, never-
fluoride5) dating can only be considered a valid dating theless, provide quite rapid answers to site-specific
method where the chronological validity of its results are problems (Hull, 2001; Rogers, 2007).
confirmed by independent chronometric data (Lyman
et al., 2012).
5.6.3.3 Some applications of obsidian
hydration dating
5.6.3 Obsidian hydration dating (OHD) Applications of OHD dating include the dating of volcanic
activity (Friedman & Obradovich, 1981), glacial events
(Pierce et al., 1976), and fluvial terrace sequences (Adams
Freshly exposed surfaces of obsidian (a form of volcanic et al., 1992). The technique has been most widely used,
glass) absorb water from their surroundings to form a however, in the dating of archaeological contexts, and
hydration layer known as perlite. This external rind should particularly of artefacts (Ambrose, 1994; Tripcevich et al.,
not, however, be confused with the patina that develops on 2012). It now appears that hydration rims may also develop
many materials as a result of chemical weathering (section on quartz and, if so, this could provide a basis for the dating
5.6.4). One factor governing the thickness of the perlite of a much wider range of artefacts (Ericson et al., 2004).
layer is time, and hence by measuring the thickness of this Indeed, the greatest value of the technique may be in the
layer, it may prove possible to estimate the relative age of establishment of relative order of antiquity of artefacts, for
exposure or the breaking of the obsidian surface, working on many archaeological sites, this may be the only dating
on the assumption that the thicker the perlite layer, the method available (Hull, 2001). As with other techniques,
older the exposed surface. This is the basic principle of greater awareness of the constraints has led to refinements
obsidian hydration dating (OHD). In addition to time, of OHD and more coherent age estimates. For example, in
however, the thickness of the hydration layer depends on a study of more than thirty sites in the southern Nazca
the rate at which hydration proceeds and this, in turn, region, Peru, estimated OHD ages were within 15 per cent
reflects a range of factors including the chemical com- of independently obtained radiocarbon ages (Eerkens et al.,
position of the obsidian, and the temperature and humid- 2008).
ity of the hydrating environment (Beck & Jones, 1994).
The thin hydration layer (sometimes less than 1.0 μm) can
be seen in section under a microscope when illuminated by
5.6.4 Weathering characteristics of rock
polarizing light, but computer-assisted imaging technology
is now also used (Ambrose, 1994). More recently, second-
surfaces
ary ion mass spectrometry (SIMS) has been employed to
measure the concentration of hydrogen as a function of
depth in obsidian samples, a technique known as obsidian A range of rock-weathering parameters has been employed
diffusion dating (Liritzis, 2006). OHD dates have been to assign ages to rock surfaces (Mills, 2005). Degree of
reported in the age range from 200–100 ka. weathering of stones or boulders, which is a reflection of
time, is indicated by a range of characteristics including
the relative decomposition or exfoliation of boulders, lack
5.6.3.2 Problems with obsidian hydration
of soluble materials (e.g. limestones) on older surfaces, the
dating
crumbly nature of sandstone or volcanic clasts on older
Numerous difficulties are encountered in the use of OHD. drifts, and the relative concentration of more durable
Some of these relate to analytical constraints, but more materials (e.g. quartz, chert, silicious rocks) at the surface.
DATING METHODS 341
The extent of weathering may also be established by

a)
recording the sound produced by boulders being struck 2 50
b ti li ibt iSl iut s c e p t i b i l i t

with a hammer. Fresh boulders produce a sharp ring and
b ti li ibt iSl iut s c e p t i b i l i t

a strong hammer rebound whereas weathered stones emit
45
a dull sound and a weaker recoil. More quantitative esti-
2
mates may be obtained from a Schmidt hammer where
s cues pc tei p
rock-hardness is measured by the distance of rebound of 40
s cues pc tei p
the controlled impact of a calibrated spring-loaded mass,
S u s c e p t i b iSl iut S
1
S u s c e p t i b iSl iut S
and which has been widely employed in geomorphological
35
research (Goudie, 2006). Indeed, the Schmidt hammer
may also offer a basis for calibrated-age dating, particularly
in Holocene contexts (Figure 5.45a; Shakesby et al., 2006). 2 30
0 2000 4000 6000 8000 10000
The thickness and chemical composition of the weathering t i m e (years)
rind, the outer layers of boulders or stones which have been r-value (Murtel) weathering rind ( G i a n d a G r i s c h a )
oxidized and discoloured by iron-bearing minerals, can weathering rind {Murtel}
also be used to establish relative order of antiquity of rock b)
1.1
surfaces (Figure 5.45a). One approach is cation-ratio
1.0
dating, which is based on the relative stability of chemical
0.9
constituents of rock varnish, the coating of clay minerals,
Susceptibilit
0.8
manganese, iron oxides, etc., that form on rock surfaces in
0.7
arid areas (Gordon & Dorn, 2005). Certain bases, notably
0.6
potassium (K) and calcium (Ca), are easily mobilized
0.5
whereas others, such as titanium (Ti), are more stable.
0.4
Hence the K + Ca/Ti ratio in rock varnish decreases with Late Neoglacial Lateglacial MIS 3
time and provides an indication of age. Magnification of Holocene
rock varnish rinds shows that they are composed of b)
microscopically thin layers, or micro-laminations, the 1 oco

S u s c e p t i b iSl iut s c e p t i b i l i t
number and structure of which also provide an index of

relative age, a technique termed varnish micro-lamination 100
(VML) dating (Liu & Broecker, 2008). Rock varnish also
contains minute quantities of organic material, and where
10
sufficient can be obtained for AMS radiocarbon dating, the
relative chronologies obtained from cation-ratio dating
can be calibrated to a radiometric timescale (Liu, 2003). 1
Late NeogIacial Lateglacial MIS 3
Where well-jointed rock crops out at the surface, variations Holocen
in depth of jointing can also be employed to determine
relative order of age, the deeper and more expanded joints Figure 5.45 Weathering and soil-forming indices used to
indicating a longer exposure to weathering agencies, for establish relative age of landform surfaces. a) Weathering
example under periglacial conditions (Ballantyne et al., rind thickness measurements and Schmidt hammer rebound
values (r-values) for boulders on the Murtèl rock glacier surface
1998).
in Switzerland (from Haeberli et al., 2003). Weathering rind
thickness measurements from the neighbouring Gianda
Grischa rock glacier reveal that weathering rates vary as a result
5.6.4.2 Problems in using surface weathering of factors such as rock type and micro-climatic conditions. b)
features as indicators of relative age Iron oxide and c) clay mineral ratios in top soils on moraines
of different age in Nepal (from Zech et al., 2003). Active non-
Although degree of weathering on rock surfaces and crystalline iron oxides (Feo) dominate in young soils but are
boulders is a function of age, it may also be affected by other progressively reduced to the crystalline state (Fed) over time.
factors such as local climate, altitude, aspect and bedrock
type. Thus variations in local lithology may influence
Schmidt hammer readings, as also might wind polishing
and case hardening of weathering rind surfaces in high
342 DATING METHODS
arctic regions (Viles et al., 2011). In glaciated regions, the 5.6.5 Pedogenesis
incorporation of older rocks and boulders into younger
glacial landforms can also result in anomalous Schmidt 5.6.5.1 General principles
hammer data (Evans et al., 1999). Cation-ratio dating has
proved to be particularly problematical, with questions Key pedological (soil) properties can be used to estimate
being raised about the precise nature of the processes that ages of Quaternary deposits from various deposition
lead to cation-ratio variations, and also about the long-term environments, estimate long-term stability and instability
stability of ultra-thin varnishes on rock surfaces (Bierman of landscapes, and make inferences about past climatic
& Gillespie, 1994). The validity of radiocarbon dates on changes (Birkeland et al., 2003). This approach rests on the
organic residues from rock varnish has also been queried concept of a soil chronosequence, that is, a series of related
(Beck et al., 1998; Watchman, 2000). However, the recent soils developed when all factors of soil formation except
time (climate, organisms, parent material and topography)
application of varnish microlamination dating to rock
are held more or less constant (Huggett, 1998). Hence
surfaces in the arid southwest USA has generated results
contrasts between soil profiles, such as grain size variations,
that not only correlate with the SPECMAP oxygen isotope
physical and chemical properties, micromorphology and
record (section 5.5.3), but accord well with independently
depth of soil development, along with other parameters,
derived CRN ages, suggesting that VML dating can pro-
such as magnetic susceptibility or clay fraction mineralogy,
vide a reliable basis for determining relative age (Liu &
can be interpreted as a function of time and so provide a
Broecker, 2008).
basis for relative dating (Birkeland, 1999). Although
qualitative assessments of soil development have proved
5.6.4.3 Some applications of surface useful in the establishment of local glacial chronologies, a
weathering dating more quantitative methodology has been adopted in recent
years. Examples include soil development indices, which
Variations in rock weathering characteristics have been
reduce soil property data from a profile to a single value that
widely employed in the establishment of Late Quaternary
reflects the degree of pedogenic development. Inter-site
glacial chronologies, where moraines and other glacial
comparisons enable temporal and regional trends in soil
features have been placed in relative order of age on the
development to be identified and profiles can be ranked
basis of, inter alia, degree of weathering of contained
in terms of relative order of age (Goodman et al., 2001).
rocks and boulders (Shakesby et al., 2004). Rock surface
Soil development indices can then be employed in soil
weathering contrasts have also been used to delimit the
chronofunctions in which changes in soil properties are
upper level of glacial erosion on mountain summits (the modelled as functions of time (Dahms et al., 2012). These
trim line: section 2.3.1), from which the former vertical provide a quantitative framework for the construction of
extent of ice sheets can be reconstructed (Ballantyne et al., timescales of soil development (Figure 5.45b).
2008), while in arid regions, rock varnish laminations
(see above) have enabled ages to be assigned to a range
of landforms and deposits, including alluvial fans, colluvial 5.6.5.2 Problems in using pedogenesis as a
boulder deposits, lava flows and rock pavements (Liu, basis for dating
2003; Liu & Broecker, 2008). Surface exposure dating The major difficulty in using degree of pedogenesis as a
has also been used in archaeological contexts, in rela- basis for chronology lies in the requirement that all soil-
tion to the physical and/or chemical weathering of rock forming factors, apart from time, must be held constant,
surfaces. Initially, this approach appeared to generate for numerous studies have shown that factors such as
equivocal results (see e.g. Watchman, 2000), but more topography and vegetation cover are frequently as import-
recent investigations have shown a close correspondence ant as time in controlling soil genesis (Almond et al., 2007;
between surface exposure and radiometrically deter- Douglass & Mickelson, 2007). Rates of pedogenesis also
mined age estimates from archaeological artefacts (e.g. appear to be influenced by regional climatic parameters
Zerboni, 2008). Finally, an increasingly important role (Birkeland, 1999) and this imposes further constraints on
for surface exposure dating is in the selection of surface models of soil evolution where time is considered to be the
samples for cosmogenic radionuclide dating (Gordon & major independent variable. An additional problem, which
Dorn, 2005). particularly affects older (pre-Holocene) glacial substrates,
is that episodic erosion and redeposition can result in soils
that may be an order of magnitude younger than underlying
DATING METHODS 343
sediments (Rodbell, 1990). As with weathering charac- fossils, are being increasingly used in Quaternary research,
teristics of rock surfaces, therefore, pedogenic development the most important methods for establishing the age of
tends to be used in an auxiliary, as opposed to a primary, Quaternary events are likely to remain the radiometric and
role in the establishment of relative chronologies (Mark incremental techniques. Because they allow events to be
et al., 2004). Indeed, with the advent of surface exposure dated in years, their application has often been referred to
dating, U-series dating of soil carbonates and OSL dating, as ‘absolute dating’. However, this term has not been used
pedogenic differentiation as a dating approach is now less here as it implies a level of exactitude that is rarely
widely used than before. attainable. Nor have we used ‘geochronometric’, which has
frequently been employed to describe dating techniques
5.6.5.3 Some applications of relative dating from all of the categories described above, including
based on degree of pedogenesis methods which, at best, only provide relative age relation-
ships at a very broad scale. Again, a degree of precision is
Pedogenic contrasts as a basis for relative chronologies implied by the use of this term that is often not reflected
have perhaps been most widely employed in North in the results produced. But irrespective of the techniques
America. In many formerly glaciated areas of the western employed, the key question that needs to be addressed in
Cordillera, for example, systematic changes in morphology formulating a sampling strategy for dating Quaternary
and soil development reflect increasing age of parent events is: how can the highest level of geochronological
material. These include the sequence and thickness of resolution be achieved on samples from a single
genetic horizons; increase in clay content of the B-horizon; stratigraphic sequence? This involves assessment of both
depth of oxidation of the B- and C-horizons; and clay stratigraphic and temporal resolution.
mineral alteration (Figure 5.45b). A clear trend is often In Figure 5.46a, cores (or monoliths) have been
apparent in pedogenesis with older soils showing evidence
obtained from bodies of Quaternary sediment, each 1 m in
of, for example, deeper profiles, thicker horizons and
length. The left-hand core is subsampled in 10 cm blocks,
increased clay accumulation. Using this type of evidence,
while the core on the right is subsampled in 2 cm incre-
it has been possible to assign relative ages to, and to establish
ments. Hence, the right-hand core is sampled at a much
correlations between, moraines in, for example, the Rocky
higher stratigraphic resolution. The former may have been
Mountains, USA (Dahms, 2002), the Peruvian Andes
sampled for coleopteran analysis (which requires greater
(Goodman et al., 2001) and the Himalayas (Zech et al.,
quantities of material), while the latter may have been
2003). An alternative approach involves the use of soil
sampled for pollen or diatom analysis, for which much
catena chronosequences, that is, the evolution of soils at
smaller samples are needed. The level of sampling
different positions on slopes of moraines of different ages,
in order to provide a more secure basis for pedologically resolution is therefore dictated by the nature of the
derived timescales of glacier activity (e.g. Mahaney et al., analytical programme. But the temporal dimension also
2009). Soil development indices and pedo-chronosequences needs to be considered (Figure 5.46b). The left-hand core
have also been used to determine the relative ages of marine may have accumulated over an interval of 1.0 ka while
terraces (Scarciglia et al., 2006), sand dunes (Botha & the right-hand spans 10 ka. Assuming a constant rate of
Porat, 2007) and river terrace sequences (Tsai et al., 2007), sedimentation in both cases, each increment from the left-
while pedogenic and weathering parameters have been hand core would have taken 100 years to accumulate, and
employed in combination with other techniques (radio- those on the right 200 years. Hence, the sequence on the
carbon, 10Be, dendrochronology) to reconstruct glacial left would have been sampled at a higher temporal
histories and other aspects of landscape evolution during resolution, albeit with a lower stratigraphic resolution. If
the Late Pleistocene and early Holocene (Zech et al., 2003; we are seeking to date these two sequences (Figure 5.46c),
Favilli et al., 2009). and a specific volume of sediment (or abundance of fossil
material) is required to satisfy the analytical constraints
of the particular dating method (e.g. radiocarbon or OSL),
then the geochronological potential of that method
5.7 STRATIGRAPHIC AND
(in theory, at least) can be determined. For example, a
TEMPORAL RESOLUTION 5 cm or 20 cm thick sample of material from the left-
Although techniques that establish age equivalence on hand core could potentially provide age estimates at a
the basis of stratigraphic markers, and methods that temporal resolution of 50 years and 200 years, respectively.
determinate the relative order of antiquity of rocks or In the right-hand core, by contrast, a sample of no more
344 DATING METHODS
a) S t r a t i g r a p h i c resolution
than 2 cm in thickness would be required to provide a
geochronological resolution of 200 years. The value for
a 20 cm sample would be 2,000 years. From a research
design perspective, therefore, if the aim is to achieve decadal
resolution or better, then only the sequence on the left can
S u s c e p tSi buisl ict e p t i b i l i t
S u s c eSput isbci lei tp t i b i l i t

provide this, but only if a small sample size is used. Finally,
S u s cSeupstci be ipl itti b i l i t

Susceptibilit
as has been shown throughout this chapter, all dating

methods have uncertainties, and so the full theoretical
geochronological potential of a sedimentary sequence will
rarely be achieved. Thus, in Figure 5.46c, the error range
2 0 0 yr
for a 5 cm thick sediment sample may be much greater than
1 0 0 yr
for the 20 cm thick sample, perhaps because there is less
material (organic carbon for radiocarbon dating; quartz
1 0 c m thick 2 c m thick
grains for OSL dating) for analysis. Hence there will have
samples samples
to be a trade-off between sample size and analytical
resolKn:
resolKn: ârse/low fine/high precision. This is geochronological resolution (as opposed
to potential). But even if it proved possible to reduce the
b) G e o c h r o n o l o g i c a l p o t e n t i a l analytical error, this might generate spurious results,
depending on how long the sediment has taken to
accumulate. This is illustrated in Figure 5.46c where a
2 0 0 yr standard deviation of ±30 years for Date 1 implies an
200
unrealistic level of temporal precision on a sediment sample
i tt i b i l i t
yr
i bpi tl i tb i l i t
that took 200 years to accumulate, whereas better age

ilp
resolution is possible for Date 2 if analytical precision can

Sputsi bc e
S u sScuespct e
be improved.
Susce
Clearly, the situation depicted in Figure 5.46 is an over-

Susceptibilit
simplification of reality, as most sedimentary sequences are

1 0 0 yr 2 0 0 yr complicated by non-linear (irregular) sedimentation rates,
50 yr
depositional hiatuses, and perhaps recycled or deformed
components. Moreover, when confronted by a stratigraphic
G r e y blocks represent the thickness sequence, it is very difficult to make an initial assess-
of s a m p l e s e x t r a c t e d for d a t i n g
ment of the geochronological potential of that sequence.
c) G e e - c h r o n o l o g i c a l r e s o l u t i o n However, given the increasing imperative in Quaternary
science to conduct analyses at the highest possible levels of
temporal resolution, it is essential that consideration be
given to these matters in the formulation of sampling and
200 Mean spurious dating strategies, and that preliminary investigations
S.D. precision
20
S u s cSeupstci beipl itti b i l i t
involving, for example, the use of range-finder dates, should

be an essential first step in the establishment of a chronology
for Quaternary sedimentary records.
unfulfilled
1 0 0 vr
50 yr
Mean
5.D. geochronological 5.8 CONCLUSIONS
potential
age
In addition to ensuring that sites are sampled at the highest
G r e y blocks represent the t h i c k n e s s
of s a m p l e s e x t r a c t e d for d a t i n g
levels of geochronological resolution, the major challenge
now in Quaternary dating would appear to be how to
Figure 5.46 Stratigraphic and temporal resolution in Quatern- achieve a greater degree of reliability in age estimates based
ary sediment sequences. S.D. – standard deviation (2σ). For on methods that, in some cases, have somewhat uncertain
explanation see text. foundations. Progress in this respect depends not only
DATING METHODS 345
on achieving higher levels of both accuracy and precision NOTES

in individual dating techniques, but also on replication of
results from a given method, and on the implementation 1 The principle of superposition states that, in the absence of
of integrated dating strategies, involving the application of evidence for disturbance or reworking, the overlying sediments
in a sequence are younger than those lying beneath them. This
more than one technique to the dating of a fossil assem-
applies both to lithological units, such as tills, solifluction
blage, body of material, stratigraphic horizon or event. deposits, etc., and to biostratigraphic units, such as pollen or
Although the approaches reviewed in this chapter have gone molluscan assemblage zones. In this way, pollen analysis, for
a considerable way towards providing a coherent dating example, can be employed as a relative dating technique.
framework, particularly for the Late Quaternary, it is 2 The secular magnetic database analysed by Jackson et al. (2000)
important that new techniques continue to be developed. contains more than 83,000 individual measurements of declin-
One area that appears to offer considerable potential in this ation for more than 64,000 locations, all collected prior to AD
regard is biomolecular dating (section 5.5.4), in particular 1800.
molecular clock analysis based on timings of genetic 3 For example, protein residues have been found in Ordovician
mutations, and which could eventually provide a complete and Devonian shells, and collagen-like proteins have been
recovered from Cretaceous dinosaur bones.
chronology for human evolution. Not only do techniques
4 The operculum (pl. opercula) in a mollusc shell is the small
such as this offer new bases for determining Quaternary lid that is attached to the upper surface of the foot and serves
time, but they provide further scope for the calibration of as a ‘trapdoor’ to close the aperture of the shell when the soft
existing methods. Bearing these points in mind, we now parts of the animal are retracted.
turn to the wider question of stratigraphic subdivision and 5 Some analyses use fluoride, which is the negative ion of the
correlation of the Quaternary record. element fluorine, as a basis for dating of fossil bone.
6
CHAPTER SIX
Approaches
to Quaternary
stratigraphy
and correlation
6.1 INTRODUCTION floor, and in certain exceptional terrestrial situations, such

as deep lakes or regions with thick loess deposits, are long
In previous chapters, the various methods employed in continuous stratigraphic records preserved, although
palaeoenvironmental reconstruction have been exam- even in those contexts the dating of events may be difficult.
ined and the techniques used in the dating of Quaternary For the most part, therefore, the Quaternary stratigrapher
events have been discussed. Of particular interest to the is confronted by a highly fragmented and partial sedi-
Quaternary scientist, however, is the way in which envir- mentary record that can be dated securely only in certain
onments have changed through time, and how such circumstances and over limited time ranges. As a conse-
information can be gained from the stratigraphic record. quence, very careful analysis of the stratigraphic relationship
There are two aspects to the interpretation of that record, between sediment sequences is required before records
namely the ordering of the evidence at any one locality into can be ordered at one locality and correctly related to
a time sequence (temporal dimension) and the relationship those at another.
of the evidence at one locality to that at another (spatial In this chapter, the principles of, and the procedures
dimension). The temporal dimension involves principles of involved in, Quaternary stratigraphy and correlation are
stratigraphy, while the spatial dimension involves principles examined, and the means through which time-stratigraphic
of correlation. A proper understanding of the procedures correlation can be achieved are explored. Although deep-
involved in these two aspects of geological investigation is sea sediment records are considered, this discussion will
fundamental to a correct interpretation of Quaternary relate primarily to the terrestrial record, for it is terrestrial
palaeoenvironmental records. evidence that provides the greatest obstacles to the
If stratigraphic sequences were complete at all places on application of conventional stratigraphic and correlation
the earth’s surface, and if every important horizon could procedures.
be dated accurately, then elements of the record could be
arranged into stratigraphic order based on increments of
time. In this way a time-stratigraphic framework for the
history of environmental change could be established, 6.2 STRATIGRAPHIC SUBDIVISION
and correlation between even the most widely separated
localities would present few problems. In practice, however,
6.2.1 Principles of Quaternary stratigraphy
this is not achievable. The terrestrial record is frequently Stratigraphy is the study of the chronological order of
incomplete, and dating control inadequate, due either to rocks and sediments, and also of the sequence of events
the absence of suitable dating materials or to limitations of reflected within them (Rawson et al., 2002). The funda-
the techniques employed (Chapter 5). Only on the ocean mental building blocks are units of geological strata that
348 STRATIGRAPHY AND CORRELATION
can be identified on the basis of visible attributes, such as 3.11) are important examples of such instrumentally
colour, grain-size variations or structural elements (bed- derived stratigraphic records.
ding, deformation features, etc.). These units comprise There are a number of other important ways in which
the basic elements of lithostratigraphy, which involves the stratigraphic units can be subdivided and ordered, however.
ranking or ordering of local rock or sediment sequences Biostratigraphy involves the classification of sediment
according to observable variations in lithology. Increasingly, units according to observable variations in fossil content,
however, a range of instruments and sensors are being and enables a sediment sequence to be divided into bio-
employed in the analysis and classification of geological stratigraphic units or biozones, each characterized by a
units, including the various remote sensing methods distinctive fossil assemblage. In Quaternary stratigraphy,
described in section 2.2.2, as well as geophysical data- landforms can be classified according to their relative
logging procedures such as digital imaging or X-ray core order of age (morphostratigraphy), with each landform
scanning (Rothwell, 2006). These enable the properties or landform suite constituting a distinct morphostrati-
of rock or sediment units to be examined in much more graphic unit. Climatostratigraphy is concerned with the
sophisticated ways than is achievable by the human eye. A division of a stratigraphic sequence into geological-climatic
wide range of physical and chemical properties, reflecting units on the basis of inferred changes in climate, while
changes in the constituent materials within rocks or chronostratigraphy involves the classification of strati-
sediments, can be analysed. Examples include magneto- graphic units according to age, the interval of time
stratigraphy (variations in strength or direction of magnetic during which a geological unit has developed being
signals), seismic stratigraphy (variations in strength of referred to as a geochronological unit (section 6.2.3.7).
penetration or reflectance of directed signal beams) and Climatostratigraphy and chronostratigraphy are essen-
chemical stratigraphy (variations in abundance of chem- tially inferential methods of stratigraphic subdivision,
ical compounds or of their isotopic ratios). Some methods whereas both lithostratigraphy and biostratigraphy are
have particular applications in Quaternary research, such more direct, being based on empirical properties of the
as oxygen isotope stratigraphy (a form of chemical sediment record.
stratigraphy) that provides the foundation for subdividing Codes of practice have been produced to aid the
the Quaternary timescale (section 3.10.2), and electrical geologist in the task of stratigraphic subdivision and
conductivity measurements (a geophysical logging classification (e.g. Salvador, 1994; Rawson et al., 2002;
method) developed specifically for the analysis of ice-core North American Commission on Stratigraphic Nomen-
sequences. Where clearly defined ‘events’ are recorded in clature, 2005). These aim to provide a clear and unam-
the stratigraphic record, such as magnetic reversals, volcanic biguous terminological framework, as well as guiding
eruptions (indicated by tephra layers) or inflections in principles on stratigraphic procedures. All of these codes,
isotopic profiles marking abrupt climatic changes, these can however, have been developed for use in pre-Quaternary
form the basis for an event stratigraphy (Alloway et al., geology, and have limitations when applied to Quatern-
2007b). As the event horizons are essentially time-parallel ary stratigraphy. This is because Quaternary sediment
(isochronous), they offer a means of correlation at a range sequences can be examined in far greater detail and at a
of spatial scales (section 6.3.2.6). much higher level of temporal resolution than their pre-
Modern stratigraphical investigations normally involve Quaternary counterparts. The result is that the Quaternary
a range of approaches, and geological sequences are usually scientist often has to deal with sedimentary sequences of
subdivided after integration of all the available evidence. unusual geological complexity, and this gives rise to
In most situations, however, it is still the visible features problems of classification, interpretation and correlation
that are of primary importance in the formal subdivision that are not encountered in earlier parts of the geological
of the stratigraphic record: the parameters derived from record. For example, boundaries between lithological units
instrumentation, such as geophysical or chemical data, are placed at positions of lithological change (section
provide supplementary information only. In some import- 6.2.3.1), but in many Quaternary sequences these cut across
ant cases, however, visible lithological variations may be the limits of fossil ranges (reflected in biozones) and the
absent or imperceptible, while important stratigraphic boundaries of other types of stratigraphic unit (Figure
changes, which may be crucial in subsequent palaeo- 6.1). Moreover, they frequently cut across time-horizons
environmental reconstructions, can be detected only by (time-transgression), and this poses particular problems in
instrumental techniques. The isotopic and trace-gas signals the correlation of stratigraphic sequences at the scale of
in deep-ocean sediments and ice cores (sections 3.10 and resolution required in Quaternary research.
STRATIGRAPHY AND CORRELATION 349
contrasts in Quaternary environments as reflected in the

biostratigraphic record and, above all, the markedly time-
transgressive nature of terrestrial stratigraphic boundaries
(Figure 6.1). In many situations, therefore, a stratotype will
have little more than local application and, as a conse-
quence, the concept of the stratotype has been less widely
adopted by Quaternary workers than by geologists deal-
ing with older segments of the stratigraphic record. How-
ever, in the interests of clarity and precision, a strong
case can be made for the more widespread adoption of
Quaternary stratotypes than has been the practice in the
Figure 6.1 Time-transgression in lithostratigraphic and bio- past. Certainly, if effective communication between sci-
stratigraphic boundaries. A till unit is overlain by a peat layer entists is to be achieved with respect to the subdivision of
which has, in turn, been buried by marine sediments. The the Quaternary stratigraphic record, the establishment
marine deposits accumulated during a gradual marine trans-
gression (t1–t2) and therefore the lithostratigraphic boundary of regional stratotypes where the lithostratigraphic and
between peat and the overlying marine sediments (X–Y) will biostratigraphic units constitute clear reference standards
be time-transgressive. During the deposition of the marine is an integral component of that process (McMillan, 2005).
sediments, changes in fossil content have occurred, repre- An example is the isotopic record in the GRIP Green-
sented by biozones Ba, Bb and Bc. The lithostratigraphic
boundary X–Y therefore cuts across the biozone boundaries.
land ice core that has been proposed as a stratotype for
the Late Quaternary in the North Atlantic region (Walker
et al., 1999). At a larger scale, global stratotypes are required
by the international geological community for defining
6.2.2 Stratotypes
major geological boundaries and, as explained in section
According to conventional geological procedures, a locality 1.3, these are referred to as Global Boundary Stratotype
where a particular stratigraphic unit is clearly and fully Sections and Points (GSSPs). Hence, as we saw in Chapter
recorded, or where its lower boundary is securely defined, 1, the GSSP for the base of the Quaternary system/period
is termed a type site or stratotype. This can then be used and Pleistocene series/epoch is located at Monte San Nicola
as a reference standard against which other sequences, in Sicily (Figure 1.3) while the GSSP for the base of the
where the equivalent unit or its boundaries are only partially Holocene series/epoch is in the NGRIP Greenland ice core
or poorly represented, can be compared. The type section (Walker et al., 2009). The Subcommission on Quaternary
should be accessible and durable, so that it may be avail- Stratigraphy (Chapter 1, note 3) is now working towards
able for further study, and perhaps reassessment where the definition of global stratotypes for the Lower–Middle
necessary. In a review of the role of stratotypes, Walsh and Middle–Upper Pleistocene boundaries (Head et al.,
(2005) defines three types: exemplary stratotypes, which 2008b; Litt et al., 2008).
provide concrete examples of important, widely recognized
stratigraphic units; boundary stratotypes, which show the
clearest exemplification of the nature of the boundaries
6.2.3 Elements of Quaternary stratigraphy
between internationally recognized stratigraphic units; and
nominal stratotypes, which are internationally agreed
6.2.3.1 Lithostratigraphy
reference sites to which the formal name of the stratotype In theory, lithostratigraphic units should be recognized
applies. Walsh draws an analogy between these procedures and defined on the basis of sediment properties alone, such
for classifying rock or sediment sequences, and the code of as colour, particle shape and size, and grain-size vari-
practice used in the naming of new biological species and ations. In practice, however, Quaternary scientists have
for establishing their hierarchical importance. also tended to classify sedimentary units on the basis of
Ideally, all stratigraphic units should be defined by inferred mode of origin (till, aeolian sand, glaciofluvial
reference to a stratotype. However, this is often difficult gravel, etc.). The problem with such genetically based
in Quaternary stratigraphy because of the highly frag- classifications is that if the mode of origin is incorrectly
mented nature of the terrestrial stratigraphic record, inferred, then the subsequent stratigraphic reconstruction
the considerable spatial variation in type and thickness may be undermined. Hence many Quaternary geologists
of Quaternary sediments, the limited lateral continuity of now prefer the use of sedimentological terms that are
many Quaternary deposits, the spatial and temporal free from such genetic connotations. One such term is
diamicton, which simply refers to any sedimentary unit finer subdivisions of the stratigraphy may be possible
with a heterogeneous mix of particle sizes from clay through where, for instance, very thin layers of perhaps 1 cm or less
to boulders. The mode of genesis of a diamicton is inferred (laminae) occur throughout a bed, each reflecting a very
at a later stage in the analysis on the basis of such diagnostic short-lived sedimentary episode. Once such individual
criteria as fabric data, clast size and shape, and micro- units have been identified, the second stage is to establish
morphological properties (section 3.3.5). When confronted mode of genesis; in other words, description and classifi-
by a complex stratigraphic sequence (Figure 6.2), the cation should always precede interpretation of the
approach should be first to identify beds, which are indi- stratigraphic record.
vidual sediment units or bodies that are considered to Where several beds can be shown to be related and have
have originated during the same or similar depositional accumulated through similar depositional processes, they
event. Careful examination of a sand unit, for example, may comprise larger lithostratigraphic units known as members.
reveal a series of large- or small-scale cross-bedded layers, Hence the stratified sediments in Figure 6.2 are grouped
each reflecting changes in the environment of deposition. together to form a member that can be differentiated
Similarly, alternations between sand and gravel beds, layers lithologically from the overlying and underlying unstratified
of silt- or clay-dominated beds, and diamicton units (which diamicton (till) members. Members that accumulated
frequently form individual beds), may indicate important sequentially during a major depositional event (e.g. a glacial
variations in mode of deposition. In some instances, even episode or a marine transgression) form lithostratigraphic
Figure 6.2 Lithostratigraphic subdivision of a glacigenic sequence.

units of higher rank, termed formations. Two or more geometry, bedding, external contacts, etc.) can be estab-
contiguous formations that possess common lithological lished for each facies, and these data can then be used to
properties (particle size characteristics, clast lithologies, define lithofacies in rock or sediment sequences, that
etc.) may be aggregated into groups. Distinctive litho- is, ancient equivalents of modern facies types. Facies may
stratigraphic units that appear regularly in sediment records, be subdivided into subfacies or grouped together into
and provide a basis for correlation between sequences (such facies associations, while at the regional scale they can be
as tephra layers), are referred to as markers. Units of considered three-dimensionally in terms of facies archi-
formation or group rank, displaying common properties or tecture. In Quaternary stratigraphy, facies analysis has
associations of members and beds, may also constitute been most widely employed in the study of glacigenic
markers. In such cases, these higher-ranking lithostrati- (including glaciofluvial) deposits (Benn & Evans, 2010),
graphic units are given formal status and may be accorded although this approach has also been employed in the
proper names. In southeast England, for example, extensive examination of other sequences, including fluvial and
spreads of glaciofluvial gravels dating from the Anglian deltaic sediments (Amorosi et al., 2003), and lacustrine
Glacial Stage have been termed the ‘Barham Formation’. records (van Leeuwen et al., 2000). As in conventional
A more extensive suite of river sediments that formed over lithostratigraphic procedure, however, it is important to
1.2 Ma prior to the Anglian Glaciation, the ‘Kesgrave Sands begin with a descriptive facies analysis, using the sort
and Gravels’ (Figure 2.51c), accumulated over many warm of criteria shown in Table 6.1. Non-genetic terms such as
and cold stages. This lithostratigraphic unit, which contains ‘diamict’ should be applied, and a shorthand notation
two major formations (Colchester and Sudbury) and which (Gm for massive gravels, Sh for horizontally laminated
can be traced over large areas of the River Thames basin sands, etc.) should be used to describe each lithofacies.
and East Anglia, is referred to as the ‘Kesgrave Group’ (Rose Again, interpretation of the sediment record in terms of
et al., 1999). Further examples of the ways in which these mode of genesis then constitutes the second stage of the
stratigraphic procedures have been applied in Quaternary analysis.
research can be found in Bowen (1999), Rose et al. (2001) In some depositional contexts, recurrent or cyclic
and Lee et al. (2004). processes lead to the repetitive formation of similar
An alternative approach to stratigraphic subdivision sedimentary ‘packages’ (discrete stacks of sediment units),
stems from detailed studies that have been undertaken on each separated by distinct unconformities or hiatuses. The
contemporary depositional environments. These have order of superposition and overall architecture of these
shown that different sedimentary contexts (estuarine, accretionary ‘prisms’ can be established by lateral tracing
glacial, fluvial, aeolian, etc.) give rise to distinctive associ- of their bounding unconformities (sequence stratigraphy),
ations of sedimentary units or facies (section 3.3.3.2). an approach that is now widely applied in geology (Miall,
Key diagnostic properties (colour, grain-size variations, 2000). This has, perhaps, been most widely employed in the
Table 6.1 Principal descriptive criteria used in defining lithofacies in glacigenic sequences (after Hambrey, 1994).
Lithology Bedding Bedding Sedimentary Boundary
characteristics geometry structures relations
Diamict(on/ite) Massive Sheet Grading: normal, Sharp
reverse, coarse-tail
Gravel Weakly stratified Discontinuous Cross-bedding: Gradational
tabular, trough
Sand(stone) Well stratified Lensoid Dropstones Disconformable
Mud(stone) Laminated Draped Clast supported Unconformable
Rhythmic lamination Prograding Matrix supported
Wispy stratification Clast concentration:
Inclined stratification layers, pockets
Ripples
Scours
Load structures
Mottling (poss. bioturbation)
HST
JHST
:lst
JHST
Flooding surfaces Transgressive Surface
M a x i m u m flooding surface Sequence Boundary
F l u v i a l o r e s t u a r i n e s a n d s t o n e s in i n c i s e d v a l l e y s Shallow-marine sandstones
Coastal-plain sandstones and mudstones Shelf mudstones
Figure 6.3 Sequence of sedimentary units deposited near continental margins during cycles of sea-level fall and rise. HST,
Highstand stage; LST, Lowstand Stage. Other terms are explained in the text.
analysis of marine deposits that have accumulated close to sequence stratigraphy, uses discontinuities and surfaces
continental margins. Repeated rises and falls in sea level lead to subdivide the sedimentary record. Hence an allostrati-
to landward and seaward migration of the zones in which graphic unit is a body of sediment or rock that is defined
shallow- and deep-water sediments accumulate. Each cycle and identified solely on the basis of its bounding dis-
has two characteristic stages: a lowstand and highstand continuities. As such, allostratigraphic units differentiate
(Figure 6.3) and two transitional phases: transgressive and between individual bodies of superposed deposits, for
falling. During a fall in sea level, marine sediments may example between a sequence of lithologically similar alluvial
become exposed and hence subject to erosion, but a and lacustrine deposits separated by a palaeosol, or between
subsequent transgressive phase will lead to the burial of the onlapping lobes of geliflucted material from different
remaining deposits by a new accretionary layer, the base of cold-climate episodes (Rawson et al., 2002). Whether this
which is marked by an obvious unconformity. Although form of stratigraphic classification represents a significant
sequence stratigraphy was initially developed for the advance over more traditional methods of classifying
analysis of pre-Quaternary rock formations, particularly in Quaternary deposits, however, is still being debated
the search for petroleum reserves, it has also been used in (Räsenen et al., 2009; Johnson et al., 2009).
the mapping and interpretation of unconsolidated Irrespective of the approach adopted, the level of litho-
Quaternary marine facies (Ridente & Trincardi, 2002; stratigraphic subdivision of a Quaternary depositional
Browne & Naish, 2003). sequence depends on the local complexity of the record, and
Lithologically similar accretionary sediment stacks can on what are frequently subjective decisions on the nature
also be generated by other depositional processes that or rank of a particular stratigraphic unit. However, inter-
operate in a cyclic manner, for example in fluvial, glacial pretation is often complicated by the fact that apparently
or aeolian systems. Some stratigraphic codes now include similar sediments and sedimentary sequences can result
the wider generic term ‘allostratigraphy’ to refer to the from different geological processes. Fluvial, glaciofluvial
classification and sequencing of the depositional stacks and aeolian sediments, for example, can appear lithologic-
(alloformations) that these systems generate (North Amer- ally similar despite the fact that they have accumulated
ican Commission on Stratigraphic Nomenclature, 2005; in markedly different depositional environments. This
Hughes, 2010). The difference between lithostratigraphy problem, referred to as equifinality, can also arise in the
and allostratigraphy is that the former often ignores signi- interpretation of Quaternary soils and landforms (Kemp,
ficant breaks in the sediment sequence (e.g. unconform- 1998; Evans, 2009). In addition, many Quaternary sedimen-
ities or flooding surfaces), whereas allostratigraphy, as in tary sequences (such as fluvial deposits) show considerable
lateral and vertical variation, and these local facies changes

Lithostratigrapic
frequently pose problems of interpretation and classifica-
record Fossil ranges
tion when using traditional lithostratigraphic criteria
(Clague, 2000; Winsemann et al., 2007). Deformation of
sediments is a further complicating factor, particularly in E Assemblage
the analysis of glacigenic deposits (Lee & Phillips, 2008). biozone
Even where the origin of the sediments can be unequivo-
cally established, the subdivision of very complex sequences
b
into lithostratigraphic units is not always straightforward. Partial
In Figure 6.2, for example, the beds of stratified sediments range
biozone
can be grouped together to form a member that can be
differentiated lithologically from the unstratified dia-
mictons above and below. Analysis shows the sediments to
have been derived from episodes of glacial and glacio- Totai
fluvial activity during a single glacial event, and hence the E range
whole sequence can be classified as a formation (scheme 1: biozone
left). Alternatively, the upper and lower diamictons and the
intervening stratified unit might be regarded as representing
Acme
discrete depositional events, and under this interpretation, biozone
each could be classified as units of formation rank (scheme
2: right). Consistency of interpretation between individual Concurrent
sb range biozone
workers may, therefore, be difficult to achieve. Nevertheless,
Consecutive
insofar as it emphasizes the need for careful analysis and range biozone
b
interpretation, this is still the best approach to the sub-
division of the Quaternary rock-stratigraphic record.
First a p p e a r a n c e R a n g e continues
Abundance
Final appearance b Speciation event
6.2.3.2 Biostratigraphy
Biostratigraphic classification organizes rock strata into Figure 6.4 Various types of biozone used in the subdivision
units based on the variety and abundance of fossils. Bio- and correlation of strata. The consecutive and concurrent
stratigraphic units are usually termed biozones, and the ranges are defined on the basis of the appearance and
following types are commonly employed in stratigraphic disappearance of taxa, normally in the context of evolutionary
lineages. Others are defined in the text.
classification (Figure 6.4):
1. Total range biozone: a group of strata containing the 6. Assemblage biozone: biozones that are defined on the
full stratigraphic and geographical range of a particular basis of a characteristic mix and relative abundance of
fossil or group of fossils. fossil types.
2. Acme biozone: a group of strata based on the acme or
maximum development of a particular taxon. In pre-Quaternary geology, the biostratigraphic record
3. Partial range biozone: the stratigraphic interval is subdivided largely on the basis of evolutionary changes
described by that part of the range of a particular in organisms, and thus acme biozones and range bio-
taxon that lies above that of a second taxon and below zones reflect those episodes of geological time where a
that of a third. Hence in Figure 6.4, PRB-b is that part species appeared, thrived for a while, and then died away.
of the range of taxon b which lies above the range of Acme and range biozones constitute valuable strati-
taxon a and below that of taxon c. graphic markers, and often form the basis for correlation.
4. Concurrent range biozone: defined on the basis of the Although there is some evidence of evolutionary develop-
overlapping ranges of several taxa. ment in lineages of rodents (van Koenigswald & van
5. Consecutive range biozone: where speciation changes Kolfschoten, 1996) and beetles (Elias, 1994), and perhaps
can be clearly established (phylogenic lineage), also in ostracods (Griffiths, 2001) and fish (Craw et al.,
biozones can be defined on the basis of the consecutive 2007), evolutionary changes are generally less common over
ranges of fossils in an evolving lineage. much of the Quaternary biological record, and thus acme
and range biozones are not as widely used as in pre- biozones as a result of the reworking and selective destruc-
Quaternary geology. Quaternary biostratigraphy is, there- tion of fossils, and because of problems of identification
fore, based almost entirely on assemblage biozones, and and ecological interpretation of fossil assemblages (see
these largely reflect the ecological response of organisms to Chapter 4).
environmental change rather than evolutionary develop-
ments in flora and fauna. The biostratigraphic approach
6.2.3.3 Morphostratigraphy
to Quaternary subdivision has perhaps been most widely
applied in the British Isles and western Europe where a Morphostratigraphy is rarely discussed in stratigraphic
range of biological indicators has been used to charac- codes, yet is an essential stratigraphic method in Quaternary
terize warm- and cold-stage sequences and to differentiate science. A morphostratigraphic unit has been defined
between them (Schreve & Thomas, 2001). Unlike evolu- as ‘a body of rock that is identified primarily from the
tionary trends, however, ecological changes are both surface form it displays’ (Willman and Frye, 1970, p. 43),
reversible and repeatable. Hence Quaternary assemblage and was intended to be used by geological surveyors unable
biozones of different ages may contain essentially the same to unravel the complexities of glacial stratigraphy on the
mix of fossils (Thomas, 2001), in which case they are basis of lithology alone. A range of geomorphological
potentially flawed as stratigraphic tools. This situation features can be recognized that possess distinctive forms and
differs from that in pre-Quaternary strata where evolution- which are diagnostic of specific geological processes. These
ary changes over much longer time periods have resulted include glacial and glaciofluvial landforms, aeolian land-
in assemblage biozones that can be unique. forms and coastal features, such as beaches. A morpho-
Because of the time-transgressive nature of climatic stratigraphic unit may therefore be defined as that part
and environmental change, the boundaries of biozones of the lithostratigraphic record represented by a particular
(pollen assemblage zones, molluscan assemblage zones, geomorphological feature. Often a series of landforms
diatom assemblage zones, etc.) cut across time horizons reflects a temporal succession of phases of formation,
and frequently transgress the boundaries of other strati- such as river-terrace sequences, a series of raised shore-
graphic units. In parts of the deep oceans, however, where lines or a suite of glacial landforms (Schellman & Radtke,
sedimentation rates are comparatively slow, assemblage 2004; Lukas, 2006). As a distinct category of stratigraphy,
biozone boundaries may appear to be time-parallel. morphostratigraphic units are integral components of the
In reality, however, all assemblage biozone boundaries Quaternary stratigraphic record.
examined at the resolution required in Quaternary science The role of morphostratigraphy and its relation to the
are time-transgressive, since it takes time for organisms lithostratigraphic record is shown schematically in Figure
to migrate and to adapt to changes in environmental 6.5. Here the landforms have been identified as glacial
conditions. Whether the degree of time-transgression can features (moraines) on the basis of morphology and
be ascertained depends on the nature of the changes and, geomorphic context. Between the moraine ridges are lake
as in all biostratigraphical investigations, on the temporal basins (B1, B2) that contain limnic sediments and peats.
resolution of the stratigraphic record. Difficulties may In order to provide a full account of the stratigraphy of
also be encountered in the establishment of Quaternary this area, the morphological evidence must be integrated
M1
M2
M3
B1
MA
B2
Figure 6.5 Schematic diagram showing a series of moraine ridges (M1 oldest to M4 youngest) between which sediments have
accumulated in lake basins (B1 and B2) following deglaciation. For further explanation see text.
with the lithostratigraphic and biostratigraphic record. will develop. Many soils are polygenetic, such as the
One interpretation is that the moraines formed sequentially Valley Farm Soil (Figure 3.24b) of East Anglia and south-
during a period of glacier retreat, with M1 being the oldest east England (Kemp et al., 1993). This particular soil is
and M4 the youngest. If this was so, the base of the lake rubified (displays a characteristic reddening) and mottled
sediments in B1 would post-date M1, but pre-date the with a high translocated clay content. It is a complex soil-
formation of M2. Similarly, the base of the limnic deposits stratigraphic unit with evidence for several episodes of soil
in B2 would post-date both M2 and M3, and the basal development spanning a number of warm and cold
sediments in B1. The complete stratigraphic record would intervals over a period (at its fullest development) of more
therefore consist of morphostratigraphic unit (m.u.) M1 > than one million years. In the North American Midwest,
(= older than) B1 base > m.u. M2 > m.u. M3 > B2 base > the Sangamon Geosol, which can be differentiated from
m.u. M4. Alternatively, all of the ridges might have formed the modern soil by virtue of a greater mineral alteration, a
in a very short period of time and hence the accumulation thicker solum, a redder colour under oxidized conditions
of the basal lake sediments in both B1 and B2 post-dates and deeper carbonate leaching, is also polygenetic and
the entire morphostratigraphic sequence. Only the appli- appears to have formed over a period of around 100 ka
cation of other methods (e.g. biostratigraphic analysis, between MIS 5 and 3, during which time the climate shifted
radiometric dating) can resolve such an issue. from cold to warm to cold again (Hall & Anderson,
2000; Grimley et al., 2003). The stratigraphic importance
of such soil-stratigraphic units is that they provide clear
6.2.3.4 Soil stratigraphy
evidence for a major hiatus in the lithostratigraphic record.
Palaeosols (section 3.5) occur in many Quaternary sedi- In addition, because the soils are very distinctive, they
ment sequences. Well-developed palaeosols that evolved constitute key stratigraphic markers, and may be used as a
during a specific soil-forming interval and which possess basis for correlation between individual sequences (see
sufficiently distinctive characteristics to enable them to below).
be traced over a wide area can be considered as soil- Where sedimentation is episodic, pedogenesis will
stratigraphic units. A soil-stratigraphic unit was initially occur during intervening periods of landscape stability
defined by the American Commission on Stratigraphic and hence, over time, sequences of soils may form (Kemp,
Nomenclature (1961, p. 654) as ‘a soil with physical fea- 2001). This is most clearly demonstrated in the loess
tures and stratigraphic relations that permit its consistent regions of the world, where numerous soil-stratigraphic
recognition and mapping as a stratigraphic unit’. A better units are found interbedded with suites of aeolian sediment,
term than ‘physical’ might be ‘pedological’, since soils for example in South America (Kemp et al., 2006), North
are now identified and categorized on the basis of a range America (Jacobs & Mason, 2007) and parts of eastern
of properties, including chemical, magnetic and micro- Europe (Marković et al., 2007). The best known and most
morphological characteristics. The North American Com- widely studied, however, are the long loess–palaeosol
mission on Stratigraphic Nomenclature (1983) subse- sequences that have developed on the Loess Plateau in
quently introduced the term pedostratigraphic unit to China (section 3.6.2). In the 160 m Baoji loess section in
describe ‘a buried, three-dimensional body of rock that the southern part of the Plateau, for example, up to thirty-
consists of one or more differentiated pedologic horizons’ seven separate palaeosols have been identified, the
(p. 864), and has recommended that the term ‘geosol’ lowermost of which has been dated to around 2.5 Ma
replace the term ‘soil’ in stratigraphic usage. Here, however, (Rutter & Ding, 1993). The palaeosols have formed through
we retain the terms ‘palaeosol’ and ‘soil-stratigraphic unit’ a combination of carbonate eluviation and illuviation, clay
to refer, respectively, to fossil soils and to the stratigraphic translocation, pseudogleization and rubification. Most
units that they represent, since these terms are still the most display Bt horizons that have developed in forest and/or
widely employed in the Quaternary literature. steppe-forest environments. This sequence, like those
Soil-forming processes will begin to operate in most elsewhere in China, reflects a long history of loess deposi-
areas immediately after surfaces become exposed to the tion with intermittent episodes of stability during which
atmosphere. The subsequent degree of pedogenesis will vegetation became established and soils formed.
depend, however, on a range of site and climatic factors,
as well as on regional history, which includes length of
6.2.3.5 Oxygen isotope stratigraphy
exposure of the land surface. In many regions, the typical
soil features of horizonation may not be present, while in It has already been shown that, in the deep oceans of the
other situations, distinct and clearly demarcated horizons world, long sequences of relatively undisturbed sediments
are preserved that may extend back to the beginning of the the isotopic stages are designated by the prefixes K-, KM-,
Quaternary and, indeed, into the Tertiary. The microfauna M- and MG-, respectively. The deep-sea isotopic signal is
and microflora within these sediments contain a record of geographically consistent, and can be replicated in cores
changing oxygen isotope ratios that not only provides from different oceanic areas, thus emphasizing the fact that
evidence for former glacial and interglacial oscillations the oxygen isotope signal provides a proxy climatic record
(section 3.10), but which also forms a basis for stratigraphic of global significance (see section 6.3.2.7).
subdivision and long-distance correlation. It must be In the main, the isotopic stage numbers equate with
appreciated, however, that marine oxygen isotope profiles conventional glacial and interglacial episodes inferred from
are not appropriate for high-resolution stratigraphic sub- the terrestrial Quaternary stratigraphic record, although
division, as the slow mixing of the deep oceans means MIS 3 is anomalous, for while it is recognized as a
that asynchronous changes are impossible to detect, and ‘warm’ stage, it is regarded only as a period of interstadial
thus at the centennial scale, the underlying concept of the status. It should also be noted that the later ‘interglacial’
oxygen isotope stratigraphic scheme becomes meaning- stages have been subdivided into separate warmer and
less (Shackleton, 2006). In other, better resolved sequences, colder episodes. MIS 5, for example, contains five substages,
however, such as ice cores, speleothems and tree-ring with 5a, 5c and 5e interpreted as warmer phases while 5b
series, oxygen isotope profiles can provide a basis for sub- and 5d reflect colder intervals. The last (Eemian, Ipswich-
division at the centennial and, in some instances, at the ian, Sangamonian) interglacial (as defined on land) is
decadal scale. represented in the isotope record by substage 5e. An alterna-
As noted in sections 1.6 and 3.10.2, oxygen isotope tive approach to substage notation is to employ a decimal
profiles from deep-ocean cores can be divided into marine system in which negative (warm) and positive (cold)
isotope stages (MIS), and these stages and substages are excursions are given odd and even numbers, respectively
fundamental units in Quaternary stratigraphy. Recall that (Figure 6.6b). In this scheme, first applied in the SPECMAP
those parts of the marine isotope curve interpreted as timescale (section 5.5.3), the warmer substages in MIS 5
representing warm stages (lighter δ18O values) are assigned that were previously referred to as 5a, 5c and 5e, are desig-
odd numbers, and the cold stages (heavier δ18O values) nated 5.1, 5.3 and 5.5 (last interglacial); the colder substages
even numbers. MIS 1 represents the Holocene period, and (5b and 5d) are numbered 5.2 and 5.4. The three warmer
higher numbers indicate successively older cold and intervals of MIS 7 are designated 7.1, 7.3 and 7.5, with
warm stages. The allocation of stage numbers is therefore the intervening colder phases denoted as 7.2 and 7.4. In the
a ‘count from the top’ division of a sinusoidal curve, case of MIS 7, however, there is some uncertainty over
the underlying assumptions being (1) that inflections in the which of the warmer intervals represents the penultimate
curve reflect changes in global ice volume and deep-ocean interglacial (sensu stricto), for the warmest climate signal
temperature (Shackleton, 2000); and (2) that no major registers in MIS 7.5 in Antarctic ice cores (Masson-
interruption in sedimentation has occurred. Twenty MISs Delmotte et al., 2010), in MIS 7.3 in pollen and benthic
can be recognized in the ocean sediment record of the last foraminiferal data (Desprat et al., 2006) and in MIS 7.1
800 ka or so (Figure 6.6a), which implies that ten inter- in some cave speleothem stable isotope records (Spötl
glacials (or near interglacials) and ten glacial episodes have et al., 2008). The boundaries between the major isotopic
occurred during that time interval. The record of isotopic stages are indicated by integers: hence the boundary
stages formally designated in Quaternary deep-ocean between MIS 5 and MIS 6 is designated as 6.0, while that
cores extends back to MIS 103 at c. 2.6 Ma (Figures 1.5 between MIS 7 and MIS 8 is 8.0. A similar numerical
and 1.6). These stages can be dated by reference to the notation has subsequently been applied to older parts
palaeomagnetic timescale (Figure 5.34) with the Brunhes– of the marine oxygen isotope record. The rationale
Matuyama boundary located in MIS 19, the reversal underlying this approach is that the exact stratigraphic level
marking the onset of the Olduvai chron at the base of of a peak or trough in the isotope curve can be defined
MIS 63, and the Gauss–Matuyama boundary in MIS 104. unambiguously in one curve and correlated with the same
This MIS subdivision has now been continued back to level in another.
5.3 Ma (Lisiecki & Raymo, 2005), but in the Pliocene As noted above (section 3.10.3) a distinctive feature of
record, the isotopic stage prefixes are designated by the most deep-ocean oxygen isotope profiles is the ‘saw-tooth’
geomagnetic subchron in which they are found. Hence appearance of the curves (Figures 1.5 and 6.6), with the
in the early Gauss polarity epoch, the warm stages are most rapid isotopic changes occurring at the end of glacial
numbered G-1 to G-21, while in the earlier part of the episodes. These seemingly rapid changes from inferred
Gauss, through the Kaena and Mammoth polarity events, glacial to interglacial conditions have been referred to as
a) 6 1 B
0 O ) b) 6 0 (%»)
1 B
0 -1 -2 -3 2 1 100 0 - 1 - 2 - 3
0 0
2 7
3.1
3.3
4 2
5 1
100 5 100 5.3
5.4
6.2 5.5
6.3
6.5
5.5
200 200 7.2 7.1
7 7 4 7 3
7.5
8.2
84 a 3
8.5
300 300 8 6
9 9.1
9.2
9 3
10.2 10.3
10.4 11.1
11 11 2 2 11.23
400 400 11.24
Susceptibilit
Susceptibilit
1
12 2
12.33
D 13.11
13.12
500 500 13.13
13.2
13.3
14 2
152 > 15.1

15 BOO 15.3
600 15.4
15.5
\\6.S 16.3
17 17.1
17.3
700 700 17.4
18.2. 17.6
18.3
18.4
19 19.1
19.2
19.3
300 20.2
800
21.1
21.2
21 21.3
21 5
22.2
800 900
Figure 6.6 a) Oxygen isotope stratigraphy in a 53 m core (core MD900963) obtained from a water depth of 2,446 m in the Indian
Ocean to the east of the Maldives Platform. The record spans c. 900 ka and extends back to MIS 22; prominent interglacial stages
are numbered. b) A stacked record of Core MD9009643 and ODP core 677 from the eastern equatorial Pacific, which has been
tuned to a model of changes in global ice volume. The isotopic stages are designated using the decimal MIS notation, as explained
in the text (based on Bassinot et al., 1994).
terminations and constitute major events (section 6.3.2.7) has been subdivided into Termination Ia (dated mid-point
in the isotope stratigraphy (Cheng et al., 2009). As with the 15 ka BP) and Termination Ib (dated mid-point 10.5 ka BP),
isotopic stages, the terminations have been numbered from and these constitute key marker horizons in the oceanic
the top down. Hence Termination I refers to the sudden isotopic record at the end of the last cold stage. These
relative decrease in 18O content of the world’s oceans at the terminations are also evident in proxy records from
end of the last glacial stage (MIS 2). However, as MIS 3 is terrestrial contexts (Figure 6.7), and therefore provide a
not considered to be a warm interval of full interglacial rank, basis for land–ocean correlations (see below). We return
Termination II marks the transition from MIS 6 to MIS 5, to the importance of terminations in the Quaternary
and not from MIS 3 to MIS 2. The most recent termination record in Chapter 7.
a)
10
Susceptibilit
-6 300
b)
Susceptibilit
260
2 T1 T2 T3 T4
c)
220
Susceptibilit
Susceptibilit
3
180
5
0 50 100 150 200 250 300 350
A g e (ka)
Figure 6.7 The last four terminations in different proxy records. a) Composite δ18O profile from speleothem carbonate from four
caves (Hulu, Dongge, Linzhu, Sanbao) in China. b) Vostok CO2 record from Antarctica. c) Benthic δ18O values from marine core
ODP 980 from the North Atlantic (after Cheng et al., 2009).
ably, although in practice this is often done. In a formerly

6.2.3.6 Climatostratigraphy
glaciated region, for example, the stratigraphic record of the
A universally employed basis for the subdivision of the presence of ice may comprise only sediments that were
Quaternary is climatic change, for the characteristics of the deposited during deglaciation (Figure 6.8). Hence the time
stratigraphic record that are distinctive, and therefore used interval reflected in the stratigraphic record by a suite of
in the identification of stratigraphic units, frequently reflect glacigenic sediments (which comprise the geologic-climatic
processes driven by former climatic conditions. This was unit) represents only a small part of the glacial or cold stage.
recognized half a century ago by the American Commission At other sites, a more protracted period of glacial
on Stratigraphic Nomenclature (1961, 1970) who proposed sedimentation may be preserved, but again this may only
a stratigraphical subdivision termed a geologic-climatic have occurred during a limited part of the glacial stage.
unit, which was ‘an inferred widespread climatic episode Hence the duration of the latter will have to be inferred
defined from a subdivision of Quaternary rocks’ (1970, from other lines of evidence, such as the biostratigraphic
p. 31). In areas affected by Quaternary glaciation, glacials record in lake sediment sequences. Here, the suite of
and interglacials constitute the principal geologic-climatic minerogenic sediments containing that record (the
units, while stadials and interstadials form units of lesser geologic-climatic unit) may span the entire cold stage.
rank. In areas not affected by glacier ice, it was anticipated Reconciling diachronous climatic signals, which may have
that other geologic-climatic units, such as pluvials and contrasting expressions in different geologic units, is
interpluvials, would form units of equivalent status. therefore a major difficulty in climatostratigraphy.
Geologic-climatic units are undoubtedly useful con- The fundamental problem, of course, is that it is not
cepts and, insofar as the Quaternary sequence in mid- and climate that is directly recorded in the stratigraphic record,
high latitudes is usually subdivided into glacials and but manifestations of climate, namely the results of climatic
interglacials (section 1.6), they form the basis for strati- influences on, for example, biota, soils, sediments and
graphic subdivision at the regional and continental scales. glaciers. Climatic reconstructions are, therefore, two steps
However, because climatic change is time-transgressive, removed from the observable data, and at each stage in the
the boundaries of geologic-climatic units are diachronous. analysis, interpretation is required. If, for example, pollen
Strictly speaking, therefore, it is not appropriate to use assemblage zones are being used as the basis for geologic-
geologic-climatic terms (glacial, interglacial, etc.) and climatic units, the first step is to infer vegetational
chronostratigraphic terms (stage, substage) interchange- communities and patterns of vegetational change from
Site 1 Site 2 Site 3

M a s s flow dlamicton Till Till
L o d g e m e n t / d e f o r m a t i o n till Organic sediments Solifluction deposits

Till Organic sediments
Glaciofluvial s e d i m e n t s Till Interglacial
e
1
e
e
2
3
Stadial
i tp t i b i l i t
Stui bs icl e
Interstadial B'
Suscep
Glacial
e interval
4
Stadia I
Interstadial A '
e
5
D i s t a n c e from ice-dispersal centre Interglacial
Till M a s s flow diamicton

Palaeosol Gravels
Till
Site 4
M a s s flow diamicton
Sand
L o d g e m e n t / d e f o r m a t i o n till
SiteS
Figure 6.8 A time–distance diagram showing the onset (o) and end (e) of glaciation at sites at increasing distance from the ice-
dispersal centre. Time is shown schematically on the y axis and distance on the x axis. Possible glacigenic sequences and relative
sediment thickness at each site are also indicated. For further explanation see text.
the pollen record, and the second is to use these recon- graphic unit that formed the basis for their definition.
structions to infer climatic changes. Inferences are therefore This may be relatively straightforward when dealing with
being made which are themselves based on inferences; pre-Quaternary successions where, because of the long
errors can enter into both stages of the analysis, but those time intervals and low temporal resolutions involved,
resulting from the first will be compounded in the second. climatic change appears in the geological record to have
These complications will arise irrespective of the form of been virtually instantaneous. In the Quaternary, however,
lithostratigraphic or biostratigraphic evidence that is being where temporal resolution of stratigraphic sequences is
employed. usually so much greater, geological boundaries based on
The American Stratigraphic Commissions (1961, 1970) evidence of climatic change are much more difficult to
originally intended the boundaries of geologic-climatic locate, and correlation based on these boundaries is more
units to follow the boundaries of the rock or biostrati- problematical. Consider, for example, Figure 6.9 which
Susceptibilit
Interglacial Interglacial
Average present day

annual temperature y y
d e f i n e d for a n i n d e x
locality
t t
Susceptibilit
x x
Glacial Glacial
Figure 6.9 Different ways of defining the onset of an interglacial. The curve is schematic and represents temperature oscillations
between interglacial and glacial cycles. The three possible points of onset for an interglacial are explained in the text.
shows temperature change through a glacial–interglacial climatic conditions may be placed between assemblage
cycle and represents, in a schematic way, one of the climatic zones 1 and 2; the pollen data might indicate a climatic
cycles reflected in the oxygen isotope signal (Figure 6.6). change between pollen zones 4 and 5, while the boundary
Where on the curve does the geologic-climatic unit of the between cold and temperate Coleoptera may occur between
interglacial begin? It could be argued that the boundary assemblage zones a and b. These contrasts reflect different
should be placed at that point on the curve where tem- response rates of biota to climatic change and also, insofar
perature increases following a thermal minimum (point X). as the taxa are derived from both marine and terrestrial
Alternatively, the boundary could be located where a environments, time delay between atmospheric and oceanic
temperature similar to that of today is achieved (point Y). temperature changes. Within a single sequence, therefore,
A third view would be to place the boundary at a point the boundary of a geologic-climatic unit could be placed
where a particular temperature threshold is crossed (point at any one of several levels. Once again, the problems that
T) as indicated, for example, by the first occurrence of a this can pose for inter-regional comparisons and
certain indicator species in the fossil record. The problem, correlations are clear.
of course, is that we are attempting to fix a point boundary The designation of geologic-climatic units is therefore
on what is a continuum of climatic change. Any of the frequently intuitive, often arbitrary, and much less precise
above could be used as a basis for defining the onset of an than other forms of stratigraphic subdivision. Indeed, this
interglacial but, because of the diachronous nature of was recognized in a subsequent code of practice produced
climatic change, it is unlikely that an interglacial established by the North American Commission on Stratigraphic
in one region will be a temporally equivalent geologic- Nomenclature (1983), where the concept of geological
climatic unit to that in another. Similar problems are climatic units was abandoned because ‘inferences regarding
encountered in determining the end of an interglacial. climate are subjective and too tenuous a basis for the
Moreover, the extent to which climatic episodes will be definition of formal geologic units’ (p. 849). That stance is
represented in the stratigraphic record will depend both unhelpful to the Quaternary researcher, however, since
on the amplitude and duration of climatic shifts, and also repeated climatic change is undoubtedly the dominant
on the sensitivity of the proxy evidence upon which climatic characteristic of the Quaternary, and it is difficult to
inferences are based. envisage a stratigraphic scheme that does not explicitly
Further complications arise when different types of acknowledge this fact. For example, in the marine oxygen
evidence are being employed in environmental recon- isotope record discussed in the preceding section, the
struction. In Figure 6.10, a transgressive sequence of marine isotopic signal mainly reflects fluctuations in global ice
sediments has been divided on the basis of different volume (section 3.10.2) and these, in turn, are driven
lithostratigraphic and biostratigraphic criteria. Each principally by changes in climate. In essence, therefore, the
particular line of evidence may define a different geologic- subdivision of the marine record into distinct isotope
climatic unit, however. On the basis of the molluscan stages is a climatostratigraphic scheme. Hence, despite the
records, for example, the boundary between warm and cold problems surrounding definition and application, geologic-
l i st c e p t i b i l i t
Lithostratigraphy Biostratigraphy Geologic-Climatic Units
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
s cpet ip t i b i l i t
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
S u s c e p t i bSi u
Susceptibilit
Susceptibilit
Susceptibilit
Susceptibilit
bilit
Susceptibilit
Environment of
deposition
ce
S uSs u
Silt with c 4 5
6 3
organic d Temperate
Salt marsh
detritus 4
1
3 Tern pe rale
6 3
2 c
2 Temperate
b Estuarine
Silt
2
Normal 2
4 Shallow water
3 polarity
b
4 3
Cool
Estuarine
2
5 1 Cool
Magnetic
1
Clay exclusion a 1
a Deep water Cool
e
1
Figure 6.10 Stratigraphic subdivision of a Quaternary sedimentary sequence based on different criteria and the geologic-climatic
units that can be inferred.
climatic units seem destined to remain an integral and horizons, even where the age of each horizon is unknown
necessary tool in Quaternary stratigraphy. and many of the rocks are missing. For example, reference
can be made to the time interval between two biozones or
two lithological boundaries. If these boundaries prove to
6.2.3.7 Chronostratigraphy
be time-parallel, they can be employed as chrono-
Chronostratigraphy is the classification of the stratigraphic stratigraphic boundaries, and the biozone or lithofacies or,
record in terms of time. Once stratigraphic units have indeed, any other stratigraphic unit, can be considered as
been established on the basis of visible, instrumental, a chronostratigraphic unit. Chronostratigraphy, like
biological or inferred climatic characteristics, and a relative climatostratigraphy, is inferential, since it is not based in
chronology determined by superposition and correlation, the first instance on the characteristics of the sediment
it is important (wherever possible) to relate the sequence record (Salvador, 1994).
of events preserved in the rock-stratigraphic record to The division of the stratigraphic record on the basis of
time. The purpose of chronostratigraphy is to divide time is termed geochronology. As explained in section
sequences of strata into chronostratigraphic units that 1.3, chronostratigraphic units represent a time period
correspond to intervals of geological time. Such units are between boundaries and, where the actual ages of such
bounded by isochronous surfaces or chronohorizons. boundaries are known, the interval of time itself is referred
Chronostratigraphic units can be defined on the basis of to as a geochronological unit. It is worth re-emphasizing
geological age, where this can be established, or in terms the distinction between a chronostratigraphic unit and a
of time intervals between isochronous horizons. In other geochronological unit since the two are frequently confused
words, where chronostratigraphic comparisons are being (Table 6.2). The term ‘chronostratigraphic unit’ refers to
effected between sites, reference can be made to the time the rock sequence laid down during a particular time
period encompassed by two designated stratigraphic interval. Hence the term ‘Quaternary system’ is used to
Table 6.2 Conventional hierarchy of chronostratigraphic and geochronological units (after Hedberg, 1976).
Chronostratigraphic Geochronological Examples
Eonothem Eon Phanerozoic
Erathem Era Cenozoic
System Period Quaternary
Series Epoch Pleistocene
Stage Age Devensian
Chronozone Chron Younger Dryas
describe the rocks and sediments that have accumulated comparatively short timespan of the Quaternary, and from
over a time range that is called the Quaternary. The the fact that the relatively fine divisions of the Quaternary
geochronological equivalent (geochronological unit), that stratigraphic sequence are of equal importance to the much
is, the time interval itself, is referred to as the ‘Quaternary coarser subdivisions of the earlier geological record. Only
period’. The same applies to the Holocene ‘series’ (chrono- a relatively small number of geochronological methods
stratigraphy) and Holocene ‘epoch’ (geochronology). It cover the whole of Quaternary time (Chapter 5), while
has been suggested, however, that this distinction in strati- dating aberrations are likely to pose particular problems in
graphic terminology between time-rock units (chrono- Quaternary research because of the more limited time
stratigraphy) and geologic time units (geochronology) range involved. Moreover, if the statistical uncertainty
should now be abandoned, and that the term ‘chrono- associated with the dates (i.e. the quoted ± value) is as great
stratigraphy’ should, henceforth, be the term used to define as or greater than the stratigraphic intervals that are under
intervals of geologic time within rock strata. As a conse- examination, then clearly these age determinations are
quence, ‘geochronology’ would revert to its original insufficiently precise either to define the boundaries of such
meaning of numerical age dating (Zalasiewicz et al., 2004). units, or to distinguish subdivisions within those units. A
Under this scheme, the terms ‘eonothem’, ‘erathem’, further difficulty arises over the recognition of isochronous
‘system’, ‘series’ and ‘stage’ (Table 6.2) would become horizons in the stratigraphic record. As we have already
redundant in favour of ‘eon’, ‘era’, ‘period’, ‘epoch’ and explained, in pre-Quaternary strata, lithostratigraphic and
‘age’. Whether this proposal, which represents a funda- biostratigraphic boundaries, although inherently time-
mental change to the traditional time-stratigraphic sub- transgressive, appear to be synchronous when set against
division of the geological record, will prove to be accept- the vast span of geological time. As such, they are frequently
able to the wider community remains to be seen. used as time-stratigraphic markers. In the Quaternary
The foregoing notwithstanding, the current basic stratigraphic record, by contrast, most boundaries are
working units in Quaternary stratigraphy are stages. manifestly time-transgressive, particularly biozones whose
Conventional stages in pre-Quaternary time typically range boundaries reflect biological response to climatic change
from 3–10 million years, but in the Quaternary, stages are which is spatially and temporally diachronous (Walker,
measured in tens of thousands of years. Stages can be 1995). Nevertheless, some Quaternary stratigraphic units
divided into smaller units or substages as, for example, in or boundaries do provide a basis for time-stratigraphic
the marine oxygen isotope record (section 6.3.2.7). The correlation, and it is to these that we now turn our attention.
basic chronostratigraphic unit is the chronozone, the
timespan of which is usually defined in terms of the 6.3 TIME-STRATIGRAPHIC
timespan of a previously designated stratigraphic unit such
as a formation, or a member, or a biozone. Chronozones
CORRELATION
have been most widely employed in Quaternary research
where biozones have been dated by radiometric methods,
6.3.1 Principles of Quaternary correlation
for example radiocarbon-dated pollen assemblage zones The stratigraphic methods outlined above all provide a
(section 4.2.4). basis for correlation; in other words, the relationship of
The division of Quaternary strata into chronostrati- stratigraphic sequences or events at one locality to those at
graphic units is seldom straightforward, however, and here another. Throughout most of the geological column,
again problems arise that are not encountered in the earlier lithostratigraphic and biostratigraphic boundaries are, at
geological record. In the main, these result from the the level of resolution applied, effectively time-parallel, and
are therefore regarded as being almost of equivalent status mode of origin, but of very different ages, are incorrectly
to chronostratigraphic units in time-stratigraphic sub- correlated, such errors of correspondence are referred to
division and subsequent correlation. This assumption as homotaxial errors.
cannot be made in the correlation of Quaternary suc- This type of error can also arise in biostratigraphy. It is
cessions, however, except perhaps at the local scale. Not entirely possible, indeed likely, that a very similar
only are stratigraphic boundaries time-transgressive (see environment existed during both interstadials A and B, and
above), but the repetitive nature of Quaternary climatic this would be reflected in the fossil evidence preserved in
change means that at any given locality, similar depositional the organic horizons at sites 2, 3 and 4. Again, the inference
sequences may be preserved that are of markedly different of a single interstadial based on the pollen evidence from
age. In view of the fragmented and highly diverse nature the three sites would be homotaxially incorrect. Finally, the
of the Quaternary stratigraphic record, effecting meaningful diagram illustrates the time-transgressive nature not only
correlations between often widely separated sites is of the lithostratigraphic and biostratigraphic units, but
therefore seldom straightforward. also of the geologic-climatic units. At site 2, the geologic-
Some of the difficulties that can arise in the correlation climatic unit (interstadial) inferred from the fossil content
of Quaternary sequences are exemplified by Figure 6.8 of the organic deposit spans the time interval a–b, while at
which shows, in a diagrammatic way, the extent of site 3, the same interstadial covers the time interval c–d.
glaciation over time, the shaded area representing the time These complications arise partly because the deposits
period during which ice covered the ground at increasing of varying ages are not arranged vertically in order of
distance from an ice-dispersal centre. Site 1 was affected by superposition, and partly because lateral correlation often
glacier ice for almost the whole of the ‘glacial’ time interval; has to be made between numerous short-lived depositional
sites 2 and 3 were occupied by ice for a shorter period, while records. The above example relates to a single glacial
site 4, and especially site 5, were ice-covered for only a small interval, yet when it is recalled that at least twenty glacial–
proportion of the glacial episode. Sites 2, 3 and 4 also interglacial cycles affected the mid- and high-latitude
experienced interstadial conditions between successive regions of the world during the last 800 ka alone, the scale
glacial advances. At each site, the environmental history is of the problem begins to emerge. Moreover, it is not only
recorded in sequences of glacigenic sediments, interbedded in formerly glaciated areas that difficulties are encountered.
at sites 2, 3 and 4, with organic deposits (peat or soil). Repeated climatic changes had a profound effect on those
At site 1, which for much of the glacial interval lay some regions that lay beyond the margins of the ice sheets, and
distance up-glacier, erosion probably dominated and the there too the complicated erosional and depositional history
only record of glacier activity is a thin layer of lodgement of the Quaternary presents the stratigrapher with major
till deposited during glacier wastage. By contrast, a complex problems of correlation. Clearly, therefore, at the regional
sequence of glacigenic sediments accumulated at site 5 and continental scales, lithostratigraphy, biostratigraphy
near the ice margin. There are two points to note here: first, and climatostratigraphy are not sufficiently sensitive tools
the thickest and most complex sequence of sediments is with which to effect meaningful time-stratigraphic cor-
preserved at the site that was covered by glacier ice for the relations. A geochronological basis is required, although as
shortest length of time; and second, although the glacier yet no single radiometric dating method has been developed
reached site 1 before site 5, on a time-stratigraphic basis, that is applicable to the whole of the Quaternary period at
the deposits that are actually preserved at site 1 may, in fact, the required level of temporal resolution. There are,
be younger than those at site 5. At sites 2, 3 and 4, two till however, a number of ways in which sequences may be
units occur, but from the stratigraphic evidence alone, it is correlated on a time-stratigraphic basis, and some of these
far from clear which till unit is the correlative of the are examined in the following section.
glacigenic sequence recorded at site 5. Moreover, because
sites 2, 3 and 4 have experienced two periods of glaciation
6.3.2 Bases for time-stratigraphic
separated by an interstadial interval, almost identical
stratigraphic sequences (till–organic sediments–till) have
correlation
developed. In the absence of other evidence, correlation
6.3.2.1 Palaeomagnetic correlation
might therefore be unwittingly effected between these
sequences. Yet, as the diagram shows, the sequence of Magnetostratigraphy utilizes stratigraphical variations in
deposits at site 4 relates to an earlier period of time and to the magnetic properties of rocks as a basis for geological
a stadial–interstadial oscillation which is different from that correlation. The palaeomagnetic record can be divided
at sites 2 and 3. Where units of similar composition and into magnetozones (units of rock with a specific magnetic
signature), the boundaries of which reflect abrupt changes marker horizons and have been used for both local and
in the earth’s magnetic field (see section 5.5.1), and are long-distance time-stratigraphic correlation. Subsequently,
therefore clearly defined. Because these changes are however, a more cautious approach has been adopted to
experienced globally, geomagnetic reversals and prominent the use of buried palaeosols in stratigraphy and correlation,
excursions provide a basis for correlating Quaternary events for it is now recognized that degree of soil development as
at the global scale. Palaeomagnetic stratigraphy has also a direct function of time can only be assumed where other
been employed to correlate marine, terrestrial and, in some soil-forming factors (parent material, climate, slope and
cases, ice-core records (see below). On the other hand, the biological factors) can be shown to have been constant
method is restricted to certain rock types (e.g. volcanic (section 5.6.5.2). Moreover, insofar as most soils are
rocks) and specific depositional environments (such as polygenetic, any buried profile may be the product of more
lake sites or aeolian contexts), while the timing of some than one phase of pedogenesis. Indeed, the Sangamon Soil
palaeomagnetic events remains to be established and the of North America, hitherto widely considered to be of last
temporal precision, particularly of older geomagnetic interglacial age, appears to have formed over a period of up
boundaries, is relatively low. At the regional scale, secular to 50 ka (MIS 5–3) during which time the climate shifted
magnetic variations (section 5.5.1.2) can be used as a basis from cold to warm and back to cold again (Hall &
for correlation between lake sequences, the distinctive Anderson, 2000).
inflections or turning points in the magnetic field meas- Despite these limitations, however, palaeosols have
urements providing a basis for correlation between different been used successfully in the development of both local
sedimentary records (Snowball et al., 2007). Thus far, histories and wider correlative schemes. In northern
however, this aspect of palaeomagnetic correlation has France and the American Midwest, fossil soils of Late
been applied routinely only to Holocene lake deposits. Pleistocene age constitute key correlative elements in the
stratigraphic record (Antoine et al., 2003; Rutter et al.,
2006). A particularly useful example in the latter region is
6.3.2.2 Correlation using tephra layers
the Farmdale Geosol of Illinois, which is radiocarbon dated
Tephra layers (section 5.5.2) constitute marker horizons to 28–25 ka and which forms an important stratigraphic
in the stratigraphic record that are essentially isochronous, marker for correlating Late Wisconsinan events at both
and hence have the potential to serve as the basis for time- local and regional scales (Jacobs et al., 2009). In areas of
stratigraphic correlation between sediments formed on the world where palaeosols are found in loess sequences,
land, in lakes and on the seabed, and between these such as those on the Chinese loess plateau, soil-stratigraphic
sequences and polar ice. Although correlation on the basis units provide a basis for correlation between different loess
of tephra deposits is not as universally applicable as that profiles (Bronger, 2003) and also, in a much wider context,
based on palaeomagnetism (e.g. the Brunhes–Matuyama between loess–palaeosol sequences, global ice volume cycles
boundary), because individual ash beds have relatively and the marine oxygen isotope record (Dearing et al.,
limited geographical ranges, tephrochronology (the dating 2001)
and correlation of deposits based on distinctive tephra
horizons) is proving to be an increasingly powerful cor-
6.3.2.4 Shoreline correlation
relative tool at the regional or sub-hemispherical scales.
Examples of the use of tephrochronology in different Marine shorelines and deposits may also, in certain con-
depositional contexts are discussed in section 5.5.2.3. texts, provide a basis for time-stratigraphic correlation. If
world sea levels are stable for long enough to allow the
development of shoreline features, then these essentially
6.3.2.3 Correlation using palaeosols
isochronous reference surfaces potentially offer a basis
Palaeosols (section 3.5) have been perhaps most widely for inter-regional correlation. Isochronous shorelines can
employed as a correlative tool in North America where it often be traced around a coastline, even in regions affected
has traditionally been assumed that pedogenesis during by glacial isostasy or tectonic uplift (Fjeldskaar et al., 2000).
interglacials and certain interstadials alternated with periods If the shoreline cuts certain sedimentary units, but is
of negligible profile development or arctic soil formation overlain by others, then the shoreline can be used as a time-
in the intervening cold phases. Hence well-developed stratigraphic reference plane for separating deposits of
buried soils (such as the Sangamon Soil of the American different age. Sedimentary units formed as a result of catas-
Midwest) have been considered to constitute stratigraphic trophic marine events, such as tsunamis, also constitute
marker horizons for correlating between marine and 6.3.2.6 Event stratigraphy and correlation
terrestrial sequences (Smith et al., 2004). At the broader
scale, evidence of former sea levels (raised beaches and Events may be broadly defined as comparatively rare
corals, submerged corals, submerged clifflines, etc.) may and geologically short-lived occurrences that have left some
provide a basis both for inter-regional correlation, and also trace in the rock records. The stratigraphic signatures of
for linking terrestrial and deep-sea records (Hearty & these sudden events are effectively isochronous markers,
Kaufman, 2000). However, problems can arise from com- and hence constitute a tool for correlating geological
sequences. This is the basis of what has been termed event
plexities in the coastal records induced by variations in
stratigraphy or event chronostratigraphy (Gale, 2009).
amounts and rates of local tectonic activity (Marquardt
The products of some events (tephra from volcanic
et al., 2004), and perhaps also from the phenomenon of
eruptions; tsunami deposits in marine sequences) have
geoidal eustasy (section 2.5).
already been discussed, but others might include the
In addition to marine shorelines and deposits, lake
geological manifestations of storms, floods, earthquakes,
shorelines can also provide timelines for correlating
mass movement and turbidity flows, although clearly not
between, and establishing the order of superposition of,
all of these will be sufficiently widespread to constitute a
local geological features (Orme & Orme, 2008). Moreover,
basis for inter-regional correlation. Other event products,
as some lake shorelines reflect climatically driven fluctu-
however, such as the layers of carbonate-rich debris in
ations in lake-water level (section 3.7.3), they provide a basis
North Atlantic Ocean sediments formed during major
for correlating climatic events in terrestrial and marine
ice-rafting events (‘Heinrich events’; section 3.10.1) of the
records (Benson et al., 1998).
last cold stage, offer a means both for inter-core correlation,
and also for linking oceanic and terrestrial sequences
6.3.2.5 Correlation on the basis of radiometric (Hemming, 2004). Other characteristics of the geological
dating record that reflect widespread and large-magnitude events
may also offer a basis for broader time-stratigraphic
These methods, which were considered in section 5.3,
correlation. These include palaeomagnetic changes pre-
are an extremely important independent means of long-
served in Quaternary sediments (see sections 6.3.2.1 and
distance correlation, for they form time-planes across the 5.5.1), and a range of geochemical indices, including
stratigraphic record against which the time-transgressive stable isotopes of oxygen, carbon, sulphur and strontium,
litho-, bio- and morphostratigraphic boundaries can that form stratigraphic markers. For example, an event
be measured. There is, however, a substantial body of stratigraphy has been proposed for the Last Termination
opinion in support of the view that the subdivision of the in the North Atlantic region based on the oxygen isotope
Quaternary should rest primarily on the stratigraphic signal in Greenland ice cores (Björck et al., 1998; Lowe
record, with radiometric dates being merely a means et al., 2008b; section 6.3.3), while a pollen-based event
whereby that record can be underpinned. Certainly, any stratigraphy for the same time interval has been developed
date applies only to the locality and to the horizon from for the Lake Suigestu sequence in Japan (Nakagawa et al.,
which it was obtained, and it can only be related to other 2005). An unusual type of event stratigraphy has involved
sequences on the basis of the observed stratigraphic record the use of the ‘plateaux’ in atmospheric 14C activity during
at the different localities. As has already been shown, no the Lateglacial and early Holocene (section 5.3.2.6) as a basis
radiometric date is free from analytical errors and, in some for time-stratigraphic correlation, working on the
cases, the error (±) associated with the age determination assumption that the individual plateaux reflect essentially
may be so great that the date cannot be used effectively in isochronous and universal marker horizons (Hajdas et al.,
time-stratigraphic correlation. Equally, all dated samples are 2003). Certain biological events may also be used for
prone to errors of contamination, some of which may go correlation purposes. For instance, the sudden decline in
undetected. Each date should not only be carefully checked Ulmus in Holocene pollen records from western Europe,
against other age determinations, therefore, but must be which has been radiocarbon dated to around 5.8 ka cal.
thoroughly evaluated in the light of its stratigraphic context BP and which appears to reflect a rapid and widespread
before it is used as an aid in correlation. Overall, radiometric biotic catastrophe (Parker et al., 2002), could be employed
dating is perhaps best regarded as a means of corroborating in time-stratigraphic correlation. In earlier Quaternary
and validating other stratigraphic and correlative proced- sequences, mammalian extinction or evolutionary events
ures, rather than as the primary basis for time-stratigraphic might also offer a basis, not only for relative dating (section
correlation. 5.5.4), but also for regional and inter-regional correlation.
6.3.2.7 Correlation using the marine oxygen individual isotopic profiles is based largely on a ‘count from
isotope record the top’ principle, and hence the possibilities of homotaxial
error are always present. In addition, the quantification of
Although the low sedimentation rate and slow mixing orbital cyclicity, which is fundamental to precise orbital
times in the deep oceans means that oxygen isotope pro- tuning, may be compromised by ‘signal noise’ in the marine
files from ocean cores cannot be used in the correlation of isotope records, resulting from time distortion (non-linear
recent marine sequences, or as a basis for high-resolution sedimentation rates), diagenesis, different sampling proto-
correlation below the centennial scale (Shackleton, 2006), cols and analytical errors (Meyers et al., 2008). A further
on millennial and longer timescales, the isotopic stage potential source of uncertainty arises from the differen-
boundaries, and especially the Terminations, are essentially tiation between interglacials and interstadials. It has already
synchronous. They therefore constitute key marker hori- been shown that MIS 3 is regarded as being of interstadial
zons within the ocean sediment sequences and offer the rather than interglacial status. The possibility cannot,
potential for the development of a correlative chronology therefore, be excluded that some previous interglacials and
that is globally applicable over most of the Quaternary time interstadials have been confused, particularly in the earlier
range (Hilgen et al., 2006). Moreover, the stage boundaries part of the Quaternary record. If this has occurred, then
can be dated by orbital tuning (section 1.4), and hence the it clearly has implications for the status attached to iso-
marine oxygen isotope record, and particularly the LR04 tope stages, and could lead to problems in correlating
‘stacked record’ (Figure 1.5), constitute the type sequence between continental records and the marine oxygen isotope
for the Quaternary against which other isotopic profiles can sequence (section 6.3.3.1). By the same token, the com-
be compared. plexities of MIS 5 and 7 (section 6.2.3.5) are surely not
Deep-ocean records undoubtedly hold a number of unique to the upper parts of the isotopic sequence.
advantages over terrestrial sequences from the point of Consistency of interpretation between different isotopic
view of stratigraphic subdivision and correlation. First, the profiles may not, therefore, be easy to achieve, particularly
sediments from which the isotopic data have been obtained in those cores where stratigraphic resolution is low.
appear to be relatively undisturbed. Second, a common Although the above difficulties have yet to be satis-
technique (oxygen isotope analysis) can be used to compare factorily resolved, there is no doubt that the isotopic trace
profiles from widely scattered localities on the deep-ocean in the ocean sediments constitutes a remarkable record of
floors. Third, the Terminations can be used as universal Quaternary climatic change, and it is now widely accepted
reference points in inter-core correlation. Fourth, although that this, and not the terrestrial sequence, provides the basic
the isotopic changes are a consequence of climatic changes, framework for a global scheme of Quaternary correlation.
and are therefore time-transgressive, this to a very large Because the isotopic stages are a reflection of climatic
extent is masked by the slow rate of sediment accumulation. change, they are essentially geologic-climatic units and, as
As a consequence, the major isotopic stage boundaries and such, should have correlatives in the terrestrial record.
Terminations can be interpreted as essentially time-parallel How these two can be linked is considered in the final
horizons. Fifth, orbital tuning enables a timeframe to be section.
established for the isotope curves, with key levels in the cores
being dated on the basis of periodicities obtained from
astronomical calculations. Moreover, marker horizons
6.3.3 Correlation between continental,
such as the Brunhes–Matuyama boundary can also be marine and ice-core records
dated by the independent method of palaeomagnetic If correlations are to be established with oceanic sequences,
stratigraphy (Hilgen et al., 2006). terrestrial records should possess certain characteristic
Interpretation of the isotopic evidence is not always features: (1) evidence of climatic change that is clear
straightforward, however. Continuity of sedimentation and unequivocal so that the climatic signal can be com-
can never be proved, and it is questionable whether, in pared directly with that in marine oxygen isotope profiles,
practice, gaps in the sedimentary record can ever be reliably or other elements of the deep-ocean sequence; (2) a history
detected. Problems also arise over poor stratigraphic of sediment accumulation that has been continuous, or
resolution, over bioturbation and reworking of sediments, more or less continuous; and (3) an independent chron-
and over the recognition and precise definition of stage ology based on radiometric, incremental, or age-equivalent
boundaries in some isotopic profiles. Moreover, although dating, and which therefore enables a time-stratigraphic
the records can be dated by orbital tuning and by reference correlative framework to be established. In practice, it is
to the palaeomagnetic timescale, correlation between perhaps the third of these that poses the greatest problem
for land–sea correlations, partly because of the technical difficulties may be resolved if independent chronological
limitations of the methods (Chapter 5), partly because data are available to underpin the records. These enable
of the restricted time ranges of some techniques, and partly time-stratigraphic equivalence to be established, in which
because many stratigraphic records lack either suitable case it may prove possible to synchronize the different
material for dating, or clearly defined time-stratigraphic records, which is the optimum situation in land–ocean
marker horizons (palaeomagnetic boundaries, tephra correlation (section 6.3.3.3).
layers, etc.). As a consequence, correlating terrestrial suc- Ideally, terrestrial sequences that are being used for
cessions to the deep-ocean sequence is frequently based correlation should span the whole Quaternary time
on record or event alignment (sometimes referred to range, as does the marine oxygen isotope signal, and thus
informally as ‘wiggle-matching’ in the literature), whereby comparisons can be made on Milankovitch (or orbital)
inflections in curves of proxy climate data form a basis for timescales. In reality, however, it is only in a few deposi-
aligning events in one record with their assumed closest tional contexts, such as tectonic basins and deep lakes,
equivalents in another. Once such matches have been where there are continuous sedimentary sequences that
established, it may be possible to tune one record against extend back to the beginning of the Quaternary and beyond.
the other. This involves changing the chronology of a But there are a number of terrestrial sites where continuous
record, either by matching it against a better-dated one, or depositional records are preserved that span several
by matching it with a known cyclicity based on an glacial–interglacial cycles, and these also enable land–
established cause-and-effect relationship, as in the tuning marine correlations to be effected at the Milankovitch
of the marine isotope signal and isotopic/trace gas records scale. In addition, there are other contexts where the sedi-
in polar ice cores to the astronomical (or orbital) timescale mentary successions are shorter and more fragmented,
(sections 5.4.3 and 5.5.3). but these may be better resolved, both stratigraphically and
Although alignment and tuning of palaeoenvironmental temporally. As such, they allow more precise correlations
records are now routinely employed in the correlation of to be established between components of terrestrial and
Quaternary sequences, these approaches are also not marine sequences, often on sub-Milankovitch timescales
without problems. The assumption in sequence alignment (section 1.5). Some examples of the types of record that can
is that the climate signal that is reflected in proxy records provide a basis for terrestrial and marine time-stratigraphic
in different terrestrial and oceanic archives is broadly correlation on long-, medium- and short-term timescales
time-parallel. But as we have already seen, climate change are considered in the next two sections.
at decadal, centennial and, sometimes, at millennial
timescales may be diachronous. Moreover, some elements
6.3.3.1 Long-term correlation on Milankovitch
of the biosphere, such as insects, respond more rapidly to
timescales
climate change than, for example, plants, and hence there
will often be a degree of time-transgression in climate In recent years, technological advances in subsurface
signals from different proxy records. In highly resolved exploration and in coring have revealed long sedimentary
stratigraphic sequences, therefore, alignment of, and tuning sequences in some continental localities which, in certain
the climate signal between, different land and ocean exceptional cases, span much or all of the Quaternary and
archives may not be a straightforward procedure. In the beyond. Proxy climate records reveal long-term climatic
deep oceans, by contrast, slow rates of sedimentation on cycles that appear to have resonated at Milankovitch
the deep-ocean floor tend to mask any spatial variation in periodicities, and hence offer a potential basis for correla-
the marine oxygen isotope signal, and hence the major tion with marine oxygen isotope records throughout
inflections in the isotopic trace, especially the Terminations, the entire Quaternary time range. Some of the longest
constitute essentially time-parallel markers for correlating sequences are preserved in the great tectonic basins of the
between the different isotopic records (section 6.3.2.7). world, such as that beneath the high plains of Bogotá in the
On the other hand, the relatively low stratigraphic resolu- Eastern Cordillera of Colombia, the pollen record from
tion of this record by comparison with some terrestrial which extends continuously back into the Pliocene (Torres
sequences, such as polar ice, means that aligning and tuning et al., 2013), and the Wanganui Basin of New Zealand,
the terrestrial record to the marine isotope signal may where great thicknesses of Plio-Pleistocene shallow-water
fail to reveal possible leads and lags in the earth–ocean– marine deposits can be linked to the marine oxygen isotope
atmosphere system. In other words, the tuning process record using a combination of biostratigraphy, magneto-
may tend to ‘blur’ or smooth out what may be a complex stratigraphy and fission track dating of interbedded rhyolitic
climate signal. As noted above, however, these and other tephra layers (Pillans et al., 2005). In the Northern
Hemisphere, the Central Graben of the southern Nether- In some of the world’s larger lakes, such as Lake Titicaca
lands contains fluvial and marine sediments extending in Bolivia/Peru (Fritz et al., 2007) and Bear Lake, Utah/
back to the Reuverian Stage of the upper Pliocene, and the Idaho (Colman et al., 2006), sediment accumulations may
quantified temperature record derived from pollen evidence span several glacial–interglacial cycles, but few lakes con-
shows clear parallels with the marine oxygen isotope record tain a record that extends continuously through the full
(Figure 6.11). Quaternary timespan. Exceptions include Lake Biwa in
b iul ist c e p t i b i l i t
Inferred
Oxygen isotope temperature
based on
Susceptibilit
stage
Susceptibilit
Susceptibilit
Susceptibilit
pollen e v i d e n c e
SusceptiS
Cold Warm
SusceptiS
SusceptiS
o O (%.)
1 a
(X)
5 4.5 4 3.5 5° 15" 25°

Hobcene 0 2
Late Weichselian
Eemian 5
Pleistocene 6¬
0.2 7
Saalian 8
5
10
Holsteinian Brunhes 11
0.4 12
Middle Elsterian
13
Pleistocene 14 15
0.6 15
Cromerian
17
18
20 19
0.8
22
21
Matuyama
25
Quaternary Bavelian
1 Jaramillo
event 30 31
34 •
Menapian
1.2 36 37
Waalian
1 4 47
Matuyama
52
55
Early 1.6
Pleistocene Eburonian
62 63
1 8
Olduvial 62
event
2 75
Tiglian
81
82
2.2
Matuyama
91
2.4
98¬
Pretiglian
100
-103
2.6 104
Pliocene Reuverian Gauss
-100 -50 -10 0

Sea level relative to present (m)
Figure 6.11 The Quaternary sequence in the Netherlands matched against the marine oxygen isotope record. Pollen-based
temperature reconstructions are shown on the right (after de Mulder et al., 2003).
6 1 8
0 (%o) BioSi0 . %
2 Diatom abundance
min, valves/gram
O D P site 6 7 7 BDP-96-2 BDP-96-2
5.5 4.5 3.5 0 20 40 0 500 1000

0 0 MIS 1
MIS 3
MIISS
M
100 MIS MIS 5
MIS
@
200 MIS
MIS MIS 7
MIS
MIS
300 MIS
MIS 9
MIS
0 100 200
Susceptibilit
400 M I S 11
500 MIS 13
a
MbI S
MIS 15
600 MIS
e
MIS 17
700
M I S 19
800
Figure 6.12 Comparison of biogenic silica and diatom records from Lake Baikal with the marine oxygen isotope signal from
ODP site 677. Diatoms were relatively less abundant in the upper part of the record, but the amplified curve (inset A) shows
that the cyclical pattern of diatom abundance continues to parallel the BioSiO2 profile throughout the MIS 9c-1 interval (after
Khursevich et al., 2001).
Japan (Miyoshi et al., 1999) and Lake Baikal in Siberia. (Kashiwaya et al., 2001). In East Africa, records extending
The latter is the deepest lake in the world and, in parts of back to the late Pliocene have been obtained from the Rift
its basin, sediments exceed 5 km in thickness and may be Valley (Trauth et al., 2005), although the sedimentary
between 20 and 40 Ma in age (Williams et al., 2001). Close regimes in these lakes reflect a complex interplay of climatic
correlations have been established between biogenic silica and tectonic factors (Bergner et al., 2009).
and diatom data from Lake Baikal and the marine oxygen There are other terrestrial sedimentary successions that
isotope sequence (Figure 6.12), a comparison that now span much or all of the Quaternary, most notably the
extends beyond 1.8 Ma (Prokopenko et al., 2006) and thick loess deposits of the continental interiors (section
which may have the potential to reach as far back as 12 Ma 3.6.5) in which aeolian and soil units provide a remarkable
5 1 8
0 (V)
4 2 0 -2 -4 Baoii Loess Section
00
2 S0
S0 0071
6
• S0
S0 0205
7 S0S0 0242
8
S0 S0
S0 0.307
10
9
S0
11
S0 0.415
12 S0
S0
0 5 13 S0
14 S0S0 0.506
0543
S0S0
15
16
17 S0 0630
18 S0 S0
B 19
S0 B
0693
S0 S0
20 0.740
21 M
22
22
23 S0
24
S0S0 0.861
S0S0 S0
25
26 0.920
27
S0
S0 S0
Susceptibilit
28 J 0972
29
S0
Susceptibilit
J 30 1.000
1.0 S0 S0
32
31
S0
S0 S0 1.058
22 S0 S0 S0 0.071
34 S0 S0
35 S0
S0 S0 1.208
36 37 S0
S0 S0 1 257
38
39
S0
S0 S0S0 1.299
40 S0 S0 S0 1.310
42
41
S0 S0 1.363
43
S0 S0 1.420
44 45 S0 S0 1.479
46 47 S0 S0 S0 1.522
48 22 S0 S0
50 S0S0 S0
51 S0
S0 S0 1.628
1 5
52 S0 S0 S0 1.690
53 S0 S0
S0 S0
54 0 1.777
55
S0 S0 S0
56
57
S0 S0S0
S0 S0
1.985
58
59
S0 S0 2077
C 60 S0 S0 S0
2.150
61 S0 S0 2.218
62 S0 S0 S0
S0
63 S0 2.404
2.50 S0 M
S0 G
2.500
valves/gra
Figure 6.13 The sequence of loess–palaeosol units preserved at the Baoji site on the Loess Plateau, China, plotted against the
marine oxygen isotope record. B/M – Brunhes–Matuyama boundary; J – Jaramillo Event; O – Olduvai Event; M/G – Matuyama–Gauss
boundary. Note the isotopic record extends back to c. 1.8 Ma (MIS 63), whereas the loess–palaeosol record goes back to 2.5 Ma
(the beginning of the Quaternary/Pleistocene) (after Rutter et al., 1996).
record of glacial–interglacial cycles. The potential of these 6.3.3.2 Correlation on sub-Milankovitch

loess–palaeosol sequences for linking terrestrial and marine timescales
records was first recognized by George Kukla who
compared the aeolian records from Moravia in eastern As we saw in Chapter 3 (section 3.11.4), the last glacial
Europe with the oxygen isotope timescale, a correlation that cycle in the Greenland ice-core record was characterized
was eventually extended back into the early Quaternary by a series of twenty-five abrupt warming events, the
(Kukla, 1977). Other loess–palaeosol sequences from Greenland Interstadials (GI) or Dansgaard–Oeschger
eastern Europe and central Asia have subsequently been (DO) events, and these have been shown to be consistent
correlated with the deep-ocean oxygen isotope signal in number and amplitude in different Greenland ice cores
(Bronger, 2003). However, it is the loess–palaeosol records (North Greenland Ice Core Project Members, 2004).
on the Loess Plateau region of central China (section 3.6.2) Equivalent features have been detected in marine and other
that offer the greatest potential for terrestrial–marine terrestrial archives (e.g. lakes and speleothem deposits) that
correlations (Ding et al., 2002b; Sun et al., 2006b). Loess span the last glacial period (see below). This sequence of
began to accumulate more than 2.6 Ma ago, and these short-lived but distinctive climatic events was not unique
sediments and their interbedded soils are another depos- to Greenland, therefore, but appears to reflect widespread,
itional record that can form a basis for time-stratigraphic perhaps global, climatic perturbations (Schulz, 2002). The
correlation throughout the entire Quaternary period same may be equally true of the Heinrich events (H), six
(Figure 6.13). prominent cooling episodes that occur in records of the
Long polar ice-core records also provide a means of mid- and later part of the last glacial cycle (section 3.10.1).
land–ocean correlation on Milankovitch timescales as the Their signatures have been detected in both marine and
δ18O values in the two archives are inter-related, the isotopic terrestrial archives and in widely separated localities,
signal in the ice cores being the reverse of that in the prompting the suggestion that these too represent short-
marine record (section 3.11). In one of the earliest attempts term, globally significant, climatic ‘pulses’ (Voelker, 2002).
to utilize this relationship as a basis for correlation, In combination, these two sets of marker events, GI1–GI25
Dansgaard et al. (1982) were able to identify the equivalents in the ice cores and H1–H6 in ocean sediments, provide the
of MIS 1–5e in the Camp Century ice core from northern basis for high-resolution correlations of marine, continental
Greenland, the isotopically ‘warmest’ 18O values being and ice-core sequences.
recorded in levels equating to MI substage 5e. With the The signatures of the Heinrich events, which are marked
recovery of much longer cores from Antarctica, however, in North Atlantic marine sediments by distinctive layers
the ice–marine isotopic correlation has been extended back
to MIS 20.2 (c. 800 ka BP), in the EPICA core from the -32
Dome C drill site (Jouzel et al., 2007).
Susceptibilit
H1 H2 H 3 H4 H5
-36
Another proxy climate record that can be used to link
GRIP
terrestrial and marine sequences is the oxygen isotope -40
signal in cave speleothems (section 3.8.4.2). In some cave
sites, particularly those located in low latitudes where -AA
0 10 20 30 40 50 60
S u s c e p t iSbui lsi tc e p t i b i l i t
speleothem growth is more likely to have been continuous

3000
over relatively long periods of time (McDermott, 2004), a
Susceptibilit
close relationship has been observed between episodes of 2000 l i tu s c e p t i bSi lui ts c e p t i b i l i t
more rapid speleothem growth and warmer (and wetter) 1000 2

periods, as inferred from other proxy climatic data. Records
0
from the Hoti Cave in northern Oman, for example, show 3
enhanced phases of speleothem growth at 132–125 ka, 4

200–180 ka and 325–300 ka, in other words during MI
stages 5e, 7 and 9, reflecting episodes of increased monsoon 5

0 10 20 30 40 50 60
rainfall (Burns et al., 2001). U-series dated stalagmite
A g e (ka)
records from the Hulu and Sanbao Caves in China show
evidence of 23 ka (precessional) cycles over the last 224 ka
Figure 6.14 Matching of marine sediments in core ENAM93-
which are linked to changes in the East Asian monsoon, and 21 from the northeast Atlantic Ocean with the GRIP δ18O
which can be correlated with MI stages back to MIS 7.3 ice-core record using Heinrich events (H1–H5) as tie-points
(Wang et al., 2008). (after Cortijo et al., 2000).
-34
G r e e n l a n d Ice
-36
Susceptibilit
-38
•40
-42
30
S u s c e p t i bSi lui ts c e p t i b i l i t
Susceptibilit
20
10
30
Susceptibilit
20
Northern 0
10 Atlantic
4 Arabian | S e a
Susceptibilit
Susceptibilit
10 Brazilian Margin
i i
Susceptibilit
Figure 6.15 Tentative correlations between records from the Greenland ice sheet, northern North Atlantic, Arabian Sea and
Brazilian Atlantic margin, using a variety of proxies to define Heinrich events (H1–H6), and the intervening DO events. YD – Younger
Dryas cold event (after Broecker & Hemming, 2001).
of ice-rafted debris (IRD: section 3.10.1), register clearly in Further linkages between marine and terrestrial events
the oxygen isotope record in the Greenland ice cores in the North Atlantic province have been established using
(Figure 6.14), and this not only enables correlations to be evidence from ocean cores from the Iberian margin of
established between the marine and ice sheet archives, but southwest Europe. These have yielded both marine micro-
also allows calendar ages to be ‘imported’ into the marine fossil data (from which sea-surface temperatures have
record from the annually counted 60 ka Greenland ice-core been reconstructed) and pollen evidence, the latter enabling
chronology (Svensson et al., 2008). This is important direct comparisons to be made with pollen-based climate
because dating of the Heinrich IRD layers by radiocarbon reconstructions from terrestrial sites. The proxy records
has proved problematical, due largely to uncertainties suggest a close link between millennial-scale oscillations
over the marine reservoir effect (section 5.3.2.4). Although in the North Atlantic and vegetation responses on the
the IRD record of Heinrich events is confined to the mid- adjacent western European landmass (Sánchez Goñi et al.,
and high latitudes of the North Atlantic, changes have 2008). This land–ocean correlation has now been extended
been observed in other marine proxies that may parallel back through the last two glacial periods (de Abreu et al.,
the Heinrich events. These include short-lived increases 2003).
in algal productivity in the sub-Antarctic Ocean, implying Attempts have also been made to compare marine
major reorganizations of ocean circulation (Sachs & and Greenland ice-core records on the basis of equivalent
Anderson, 2005), variations in Fe/Ca ratios in cores from Interstadial (IS) or DO events. For example, in a sequence
near the Brazilian margin of the South Atlantic which of sediments from the Cariaco Basin, located in the tropical
reflect enhanced sediment supply (increased run-off), and Atlantic off the Venezuelan coast, the colour reflectance
reductions in the organic carbon content of sediments in record (which indicates changes in marine productivity) has
the Arabian Sea, indicating colder water temperatures been tuned to the sequence of interstadials and stadials in
(Broecker & Hemming, 2001). If these are indeed the the GISP2 ice core, using prominent IS events as tie-points
correlatives of the North Atlantic Heinrich events, then the (Figure 6.16). A similar correlation has been suggested
HE marker horizons can be used to link marine sequences between the oxygen isotope profile from speleothem in
in different ocean basins (Figure 6.15). Hulu Cave, China, which has also been tuned to the GISP2
7
A M S " C dates Laminated
sediment Dark
S u s c e p t i b i lSi tu s c e p t i b i l i t
12
11 5 7 8 iqn 14 ie 19
13 15 ,17 20 21
2 3 4 18
9
15
1002C
19
Light
Tie points
23 Dolomite
1
i l iut s c e p t i b i l i t
•35
19 20 21
3 12 14
8
4 •37
2 5
67 1011 15 16.17
13 '8
S
l i st c e p t i bS
-39
G I S P II
-41
-43
MIS-1 MIS-2 MIS-3 MIS-4 MIS-5a -45

0 10 20 30 40 50 60 70 80 90
Aae (ka)
Figure 6.16 Comparison between colour reflectance of Cariaco Basin sediments and the GISP2 δ18O record over the last glacial
cycle. The interstadial (IS) events correspond with darker layers in the sediment sequence which reflect higher levels of organic
productivity. The Cariaco record has been tuned to the GISP2 isotopic signal on the basis of matching (equivalent) interstadial
events, with the key tie-points shown by dashed lines (after Peterson et al., 2000).
ice-core sequence, while some of the most prominent shifts 6.3.3.3 Synchronizing records of past
in the isotopic signal are considered to reflect the North environmental change
Atlantic Heinrich events (Figure 6.17). There the tuning is
based on eighteen distinctive marker events and the In the foregoing examples, correlation between the different
similarity between the two isotopic records suggests a link land and ocean archives is based on alignment and, in some
between the East Asian monsoon and North Atlantic cases on tuning, of curves from proxy climate records.
atmospheric circulatory systems (Wang et al., 2001). Other Where correlations can be underpinned by independent
speleothem sequences that provide the potential for this dating, however, it may be possible to synchronize records
degree of close matching to the ice-core records include of past environmental change. This is important for not
those from cave sites in Europe (Genty et al., 2003), the only does it enable more secure comparisons to be effected
American Midwest (Serefiddin et al., 2004) and Israel between marine and terrestrial records, it also provides a
(Frumkin et al., 1999). means whereby leads and lags between different compo-
Correlations have also been made between the central nents of the earth–atmosphere system can be detected.
European loess–palaeosol record and Heinrich Events (Shi The principles and some potential problems of aligning
et al., 2003), while in the Malan Loess of central China, and synchronizing stratigraphic records are illustrated in
high-frequency fluctuations in dust influx include peaks Figure 6.18. Here, there are two sediment cores, one of
that appear to correlate with events in the North Atlantic which is from a lake sequence (core 1), while core 2 contains
(Porter, 2001). In addition, loess sequences have been a marine record. Six common marker horizons or tie-
correlated with the Greenland ice-core record. In the points (M1–M6) are shown; these might be Heinrich events
American Midwest, for example, a series of palaeosols in the marine core and their inferred terrestrial equivalents
(reflecting warmer episodes) have been correlated with the in the terrestrial lake sequence. The markers provide the
IS events in Greenland ice cores (Wang et al., 2003), while basis for aligning the proxy climate records from the two
a link has also been suggested between features of the cores (Figure 6.18a). However, in order to establish a
Chinese loess–palaeosol sequences and Greenland IS events precise time-stratigraphic correlation (synchronization)
(Huang et al., 2000). and to show that the alignment is valid, additional
-26 -9
17 14 13 12 10 a 0 4
21 15
20
7 6i
-26 7 Hulu
19
IS Cave
-30
-7
2
-32
-34
-5
H6 H5 H4 H • H2 H1 YD
-36
20 19 14 12 2
1716 7
15 .10 7 7 5 4
21 11 2
-36 18 13
7 GISP2
-3
-40
-42
-44
80 70 60 50 40 30 20 10 0
A g e (ka)
Figure 6.17 Correlation between the δ18O record from Hulu Cave, China, and the δ18O profile from the GISP2 Greenland ice
core. The numbers above the isotopic traces denote Greenland Interstadial (DO) intervals. The vertical bars mark the positions of
Heinrich events 1–6 in the two sequences; YD – Younger Dryas (after Wang et al., 2001 and Genty et al., 2003).
a) In this instance, therefore, the records cannot be syn-

M6 M5 M4 M3 M2 M1
chronized and, indeed, there may be doubts over the
Core 1 Top reliability of the proposed alignment.
One of the most effective approaches to the synchron-
ization of records from different depositional archives is
Core 2 Top
the use of event stratigraphy (section 6.3.2.5), such as that
M6 M5 M4 M3 M2 M1 proposed for the North Atlantic region by the International
Working Group known as INTIMATE (INTegration of
Ice-core, MArine and TErrestrial records). INTIMATE
b) initially outlined a North Atlantic Event Stratigraphy for
47+/- 7 k a 1B+/-0.9 k a
the Last Termination (15–8 ka) based on the annually
M6 T2M5 M4 P1 M 3 M 2 T1 M1 dated isotopic signal in the GRIP Greenland ice core (Björck
Core 1 Top
et al., 1998), but with the subsequent construction of
the high-resolution GICC05 timescale from the NorthGRIP
ice core (Rasmussen et al., 2006), NGRIP has now been
Core 2 Top adopted as the regional stratotype and the Event Strati-
M6 M5T2 M4 P1 M3 M2 rjM1 graphy extended back initially to 30 ka (Lowe et al., 2008b),
48+/- 5.5 k a
and subsequently to 48 ka (Blockley et al., 2012). The
18+/-0.9ka
INTIMATE group has recommended a series of protocols
M - marker horizon Mean and 1 S D range for synchronizing Last Termination records from the North
T- t e p h r a l a y e r of a g e e s t i m a t e Atlantic region, including the use of quantified proxy
P - p a l a e o m a g n e t i c feature
climate data from both marine and terrestrial contexts,
Figure 6.18 Schematic diagram showing the principles and secure 14C dating frameworks based on high-resolution
limitations of tuning palaeoenvironmental records from two sampling and age modelling, and tephrochronology, and
cores. For explanation see text. with inter-archive correlations established on the basis of
a high-quality 14C chronology and tephrochronology. This
chronological data are needed. In Figure 6.18b, as well as approach provides a secure time-stratigraphic framework
the six tie-points, there are two tephra horizons (T1 and for inter-regional comparisons as, for instance, in the
T2), a palaeomagnetic boundary (P1) and three radiometric synchronization of varved lake sediment records from sites
dates. T1 has been provenanced and securely dated in in north-central Europe which is based on a combination
another lake profile, and hence that age can be ‘imported’ of pollen stratigraphy and tephrochronology (Figure 6.19).
into both cores. By contrast, T2 has not been radio- As noted above, this approach is also important in palaeo-
metrically dated elsewhere, and the only available age environmental reconstruction as it may reveal temporal
estimates for this tephra are those that have been obtained variations in terrestrial and marine responses to changes in
from cores 1 and 2. The stratigraphical relationships the climate system. An example is provided by the detailed
between T1 and P1 and the marker horizons M1–M4 are pollen record and associated high-resolution 14C age model
consistent in the two cores, and the imported radiometric from Crystal Lake, Illinois, which shows that although
date with a small standard error provides a precise age for vegetational changes in the American Midwest were broadly
T1. The alignment of M1 with M4 is therefore underpinned coeval with millennial-scale trends in the NGRIP ice-core
by superposition, by two time-stratigraphic markers and by record over the period 17–11 ka, major shifts in vegetation
radiometric dating; hence the two records are effectively lag the NGRIP record by 300–400 years (Gonzalez &
synchronized. This is not the case with the alignment at M5 Grimm, 2009). This probably reflects the proximity of the
where the tephra T2 occurs above M5 in core 1 and below Laurentide ice sheet, which could have affected regional
M5 in core 2. Several explanations are possible for this: climate through its influence on atmospheric frontal
(1) M5 has been incorrectly identified or positioned within boundaries and the position of the jet stream.
either core 1 or 2 (or both); (2) M5 is diachronous between In the Southern Hemisphere, the New Zealand
the two records; or (3) different tephras, of different age, INTIMATE Group has also developed a climate event
with virtually identical chemical signatures occur in the stratigraphy for the past 30 ka (Alloway et al., 2007b). This
two cores. Note that the radiometric dates for T2 both is based on a series of well-dated, high-resolution onshore
have large standard errors, and as the age range brackets and offshore proxy records from a variety of latitudes and
M5 in both cores, they do not help in resolving the issue. elevations. These include long peat and lacustrine sediment
Muggesfelder S e e BelauerSee Plupsee Hamelsee
Susceptibilit Susceptibilit
S u s c e pSt ui bsicl iet p t i b i l i t
27m
Susceptibilit
Susceptibilit
VIII
10030 elm decline
VII
30m Susceptibilit
20m
VI
VI 17m
VII 37,3m VI
9218
c.
Susceptibilit
VI c
V 37,5m V
"25m b V b
V 3 18m Boreal
b
10030 Saksunarvatn
S u s cSeupstci beipl itti b i l i t S u s c e p t i b i l i t

a a
Susceptibilit
V c 10640
3 10640
33,41m- III IV c IV Y o u n gYeor u nDgr ye ar D r y a
33.60m II IV b
11560 Younger Drya
I a 11560 Younger Drya
III s
34,0m 39m III 39m"
Susceptibilit
30rrr b
II III
Younger Dryas
1156
s
S u s c e p tSi buisl ict e p t i bS
12700
b 1S7 Allerad
II L a a c h e r s e e Tephra
20m a
13325 Older Dryas
c
D b Balling
SusceptibS
21m
140 Oldest Dryas
535 Melandorf
14420
Pleniglacial
Figure 6.19 Synchronization of four Lateglacial and early mid-Holocene lake records from northern Germany, using a combination
of pollen analyses (key pollen zone boundaries such as the ‘Elm Decline’), tephra isochrones (Saksunarvatn Ash and Laacher See
Tephra) and varve chronology (after Litt et al., 2001).
accumulations, speleothem and marine-core records, and principal means of synchronizing the different stratigraphic
more fragmented glacial, fluvial and aeolian sequences. sequences (Lowe et al., 2008a). Further south in Antarctica,
The records are linked by a combination of high-precision older tephras, dating from MIS 5 and 6, have been detected
radiometric dating (radiocarbon, U-series, OSL) and in both ice and marine cores, and while the dating and
(with the exception of the speleothem) by well-dated provenancing of these tephras is currently uncertain, they
tephra horizons. Indeed, the widespread and numerous may in due course provide a basis for synchronizing
tephra marker beds (twenty-two in all) provide the Antarctic and marine records (Hillebrand et al., 2008).
Volcanic events marked by acidity or sulphate peaks, or
GPTS Standard o x y g e n isotope Mediterranean
by the presence of tephra particles, have also been used to
Susceptibilit
stratigraphy a n d calcareous sap rope I
Susceptibilit
Susceptibilit
synchronize different ice-core records. In Greenland, for plankton biostratigraphy layers
example, volcanic reference horizons were used to link 2.0 0.0 -20
the Holocene sections of the Dye-3, GRIP and NGRIP 51 <2)
cores for the construction of the GICC05 timescale (Vinther 51 52 (6)
0.1 55 55 S5(12)
et al., 2006), and also to synchronise proxy records from 6 3
6.5 S6 (16)
those ice cores spanning the later part of the last cold 0.2 7.1
7.5 73
S9 (22)
stage (Rasmussen et al., 2008). At the global scale, Green- 3 5L S' (26)
Susceptibilit
0.3 9.1
Susceptibilit
93 510 (30)
land and Antarctic ice-core records have been synchronized
using isotopic and trace gas data (Blunier et al., 2007). 0.4 f' 5 511 (38)
Millennial-scale variations in water isotope ratios (δ18O) 13 1 512 (44)

0.5
are evident in both NGRIP (Greenland) and EPICA 13 3 Sa (50)
(Antarctica) ice cores, and these can be coupled to the 0.6 is.|!-: Sb(56)
15.5
17 1
globally applicable methane (CH4) profiles to align 0.7 i5
"17 5
Greenland and Antarctic records, for example in the period 13 2
^19.3.
1
0.8 SI 1
of MIS 5 between 123 and 80 ka (Capron et al., 2010a). 21 3
21.5
Other approaches to linking Greenland and Antarctic ice 0.9 23
cores involve the use of cosmogenic 10Be at the Laschamp 25 S" (90)
1.0 27 Sc(92)
palaeomagnetic excursion around 41 ka (Raisbeck et al., J
2007), and bipolar volcanic markers, such as the acidity 1.1 51 Sd(ioo)
55
spikes in the cores that reflect the major Toba eruption 12
lb
around 74 ka (Svensson et al., 2012), evidence that is dis- 57
Susceptibilit
1.3 Susceptibilit v(122)

cussed in more detail in Chapter 7. SI
u (126)
On longer timescales, the sapropels of the Mediter- 45 1(130)
i.4 45 '(134)
ranean region offer a remarkable basis for correlating 47 s<138)
terrestrial and marine sequences, and for synchronizing 15 q(142)
51
these different records. As explained in Chapter 1 (note 6), 1.6 52 p(1521

55 0(156)
sapropels are layers rich in organic material which 57 n (150)
1.7 5!>
accumulate under anoxic conditions in ocean basins. In the 11(168)
65 •(172)
Mediterranean, sapropels have formed episodically from 1 8 e(l76)
Pliocene times onwards (Emeis et al., 2000), and are found C 65 C5(17B
C3(182)
in both marine cores and in uplifted marine sequences, 1.9 M 67 C2 (186)
•J<J.
particularly around the coastlines of Italy and Sicily, where CO (190)
2.0 71
they occur in repetitive cycles of organic and inorganic
Susceptibilit
2.1 79 B7 (204)
sediment accumulations (Figure 1.3). The sapropels are G SI B5 (208)
believed to reflect near-synchronous oceanographical 83 B4(212)
2.2
changes across the Mediterranean region, and have been BX (218)
related to increased run-off or river discharge, and 2.3 B1 (222)
associated increases in nutrient input during periods of 2.4 35

higher precipitation (Cane et al., 2002). These wetter 27
episodes coincide closely with minima in the Milankovitch 2 5
4.0 3.0
precessional index (c. 21 ka), which allows the MPRS 2.6 B5 (250)
(Mediterranean precession-related sapropels) sequence

to be tuned to the astronomical timescale (Lourens, 2004). Figure 6.20 Stratigraphic framework for Quaternary marine
The sapropel layers can be correlated with microfaunal and sequences in the Mediterranean, showing regionally iso-
microfloral biozones, and with the marine δ18O signal, chronous sapropel units, foraminiferal biozones and magneto-
zones, linked to the astronomical timescale. The three principal
while further tie-points are provided by radiometric dating magnetic excursions of the Matuyama chron are indicated:
of younger parts of the sequence, and magnetostratigraphy J – Jaramillo; CM – Cob Mountain; G – Gilse (after Massari
on earlier parts of the record (Figure 6.20). These unique et al., 2004).
stratigraphic sequences therefore provide a basis for ocean–
ocean, and land–ocean correlation and synchronization data from chrono- and climatostratigraphy. Ultimately,
within the Mediterranean region throughout the entire geologic-climatic units will be established, and these form
Quaternary period and beyond. the basis for correlation at the regional and continental
scales. Geologic-climatic units are also the basis for cor-
relating marine and terrestrial sequences, while the essential
6.4 CONCLUSIONS reference standard for global correlation at Milankovitch
There is now a broad consensus that subdivision of the frequencies is the oxygen isotope stratigraphy of the
Quaternary stratigraphic record should follow as closely deep-ocean sediments. At sub-Milankovitch frequencies,
as possible conventional geological procedures, although well-defined events in marine, terrestrial and ice-core
the higher levels of resolution and analytical precision records provide a basis for correlation, with the event
that are required in the classification of the most recent part stratigraphic approach offering perhaps the best prospects
of the geological record mean that a strict adherence to these for establishing time-stratigraphic correlation between
principles may pose particular problems for the Quaternary these different archives. Secure correlations are an essential
stratigrapher. Nevertheless, it is generally acknowledged prerequisite for reconstructing spatial and temporal
that, wherever possible, lithostratigraphy should comprise patterns of climate change, and for understanding the
the basic building blocks of the Quaternary terrestrial forcing factors that drive those changes, and it is to these
stratigraphic sequence, augmented by empirical evidence aspects of the Quaternary record that we turn our attention
from bio- and morphostratigraphy along with inferential in the final chapter.
7
CHAPTER SEVEN
Global environmental
change during the
Quaternary
7.1 INTRODUCTION provided the first high-resolution records of climatic

changes during the last glacial cycle (section 3.11.1), and
Thus far, we have considered the different types of evidence partly because North Atlantic Ocean circulation was
that can be used to reconstruct Quaternary environ- considered by many to be the principal driver behind these
ments, the means by which a timescale for environmental changes. Since 1997, however, data from other regions of
change can be established, and the stratigraphic procedures the world have begun to reveal the true scale and complexity
that enable sedimentary records to be interpreted and of Quaternary environmental changes, not only during
meaningful correlations to be effected between often widely the past 130 ka, but also over earlier glacial–interglacial
scattered localities. In this final chapter, we will show cycles. The evidence shows that while the North Atlantic
how results obtained using these different methods and does indeed play an important role in the world’s ocean
approaches can be synthesized to produce an overview of circulation regime and has a profound effect on Northern
global environmental change and, in particular, to reveal Hemisphere climate, it is but one component of a highly
the forcing factors behind these changes. Data compilation complex pan-global ocean–climate system. To understand
and syntheses at regional, continental or, indeed, global that system and its role in Quaternary climatic and envir-
scales are important aspects of contemporary Quaternary onmental change, it is now clear that a global perspective
science for, in addition to providing snapshots of past is necessary, and it is such an overview that this final
environmental conditions, they offer a means of cross- chapter aims to provide. We freely acknowledge, however,
checking evidence from different proxy sources, they focus that this is a daunting task and is worthy of a book in its
attention on the linkages between processes and com- own right. Hence, we will not be able to cover all aspects
ponents of the global environmental system (e.g. between of the subject; nor will we have the space to go into the level
the glacial, oceanic and terrestrial realms) and they may of detail that the topic really demands. As a consequence,
provide new insights into the causes of environmental we have opted to focus the discussion on what we consider
changes. All of these aspects were exemplified by the revo- to be some of the key questions about the nature, causes
lutionary CLIMAP and COHMAP projects of the 1970s and and consequences of global environmental change, and
1980s (section 4.10.7), and they underpin more recent to show how these issues are central to many areas of
research initiatives, some of which are considered in this contemporary Quaternary research.
chapter. A recurrent theme throughout this book has been the
In the second edition of this book (published in 1997), fundamental imprint on the Quaternary environmental
we examined these various interlocking themes by focusing record of climate change, operating at Milankovitch,
on the interval from c. 130–10 ka, the last interglacial– sub-Milankovitch and much shorter (decadal) timescales.
glacial cycle of the Quaternary record, and we centred our It should not be assumed, however, that climate is always
discussions on the North Atlantic. We chose this region the prime instigator of change, with other environmental
and this time period partly because the evidence obtained processes (soil development or erosion, glacier expansion
from the Greenland ice-core records in the early 1990s had or contraction, changes in atmospheric gas content, etc.)
380 GLOBAL ENVIRONMENTAL CHANGE
being mere responders. The climate system is itself reactive The sensitivity of a model to changes in particular
to precursor and contemporaneous environmental condi- parameters can then be tested, and their relative importance
tions: one of the main challenges confronting Quaternary deduced. Because ESMs are now so widely employed in
scientists, therefore, is that of the ‘chicken-and-egg’ Quaternary research, we preface our discussion of
problem, of identifying within continuous and some- timescales of change with a consideration of the bases and
times cyclic proxy environmental records those factors capabilities of ESMs, and of their importance in the study
that are causally connected, and those that are initiators of of land–ice–ocean–atmosphere interactions over the course
change. To meet that challenge, it is important to define the of the Quaternary record.
precise order and timing of key signals in the proxy data,
as these may point to leads and lags between environmental
variables, distinguishing forcing factors from responders.
7.2.2 Box models
The degree to which leads and lags can be clearly resolved, The simplest ESMs are box models that reduce com-
however, in part depends on the temporal resolution at plex components of the global environmental system (e.g.
which the stratigraphical record can be examined. It also the global ocean, atmosphere or polar ice) to artificial
depends on the quality of the proxy data available, and ‘reservoirs’, within which conditions are homogenized;
on the nature of the environmental processes inferred as the reservoirs are connected by ‘pipes’ that control
these operate over a range of timescales. We illustrate the mass transfer between the reservoirs (see McGuffie &
problems of defining leads and lags, and of establishing Henderson-Sellars, 2005). The principal properties of each
causal relationships between proxy indicators, by consid- reservoir (e.g. volume, density, salinity and temperature
ering the evidence at four temporal scales: the Milankovitch changes) can be varied in order to see the effect that each
timescale (section 7.3), the sub-Milankovitch (centennial change imposes on the operation of the model. These
to millennial) timescale (section 7.4), the sub-centennial models are not meant to represent the complexity of the
timescale, for which we focus on evidence from the Last real environmental system; rather, they are deliberately
Termination (section 7.5), and finally at a decadal to annual simplified to allow the operator to study the feedback links
resolution, with a particular emphasis on the Holocene between selected components of the system (Claussen et al.,
(section 7.6). 2002). Figure 7.1, for example, shows a schematic of the
elements of a box model that represents (proportionately)
7.2 ENVIRONMENTAL SIMULATION the global atmosphere, polar ice, sea ice and ocean.
When the model was run, the output generated a 100 ka
MODELS (ESMs) climatic cycle that is self-sustaining (i.e. it works without
any external forcing), and the key driver was found to be
7.2.1 Introduction the extent of sea-ice cover in the polar oceans, which
While the geological record can reveal the precise order of regulates the growth and decay of polar ice sheets (Gildor
events or environmental changes this, by itself of course, & Tziperman, 2001; Gildor et al., 2002). The degree to
does not demonstrate causality. In the experimental which this reflects reality can be explored through a
sciences, causality can be established by controlled manipu- combination of further modelling runs calibrated against
lation of selected properties or processes and measure- empirical data. Other simple box model experiments
ment of their effects, an approach that is not possible with have been used to explore ideas about the operation
geological evidence. Instead, causality in geology is inferred and impacts of the North Atlantic limb of the global
from observations of repeated associations of, or sequen- thermohaline circulation (Rahmstorf, 1996). They were also
tial changes in, proxy variables that are common to a used for some of the earliest ‘freshwater hosing experi-
number of stratigraphical records. A causal link is diffi- ments’, which simulate the injection of glacial meltwaters
cult to prove, however, from such observations alone, into the ocean in order to assess their climatic signifi-
since it may arise by chance or may reflect the operation cance (Kageyama et al., 2010). Over longer timescales, box
of other factors or processes not yet detected. An alternative models have been used, inter alia, to simulate bottom
approach is the simulation of natural environmental water salinity changes in the Southern Ocean, the outputs
processes by numerical modelling. The construction of suggesting that these processes could have played a key role
environmental simulation models (ESMs) enables inputs in modulating the rhythm of climate change during the
to be varied in a controlled manner and ‘what-if?’ experi- Quaternary, and particularly the change from the 41 ka
ments to be conducted by exaggerating or suppressing to 100 ka cycle around 900–800 ka (sections 1.7 and 7.3;
selected components, or by omitting or replacing variables. Paillard & Parrenin, 2004).
GLOBAL ENVIRONMENTAL CHANGE 381
b) Meridional extent of land ice

a} in the Northern polar box
0 4
0.2
0
300 200 100 0
A t m o s p h e r i c temperature in the northern

polar box (C)
0
-10'
-20
300 200 100 0
HP
45S Meridional extent of s e a ice
in the Northern polar box
0.9
45S
SP 0.6
0.3
0
300 200 100 0
A c c u m u l a t i o n a n d ablation of ice in
the Northern polar box (Sv)
0.12
0.09
006
003
300 200 100 0
Figure 7.1 a) Schematic representation of a coupled box model showing atmospheric boxes (upper slab), thermohaline
circulation (black arrows), parabolic land ice sheets (grey semi-cylinders), and a partial sea-ice cover (grey slabs) in the polar ocean
boxes. All of these components are allowed to vary in model equations. Ocean biogeochemistry and atmospheric CO2 interact
with physical climate components. NP – North Pole; SP – South Pole; 45°N and 45°S indicated (from Gildor & Tziperman, 2001).
b) Box model results simulating the 100 ka Milankovitch climatic cycle. The horizontal axis is model run-time in simulated ka (from
Gildor et al., 2002). For further explanation see text.
The advantage of box models is that they are com- attention is increasingly being directed towards more
putationally simple, and hence results can be obtained sophisticated models, which we now move on to consider.
quickly. They are useful, therefore, for testing the efficacy
of selected global environmental feedback loops, with the
results feeding into the design of more complex models,
7.2.3 General circulation models (GCMs)
often leading to considerable saving of time and resources. Some of the most advanced environmental models yet
The principal disadvantage of box models, however, is developed are those that simulate the global climate system,
that they tend to underestimate the complexity of global and have been generated primarily as a response to the
systems and their processes. In nature, the oceans and demand for improved weather forecasting and climate
atmosphere are dynamic systems with complicated zonal prediction capabilities (Randall et al., 2007). They are
and non-linear flow mechanisms, which in the reservoirs widely referred to as general circulation models (GCMs)
and slabs of box models are reduced to uniform mass but technically, perhaps, should be called climatic general
properties. Box models also tend to isolate elements of circulation models, to distinguish them from those that
the global environmental system from other feedback incorporate simulations of other environmental processes
mechanisms that could modulate their operation. As a (sections 7.2.4 and 7.2.5). GCMs are designed to simulate
consequence, therefore, box models may simplify natural the complex three-dimensional structure and dynamic
processes to the extent that they can deliver mislead- flows of the earth’s atmosphere and oceans. The input data
ing outputs (Marotzke, 2000; Wunsch, 2010), and thus fall into two categories. The first are boundary conditions,
which are the prescribed surface values for such physical In the
parameters as sea-surface temperatures (SSTs), surface Atmospheric
albedo, radiation receipt, atmospheric transparency, sea-ice Column
Vertical e x c h a n g e
cover and topography. For simulations of modern climate, between levels / Wind vectors
Humidity
the input values are based on direct measurements of
Clouds
current physical parameters, while boundary conditions Temperature
that prevailed during earlier periods are based on estimates Height
derived from proxy (geological) data. Dynamic conditions
(fluxes or flows) can then be added by parameterization of
At the Surface i
surface and ocean processes. For the atmosphere, these Ground temperature Horizontal e x c h a n g e
might include heat and moisture exchange gradients water and energy \ between c o l u m n s
fluxes^ '
between surfaces, moisture convection, Coriolis and shear
constants, atmospheric pressure equilibria, and cloud
formation and effects; and for the oceans, salinity flux,
surface and subsurface currents, upwelling and overturn-
ing, eddy interference and density mixing (Randall et al.,
2007).
GCMs vary in their construction in three main ways.
First, the earth’s surface is represented by a geometrical
Cartesian grid, the size of the grid cells depending upon the
type and amount of information to be computed, and the
capabilities of the computer employed (Figure 7.2). The
grid in early GCMs varied between 4° × 5° and 11.5° ×
11.25°, but recent advances in computer capabilities and Time step - 3 0 minutes G r i d s p a c i n g - 3 ° x 3°
computational algorithms have allowed cell sizes of

1° × 1° or less to be employed. Second, the atmosphere and Figure 7.2 Cartesian grid arrangement for a GCM using a six-
layered atmosphere and a grid cell size of 13° × 13° (latitude
oceans can be represented as a series of layers (or slabs), and longitude). Physical parameters (wind vectors, humidity,
enabling both horizontal and vertical flows to be calculated clouds, etc.) are assigned to each grid block, and horizontal and
(Figure 7.2). Third, the number of properties measured for vertical flows between adjacent blocks also have to be
each grid cell and the type of algorithms used for dynamic parameterized for each interface (i.e. for six interfaces for
each cube within the stack).
flow calculations also vary between GCMs, and these
inevitably affect model performance.
The most sophisticated GCMs operating at the present
time have been developed at specialist research centres develop a ‘Fast Ocean’ GCM, with improved representation
sponsored by national organizations with research budgets of ocean gyres (Jones, 2003).
large enough to meet the heavy technical and resource The performance of individual GCMs can be tested by
demands. Examples are those developed by NASA-GISS comparing the extent to which they simulate modern
(Goddard Institute of Space Studies) and NOAA (National weather patterns and conditions (Phillips et al., 2004),
Oceanographic and Atmospheric Administration) in or longer-term climate trends for which instrumentally
the USA, the Hadley Centre for Climate Prediction and measured data are available (Randall et al., 2007). The
Research of the Meteorological Office in the UK, the latter approach is shown in Figure 7.3, where the model
Meteorological Research Institute in Japan, the Bjerknes simulations tend to track the long-term instrumental
Centre for Climate Research in Norway and the Max Planck temperature record, and the majority also pick out
Institute for Meteorology in Germany. The models have four short-term climatic downturns that immediately
been developed independently, although there is regular followed significant volcanic eruptions. The scatter in the
collaboration and cross-checking between research teams, model outputs (the ‘noise’ in the data) reflects differ-
and sharing of technical developments (e.g. Covey et al., ences of design between each GCM, for example Cartesian
2003; Jacob et al., 2007). Over time, a degree of special- grid cell size, number of atmospheric and oceanic layers
ization has developed, with different teams focusing on involved, and the computational algorithms employed.
improving the performance of specific components of the Inevitably, all models are characterized by a degree of
climate system, such as the UK Hadley Centre’s project to uncertainty, since even the most powerful of computers
1 0
b si l ci te p t i b i lSi tu s c e p t i b i l i t 0 5
0 0
tiu
S u s c e pS
-0.5
Pinatubo
Santa Maria Agung El C h i c h o n

-1.0
1900 1920 1940 1960 1980 2000
Year
Figure 7.3 Comparison of fifty-eight individual GCM simulations of global temperature change since AD 1950 (thin light-grey curves,
with mean trend shown in thicker, dark-grey line) compared with instrumentally measured temperature variations over the same
period (black curve). The data are shown as deviations (anomalies) from the 1901–1950 mean value. Vertical grey lines indicate
the timing of four major volcanic eruptions that occurred during the period (from Randall et al., 2007). For further explanation
see text.
cannot adequately represent the true scale and complexity data; glacier ice and sea-ice cover from glacial geomorpho-
of the global climate system. logical and palaeoceanographic data; and atmospheric
Climate GCMs have generally been tuned to simulate aerosol content from ice-core records. Orbitally induced
modern climatic conditions especially, as noted above, in solar radiation receipt can be calculated from radiation
climatology and weather forecasting. Of particular interest tables. Figure 7.4 shows, in a schematic way, how these
to the Quaternary scientist, however, are environmental boundary conditions may have varied at a global scale over
simulations for times in the past when the earth’s surface the last 18 k 14C yr (Kutzbach & Webb, 1993). The figure
boundary conditions were markedly different, and the also illustrates the ways in which the earth’s surface can be
information that such experiments can provide about depicted in contrasting states or ‘modes’: (1) a ‘glacial
the factors driving long-term climate change. In order to mode’ at c. 18 k 14C yrs BP (the Last Glacial Maximum:
address these issues fully, however, additional environ- LGM), when the earth’s orbital configuration was similar
mental parameters need to be considered, as is explained to that of today, but ice cover was much greater, sea level
in the following section. lower, and SSTs well below those of the present; (2) a period
of enhanced seasonality around 11–10 k 14C yrs BP; (3) a
period when ice cover was significantly reduced and sea
7.2.4 Earth system models of intermediate
levels were close to those of the present day, but when
complexity (EMICs) seasonal differences remained high (at c. 6 k 14C yrs BP);
Attempts to model the global climate system under different and (4) the present-day situation. The figure emphasizes
boundary conditions have tended to focus on the last 18 ka, how the climate system is constantly adapting to transient
the period for which the appropriate proxy data are most boundary conditions, which are very difficult to simulate:
abundant (e.g. COHMAP Members, 1988). Atmospheric indeed, it has taken several decades to develop GCMs to the
temperatures over land areas can be reconstructed using point where they can successfully model the modern
palaeoecological evidence (botanical records, coleopteran climatic situation, and it is likely to take another step-change
records, etc.); SSTs from marine micropalaeontological in computer power, as well as greatly increased operator
18 15 12 9 6 3 Ok COHMAP project highlighted include (1) the close cor-

respondence between enhanced summer insolation and the
S u s c eSputsi bc iel ipt t i b i l i t 8 strength of the monsoon cells in the Northern Hemisphere
100 (Kutzbach & Guetter, 1986); (2) the deflection (or splitting)
Susceptibilit
u tsi cbei lpi tt i b i l i t

Aeroso
of the northern jet stream by the build-up of the Lauren-
4
50 tide ice sheet (Figure 7.5), leading to marked changes in
regional moisture distribution throughout the Northern
ep
S u s c eSputsi bc iel ipt t iSbui lsi tc S
330 0 0 Hemisphere (Kutzbach et al., 1993); (3) the downstream
modification of climate in Europe caused by sea-surface
Susceptibilit
-1
Susceptibilit
-2 temperature changes in the North Atlantic (Rind et al.,

265 -4
-3 1986); and (4) the delay (by some 2–2.5 ka) of the Northern
-4 Hemisphere interglacial thermal maximum following
Aeroso
200 -5 -8 the summer insolation peak at c. 10 k 14C yrs BP (Figure 7.4)
as a consequence of the delayed melting of the great
18 15 12 9 6 3 Ok northern ice sheets (Webb et al., 1993).
Aerosol SST Ice co 2 Although the situation has changed considerably since
the COHMAP experiments, particularly as more compre-
Figure 7.4 Boundary conditions for the COHMAP simulation hensive palaeo datasets have become available and models
for the last 18 k 14C yrs BP. External forcing is shown for of much higher spatial resolution have been developed,
Northern Hemisphere solar radiation in June–August (SJJA) and
December–February (SDJF) as the percentage difference from two major problems remain. First, it is still not possible to
present-day radiation receipts. Internal boundary conditions generate circulation models for times in the past that are
include land ice as a percentage of 18 kyr ice volume; global as sophisticated as the GCMs employed to simulate the
mean SSTs are expressed as a difference from present-day modern climate, since quantified values for past boundary
SSTs. The excess glacial-age atmospheric aerosol content is
conditions (inferred from proxy data) have varying degrees
shown on an arbitrary scale, while atmospheric CO2 concen-
tration is shown in parts per million by volume (ppmv). The of uncertainty (as discussed in Chapters 2–4), and there are
horizontal scale shows the time interval of seven sets of ‘time- still many areas of the world from which the requisite
slice’ simulation experiments conducted by COHMAP (after high-resolution data are simply not available. Second, the
Kutzbach & Webb, 1993). climate system is affected by other global circulation
processes, not routinely incorporated into climate GCMs,
for example the regulation of atmospheric gas content by
time, before GCMs will be capable of simulating moving the biogeochemical cycle. The amount of CO2 and CH4
boundary conditions at the global scale (see section 7.2.4). present in the atmosphere at any one time reflects a balance
In order to circumvent this problem, a time-slice between, on the one hand, release of these gases through a
approach has often been adopted, whereby the prevailing variety of processes and routeways (from terrestrial
(static) global climate state is simulated for specified time vegetation, disturbed peatlands, wildfires, permafrost
periods or intervals, effectively the approach that was melting, etc.) and, on the other, the amount that is absorbed
adopted by COHMAP (Figure 7.5). Such models employed into (and effectively trapped in) various reservoirs, for
fixed SSTs, which is clearly unrealistic since the real oceans example the ocean (carbon sinks). A further influence is the
and atmosphere interact in a dynamic and mutually amount and distribution of glacial and sea ice at any
responsive way. They also adopted a relatively coarse specific time, as this affects albedo. Other surface features
grid size without any topographic detail, did not build that influence global climate include the extent and type of
in changes in global sea level, vegetation cover or ocean terrestrial biomes, ocean biomass, distribution and
circulation, and the amount and quality of palaeodata thickness of permafrost, ocean circulation and sea level. In
available was far less than is the case today. Nevertheless, addition, the climate system may experience a degree of
these models represented a significant breakthrough in inertia, with delayed response to effects inherited from these
Quaternary science, for they brought sharply into focus long-term processes.
the scale and complexity of the changes in boundary The need to capture more of these processes and
conditions that typically occur during the course of a influences has led to the development of a new set of
glacial–interglacial cycle and which would, in turn, have advanced models, which are referred to as earth system
modulated the various feedbacks within the global environ- models (Figure 7.6). These attempt to identify and simulate
mental system. Examples of important linkages that the as many as possible of both the short- and long-term
Stronger
Aleutian low
colder,
glacial drier than
anticyclone present
(July) i
weaker glacial
subtropical anticyclone
high colder,
drier than
stronaer present
flow - colder
(Jan) than i
Cooler, moister very strong
present
than present 18k
oolder
than stronqer flow
present . (July)
•'weaker-
subtropical anticyclon
anticyclon
high glacial (Jan j
<Julv)- anticyclone colder,
(July) drier than
present
colder than
present
very strong
westerly flow 12k
warmer,
drier warmer, drier
strong er> \ than than present
subtropical present I.July!
high Duly)
(July) stronger flow
warmer, drier
(Jan) , .
than present
(July) j glacial j
coole* anticyclone
.(July) v
Lhan present
stronger ' warmer tharT"

onshore high present (July)
(July) 9k
stronger
subtropical
high
(July)
warmer, drier
than present
stronger
(July) •
flow
stronger flow
. (July) (July)
warmer, drier
than present
(July)
stronger flow (July) Gk
subtropical
high (July) i
westerly
surface
winds
January
jet stream
southerly surface winds (July) Present
Figure 7.5 Palaeoclimatic model simulations for the last 18 k 14C yrs BP (after Kutzbach et al., 1993).
processes acting on the atmosphere and oceans (Flato, in the USA (Prinn et al., 2011). A second example is the
2011). However, because it is still not feasible to build into GENIE initiative (Grid ENabled Integrated Earth system),
the models all of the environmental parameters that exert developed in the UK for experiments designed to identify
an influence on the atmosphere, ‘sub-models’ have been the ‘tipping points’ that affect Atlantic thermohaline
constructed, which include a degree of simplification in circulation; in other words, those critical conditions that
terms of the inputs required for the simulation of long-term can tip the system into instability (Lenton et al., 2009).
environmental feedback processes (Prinn, 2011). These Because EMICs can simulate changes over long time-
‘reduced’ models, intermediate in terms of computer power scales, including the 100 ka cycles (Weber, 2010), they
and complexity between conceptual box models on the are becoming increasingly important in the study of
one hand and top-of-the-range climate GCMs on the other, Quaternary environmental change, as is illustrated in later
are termed earth system models of intermediate com- sections.
plexity (EMICs). Since the computations performed are
simpler than in GCMs, EMICs are able to generate results
reasonably quickly, while several of these lower-order
7.2.5 Transient simulations
models can be coupled to each other or to GCMs in order A further difficulty with full-scale GCMs is that they are
to test the ways in which different subsystems integrate and most often used to simulate the climate at equilibrium, and
feed in to the climate system. Figure 7.6 shows a schematic the results therefore provide ‘snap-shots’ of the climate state
of the type of integrated earth system model array being at particular points in time. In order to examine change in
developed by the Massachusetts Institute of Technology the system, new boundary conditions and feedback fluxes
Human Activity (EPPA)

Human
N a t i o n a l andfor R e g i o n a l E c o n o m i c
health
Development, Emissions, Land Use
effects
C0 , CH , 2 4
Trace gas
Agriculture, C O , N2O
fluxes
forestry, Climate/ NOx, S O x Sea Land
(C0 , CH ,
bio-energy, NH , CFCs.
2 4
energy 3 level use
N2O) a n d
ecosystem demand HFCs, PFCs, change change
policy
productivity SF , VOCs,
4
constraints
B C , etc.
Hydrology/
water
resources
E x a m p l e s of
Model Outputs
EARTH SYSTEM G D P growth,

energy use,
policy c o s t s ,
Atmosphere Urban
agriculture a n d
2-Dimensional Chemical A i r Pollution
Solar
& Dynamical P r o c e s s e s Processes health i m p a c t s -
Forcing
Global mean
a n d latitudinal
temperature and
Coupled Ocean - Atmosphere & Land
precipitation,
s e a level rise...
Ocean Land
Volcanic 2 - o r 3- D i m e n s i o n a l Water & Energy
Permafrost area,
Forcing Dynamics, Biological, Budgets (CUM)
vegetative a n d
C h e m i c a l a n d Ice Biogeochemical Processes
soil cartoon. T r a c e
P r o c e s s e s (MITgom) (TEM & NEM)
g a s e m i s s i o n s from
ecosystems...
Figure 7.6 Feedback framework for the Integrated Global System Model of MIT (Massachusetts Institute of Technology, USA;
http://mit.edu/globalchange).
must be prescribed, and the model set to run until it spanning a specific time interval, for example from the
achieves equilibrium with the new controls. In other words, LGM to the present, boundary conditions for selected
the model effectively ‘jumps’ from one climate state to the times can be inferred and fed into the ESM. If the changes
next, unlike the real climate system which is continually in boundary conditions for the early stages only are fed into
responding to a range of forcing factors and feedback the model, then it can be set to run with these prescribed
loops, some of which change gradually, while others may boundary conditions, and allowed to continue to simulate
cross thresholds that trigger more sudden adjustments how the climate might evolve at a later stage. The model
(tipping points). Of particular interest to Quaternary outputs can then be compared with independent empirical
scientists are the transitions between climate states, for data for the later period in order to determine how well they
example those that characterize the last glacial cycle (DO match. Because the ‘predicted’ (forecast) outcome in this
cycles), and the factors that initiate a change from one instance is for some time in the past, this procedure is often
climatic episode to another. Current GCMs are not able referred to as ‘post-diction’ (or ‘hind-casting’). This is
to simulate these non-equilibrated (transient) climatic extremely important in Quaternary research, for it is the
shifts, due to excessive computational demand (Timm & only way in which ESMs (including GCMs) can be tested
Timmermann, 2007). But by improving the design and for their ability to predict climate states not experienced
performance of EMICs, which have fewer climatic com- within the period of instrumental records, which at most
ponents but include computations for geological variables extend over the past 150–200 years. However, the data have
such as ice-sheet variations, ocean circulation or volcanic to satisfy a number of criteria if they are going to deliver
forcing, transient models can be constructed which are reliable information that can be used for this purpose
capable of responding to continuously adjusting vari- (Kohfeld & Harrison, 2000). First, they should be con-
ables and which can deliver transient accelerated palaeo- tinuous or near-continuous over the period concerned,
simulations (Lunt et al., 2006). These require specialized although small gaps can be filled by data interpolation
sub-models representing different components of the techniques. Second, they should be reliably dated, to ensure
environmental system, which are coupled together in order that the boundary conditions prescribed for specific
to test their interactions. One example is the EC–Bilt–CLIO times are derived from contemporaneous evidence. Third,
model (http://csdms.colorado.edu/wiki/Model:ECBILT- they should be spatially extensive, to discriminate local
CLIO), which consists of three coupled sub-models anomalies from more general changes. Fourth, the former
simulating the global atmosphere, ocean dynamism and environmental conditions should preferably be quantified
sea-ice variations (Renssen et al., 2007b). This transient and in a form that can be directly compared with model
model features in a range of model experiments designed output data (e.g. seasonal or annual mean temperatures,
to test the operation and regional impacts of thermohaline precipitation or moisture change, etc.). Fifth, realistic
circulation (Timmermann et al., 2005). EC–Bilt–CLIO statistical uncertainties should be integral to the data
can also be coupled with other sub-models, for example (Kohfeld & Harrison, 2000).
those that simulate changing vegetation cover, to include Insights into Late Quaternary climates from palaeodata-
additional feedbacks in palaeoenvironmental experiments model comparisons, therefore, require a close dialogue
(Renssen et al., 2005). Further examples of the EMIC arrays between climate modellers on the one hand, and Quatern-
used in transient models and an assessment of their ary scientists who collect and interpret proxy data on the
simulation performance and limitations can be found in other. The models are, of course, only as secure as the
Petoukhov et al. (2005). data upon which they are based, and against which they are
ultimately to be tested. The onus therefore falls squarely
on the Quaternary community to provide the necessary
7.2.6 Palaeodata-model comparisons high-quality, high-resolution proxy records to enable
Palaeoenvironmental proxy data (commonly shortened climatic modelling to begin to provide answers to ques-
to ‘palaeodata’) make important contributions to the tions, not only about the pattern of past climatic changes,
development of ESMs in two main ways: first, in providing but also about their causes. To meet these objectives,
estimates of the boundary conditions that prevailed in global palaeo-datasets are now usually assembled through
former times, and second, in evaluating the results of ESM collaborative (community) research initiatives that are
experiments that are designed to ‘predict’ (or ‘retrodict’) subject to strict quality control measures, with the resulting
past climate states. The former application was discussed archives being held in publicly accessible data-centres.
above, while the latter works in the following way. If a These include, for example, DIRTMAP (Dust Indicators
number of site records provide continuous palaeodata and Records from Terrestrial and Marine Palaeoenviron-
ments), an integrated record of variations in quantified the greater the number of variables, the greater the
rates of aeolian dust accumulation (Bullard, 2010), the complexity in terms of links between different sub-
Global Lake Status Data-base, which collates information systems. This almost inevitably means that the
on past variations in lake level (Qin et al., 1998; section outcomes are less predictable (Schellnhuber, 1999).
3.7.3), and BIOME-6000, a global synthesis of palaeo- 4. Only very small changes in one or more parameters can
vegetation records based mainly on pollen data which are have significant effects on model output. Running the
classified into a series of different biome types, an initiative same model twice using exactly the same parameters
established primarily to test ESM simulations (Wu et al., and boundary conditions can generate quite different
2007). Global databases are also being assembled for marine outputs (Sivakumar, 2004), which may help explain
records, for example the multiproxy MARGO database why GCMs often tend to produce different predictions
referred to in section 4.10.7. There are also major inter- of future climatic scenarios.
disciplinary initiatives that foster closer collaboration
between scientists who collect palaeodata and modellers Despite these limitations, however, if an element of
interested in performing ‘hindcasting’ experiments, such as chaotic behaviour is involved in all natural processes, then,
PMIP (the Paleoclimate Model Intercomparison Project), paradoxically perhaps, the only viable way forward in the
which focuses on experimental design and capabilities search for an understanding of the global climate system,
(Weber et al., 2007). and predicting its behaviour, is through numerical model-
ling. If this challenge is to be met, future generations of
numerical models must become even more sophisticated.
7.2.7 Limitations of ESMs
They will have to incorporate a much wider range of
Although numerical modelling has provided important new variables and linkages to reflect more realistically the
insights into the workings of the global environmental complexity of the global environmental system.
system, even the most sophisticated of models continue to
have limitations. Some of the remaining areas of concern
7.2.8 The importance of ESMs in
are the following:
Quaternary research
1. Models invariably (and inevitably) oversimplify Despite their limitations, ESMs of all categories are playing
environmental complexity for, as we have seen, they an increasingly important role in Quaternary science.
cannot represent every component and synergistic They provide a means of synthesizing a wide range of
interaction that occurs in nature. New discoveries environmental information at the global scale, of quan-
continue to show how small-scale processes, such as tifying the major processes influencing global climate, and
feedback mechanisms involving cloud formation, for of conceptualizing the cause-and-effect pathways involved
example, can have far bigger impacts on climate (and in climate change, and their environmental impacts. ESMs
also model output) than might have been anticipated have also focused attention on the important time lags that
(McGuffie & Henderson-Sellars, 2005). appear to have occurred between climate-forcing signals on
2. Because numerical models can only handle a finite the one hand, and the response of climatic circulation and
number of calculations, all the global processes and other environmental processes on the other, examples of
interactions represented in the models must be scaled which are discussed in later sections. Palaeodata-modelling
down in order to be manageable, and this inevit- collaborations are critical for testing and improving ESMs,
ably leads to a difficulty in scaling-up when it comes and hence for reducing uncertainties in their results, which
to interpreting the output. This is the reverse-logic is one of the key issues highlighted by the IPCC (Inter-
problem of knowing whether or not the scaled- governmental Panel on Climate Change) in its latest reports
down, almost artificial, microcosm created in the (Schmidt, 2010). The threat of accelerating anthropo-
model truly reflects the operation of the system at genically induced climate change, with its environmental,
the global scale. economic, political and societal ramifications, and the
3. Interactions between global subsystems (the atmos- need for appropriate remedies and policies, places an
phere, hydrosphere, cryosphere, biosphere, geosphere) increasing responsibility on environmental scientists to
are invariably non-linear. Hence, while it may be provide a realistic context within which future climatic
possible to predict responses within a simplified scenarios can be assessed. Quaternary science has a key role
subsystem when only a few variables are involved, to play in this regard, for the Quaternary stratigraphic
record provides the all-important data archive against of data-smoothing and tuning of the results to the
which climate models can be tested and evaluated. With this Milankovitch timescale, a procedure frequently adopted in
in mind, we now turn our attention to the record of the the analysis of isotopic profiles from deep-sea sediments,
past, and begin with environmental change over longer but which may conceal differences in detail between records
timescales. by assuming, rather than testing for, contemporaneity
between inferred climatic events (Blaauw, 2012; Berger
7.3 CLIMATIC CHANGE OVER 2013a, 2013b; section 6.3.3.3).
Accounting for the amplitude and frequency of the
MILANKOVITCH TIMESCALES
climate signal in Quaternary stratigraphic sequences in
terms of the Milankovitch rhythms is also often difficult.
7.3.1 Introduction
For example, while there may be strong empirical evidence
Although there is a general consensus that the precessional pointing to the influence of precession and obliquity in
(c. 21 ka), obliquity (c. 41 ka) and eccentricity (c. 100 ka) proxy climate records, the influence of the 100 ka cycle is
Milankovitch cycles are key elements in explanations of more difficult to explain, because calculations have shown
long-term climate change (section 1.7), it is now clear that variations in insolation forcing resulting from
that these alone cannot account for all aspects of the eccentricity changes are, by themselves, too small (of the
Quaternary climate record. Indeed, despite the fact that order of 0.1 per cent) to be the direct cause of glacial–
the astronomical theory has been a central tenet of Quater- interglacial changes over the past 900 ka (Imbrie et al.,
nary research ever since the publication of the seminal 1993). This ‘100 kyr problem’ of Milankovitch theory is
paper by Hays et al. (1976), it is a curious fact that there is further compounded by the fact that the longer 413 ka
still no universally accepted account of the individual eccentricity cycle, which appears to be the largest com-
and combined effects of the Milankovitch variables on ponent of eccentricity forcing, does not register clearly in
global climate (Elkibbi & Rial, 2001). Moreover, while a the marine oxygen isotope records (Rial & Anaclerio,
Milankovitch signal is clearly evident in many proxy 2000). The pacing of recent glacial–interglacial cycles is also
records, in some the astronomical rhythms are more problematic, for although the climatic mode (as reflected
subdued, while in others there are clear off-sets between in marine isotope records) of the last 800–900 ka is around
astronomically predicted and empirically based climatic 100 ka, the duration of consecutive glacial periods actually
cycles. One reason for this may be because the former are varies from c. 80–120 ka. The interval between the last two
approximations of complex planetary harmonics: the c. 21 interglacials is 120 ka, while around 400 ka, the interval was
ka cycle, for example, is an average of two separate but c. 80 ka, with three successive interglacials occurring in less
interlocked cycles, while the c. 100 ka cycle is a derivative than 200 ka (Raymo, 1997). These and other apparent
of at least five separate cycles of varying power, ranging in inconsistencies between observation and theory have
frequency between 95 and 107 ka, and each reflecting prompted questions about whether we can indeed match
varying orbital interference between Mars, Mercury, Venus, the geological record to the astronomical variables (Meyers
Earth and Jupiter (Berger et al., 2005). And there are other et al., 2008) and, moreover, whether the earth’s climate
orbital factors that might influence global insolation, for system does actually respond in a linear fashion to the
example changes in the Earth’s orbital plane (the imaginary Milankovitch influences, or that the nature of the response
surface marked out by the Earth’s orbital path) which is is non-linear or stochastic (Wunsch, 2004; Huybers &
known to vary in terms of its inclination over a quasi-100 Wunsch, 2005). While this is still perhaps a minority view,
ka cycle (Muller & MacDonald, 1997). Off-sets between it does highlight the fact that an understanding of long-
empirically derived and astronomically driven climatic term climate change requires additional explanatory
cycles may also reflect chronological and stratigraphical mechanisms.
uncertainties associated with geological age models (Ao In this section we look again at the Astronomical Theory
et al., 2011), or the fact that climatic responses may lag in the light of some of these difficulties and we focus on two
changes in insolation because of delays in environmental particular aspects of the Quaternary record: the switch
feedbacks, for example mass balance changes of large from 41 ka to 100 ka cycles at the Middle Pleistocene
continental ice sheets (see Ruddiman, 2006) which influ- Transition (MPT), and the factors driving glacial–
ence the δ18O signal in the oceans through variations in interglacial cycles during the last 800 ka. The latter, in
ice mass volume (section 3.10.2). Furthermore, the detail particular, have importance for evaluating developments
in empirical records may be blurred by the application during the present interglacial (Holocene), including the
extent to which humans may have exerted an influence over Pacific shown in Figure 7.7, for example, the isotopic
natural environmental processes, and to which we return signal, reconstructed sea-level history, and inferred sea-
later in the chapter (section 7.6). surface temperature records (SSTs) are clearly dominated
by the 41 ka cycle before c. 1.25 Ma, but this is followed
by a period up to c. 700 ka where the pattern is much less
7.3.2 The Mid-Pleistocene Transition
clear (Siddall et al., 2010a). Thereafter, the 100 ka cycle
(MPT) becomes dominant, and this accords with the majority of
As we noted in Chapter 1, the terms ‘Middle Pleistocene other records which show that by c. 650 ka, the start of what
Transition’ (MPT) or ‘Mid-Pleistocene Revolution’ refer to has been termed the Milankovitch chron (Berger et al.,
a shift in spectral frequency in oxygen isotope records, 1994), global climate was locked into the 100 ka rhythm of
between the ‘41 kyr world’ that characterizes the early change (Figure 7.8a). It is important to remember, however,
to Mid-Pleistocene and the ‘100 kyr world’ of the late that what we are seeing in Figure 7.8a are variations in the
Pleistocene (Raymo & Nisancioglu, 2003). For the sake of strength of the global climatic response to astronomical
convenience, this change in spectral dominance is often forcing factors, and this record appears to show that the
described as having occurred around 900–800 ka but, in global climate system became much more responsive to the
practice, there is considerable variability between differ- 100 ka insolation cycle after the MPT. But herein lies one
ent proxy records in the timing of the onset of the MPT of the problems referred to above, namely that despite this
(e.g. Head et al., 2008b; Elderfield et al., 2012). Moreover, increased sensitivity of global climate to the Milankovitch
the transition is often gradual, spanning several hundred rhythms post the MPT, astronomical calculations suggest
thousand years. In the marine isotope record from the that eccentricity variations are insufficient in themselves
a) MPT
S u s c eSput isbci el i pt t i b i l i t
Sput isbci el i pt t i b i l i t
0 0.5 1 1.5 2 2.5 3 35 4 45 5
b) A g e (Ma)
100
S u s c eSput isbci el i pt t i b i l iSt u s c e

0
-100
0 0.5 1 1.5 2 2.5 3 3.5 4 45 5

A g e (Ma)
4
2
0
-2
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
A g e (Ma)
d} 30
25
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

A g e (Ma)
Figure 7.7 Deep Pacific records of environmental change during the past 5 Ma. a) Composite oxygen isotope record based on
benthic Foraminifera. b) Variations in global sea level interpolated from the isotopic data. c) Temperature residuals of the isotopic
data that cannot be explained by the sea-level component. d) Surface ocean temperature variations based on Mg/Ca analysis of
marine microfossils. The crosses are the primary data, the curve is a moving average trend-line, and the two dashed lines represent
an arbitrary 2°C interval used as a guide to long-term trends in SST. For further detail see Siddall et al. (2010a).
to account for the large-amplitude oscillations that charac- periods of more agitated or excited behaviour. The latter
terize the late Pleistocene mode of climatic behaviour; in appear to be exaggerated responses to the low-amplitude
other words, the repeated shifts between glacial and eccentricity signal, as for example during the last 400 ka, a
interglacial conditions (Paillard, 2001). Furthermore, period informally termed the Emiliani chron (Figure 7.8b).
when considered over the timescale of the last 1 Ma or so, Since these marked shifts from muted to excited behaviour
the strength of the correlation between marine records of the climate system cannot be explained by Milankovitch
(reflecting climate response) and insolation data (possible forcing alone, they must reflect the influence of additional
forcing) indicates a considerable (and surprising) variability factors. Efforts to establish what these factors might be
in sensitivity of the climate system to changes in the and how they might have influenced global climate rhythms
Milankovitch variables (Figure 7.8b). Particularly curious prior to and post the MPT have involved three main
is an interval of unusually weak correlation between the approaches: the analysis of terrestrial, oceanic and atmos-
proxy records and insolation data, which Berger (2013b) pheric records; investigations of what is termed ‘phase
has described as a ‘deaf zone’ (a period when the global interference’ between the astronomical cycles; and the use
climate system seems to have been relatively inert to of numerical models. Each of these will be considered in
astronomical forcing), and which is interposed between turn.
a) 3.5
Brunhes Matuyama
2 5
Susceptibilit
2 Chro
1.5
Milankovitch C h r o n
0.5
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
b) A g e (ka)
1
vV=12 W=25
0.8
Deaf zone
Susceptibilit
0.6 460
0.4
0.2
0
0 200 400 600 800 1000
A g e (ka)
Figure 7.8 a) Marine oxygen isotope record for the last 2.1 Ma (ODP Site 806) and the spectral signal of the 41 ka and 100 ka
cycles from this record. b) A measure of the strength of correlation between the stacked oxygen isotope record and Milankovitch
forcing over the past 1 Ma. The method employs ‘sliding windows’ (W) of 50 ka (darkest grey and thickest line), 25 ka (mid-grey;
W = 25) and 12 ka (lightest grey; W = 12) intervals in the correlations. The data reveal a period (centred on 460 ka) of particularly
low correlation coefficients (r-squared values) suggesting a time of low sensitivity of the earth system to insolation, which Berger
(2013b) refers to as a ‘deaf zone’ (from Berger, 2013b). For further explanation see text.
A number of potential amplification agents have been ka climate rhythm, pointing to the uplift of the submarine
proposed from an analysis of geological records. Consid- Greenland–Scotland ridge which occurred between 900 and
erable attention has focused on variations in atmospheric 650 ka BP, and which may have led to marked changes in
CO2 and other greenhouse gases, as ice-core data in North Atlantic deep-water circulation.
particular have shown a close relationship between the As noted above, a characteristic of the Mid-Pleistocene
Milankovitch variables, atmospheric CO2 and inferred Transition (MPT) is that post the MPT, global ice volume
temperature changes (see Figure 7.13). Physical processes was greater and ice sheets appear to have been substantially
within the atmosphere have also been suggested as forcing thicker than in pre-transition times. Clark and Pollard
factors for the MPT, such as intensification of the Walker (1998) attributed this change in ice thickness (and volume)
circulation cells leading to increased moisture transport to the gradual removal by ice-sheet erosion of a thick
to high latitudes, thereby accelerating the growth of the regolith. Prior to the MPT, this ‘low-friction’ regolith led
northern continental ice sheets (McClymont & Rosell- to the development of extensive, but relatively thin ice
Melé, 2005). In addition, there are indications of a signifi- sheets that responded linearly to the 41 ka obliquity cycle.
cant increase in Antarctic ice volume at around 900 ka As this mantle of debris was removed by successive
which some have seen as a potential trigger for the MPT glaciations, large areas of Precambrian Shield crystalline
(Elderfield et al., 2012), while others have invoked vari- bedrock were exposed, and this ‘high-friction’ substrate
ations in atmospheric dust flux, which alters surface albedo induced thicker ice sheets with a fundamentally different
(especially over ice sheets) and insolation receipt (Bar-Or response to orbital forcing after the MPT. In addition,
et al., 2008; Martinez-Garcia et al., 2011). A widely held regolith erosion and exposure of crystalline bedrock could
view is that the change in climatic periodicity around the have caused an increase in silicate weathering rates (section
MPT and the concomitant intensification of Northern 1.7), and carbon-cycle modelling suggests that the draw-
Hemisphere glaciation could be due to tectonic activity, down in atmospheric CO2 that would have accompanied
with progressive land uplift (e.g. in the Himalayas) leading this process could have led to a decrease of around 7–12
to the transgression of critical altitudinal thresholds and ppm in atmospheric CO2 (Clark et al., 2006). If this was
accompanying changes in atmospheric circulation regime indeed the case, then the global cooling resulting from
(Raymo et al., 1992). Other explanations for the shift in this secular decrease in atmospheric CO2 may have been
climatic phasing prior to and post the MPT, however, an important feedback in causing the MPT (Mudelsee &
involve a complex interplay of Milankovitch forcing Schultz, 1997; Raymo, 1997). In a radically different inter-
(principally on precessional and obliquity timescales), and pretation, it has been suggested that the apparent shift
feedback mechanisms involving ocean circulation changes, in pacing from c. 40 to c. 100 ka cycles that characterized
ice-sheet build-up and atmospheric gases. For example, the MPT can be seen not as a response to changes in
Raymo (1997) has suggested that prior to 900 ka, ice sheets atmospheric CO2, continental regolith exposure or other
melted during each precession- or obliquity-induced warm external control, but rather as a reflection of glacial vari-
stage, but around 900 ka, Northern Hemisphere tempera- ability spontaneously switching between long and short
tures had cooled to a critical threshold which allowed period modes; in other words, it is another aspect of the
ice to persist through weaker insolation maxima and hence chaotic or non-linear response to orbital forcing referred
to grow to a larger size over each successive cycle. Because to above (Huybers, 2009).
high-amplitude precession maxima only occur in every There is, therefore, a range of factors within the earth’s
fourth cycle (100 ka), due to eccentricity modulation, ice terrestrial, oceanic and climate systems that could, through
sheets would tend to grow and melt with a periodicity of a complex network of linkages and feedback mechanisms,
100 ka. Such effects would have been increasingly ampli- have affected the astronomical pacing of global climate
fied by CO2 feedback (Ruddiman, 2003a). Denton (2000) and instigated the MPT, and the subsequent shift to the
has pointed to the effects of the ocean in such feedback 100 ka mode (McClymont et al., 2013). It is, however,
loops. He envisages ice sheets growing steadily over the difficult to establish which, if any, of these individual
course of a 100 ka cycle, extracting waters from the world’s factors could have been driving climate change, as opposed
oceans and leading to major reorganization of deep- to being secondary effects, an issue we return to in later
water circulation from an ‘interglacial’ to ‘glacial’ mode. sections of this chapter.
The Milankovitch variables will eventually trigger ice-sheet The second approach to resolving the conundrum of the
collapse which then shifts circulation back into an inter- MPT focuses on the analysis of the amplitude, frequency
glacial mode and the cycle begins again. He also invokes a and phasing (concordancy) of the different Milankovitch
possible tectonic component in the development of the 100 rhythms, and their integrated effects, the principles and
language used in these analyses being borrowed from the than previously realized, with the global climate system
study of electromagnetic wave-forms. The process of having apparently synchronized to this frequency around
modifying a wave’s amplitude while keeping frequency 1.2 Ma, after which an increasing alignment between the
constant is termed amplitude modulation, while the 41 ka, 100 ka and 400 ka cycles culminated in enhanced
opposite (constant amplitude but variable frequency) is amplitude of the 100 ka cycle.
referred to as frequency modulation. In theory, natural The third approach uses numerical modelling to test
earth surface processes could modulate insolation waves observations or ideas obtained from empirical evidence or
in both ways. On the other hand, the wave-forms them- from statistical analyses of long Quaternary records.
selves could self-regulate through a process termed phase- Reference has already been made (section 7.2.1) to the use
locking: two waveforms can be locked into harmonious of simple box models to evaluate the hypothesis that growth
frequencies, but shifted in phase, by the operation of a of sea ice initiated and regulated the late Pleistocene 100 ka
feedback loop. The combined effect of this could either oscillations (Tziperman & Gildor, 2003). A more compre-
amplify or dampen the output signal, and could do so in hensive ice-sheet–ocean temperature model was developed
either a linear or non-linear fashion, the latter having less by Bintanja & van de Wal (2008) to quantify changes in
predictable outcomes. global surface air temperature, ice volume and sea level
As we have seen, time-series data from the Quaternary from marine benthic δ18O2 measurements spanning the
record, whether they be isotopic signals from ice cores last 3 Ma. The results suggest that the evolution of the
or ocean cores, long trace-gas records, or palaeotemperature 100 ka cycles coincided with a change in the relative import-
reconstructions, are characterized by wave amplitude ance of the North American and European ice sheets, with
and frequency variations, and are amenable to a number the latter apparently dominating before the MPT, but the
of mathematical or statistical methods for analysing the former exerting an increasingly greater influence during the
strength of, and resonance between, interacting wave forms.
Milankovitch chron (i.e. post-650 ka BP; Figure 7.9). This
Perhaps the most widely applied has been spectral analysis
was probably a result of the merging of the Cordilleran and
(see Chapter 1, note 7), but other techniques have also
Laurentide ice sheets, enabling the combined ice mass to
been employed. For example, Huybers (2007) applied a
survive insolation maxima during each successive glacial
probability statistic (Rayleigh’s R)1 in an analysis of the
cycle. Ultimately, this compound ice sheet grew to the
relationship between the timing of deglacial events and
extent that it became unstable, initiating rapid terminations
orbital variations. He focused on thirty-six deglaciation
(Bintanja & van de Wal, 2008).
signals identified in the records for the past 2 Ma and
Ganopolski & Calov (2011) used an earth system model
found that thirty-three of them occurred when the obliquity
of intermediate complexity to demonstrate that both
signal was anomalously large: during the early Pleistocene,
they align closely with the obliquity signal, and hence have strong 100 ka periodicity in the ice-volume variations and
a 40 ka rhythm, while in the late Pleistocene, deglaciation the timing of glacial terminations during the past 800 ka
occurred every 80 or 120 ka, giving an average frequency can be successfully simulated as a direct, strongly non-
of 100 ka. Maslin & Ridgwell (2005) have linked the 100 linear response of the climate-cryosphere system to orbital
ka glacial–interglacial mode of behaviour to resonance forcing, providing that the atmospheric CO2 concentration
between the precession and eccentricity cycles, with eccen- stays below its typical interglacial value. Their modelling
tricity pacing the changes, but the climate response being results suggest that the existence of long glacial cycles
enhanced by concordant phasing with the obliquity cycle. during the late Pleistocene is primarily attributed to the
Other studies have suggested that strengthening of a semi- North American ice sheet, and requires the presence of a
precession signal (periods of about 11.5 ka) during the late large continental area with exposed rocks, and that tem-
Pleistocene was responsible for the development of 100 ka poral variability in CO2 concentration plays an important
cycles (Rutherford & D’Hoydt, 2000). Possible support for role in the amplification of the 100 ka cycles. These
this contention comes from Chinese loess records, which modelling results accord with empirical evidence referred
reveal evidence of well-defined semi-precessional cycles to above for increased ice volume during the later Pleisto-
during the most recent interglacial–glacial cycle, and which cene, and with amplification of the astronomical signal by
are attributed to periodic strengthening and weakening of atmospheric CO2 variations. They also reinforce the view
the East Asia summer monsoon (Sun & Huang, 2006). that non-linear amplification of astronomical forcing may
More recently, however, spectral analysis by Rial et al. be necessary for the generation of the 100 ka cycle that has
(2013) of climatic variations over the past 4 Ma suggests dominated climatic behaviour during the Milankovitch
that the 400 ka eccentricity cycle has been more important chron (see for example Ditlevsen, 2009).
0
-20
^to
S u s c e pSt iubsi lci te pSt iubsi lci te p t iSbui lsi tc e p t i S
-60
S u s c e p t i b i l i tS u s c e pSt iubsi lci te p t i b i l i t

S u s c e pSt ui bsicl iet SSuus sc ceeppt itbi bi liilti t
-80 10
S u s c e pSt iubsi lci ptetpi bt iibl iitl i t

-100 0
-120 -10
-14D -20
0.5 0.7 0,9 1.1 1.3 1.5
Time (Ma)
E u r o p e a n Ice S h e e t North A m e r i c a n Ice S h e e t Global
Temperature M e a n s e a level equivalent = 4 5 S e e caption
Figure 7.9 Modelled time-series of ice volume (top three curves) and Northern Hemisphere subarctic surface air temperatures
(green) from 1.5–0.5 Ma, encompassing the Mid-Pleistocene Transition. The model simulates changes in global sea level (black
curve) and contributions to global sea-level variations made by the European (red) and North American (blue) ice sheets, all shown
relative to long-term mean sea level (msl: pale blue horizontal line). The vertical black arrows indicate positions of msl maxima
(interglacials), approximately every 100 ka. The model results suggest an increasing influence of the North American ice sheet on
global sea levels, with progressively lower global sea level minima (orange bars), over the course of the last 1 Ma (from Bintanja
& van de Wal, 2008).
While these different approaches to the problem have 7.3.3 The glacial–interglacial cycles of the
undoubtedly thrown new light on the nature and causes last 800 ka
of the MPT, there are many questions about this enig-
The high-resolution palaeoenvironmental records for the
matic section of the Quaternary record that remain
last 800 ka obtained from recent deep-ice drilling of the
unresolved. These are exemplified by a recent analysis of
global sea-surface temperature variations over the past Antarctic ice sheet (section 3.11) have provided important
2 Ma (McClymont et al., 2013). This shows that although templates for climatic variability over the last eight glacial–
a common cooling trend is manifest in all records between interglacial cycles. These records have revolutionized the
1.2 Ma and 800 ka, and that the majority are character- study not only of long-term climatic change, but also of
ized by significant cooling steps at 1.2 Ma and 900 ka, potential forcing factors that drive those fluctuations,
some oceanic records appear to show earlier responses. because they contain information on atmospheric condi-
Moreover, for some regions, the coldest glacial stages tions (notably greenhouse gas [GHG] content and dust
occur after 900 ka, whereas in other regions, predominantly flux); on global temperature, principally through the
in the Southern Hemisphere, the glacial stages post- proxy of deuterium (δD); on ice-volume changes; and on
900 ka were warmer than those in the mid- and early a range of other environmental variables. Importantly,
Pleistocene. In addition, the shift towards dominant 100 ka these are co-registered in the same ice core (Figure 7.10),
cyclicity occurs gradually in some records, emerging as and hence precise temporal relationships between each of
early as 1.2 Ma, while in others it is not evident until after them can be determined. Moreover, records are available
800 ka. We return to these and related matters in section from ice from a number of different drilling sites (e.g.
7.3.3, after first examining the nature and form of the Vostok, Dome C, Dome F and Dronning Maud Land),
100 ka oscillations reflected in the fossil record of the which enables local influences to be identified and quan-
past 800 ka. tified. The remarkable degree of overall compatibility
Susceptibilit 0
Susceptibilit
-2
-2
Susceptibilit
Susceptibilit
0
2
Susceptibilit
-2
-2 2
Susceptibilit
0
2
-2
Susceptibilit
-2
0 200 400 600 800

A g e (ky B P )
Figure 7.10 Records of deuterium (δD), dust flux and aerosols (CO2, CH4, Ca and Na flux) over the last 800 ka obtained from
the EPICA Dome C (EDC) ice core, East Antarctica, compared with sea-level changes inferred from the LR04 marine isotope record
(Figure 1.5). The data are shown as variations in standard deviation units from the long-term means; the thick lines indicate forcing
on orbital timescales, while the shaded areas reflect non-orbital influences – a combination of data uncertainty and feedback variability
(after Masson-Delmotte et al., 2010).
between different Antarctic ice-core records is illustrated impurities (e.g. dust), and in other variables, can alter the
in Figure 7.11. atmosphere’s radiative balance and, if sustained, can lead
Antarctic ice-core data have provided new insights into to a new climatic regime. As we saw in Chapter 3, variations
glacial–interglacial climate dynamics, particularly in regard in atmospheric gas content can be determined from ice-core
to radiative and climate feedback forcing mechanisms, records, and their radiative forcing impacts can then be
and also in the estimations of time lags between forcing estimated from laboratory measurements of the radiative
factors and climatic response. Radiative forcing is any effects of individual gases (e.g. CO2, CH4, measured in
change in the intensity of the solar radiation signal trans- Watts per square metre: W m–2), or from records of the
mitted through the atmosphere to the earth’s surface. impacts of recent changes in atmospheric greenhouse gas
Changes in greenhouse gas content, cloud properties, content on modern temperatures (see for example the
40 5 1 1
Dome C
Susceptibilit
30 Dome F 3.8 0
Vostok a)
2C 2 6 -1
10 1 3 -2
sScpuetsipbctieilbiptitliibt i l i t
Susceptibilit
Susceptibilit
5
0 0 -3
0
b)
-10 -1.3
SusceptiS
•5
ce
-20 -2.6 4
S uSs u
-10
2
•30 -3.8
0
-40 •5.1
-2
-50 -6.4 -4
0 50 100 150 200 250 300
Susceptibilit
24.0 2
A g e (kyr)
23.5 0
Figure 7.11 Isotope time-series data for the past c. 360 ka 23.0 -2
from three East Antarctic ice-core records: EPICA Dome C -4
(EDC; δD), Dome F (Fuji; δ18O) and Vostok (δD). Although the
data have been converted to a common timescale (an EDC age 0 100 200 300 400 500 600 700 800
model), nevertheless the degree of similarity between all three A g e E D C 3 (ka)
suggests synchronous changes within the East Antarctic ice
sheet (from Sime et al., 2009). Figure 7.12 Radiative forcing, temperature profile and the
obliquity signal in the EPICA Dome C (EDC) ice core over the
past 800 ka. a) The estimated radiative forcing component of
NOAA Annual Greenhouse Gas Index [AGGI]: http://www. the EDC record. b) The EDC temperature profile reconstructed
esrl.noaa.gov/gmd/aggi/). In Figure 7.12, a radiative forcing from δD. c) The residual temperature component that cannot
component has been estimated for the 800 ka long EPICA be explained by either radiative forcing or obliquity (thin grey
Dome C (EDC) ice core (7.12a), and subtracted from the line); the thick orange line is the long-term orbital trend. d) The
original temperature record obtained from the δD signal orbital component of the EDC record superimposed on the
obliquity cycle (from Masson-Delmotte et al., 2010).
(7.12b). The residual values are plotted against the orbital
element of the temperature record (Figure 7.12c), which
reveals the component of the EDC temperature record that
cannot be explained by radiative or orbital forcing. This mechanisms were clearly important, their effects appear
represents the amount of unexplained orbital forcing which to have consistently lagged the obliquity signal by c. 5 ka.
must therefore be caused by feedback processes. Finally, the EDC record reveals evidence for a gradual
To be able to quantify and compare potential atmos- long-term cooling trend between 700 and 400 ka (Figure
pheric forcing factors with such a high temporal resolution 7.12c). These and other Antarctic ice-core data provide a
over a period of eight glacial–interglacial cycles represents further confirmation that the Milankovitch rhythms are
a major advance in the search for causes of Quaternary mediated by other factors. This is further underlined by the
climate change and, in particular, of possible linkages contrast between the symmetrical wave-form of astro-
between long-term climatic patterns and forcing factors. nomical forcing (Figure 7.13a) compared with the strongly
In Figure 7.12, for example, it is apparent that the EDC asymmetrical form of the proxy records of glacial–
temperature record for the last 800 ka has a strong obliquity interglacial climatic oscillations reflected in the ice-core
component (Figure 7.12d), and that climate sensitivity (as data (Figure 7.13b). This asymmetry in climate signal first
reflected by the amplitude of variations in Figure 7.12b and appears at around 2.5 Ma and becomes increasingly
c) was significantly higher during the last 450 ka (the accentuated in the late Pleistocene (Lisiecki & Raymo,
Emiliani chron) than in the period 800–450 ka. In addition, 2007), particularly during the Emiliani chron (i.e. post-400
it appears that the pattern of climatic variability differed ka). As explained earlier (section 3.10.2.2), the end of each
significantly between individual glacial and interglacial cold or glacial stage in the isotopic records (terminations)
stages: MIS 11 and 5.5 appear to have been the warmest is often abrupt, but the cooling trends that precede each
interglacials, MIS 16, 12 and 2 had the strongest glacial termination are more gradual (Figure 7.14), a pattern that
maxima, while glacials MIS 13 and MIS 17 were notably again cannot be explained by astronomical forcing alone,
weaker in intensity. Moreover, the high residual values but which must reflect feedback processes within the global
(Figure 7.12c) indicate that although climate feedback climate system once a critical threshold has been trans-
600
Susceptibilit
Susceptibilit
Susceptibilit
550
a)
500
450
0
-4
-8
350
300
Susceptibilit
250
200 ;
150
Susceptibilit
15
Susceptibilit
d) 10
5
0
BOO 600 400 200 0
A g e ka
Figure 7.13 Milankovitch insolation record for the past 800 ka a), compared with temperature b), CO2 c) and dust profiles d)
from the EPICA Dome C (EDC) ice core, Antarctica. The dashed vertical lines show correspondence between peak interglacial
temperatures (curve b) and insolation power (curve a) (data from http://www.climatedata.info).
gressed. Possible feedbacks that could trigger terminations insolation in both hemispheres over the past 2 Ma indicates
include changes in snow-ice albedo, which feeds the growth that at each termination insolation increased in both
of ice in the early stages until the ice sheets become too large hemispheres in concert, but with the Southern Hemisphere
to sustain the process (Stott et al., 2007); the release of having a slight lead (Schulz & Zeebe, 2006). This has been
CO2 stored in the deep ocean during cold stages but released supported by records from Antarctica which suggest that
as the ocean warms, the process being subsequently accel- warmings or terminations are led from the Southern Hemi-
erated by a positive feedback greenhouse effect (Skinner sphere, although specific conditions (ice-volume changes,
et al., 2010); or changes in ocean circulation in the Southern ocean circulation, etc.) in the Northern Hemisphere are
Ocean which initiates positive feedbacks through altered needed in order for the climate state to complete its shift
albedo or CO2 release in both hemispheres and which, in from glacial to full interglacial conditions (Wolff et al.,
turn, leads to what has been referred to as ‘runaway global 2009). Hence, although it is clear that the Northern
warming’ (Wolff et al., 2009). Hemisphere generated the greatest volume increase in
Data from Antarctic ice-core records are also beginning glacial ice during successive glacial stages, it now seems that
to challenge some long-held views concerning the dominant processes operating in the Southern Hemisphere played the
role that the Northern Hemisphere has played in global key role in terminating the glacial stages, a subject that we
climate change. Following Milankovitch’s initial proposal, return to in section 7.4, where we consider climatic
it has tended to be assumed that climatic patterns in the oscillations operating over shorter timescales.
Southern Hemisphere were largely controlled by Northern
Hemisphere insolation, a view that is still current in some
quarters today (e.g. Kawamura et al., 2007), although one
7.3.4 Overview
that has become increasingly contentious (Laepple et al., From the foregoing accounts, it might be thought that
2011). For example, statistical analysis of midsummer feedback effects within the global terrestrial–oceanic–
0
5 5
ep
S u s cS
-5 T,
4 5 1 5.3
2
3
5 2 fl 4
Marine Isotope Staqes
-10
20 40 60 80 100 120
5
7.5 93 11 3
7.1 7.3
0
11.1
7.4
-5 6 3 10 72.3
72.2 12.4
7 2
ep
11.2
S u s cS
9.2
-10
150 200 250 300 350 400 450
13.3 14.3 15.3

18.3
15.1 15.5 17.3' 19.3
•5 Jvni
13 1
16.2 $8 2
ep
13.2
S u s cS
145 144 15.215.4

-10
500 550 600 650 700 750 800
A g e (ka)
Figure 7.14 Temperature anomaly record (deviation from late Holocene mean, the zero line in the figure) for the last 800 ka
derived from the EPICA Dome C (EDC) ice-core deuterium record. The record matches closely the LR04 benthic marine record
(Figure 1.5) so that marine isotope stages and substages (numbered) back to MIS 19.3 can be readily identified. Also shown are
the nine terminations (T-I–T-IX) that are prominent features in the record (from Jouzel et al., 2007).
climatic system have assumed such prominence in inter- Modulation of the underlying astronomical signal by other
pretations of the Quaternary climate record that the factors is evident throughout the Quaternary record, and
Milankovitch hypothesis as an explanatory vehicle for is implicated not only in the long-term cooling trend
long-term climate change no longer has currency. This is reflected in the marine oxygen isotope profiles (Figure
most certainly not the case. Modulation of the astronomical 1.5), but also in shifts in climatic phasing in both marine
cycles fails to obscure the characteristic astronomical and ice-core time-series data (such as the MPT), as well as
climatic rhythms, which persist as indelible hallmarks in in the amplitude and frequency of the climatic signal in
marine and ice-core records, and which therefore remain different proxy records. The increase in prominence of the
as fundamental components in the formulation of any 100 ka glacial–interglacial cycles and their abrupt termin-
comprehensive theory of Quaternary climate change (Roe, ations are further signs of signal amplification, operating
2006; Meyers et al., 2008; Berger & Loutre, 2010). But while over shorter timescales. But uncertainties continue to
the Milankovitch theory remains a key element in the obscure the precise physical processes and linkages in-
explanation of glacial–interglacial cycles, it must now be volved in the modulation process: for example, the MPT
acknowledged that it provides only a partial explanation of appears to have occurred without any significant change in
the pattern and magnitude of long-term climate change. insolation forcing; the prominence of the 100 ka cycle
during the Milankovitch chron remains enigmatic; and the long-term context. Global climates were generally
additional (non-Milankovitch) spectral peaks have been colder, sea levels lower (at times much lower) and a great
detected in some Quaternary records which cannot yet be deal more glacier ice existed than is the case today. If this
explained (Huybers, 2007). A key challenge for Quaternary was so, then perhaps the maxim ‘the present is the key to
science in the future, therefore, is to try to understand not the past’, which underpins much of our work in Quatern-
only what the key modulation/amplification processes are, ary science (section 1.5), might therefore require some
but how they operate, how they are interlinked and, above qualification!
all, how they complement insolation variations.
These problems and difficulties notwithstanding,
there can be little doubt that our knowledge of climate 7.4. ENVIRONMENTAL CHANGE
evolution at the Milankovitch timescale has advanced OVER SUB-ORBITAL (MILLENNIAL)
considerably in recent years. We are now able to match
the high-resolution ice-core isotopic variations with those
TIMESCALES
of the marine isotope record for the past 800 ka or so
(Figure 7.14), which clearly suggests a close coupling
7.4.1 Introduction
between ice-sheet growth and decay and ocean tempera- As we saw in Chapter 3, the most highly resolved and
ture changes over that time period. This further confirms continuous records spanning the last glacial cycle are to be
a linkage that has long been considered, namely the reci- found in the Greenland ice cores. Of these, perhaps the
procal relationship between ice sheet, ocean temperature most important in terms of the last cold stage and the Last
and, by implication, air temperature variations. Some Termination (section 7.5) is the core from NorthGRIP
general trends are discernible throughout this period, for (NGRIP), as this has been dated by multiparameter annual
example the greater amplitude of the climatic cycles layer counting (section 5.4.3.3) and therefore constitutes a
(differences between thermal maxima and minima) post- stratotype against which other records can be compared
400 ka compared with earlier in the record, but what is also (Svensson et al., 2008; Blockley et al., 2012; section 7.5).
apparent is that over the past 800 ka, no two glacial or The most distinctive feature of the NGRIP record is the
interglacial episodes show precisely the same climatic sequence of twenty-five short-lived stadial–interstadial
pattern. This is very curious given that the same underlying oscillations, the so-called Dansgaard–Oeschger (DO)
climatic rhythms (astronomical variables) and modulating events (section 3.11.4). The onset of each event was abrupt,
factors were operating throughout that period. Lang & with peak warmth achieved in some cases in less than
Wolff (2011) attribute this variability to preconditioning fifty years (Steffensen et al., 2008), while the amplitude
of the environment, in that the climatic evolution of an of warming ranged between 5 and 16°C (Figure 7.15). This
individual isotopic stage is determined by three principal sequence of DO events reflects repeated instability in the
influences: insolation variations, natural feedbacks (modu- climate–ocean system of the North Atlantic region over
lation), and baseline conditions inherited from the millennial timescales, and is unusual in the Quaternary
preceding stage. As in a game of cards, although the rules record in terms of the amplitude and frequency and, above
remain the same, each hand unfolds differently, according all, in the resolution of the climatic oscillations.
to the cards that are dealt at the start. Although the DO events are most clearly reflected in the
Finally, it is worth noting one particularly enigmatic Greenland ice cores, climatic instability during the last
feature of the record of the last 800 ka, and that is the cold stage was not only a feature of the Greenland record,
very limited time that temperatures have been as warm as for it is evident in proxy data from a range of terrestrial
those of the present. In Figure 7.14, where the black line contexts and is apparent in a number of profiles from the
across the diagram equates with the mid- to late Holocene deep ocean (section 6.3.3.2; Figures 6.14–6.16). Evidence
thermal maximum, it is clear that in only four earlier for climatic instability is also found in Antarctic ice-core
interglacials did temperatures exceed this value: MIS 7 and records. The question arises, therefore, as to whether these
MIS 9 for a short period only, while in MIS 5 and MIS 11 various archives are related and reflect contemporaneous
the thermal maxima were sustained for longer. Indeed, climatic events, or whether there are spatial and temporal
from the data in Figure 7.14, it would appear that the total variations in the respective climate signals.
amount of time during which temperatures were as warm The linkage of climatic anomalies over great distances
as or warmer than the present day appears to represent, at (i.e. at the global scale) is referred to as teleconnection.
most, only 10 per cent of the last 800 ka. In other words, These linkages are a result of energy transport and wave
present-day conditions are highly unusual, when viewed in propagation in the atmosphere and ocean, enabling the
-34 Younger Dryas Warm
Dansgaard-Oeschger Events
1
2 3
, 24 25
20 21
12 22
19
8 14 16
Susceptibilit
-38 3 11
4 *6 7
10 15
13
• 18
1 7
-42
N G R I P Ice C o r e
-46 Cold
H1 H2 H3 H4 H5 H6
0 50 100
A g e (ka)
Figure 7.15 The δ18O record from the NorthGRIP ice core over the last 125 ka. Short-lived warming (Dansgaard–Oeschger) events
are numbered from most recent (1) to oldest (25). H1–H6 are the six Heinrich events, explained in the text (from Clement & Peterson,
2008).
atmosphere to act as a ‘bridge’ between different parts of Atlantic where sediment accumulation rates were excep-
the ocean, and the ocean to act as a ‘tunnel’ linking differ- tionally high (Bond et al., 1993). Sea-surface temperature
ent atmospheric regions (Liu & Alexander, 2007). In this variations reconstructed from these cores, based on the
section we examine the degree to which the marked abundance of the left-coiling cold-water foram species N.
climatic, oceanographic and cryospheric changes that pachyderma (section 4.10.7; Figure 4.44), closely mirror the
characterize the last cold stage do indeed reflect telecon- Greenland isotope records in the number and relative
nected events. We first consider the role of ocean–ice duration of inferred DO events; in the abrupt onset and
interactions in the North Atlantic region, long regarded as asymmetric form of each DO event, particularly noticeable
having had a dominant influence on climate during the last for events of longer duration; and in the tendency for
cold stage, then assess the evidence for oceanic effects at the interstadials to be grouped into longer-term cooling
global scale, and finally evaluate other factors that may have cycles, in which successive DO events have progressively
generated rapid and repeated climatic fluctuations over lower thermal maxima (Figure 7.16). These longer-term
millennial to centennial timescales. composite cycles have subsequently been termed Bond
cycles (Rasmussen & Thomsen, 2004). Initially, it was
believed that they approximated a quasi-regular ~1,500-
7.4.2 Ice–ocean–climate interplay in the
year periodicity (Bond et al., 1997). However, subsequent
North Atlantic high-resolution dating on both the ice-core records and
Prior to the publication of the Greenland ice-core records, on marine sequences, and the application of new age
interpretations of the climatic history of the last cold stage models to one of the key profiles investigated by Bond
were based largely on marine oxygen isotope records. et al., suggest that the 1,500-year cycle may, in fact, be an
While these have proved invaluable in Quaternary science admixture of ~1,000- and ~2,000-year cycles. Overall, the
(section 3.10), many were obtained from areas of the 1,500-year number, which has been widely cited in the
seabed where sediment accumulation was slow. As a result, literature, now appears to be an artefact of averaging and
the temporal resolution of the last glacial cycle was relatively seems to have little statistical justification (Obrochta et al.,
low and, while the major isotopic stages could usually be 2012). This is not to say that the Bond cycles do not exist
discerned, finer structures of the ocean–climate record (for we return to these below), but that there remains a
were often obscured. In the 1990s, however, Gerard Bond question over their periodicity.
and his colleagues showed, for the first time, that high- Ocean-core records obtained by Bond and others show
resolution ocean–climate records could be obtained from that both surface and deeper circulation in the North
deep-ocean sediments by sampling areas of the North Atlantic varied markedly in strength and direction during
Stage 5B
20
u tsi cbei lpi tt i b i l i t S u s c e p t i b i l i t
40
60 % N. pachyderma (s.)
80 at D S D P site 6 0 9
100
20
40
% N. pachyderma (s.)
ep
60
S u s cS
in V e m a 2 3 - 8 1
80
100
fee
rating
-36 YD HI H2 H3 H4 H5 H6
Susceptibilit
-38
-40
5 O in S u m m i t
1 s
-42 Inteirstadiat
Ice C o r e ( G R I P )
number f e
o Longer-term 14 IS 19 ?1
2 4 "cooling cycles 14
Bond cycles
Depth (m) in Summit ice core (GRIP)
1,600 1,8001,900 2,000 2,100 2,200 2,300 2,400 2,500 2,600 2.700
10 20 30 40 5 0 60 70 80 90
A g e (ka)
Figure 7.16 Correlation between variations in abundance of N. pachyderma, a cold-tolerant foraminiferal species, from North
Atlantic deep marine sites DSDP-609 and V23-81, and the δ18O record from the GRIP Summit ice core. Also shown are the long-
term Bond cycles, defined by grouping the Dansgaard–Oeschger cycles and Heinrich events (H1–H6) (after Bond et al., 1993).
the last cold stage (Rahmstorf, 2002). The principal in Figure 7.17 (see Summerhayes & Thorpe, 2002, or visit
circulation system that operates at the global scale is known the NASA web site [or YouTube] for animated films of the
as the ‘ocean’ or ‘global conveyor’ (Figure 7.17) and is system in operation).
driven by variations in water temperature and density The surface flow of the North Atlantic is dominated
(thermohaline circulation: THC). A number of processes by the Gulf Stream, warm salty water that is driven north-
contribute to the global THC flow pattern, but a key driver eastwards by prevailing winds, keeping northwest Europe
is surface wind. The trade winds, which blow constantly significantly warmer than equivalent latitudes on the
from east to west, lead to evaporation of surface waters in western Atlantic seaboard. As this salty water cools, its
the tropical Atlantic and the net export of moisture into density increases and in the vicinity of south Greenland
the Pacific where it is precipitated. As a result, the surface and the Norwegian Sea it sinks (Figure 7.18) to feed a
Atlantic is saltier and hence more dense than the Pacific. North Atlantic Deep Water (NADW) current that flows
Density variations also arise from the cooling of waters in southwards to compensate for the north-flowing Gulf
subpolar regions, aided by expulsion of brine as sea ice Stream; in combination, these currents comprise the
forms, and are particularly evident in the North Atlantic Atlantic Meridional Overturning Circulation (AMOC).
and circum-Antarctic (Clark et al., 2002). The result is a Cold water sinking from the margins of Antarctica forms
sinking of surface waters and the formation of deep-water an Atlantic Bottom Water (ABW) current that is denser
currents that link to surface flows by upwelling, for example than NADW, and which feeds an overturning cell in the
in the eastern Pacific close to Chile and Peru (Humboldt Southern Ocean that is generally weaker than the AMOC.
Current) and the west coast of the USA (California Although other factors serve to complicate this picture, such
Current), and in the eastern Atlantic close to the coasts of as episodic deep-water flow (Praetorius et al., 2008), eddy
the Cape of Good Hope (Benguela Current) and northwest and wind-field influences (Lozier, 2010), and the outflow
Africa (Canary Current). A schematic for the THC is shown into the Atlantic of more saline Mediterranean waters
sea-to-air
transfer warm shallow
current
cold a n d salty
d e e p current
Figure 7.17 Schematic model of the main currents of the global ‘ocean conveyor’: surface (warm) currents are red, bottom (cold)
currents are blue; upwelling of deep water and sinking of surface waters through density variations complete the circuit (modified
from National Earth Science Teachers Association web site, http://www.windows2universe.org/earth/Water/deep_ocean.html).
For further explanation see text.
H e a t input
25
1000
S u s c e p t i bSi lui ts c e p t i b iSl iut s c e p t i b i l i t

20
2000
Susceptibilit
15
3000
10
4000
5
5000
6000 0
40°S 20°S EQ 20°N 40°N 60°N

Latitude
Figure 7.18 Cross-section of the Atlantic Ocean basin showing thermal stratification, principal flows and surface heat exchange.
The black arrows represent, schematically, net gain to, or heat loss from, the ocean surface. Heat penetrates into the ocean surface
layer and is subducted into the upper ocean along surfaces of equal density. The isolines and colour coding show pressure-adjusted
(potential) temperature variations. A combination of winds, sinking processes and ocean circulation leads to heat transport (white
arrows) and Atlantic Meridional Overturning Circulation (AMOC) (from Hegerl & Bindoff, 2005, reprinted with permission of the
AAAS; based on Schlitzer, 2003).
(Rogerson et al., 2012), this general model of the AMOC

a) 0
remains a useful vehicle for explaining North Atlantic
Ocean circulation. 1
Susceptibilit
It now appears that there were three dominant 2
circulation modes in the North Atlantic during the last cold
stage (Rahmstorf, 2002; Clark et al., 2002). During warmer 3
intervals, the AMOC dominated, as it does at the present 4

Warm'
time (Figure 7.19a); during colder (‘glacial’) episodes, the
5
AMOC still operated, but through reduced density and
buoyancy, was confined to shallower depths with a more b) 0
sluggish flow (7.19b); and during Heinrich events (episodes
1
Susceptibilit
of iceberg and freshwater injection following ice-sheet
collapse: see below), water derived from the Antarctic 2
filled the North Atlantic basin, except perhaps for the top
3
1,000 m (Figure 7.19c) when flow was almost switched off
(Clark et al., 2002). 4 'Cold'
In the late 1980s–early 1990s, Wally Broecker and
5
colleagues suggested that the abrupt climate changes of the
last cold stage evident in Greenland records could be C) 0
explained by salinity-driven variations in the strength of the
1
Susceptibilit
AMOC (Broecker & Denton, 1989; Broecker et al., 1990).
This ‘salt oscillator hypothesis’ envisaged two competing 2
influences that drove the surface ocean to switch between 3
climate states. The first is when too much freshwater is
injected into the Atlantic near sites of deep-water formation, 4 'Off
which reduces surface-water density to the point where 5
sinking can no longer take place, thereby slowing or 30°S 0 3(TN 60'N 90 N
shutting down the AMOC. The effect would be to reduce
the amount of heat being transferred northwards from the Figure 7.19 Schematic model of three dominant types of
tropics because the Gulf Stream would be less effective, Atlantic circulation that recurred throughout the last cold
thereby enabling northern ice sheets to grow. However, stage. The black line indicates North Atlantic overturning; the
grey line, Antarctic Bottom Water; the infill at the base of each
evaporation would still be taking place in more southerly block represents bottom topography, with the rise in the north
and temperate latitudes, so salt would accumulate there representing the shallow sill between Scotland and Norway
during times of low AMOC strength, and would do so until near which deep-water formation occurs. Circulation is repre-
the surface water was sufficiently dense to sink, starting up sented for a) a warm interval, b) a cold interval, and c) a
Heinrich event (based on Rahmstorf, 2002). For further
the AMOC once again (the second competing influence).
explanation see text.
The sudden onsets of DO events could therefore be
explained by the latter process reaching a threshold, or
‘tipping point’, and initiating a positive feedback process The records also show that episodic injections of freshwater
through which northward heat transport would, for occurred throughout the last cold stage, either from ice
example, check or reverse the build-up of sea ice or ice sheet collapse (see below) or release from ice-dammed
sheets. lakes, which led to temporary shutdowns of the AMOC
The salt oscillator hypothesis is now an accepted (Clark et al., 2001). Indeed, simulations of the AMOC in
paradigm in Quaternary palaeoclimatology and palae- coupled numerical models indicate that the Atlantic Ocean
oceanography (Alley, 2007). It is supported by a substantial is sensitive to relatively small changes in freshwater supply
body of data from ocean-core records that enable changes (Otto-Bliesner & Brady, 2010). Contemporary monitor-
in strength of North Atlantic circulation during the last ing of modern flow rates and physical properties in the
glacial stage to be reconstructed in considerable detail Atlantic show that the AMOC varies markedly on time-
(Gherardi et al., 2009; dos Santos et al., 2010; Thornalley scales of weeks to months, also suggesting a high degree of
et al., 2010). These allow variations in temperature, density sensitivity to minor variations in water density and pressure
and flow speed associated with the AMOC to be inferred. (Rayner et al., 2011). These may be contributory factors to
present-day short-term variations observed in Atlantic the Laurentide ice sheet (e.g. Alley & McAyeal, 1994),
circulation, such as the North Atlantic Oscillation (section subsequent studies have revealed a more complex history
7.6.4.3). of IRD deposition in the North Atlantic during the last cold
While changes in AMOC strength may have been stage, reflecting non-synchronous behaviour between the
contributory factors in DO cycles, the sequence and pacing Laurentide, Greenland, British and Fennoscandian ice
of the cycles is more difficult to explain. External factors, such sheets (Dowdeswell et al., 1999; Scourse et al., 2009). This
as orbital or solar forcing (section 7.6.4.1) seem unlikely is an important discovery because it means that ice-sheet
to have driven the changes, and although the ~1,000- and instability and collapse, which may have been a factor in
~2,000-year periodicity that now seems to have been the termination of the Bond cycles (see above), was not
associated with the Bond cycles (see above) matches evidence necessarily a circum-North Atlantic phenomenon reflecting
for changes in solar irradiance at around these frequencies, a common forcing factor (or factors); in other words, the
the amplitude of the cycles in terms of inferred temperature ice-rafting events are ice-sheet specific, and are driven by
change seems unlikely to be accounted for by solar forcing internal as opposed to external processes. However, it is
alone. The answer, therefore, seems to lie in some sort of possible that collapse of one ice sheet could have led to the
reorganization and associated feedbacks within the ocean– break-up of others, for as one ice sheet failed and surged,
cryosphere–climate systems. Precisely how this might the rise in sea level could have destabilized ice masses
operate, however, remains unclear, but ice-sheet activity elsewhere. For example, it has been suggested that during
appears to be an important part of the process, for the simi- Heinrich events 1, 2, 4 and 5, European ice break-up
larity between the oceanic and cryospheric records suggests appears to have preceded the onset of ice-sheet melting by
that ocean and ice-sheet behaviour were strongly coupled. ~1,500 years (Maslin et al., 2001).
This linkage is underlined by the record in North A key element in ice-sheet collapse was not simply the
Atlantic cores of the six Heinrich events (H1–H6). As release of icebergs, but the injection of large quantities of
explained in Chapter 3, these are reflected in layers of ice- freshwater into the northern oceans. As we saw above, the
rafted debris (IRD) that mark episodic deposition of North Atlantic is highly sensitive to freshwater input, and
carbonate-rich glacially derived sediments from icebergs even relatively minor quantities of freshwater could cause
drifting mainly eastwards across the Atlantic from the profound changes in ocean circulation (Rahmstorf et al.,
margins of the Laurentide ice sheet (section 3.10.1). The 2005). One possible scenario that follows from this is that
IRD layers tend to coincide with the final cold phase of a as each DO event was associated with ice-sheet melting and
Bond cycle, immediately preceding an abrupt increase in freshwater release into the Nordic Seas, this could have been
temperature (Hulbe et al., 2004). Periodic collapse of the sufficient to reduce the strength of NADW, weaken the
ice margins is indicated by repeated IRD layers after which Northern Hemisphere circulation and thus conserve heat
the ice was able to stabilize and build again, a process in the southern oceans. The resultant warming around
MacAyeal (1992) has termed ‘binge-purge’. This process Antarctica would trigger accelerated ice-sheet melting and
appears to reflect a complex interplay between ice-sheet contraction of sea-ice extent in the Southern Hemisphere,
volume, subglacial processes and sea level. As the ice sheets a reduction in strength of the Antarctic Bottom Water
reached a critical size a ‘tipping point’ was reached, caused, current, and eventually the re-establishment of NADW
for example, by increased basal melting and lubrication formation, initiating the next DO cycle. Ice-sheet build-up
beneath thicker ice margins, and/or by advance of the over a number of DO cycles would, in due course, lead to
ice fronts beyond the grounding line near the edges ice-sheet destabilization and a Heinrich event. This would
of Laurentide continental shelves, enabling seawater to bring an end to the final DO event within a Bond cycle and
penetrate beneath the ice margin (Hemming, 2004; Marcott the resetting of the DO ‘clock’ (Maslin et al., 2001).
et al., 2011). This combination of factors could have been While the above account may explain some aspects of
sufficient to destabilize ice margins, leading to collapse and the North Atlantic record during the last cold stage, other
release of meltwater and icebergs into the northern oceans. issues remain to be resolved. For example, the scenario does
Moreover, as the ice sheets expanded, their margins not offer an explanation for the amplitude, pacing and
encroached closer to the places where deep-water formation frequency of the DO events in the ice cores. In the first part
occurs, and hence where meltwaters could more directly of the DO sequence in Figure 7.15 (c. 110–65 ka), only eight
affect AMOC strength, again contributing to ice-sheet events occurred, and these tend to have higher temperature
destabilization. maxima and minima and, on average, longer durations than
Although the early work on ice-rafting in the North later ones. Double this number of DO events occur post-
Atlantic suggested that the IRD layers were derived from 65 ka, each of which was shorter. It is also notable that
Heinrich events are confined only to the middle and later explanation of what initiates and sustains DO behaviour
part of the record, and there is no evidence of major IRD (Clement & Peterson, 2008).
events pre-65 ka. The period from DO event 17 onwards
therefore appears to have been more climatically unstable
as reflected in increasing DO cycle frequency, which may
7.4.3 A bipolar teleconnection
reflect more frequent ‘binge–purge’ episodes around the Ice-core records from Antarctica show a pattern of climatic
much larger northern ice sheets that developed during variability remarkably similar to, at least in terms of
that period. But if the absence of Heinrich events from the rhythm, those recorded in the Greenland ice cores (Figure
pre-65 ka record is indicative of more restricted ice-sheet 7.20). A major difficulty in comparing the records, however,
growth, then ice-sheet collapse may not have been a is that Antarctic ice cores are seldom annually resolved
primary mechanism for driving these earlier DO cycles. because the layers are thinner than those encountered
Finally, there is the question of the pacing of the DO cycles. in Greenland and diffuse more rapidly with depth. An
Although some have argued that the abrupt climatic events exception is the upper part of the core from Dronning
reflected in the ice cores are paced by a regular cycle with Maud Land (EDML) in which annual layers can be
a periodicity of ~1,500 years (e.g. Rahmstorf, 2003; Braun identified (Fischer et al., 2004), although how far back in
et al., 2005), others have suggested that the recurrence times time these can be traced has not yet been determined. In
of the DO events are indistinguishable from a random the absence of annual chronologies, therefore, synchroniza-
occurrence (Ditlevsen et al., 2007). Much clearly still tion of Antarctic records with the NGRIP record relies on
remains to be learned about the nature and origin of the other methods.
climatic changes that occurred around the North Atlantic As explained in Chapter 6, Greenland and Antarctic ice-
during the last cold stage. Indeed, despite the fact that the core records have been linked using atmospheric trace
Dansgaard–Oeschger events were first identified in the ice- gases, particularly methane (CH4). Changes in the methane
core records as long ago as 1993, there is still no coherent content of the atmosphere are quickly equilibrated globally,
a)
-38 MIS5.5 -36
MIS1
Susceptibilit
MIS4 EDML -40
-43
MIS3 -44
LGM
Susceptibilit
-48 -48
-52
-53 Dome F
-330
-58
-370
ptibilit
-63 EDC
-410
S u s cSeupst icbei l i t
b)
-32 -450
D01
20 21
2 3
24 25 -490
-36 11 12 14 ' 2
Susceptibilit
6 8 10 17 19 22
s 7
15
45 13 13 -530
2 9
Susceptibilit
-40
-44 NGRIP
-48
0 50,000 100,000 150,000
E D M L a g e (ka)
Figure 7.20 Stable isotopic records spanning the last glacial cycle from Dronning Maud Land (EDML), Dome F and EPICA Dome
C (EDC), Antarctica, aligned with the NGRIP isotopic record from Greenland. All four records are plotted on a common timescale
(explained in text). MIS – marine oxygen isotope stages; LGM – Last Glacial Maximum; DO – Dansgaard–Oeschger events, 1–25
(from EPICA Community Members, 2006).
and hence variations in atmospheric methane concen- For the part of the record shown in Figure 7.21, linking
tration should register contemporaneously in both polar ice the Greenland and Antarctic ice cores using methane
sheets and should serve as a basis for correlating Green- fluctuations and marker horizons also provides a basis for
land and Antarctic records (Barker et al., 2009). Although comparing the stable isotope records (Figure 7.21 upper).
this approach may be compromised to some extent by But this produces a surprising result: the Antarctic and
ice age–gas age differences in ice cores (section 5.4.3.3), Greenland isotopic variations are in anti-phase with each
comparison of the records shows that there is a close other (Figure 7.21), suggesting that cooling episodes in
correlation between the methane profiles in Antarctica Greenland coincide with warming in Antarctica and vice
and Greenland over the period between c. 10 and 52 ka versa. It is worth noting in passing that a quite different (and
(Figure 7.21 lower). The matched record is currently erroneous) conclusion might have been drawn had the
confined to this period because the age uncertainties for polar ice cores been linked (or tuned) using stable isotope
older layers are much larger (EPICA Community Members, variations alone (section 6.3.3).
2006). In addition to the methane signal, ice-core profiles Consistent anti-phase behaviour between the two polar
can be linked using atmospherically generated cosmogenic ice sheets over several millennia indicates that they are
nuclides (10Be); geomagnetic events, such as the Laschamp coupled to a driving mechanism that transfers energy
excursion at c. 41 ka (Raisbeck et al., 2007); and volcanic alternately between the two hemispheres, which has been
eruptions, such as Mt Toba at c. 74 ka (Williams, 2012; termed the ‘bipolar seesaw’ (Broecker, 1998; Stocker, 1998).
section 6.3.3.3). Central to this process are changes affecting AMOC
-34
LGM MIS3
Susceptibilit
-43
-37
A2
A1
•44
-46
-40
Susceptibilit
Susceptibilit
AIM1 A I M 12
AIMS 48
AfM4 7 ion
-49 .5
-410 4.1 8 9
3
Susceptibilit
AIM2
ACR EDC -52
-430 -52
33
-450 D01
DQ12 NGRIP
10 -1 -37
Susceptibilit
H1 3 4 5 6 7 ESS
2 9
H2 H3 M4 H5
4 ^11
800
s cues pc tei pb ti li ibt i l i t
-45
650
composite
500
S uS
EDML
350
10,000 20,000 30,000 40,000 50,000 60,000
N G R I P a g e (yr B P )
Figure 7.21 Synchronization of isotopic records from Antarctic and NGRIP ice cores between 10 and 52 ka using common
variations in methane (CH4; lower curve). All records have been converted to the Greenland GICC05 timescale. H1–H5 – Heinrich
layers; DO – Dansgaard–Oeschger events; A and AIM – Antarctic Isotope Maximum Events (distinct warming events); ACR –
Antarctic Cold Reversal (after EPICA Community Members, 2006). For further explanation see text.
behaviour: as explained earlier, when the AMOC is closed hypothesis, for if the AMOC was the sole or dominant
down by injections of cold water into the northern North driver, then the opposing polar climate shifts should be
Atlantic, heat can no longer be carried northwards by the broadly synchronous (Seidov et al., 2001). It may be,
Gulf Stream, and is consequently stored in the southern therefore, that while the essence of the theory is correct, as
oceans. Hence cooling in the North Atlantic coincides with with the Milankovitch theory discussed above, other factors
a warming trend in the south. Sustained heating of the serve to influence or modulate climatic response. This is
south would, however, increase evaporation rates and seemingly confirmed by numerical simulation experiments
hence sea-surface salinity, which ultimately leads to re- which reveal the sensitivity of AMOC behaviour to a range
establishment of the AMOC, thereby transferring stored of additional environmental factors. These include the
heat northwards and allowing the southerly latitudes to much larger ocean volume in the Southern Hemisphere
cool. Once initiated, this process of alternating surface compared with the north, the former taking much longer
heat and mass transfer between the hemispheres, which to warm up and cool down; variable rates, durations and
Seidov & Maslin (2001) refer to as ‘heat piracy’, becomes loci of the injection of cold water into the North Atlantic;
a self-perpetuating system, although how it is initiated is similar variations in the timing and volume of freshwater
still not yet fully understood (Severinghaus, 2009; Rial, release into the Southern Ocean from the Antarctic ice
2012). sheet; and atmospheric processes and feedbacks that
The bipolar seesaw hypothesis has gained wide support amplify the effects of ocean circulation changes (Dokken
because it provides an elegant explanation of the anti- & Nisancioglu, 2004; Knutti et al., 2004; Barreiro et al., 2008;
phase relationship reflected in the aligned polar ice-core Swingedouw et al., 2009). In combination, these may
records, and it is a concept that accords with reconstruc- suppress or delay the transfer of energy and mass via the
tions of Atlantic circulation reorganization during the last bipolar teleconnection, in particular slowing the rate of
cold stage (Hoogakker et al., 2007; Barker et al., 2009). response of the Southern Ocean to abrupt climate shifts in
Moreover, simple box models (Stocker & Johnsen, 2003) the north (Severinghaus, 2009).
and some GCM experiments (Manabe & Stouffer, 1997)
reproduce this oscillatory bipolar climatic behaviour when
7.4.4 Global teleconnections: linking
set to run from a fixed set of input variables, suggesting that
it is an integral component of the ocean system during a
mechanisms
glacial stage (Rial, 2012). It could also operate in warmer There is now increasing evidence that the millennial-scale
periods, however, for twentieth-century Arctic and Ant- climate oscillations reflected in the polar ice-core records
arctic temperatures have been shown to vary in an anti- affected not just the Atlantic region, but all parts of the globe
phase seesaw pattern, with warming in the Arctic being (Voelker, 2002; Fritz et al., 2010b). The question, therefore,
accompanied by cooling in the Antarctic and vice versa is precisely how the global climate system responded to the
(Chylek et al., 2010). abrupt climatic shifts of the last cold stage reflected in the
While the concept of the bipolar seesaw has gained North Atlantic DO cycles and, in particular, whether other
widespread acceptance in recent years, there are several regions responded synchronously to these changes, or
aspects of the polar ice-core records that are not adequately whether there were significant leads and lags between them.
explained by this mechanism. For instance, there appear to First, however, it is necessary to examine some of the
have been significant time lags in temperature shifts processes that are involved in the propagation of North
between the two poles, though it is uncertain as to which Atlantic-driven climatic fluctuations to other regions of the
pole leads in the process. Some have concluded that tem- globe, and how these may relate to the operation of the
perature shifts in Greenland preceded the corresponding bipolar seesaw. These include CO2 storage in the oceans,
opposite trends in Antarctica by around 400–800 years displacements of the Inter-Tropical Convergence Zone
(Schmittner et al., 2003), while others have suggested that (ITCZ), atmospheric dust supply and fluctuations in sea
temperature changes in the Antarctic preceded those in level.
Greenland by as much as 1,000–2,000 years (Hinnov et al., It has long been recognized that the oceans have a
2002). Brook et al. (2005) support the view that Antarctica capacity to store CO2 and that the amount stored is
led the changes, but found that the lead time was not dependent on water temperature. High-resolution ice-
constant. Moreover, Capron et al. (2010b) have noted that core and palaeoceanographic records now enable the rate
when the northern ice sheets were extensive, Antarctica of exchange of CO2 between different carbon reservoirs
did not always stay warm during entire Greenland to be quantified, in turn informing the design of GCM
stadials. These observations do not fit well with the seesaw experiments used to test the climatic significance of CO2
variations (Meissner et al., 2007). Changes in the CO2 An example of the kinds of impact that millennial-scale
content of Antarctic Bottom Water (ABW), the largest climate changes could have had on global ocean circulation
single ocean store of CO2, appear to have played an systems and, in turn, on CO2 exchange between the oceans
important role in modulating the bipolar seesaw during and the atmosphere is provided by Bereiter et al. (2012).
the last cold stage, particularly during MIS 3 (Martin et al., They noted an important difference in CO2 and Greenland
2005). A number of factors regulate the CO2 content of temperature in MIS 5a–d on the one hand, and in MIS 3
the Southern Ocean (the source of ABW), including SST on the other: during the former, peak CO2 values lag the
and salinity, marine biomass, deep-ocean temperature and onset of Greenland DO events by around 250 ± 190 years,
circulation, storage of CO2 in the ocean, sea-ice extent and whilst in MIS 3 this lag increases to around 850 ± 90 years,
fertilization of the ocean surface by Fe (derived from dust and peak CO2 values are also higher (Figure 7.22). It
influx). These various components account for about 80 per appears that during MIS 5 a–d, the North Atlantic was not
cent of the calculated millennial-scale fluctuations in CO2 much cooler than during interglacial times (MIS 5e), with
exchange during the last cold stage, which implies that some a circulation similar to that of the present (note that NADW
important sources and processes are not yet accounted for extends to the ocean floor). After the onset of a DO event,
(Fischer et al., 2010b). It is also clear that no single factor therefore, the AMOC remained strong enough to prevent
can explain all of the observed patterns of CO2 exchange the northward spread of ABW. During the colder MIS 3
during the last cold stage, which suggests that different period, however, the starting conditions were very different
factors assumed greater importance at different times. as ABW occupied the whole of the Atlantic deep (Figure
a) Time
Susceptibilit
Susceptibilit
b)
Susceptibilit
NADW NADW NADW
AABW AABW AABW

s N S N S N
<=)
Susceptibilit
GNAIW GNAIW
NADW
AABW AABW AABW

S N S N s N
Atlantic O c e a n
d)
Susceptibilit
MIS 3
MIS 5
Figure 7.22 Schematic representation of Atlantic water mass changes during DO cycles. a) Simplified, schematic asymmetric
temperature oscillation. b) Time slices of water mass distribution and circulation before, during and after a DO event during MIS
5. c) As for b), but representing the sequence of changes during MIS 3. d) Idealized changes in atmospheric CO2 concentration
leading into, and following on from, a DO event during MIS 5 (in black) and during MIS 3 (black plus grey dashed line). NADW –
North Atlantic Deep Water; ABW – Atlantic Bottom Water; GNAIW – Glacial North Atlantic Intermediate Water (from Bereiter
et al., 2012). For further explanation see text.
7.22). This would have restricted the North Atlantic 2006). These provide further confirmation that a southward
overturning to a shallower circuit, termed Glacial North migration of the ITCZ coincided with periods of high
Atlantic Intermediate Water (GNAIW). As a consequence, northern-latitude cooling and, moreover, that North
the surface area of ocean where deep water could exchange Atlantic forcing was a primary driver of millennial-scale
directly with the atmosphere would have been reduced, so monsoonal variability during the last cold stage (Denniston
that the CO2 storage capacity of the ABW was increased. et al., 2013).
At the start of a DO event, when NADW circulation was That these shifts are closely associated with the operation
reinstated, this enhanced deep-water store of CO2 could of the bipolar seesaw tends to be generally assumed, because
then be released (Figure 7.22). The effect would have been they approximate the temporal sequence of millennial-
not only to return much greater quantities of CO2 to the scale climate variations evident in polar ice-core and other
atmosphere compared with MIS 5, but also to prolong that high-latitude records. However, the extent to which low-
release by 500–1,000 years. Since CO2 is a greenhouse gas latitude records align with high-latitude datasets appears to
that mixes and diffuses rapidly in the atmosphere, this vary. In some cases, these appear to match the Greenland
should have had an immediate and significant feedback signals quite closely, for example climatic oscillations
effect on climate at a global scale. reflected in a sediment record from Lake Malawi (Brown
Whether the above example adequately explains all et al., 2007b), and wet phases seemingly related to intensifi-
aspects of CO2 storage and release from the ocean reservoir cation of the South American summer monsoon recorded
during the last cold stage is open to question, but it does in sediments in Lake Titicaca (Fritz et al., 2010b). The
serve to highlight the various linkages and feedback implication is that in these examples, excursions of the
mechanisms involved in ocean circulation, CO2 storage and ITCZ were being driven by climatic variability in
release, and the possible climatic effects. It also highlights Greenland. By contrast, other archives seem to bear the
the important role of the oceans in regulating the carbon imprint of Antarctic signals, such as climatic variations
cycle (Doney et al., 2009). For example, it appears that revealed in South African speleothems (Holmgren et al.,
increased CO2 storage may have led to acidification of the 2003), and in marine sediment records from the subtropical
deep ocean while the shallower layers became more alkaline; Indian and Southern Atlantic Oceans (De Deckker et al.,
the latter process would increase the rate of uptake of 2012). In yet other cases from the low latitudes, however,
atmospheric CO2 by the surface ocean layer and its transfer the imprints of both polar signals have been detected. For
to the deep, thereby enhancing the capacity of the deep instance, planktonic and benthic microfossil records
ocean to act as a reservoir for storing (and subsequently from MIS 3 from the western tropical Pacific indicate that
releasing) CO2 during millennial-scale climatic fluctuations while variations in sea-surface temperatures and salinities
(Sigman et al., 2010). show close affinities with the Greenland rhythm, changes
The second mechanism that is key to understanding in deep-water conditions more closely resemble the South-
global climate shifts over millennial timescales is the ern Ocean–Antarctic sequence (Saikku et al., 2009).
displacement of the ITCZ. Proxy records from marine The position of the ITCZ is also a barometer of changes
and terrestrial sites located within, or close to, the Tropics in position and strength of the monsoon belts. During
suggest that the mean position of the ITCZ shifted the last cold stage, there appear to have been significant
southwards during Greenland DO and Heinrich events, latitudinal displacements of the monsoon cells which, in
which reduced the strength of the East Asian and Indian turn, brought about major regional changes in seasonal and
monsoons (Fritz et al., 2010b; Weldeab, 2012). In a annual precipitation. Such changes, which reflect repeated
speleothem isotopic time-series from northern Australia, shifts in the ITCZ, are evident, inter alia, in the marked
for example, higher summer monsoon rainfall coincides fluctuations in low-latitude lake levels throughout the
with Heinrich stadials and also with the Younger Dryas, tropical zone (section 3.7). The recurrent changes in pre-
while decreased rainfall characterizes the DO interstadials cipitation, which are often accompanied by changes in
(Denniston et al., 2013). This precipitation response is meridional temperature gradients, are indicative of major
anti-phased with sites from around the Indo-Pacific warm reorganizations of atmospheric circulation at the global
pool and with Chinese records of the East Asian summer scale. This may have been driven by (or at least related to)
monsoon. It also finds parallels with anti-phase behaviour the bipolar seesaw. One possible consequence of the
between the Northern and Southern Hemispheres inferred climatic reorganization that follows a shift in the ITCZ
from other records, for example opposing trends in is the replacement of the zonal (latitudinal) pattern of
palaeomonsoon strength inferred from isotopic signals in monsoon belts that dominates at the present time by a more
speleothems from sites in China and Brazil (Wang et al., meridionally dominated circulation (Fritz et al., 2010b).
If so, this would weaken the northern monsoon cells, sudden shifts in the trajectory of dust-bearing winds over
preventing, for example, moisture from the Indian and the polar regions, caused by adjustments in global atmos-
Pacific Oceans reaching the Himalayas and Tibet (Aizen & pheric circulation, but is more likely to be the result of
Aizen, 1996), and leading to greater aridification in those variation in the quantities of mineral dust removed from
regions (Herzschuh, 2006). This, in turn, would lead to an the source regions and advected into the upper atmosphere.
expansion in mid-latitude continental deserts, thereby In addition to the mid-latitude inland deserts referred to
increasing the potential source of global dust. above, which increased in area and dust transport capacity
We have already seen that the dust content in polar ice during colder episodes (Maher et al., 2010; Muhs, 2013),
cores varies inversely with ambient temperature over additional sources of aeolian material include outwash
Milankovitch timescales (section 7.3.3), and that atmos- deposits near glacier and ice sheet margins (Sugden et al.,
pheric dust transport increases during colder periods. 2009) and continental shelves that became exposed during
This relationship also appears to hold true on millennial times of lower global sea level (Gaiero, 2007). A further
timescales. Over Greenland, for example, there was a signifi- factor leading to enhancement of atmospheric dust flux
cant increase in dust flux during the stadial events of the could have been higher wind variability during cold
last 80 ka (Figure 7.23), while the same association between intervals, that is, an increase in frequency and intensity of
dust and cold intervals is also a feature of Antarctic ice- storms due to steepening of meridional temperature
core records from the last cold stage (Fischer et al., 2007). gradients (McGee et al., 2010). Overall, it seems that global
The most remarkable feature of Figure 7.23, however, is the cooling is invariably associated with an increase in aeolian
abruptness of the changes in the dust record: it appears activity, with the higher atmospheric dust loading leading
as though the dust flux to the polar ice sheets could be to important albedo feedback effects (Arimoto, 2001). The
switched on and off quite suddenly. This might reflect abrupt changes in atmospheric dust content indicated in
-35
1
12 20
19
14 17
8 16
Susceptibilit
5 7 10" H5
34 6 13 18
-40 2 I
-45
8000
6000
S u s c e p t i b i l i t S u s c e p t i bS
4000
2000
0
20 40 60 80
A g e / ka b2k { G I C C 0 5 e x t e n d e d )
Figure 7.23 Variations in concentration of terrestrial dust (lower panel) over the last 80 ka shown against δ18O variations, a proxy
for temperature fluctuations (upper panel), in the NGRIP ice core; Greenland DO events are numbered in the upper panel (from
Wolff et al., 2010).
Figure 7.23 suggest that dust emission and flux are closely 36 38 4 0 42 4 4 4 6 4 8 50 52 54 56 58 6 0 6 2 6 4
coupled to other environmental factors, a point we return
a) chronology by correlation to BYRD (+ = ties)
to below. -40
Susceptibilit
The final factor to consider is sea-level change (section
Susceptibilit
-38
-60
2.5). The trend of falling sea level between MIS 5 and MIS
2 (Figure 2.33), which reflects the effects of expanding ice -80 -40
sheets, was interrupted by a number of short-lived rises
S u s c e p t i bSiul ist c e p t i b i l i t
100
-42
in sea level, some of which are linked to the climatic vari- — 3pt (~500yr mov. avg.
ability reflected in the polar ice-sheet records. For example, b) comparison sea-level with GISP2
four distinct fluctuations in sea level of around 20–30 m -40 - H4 H5 • H6? -36
Susceptibilit
Susceptibilit
amplitude during MIS 3 have been linked to millennial- -60
scale warming phases that have been recorded in Antarctic -40
-80
Susceptibilit
records (Siddall et al., 2008; see below), while short-lived
rises in sea level in the China Sea appear to coincide with 100
-44
North Atlantic Heinrich events (Huang & Tian, 2012). c) KL11 co-registered aeolian dust signal 60
Whether these fluctuations in sea level instigated, or 40
Susceptibilit
Susceptibilit
resulted from, ice-sheet mass balance changes is not yet -60
12C
clear, and may in any case be difficult to determine, as they 14

-80 0
are interdependent processes.
Sea-level events in the two hemispheres may be further 100 180
complicated by the operation of the bipolar seesaw. As we Susceptibilit Susceptibilit d) dust signal comparison with GISP2
saw above, under this hypothesis, colder conditions in -40

-36
Susceptibilit
Susceptibilit
Susceptibilit
Greenland were accompanied by warming in Antarctica,
and vice versa. That this was indeed the case is demon- 0
-40
strated by a core sequence (GeoTü-KL11) from the Red Sea,
where the terrestrial dust record matches that in the GISP- 40
-44
2 Greenland ice core (Figure. 7.24), while variations in sea
3 6 38 4 0 4 2 4 4 4 6 4 8 50 52 5 4 56 58 6 0 6 2 6 4
level based on independent proxies match more closely the A g e (ka) to the G I S P 2 a g e model
Antarctic Byrd Station isotope profile (Rohling et al., 2008). of Blunier et al (1998)
Since the different proxies were co-registered in the same
core sequence and their separate signal affiliations are Figure 7.24 Terrestrial dust flux and sea-level reconstruction
consistent throughout the core sequence, it seems reason- for the period 36–64 ka BP based on analysis of proxy records
obtained from core GeoTü-KL11 in the Red Sea. a) Sea-level
able to conclude that a regional terrestrial process (dust fluctuations (KL11; thin line) compared with the corresponding
influx) was responding to the north polar climatic signal, part of the Byrd Station (Antarctic) isotope record (grey shade).
while a global index (eustatic sea-level change) was forced A1–A4 are Antarctic warm periods (AIMS). b) KL11 sea-level
by a south polar signature. A number of implications fluctuations compared with the GISP2 isotopic record (grey
shade); H4, H5, H6 are Heinrich events. c) KL11 sea-level
follow from this. First, the Antarctic ice mass appears to fluctuations compared with the KL11 magnetic susceptibility
have had the largest influence on millennial-scale sea-level data (grey shade), a proxy for dust influx. d) KL11 dust influx
variations, and could have instigated changes that affected record (thin line) compared with the GISP2 isotope record (grey
the north polar region, such as sea-level rise that may shade). Arrows indicate places of particularly strong agreement
have been a factor in triggering Heinrich events (Flückiger (from Rohling et al., 2008).
et al., 2006). Second, the Red Sea record underlines
the importance of being able to discriminate between, and
to quantify, the separate contributions that the south and tuned a sediment record of sea-level change from the
north polar ice masses made to millennial-scale sea-level Mediterranean continental shelf to the Greenland ice-core
variations, but a satisfactory means of achieving this has record, from which they inferred that a series of marine
yet to be developed (Rohling et al., 2004). Third, care flooding events in the Mediterranean was generated by
needs to be exercised when tuning sea-level records to rapid melting of Northern Hemisphere ice sheets. Had
climate curves in order to establish phase relationships the sea-level record been tuned to the Antarctic ice
between them, for the outcome will depend upon which cores, a somewhat different conclusion might have been
polar record is selected. For example, Sierro et al. (2009) reached!
7.4.5 Overview
Although we have focused our discussion on the last cold
stage, it is now apparent that millennial-scale perturbations •global
ice v o l u m e
appear to be a persistent feature of earlier glacial stages also
(Poli et al., 2000; Loulerge et al., 2008; Margari et al., 2010),
which suggests that processes such as the salt oscillator and
the bipolar seesaw are integral components of the earth–
atmosphere system. While some tentative explanations
have been advanced to account for the series of climatic
.hydrologica™ vegetation
changes that took place during the last cold stage, it is clear cover
c y c l e strength
that we are still some way from formulating a theory that
explains all aspects of the climate record for this time
period. A particular difficulty relates to the establishment
of the precise order of events, and hence the determination temperature; dust deposition
of leads and lags in the system. There are a number of
reasons for this. First, there are problems of chronology, for
with the exception of annually resolved ice-core records and
some (relatively rare) varved sequences, the age uncertain-
ties on radiometric dates are often large (Chapter 5). In the
atmospheric C O Southern Ocean
absence of close dating control, the tendency has often mixing ratio
;
productivity
been to tune individual records to selected master sequences
but, as we have seen, this has its problems, because differ- Figure 7.25 Schematic representation of the role of dust in
ent models (Greenland, Antarctic, Hulu Cave, etc.) are the earth feedback system. The components link in three
possible ways: by positive response, where an increase in one
employed in the tuning process, some records contain the
component results in a corresponding increase in another
imprint of more than one climatic rhythm, while in others (grey arrows); by negative response, where an increase in one
there may be compound signals with ‘overprints’ from component leads to a decrease in others (black arrows); and
different regional or global contexts. In relation to the last neutral or no influence. If the sum of the connections between
point, for example, if it is indeed the case that Antarctic components is positive, this gives rise to a positive feedback,
where an initial perturbance is amplified, and vice versa for
warmings are implicated in DO and possibly also Heinrich negative feedback, which dampens the initial change. Solid
events, then North Atlantic and Greenland records reflect arrows in the diagram are primary interactions or links, and
a compound signal showing the influence of both north dashed lines are additional linkages (from Ridgwell, 2002).
(ice-sheet fluctuations, ocean circulation changes, climate
change) and south (ocean circulation change, sea-level
change) polar regions. But if we are to understand fully the of the causes of millennial-scale climatic perturbations
sequence of changes reflected in these signals and establish can seem contradictory, for example the contrasting views
the leads and lags (and causes and effects), then we need over which pole takes the lead in driving the bipolar see-
to improve the precision and accuracy of dating and saw, or the nature and sequence of processes that are
correlation methods, so that records can be compared involved in the generation of Heinrich events. A major aim
more objectively. for Quaternary science over the coming years, therefore,
A second difficulty concerns the relatively limited is to develop more comprehensive databases of high-
number of high-resolution records that extend through the resolution palaeo-records as a basis for meaningful simula-
last cold stage. Although the number is steadily increasing, tions of millennial-scale climate fluctuations operating at
as yet they are barely sufficient to develop the databases that the global scale.
are required to underpin global-scale modelling and A third problem arises from the interdependency of the
simulation programmes (Kageyama et al., 2010). Further- processes and agencies that contribute to millennial-scale
more, much of the published work on millennial-scale climate variation. This point is exemplified in Figure 7.25
oscillations during the last cold stage has focused on which shows how an increase in the earth’s dust flux, once
individual components of the system, such as AMOC initiated, can induce a number of negative responses,
behaviour, the influence of sea-ice cover, or ice-sheet some, for instance, relating to albedo effects and leading to
fluctuations. Maybe because of this, some interpretations an overall negative feedback in the climate system. This is
an example of the close coupling that often exists between undoubtedly the best resolved of any that cover one of these
key environmental variables. But something has to initiate major global climate shifts, and hence provide unique
the change in dust flux in the first instance, perhaps insights into the sequence, spatial pattern and timing of
vegetation change, and that in turn must be responding events that occurred during Quaternary terminations. In
to some other causal factor. Whichever variable we choose, this section, we examine some of the evidence for Termin-
this will always be the case: all of the variables are mutually ation 1, and consider the implications for our under-
interacting in some way, and hence it is difficult to isolate standing of how these major climatic transitions occurred
one as the genuine instigator of change. It could be argued at the regional, hemispherical and global scales.
that astronomical forcing, which is external to the earth
system, initiates change at Milankovitch timescales, but, as
we have seen, millennial-scale climate fluctuations during
7.5.2 Definition of the Last Termination
the last cold stage appear to have been induced by internal The term ‘Last Termination’ has not so far been formally
regulatory mechanisms, such as the bipolar seesaw, and defined, but is widely used to refer to the interval between
hence could reflect some form of what has been termed the Last Glacial Maximum (LGM: also not formally defined,
stochastic resonance2 (Claussen et al., 2003). Once the but generally acknowledged to have occurred around 21 ka:
resonance is set, then it is difficult to single out any one Mix et al., 2001) and 11.7 ka, the onset of the Holocene
point in the sequence as being of particular causal signifi- (Walker et al., 2009). This time interval includes the period
cance, since all components in the process are respond- widely referred to as the Lateglacial (or Late Glacial), and
ing to precursor conditions and events. The challenge, which, from a Northern Hemisphere perspective at least,
therefore, is to establish how the cycle starts in the first has been considered to encapsulate the Last Termination.
place, the key feedbacks that keep the cycle going, and how Indeed, the Lateglacial, the onset of which is marked in
quickly global teleconnections operate, that is, how rapidly many Northern Hemisphere proxy records by an abrupt
the changes in climatic conditions are transmitted through rise in temperature at ~14.7 ka (e.g. Walker et al., 2003;
the global system. To answer these questions, strategic Steffensen et al., 2008), has frequently been referred to as
collaboration between the palaeodata and climate model- the Last Glacial–Interglacial Transition (e.g. Björck et al.,
ling communities will be required, and the former will need 1998).
to supply time-series of palaeoenvironmental data based on As we saw in Chapter 1 (section 1.6), the Lateglacial
improved age models. This may seem a tall order, but may was initially subdivided on the basis of inferred climatic
not be beyond the bounds of feasibility! events in northwest Europe. An initial amelioration (the
Bølling Interstadial) was followed, in turn, by the short-
lived Older Dryas (OD) cold oscillation, a further mild
climatic period, the Allerød Interstadial (interrupted by
7.5 THE LAST TERMINATION the short-lived Inter-Allerød Cold Phase: IACP) and
the much colder Younger Dryas (YD) Stadial; this last
7.5.1 Introduction episode spanned the time interval from 12.9–11.7 ka,
At several points in this book, we have made reference to and was terminated by the abrupt warming at the start
terminations, the abrupt shifts in the marine oxygen iso- of the Holocene (Figure 7.26). This terminology formed
tope signal that mark the change in the global climate the basis for a climato-stratigraphic subdivision of terres-
system from glacial to interglacial mode (sections 3.10.2 trial records not only from northern Europe but, some-
and 6.2.3.5). Although the terminations are generally what paradoxically, from many other regions for which
considered to reflect a combination of astronomical forc- the subdivision was never initially intended (see below). It
ing and associated climate feedback processes (Schulz was, moreover, never related to a type site or stratotype
& Zeebe, 2006; Cheng et al., 2009), the relatively low reso- (section 6.2.2). As explained previously (sections 1.6 and
lution of the records means that the precise sequence of 6.3.3.3), however, the Bølling–Allerød–Younger Dryas
events and the interplay between the different components terminology has, to a large degree, been superseded by
of the earth–atmosphere system during the course of an event stratigraphy based on the oxygen isotope signal in
each termination is not well understood. The situation the Greenland cores, with the NGRIP record now con-
is very different for the Last Termination (Termination 1), stituting a stratotype sequence for the Last Termination in
the transition from the last cold stage (MIS 2) to the the North Atlantic region (Lowe et al., 2008b). The parallel
present interglacial (MIS 1: the Holocene). The various set of isotopically defined terms that is now increasingly
stratigraphical records that span this time interval are being employed for events during the Last Termination,
YD Allerad Belling
2 (GS-1) (Gl-1a-c) (Gi-ie) (GS-2)
Susceptibilit
North G r e e n l a n d
Antarctic c o m p o s i t e
1
0
ep
S u s cS
-1
ACR
-2
11 12 13 14 15 16 17 18 19 20
A g e (ka b2k. G I C C 0 5 )
S u s t a i n e d warming in Antarctica P a u s e in Antarctic warming Significant Antarctic cooling
Figure 7.26 Isotopic records from Antarctica (black line: a composite of four Antarctic ice-core records) and Greenland (NGRIP,
grey line) for the Last Termination, defined here as the interval between 20 and 11.7 ka. Vertical light-grey shading denotes an
episode of sustained warming in Antarctica, the vertical medium-grey band a pause in Antarctic warming, and the dark-grey band
a significant Antarctic cooling (the Antarctic Cold Reversal: ACR). The NGRIP record shows the marked warming in Greenland at
c. 14.7 ka that defines the start of the Lateglacial period. b2k – years before AD 2000 (from Pedro et al., 2011). For further explanation
see text.
based on Greenland stadial (cold: GS) and interstadial the two hemispheres experienced very different climatic
(warm: GI) units, is shown in Figure 7.27. In this scheme, histories during the Last Termination. Whereas conditions
GS-2a denotes the cold period between the LGM and the remained cold in the north until the abrupt rise in
onset of Lateglacial warming (sometimes also referred to temperatures at c. 14.7 ka, warming in Antarctica began
as the Oldest Dryas period), while the succeeding warmer perhaps as early as c. 19 ka, and was more gradual. Other
period encompassing the Bølling and Allerød interstadials differences between northern and southern records include
equates with GI-1 in the NGRIP sequence. GI-1 is a significant period of cooling in Antarctica, the Atlantic
subdivided into five subunits on the basis of inferred Cold Reversal (ACR), which coincided with the warmest
climatic variability: GI-1e broadly corresponds with the phase of the Lateglacial period in the north; the absence of
Bølling, GI-1d with the OD and GI-1c to GI-1a with the a reciprocal change in Antarctica when the IACP (GI-1b)
Allerød, the IACP seemingly marked by a short-lived cooler affected Greenland; and gradual warming in Antarctica
episode (GI-1b). GS-1 corresponds with the YD Stadial. during the North Atlantic YD (GS-1) cold event. In this
Strictly speaking, however, these terms are not synonymous, context, therefore, the YD/GS-1 cold episode, which had
nor time-parallel, and should not be assumed to be so, as such a profound effect on the environment of large areas
explained in section 7.5.5. of North America and Eurasia, can be seen as a uniquely
Although the NGRIP ice core constitutes a viable Northern Hemisphere event. Moreover, while from a north
regional (and indeed hemispherical) stratotype for the Last polar perspective, the marked rise in temperatures at 14.7
Termination in the North Atlantic region, it is evident from ka constitutes the first unequivocal sign that the last glacial
Figure 7.26 that it cannot represent the sequence of events period had ended, this major event is signalled much earlier
in both the north and south polar regions. This is because in the south, and accords with the marine isotope evidence
Events Current recommendation

Depth in NGRIP GICC05 age
NGRIP GRIP GICC05 age core (m) (yr b2k)
depth (m) depth (m) ka b2k
3.2 ka BP event 1219.47 8140
Even is
1000 Volcanic peak inside 1228.67 8236
and episodes
8.2 ka
1200
3.2 ka BP event 1234 78 8300
1200 9.3 ka BP event 1322.33 9240
5.2 ka event
9.3 ka BP event 1331.65 9350
1400 PBO even!
9.3 ka event
Start of Holocene 1492.45 11703
1500 It Start of GS-1 1526.52 12696
1400
Start of G l - l a 1534.5 13099
Pre tores I Oscillation Start o f G M b 1542.1 13311
1500 Start of G l - t c 1570.5 13954
1600 GS-1 Start of Gl-1d 1574.80 14075
Gl-Ta Start of Gl-1e 1604.64 14692
t)
Gl-lc Start of GS-2a Feature not clear in NGRIP profile
d
1600 1700 Gl-1e
15
GS-2a
1800
-46 -44 -42 -40 -33 -36 -34

6 0 (permit)
1 6
Figure 7.27 Stratotype scheme for the Lateglacial–early Holocene in the North Atlantic region. The figure on the left shows the
NGRIP (blue) and GRIP (red) isotope records and corresponding isotopic events and sub-events: GI – Greenland Interstadial; GS
– Greenland stadial. The table on the right defines the boundaries between each isotopic event in years before 2000 (b2k) (from
Lowe et al., 2008b).
that points to a significant decline in global ice cover from and Southern Hemisphere insolation changes suggest that
c. 20 ka onwards. On a global scale, therefore, the Last temperatures in the southern Pacific (both deep-water and
Termination did indeed begin as early as early as 19 or 20 sea-surface temperatures: SSTs) started to warm around
ka, and markedly predated the beginning of the Northern 1 ka before warming of tropical Pacific SSTs and the rise in
Hemisphere Lateglacial. atmospheric CO2 content as reflected in the ice cores
(Figure 7.28). This would appear to rule out increased
atmospheric CO2 or tropical climate circulation as the
7.5.3 Onset of the Last Termination forcing agents that initiated the onset of the last termination
In section 6.2.3.5 we noted that a distinctive feature of the (Peeters et al., 2004; Timmermann et al., 2009). One
deep-ocean marine isotope profiles is the ‘saw-tooth’ plausible scenario is that increased solar radiation over
appearance of the curves reflecting the fact that while Antarctica initiated a retreat of sea ice which, in turn, led
glacial inceptions are usually progressive over tens of to increased ventilation of the Southern Ocean, and
thousands of years, the transitions from full glacial to full therefore higher atmospheric CO2 levels, which enhanced
interglacial conditions (the terminations) occur within the warming trend (Stott et al., 2007).
maybe 10 ka or less. Why, after a long period of sustained Others have suggested, however, that the onset of the
cold climate, glaciations should end so abruptly remains a Last Termination was triggered by processes operating in
puzzle, because the orbitally forced increase in solar the north. Denton et al. (2010), for example, note that as
insolation (Figure 7.28) cannot alone account for such a the northern ice sheets reached their maximum extent
sudden and pronounced climatic shift (Denton et al., during the closing stages of the last glacial period, the
2010). One possibility is that the trigger may lie in the Greenland stadial events became more prolonged (see
Southern Hemisphere, because of the early manifesta- Figure 7.15). This in turn would have led to longer periods
tion of an increase in temperatures shortly after the LGM of warming of more southerly latitudes (because of the
(Stott et al., 2007). For example, comparisons between effects of the bipolar seesaw: section 7.4.3). Warming was
Antarctic ice-core records, Pacific Ocean marine records further induced by an associated increase in ventilation in
270
230
Insolation 65°S
Susceptibilit
280.0
Susceptibilit
250
2400
-V—- i
EP1CA Dome C C 0 2
30
240
200.0
Susceptibilit
MD81
16 Tropical Pacif c S S T 2S MD7E
MD70
Susceptibilit
14
26
12
South Pacific S S T
3 5
O O P Site 1233
10
" 0.0
4 0
Susceptibilit
-2.0
Deep Pacif c
MD81 Benthic
Susceptibilit
-4.0
Dome Fuji 2C
4.5
Temperature
-GO
6.0
-s.o
•10.0
10.000 12,000 14,000 1S.O0O 18,000 20,000 22.000
Year B P
Figure 7.28 Temperature variations in the Pacific Ocean at the surface (SST) and at depth (benthic) during the Last Termination
inferred from foraminiferal assemblages, shown against ice-core CO2 data (EPICA Dome C), ice-core temperature reconstructions
(Dome Fuji) and summer insolation in the Southern Hemisphere (upper curve). The vertical grey shading highlights the difference
in timing between the onset of deep-water warming (close to 19 ka) and Tropical Pacific SST warming. The grey shading on the
MD81 benthic curve represents a 200-year uncertainty on the ages obtained from benthic foraminiferal samples (from Stott et
al., 2007). For further explanation see text.
the southern oceans and in the release of CO2 to the in this case on the analysis of variations in strength of the
atmosphere. These combined effects led eventually to the northern monsoon. In their scenario, periods of northern
transgression of a critical threshold, when stored warmth ice-sheets collapse and associated sea-level rise coincide with
in the south grew to a level sufficient to prevent a return times of low monsoon strength and an increase in
to stadial conditions in the north, save for some short-lived atmospheric CO2. The relative strength of the northern
reversals such as the IACP and YD (see above). Cheng monsoons appears to have been a critical factor in the
et al. (2009) arrived at a broadly similar conclusion, based prolongation of stadials in the North Atlantic region, again
a)
i l ti ti b i l i t
i bui lsi tc e p t i b i l i t
260 0
bp
1
c eups tci e
ptibilit
240 -1
t u sS
i l iut s c e p t S
it
Stui bs icl e
b iul ist c e p t i b i l iS
0
220 -2
Suscep
S u s c e p t i bS
200
-3 -1
SusceptiS
180
-4 LGM OD B-A YD Holocene
22 20 18 16 14 12 10 8
b) Age (ka)
1
u isl ci te p t i b i l i t
O n s e t of
0.75
S u s c eSputsi bc iel ipt t i b i l i t
C0 2 rise
O n s e l of
tib
60-90° N
S u s c e p t i bSiul ist c e p S
0.5
seesaw 30-60° N
0-30° N
0.25 0-30° S
30-60° S
C 60-90° S
-0.25
22 20 18 16 14 12 10 8
Age (ka)
Figure 7.29 a) Global temperature stack for the Last Termination based on eighty globally distributed records (blue curve),
compared with temperature (red curve) and atmospheric CO2 concentration (yellow dots) records from Antarctic ice-cores. All data
are shown with 1σ error ranges. LGM – Last Glacial Maximum; OD – Oldest Dryas; B–A – Bølling–Allerød interval; YD – Younger
Dryas. b) Mean temperature variations (with 1σ error ranges) for stacks of proxy records arranged in 30° latitude bands. The mean
values have been normalized to fractions or percentages of the total glacial–interglacial (G–IG) temperature range (from Shakun
et al., 2012, reprinted with permission from Macmillan Publishers Ltd: Nature (Shakun, J.D., Clark, P.U., Feng, H. et al.) copyright
2012). For further explanation see text.
allowing longer periods of CO2 release from the southern Northern Hemisphere records at around 19 ka, preceding
oceans, and resulting in an amplification of the solar signal the more marked warming in the Southern Hemisphere,
and the demise of the last great ice sheets. again implicating events in the Northern Hemisphere as a
A compilation of global temperature records for the Last possible trigger for the Last Termination (He et al., 2013).
Termination (Figure 7.29) suggests that air temperature A possible sequence of events during the Last Termin-
over Antarctica had started to rise by at least 18 ka, before ation is shown in Figure 7.30. An increase in boreal summer
atmospheric CO2 concentrations rose, but then increased insolation beginning c. 22 ka (1 in Figure 7.30) caused the
in concert with CO2 after c. 17 ka (Figure 7.29a). The com- northern ice sheets to melt, reducing surface reflectivity,
posite for mean global temperature, however, clearly allowing Greenland to warm (2). Meltwater released from
lags these developments during the period 18 to c. 11 ka, the wasting ice sheets and glaciers caused a shutdown in the
except during the Bølling–Allerød (B–A) interval. All three Atlantic Meridional Overturning Circulation (AMOC),
curves converge near the start of the Holocene. When the and a cooling in the North Atlantic. Southwards heat
data are resolved into latitudinal belts, however, it can be transfer led to a warming of Antarctica and the Southern
seen that the global composite temperature curve shown Oceans (3) and subsequently the release of CO2 into the
in Figure 7.29a is biased by the data obtained from sites atmosphere (4). A consequent increase in global-scale
located between 30°N and 90°N (Fig. 7.29b). These warming accelerated northern ice-sheet melting, keeping
partitioned data show that the first net increase in tem- the north cool, the south warm and further CO2 release
perature, albeit small, is recorded in the high-latitude (a positive feedback loop). Ultimately, this led to the
250
Susceptibilit
Insolation
1 0
200 c o 4
z
-5
I n s o l a tI inosno l a t i o n
4
-10
S u s c e p tSi buisl ict e p t i bS

2
-15
-20
Greenland temp.
-25
24 22 20 18 16 14 12 10
A g e (ka)
Figure 7.30 Hypothesized chain of events during the Last Termination; the sequence (1–4) is explained in the text (from
Severinghaus, 2009).
termination of the last glacial stage (Severinghaus, 2009). the course of the Last Termination. In the Southern
This scenario, based initially on empirical evidence, has Hemisphere, for example, records that are closely in-phase
been supported by transient simulation modelling which with the Antarctic climate rhythms include glacier oscilla-
suggests that the last deglaciation was initiated by rising tions in the Southern Alps, New Zealand and Patagonia,
insolation during spring and summer in the mid- and including a marked readvance during the ACR (Putnam
high northern latitudes. AMOC changes associated with this et al., 2010; Garcia et al., 2012); sea-surface temperature
orbitally induced retreat of Northern Hemisphere ice sheets (SST) variations along the Chilean coastal margin (Lamy
led to early Southern Hemisphere deglacial warming, with et al., 2004), the tropical Indian Ocean (Naidu & Govil,
the ensuing rise in atmospheric CO2 providing the critical 2010), the Arabian Sea (Huguet et al., 2006) and around
feedback for global deglaciation (He et al., 2013). South Australia (Calvo et al., 2007); air temperature records
from South Africa (Holmgren et al., 2003); and palaeo-
hydrological trends in the Dry Valley lakes of Antarctica
7.5.4 Global teleconnections during the
(Hall et al., 2010). Throughout most of the Northern
Last Termination Hemisphere, by contrast, palaeoenvironmental records
In previous sections, we examined the evidence for tend to align closely with the Greenland ice-core climate
teleconnections between the polar regions, monsoon cells, signal. Examples include variations in strength of the
the Inter-Tropical Convergence Zone (ITCZ) and other Indian, Asian and African monsoons (Chang et al.,
elements of the global circulation system over the last 2008b); lake-level variations in North America (Benson
glacial cycle. Here we consider further evidence for the close et al., 1998); SST variations and circulation in the central
coupling of these regional elements of global climate during Mediterranean Sea (Cacho et al., 1999) and tropical
the Last Termination. Atlantic (Lea et al., 2003); air temperature variations in
As we saw in section 7.4, a distinctive feature of the Japan (Kossler et al., 2011); and glacial and aridity cycles
ocean–atmosphere system over the course of the last glacial in the Spanish Pyrenees (González-Sempériz et al., 2006).
cycle was the operation of the bipolar seesaw, and this also Some low- and mid-latitude regions, however, appear
appears to constitute a working hypothesis for the pattern to have been affected by both Northern and Southern
of environmental changes in the two hemispheres during hemisphere factors. For example, warming of the Cali-
fornian margin of the North Pacific appears to have potentially dated more precisely, than is the case for earlier
predated the start of the GI-1e warming in Greenland by Quaternary phases (Walker, 2001). Realizing the chrono-
c. 2 ka, but post-dates the start of warming in Antarctica, logical potential of these records, however, has proved
perhaps reflecting both northern and southern circula- to be a major challenge, for the ice cores and their
tion influences in this region (Hill et al., 2006). Similarly, accompanying timescales set new standards in terms of
large-scale retreat of mid-latitude LGM glaciers in both stratigraphic resolution and chronological precision that
hemispheres seems to have commenced at broadly the other archives have not always been able to match. In this
same time, but the timing of retreat is consistent with section, we consider the development of new time-
the onset of temperature and CO2 increases in Antarctic stratigraphic frameworks for the Lateglacial that are
ice cores, suggesting a Southern Hemisphere driver for increasingly enabling us not only to distinguish spatial
deglaciation in some Northern Hemisphere regions and temporal patterns of environmental change during
(Schaefer et al., 2006). There is also evidence suggesting the this key interval of the Quaternary record but, equally
operation of an Atlantic–Pacific seesaw teleconnection importantly, to begin to quantify leads and lags in climate
during the Last Termination, with deep-water formation change and accompanying environmental response.
in the northern Pacific when the AMOC was closed down
in the North Atlantic (Okazaki et al., 2010). More puzzling
7.5.5.2 Lateglacial stratigraphy and
are some records from the Southern Ocean that appear to
chronology
accord more closely with the Greenland climatic record, for
example a cooling phase in the Great Australian Bight The subdivision of the Lateglacial outlined above (section
and the South Atlantic that was contemporaneous with the 7.5.2) was based originally on pollen-based biozones
GS-1 event (Kim et al., 2002; Andres et al., 2003). Clearly, (section 6.2.3.2) identified in southern Scandinavia (the
therefore, some events during the Last Termination cannot ‘Norden model’: Mangerud et al., 1974). The pollen zone
be accounted for by the bipolar seesaw alone (Seidov et al., boundaries were dated by radiocarbon, and hence assumed
2005). Indeed, despite the numerous proxy records now the status of chronozones (section 6.3.7). This has since
available from the Last Termination, it is not clear how proved problematical, however, partly because of errors
climate changes initiated at the poles are propagated that can affect radiocarbon dates (section 5.3.2.4), but
throughout the global climate system. There is, neverthe- also because biozone boundaries (which reflect response
less, a considerable body of evidence to support the view of vegetation to climate/environment thresholds) are
that the bipolar seesaw remained a key element of the time-transgressive, whereas chronozone boundaries, by
ocean–atmosphere system during the Last Termination definition, are time-parallel (Walker, 1995). Indeed, this
(e.g. Stenni et al., 2011), and we return to this matter was one of the principal reasons why the INTIMATE
below. Working Group (section 6.3.3.3) advocated replacement of
the Norden Model by the Greenland event stratigraphy
(Björck et al., 1998). Yet the terms Bølling, Older Dryas,
7.5.5 Synchronizing records of Lateglacial
Allerød and Younger Dryas (section 7.5.2; Figure 7.26) were
age widely adopted, and are still applied today, often in parts
of the world distant from Scandinavia, to stratigraphic
7.5.5.1 Introduction
units that are considered to approximate, in time or climatic
Thus far, we have been considering the Last Termination characteristics, those in the Norden Model (Peteet, 1995).
in its broadest sense; in other words, the whole time period The use of the term ‘Younger Dryas’ (YD) to define climato-
from the end of the LGM to the onset of the Holocene. We stratigraphic units in sites remote from northwest Europe,
now turn our attention to the Lateglacial, that 3 ka interval especially those in the high-latitude Southern Hemisphere,
of the Last Termination when, in the Northern Hemisphere appears particularly incongruous, given that that part of
at least, profound and often extremely rapid changes the globe was warming at the time of the YD cold period
occurred in the climate–environment system. in the north!
The study of the Lateglacial period has a long history, As we saw in Chapter 5, the Greenland ice-core record
and it is undoubtedly one of the most intensively investi- is underpinned by the high-resolution GICC05 timescale,
gated episodes in the entire Quaternary record. This is an annually resolved record based on multiparameter
because the geomorphological, lithostratigraphical and layer counting, with a maximum counting error (MCE)
biostratigraphical evidence from the Lateglacial can be in the Lateglacial of ~3 per cent (section 5.4.3.3). Thus at
examined at a much higher temporal resolution, and the GS-1–Holocene boundary, the MCE is of the order of
99 years, while at the onset of the Lateglacial (onset of to the continuous dendrochronological record, calibration
GI-1e), it is 190 years (Rasmussen et al., 2006). It is worth may in due course be extended back to the early Lateglacial
recalling that the MCE is broadly equivalent to a 2σ error (GI-1e: Kaiser et al., 2012). Other developments involve
in a Gaussian probability distribution, and hence the MCE the use of more reliable dating media, for example through
in the GICC05 chronology equates approximately to a the careful targeting of fossil types or chemical fractions
68 per cent probability value (1σ) on a conventional radio- most likely to yield coherent results (Blockley et al., 2007a;
carbon date. At the GS-1–Holocene boundary, therefore, Howarth et al., 2013), and by refining the procedures
the MCE of 99 years corresponds broadly to a c. 50-year employed for sample pretreatment and 14C activity
error on a radiocarbon date. However, many dates from measurement so that samples with low 14C content yield
the Lateglacial have standard errors that are in excess of the more accurate age estimates (Watanabe et al., 2009).
MCE in GICC05, and hence, until comparatively recently, A second important advance has been refinements in
there were considerable chronological uncertainties in age tephrochronology, especially the employment of crypto-
models for the Lateglacial sequence, sometimes millennial tephra layers (section 5.5.2.1) to identify distal volcanic ash
in scale (Lowe et al., 2007). This meant that the degree of horizons and to use these to correlate Lateglacial records
correspondence between Lateglacial terrestrial and marine and to synchronize climatic events (Figure 7.31). Some
sequences, and between these archives and the Greenland distal tephras detected in marine cores are also found in
ice cores, could not be rigorously tested, and correlations continental records, and so terrestrially based age estimates
therefore tended to be assumed rather than convincingly can be imported into marine sequences, circumventing
demonstrated. This was especially the case where sequences the problems caused by marine reservoir effects on 14C dates
were either poorly dated or where radiocarbon dates were (section 5.3.2.5) and thereby ensuring more secure marine–
lacking. In these instances, alignment of climatically derived land correlations (Thornalley et al., 2011). Of particular
signals from fossil records with the oxygen-isotope record value are those tephra layers that are co-registered in the
from the Greenland ice cores became standard procedure polar ice cores and in continental or marine records, for
(section 6.3.3.2). But this brought with it additional they provide well-dated tie-points. For example, the widely
problems, particularly in terrestrial sites (lakes and bogs) dispersed Vedde Ash, which originates in Iceland (Katla
where sedimentation was slow and hence stratigraphic volcano), is found in the Greenland ice-core sequences
resolution was poor, for contemporaneity of events could (Figure 5.26), where it is dated on the GICC05 timescale
never be conclusively demonstrated. to 12,171 years with an MCE of 114 years (c. 57 years at 1σ),
In recent years, therefore, there has been a concerted placing it within the YD/GS-1 Stadial (Lane et al., 2012).
attempt to refine dating procedures and to establish It occurs in marine sediments in the North Atlantic, and
stratigraphic protocols in order to provide a more secure can be traced throughout much of Europe, from the British
basis for correlating between, and synchronizing, Lateglacial Isles in the west to the Ural Mountains in the east, and from
records. We examine some of these initiatives in the Norway to as far south as Slovenia and North Italy (Lane
following sections. et al., 2011). It therefore constitutes a key time-stratigraphic
marker horizon for the Lateglacial across a large area of the
Northeast Atlantic region.
7.5.5.3 Lateglacial age models and correlation
A third key innovation has been the analysis of
procedures
increasing numbers of varved sediment records and the
A number of parallel developments have helped to refine construction of varve chronologies (section 5.4.2). Since
Lateglacial age models and the synchronization of ice- these are based on annual-layer counting, they perhaps
core, marine and continental records. The first of these come closest to ice-core records in terms of the potential
is the gradual fine-tuning of the IntCal radiocarbon they offer for discriminating environmental changes
calibration model (section 5.3.2.6), which has significantly with a high temporal resolution (e.g. Litt et al., 2001;
reduced the uncertainties of calibrated ages within the Brauer et al., 2008). Some Lateglacial varved records also
Lateglacial time frame. For example, although the latest contain distal tephra layers (Blockley et al., 2007b), which
iteration, IntCa13, employs a tree-ring dataset that extends have proved particularly valuable for refining the timing of
back to 14.2 ka BP, the dendrochronologically based regional events, and showing how they align with the polar
calibration model has been terminated at 13.9 ka because ice-core records.
of sparse measurements in the early part of the record The fourth important development has been the
(Reimer et al., 2013). However, if these problems can be emergence of statistical tools that aid the evaluation of
resolved, and currently floating chronologies can be linked complex chronological data and form the basis for the
a ) 6 O uariations
1£i
b) Abernethy Forest c) Loch Ashik d) WNtrig Bog e)Kr£kenes Tephra
recorded in NGRIP (this study) (this study) (Brooks and Birks, (Brooks and Birks, layers
Rasmussen et al. (2006) 2000) 2000)
Chiron Dm id-inferred mean July air temperatures f C)
5 O (%«)
, e
-44 -42 -40 -38 -36 -34 4 6 e 10 12 14

33C
4 6 8 10 12 14 4 6 e 10 12 14
520 65 C Saksunarvatn Ash
10.5 520 630 10 3510 089
54 0 700 GIGC05 ka BP
720
34 C 74 C
4 6 8 10 12 14
563 60 76 C
11.5
780
550 80
800
530 Vedde Ash
100 320
12.1210.114
SBC oft
120 GICC05 ka BP
12.5 600 mo
140 680
620 600 900
180
920
94 a
13.5
64 0 620
z 960 Penifiler Lephra
14.09-13.65
cal ka BP
66 a 640 220
Borrobol tephra
14 5 240 14 14-13 96
660 660 calka BP
260
700 730
Figure 7.31 Quantified palaeotemperature records for the Lateglacial and early Holocene from three sites in Scotland and one
(Kråkenes) in Norway aligned using four tephra isochrons. The temperature reconstructions are based on the analysis of fossil
chironomid assemblages. Two of the isochrons, the Saksunarvatn Ash and Vedde Ash, occur in the Greenland ice cores (left) and
are therefore dated by ice-core chronology. The ages of the Borrobol and Penifiler tephras are based on calibrated radiocarbon-
based ages. The grey shading indicates relatively warmer phases (Bølling, Allerød, early Holocene). The timescales for b)–e) are
in ka cal. BP (from Brooks et al., 2012). For further explanation see text.
construction of age models (section 5.3.2.6). These take into is that individual site records should be subdivided on the
account all information concerning the stratigraphic order basis of local or regional criteria, and should be assigned
of dated samples and additional evidence relating to the local stratigraphic terms, with ages and durations deter-
stratigraphic sequence and its context, and combine this mined by independent dating, prior to correlation with the
information using Bayesian probability procedures ice-core records. It must be recognized, however, that the
(Blockley et al., 2007a; Bronk Ramsey, 2008). The models NGRIP stratotype is applicable only to the North Atlantic
can be adjusted using different assumptions (e.g. about province, and will not provide a suitable template for
changes in sedimentation rate, possible hiatuses in the regions that have experienced a quite different sequence of
sequence) until an optimized age–depth profile has been climatic events during the Last Termination. Finally, a
achieved. This approach is particularly useful where more common misconception is that INTIMATE has simply
than one chronological method, for example radiocarbon replaced one set of general stratigraphic terms (Bølling,
dating and tephrochronology, is being employed. Allerød etc.) with another (GI-1e, GS-1, etc.). This is not
Finally, there has been the formulation of protocols for the case, as that defeats the purpose of stratigraphy and
the stratigraphic subdivision and correlation of Lateglacial correlation, which is to test how each sequence has evolved,
sequences. As we explained in Chapter 6, the INTIMATE and how the inferred climate signal correlates with that in
Group has proposed that the NGRIP ice-core record should other stratigraphic records.
be the stratotype for the Lateglacial period for the North
Atlantic region (section 6.3.3.3), and that the ages of the
7.5.5.4 Rapid environmental change during
event boundaries should be dated using the GICC05
the Lateglacial
timescale (Figure 7.27; Lowe et al., 2008b; Blockley et al.,
2012). While the NGRIP isotopic signal constitutes the There are strong indications in the ice-core records that the
regional stratigraphic template, the INTIMATE proposal bipolar seesaw was influential in determining the pattern
of climate changes during the course of the Lateglacial record from Lake Suigetsu in Japan (section 5.4.2.7)
(Stenni et al., 2011) although, as we saw above, there suggests that the onset of the warming at the beginning of
remains some uncertainty over how, and how rapidly, the the Lateglacial occurred 300–500 years earlier than in
climate signals were transmitted from high- to low-latitude Greenland (Figure 7.32; Nakagawa et al., 2005), suggesting
regions. Here we examine some recent evidence of the that other areas of the Northern Hemisphere were
processes involved, and of their response times. The focus experiencing climate shifts that differed significantly in
will be on the Northern Hemisphere and on three key character and timing from those in the North Atlantic.
events or episodes: (1) the marked warming at the start of Peak thermal conditions in Greenland during GI-1e
GI-1e (Bølling); (2) the cooling trend between GI-1d and lasted for perhaps only 100 years or so, before tempera-
GI-1a; and (3) the YD Stadial (GS-1; Figure 7.26). tures started to decline. A long-term cooling trend between
Prior to the GI-1e event, the North Atlantic region c. 14.5 and 12.9 ka was interrupted by several short-
experienced markedly colder conditions with the shutdown lived, and colder excursions. Two of these are defined in
of the AMOC and widespread formation of sea ice leading the Greenland stratotype sequence (Figure 7.27) – events
to more severe winter conditions, not only in the North GI-1d (onset 14.075 ka; MCE 169 years; duration 121
Atlantic region but also in Asia (Denton et al., 2005). This years) and GI-1b (onset 13.111 ka; MCE 149 years; duration
period of extreme cold was followed by the re-establishment 203 years) – while a third, of much shorter duration (< 100
of the AMOC and rapid warming at the onset of the years), is sandwiched between them. In each of these cases,
Lateglacial (GI-1e). Precisely what caused the shutdown both the cooling and subsequent warming associated with
and subsequent re-establishment of the AMOC remains each of these events appears, from the Greenland isotope
unclear, but it could reflect retreat of sea ice in the Southern signal at least, to have occurred relatively rapidly. Many
Ocean leading to export of heat northwards (Bianchi & palaeoclimatic records from lake sediment sequences
Gersonde, 2004), or rapid melting of the Antarctic ice around the North Atlantic region tend to show a similar
sheet leading to sudden sea-level rise which destabilized pattern of climate variation during the Bølling–Allerød
the ice margins in the north (Deschamps et al., 2012). interval (GI-1), although the number of clearly identified
Alternatively, it may have been triggered by events in the cold intervals varies between sites. Frequently only one
Northern Hemisphere, most probably the release into the of these events can be detected, usually the Older Dryas
northern oceans of large quantities of meltwater from (GI-1d) in European sites, but the Inter-Allerød Cold
the wasting ice sheets. As we saw above, the first signs of Phase (GI-1b) is evident in some North American records
warming in the northern high latitudes occurred c. 22 ka, (Yu & Wright, 2001). In a number of high-resolution
and this may have been the trigger for ice-sheet wastage. records from Europe, however, up to four such cold epi-
Oceanographic records from the North Atlantic certainly sodes have been recognized (Brauer et al., 1999; van Raden
show that the onset of meltwater release from the collapsing et al., 2013). There may be a number of explanations for
ice sheets began c.19 ka, and was followed by the deposition this. Slow rates of sediment accumulation in many lake
of ice-rafted debris (Heinrich event 1) between c. 17.5 ka basins could have compressed the stratigraphic record and
and 16.5–15 ka. This appears to have been accompanied made these short-lived climatic phases difficult to detect,
by a ‘pooling’ of freshwater in the Nordic Seas which was a problem that may have been further compounded by
subsequently purged into the North Atlantic, thereby inadequate sampling resolution. In other cases, the climatic
preconditioning the Nordic Seas for convective deep- signals might not have registered in proxy data, either
water formation (Stanford et al., 2011). This, in turn, because of the low amplitude of the climatic fluctuations,
allowed an abrupt restart of the AMOC and North Atlantic or because the changes in temperature and/or precipitation
Deep Water (NADW) formation at the GI-1e warming regime were insufficient to overcome critical geomorphic
(14.7 ka). Greenland ice-core records suggest that the and ecological thresholds to produce, for example, changes
major reorganization in North Atlantic hemispherical in lake sedimentation or in regional plant communities. As
circulation that accompanied this warming occurred a consequence, and perhaps surprisingly as many hundreds
within one to three years and resulted in a centennial-scale of Lateglacial sites have now been investigated in Europe
rise in temperatures of up to 10°C (Figure 7.27; Steffensen and North America, a complete climatic picture for the
et al., 2008). In the Southern Hemisphere, by contrast, Ant- Lateglacial Interstadial around the North Atlantic region has
arctic ice-core records show a gradual warming throughout yet to be established.
this period, while the North Pacific experienced three short Antarctic records (Figure 7.26), by contrast, do not show
warmings, each lasting between 500 and 1,500 years any matching reciprocal signals for these short-lived cooler
(Sarnthein et al., 2006). Moreover, the high-resolution events, while the Lake Suigetsu record (Figure 7.32) reveals
(Grey scale, 0-255)

SGP2003-A N.Atlantic event
T nn f X ) E v e n t
Pollen 160 220
a n n
' stratigraphy stratigraphy stratigraphy
10,000
Cariaco GRIP
SGP2003-A1
SGPH Holocene
11,000
SGP2003-A2
SGP2003-B1
SGPS-1
12,000
SGP2003-B2 GS-1 (YD)
Susceptibilit
SGP2003-C1 Gl-1a
13,000 Gl-lb
SGPI-1a
SGP2003-C2 Gl-1c
?
SGPI-1C ;sgp;2003-c3; Gl-1d
14,000
Gl-1e
SGP2003-C4
SGPI-1C
SGP2003-C5
15,000
GS-2
SGPS-2 SGP2003-D
16,000
0 15 -42 -34
Age (ka)
; 5 8 C \ %o
1
Figure 7.32 Lateglacial temperature record from Lake Suigetsu Japan, derived from a pollen-based transfer function. The timescale
is based on varve chronology and high-precision radiocarbon dating (section 5.4.2.7; Figure 5.25). The black (thick) curve on the
left is a running mean through individual temperature uncertainty ranges. The record is plotted against the Cariaco Basin (Venezuela)
grey-scale record (also dated by a combination of radiocarbon and varve chronology), the Greenland (GRIP) ice-core record, and
the North Atlantic event stratigraphy scheme dated using independent chronologies (from Nakagawa et al., 2005).
only one similar event, the age of which lies between those YD is occasionally referred to as ‘Heinrich event 0’ (H-0),
of GI-1d and GI-1b. The pattern of climate variability because a detrital sediment layer has been identified in
during GI-1 described above appears, therefore, to be North Atlantic marine sediments from the YD interval
confined largely to the North Atlantic region, and is (Andrews et al., 1995). This term has not been widely
probably attributable to episodic phases of meltwater influx adopted, however, as the detrital layers appear to be
released during the continuing demise of the northern ice much more limited in terms of their spatial extent, unlike
sheets (Clark et al., 2001). the widespread IRD deposits that characterize H1–H6
The final climatic episode of the Lateglacial is the clearly (Hemming, 2004).
defined Younger Dryas (YD) Stadial, dated in the Green- In the Greenland ice-core records, the YD is the ‘end
land stratotype sequence to between 12.9 and 11.7 ka member’ of the progressive cooling trend that characterizes
(GICC05: MCE of 138 and 99 years, respectively: Figure the GI-1 interval (Figure 7.26). While its onset is marked
7.27). This is one of the best-known and most intensively by a marked shift in isotopic values (reflecting a fall in
studied intervals in the entire Quaternary record, partly temperature), this is no steeper or greater in magnitude
because it constitutes a clearly defined time-stratigraphic than the two preceding thermal declines (GI-1e to GI-1d
unit in both lithological and pollen-stratigraphic sequences and GI-1c to GI-1b). The records therefore suggest that the
(e.g. Figure 3.48), and partly because it constitutes the YD is the product of a longer, incremental cooling trend,
most recent and accessible example in the recent geo- though many consider it to be a unique event. The most
logical record of a high-resolution millennial- (or even widely accepted explanation for the onset of YD cooling
centennial-) scale stadial event and, as such, is readily again involves the injection of meltwater into the North
amenable to climate modelling (Broecker et al., 2010). The Atlantic, this time on a sufficient scale and for a sufficiently
long period to lead to a significant decrease in the rate of as a driver of YD climate changes, but they also reveal the
NADW formation and a weakening (or complete remarkable temporal resolution that is now possible in
disruption) of the AMOC, with the resulting spread of sea the reconstruction of Lateglacial climatic events. Two
ice and a hemispherical drop in atmospheric temperature. lake records serve to illustrate this point (Figure 7.33).
This hypothesis has gained a great deal of support, because In Meerfelder Maar in the Eifel district, Germany, the
there is clear evidence of extensive ice-dammed lakes sediments are varved throughout the entire YD interval, and
trapped by the retreating ice margin and of episodic the series has been tied to an absolute timescale by the
catastrophic drainage from these lakes (Leverington et al., well-dated Laacher See Tephra (LST), and by a calibrated
2002), while marine cores show clear evidence of lower radiocarbon chronology based on sixty-nine AMS radio-
salinity and bottom-water flow rates during the YD interval carbon dates (Brauer et al., 2008). Evidence from sub-
(Roberts et al., 2010). In addition, ‘hosing’ simulation annual microfacies and geochemical variations within
experiments have confirmed that meltwater influx on a the varve layers indicates that the lake experienced a
sufficiently large scale would seriously disrupt AMOC significant increase in stormy conditions at the onset of
behaviour (Tarasov & Peltier, 2004). The routing of melt- the YD. This is attributed to a reduction in strength of the
water into the North Atlantic appears to have been via the AMOC, which enabled North Atlantic sea ice to expand,
St Lawrence Seaway and the Champlain Valley of the driving the oceanic Polar Front southwards (Figure 7.33),
northeast USA (Barber et al., 1999; Rayburn et al., 2011), blocking off ocean heat, and diverting strong westerly
although there may well have been a more northerly flux winds across central Europe. The registration of this marked
of meltwater and icebergs from the Arctic Ocean via the climatic shift in the Meerfelder Maar sequence is precisely
Fram Strait and into the Greenland–Iceland–Norwegian dated at 12,679 varve years. As such, it post-dates the onset
Seas (Tarasov & Peltier, 2004). Re-diversion of North of the YD in Greenland by over 200 years, reflecting the
American meltwater drainage and a reduction in iceberg length of time it took for the sea ice to build up and the
meltwater influx into the North Atlantic led to a density Polar Front to migrate southwards to the latitude of the Eifel
increase in surface waters and reactivation of the AMOC. region of Germany.
The associated heat release generated a hemispherical- A similar high-precision Lateglacial record is available
scale warming and brought an end to the YD (Alley, 2000). from Kråkenes Lake in western Norway. There the age
An alternative, and more recent, hypothesis attributes model is based on ninety-six calibrated radiocarbon dates,
the onset of the YD to the impact of an extraterrestrial
object (bolide) at c.12.9 ka that exploded above North
America and partially destabilized the Laurentide ice sheet.
This, it has been argued, caused, inter alia, continental-scale
wildfires, extinction of many megafaunal species, the
disappearance of the Clovis (palaeo-Indian) culture and,
perhaps most significantly, the onset of YD cooling (e.g.
Firestone et al., 2007; Bunch et al., 2012). The impact
hypothesis has proved to be highly controversial, however,
because of questions over the stratigraphical evidence,
consistency in analytical results, and the nature of the
physical mechanisms proposed (e.g. Pinter et al., 2011). It
is also the case that the Younger Dryas is not a unique event
in the recent geological record, and hence does not require
a particular (and in this instance, catastrophist) explanation
to account for its onset and character. Rather, as we have
seen, it is the latest of a series of millennial- or centennial-
scale climatic perturbations that occurred during the last
cold stage and shares a number of common characteristics
(and therefore, by implication, causal mechanisms) with its
predecessors (Broecker et al., 2010). Figure 7.33 Location of the Kråkenes and Meerfelder Maar
(MFM) sites that contain high-resolution records of the GS-1
Recent studies of the YD based on high-resolution (Younger Dryas Stadial) interval (see Figures 7.34 and 7.35 and
analysis of European lake deposits not only provide further text) and inferred positions of the Polar Front (from Lane et al.,
support for the North Atlantic, and particularly the AMOC 2013).
Figure 7.34 a) The end-moraine formed by a cirque glacier that blocked local
drainage to create Kråkenes Lake (foreground) on the western coast of Norway
(photograph by Jan Mangerud, Bergen University, Norway). The lake was drained
a century ago for cultivation, but the brownish vegetation marks the top of the
exposed lake sediments. b) Still preserved in the sequence is a varved record
of the YD interval which contains the Vedde Ash layer. AL/YD marks the base
of the YD, below which are organic-rich deposits of interstadial (Allerød) age.
Detailed analysis of this record has provided precise ages for climatic shifts in
the northeast Atlantic (see text and Figure 7.35) (after Lohne et al., 2013; image
supplied by Øystein Lohne, Swecos Norway AS, Bergen, Norway).
and tied to the NGRIP stratotype by the Vedde Ash, which increased glacier melting. Evidence from marine cores
is present in both the Kråkenes Lake and in Greenland off western Norway indicates that periodic incursions of
ice-core sequences (Bakke et al., 2009). The lake is located warm salty water began at around the same time. The
close to a cirque that was occupied by a glacier during the evidence suggests that the Nordic seas were permanently
YD and which delivered sediment directly into the Kråkenes frozen during the early YD but, as the Polar Front retreated
basin (Figure 7.34). A significant increase in Ti content northwards, sea ice started to break up with largely ice-
(followed by greater variability in the record) and also free conditions being achieved around 12.15 ka BP (Bakke
in sediment heterogeneity (as reflected in the XRF data) et al., 2009). This is around 400 years before the end of
occurs around 12.5 ka cal. BP (Figure 7.35), these trends the YD as recorded in the Greenland ice cores (Walker
reflecting a more erratic sedimentation pattern due to et al., 2009).
MFM varve thickness MFM Ti Krakenes Ti NGRIP dust NGRIP

(mm) (count rate) (count rata) (particles >1 pm per ml) p1SO)
0 1 2 3 0 2000 4000 6000 8000 0 50 100 150 200 250 0 ZOO 400 600 -15 4 3 -11 -39 -37 -3S
11400
11600
i l iut s c eSput isbci el i pt t i b i l i t

11300
12000
i t p t i bSi lui ts c e p t iSbui lsi tc e p t i bS

Susceptibilit
Vedde Ash
Ice-free sea
12200
Northward shift
- ofwesterlies
12400
Increased
glacier
melting
Stui bs icl e
12600
Suscep
Increased storm in ess
12300
13000
Figure 7.35 Comparison of the timing of the mid-YD transitions reflected in the Kråkenes Ti, and Meerfelder Maar (MFM) Ti
and varve thickness records. The age models for the two sites agree within twenty years at the level of the co-registered Vedde
Ash (dashed line). The position of the Vedde Ash in the NGRIP ice core is shown on the right. The main mid-YD transition in the
lake records is marked by the colour change from blue to yellow (from Lane et al., 2013). For further explanation see text.
The Kråkenes record is one of a number from both transition was not a single synchronous event, but rather
terrestrial and marine contexts that point to a marked was a time-transgressive northward shift of polar air masses
change in atmospheric and oceanic circulation regimes not whose average rate of retreat was around 10 km per yr (Lane
just at the onset and end of the YD, but as a transitional et al., 2013).
episode in the mid-YD (e.g. Isarin et al., 1998; Elmore & These examples show the extraordinary level of tem-
Wright, 2011). It has been suggested that this was a broadly poral precision that can now be achieved when Lateglacial
synchronous event across the North Atlantic realm (e.g. records are dated and correlated using the approaches
Bakke et al., 2009), a hypothesis that can now be tested by outlined in section 7.5.5.2. Given suitable sites and the
comparing the records from the Kråkenes and Meerfelder appropriate evidence (varved sequences, media for high-
Maar sites. The key stratigraphic marker linking the two resolution and high-precision radiocarbon dating, tephra
records is the Vedde Ash which is dated at 12,140 ± 40 years isochrones), it now becomes possible to compare and to
on the Meerfelder Maar varve chronology. In that lake correlate records that are greater than 10 ka in age, at a
sequence, an abrupt climatic amelioration, marked by a centennial or, in some instances as we have seen, at a
northward shift of westerly wind systems (reflected in the decadal scale. It is by using such high-precision and high-
XRF Ti record: Figure 7.35), occurs at 12,240 varve years, resolution archives that we will, in due course, be able to
which is precisely 100 years before the deposition of the distinguish, with a degree of confidence, those records that
Vedde Ash. In Kråkenes Lake, a similar shift in the Ti signal are synchronous from those that are not. In this way, we
occurs c. 20 years after the deposition of the Vedde Ash, in will be able to begin to resolve one of the longest-standing
other words c. 120 years after the reduction in storminess problems in Quaternary palaeoclimate research, namely
recorded in the Meerfelder Maar sequence. This temporal whether there is indeed a temporal order in events, and
offset between the two records must reflect the time taken which of those events constitute the leads and which the
for the Polar Front to move northwards from the latitude lags; in other words, we will be much better equipped to
of central Europe to that of southwestern Norway. tackle the ‘chicken-and-egg’ conundrum to which we have
Moreover, it clearly demonstrates that the mid-YD climatic referred in earlier sections of this chapter.
7.6 CLIMATE AND THE HOLOCENE 10000 8000 6000 4000 2000 0
08
Susceptibilit
7.6.1 Introduction
0.4
The Holocene is the most recent period within the
Susceptibilit
geological record and covers the time interval from 11.7 ka 0
until the present day. The term Holocene, which means
s cuespctei bp itliibt i l i t
‘entirely recent’ (or ‘wholly modern’), refers to the warm -0.4
episode that began with the end of the last glacial period
(the Younger Dryas). In many ways, the Holocene is no -0.8
SuS
different in terms of climatic trends from previous
-1.2
interglacial periods of the Quaternary record, but what
10000 8000 6000 4000 2000 0
distinguishes the Holocene from these earlier warm stages Years (BP)
is the evolution of the human environment. Indeed, it is
this anthropogenic signature that is the hallmark of the Figure 7.36 Temperature anomalies for the Holocene derived
Holocene and that not only sets it apart from previous from the mean of seventy-three globally distributed
interglacials, but that justifies its status as a time- temperature records (dark line) with its 1σ uncertainty (grey
stratigraphic unit of series/epoch rank in the geological band) (after Marcott et al., 2013).
timescale (section 1.3). Here we examine some of the key
climatic events that have occurred during the course of the
present interglacial, the evidence for cyclical climate changes 11.0–7 ka followed by relatively constant temperatures
during the later Holocene and their possible causes, and the over subsequent millennia (Marcott et al., 2013). Early
anthropogenic effects on Holocene climate. But we begin Holocene warmth in the Southern Hemisphere at a time
by considering the major trends that are evident in the when Northern Hemisphere climate was still to recover fully
Holocene climate record. from glacial conditions may reflect the persistence of the
bipolar seesaw (section 7.4) into the Holocene epoch.
This broad tripartite pattern of Holocene climate
7.6.2 Holocene climate trends
change, which is reflected in proxy climate records from
When viewed in the long-term perspective of the both marine and terrestrial realms, can be largely attri-
Quaternary, the Holocene appears as a relatively stable buted to orbital forcing. For example, in the high northern
climatic episode. The Greenland oxygen isotope records, latitudes, the Milankovitch variables for 65°N predict
for example, show relatively few significant fluctuations maximum insolation values around 11–10 ka, due to a
during the course of the Holocene, and certainly nothing combination of summer perihelion and enhanced obliquity,
to compare with the regular, pronounced and often abrupt with a subsequent decline by a factor of 10 per cent since
climatic swings of the last cold stage and the Last Ter- that time interval (Ritchie et al., 1983). However, it is now
mination. Global temperature reconstructions suggest a apparent that, when examined at a finer scale of resolution,
warming of around 0.6°C from the early Holocene to a Holocene climate records were more temporally and
temperature plateau extending from 9.5–5.5 ka (Figure spatially variable than the above generalized patterns might
7.36), a period often referred to as the ‘Holocene Thermal suggest. Indeed, there appear to have been significant
Maximum’ (or the ‘Hypsithermal’, ‘Altithermal’, or changes in climate often over very short timescales. For
‘Climatic Optimum’), this warm interval being followed by example, Mayewski et al. (2004) have identified six distinct
a long-term 0.7°C cooling from 5.5 ka to c. 100 years ago. periods of rapid climate change over the course of the last
This general pattern appears to have been globally variable, 11.5 ka (9.0–8.0 ka, 6.0–5.0 ka, 4.3–3.8 ka, 3.5–2.5 ka,
however, with the extra-tropical Northern Hemisphere 1.2–1.0 ka and 600–150 years ago), each characterized by
(north of 30°N) showing a cooling of c. 0.7°C from 7.0 ka polar cooling, tropical aridity and major atmospheric
onwards. In the low latitudes (30°N–30°S), by contrast, circulation changes. More recently, Wanner et al. (2011)
there appears to have been a lesser warming of 0.4°C from also recognized six distinct episodes during the Holocene
11.0–5.0 ka, with temperatures levelling off thereafter. In when periods of more stable and warmer climate were
the Southern Hemisphere (south of 30°S), however, where interrupted by clearly defined cold relapses (c. 8.2 ka, 6.3
the thermal maximum occurred near the beginning of the ka, 4.7 ka, 2.7 ka, 1.55 ka and 550 years ago). We consider
Holocene, cooling (of around 0.4°C) is recorded from some of these climatic events in the following sections.
7.6.3 Holocene climatic events in the high-resolution records from the Greenland ice
sheet, as these indicate that the major climatic shift at the
beginning of the Holocene occurred within a century or
7.6.3.1 The Pleistocene–Holocene transition
less.
The transition from the last cold stage (Younger Dryas) Figure 7.37 shows that part of the Greenland NGRIP
to the Holocene, defined in stratigraphic records by the ice core that spans the Pleistocene–Holocene boundary,
Pleistocene–Holocene boundary (Walker et al., 2009), is while Figure 7.38 presents a multiparameter dataset from
manifest in a range of proxy records. These include, the same section of core. The stratigraphic record of
for example, the replacement of cold-climate elements impurities and chemicals within the ice is at subannual
by more thermophilous taxa in pollen profiles; the (often resolution, and provides a unique insight into the onset
abrupt) lithostratigraphic change from minerogenic to and evolution of the rapid climatic shift that occurred
organic sediment in limnic sequences; rapid isotopic shifts at the end of the Pleistocene (Steffensen et al., 2008). The
(reflecting changing temperature values) in isotopic records δ18O profile, which is a proxy for past air temperature at
from lake sediment and from speleothems; and the change the coring site, records a warming during the Pleistocene–
from cold- to warm-dominated microfaunal and micro- Holocene transition of the order of 10°C over 60 years.
floral assemblages in marine sediment cores. All of these It was accompanied by a decline (by a factor of 5–7) in dust
indicate a marked reorganization in the global climate concentration and in calcium, both reflecting a marked
system, and there are indications that the transition from change in atmospheric circulation regime, notably a
a glacial to an interglacial mode was very rapid. However, reduction in dust flux from the low-latitude Asian deserts.
the most impressive manifestation of the nature and speed But the speed of the climatic transition is most clearly
of climate change at the onset of the Holocene is to be found evident in the curve for deuterium excess (d-excess). This
Figure 7.37 a) Visual stratigraphy of the NGRIP ice core across the Pleistocene–Holocene boundary obtained using digital line
scanning (section 5.4.3.1). As explained in Figure 5.26, this process produces a ‘reversed’ image, so that clear ice shows up black,
whereas the cloudy bands, which contain relatively large quantities of impurities, in particular micrometre-sized dust particles from
dry areas in eastern Asia, appear white. This is essentially a seasonal signal that is reflected in annual banding in the ice. Note
that the dust concentrations are much higher in the Pleistocene (GS-1) section of the core than in the Holocene, reflecting a markedly
increased dust flux during colder periods. b) The precise location of the Pleistocene–Holocene boundary at 1,492.45 m is shown
in the enlarged lower image (after Walker et al., 2009; image provided by Sune Rasmussen, Niels Bohr Institute, University of
Copenhagen, Denmark).
value, which is derived from the relationship between a 40 per cent increase in ice-layer thickness (Figure 7.38),
deuterium and oxygen, is an important isotopic tracer of reflecting an abrupt increase in precipitation, the two
precipitation and, in the Greenland ice-core records, is proxies marking a sudden and major reorganization of
indicative of changes in the moisture source areas of the Northern Hemisphere atmospheric circulation, which is
precipitation that falls on the ice sheet. It can also be believed to be related to the rapid northward movement
considered as a proxy for sea-surface temperatures in those of the oceanic Polar Front at the end of the Younger Dryas
moisture source areas. The NGRIP d-excess record shows Stadial/Greenland Stadial 1. This extraordinary record
a 2–3 per mil decrease at the onset of the Holocene, cor- from the Greenland ice sheet shows that an abrupt shift
responding to an ocean-surface temperature decline of between two radically different climatic states can occur
2–4°C in the regions from which the precipitation was being within a matter of a few years.
advected; in other words, it reflects a change in the source
of Greenland precipitation from the warmer mid-Atlantic
7.6.3.2 The 8.2 ka event
during glacial times to colder high latitudes in the early
Holocene (Masson-Delmotte et al., 2005). The NGRIP During the course of the early Holocene, three climatic
record shows that this change in d-excess occurred over a cooling events are evident in the Greenland ice cores at
timescale of no more than 1–3 years! It is accompanied by c. 11.5 ka, 9.3 ka and 8.2 ka (Rasmussen et al., 2007). These
a) G I C C 0 5 t i m e s c a l e (ka b2k)
8 9 10 11 12 13 14 15
-34 GS-1 GI-1, GS-2

3 b c 5 e
Susceptibilit
-40 Holocene
-45
34
Susceptibilit
b) -36
-38
3
-40
Susceptibilit
-42
2
44
1
i tp t i b i l i t
10
0
Stui bs icl e
-1
5
p t i b i l i tS u s c e p
150
ptibilit Susceptibilit
1 CO 0
50 150
S u s c eSputsi bc iel i t
0 100
15 50
S u s c e p t iSbui lsi tc e
0
10
5 11.6 11.65 11.7 11.75 11.8

G I C C 0 5 t i m e s c a l e (ka b2k)
Figure 7.38 a) The δ18O record through the last glacial–interglacial transition, showing the position of the Pleistocene–Holocene
boundary in the NGRIP core. b) High-resolution multiparameter record across the Pleistocene–Holocene boundary: δ18O, electrical
conductivity (ECM), annual layer thicknesses corrected for flow-induced thinning (λcorr) in arbitrary units, Na concentration, dust
content, and deuterium excess (after Walker et al., 2009.Copyright © 2008 John Wiley and Sons, Ltd).
Dongge Cave, East China -9.5
Susceptibilit
-8.5
Arabian Sea (Pakistan margin)
Susceptibilit
-2.2
-1.8
Hoti Cave, North Oman
Susceptibilit
•6
-5
-4
Aegean Sea
100
80
60
Susceptibilit
Ammersee, Germany
-4
-4.8
Susceptibilit
200 G e r m a n tree-ring series

•5.6
120
C a r i a c o b a s i n , nr. V e n e z u e l a
Susceptibilit
04
0.32 200
0.24
180
Cariaco basin, nr. Venezuela {greyscale}
1 Norwegian Sea
Susceptibilit
summer
2 winter
3
GISP2, Greenland
-34
Susceptibilit
-35
-36
Terminal A g a s s i z a n d
I O j i b w a y l a k e s outburst
NGRIP, Greenland
-34
Susceptibilit
-35
-36
6000 7000 8000 9000 10000

A g e f c a l . yr BP}
Figure 7.39 Proxy records for the 8.2 ka event. The vertical grey bar marks the approximate duration of the climatic anomaly
associated with the 8.2 ka event (after Walker et al., 2012. Copyright © 2012 John Wiley and Sons, Ltd).
lasted for no more than c. 150 years, and each is marked In higher latitudes, by contrast, there are indications of
by a pronounced δ18O anomaly (abrupt shift to low isotopic colder conditions that appear to be coeval with the lower-
values) in the ice-core records (Figure 7.39 lower two latitude aridification event. In the Southern Hemisphere,
panels). The most significant of these cooling episodes in these include deuterium-derived temperature reconstruc-
terms of a global climate signal is that which occurred tions from Antarctica (Masson-Delmotte et al., 2004) and
around 8.2 ka. ocean-core records off South Australia (Moros et al., 2009),
This event is also registered in the ice cores in a decline while in the mid- and high latitudes of the Northern
in ice-layer thickness and deuterium excess, in a conspic- Hemisphere, a climatic shift to cooler and wetter conditions
uous minimum in atmospheric methane (a ‘global event’), is reflected, for example, in European peat sequences
and in a subsequent increase in atmospheric CO2, and the (Barber et al., 2003), in glacier readvances in northwest
duration is estimated to be 160 ± 10 years. It is generally Canada (Menounos et al., 2008) and in ice-core records
considered that the event reflects curtailment of North from the Yukon (Fisher et al., 2008).
Atlantic Deep Water (NADW) formation and its associated The forcing mechanisms behind the global reorgan-
northward heat transport, due to catastrophic meltwater ization of climate that is marked by the 4.2 ka event are less
release from glacial lakes Agassiz and Ojibway into the obvious than is the case with the climatic shift at 8.2 ka.
North Atlantic during the later stages of the wastage of the There is, for example, no evidence for massive meltwater
Laurentide ice sheet (Kleiven et al., 2008). It is possible that releases into mid- and high-latitude oceans; nor is there
the earlier Holocene oscillations resulted from a similar evidence for major increases in atmospheric trace gases
process, as climate modelling experiments have shown or volcanic aerosols which might have served to trigger
that freshwater fluxes into the North Atlantic can also the event. Similarly, there is nothing in the Holocene
trigger centennial-scale cooling events with temperature solar irradiance record to suggest that solar activity might
anomalies resembling proxy evidence for the cooling event have played a part (Figure 7.41). Mayewski et al. (2004)
at 8.2 ka (Renssen et al., 2007b). The 8.2 ka event has been have suggested that the southward migration of the Inter-
detected in a range of proxy records (Figure 7.39), Tropical Convergence Zone (ITCZ) might account for
particularly from round the North Atlantic region, but also enhanced low-latitude aridity, and would be consistent with
in Africa, Eastern Asia, South America and perhaps New the increase in strength of the westerlies over the North
Zealand (Walker et al., 2012). In the South Atlantic, there Atlantic, and increased precipitation and the consequent
are indications that the event may be associated with an glacier advances in western North America. The onset of
increase in sea-surface temperatures, and coupled climate aridification also coincides with a 1–2°C cooling of North
model simulations appear to show a warm response at Atlantic surface waters (Bond et al., 1997), while in the
around 8.2 ka in the Southern Oceans (Wiersma et al., Pacific, tropical deep waters may also have cooled suffi-
2011), again perhaps reflecting the continued operation of ciently to allow the switch-on of the modern El Niño
the bipolar seesaw referred to above. Overall, therefore, Southern Oscillation (ENSO) regime. The climatic signifi-
although the trigger mechanism for the 8.2 ka event lies in cance of ENSO and associated La Niña climatic regimes is
the North Atlantic, it is unusual in late Quaternary records discussed further in section 7.6.4.2.
in being a climatic event that is near global in nature The global impacts of the 8.2 and 4.2 ka events and their
(Rohling & Pälike, 2005). registration in a range of stratigraphic and proxy climate
records has led to the suggestion that these might consti-
tute time-stratigraphic marker horizons for a subdivision
7.6.3.3 The 4.2 ka event
of the Holocene (Walker et al., 2012). Hitherto, there has
Throughout the middle and low-latitude regions of the been no formal subdivision of the Holocene epoch (the only
world, there is widespread evidence for a pronounced, epoch in the geological record so far not to be so divided),
but relatively short-lived (200–300 years), aridification and it has been proposed that the Holocene be formally
event which occurred around 4.2 ka. This is again subdivided into three sub-epochs: the Early Holocene
reflected in a range of proxy climate records (Figure 7.40), (11.7–8.2 ka); the Middle Holocene (8.2–4.2 ka); and the
including fluvial archives, lake sediment sequences and Late Holocene (4.2 ka to the present). This is currently
cave stalagmite profiles from North America, through the (2014) under consideration by the Subcommision on Qua-
Middle East to China; and from Africa, parts of South ternary Stratigraphy and the International Commission on
America to Antarctica where there is also evidence from Stratigraphy (section 1.3) and, if approved, will be for-
ice-core records of drier conditions (Mayewski et al., warded to the International Commission on Stratigraphy
2004). for formal ratification.
12
Susceptibilit
45
Susceptibilit
Susceptibilit
Susceptibilit
30 Gulf of O m a n CaC0 3 8
15 4
0 Dolomite 12
Susceptibilit
Susceptibilit
Lake Van
9
Susceptibilit
0 01 60
Susceptibilit
Gdlhisar Golii
Susceptibilit
Susceptibilit
-2.5
40
-5 0
20
-7.5
ptibilit
370 0
380
t usce
ptibilit
390
Dead S e a
S u s c e p t i b i l iS
400
1
S u s c e p t i b i lSi tu s c e
410
Susceptibilit
420 0
Shaban Deep -1
-4.5
-10
Susceptibilit
Susceptibilit
-5.1 Soreq
-11
-5.7
-12
-6.3
Renella 1
Susceptibilit
0
Susceptibilit
-1
-6
-2
Kilimanjaro
-9
1000
i bt uiiblsiitcl iet p t i b i l i t
Susceptibilit
-12 100
Susceptibilit
10
-15 Kilimanjaro 1
u sucsecpetpS
0.1
0.01
Mount Logan
SS
28
Susceptibilit
30
32
34
36
38
-4.5
Susceptibilit
-5.5 Mawmlull Cave
-6 5
-7.5
2000 3000 4000 5000 6000
A g e (cal. yr B P )
Figure 7.40 Proxy records for the 4.2 ka event. The vertical grey bar marks the likely onset and termination of the event (after
Walker et al., 2012. Copyright © 2012 John Wiley and Sons, Ltd).
800
l i st c e p t i b i l i tS u s c e p t i b i l i t
600
400
200
-200
-400
-600
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
A g e (yr B P )
Figure 7.41 Solar activity throughout the Holocene. The solar modulation parameter (Φ) is based on 14C data from IntCal04 (orange
curve) and 10Be from the GRIP ice cores (blue) (after Knudsen et al., 2009). Note the significant decline in solar irradiance at around
2.8 ka; note also that there is no similar decline at c. 4.2 ka suggesting that solar forcing was unlikely to be implicated in this event.
7.6.3.4 The 2.8 ka event earth’s geomagnetic field. At times of increased solar activity
when the solar magnetic field is strengthened, fewer cosmic
An abrupt climatic deterioration marked by the onset of rays reach the earth’s atmosphere, and thus the production
cooler and wetter conditions and beginning at around of radionuclides such as 14C and 10Be is reduced; the reverse
2.8 ka cal. BP has been widely recorded in both proxy obtains at times of lower solar intensity. Holocene records
climate and archaeological records from Europe (e.g. van show a significant increase in radionuclide concentrations
Geel et al., 1996; Speranza et al., 2002). For example, at c. 2.8 ka, and hence a decline in solar intensity (Figure
wet-shifts at about that time have been found in peat- 7.41). This, coupled with the fact that evidence for the 2.8
land sites in Britain (Charman et al., 2006) and Ireland ka event is found in both hemispheres and is broadly
(Swindles et al., 2007), while a similar change to wetter synchronous, provides strong supporting evidence for the
conditions has also been reported from Germany (Barber hypothesis that a worldwide climatic shift at 2.8 ka is
et al., 2004) and Siberia (van Geel et al., 2004), and in driven principally by solar forcing.
the Southern Hemisphere from the Andean region of
South America (Chambers et al., 2007). In addition, cooler
ocean surface waters and an increase in ice-rafting at 7.6.3.5 The Little Ice Age
around 2.8–2.7 ka is evident in marine cores from the Of all the climatic events in the Holocene, none has
North Atlantic (Bond et al., 1997). Over the broader time attracted more attention than the Little Ice Age (LIA), the
interval of 3.5–2.5 ka, there is evidence for a weakening short-lived cold interval from approximately the sixteenth
of westerly winds over North America, cooling over the to the nineteenth centuries. One of the reasons for this is
northeast Mediterranean related to winter-time conti- that although the Little Ice Age was not a true ‘ice age’ in
nentality/polar air outbreaks, advances of alpine glaciers the conventional sense, it was nevertheless a colder episode
and a decline in the treeline limit in Scandinavia (Mayewski that, particularly in northern Europe, had considerable
et al., 2004). effects on late- and post-Medieval societies. In alpine areas,
The principal factor driving this climatic change is glaciers readvanced, in some cases destroying farms;
widely considered to be a reduction in solar activity (Bond crop failure and resultant famine were widespread, while
et al., 2001b). Long-term variations in solar activity can expanded sea ice disrupted coastal communications and
be determined from the record of cosmogenic radio- fisheries (Grove, 1988; Fagan, 2000). When viewed at
nuclides (14C and 10Be) preserved in tree rings and ice the global scale, however, the Little Ice Age appears to have
cores (Knudsen et al., 2009; Steinhilber et al., 2012). We been a relatively modest cooling of the Northern Hemi-
have already seen (sections 5.3 and 5.6) that both of these sphere with an overall temperature decline of less than
radionuclides are produced in the atmosphere from 1°C relative to late twentieth-century levels. On the other
incoming cosmic rays, but the cosmic ray flux reaching the hand, these generalized values conceal considerable regional
earth is modulated by both the solar magnetic field and the and temporal differences (Crowley & Lowery, 2000).
For example, in Europe, the coldest periods appear to have The second hypothesis, involving volcanic forcing has also
occurred in the eighteenth century, with temperatures in been long espoused. During an eruption, large quantities
some regions 2°C below twentieth-century levels, while in of ash and sulphur are ejected into the atmosphere, the
North America, the coldest decades were experienced former blocking the sun’s rays beneath the spreading
during the nineteenth century and were close to 1.5°C ash cloud, while the latter interacts with atmospheric con-
colder (Mann et al., 1998). stituents to create sulphuric acid particles that are dis-
Whether the Little Ice Age was largely a Northern seminated globally and which also screen out incoming
Hemisphere, and indeed a North Atlantic phenomenon, solar radiation (Oppenheimer, 2011). Recent work in Arctic
has been much debated (and indeed the global relevance Canada and Iceland, using a combination of empirical
of the term ‘Little Ice Age’ itself has been questioned: evidence and climate simulation modelling, has shown
Jones & Mann, 2004), principally because of the spatial that the onset of the LIA can be linked to an unusual
bias towards Northern Hemisphere proxy and meteoro- 50-year-long episode with four large sulphur-rich explosive
logical records (Jones et al., 2009). However, there is now eruptions around the end of the thirteenth century, the
an emerging body of evidence from the Southern Hemi- subsequent cold being maintained by sea ice–ocean feed-
sphere to suggest a climate cooling broadly coterminous backs long after the volcanic aerosols were removed (Miller
with the LIA in the Northern Hemisphere. There are, for et al., 2012). Under this scenario, the climatic patterns of
example, indications of glacier advances in New Zealand the LIA can be explained without recourse to changes in
during the eighteenth century (Winkler, 2000); stalag- solar irradiance.
mite data from South Africa point to a cold period between
AD 1500 and 1800 (Holmgren et al., 2001); borehole records
from Australia suggest that the seventeenth century was
7.6.4 Holocene climatic cycles
the coldest in Australia over the last 500 years (Pollack A distinctive feature of many Holocene records from both
et al., 2006); while stable isotope data from a coastal ice marine and terrestrial realms is the apparently rhythmical
core from Antarctica indicate that the area experienced or cyclic nature of climatic fluctuations over the course
1.6–1.4°C cooler average temperatures prior to AD 1850 of the past 11.7 ka or so. Unlike the longer-term climatic
compared with the last 150 years (Rhodes et al., 2012). cycles that characterize the Pleistocene record, however,
Nonetheless, the question still remains as to whether these these are not driven by orbital influences; rather, they
events are indeed synchronous with periods of anomalous appear to reflect external forcing principally through
cold further north and whether they reflect a globally variations in solar irradiance, or to be the consequence of
significant climate signal. internal reorganization and feedback within the
The LIA occurred against a backdrop of long-term ocean–atmosphere system. Some of these cyclical climate
Northern Hemisphere temperature decline driven by changes are discussed in this section.
orbital forcing. In the Arctic, for example, it has been
estimated that the rate of cooling has been of the order of
7.6.4.1 Late Holocene solar cycles
0.02°C per century (Kaufman et al., 2009). But this alone
would be insufficient to generate the levels of cooling We have already seen that a proxy record of past solar
evident in Europe and North America during the LIA. variability can be obtained from cosmogenic nuclides in tree
A possible causal factor might have been a reduction in rings and ice cores (Figure 7.41), and indeed that some
North Atlantic Deep Water Formation and the disruption events in the Late Holocene palaeoclimate record, such as
of thermohaline circulation, although so far the oceano- the 2.8 ka event and the Little Ice Age, have been linked to
graphic evidence in support of this hypothesis is equivocal changes in solar irradiance. It is also apparent from these
(Broecker, 2000). More likely explanations appear to lie in proxy records, and also from observational and instru-
climatic forcing from solar or volcanic activity (Grove, mental measurements, that there is a cyclical pattern in solar
1988). It has long been known that two significant epi- variability. These periodicities include the 11-year Schwabe
sodes of reduced sunspot activity (reflecting a reduction in sunspot cycle, the 22-year Hale cycle, the 88-year Geisberg
solar output) occurred during the LIA: the Maunder cycle, the ~200-year Suess cycle, and the ~1,000-year and
Minimum between 1645 and 1715, and the Dalton Mini- ~2,200-year (Hallstattzeit) cycles (Chambers & Blackford,
mum from 1790–1830, and these are well represented in 2001).
the radionuclide records of 14C and 10Be from tree rings Until relatively recently, an influence of solar variability
and ice cores (Steinhilber et al., 2012). The reduction in on climate, whether through cycles or trends, was usually
solar irradiance during the LIA is apparent in Figure 7.41. dismissed because climate simulations with simple energy
balance models indicated that the responses to the decadal periods, such as the late nineteenth and early twentieth
solar cycles would be too small to be detected in obser- centuries, are marked by high-amplitude oscillations, but
vations and hence in proxy records. However, more recent others (such as the 1930s) are more quiet (Cane, 2004).
modelling studies have found indications of positive The causes of the oscillation are not fully understood,
feedbacks in the ocean–atmosphere system that may but appear to be associated with a reduction in upwell-
amplify the response to solar irradiance variations, and now ing of colder waters off the South American coast as a result
solar cycles and trends are widely recognized as important of the weakening of the trade winds (part of the atmospheric
components of natural climatic variability on decadal to Walker circulation). There is a surface-water heat transfer
centennial timescales (Lean, 2009). For example, although from the western to the eastern Pacific. As the Pacific
the variation in solar irradiance that is associated with a Ocean is a major heat reservoir that drives global wind
Schwabe cycle is only of the order of ~1 W m–2 between patterns, the changes in temperature that result from
solar maximum and minimum, winter and spring tem- this redistribution of surface water exert an influence on
peratures in the Northern Hemisphere nevertheless show weather patterns at the global scale. For example, rainfall
a response even to this small-scale variability (Engels & van shifts from the western Pacific towards the Americas, while
Geel, 2012). Australia, Indonesia and India experience droughts.
Over the course of the Holocene, perhaps the most The ENSO cycle is apparent in proxy climate records
pervasive cycles have been those at ~1,000 and maybe (such as tropical corals) extending back to the last inter-
also ~2,000 years. Of these, the ~1,000-year cycle has been glacial (Tudhope et al., 2001). During the present inter-
the most widely detected, registering in such diverse glacial, it appears that the ENSO cycle was markedly weaker
archives as lake sediments in Alaska (Hu et al., 2003), during the early and middle Holocene, but proxy records
varved sediments in the Santa Barbara Basin off Cali- indicate that a major climatic transition in the Pacific
fornia (Nederbragt & Thurow, 2005), foraminiferal records region, characterized by an increasing ENSO influence, is
and sea-surface reconstructions off the east coast of evident from c. 5–3 ka onwards (e.g. Gomez et al., 2004).
North America (Cléroux et al., 2012) and loess–palaeosol An enhanced ENSO regime between 4 and 2 ka is also
sequences in southern Siberia (Kravchinsky et al., 2013). evident along the Pacific coastline of North America
The extent to which these solar cycles have influenced (Barron & Anderson, 2010). This increase in ENSO
Holocene climatic variability remains uncertain, but they variability during the later Holocene may be an oceanic
may have served to amplify or modulate climate changes response to changing insolation (Loubere et al., 2013).
arising from other forcing mechanisms such as variations Over the past 1.0 ka, data from corals show both decadal
in deep-water flow, sea-surface temperature changes, sea and longer variations in the strength of the ENSO cycle,
ice formation and the effects of the ENSO regime (see possibly due to a combination of volcanic and solar forcing,
below). and amplified by internal feedbacks within the ocean–
atmosphere system (Cane, 2004). Atmospheric general
circulation modelling suggests that with rising temperatures
7.6.4.2 El Niño–Southern Oscillation (ENSO)
from global warming (section 7.6.5.1), ENSO-induced
The term ‘El Niño’ was first used by Peruvian fishermen to drought and floods are likely to be more intense in future
refer to the annual warming of coastal waters that often El Niño years (Power et al., 2013).
occurs around Christmas time, but scientists now reserve
the term for an anomalous periodic warming, reflected
7.6.4.3 Late Holocene Atlantic and Pacific
in an increase in sea-surface temperatures (SSTs) of the
Oscillations
eastern equatorial Pacific. The ‘Southern Oscillation’ is a
seesawing of atmospheric mass, and hence of sea-level The North Atlantic Oscillation (NAO)
pressure (SLP) between the eastern and western Pacific. This is one of the principal modes of climatic variability
Records over the past 100 years or so show a striking in the North Atlantic region, and affects surface air tem-
similarity between the SST and SLP indices, one oceanic and peratures, winds, storminess and precipitation (Hurrell &
the other atmospheric, despite the fact that they are widely Deser, 2009). It is driven by variations in atmospheric
separated in space: El Niño accompanies high surface pressure at sea level between the Icelandic low and the
pressure in the western Pacific, while the cold phase (known Azores anticyclone. During a positive NAO phase, streng-
as La Niña) coincides with low surface pressure in the thening of the high-pressure cell over the subtropical
western Pacific. Oscillations occur approximately every Atlantic and a weakening of the Icelandic low accentuates
4 years (2–7 years normally defines the ENSO band); some the pressure gradient between the two cells, and this results
in warmer and wetter conditions over northwest Europe ‘negative’ phase, the opposite pattern occurs. This Pacific
and the eastern USA, whereas Greenland, northern Canada Decadal Oscillation operates over a timescale of 20–30
and the Mediterranean tend to be cooler and drier. In years, and has been linked to ENSO, marked sea-surface
negative NAO phases, when the gradient between the temperature changes in subsequent winters and atmos-
Azores high and the Icelandic low is reduced, this situation pheric forcing (Newman et al., 2003). During positive
is reversed. The oscillation is driven entirely by changes phases, warmer conditions are experienced in the Pacific
within the atmosphere and there is no discernible period- Northwest and Alaska, whereas lower temperatures are
icity; indeed, it displays considerable inter-decadal and recorded in Mexico and the southeastern USA (Mantua &
inter-annual variability (Thompson et al., 2003). The NAO Hare, 2002). Droughts over the northern and southwestern
has been detected in both instrumental and proxy records. USA have been associated with a positive AMO, especially
For example, the exceptionally cold winter of 2009–10 in when linked to a negative PDO (McCabe et al., 2004). In
Britain and western Europe coincided with an extremely India, increased summer rainfall and decreased tempera-
negative phase of the NAO, while the observed shifts during tures have been associated with the negative phase of the
the 1990s towards a pattern of anomalously wet summers PDO (Krishnan & Sugi, 2003). Tree-ring records from
in northern Europe and concomitant hot and dry summers western North America provide a history of the PDO, for
in southern Europe have been related to warming of the example a combined tree-ring record from California and
North Atlantic following a positive phase of the NAO Alberta reveals PDO changes over the course of the last
(Robson et al., 2012). In southern Norway, rapid glacier millennium (MacDonald & Case, 2005).
advances during the early eighteenth century have been
attributed to increased precipitation in milder and wetter
winters during a positive NAO mode (Nesje & Dahl, 2003).
7.6.5 People and climate
On the east coast of the USA, marine Foraminifera inter- Thus far in this chapter, we have considered climate change
bedded in littoral marsh sediments reflect coastal inunda- on a range of temporal and spatial scales, and have exam-
tion resulting from hurricanes, and have been linked to ined the various forcing factors and feedback mechanisms
changes in the position of the Bermuda (or Azores) high- that have fashioned global climate over the course of the
pressure cell, which influences the direction of storm paths middle and later part of the Quaternary. To date, however,
for North Atlantic hurricanes (Scott et al., 2003). we have said relatively little about people, and have focused
our attention largely on the impacts of climate change on
The Atlantic Multidecadal Oscillation (AMO) the natural environment. We have tended also to consider
This was first identified in 1994 and is marked by vari- the larger-scale changes in climate as reflected, for example,
ations in North Atlantic sea-surface temperature. These in the high-amplitude temperature fluctuations that
have been linked to small changes in the North Atlantic occurred during the DO cycles. But it is evident that these
branch of the thermohaline circulation. The AMO appears temperature (and precipitation) changes must, equally,
to have a quasi-cycle of around 50–90 years, with past have impacted on plants and animals and, especially, on
recorded peaks around AD 1880 and 1950, and the next humans, where even comparatively small variations in
likely peak between 2000 and 2040 (Enfield & Cid-Serrano, climatic regime may have had a profound impact on human
2010). The AMO has been linked to variations in air societies. This is particularly so as globally averaged
temperature and rainfall throughout the Atlantic region. temperatures (with which we have largely been dealing)
Warmer waters in the North Atlantic have been associated, tend to mask the spatial variations of climate shifts, so that
for example, with increased rainfall in southern Florida, the some regions would have experienced more pronounced
African Sahel and India, and with increased Atlantic climatic effects than others. The extent to which climate,
hurricane activity (Zhang & Delworth, 2006). The AMO has and especially the DO cycles and Heinrich events, impacted
also been linked with drought conditions in the American on the lives of Neanderthals and early modern humans in
Midwest and over large areas of West Africa (Shanahan the Old World during the last cold stage has been much
et al., 2009). debated (e.g. Gilligan, 2007; Jiménez-Espejo et al., 2007) as,
of course, have the climatic and other factors (such as ice-
The Pacific Decadal Oscillation (PDO) sheet extent) that determined the timing of the arrivals of
Periodic spatial changes in sea-surface temperatures have humans in the Americas (e.g. Davidson, 2013; Rabassa &
been detected in the Pacific Ocean north of 20°C; a cooling Ponce, 2013).
in the west Pacific is accompanied by a warming in the The most recent of these pronounced climatic shifts,
east (‘warm’ or ‘positive’ phase), while during a ‘cool’ or the transition from the last cold stage to the Holocene,
coincides with the beginnings of agriculture, one of the Callendar, showed that the planet had warmed by about
most important innovations in human history, and the 0.3°C over the previous 50 years and suggested that carbon
transition from the hunter-gatherer economies of the last dioxide (CO2) could be partly to blame (Hawkins & Jones,
cold stage to the sedentary cultivation-based societies of the 2013). Remarkably, Callendar’s global land temperature
Holocene. And with that cultural change began the first estimates agree well with more recent analyses (Figure
significant impacts of humans on the natural environment, 7.42) which show a sustained rise in the earth’s temperature
with widespread forest clearance, the development of urban over the course of the last 100 years or so. Although some
societies and industrialization. One consequence of this, of still remain sceptical about the nature and extent of the
course, was to reverse the former scenario in which climate greenhouse effect (for both scientific and political reasons!),
was a limiting (in some cases, perhaps, even a forcing) factor a combination of empirical data from instrumental and
in human activity and behaviour, to one where humans proxy sources, coupled with the results of increasingly
themselves are exerting an increasing influence on the sophisticated climate modelling, have provided a powerful
global climatic environment, and it is this human dimen- confirmation of the hypothesis that the recent rise in global
sion in recent climate and environmental change that we temperatures can be attributed, in very large measure, to
address in the final sections of this chapter. the effects of human activity. Indeed, the most recent
report by the Intergovernmental Panel on Climate Change
(IPCC) published in September 2013 states that scientists
7.6.5.1 The greenhouse effect
are now 95 per cent certain that humans are the dominant
During the later years of the nineteenth century, pioneering cause of global warming since the 1950s.
experimental work by John Tyndall had shown that gases, The principal contributors to atmospheric green-
such as carbon dioxide, can effectively absorb infrared house warming are the trace gases CO2 (carbon dioxide),
radiation. Later, the Swedish scientist, Svante Arrhenius, CH4 (methane) and N2O (nitrous oxide), along with
used these results to estimate the sensitivity of global CFCs (chlorofluorocarbons). The most valuable archives
temperatures to increases in carbon dioxide. This was the of changes in these greenhouse gases are the polar ice
first articulation of what we now refer to as the ‘green- sheets from which records of both long-term (section 3.11)
house effect’, although in the early 1900s, the theory that and more recent (post-industrial) atmospheric concen-
an increase in atmospheric carbon dioxide could change trations have been obtained (Figure 7.43). These show
climate was not widely accepted. In a seminal paper that for CO2, the mean concentration of late Holocene pre-
published in 1938, however, an amateur scientist, Guy industrial levels was around 280 ppmv (parts per million
1,2 C R U T E M 4 (Jones et al. 2 0 1 2 ) - 6 0 ° S - 6 0 ° N

Catlendar(1938)
1
Callertdar(1961)
0.8
06
0.4
S u s c eSputsi bc iel ipt t i b S
02
-0 2
-0.4
-06
1850 1900 1950 2000

Year
Figure 7.42 Comparing historical reconstructions of near-global land temperatures (‘CRUTEM4’: Jones et al., 2012) with
Callendar (1938) and Callendar (1961), using a reference period of 1880–1935. The CRUTEM4 estimates are for 60°S–60°N (to
accord with Callendar’s series), with grey shading representing the 95 per cent uncertainty.
400 2000
add measurably to the natural greenhouse gas levels already
in the atmosphere. But this interpretation of the atmos-
Carbon dioxide ( C 0 ) 2
1800 pheric trace-gas record was challenged by Bill Ruddiman,
b si l ci te p t i b i l i t
Methane (CH ) 4
Nitrous oxide ( N 0 )
who argued that anthropogenic influences on Holocene
2 1600
350 climate predated by several millennia the eighteenth-
Susceptibilit
1400 century increase in greenhouse gases (Ruddiman, 2005a,
tiu
i bui lsi tc e p S
2005b). He first proposed that the natural decrease in

1200
atmospheric CH4 that accompanied the decline in orbitally
300
S u s c e p tS
1000 driven solar insolation since the early Holocene was

reversed around 5 ka, principally as a consequence of the
800
spread of rice farming throughout southeast Asia (Ruddi-
250 600 man & Thomson, 2001). The irrigated rice paddies with
0 500 1000 1500 2000
their rich biomass would have been a natural source of
Year
methane, and the release of these gases to the atmosphere
was, Ruddiman suggested, largely responsible for the
Figure 7.43 Atmospheric concentrations of greenhouse
gases over the past 2 ka. The increase since c. AD 1750 is
progressive upward trend in methane concentration that
largely attributable to human activities in the industrial era (from is evident in the ice-core records from the mid-Holocene
Intergovernmental Panel on Climate Change, 2007). onward (Figure 7.44a). He further proposed that humans
also reversed the natural decrease in atmospheric CO2
values around 8 ka by beginning the process of forest
by volume). Following the Industrial Revolution in Europe clearance (Ruddiman, 2003b), again the resultant release
and North America in the eighteenth century, the records
show a progressive rise to values of around 315–320 ppmv
by the 1950s, and to around 380 ppmv by the year 2000.
a)
In May 2013, continuous monitoring of atmospheric CO2 Observed C H 4 trend
at the Maua Loa research facility in Hawaii revealed that 700
the gas had reached an average daily level of 400 ppmv,

l i st cSeupstci be ipl itti bSi u
higher than at any time in the past 3 Ma. Since c. AD 1750,

atmospheric concentrations of CH4 have risen from a pre- 600 250 ppb
industrial level of around 700 ppbv (parts per billion
by volume) to a value in excess of 1,950 ppbv by 2000, and

N2O has increased from around 275 ppbv to more than 500 S o l a r radiation
310 ppbv. The consequences of these increases on global
temperatures have been dramatic. For example, in the 10,000 5,000 0
NOAA National Climatic Data Center annual list Years ago
(http://www.ncdc.noaa.gov), each of the twelve years from b)
290
2001–12 features as one of the warmest in a record that
Observed C 0 trend
began in 1880. There were only two years in the pre-AD 2
280
2000 list where broadly comparable global temperatures

were achieved. Indeed, we have to go back to 1911 to find 270

the coolest year in the NOAA record. These data are
compelling, and it seems very difficult to deny the likelihood 260

40 ppm
that human influence on the global climatic environment bilit
epti i st c
is now all-pervasive. 250 cep
t i bSi ul
Sus
240
7.6.5.2 Early human impact? 10,000 5,000 0
Years ago
Much of the literature on global warming focuses on the
post-industrial period, with the ‘tipping point’ in the Figure 7.44 Concentration of CH4 a) and CO2 b) over the last
anthropogenic impact scenario occurring around AD 1750, 11 ka, showing departures (increases) from the natural trace-
when the byproducts of the industrial revolution began to gas trends (after Ruddiman, 2005a).
of CO2 to the atmosphere being marked in the CO2 record cultivation methods, yield CO2 emission estimates that are
in polar ice-core records (Figure 7.44b). If these arguments in line with pre-industrial trajectories (Ruddiman & Ellis,
are correct, then human influence on the atmosphere, in 2009), while natural explanations for the CO2 increase still
terms of increased trace-gas concentration, does indeed confront the problem that, when compared with the
long precede the industrial era (Ruddiman, 2013). previous four interglacial episodes, the late Holocene CO2
increase is anomalous (Ruddiman et al., 2011).
7.6.5.3 Delayed glaciation?
A second element in the Ruddiman scenario concerns
7.6.6 The Anthropocene
the extent to which trace-gas emission as a consequence The term ‘Anthropocene’ (meaning anthrõpos ‘human
of human activity during the course of the Holocene being’ and kainos ‘new’) has become increasingly widely
could have caused an anthropogenic warming sufficient to used since it was proposed by Crutzen & Stoermer
counter the orbitally driven natural cooling trend, and (2000) to denote the present time interval during which the
to prevent the onset of the next glaciation. For the pre- earth’s climate and many geologically significant processes
industrial period, Ruddiman estimated that global climate (ranging from erosion and sediment transport to sea-level
would have warmed by around 0.8°C. This is around 15 per rise) have been, and continue to be, affected and altered by
cent of the global cooling at the Last Glacial Maximum, human activities. These anthropogenic effects, which have
and when amplified polewards, this would have led to become especially marked since the Industrial Revolution
a temperature decrease of ~2°C which, modelling results in northern Europe, have resulted in marked changes not
suggest, would be more than sufficient to generate only to global climate (see above), but also to the earth’s
permanent snow and ice cover over parts of northeastern surface, and almost every aspect of the contemporary global
Canada (Ruddiman et al., 2005); in other words, to initiate environment is now modified, or at least influenced, by
the next glaciation. Ruddiman also noted that in all of the human activity (Steffen et al., 2011).
last interglacial stages (MIS 1, 5, 7, 9 and 11), summer Although the term ‘Anthropocene’ was initially invoked
insolation reached late deglacial/early glacial peaks and in an informal sense to describe this period of expand-
then fell for the next 11.0 ka or so. Through each of those ing human influence on the global environment (Crutzen,
intervals, CO2 and CH4 concentrations also declined in 2002), there is a body of opinion within the geological
every case but one: the Holocene, where the trend is community that this accelerated human impact may be
reversed (Figure 7.13). Indeed, modelling evidence and reflected in recent stratigraphic sequences, where it can
analogies with earlier interglacials suggest that had CO2 be distinguished from natural ‘background’ conditions
levels been lower (below 240 ppmv, the baseline in Figure (Zalasiewicz et al., 2008). As a consequence, the term is
7.44b), the end of the current interglacial would occur increasingly being applied to the geological record, where
within the next 1,500 years (Tzedakis et al., 2012). The it is viewed by some as being equivalent to those used as
implication, therefore, is that human-induced increases in series-status divisions of the Cenozoic Erathem, notably
atmospheric trace gases have delayed the onset of the next ‘Pleistocene’ and ‘Holocene’. Indeed, a Working Group of
glaciation. the Subcommission on Quaternary Stratigraphy is currently
Both of these ideas, namely that early and mid-Holocene looking into the matter to determine whether the strati-
increases in atmospheric trace gases have resulted from graphic signature of the Anthropocene is now sufficiently
prehistoric human activity, and that these have been clearly defined to warrant its formal definition as a new
sufficient to prevent the onset of glaciation, have proved period of geological time that is additional to, and separate
controversial, some critics questioning whether there from, the Holocene (Zalasiewicz et al., 2011). The Group
were sufficient people in prehistory capable of clearing and aims to conclude its deliberations by 2016.
burning sufficient forest in order to create the observed There are, however, a number of problems relating
CO2 anomaly in the ice-core records, while others have to formal definition of the Anthropocene in this way.
argued that the later Holocene increase in CO2 could be Where to place the onset of the Anthropocene, for example,
accounted for by delayed ocean carbonate compensation has proved to be a contentious issue. Many have advo-
and atmospheric CO2 release, in part associated with coral cated the increase in trace-gas concentrations at around
reef formation (Joos et al., 2004; Elsig et al., 2009). However, AD 1750 in the ice-core records as providing the key marker
preliminary reconstructions of possible rates of early and horizon for the first significant impact of humans on the
mid-Holocene forest clearance, based partly on analogies climatic environment (Gibbard & Walker, 2013). But it is
with contemporary cultures that still employ shifting now clear that human influence on global climate may
considerably predate this event (see above). Alternative the earth can switch from a ‘glacial’ to an ‘interglacial’ state
suggestions include the ‘Great Acceleration’ of the second within a matter of hundreds, as opposed to thousands, of
half of the twentieth century when many human activities years, a timescale that was considered almost inconceivable
reached their take-off points in terms of their environ- prior to the recent findings in the polar ice cores. Indeed,
mental impacts (Steffen et al., 2011), and the short-lived there is clear evidence that wholesale atmospheric reorgan-
atmospheric radiogenic nuclide peak around AD 1963 izations, involving temperature changes of more than 10°C,
resulting from nuclear bomb testing of the mid-twentieth can occur within centuries (in some cases decades) during
century (Zalasiewicz et al., 2011). Again, however, these ‘glacial’ stages, and perhaps did so on more than twenty-
simply mark one of the many stages in human techno- five separate occasions during the last glacial stage.
logical development; they do not coincide with the onset Until recently, major climate shifts appeared to be more
of environmental changes resulting from human activity or less synchronous events in proxy records from ocean
and hence, it has been suggested, have little value in defin- basin, polar ice and terrestrial contexts, but improvements
ing the initiation of human impact (Gale & Hoare, 2012). in dating precision, coupled with the analysis of more
A further difficulty is that the impact of human activity highly resolved stratigraphical archives, are increasingly
is time-transgressive, the effects of industrialization, for revealing that this is not the case, but there are clear
example, occurring at different times in different parts leads and lags in the climate system. This work has also
of the world. Hence it is difficult to isolate a single boundary highlighted the importance of coupled variables, such as
within the geological record that marks the global ‘tipping atmospheric gas and dust content, sea ice and ice-sheet
point’ where natural processes have been overtaken by extent, sea-level change, and ocean circulation, and their
human-induced environmental changes (Gibbard & respective roles in influencing global climate. Although
Walker, 2013; Ruddiman, 2013). the inter-relationships are complex, simulation modelling
There is no doubt that the term ‘Anthropocene’ has is beginning to tease out key cause-and-effect linkages
caught the imagination, and it is being widely used in both that drive the global climate system. Moreover, while it
the popular and scientific literature. The problem, however, still remains the case that Milankovitch-induced variations
is over the status of the term in the geological record, in insolation are the principal forcing factors underlying
and whether a formal designation of the Anthropocene as glacial–interglacial cycles, it is equally clear that internal
a time-stratigraphic unit that is equivalent to a series or feedback mechanisms modulate the astronomical cycles,
epoch division within the geological timescale would and may promote irregular, non-symmetric climatic behav-
prove to be acceptable to the wider geological commun- iour. The bipolar seesaw is but one example of a number
ity. If not, then the term might simply become another name of important mechanisms that we need to understand if
for the Late Holocene, with a usage similar, perhaps, to ‘Dark we are to explain (often abrupt) climate change at the global
Ages’ or ‘Middle Ages’. In a geological sense, therefore, its scale. A future research agenda will need to examine the
status would remain unchanged, and it would continue to ways in which other components of global climate, such as
be employed informally to denote the recent period of the ITCZ, monsoon cells and ENSO, interconnect with the
accelerated human influence on the global environment. bipolar rhythm of climate change.
We know that the global climate system is in a con-
stant state of flux, with a variability that is hard to predict
7.7 CONCLUDING REMARKS (Figure 7.45), and this is one of the reasons why models
In this chapter we have reviewed the evidence for the pat- that attempt to postdict or predict the climate system
tern and tempo of climate change during the Quaternary, have large uncertainties. The new polar ice-core records
and considered some of the more important factors that have revealed that while long-term cycles are clearly evident,
appear to have influenced global climates over this time no two glacials or interglacials, or even DO events for that
period. In this final section of the book, we highlight matter, were exactly the same, in terms of either duration
some of the key themes that have emerged from recent or climate signal. Indeed, there is no precise analogue for
investigations into Quaternary environmental change, and the current interglacial or for any of its substages. This is
briefly consider their relevance to current concerns over partly because each climatic stage contains the imprint of
potential future climate and environmental change. a series of superimposed climatic rhythms (Figure 7.45).
Perhaps the most striking discovery to emerge from The phasing between these climate frequencies and their
recent palaeoclimatic research is the apparent rapidity with integrated product is constantly changing. In addition,
which the global climate system is able to shift from one each individual climatic event starts from a different
dominant mode to another. It is now widely accepted that combination of initial boundary conditions, for the earth’s
-35
-37
Susceptibilit
Susceptibilit
-39
-41
-43
-43
11400 11600 11800 12000 12200 12400 12600 12800 13000
A g e (ka)
1
S u s c eSp ut isbci lei tp t i b i l i t
-1
-2
11 12 13 14 15 16 17 18 19 20
A g e (ka b2k, G I C C 0 5 )
-34
Susceptibilit
-38
-42
-4S
5 0
Age(ka) 1 0 0
^5
Susceptibilit
Susceptibilit
-5
ep
S u s cS
•10
0 100 200 300 400 500 600 700 800

A g e E D C 3 (ka)
Figure 7.45 Schematic diagram showing superimposition of global climatic variations at different frequencies, from
glacial–interglacial (bottom) to centennial scale (top).
surface is also in a constant state of flux. Modelling these The impact that humans have had on earth surface
changes, and deriving predictions from them, represents processes and biota over what is a very small interval
one of the biggest challenges facing the science community, of geological time cannot be overestimated (Figure 7.46).
and it is one that will require improved stratigraphical While it is difficult to gauge human population numbers
records, more precise geochronological methods, more in prehistoric times, these are generally considered to be
robust and reliable proxy data, integrated global databases, no more than 2 million at around 50 ka. By the start of
and collaborations with the numerical modelling com- the Holocene, population may have increased to 5 million,
munity. As we hope this book has shown, these tools and and by the start of the Christian (or Common) Era was
their associated approaches have been dramatically refined around 300 million, the steep rise over the course of the
since 1990, and the foundations have been laid for the Holocene reflecting the combined benefits of farming,
multidisciplinary cooperation that is increasingly needed settlement and metal tool production. Figures are more
if we are to understand fully the workings, both spatially certain for recent historical times: global population has
and historically, of the global climate–environmental grown from 0.5 billion in AD 1650 to today’s total of nearly
system. 7 billion (Population Reference Bureau, 2011). At a number
A final point concerns the possible link between climate of points in this book, we have touched upon the various
change and humans, which we considered in section 7.6. impacts that this inexorable rise in global human popula-
The Quaternary record shows that we live in unusual times, tion may have had, especially with respect to the extinc-
with global atmospheric levels of CO2 and CH4 higher than tion, forced migration and range restriction of plant and
at any time during the last 800 ka. While some concerns that animal species, over-exploitation of resources (especially
have been raised about this matter could well prove to be water), forest destruction, soil degradation and so on. It is
exaggerated, they most certainly cannot be ignored, as they scarcely surprising, therefore, that this level of environ-
might prove prescient! For example, one prediction suggests mental impact, combined with the polluting effects of the
that the earth’s surface will be heated beyond what humans industrial revolution, and successive regional industrial-
can tolerate by AD 2300 (Sherwood & Huber, 2010). Other izations, has contributed to the current abnormal levels of
assessments suggest that atmospheric greenhouse gas greenhouse gases in our atmosphere (see Sapart et al.,
concentrations are approaching levels last seen during the 2012). Humans are arguably a product of climate change,
Pliocene, when the world was much warmer and sea levels stimulated into learning to cope with its constant upheavals
significantly higher than today (Raymo et al., 2011). and associated stresses (Figure 7.46), but they may also
Younge
Birth of A g r i c u l t u r e
Younge
Anatomically modern
h u m a n s leave Africa
Terminal decline
7
of N e a n d e r t h a l s -34
Younger Dryas Warm
l i st c e p t i b i l i t S u s c e p t i b i l i t
6
23 24 25
1 21
20
12 22
5 14 1 6 19
3 8 11 -38
7 10 15 17
4 56
13
4 2 9 18
3
i bui lsi tc e p t i bSi u
-42
2
S u s c e p tS
1 N G R I P Ice C o r e
Cold -45
H1 H2 H3 H4 H5 H6
0
0 50 100
A g e (ka)
Figure 7.46 Significant developments in human history plotted against the NGRIP climate record for the last glacial–interglacial
cycle. The total global human population (Population Reference Bureau, 2011, http://www.prb.org) is shown on the left of the
diagram.
be the architects of future distress, if some of the predic- NOTES

tions about future climate change and its consequences
are proved correct. It is here that Quaternary science can 1 The Rayleigh test is a statistical procedure for determining
play a leading role, for it not only informs the debate in whether a circular distribution is random or non-random; in
other words, whether the azimuths of a distribution are clumped
terms of the land–air–ocean–cryosphere processes, but it
in a particular direction.
also provides the all-important historical dimension, best 2 Stochastic resonance is a term used by physicists to describe a
exemplified by the contribution of the Intergovernmental situation where a signal that is normally too weak to be detected
Panel for Climate Change (IPCC), on international policy by a sensor can be boosted by adding white noise (a random
making. One of the basic dictums of Quaternary science, signal) that contains equal power across a range of frequencies.
as we noted in Chapter 1, is that the ‘present is the key to The frequencies within the white noise that correspond to those
the past’. But the past is also the key to the future, and it is in the original signal’s frequencies will resonate with each other,
by providing that key that Quaternary science can perhaps thereby amplifying the signal and making it easier to detect.
make its greatest contribution to the evaluation of future
risks to humankind.
References
Abbott, M.B., Wolfe, B.B., Aravena, R., et al. 2000. Holocene hydro- Agenbroad, L.D. 2012. Giants and pygmies: mammoths of Santa Rosa
logical reconstructions from stable isotopes and paleolimnology, Island, California. Quaternary International 255, 2–8.
Cordillera Real, Bolivia. Quaternary Science Reviews 19, Aguire, E., Pasini, G. 1985. The Pliocene–Pleistocene boundary.
1801–1820. Episodes, 8, 116–120.
Abell, P.I., Hoelzmann, P. 2000. Holocene palaeoclimates in Aguirre, M.L., Donato, M., Richiano, S., et al. 2011. Pleistocene and
northwestern Sudan: stable isotope studies on molluscs. Global Holocene interglacial molluscan assemblages from Patagonian
and Planetary Change 26, 1–12. and Bonaerensian littoral (Argentina, SW Atlantic):
Abellán, P., Benetti, C.J., Angus, R.B., et al. 2011. A review of palaeobiodiversity and palaeobiogeography. Palaeogeography,
Quaternary range shifts in European aquatic Coleoptera. Global Palaeoclimatology, Palaeoecology 308, 277–292.
Ecology and Biogeography 20, 87–100. Ahr, S.W., Nordi, L.C., Forman, S.L. 2013. Soil genesis, optical
Abelmann, A., Gowing, M.M. 1997. Spatial distribution pattern of dating, and geoarchaeological evaluation of two upland Alfisol
living polycystine radiolarian taxa – baseline study for paleo- pedons within the Tertiary Gulf Coastal Plain. Geoderma 192,
environmental reconstructions in the Southern Ocean (Atlantic 211–226.
sector). Marine Micropaleontology 30, 3–28. Aitken, M.J. 1990. Science-based Dating in Archaeology. Longman,
Abril, J.M. 2003. Constraints on the use of 137Cs as a time-marker to London.
support CRS and SIT chronologies. Environmental Pollution Aitken, M.J. 1998. An Introduction to Optical Dating. Oxford
129, 31–37. University Press, Oxford.
Adams, J.B., Mann, M.E., Ammann, C.M. 2003. Proxy evidence for Aizen, V.B., Aizen, E.M. 1996. Glaciers and snow cover in Central
an El Niño-like response to volcanic forcing. Nature 426, Asia as indicators of climate change in the Earth–Ocean–
274–278. Atmosphere system. In Regional Hydrological Response to Climate
Adams, J.M., Faure, H. 1998. A new estimate of changing carbon Change (edited by J.J.A. Jones). Kluwer Academic Publishers,
storage on land since the last glacial maximum, based on global Dordrecht, 269–286.
land ecosystem reconstruction. Global and Planetary Change Alfimov, A.V., Berman, D.I. 2009. Possible errors of the Mutual
16–17, 3–24. Climatic Range (MCR) method in reconstructing the Pleistocene
Adams, J.W., Herries, A.I.R., Kuykendall, K.L., et al. 2007. climate of Beringia. Entomological Review 89, 487–499.
Taphonomy of a South African cave: geological and hydrological Alho, P., Russell, A.J., Carrivick, J.L., et al. 2005. Reconstruction of
influences on the GD 1 fossil assemblage at Gondolin, a the largest Holocene jökulhlaup within Jökulsáá Fjöllum, NE
Plio–Pleistocene paleocave system in the Northwest Province, Iceland. Quaternary Science Reviews 24, 2319–2334.
South Africa. Quaternary Science Reviews 26, 2526–2543. Ali, A.A., Carcaillet, C., Guendon, J.-L., et al. 2003. The Early
Adams, K.D. 2010. Lake levels and sedimentary environments during Holocene treeline in the southern French Alps: new evidence from
deposition of the Trego Hot Springs and Wono tephras in the travertine formations. Global Ecology and Biogeography 12,
Lake Lahontan basin, Nevada, USA. Quaternary Research 73, 411–419.
118–129. Al-juaidi, F., Millington, A.C., Mclaren, S.J. 2003. Merged remotely
Adams, K.D., Locke, W.W., Rossi, R. 1992. Obsidian-hydration sensed data for geomorphological investigations in deserts:
dating of fluvially reworked sediments in the West Yellowstone examples from central Saudi Arabia. The Geographical Journal
region, Montana. Quaternary Research 38, 180–195. 169, 117–130.
Affek, H.P., Bar-Matthews, M., Ayalon, A., et al. 2008. Glacial/ Allard, G., Roy, M., Richard, P.J.H., et al. 2012. Constraining the age
interglacial temperature variations in Soreq Cave speleothems of the last interglacial–glacial transition in the Hudson Bay
as recorded by ‘clumped isotope’ thermometry. Geochimica et lowlands (Canada) using U–Th dating of buried wood.
Cosmochimica Acta 72, 5351–5360. Quaternary Geochronology 7, 37–47.
446 REFERENCES
Allen, J.R.M., Scourse, J.D., Hall, A.R., et al. 2009. Palaeo- stable isotopes and Sr and Mg chemistry of ostracod shells.
environmental context of the Late-glacial woolly mammoth Journal of Paleolimnology 42, 261–279.
(Mammuthus primigenius) discoveries at Condover, Shropshire, Anand, P., Elderfield, H., Conte, M.H. 2003. Calibration of Mg/Ca
UK. Geological Journal 44, 414–446. thermometry in planktonic foraminifera from a sediment trap
Allen, K., Drew, D.M., Downes, G.M. 2012. Ring width, climate and time series. Paleoceanography 18, doi: 10.1029/2002PA000846.
wood density relationships in two long-lived Tasmanian tree Andersen, B.G., Mangerud, J., Sørensen, R., et al. 1995. Younger Dryas
species. Dendrochronologia 30, 167–177. ice-marginal deposits in Norway. Quaternary International 28,
Alley, R.B. 2000. Ice-core evidence of abrupt climate changes. 147–169.
Proceedings of the National Academy of Sciences of the United States Andersen, K.K., Svensson, A., Rasmussen, S.O., et al. 2006. The Green-
of America 97, 1331–1334. land Ice Core Chronology 2005, 15–42 ka. Part 1: constructing
Alley, R.B. 2007. Wally was right: predictive ability of the North the time scale. Quaternary Science Reviews 25, 3246–3257.
Atlantic ‘conveyor belt’ hypothesis for abrupt climate change. Anderson, D. 2001. Attenuation of millennial-scale events by
Annual Review of Earth and Planetary Sciences 35, 241–272. bioturbation in marine sediments. Paleoceanography 16, doi:
Alley, R.B., MacAyeal, D.R. 1994. Ice-rafted debris associated with 10.1029/2000PA000530.
binge/purge oscillations of the Laurentide Ice Sheet. Anderson, J. 2000. What is the origin of the carabid beetle fauna of
Paleoceanography 9, 503–511. dry, anthropogenic habitats in western Europe? Journal of
Alley, R.B., Gow, A.J., Meese, D.A., et al. 1997. Grain-scale processes, Biogeography 27, 795–806.
folding, and stratigraphic disturbances in the GISP2 ice core. Anderson, R.F., Ali, S., Bradtmiller, L.I., et al. 2009. Wind-driven
Journal of Geophysical Research 102, 26819–26830. upwelling in the Southern Ocean and the deglacial rise in atmos-
Alley, R.B., Clark, P.U., Huybrechts, P., et al. 2005. Ice-sheet and sea- pheric CO2. Science 323, 1443–1448.
Andersson, S., Schoning, K. 2010. Surface wetness and mire
level changes. Science 310, 456–460.
development during the late Holocene in central Sweden. Boreas
Alloway, B.V., Larsen, G., Lowe, D.J., et al. 2007a. Tephrochronology.
39, 749–760.
In Encyclopedia of Quaternary Science (edited by S.A. Elias).
André, M.-F. 2009. From Climatic to Global Change Geomorphology:
Elsevier, Amsterdam, 2869–2898.
Contemporary Shifts in Periglacial Geomorphology. Geological
Alloway, B.V., Lowe, D.J., Barrell, D.J.A., et al. 2007b. Towards a
Society of London, Special Publication 320. Blackwell Scientific
climate event stratigraphy for New Zealand over the past 30 000
Publications, Oxford, 5–28.
years (NZ-INTIMATE project). Journal of Quaternary Science
Andrén, T., Björck, J., Johnsen, S. 1999. Correlation of Swedish glacial
22, 9–35.
varves with the Greenland (GRIP) oxygen isotope record. Journal
Almond, P., Roering, J., Hales, T.C. 2007. Using soil residence time of Quaternary Science 14, 361–371.
to delineate spatial and temporal patterns of transient landscape Andres, M.S., Bernasconi, S.M., McKenzie, J.A., et al. 2003. Southern
response. Journal of Geophysical Research 112, F03S17, doi: Ocean deglacial record supports global Younger Dryas. Earth and
10.1029/2006JF000568. Planetary Science Letters 216, 515–524.
Alonso-Zarza, A.M., Wright, V.P. 2010. Calcretes. Developments in Andrews, J.E., Brasier, A.T. 2005. Seasonal records of climatic change
Sedimentology 61, 225–267. in annually laminated tufas: short review and future prospects.
Ambrose, S.H. 1998. Late Pleistocene human population bottlenecks, Journal of Quaternary Science 20, 411–421.
volcanic winter, and differentiation of modern humans. Journal Andrews, J.T. 1970. A geomorphological study of Postglacial uplift
of Human Evolution 34, 623–651. with particular reference to Arctic Canada. Institute of British
Ambrose, W.R. 1994. Obsidian hydration dating of a Pleistocene age Geographers, Special Publication 2.
site from the Manus Islands, Papua New Guinea. Quaternary Andrews, J.T., Tedesco, K. 1992. Detrital carbonate-rich sediments,
Science Reviews 13, 137–142. northwestern Labrador Sea: implications for ice-sheet dynamics
American Commission on Stratigraphic Nomenclature. 1961. Code and iceberg rafting (Heinrich) events in the North Atlantic.
of Stratigraphic Nomenclature. American Association of Petroleum Geology 20, 1087–1090.
Geologists Bulletin 45, 645–665. Andrews, J.T., Principato, S.M. 2002. Grain-size Characteristics and
American Commission on Stratigraphic Nomenclature. 1970. Code Provenance of Ice-proximal Glacial Marine Sediments. Geological
of Stratigraphic Nomenclature, 2nd edn. American Association Society of London, Special Publication 203. Blackwell Scientific
of Petroleum Geologists Bulletin 60, 1–45. Publications, Oxford, 305–324.
Amorosi, A., Centineo, M.C., Colalongo, M.L., et al. 2003. Facies Andrews, J.T., Dyke, A.S. 2007. Late Quaternary in North America.
architecture and latest Pleistocene–Holocene depositional In Encyclopedia of Quaternary Science (edited by S.A. Elias).
history of the Po Delta (Comaccio area), Italy. Journal of Geology Elsevier, Amsterdam, 1095–1101.
111, 39–56. Andrews, J.T., Jennings, A.E., Kirby, K.M., et al. 1995. A Heinrich-
An, Z. 2000. The history and variability of the East Asian paleo- like event, H-0 (DC-0): source(s) for detrital carbonate in the
monsoon climate. Quaternary Science Reviews 19, 171–187. North Atlantic during the Younger Dryas chronozone.
Anadón, P., Gabàs, M. 2009. Paleoenvironmental evolution of Paleoceanography 10, 943–952.
the Early Pleistocene lacustrine sequence at Barranco León Andrews, J.T., Hardadottir, J., Stoner, J.S., et al. 2003. Decadal to
archeological site (Orce, Baza Basin, Southern Spain) from millennial-scale periodicities in North Iceland shelf sediments
REFERENCES 447
over the last 12 000 cal yr: long-term North Atlantic oceano- Ascough, P., Cook, G., Dugmore, A. 2005. Methodological
graphic variability and solar forcing. Earth and Planetary Science approaches to determining the marine radiocarbon reservoir
Letters 210, 453–465. effect. Progress in Physical Geography 29, 532–547.
Andrus, C.F.T. 2011. Shell midden sclerochronology. Quaternary Ashworth, P.C., Buckland, P.C., Sadler, J.P. (eds). 1997. Studies in
Science Reviews 30, 2892–2905. Quaternary Entomology: An Inordinate Fondness for Insects.
Anisimov, O.A., Nelson, F.E. 1996. Permafrost distribution in the Quaternary Proceedings 5, John Wiley, Chichester and New
Northern Hemisphere under scenarios of climatic change. Global York.
and Planetary Change 14, 59–72. Atkinson, T.C., Rowe, P.J. 1992. Applications of dating to denuda-
Anovitz, L.M., Elam, J.M., Riciputi, L.R., et al. 1999. The failure of tion chronology and landscape evolution. In Uranium-Series
obsidian hydration dating: sources, implications and new Disequilibrium: Applications to Earth, Marine and Environmental
directions. Journal of Archaeological Science 26, 735–752. Science, 2nd edn (edited by M. Ivanovich, R.S. Harmon).
Anthony, D.M., Granger, D.E. 2007. A new chronology of Appala- Clarendon, Oxford, 669–702.
chian erosional surfaces determined by cosmogenic nuclides Atkinson, T.C., Briffa, K.R., Coope, G.R. 1987. Seasonal temperatures
in cave sediments. Earth Surface Processes and Landforms 32, in Britain during the last 22,000 years, reconstructed using beetle
874–887. remains. Nature 325, 587–592.
Antoine, P., Lautridou, J.-P., Laurent, M. 2000. Long-term fluvial Auban, J.B., Barton, C.M., Rapol, M.P. 2001. A taphonomic
archives in NW France: response of the Seine and Somme perspective on Neolithic beginnings: theory, interpretation
rivers to tectonic movements, climatic variations and sea-level and empirical data in the western Mediterranean. Journal of
changes. Geomorphology 33, 183–207. Archaeological Science 28, 597–612.
Antoine, P., Catt, J., Lautridou, J.-P., et al. 2003. The loess and Aubry, S., Magnin, F., Bonnet, V., et al. 2005. Multi-scale altitudinal
coversands of northern France and southern England. Journal patterns in species richness of land snail communities in south-
of Quaternary Science 18, 309–318.
eastern France. Journal of Biogeography 32, 985–998.
Antonioli, F., Bard, E., Potter, E.-K., et al. 2004. 215-ka history
Auler, A.S., Wang, X., Edwards, R.L., et al. 2004. Quaternary
of sea-level oscillations from marine and continental layers in
ecological and geomorphic changes associated with rainfall
Argentarola Cave speleothems (Italy). Global and Planetary
events in presently semi-arid northeastern Brazil. Journal of
Change 43, 57–78.
Quaternary Science 19, 693–701.
Ao, H., Dekkers, M.J., Qin, L., et al. 2011. An updated astronomical
Ayalon, A., Bar-Matthews, M., Kaufman, A. 1999. Petrography,
timescale for the Plio–Pleistocene deposits from South China
strontium, barium and uranium isotope ratios in speleothems
Sea and new insights into Asian monsoon evolution. Quaternary
a paleoclimate proxies: Soreq Cave, Israel. The Holocene 9,
Science Reviews 30, 1560–1575.
715–722.
Appleby, P.G. 2008. Three decades of dating recent sediments by
fallout nuclides. The Holocene 18, 83–93. Ayalon, A., Bar-Matthews, M., Kaufman, A. 2002. Climatic con-
Appleby, P.G., Oldfield, F. 1992. Applications of lead-210 to ditions during marine oxygen isotope stage 6 in the eastern
sedimentation studies. In Uranium Series Disequilibrium, Appli- Mediterranean region from the isotopic composition of speleo-
cations to Earth, Marine and Environmental Sciences (edited thems of Soreq Cave, Israel. Geology 30, 303–306.
by M. Ivanovich, R.S. Harmon). Clarendon Press, Oxford, Aycliffe, L.K., Marianelli, P.C., Moriarty, K.C., et al. 1998. 500 ka
731–779. precipitation record from southeastern Australia: evidence for
Applegarth, M.T., Dahms, D.E. 2001. Soil catenas of calcareous interglacial relative aridity. Geology 26, 147–150.
tills, Whiskey Basin, Wyoming, USA. Catena 42, 17–38. Bacon, C.R., Lanphere, M.A. 2006. Eruptive history and geo-
Aref, M.A.M. 2003. Classification and depositional environments chronology of Mount Mazama and the Crater Lake region,
of Quaternary pedogenic gypsum crusts (gypcrete) from east Oregon. Geological Society of America Bulletin 118, 1331–1359.
of the Fayum, Depression (Egypt). Sedimentary Geology 155, Badding, M.E., Briner, J.P., Kaufman, D.S. 2013. 10Be ages of late
87–108. Pleistocene deglaciation and Neoglaciation in the north-central
Arge, S.L., Sveinbjarnardótti, G., Edwards, K.J., et al. 2005. Viking Brooks Range, Arctic Alaska. Journal of Quaternary Science 28,
and Medieval settlement in the Faroes: people, place and 95–102.
environment. Human Ecology 33, 597–620. Baier, J., Lücke, A., Negendank, J.F.W., et al. 2004. Diatom and
Arimoto, R. 2001. Eolian dust and climate: relationships to sources, geochemical evidence of mid- to late-Holocene climatic changes
tropospheric chemistry, transport and deposition. Earth-Science as Lake Holzmaar, West-Eifel (Germany). Quaternary Inter-
Reviews 54, 29–42. national 113, 81–96.
Armbrust, E.V. 2009. The life of diatoms in the world’s oceans. Nature Bailey, G.N., Flemming, N.C. 2008. Archaeology of the conti-
459, 185–192. nental shelf: marine resources, submerged landscapes and
Armstrong, H., Brasier, M. 2005. Microfossils. Blackwell Publishing, underwater archaeology. Quaternary Science Reviews 27,
Oxford and Malden, MA. 2153–2165.
Arnold, N., Sharp, M. 2002. Flow variability in the Scandinavian Baillie, M.G.L. 1982. Tree Ring Dating and Archaeology. Croom
ice sheet: modelling the coupling between ice sheet flow and Helm, London.
hydrology. Quaternary Science Reviews 21, 485–502. Baillie, M.G.L. 1995. A Slice Through Time. Batsford, London.
448 REFERENCES
Baily, B., Collier, P., Farres, P., et al. 2002. Comparative assessment Ballantyne, C.K. 2002b. The Loch Lomond Readvance on the Isle
of analytical and digital photogrammetric methods in the of Mull, Scotland: glacier reconstruction and palaeoclimatic
construction of DEMs of geomorphological forms. Earth Surface implications. Journal of Quaternary Science 17, 759–771.
Processes and Landforms 28, 307–320. Ballantyne, C.K. 2007a. Loch Lomond Stadial glaciers in North
Bajo, P., Drysdale, R., Woodhead, J., et al. 2012. High-resolution Harris, Outer Hebrides, north-west Scotland: glacier reconstruc-
U–Pb dating of an Early Pleistocene stalagmite from Corchia tion and palaeoclimatic implications. Quaternary Science Reviews
Cave (central Italy). Quaternary Geochronology 14, 5–17. 26, 3134–3149.
Baker, A., Caseldine, C.J., Gilmour, M.A., et al. 1999. Stalagmite Ballantyne, C.K. 2007b. The Loch Lomond Readvance on north
luminescence and peat humification records of palaeo- Arran, Scotland: glacier reconstructions and palaeoclimatic
moisture for the last 2500 years. Earth and Planetary Science Letters implications. Journal of Quaternary Science 22, 343–359.
165, 157–162. Ballantyne, C.K. 2010. Extent and deglacial chronology of the last
Baker, A., Asrat, A., Fairchild, I.J., et al. 2007. Analysis of the British–Irish Ice Sheet: implications of exposure dating using
climate signal contained within δ18O and growth rate parameters cosmogenic isotopes, Journal of Quaternary Science 25, 515–534.
in two Ethiopian stalagmites. Geochimica et Cosmochimica Acta Ballantyne, C.K., Kirkbride, M.P. 1986. The characteristics and
71, 2975–2988. significance of some Lateglacial protalus ramparts in upland
Baker, A.C., Glynn, P.W., Riegl, B. 2008a. Climate change and coral Britain. Earth Surface Processes and Landforms 11, 659–671.
reef bleaching: An ecological assessment of long-term impacts, Ballantyne, C.K., Benn, D.I. 1994. Glaciological constraints on
recovery trends and future outlook. Estuarine, Coastal and Shelf protalus rampart development. Permafrost and Periglacial
Science 80, 435–471. Processes 5, 145–153.
Baker, A., Smith, C.J., Jex, C., et al. 2008b. Annually laminated Ballantyne, C.K., Harris, C. 1994. The Periglaciation of Great Britain.
speleothems: a review. International Journal of Speleology 37, Cambridge University Press, Cambridge.
193–206. Ballantyne, C.K., McCarroll, D., Nesje A., et al. 1998. The last ice sheet
Bakke, J., Dahl, S.O., Nesje, A. 2005. Lateglacial and early Holocene in north-west Scotland: reconstruction and implications.
palaeoclimatic reconstruction based on glacier fluctuations and Quaternary Science Reviews 17, 1149–1184.
equilibrium-line altitudes at northern Folgefonna, Hardanger, Ballantyne, C.K., Stone, J.O., McCarroll, D. 2008. Dimensions and
western Norway. Journal of Quaternary Science 20, 179–198. chronology of the last ice sheet in Western Ireland. Quaternary
Bakke, J., Lie, Ø., Dahl, S.O., et al. 2008. Strength and spatial patterns Science Reviews 27, 185–200.
of the Holocene wintertime westerlies in the NE Atlantic region. Ballantyne, C.K., Schnabel, C., Xu, S. 2009. Readvance of the last
Global and Planetary Change 60, 28–41. British–Irish Ice Sheet during Greenland Interstade 1 (GI-1): the
Bakke, J., Lie, Ø., Heegaard, E., et al. 2009. Rapid oceanic and atmos- Wester Ross Readvance, NW Scotland. Quaternary Science
pheric changes during the Younger Dryas cold period. Nature Reviews 28, 783–789.
Geoscience 2, 202–205. Balsam, W., Ji, J., Chen, J. 2004. Climatic interpretation of the
Balakrishnan, M., Yapp, C.J., Meltzer, D.J., et al. 2005. Paleo- Luochan and Lingtai loess sections, China, based on changing
environment of the Folsom archaeological site, New Mexico, iron oxide mineralogy and magnetic susceptibility. Earth and
USA, approximately 10,500 14C yr B.P. as inferred from the Planetary Science Letters 223, 335–348.
stable isotope composition of fossil land snail shells. Quaternary Bamber, J.L., Rivera, A. 2007. A review of remote sensing methods
Research 63, 31–44. for glacier mass balance determination. Global and Planetary
Balascio, N.L., Kaufman, D.S., Briner, J.P., Manley, W.F. 2005. Change 59, 138–148.
Equilibrium-line altitudes during the Last Glacial Maximum Bamber, J.L., Alley, R.B., Joughin, I. 2007. Rapid response of modern
across the Brooks Range, Alaska. Journal of Quaternary Science day ice sheets to external forcing. Earth and Planetary Science
20, 821–838. Letters 257, 1–13.
Balco, G. 2011. Contributions and unrealised potential contributions Bannister, A., Raymond, S., Baker, R. 1998. Surveying, 7th edn.
of cosmogenic-nuclide exposure dating to glacier chronology. Prentice Hall, Harlow.
Quaternary Science Reviews 30, 3–27. Barbante C., Fischer., Masson-Delmotte, V., et al. 2010. Climate of
Balco, G., Stone, J.O., Lifton, N., et al. 2008. A complete and easily the last million years: new insights from EPICA and other
accessible means of calculating surface exposure ages or erosion records. Quaternary Science Reviews 29, 1–7.
rates from 10Be and 26Al measurements. Quaternary Geo- Barber, D.C., Dyke, A., Hillaire-Marcel, C., et al. 1999. Forcing of
chronology 3, 174–195. the cold event of 8,200 years ago by catastrophic drainage of
Balco, G., Schaefer, J.M., LARISSA Group. 2013. Exposure-age Laurentide lakes. Nature 400, 344–348.
record of Holocene ice sheet and ice shelf change in the north- Barber, K.E. 1981. Peat Stratigraphy and Climatic Change: A
east Antarctic peninsula. Quaternary Science Reviews 59, 101–111. Palaeoecological Test of the Theory of Cyclic Peat Bog Regeneration.
Ballantyne, C.K., 1989. The Loch Lomond Advance on the Isle of Skye, Balkema, Rotterdam.
Scotland: glacier reconstruction and palaeoclimatic implications. Barber, K.E., Langdon, P.G. 2007. What drives the peat-based
Journal of Quaternary Science 4, 95–108. palaeoclimate record? A critical test using multi-proxy climate
Ballantyne, C.K. 2002a. Paraglacial geomorphology. Quaternary records from northern Britain. Quaternary Science Reviews 26,
Science Reviews 21, 1935–2017. 3318–3327.
REFERENCES 449
Barber, K.E., Chambers, F.M., Maddy, D., et al. 1994. A sensitive Barnes, I., Barnett, R. Shapiro, B. 2007. Ancient DNA. In Encyclopedia
high resolution record of Late Holocene climatic change from a of Quaternary Science (edited by S.A. Elias). Elsevier, Amsterdam,
raised bog in Northern England. The Holocene 4, 198–205. 3249–3253.
Barber, K.E., Maddy, D., Rose, N., et al. 2000. Replicated proxy-climate Barnosky, A.D., Matzke, N., Tomiya, S., et al. 2011. Has the Earth’s
signals over the last 2000 years from two distant UK peat bogs: sixth mass extinction already arrived? Nature 471, 51–57.
new evidence for regional palaeoclimatic teleconnections. Bar-Or, R., Erlick, C., Gildor, H. 2008. The role of dust in glacial–
Quaternary Science Reviews 19, 481–487. interglacial cycles. Quaternary Science Reviews 27, 201–208.
Barber, K.E., Chambers, F.M., Maddy, D. 2003. Holocene palaeo- Bar-Oz, G., Nadel, D., 2013. Worldwide large-scale trapping and
climates from peat stratigraphy: macrofossil proxy climate hunting of ungulates in past societies. Quaternary International
records from three oceanic raised bogs in England and Ireland. 297, 1–7.
Quaternary Science Reviews 22, 521–539. Barraclough, T.G., Vogler, A.P. 2002. Recent diversification rates
Barber, K.E., Chambers, F.M., Maddy, D. 2004. Late Holocene in North America tiger beetles estimated from a dated mtDNA
climatic history of northern Germany and Denmark: peat macro- phylogenetic tree. Molecular Biology and Evolution 19, 1706–
fossil investigations at Dosenmoor, Schleswig-Holstein, and 1716.
Svanemose, Jutland. Boreas 33, 132–144. Barreiro, M., Fedorov, A., Pacanowski, R., et al. 2008. Abrupt climate
Barbour, M.M. 2007. Stable oxygen isotope composition of plant changes: how freshening of the northern Atlantic affects the
tissue: a review. Functional Plant Biology 34, 83–94. thermohaline and wind-driven oceanic circulations. Annual
Barclay, D.J., Wiles, G.C., Calkin, P.E. 2009. Holocene glacier Review of Earth and Planetary Sciences 36, 33–58.
fluctuations in Alaska. Quaternary Science Reviews 28, 2034– Barreto, A.M.F., Bezerra, F.H.R., Suguio, K., et al. 2002. Late
2948. Pleistocene marine terrace deposits in northeastern Brazil: sea-
Bard, E., Arnold, M., Duplessy, J.-C. 1991. Reconciling the sea level change and tectonic implications. Palaeogeography,
Palaeoclimatology, Palaeoecology 179, 57–69.
level record of the last deglaciation with 18O spectra from deep
Barrett, B.E., Nicholls, K.W., Murray, T., et al. 2009. Rapid recent
sea cores. In Radiocarbon Dating: Recent Applications and Future
warming on Rutford Ice Stream, West Antarctica, from borehole
Potential (edited by J.J. Lowe). Quaternary Proceedings 1,
thermometry. Geophysical Research Letters 36, doi: 10.1029/
Quaternary Research Association, Cambridge, 67–73.
2008GL036369.
Bard, E., Hamelin, B., Delanghe-Sabatier, D. 2010. Deglacial
Barron, J.A., Bukry, D. 2007. Development of the California
meltwater pulse 1B and Younger Dryas sea levels revisited with
Current during the past 12,000 yr based on diatoms and silico-
boreholes at Tahiti. Science 327, 1235–1237.
flagellates. Palaeogeography, Palaeoclimatology, Palaeoecology
Barker, G.M. (ed.). 2001. The Biology of Terrestrial Molluscs. CABI
248, 313–338.
Publishing, Wallingford.
Barron, J.A., Anderson, L. 2010. Enhanced Late Holocene ENSO/
Barker, S., Greaves, M., Elderfield, H. 2003. A study of cleaning PDO expression along the margins of the eastern North Pacific.
procedures used for foraminiferal Mg/Ca paleothermometry. Quaternary International 235, 3–12.
Geochemistry, Geophysics, Geosystems 4, doi: 10.1029/2003 Barrows, T.T., Juggins, S. 2005. Sea-surface temperatures around the
GC000559. Australian margin and Indian Ocean during the Last Glacial
Barker, S., Cacho, I., Benway, H., et al. 2005. Planktonic foraminiferal Maximum. Quaternary Science Reviews 24, 1017–1047.
Mg/Ca as a proxy for past oceanic temperatures: a methodo- Bartoli, G. Sarnthein, M., Weinelt, M., et al. 2005. Final closure of
logical overview and data compilation for the Last Glacial Panama and the onset of northern hemisphere glaciation. Earth
Maximum. Quaternary Science Reviews 24, 821–834. and Planetary Science Letters 237, 33–44.
Barker, S., Diz, P., Vautravers, M.J., et al. 2009. Interhemispheric Barton, R.N.E., Berridge, P.J., Walker, M.J.C., et al. 1995. Persistent
Atlantic seesaw response during the last deglaciation. Nature 457, places in the Mesolithic landscape: an example from the Black
1007–1102. Mountain Uplands of south Wales. Proceedings of the Prehistoric
Barletta, F., St-Onge, G., Channell, J.E.T., et al. 2008. High-resolution Society 61, 81–116.
palaeomagnetic secular variation and relative palaeointensity Barton, R.N.E., Bouzouggar, A., Collcutt, S.N., et al. 2009. OSL
records from the eastern Canadian Arctic: implications for dating of the Aterian levels at Dar es-Soltan 1 (Rabat, Morocco)
Holocene stratigraphy and geomagnetic field behaviour. and implications for the dispersal of modern Homo sapiens.
Canadian Journal of Earth Sciences 45, 1265–1281. Quaternary Science Reviews 28, 1914–1931.
Bar-Matthews, M., Ayalon, A., Kaufman, A. 1997. Late Quaternary Bassinot, F.C., Labeyrie, L., Vincent, E., et al. 1994. The astronomical
paleoclimate in the eastern Mediterranean region from stable theory of climate and the age of the Brunhes–Matuyama
isotope analysis of speleothems at Soreq Cave, Israel. Quaternary magnetic reversal. Earth and Planetary Science Letters 126, 91–108.
Research 47, 155–168. Bateman, M.D., Carr, A.S., Murray-Wallace, C.V., et al. 2008.
Bar-Matthews, M., Ayalon, A., Gilmour, M., et al. 2003. Sea-land A dating intercomparison study of Late Stone Age midden
oxygen isotope relationships from planktonic foraminifera and deposits, South Africa. Geoarchaeology 23, 715–741.
speleothems in the Eastern Mediterranean region and their Bathurst, R.R., Zori, D., Byock, J. 2010. Diatoms as bio-indicators
implication for palaeorainfall during interglacial intervals. of site use: locating turf structures from the Viking Age. Journal
Geochimica et Cosmochimica Acta 67, 3181–3199. of Archaeological Science 37, 2920–2928.
450 REFERENCES
Batt, C.M. 1997. The British archaeomagnetic calibration curve: Behre, K.-E., van der Plicht, J. 1992. Towards an absolute chron-
an objective treatment. Archaeometry 39, 153–168. ology for the last glacial period in Europe: radiocarbon dates from
Battarbee, R.W. 2000. Palaeolimnological approaches to climate Oerel, northern Germany. Vegetation History and Archaeobotany
change, with special regard to the biological record. Quaternary 1, 111–117.
Science Reviews 19, 107–124. Behrensmeyer, A.K. 1988. Vertebrate preservation in fluvial chan-
Battarbee, R.W., Bennion, H. 2011. Palaeolimnology and its develop- nels. Palaeogeography, Palaeoclimatology, Palaeoecology 63,
ing role in assessing the history and extent of human impact on 183–199.
lake ecosystems. Journal of Paleolimnology 45, 399–404. Belanger, C.L., Jablonski, D., Roy, K., et al. 2012. Global environmental
Battarbee, R.W., Jones, V.J., Flower, B.P., et al. 2001. Diatoms. In predictors of benthic marine biogeographic structure. Proceedings
Tracking Environmental Change Using Lake Sediments. Volume of the National Academy of Sciences of the United States of America
3: Terrestrial, Algal and Siliceous Indicators (edited by J.P. Smol., 109, 14046–14051.
H.J.B. Birks, E.F. Stoermer). Kluwer, Dordrecht, 155–201. Bell, M., Walker, M.J.C. 2005. Late Quaternary Environmental
Battarbee, R.W., Charles, D.F., Bigler, C., et al. 2010. Diatoms as Change: Physical and Human Perspectives. Pearson: Prentice
indicators of surface-water acidity. In The Diatoms: Applica- Hall, London.
tions for the Environmental and Earth Sciences, 2nd edn (edited Belt, S.T., Müller, J. 2013. The Arctic sea ice biomarker IP25:
by J.P. Smol, E.F. Stoermer). Cambridge University Press, a review of current understanding, recommendations for
Cambridge, 98–121. future research and applications in palaeo sea ice reconstruc-
Baumann, K.-H., Andruleit, H., Bockel, B., et al. 2005. The signifi- tions. Quaternary Science Reviews doi: 10.1016/j.quascirev.2012.
cance of extant coccolithophores as indicators of ocean water 12.001.
masses, surface water temperature, and palaeoproductivity: Benedict, J.B. 2009. A review of lichenometric dating and its appli-
a review. Paläontologische Zeitschrift 79, 93–112. cations to archaeology. American Antiquity 74, 143–172.
Baumgartner, P.O., Aitchison, J.C., De Wever, P., et al. (eds). 2006. Benn, D.I. 1994. Fluted moraine and till genesis beneath a temperate
Radiolaria: Siliceous Plankton Through Time. Eclogae Geologicae glacier: Slettmarkbreen, Jotunheimen, Norway. Sedimentology 41,
Helvetiae Supplement 1, S63–S66, Birkhäuser, Basel. 279–292.
Bay, R.C., Bramall, N., Price, P.B. 2004. Bipolar correlation of Benn, D.I. 1997. Glacier fluctuations in western Scotland. Quaternary
volcanism with millennial climate change. Proceedings of the International 38–39, 137–147.
National Academy of Sciences of the United States of America 101, Benn, D.I. 2004. Macrofabric. In A Practical Guide to the Study of
6341–6345. Glacial Sediments (edited by D.J.A. Evans, D.I. Benn). Arnold,
Bayer, M.M., Juggins, S. 2002. ADIAC imaging techniques and London, 93–114.
databases. In Automated Diatom Identification (edited by H. du Benn, D.I., Lemkuhl, F. 2000. Mass balance and equilibrium-line
Buf, M.M. Bayer). World Scientific Press, Singapore, 55–73. altitudes of glaciers in high-mountain environments. Quaternary
Bazin, L., Landais, A., Lemiuex-Dudon, B., et al. 2012. An optimised International 65/66, 15–29.
multi-proxy, multi-site Antarctic ice and gas orbital chronology Benn, D.I., Ringrose, S. 2001. Random variation of fabric eigenvalues:
(AICC2012): 120–180 ka. Climate of the Past Discussions 8, implications for the use of a-axis fabric data to differentiate till
5963–6009. facies. Earth Surface Processes and Landforms 26, 295–306.
Beaudoin, A.B. 2007. On the laboratory procedure for processing Benn, D.I., Ballantyne, C.K. 2005. Palaeoclimatic reconstruction
unconsolidated sediment samples to concentrate subfossil from Loch Lomond Readvance glaciers in the West Drumochter
seed and other plant macro-remains. Journal of Paleolimnology Hills, Scotland. Journal of Quaternary Science 20, 577–592.
37, 301–308. Benn, D.I., Lukas, S. 2006. Younger Dryas glacial landsystems in
Beck, C., Jones, G.T. 1994. Dating surface assemblages using obsidian north west Scotland: an assessment of modern analogues and
hydration. In Dating in Exposed and Surface Contexts (edited by palaeoclimatic implications. Quaternary Science Reviews 25,
C. Beck). University of New Mexico Press, Albuquerque, 47–75. 2390–2408.
Beck, J.W., Donahue, D.J., Jull, A.J.T., et al. 1998. Ambiguities in direct Benn, D.I., Evans, D.J.A. 2010. Glaciers and Glaciation, 2nd edn.
dating of rock surfaces using radiocarbon measurements. Science Hodder-Arnold, London.
280, 2132–2135. Benn, D.I., Hulton, N.R.J. 2010. An Excel TM spreadsheet program
Beck, J.W., Richards, D.A., Edwards, R.L., et al. 2001. Extremely for reconstructing the surface profile of former mountain glaciers
large variations in atmospheric 14C concentrations during the last and icecaps. Computers and Geosciences 36, 605–610.
glacial period. Science 292, 2453–2458. Benn, D.I., Kirkbride, M.P., Owen, L.A., et al. 2003. Glaciated valley
Beerling, D.J., Royer, D.L. 2002. Reading a CO2 signal from fossil landsystems. In Glacial Landsystems (edited by D.J.A. Evans).
stomata. New Phytologist 153, 387–397. Arnold, London, 372–406.
Begon, M., Townsend, C.R., Harper, J.L. 2006. Ecology: From Benn, D.I., Owen, L.A., Osmaston, H.A., et al. 2005. Reconstruction
Individuals to Ecosystems, 4th edn. Blackwell Publishing, Malden, of equilibrium-line altitudes for tropical and sub-tropical glaciers.
MA and Oxford. Quaternary International 138–139, 8–21.
Behre, K.-E. (ed.). 1986. Anthropogenic Indicators in Pollen Diagrams. Bennett, K.D. 1996. Determination of the number of zones in a
Balkema, Rotterdam. biostratigraphical sequence. New Phytologist 132, 155–170.
REFERENCES 451
Bennett, K., Willis, K.J. 2001. Pollen. In Tracking Environmental richness, vegetation dynamics and land-use. Vegetation History
Change Using Lake Sediments. Volume 3: Terrestrial, Algal and and Archaeobotany 17, 573–583.
Siliceous Indicators (edited by J.P. Smol., H.J.B. Birks, E.F. Bergman, J., Wastegård, S., Hammarlund, D., et al. 2004. Holocene
Stoermer). Kluwer Academic Publishers (Springer), Dordrecht tephra horizons at Klocka Bog, west-central Sweden: aspects of
and Boston, 5–30. reproducibility in subarctic peat deposits. Journal of Quaternary
Bennett, M.R., Glasser, N.F. 2009. Glacial Geology: Ice Sheets and Science 19, 241–249.
Landforms, 2nd edn. Wiley-Blackwell, Chichester and Oxford. Bergner, A.G.N., Trauth, M.H., Bookhagen, B. 2003. Paleo-
Bennike, O., Anderson, N.J., McGowan, S. 2010. Holocene palaeoprecipitation estimates for the Lake Naivasha basin (Kenya)
ecology of southwest Greenland inferred from macrofossils in during the last 175 k.y. using a lake-balance model. Global and
sediments of an oligosaline lake. Journal of Paleolimnology 43, Planetary Change 36, 117–136.
787–798. Bergner, A.G.N., Strecker, M.R., Trauth, M.H., et al. 2009. Tectonic
Benson, L., Paillet, F.L. 2002. HIBAL: a hydrologic-isotopic-balance and climate control on evolution of rift lakes in the Central
model for application to paleolake systems. Quaternary Science Kenya Rift, East Africa. Quaternary Science Reviews 28, 2804–
Reviews 21, 1521–1539. 2816.
Benson, L.V., Lund, S.P., Burdett, J.W., et al. 1998. Correlation of Berkeley, A., Perry, C.T., Smithers, S.G., et al. 2007. A review of the
late-Pleistocene lake-level oscillations in Mono Lake, California, ecological and taphonomic controls on foraminiferal assem-
with North Atlantic climate events. Quaternary Research 49, blage development in intertidal environments. Earth-Science
1–10. Reviews 83, 205–230.
Benson, L., Kashgarian, M., Rye, R., et al. 2002. Holocene multi- Bernasconi, M.P., Stanley, J.-D., Caruso, C. 2010. Sybaris–
decadal and multicentennial droughts affecting northern Cali- Thuri–Copia deltaic settings in Calabria, Italy: molluscs, associ-
fornia and Nevada. Quaternary Science Reviews 21, 659–682. ated biogenic components and ecobiostratigraphy applied to
Benson, L., Lund, S., Negrini, R., et al. 2003. Response of North archaeology. Journal of Coastal Research 26, 377–390.
American Great Basin Lakes to Dansgaard–Oeschger oscillations.
Bernet, M., Brandon, M.T., Garver, J.I., et al. 2004. Downstream
Quaternary Science Reviews 22, 2239–2251.
changes of Alpine zircon fission-track ages in the Rhône and
Benson, L., Madole, R., Phillips, W., et al. 2004. The probable
Rhine Rivers. Journal of Sedimentary Research 74, 82–94.
importance of snow and sediment shielding on cosmogenic ages
Berthier, E., Schiefer, E., Clarke, G.K.C., et al. 2010. Contribution of
of north-central Colorado Pinedale and pre-Pinedale moraines.
Alaskan glaciers to sea-level rise derived from satellite imagery.
Nature Geoscience 3, 92–95.
Bera, S.K. 2004. Late Holocene palaeo-winds and climatic changes
Bettis III, E.A., Muhs, D.R., Roberts, H.M., et al. 2003. Last Glacial
in Eastern Antarctica as indicated by long-distance transported
loess in the coterminous USA. Quaternary Science Reviews 22,
pollen-spores and local microbiota in polar lake core sediments.
1907–1946.
Current Science 86, 1485–1488.
Bereiter, B., Lüthi, D., Seagrist, M., et al. 2012. Mode change of Bhiry, N., Filion, L. 1996. Mid-Holocene hemlock decline in eastern
millennial CO2 variability during the last glacial cycle associ- North America linked with phytophagous insect activity.
ated with a bipolar marine carbon seesaw. Proceedings of the Quaternary Research 45, 312–320.
National Academy of Sciences of the United States of America 109, Bianchi, C., Gersonde, R. 2004. Climate evolution at the last
9755–9760. deglaciation: the role of the Southern Ocean. Earth and Planetary
Berger, A., Loutre, M.F. 2010. Modeling the 100-kyr glacial– Science Letters 228, 407–424.
interglacial cycles. Global and Planetary Change 72, 275–281. Bierman, P.R., Gillespie, R.A. 1994. Evidence suggesting that methods
Berger, A., Mélice, J.J., Loutre, M.F. 2005. On the origin of the 100- of rock-varnish cation-ratio dating are neither comparable nor
kyr cycles in the astronomical forcing. Paleoceanography 20, consistently reliable. Quaternary Research 41, 82–90.
PA4019, doi: 10.1029/2005PA001173 Bintanja, R., van de Wal, R.S.W. 2008. North American ice-sheet
Berger, W.H. 2013a. Milankovitch tuning of deep-sea records: dynamics and the onset of the 100,000-year glacial cycles. Nature
Implications for maximum rates of change of sea level. Global 454, 869–872.
and Planetary Change 101, 131–143. Bintanja, R., van de Wal, R.S.W., Oerlemans, J. 2005. Modelled
Berger, W.H. 2013b. On the Milankovitch sensitivity of the atmospheric temperatures and global sea levels over the past
Quaternary deep-sea record. Climate of the Past Discussions 9, million years. Nature 437, 125–128.
1237–1257. Bird, M.I., Fifield, L.K., Santos, G.M., et al. 2003. Radiocarbon
Berger, W.H., von Rad, U. 2002. Decadal to millennial cyclicity in dating from 40 to 60 ka BP at Border Cave, South Africa.
varves and turbidites from the Arabian Sea: hypothesis of tidal Quaternary Science Reviews 22, 943–947.
origin. Global and Planetary Change 34, 313–325. Bird, M.I., Taylor, D., Hunt, C. 2005. Palaeoenvironments of
Berger, W.H., Yasuda, M.K., Bickert, T., et al. 1994. Quaternary time insular Southeast Asia during the Last Glacial Period: a savanna
scale for the Ontong Java Plateau: Milankovitch template for corridor in Sundaland? Quaternary Science Reviews 24, 2228–
Ocean Drilling Program Site 806. Geology 22, 463–467. 2242.
Berglund, B.E., Gaillard, M.-J., Björkman, L., et al. 2008. Long-term Birkeland, P.W. 1999. Soils and Geomorphology. Oxford University
changes in floristic diversity in southern Sweden: palynological Press, Oxford.
452 REFERENCES
Birkeland, P.W., Shroba, R.R., Burns, S.F., et al. 2003. Integrating soils Blackwell, B.A.B. 2001a Electron spin resonance (ESR) dating in
and geomorphology in mountains – an example from the Front lacustrine environments. In Tracking Environmental Change
Range of Colorado. Geomorphology 55, 329–344. using Lake Sediments. Volume 1: Basin Analysis, Coring and
Birks, H.H. 2001. Plant macrofossils. In Tracking Environmental Chronological Techniques, 2nd edn (edited by W.M. Last, J.P.
Change Using Lake Sediments. Volume 3: Terrestrial, Algal and Smol). Kluwer, Dordrecht, 283–369.
Silicious Indicators (edited by J.P. Smol, H.J.B. Birks, E.F. Blackwell, B.A.B. 2001b. Amino acid racemisation (AAR) dating and
Stoermer). Kluwer Academic Publishers, Dordrecht, 49–74. analysis in lacustrine environments. In Tracking Environmental
Birks, H.H. 2007. Plant macrossil introduction. In Encyclopedia of Change Using Lake Sediments. Volume 1: Basin Analysis, Coring
Quaternary Science (edited by S.A. Elias). Elsevier, Amsterdam, and Chronological Techniques, 2nd edn (edited by W.M. Last,
2266–2287. J.P. Smol). Kluwer, Dordrecht, 391–450.
Birks, H.H. 2008. The Late-Quaternary history of arctic and alpine Blackwell, B.A.B., Skinner, A.R., Mashriqi, F., et al. 2012. Challenges
plants. Plant Ecology & Diversity 1, 135–146. in constraining pluvial events and human activity: examples of
Birks, H.H. and Wright, H.E. (eds). 2000. The reconstruction of the ESR dating molluscs from the Western Desert, Egypt. Quaternary
late-glacial and early-Holocene aquatic ecosystems at Kråkenes Geochronology 10, 430–435.
Lake, Norway. Journal of Paleolimnology 23, 1–114. Blanchet CL., Tjallingii R., Frank M., et al. 2013. High- and low-
Birks, H.H., Birks, H.J.B. 2006. Multi-proxy studies in palaeo- latitude forcing of the Nile River regime during the Holocene
limnology. Vegetation History and Archaeobotany 15, 235–251. inferred from laminated sediments of the Nile deep-sea fan. Earth
Birks, H.J.B. 1989. Holocene isochrone maps and patterns of and Planetary Science Letters 364, 980–110.
tree-spreading in the British Isles. Journal of Biogeography 16, Blockley, S.P.E., Lowe, J.J., Walker, M.J.C., et al. 2004. Bayesian
503–540. analysis of radiocarbon chronologies: examples from the
Birks, H.J.B., Birks, H.H. 1980. Quaternary Palaeoecology. Edward European Late-glacial. Journal of Quaternary Science 19, 159–175.
Arnold, London. Blockley, S.P.E., Blaauw, M., Bronk Ramsey, C., et al. 2007a. Building
Birks, H.J.B., Seppä, H. 2004. Pollen-based reconstructions of late- and testing age models for radiocarbon dates in Lateglacial and
Quaternary climate in Europe – progress, problems, and pitfalls. Early Holocene sediments. Quaternary Science Reviews 26,
Acta Palaeobotanica 44, 317–334. 1915–1926.
Bitinas, A. 2012. New insights into the last deglaciation of the south- Blockley, S.P.E., Bronk Ramsey, C., Pyle, D.M. 2007b. Improved age
eastern flank of the Scandinavian Ice Sheet. Quaternary Science modelling and high-precision age estimates of late Quaternary
Reviews 44, 69–80. tephras, for accurate palaeoclimate reconstruction. Journal of
Bitinas, A., Karmaziené, D., Jusiené, A. 2004. Glaciolacustrine kame Volcanology and Geothermal Research 177, 251–262
terraces as an indicator of deglaciation in Lithuania during the Blockley, S.P.E., Lane, C.S., Lotter, A.F., et al. 2007c. Evidence for
Last Deglaciation. Sedimentary Geology 165, 285–294. the presence of the Vedde Ash in central Europe. Quaternary
Bittman, F., Müller, H. 1996. The Kärlich Interglacial site and its Science Reviews 26, 3030–3036.
correlation with the Bilshausen sequence. In The Early–Middle Blockley, S.P.E., Lane, C.S., Hardiman, M., et al. 2012. Synchron-
Pleistocene in Europe (edited by C. Turner). Balkema, Rotterdam, isation of palaeoenvironmental records over the last 60,000
187–193. years, and an extended INTIMATE event stratigraphy. Journal
Björck, S. 1995. A review of the history of the Baltic Sea, 13.0–8.0 ka of Quaternary Science 36, 2–10.
BP. Quaternary International 27, 19–40. Blois, J.L., Hadly, E.A. 2009. Mammalian response to Cenozoic
Björck, S., Walker, M.J.C., Cwynar, L.C., et al. 1998. An event climatic change. Annual Review of Earth and Planetary Sciences
stratigraphy for the Last Termination in the North Atlantic 37, 181–208.
region based on the Greenland ice-core record. Journal of Blois, J.L., McGuire, J.L., Hadly, E.A. 2010. Small mammal diversity
Quaternary Science 13, 283–292. loss in response to late-Pleistocene climatic change. Nature 465,
Björck, S., Koç, N., Skog, G. 2003. Consistently large marine reser- 771–774.
voir ages in the Norwegian Sea during the last deglaciation. Blott, S.J. and Pye, K., 2008, Particle shape: a review and new methods
Quaternary Science Reviews 22, 429–435. of characterization and classification. Sedimentology 55, 31–63.
Blaauw, M., 2012. Out of tune: the dangers of aligning proxy archives. Blott, S.J., Croft, D.J., Pye, K., et al. 2004. Particle Size Analysis by Laser
Quaternary Science Reviews 36, 38–49. Diffraction. Geological Society of London, Special Publication 232.
Blaauw, M., Heuvelink, G.B.M., Mauquoy, D., et al. 2003. A Blackwell Scientific Publications, Oxford, 63–73.
numerical approach to C-14 wiggle-match dating of organic Blum, M.D., Törnqvist, T.E. 2000. Fluvial responses to climate and
deposits: best fits and confidence intervals. Quaternary Science sea-level change: a review and look forward. Sedimentology 47
Reviews 22, 1485–1500. (Suppl. 1), 2–48.
Blaauw, M., Van Geel, B., Mauquoy, D., et al. 2004. Radiocarbon Blunier, T.J., Chappellaz, J., Schwander, J., et al. 1998. Asynchrony
wiggle-match dating of peat deposits: advantages and limitations. of Antarctic and Greenland climate change during the last glacial
Journal of Quaternary Science 19, 177–181. period. Nature 394, 739–743.
Blackford, J., Chambers, F.M. 1995. Proxy climate record for the last Blunier, T., Spahni, R., Barnola, J.-M., et al. 2007. Synchronisation
1000 years from Irish blanket peat and a possible link to solar of ice core records via atmospheric gases. Climate of the Past 3,
variability. Earth and Planetary Science Letters 133, 145–150. 325–330.
REFERENCES 453
Blyth, A.J., Asrat, A., Baker, A., et al. 2007. A new approach to Bondevik, S., Mangerud, J. 2002. A calendar age estimate of a very
detecting vegetation and land-use change using high-resolution late Younger Dryas ice sheet maximum in western Norway.
lipid biomarker records in stalagmites. Quaternary Research 68, Quaternary Science Reviews 21, 1661–1676.
314–324. Bonny, A.P. 1980. Seasonal and annual variations over 5 years in
Blyth, A.J., Baker, A., Collins, M.J., et al. 2008. Molecular organic contemporary airborne pollen trapped at a Cumbrian lake.
matter in speleothems and its potential as an environmental Journal of Ecology 68, 421–441.
proxy. Quaternary Science Reviews 27, 905–921. Boomer, I., Wünnemann, B., Mackay, A.W., et al. 2009. Advances
Boardman, J., Poesen, J. (eds). 2006. Soil Erosion in Europe. Wiley, in understanding the late Holocene history of the Aral Sea region.
Chichester and New York. Quaternary International 194, 79–90.
Bockheim, J.G. 1995. Permafrost distribution in the Southern Boomer, I., Guichard, F., Lericolais, G. 2010. Late Pleistocene to
Circumpolar Region and its relation to the environment: a Recent ostracod assemblages from the western Black Sea. Journal
review and recommendation for further research. Permafrost and of Micropalaeontology 29, 119–133.
Periglacial Processes 6, 27–45. Booth, R.K. 2002. Testate amoebae as paleoindicators of surface-
Boehrer, B., Schultze, M. 2008. The stratification of lakes. Review of moisture changes on Michigan peatlands: modern ecology
Geophysics 46, doi: 10.1029/2006RG000210. and hyrdological calibration. Journal of Paleolimnology 28,
Boenigk,W., Frechen, M. 2006. The Pliocene and Quaternary fluvial 329–348.
archives of the Rhine system. Quaternary Science Reviews 25, Booth, R.K., Jackson, S.T., Gray, C.E.D. 2004. Paleoecology and
550–574. high-resolution paleohydrology of a kettle peatland in upper
Boës, X., Fagel, N. 2008. Relationship between southern Chilean Michigan. Quaternary Research 61, 1–13.
varved lake sediments, precipitation and ENSO for the last 600 Booth, R.K., Notaro, M., Jackson, S.T., et al. 2006. Widespread
years. Journal of Paleolimnology 39, 237–252. drought episodes in the western Great Lakes region during the
Boettger, T., Hiller, A., Kremenetski, K. 2003. Mid-Holocene past 2000 years: geographic extent and potential mechanisms.
warming in the northwest Kola Peninsula, Russia: northern Earth and Planetary Science Letters 242, 415–427.
pine-limit movement and stable isotope evidence. The Holocene Bos, J.A.A., Bohncke, S.J.P., Jannsen, C.R. 2006. Lake-level
13, 403–410. fluctuations and small-scale vegetation patterns during the
Bohacs, K.M., Carroll, A.R., Neal, J.E. 2003. Lessons from large lake Late Glacial in the Netherlands. Journal of Paleolimnology 35,
systems – thresholds, nonlinearity, and strange attractors. 211–238.
Geological Society of America, Special Paper 370, 75–90. Böse, M. 2005. The Last Glaciation and geomorphology. In The
Bohncke, S., Vandenberghe, J. and Huijzer, A.S. 1993. Periglacial Physical Geography of Western Europe (edited by E.A. Koster).
environments during the Weichselian Late Glacial in the Maas Oxford University Press, Oxford, 61–74.
valley, the Netherlands. Geologie en Mijnbouw (Netherlands Boslough, M., Nicoll, K., Holliday, V., et al. 2012. Arguments and
Journal of Geosciences) 72, 193–210. evidence against a Younger Dryas impact event. Geophysical
Bokhorst, M.P., Vandenberghe, J., Sümegi, P., et al. 2011. Atmos- Monographs Series 198, 13–26.
pheric circulation patterns in central and eastern Europe during Boswijk, G., Folwer, A., Lorrey, A., et al. 2006. Extension of the New
the Weichselian Pleniglacial inferred from loess grain-size Zealand kauri (Agathis australis) chronology to 1724 BC. The
records. Quaternary International 234, 62–74. Holocene 16, 1–12.
Boles, J.R., Clark, J.F., Leifer, I., et al. 2001. Temporal variation Botha, G., Porat, N. 2007. Soil chronosequence development in
in natural methane seep rate due to tides, Coal Oil Point dunes on the southeast African coastal plain, Maputaland, South
area, California. Journal of Geophysical Research 106, 27,077– Africa. Quaternary International 162/163, 111–132.
27,086. Bouchet, V.M.P., Alve, E., Rygg, B., et al. 2012. Benthic foraminifera
Böllhofer, A., Mangini, A., Lenhard, A., et al. 2004. High-resolution provide a promising tool for ecological quality assessment of
210
Pb dating of Lake Constance sediments: stable lead in Lake marine waters. Ecological Indicators 23, 66–75.
Constance. Environmental Geology 24, 267–274. Boulton, G.S. 1980. Classification of till. Quaternary Newsletter 31,
Bond, G., Broecker, W., Johnsen, S., et al. 1993. Correlations between 1–12.
climatic records from North Atlantic sediments and Greenland Boulton, G.S., Eyles, N. 1979. Sediimentation by valley glaciers: a
ice. Nature 365, 143–147. model and genetic classification. In Moraines and Varves (edited
Bond, G., Showers, W., Cheseby, M., et al. 1997. A pervasive by C. Schlüchter). Balkema, Rotterdam, 11–23.
millennial-scale cycle in North Atlantic Holocene and glacial Boulton, G., Hagdorn, M. 2006. Glaciology of the British Isles Ice Sheet
climates. Science 278, 1257–1266. during the last glacial cycle: form, flow, streams and lobes.
Bond, G.C., Mandeville, C., Hoffman, S. 2001a. Were rhyolitic Quaternary Science Reviews 25, 3359–3390.
glasses in the Vedde Ash and in the North Atlantic’s Ash Zone 1 Boulton, G.S., Dobbie, K.E., Zatsepin, S. 2001. Sediment deformation
produced by the same volcanic eruption? Quaternary Science beneath glaciers and its coupling to the subglacial hydraulic
Reviews 20, 1189–1199. system. Quaternary International 86, 3–28.
Bond, G., Kromer, B., Beer J., et al. 2001b. Persistent solar influence Bourdon, B., Henderson, G.M., Lundstrom, C.C., et al. (eds). 2003.
on North Atlantic climate during the Holocene. Science 294, Uranium Series Geochemistry. Reviews in Mineralogy and
2130–2136. Geochemistry 52, 1–656.
454 REFERENCES
Bourgeois, J.C., Koerner, R.M., Gajewski, K., et al. 2000. A Holocene Brand, U., McCarthy, F.M.G. 2005. The Allerød–Younger Dryas–
ice-core pollen record from Ellesmere Island, Nunavut, Canada. Holocene sequence in the west-central Champlain Sea, eastern
Quaternary Research 54, 275–283. Ontario: a record of glacial, oceanographic, and climatic changes.
Bowen, D.Q. 1981. The ‘South Wales End Moraine’: fifty years after. Quaternary Science Reviews 24, 1463–1478.
In The Quaternary in Britain (edited by J. Neale, J. Flenley). Brauer, A., Casanova, J. 2001. Chronology and depositional processes
Pergamon Press, Oxford, 60–67. of the laminated sediment record from Lac d’Annecy, French Alps.
Bowen, D.Q. 1991. Time and space in the glacial sediment systems Journal of Paleolimnology 25, 163–177.
of the British Isles. In Glacial Deposits in Great Britain and Brauer, A., Negendank, J.F.W. (eds). 2004. High resolution lake
Ireland (edited by J. Ehlers, P.L. Gibbard, J. Rose). Balkema, sediment records in climate and environment variability studies:
Rotterdam, 3–11. Euopean lake drilling program. Quaternary Proceedings 122,
Bowen, D.Q. 1999. On the correlation and classification of 1–133.
Quaternary deposits and land–sea correlations. In A Revised Brauer, A., Endres, C., Gunter, C., et al. 1999. High resolution and
Correlation of Quaternary Deposits in the British Isles (edited by sediment and vegetation responses to Younger Dryas climate
D.Q. Bowen). Geological Society of London, Special Report 23. change in varved lake sediments from Meerfelder Maar,
Geological Society, Bath, 1–9. Germany. Quaternary Science Reviews 18, 321–329.
Bowen, D.Q., Gibbard, P.L. 2007. The Quaternary is here to stay. Brauer, A., Allen, J.R.M., Mingram, J., et al. 2007. Evidence for last
Journal of Quaternary Science 22, 3–8. interglacial chronology and environmental change from
Bowen, D.Q., Pillans, B., Sykes, G.A., et al. 1998. Amino acid Southern Europe. Proceedings of the National Academy of Sciences
geochronology of Pleistocene marine sediments in the Wanganui of the United States of America 104, 450–455.
Basin: a New Zealand framework for correlation and dating. Brauer, A., Haug, G.H., Dulski, P., et al. 2008. An abrupt wind shift
Journal of the Geological Society of London 155, 439–446. in western Europe at the onset of the Younger Dryas cold period.
Bowen, D.Q., Phillips, F.M., McCabe, A.M., et al. 2002. New data for
Nature Geoscience 1, 520–523.
the Last Glacial Maximum in Great Britain and Ireland.
Braun, H., Christl, M., Rahmstorf, S., et al. 2005. Possible solar origin
of the 1470-year glacial climate cycle demonstrated in a coupled
Bowman, S. 1990. Radiocarbon Dating. British Museum, London.
model. Nature 438, 208–211.
Boyd, R. (ed.). 1999. Indians, Fire and the Land. Oregon State
Bray, P.J., Blockley, S.P.E., Coope, G.R., et al. 2006. Refining mutual
University Press, Corvallis, OR.
climatic range (MCR) quantitative estimates of palaeo-
Boyle, E.A. 2000. Is ocean thermohaline circulation linked to abrupt
temperature using ubiquity analysis. Quaternary Science Reviews
stadial/interstadial transitions? Quaternary Science Reviews 19,
25, 1865–1876.
255–272.
Brazier, V. Kirkbride, M., Owens, I. 1998. The relationship between
Brace, S., Palkopoulou, E., Dalén, L., et al. 2012. Serial population
extinctions in a small mammal indicate Late Pleistocene climate and rock glacier distribution in the Ben Ohau Range, New
ecosystem instability. Proceedings of the National Academy of Zealand. Geografiska Annaler 80, 193–207.
Sciences of the United States of America 109, 20532–20536. Breemer, C.W., Clark, P.U., Haggerty, R. 2002. Modeling the
Bradley, R.S. 1999. Palaeoclimatology. Reconstructing Climates of the subglacial hydrology of the late Pleistocene Lake Michigan Lobe,
Quaternary, 2nd edn. Academic Press, London. Laurentide Ice Sheet. Geological Society of America Bulletin 114,
Bradshaw, E.G., Rasmussen, P., Nielsen, H., et al. 2005. Mid- to 665–674.
late-Holocene land-use change and lake development at Dallund Breitenlechner, E., Hilber, M., Lutz, J., et al. 2010. The impact of
Sø, Denmark: trends in lake primary production as reflected by mining activities on the environment reflected by pollen, charcoal
algal and macrophyte remains. The Holocene 15, 1130–1142, and geochemical analyses. Journal of Archaeological Science 37,
Bradtmiller, L.I., Anderson, R.F., Fleisher, M.Q., et al. 2006. Diatom 1458–1467.
productivity in the equatorial Pacific Ocean from the last glacial Bridgland, D.R. (ed.). 1994. Quaternary of the Thames. Chapman and
period to the present: a test of the silicic acid leakage hypothesis. Hall, London.
Paleoceanography 21, PA4201, doi: 10.1029/2006PA001282. Bridgland, D.R. 2000. River terrace systems in northwest Europe: an
Bradwell, T., Stoker, M.S., Larter, R. 2007. Geomorphological archive of environmental change, uplift and early human
signature and flow dynamics of The Minch palaeo-ice stream, occupation. Quaternary Science Reviews 19, 1293–1303.
NW Scotland. Journal of Quaternary Science 22, 609–622. Bridgland, D.R. 2006. The Middle and Upper Pleistocene sequence
Bradwell, T., Stoker, M.S., Golledge, N.R., et al. 2008a. The northern in the Lower Thames: a record of Milankovitch climate fluc-
sector of the last British Ice Sheet: maximum extent and demise. tuation and early human occupation of southern Britain.
Earth-Science Reviews 88, 207–226. Proceedings of the Geologists’ Association 117, 281–305.
Bradwell, T., Fabel, D., Stoker, M., et al. 2008b. Ice caps existed Bridgland, D.R. 2010. The record from British Quaternary river
throughout the Lateglacial Interstadial in northern Scotland. systems within the context of global fluvial archives. Journal of
Journal of Quaternary Science, 23, 410–407. Quaternary Science 25, 433–446.
Bradwell, T., Stoker, M.S., Krabbendam, M. 2008c. Megagrooves and Bridgland, D.R., Westaway, R. 2008a. Climatically controlled river
streamlined bedrock in NW Scotland: the role of ice streams in terrace staircases: a worldwide Quaternary phenomenon.
landscape evolution. Geomorphology 97, 135–156. Geomorphology 98, 285–315.
REFERENCES 455
Bridgland, D.R., Westaway, R. 2008b. Preservation patterns of Broecker, W.S., Hemming, S. 2001. Climate swings come into focus.
Late Cenozoic fluvial deposits and their implications: results from Science 294, 2308–2309.
IGCP 449. Quaternary International 189, 5–38. Broecker,W.S., Bond, G., Klas, M.A. 1990. A salt oscillator in the glacial
Bridgland, D.R., Schreve, D.C. 2009. Implications of new Quaternary Atlantic? 1. The concept. Paleoceanography 5, 469–77.
uplift models for correlation between the Middle and Upper Broecker, W.S., Denton, G.H., Edwards, R.L., et al. 2010. Putting the
Thames terrace sequences, UK. Global and Planetary Change 68, Younger Dryas cold event into context. Quaternary Science
346–356. Reviews 29, 1078–1081.
Bridgland, D.R., Schreve, D.C., Keen, D.H., et al. 2004. Bio- Bromham, L., Penny, D. 2003. The modern molecular clock. Nature
stratigraphical correlation between the late Quaternary sequence Reviews Genetics 4, 216–224.
of the Thames and key fluvial localities in central Germany. Bronger, A. 2003. Corrrelation of loess–paleosol sequences in East
Proceedings of the Geologists’ Association 115, 125–140. and Central Asia with SE Central Europe: towards a conti-
Bridgland, D.R., Harding, P., Allen, P., et al. 2013. An enhanced record nental Quaternary pedostratigraphy and paleoclimatic history.
of MIS 9 environments, geochronology and geoarchaeology: Quaternary International 106/107, 11–32.
data from the construction of the High Speed 1 (London– Bronger, A., Winter, R., Sedov, S. 1998. Weathering and clay mineral
Channel Tunnel) rail-link and other recent investigations at formation of two Holocene soils and of buried paleosols in
Purfleet, Essex, UK. Proceedings of the Geologists’ Association Tadjikistan: towards a Quaternary climatic record in Central Asia.
124, 417–476. Catena 34, 19–36.
Briffa, K.R. 2000. Annual climatic variability in the Holocene: Bronk Ramsey, C. 2008. Depositional models for chronological
interpreting the message of ancient trees. Quaternary Science records. Quaternary Science Reviews 27, 42–60.
Reviews 19, 87–105. Bronk Ramsey, C., Staff, R.A., Bryant, C.L., et al. 2012. A complete
Briggs, D.E.G. 2003. The role of decay and mineralization in the terrestrial radiocarbon record for 11.2 to 52.8 kyr BP. Science 338,
preservation of soft-bodied fossils. Annual Review of Earth and 370–374.
Planetary Science 31, 275–301. Brook, E.J., White, J.W.C., Schilla, A.S.M., et al. 2005. Timing of
Briggs, D.J., Gilbertson, D.D., Harris, A.L. 1990. Molluscan millennial-scale climate change at Siple Dome, West Antarctica,
taphonomy in a braided river environment and its implications during the last glacial period . Quaternary Science Reviews 24,
for studies of Quaternary cold-stage river deposits. Journal of 1333–1343.
Biogeography 17, 623–637. Brook, M.S., Williams, J. 2012. A relict pronival (protalus) rampart
Bright, J., Kaufman, D.S., Forester, R.M., et al. 2006. A continuous in the Tararua Range, North Island, New Zealand. Permafrost and
250,000 yr record of oxygen and carbon isotopes in ostracode Periglacial Processes 24, 67–74.
and bulk-sediment carbonate from Bear Lake, Utah–Idaho. Brookes, L.A. 1993. Geomorphology and Quaternary geology of the
Quaternary Science Reviews 25, 2258–2270. Dakhla Oasis region, Egypt. Quaternary Science Reviews 9,
Briner, J.P. 2007. Supporting evidence from the New York drumlin 273–280.
field that elongate subglacial bedforms indicate fast ice flow. Boreas Brookes, L.A. 2003. Geomorphic indicators of Holocene winds in
36, 143–147. Egypt’s Western Desert. Geomorphology 56, 155–166.
Briner, J.P., Young, N.E., Goehring, B., et al. 2012. Constraining Brookhagen, B., Fleitmann, D., Nishiizumi, K., et al. 2006. Holocene
Holocene 10Be production rates in Greenland. Journal of monsoonal dynamics and fluvial terrace formation in the
Quaternary Science 27, 2–6. northwest Himalaya, India. Geology 34, 60–1604.
Brocklehurst, S.H. 2010. Tectonics and geomorphology. Progress in Brooks, A.J., Bradley, S.L., Edwards, R.J., et al. 2007a. Postglacial
Physical Geography 34, 357–383. relative sea-level observations from Ireland and their role in
Brodersen, K.P., Quinlan, R. 2006. Midges as palaeoindicators of glacial rebound modelling. Journal of Quaternary Science 23,
lake productivity, eutrophication and hypolimnetic oxygen. 175–192.
Quaternary Science Reviews 25, 1995–2012. Brooks, A.S. Hare, P.E., Kokis, J.E., et al. 1990. Dating Pleistocene
Brodzikowski, K., Van Loon, A.J. 1987. A systematic classification archaeological sites by protein diagenesis in ostrich eggshell.
of glacial and periglacial environments, facies and deposits. Science 248, 60–64.
Earth-Science Reviews 24, 297–381. Brooks, S.J. 2006. Fossil midges (Diptera: Chironomidae) as
Broecker, W.S. 1998. Paleocean circulation during the last palaeoclimatic indicators for the Eurasian region. Quaternary
deglaciation: a bipolar seesaw? Paleoceanography 13, 119–121. Science Reviews 25, 1894–1910.
Broecker, W.S. 2000. Was a change in thermohaline circulation Brooks, S.J., Birks, H.J.B. 2000. Chironomid-inferred Late-glacial air
responsible for the Little Ice Age? Proceedings of the National temperatures at Whitrig Bog, southeast Scotland. Journal of
Academy of Sciences of the United States of America 97, 1339–1342. Quaternary Science 15, 759–764.
Broecker W.S., Denton, G.H. 1989. The role of ocean-atmosphere Brooks, S.J., Langdon, P.G., Heiri, O. 2007b. The Identification and
reorganizations in glacial cycles. Geochimica et Cosmochimica Acta Use of Palearctic Chironomidae Larvae in Palaeoecology. Technical
53, 2465–2501. Guide 10. Quaternary Research Association, London.
Broecker, W.S., Denton, G.H. 1990. What drives glacial cycles? Brooks, S.J., Matthews, I.P., Birks, H.H., et al. 2012. High resolu-
Scientific American 262, 42–50. tion Lateglacial and early-Holocene summer air temperature
456 REFERENCES
records from Scotland inferred from chironomid assemblages. Burns, S.J., Fleitmann, D., Matter, A., et al. 2001. Speleothem evidence
Quaternary Science Reviews 41, 67–82. from Oman for continental pluvial events during interglacial
Brown, A.G., Carpenter, R.G., Walling, D.E. 2007a. Monitoring periods. Geology 29, 623–262.
fluvial pollen transport, its relationship to catchment vegetation Burrough, S.L., Thomas, D.S.G., Baillie, R.M. 2009. Mega-Lake in
and implications for palaeoenvironmental studies. Review of the Kalahari: A Late Pleistocene record of the Palaeolake
Palaeobotany and Palynology 147, 60–76. Makgadikgadi system. Quaternary Science Reviews 28, 1392–1411.
Brown, D.M., Baillie, M.G.L. 1992. Construction and dating of a Busschers, F.S., Kasse, C., van Balen, R.T., et al. 2007. Late Pleistocene
5000-year English bog oak tree-ring chronology. Lundqua Report evolution of the Rhine–Meuse system in the southern North
34, 72–75. Sea basin: imprints of climate change, sea-level oscillation and
Brown, E.T., Johnson, T.C., Scholz, C.A., et al. 2007b. Abrupt change glacio-isostacy. Quaternary Science Reviews 26, 3216–3248.
in tropical African climate linked to the bipolar seesaw over the Butler, P.G., Wanamaker, A.D., Scourse, J.D., et al. 2013. Variability
past 55,000 years. Geophysical Research Letters 34, doi: of marine climate on the North Icelandic Shelf in a 1357-year
proxy archive based on growth increments in the bivalve Arctica
10.1029/2007GL031240.
islandica. Palaeogeography, Palaeoclimatology, Palaeoecology 373,
Browne, G.H., Naish, T.R. 2003. Facies development and sequence
141–151.
architecture of a late Quaternary fluvial–marine transition,
Cacho, I., Grimalt, J.O., Pelerejo, C., et al. 1999. Dansgaard–Oeschger
Canterbury Plains and shelf, New Zealand: implications for
and Heinrich event imprints in Alboran Sea paleotemperatures.
forced regressive deposits. Sedimentary Geology 158, 57–86. Paleoceanography 14, 698–705.
Brugal, J.-P., Croitor, R. 2007. Evolution, ecology and biochron- Cacho, I., Grimalt, J.O., Sierro, F.J., et al. 2000. Evidence for enhanced
ology of herbivore associations in Europe during the last 3 Mediterranean thermohaline circulation during rapid climatic
million years. Quaternaire 18, 129–151. coolings. Earth and Planetary Science Letters 183, 417–420.
Buck, C.E., Blackwell, P.G. 2004. Formal statistical models for Callendar, G. 1938. The artificial production of carbon dioxide and
estimating radiocarbon calibration curves. Radiocarbon 46, its influence on temperature. Quarterly Journal of the Royal
1093–1102. Meteorological Society 64, 223–240.
Buckland, P.C., Edwards, K.J., Panagiotakopulu, E., et al. 2009. Callendar, G. 1961. Temperature fluctuations and trends over the
Palaeoecological and historical evidence for manuring and earth. Quarterly Journal of the Royal Meteorological Society 87,
irrigation at Garðar (Igaliku), Norse Eastern Settlement, 1–12.
Greenland. The Holocene 19, 105–116. Calvo, E., Pelejero, C., De Deckker, P., et al. 2007. Antarctic deglacial
Buckland, P.I., Buckland, P.C. 2006. Bugs coleopteran ecology pattern in a 30 kyr record of sea surface temperature offshore
package, v7.63, http://www.bugscep.com. South Australia. Geophysical Research Letters 34, doi: 10.1029/
Buiron, D., Chapellaz, J., Stenni, B., et al. 2011. TALDICE-1 age scale 2007GL029937.
of the Talos Dome deep ice core, East Antarctica. Climate of the Cameron, R. 2003. Land Snails in the British Isles. FSC Publications,
Past 7, 1–16. Preston Montford.
Bull, W.B. 2003. Lichenometry dating of coseismic changes to a Campbell, I.D. 1999. Quaternary pollen taphonomy: examples of
New Zealand landslide complex. Annals of Geophysics 46, differential redeposition and differential preservation. Palaeo-
1155–1167. geography, Palaeoclimatology, Palaeoecology 149, 245–256.
Bullard, J. 2010. Bridging the gap between field data and global models: Camoin, G.F., Ebren, Ph., Eisenhauer, A., et al. 2001. A 300,000-yr
current strategies in aeolian research. Earth Surface Processes and coral reef record of sea level changes, Mururoa atoll (Tuamotu
Landforms 35, 496–499. archipelago, French Polynesia). Palaeogeography, Palaeo-
climatology, Palaeoecology 175, 325–341.
Bunch, T.E., Hermes, R.E., Moore, A.M.T., et al. 2012. Very high-
Camoin, G.F., Montaggioni, L.F., Braithwaite, C.J.R. 2004. Late
temperature impact melt products as evidence for cosmic
glacial to post glacial sea levels in the Western Indian Ocean.
airbursts and impacts 12,900 years ago. Proceedings of the National
Marine Geology 206, 119–146.
Academy of Sciences of the United States of America 109, doi:
Camoin, G.F., Seard, C., Deschamps, P. et al. 2012. Reef response
10.1073/pnas.1204453109.
to sea-level and environmental changes during the last de-
Büntgen, U., Frank, D.C., Nievergelt, D., et al. 2006. Summer glaciation: Integrated Ocean Drilling Program Expedition 310,
temperature variations in the European Alps, AD 755–2004. Tahiti Sea Level. Geology 40, 643–646.
Journal of Climate 19, 5606–5623. Cande, S.C., Kent, D.V. 1995. Revised calibration of the geomagnetic
Burbank, D.W., Blythe, A.E., Putkonen, J., et al. 2003. Decoupling polarity timescale for the Late Cretaceous and Cenozoic. Journal
of erosion and precipitation in the Himalayas. Nature 426, of Geophysical Research 100, 6093–6095.
652–655. Candy, I., Black, S. 2009. The timing of Quaternary calcrete
Burckle, L. 1998. Marine diatoms. In Introduction to Marine development in semi-arid sotheast Spain: the role of climate on
Micropaleontology (edited by B.U. Haq, A. Boersma). Elsevier, calcrete genesis. Sedimentary Geology 218, 6–15.
Amsterdam, 245–266. Candy, I., Black, S., Sellwood, B.W. 2005. U-series isochron dating
Burney, D.A., Burney, L.P., Godfrey, L.R., et al. 2004. A chronology of immature and mature calcretes as a basis for constructing
for late prehistoric Madagascar. Journal of Human Evolution 47, Quaternary landform chronologies for the Sorbas basin,
25–63. southeast Spain. Quaternary Research 64, 100–111.
REFERENCES 457
Candy, I., Coope, G.R., Lee, J.R., et al. 2010. Pronounced warmth Carr, S.J., Holmes, R., van der Meer, J.J.M., Rose, J. 2006. The Last
during early Middle Pleistocene interglacials: investigating the Glacial Maximum in the North Sea Basin: micromorphological
Mid-Brunhes Event in the British terrestrial sequence. Earth- evidence of extensive glaciation. Journal of Quaternary Science 21,
Science Reviews 103, 183–196. 131–153.
Cane, M.A. 2004. The evolution of El Niño, past and future. Earth Carr, A.S., Bateman, M.D., Roberts, D.L., et al. 2010. The last
and Planetary Science Letters 164, 1–10. interglacial sea-level high stand on the southern Cape coastline
Cane, A., Molnar, P. 2001. Closing of the Indonesian seaway as a of South Africa. Quaternary Research 73, 351–363.
precursor to East African aridification around 3–4 million years Carré, M., Bentaleb, I., Fontugne, M., et al. 2005. Strong El Niño
ago. Nature 411, 157–162. events during the early Holocene: stable isotope evidence from
Cane, T., Rohling, E.J., Kemp. A.E.S., et al. 2002. High-resolution Peruvian sea shells. The Holocene 15, 42–47.
stratigraphic framework for Mediterranean saropel S5: defining Carrión, J.S., Mumuera, M., Navarro, C., et al. 1999. The palaeo-
temporal relationships between records of Eemian climate ecological potential of pollen records in caves: the case of
variability. Palaeogeography, Palaeoclimatology, Palaeoecology Mediterranean Spain. Quaternary Science Reviews 18, 1061–1073.
87–101. Carter, L., Neil, H.L., McCave, I.N. 2000. Glacial to interglacial
Cannariato, K.G., Kennett, J.P., Behl, R.J. 1999. Biotic response to changes in non-carbonate and carbonate accumulation in the
late Quaternary rapid climate switches in Santa Barbara Basin: SW Pacific Ocean, New Zealand. Palaeogeography, Palaeo-
ecological and evolutionary implications. Geology 27, 63–66. climatology, Palaeoecology 162, 333–356.
Canti, M.G. 1998. The micromorphological identification of faecal Carter, R.W.G. 1992. Sea-level changes: past, present and future.
spherulites from archaeological and modern materials. Journal In The Applications of Quaternary Research (edited by J.M.
of Archaeological Science 25, 435–444. Gray). Quaternary Proceedings 2, Quaternary Research Associ-
Cao, L., Fairbanks, R.G., Mortlock, R.A., et al. 2007. Radiocarbon ation, Cambridge, 111–130.
reservoir age of high latitude North Atlantic surface water during Carter, S.P. 1990. The stratification and taphonomy of shells
the last deglacial. Quaternary Science Reviews 26, 732–742. in calcareous soils: implications for land snails analysis in
CAPE-Last Interglacial Project Members. 2006. Last Interglacial archaeology. Journal of Archaeological Science 17, 495–507.
Arctic warmth confirms polar amplification of climate change. Casana, J. 2008. Mediterranean valleys revisited: linking soil erosion,
Quaternary Science Reviews 25, 1383–1400. land use and climate variability in the Northern Levant.
Capotondi, L., Borsetti, A.M. Morigi, C. 1999. Foraminiferal Geomorphology 101, 429–442.
ecozones, a high resolution proxy for the late Quaternary Catani, F., Farina, P., Moretti, S., et al. 2004. On the application of
biochronology of the central Mediterranean Sea. Marine Geology SAR interferometry to geomorphological studies: estimation
153, 253–274. of landform attributes and mass movements. Geomorphology 66,
Capron, E., Landais, A., Lemieux-Dudon, B., et al. 2010a. 119–131.
Synchronising EDML and NorthGRIP ice cores using δ18O of Cato, I. 1985. The definitive correction of the Swedish geo-
atmospheric oxygen (δ18Oatm) and CH4 measurements over chronological time scale with the present, and the new date of
MIS5 (80–123 kyr). Quaternary Science Reviews 29, 222–234. the zero year in Döviken, northern Sweden. Boreas 14, 117–122
Capron, E., Landais, A., Chappellaz, J., et al. 2010b. Millennial and Cattaneo, A., Couillard, Y., Wunsam, S., et al. 2004. Diatom
sub-millennial scale climatic variations recorded in polar ice cores taxonomic and morphological changes as indicators of metal
over the last glacial period. Climate of the Past 6, 345–365. pollution and recovery on Lac Dufault (Québec, Canada). Journal
Carboni, M.G., Bergamin, L., Di Bella, L. 2010. Palaeoenvironmental of Paleolimnology 32, 163–175.
reconstruction of late Quaternary foraminifera and molluscs from Catto, N.R. 1990. Clast fabric of diamictions associated with some
the ENEA borehole (Versilian plain, Tuscany, Italy). Quaternary roches moutonnées. Boreas 19, 289–296.
Research 74, 265–276. Catto, N.R. 1998. Comparative study of striationa and basal till clast
Carew, J.L., Mylroie, J.E. 1995. Quaternary tectonic stablity of the fabrics, Malpeque-Bedeque region Prince Edward Island,
Bahamian Archipelago: evidence from fossil coral reefs and Canada. Boreas 27, 259–274.
flank margin caves. Quaternary Science Reviews 14, 145–154. Cellot, B., Mouillot, F., Henry, C.P. 1998. Flood drift and propagule
Carlquist, S.J. 2001. Comparative Wood Anatomy: Systematic, bank of aquatic macrophytes in a riverine wetland. Journal of
Ecological and Evolutionary Aspects of Dicotyledon Wood. Springer, Vegetation Science 9, 631–640.
Berlin. Chambers, D.P. 2006. Observing seasonal steric sea level variations
Carr, S.J. 2004. The North Sea basin. In Developments in Quaternary with GRACE and satellite altimetry. Journal of Geophysical
Science. Volume 2: Quaternary Glaciations: Extent and Chronology Research 111, doi: 10.1029/2005JC002914.
(edited by J. Ehlers, P.L. Gibbard). Elsevier, Amsterdam, 261–270. Chambers, F.M. 2002. The environmental applications of pollen
Carr, S.J., Rose, J. 2003. Till fabric patterns and significance: particle analysis. In Quaternary Environmental Micropalaeontology
response to subglacial stress. Quaternary Science Reviews 22, (edited by S.K. Haslett). Arnold, Oxford and New York, 241–258.
1415–1426. Chambers, F.M., Blackford, J.J. 2001. Mid- and Late-Holocene
Carr, S.J., Coleman, C. 2007. An improved technique for the climatic change: a test of periodicity and solar forcing in proxy-
reconstruction of former glacier mass-balance and dynamics. climate data from blanket peat bogs. Journal of Quaternary
Geomorphology 92, 76–90. Science 16, 329–338.
458 REFERENCES
Chambers, F.M., Barber, K.E., Maddy, D., Brew, J., 1997. A 5500- Chen, M.-T., Huang, C.-C., Pflaumann, U., et al. 2005. Estimating
year proxy-climate and vegetation record from blanket mire at glacial western Pacific sea-surface temperature: methodological
Talla Moss, Borders, Scotland. The Holocene 7, 391–399. overview and data compilation of surface sediment planktic
Chambers, F.M., Mauquoy, D., Brain, S.A., et al. 2007. Globally foraminifer faunas. Quaternary Science Reviews 24, 1049–1062.
synchronous climate change 2800 years ago: proxy data from Cheng, H., Ewards, R.L., Broecker, W.S., et al. 2009. Ice age
peat in South America. Earth and Planetary Science Letters 253, terminations. Science 326, 248–252.
439–444. Chepalyga, A.L. 2007. The late-glacial Great Flood in the Ponto-
Chambers, F.M., Beilman, D.W., Yu, Z. 2011. Methods for deter- Caspian Basin. In The Black Sea Flood Question (edited by
mining peat humification and for quantifying peat bulk density, V. Yanko-Hombach, A.S. Gilbert, N. Panin, et al.). Springer,
organic matter and carbon content for palaeostudies of climate Dordrecht, 119–148.
and peatland carbon dynamics. Mires and Peat 7(art. 7), 1–10. Chernyshev, I.V., Lebedev, V.A., Arakelyants, M.M. 2006. K–Ar
Chambers, F.M., Booth, R.K., De Vleeschouwer, F., et al. 2012. dating of Quaternary volcanics: methodology and interpretation
Development and refinement of proxy-climate indicators from of results. Petrology 14, 1556–2085.
peats. Quaternary International 268, 21–33. Chiverrell, R.C., 2001. A proxy record of late Holocene climate
Chang, F., Li,T., Zhuang, L., Yan, J. 2008a. A Holocene paleo- change from May Moss, northeast England. Journal of Quaternary
temperature record based on radiolaria from the northern Science 16, 9–29.
Okinawa Trough (East China Sea). Quaternary International 183, Chiverrell, R.C., Thomas, G.S.P. 2010. Extent and timing of the Last
115–122. Glacial Maximum (LGM) in Britain and Ireland: a review.
Chang, P., Zhang, R., Hazeleger, W., et al. 2008b. Oceanic link Journal of Quaternary Science 25, 535–549.
between abrupt changes in the North Atlantic Ocean and the Chiverrell, R.C., Thranser, I.M., Thomas, G.S.P., et al. 2013. Bayesian
African monsoon. Nature Geoscience 1, 444–448. modelling the retreat of the Irish Sea ice stream. Journal of
Chappell, J. 2002. Sea level changes forced ice breakouts in the Last
Chlachula, J. 2003. The Siberian loess record and its significance for
Glacial cycle: new results from coral terraces. Quaternary Science
reconstruction of Pleistocene climate change in north-central
Reviews 21, 1229–1240.
Asia. Quaternary Science Reviews 22, 1879–1906.
Charman, D.J. 2001. Biostratigraphic and palaeoenvironmental
Christiansen, H.H. 2005. Thermal regime of ice-wedge cracking
applications of testate amoebae. Quaternary Science Reviews
in Adventdalen, Svalbard. Permafrost and Periglacial Processes 16,
20, 1753–1764.
87–98.
Charman, D. 2004. Peatlands and Environmental Change. John
Chu, G., Sun, Q., Rioual, P., et al. 2008. Dinocyst microlaminations
Wiley, Chichester and New York.
and freshwater ‘red tides’ recorded in Lake Xiaolongwan,
Charman, D.J., Hendon, D., Packman, S. 1999. Multiproxy surface
northeastern China. Journal of Paleolimnology 39, 319–333.
wetness records from replicate cores on an ombrotrophic mire: Chu, G., Sun, Q., Zhaoyan, G., et al. 2009. Dust records from varved
implications for Holocene palaeoclimate records. Journal of lacustrine sediments of two neighbouring lakes in northeastern
Quaternary Science 14, 451–463. China over the last 1400 years. Quaternary International 194,
Charman, D.J., Hendon, D., Woodland, W.A. 2000. The Identification 108–118.
of Testate Amoebae (Protozoa: Rhizopoda) in Peats. Technical Church, J.A., White, N.J., Aarup, T., et al. 2008. Understanding global
Guide 9. Quaternary Research Association, London. sea levels: past, present and future. Sustainability Science 3, 9–22.
Charman, D.J., Bludell, A., Chiverell, R.C., et al. 2006. Compilation Church, M., Ryder, J.M. 1972. Paraglacial sedimentation, a
of non-annually resolved Holocene proxy climate records: consideration of fluvial processes conditioned by glaciation.
stacked Holocene peatland palaeo-water table reconstructions Geological Society of America Bulletin 83, 3059–3072.
from northern Britain. Quaternary Science Reviews 25, 336–350. Chylek, P., Folland, C.K., Lesins, G., et al. 2010. Twentieth century
Charman, D., Barber K.E., Blaauw, M., et al. 2009. Climate drivers bipolar seesaw of the Arctic and Antarctic surface air tempera-
for peatland palaeoclimate records. Quaternary Science Reviews tures. Geophysical Research Letters 37, L08703, doi: 10.1029/
28, 1811–1819. 2010GL042793
Charman, D.J., Gehrels, W.R., Manning, C., et al. 2010. Recon- Ciliberto, E., Spoto, G. 2000. Modern Analytical Methods in Art and
struction of recent sea-level change using testate amoebae. Archaeology. John Wiley, Chichester and New York.
Quaternary Research 73, 208–219. Claessens, L., Veldkamp, A., ten Broeke, E.M., et al. 2009. A
Cheetham, P. 2005. Forensic geophysical survey. In Forensic Quaternary uplift record for the Auckland region, North Island,
Archaeology: Advances in Theory and Practice (edited by J. Hunter, New Zealand, based on marine and fluvial terraces. Global and
M. Cox). Routledge, Abingdon and New York. Planetary Change 68, 383–394.
Chen, F., Cheng, B., Zhao, Y., et al. 2006. Holocene environmental Clague, J. 2000. Recognizing order in chaotic sequences of
change inferred from a high-resolution pollen record, Lake Quaternary sediments in the Canadian Cordillera. Quaternary
Zhuyeze, arid China. The Holocene 16, 675–684. International 68–71, 29–38.
Chen, F., Yu, Z., Yang, M., et al. 2008. Holocene moisture evolution Clague, J.J., James, T.S. 2002. History and isostatic effects of the last
in arid central Asia and its out-of-phase relationship with Asian ice sheet in southern British Columbia. Quaternary Science
monsoon history. Quaternary Science Reviews 27, 351–364. Reviews 21, 71–87.
REFERENCES 459
Clague, J.J., Menounos, B., Osborn G., et al. 2009. Nomenclature and eggshells from southern Madagascar. Quaternary Science Reviews
resolution in Holocene glacial chronologies. Quaternary Science 25, 2343–2356.
Reviews 28, 2231–2238. Claussen, M. 2009. Late Quaternary vegetation–climate feedbacks.
Clark, C.D. 1997. Reconstructing the evolutionary dynamics of Climate of the Past Discussions 5, 635–670.
former ice sheets using multi-temporal evidence, remote sensing Claussen, M., Mysak, L.A., Weaver, A.J., et al. 2002. Earth system
and GIS. Quaternary Science Reviews 16, 1067–1092. models of intermediate complexity: closing the gap in the
Clark, C.D., Evans, D.J.A., Khatwa, A., et al. 2004. Map and GIS spectrum of climate system models. Climate Dynamics 18,
database of glacial landforms and features related to the last 579–585.
British Ice Sheet. Boreas 33, 359–375. Claussen, M., Ganopolski, A., Brovkin, V., et al. 2003. Simulated
Clark, C.D., Hughes, A.L.C., Greenwood, S.L., et al. 2009b. Size global-scale response of the climate system to Dansgaard/
and shape characteristics of drumlins, derived from a large Oeschger and Heinrich Events. Climate Dynamics 21, 361–370.
sample, and associated scaling laws. Quaternary Science Reviews Cleaveland, M.K. 2000. A 963-year reconstruction of summer (JJA)
28, 677–692. streamflow in the White River, Arkansas, USA, from tree-rings.
Clark, C.D., Hughes, A.L.C., Greenwood, S.L., et al. 2012. Pattern The Holocene 10, 33–41.
and timing of retreat of the last British–Irish Ice Sheet. Quaternary Clement, A.C., Peterson, L.C. 2008. Mechanisms of abrupt climate
Science Reviews 44, 112–146. change of the last glacial period. Review of Geophysics 46, doi:
Clark, I.D., Fontes, J.-C. 1990. Palaeoclimatic reconstruction in 10.1029/2006RG000204.
Northern Oman based on carbonates from hypersaline ground- Clemmensen, L.B., Murray, A. 2006. The termination of the last major
waters. Quaternary Research 33, 320–336. phase of aeolian sand movement, coastal dunefields, Denmark.
Clark, J., McCabe, A.M., Schnabel., C., et al. 2009a. Cosmogenic Earth Surface Processes and Landforms 31, 795–808.
10
Be chronology of the last deglaciation of western Ireland, and Cléroux, C.M., Debret, M., Cortijo, E., et al. 2012. High-resolution
implications for sensitivity of the Irish Ice Sheet to climate sea surface temperature reconstructions off Cape Hatteras
change. Geological Society of America Bulletin 121, 3–16. over the last 10 ka. Paleoceanography 27, doi: 10.1029/2011
Clark, P.U. 1994. Unstable behaviour of the Laurentide Ice Sheet over PA002184.
deforming sediment and its implications for climate change. CLIMAP Project Members. 1976. The surface of ice-age earth.
Quaternary Research 41, 19–25. Science 191, 1131–1137.
Clark, P.U., Pollard, D. 1998. Origin of the middle Pleistocene CLIMAP Project Members. 1981. Seasonal reconstructions of the
transition by ice sheet erosion of regolith. Paleoceanography 13, earth’s surface at the last glacial maximum. Geological Society
1–9. of America Map and Chart Series, MC-36.
Clark, P.U., Marshall, S.J., Clarke, G.K.C., et al. 2001. Freshwater Cline, R.M., Hays, J.D. 1976. Investigation of Late Quaternary
forcing of abrupt climate change during the last glaciation. Paleoceanography and Paleoclimatology. Geological Society of
Science 293, 283–287. America Memoirs, 145.
Clark, P.U., Pisias, N.G., Stocker, T.F., et al. 2002. The role of the Clymo, R.S. and Mackay, D. 1987. Upwash and downwash of pollen
thermohaline circulation in abrupt climate change. Nature and spores in the unsaturated surface layer of Sphagnum-
415, 863–869. dominated peat. New Phytologist 105, 175–183.
Clark, P.U., Brook, E.J., Raisbeck, G.M., et al. 2003. Cosmogenic 10Be Cockburn, H.A.P., Summerfield, M.A. 2004. Geomorphological
ages of the Saglek Moraines, Torngat Mountains, Labrador. applications of cosmogenic isotope analysis. Progress in Physical
Geology 31, 617–620. Geography 28, 1–42.
Clark, P.U., Archer, D., Pollard, D., et al. 2006. The middle Pleistocene Coe, M.T., Harrison, S.P. 2002. The water balance of northern Africa
transition: characteristics, mechanisms, and implications for during the mid-Holocene: an evaluation of the 6 ka BP PMIP
long-term changes in atmospheric pCO2. Quaternary Science simulations. Climate Dynamics 19, 155–166.
Reviews 25, 3150–3184. Cohen, A .C., Peterson, D.E., Maddocks, R.F. 2007. Ostracoda. In
Clarke, A.L., Weckström, K., Conley, D.J., et al. 2006a. Long-term The Light & Smith Manual: Intertidal Invertebrates from Central
trends in eutrophication and nutrients in the coastal zone. California to Oregon, 4th edn (edited by J.T. Carlton). University
Limnology and Oceanography 51, 385–397. of California Press, Berkeley and Los Angeles, 417–446.
Clarke, J. 2003. The occurrence and significance of biogenic opal in Cohen, A.S. 2003. Paleolimnology: The History and Evolution of Lake
the regolith. Earth-Science Reviews 60, 175–194. Systems. Oxford University Press, Oxford and New York.
Clarke, M.J., Vogel, J.C., Botha, G.A., et al. 2003. Late Quaternary Cohen, K.M., Gibbard, P.L. 2011. Global Chronostratigraphical
hillslope evolution recorded in eastern South African colluvial Correlation Table for the Last 2.7 million years. V. 2011. Sub-
badlands. Palaeogeography, Palaeoecology, Palaeoclimatology commission on Quaternary Stratigraphy, Cambridge, http://
197, 199–212. www.qpg.geog.cam.uk/.
Clarke, S.J., Murray-Wallace, C.V. 2006. Mathematical expressions Cohen, T.J., Nanson, G.C., Jansen, J.D., et al. 2011. Continental
used in amino acid racemisation geochronology – a review. aridification and the vanishing of Australia’s megalakes. Geology
Quaternary Geochronology 1, 261–278. 39, 167–170.
Clarke, S.J., Miller, G.F., Fogel, M.L., et al. 2006b. The amino acid COHMAP Members. 1988. Climatic changes of the last 18,000 years:
and stable isotope biogeochemistry of elephant bird (Aepyornis) observations and model simulations. Science 241, 1043–1052.
460 REFERENCES
Colgan, P.M. 1999. Early Middle Pleistocene glacial sediments Coope, G.R. 2004. Several million years of stability among insect
(780000–620000 BP) near Kansas City, northeastern Kansas and species because of, or in spite of, ice age climatic instability?
northwestern Missouri, USA. Boreas 28, 477–489. Philosophical Transactions of the Royal Society of London B 359,
Colhoun, E.A., Kiernan, K.W., McConnell, A., et al. 2009. Late 209–214.
Pliocene age of glacial deposits at Heidemann Valley, East Coope, G.R. 2010. Coleopteran faunas as indicators of interglacial
Antarctica: evidence for the last major glaciation in the Vestfold climates in central and southern England. Quaternary Science
Hills. Antarctic Science 2009, doi: 10.1017/S0954. Reviews 29, 1507–1514.
Colinvaux, P., de Oliveira, P.E., Moreno Patino, J.E. 1999. Amazon Coope, G.R. Brophy, J.A. 1972. Late Glacial environmental changes
Pollen Manual and Atlas. Harwood Academic Publishers, indicated by a coleopteran succession from North Wales. Boreas
Amsterdam. 1, 97–142.
Collins, M.J., Gernaey, A.M., Nielsen-March, C.M., et al. 2000. Slow Coope, G.R., Lemdahl, G. 1995. Regional differences in the Lateglacial
rates of degradation of osteocalcin: green light for fossil bone climate of northern Europe based on coleopteran analysis.
protein? Geology 28, 1139–1142. Journal of Quaternary Science 10, 391–395.
Collins, M.J., Penkman, K.E.H., Rohland, N., et al. 2009. Is amino Coope, G.R., Lemdahl, G., Lowe, J.J., et al. 1998. Temperature
acid racemization a useful tool for screening for ancient DNA? gradients in Northern Europe during the Last Glacial–Interglacial
Proceedings of the Royal Society B 276, 2971–2977 Transition (14–9 14C ka BP) interpreted from coleopteran
Colman, S.M., Kaufman, D.S., Bright, J., et al. 2006. Age model for assemblages. Journal of Quaternary Science 13, 419–433.
a continuous, ca 250–ka Quaternary lacustrine record from Corrège, T. 2006. Sea surface temperature and salinity reconstruction
Bear Lake, Utah–Idaho. Quaternary Science Reviews 25, 2271– from coral geochemical tracers. Palaeogeography, Palaeo-
2282. climatology, Palaeoecology 232, 408–428.
Colmenero-Hidalgo, E., Flores, J.-A., Sierro, F.J., et al. 2004. Ocean Cortijo, E., Labeyrie, L., Elliot, M., et al. 2000. Rapid climatic
surface water response to short-term climate changes revealed
variability of the North Atlantic Ocean and global climate: a focus
by coccolithophores from the Gulf of Cadiz (NE Atlantic) and
of the IMAGES program. Quaternary Science Reviews 19,
Alboran Sea (W Mediterranean). Palaeogeography, Palaeo-
227–241.
climatology, Palaeoecology 205, 317–336.
Covey, C., AchutaRao, K.M., Cubasch, U., et al. 2003. An overview
Colonese A.C., Mannino , M.A., Bar-Yosef, D.E., et al. 2011. Marine
of results from the Coupled Model Intercomparison Project.
mollusc exploitation in Mediterranean prehistory: an overview.
Global and Planetary Change 37, 103–133.
Quaternary International 239, 86–103.
Craddock, R.A., Hutchinson, M.F., Stein, J.A. 2010. Topographic data
Conway, H., Hall, B.L., Denton, G.H., et al. 1999. Past and future
reveal a buried fluvial landscape in the Simpson Desert, Australia.
grounding-line retreat of the West Antarctic Ice Sheet. Science
Australian Journal of Earth Sciences 57, doi: 10.180/0812009
286, 280–283.
Cook, E.R., Buckley, B.M., Palmer, J.G., et al. 2006. Millennia-long 0903416278
tree-ring records from Tasmania and New Zealand: a basis for Craw, D., Burridge, C., Anderson, L., et al. 2007. Late Quaternary
modelling climate variability and forcing, past, present and river drainage and fish evolution, Southland, New Zealand.
future. Journal of Quaternary Science 21, 689–699. Geomorphology 84, 98–110.
Coope, G.R. 1959. A late Pleistocene insect fauna from Chelford. Craw, D., Burridge, C., Norris, R., et al. 2008. Genetic ages for
Proceedings of the Royal Society B 151, 70–86. Quaternary topographic evolution: a new dating tool. Geology
Coope, G.R. 1967. The value of Quaternary insect faunas in the 36, 19–22.
interpretation of ancient ecology and climate. In Quaternary Croitor, R., Brigal, J.-P. 2010. Ecological and evolutionary dynamics
Palaeoecology (edited by E.J. Cushing, H.E. Wright Jr.). Yale of the carnivore community in Europe during the last 3 million
University Press, New Haven, 359–380. years. Quaternary International 212, 98–108.
Coope, G.R. 1977. Fossil coleopteran assemblages as sensitive Cronin, T. 2009. Paleoclimates: Understanding Climate Change Past
indicators of climatic changes during the Devensian (last) cold and Present. Columbia University Press, New York.
stage. Philosophical Transactions of the Royal Society of London B Crowley T.J., Lowery, T.S. 2000. How warm was the Medieval Warm
280, 313–340. Period? AMBIO: A Journal of the Human Environment 29, 51–54.
Coope, G.R. 1987. The response of Late Quaternary insect Crutzen, P.J. 2002. The geology of mankind. Nature 415, 23.
communities to sudden climatic changes. In Organization of Crutzen, P.J., Stoermer, E.F. 2000. The Anthropocene. Global Change
Communities, Past and Present (edited by J.H.R. Gee, P.S. Miller). Newsletter 41, 17–18.
Blackwell, Oxford and Boston, 421–438. Cuenca-Bescós, G., Rofes, J., Lópes-García, J.M., et al. 2010.
Coope, G.R. 1990. The invasion of Northern Europe during the Biochronology of Spanish Quaternary small vertebrate faunas.
Pleistocene by Mediterranean species of Coleoptera. In Biological Quaternary International 212, 109–119.
Invasions in Europe and the Mediterranean Basin (edited by F. di Cugny, C., Mazier, F., Galop, D. 2010. Modern and fossil non-
Castri, A. Hansen, M. Debissche). Kluwer, Dordrecht, 203–215. pollen palynomorphs from the Basque mountains (western
Coope, G.R. 2000. The climatic significance of coleopteran assem- Pyrenees, France): the use of coprophilous fungi to reconstruct
blages from the Eemian deposits in southern England. Geologie pastoral activity. Vegetation History and Archaeobotany 19,
en Mijnbouw (Netherlands Journal of Geosciences) 79, 257–268. 391–408.
REFERENCES 461
Cullingford, R.A., Smith, D.E. 1966. Late-glacial shorelines in eastern Dahms, D., Favilli, F., Krebs, R., et al. 2012. Soil weathering and
Fife. Transactions of the Institute of British Geographers 39, 31–51. accumulation rates of oxalate-extractable phases derived from
Cullingford, R.A., Caseldine, C.J., Gotts, P.E. 1980. Early Flandrian alpine chronosequences of up to 1 Ma in age. Geomorphology
land and sea-level changes in Lower Strathearn. Nature 284, 151–152, 99–113.
159–161. Dale, A.L., Dale, B. 2002a. Application of ecologically-based statistical
Cummins, R.H. 1994. Taphonomic processes in modern freshwater treatments for micropalaeontology. In Quaternary Environmental
molluscan death assemblages: implications for the freshwater Micropalaeontology (edited by S. Haslett). Arnold, London,
fossil record. Palaeogeography, Palaeoclimatology, Palaeoecology 259–286.
108, 55–73. Dale, B., Dale, A.L. 2002b. Environmental applications of dino-
Currant, A.P., Jacobi, R. 2011. The mammal faunas of the British flagellate cysts and Acritarchs. In Quaternary Environmental
Late Pleistocene. In The Ancient Human Occupation of Britain Micropalaeontology (edited by S. Haslett). Arnold, London,
(edited by N. Ashton, S.G. Lewis, C. Stringer). Elsevier, 207–240.
Amsterdam, 165–180. Daley, T.J., Street-Perrott, F.A., Loader, N.J., et al. 2009. Terrestrial
Curry, A.M., Cleasby, V., Zokowskyj, P. 2006. Paraglacial response climate signal of the ‘8200 yr B.P.cold event’ in the Labrador Sea
of steep, sediment-mantled slopes to post-‘Little Ice Age’ glacier region. Geology 37, 831–834.
recession in the central Swiss Alps. Journal of Quaternary Science Daley, T.J., Barber, K.E. Street-Perrott, F.A., et al. 2010. Holocene
21, 211–225. climate variability revealed by oxygen isotope analysis of
Curry, A.M., Porter, P.R., Irvine-Fynn, T.D.L., et al. 2009. Sphagnum cellulose from Walton Moss, northern England.
Quantitative particle size, microtextural and outline shape Quaternary Science Reviews 29, 1590–1601.
analyses of glacigenic sediment reworked by paraglacial debris Dalsgaard, K., Odgaard, B.V. 2001. Dating sequences of buried
flows. Earth Surface Processes and Landforms 34, 48–62. horizons of podzols developed in wind-blown sand at Ulfborg,
Curry, B.B., Baker, R.G. 2000. Palaeohydrology, vegetation, and western Jutland. Quaternary International 78, 53–60.
climate since the late Illinois Episode (~130 ka) in south-central Damon, P.E., Donahue, D.J., Gore, B.H., et al. 1989. Radiocarbon
Illinois. Palaeogeography, Palaeoclimatology, Palaeoecology 155, dating of the Shroud of Turin. Nature 337, 611–615.
59–81. D’Anjou, R.M., Bradley, A.S., Balascio, N.L., et al. 2012. Climate
Curry, B.B., Grimley, D.A., McKay III, E.D. 2011. Quaternary impacts on human settlement and agricultural activities in
glaciations in Illinois. In Developments in Quaternary Science. northern Norway revealed through sediment biogeochemistry.
Volume 15: Quaternary Glaciations – Extent and Chronology: Proceedings of the National Academy of Sciences of the United States
A Closer Look (edited by J. Ehlers, P.L. Gibbard, P.D. Hughes). of America 109, 20332–20337.
Elsevier, Amsterdam, 467–488. Dansgaard, W., Johnsen, S.J., Moller, J., et al. 1969. One thousand
Cushing, E.J. 1967. Late Wisconsin pollen stratigraphy and the centuries of climatic record from Camp Century on the
glacial sequence in Minnesota. In Quaternary Palaeoecology Greenland ice sheet. Science 166, 499–502.
(edited by E.J. Cushing, H.E. Wright Jr.). Yale University Press, Dansgaard, W., Clausen, H. B., Gundestrup, N., et al. 1982. A New
New Haven, 59–88. Greenland Deep Ice Core. Science 218, 1273–1277.
Cutler, P.M., Mickelson, D.M., Colgan, P.M., et al. 2000. A numerical D’Argenio, A., Pescatore, T., Senatore, M.R. 2004. Sea-level change
investigation of ice-lobe–permafrost interaction around the and volcano-tectonic interplay. The Gulf of Pozzuoli (Campi
southern Laurentide ice sheet. Journal of Glaciology 46, 311–325. Flegrei, Eastern Tyrrhenian Sea) during the last 39 ka. Journal
Cutler, P.M., MacAyeal, D.R., Mickelson, D.M., et al. 2001. Influence of Volcanology and Geothermal Research 133, 105–121.
of the Great Lakes on the dynamics of the southern Lauren- Davidson, D.A., Simpson, I.A. 2001. Archaeology and soil micro-
tide ice sheet: numerical experiments. Geology 29, 1039–1042. morphology. In Handbook of Archaeological Sciences (edited by
Cuttelod, A., Seddon, M., Neubert, E. 2011. European Red List of D.R. Brothwell, A.M. Pollard). John Wiley, Chichester and New
Non-marine Molluscs. Publications Office of the European York, 166–177.
Union, Luxembourg. Davidson, D.A., Grieve, I.C., Tyler, A.N., et al. 1998. Archaeological
Dahl, S.O., Nesje, A. Lie , Ø., et al. 2002. Timing, equilibrium-line sites: assessment of erosion risk. Journal of Archaeological Science
altitudes and climatic implications of two early-Holocene glacier 25, 857–860.
advances during the Erdalen Event at Jostedalsbreen, western Davidson, D.A., Carter, S., Boag, B., et al. 1999. Analysis of
Norway. The Holocene 12, 17–25. pollen in soils: processes of incorporation and redistribution
Dahl-Jensen, D., Mosegaard, K., Gundestrup, N., et al. 1998. Past of pollen in five soil profile types. Soil Biology and Biochemistry
temperatures directly from the Greenland ice sheet. Science 282, 31, 643–653.
268–271. Davidson, I. 2013. Peopling the last new worlds: the first colonisation
Dahl-Jensen, D., Gundestrup, N., Gogineni, S.P., et al. 2003. Basal of Sahul and the Americas. Quaternary International 285, 1–29.
melt at NorthGRIP modeled from borehole, ice-core and Davidson, T.A., Sayer, C.D., Bennion, H., et al. 2005. A 250 year
radio-echo sounder observations. Annals of Glaciology 37, comparison of historical, macrofossil and pollen records of
207–212. aquatic plants in a shallow lake. Freshwater Biology 50, 1671–1686.
Dahms, D.E. 2002. Glacial stratigraphy of Stough Creek Basin, Wind Davies, P. 2008. Snails: Archaeology and Landscape Change. Oxbow
River Range, Wyoming. Geomorphology 42, 59–83. Books, Oxford.
462 REFERENCES
Davies, P., Griffiths, H.I. 2005. Molluscan and ostracod De Francesco, C.G., Hassan, G.S. 2008. Dominance of reworked
biostratigraphy of Holocene tufa in the Test valley at Bossington, fossil shells in modern estuarine environments: implications
Hampshire, UK. The Holocene 15, 97–110. for paleoenvironmental reconstructions based on biological
Davies, S.J., Metcalfe, S.E., Caballero, M.E., et al. 2002a. Developing remains. Palaios 23, 14–23.
diatom-based transfer functions for Central Mexican lakes. De Geer, G. 1912. A geochronology of the last 12,000 years. XIth
Hydrobiologia 467, 199–213. International Geological Congress, Stockholm, 1, 241–253.
Davies, S.M., Branch, N.P., Lowe, J.J., et al. 2002b. Towards a Degenhardt Jr., J.J. 2009. Development of tongue-shaped and
European tephrochronological framework for Termination 1 and multilobate rock glaciers in alpine environments – interpreta-
the Early Holocene. Philosophical Transactions of the Royal Society tions from ground penetrating radar surveys. Geomorphology 109,
of London A 360, 767–802. 94–107.
Davies, S.M., Wastegård, S., Rasmussen, T.L., et al. 2008. Dehnert, A., Schlüchter C. 2008. Sediment burial dating using
Identification of the Fugloyarbanki tephra in the NGRIP ice core: terrestrial cosmogenic nuclides. Eiszeitalter und Gegenwart
a key tie point for marine and ice-core sequences during the last (Quaternary Science Journal) 57, 210–225.
glacial period. Journal of Quaternary Science 23, 409–414. Dehnert, A., Kracht, O., Preusser, F., et al. 2011. Cosmogenic isotope
Davies, S.M., Larsen, G., Wastegård, S., et al. 2010. Widespread burial dating of fluvial sediments from the Lower Rhine Embay-
dispersal of Icelandic tephra: how does the Eyjafjöll eruption of ment, Germany. Quaternary Geochronology 6, 313–325.
2010 compare to past Icelandic events? Journal of Quaternary De Jong, R., Kamenik, C. 2011. Validation of a chrysophyte
Science 25, 605–611. stomatocyst-based cold-season climate reconstruction from
Davis, S.R., Brown, A.G., Dinnin, M.H. 2007. Floodplain high-Alpine Lake Silvaplana, Switzerland. Journal of Quaternary
connectivity, disturbance and change: a palaeoentomological Science 26, 268–275.
investigation of floodplain ecology from south-west England. Delmonte, B., Andersson P.S., Schöberg, H., et al. 2010. Geographic
Journal of Animal Ecology 76, 276–288. provenance of aeolian dust in East Antarcica during Pleisto-
Dawson, S., Smith, D.E. 2000. The sedimentology of Middle cene glaciations: preliminary results from Talos Dome and
Holocene tsunami facies in northern Sutherland, Scotland, comparison with East Antarctic and new Andean ice core data.
UK. Marine Geology 170, 69–79. Quaternary Science Reviews 29, 256–264.
Dayan, T., Wool, D., Simberloff, D. 2002. Variation and co-variation Demarchi, B., Williams, M.G., Milner, N., et al. 2011. Amino acid
of skulls and teeth: modern carnivores and the interpretation of racemization dating of marine shells: a mound of possibilities.
fossil mammals. Paleobiology 28, 508–526. Quaternary International 239, 114–124.
D’Costa, D.M., Palmer, J., Hogg A., et al. 2009. Stratigraphy, pollen Demetsopolous, J.C., Burleigh, R., Oakley, K.P. 1983. Relative and
and 14C dating of Johnston’s Gum Hole, a late Quaternary fossil absolute dating of the human skeleton from Galley Hill, Kent.
kauri (Agathus australis) site, Northland, New Zealand. Journal Journal of Archaeological Science 10, 129–134.
of Quaternary Science 24, 47–59. De Mulder, E., Geluk, M.C., Ritsma, I., et al. 2003. De Ondergrond
De Abreu, L., Shackleton, N.J., Schonfeld, J., et al. 2003. Millennial- van Nederland. Geologie van Nederland, deel 7. Nederlandse
scale oceanic climate variability off the western Iberian margin Organisatie voor toegepast-natuurwetenschappelijk onderzoek
during the last two glacial periods. Marine Geology 196, 1–20. – TNO, Utrecht.
Dean, W.E 1997. Rates, timing and cyclicity of Holocene eolian activity Deng, W., Wei, G., Li, X., et al. 2009. Paleoprecipitation record from
in north-central United States: evidence from varved lake coral Sr/Ca and d18O during the mid Holocene in the northern
sediments. Geology 25, 331–334 South China Sea. The Holocene 19, 811–821.
Dearing, J.A., Livingstone, I.P., Bateman, M.D., et al. 2001. Denniston, R.F., Wyrwoll, K.-H., Asmerom, Y., et al. 2013. North
Palaeoclimatic records from OIS 8.0–5.4 recorded in loess– Atlantic forcing of millennial-scale Indo-Australian monsoon
palaeosol sequences on the Matmata Plateau, southern Tunisia, dynamics diuring the Last Glacial period. Quaternary Science
based on mineral magnetism and new luminescence dating. Reviews 72, 159–168.
Quaternary International 76/77, 43–56. Denton, G. 2000. Does an asymmetric thermohaline–ice-sheet
DeBusk Jr., G.H. 1997. The distribution of pollen in the surface oscillator drive 100,000-yr glacial cycles? Journal of Quaternary
sediments of Lake Malawi, Africa, and the transport of pollen in Science 15, 301–318.
large lakes. Review of Palaeobotany and Palynology 97, 123–153. Denton, G.H., Hughes, T.J. 2002. Reconstructing the Antarctic Ice
De Deckker, P. 2001. Late Quaternary cyclic aridity in tropical Sheet at the Last Glacial Maximum. Quaternary Science Reviews
Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 170, 21, 193–202.
1–9. Denton, G.H., Alley, R.B., Comer, G.C., et al. 2005. The role of
De Deckker, P., Moros, M., Perner, K., et al. 2012. Influence of the seasonality in abrupt climate change. Quaternary Science Reviews
tropics and southern westerlies on glacial inter-hemispheric 24, 1159–1182.
asymmetry. Nature Geoscience 5, 266–269. Denton, G.H., Anderson, R.F., Toggweiler, J.R., et al. 2010. The last
Dee, M., Wengrow, D., Shortland, A., et al. 2013. An absolute glacial termination. Science 328, 1652–1656.
chronology for early Egypt using radiocarbon dating and Derbyshire, E. (ed.). 2003. Loess and dust indicators and records
Bayesian statistical modelling. Proceedings of the Royal Society A of terrestrial and marine palaeoenvironments (DIRTMAP)
469(2159), 20130395. database. Quaternary Science Reviews 22, 1813–2052.
REFERENCES 463
Derbyshire, E., Keen, D., Kemp, R.A., et al. 1995. Loess–palaeosol Ding, Z.L., Derbyshire, E., Yang, S.L., et al. 2002b. Stacked 2.6–Ma
sequences as recorders of palaeoclimatic variations during the grain size record from the Chinese loess based on five sections
last glacial–interglacial cycle: some problems of correlation in and correlated with the deep-sea δ18O record. Paleoceanography
North-Central China. Quaternary Proceedings 4, 7–18. 17, 1–21.
Derbyshire, E., Meng, X.M., Dijkstra, T.A. (eds). 2000. Landslides in Ding, Z.L., Derbyshire, E., Yang, S.L., et al. 2005. Stepwise expansion
Thick Loess Terrain of Northwest China. John Wiley, Chichester of desert environment across northern China in the past 3.5 Ma
and New York. and implications for monsoon evolution. Earth and Planetary
Deschamps, P., Durand, N., Bard, E., et al. 2012. Ice-sheet collapse Science Letters 237, 45–55.
and sea-level rise at the Bølling warming 14,600 years ago. Nature Ditlevsen, P.D. 2009. Bifurcation structure and noise-assisted
483, 559–563. transitions in the Pleistocene glacial cycles. Paleoceanography
Desmarchelier, J.M., Goede, A., Aycliffe, L.K., et al. 2000. Stable 24, PA3204, doi: 10.1029/2008PA001673.
isotope record and its palaeoenvironmental interpretation for a Ditlevsen, P.D., Andersen, K.K., Svensson, A. 2007. The DO-climate
late Middle Pleistocene speleothem from Victoria Fossil Cave, events are probably noise induced: statistical investigation of the
Naracoorte, South Australia. Quaternary Science Reviews 19, claimed 1470 years cycle. Climate of the Past 3, 129–134.
763–774. Dixon, E.J. 2001. Human colonization of the Americas: timing,
Desprat, S., Sánchez-Goñi, M.F., Turon, J.-L., et al. 2006. Climatic technology and process. Quaternary Science Reviews 20, 277–
variability of Marine Isotope Stage 7: direct land–sea–ice 299.
correlation from a multiproxy analysis of a north-western Iberian Dixon, J.C., McLaren, S.J. 2009. Duricrusts. In Geomorphology of
marine deep-sea core. Quaternary Science Reviews 25, 1010–1026. Desert Environments, 2nd edn (edited by A.J. Parsons, A.
De Vernal, A., Henry, M., Bilodeau, G. 2010. Micropaleontological Abraham). Springer, Dordrecht, 123–150.
preparation techniques and analyses. Les Cahiers du GEOTOP Dogan, U. 2010. Fluvial response to climate change during and after
3, http://www.geotop.ca/.
the Last Glacial Maximum in Central Anatolia, Turkey.
DeVogel, S.B., Magee, J.W., Manley, W.F., et al. 2004. A GIS-based
Quaternary International 222, 221–229.
reconstruction of late Quaternary paleohydrology: Lake Eyre,
Dokken, T.M., Nisancioglu, K.H. 2004. Palaeoclimatology: fresh angle
arid central Australia. Palaeogeography, Palaeoclimatology,
on the polar seesaw. Nature 430, 842–843.
Palaeoecology 204, 1–13.
Domack, E.W., Jacobson, E.A., Ship, S., et al. 1999. Late Pleistocene–
Dickinson, W.R. 2000. Hydro-isostatic and tectonic influences on
Holocene retreat of the West Antarctic Ice-Sheet system in the
emergent Holocene paleoshorelines in the Mariana Islands,
Ross Sea: Part 2 – Sedimentologic and stratigraphic signature.
western Pacific Ocean. Journal of Coastal Research 16, 735–746.
Geological Society of America Bulletin 111, 1517–1536.
Dickinson, W.R. 2004. Impacts of eustasy and hydro-isostasy on the
Donahue, D.J., Olin, J.S., Harbottle, G. 2002. Determination of
evolution and landforms of Pacific atolls. Palaeogeography,
Palaeoclimatology, Palaeoecology 213, 251–269. the radiocarbon age of the parchment of the Vinland Map.
Dieffenbacher-Krall, A.C. 2007. Plant macrofossil methods and Radiocarbon 44, 45–52.
studies. Surface samples, taphonomy, representation. In Doney, S.C., Fabry, V.J., Feely, R.A., et al. 2009. Ocean acidification:
Encyclopedia of Quaternary Science (edited by S.A. Elias). Elsevier, the other CO2 problem. Annual Reviews of Marine Science 1,
Amsterdam, 2367–2374. 169–192.
Dilcher, D.L., Kowalski, E.A., Wiemann, M.C., et al. 2009. A climatic Dorale, J.A., Edwards, L., Alexander Jr., E.C., et al. 2007. Uranium
and taxonomic comparison between leaf litter and standing series dating of speleothems: current techniques, limits and
vegetation from a Florida swamp woodland. American Journal applications. In Studies of Cave Sediments, 2nd edn (edited by
of Botany 96, 1108–1115. I.D. Sasowsky, J. Mylroie). Springer, Dordrecht, 177–197.
Dillehay, T.D., Rossen, J., Andres, T.C., et al. 2007. Preceramic Dorighel, O., Poupeau, G., Bellot-Gurlet, L., et al. 1998. Fission track
adoption of peanut, squash, and cotton in Northern Peru. Science dating and provenience of archaeological obsidian artefacts in
316, 1890–1893. Colombia and Ecuador. In Advances in Fission-Track Geo-
Dillon, R.T. 2000. The Ecology of Freshwater Molluscs. Cambridge chronology (edited by P. van den Haute, F. de Corte). Kluwer,
University Press, Cambridge. Dordrecht, 313–324.
Ding, Z., Yu, Z., Rutter, N.W., et al. 1994. Towards an orbital time Dortch, J.M., Owen, L.A., Cafee, M.W. (2013) Timing and climate
scale for Chinese loess deposits. Quaternary Science Reviews 13, drivers for glaciation across semi-arid western Himalayan-
39–70. Tibetan orogen. Quaternary Science Reviews 78, 188–208.
Ding, Z.L., Xiong, S.F., Sun, J.M., et al. 1999. Pedostratigraphy and Dos Santos, R.A.L., Prange, M., Castañeda, I.S., et al. 2010.
palaeomagnetism of a similar to 7.0 Ma eolian loess–red clay Glacial–interglacial variability in Atlantic meridional overturning
sequence at Lingtai, Loess Plateau, north central China and the circulation and thermocline adjustments in the tropical North
implications for paleomonsoon evolution. Palaeogeography, Atlantic. Earth and Planetary Science Letters 300, 407–414.
Palaeoclimatology, Palaeoecology 152, 49–66. Douglas, M.E., Douglas, M.R., Schuett, G.W., et al. 2006. Evolution
Ding, Z.L., Ranov, V., Yang, S.L., et al. 2002a. The loess record in of rattlesnakes (Viperidae: Crotalus) in the warm deserts of
southern Tajikistan and correlation with Chinese loess. Earth and western north America shaped by Neogene vicariance and
Planetary Science Letters 200, 387–4. Quaternary climate change. Molecular Ecology 15, 3353–3374.
464 REFERENCES
Douglass, D.C., Mickelson, D.M. 2007. Soil development and glacial Milankovitch climate oscillations. Proceedings of the National
history, West Fork of Beaver Creek, Uinta Mountains, Utah. Arctic Academy of Sciences of the United States of America 97, 9115–
and Alpine Research 39, 592–602. 9120.
Dowdeswell, J.A., Elverhøi, A., Spielhagen, R., 1998. Glacimarine Ebbesen, H., Hald, M. 2004. Unstable Younger Dryas climate in the
sedimentary processes and facies on the polar North Atlantic northeast Atlantic. Geology 32, 673–676.
margins. Quaternary Science Reviews 17, 243–272. Ebert, K., Willenbring, J., Norton, K.P., et al. 2012. Meteoric 10Be
Dowdeswell, J.A., Elverhøi, A., Andrews, J.T., et al. 1999. concentrations from saprolite and till in northern Sweden:
Asynchronous deposition of ice-rafted layers in the Nordic seas implications for glacial erosion and age. Quaternary Geo-
and North Atlantic Ocean. Nature 400, 348–351. chronology 12, 11–22.
Dowdeswell, J.A., Ó Cofaigh, C., Pusey, C.J. 2004. Thickness and Eberwein, A., Mackensen, A. 2008. Last Glacial Maximum paleo-
extent of the subglacial till layer beneath an Antarctic paleo-ice productivity and water masses off NW-Africa: evidence from
stream. Geology 32, 13–16. benthic foraminifera and stable isotopes. Marine Micro-
Dowdeswell, J.A., Hogan, K.A., Evans, J., et al. 2010. Past ice-sheet paleontology 67, 87–103.
flow east of Svalbard inferred from streamlined subglacial Edwards, K.J., Rowntree K.M. 1980. Radiocarbon and palaeo-
landforms. Geology 38, 163–166. environmental evidence for changing rates of erosion at a
Drake, N., Bristow, C. 2006. Shorelines in the Sahara: geo- Flandrian stage site in Scotland. In Timescales in Geomorphology
morphological evidence for an enhanced monsoon from (edited by R.A. Cullingford, D.A. Davidson, J. Lewin). John Wiley,
palaeolake Megachad. The Holocene 16, 901–911. Chichester and New York, 207–223.
Drake, N.A., Blench, R.M., Armitage, S.J., et al. 2010. Ancient Edwards, K.J., Whittington, G. 2001. Lake sediments, erosion and
watercourses and biogeography of the Sahara explain the landscape change during the Holocene in Britain. Catena 42,
peopling of the desert. Proceedings of the National Academy of 143–173.
Sciences of the United States of America 108, 458–462. Edwards, M.E., Brubaker, L.E., Lozhkin, A.V., et al. 2005. Structurally
Drysdale, R.N., Spötl, C., Hellstrom, J.C., et al. (eds). 2012. New novel biomes: a response to past warming in Beringia. Ecology
advances in the dating of speleothem. Quaternary Geochronology 86, 1696–1703.
14, 1–114. Edwards, R. 2007. Sea levels: resolution and uncertainty. Progress in
Duller, G.A.T. 2004. Luminescence dating of Quaternary sediments: Physical Geography 31, 621–632.
recent advances. Journal of Quaternary Science 19, 183–192. Edwards, R., Horton, B.P. 2006. Developing detailed records of
Duller, G.A.T., Wintle, A.G. 2012. A review of the thermally relative sea-level change using a foraminiferal transfer func-
transferred optically stimulated luminescence signal from tion: an example from North Norfolk, UK. Philosophical
quartz for dating sediments. Quaternary Geochronology 7, Transactions of the Royal Society of London A 364, 973–991.
6–20. Edwards, R.L., Gallup, C.D., Cheng, H. 2003. Uranium-series dating
Dumitrica, P. 2013. Cleveiplegma nov. gen., a new generic name for of marine and lacustrine carbonates. In Uranium Series Geo-
the radiolarian species Rhizoplegma boreale (Cleve, 1899). Revue chemistry (edited by B. Bourdon, G.M. Henderson, C.C.
de Micropaléontologie 56, 21–25. Lundstrom, et al.). Reviews in Mineralogy and Geochemistry 52,
Dunai, T. 2010. Cosmogenic Nuclides: Principles, Concepts, and 363–405.
Applications in the Earth Sciences. Cambridge University Press, Eerkens, J.W., Vaughn, K.J., Carpenter, T.R., et al. 2008. Obsidian
Cambridge. hydration dating on the South Coast of Peru. Journal of
Dutta, K., Bushan, T., Somayajulu, B.L.K. 2001. Delta R correction Archaeological Science 35, 2231–2239.
values for the northern Indian Ocean. Radiocarbon 43, 483–488. Eggermont, H., Martens, K. 2011. Preface: Cladocera crustaceans:
Dwyer, G.S., Cronin, T.M., Baker, P.A., et al. 1993. North Atlantic sentinels of environmental change. Hydrobiologia 676, 1–7.
deepwater temperature change during Late Pliocene and Late Eggermont, H., Heiri, O. 2012. The chironomid–temperature
Quaternary climatic cycles. Science 270, 1347–1351. relationship: expression in nature and palaeoenvironmental
Dyke, A.S. 2004. An outline of North American deglaciation with implications. Biological Reviews 87, 430–456.
emphasis on central and northern Canada. In Developments in Eggins, S.M., Grün, R., McCullough, M.T., et al. 2005. In situ U-series
Quaternary Science. Volume 2: Quaternary Glaciations: Extent and dating by laser-ablation multi-collector ICPMS: new prospects
Chronology (edited by J. Ehlers, P.L. Gibbard). Elsevier, for Quaternary geochronology. Quaternary Science Reviews 24,
Amsterdam, 373–424. 2523–2538.
Dyke, A.S., Dale, J.E., McNeely, R.N. 1996. Marine mollusks as Eglinton, G., Calvin, M. 1967. Chemical fossils. Scientific American
indicators of environmental change in glaciated North America 216, 32–43.
and Greenland during the last 18 000 years. Géographie physique Ehlers, J., Gibbard, P.L. 2003. Extent and chronology of glaciations.
et Quaternaire 50, 125–184. Quaternary Science Reviews 15–17, 1561–1568.
Dyke, A.S., Andrews, J.T., Clark, P.U., et al. 2002. The Laurentide Ehlers, J., Gibbard, P.L. 2007. The extent and chronology of Cenozoic
and Innuitian ice sheets during the Last Glacial Maximum. Global Glaciation. Quaternary International 164–165, 6–20.
Quaternary Science Reviews 21, 9–31. Ehlers, J., Gibbard, P.L., Hughes, P.D. (eds). 2011a. Developments in
Dynesius, M., Jansson, R. 2000. Evolutionary consequences Quaternary Science. Volume 15: Quaternary Glaciations – Extent
of changes in species’ geographical distributions driven by and Chronology: A Closer Look. Elsevier, Amsterdam.
REFERENCES 465
Ehlers, J., Grube, A., Stephan, H.-G., et al. 2011b. Pleistocene Elliott, S.M., Roe, H.M., Patterson, R.T. 2012. Testate amoebae
Glaciations of North Germany – new results. In Developments as indicators of hydroseral change: an 8500 year record from
in Quaternary Science. Volume 15: Quaternary Glaciations – Mer Bleue Bog, eastern Ontario, Canada. Quaternary Inter-
Extent and Chronology: A Closer Look (edited by J. Ehlers, P.L. national 268, 128–144.
Gibbard, P.D. Hughes). Elsevier, Amsterdam, 149–162. Ellis, B.S., Mark, D.,F., Pritchard, C.J., et al. 2012. Temporal dissection
Ehrmann, W., Hillenbrand, C.-D., Smith, J.A., et al. 2011. Provenance of the Huckleberry Ridge Tuff using the 40Ar/39Ar dating tech-
changes between recent and glacial-time sediments in the nique. Quaternary Geochronology 9, 34–41.
Amundsen Sea embayment, West Antarctica: clay mineral Ellis, C.J., Tallis, J.H., 2000. Climatic control of blanket mire
assemblage evidence. Antarctic Science 23, 471–486. development at Kentra Moss, north-west Scotland. Journal of
Eichler, A., Tinner, W., Brütsch, S., et al. 2011. An ice-core based Ecology 88, 869–889.
history of Siberian forest fires since AD 1250. Quaternary Science Ellis, G.L., Goodfriend, G.A., Abbott, J.T., et al. 1996. Assessment
Reviews 30, 1027–1034. of integrity and geochronology of archaeological sites using
Eide, W., Birks, H.H., Bigelow, N.H., et al. 2006. Holocene forest amino acid racemization in land snail shells: examples from
development along the Setesdal valley, southern Norway, Central Texas. Geoarchaeology 11, 189–213.
reconstructed from macrofossil and pollen evidence. Vegetation Ellwood, B.B., Harrold, F.B., Benoist, S.L., et al. 2004. Magnetic
History and Archaeobotany 15, 65–85. susceptibility applied as an age-depth-climate relative dating
Eiríksson, J., Larsen, G., Knudsen, K.-L., et al. 2004. Marine reservoir technique using sediments from Scladina Cave, a Late Pleistocene
age variability and water mass distribution in the Icelandic Sea. site in Belgium. Journal of Archaeological Science 31, 283–293.
Quaternary Science Reviews 23, 2247–2268. Elmore, A.C., Wright, J.D. 2011. North Atlantic deep water and
Eitel, B., Blümel, W.D., Blümel, K., et al. 2001. Dust and loessic alluvial climate variability during the Younger Dryas cold period. Geology
deposits in Northwestern Namibia (Damaraland, Kaokoveld): 39, 107–110.
sedimentology and palaeoclimatic evidence based on lumin-
Elsig, J., Schmitt, J., Leuenberger, D., et al. 2009. Stable isotope
escence dates. Quaternary International 76/77, 57–66.
constraints on Holocene carbon cycle changes from an Antarctic
Ekdahl, E.J., Teranes, J.L., Guilderson, T.P., et al. 2004. Prehistorical
ice core. Nature 461, 507–510.
record of cultural eutrophication from Crawford Lake, Canada.
Emeis, K.C., Sakamoto, T., Wehausen, R., et al. 2000. The sapropel
Geology 32, 745–748.
record of the eastern Mediterranean Sea – results of Ocean
Elderfield, H. (ed.). 2006. The Oceans and Marine Geochemistry.
Drilling Program Leg 160. Palaeogeography, Palaeoclimatology,
Elsevier, Oxford and Amsterdam.
Elderfield, H., Rickaby, R.E.M. 2000. Oceanic Cd/P ratio and nutrient
Enfield, D.B., Cid-Serrano, L. 2010. Secular and multi-decadal
utilization in the glacial Southern Ocean. Nature 405, 305–310.
warmings in the North Atlantic and their relationships with major
Elderfield, H., Ferretti, P., Greaves, M., et al. 2012. Evolution of ocean
temperature and ice volume through the Mid-Pleistocene hurricane acivity. International Journal of Climatology 30,
Climate Transition. Science 337, 704–709. 174–184.
Elenga, H., de Namur, C., Vincens, A., et al. 2000. Use of plots to define Engelhardt, H. 2004. Ice temperature and high geothermal flux at
pollen–vegetation relationships in densely forested ecosystems Siple Dome, West Antarctica, from borehole measurements.
of Tropical Africa. Review of Palaeobotany and Palynology 112, Journal of Glaciology 50, 251–256.
79–96. Engelhart, S.E., Horton, B.P., Kemp. A.C. 2011. Holocene sea level;
Elias, S.A. 1994. Quaternary Insects and their Environments. changes along the United States’ Atlantic coast. Oceanography 24,
Smithsonian Institution Press, Washington, DC. 70–79.
Elias, S.A. 1997. The Mutual Climatic Range method of palaeoclimate Engels, S., van Geel, B. 2012. The effects of changing solar activity
reconstruction based on insect fossils: new applications and on climate: contributions from paleoclimatological studies.
inter-hemispheric applications. Quaternary Science Reviews 16, Journal of Space Weather and Space Climate 2, doi: org/10.1051/
1217–1225. swsc/2012009.
Elias, S.A. 2000. Late Pleistocene climates of Beringia, based on England, J. 1999. Coalescent Greenland and Innuitian ice during
analysis of fossil beetles. Quaternary Science Reviews 53, 229–235. the Last Glacial Maximum: revising the Quaternary of the
Elias, S.A. 2007a. Beetle records: overview. In Encyclopedia of Canadian High Arctic. Quaternary Science Reviews 18, 421–456.
Quaternary Science (edited by S.A. Elias). Elsevier, Amsterdam, England, J., Smith, I.R., Evans, D.J.A. 2000. The last glaciation of east-
153–163. central Ellesmere Island, Nunavut: ice dynamics, deglacial
Elias, S.A. 2007b. Postglacial North America. In Encyclopedia of chronology, and sea level change. Canadian Journal of Earth
Quaternary Science (edited by S.A. Elias). Elsevier, Amsterdam, Sciences 37, 1355–1371.
275–286. England, J., Atkinson, N., Bednarski, J., et al. 2006. The Innuitian Ice
Elias, S.A. 2010. Developments in Quaternary Science. Volume 12: Sheet: configuration, dynamics and chronology. Quaternary
Advances in Quaternary Entomology. Elsevier, Amsterdam. Science Reviews 25, 689–703.
Elkibbi, M., Rial, J.A. 2001. An outsider’s review of the astronomical English, P., Spooner, N.A., Chappell, J., et al. 2001. Lake Lewis basin,
theory of the climate: is the eccentricity-driven insolation the central Australia: environmental evolution and OSL chronology.
main driver of the ice ages? Earth-Science Reviews 56, 161–177. Quaternary International 83–85, 81–102.
466 REFERENCES
EPICA Community Members. 2004. Eight glacial cycles from an terminating ice streams: case study of the western Canadian
Antarctic ice core. Nature 429, 823–828. prairies and high plains. Quaternary Science Reviews 46, 80–125.
EPICA Community Members. 2006. One-to-one coupling of glacial Evans, M.E., Heller, F. 2003. Environmental Magnetism: Principles and
climate variability in Greenland and Antarctica. Nature 444, Applications of Enviromagnetics. Academic Press, NewYork.
195–198. Evans, M., Slaymaker, O. 2004. Spatial and temporal variability of
Erdtman, G. 1969. Handbook of Palynology. Munksgaard, Copen- sediment delivery from alpine lake basins, Cathedral Provincial
hagen. Park, southern British Columbia. Geomorphology 61, 209–224.
Ericson, D.B., Wollin, G. 1968. Pleistocene climates and chronology Evans, M., Warburton, J. 2007. Geomorphology of Upland Peat:
in deep-sea sediments. Science 162, 1227–1234. Erosion, Form and Landscape Change. Blackwell, Malden, MA
Ericson, J.E., Dersch, O., Rauch, F. 2004. Quartz hydration dating. and Oxford.
Journal of Archaeological Science 31, 883–902. Everest, J., Bradwell, T., Fogwell, C.J., et al. 2006. Cosmogenic 10Be
Eronen, M., Zetterberg, P., Briffa, K.R., et al. 2002. The supra- age constraints for the Wester Ross Readvance Moraine: insights
long Scots pine tree-ring record for Finnish Lapland: Part 1, into British ice-sheet behaviour. Geografiska Annaler. A. Physical
chronology construction and initial inferences. The Holocene 12, Geography 88, 9–17.
673–680. Everest, T., Bradwell, T., Stoker, M., et al. 2013. New age constraints
Erthal, F., Kotzian, C.B., Simões, M.G. 2011. Fidelity of molluscan for the maximum extent of the last British–Irish Ice Sheet (NW
assemblages from the Touro Passo Formation (Pleistocene– sector). Journal of Quaternary Science 28, 2–7.
Holocene), southern Brazil: taphonomy as a tool for discovering Eyles, N., Lazorek, M. 2007. Glacial landforms, sediments. Glacigenic
natural baselines for freshwater communities. Palaios 26, Lithofacies. In Encyclopedia of Quaternary Science (edited by S.A.
433–446. Elias). Elsevier, Amsterdam, 920–932.
Esat, T.M., McCulloch, M.T., Chappell, J., et al. 1999. Rapid Eyles, N., Boyce, J.I., Barendregt, R.W. 1999. Hummocky moraine:
fluctuations in sea level recorded at Huon Peninsula during the sedimentary record of stagnant Laurentide ice sheet lobes resting
penultimate deglaciation. Science 283, 197–201.
on soft beds. Sedimentary Geology 123, 163–174.
Esper, O., Zonneveld, K.A.F. 2007. The potential of organic-walled
Eyles, N., Doughty, M., Boyce, J.I., et al. 2003. Acoustic architecture
dinoflagellate cysts to reconstruct past sea-surface conditions in
of glaciolacustrine sediments deformed during zonal stagnation
the Southern Ocean. Marine Micropaleontology 63, 185–212.
of the Laurentide Ice Sheet; Mazinaw Lake, Ontario, Canada.
Espizua, L.E., Bigazzi, G., Junes, P.J., et al. 2002. Fission-track dating
Sedimentary Geology 157, 133–151.
of a tephra layer related to Poti-Malal and Seguro drifts in the
Fabel, D., Ballantyne, C.K., Xu, S. 2012. Trimlines, blockfields,
Río Grande basin, Mendoza, Argentina. Journal of Quaternary
mountain-top erratics and the vertical dimensions of the last
Science 17, 781–788.
British–Irish ice sheet in NW Scotland. Quaternary Science
Evans, D.J.A. 2003. Ice-marginal terrestrial landsystems: active
Review 55, 91–102.
temperate glacier margins. In Glacial Landsystems (edited by
D.J.A. Evans). Arnold, London, 259–288. Faegri, K. Iversen, J. 1989. Textbook of Pollen Analysis, 4th edn, with
Evans, D.J.A. 2009. Controlled moraines: origins, characteristics K. Krzywinski. John Wiley, Chichester and New York.
and palaeoglaciological implications. Quaternary Science Reviews Fagan, B. 2000. The Little Ice Age. Basic Books, New York.
28, 181–260. Fagel, N., Boës, X. 2008. Clay-mineral record in Lake Baikal
Evans, D.J.A., Twigg, D.R. 2002. The active temperate glacial sediments: the Holocene and Late Glacial transition. Palaeo-
landsystem: a model based on Breiðamerkurjökull and Fjalls- geography, Palaeoclimatology, Palaeoecology 259, 230–243.
jökull, Iceland. Quaternary Science Reviews 21, 2143–2177. Fairbanks, R.G., Mortlock, R.A., Chiu, T.-C., et al. 2005. Radiocarbon
Evans, D.J.A., Benn, D.I. 2004. A Practical Guide to the Study of Glacial calibration curve spanning 0 to 50,000 years BP based on paired
238
Sediments. Arnold, London. Th/234U/238U and 14C dates on pristine corals. Quaternary
Evans, D.J.A., Owen, L.A., Roberts, D. 1995. Stratigraphy and Science Reviews 24, 1781–1796.
sedimentology of Devensian (Dimlington Stadial) deposits, east Fairchild, I.J. 2009. Trace elements in speleothems as recorders of
Yorkshire, England. Journal of Quaternary Science 10, 241–265. environmental change. Quaternary Science Reviews 28, 449–468.
Evans, D.J.A., Archer, S., Wilson, D.J.H. 1999. A comparison of the Fairchild, I.J., McMillan, A. 2007. Speleothems as indicators of wet
lichenometric and Schmidt hammer dating techniques based on and dry periods. International Journal of Speleology 36, 69–74.
data from the proglacial areas of some Icelandic glaciers. Fairchild, I.J., Baker, A. 2012. Speleothem Science. From Process to Past
Quaternary Science Reviews 18, 13–41. Environments. Wiley Blackwell, Chichester.
Evans, D.J.A., Clark, C.D., Mitchell, W.A. 2005. The last British Ice Faith, J.T., Surovell, T.A. 2009. Synchronous extinction of North
Sheet: a review of the evidence utilized in the compilation of the America’s Pleistocene mammals. Proceedings of the National
Glacial Map of Britain. Earth-Science Reviews 70, 253–312. Academy of Sciences of the United States of America 106,
Evans, D.J.A., Phillips, E.R., Hiemstra, J.F., et al. 2006. Subglacial till: 20641–20645.
formation, sedimentary characteristics and classification. Earth- Faith, J.T., Domínguez-Rodrigo, M., Gordon, A.D. 2009. Long-
Science Reviews 78, 115–176. distance carcass transport at Olduvai Gorge? A quantitative
Evans, D.J.A., Hiemstra, J.F., Boston, C.M., et al. 2012. Till examination of Bed I skeletal element abundances. Journal of
stratigraphy and sedimentology at the margins of terrestrially Human Evolution 56, 247–256.
REFERENCES 467
Farmer, G.L., Licht, K., Swope, R.J., et al. 2006. Isotopic constraints Firth, C.R., Smith, D.E., Cullingford, R.A. 1993. Late Devensian
on the provenance of fine-grained sediment in LGM tills from and Holocene glacio-isostatic uplift patterns in Scotland. In
the Ross Embayment, Antarctica. Earth Surface Processes and Neotectonics: Recent Advances (edited by L.A. Owen, I. Stewart,
Landforms 249, 90–107. C. Vita-Finzi). Quaternary Proceedings 3, Quaternary Research
Farmer, J.G., Anderson P., Cloy, J.M., et al. 2009. Historical Association, Cambridge, 1–13.
accumulation rates of mercury in four Scottish ombrotrophic peat Fischer, H., Traufetter, F., Oerter, H., et al. 2004. Prevalence of the
bogs over the past 2000 years. Science of the Total Environment Antarctic Circumpolar Wave over the last two millenia recorded
407, 5578–5588. in Dronning Maud Land ice. Geophysical Research Letters 31, doi:
Favilli, F., Egli, M., Brandová, D., et al. 2009. Combined use of 10.1029/2003GL019186.
relative and absolute dating techniques for detecting signals of Fischer, H., Siggaard-Andersen, M.-L., Ruth, U., et al. 2007.
Alpine landscape evolution during the late Pleistocene and early Glacial/interglacial changes in mineral dust and sea-salt records
Holocene. Geomorphology 112, 48–66. in polar ice cores: sources, transport, and deposition. Reviews of
Fedele, F.G., Giaccio, B., Isaia, R., et al. 2002. Ecosystem impact of Geophysics 45, doi: 10.1029/2005RG000192.
the Campanian Ignimbrite eruption in Late Pleistocene Europe. Fischer, H., Masson-Delmotte, V., Waelbroeck, C., et al. (eds).
Quaternary Research 57, 420–424. 2010a. Climate of the last million years: new insights from
Fei, W., Hongchun, L., Rixiang, Z., et al. 2004. Late Quaternary EPICA and other records. Quaternary Science Reviews 29, 1–365.
downcutting rates of the Qianyou River from U/Th speleothem Fischer, H., Schmitt, J., Lüthi, D., et al. 2010b. The role of Southern
dates, Qinling Mountains, China. Quaternary Research 62, Ocean processes in orbital and millennial CO2 variations – a
194–200. synthesis. Quaternary Science Reviews 29, 193–205.
Feng, Z.-D., Johnson, W.C., Sprowl, D.R., et al. 1994. Loess Fisher, D., Osterberg, E., Dyke, A., et al. 2008. The Mt Logan
accumulation and soil formation in central Kansas, United Holocene–late Wisconsian isotope record: tropical Pacific–
Yukon connections. The Holocene 18, 1–11.
States, during the past 400,000 years. Earth Surface Processes and
Fisher, D.C., Tikhonov, A.N., Kosintsev, P.A., et al. 2012. Anatomy,
Landforms 19, 55–67.
death, and preservation of a woolly mammoth (Mammuthus
Ferguson, C.W., Graybill, D.A. 1983. Dendrochronology of bristle-
primigenius) calf, Yamal Peninsula, northwest Siberia. Quatern-
cone pine; a progress report. Radiocarbon 25, 287–88.
ary International 255, 94–105.
Fernández, E., Ortiz, J.E., Pérez-Pérez, A., et al. 2009. Aspartic acid
Fitzsimmons, K.E., Magee, J.W., Amos, K.J. 2009. Characterisation
racemization variability in ancient human remains: implica-
of aeolian sediments from the Strzelecki and Tirari Deserts,
tions in the prediction of ancient DNA recovery. Journal of
Australia: implications for reconstructing palaeoenvironmental
Archaeological Science 36, 965–972.
conditions. Sedimentary Geology 218, 61–73.
Fernández, M.H. 2006. Rodent paleofaunas as indicators of climatic
Fitzsimmons, K.E., Marković, S.B., Hambech, U. 2012. Pleistocene
change in Europe during the last 125,000 years. Quaternary environmental dynamics recorded in the loess of the middle
Research 65, 308–323. and lower Danube basin. Quaternary Science Reviews 41,
Fernández, M.H., Vrba, E.S. 2006. Plio–Pleistocene climatic change 104–128.
in the Turkana Basin (East Africa): evidence from large mammal Fjeldskaar, W., Lindholm, C., Dehls, J.F., et al. 2000. Postglacial
faunas. Journal of Human Evolution 50, 595–626. uplift, neotectonics and seismicity in Fennoscandia. Quaternary
Fiedel, S.J. 2005. Man’s best friend – mammoth’s worst enemy? A Science Reviews 19, 1413–1422.
speculative essay on the role of dogs in Paleoindian colonization Flato, G.M. 2011. Earth system models: an overview. Wiley Inter-
and megafaunal extinction. World Archaeology 37, 11–25. disciplinary Reviews: Climate Change 2, 783–800.
Field, J., Fillios, M., Wroe, S. 2008. Chronological overlap between Fleitmann, D., Burns, S.J., Neff, U., et al. 2004. Palaeoclimatic
humans and megafauna in Sahul (Pleistocene Australia–New interpretation of high-resolution oxygen isotope profiles derived
Guinea): a review of the evidence. Quaternary Science Reviews 89, from annually laminated speleothems from southern Oman.
97–115. Quaternary Science Reviews 23, 935–945.
Fifield, L.K., Morgenstern, U. 2009. Silicon-32 as a tool for dating Fleming, K., Johnston, P., Zwartz, D., et al. 1998. Refining the
the recent past. Quaternary Geochronology 4, 400–405. eustatic sea-level curve since the Last Glacial Maximum using
Finlayson, A., Merritt, J., Browne, M., et al. 2010. Ice sheet advance, far- and intermediate-field sites. Earth and Planetary Science
dynamics, and decay configurations: evidence from west central Letters 163, 327–342.
Scotland. Quaternary Science Reviews 29, 969–988. Flint, R.F. 1957. Glacial and Pleistocene Geology. John Wiley, New
Finsinger, W., Heiri, O., Valsecchi, V., et al. 2007. Modern pollen York.
assemblages as climate indicators in southern Europe. Global Flores, J.-A., Sierro, F.J. 2007. Coccoliths. In Encyclopedia of
Ecology and Biogeography 16, 567–582. Quaternary Science (edited by S.A. Elias). Elsevier, Amsterdam,
Firestone, R.B., West, A., Kennett, J.P., et al. 2007. Evidence for an 1634–1646.
extra-terrestrial impact 12,900 years ago that contributed to the Flores, J.-A., Colmenero-Hidalgo, E., Mejía-Melina, A.E., et al. 2010.
megafaunal extinctions and the Younger Dryas cooling. Distribution of large Emiliania huxleyi in the Central and
Proceedings of the National Academy of Sciences of the United States Northeast Atlantic as a tracer of surface ocean dynamics during
of America 104, 16016–16021. the last 25,000 years. Marine Micropaleontology 76, 53–66.
468 REFERENCES
Flores, J.-A., Filippelli, G.M., Sierro, F.J., et al. 2012. The ‘White Ocean’ French, H.M., Shur, Y. 2010. The principles of cryostratigraphy. Earth-
hypothesis: a late Pleistocene Southern Ocean governed by Science Reviews 101, 190–206.
coccolithophores and driven by phosphorus. Frontiers in Frenzel, P., Boomer, I. 2005. The use of ostracods from marginal
Microbiology 3, doi: 10.3389/fmicb.2012.00233. marine, brackish waters as bioindicators of modern and
Florinski, V., Axford, W.I., Zank, G.P. 2004. The cosmic ray increases Quaternary environmental change. Palaeogeography, Palaeo-
at 35 and 60 kyr BP. Radiocarbon 46, 683–690. climatology, Palaeoecology 225, 68–92.
Flower, B.P., Oppo, D.W., McManus, J.F., et al. 2000. North Atlantic Fricke, H.C., O.Neill, J.R. 1999. The correlation between 18O/16O
intermediate to Deep Water circulation and chemical stratifi- ratios of meteoric water and surface temperature: its use in
cation during the past 1 Myr. Paleoceanography 15, 388–403. investigating terrestrial climatic change over geologic time. Earth
Flower, B.P., Hastings, D.W., Randle, N.J. 2009. Paired AMS 14C dates and Planetary Science Letters 170, 181–196.
on planktic foraminifera from a Gulf of Mexico sediment core: Friedman, I., Obradovich, J. 1981. Obsidian hydration dating of
an assessment of stratigraphic continuity. Radiocarbon 53, volcanic events. Quaternary Research 16, 37–47.
337–344. Friedrich, M., Remmele, S., Kromer, B., et al. 2004. The 12,460-year
Flückiger, J., Knutti, R., White, J.W.C. 2006. Oceanic processes as Hohenheim oak and pine tree-ring chronology from Central
potential trigger and amplifying mechanisms for Heinrich events. Europe – a unique annual record for radiocarbon calibration
Paleoceanography 21, doi: 10.1029/2005PA001204. and paleoenvironment reconstructions. Radiocarbon 46,
Ford, D., 2006. Karst geomorphology, caves and cave deposits: a 1111–1122.
review of North American contributions during the past half Fritts, H.C. 2001. Tree Rings and Climate, 2nd edn. Blackburn Press,
century. Geological Society of America, Special Paper 404, 1–13. Caldwell, NJ.
Forester, R.M., Smith, A.J., Palmer, D.F., et al. 2005. North American Fritz, S.C. 2008. Deciphering climatic history from lake sediments.
Non-marine Ostracode Database ‘NANODe’ version 1, http:// Journal of Paleolimnology 39, 5–16.
www.kent.edu/NANODe, Kent State University, Kent, OH. Fritz, S.C., Baker, P.A., Seltzer, G.O., et al. 2007. Quaternary glaciation
Forman, S.L., Marin, L., van der Veen, C., et al. 2007. Little Ice Age and hydrologic variation in the South American tropics as
and neoglacial landforms at the Inland Ice margin, Isunguata reconstructed from the Lake Titicaca drilling project. Quaternary
Sernia, Kangerlussuaq, west Greenland. Boreas 36, 341–351. Research 68, 410–420.
Forman, S.L., Marín, L., Gomez, G., et al. 2008. Late Quaternary eolian Fritz, S.C., Cumming, B.F., Gasse, F., et al. 2010a. Diatoms as
sand depositional record for southwestern Kansas: landscape indicators of hydrologic and climatic change in saline lakes. In
sensitivity to droughts. Palaeogeography, Palaeoclimatology, The Diatoms: Applications for the Environmental and Earth
Palaeoecology 265, 107–120. Sciences, 2nd edn (edited by J.P. Smol, E.F. Stoermer). Cambridge
Fornós J.J., Gelabert B., Ginés A., et al. 2002. Phreatic overgrowths University Press, Cambridge, 186–208.
on speleothems: a useful tool in structural geology in littoral Fritz, S.C., Baker, P.A., Ekdahl, E., et al. 2010b. Millennial-scale climate
karstic landscapes. The example of eastern Mallorca (Balearic variability during the Last Glacial period in the tropical Andes.
Islands). Geodinamica Acta 15, 113–125. Quaternary Science Reviews 29, 1017–1024.
Foss, P. 1987. The distribution and formation of Irish peatlands. Froese, D.G., Lowe, D.J., Knott, J.R., et al. (eds). 2008. Global tephra
In The IPCC Guide to Irish Peatlands (edited by C. O’Connell), studies: John Westgate and Andrei Sarna-Wojcicki com-
Irish Peatland Conservation Council, Ireland, 5–12. memorative volume. Quaternary International 178, 1–319.
Foulds, S.A., Macklin, M.G. 2006. Holocene land-use change and its Froyd, C.A., Willis, K.J. 2008. Emerging issues in biodiversity &
impact on river basin dynamics in Great Britain and Ireland. conservation management: the need for a palaeoecological
Progress in Physical Geography 30, 589–604. perspective. Quaternary Science Reviews 17–18, 1723–1732.
Frauenfelder, R., Haeberli, W., Hoelzle, M., et al. 2001. Using relict Frumkin, A., Ford, D.C., Schwarcz, H.P. 1999. Continental oxygen
rockglaciers in GIS-based modelling to reconstruct Younger isotopic record of the last 170,000 years in Jerusalem. Quaternary
Dryas permafrost distribution patterns in the Err-Julier area, Swiss Research 51, 317–327.
Alps. Norsk Geografisk Tidsskrift 55, 195–205 Fuchs, M., Lang. A., Wagner, G.A. 2004. The history of Holocene soil
Frechen, M., Vanneste, K., Veerbeeck, K., et al. 2001. The deposition erosion in the Phlious Basin, NE Peleponnese, Greece, based on
history of the coversands along the Bree Fault escarpment, NE optical dating. The Holocene 14, 334–345.
Belgium. Geologie en Mijnbouw (Netherlands Journal of Fuhrmann, A., Fischer, T., Lücke, A., et al. 2004. Late Quaternary
Geosciences) 80, 171–185. environmental and climatic changes in central Europe as in-
Frechen, M., Serrialta, M., Oezen, D., et al. 2007. Uranium-series ferred from the composition of organic matter in annually
dating of peat from central and northern Europe. In The Climate laminated lake maar sediments. Geochemistry, Geophysics,
of Past Interglacials (edited by F. Sirocko, M. Claussen, M.F. Geosystems (G3), 5, Q11015.
Sanchez-Goñi, et al.). Elsevier, Amsterdam, 93–118. Fujioka, T., Chappell, J., Fifield, L.K., et al. 2009. Australian desert
French, C. 2003. Geoarchaeology in Action. Routledge, London. dune fields initiated with Pliocene–Pleistocene global climatic
French, H.M. 2007. The Periglacial Environment, 3rd edn. John shift. Geology 37, 51–54.
Wiley, Chichester and New York. Fuller, S., Murray, T. 2002. Sedimentological investigations in the
French, H.M. 2008. Recent contributions to the study of past forefield of an Icelandic surge-type glacier: implications for
permafrost. Permafrost and Periglacial Processes 19, 179–194. the surge mechanism. Quaternary Science Reviews 21, 1503–1520.
REFERENCES 469
Fullerton, D.S., Clayton, R.B. 1986. Stratigraphy and correlation of responses to climate change. Palaeogeography, Palaeoclimatology,
the glacial deposits on the Montana Plains. Quaternary Science Palaeoecology 312, 98–114.
Reviews 5, 161–169. Garnett, E.R., Gilmour, M.A., Rowe, P.J., et al. 2004. 230Th/234U dating
Funder, S., Demidov, I., Yelovicheva, Y. 2002. Hydrography and of Holocene tufas: possibilities and problems. Quaternary Science
mollusc faunas of the Baltic and the White Sea–North Sea seaway Reviews 23, 947–958.
in the Eemian. Palaeogeography, Palaeoclimatology, Palaeoecology Gaschen, A.A.-M., Döbeli, M., Markwitz, A., et al. 2008. Restrictions
184, 275–304. on fluorine depth profiling for exposure age dating in archaeo-
Funnell, B.F. 1995. Global sea level and the (pen)insularity of late logical bones. Journal of Archaeological Science 35, 535–552.
Cenozoic Britain. In Island Britain: A Quaternary Perspective Gauthier, E., Bichet, V., Petit, C., et al. 2010. Pollen and non-pollen
(edited by R.C. Preece). Geological Society of London, Special palynomorph evidence of medieval farming activities in
Publication 96. Geological Society, London, 3–14. southwestern Greenland. Vegetation History and Archaeobotany
Gagan, M.K., Ayliffe, L.K., Beck, J.W., et al. 2000. New views of 19, 427–438.
tropical paleoclimates from corals. Quaternary Science Reviews Gavin, D.G., Hallett, D.J., Hu, F.S., et al. 2007. Forest fire and climate
19: 45–64. change in western North America: insights from sediment
Gaiero, D.M. 2007. Dust provenance in Antarctic ice during glacial charcoal records. Frontiers in Ecology and the Environment 5,
periods: from where in southern South America? Geophysical 499–506.
Research Letters 34, doi: 10.1029/2007GL030520. Gearey, B.R., Caseldine, C.J. 2006. Archaeological applications of
Gajewski, K. 2008. The Global Pollen Database in biogeographical testate amoebae analyses: a case study from Derryville, Co.
and palaeoclimatic studies. Progress in Physical Geography 32, Tipperary, Ireland. Journal of Archaeological Science 33, 49–55.
379–402. Gedye, S.J., Jones, R.T., Tinner, W., et al. 2000. The use of mineral
Gale, S.J. 2009. Event chronostratigraphy: a high-resolution tool for magnetism in the reconstruction of fire history: a case study from
dating the past. Quaternary Geochronology 4, 391–399. Lago di Origlio, Swiss Alps. Palaeogeography, Palaeoclimatology,
Gale, S.J., Hoare, P.J. 2011. Quaternary Sediments: Petrographic
Methods for the Study of Unlithified Rocks, 2nd edn. Blackburn
Gehrels, R. 2010. Sea-level changes since the Last Glacial Maximum:
Press, Caldwell, NJ.
an appraisal of the IPCC Fourth Assessment Report. Journal of
Gale, S.J., Hoare, P.G. 2012. The stratigraphic status of the
Anthropocene. The Holocene 23, 397–414.
Gehrels, W.R., Milne, G.A., Kirby, J.R., et al. 2004. Late Holocene
Galili, E., Zviely, D., Ronen, A., et al. 2007. Beach deposits of MIS
sea-level changes and isostatic crustal movements in Atlantic
5e high sea stand as indicators for tectonic stability of the
Canada. Quaternary International, 120, 79–89.
Carmel coastal plain, Israel. Quaternary Science Reviews 26,
Gehrels, W.R., Marshall, W.A., Gehrels, M.J., et al. 2006. Rapid
2544–2557.
sea-level rise in the North Atlantic Ocean since the first half of
Gallet, Y., Genevey, A., Courtillot, V. 2003. On the possible
occurrence of ‘archaeomagnetic jerks’ in the geomagnetic field the 19th century. The Holocene 16, 948–964.
over the past three millennia. Earth and Planetary Science Letters Geiss, C.E., Umbanhowar, C.E., Camill, P., et al. 2003. Sediment
214, 237–242. magnetic properties reveal Holocene climate change along the
Ganopolski, A., Calov, R. 2011. The role of orbital forcing, carbon Minnesota prairie–forest ecotone. Journal of Paleolimnology 30,
dioxide and regolith in 100 kyr glacial cycles. Climate of the 151–166.
Past 7, 1415–1425. Genty, D., Blamart, D., Ouahdi, R., et al. 2003. Precise dating of
Gao, J., Liu, Y. 2001. Applications of remote sensing, GIS and GPS Dansgaard–Oeschger climatic oscillations in western Europe
in glaciology: a review. Progress in Physical Geography 25, from stalagmite data. Nature 421, 833–837.
520–540. Georgetti, G., Talarico, F., Sandroni, S., et al. 2009. Provenance
Garcia, J.L., Kaplan, M.R., Hall, B.L. 2012. Glacier expansion in of Pleistocene sediments in the ANDRILL AND-1B drillcore:
southern Patagonia throughout the Antarctic cold reversal. clay and heavy mineral data. Global and Planetary Change 69,
Geology 40, 859–862. 94–102.
Garcia, T., Féraud, G., Falguères C., et al. 2010. Earliest human Gersonde, R., Zielinski, U. 2000. The reconstruction of late
remains in Eurasia: new 40Ar/39Ar dating of the Dmanisi hominid- Quaternary Antarctic sea-ice distribution – the use of diatoms
bearing levels, Georgia. Quaternary Geochronology 5, 443–451. as a proxy for sea-ice. Palaeogeography, Palaeoclimatology,
Garcin, Y., Junginger, A., Melnick, D., et al. 2009. Late Pleistocene– Palaeoecology 162, 263–286.
Holocene rise and collapse of Lake Suguta, northern Kenya Rift. Gersonde, R., Crosta, X., Abelmann, A., et al. 2005. Sea-surface
Quaternary Science Reviews 28, 911–925. temperature and sea ice distribution of the Southern Ocean at
Gardner, T., Webb, J., Pezzia, C., et al. 2008. Episodic intraplate the EPILOG Last Glacial Maximum – a circum-Antarctic view
deformation of stable continental margins: evidence from based on siliceous microfossil records. Quaternary Science Reviews
Late Neogene and Quaternary marine terraces, Cape Liptrap, 24, 869–896.
Southeastern Australia. Quaternary Science Reviews 28, 39–53. Gervais B.R., MacDonald G.M. 2001. Tree-ring and summer-
Garilli, V. 2011. Mediterranean Quaternary interglacial molluscan temperature response to volcanic aerosol forcing at the northern
assemblages: Palaeobiogeographical and palaeoceanographical tree-line, Kola Peninsula, Russia. The Holocene 11, 499–505.
470 REFERENCES
Geyh, M., Schotterer, U., Grosjean, M. 1998. Temporal changes of Gladstone W. 2002. The potential value of indicator groups in the
the 14C reservoir effect in lakes. Radiocarbon 40, 921–931. selection of marine reserves. Biological Conservation 104,
Gherardi, J.-M., Labeyrie, L., Nave, S., et al. 2009. Glacial–interglacial 211–220.
circulation changes inferred from 231Pa/230Th sedimentary record Glasser, N.F., Hambrey, M.J. 2001. Styles of sedimentation
in the North Atlantic region. Paleoceanography 24, PA2204, doi: beneath Svalbard valley glaciers under changing dynamic and
10.1029/2008PA001696. thermal regimes. Journal of the Geological Society of London
Ghilani, C.D., Wolf, P.R. 2008. Elementary Surveying: An Intro- 158, 697–707.
duction to Geomatics, 12th edn. Prentice Hall, Upper Saddle Glasser, N.F., Bennett, M.R. 2004. Glacial erosional landforms:
River, NJ. origins and significance for palaeoglaciology. Progress in Physical
Ghoneim, E., El-Baz, F. 2007. The application of radar topographic Geography 28, 43–75.
data to mapping of a mega-paleodrainage in the Eastern Sahara. Glasser, N., Harrison, S., Jansson, K.N. 2009. Topographic controls
Journal of Arid Environments 69, 658–675. on glacier sediment–landform associations around the temperate
Giardino, J.R., Vitek, J.D. 1988. The significance of rock glaciers North Patagonian Icefield. Quaternary Science Reviews 28,
in the glacial–periglacial landscape continuum. Journal of 2817–2832.
Quaternary Science 3, 97–103. Gob, F., Petit, F., Bravard J.-P., et al. 2003. Lichenometric application
Gibbard, P.L. 1994. Pleistocene History of the Lower Thames Valley. to historical and subrecent dynamics and sediment transport
Cambridge University Press, Cambridge. of a Corsican stream (Figarella River – France). Quaternary
Gibbard, P.L., van Kolfschoten, Th. 2005. The Pleistocene and Science Reviews 22, 2111–2124.
Holocene Series. In A Geologic Timescale 2004 (edited by Goble, R.J., Mason, J.A., Loope, D.B., et al. 2004. Optical and
F. Gradstein, J. Ogg, A.G. Smith). Cambridge University Press, radiocarbon ages of stacked paleosols and dune sands in the
Cambridge, 441–452. Nebraska Sand Hills, USA. Quaternary Science Reviews 23,
Gibbard, P.L., Lewin, J. 2009. River incision and terrace formation 1173–1182.
in the Late Cenozoic of Europe. Tectonophysics 474, 41–55. Godfrey-Smith, D.I., Casey, J.L. 2003. Direct thermoluminescence
Gibbard, P.L., Walker, M.J.C. 2013. The Term ‘Anthropocene’ in the chronology for Early Iron Age smelting technology on the
Context of Formal Geological Classification. Geological Society of Gambaga Escarpment, northern Ghana. Journal of Archaeological
London, Special Publication, 395, doi: 10.1144/SP395.1.
Science 30, 1037–1050.
Gibbard, P.L., West, R.G., Zagwijn, W.H., et al. 1991. Early and Middle
Godsey, H.S., Currey, D.R., Chan, M.A. 2005. New evidence for an
Pleistocene correlations in the Southern North Sea Basin.
extended occupation of the Provo shoreline and implications
for regional climate change, Pleistocene Lake Bonneville, Utah,
Gibbard, P.L., Smith, A.G., Zalasiewicz, J., et al. 2005. What status
USA. Quaternary Research 63, 212–223.
for the Quaternary? Boreas 34, 1–6.
Goehring, B.M., Lohne Ø.S., Mangerud, J., et al. 2012. Late Glacial
Gibbard, P.L., Head, M.J., Walker, M.J.C., et al. 2010. Formal
and Holocene beryllium-10 production rates for western
ratification of the Quaternary System/Period and the Pleistocene
Norway. Journal of Quaternary Science 27, 89–96.
Series/Epoch with a base at 2.58 Ma. Journal of Quaternary
Gögen K., Wagner, G.A. 2002. Alpha-recoil track dating of
Science 25, 96–102.
Giesecke, T., Fontana, S.L. 2008. Revisiting pollen accumulation rates Quaternary volcanics. Chemical Geology 166, 127–137.
from Swedish lake sediments. The Holocene 18, 293–305. Goldstein, S.J., Stirling, C.H. 2003. Techniques for measuring
Gildor, H., Tziperman, E. 2001. A sea ice climate switch mechanism uranium-series nuclides: 1992–2002. In Uranium Series Geo-
for the 100-kyr cycles. Journal of Geophysical Research 106, chemistry (edited by B. Bourdon, G.M. Henderson, C.C.
9117–9133. Lundstrom, et al.). Reviews in Mineralogy and Geochemistry
Gildor, H., Tziperman, E., Toggweiler, J.R. 2002. Sea ice switch 52, 23–57.
mechanism and glacial–interglacial CO2 variations. Global Golledge, N.R. 2010. Glaciation of Scotland during the Younger Dryas
Biogeocehmical Cycles 16, 6–1 to 6–14. stadial: a review. Journal of Quaternary Science 25, 550–566.,
Gill, J.L., Williams, J.W., Jackson, S.T., et al. 2009. Pleistocene Goman, M. 2001. Statistical analysis of modern seed assemblages
megafaunal collapse, novel plant communities, and enhanced fire from the San Francisco Bay: applications for the reconstruction
regimes in North America. Science 326, 1100–1103. of paleo-salinity and paleo-tidal inundation. Journal of Paleo-
Gilligan, I. 2007. Neanderthal extinction and modern human limnology 24, 393–409.
behaviour: the role of climate change and clothing. World Gomez, B., Carter, L., Trustrum, N.A., et al. 2004. El Niño–Southern
Archaeology 39, 499–514. Oscillation signal associated with middle Holocene climatic
Giraudeau, J., Grelaud, M., Solignac, S., et al. 2010. Millennial-scale changes in intercorrelated terrestial and marine sediment cores,
variability in Atlantic water advection to the Nordic Seas derived New Zealand. Geology 32, 653–656.
from Holocene coccolith concentration records. Quaternary Gonzalez, L.M., Grimm, E.C. 2009. Synchronisation of late-
Science Reviews 29, 1276–1287. glacial vegetation changes at Crystal Lake, Illinois, USA, with the
Gischler, E., Hudson, J.H., Storz, D. 2009. Growth of Pleistocene North Atlantic Event Stratigraphy. Quaternary Research 72,
massive corals in south Florida: low skeletal extension-rates and 234–245.
possible ENSO, decadal, and multi-decadal cyclicities. Coral González-Sempériz, P., Valero-Garcés, Moreno, A., et al. 2006.
Reefs 28, 823–830. Climate variability in the Spanish Pyrenees during the last
REFERENCES 471
30,000 yr revealed by the El Portalet sequence. Quaternary Green, D.G. 1995. Time and spatial analysis. In Statistical Modelling
Research 66, 38–52. and Quaternary Science Data (edited by D. Maddy, J.S.
Goodfellow, B.W. 2007. Relict non-glacial surfaces in formerly Brew). Technical Guide 5. Quaternary Research Association,
glaciated landscapes. Earth-Science Reviews 80, 47–73. Cambridge, 65–105.
Goodfriend, G.A., Mitterer, R.M. 1993. A 45,000 year record of a Green, O.R. 2001. A Manual of Practical Laboratory and Field
tropical lowland biota: the land snail fauna from cave sediments Techniques in Palaeobiology. Springer, Dordrecht.
at Coco Ree, Jamaica. Geological Society of America Bulletin 105, Greenwood, M.T., Agnew, M.D., Wood, P.J. 2003. The use of
18–29. caddisfly fauna (Insecta: Trichoptera) to characterise the Late-
Goodfriend, G., Collins, M.J., Fogel, M.L., et al. (eds). 2000. glacial River Trent, England. Journal of Quaternary Science 18,
Perspectives in Amino Acid and Protein Geochemistry. Oxford 645–661.
University Press, Oxford. Greenwood, S.L., Clark, C.D. 2009. Reconstructing the last Irish Ice
Goodman, A.Y., Rodbell, D.T., Seltzer, G.O., et al. 2001. Subdivision Sheet 1: changing flow geometries and ice flow dynamics
of glacial deposits in southeastern Peru based on pedogenic deciphered from the glacial landform record. Quaternary Science
development and radiometric ages. Quaternary Research 56, Reviews 28, 3085–3100.
31–50. Gregory, J.M., Lowe, J.A., Tett, S.F.B. 2006. Simulated global-mean
Goodwin, D.H., Flessa, K.W., Schöne, B., et al. 2001. Cross- sea level changes over the last half-millennium. Journal of Climate
calibration of daily growth increments, stable isotope variation, 19, 4576–4591
and temperature in the Gulf of California bivalve mollusk Griffiths, H.I. 2001. Ostracod evolution and extinction – its
Chione coretzi: implications for paleoenvironmental analysis. biostratigraphic value in the European Quaternary. Quaternary
Palaios 16, 387–398. Science Reviews 20, 1743–1751.
Gordon, S.J., Dorn, R.I. 2005. In situ weathering rind erosion. Griffiths, J.S. (ed.). 2002. Mapping in Engineering Geology. (‘Key Issues
Geomorphology 67, 97–113. in Earth Sciences’ Series), Geological Society of London, London.
Goslar, T., Arnold, M., Tisnerat-Laborde, N., et al. 2000. Variations Griffiths, J.S., Brunsden, D., Lee, E.M., et al. 1995. Geomorphological
of Younger Dryas atmospheric radiocarbon explicable without investigations of the Channel Tunnel Terminal and Portal. The
ocean circulation changes. Nature 403, 877–880. Geographical Journal 161, 275–284.
Gosling, W.D., Mayle, F.E., Killeen, T.J., et al. 2003. A simple and Griggs, C.B., Kromer, B. 2008. Wood macrofossils and dendro-
effective methodology for sampling modern pollen rain in
chronology of three mastodon sites in upstate New York.
tropical environments. The Holocene 13, 613–618.
Palaeontologica Americana 61, 49–61.
Gosse, J.C., Phillips F.M. 2001. Terrestrial in situ cosmogenic
Grimley, D.A., Follmer, L.R., Hughes, R.E., et al. 2003. Modern
nuclides: theory and applications. Quaternary Science Reviews 20,
Sangamon and Yarmouth soil development in loess of unglaci-
1475–1560.
ated southwestern Illinois. Quaternary Science Reviews 22,
Gosse, J.C., Evenson, E.B., Klein, J., et al. 1995. Precise cosmogenic
10 225–244.
Be measurements in western North America: support for a
Gröcke, D.R., Gilliken, D.P. 2008. Advances in mollusc
global Younger Dryas cooling event. Geology 23, 877–880.
sclerochronology and sclerochemistry: tools for understanding
Goudie, A.S. 2006. The Schmidt hammer in geomorphological
climate and environment. Geo-Marine Letters 28, 265–268.
research. Progress in Physical Geography 30, 703–718.
Grab, S. 2005. Aspects of the geomorphology, genesis and environ- Gröcke, D.R., Schimmelmann, A., Elias, S., et al. 2006. Stable
mental significance of earth hummocks (thúfur, pounus): hydrogen-isotope ratios in beetle chitin: preliminary European
miniature cryogenic mounds. Progress in Physical Geography data and re-interpretation of North American data. Quaternary
29, 139–155. Science Reviews 25, 1850–1864.
Gradstein, F.M., Ogg, J.G., Smith, A.G., et al. 2005. A Geological Time Gröcke, D.R., Van Hardenbroek, M., Sauer, P. E., et al. 2011.
Scale 2004. Cambridge University Press, Cambridge. Hydrogen isotopes in beetle chitin. Topics in Geobiology, 34,
Gradstein, F.M., Ogg, J.G., Ogg, G., et al. 2012. A Geologic Time Scale 105–116.
2012. Elsevier, Amsterdam. Grootes, P.M. 1978. Carbon-14 timescale extended: comparison of
Graham, A.G.C., Lonergan, L., Stoker, M.S. 2007. Evidence for Late chronologies. Science 200, 11–15.
Pleistocene ice stream activity in the Witch Ground Basin, central Gross, R.S., Fukumori, I., Menemenlis, D. 2005. Atmospheric and
North Sea, from 3D seismic reflection data. Quaternary Science oceanic excitation of decadal-scale Earth orientation vari-
Reviews 26, 627–643. ations. Journal of Geophysical Research 110, doi: 10.1029/2004
Graham, I.J., Ditchburn, R.G., Whitehead, N.E. 2001. Be isotope JB003565.
analysis of a 0–500 ka loess–paleosol sequence from Rangitatau Grosse, G., Jones, B.M. 2011. Spatial distribution of pingos in
East, New Zealand. Quaternary International 76/77, 29–42. northern Asia. The Cryosphere 5, 13–33.
Graham R.W., Lundelius Jr., E.L., Graham, M.A., et al. 1996. Spatial Grosse, G., Schirrmeister, L., Kunitsky, V.V., et al. 2005. The use of
response of mammals to Late Quaternary environmental CORONA images in remote sensing of periglacial geomor-
fluctuations. Science 272, 1601–1606. phology: an illustration from the NE Siberian coast. Permafrost
Gray, J.M. 1984. The Quaternary geomorphology of Scotland: the and Periglacial Processes 16, 163–172.
research contribution of J.B. Sissons. Quaternary Science Reviews Grottoli, A.G., Eakin, C.M. 2007. A review of modern coral δ18O and
3, 259–289. Δ14C proxy records. Earth-Science Reviews 81, 67–91.
472 REFERENCES
Grove, J.M. 1988. The Little Ice Age. Routledge, London. Conference on Permafrost (edited by M. Phillips, S.M. Springman,
Grudd, H., Briffa, K.R., Karlén, W., et al. 2002. A 7400-year tree-ring L.U. Arenson). Taylor & Francis Group, 343–348.
chronology in northern Swedish Lapland: natural climatic Haesaerts, P., Mestdagh, H. 2000. Pedosedimentary evolution of the
variability expressed on annual to millennial timescales. The last interglacial and early glacial sequence in the European loess
Holocene 12, 657–666. belt from Belgium to central Russia. Geologie en Mijnbouw
Grün, R. 2001. Trapped charge dating (ESR, TL, OSL). In Handbook (Netherlands Journal of Geosciences) 79, 313–324.
of Archaeological Science (edited by D.E. Brothwell, A.M. Pollard). Haflidason, H., Eiriksson, J., van Kreveld, S. 2000. The tephro-
John Wiley, Chichester and New York, 47–62. chronology of Iceland and the North Atlantic region during
Grün, R. (ed.). 2010. 12th International Conference on Luminescence the Middle and Late Quaternary: a review. Journal of Quaternary
and Electron Spin Resonance Dating (LED 2008). Quaternary Science 15, 3–22.
Geochronology 5, 83–390. Hailer, F., Kutschera, V.E., Hallström, B.M., et al. 2012. Nuclear
Grün, R., Schwarcz, H.P. 2000. Revised open-system U-series/ESR genomic sequences reveal that polar bears are an old and distinct
age calculations for teeth from Stratum C at the Hoxnian bear lineage. Science 336, 344–347.
Interglacial type locality, England. Quaternary Science Reviews 19, Hajdas, I., Bonani, G., Moreno, P.I., et al. 2003. Precise radiocarbon
1151–1154. dating of Late-Glacial cooling in mid-latitude South America.
Grün, R., Preusser, F. (eds). 2012. 13th International Conference on Quaternary Research 59, 70–78.
Luminescence and Electron Spin Resonance (LED 2011). Håkansson, L., Jansson, M. 1983. Principles of Lake Sedimentology.
Quaternary Geochronology 10, 1–442. Springer, Berlin.
Grün, R., Aubert, M., Hellstrom, J., et al. 2010. The challenge of direct Hall, A.R. 1996. A survey of palaeobotanical evidence for dyeing and
dating of old human bones. Quaternary International, 223–224, mordanting from British archaeological excavations. Quaternary
87–93. Science Reviews 15, 635–640.
Gubbins, D. 2008. Earth science: Geomagnetic reversals. Nature Hall, B.L., Denton, G.H., Fountain, A.G., et al. 2010. Antarctic lakes
452, 165–167. suggest millennial reorganizations of Southern Hemisphere
Gunn, J. (ed.). 2003. Encyclopedia of Caves and Karst Science. atmospheric and oceanic circulation. Proceedings of the National
Routledge, London. Academy of Sciences of the United States of America 107,
Gunnarson, B.E., Linderholm, H.W. 2002. Low-frequency summer 21355–21359.
temperature variation in central Sweden since the tenth century Hall, R.D., Anderson, A.K. 2000. Comparative soil development of
inferred from tree rings. The Holocene 12, 667–672. Quaternary paleosols of the central United States. Palaeo-
Gunnell, Y., Jarman, D., Braucher, R., et al. 2013. The granite tors geography, Palaeoclimatology, Palaeoecology 158, 109–145.
of Dartmoor, southwest England: rapid and recent emergence Hall, R.I., Smol, J.P. 2010. Diatoms as indicators of lake eutroph-
revealed by Late Pleistocene cosmogenic apparent exposure ication. In The Diatoms: Applications for the Environmental and
ages. Quaternary Science Reviews 61, 62–76. Earth Sciences, 2nd edn (edited by J.P.Smol, E.F. Stoermer).
Günster, N., Eck, P., Skowronek, A., et al. 2001. Late Pleistocene loess Cambridge University Press, Cambridge, 122–151.
and their paleosols in the Granada Basin, southern Spain. Hambley, G.W., Lamoureux, S.F. 2006. Recent summer climate
Quaternary International 76/77, 241–245. recorded in complex varved sediments, Nicolay Lake, Cornwall
Gupta, A., Hock, L., Xiaojing, H., et al. 2002. Evaluation of part of Island, Nunavut, Canada. Journal of Paleolimnology 35, 629–640.
the Mekong River using satellite imagery. Geomorphology 44, Hambrey, M. 1994. Glacial Environments. UCL Press, London.
221–239. Hambrey, M.J., Fitzsimons, S.J. 2010. Development of sediment–
Gupta, S., Collier, J.S., Palmer-Felgate, A., et al. 2007. Catastrophic landform associations at cold glacier margins, Dry Valleys,
flooding origin of shelf valley systems in the English Channel. Antarctica. Sedimentology 57, 857–882.
Nature 448, 342–345. Hambrey, M.J., Glasser, N. 2012. Discriminating glacier thermal
Gurney, S. 2003. Relict cryogenic mounds in the UK as evidence of and dynamic regimes in the sedimentary record. Sedimentary
climate change. In Linking Climate Change to Land Surface Geology 251–252, 1–33.
Change (edited by S.J. McLaren, D.R. Kniveton). Kluwer Hambrey, M.J., Davies, J.R., Glasser, N.F., et al. 2001. Devensian
Academic Publishers, Dordrecht, 209–229. glacigenic sedimentation and landscape evolution in the
Guyard, H., Chapron, E., St-Onge, G., et al. 2007. High-altitude varve Cardigan area of southwest Wales. Journal of Quaternary Science
records of abrupt environmental change and mining activity over 16, 455–482.
the last 4000 years in the western French Alps (Lake Bramant, Hamilton, S.J., Whalley, W.B. 1995. Rock glacier nomenclature: a
Grandes Rousses Massif). Quaternary Science Reviews 26, re-assessment. Geomorphology 14, 73–80.
2644–2660. Hammarlund, D., Edwards, T.W.D., Björck, S., et al. 1999. Climate
Haase, D., Fink, J., Haase G., et al. 2007. Loess in Europe – its spatial and environment during the Younger Dryas (GS-1) as reflected
distribution based on a European Loess Map, scale 1:2,500,000. by composite stable isotope records of lacustrine carbonates at
Quaternary Science Reviews 26, 1301–1312. Torreberga, southern Sweden. Journal of Quaternary Science 14,
Haeberli, W., Brandová, D., Burga, C., et al. 2003. Methods for 17–28.
absolute and relative age dating of rock-glacier surfaces in alpine Hammarlund, D., Barnekow, L., Birks, H.J.B., et al. 2002. Holocene
permafrost. In Permafrost: Proceedings of the 8th International changes in atmospheric circulation recorded in the oxygen-
REFERENCES 473
isotope stratigraphy of lacustrine carbonates from southern Pleistocene mountain geomorphology and palaeoclimate.
Sweden. The Holocene 12, 339–351. Journal of Quaternary Science 23, 287–304.
Hammarlund, D., Björk, S., Buchardt, B., et al. 2003. Rapid Hart, J.K. 1994. Till fabric associated with deformable beds. Earth
hydrological changes during the Holocene revealed by stable Surface Processes and Landforms 19, 15–32.
isotope records of lacustrine carbonates from Lake Igelsjön, Hart, J.K. 1997. The relationship between drumlins and other forms
southern Sweden. Quaternary Science Reviews 22, 353–370. of subglacial glaciotectonic deformation. Quaternary Science
Hammer, C.U., Mayewski, P.A., Peel, D., et al. (eds). 1997. Greenland Reviews 16, 93–107.
Summit Ice Cores. Greenland Ice Sheet Project 2/Greenland Hart, J., Rose, J. 2001. Approaches to the study of glacier bed
Ice Core Project. Journal of Geophysical Research 102, deformation. Quaternary International 86, 45–58.
26315–26886. Hart, J.K., Rose, K.C., Martinez, K., et al. 2009. Subglacial clast
Hancock, G.S., Anderson, R.S. 2002. Numerical modeling of fluvial behaviour and its implication for till fabric development: new
strath-terrace formation in response to oscillating climate. results derived from wireless subglacial probe experiments.
Geological Society of America Bulletin 114, 1131–1142. Quaternary Science Reviews 28, 597–607.
Hansen, E.W.S. 2008. The application of lichenometry in dating of Harvey, A., Foster, G., Hannam, J., et al. 2003. The Tabernas alluvial
glacier deposits. Danish Journal of Geography 108, 143–151. fan and lake system, southeast Spain: applications of mineral
Hansen, E.S. 2010. A review of lichen growth and applied lichen- magnetic and pedogenic iron oxide analyses towards clarifying
ometry in southwest and southeast Greenland. Geografiska the Quaternary sediment sequences. Geomorphology 151–171.
Annaler 92, 65–79. Haslett, S.K. (ed.). 2002. Quaternary Environmental Micro-
Hansen, J. 2001. Macroscopic plant remains from Mediterranean palaeontology. Arnold, London and New York.
caves and rockshelters: avenues and interpretation. Geoarchae- Haslett, S.K. 2004. Late Neogene–Quaternary radiolarian bio-
ology 16, 401–432. stratigraphy: a brief review. Journal of Micropalaeontology 23,
Hanson, P.R., Arbogast, A.F., Johnson, W.C., et al. 2010. Mega- 39–47.
droughts and late Holocene dune activation at the eastern margin
Haslett, S.K., Smart, C.W. 2006. Late Quaternary upwelling off
of the Great Plains, north-central Kansas, USA. Aeolian Research
tropical NW Africa: new micropalaeontological evidence from
1, 101–110.
ODP Hole 658C. Journal of Quaternary Science 21, 259–269.
Haq, B.U., Boersma, A. 1998. Introduction to Marine Micro-
Hather, J.G. 2000. The Identification of the Northern European
palaeontology. Elsevier Science, Singapore.
Woods: A Guide for Archaeologists and Conservators. Archetype,
Harkness, D.D. 1975. The role of the archaeologist in C-14
London.
measurement. In Radiocarbon Calibration and Prehistory
Hatzopoulos, J. 2008. Topographic Mapping. Universal Publishers,
(edited by T. Watkins). Edinburgh University Press, Edinburgh,
Boca Raton, FL.
128–135.
Haug, G.H., Günther, D., Peterson, L.C., et al. 2003. Climate and the
Harlan, T., Robertson, I. 2006. Confirmation of the extended
Bristlecone pine chronology. Quaternary Newsletter 110, 37–39. collapse of the Maya civilization. Science 299, 1731–1735.
Harland, E. 1988. Dinoflagellates, their cysts and Quaternary Haug, G.H., Ganopolski, A., Sigman, D.M., et al. 2005. North Pacific
stratigraphy. New Phytologist 108, 111–120. seasonality and the glaciation of North America 2.7 million
Harland, R., FitzPatrick, M.E.J., Pudsey, C.J. 1999. Latest Quaternary years ago. Nature 433, 821–825.
dinoflagellate cyst climatostratigraphy for three cores from the Havinga, A.J. 1964. Investigation into the differential corrosion
Falkland Trough, Scotia and Weddell seas, Southern Ocean. susceptibility of pollen and spores in various soil types. Pollen et
Review of Palaeobotany and Palynology 107, 265–281. Spores 4, 621–635.
Harris, C., Kern-Luetschg, M., Smith, F., et al. 2008. Solifluction Hawkins, E., Jones, P.D. 2013. On increasing global temperatures:
processes in an area of seasonal ground freezing, Dovrefjell, 75 years after Callendar. Quarterly Journal of the Royal Meteoro-
Norway. Permafrost and Periglacial Processes 19, 31–47. logical Society, doi: 10.1002/qj.2178.
Harris, C., Arenson, L.U., Christiansen, H.H., et al. 2009. Permafrost Hay, R.L., Kyser, T.K. 2001. Chemical sedimentology and palaeo-
and climate in Europe: monitoring and modelling thermal, environmental history of Lake Olduvai, a Pliocene lake in
geomorphological and geotechnical responses. Earth-Science northern Tanzania. Geological Society of America Bulletin 113,
Reviews 92, 117–171. 1505–1521.
Harrison, S.P., Dodson, J. 1993. Climates of Australia and New Hayes, A., Kucera, M., Kallel, N., et al. 2005. Glacial Mediterranean
Guinea since 18,000 yr BP. In Global Climates Since the Last Glacial sea surface temperatures based on planktonic foraminiferal
Maximum (edited by H.E. Wright Jr., J.E. Kutzbach, T. Webb III, assemblages. Quaternary Science Reviews 24, 999–1016.
et al.). University of Minnesota Press, Minneapolis and London, Haynes, G. (ed.). 2009. American Megafaunal Extinctions at the
265–292. End of the Pleistocene. Springer Science and Business Media,
Harrison, S.P., Prentice, C., Guiot, J. 1993. Climatic controls on Dordrecht.
Holocene lake-level changes in Europe. Climate Dynamics 8, Haynes, G. 2013. Extinctions in North America’s Late Glacial
189–200. landscapes. Quaternary International 285, 89–98.
Harrison, S., Whalley, B., Anderson, E. 2008. Relict rock glaciers Hays, J.D., Imbrie, J., Shackleton, N.J. 1976. Variations in the earth’s
and protalus lobes in the British Isles: implications for Late orbit: pacemaker of the Ice Ages. Science 194, 1121–1132.
474 REFERENCES
Haywood, A.M., Valdes, P.J. 2004. Modelling Pliocene warmth: Heiri, O., Brooks, S.J., Birks, H.J.B., et al. 2011. A 274-lake calibration
contribution of atmosphere, oceans and cryosphere. Earth and data-set and inference model for chironomid-based summer
Planetary Science Letters 218, 363–377. air temperature reconstruction in Europe. Quaternary Science
He, F., Shakun, J.D., Clark, P.U., et al. 2013. Northern Hemisphere Reviews 30, 3445–3456.
forcing of Southern Hemisphere climate during the last Helama, S., Lindholm, M., Timonen, M., et al. 2002. The supra-long
deglaciation. Nature 494, 81–85. Scots pine tree-ring record for Finnish Lapland: Part 2, inter-
Head, M.J., Gibbard, P.L., Salvador, A. 2008a. The Quaternary: its annual to centennial variability in summer temperatures for 7500
character and definition. Episodes 31, 234–238. years. The Holocene 12, 681–687.
Head, M.J., Gibbard, P.L., Salvador, A. 2008b. The Early–Middle Hellman, S., Bunting, M.J., Gaillard, M.-J. 2009. Relevant source
Pleistocene Transition: characterisation and proposed guide for area of pollen in patchy cultural landscapes and signals of
defining the boundary. Episodes 31, 255–259. anthropogenic landscape disturbance in the pollen record:
Hearty, P.J. 2003, Stratigraphy and timing of aeolianite deposition a simulation approach. Review of Palaeobotany and Palynology
on Rottnest Island, Western Australia. Quaternary Research 60, 153, 245–258.
211–222. Hemming, S.R. 2004. Heinrich events: massive Late Pleistocene
Hearty, P.J., Kindler, P. 1995. Sea level highstand chronology from detritus layers of the North Atlantic and their global climate
stable carbonate platforms (Bermuda and the Bahamas). Journal imprint. Review of Geophysics 42, 1–43.
of Coastal Research 11, 675–689. Henderson, G.M. 2002. New oceanic proxies for paleoclimate. Earth
Hearty, P.J., Kaufman, D.S. 2000. Whole-rock aminostratigraphy and and Planetary Science Letters 203, 1–13.
Quaternary sea-level history of the Bahamas. Quaternary Research Hendon, D., Charman, D. 2004. High-resolution peatland water-
54, 163–173 table changes for the past 200 years: the influence of climate and
Hearty, P.J., O’Leary, M.J. 2007. Carbonate eolianites, quartz sands, implications for management. The Holocene 14, 125–134.
and Quaternary sea-level cycles, Western Australia: a chrono- Hendon, D., Charman, D.J., Kent, M. 2001. Palaeohydrological
stratigraphic approach. Quaternary Geochronology 3, 26–55. records derived from testate amoebae analysis from peatlands
Hearty, P.J., Olson, S.L., Kaufman, D.S., et al. 2004. Stratigraphy and in northern England: within-site variability, between-site com-
geochronology of pitfall accumulations in caves and fissures, parability and palaeoclimatic implications. The Holocene 11,
Bermuda. Quaternary Science Reviews 23, 1151–1171. 127–148.
Hebdon, N.J., Atkinson, T.C., Lawson T.J., et al. 1997. Rate of glacial Hendy, E.J., Gagan, M.K., Lough, J.M. 2003. Chronological control
valley deepening during the late Quaternary in Assynt, Scotland. of coral records using luminescent lines and evidence for
Earth Surface Processes and Landforms 22, 307–315. non-stationary ENSO teleconnections in northeastern Australia.
Hedberg, H.D. (ed.). 1976. International Stratigraphic Guide. John The Holocene 13, 187–199.
Wiley, London and New York. Herzschuh, U. 2006. Palaeo-moisture evolution in monsoonal
Hedding, D.W., Sumner, P.D., Holness, S.D., et al. 2007. Formation Central Asia during the last 50,000 years. Quaternary Science
of a pronival rampart on sub-Antarctic Marion Island. Antarctic Reviews 25, 163–178.
Science 19, 443–450. Hess, D.P., Briner, J.P. 2009. Geospatial analysis of controls on
Hedges, R.E.M. 2001. The future of the past. Radiocarbon 43, subglacial bedform morphometry in the New York Drumlin
141–148. Field – implications. Earth Surface Processes and Landforms
Hegerl, G.C., Bindoff, N.L. 2005. Warming the world’s oceans. 34, 1126–1135.
Science 309, 254–255. Hesse, P.P., McTainsh, G.H. 2003. Australian dust deposits: modern
Heine, K., Reuther, A.U., Thieke, H.U., et al. 2009. Timing of processes and the Quaternary record. Quaternary Science Reviews
Weichselian ice marginal positions in Brandenburg (north- 22, 2007–2035.
eastern Germany) using cosmogenic in situ 10Be. Zeitschrift Hesse, P.P., Magee, J.W., van der Kars, S. 2004. Late Quaternary
für Geomorphologie N.F., 53, 433–454. climates of the Australian arid zone: a review. Quaternary
Heinemeier, J., Ringbom, Å., Lindroos, A., et al. 2010. Successful AMS International 118–119, 103–126.
14 Hessler, I., Duppont, L., Bonnefille, R., et al. 2010. Millennial-scale
C dating of non-hydraulic lime mortars from the Medieval
churches of the Åland Islands, Finland. Radiocarbon 52, 171–204. changes in vegetation records from tropical Africa and South
Heinrich, H. 1988. Origin and consequences of cyclic ice-rafting in America during the last glacial. Quaternary Science Reviews 29,
the Northeast Atlantic Ocean during the past 130,000 years. 2882–2899.
Quaternary Research 29, 142–152. Hetherington, A.M., Woodward, F.I. 2003. The role of stomata
Heinrichs, M.L., Walker, I.R. 2006. Fossil midges and palaeosalinity: in sensing and driving environmental change. Nature 424,
potential as indicators of hydrological balance and sea-level 901–908.
change. Quaternary Science Reviews 25, 1948–1965. Hetherington, R., Barrie, J.V. 2004. Interaction between local
Heiri, O., Lotter, A.F., Lemeke, G. 2001. Loss on ignition as a method tectonics and glacial unloading on the Pacific margin of Canada.
for estimating organic and carbonate content in sediments: Quaternary International 120, 65–77.
reproducibility and comparability of results. Journal of Paleo- Hewitt, G.M. 2011. Quaternary phylogeography: the roots of
limnology 25, 101–110. hybrid zones. Genetica 139, 617–638.
REFERENCES 475
Hicks, S., Tinsley, H., Huusko, A., et al. 2001. Some comments Hocknull, S.A., Zhao, J.-X., Feng, Y.-X., et al. 2007. Responses of
on spatial variation in arboreal pollen deposition: first records Quaternary rainforest vertebrates to climate change in Australia.
from the Pollen Monitoring Programme (PMP). Review of Earth and Planetary Science Letters 264, 317–331.
Palaeobotany and Palynology 117, 183–194. Hodder, K.R. 2009. Flocculation: a key process in the sediment flux
Hicock, S.R., Goff, J.R., Lian, O.B., et al. 1996. On the interpretation of a large, glacier-fed lake. Earth Surface Processes and Landforms
of subglacial till fabric. Journal of Sedimentary Research 66, 34, 1151–1163.
928–934. Hodgson, D.A., Convey, P. 2005. A 7000-year record of oribatid
Hiemstra, J.F., Evans, D.J., Scourse, J.D., et al. 2006. New evidence mite communities on a maritime-Antarctic island: responses
for a grounded Irish Sea glaciation of the Isles of Scilly. Quaternary to climate change. Arctic, Antarctic, and Alpine Research 37,
Science Reviews 25, 299–309. 239–245.
Higham, T.F.G., Hedges, R.E.M., Andersen, A.J., et al. 2004. Problems Hoek, W.Z. 1997. Late-glacial and Early Holocene climatic events
associated with the AMS dating of small bone samples: the and chronology of vegetation development in the Netherlands.
question of the arrival of Polynesian rats to New Zealand. Vegetation History and Archaeobotany 6, 197–213.
Radiocarbon 46, 207–218. Hoey, T.B., 2004. The size of sedimentary particles. In A Practical
Higham, T.F.G., Jacobi, R.M., Bronk Ramsey, C., 2006. AMS radio- Guide to the Study of Glacial Sediments (edited by D.J.A. Evans,
carbon dating of ancient bone using ultrafiltration. Radiocarbon D.I. Benn). Edward Arnold, London, 52–77.
48,179–195 Hof, C., Levinsky, I., Araújo, M.B., et al. 2011. Rethinking species’
Hilgen, F., Brinkhuis, H., Zachariasse, W.-J. 2006. Unit stratotypes ability to cope with rapid climate change. Global Change Biology
for global stages: the Neogene perspective. Earth-Science Reviews 17, 2987–2990.
74, 113–125. Hoffmann, T., Thorndycraft, V.R., Brown, A.G., et al. 2010. Human
Hill, J.K., Griffiths, H.M., Thomas, C.D. 2011. Climate change and impact on fluvial regimes and sediment flux during the Holocene:
evolutionary adaptations at species’ range margins. Annual
review and future research agenda. Global and Planetary Change
Review of Entomology 56, 143–59.
72, 87–98.
Hill, T.M., Kennett, J.P., Behl, R.J., et al. 2006. Pre-Bølling warming
Hofreiter, M. 2008. Long DNA sequences and large data sets:
in Santa Barbara Basin, California: surface and intermediate water
investigating the Quaternary via ancient DNA. Quaternary
records of early deglacial warmth. Quaternary Science Reviews 25,
Science Reviews 27, 2586–2592
2835–2845.
Hofreiter, M., Stewart, J. 2009. Ecological change, range fluctuations
Hillaire-Marcel, C., Ochietti, S. 1980. Chronology, palaeogeography
and population dynamics during the Pleistocene (Review).
and palaeoclimatic significance of the late and post-glacial events
Current Biology 19, R584–R594.
in eastern Canada. Zeitschrift für Geomorphologie 24, 373–392.
Hofreiter, M., Barnes, I. 2010. Diversity lost: are all Holarctic large
Hillaire-Marcel, C., De Vernal, A. (eds). 2007. Proxies in Late Cenozoic
Paleoceanography. Elsevier, Amsterdam. mammal species just relict populations? BMC Biology 8, 46.
Hillam, J., Groves, C.M., Brown, D.M., et al. 1990. Dendrochronology Hofreiter, M., Collins, M., Stewart, J.R. 2012. Ancient biomolecules
of the British Neolithic. Antiquity 64, 210–220. in Quaternary palaeoecology. Quaternary Science Reviews 33,
Hillebrand, C.-D., Moreton, S.G., Caburlotto, A., et al. 2008. Volcanic 1–13.
time-markers for Marine Isotope Stages 6 and 5 in the Southern Hogg, A.G., Higham, T.F.G., Lowe, D.J., et al. 2003. A wiggle-match
Ocean sediments and Antarctic ice cores: Implications for date for Polynesian settlement of New Zealand. Antiquity 77,
correlations between palaeoclimatic records. Quaternary Science 116–125.
Reviews 27, 518–540. Hogg, A.G., Fifeld, L.K., Turney, C.S.M., et al. 2006. Dating ancient
Hillebrand, C.-D., Ehrmann, W.U., Larter, R.D., et al. 2009. Clay wood by high-sensitivity liquid scintillation counting and accel-
mineral provenance of sediments in the southern Bellinghausen erator mass spectrometry – pushing the boundaries. Quaternary
Sea reveals drainage changes of the West Antarctic Ice Sheet during Geochronology 1, 241–248.
the Late Quaternary. Marine Geology 265, 1–18. Holden, P.B., Mackay, A.W., Simpson, G.L. 2008. A Bayesian palaeo-
Hillman, G., Hedges, R., Moore, A., et al. 2001. New evidence of environmental transfer function model for acidified lake. Journal
Lateglacial cereal cultivation at Abu Hureyra on the Euphrates. of Paleolimnology 39, 551–566.
The Holocene 11, 383–393. Höll, C., Karwath, B., Rühlemann, C., et al. 1999. Palaeo-
Hinnov, L.A., Schulz, M., Yiou, P. 2002. Interhemispheric space–time environmental information gained from calcareous dinoflagel-
attributes of the Dansgaard–Oeschger oscillations between 100 lates: the late Quaternary eastern and western tropical Atlantic
and 0 ka. Quaternary Science Reviews 21, 1213–1228. Ocean in comparison. Palaeogeography, Palaeoclimatology,
Hirmas, D.R., Allen, B.L. 2007. Degradation of pedogenic calcretes Palaeoecology 146, 147–164.
in West Texas. Soil Science Society of America Journal 71, Hollands, C.B., Nanson, G.C., Jones, B.G., et al. 2006. Aeolian–fluvial
1878–1888. interaction: evidence for Late Quaternary channel change and
Hiscock, K., Southward, A., Tittley, I., et al. 2004. Effects of changing wind-rift linear dune formation in the northwestern Simpson
temperature on benthic marine life in Britain and Ireland. Desert, Australia. Quaternary Science Reviews 25, 142–162.
Aquatic Conservation: Marine and Freshwater Ecosystems 14, Holliday, V.T. 2004. Soils in Archaeological Research. Cambridge
333–362. University Press, Oxford and New York.
476 REFERENCES
Holmes, J.A. 2002. Ostracoda. In Tracking Environmental Change Horsák, M., Chitrý, M., Pokryszko, B.M., et al. 2010. Habitats of
Using Lake Sediments. Volume 4: Zoological Indicators (edited by relict terrestrial snails in southern Siberia: lessons for the
J.P. Smol, H.J.B. Birks, W.M. Last). Springer, Berlin, 125–151. reconstruction of palaeoenvironments of full-glacial Europe.
Holmes, J.A., Chivas, A. 2002. The Ostracoda: Applications in Journal of Biogeography 37, 1450–1462.
Quaternary Research. Geophysical Monograph 131. American Horton, B., Culver, S.J., Hardbattle, M.I.J., et al. 2007. Reconstructing
Geophysical Union, Washington, DC. Holocene sea-level change for the central Great Barrier Reef
Holmes, J.A., Atkinson, T., Darbyshire, D.B.F., et al. 2010. Middle (Australia) using subtidal Foraminifera. Journal of Foraminiferal
Pleistocene climate and hydrological environment at the Box- Research 37, 327–343.
grove hominin site (West Sussex, UK) from ostracod records. Houmark-Nielsen, M. 1989. The last interglacial–glacial cycle in
Quaternary Science Reviews 29, 1515–1527. Denmark. Quaternary International 3/4, 31–39.
Holmgren, K., Tyson, P.D., Moberg, A., et al. 2001. A preliminary Howard, L.C., Wood, P.J., Greenwood, M., et al. 2010. Sub-fossil
3000-year regional temperature reconstruction for South Africa. Chironomidae as indicators of palaeoflow regimes: integration
South African Journal of Science 97, 49–51. into the PalaeoLIFE flow index. Journal of Quaternary Science 25,
Holmgren, K., Lee-Thorp, J.A., Cooper, G.R.J., et al. 2003. Persist- 1270–1283.
ent millennial-scale climatic variability over the past 25,000 Howarth, J.D., Fitzsimons, S.J., Jacobsen, G.E., et al. 2013. Identifying
years in Southern Africa. Quaternary Science Reviews 22, a reliable target fraction for radiocarbon dating sedimentary
2311–2326. records from lakes. Quaternary Geochronology 17, 68–80.
Hölzel, N., Otte, A. 2004. Inter-annual variation in the soil seed Howell, D., Siegert, M.J., Dowdeswell, J.A. 2000. Modelling the
bank of flood-meadows over two years with different flooding influence of glacial isostasy on Late Weichselian ice-sheet growth
patterns. Plant Ecology 174, 279–291. in the Barents Sea. Journal of Quaternary Science 15, 475–486.
Hölzer, A., Hölzer, A. 1998. Silicon and titanium in peat profiles as Hu, F.S., Kaufman, D., Yoneji, S., et al. 2003. Cyclic variation and
indicators of human impact. The Holocene 8, 685–696. solar forcing of Holocene climate in the Alaskan Subarctic.
Holzhauser, H., Magny, M., Zumbühl, H.J. 2005. Glacier and lake- Science 301, 1890–1893.
level variations in west-central Europe over the last 3500 years. Hua, Q., Barbetti, M., Fink, D., et al. 2009. Atmospheric 14C variations
The Holocene 15, 789–801. derived from tree rings during the early Younger Dryas.
Hongve, D. 2003. Chemical stratigraphy of recent sediments from Quaternary Science Reviews 28, 2982–2990.
a depth gradient in a meromictic lake, Nordbytjernet, SE Norway, Huang, C.C., Pang, J., Zhao, J. 2000. Chinese loess and the evolution
in relation to variable external loading and sedimentary fluxes. of the east Asian monsoon. Progress in Physical Geography 24,
Journal of Paleolimnology 30, 75–93. 75–96.
Hoogakker, B.A.A., McCave, I.N., Vautravers, M.J. 2007. Antarctic Huang, E., Tian, J. 2012. Sea-level rises at Heinrich stadials of early
link to deep flow speed variation during Marine Isotope Stage 3 Marine Isotope Stage 3: evidence of terrigenous n-alkane input
in the western North Atlantic. Earth and Planetary Science Letters in the southern South China Sea. Global and Planetary Change
257, 463–473. 94–95, 1–12.
Hooghiemstra, H., van Geel, B. 1998. World list of Quaternary Hubbard, B., Glasser, N. 2005. Field Techniques in Glaciology and
pollen and spore atlases. Review of Palaeobotany and Palynology Glacial Geomorphology. John Wiley, Chichester.
104, 157–182. Hubeny, J.B., King, J.W., Santos, A. 2006. Subdecadal to multidecadal
Hooghiemstra, H., van der Hammen, T. 2004. Quaternary ice-age cycles of Late Holocene North Atlantic climate variability
dynamics in the Colombian Andes: developing an understanding preserved by estuarine fossil pigments. Geology 34, 569–572.
of our legacy. Philosophical Transactions of the Royal Society of Huggett, R.J. 1998. Soil chronosequences, soil development, and soil
London B 359, 173–181. evolution: a critical review. Catena 32, 155–172.
Hooke, J.M. 2006. Human impacts on fluvial systems in the Hughen, K., Lehman, S., Southon, J., et al. 2004a. 14C activity and
Mediterranean region. Geomorphology 79, 311–335. global carbon cycle changes over the past 50,000 years. Science
Hooyer, T.S., Iverson, N.R. 2002. Flow mechanism of the Des 303, 202–207.
Moines lobe of the Laurentide ice sheet. Journal of Glaciology Hughen, K.A., Eglinton, T.I., Xiu, L., et al. 2004b. Abrupt tropical
48, 575–586 vegetation response to rapid climatic changes. Nature 304,
Horne, D.J., Holmes, J.A., Rodriguez-Lazaro, J., et al. (eds). 2012. 1955–1959.
Developments in Quaternary Science. Volume 17: Ostracoda as Hughen, K., Southon, J., Lehman, S., et al. 2006. Marine-derived
Proxies for Quaternary Climate Change. Elsevier, Amsterdam. C-14 calibration and activity record for the past 50,000 years
Horrocks, M., Jones, M., Scott, N., et al. 2002. Wetland microfossils updated from the Cariaco Basin. Quaternary Science Reviews 25,
in soil: implications for the study of land use on archaeological 3216–3227.
landscapes. Environmental Archaeology 7, 101–106. Hughes, G.R., Kennedy, D.M. 2009. Late Pleistocene sea-level
Horsák, M. 2005. Mollusc community patterns and species response oscillations (MIS 10-2) recorded in shallow marine and coastal
curves along a mineral richness gradient: a case study in fens. plain sediments of the southern Wanganui Basin, New Zealand.
Journal of Biogeography 33, 98–107. Quaternary Research 71, 477–489.
Horsák, M. 2011. Mollusc assemblages in palaeoecological recon- Hughes, M.K., Swetnam, T.W., Diaz, H.F. (eds). 2011. Dendro-
structions: an investigation of their predictive power using climatology: Progress and Prospects. Developments in Palaeo-
transfer function models. Boreas 40, 459–467. environmental Research, 11. Springer, Berlin.
REFERENCES 477
Hughes, P.D. 2010. Geomorphology and Quaternary stratigraphy: Hurford, A.J. 1991. Fission Track Dating. In Quaternary Dating
the roles of morpho-, litho-, and allostratigraphy. Geomorphology Methods – A User’s Guide (edited by P.L. Smart, P.D. Frances).
123, 189–199. Technical Guide 4. Quaternary Research Association,
Hughes, P.D.M., Mauquoy, D., Barber, K.E., et al. 2000. Mire- Cambridge, 84–107.
development pathways and palaeoclimatic records from a full Hurrell, J., Deser, C. 2009. North Atlantic climate variability: the
Holocene peat archive at Walton Moss, Cumbria, England. The role of the North Atlantic Oscillation. Journal of Marine Science
Holocene 10, 465–479. 78, 28–41.
Hughes, P.D., Gibbard, P.L., Woodward, J.C. 2003. Relict rock Hutchinson, I., James, T.S., Clague, J.J., et al. 2004. Reconstruction
glaciers as indicators of Mediterranean palaeoclimate during the of late Quaternary sea-level change in southwestern British
Last Glacial Maximum (Late Würmian) in northwest Greece. Columbia from sediments in isolation basins. Boreas 33, 183–194.
Journal of Quaternary Science 18, 431–440. Hutri, K.-L., Kotilainen, A. 2007. An acoustic view into Holocene
Hughes, P.D., Gibbard, P.L., Ehlers, J. 2013. Timing of glaciation palaeoseismicity offshore southwestern Finland, Baltic Sea.
during the last glacial cycle: evaluating the concept of a global ‘Last Marine Geology 238, 45–59.
Glacial Maximum’ (LGM). Earth-Science Reviews 125, 171–198. Huybers, P. 2007. Glacial variability over the last two million years:
Huguet, C., Kim, J.-H., Damsté, J.S.S., et al. 2006. Reconstruction an extended depth-derived age model, continuous obliquity
of sea surface temperature variations in the Arabian Sea over pacing, and the Pleistocene progression. Quaternary Science
the last 23 kyr using organic proxies (TEX86 and U37K′). Reviews 26, 37–55.
Paleoceanography 21, doi: 10.1029/2005PA001215. Huybers, P. 2009. Pleistocene glacial varibility as a chaotic response
Huijzer, A.S., Isarin, R.F.B. 1997. The reconstruction of past climates to obliquity forcing. Climate of the Past 5, 481–488.
using multi-proxy evidence: an example of the Weichselian Huybers, P., Wunsch, C. 2005. Obliquity pacing of the late Pleistocene
Pleniglacial in northwestern and central Europe. Quaternary terminations. Nature 434, 491–494.
Huybrechts, P. 2006. Numerical modeling of ice sheets through
time. In Glacier Science and Environmental Change (edited by
Huijzer, B., Vandenberghe, J. 1998. Climatic reconstruction of the
P.G. Knight). Blackwell Publishing, Oxford, 406–412.
Weichselian Pleniglacial in northwestern and central Europe.
Ikehara, K., Itaki, T. 2007. Millennial-scale fluctuations in seasonal
Journal of Quaternary Science 13, 391–417.
sea-ice and deep-water formation in the Japan Sea during the late
Hulbe, C.L., MacAyeal, D.R., Denton, G.H., et al. 2004. Catastrophic
Quaternary. Palaeogeography, Palaeoclimatology, Palaeoecology
ice shelf breakup as the source of Heinrich event icebergs.
247, 131–143.
Paleoceanography 19, PA1004, doi: 10.1029/2003PA000890.
Iliyashuk, B., Gobet, E., Heiri, O., et al. 2009. Lateglacial environ-
Hull, K.L. 2001. Reasserting the utility of obsidian hydration
mental changes at the Maloja Pass, Central Swiss Alps, as recorded
dating: a temperature-dependent empirical approach to practical
by chironomids and pollen. Quaternary Science Reviews 28,
temporal resolution with archaeological obsidians. Journal of 1340–1353.
Archaeological Science 28, 1025–1040. Imbrie, J., Kipp, N.G. 1971. A new micropalaeontological method
Hüls, C.M., Grootes, P.M., Nadeau, M.-J., et al. 2004. AMS of quantitative palaeoclimatology: application to a Late Pleisto-
radiocarbon dating of iron artefacts. Nuclear Instruments and cene Caribbean core. In The Late Cenozoic Glacial Ages (edited
Methods in Physics Research B 223–224, 709–715. by K.K. Turekian). Yale University Press, New Haven, 71–181.
Humlum, O., Christiansen, H.H. 2008. Lowland periglacial research: Imbrie, J., Imbrie, K.P. 1979. Ice Ages: Solving the Mystery. Macmillan,
a review of published advances 2003–2007. Permafrost and London.
Periglacial Processes 19, 211–235. Imbrie, J., Hays, J.D., Martinson, D.G., et al. 1984. The orbital theory
Humlum, O., Christiansen, H.H., Juliussen, H. 2007. Avalanche- of Pleistocene climate: support from a revised chronology of the
derived rock glaciers in Svalbard. Permafrost and Periglacial marine δ18O record. Terminations. In Milanovitch and Climate
Processes 18, 75–88. (edited by A. Berger, J. Imbrie, J. Hays, et al.). Reidel, Dordrecht,
Hüneke, H., Mulder, T. (eds). 2011. Deep-Sea Sediments. Elsevier, 269–306.
Amsterdam. Imbrie, J., Berger, A., Shackleton, N.J. 1993. Role of orbital forc-
Hunt, C.O. 1993. Mollusc taphonomy in caves: a conceptual model. ing: a two-million year perspective. In Global Changes in the
Cave Science 20, 45–49. Perspective of the Past (edited by J.A. Eddy, H. Oeschger). John
Hunt, C.O. 2005. Pollen taphonomy and airfall sedimentation in a Wiley, Chichester and New York, 263–277.
tropical cave: the West Mouth of The Great Cave of Niah in Ingólfsson, O. 2004. Quaternary glacial and climatic history of
Sarawak, Malaysian Borneo. Journal of Archaeological Science 32, Antarctica. In Developments in Quaternary Science. Volume 2:
465–473. Quaternary Glaciations: Extent and Chronology (edited by
Huntley, B. 1992. Rates of change in the European palynological J. Ehlers, P.L. Gibbard). Elsevier, Amsterdam, 3–43.
record of the last 13,000 years and their climatic interpretation. Ingólfsson, O., Landvik, J.Y. 2013. The Svalbard–Barents Sea ice
Climate Dynamics 6, 185–191. sheet – historical, current and future perspectives. Quaternary
Huntley D.J., Lamothe, M. 2001. Ubiquity of anomalous fading in Science Reviews 64, 33–60.
K-feldspars, and the measurement and correction for it in optical Ingólfsson, O., Björck, S., Haflidason, H., et al. 1997. Glacial and
dating. Canadian Journal of Earth Sciences 38, 1093–1106. climatic events in Iceland reflecting regional North Atlantic
478 REFERENCES
climatic shifts during the Pleistocene–Holocene transition. Izett, G.A., Naeser, C.W. 1976. Age of the Bishop Tuff of eastern
Quaternary Science Reviews 16, 1135–1144. California as determined by the fission-track method. Geology
Innes, J.L. 2006. Lichenometric dating of debris-flow deposits in the 4, 587–590.
Scottish Highlands. Earth Surface Processes and Landforms 8, Jackson, A., Jonkers, A.R.T., Walker, M.R. 2000. Four centuries of
579–588. geomagnetic secular variation from historical records. Philo-
Intergovernmental Panel on Climate Change. 2007. Climate Change sophical Transactions of the Royal Society of London A 358,
2007: Synthesis Report. Contribution of Working Groups I, II and 957–990.
III to the Fourth Assessment Report of the Intergovernmental Jackson, L.E., Phillips, F.M., Shimamura, K., et al. 1997. Cosmogenic
Panel on Climate Change. IPCC, Geneva. 36
Cl dating of the Foothills erratics train, Alberta, Canada. Geology
Intergovernmental Panel on Climate Change. 2013. Climate Change 25, 195–198.
2013: The Physical Science Basis. Contribution of Working Jackson, S.T., Williams, J.W. 2004. Modern analogs in Quaternary
Group I to the Fifth Assessment Report of the Intergovernmental paleoecology: here today, gone yesterday, gone tomorrow?
Panel on Climate Change (edited by T.F. Stocker, D. Qin, Annual Review of Earth and Planetary Sciences 32, 495–537.
G.-K. Plattner, et al.). Cambridge University Press, Cambridge Jackson, S.T., Betancourt, J.L., Lyford, M.E., et al. 2005. A 40,000-
and New York. year woodrat midden record of vegetational and biogeographical
International Atomic Energy Agency. 2009. Reference Sheet for dynamics in north-eastern Utah, USA. Journal of Biogeography
International Measurement Standards VSMOW2, SLAP2, Water, 32, 1085–1106.
http://nucleus.iaea.org/rpst/ReferenceProducts. Jacob, D., Bärring, L., Christensen, O.B., et al. 2007. An inter-
Isarin, R.F.B. 1997. Permafrost distribution and temperatures in comparison of regional climate models for Europe: model per-
Europe during the Younger Dryas. Permafrost and Periglacial formance in present-day climate. Climatic Change 81, 31–52.
Processes 8, 313–333. Jacobs, P.M., Mason, J.D. 2007. Late Quaternary climate change,
Isarin, R.F.B., Bohncke, S.J.P. 1999. Mean July temperatures during loess sedimentation, and soil profile development in the central
the Younger Dryas in Northwestern and Central Europe as Great Plains: a pedosedimentary model. Geological Society of
inferred from climate indicator plant species. Quaternary America Bulletin 119, 462–475.
Research 51, 158–173. Jacobs, P.M., Konen, M.E., Curry, B.B. 2009. Pedogenesis of a
Isarin, R.F.B., Renssen, H. 1999. Reconstructing and modelling catena of the Farmdale–Sangamon Geosol complex in the north
Late Weichselian climates: the Younger Dryas in Europe as a case central United States. Palaeogeography, Palaeoclimatology, Palaeo-
study. Earth-Science Reviews 48, 1–38. ecology 282, 119–132.
Isarin, R.F.B., Renssen, H., Koster, E.A. 1997. Surface wind climate Jacomet, S. 2007. Plant macrofossil methods and studies: use in
during the Younger Dryas in Europe as inferred from aeolian environmental archaeology. In Encyclopedia of Quaternary
records and model simulations. Palaeogeography, Palaeoclimat- Science (edited by S.A. Elias). Elsevier, Amsterdam, 2384–2412.
ology, Palaeoecology 134, 127–148. James, R.H., Austin, W.E.N. 2008. Biogeochemical Controls on
Isarin, R.F.B., Renssen, H., Vandenberghe, J. 1998. The impact of the Palaeoceanographic Environmental Proxies: A Review. Geological
North Atlantic Ocean on the Younger Dryas climate in north Society of London, Special Publication 303. Geological Society,
western and central Europe. Journal of Quaternary Science 13, London, 3–32.
447–453. James, T.S., Clague, J.J., Wang, K., et al. 2000. Postglacial rebound
Itaki, T., Komatsu, N., Motoyama, I. 2007. Orbital- and millennial- at the northern Cascadia subduction zone. Quaternary Science
scale changes of radiolarian assemblages during the last 220 kyrs Reviews 19, 1527–1541.
in the Japan Sea. Palaeogeography, Palaeoclimatology, Palaeo- Janke, J., Frauenfelder, R. 2008. The relationship between rock
ecology 247, 115–130. glacier and contributing area parameters in the Front Range of
Iturrizaga, L. 2008. Paraglacial landform assemblages in the Colorado. Journal of Quaternary Science 23, 153–163.
Hindukush and Karakoram Mountains. Geomorphology 95, Jans, M.M.E., Kars, H., Nielsen-Marsh, C.M., et al. 2002. In situ
27–47. preservation of archaeological bone: a histological study within
Ivanova, E.V., Beaufort, L., Vidal, L., et al. 2012. Precession forcing a multidisciplinary approach. Archaeometry 44, 343–352.
of productivity in the Eastern Equatorial Pacific during the last Jarman, D., Wilson, P., Harrison, S. 2013. Are there any relict rock
glacial cycle. Quaternary Science Reviews 40, 64–77. glaciers in the British mountains? Journal of Quaternary Science
Ivanovich, M., Harmon, R.S. (eds). 1995. Uranium-series Dis- 28, 131–143.
equilibrium: Applications to Earth, Marine and Environmental Jass, C.N., George, C.O. 2010. An assessment of the contribution of
Sciences, 2nd edn. Clarendon Press, Oxford. fossil cave deposits to the Quaternary paleontological record.
Ivy-Ochs, S., Kober, F. 2008. Surface exposure dating with Quaternary International 217, 105–116.
cosmogenic nuclides. Eiszeitalter und Gegenwart (Quaternary Jedoui, Y., Reyss, J.-L., Kallel, N., et al. 2003. U-series evidence for
Science Journal) 57, 179–209. two high Late Interglacial sea levels in southeastern Tunisia.
Ivy-Ochs, S., Poschinger, A.V., Syndal, H.-A., et al. 2009. Surface Quaternary Science Reviews 22, 343–351.
exposure dating of the Flims landslide, Grabünden, Switzerland. Jellinek, A.M., Manga, M., Saar, M.O. 2004. Did melting glaciers cause
Geomorphology 103, 104–112. volcanic eruptions in eastern California? Probing the mechanics
REFERENCES 479
of dike formation. Journal of Geophysical Research 109, B09206, Johnsson, K. 1997. Chemical dating of bones based on diagenetic
doi: 10.1029/2004JB002978. changes in bone apatite. Journal of Archaeological Science 24,
Jenkins, D.L., Davis, L.G., Stafford Jr., T.W, et al. 2012. Clovis age 431–437.
western stemmed projectile points and human coprolites at the Johnston, P., Lambeck, K. 2002. Postglacial rebound and sea level
Paisely Caves. Science 337, 223–228. contributions to changes in the geoid and the Earth’s rotation
Jennings, A.E., Tedesco, K.A., Andrews, J.T., et al. 1996. Shelf Erosion axis. Geophysical Journal International 136, 537–555.
and Glacial Ice Proximity in the Labrador Sea During and After Jones, A.P., Tucker, M.E., Hart, J.K. 1999. The Description and
Heinrich Events (H-3 or 4 to H-0) as Shown by Foraminifera. Analysis of Quaternary Stratigraphic Field Sections. Technical
Geological Society of London, Special Publication 111. Blackwell Guide 7. Quaternary Research Association, London
Scientific Publications, Oxford, 29–49. Jones, C. 2003. A Fast Ocean GCM without flux adjustments. Journal
Jennings, C. 2004. The status of glacial mapping in Minnesota. In of Atmospheric and Oceanic Technology 20, 1857–1868.
Developments in Quaternary Science. Volume 2: Quaternary Jones, J., Tinsley, H., Brunning, R. 2007a. Methodologies for assess-
Glaciations: Extent and Chronology (edited by J. Ehlers, P.L. ment of the state of preservation of pollen and plant macrofossil
Gibbard). Elsevier, Amsterdam, 119–124. remains in waterlogged deposits. Environmental Archaeology 12,
Jennings, C.E., Johnson, M.D. 2011. The Quaternary of Minnesota. 71–86.
In Developments in Quaternary Science. Volume 15: Quaternary Jones, M.D., Roberts, C.N., Leng, M.J., et al. 2006. A high-resolution
Glaciations – Extent and Chronology: A Closer Look (edited by J. late Holocene lake isotope record from Turkey and links to
Ehlers, P.L. Gibbard, P.D. Hughes). Elsevier, Amsterdam, North Atlantic and monsoon climate. Geology 34, 361–364.
499–511. Jones, M.D., Roberts, C.N., Leng, M.J. 2007b. Quantifying climatic
Jensen, J.R. 2007. Remote Sensing of the Environment: An Earth change through the last glacial–interglacial transition based
Resource Perspective, 2nd edn. Prentice Hall, Upper Saddle River, on lake isotope palaeohydrology from central Turkey. Quaternary
NJ. Science Reviews 67, 463–473.
Jewell, P.W. 2007. Morphology and paleoclimatic significance of Jones, P.D., Mann, M.E. 2004. Climate over past millennia. Review
Pleistocene Lake Bonneville spits. Quaternary Research 68, of Geophysics 42, doi: 10.129/2003RG000143.
421–430. Jones, P.D., Briffa, K.R., Osborn, T.J., et al. 2009. High-resolution
Jex, C.N., Baker, A., Fairchild, I.J., et al. 2010. Calibration of palaeoclimatology of the last millennium: a review of current
speleothem δ18O with instrumental climate records from Turkey. status and future prospects. The Holocene 19, 3–49.
Global and Planetary Change 71, 207–217. Jones, P.D., Lister, D.H., Osborn, T.J., et al. 2012. Hemispheric and
Jex, C.N., Baker, A., Eden, J.M., et al. 2011. A 500 yr speleothem- large-scale land-surface air temperature variations: an extensive
derived reconstruction of late autumn–winter precipitation, revision and an update to 2010. Journal of Geophysical Research
northeast Turkey. Quaternary Research 75, 399–405. 117, D05 127, doi: 10.1029/2011JD017139.
Jiménez-Espejo, F., Martínez-Ruiz, F., Finlayson, C., et al. 2007. Jones, R.T., Marshall, J.D., Crowley, S.F., et al. 2002. A high resolution
Climate forcing and Neanderthal extinction in Southern Iberia: multiproxy late-glacial record of climate change and intrasystem
insights from a multi-proxy marine record. Quaternary Science responses in NW England. Journal of Quaternary Science 17,
Reviews 26, 836–852. 329–340.
Jin, Z., Bickle, M., Chapman, H., et al. 2006. An experimental Jones, V. 2007. Diatom introduction. In Encyclopedia of Quaternary
evaluation of cleaning methods for fossil ostracod Mg/Ca and Science (edited by S.A. Elias). Elsevier, Amsterdam, 476–484.
Sr/Ca determination. Journal of Paleolimnology 36, 211–218. Joos, F., Gerber, S., Prentice, I.C., et al. 2004. Transient simulations
Johnsen, S.J., Dahl-Jensen, D., Dansgaard, W., et al. 1995. Greenland of Holocene atmospheric carbon dioxide and terrestrial carbon
palaeotemperatures derived from GRIP bore hole temperature since the Last Glacial Maximum. Global Biogeochemical Cycles 18,
and ice core isotope profiles. Tellus 47B, 624–629. GB2002, doi: 10.1029/2003GB002156.
Johnsen, S., Dahl-Jensen, D., Gundestrup, N., et al. 2001. Oxygen Jordan, B.R., Sigurdsson, H., Carey, S.N., et al. 2006. Geochemical
isotope and palaeotemperature records from six Greenland correlation of Caribbean Sea tephras with ignimbrites in central
ice-core stations: Camp Century, Dye-3, GRIP, GISP2, Renland America. Geological Society of America, Special Paper 402,
and NorthGRIP. Journal of Quaternary Science 16, 299–308. 175–208.
Johnson, A. 1999. The Ancient Bristlecone Pine Forest. Community Josephs, R.L. 2010. Micromorphology of an early Holocene loess–
Printing and Publishing, Bishop, CA. paleosol sequence, central Alaska, USA. Arctic, Antarctic and
Johnson, M.D., Mickelson, D.M., Clayton, L., et al. 1995. Composi- Alpine Research 42, 67–75.
tion and genesis of glacial hummocks, western Wisconsin, USA. Jouzel, J., Alley, R.B., Cuffey, K.M., et al. 1997. Validity of the
Boreas 24, 97–116. temperature reconstruction from water isotopes in ice cores.
Johnson, M.D., Jennings, C., Hobbs, H. 2009. Lithostratigraphy for Journal of Geophysical Research 102, 26471–26487.
Quaternary glacial deposits: ‘If it ain’t broke, don’t fix it!’. GSA Jouzel, J., Masson-Delmotte, V., Cattani, O., et al. 2007. Orbital and
Today 19, 15–16. millennial Antarctic climate variability over the past 800,000 years.
Johnson, M.R., Smith, A.M. 1997. Seabed topography under the Science 317, 793–796.
southern and western Ronne Ice Shelf, derived from seismic Juggins, S., Cameron, N.G. 2010. Diatoms and archeology. In The
surveys. Antarctic Science 9, 201–208. Diatoms: Applications for the Environmental and Earth Sciences,
480 REFERENCES
2nd edn (edited by J.P. Smol, E.F. Stoermer). Cambridge Kaspari, S.D.S., Schwikowski, M., Gysel, M., et al. 2011. Recent
University Press, Cambridge, 514–522. increases in black carbon concentrations from a Mt Everest ice
Juyal, N., Chamyal, L.S., Bhandari, S., et al. 2006. Continental record core spanning 1860–2000 AD. Geophysical Research Letters 38, doi:
of the southwest monsoon during the last 130 ka: evidence from 10.1029/2010GL046096.
the southern margin of the Thar Desert, India. Quaternary Kasse C., Vandenberghe J., De Corte F., et al. 2007. Late Weichselian
Science Reviews 25, 2632–2650. fluvio-aeolian sands and coversands of the type locality Grubben-
Kageyama, M., Paul, A., Roche, D.M., et al. 2010. Modelling glacial vorst (southern Netherlands): sedimentary environments,
climatic millennial-scale variability related to changes in the climate record and age. Journal of Quaternary Science 22, 695–708.
Atlantic meridional overturning circulation: a review. Quaternary Kattel, G.R., Battarbee, R.W., Mackay, A.W., et al. 2008. Recent
Science Reviews 29, 2931–2956. ecological change in a remote Scottish mountain loch: an
Kaiser, K., Hilgers, A., Schlaak, N., et al. 2009. Palaeopedological evaluation of a Cladocera-based temperature transfer-function.
marker horizons in northern central Europe: characteristics of Palaeogeography, Palaeoclimatology, Palaeoecology 259, 51–76.
Lateglacial Usselo and Finow soils. Boreas 38, 591–609. Katz, M.E., Cramer, B.S., Franzese, A., et al. 2010. Traditional and
Kaiser, K.F., Friedrich, M.N., Miramont, C., et al. 2012. Challenging emerging geochemical proxies in Foraminifera. Journal of
process to make Lateglacial tree-ring chronologies from Europe Foraminiferal Research 40, 165–192.
absolute – an inventory. Quaternary Science Reviews 36, 78–90. Kaufman, D. 2003a. Dating deep-lake sediments by using amino acid
Kale, V.S. 2007. Fluvio–sedimentary response of the monsoon-fed racemization in fossil ostracodes. Geology 31, 1049–1052.
Indian rivers to Late Pleistocene–Holocene changes in monsoon Kaufman, D.S. 2003b. Amino acid paleothermometry of Quaternary
strength: reconstruction based on existing 14C dates. Quaternary ostracodes from the Bonneville Basin, Utah. Quaternary Science
Science Reviews 26, 1610–1620. Reviews 22, 899–914.
Kalis, A.J., van der Knaap, W.O., Schweizer, A., et al. 2006. A three Kaufman, D.S., Manley, W.F. 1998. A new procedure for determining
thousand year succession of plant communities on a valley DL amino acid ratios in fossils using reverse phase liquid
chromotography. Quaternary Science Reviews 17, 987–2000.
bottom in the Vosges Mountains, NE France, reconstructed
Kaufman, D., Schneider, D., McKay, N., et al. 2009. Recent warming
from fossil pollen, plant macrofossils, and modern phyto-
reverses long-term arctic cooling. Science 325, 1236–1239.
sociological communities. Vegetation History and Archaeobotany
Kaufmann, P., Fundel, F., Fischer, H., et al. 2010. Ammonium and
15, 377–390.
non-sea salt sulfate in the EPICA ice cores as indicator of
Kamenik, C., Schmidt, R. 2005. Chrysophyte resting stages: a
biological activity in the Southern Ocean. Quaternary Science
tool for reconstructing winter/spring climate from Alpine lake
Reviews 29, 313–323.
sediments. Boreas 34, 477–489.
Kawamura, K., Parrenin, F., Lisiecki, L., et al. 2007. Northern
Kandiano, E.S., Bauch, H.A., Kirsten, F., et al. 2012. The meridional
Hemisphere forcing of climatic cycles in Antarctica over the past
temperature gradient in the eastern North Atlantic during MIS
360,000 years. Nature 448, 912–916.
11 and its link to the ocean–atmosphere system. Palaeogeography, Kearey, P., Brooks, M., Hill, I. 2002. An Introduction to Geophysical
Palaeoclimatology, Palaeoecology 333/334, 24–39. Exploration. Blackwell Publishing, Oxford.
Karali, L., Mavridis, F., Kjormazopolou, L. 2005. Cultural landscapes Kebede, S., Travi, Y., Alemayehu, T., et al. 2006. Water balance of
during the Late and Final Neolithic of the Aegean. A case study Lake Tana and its sensitivity to fluctuations in rainfall, Blue Nile
from Leontari Cave, Mt Hymettos, Athens, Greece. Antiquity 79, basin, Ethiopia. Journal of Hydrology 316, 233–247.
303. Keen, D.H. 2001. Towards a late Middle Pleistocene non-marine
Karg, S., Märkle, T. 2002. Continuity and changes in plant resources molluscan biostratigraphy for the British Isles. Quaternary Science
during the Neolithic period in western Switzerland. Vegetation Reviews 20, 1657–1665.
History and Archaeobotany 11, 169–176. Kehl, M., Frechen, M., Skowronek, A. 2005. Paleosols derived from
Karner, D.B., Levine, J., Medeiros, B.P. 2002. Constructing a stacked loess and loess-like sediments in the Basin of Persepolis, Southern
benthic δ18O record. Paleoceanography 17, 1030, doi: 10.1029/ Iran. Quaternary International 140/141, 135–149.
2001PA000667. Kehrwald, N.M., McCoy, W.D., Thiebeault, J., et al. 2010.
Karney, G.B., Butler, P.G., Scourse, J.D., et al. 2011. Identification Paleoclimatic implications of the spatial patterns of modern
of growth increments in the shell of the bivalve mollusc Arctica and LGM European land-snail shell δ18O. Quaternary Research
islandica using backscattered electron imaging. Journal of 74, 166–176.
Microscopy 241, 29–36. Kelley, S. 2002a. Excess argon in K–Ar and Ar–Ar geochronology.
Kars, R.H., Wallinga, J., Cohen, K. 2008. A new approach towards Chemical Geology 188, 1–22.
anomalous fading corection for feldspar IRSL dating – tests on Kelley, S. 2002b. K–Ar and Ar–Ar dating. Reviews in Mineralogy and
samples in field saturation. Radiation Measurements 43, 786–790. Geochemistry 47, 785–818.
Karte, J. 1983. Periglacial phenomena and their significance as Kellogg, T.B. 1976. Late Quaternary climatic changes: evidence from
climatic and edaphic indicators. GeoJournal 7, 329–340. deep-sea cores of Norwegian and Greenland seas. In Investigation
Kashiwaya, K., Ochiai, S., Sakai, H., et al. 2001. Orbit-related long- of Late Quaternary Paleoceanography and Palaeoclimatology
term climate cycles revealed in a 12-Myr continental record (edited by R.M. Cline, J.D. Hays). Geological Society of America
from Lake Baikal. Nature 410, 71–74. Memoir 145, 77–110.
REFERENCES 481
Kelly, M., Kubik, P.W., von Blanckenburg, F., et al. 2004. Surface Kenward, H., Carrott, J. 2006. Insect species associations characterise
exposure dating of the Great Aletsch Egesen moraine system, past occupation sites. Journal of Archaeological Science 33,
western Swiss Alps, using the cosmogenic nuclide 10Be. Journal 1452–1473.
of Quaternary Science 19, 431–442. Kenward, H., Hall, A.R., Jones, A.K.G. 2012. Turf roofs and urban
Kemp, A.C., Horton, B.P., Corbett, D.R., et al. 2009. The relative utility archaeological build-up. Environmental Archaeology 17, 66–79.
of foraminifera and diatoms for reconstructing late Holocene sea- Kerney, M. 1999. Atlas of Land and Freshwater Molluscs of Britain and
level change in North Carolina, USA. Quaternary Research 71, Ireland. Harley Books, Colchester.
9–21. Kerschner, H., Kaser, G., Sailer, R. 2000. Alpine Younger Dryas
Kemp, A.C., Sommerfield, C.K., Vane, C.H., et al. 2012b. Use of lead glaciers as palaeo-precipitation gauges. Annals of Glaciology 31,
isotopes for developing chronologies in recent salt-marsh 80–84.
sediments. Quaternary Geochronology 12, 40–49. Khan, K., Khan, S.A., Krasgaard, N.J., et al. 2012. Aerial photographs
Kemp, A.E.S. 2003. Evidence for abrupt climatic changes in annually reveal late-20th century dynamic ice loss in northwestern
laminated marine sediments. Philosophical Transactions of the Greenland. Science 337, 586–573.
Royal Society of London, London A 361, 1851–1870. Khim, B.K., Ikehara, K., Irino, T. 2012. Orbital- and millennial-scale
Kemp, J., Radke, L.C., Olley, J., et al. 2012a. Holocene lake salinity paleoceanographic changes in the north-eastern Japan basin, East
changes in the Wimmera, southeastern Australia, provide Sea/Japan Sea during the late Quaternary. Journal of Quaternary
evidence for millennial-scale climate variability. Quaternary Science 27, 328–335.
Research 77, 65–76. Khursevich, G.K., Karabanov, E.B., Prokopenko, A.A., et al. 2001.
Kemp, R.A. 1987. The interpretation and environmental significance Insolation regime in Siberia as a major factor controlling diatom
of a buried Middle Pleistocene soil near Ipswich Airport, Suffolk, production in Lake Baikal during the past 800,000 years.
England, Philosophical Transactions of the Royal Society of London Quaternary International 80/81, 47–58.
B 317, 365–391.
Kidwell, S.M. 2002. Mesh-size effects on the ecological fidelity of
Kemp, R.A. 1998. Role of micromorphology in paleopedological
death assemblages: a meta-analysis of molluscan live–dead
research. Quaternary International 51–52, 133–141.
studies. Geobios Memoire Special 24, 107–119.
Kemp, R.A. 2001. Pedogenic modification of loess: significance
Kilfeather, A.A., Ó Cofaigh, C., Dowdeswell, J.A., et al. 2010.
for palaeoclimatic reconstructions. Earth-Science Reviews 54,
Micromorphological characteristics of glacimarine deposits:
145–156.
implications for distinguishing genetic processes of massive
Kemp, R.A., Derbyshire, E. 1998. The loess soils of China as records
diamicts. Geo-marine Letters 30, 77–97.
of climate change. European Journal of Soil Science 49, 525–539.
Kim, J., Park, C., Kim, T.-H., et al. 2003. Effects of various
Kemp, R.A., Whiteman, C.A., Rose, J. 1993. Palaeoenvironmental
pretreatments for enhanced anaerobic digestion with waste-
and stratigraphic significance of the Valley Farm and Barham Soils
in eastern England. Quaternary Science Reviews 12, 833–848. activated sludge. Journal of Bioscience and Bioengineering 95,
Kemp, R.A., Jerz, H., Grottenhaler, W., et al. 1994. Pedosedimentary 271–275.
fabrics of soils within loess and colluvium in southern England Kim, J.-H., Schneider, R.R., Müller, P.J., et al. 2002. Interhemispheric
and southern Germany. In Soil Micromorphology: Studies in comparison of deglacial sea-surface temperature patterns in
Management and Genesis (edited by A.J. Ringrose-Voase, G.S. Atlantic eastern boundary currents. Earth and Planetary Science
Humphreys). Elsevier, Amsterdam, 207–219. Letters 194, 383–393.
Kemp, R.A., Derbyshire, E., Meng, X.M., et al. 1995. Pedosedimentary Kim, J.-H., Schouten, S., Hopmans, E.C., et al. 2008. Global sediment
reconstruction of a thick loess–paleosol sequence near Lhanzou core-top calibration of the TEX86 paleo-thermometer in the
in north-central China. Quaternary Research 43, 30–45. ocean. Geochimica et Cosmochimica Acta 72, 1154–1173.
Kemp, R.A., Derbyshire, E., Meng, X. 2001. A high-resolution Kindermann, K., Bubenzer, O., Nussbaum, S., et al. 2006. Palaeo-
micromorphological record of changing landscapes and climates environment and Holocene land use of Djara, Western Desert
on the western Loess Plateau of China during oxygen isotope of Egypt. Quaternary Science Reviews 25, 1619–1637.
stage 5. Palaeogeography, Palaeoclimatology, Palaeoecology 170, Kindler, P., Hearty, P.J. 2000. Elevated marine terraces from
157–169. Eleuthera (Bahamas) and Bermuda: sedimentological, petro-
Kemp, R.A., Toms, P.S., Sayago, J.M., et al. 2003. Micromorphology graphic and geochronological evidence for important deglaci-
and OSL dating of the basal part of the loess–palaeosol sequence ation events during the middle Pleistocene. Global and Planetary
at La Mesada in Tucumán province, Northwest Argentina. Change 24, 41–58.
Quaternary International 106/107, 111–117. King, E.C., Hindmarsh, R.C.A., Stokes, C.R. 2009a. Formation of
Kemp, R.A., King, M., Toms, P., et al. 2004. Pedosedimentary devel- mega-scale glacial lineations observed beneath a West Antarctic
opment of part of a Late Quaternary loess–palaeosol sequence ice stream. Nature Geoscience 2, 585–588.
in northwest Argentina. Journal of Quaternary Science 19, King, G., Bailey, G. 2006. Tectonics and human evolution. Antiquity
567–576. 80, 1–22.
Kemp, R.A., Zárate, M., Toms, P., et al. 2006. Late Quaternary King, G.A. 2012. Isotopes as palaeoeconomic indicators: new
paleosols, stratigraphy and landscape evolution in the Northern applications in archaeoentomology. Journal of Archaeological
Pampa, Argentina. Quaternary Research 66, 119–132. Science 39, 511–520.
482 REFERENCES
King, G.A., Gilbert, M.T.P., Willerslev, E., et al. 2009b. Recovery of Knudsen, M.F., Riisager, P., Jacobsen, B.H., et al. 2009. Taking
DNA from archaeological insect remains: first results, problems the pulse of the sun during the Holocene by joint analysis of 14C
and potential. Journal of Archaeological Science 36, 1179–1183. and 10Be. Geophysical Research Letters 36, doi: 10.1029/2009
Kiss, L., Magnin, F., Torre, F. 2004. The role of landscape history and GL039439
persistent biogeographical patterns in shaping the responses of Knutti, R., Flückiger, J., Stocker, T.F., et al. 2004. Strong hemispheric
Mediterranean land snail communities to recent fire disturb- coupling of glacial climate through freshwater discharge and
ances. Journal of Biogeography 31, 145–157. ocean circulation. Nature 430, 851–856.
Kjaer, K.H., Houmark-Nielsen, M., Richardt, N. 2003. Ice-flow Kober, B., Schwalb, A., Schettler, G., et al. 2007. Constraints on
patterns and dispersal of erratics at the southwestern margin of paleowater dissolved loads and on catchment weathering
the last Scandinavian Ice Sheet: signature of palaeo-ice streams. over the past 16 ka from 87Sr/86Sr ratios and Ca/Mg/Sr chemistry
Boreas 32, 130–148. of freshwater ostracode tests in sediments of Lake Constance,
Klassen, R.A. 2001. A Quaternary Geological Perpective on Geochemical Central Europe. Chemical Geology 240, 361–376.
Exploration in Glaciated Terrain. Geological Society of London, Koç, N. 2007. Diatom records – North Atlantic and Arctic. In
Special Publication 185. Blackwell Scientific Publications, Encyclopedia of Quaternary Science (edited by S.A. Elias). Elsevier,
Oxford, 1–17. Amsterdam, 567–576.
Kleiven H.F., Jansen, E., Fronval, T., et al. 2002. Intensification of Koch, P.L., Barnosky, A.D. 2006. Late Quaternary extinctions: state
North Hemisphere glaciations in the circum Atlantic region of the debate. Annual Review of Ecology, Evolution, and Systematics
(3.5–2.4 Ma) – ice-rafted detritus evidence. Palaeogeography, 37, 215–250.
Palaeoecology, Palaeooclimatology 184, 213–223. Kohfeld, K.E., Harrison, S.P. 2000. How well can we simulate past
Kleiven, H.F., Kissel, C., Laj, C., et al. 2008. Reduced North Atlantic climates? Evaluating the models using global palaeoenviron-
deep water coeval with the Glacial Lake Agassiz freshwater mental datasets. Quaternary Science Reviews 19, 321–346.
outburst. Science 319, 60–64. Kohfeld, K.E., Harrison, K.E. 2003. Glacial–interglacial changes in
Kligmann, D.M., Calderari, M. 2012. Diatoms and ceramic proven- dust deposition on the Chinese Loess Plateau. Quaternary Science
ance: a cautionary tale. Archaeometry 54, 129–143. Reviews 22, 1859–1878.
Klok, E.J., Oerlemans, J. 2004. Climate reconstructions derived from Kohn, M.J., McKay, M.P. 2012. Paleoecology of late Pleistocene–
global glacier length records. Arctic, Antarctic and Alpine Research Holocene faunas of eastern and central Wyoming, USA, with
36, 573–583. implications for LGM climate models. Palaeogeography,
Klotz, S., Müller, U., Mosbrugger, V., et al. 2004. Eemian to early Palaeoclimatology, Palaeoecology 326–328.
Würmian climate dynamics: history and pattern of changes in Kolstrup E. 1993. Complex frost wedge casts as indicators of ice age
Central Europe. Palaeogeography, Palaeoclimatology, Palaeo- environmental change. Würzburger Geographische Arbeiten 87,
ecology 211, 107–126. 269–282.
Klütsch, C.F.C., Manseau, M., Wilson, P.J. 2012. Phylogeographical Kolstrup, E. 2004. Stratigraphic and environmental implications of
analysis of mtDNA data indicates Postglacial expansion from a large ice-wedge cast at Tjæreborg, Denmark. Permafrost and
multiple glacial refugia in Woodland Caribou (Rangifer tarandus Periglacial Processes 15, 31–40.
caribou). PLoS ONE 7(12): e52661, doi: 10.1371/journal. Konopinski, D.I., Hudziak, S., Morgan, R., et al. 2012. Investigation
pone.0052661 of quartz surface textures by atomic force microscopy for forensic
Kneisel, C. 2010. Frozen ground conditions in a subarctic mountain science. Forensic Science International 223, 245–255.
environment, Northern Sweden. Geomorphology 118, 80–92. Korhola, A. 2007. Diatom methods: data interpretation. In Encyclo-
Knight, J. 2012. Glacitectonic sedimentary and hydraulic processes pedia of Quaternary Science (edited by S.A. Elias). Elsevier,
at an oscillating ice margin. Proceedings of the Geologists’ Amsterdam, 494–507.
Association 123, 714–727. Korsun, S., Hald, M. 1998. Modern benthic Foraminifera off Novaya
Knight, J., Harrison, S. (eds). 2009. Periglacial and Paraglacial Zemlya tidewater glaciers, Russian Arctic. Arctic and Alpine
Processes and Environments. Geological Society of London, Special Research 30, 61–77.
Publication 320. Geological Society, London. Kos, A.M. 2003. Characterisation of post-depositional taphonomic
Knight, P.J. 1999. Glaciers. Stanley Thornes, Cheltenham. processes in the accumulation of mammals in a pitfall cave
Knox, R.W.O’B., Pearson, P.N., Barry, T.L., et al. 2012. Examining deposit in southeastern Australia. Journal of Archaeological Science
the case for the use of the Tertiary as a formal period or informal 30, 781–796.
unit. Proceedings of the Geologists’ Association 123, 390–393. Kosnik, M.A., Kaufman, D.S., Hua, Q. 2008. Identifying outliers
Knudsen, K.-L., Seidenkrantz, M.-S., Kristensen, P. 2002. Last and assessing the accuracy of amino acid racemization measure-
interglacial and early glacial circulation in the northern North ments for geochronology: I. Age calibration curves. Quaternary
Atlantic Ocean. Quaternary Research 58, 22–26. Geochronology 3, 308–327.
Knudsen, K.L., Eiríksson, J., Bartels-Jónsdóttir, H.B. 2012. Oceano- Kossler, A., Tarasov, P., Schlolaut, G., et al. 2011. Onset and
graphic changes through the last millennium off North termination of the late-glacial climate reversal in the high-
Iceland: temperature and salinity reconstructions based on resolution diatom and sedimentary records from the annually
foraminifera and stable isotopes. Marine Micropaleontology laminated SG06 core from Lake Suigetsu, Japan. Palaeogeography,
84–85, 54–73. Palaeoclimatology, Palaeoecology 306, 103–115.
REFERENCES 483
Kovanen, D.J., Easterbrook, D.J. 2002. Timing and extent of Allerød in the southern North Sea region. Geologie en Mijnbouw
and Younger Dryas Age (ca. 12,500–10,000 14C yr B.P.) (Netherlands Journal of Geosciences) 85, 19–35.
oscillations of the Cordilleran Ice Sheet in the Fraser Lowland, Kuhn, B.F., Berger, L.R., Skinner, J.D. 2010. Examining criteria for
Western North America. Quaternary Research 57, 208–224. identifying and differentiating fossil faunal assemblages
Kovanen, D.J., Slaymaker, O. 2004. Relict shorelines and ice flow accumulated by hyenas and hominins using extant hyenid
patterns of the Northern Puget Lowland from lidar data and accumulations. International Journal of Osteoarchaeology 20,
digital terrain modelling. Geografiska Annaler A86, 385–400. 15–35.
Krachler, M., Zheng, J., Fisher, D., et al. 2009. Global atmospheric Kühn, P., Terhorst, B., Ottner, F. 2006. Micromorphology of middle
As and Bi contamination preserved in 3000 year old Arctic ice. Pleistocene palaeosols in northern Italy. Quaternary International
Global Biogeochemical Cycles 23, doi: 10.1029/209/GB003471. 156/157, 156–166.
Kramer, A., Herzschuh, U., Mischke, S., et al. 2010. Late Quaternary Kukla, G.J. 1977. Pleistocene land–sea correlations – 1. Europe.
environmental history of the south-eastern Tibetan Plateau Earth-Science Reviews 13, 307–374.
inferred from the Lake Naleng non-pollen palynomorph record. Kumambala, P.G., Ervine, A. 2010. Water balance model of Lake
Vegetation History and Archaeobotany 19, 453–468. Malawi and its sensitvity to climate change. The Open Hydrology
Krasilnikov, P., García-Calderón, N.E. 2006. A WRB-based buried Journal 4, 152–162.
paleosol classification. Quaternary International 156–157, Kumar, S. 2005. Molecular clocks: four decades of evolution. Nature
176–188. Reviews Genetics 6, 654–662.
Kravchinsky, V.A., Langereis, C.G., Walker, S.D., et al. 2013. Kurtén, B. 2009. Pleistocene Mammals of Europe, 2nd edn.
Discovery of Holocene millennial climate cycles in the Asian Transaction Publishers, New Brunswick, NJ.
continental interior: has the sun been governing the conti- Kurth, G., Phillips, F.M., Rehels, M., et al. 2010. Cosmogenic nuclide
nental climate? Global and Planetary Change doi: 10.1016/j. and uranium-series dating of old, high shorelines in the western
gloplacha2013.02.011. Great Basin, USA. Geological Society of America Bulletin 123,
Krishnan, R., Sugi, M. 2003. Pacific decadal oscillation and vari- 744–768.
ability of the Indian summer monsoon rainfall. Climate Dynamics Kutzbach, J.E., Guetter, P.J. 1986. The influence of changing orbital
21, 233–242. parameters and surface boundary conditions on climate
Kromer, B. 2009. Radiocarbon and dendrochronology. Dendro- simulations for the past 18,000 years. Journal of Atmospheric
chronologia 27, 15–19. Sciences 43, 1726–1759.
Kubala-Kukus, A., Ludwikowska-Kedzia, M., Banas, D., et al. 2013. Kutzbach, J.E., Webb III, T. 1993. Conceptual basis for understand-
Application of the X-ray fluorescence analysis and X-ray ing Late Quaternary climates. In Global Climates Since the Last
diffraction in geochemical studies of the Pleistocene tills from Glacial Maximum (edited by H.E.Wright Jr., J.E.Kutzbach,
Holy Cross Mountains. Radiation Physics and Chemistry 93, T. Webb III, et al.). University of Minnesota Press, Minneapolis
92–98. and London, 1–5.
Kubiak-Martens, L. 1999. The plant food component of the diet Kutzbach, J.E., Guetter, P.J., Behling, P.J., et al. 1993. Simulated
at the late Mesolithic (Ertebolle) settlement at Tybrind Vig, climatic changes: results of the COHMAP climate-model experi-
Denmark. Vegetation History and Archaeobotany 8, 117–127. ments. In Global Climates Since the Last Glacial Maximum (edited
Kucera, M. 2007. Planktonic Foraminifera as tracers of past oceanic by H.E.Wright Jr., J.E. Kutzbach, T. Webb III, et al.). University
environments. In Developments in Marine Geology, Volume 1: of Minnesota Press, Minneapolis and London, 24–93.
Proxies in Late Cenozoic Paleoceanography (edited by C. Hillaire- Kuzmin, Y.V., Burr, G.S., Jull, A.H.T. 2001. Radiocarbon reservoir
Marcel, A. De Vernal). Elsevier, Amsterdam, 213–262. correction ages in the Peter the Great Gulf, Sea of Japan, and
Kucera, M., Rosell-Melé, A., Schneider, R., et al. 2005a. Multiproxy eastern coast of the Kunashir, southern Kuriles (northwestern
approach for the reconstruction of the glacial ocean surface Pacific). Radiocarbon 43, 477–481.
(MARGO). Quaternary Science Reviews 24, 813–819. Kuzmina, S.A., Sher, A.V., Edwards, M.E., et al. 2011. The late
Kucera, M., Schneider, R., Weinelt, M. (eds). 2005b. MARGO: Pleistocene environment of the Eastern West Beringia based on
Multiproxy approach for the reconstruction of the glacial ocean the principal section at the Main River, Chukotka. Quaternary
surface. Quaternary Science Reviews 24, 813–1107. Science Reviews 30, 2091–2106.
Kuehn, S.C., Froese, D.G., Carrara, P.E., et al. 2009. Major- and Kwiecien, O., Arz, H.W., Lamy, F., et al. 2008. Estimated reservoir
trace-element characterization, expanded distribution, and a ages of the Black Sea since the Last Glacial. Radiocarbon 50,
new chronology for the latest Pleistocene Glacier Peak tephras 99–118.
in western North America. Quaternary Research 71, 201–216. Lachniet, M.S., Vazquez-Selem, L. 2005. Last Glacial Maximum
Kühl, N., Gebhardt, C., Litt, T., et al. 2002. Probability density equilibrium line altitudes in the circum-Caribbean (Mexico,
functions as botanical–climatological transfer functions for Guatemala, Costa Rica, Colombia, and Venezuela). Quaternary
climate reconstruction. Quaternary Research 58, 381–392. International 138/139, 129–144.
Kuhlmann, G., Langereis, C.G., Munsterman, D., et al. 2006. Laepple, T., Werner, M., Lohmann, G. 2011. Synchronicity of
Integrated chronostratigraphy of the Pliocene–Pleistocene Antarctic temperatures and local solar insolation on orbital
interval and its relation to the regional stratigraphical stages timescales. Nature 471, 91–94.
484 REFERENCES
Lagerlund, E., Houmark-Nielsen, M. 2008. Timing and pattern Lane, C.S., Brauer, A.C., Blockley, S.P.E., et al. 2013. Volcanic ash
of the last deglaciation in the Kattegat region, southwest reveals a time-transgressive abrupt climate change during the
Scandinavia. Boreas 22, 337–347. Younger Dryas. Geology 41, 1251–1254.
Lagoe, M.B., Eyles, C.H., Eyles, N., et al. 1993. Timing of Late Lane, S.N., Chandler, J.H. 2003. The generation of high quality
Cenozoic tidewater glaciation in the far North Pacific. Geological topographic data for hydrology and geomorphology: new data
Society of America Bulletin 105, 1542–1560. sources, new applications and new problems. Earth Surface
Laird, K.R., Michels, A., Stuart, C.T.L., et al. 2007. Examination of Processes and Landforms 28, 229–230.
diatom-based changes from a climatically sensitive prairie lake Lang, B., Bedford, A.P., Richardson, N., et al. 2003. The use of ultra-
(Saskatchewan, Canada) at different temporal perspectives. sound in the preparation of carbonate and clay sediments for
Quaternary Science Reviews 26, 3328–3343. chironomid analysis. Journal of Paleolimnology 30, 451–460.
Laj, C., Kissel, C., Mazaud, A., Michel, E., et al. 2002. Geomagnetic Lang, N., Wolff, E.W. 2011. Interglacial and glacial variability from
field intensity, North Atlantic Deep Water circulation and the last 800 ka in marine, ice and terrestrial archives. Climate of
atmospheric Δ14C during the last 50 k yr. Earth and Planetary the Past 7, 361–380.
Science Letters 200, 177–190. Langdon, P.G., Barber, K.E. 2004. Snapshots in time: precise
Lakeman, T.R. Clague, J.J., Menonous, B. 2008. Advance of alpine correlations of peat based proxy climate records in Scotland using
glaciers during final retreat of the Cordilleran ice sheet in mid-Holocene tephras. The Holocene 14, 21–33.
the Finlay River area, northern British Columbia, Canada. Langdon, P.G., Barber, K.E. 2005. The climate of Scotland over the
Quaternary Research 69, 188–200. last 5000 years inferred from multiproxy peatland records: inter-
Lambeck, K., Chappell, J. 2001. Sea level change through the last site correlations and regional variability. Journal of Quaternary
glacial cycle. Science 292, 679–686. Science 20, 549–566.
Lambeck, K., Purcell, A., Johnston, P., et al. 2003. Water-load defini- Langdon, P.G., Barber, K.E., Hughes, P.D.M. 2003. A 7500-year peat-
tion in the glacio-hydro-isostatic sea-level equation. Quaternary based palaeoclimatic reconstruction and evidence for an 1100-
Science Reviews 22, 309–318. year cyclicity in bog surface wetness from Temple Hill Moss,
Lambert, D.M., Ritchie, P.A., Millar, C.D., et al. 2002. Rates of Pentland Hills, southeast Scotland. Quaternary Science Reviews
evolution in ancient DNA from Adélie penguins. Science 295, 22, 259–274.
2270–2273. Lanphere, M.A. 2000. Comparison of conventional K–Ar and
40
Lambert, F., Delmonte B., Petit, J.R., et al. 2008. Dust record from Ar/39Ar dating of young mafic volcanic rocks. Quaternary
the EPICA Dome C ice core, Antarctica, covering 0 to 800 kyr Research 53, 294–301.
BP. Nature 452, 616–619. Larkin, N.R. 2010. Literally a ‘mammoth task’: the conservation,
Lamoureux, S.F., England, J.H., Sharp, M.J., et al. 2001. A varve record preparation and curation of the West Runton Mammoth
of increased ‘Little Ice Age’ rainfall associated with volcanic skeleton. Quaternary International 228, 233–240.
activity, Arctic Archipelago, Canada. The Holocene 11, Larocque, I., Hall, R.I. 2003. Chironomids as quantitative indicators
243–249. of mean July air temperature: validation by comparison with
Lamy, F., Kaiser, J., Ninnemann, U., et al. 2004. Antarctic timing century-long meteorological records from northern Sweden.
of surface water changes off Chile and Patagonian ice sheet Journal of Paleolimnology 29, 475–493.
response. Science 304, 1959–1962. Larocque, S.J., Smith, D.J. 2003. Little Ice Age glacial activity in the
Lancaster, N. 2000. How dry was dry? Late Pleistocene palaeo- Mt. Waddington area, British Columbia Coast Mountains,
climates in the Namib Desert. Quaternary Science Reviews 21, Canada. Canadian Journal of Earth Science 40, 1413–1436.
769–782. Larsen, N.K., Piotrowski, J.A. 2003. Fabric pattern in a basal till
Lancaster, N. 2008. Desert dune dynamics and development: insights succession and its significance for reconstructing subglacial
from luminescence dating. Boreas 37, 559–573. processes. Journal of Sedimentary Research 73, 725–734.
Lancaster, N., Kocurek, G., Singhvi, A., et al. 2002. Late Pleistocene Larson, G.J. 2011. Ice-margin fluctuations at the end of the Wisconsin
and Holocene dune activity and wind regimes in the western episode, Michigan, USA. In Developments in Quaternary Science.
Sahara Desert of Mauritania. Geology 30, 991–994. Volume 15: Quaternary Glaciations – Extent and Chronology:
Lane, C.S., Blockley S.P.E., Lotter, A.F., et al. 2010. A regional A Closer Look (edited by J. Ehlers, P.L. Gibbard, P.D. Hughes).
tephrosratigraphical framework for central and southern Euro- Elsevier, Amsterdam, 489–497.
pean climate archives during the Last Glacial to Interglacial Last, W.M., Smol, J.P. (eds). 2001. Tracking Environmental Change
transition: comparisons north and south of the Alps. Quaternary Using Lake Sediments. Volume 1: Basin Analysis, Coring, and
Science Reviews 36, 50–58. Chronological Techniques. Kluwer, Dordrecht.
Lane, C.S., Andric, M., Cullen, V.L., et al. 2011. The occurrence of Latalowa, M., Van der Knaap, W.O. 2006. Late Quaternary expansion
distal Icelandic and Italian tephra in the Lateglacial of Lake of Norway spruce Picea abies (L.) Karst. in Europe according to
Bled, Slovenia. Quaternary Science Reviews 30, 1013–1018. pollen data. Quaternary Science Reviews 25, 2780–2805.
Lane, C.S., Blockley, S.P.E., Mangerud, J., et al. 2012. Was the 12.1 Latham, A.G. 2001. Uranium-series dating. In Handbook of
ka Icelandic Vedde Ash one of a kind? Quaternary Science Reviews Archaeological Sciences (edited by D.E. Brothwell, A.M. Pollard).
33, 87–99. John Wiley, Chichester and New York, 63–72.
REFERENCES 485
Latrubesse, E.M., Nelson, B.W. 2001. Evidence for Late Quaternary Lehtonen, M.L., Marmo, J.S., Nissinen, A.J., et al. 2005. Glacial
aeolian activity in the Roraima–Guyana Region. Catena 43, dispersal studies using indicator minerals and till geochemistry
63–80. around two eastern Finland kimberlites. Journal of Geochemical
Lauritzen, S.-E. 1993. Natural environmental change in karst: the Exploration 87, 19–43.
Quaternary record. In Karst Terrains: Environmental Changes and Leier, A.L., DeCelles, P.G., Pelletier, J.D. 2005. Mountains, monsoons,
Human Impact (edited by P.W. Williams). Catena Suppplement and megafans. Geology 33, 289–292.
25, Catena, Cremlingen-Destedt, 21–40. Lekens, W.A.H., Sejrup, H.P., Haflidason, H., et al. 2005. Laminated
Lauritzen, S.-E., Lundberg, J. (eds). 1999. Speleothems as high- sediments preceding Heinrich Event 1 in the northern North Sea
resolution palaeoclimatic archives. The Holocene 9, 643–726. and southern Norwegian Sea: origin, processes and regional
Lawson, I., Frogley, M., Bryant, C., et al. 2004. The Lateglacial and linkage. Marine Geology 216, 27–50.
Holocene environmental history of the Ioannina basin, north- Lemieux-Dudon, B., Blayo, E., Petit, J.-R., et al. 2010. Consistent
west Greece. Quaternary Science Reviews 23, 1599–1625. dating for Antarctic and Greenland ice cores. Quaternary Science
Laxon, S.W., Giles, K.A., Ridout, A.L., et al. 2013. CryoSat-2 estimates Reviews 29, 8–20.
of Arctic sea ice thickness and volume. Geophysical Research Letters Leng, M.J. (ed.). 2004. Isotopes in Quaternary palaeoenviron-
doi: 10.1002/grl.50193. mental reconstruction (ISOPAL). Quaternary Science Reviews
Lea, D.A., Martin, P.A., Pak, D.K., et al. 2002. Reconstructing a 350 23, 739–991.
ky history of sea level using planktonic Mg/Ca and oxygen Leng, M.J., Marshall, J.D. 2004. Palaeoclimatic interpretation of
isotope records from a Cocos Ridge core. Quaternary Science stable isotope data from lake sediment archives. Quaternary
Reviews 21, 283–293. Science Reviews 23, 811–831.
Lea, D.W., Mashiotta, T.A., Spero, H.J. 1999. Controls on magnesium Leng, M.J., Heaton,T.H.E., Lamb, H.F., et al. 1998. Carbon and oxygen
and strontium uptake in planktonic foraminifera determined isotope variations within the shell of an African land snail
by live culturing. Geochimica et Cosmochimica Acta 63, (Limicolaria kambeul chudeaui Germain): a high-resolution
2369–2379. record of climate seasonality? The Holocene 8, 407–412.
Lea, D.W., Pak, D.K., Peterson, L.C., et al. 2003. Synchroneity of Leng, M.J., Barker, P.A., Schwalb, A. 2013. Isotopes and lakes.
tropical and high-latitude Atlantic temperatures over the Last Quaternary Science Reviews 66, 1–36.
Glacial Termination. Science 301, 1361–1364. Lenton, T.M., Myerscough, R.J., Marsh, R., et al. 2009. Using GENIE
Lean, J. 2009. Cycles and trends in solar irradiance and climate. to study a tipping point in the climate system. Philosophical
Climate Change 1, 111–122. Transactions of the Royal Society of London A 367, 871–884.
Leblanc, M.J., Leduc, C., Stagnitti, F., et al. 2006. Evidence for Leroy, S., Dupont, L. 1994. Development of vegetation and con-
Megalake Chad, north-central Africa, during the late Quaternary tinental aridity in northwestern Africa during the Late Pliocene:
from satellite data. Palaeogeography, Palaeoclimatology, Palaeo- the pollen record of ODP Site 658. Palaeogeography, Palaeo-
ecology 230, 230–242. climatology, Palaeoecology 109, 295–316.
Lécuyer, C., Daux, V., Moisette, P., et al. 2012. Stable carbon and Leverington, D.W., Mann, J.D., Teller, J.T. 2002. Changes in the
oxygen isotope compositions of invertebrate carbonate shells bathymetry and volume of glacial Lake Agassiz between 9200 and
and the reconstruction of paleotemperatures and paleosalinities 7700 14C yr BP. Quaternary Research 57, 244–252.
– A case study of the early Pleistocene of Rhodes, Greece. Lewis, S.E., Sloss, C.R., Murray-Wallace, C.V., et al. 2012. Post-glacial
Palaeogeography, Palaeoclimatology, Palaeoecology 350–352, sea-level changes around the Australian margin: a review.
39–48. Quaternary Science Reviews 74, 115–138.
Lee, J.R., Phillips, E.R. 2008. Progressive soft sediment deformation Lewis, S.G., Maddy, D., Buckingham, C., et al. 2006. Pleistocene fluvial
within a subglacial shear zone – a hybrid mosaic–pervasive sediments, palaeontology and archaeology of the upper River
deformation model for Middle Pleistocene glaciotectonised Thames at Latton, Wiltshire, England. Journal of Quaternary
sediments from eastern England. Quaternary Science Reviews 27, Science 21, 181–205.
1350–1362. Li, Y., Morrill, C. 2010. Multiple factors causing Holocene lake-level
Lee, J., Rose, J., Hamblin, R.J.O., et al. 2004. Dating the earliest lowland change in monsoonal and arid central Asia as identified by
glaciation of eastern England: a pre-MIS 12 early Middle model experiments. Climate Dynamics 35, 1119–1132.
Pleistocene Happisburgh glaciation. Quaternary Science Reviews Lian, O., Huntley, D.J. 2001. Luminescence dating. In Tracking
23, 1551–1566. Environmental Change Using Lake Sediments. Volume 1: Basin
Lee, J.R. Rose, J., Hamblin, R.J.O., et al. 2011. The glacial history of Analysis, Coring and Chronological Techniques (edited by W.M.
the British Isles during the Early and Middle Pleistocene: Last, J.P. Smol). Kluwer, Dordrecht, 261–282.
implications for the long-term development of the British Libby, W.F. 1955. Radiocarbon Dating, 2nd edn. University of
Ice Sheet. In Developments in Quaternary Science. Volume 15: Chicago Press, Chicago.
Quaternary Glaciations – Extent and Chronology: A Closer Look Licciardi, J.M. 2001. Chronology of latest Pleistocene lake-level
(edited by J. Ehlers, P.L. Gibbard, P.D. Hughes). Elsevier, fluctuations in the pluvial Lake Chewaucan basin, Oregon, USA.
Amsterdam, 59–74. Journal of Quaternary Science 16, 545–553.
Lehmann, G., Röttger, R. 1997. Techniques for the concentration Lillesand, T.M., Kiefer, R.W., Chipman, J. 2004. Remote Sensing and
of foraminifera from coastal salt meadow sediments. Journal of Image Interpretation, 5th edn. John Wiley, Chichester and
Micropalaeontology 16, 144. Hoboken, NJ.
486 REFERENCES
Lima, F.P., Ribeiro, P.A., Queiroz, N., et al. 2007. Modelling past and Liu, T. 2003. Blind testing of rock varnish microstratigraphy as a
present geographical distribution of the marine gastropod Patella chronometric indicator: results on late Quaternary lava flows in
rustica as a tool for exploring responses to environmental change. the Mojave Desert, California. Geomorphology 53, 209–234.
Global Change Biology 13, 2065–2077. Liu, T., Ding, Z. 1998. Chinese loess and the paleomonsoon. Annual
Limondin-Lozouet, N., Antoine, P. 2006. A new Lyrodiscus Reviews of Earth and Planetary Sciences 26, 111–145.
(Mollusca, Gastropoda) assemblage from Saint-Acheul (Somme Liu, T., Broecker, W.S. 2008. Rock varnish microlamination
Valley): a reappraisal of MIS 11 malacofaunas from northern dating of late Quaternary geomorphic features in the drylands
France. Boreas 35, 622–633. of western USA. Geomorphology 93, 501–523.
Linderholm, H.W., Bjorklund, J.A., Seftigen, K., et al. 2010. Dendro- Liu X.-J., Lai Z.-P. 2013. Optical dating of sand wedges and ice-wedge
climatology in Fennoscandia – from past accomplishments to casts from Qinghai Lake area on the northeastern Qinghai-
future potential. Climate of the Past 6, 93–114. Tibetan Plateau and its palaeoenvironmental implications.
Lindqvist, C., Schuster, S.C., Sun, Y., et al. 2010. Complete mito- Boreas 42, 333–341.
chondrial genome of a Pleistocene jawbone unveils the origin Liu, Z., Alexander, M. 2007. Atmospheric bridge, oceanic tunnel, and
of polar bear. Proceedings of the National Academy of Sciences of global climatic teleconnections. Reviews of Geophysics 45, doi:
the United States of America 107, 5053–5057. 2005RG000172.
Linge, H., Lauritzen, S.-E., Lundberg, J. 2001. Stable isotope strati- Livingstone, S.J., Evans, D.J.A., Ó Cofaigh, C. 2010. Re-advance of
graphy of a late last interglacial speleothem from Rana, northern Scottish ice into the Solway Lowlands (Cumbria, UK), during the
Norway. Quaternary Research 56, 155–164. Main Late Devensian deglaciation. Quaternary Science Reviews
Lingle, C.S., Fatland, D.R. 2003. Does englacial water storage drive 29, 2544–2570.
temperate glacier surges? Annals of Glaciology 36, 14–20. Lloyd, G.T., Pearson, P.N., Young, J.R., et al. 2012. Sampling bias and
Liritzis, I. 2006. SIMS-SS, a new obsidian hydration dating method: the fossil record of planktonic foraminifera on land and in the
analysis and theoretical principles. Archaeometry 48, 533–547. deep sea. Paleobiology 38, 569–584.
Lisé-Pronovost, A., St-Onge, G., Gogorza, C. et al. 2013. High- Loader, N.J., McCarroll, D., van der Knaap, W.O., et al. 2007.
resolution paleomagnetic secular variations and relative paleo- Characterising carbon isotopic variability in Sphagnum. The
intensity since the Late Pleistocene in southern South America. Holocene 17, 403–410.
Quaternary Science Reviews 71, 91–108. Loader, N.J., Young, G.H.F., Grudd, H., et al. 2013. Stable carbon
Lisiecki, L.E., Raymo, M.E. 2005. A Pliocene–Pleistocene stack of isotopes from Torneträsk, northern Sweden provide a millennial
57 globally distributed benthic δ18O records. Paleoceanography length reconstruction of summer sunshine and its relationship
20, PA1003, doi: 10.1029/2004PA001071. to Arctic circulation. Quaternary Science Reviews 62, 97–113.
Lisiecki, L.E., Raymo, M.E. 2007. Plio–Pleistocene climate evolution: Lohne, Ø.S., Bondevik, S., Mangerud, J., et al. 2007. Sea-level
trends and transitions in glacial cycle dynamics. Quaternary fluctuations imply that the Younger Dryas ice-sheet expansion
Science Reviews 26, 56–69. in western Norway commenced during the Allerød. Quaternary
Lisker, S., Vaks, A., Bar-Matthews, M., et al. 2009. Stromatolites in Science Reviews 26, 2128–2151.
caves of the Dead Sea Fault Escarpment: implications to latest Lohne, Ø.S., Mangerud, J., Birks, H.H. 2013. Precise 14C ages of the
Pleistocene lake levels and tectonic susidence. Quaternary Science Vedde and Saksunarvatn ashes and the Younger Dryas bound-
Reviews 28, 80–92. aries from western Norway and their comparison with the
Lister, A.M. 2004. The impact of Quaternary ice ages on mammalian Greenland Ice Core (GICC05) chronology. Journal of Quaternary
evolution. Philosophical Transaction of the Royal Society of London Science 28, 490–500.
B 359, 221–241. Lomas-Clarke, S.H., Barber, K.E. 2006. Human impact signals from
Lister, A.M. 2009. Late-glacial mammoth skeletons (Mammuthus peat bogs: a combined palynological and geochemical approach.
primigenius) from Condover (Shropshire, UK): anatomy, Vegetation History and Archaeobotany 16, 419–429.
pathology, taphonomy and chronological significance. Geological Long, A.J., Roberts, D.H., Simpson, M.J.R., et al. 2008. Late
Journal 44, 447–479. Weichselian relative sea-level changes and ice sheet history in
Lister, A.M., Bahn, P. 2009. Mammoths: Giants of the Ice Age. southeast Greenland. Earth and Planetary Science Letters 272, 8–18.
University of California Press, Berkeley. Long, A.J., Woodroffe, S., Dawson, S., et al. 2009. Late Holocene
Litt, T. Gibbard, P.L. 2008. A proposed Global Stratotype Section and relative sea level rise and the Neoglacial history of the Greenland
Points (GSSP for the base of the Upper (Late) Pleistocene Ice Sheet. Journal of Quaternary Science 24, 345–359.
Subseries (Quaternary System/Period). Episodes 31, 260–263. Lord, A., Cabral, M.C., Dambeck, R., et al. 2011. Ostracod evidence
Litt, T., Brauer, A., Goslar, T., et al. 2001. Correlation and syn- for the Neolithic environment of Rio Sizandro, Portugal.
chronisation of Lateglacial continental sequences in northern Palaeobiodiversity and Palaeoenvironments 91, 215–228.
Central Europe based on annnually laminated lacustrine Lord, A.R., Boomer, I., Brouwers, E., et al. 2012. Ostracod taxa as
sediments. Quaternary Science Reviews 20, 1233–1249. palaeoclimatic indicators in the Quaternary. In Developments in
Liu, K.-b., Yao, Z., Thompson, L.G. 1998. A pollen record of Quaternary Science. Volume 17: Ostracoda as Proxies for
Holocene climatic changes from the Dunde ice cap, Qinghai- Quaternary Climate Change (edited by D.J. Horne, J.A. Holmes,
Tibetan Plateau. Geology 26, 135–138. J. Rodriguez-Lazaro, et al.). Elsevier, Amsterdam, 37–46.
REFERENCES 487
Loso, M.G., Doak, D.F. 2006. The biology behind lichenometric dating Lowe, J.J., Rasmussen, S.O., Björck, S., et al. 2008b. Synchronisation
curves. Oecologia 147, 223–229. of palaeoenvironmental events in the North Atlantic region
Lotter, A.F., Heiri, O., Brooks, S., et al. 2012. Rapid summer during the Last Termination: a revised protocol recommended
temperature changes during Termination 1a: high-resolution by the INTIMATE group. Quaternary Science Reviews 27,
multi-proxy climate reconstructions from Gerzensee (Switzer- 6–17.
land). Quaternary Science Reviews 36, 103–113. Lowe, J., Barton, N., Blockley, S., et al. 2012. Volcanic ash layers
Loubere, P., Creamer, W., Haas, J. 2013. Evolution of the El illuminate the resilience of Neanderthals and early modern
Niño–Southern Oscillation in the late Holocene and insolation humans to natural hazards. Proceedings of the National Academy
driven change in the tropical annual SST cycle. Global and of Sciences of the United States of America 109, 13532–13537.
Planetary Change doi: 10.1016/j.gloolacha.2012.10.007. Lowell, T.V., Hayward, R.K., Denton, G.H. 1999. Role of climate
Louchart, A., Wesselmann, H., Blumenschine, R.J., et al. 2009. oscillations in determining ice-margin position: hypothesis,
Taphonomic, avian, and small-vertebrate indicators of Ardi- example and implication. Geological Society of America, Special
pithecus ramidus habitat. Science 326, 66 (66e1–66e4). Paper 337, 192–203.
Lough, J.M., Cooper, T.F. 2011. New insights from coral growth band Lozier, M.S. 2010. Deconstructing the conveyor belt. Science 328,
studies in an era of rapid environmental change. Earth-Science 1507–1511.
Reviews 108, 170–184. Lücke, A., Schleser, G.H., Zolitschka, B., et al. 2003. A Lateglacial and
Loulerge, L., Parrein, F., Blnier, T., et al. 2007. New constraints on Holocene organic carbon isotope record of lacustrine
the gas age-ice age difference along the EPICA ice cores, 0–50 kyr. palaeoproductivity and climate change derived from varved lake
Climate of the Past 3, 527–540. sediments of Lake Holzmaar, Germany. Quaternary Science
Loulerge, L., Schilt, A., Spahni, R., et al. 2008. Orbital and millennial- Reviews 22, 569–580.
scale features of atmospheric CH4 over the past 800,000 years. Luckman, B.H. 2000. The Little Ice Age in the Canadian Rockies.
Nature 453, 383–386.
Geomorphology 32, 357–384.
Lourens, L. 2004. Revised tuning of Ocean Drilling Program Site 964
Luckman, B.H., Holdsworth, G., Osborn G.D. 1993. Neoglacial
and KC01B (Mediterranean) and implications for the δ18O,
glacier fluctuations in the Canadian Rockies. Quaternary Research
tephra, calcareous nannofossil, and geomagnetic reversal
39, 144–153.
chronologies of the past 1.1 Myr. Paleoceanography 19, doi:
Lüer, V., Hollis, C.J., Willems, H. 2008. Late Quaternary radiolarian
10.1029/2003PA000997.
assemblages as indicators of paleoceanographic changes north
Lourens, L.J. 2008. On the Neogene–Quaternary debate. Episodes 31,
of the subtropical front, offshore eastern New Zealand, southwest
239–242.
Pacific. Micropaleontology 54, 49–69.
Lourens, L.J., Hilgen, F.J. 1997. Long-period variations in the
Lukas, S. 2006. Morphostratigraphic principles in glacier recon-
Earth’s obliquity and their relation to third-order eustatic
cycles and late Neogene glaciations. Quaternary International struction – a perspective from the British Younger Dryas. Progress
40, 43–52. in Physical Geography 30, 719–736.
Lourens L., Hilgen, F., Shackleton, N.J., et al. 2005. The Neogene Lukas, S., Benn, D.I. 2006. Retreat dynamics of Younger Dryas
Period. In A Geologic Time Scale 2004 (edited by F. Gradstein, glaciers in the far NW Scottish Highlands reconstructed from
L.G. Ogg, A.G. Smith). Cambridge University Press, Cambridge, moraine sequences. Scottish Geographical Journal 122, 308–325.
409–440. Lukas, S., Nicholson, L.I., Ross, F.H., et al. 2005. Formation, melt-
Lowe, D.J. 2011. Tephrochronology and its applications. Quaternary out processes and landscape alteration of high-Arctic ice-cored
Geochronology 6, 107–153. moraines – examples from Nordenskiold Land, Central Spits-
Lowe, D.J., Shane, P.A.R., Alloway, B.V., et al. 2008a. Fingerprints bergen. Polar Geography 29, 157–187.
and age models for widespread New Zealand tephra marker beds Lukas, S., Spencer, J.Q.G., Robinson, R.A.J., et al. 2007. Problems
erupted since 30,000 years ago: a framework for NZ-INTIMATE. associated with luminescence dating of Late Quaternary glacial
Quaternary Science Reviews 27, 95–126. sediments in the NW Scottish Highlands. Quaternary Geo-
Lowe, J.J., Lowe, S., Fowler, A.J., et al. 1988. Comparison of chronology 2, 243–248.
accelerator and radiometric age measurements obtained from Lundqvist, J. 1986. Late Weichselian glaciation and deglaciation in
Late Devensian Lateglacial lake deposits from Llyn Gwernan, Scandinavia. Quaternary Science Reviews 5, 269–292.
North Wales. Boreas 17, 355–369. Lundqvist, J., Wohlfarth, B. 2001. Timing and east–west correlation
Lowe, J.J., Hoek, W., INTIMATE Group. 2001. Inter-regional of south Swedish ice marginal lines during the Late Weichselian.
correlation of palaeoclimatic records for the Last Glacial– Quaternary Science Reviews 20, 1127–1148.
Interglacial Transition: a protocol for improved precision Lunt, D.J., Williamson, M.S., Valdes, P.J.I. 2006. Comparing
recommended by the INTIMATE project group. Quaternary transient, accelerated, and equilibrium simulations of the last
Science Reviews 20, 1175–1187. 30000 years with the GENIE-1 model. Climate of the Past 2,
Lowe, J.J., Blockley, S., Trincardi, F., et al. 2007. Age modelling of 221–235.
late Quaternary marine sequences in the Adriatic: towards Luoto, M., Seppälä, M. 2003. Thermokarst ponds as indicators of the
improved precision and accuracy using volcanic event strati- former distribution of palsas in Finnish Lapland. Permafrost and
graphy. Continental Shelf Research 27, 560–582. Periglacial Processes 14, 19–27.
488 REFERENCES
Luoto, M., Fonzek, S., Zuidhoff, F.S. 2004. Spatial modelling of palsa McCarroll, D., Shakesby, R.A., Matthews, J.A. 1998. Spatial and
mires in relation to climate in northern Europe. Earth Surface temporal patterns of Late Holocene rockfall activity on a Nor-
Processes and Landforms 29, 1373–1387. wegian talus slope: a lichenometric and simulation-modelling
Lusch, D.P., Stanley, K.E., Schaetzl, R.J., et al. 2009. Characterization approach. Arctic and Alpine Research 30, 51–60.
and mapping of patterned ground in the Saginaw Lowlands, McCarroll, D., Knight, J., Rijsdijk, K. (eds). 2001. The glaciation of
Michigan: possible evidence for Late-Wisconsin permafrost. the Irish Sea Basin. Journal of Quaternary Science 16, 391–506.
Annals of the Association of American Geographers 99, 445–466. McCarroll, D., Jalkanen, R., Hicks, S., et al. 2003. Multiproxy
Lydeard, C., Cowie, R.H., Ponder, W.F., et al. 2004. The global dendroclimatology: a pilot study in northern Finland. The
decline of nonmarine mollusks. BioScience 54, 321–330. Holocene 13, 829–838.
Lykke-Andersen, H. 1998. Neogene–Quaternary depositional history McCarthy, L., Head, L. 2001. Holocene variability in semi-arid
of the East Greenland shelf in the vicinity of Leg 152 shelf sites. vegetation: new evidence from Leporillus middens from the
Proceedings of the Ocean Drilling Program, Scientific Results 152, Flinders Ranges, South Australia. The Holocene 11, 681–689.
29–38. McCave, I.N., Hall, I.R., Bianchi, G.G. 2006. Laser vs. settling velocity
Lyman, R.L., Rosania, C.N., Boulanger, M.T. 2012. Comparison of differences in silt grainsize measurements: estimation of
fluoride and direct AMS radiocarbon dating of black bear bone palaeocurrent vigour. Sedimentology 53, 919–928.
from Lawson Cave, Missouri. Journal of Field Archaeology 37, McClymont, E.L., Rosell-Melé, A. 2005. Links between the onset of
226–237. modern Walker circulation and the mid-Pleistocene climate
Lynch-Stieglitz, J., Adkins, J.F., Curry, W.B., et al. 2007. Atlantic transition. Geology 33, 389–392.
meridional overturning circulation during the Last Glacial McClymont, E.L., Sosdian, S.M., Rosell-Melé, A., et al. 2013.
Maximum. Science 316, 66–69. Pleistocene sea-surface temperature evolution: early cooling,
Lyons, S.K. 2005. A quantitative model for assessing community delayed glacial intensification, and implications for the mid-
dynamics of Pleistocene mammals. The American Naturalist
Pleistocene climate transition. Earth-Science Reviews doi:
165, E168–E185.
10.1016/j.earscirev.2013.04.006.
Lyons, S.K., Wagner, P.J., Dzikiewicz, K. 2010. Ecological correlates
McCoy, W.D. 1987. The precision of amino acid geochronology and
of range shifts of Late Pleistocene mammals. Philosophical
palaeothermometry. Quaternary Science Reviews 6, 43–54.
Transactions of the Royal Society of London B 365, 3681–3693.
McCullagh, J.S.O., Marom, A., Hedges, R.E. 2010. Radiocarbon
Ma, Z., Wang, Z., Liu, J., et al. 2004. U-series chronology of sediments
dating of individual amino acids from archaeological bone
associated with Late Quaternary fluctuations, Balikun Lake,
collagen. Radiocarbon 52, 620–634.
northwestern China. Quaternary International 121, 89–98.
McCullough, M.T., Esat, T. 2000. The coral record of last interglacial
Maat, P.B., Johnson, W.C. 1996. Thermoluminescence and new 14C
sea levels and sea surface temperaures. Chemical Geology 169,
age estimates for the late Quaternary loesses in southwestern
Nebraska. Geomorphology 17, 115–128. 107–129.
McAndrews, J.H., Turton, C.L. 2010. Fungal spores record Iroquoian McDermott, F. 2004. Palaeo-climate reconstruction from stable
and Canadian agriculture in 2nd millennium AD sediment of isotope variations in speleothems: a review. Quaternary Science
Crawford Lake, Ontario, Canada. Vegetation History and Reviews 23, 901–918.
Archaeobotany 19, 495–501. McDermott, F., Frisia, S., Huang, Y., et al. 1999. Holocene climatic
MacAyeal, D.R. 1992. Irregular oscillations of the West Antarctic ice variability in Europe: evidence from δ18O and textural variations
sheet. Nature 359, 29–32. in speleothems. Quaternary Science Reviews 18, 1021–1038.
McCabe, G.J., Palecki, M.A., Betancourt, J.L., 2004. Pacific and MacDonald, G.M., Case, R.A. 2005. Variations in the Pacific
Atlantic Ocean influences on multidecadal drought frequency in Decadal Oscillation over the past millennium. Geophysical
the United States. Proceedings of the National Academy of Sciences Research Letters 32.
of the United States of America 101, 4136–4141. MacDonald, G.M., Bennett, K.D., Jackson, S.T., et al. 2008. Impacts
McCabe, M., Clark, P.U., Clark, J., et al. 2007. Radiocarbon of climate change on species, populations and communities:
constraints on readvances of the British–Irish Ice Sheet in the palaeobiogeographical insights and frontiers. Progress in Physical
northern Irish Sea Basin during the last deglaciation. Quaternary Geography 32, 139–72.
Science Reviews 26, 1204–1211. Macdougall, D. 2008. Nature’s Clocks: How Scientists Measure the
McCarroll, D. 2002. Amino-acid geochronology and the British Age of Almost Everything. University of California Press, Berkeley,
Pleistocene: secure stratigraphical framework or a case of Los Angeles and London.
circular reasoning? Journal of Quaternary Science 17, 647–651. McDougall, I., Harrison, T.M. 1999. Geochronology and Thermo-
McCarroll, D.M., Harris, C. 1992. The glacigenic deposits of western chronology by the 40Ar/39Ar Method, 2nd edn. Oxford University
Lleyn, north Wales. Journal of Quaternary Science 7, 19–30. Press, Oxford.
McCarroll, D.M., Ballantyne, C.K. 2000. The last ice sheet in McDougall, I., Brown, F.H., Fleagle, J.G. 2005. Stratigraphic
Snowdonia. Journal of Quaternary Science 15, 765–778. placement and age of modern humans from Kibish, Africa.
McCarroll, D., Rijsdijk, K. 2003. Deformation styles as a key for Nature 433, 733–735.
interpreting glacial depositional sequences. Journal of Quaternary Mcelwain, J.C., Mayle, F.E., Beerling, D.J. 2002. Stomatal evidence
Science 18, 473–489. for a decline in atmospheric CO2 concentration during the
REFERENCES 489
Younger Dryas stadial: a comparison with Antarctic ice core Australia, identifies MIS 3 and MIS 2 glacial advances and
records. Journal of Quaternary Science 17, 21–29. climatic variability. Journal of Quaternary Science 21, 363–376.
McElwhinney, M.W., McFadden, P.L. 2000. Palaeomagnetism: McLaren, S.J., Rowe, P.J. 1996. The reliability of uranium-series
Continents and Oceans. Academic Press, London and San Diego. mollusc dates from the western Mediterranean basin. Quaternary
McFarlane, D.A., Blake, J. 2005. The late Pleistocene hutlas (Geo- Science Reviews 15, 709–717.
capromys brownii) of Red Hills Fissure, Jamaica. Geological MacLeod, A., Palmer, A., Lowe, J., Rose, J., Bryant, C., Merritt, J. 2011.
Journal 40, 3899–404. Timing of glacier response to Younger Dryas climatic cooling in
McGarry, S., Baker, A. 2000. Organic acid fluorescence: applications Scotland. Global and Planetary Change 79, 264–274.
to speleothem palaeoenvironmental reconstruction. Quaternary MacLeod, D.M., Osborn, G., Spooner, I. 2006. A record of post-glacial
Science Reviews 19, 1087–1101. moraine deposition and tephra stratigraphy from Otokomi Lake,
McGarry, S.F., Caseldine, C.J. 2004. Speleothem palynology: an Rose Basin, Glacier National Park, Montana. Canadian Journal
undervalued tool in Quaternary studies. Quaternary Science of Earth Sciences 43, 447–460.
Reviews 23, 2389–2404. Macklin, M., Lewin, J. 2008. Alluvial responses to the changing
McGarry, S., Bar-Matthews, M., Matthews, A., et al. 2004. Constraints Earth system. Earth Surface Processes and Landforms 33,
on hydrological and paleotemperature variations in the Eastern 1374–1395.
Mediterranean region in the last 140 ka given by the δD values Macklin, M., Woodward, J. 2009. River systems and environmental
of speleothem fluid inclusions. Quaternary Science Reviews 23, change. In The Physical Geography of the Mediterranean (edited
919–934. by J. Woodward). Oxford University Press, Oxford, 320–351.
McGee, D., Broecker, W.S., Winckler, G. 2010. Gustiness: The driver McMartin, I., Henderson, P.J. 2004. Evidence from Keewatin (central
of glacial dustiness? Quaternary Science Reviews 29, 2340–2350. Nunavut) for paleo-ice divide migration. Géographie physique et
McGlone, M.S., Wilmshurst, J.M. 1999. A Holocene record of Quaternaire 58, 163–186.
climate, vegetation change and peat bog development, east McMillan, A.A. 2005. A provisional Quaternary and Neogene
Otago, South Island, New Zealand. Journal of Quaternary Science lithostratigraphical framework for Great Britain. Geologie en
14, 239–254. Mijnbouw (Netherlands Journal of Geosciences) 84, 87–107.
McGuffie, K., Henderson-Sellars, A. 2005. A Climate Modelling McMillan, E.A., Fairchild, I.J., Frisia, S., et al. 2005. Annual trace
Primer, 3rd edn. John Wiley, Chichester and New York. element cycles in calcite–aragonite speleothems: evidence of
McGuire, W.J., Howarth, R.J., Firth, C.R., et al. 1997. Correlation drought in the western Mediterranean 1200–1100 yr BP. Journal
between rate of sea-level change and frequency of explosive of Quaternary Science 20, 423–433.
volcanism in the Mediterranean. Nature 389, 473–476 Maddy, D., 1997. Uplift-driven valley incision and river terrace
Machette, M.N., Slate, J.L., Phillips, F.M. 2008. Terrestrial formation in southern England. Journal of Quaternary Science 12,
cosmogenic-nuclide dating of alluvial fans in Death Valley, 539–545.
California, USA. U.S. Geological Survey Professional Paper 1745, Maddy, D., Demir, T., Bridgland, D.R., et al. 2005. An obliquity-
1–45. controlled Early Pleistocene river terrace record from Western
Machida, H. 1999. The stratigraphy, chronology and distribution Turkey? Quaternary Research 63, 339–346.
of distal marker-tephras in and around Japan. Global and Magnani, G., Bartolomei, P., Cavulli, F., et al. 2007. U-series and
Planetary Change 21, 71–94. radiocarbon dates on mollusc shells from the uppermost layer
McIntyre, A., Ruddiman, W.F. 1972. North-east Atlantic post- of the archaeological site of KHB-1, Ra’s al Khabbah, Oman.
Eemian paleoceanography: a predictive analog for the future. Journal of Archaeological Science 34, 749–755.
Quaternary Research 2, 350–354. Magny, B. 2007. Lake level studies: west-central Europe. In
McIntyre, A., Ruddiman, W.F., Jantzen, R. 1972. Southward Encyclopedia of Quaternary Science (edited by S.A. Elias). Elsevier,
penetration of the North Atlantic Polar Front and faunal and Amsterdam, 1389–1399.
floral evidence of large scale surface water mass movements Magny, M. 2001. Palaeohydrological; changes as reflected by lake-
over the past 225,000 years. Deep Sea Research 19, 61–77. level fluctuations in the Swiss Plateau, the Jura Mountains and
Mackay, A.W. 2007. The palaeoclimatology of Lake Baikal: a diatom the northern French Pre-Alps during the Last Glacial–Holocene
synthesis and prospectus. Earth-Science Reviews 82, 181–215. transition: a regional synthesis. Global and Planetary Change 30,
Mackay, A., Battarbee, R.W., Birks, J., et al. (eds). 2003. Global 85–101.
Change in the Holocene. Arnold, London. Magny, M. 2004. Holocene climate variability as reflected by mid-
Mackay, J.R., Burns, C.R. 2002. The first 20 years (1978–1979 to European lake-level fluctuations and its probable impact on
1998–1999) of ice-wedge growth at the Illisarvik experimental prehistoric human settlement. Quaternary International 113,
drained lake site, western Arctic coast, Canada. Canadian Journal 65–79.
of Earth Sciences 39, 95–111. Magny, M., Bégeot, C., Guiot, J., et al. 2003. Reconstruction and
Mackereth, F.J.H. 1965. Chemical investigations of lake sediments palaeoclimatic interpretation of mid-Holocene vegetation and
and their interpretation. Proceedings of the Royal Society B 161, lake-level changes at Saint-Jorioz, Lake Annecy, French Pre-Alps.
295–3090. The Holocene 13, 265–275.
Mackintosh, A.N., Barrows, T.T., Colhoun, E.A., et al. 2006. Exposure Magri, D. 2010. Persistence of tree taxa in Europe and Quaternary
dating and glacial reconstruction at Mt. Field, Tasmania, climate changes. Quaternary International 219, 145–151.
490 REFERENCES
Mahaney, W.C., Kalm, V., Kapran, B., et al. 2009. A soil chrono- Mantua, N., Hare, S.R. 2002. The Pacific Decadal Oscillation. Journal
sequence in Late Glacial and Neoglacial moraines, Humboldt of Oceanography 58, 35–44.
Glacier, northwestern Venezuelan Andes. Geomorphology 109, Marchant, D.R., Lewis, D.R., Phillips, W.M., et al. 2002. Formation
236–245. of patterned ground and sublimation till over Miocene glacier
Maher, B. 2007. Environmental magnetism and climate change. ice in Beacon Valley, southern Victoria Land, Antarctica. Geo-
Contemporary Physics 48, 247–274. logical Society of America Bulletin 114, 718–730.
Maher, B.A., Thompson, R. (eds). 1999. Quaternary Climates, Marchetti, M. 2002. Environmental changes in the central Po Plain
Environments and Magnetism. Cambridge University Press, (northern Italy) due to fluvial modifications and anthropogenic
Cambridge. activities. Geomorphology 44, 361–373.
Maher, B. A., Hallam, D.F. 2005. Palaeomagnetic correlation and Marcolini, F., Bigazzi, G., Bonadonna, F.P., et al. 2003. Tephro-
dating of Plio/Pleistocene sediments at the southern margins chronology and tephrostratigraphy of two Pleistocene conti-
of the North Sea Basin. Journal of Quaternary Science 20, 67–77. nental fossiliferous successions from central Italy. Journal of
Maher, B.A., Prospero, J.M., Mackie, D., et al. 2010. Global connec- Quaternary Science 18, 545–556.
tions between aeolian dust, climate and ocean biogeochemistry Marcott, S.A., Clark, P.U., Padman, L., et al. 2011. Ice-shelf collapse
at the present day and at the last glacial maximum. Earth-Science from subsurface warming as a trigger for Heinrich events.
Reviews 99, 61–97. Proceedings of the National Academy of Sciences of the United States
Major, C.O., Goldstein, S.L., Ryan, W.B.F., et al. 2006. The co- of America 108, 13415–13419.
evolution of Black Sea level and composition through the last Marcott, S.A., Shakun, J.D., Clark, P.U., et al. 2013. A reconstruction
deglaciation and its paleoclimatic significance. Quaternary Science of regional and global temperature for the past 11,300 years.
Reviews 25, 2031–2047. Science 339, 1198–1201.
Maldonado, A., Betancourt, J.L., Latorre, C., et al. 2005. Pollen Margari, V., Skinner, L.C., Tzedakis, P.C., et al. 2010. The nature of
analyses from a 50,000-yr rodent midden series in the south- millennial-scale climate variability during the past two glacial
ern Atacama Desert (25˚30′S). Journal of Quaternary Science 20,
periods. Nature Geoscience 3, 127–131.
493–507.
MARGO Project Members. 2009. Constraints on the magnitude and
Mallick, R., Frank, N. 2002. A new technique for precise uranium-
patterns of ocean cooling at the Last Glacial Maximum. Nature
series dating of travertine micro-samples, Geochimica et Cosmoc-
Geoscience 2, 127–132.
himica Acta 66, 4261–4272.
Marín-Arroyo, A.B., Fosse, P., Vigne, J.-D. 2009. Probable evidence
Mallol, C., Cabanes, D., Baene, J. 2010. Microstratigraphy and
of bone accumulation by Pleistocene bearded vulture at the
diagenesis at the upper Pleistocene site of Esquilleu Cave
archaeological site of El Mirón Cave, Spain. Journal of Archaeo-
(Cantabria, Spain). Quaternary International 214, 70–81.
logical Science 36, 284–296.
Manabe, S., Stouffer, R.J. 1997. Coupled ocean–atmosphere model
response to freshwater input: comparison to Younger Dryas Mark, B.G., Seltzer, G.O., Rodbell, D.T. 2004. Late Quaternary
Event. Paleoceanography 12, 321–336. glaciations of Ecuador, Peru and Bolivia. In Developments in
Mangerud, J. 1977. Late Weichselian marine sediments contain- Quaternary Science. Volume 2: Quaternary Glaciations: Extent
ing shells, Foraminifera and pollen at Agotnes, western Norway. and Chronology (edited by J. Ehlers, P.L. Gibbard). Elsevier,
Norsk Geologiske Tidsskrift 57, 23–54. Amsterdam, 151–163.
Mangerud, J., Andersen, S.T., Berglund, B.E., et al. 1974. Quaternary Mark, D.F., Petraglia, M., Smith, V.C. 2013. A high-precision
40
stratigraphy of Norden, a proposal for terminology and classifi- Ar/39Ar age of the Young Toban Tuff and dating of ultra-
cation. Boreas 3, 109–128. distal tephra: forcing of Quaternary climate and implications for
Mangerud, J., Jakobsson, M., Alexanderson, H., et al. 2004. Ice- hominin occupation of India. Quaternary Geochronology doi:
dammed lakes and rerouting of the drainage of northern Eurasia 10.1016/j.quageo.2012.12.004.
during the Last Glaciation. Quaternary Science Reviews 23, Markova, A.K. 2005. Eastern European rodent (Rodentia, mam-
1313–1332. malia) faunas from the Early–Middle Pleistocene transition.
Mangini, A., Spötl, C., Verdes, P. 2005. Reconstruction of tem- Quaternary International 131, 71–77.
perature in the Central Alps during the past 2000 yr from a δ18O Marković, S.B., Bokhorst, M.P., et al. 2007. Late Pleistocene loess–
stalagmite record. Earth and Planetary Science Letters 235, palaeosol sequences in the Vojvodina region, north Serbia.
741–751. Journal of Quaternary Science 23, 73–84.
Mann, M.E., Bradley, R.S., Hughes, M.K. 1998. Global-scale tem- Marks, L. 2012. Timing of the Late Vistulian (Weichselian) glacial
perature patterns and climate forcing over the past six centuries. phases in Poland. Quaternary Science Reviews 44, 81–88.
Nature 392, 779–787. Marotzke, J. 2000. Abrupt climate change and thermohaline
Mannes, D., Lehmann, E., Cheubini, P., et al. 2007. Neutron imag- circulation: mechanisms and predictability. Proceedings of the
ing versus standard X-ray densitrometry as method to measure National Academy of Sciences of the United States of America 97,
tree-ring wood density. Trees – Structure and Function 21, 1347–1350.
605–612. Marquardt, C., Lavenu, A., Ortlieb, L., et al. 2004. Coastal
Manning, A., Birley, R., Tipping, R. 2007. Roman impact on the neotectonics in Southern Central Andes: uplift and deforma-
environment at Hadrian’s Wall: precisely dated pollen analysis tion of marine terraces in Northern Chile (27°S). Tectonophysics
from Vindolanda, northern England. The Holocene 17, 527–538. 394, 193–219.
REFERENCES 491
Marquez-Grant, N., Fibiger, L. (eds). 2011. The Routledge Handbook Maslin, M., Seidov, D., Lowe, J.J. 2001. Synthesis of the nature and
of Archaeological Human Remains and Legislation. An inter- causes of rapid climate transitions during the Quaternary. In
national guide to laws and practice in the excavation and treatment The Oceans and Rapid Climatic Change: Past, Present and Future
of archaeological human remains. Routledge, London. (edited by D. Seidov, M. Maslin, B.J. Haupt). American
Marra, F., Karner, D.B., Freda, C., et al. 2009. Large mafic eruptions Geophysical Union, Geophysical Monograph 126, 9–52.
at Alban Hills Volcanic District (Central Italy): chronostrati- Mason, B.G., Pyle, D.M., Oppenheimer, C. 2004. The size and
graphy, petrography and eruptive behaviour. Journal of Volcan- frequency of the largest explosive eruptions on Earth. Bulletin of
ology and Geothermal Research 179, 217–232. Vulcanology 66, 735–478.
Marra, M.J., Leschen, R.A.B. 2011. Persistence of New Zealand Mason, J.A. 2001. Transport direction of Peoria loess in Nebraska
Quaternary beetles. New Zealand Journal of Geology and and implications for loess sources on the central Great Plains.
Geophysics 54, 403–413. Quaternary Research 56, 79–86.
Marret, F., Scourse, J. 2002. Control of modern dinoflagellate cyst Massari, F., Rio, D., Serandrei Barbero, R., et al. 2004. The environ-
distribution in the Irish and Celtic seas by seasonal stratification
ment of Venice area in the past two million years. Palaeo-
dynamics. Marine Micropaleontology 47, 101–116.
geography, Palaeoclimatology, Palaeoecology 202, 273–308.
Marret, F., Zonneveld, K.A.F. 2003. Atlas of modern organic-walled
Masson-Delmotte, V., Stenni, B., Jouzel, J. 2004. Common
dinoflagellate cyst distribution. Review of Palaeobotany and
millennial-scale variability of Antarctic and Southern Ocean
Palynology 125, 1–200.
Marshall, S.J. 2005. Recent advances in understanding ice sheet temperatures during the past 5000 years reconstructed from the
dynamics. Earth and Planetary Science Letters 240, 191–204. EPICA Dome C ice core. The Holocene 14, 145–151.
Marshall, S.J., James, T.S., Clarke, G.K.C. 2002. North American Ice Masson-Delmotte, V., Jouzel J., Landais A., et al. 2005. GRIP
Sheet reconstructions at the Last Glacial Maximum. Quaternary deuterium excess reveals rapid and orbital-scale changes in
Science Reviews 21, 175–192. Greenland moisture origin. Science 209, 118–121.
Marshall, W.A., Chalmers, M.O. 1997. Airborne dispersal of Antarctic Masson-Delmotte, V., Stenni, B., Pol, K., et al. 2010. Epica Dome C
terrestrial algae and cyanobacteria. Ecography 20, 585–594. record of glacial and interglacial intensities. Quaternary Science
Martens, K., Schön, I., Meisch, C., et al. 2008. Global diversity of Reviews 29, 113–128.
ostracods (Ostracoda, Crustacea) in freshwater. Hydrobiologia Mastronuzzi, G., Sanso, P., Murray-Wallace, C.V., et al. 2005.
595, 185–193. Quaternary coastal morphology and sea-level changes – an
Martin, P., Archer, D., Lea, D.W. 2005. Role of deep sea temperature introduction. Quaternary Science Reviews 24, 1963–1968.
in the carbon cycle during the last glacial. Paleoceanography Matsuoka, N. 2011. Climate and material controls on periglacial soil
20, PA2015, doi: 10.1029/2003PA000914. processes: toward improving periglacial climate indicators.
Martínez-Cortizas, A., Mighall, T., Pontevedra-Pombal, X., et al. 2005. Quaternary Research 75, 356–365.
Linking changes in atmospheric dust deposition, vegetation Matthews, J.A. 1992. The Ecology of Recently Deglaciated Terrain.
change and human activities in northwest Spain during the last Cambridge University Press, Cambridge.
5300 years. The Holocene 15, 698–706. Matthews, J.A. 2005. ‘Little Ice Age’ glacier variations in Jotunheimen,
Martínez-Cortizas, A., Pontevedra-Pombal, X., Nóvoa Muñoz, J.C. southern Norway: a study in regionally controlled lichenometric
et al. 1997. Four thousand years of atmospheric Pb, Cd, and dating of recessional moraines with implications for climate
Zn deposition recorded by the ombrotrophic peat bog of Penido and lichen growth. The Holocene 15, 1–29.
Vello (Northwestern Spain). Water, Air & Soil Pollution 100, Matthews, J.A., Dresser, P.Q. 2008. Holocene glacier variation
387–403. chronology of the Smørstabbtindan massif, Jotunheimen,
Martinez-Garcia, A., Rosell-Mele, A., Jaccard, S.L., et al. 2011.
southern Norway, and the recognition of century- to millennial-
Southern Ocean dust–climate coupling over the past four
scale European Neoglacial Events. The Holocene 18, 181–201.
million years. Nature 476, 312–315.
Matthiessen, J., Knies, J., Nowaczyk, N.R., et al. 2001. Late Quaternary
Martini, I.P., French, H.M., Alberti, A.P. (eds). 2011. Ice-Marginal
dinoflagellate cyst stratigraphy at the Eurasian continental
and Periglacial Processes and Sediments. Geological Society of
margin, Arctic Ocean: indications for Atlantic water inflow in the
London, Special Publication 354. Geological Society, London.
Martinson, D.G., Pisias, N.G., Hays, J.D., et al. 1987. Age dating and past 150,000 years. Global and Planetary Change 31, 65–86.
the orbital theory of ice ages: development of a high-resolution Maul, L.C., Parfitt, S.A. 2010. Micromammals from the 1995
0–300 000 year chronostratigraphy. Quaternary Research 27, Mammoth Excavation at West Runton, Norfolk, UK: mor-
1–29. phometric data, biostratigraphy and taxonomic reappraisal.
Maslin, M.A., Ridgwell, A.J. 2005. Mid-Pleistocene Revolution and Quaternary International 228, 91–115.
the ‘eccentricity myth’. In Early–Middle Pleistocene Transitions: Mauquoy, D., Barber, K.E., 1999a. A replicated 3000 year proxy-
The Land–Ocean Evidence (edited by M.J. Head, P.L. Gibbard). record from Coom Rigg Moss and Felecia Moss, the Border
Geological Society of London, Special Publication 247. Geological Mires, northern England. Journal of Quaternary Science 14,
Society, Bath, 19–34. 263–275.
Maslin, M.A., Li, X.S., Loutre, M.F., et al. 1998. The contribution of Mauquoy, D., van Geel, B., Blaauw, M., et al. 2004. Changes in solar
orbital forcing to the progressive intensification of Northern activity and Holocene climatic shifts derived from 14C wiggle-
Hemisphere glaciation. Quaternary Science Reviews 17, 411–426. match dated peat deposits. The Holocene 14, 45–52.
492 REFERENCES
Mauquoy, D., Hughes, P.D.M., van Geel, B. 2010. A protocol for individual to ecosystem: a review of recent developments and
plant macrofossil analysis of peat deposits. Mires and Peat 7 future potential. In Developments in Quaternary Science. Volume
(art. 6), 1–5. 17: Ostracoda as Proxies for Quaternary Climate Change (edited
Mauz, B., Felix-Henningsen, P. 2005. Palaeosols in Saharan and by D.J. Horne, J.A. Holmes, J. Rodriguez-Lazaro, et al.). Elsevier,
Sahelian dunes of Chad: archives of Holocene and North African Amsterdam, 15–35.
climate changes. The Holocene 15, 453–458. Meyer, E., Sarna-Wojcicki, A.M., Hillhouse, J.W., et al. 1991. Fission-
Maxwell, T.A., Issawi, B., Haynes Jr., C.V. 2010. Evidence for track age (400,000 yr) of the Rockland tephra, based on inclusions
Pleistocene lakes in the Tushka region, south Egypt. Geology 38, of zircon grains lacking fossil fission tracks. Quaternary Research
1135–1138. 39, 54–162.
Mayewski, P.A., Rohling, E.E., Stager, J.C., et al. 2004. Holocene Meyer, M.C., Spötl, C., Mangini, A. 2008. The demise of the Last
climate variability. Quaternary Research 62, 243–255. Interglacial recorded in isotopically dated speleothems from the
Meese, D.A., Gow, A.J., Alley, R.B., et al. 1997. The Greenland Ice Alps. Quaternary Science Reviews 27, 476–496.
Sheet Project 2 depth-age scale: methods and results. Journal of Meyers, P.A. 2006. Paleoceanographic and paleoclimatic similar-
Geophysical Research 102, 26411–26423. ities between Mediterranean sapropels and Cretaceous black
Megaloudi, F. 2005. Burnt sacrificial plant offerings in Hellenistic shales. Palaeogeography, Palaeoclimatology, Palaeoecology 235,
times: an archaeobotanical case study from Messene, Pelopon- 305–320.
nese, Greece. Vegetation History and Archaeobotany 14, 329–340. Meyers, P.A., Teranes, J.J. 2001. Sediment organic matter. In Tracking
Meijer, T., Preece, R.C. 1995. Malacological evidence relating to the Environmental Changes Using Lake Sediments. Volume II: Physical
insularity of the British Isles during the Quaternary. In Island and Chemical Techniques (edited by W.M. Last, J.P. Smol).
Britain: A Quaternary Perspective (edited by R.C. Preece). Kluwer, Dordrecht, 239–269.
Geological Society of London, Special Publication 96. Geological Meyers, S.R., Sageman, B.B., Pagani, M. 2008. Resolving Milank-
Society, London, 89–110. ovitch: consideration of signal and noise. American Journal of
Meissner, K.J., Eby, M., Weaver, A.J., et al. 2007. CO2 threshold for Science 308, 770–786.
millennial-scale oscillations in the climate system: implications Meyrick, R. 2003. Holocene Molluscan faunal history and
for global warming scenarios. Climate Dynamics 30, 161–174. environmental change at Kloster Mühle, Rheinland-Pfalz,
Meland, M.Y., Jansen, E., Elderfield, H. 2005. Constraints on SST western Germany. Journal of Quaternary Science 18, 121–132.
estimates for the northern North Atlantic/Nordic Seas during the Meyrick, R.A., Karrow, P.F. 2007. Three detailed, radiocarbon-
LGM. Quaternary Science Reviews 24, 835–852. dated, Holocene tufa and alluvial fan mollusc successions
Menounos, B., Clague, J.J., Osborn, G., et al. 2008. Western Canadian from southern Ontario: the first in north-eastern North
glaciers advance in concert with climate change circa 4.2 ka. America. Palaeogeography, Palaeoclimatology, Palaeoecology 243,
Geophysical Research Letters 35, doi: 10.1029/2008GL033172. 250–271.
Menzies, J., Zaniewski, K. 2003. Microstructures within a modern Meyrick, R.A., Preece, R.C. 2001. Molluscan successions from
debris flow deposit derived from Quaternary glacial diamicton two Holocene tufas near Northampton, English Midlands. Boreas
– a comparative micromorphological study. Sedimentary Geology 28, 77–93.
157, 31–48. Miall, A.D. 2000. Principles of Sedimentary Basin Analysis, 3rd edn.
Menzies, J., van der Meer, J.J.M., Rose, J. 2006. Till – as a glacial Springer, Berlin.
‘tectomict’, its internal architecture, and the development of a Miallier, D., Condomines, M., Pilleyre, T., et al. 2004. Concordant
‘typing’ method for till differentiation. Geomorphology 75, thermoluminescence and 238U–230Th ages for a trachytic dome
172–200. (Grand Sarcoui) from the Chaîne des Puys (French Massif
Merritt, D. 2007. Fluvial environments: terrace sequences. In Central). Quaternary Science Reviews 23, 709–715.
Encyclopedia of Quaternary Science (edited by S.A. Elias). Elsevier, Mighall, T.M., Timberlake, S., Clark, S.H.E., et al. 2002. A palaeo-
Amsterdam, 694–704. environmental investigation of sediments from the prehistoric
Mertens, K.N.J.M., Lynn, M., Aycard, M., et al. 2009. Coccolitho- mine of Copa Hill, Cwmystwyth, mid-Wales. Journal of Archaeo-
phores as palaeoecological indicators for shifts of the ITCZ in logical Science 29, 1161–1188.
the Cariaco Basin during the late Quaternary. Journal of Migowski, C., Agnon, A., Bookman, R., et al. 2004. Recurrence
Quaternary Science 24, 159–174. pattern of Holocene earthquakes along the Dead Sea trans-
Mesolella, K.J., Matthews, R.K., Broecker, W.S., et al. 1969. The form revealed by varve counting and radiocarbon dating of
astronomical theory of climatic change: Barbados data. Journal lacustrine sediments. Earth and Planetary Science Letters 222,
of Geology 77, 250–277. 301–314.
Mesquita, F., Roca, J.R., Reed, J.M., et al. 2005. Quantifying Milker, Y., Schmiedl, G., Betzler, C. 2011. Paleobathymetric
species–environment relationships in non-marine Ostracoda history of the Western Mediterranean Sea shelf during the latest
for ecological and palaeoecological studies: examples using glacial period and the Holocene: quantitative reconstructions
Iberian data. Palaeogeography, Palaeoclimatology, Palaeoecology based on foraminiferal transfer functions. Palaeogeography,
225, 93–117. Palaeoclimatology, Palaeoecology 307, 324–338.
Mesquita-Joanes, F., Smith, A.J., Viehberg, F.A. 2012. The ecology Millard, A.R. 2006. Bayesian analysis of Pleistocene chronometric
of Ostracoda across levels of biological organization from methods. Archaeometry 48, 359–375.
REFERENCES 493
Millard, A.R., Hedges, R.E.M. 1996. A diffusion–adsorption model Mills, H.H. 2005. Relative-age dating of transported regolith and
of uranium uptake by archaeological bone. Geochimica et Cosmo- application to study of landform evolution in the Appalachians.
chimica Acta 60, 2139–2152. Geomorphology 67, 63–96.
Miller, B.B., Tevesz, M.J.S. 2001. Freshwater molluscs. In Tracking Milne, G.A., Mitrovica, J.X. 2008. Searching for eustasy in deglacial
Environmental Change Using Lake Sediments. Volume 4: Zoolog- sea-level histories. Quaternary Science Reviews 27, 2292–2302.
ical Indicators (edited by J.P. Smol, H.J.B. Birks, W.M. Last). Milne, G.A., Gehrels, W.R., Hughes, C.W., et al. 2009. Identifying
Kluwer, Dordrecht, 153–171. the causes of sea-level change. Nature Geoscience 2, 471–478.
Miller, G.H., Clarke, S.J. 2007. Amino-acid dating. In Encyclopedia Milner, N., Conneller, C., Elliott, B., et al. 2011. From riches to rags:
of Quaternary Science (edited by S.A. Elias). Elsevier, Amsterdam, organic deterioration at Star Carr. Journal of Archaeological
41–52. Science 38, 2818–2832.
Miller, G.H., Beaumont, P.B., Jull, A.J.T., et al. 1992. Pleistocene Mishra, S., White, M.J., Beaumont, P., et al. 2007. Fluvial deposits
geochronology and palaeothermometry from protein dia- as an archive of early human activity. Quaternary Science Reviews
genesis in ostrich eggshells: implications for the evolution of 26, 2996–3016.
modern humans. Philosophical Transactions of the Royal Society Mithen, S., Reed, M. 2002. Stepping out: a computer simulation of
of London B 337, 149–157. hominid dispersal from Africa. Journal of Human Evolution 43,
Miller, G.H., Magee, J.W., Jull, A.J.T. 1997. Low-latitude glacial 433–462.
cooling in the Southern Hemisphere from amino acids in emu Mitrovica, J.X. 2003. Recent controversies in predicting post-glacial
eggshells. Nature 385, 241–244. sea-level change. Quaternary Science Reviews 22, 127–133.
Miller, G.H., Beaumont, P.B., Deacon, H.J., et al. 1999a. Earliest Mitrovica, J.X., Peltier, W.R. 1991. On postglacial geoid subsidence
modern humans in southern Africa dated by isoleucine epimer- over the equatorial oceans. Journal of Geophysical Research, 96,
ization in ostrich eggshell. Quaternary Science Reviews 18, 20053–20071.
1537–1548. Miura, O. 1987. Vertical migration of pollen in soils. Ecological
Miller, G.H., Magee, J.W., Johnson, B.J., et al. 1999b. Pleistocene Review 21, 125–132.
extinction of Genyornis newtoni: human impact on Australian Mix, A.C., Bard, E., Schneider, R. 2001. Environmental processes of
megafauna. Science 283, 205–208. the ice age: land, oceans, glaciers (EPILOG). Quaternary Science
Miller, G.H., Fogel, M.L., Magee, J.W., et al. 2005. Ecosystem collapse Reviews 20, 627–657.
in Pleistocene Australia and a human role in megafaunal Miyahara, H., Masuda, K., Nagaya, K., et al. 2007. Variation of solar
extinction. Geology 309, 287–290. activity from the Spoerer to the Maunder minima indicated
Miller, G.H., Geirsdottir, A., Zhong, Y., et al. 2012. Abrupt onset by radiocarbon content in tree-rings. Advances in Space Research
of the Little Ice Age triggered by volcanism and sustained by 40, 1060–1063.
sea-ice/ocean feedbacks. Geophysical Research Letters 39, doi: Miyoshi, N., Fujiki, T., Morita, Y. 1999. Palynology of a 250-m core
10.1029/2011/GL050168. from Lake Biwa: a 430,000-year record of glacial–interglacial
Miller, K.G., Kominz, M.A., Browning, J.V., et al. 2005. The vegetation change in Japan. Review of Palaeobotany and Palyn-
Phanerozoic record of global sea-level change. Science 310, ology 104, 267–283.
1293–1298. Moine, O., Rousseau, D.-D., Jolly, D., et al. 2002. Paleoclimatic
Miller, K.G., Mountain, G.S., Wright, J.D., et al. 2011. A 180-million- reconstruction using Mutual Climatic Range on terrestrial
year record of sea level and ice volume variations from mollusks. Quaternary Research 57, 162–172.
continental margins and deep-sea isotopic records. Ocean- Molinaroli, E., De Falco, G., Rabitti, S., et al. 2000. Stream-scanning
ography 24, 40–53. laser system, electric sensing counter and settling grain size
Miller, R.F., Elias, S.A. 2000. Late-glacial climate in the Maritimes analysis: a comparison using reference materials and marine
Region, Canada, reconstructed from mutual climatic range sediments. Sedimentary Geology 130, 269–281.
analysis of fossil Coleoptera. Boreas 29, 79–88. Mollard, J.D. 2000. Ice-shaped ring-forms in Western Canada: their
Miller, W., Schuster, S.C., Welch, A.J., et al. 2013. Polar and brown airphoto expressions and manifold polygenetic origins.
bear genomes reveal ancient admixture and demographic Quaternary International 68–71, 187–198.
footprints of past climate change. Proceedings of the National Molloy, K., O’Connell, M. 2004. Holocene vegetation and land-
Academy of Sciences of the United States of America, doi: use dynamics in the karstic environment of Inis Oírr, Aran
101073/pnas.1210506109. Islands, western Ireland: pollen analytical evidence evaluated
Millien, V., Lyons, S.K., Olson, L., et al. 2006. Ecotypic variation in in light of the archaeological record. Quaternary International
the context of global climate change: revisiting the rules. Ecology 113, 41–64.
Letters 9, 853–869. Montaggioni, L.F., Braithwaite, C.J.R. 2009. Developments in
Millière, L., Spangenberg, J.E., Bindschedler, S., et al. 2011. Reliability Marine Geology. Volume 5: Quaternary Coral Reef Systems:
of stable carbon and oxygen isotope compositions of pedo- History, Development Processes and Controlling Factors. Elsevier
genic needle fibre calcite as environmental indicators: examples Science, Burlington, MA.
from Western Europe. Isotopes and Environmental Health Studies Mooers, H.D. 1989. Drumlin formation: a time-transgressive model.
47, 341–358. Boreas 18, 99–107.
494 REFERENCES
Mooney, S.D., Harrison, S.P., Bartlein, P.J., et al. 2011. Late assemblages, morphology and paleosalinity. Marine Micro-
Quaternary fire regimes of Australasia. Quaternary Science paleontology 43, 155–178.
Reviews 30, 28–46. Muhs, D.R. 2013. The geologic records of dust in the Quaternary.
Moore, A.E., Cotterill, F.P.D. (Woody), Eckartd, F.D. 2012. The Aeolian Research 9, 3–48.
evolution and ages of Makgadikgadi Palaeo-lakes: consilient Muhs, D.R., Bettis III, E.A. 2000. Geochemical variations in Peoria
evidence from Kalahari drainage evolution South-central Africa. Loess of western Iowa indicate Last Glacial paleowinds of mid-
South African Journal of Geology 1156, 385–423. continental North America. Quaternary Research 53, 49–61.
Moore, P.D., Webb, J.A., Collinson, M.D. 1991. Pollen Analysis. Muhs, D.R., Roskin, J., Tsoar, H., et al. 2013. Origin of the Sinai–
Blackwell, Oxford. Negev erg, Egypt and Israel: mineralogical and geochemical
Moos, M.T., Laird, K.R., Cumming, B.F. 2009. Climate-related evidence for the importance of the Nile and sea level history.
eutrophication of a small boreal lake in northwestern Ontario: Quaternary Science Reviews 69, 28–48.
a palaeolimnological perspective. The Holocene 19, 359–367. Muller, R.A., MacDonald, G.J. 1997. Glacial cycles and astronomical
Moreno, A., Cacho, I., Canals, M., et al. 2005. Links between marine forcing. Science 277, 215–218.
and atmospheric processes oscillating on a millennial time- Müller, R.D., Sdrolius, M., Gaina, C., et al. 2008. Long-term sea-level
scale. A multi-proxy study of the last 50,000 yr from the Alboran fluctuations driven by ocean basin dynamics. Science 319,
Sea (western Mediterranean Sea). Quaternary Science Reviews 24, 1357–1362.
1623–1636. Müller, U.C., Klotz, S., Geyh, M.A., et al. 2005. Cyclic climate
Morgenstern, U., Fifield, L.K., Zondervan, A. 2000. New frontiers fluctuations during the last interglacial in central Europe. Geology
in glacial ice dating: measurement of natural 32Si by AMS. 33, 449–452.
Nuclear Instruments and Methods B172, 605–609. Munroe, J.S., Mickelson, D.M. 2002. Last Glacial Maximum
Morgenstern, U., Geyh, M.A., Kudrass, H.R., et al. 2001. 32Si equilibrium-line altitudes and paleoclimate, northern Uinta
dating of marine sediments from Bangladesh. Radiocarbon 43, Mountains, Utah, USA. Journal of Glaciology 48, 257–266.
909–916.
Murray, J.W. 2006. Ecology and Applications of Benthic Foraminifera.
Mörner, N.-A. 1980. Eustasy and geoid changes as a function of
Cambridge University Press, Cambridge.
core/mantle changes. In Earth Rheology, Isostasy and Eustasy
Murray, J.W., Alve, E. 1999. Natural dissolution of modern shallow
(edited by N.-A. Mörner). John Wiley, Chichester and New
water benthic foraminifera: taphonomic effects on the palaeo-
York, 535–553.
ecological record. Palaeogeography, Palaeoclimatology, Palaeo-
Moros, M., De Dekker, P., Jansen, E., et al. 2009. Holocene climatic
ecology 146, 195–209.
variability in the Southern Ocean recorded in a deep-sea sedi-
Murray-Wallace, C.V. 2002. Pleistocene coastal stratigraphy, sea-
ment core off South Australia. Quaternary Science Reviews 28,
level highstands and neotectonism of the southern Australian
1932–1940.
passive continental margin – a review. Journal of Quaternary
Morrill, C., Overpeck, J.T., Cole, J.E., et al. 2006. Holocene vari-
ations in the Asian monsoon inferred from the geochemistry Science 17, 469–489.
of lake sediments in central Tibet. Quaternary Research 65, Murray-Wallace, C.V., Bouman, R.P., Prescott, J.R., et al. 2010.
232–243. Aminostratigraphy and thermoluminescence dating of coastal
Mortensen, A.K., Bigler, M., Grönvold, K., et al. 2005. Volcanic ash aeolianites and the later Quaternary history of a failed delta:
layers from the Last Glacial termination in the NGRIP ice core. the River Murray mouth region, South Australia. Quaternary
Journal of Quaternary Science 20, 209–219. Geochronology 5, 28–49.
Morwood, M.J., O’Sullivan, P.B., Aziz, E., et al. 1998. Fission-track Murray-Wallace, C.V., Woodroffe, C.D. 2014. Quaternary Sea-Level
ages of stone tools and fossils on the east Indonesian island of Changes: A Global Perspective. Cambridge University Press,
Flores. Nature 392, 173–176. Cambridge.
Moucha, R., Forte, A.M., Mitrovica, J.X., et al. 2008. Dynamic Murton, J.B., Worsley, P., Gozdzik, J. 2000. Sand veins and wedges
topography and long-term sea-level variations: There may be no in cold aeolian environments. Quaternary Science Reviews 19,
such thing as a stable continental platform. Earth and Planetary 899–922.
Science Letters 271, 101–108. Muscheler, R., Beer, J., Qwagner, G., et al. 2004. Changes in the carbon
Mroczek, P. 2013. Recycled loess – a micromorphological approach cycle during the last deglaciation as indicated by the comparison
to the determination of local source areas of loess. Quaternary of Be-10 and C-14 records. Earth and Planetary Science Letters
International 296, 241–250. 219, 325–340.
Mudelsee, M., Schultz, M. 1997. The Mid-Pleistocene climate Muscheler, R., Beer, J., Kubik, P.W., et al. 2005. Geomagnetic field
transition: onset of the 100 ka cycle lags ice volume build-up by intensity during the last 60,000 years based on 10Be and 36Cl
280 ka. Earth and Planetary Science Letters 151, 117–123. from the Summit ice cores and 14C. Quaternary Science Reviews
Mudelsee, M., Raymo, M.E. 2005. Slow dynamics of the Northern 24, 1849–1860.
Hemisphere glaciation. Paleoceanography 20, PA4022, doi: 10. Nádor, A., Lantos, M., Tóth-Makk, Á., et al. 2003. Milankovitch-scale
1029/2005/PA001153. multi-proxy records from fluvial sediments of the last 2.6 Ma,
Mudie, P.J., Aksu, A.E., Yasar, D. 2001. Late Quaternary dinoflagellate Pannonian Basin, Hungary. Quaternary Science Reviews 22,
cysts from the Black, Marmara and Aegean seas: variations in 2157–2175.
REFERENCES 495
Naegler, T., Levin, I. 2006. Closing the global radiocarbon budget NEEM Community Members. 2013. Eemian interglacial recon-
1945–2005. Journal of Geophysical Research 111, D12311. structed from a Greenland folded ice core. Nature 493, 489–494.
Naidu, P.D., Govil, P. 2010. New evidence on the sequence of Nelson, K.J.P., Nelson, F.E., Walegur, M.T. 2007. Periglacial
deglacial warming in the tropical Indian Ocean. Journal of Appalachia: paleoclimatic significance of blockfield elevation
Quaternary Science 25, 1138–1143. gradients, eastern USA. Permafrost and Periglacial Processes 18,
Nakagawa, T., Kitagawa, H., Yasuda, Y., et al. 2005. Pollen/event 61–73.
stratigraphy of the varved sediment of Lake Suigetsu, central Japan Nelson, S.T., Wood, M.J., Mayo, A.L., et al. 2005. Shoreline tufa and
from 15,701 to 10,217 SG vyr (Suigetsu varve years before tufaglomerate from Pleistocene Lake Bonneville, Utah, USA:
present): description, interpretation and correlation with other stable isotopic and mineralogical records of lake conditions,
regions. Quaternary Science Reviews 24, 1691–1701. processes, and climate. Journal of Quaternary Science 20, 3–19.
Nakagawa, T., Nakagawa, T., Gotanda, K., et al. 2012. SG06, a fully Nesje, A., Sejrup, H.-P. 1988. Late Weichselian/Devensian ice sheets
continuous and varved sediment core from Lake Suigetsu, Japan: in the North Sea and adjacent land areas. Boreas 17, 371–384.
stratigraphy and potential for improving the radiocarbon Nesje, A., Dahl, S.O. 2003. The ‘Little Ice Age – only temperature?
calibration model and understanding late Quaternary climate The Holocene 13, 39–145.
changes. Quaternary Science Reviews 36, 164–176. Nesje, A., Dahl., S.O., Andersson, C., et al. 2000. The lacustrine
Namiotko, T., Danielopol, D.L., Baltanás, A. 2011. Soft body sedimentary sequence in Sygneskardvatnet, western Norway: a
morphology, dissection and slide preparation of Ostracoda: continuous high-resolution record of the Jostedalsbreen ice cap
a primer. Joannea-Geologie und Paläontologie 11, 327–343. during the Holocene. Quaternary Science Reviews 19, 1047–1065.
Nanson, G.C., Jones, B.J., Price, D.M., et al. 2005. Rivers turned to Nesje, A., Matthews, J.A., Dahl, S.O., et al. 2001. Holocene glacier
rock: Late Quaternary alluvial induration influencing the fluctuations of Flatebreen and winter-precipitation changes in
behaviour and morphology of an anabranching river in the the Jostedalsbreen region, western Norway, based on glacio-
Australian monsoon tropics. Geomorphology 70, 398–420.
lacustrine sediment records. The Holocene 11, 267–280.
Nanson, G.C., Price, D.M., Jones, B.G., et al. 2008. Alluvial evidence
Nesje, A., Bakke, J., Dahl., S.O., et al. 2008. Norwegian mountain
for major climate and flow regime changes during the middle and
glaciers in the past, present and future. Global and Planetary
late Quaternary in eastern central Australia. Geomorphology 101,
Change 60, 10–27.
109–129.
Nevalainen, L., Sarmaja-Korjonen, K., Luoto, T.P. 2011. Sedimentary
Narcisi, B., Petit, J.R., Delmonte, B., et al. 2005. Characteristics and
Cladocera as indicators of past water-level changes in shallow
sources of tephra layers in the EPICA-Dome C ice record (East
northern lakes. Quaternary Research 75, 430–437.
Antarctica): implications for past atmospheric circulation and
Newman, M., Compo, G.P., Alexander, M. 2003. ENSO-forced
ice core stratigraphic correlations. Earth and Planetary Science
variability of the Pacific Decadal oscillation. Journal of Climate
Letters 239, 253–265.
Narcisi, B., Petit, J.R., Delmonte, B. 2010. Extended East Antarctic 16, 3853–3857.
ice-core tephrostratigraphy. Quaternary Science Reviews 29, Newnham, R.M., Eden, D.N., Lowe, D.J., et al. 2003. Rerewhakaaitu
21–27. Tephra, a land–sea marker for the last termination in New
Nash, D.J., Thomas, D.S.G., Shaw, P.A. 1994. Silicious duricrusts as Zealand, with implications for global climate change. Quaternary
palaeoclimatic indicators: evidence from the Kalahari desert of Science Reviews 22, 289–308.
Botswana. Palaeogeography, Palaeoclimatology, Palaeoecology Ng, F.S.L., Barr, I.D., Clark, C.D. 2010. Using the surface profiles of
112, 279–295. modern ice masses to inform palaeo-glacier reconstructions.
Nathan, R., Casagrandi, R. 2004. A simple mechanistic model of seed Quaternary Science Reviews 29, 3240–3255.
dispersal, predation and plant establishment: Janzen-Connell and Nichols, J.E., Booth, R.K., Jackson, S.T., et al. 2006. Paleohydrologic
beyond. Journal of Ecology 92, 733–746. reconstruction based on n-alkane distributions in ombrotrophic
Navarro, C., Carrión, J.S., Navarro, J., et al. 2000. An experimental peat. Organic Geochemistry 37, 1505–1513.
approach to the palynology of cave deposits. Journal of Nichols, J.E., Booth, R.K., Jackson, S.T., et al. 2010. Differential
Quaternary Science, 15, 603–619. hydrogen isotopic ratios of Sphagnum and vascular plant
Nayling, N., McGrail, S. 2004. The Barland’s Farm Romano-Celtic biomarkers in ombrotrophic peatlands as a quantitative proxy
Boat. CBA Research Report 138, Council for British Archaeology, for precipitation–evaporation balance. Geochimica et Cosmo-
York. chimica Acta 74, 1407–1416.
Naysmith, P., Cook, G.T., Phillips, M., et al. 2004. Preliminary Nicoll, K. 2001. Radiocarbon chronologies for prehistoric human
results for the extraction and measurement of in situ 14C from occupation and hydroclimatic change in Egypt and Northern
quartz. Radiocarbon 46, 210–208. Sudan. Geoarchaeology 16, 47–64.
Neal, A. 2004. Ground-penetrating radar and its use in Nicoll, K. 2004. Recent environmental change and prehistoric human
sedimentology: principles, problems and progress. Earth-Science activity in Egypt and Northern Sudan. Quaternary Science Reviews
Reviews 66, 261–330. 23, 561–580.
Nederbragt, A.J., Thurow, J. 2005. Geographic coherence of Nielsen, A.B., Odgaard, B.V. 2004. The use of historical analogues
millennial-scale climatic cycles during the Holocene. Palaeo- for interpreting fossil pollen records. Vegetation History and
geography, Palaeoclimatology, Palaeoecology 221, 313–324. Archaeobotany,13, 33–43.
496 REFERENCES
Nielsen-Marsh, C., Hedges, R.E.M. 2000. Patterns of diagenesis in Ó Cofaigh, C., Dowdeswell, J.A. 2001. Laminated sediments in
bone 1: the effects of site environments. Journal of Archaeological glacimarine environments: diagnostic criteria for their inter-
Science 27, 1139–1150. pretation. Quaternary Science Reviews 20, 1411–1436.
Nijampurkar, V.N., Rao, D.K., Oldfield., F., et al. 1998. The half-life Ó Cofaigh, C., Evans, D.J.A. 2007. Radiocarbon constraints on the
of 32Si: a new estimate based on varved lake sediments. Earth and age of the maximum advance of the British–Irish Ice Sheet in the
Planetary Science Letters 163, 191–196. Celtic Sea. Quaternary Science Reviews 26, 1197–1203.
Nivière, B., Bruestle, A., Bertrand, G., et al. 2008. Active tectonics of Ó Cofaigh, C., Evans, D.J.A., Smith, I.R. 2009. Large-scale reorgan-
the southeastern Upper Rhine Graben, Freiburg area (Germany). isation and sedimentation of terrestrial ice streams during late
Quaternary Science Reviews 27, 541–555. Wisconsinan Laurentide Ice Sheet deglaciation. Geological Society
North American Commission on Stratigraphic Nomenclature. 1983. of America Bulletin 122, 743–756.
North American Stratigraphic Code. American Association of Ó Cofaigh, C., Telfer, M.W., Bailey, R.M., et al. 2012a. Late
Petroleum Geologists Bulletin 67, 841–875. Pleistocene chronostratigraphy and ice sheet limits, southern
North American Commission on Stratigraphic Nomenclature. 2005. Ireland. Quaternary Science Reviews 44, 160–179.
North American Stratigraphic Code. American Association of Ó Cofaigh, C., Dowdeswell, J.A., Jennings, A.E., et al. 2012b. An
Petroleum Geologists Bulletin 89, 1547–1591. extensive and dynamic ice sheet on the West Greenland shelf
North Greenland Ice Core Project Members. 2004. High-resolution during the last glacial cycle. Geology 41, 219–222.
record of Northern Hemisphere climate extending into the last O’Connor, T.P. (ed.). 2005. Biosphere to Lithosphere – New Studies
interglacial period. Nature 431, 147–151. in Vertebrate Taphonomy. Oxbow Books, Oxford.
Notebaert, B., Verstraeten, G. 2010. Sensitivity of West and Central Odgaard, B.V. 1999. Fossil pollen as a record of past biodiversity.
European river systems to environmental changes during the Journal of Biogeography 26, 7–17.
Holocene: a review. Earth-Science Reviews 103, 163–182. O’Donnell, T.H., Macko, S.A., Wehmiller, J.F. 2006. Stable carbon
O’Brien, C., Selby, K., Ruiz, Z., et al. 2005. A sediment-based isotope composition of amino acids in modern and fossil
multiproxy palaeoecological approach to the environmental Mercenaria. Organic Geochemistry 38, 485–498.
archaeology of lake dwellings (crannogs), central Ireland. The Oerlemans, J., 2001. Glaciers and Climate Change. Balkema, Lisse.
Holocene 15, 707–719. Ó Foighil, D., Lee, T., Slapcinsky, J. 2011. Prehistoric anthropogenic
Obrochta, S.P., Miyahara, H., Yokoyama, Y., et al. 2012. A re- introduction of partulid tree snails in Papua New Guinean
examination of evidence for the North Atlantic ‘1500-year cycle’ archipelagos. Journal of Biogeography 38, 1625–1632.
at Site 609. Quaternary Science Reviews 55, 23–33. Ohmura, A., Kasser, P., Funk, M. 1992. Climate at the equilibrium
Occhietti, S. 2007. The Saint-Narcisse morainic complex and early line of glaciers. Journal of Glaciology 38, 397–411.
Younger Dryas events on the southeastern margin of the Ojala, A.E.K., Tiljander, M. 2003. Testing the fidelity of sediment
Laurentide Ice Sheet. Géographie physique et Quaternaire 61, chronology: comparison of varve and palaeomagnetic results from
89–117. Holocene lake sediments from central Finland. Quaternary
Oches, E.A., McCoy, W.D. 1995. Aminostratigraphic evaluation of Science Reviews 22, 1787–1803.
conflicting age estimates for the ‘Younger Loess’ of Hungary. Ojala, A.E.K., Alenius, T. 2005. 10 000 years of interannual
Quaternary Research 44, 160–170. sedimentation recorded in the Lake Nautajärvi (Finland) clastic-
Oches, E.A., McCoy, W. 2001. Historical developments and recent organic varves. Palaeogeography, Palaeoclimatology, Palaeoecology
advances in amino acid geochronology applied to loess research: 219, 285–302.
examples from North America, Europe and China. Earth-Science Ojala, A.E.K., Kubischta, F., Kaakinen, A., et al. 2011a. Charac-
Reviews 54, 173–192. terization of diamictons on the basis of their mineral magnetic
Oches, E.A., McCoy, W.D., Clark, P.U. 1996. Amino acid estimates properties in Murchisonfjorden, Nordaustlandet, Svalbard.
of latitudinal temperature gradients and geochronology of Sedimentary Geology 233, 150–158.
loess deposition during the last glaciation, Mississippi Valley, Ojala, A.E.K., Francus, P., Zolitschka, B., et al. 2011b. Characteristics
United States. Geological Society of America Bulletin 108, of sedimentary varve chronologies – a review. Quaternary Science
892–903 Reviews 43, 45–60.
Oches, E.A., McCoy, W.D., Gnieser, D. 2000. Aminostratigraphic Okazaki, Y., Timmermann, A., Menviel, L., et al. 2010. Deepwater
correlations of loess–paleosol sequences across Europe. In formation in the North Pacific during the Last Glacial
Perspectives in Amino Acid and Protein Geochemistry (edited Termination. Science 329, 200–204.
by G.A. Goodfriend, M.J. Collins, M.L. Fogel, et al.). Oxford Oldfield, F., Richardson, N., Appleby, P.G. 1995. Radiometric dating
University Press, New York, 331–348. (210Pb, 137Cs, 241Am) of recent ombrotrophic peat accumula-
Ó Cofaigh, C. 2007. Glacimarine sediments and ice-rafted debris. tion and evidence for changes in mass balance. The Holocene 5,
In Encyclopedia of Quaternary Science (edited by S.A. Elias). 141–148.
Elsevier, Amsterdam, 932–945. Oldfield, F., Wake, R., Boyle, J., et al. 2003. The late-Holocene
Ó Cofaigh, C. 2013. Glacimarine sediments and ice-rafted debris. history of Gormire Lake (NE England) and its catchment: a
In Encyclopedia of Quaternary Science, volume 2, 2nd edn (edited multiproxy reconstruction of past human impact. The Holocene
by S.A. Elias). Elsevier, Amsterdam, 30–42. 13, 677–690.
REFERENCES 497
Oliver, K.I.C., Hoogakker, B.A.A., Crowhurst, S., et al. 2009. A Osterberg, E., Mayewski, P.A., Kreutz, K.J., et al. 2008. Ice core
synthesis of marine sediment core δ13C data over the last 150 000 record of rising lead pollution in the North Pacific Atmosphere.
years. Climate of the Past Discussions 5, 2497–2554. Geophysical Research Letters 35, doi: 10.1029/2007/GL032680.
Ollier, C., Sheth, H.C. 2008. The High Deccan duricrusts of India O’Sullivan, P.B., Morwood, M., Hobbs, D., et al. 2001. Archaeological
and their significance for the ‘laterite’ issue. Journal of Earth System implications of the geology and chronology of the Soa basin,
Science 117, 537–551. Flores, Indonesia. Geology 29, 607–610.
Olson, C.G., Nettleton, W.D. 1998. Paleosols and the effects of Ota, Y., Yamaguchi, M. 2004. Holocene coastal uplift in the western
alteration. Quaternary International 51/52, 185–194. Pacific Rim in the context of late Quaternary uplift. Quaternary
Olsson, F., Lemdahl, G. 2009. A continuous Holocene beetle record International 120, 105–117.
from the site Stavsåkra, southern Sweden: implications for the Otto-Bliesner, B.L., Brady, E.C. 2010. The sensitivity of the climate
last 10 600 years of forest and land use history. Journal of response to the magnitude and location of freshwater forcing:
Quaternary Science 24, 612–626. last glacial maximum experiments. Quaternary Science Reviews
Onac, B., Constantin, S., Lundberg, J., et al. 2002. Isotopic climate 29, 56–73.
record in a Holocene stalagmite from Usilor Cave (Romania). Otvos, E.G. 2004. Prospects for interregional correlations using
Journal of Quaternary Science 17, 319–327. Wisconsin and Holocene aridity episodes, northern Gulf of
O’Neal, M.A., Schoenenberger, K.R. 2003. A Rhizocarpon geo- Mexico coastal plain. Quaternary Research 61, 105–118.
graphicum growth curve for the Cascade Range of Washington Overpeck, J.T., Webb, R.S. 1992. Mapping eastern North American
and northern Oregon, USA. Quaternary Research 60, 233–241. vegetation change of the past 18 ka: no-analogs and the future.
Opgenoorth, L., Vendramin, G.C., Mao, K., et al. 2010. Tree Geology 20, 1071–1074.
endurance on the Tibetan Plateau marks the world’s highest Overpeck, J.T., Otto-Bliesner, B., Miller, G.H., et al. 2006.
known tree line of the Last Glacial Maximum. New Phytologist Paleoclimatic evidence for future ice-sheet instability and rapid
185, 332–342. sea-level rise. Science 311, 1747–1750.
Oppenheimer, C. 2011. Eruptions that Shook the World. Cambridge Oviatt, C.G. 1997. Lake Bonneville fluctuations and global climate
University Press, Cambridge. change. Geology 25, 155–158.
Ori, G. G., Diachille, G., Komatsu, G., et al. 2009. River morphologies Oviatt, C.G., Thompson, R.S., Kaufman, D.S., et al. 1999.
and palaeodrainages of Western Africa (Sahara and Sahel) during
Reinterpretation of the Burmester Core, Bonneville Basin, Utah.
humid climatic conditions. In Sedimentary Processes, Envir-
Quaternary Research 52, 180–184.
onments and Basins: A Tribute to Peter Friend (edited by
Owen, L.A., Richards, B., Rhodes, E.J., et al. 1998. Relic permafrost
G. Nichols, E. Williams, C. Paola). Blackwell Publishing, Oxford,
structures in the Gobi of Mongolia: age and significance. Journal
519–534.
of Quaternary Science 13, 539–547.
Orme, A.R. 2008. Pleistocene Pluvial Lakes of the American West:
Owen, L.A., Finkel, R.C., Minnich, R.A., et al. 2003. Extreme
A Short History of Research. Geological Society of London,
southwestern margin of late Quaternary glaciation in North
Special Publication 301. Blackwell Scientific Publications,
America: timing and controls. Geology 31, 729–732.
Oxford, 51–78.
Owen, L.A., Caffee, M.W., Bovard, K.R., et al. 2006. Terrestrial
Orme, A.R., Orme, A.J., 2008. Late Pleistocene shorelines of Owens
Lake, California, and their hydroclimatic and tectonic impli- cosmogenic nuclide surface exposure dating of the oldest glacial
cations. In Late Cenozoic Drainage History of the Southwestern succession in the Himalayan orogen: Ladakh Range, northern
Great Basin and Lower Colorado River Region: Geologic and Biotic India. Geological Society of America Bulletin 118, 383–392.
Perspectives (edited by M.C. Reheis, R. Hershler, D.M. Miller). Owen, L.A., Bright, J., Finkel, R.C., et al. 2007. Numerical dating of
Geological Society of America, Special Paper 439, 207–225. a Late Quaternary spit-shoreline complex at the northern end
Ortiz, J.E., Torres, T., Julia, R., et al. 2004. Numerical dating of Silver Lake playa, Mojave Desert, California: a comparison of
algorithms of amino acid racemization ratios from continental the applicability of radiocarbon, luminescence, terrestrial
ostracodes. Application to the Guadix-Baza Basin (southern cosmogenic nuclide, electron spin resonance, U-series and amino
Spain). Quaternary Science Reviews 23, 717–730. acid racemization methods. Quaternary International 166,
Osborn, G., Bevis, K. 2001. Glaciation in the Great Basin of the 87–110.
Western United States. Quaternary Science Reviews 20, 1377– Owen, L.A., Thackray, G., Anderson, R.S., et al. 2009. Integrated
1410. research on mountain glaciers: current status, priorities and
Osborn, G., Menounos, B., Koch, J., et al. 2007. Multi-proxy record future prospects. Geomorphology 103, 158–171.
of Holocene glacial history of the Spearhead and Fitzsimmons Paillard, D. 2001. Glacial cycles: toward a new paradigm. Reviews of
ranges, southern British Columbia. Quaternary Science Reviews Geophysics 39, 325–346.
26, 479–493. Paillard, D., Parrenin, F. 2004. The Antarctic ice sheet and the
Oschmann, W. 2009. Editorial (special issue on sclerochronology). triggering of deglaciations. Earth and Planetary Science Letters 227,
International Journal of Earth Sciences (Geologische Rundschau) 263–271.
98, 1–2. Palmer, A.P., Rose J., Lowe, J.J., et al. 2010. Annually-resolved dating
Osipov, E.U. 2004. Equilibrium-line altitudes on reconstructed of Younger Dryas glaciation and proglacial lake duration in
LGM glaciers of the northwest Barguzinsky Ridge, Northern Glen Roy and Glen Spean, western Scottish Highlands. Journal
Baikal, Russia. Quaternary International 209, 219–226. of Quaternary Science 25, 581–596.
498 REFERENCES
Palmer, A.P., Rose, J., Rasmussen, S.O. 2012. Evidence for phase- Norse colonies. Proceedings of the National Academy of Sciences
locked changes in climate between Scotland and Greenland of the United States of America 107, 5306–5310.
during GS-1 (Younger Dryas) using micromorphology of Paulsen, D.E., Li, H.-C., Ku, T.-L. 2003. Climate variability in central
glaciolacustrine varves in Glen Roy. Quaternary Science Reviews China over the last 1270 years revealed by high-resolution
36, 114–123. stalagmite records. Quaternary Science Reviews 22, 691–701.
Palmer, J., Lorrey, A., Turney, C.S.M., et al. 2006. Extension of New Pawley, S.M., Rose, J., Lee, J.R., et al. 2004. Middle Pleistocene
Zealand kauri (Agathus australis) tree-ring chronologies into sedimentology and lithostratigraphy of Weybourne, northeast
Oxygen Isotope Stage (OIS) 3. Journal of Quaternary Science 21, Norfolk, England. Proceedings of the Geologists’ Association 115,
779–787. 25–42.
PALSEA (PALeo SEA Level Working Group). 2010. The sea-level Pawley, S.M., Toms, P., Armitage, S.J., et al. 2010. Quartz
conundrum: case studies from palaeo-archives. Journal of luminescence dating of Anglian Stage (MIS12) fluvial sediments:
Quaternary Science 25, 19–25. comparison of SAR age estimates to the terrace chronology of
Pan, B., Burbank, D., Wang, Y., et al. 2003. A 900 k.y. record of strath the Middle Thames Valley, UK. Quaternary Geochronology 5,
terrace formation during glacial–interglacial transitions in 569–582.
northwest China. Geology 31, 957–960. Payne, A.J. 1999. A thermomechanical model of ice flow in West
Panagiotakopulu, E., Greenwood, M.T., Buckland, P.C. 2012. Insect Antarctica. Climate Dynamics 15, 115–125.
fossils and irrigation in medieval Greenland. Geografiska Annaler Payne, R.J., Blackford, J.J. 2008. Peat humification and climate
Series A, Physical Geography 94, 531–548. change: a multi-site comparison from mires in south-east Alaska.
Pancost, R.D., McClymont, L., Bingham, E.M., et al. 2011. Archaeol Mires and Peat 3(art. 9), 1–11.
as a methanogen biomarker in ombrotrophic bogs. Organic Pazzaglia, F.J., Brandon, M.T. 2001. A fluvial record of long-term
Geochemistry 42,1279–1287. steady-state uplift and erosion across the Cascadia Forearc High,
Panieri, G., Camerlenghi, A., Cacho, I., et al. 2012. Tracing seafloor Western Washington State. American Journal of Science 301,
methane emissions with benthic foraminifera: results from the
385–431.
Ana submarine landslide (Eivissa Channel, Western Mediter-
Peacock, J.D. 1993. Late Quaternary marine mollusca as palaeo-
ranean Sea). Marine Geology 291/294, 97–112.
environmental proxies: a compilation and assessment of basic
Panyushkina, I., Sljusarenko, I. Bikov, N., et al. 2007. Floating
numerical data for NE Atlantic species found in shallow water.
larch tree-ring chronologies from archaeological timbers in the
Russian Altai between about 800 BC and AD 800. Radiocarbon
Pearce, N.J.G., Westgate, J.A., Preece, S.J., et al. 2004. Identifica-
49, 693–702.
tion of Aniakchak (Alaska) tephra in Greenland ice core
Parfitt, S.A., Barendregt, R.W., Breda, M., et al. 2005. The earliest
challenges the 1645 BC date for Minoan eruption of Santorini.
record of human activity in northern Europe. Nature 438,
Geochemistry, Geophysics, Geosystems 5, doi: 10.1029/2003
1008–1012.
Parham, P.R., Riggs, S.R., Culver, S.J., et al. 2007. Quaternary GC000672.
depositional patterns and sea-level fluctuations, northeastern Pearson, P.N. 2012. Oxygen isotopes in Foraminifera: overview
North Carolina. Quaternary Research 67, 83–99. and historical review. In Reconstructing Earth’s Deep-Time
Parker, A.G., Goudie, A.S., Anderson, D.E., et al. 2002. A review of Climate – The State of the Art in 2012 (edited by L.C Ivany,
the mid-Holocene elm decline in the British Isles. Progress in B.T Huber). The Palaeontological Society Special Papers 18.
Physical Geography 26, 1–45. The Paleontological Society, London, 1–38.
Parmalee, P.W. 2005. Ice Age cave faunas of North America. Pedoja, K., Regard, V., Husson, L., et al. 2011. Uplift of Quaternary
Geoarchaeology 20, 331–334. shorelines in eastern Patagonia: Darwin revisited. Geomorphology
Parmesan, C. 1999. Coping with modern times? Insect movement 127, 121–142.
and climate change. In Insect Movement: Mechanisms and Pedro, J.B., van Ommen, T.D., Rasmussen, S.O., et al. 2011. The last
Consequences (edited by I.P. Woiwod, D.R. Reynolds, C.D. deglaciation: timing the bipolar seesaw. Climate of the Past 7,
Thomas). CABI Publishing, Oxford and New York, 387–414. 671–683.
Parrenin, F., Barnola, J.M., Beer, J., et al. 2007. The ED3 chronology Peeters, F.J.C., Acheson, R., Brummer, G.-J.A., et al. 2004. Vigorous
for the EPICA Dome C ice core. Climate of the Past 3, 485–497. exchange between the Indian and Atlantic oceans at the end of
Parrenin F., Petit J.-R., Masson-Delmotte V., et al. 2012. Volcanic the past five glacial periods. Nature 430, 661–665.
synchronisation between the EPICA Dome C and Vostok ice cores Peglar, S.M. 1993. The mid-Holocene Ulmus decline at Diss, Norfolk,
(Antarctica) 0–145 kyr BP. Climate of the Past 8, 1031–1045. UK: a year-by-year pollen stratigraphy from annual laminations.
Paterson, W.S.B. 1994. The Physics of Glaciers, 3rd edn. Elsevier, The Holocene 3, 1–13.
Amsterdam. Peglar, S.M., Birks, H.J.B. 1993. The mid-Holocene Ulmus fall at Diss
Patterson, R.T., Kumar, A. 2002. A review of current testate rhizopod Mere, south-east England – disease and human impact? Vegeta-
(thecamoebian) research in Canada. Palaeogeography, Palaeo- tional History and Archaeobotany 2, 61–68.
climatology, Palaeoecology 180, 225–251. Peltier, W.R. 2002. Global glacial isostatic adjustment: palaeogeodetic
Patterson, W.P., Dietrich, K.A., Holmden, C., et al. 2010. Two and space-geodetic tests of the ICE-4G (VM2) model. Journal of
millennia of North Atlantic seasonality and implications for Quaternary Science 17, 491–510.
REFERENCES 499
Peltier, W.R. 2005. On the hemispheric origins of meltwater pulse Petsko, G.A., Ringe, D. 2004. Protein Structure and Function. New
1a. Quaternary Science Reviews 24, 1655–1671. Science Press, London.
Peltier, W.R., Fairbanks, R.G. 2006. Global glacial ice volume and Petterson, G. 1996. Varved sediments in Sweden: a brief review. In
Last Glacial Maximum duration from an extended Barbados Palaeoclimatology and Palaeoceanography from Laminated
sea level record. Quaternary Science Reviews 25, 3322–3337. Sediments (edited by A.E.S. Kemp). Geological Society, London,
Pelto, M.S., Hedlund, C. 2001. Terminus behavior and response 73–77.
time of North Cascade glaciers, Washington, USA. Journal of Péwé, T.L. 1966. Palaeoclimatic significance of fossil ice wedges.
Glaciology 47, 497–506. Biuletyn Peryglacjalny 15, 65–73.
Peña-Chocarro, L., Zopata, L., Iriarte, M.J., et al. 2005. The oldest Pflaumann, U., Duprat, J., Pujol, C., et al. 1996. SIMMAX: A modern
agriculture in northern Atlantic Spain: new evidence from analog technique to deduce Atlantic sea surface temperatures
El Mirón Cave (Ramales de la Victoria, Cantabria). Journal of from planktonic foraminifera in deep-sea sediments. Paleoceano-
Archaeological Science 32, 579–587. graphy 11, 15–35.
Penkman, K.E.H. 2010. Amino acid geochronology: its impact on Phillips, E., Lee, J.R., Burke, H. 2008a. Progressive proglacial to sub-
our understanding of the Quaternary stratigraphy of the British glacial deformation and syntectonic sedimentation at the margins
Isles. Journal of Quaternary Science 25, 501–514. of the Mid-Pleistocene British Ice Sheet: evidence from north
Penkman, K.E.H., Preece, R.C., Keen, D.H et al. 2007. Testing the Norfolk, UK. Quaternary Science Reviews 27, 1848–1871.
aminostratigraphy of fluvial archives: the evidence from intra- Phillips, T.J., Potter, G.L., Williamson, D.L., et al. 2004. Evaluating
crystalline proteins within freshwater shells. Quaternary Science parameterizations in General Circulation Models: climate
Reviews 26, 2958–2969. simulation meets weather prediction. Bulletin of the American
Penkman, K.E.H., Kaufman, D.S., Maddy, D., et al. 2008. Closed- Meteorological Society 85, 1903–1915.
system behaviour of the intra-crystalline fraction of amino acids Phillips, W.M., Hall, A.R., Mottram, R., et al. 2006. Cosmogenic
10
in mollusc shells. Quaternary Geochronology 3, 2–25. Be and 26Al exposure ages of tors and erratics, Cairngorm
Penkman, K.E.H., Preece, R.C., Bridgland, D.R., et al. 2013. An Mountains, Scotland: timescales for the development of a classic
aminostratigraphy for the British Quaternary based on Bithynia landscape of selective glacial erosion. Geomorphology 73, 222–245.
opercula. Quaternary Science Reviews 61, 111–134. Phillips, W.M., Hall, A.M., Ballantyne, C.K., et al. 2008b. Extent of
Pennington, W., Haworth, E.Y., Bonny, A.P., et al. 1972. Lake the last ice sheet in northern Scotland tested with cosmogenic
10
sediments in northern Scotland. Philosophical Transactions of Be exposure ages. Journal of Quaternary Science 23, 101–108.
the Royal Society of London B 264, 191–294. Pickering, R., Jacobs, Z., Herries, A.I.R., et al. 2013. Palaeo-
Pentecost A. 2005. Travertine. Springer, Berlin. anthropologically significant South African sea caves dated to
Perez, K.E., Cordeiro, J.R., Coppolino, L. 2008. A Guide for Terrestrial 1.1–1.0 million years using a combination of U–Pb, TT-OLS and
Gastropod Identification. American Malacological Society, palaeomagnetism. Quaternary Science Reviews 65, 39–52.
Carnondale, IL. Pierce, K.L. 2003. Pleistocene glaciations of the Rocky Mountains.
Perner, K., Moros, M., Lloyd, J.M., et al. 2011. Centennial scale In Developments in Quaternary Science. Volume 1: The Quaternary
benthic foraminiferal record of late Holocene oceanographic Period in the United States (edited by A.R. Gillespie, S.C. Porter,
variability in Disko Bugt, West Greenland. Quaternary Science B.F. Atwater). Elsevier, Amsterdam, 63–76.
Reviews 30, 2815–2826. Pierce, K.L., Obradovich, J.D., Friedman, I. 1976. Obsidian hydration
Pessenda, L.R.C., Gouveia, S.E.M., Aravena, R. 2001. Radiocarbon dating and correlation of Bull Lake and Pinedale Glaciations
dating of total soil organic matter and humin fractions and its near west Yellowstone, Montana. Geological Society of America
comparison with 14C ages of fossil charcoal. Radiocarbon 43, Bulletin 87, 703–710.
595–601. Pike A.W.G., Pettitt, P.B. 2003. U-series dating and human evolution.
Peteet, D. 1995. A global Younger Dryas? Quaternary International In Uranium Series Geochemistry (edited by B. Bourdon, G.M.
28, 93–104. Henderson, C.C. Lundstrom, et al.). Reviews in Mineralogy and
Peterson, L.C., Haug, G.H., Hughen, K.A., et al. 2000. Rapid changes Geochemistry 52, 607–630.
in the hydrologic cycle of the Tropical Atlantic during the Last Pike, A.W.G., Hedges, R.E.M., van Calsteren, P. 2002. U-series
Glacial. Science 290, 1947–1951. dating of bone using the diffusion-absorption model. Geo-
Petit, J.R., Jouzel, J., Raynaud, D., et al. 1999. 420,000 years of chimica et Cosmochimica Acta 66, 4273–4286.
climate and atmospheric history revealed by the Vostok deep Pike, A.W.G., Hoffmann, D.L., García-Diez, M., et al. 2012. U-series
Antarctic ice core. Nature 399, 429–436. dating of Palaeolithic art in 11 caves in Spain. Science 336,
Petoukhov, V., Claussen, M., Berger, A., et al. 2005. EMIC Inter- 1409–1413.
comparison Project (EMIP–CO2): comparative analysis of EMIC Pike, J., Stickley, C.E. 2007. Diatom records: marine laminated
simulations of climate, and of equilibrium and transient sediments. In Encyclopedia of Quaternary Science (edited by S.A.
responses to atmospheric CO2 doubling. Climate Dynamics 25, Elias). Elsevier, Amsterdam, 557–566.
363–385. Pike, R.J. 2000. Geomorphometry – diversity in quantitative surface
Petraglia, M., Korisettar, R., Boivin, N., et al. 2007. Middle Palaeo- analysis. Progress in Physical Geography 24, 1–20.
lithic assemblages from the Indian subcontinent before and Pillans, B. 2004. Defining the Quaternary. Quaternary Science Reviews
after the Toba super-eruption. Science 317, 114–116. 23, 2271–2282.
500 REFERENCES
Pillans, B., Gibbard, P.L. 2012. The Quaternary Period. In The Pollard, A.M., Blockley, S.P.E., Lane, C.S. 2006. Some numerical
Geologic Time Scale 2012 (edited by F. Gradstein, J. Ogg, M. considerations in the geochemical analysis of distal microtephra.
Schmitz, et al.). Elsevier, Amsterdam, 980–1009. Applied Geochemistry 21, 1692–1714.
Pillans, B.J., Alloway, B.V., Naish, T.R., et al. 2005. Silicic tephras in Pollard, D., DeConto, R.M. 2009. Modelling West Antarctic ice
Pleistocene sediments of Wanganui Basin, New Zealand. Journal sheet growth and collapse through the past five million years.
of the Royal Society of New Zealand 35, 43–90. Nature 458, 329–332.
Pint, A., Frenzel, P., Fuhrmann, R., et al. 2012. Distribution of Polly, P.D., Eronen, J.T. 2011. Mammal associations in the Pleisto-
Cyprideis torosa (Ostracoda) in Quaternary athalassic sediments cene of Britain: implications of Ecological Niche Modelling
in Germany and its application for palaeoecological reconstruc- and a method for reconstructing palaeoclimate. In The Ancient
tions. International Review of Hydrobiology 97, 330–355. Human Occupation of Britain (edited by N. Ashton, S.G. Lewis,
Pinter, N., Scott, A.C., Daulton, T.L et al. 2011. The Younger C. Stringer). Elsevier, Amsterdam, 279–304.
Dryas impact hypothesis: a requiem. Earth-Science Reviews 106, Ponel, P., Coope, G.R., Antoine, P., et al. 2005. Lateglacial palaeo-
247–264. environments and palaeoclimates from Cont and Houdancourt,
Piotrowski, J.A., Larsen, N.K., Junge, F.W. 2004. Reflections on soft northern France, reconstructed from beetle remains. Quaternary
subglacial beds as a mosaic of deforming and stable spots. Science Reviews 24, 2449–2465.
Quaternary Science Reviews 23, 993–1000. Poole, G.C., Stanford, J.A., Frissell, C.A., et al. 2002. Three-
Piro, S., Campana, S. 2012. GPR investigations in different dimensional mapping of geomorphological controls on flood-
archaeological sites in Tuscany (Italy). Analysis and comparison plain hydrology and connectivity from aerial photos.
of the obtained results. Near Surface Geophysics 10, 47–56. Geomorphology 48, 329–347.
Pisaric, M.F.J. 2002. Long-distance transport of terrestrial plant Popov, D., Vandenberghe, D.A., Marković, V.C. 2012. Luminescence
material by convection resulting from forest fires. Journal of dating of fluvial deposits in Vojvodina, N. Serbia: first results.
Paleolimnology 28, 349–354. Quaternary Geochronology 13, 42–51.
Pissart, A. 2001. Remnants of lithalsas of the Hautes Fagnes, Belgium: Porch, N. 2010. Climate space, bioclimatic envelopes and coexistence
a summary of present-day knowledge. Permafrost and Periglacial methods for the reconstruction of past climates: a method using
Processes 11, 327–355. Australian beetles and significance for Quaternary reconstruc-
Pissart, A. 2002. Palsas, lithalsas and remnants of these periglacial
tion. Quaternary Science Reviews 29, 633–647.
mounds: a progress report. Progress in Physical Geography 26,
Porinchu, D.F., MacDonald, G.M. 2003. The use and application
605–621
of freshwater midges (Chironomidae: Insecta: Diptera) in
Pla, S., Anderson, N.J. 2005. Environmental factors correlated with
geographical research. Progress in Physical Geography 27,
chrysophyte cyst assembages in low Arctic lakes of Southwest
378–422.
Greenland. Journal of Phycology 41, 957–974.
Porinchu, D.F., Rolland, N., Moser, K. 2009. Development of a
Pla-Rabes, S., Catalan, J. 2011. Deciphering chrysophyte responses
chironomid-based air temperature inference model for the
to climate seasonality. Journal of Paleolimnology 46, 139–150.
central Canadian Arctic. Journal of Paleolimnology 41, 349–368.
Plag, H.-P., Engen, B., Clark, T.A., et al. 1998. Post-glacial rebound
Porter, S.C. 2000. Snowline depression in the tropics during the Last
and present-day three-dimensional deformations. Journal of
Geodynamics 25, 263–301 Glaciation. Quaternary Science Reviews 20, 1067–1091.
Plagnes, V., Causse, C., Genty, D., et al. 2002. A discontinuous Porter, S.C. 2001. Chinese loess record of monsoon climate during
climatic record from 187 to 74 ka from a speleothem of the the last glacial–interglacial cycle. Earth-Science Reviews 54,
Clamouse Cave (south of France). Earth and Planetary Science 115–128.
Letters 201, 87–103. Porter, S.C., Singhvi, A., Zhishng, A., et al. 2001. Luminescence age
Plug L.J., Twerner B.P. 2008. Modelling ice-wedge networks. and palaeoenvironmental implications of a late Pleistocene
Permafrost and Periglacial Processes 19, 63–69. ground wedge on the Northeastern Tibetan Plateau. Permafrost
Plunkett, G.M., Whitehouse, N.J., Hall, V.A., et al. 2004. A precisely- and Periglacial Processes 12, 203–210.
dated lake-level rise marked by diatomite formation in Potter, E.-K. Lambeck, K. 2003. Reconciliation of sea-level
northeastern Ireland. Journal of Quaternary Science 19, 3–7. observations in the Western North Atlantic during the last glacial
Poli, M.S., Thunnell, R.C., Rio, D. 2000. Millennial-scale changes cycle. Earth and Planetary Science Letters 217, 171–181.
in North Atlantic Deep Water circulation during marine iso- Povinec, P.P., Litherland, A.E., von Reden, K.F. 2009. Developments
tope stages 11 and 12: linkage to Antarctic climate. Geology 28, in radiocarbon technologies: from the Libby Counter to
807–810. compound-specific AMS analyses. Radiocarbon 51, 45–78.
Pollack, H.N., Huang, S., Smerdon, J.E. 2006. Five centuries of Powell, E.N., Staff, G.M., Callender, W.R., et al. 2011. The influence
climate change in Australia: the view from underground. Journal of molluscan taxon on taphofacies development over a broad
of Quaternary Science 21, 701–706. range of environments of preservation: the SSETI experi-
Pollard, A.M., Heron, C. 1996. Archaeological Chemistry. Royal ence. Palaeogeography, Palaeoclimatology, Palaeoecology 312,
Society of Chemistry, Cambridge. 233–264.
Pollard, A.M., Blockley, S.P.E., Ward, K.R. 2003. Chemical alteration Power, S., Delage, F., Chung, C., et al. 2013. Robust twenty-first
of tephra in the depositional environment: theoretical stability century projections of El Niño and related precipitation
modelling. Journal of Quaternary Science 18, 385–394. variability. Nature doi: org/10.1038/nature 12580.
REFERENCES 501
Praetorius, S.K., McManus, J.F., Oppo, D.W., et al. 2008. Episodic Prueher, L.M., Rea, D.K. 1998. Rapid onset of glacial conditions in
reductions in bottom-water currents since the last ice age. Nature the subarctic North Pacific region at 2.67 Ma: clues to causality.
Geoscience 1, 449–452. Geology 26, 1027–1030.
Prasad, S., Negendank, J.F.W., Stein, M. 2009. Varve counting reveals Pudsey, C.J., Murray, J.W., Appleby, P., et al. 2006. Ice shelf history
high resolution radiocarbon reservoir variations in palaeolake from petrographic and foraminiferal evidence, Northeast Ant-
Lisan. Journal of Quaternary Science 24, 690–696. arctic Peninsula. Quaternary Science Reviews 25, 2357–2379.
Preece, R.C. 2001. Molluscan evidence for differentiation of inter- Pugh, R.S., McCave, I.N., Hillebrand, C.-D., et al. 2009. Circum-
glacials within the ‘Cromerian Complex’. Quaternary Science Antarctic age modelling of Quaternary marine cores under the
Reviews 20, 1643–1656. Antarctic Circumpolar Current: ice-core dust–magnetic cor-
Preece, R.C., Bridgland, D.R. 1999. Holywell Coombe, Folkestone: relation. Earth and Planetary Science Letters 284, 113–123.
a 13,000 year history of an English chalkland valley. Quaternary Punt, W., Blackmore, S., Hoen, P.P., et al. 2009. The Northwest
Science Reviews 18, 1075–1125. European Pollen Flora IX. Elsevier, Amsterdam.
Preece, R.C., Parfitt, S.A. 2012. The Early and early Middle Pleistocene Puthusserry, J.T., Murray, A.S., Kjaer, K.H., et al. 2008. Optically
context of human occupation and lowland glaciation in Britain stimulated luminescence (OSL) dating of glacial sediments
and northern Europe. Quaternary International 271, 6–28. from Arctic Russia – depositional bleaching and methodological
Preece, R.C., Gowlett, J.A.J., Parfitt, S.A., et al. 2006. Humans in aspects. Boreas 35, 587–599.
the Hoxnian: habitat, context and fire use at Beeches Pit, West Putkonen, J., Swanson, T. 2003. Accuracy of cosmogenic ages for
Stow, Suffolk, UK. Journal of Quaternary Science 21, 485–496. moraines. Quaternary Research 59, 255–261.
Preece, R.C., Parfitt, S.A., Coope, G.R., et al. 2009. Biostratigraphic Putnam, A.E., Denton, G.H., Schaefer, J.M., et al. 2010. Glacier
and aminostratigraphic constraints on the age of the Middle advance in southern middle-latitudes during the Antarctic
Pleistocene glacial succession in north Norfolk, UK. Journal of Cold Reversal. Nature Geoscience 3, 700–704.
Quaternary Science 24, 557–580. Pyle, D.M., Ricetts, G.M., Margari, V., et al. 2006. Wide dispersal
Prentice, I.C., Farquar, G.D., Fasham, M.J.R., et al. 2001. The carbon and deposition of distal tephra during the Pleistocene ‘Campan-
cycle and atmospheric carbon dioxide. In Climate Change ian Ignimbrite/Y5’ eruption, Italy. Quaternary Science Reviews
2001: The Scientific Basis (edited by J.T. Houghton, Y. Ding, 25, 2713–2728.
D.J. Griggs, et al.). Cambridge University Press, Cambridge, Qin, B., Harrison, S.P., Kutzbach, J.E 1998. Evaluation of modelled
185–235. regional water balance using lake status data: a comparison of
Previsic, A., Walton, C., Kucinic, M., et al. 2009. Pleistocene diver- 6 ka simulations with the NCAR CCM. Quaternary Science
gence of Dinaric Drusus endemics (Tricoptera, Limnephilidae) Reviews 17, 659–668.
in multiple microrefugia within the Balkan Peninsula. Molecular Quigley, M.C., Horton, T., Helstrom, J.C., et al. 2010. Holocene
Ecology 18, 634–647. climate change in arid Australia from speleothem and alluvial
Prevosti, F.J., Tonni, E.P., Bidegain, J.C. 2009. Stratigraphic range records. The Holocene 20, 1093–1104.
of the large canids (Carnivora, Canidae) in South America and Rabassa, J., Ponce, J.F. 2013. The Heinrich and Dansgaard–Oeschger
its relevance to quaternary biostratigraphy. Quaternary Inter- climatic events during Marine Isotope Stage 3: searching for
national 210, 76–81. appropriate times for human colonisation of the Americas.
Price, G.J., Webb, G.E. 2006. Late Pleistocene sedimentology, Quaternary International 299, 94–105.
taphonomy and megafauna extinction on the Darling Downs, Radtke, U., Schellman, G. Scheffers, A., et al. 2003. Electron spin
southeastern Queensland. Australian Journal of Earth Sciences 53, resonance and radiocarbon dating of coral deposited by Holocene
947–970. tsunami events on Curaçao, Bonaire and Aruba (Netherlands
Prichard, S.J., Gedalof, Z., Oswald, W.W., et al. 2009. Holocene fire Antilles). Quaternary Science Reviews 22, 1309–1315.
and vegetation dynamics in a montane forest, North Cascade Rahmstorf, S. 1996. On the freshwater forcing and transport of the
Range, Washington, USA. Quaternary Research 72, 57–67. Atlantic thermohaline circulation. Climate Dynamics 12,
Prinn, R.G. 2011. Development and application of earth system 799–811.
models. Proceedings of the National Academy of Sciences of the Rahmstorf, S. 2002. Ocean circulation and climate during the past
United States of America 110, 3673–3680. 120,000 years. Nature 419, 207–214.
Prinn, R., Paltsev, S., Sokolov, A., et al. 2011.Scenarios with MIT Rahmstorf, S. 2003. Timing of abrupt climate change: a precise
integrated global systems model: significant global warming clock. Geophysical Research Letters 30, 1510, doi: 10.1029/2003
regardless of different approaches. Climatic Change 104, 515–537. GL017115.
Prokopenko, A.A., Hinnov, L.A., Williams, D.F., et al. 2006. Orbital Rahmstorf, S., Cricifix, M., Ganopolski, A., et al. 2005. Thermohaline
forcing of continental climate during the Pleistocene: a complete circulation hysteresis: A model intercomparison. Geophysical
astronomically tuned climatic record from Lake Baikal, SE Research Letters 32, L23605, doi: 10.1029/2005GL023655.
Siberia. Quaternary Science Reviews 25, 3431–3457. Raia, P., Meiri, S. 2006. The Island Rule in large large mammals:
Pross, J., Klotz, S., Mosbrugger, V. 2000. Reconstructing palaeo- paleontology meets ecology. Evolution 60, 1731–1742.
temperatures for the Early and Middle Pleistocene using the Rainio, H., Saarnisto, M., Ekman, I. 1995. Younger Dryas end
mutual climatic range method based on plant fossils. Quaternary moraines in Finland and NW Russia. Quaternary International
Science Reviews 19, 1785–1799. 28, 179–192.
502 REFERENCES
Raisbeck, G.M., Yiou, F., Jouzel, J., et al. 2007. Direct North–South Raymo, M.E. 1997. The timing of major climatic terminations.
synchronisation of abrupt climate change record in ice cores using Paleoceanography 12, 577–585.
beryllium 10. Climate of the Past Discussions 3, 541–547. Raymo, M.E., Ruddiman, W.F. 1992. Tectonic forcing of late
Ram, M., Donarummo, J., Stolz, M.R., et al. 2012. Calibration of laser- Cenozoic climate. Nature 359, 117–122.
light scattering measurements of dust concentration for Raymo, M., Nisancioglu, K. 2003. The 41 kyr world: Milankovitch’s
Wisconsinan GISP2 ice using instrumental neutron activation other unsolved mystery. Paleoceanography 18, doi: 10.1029/
analysis of aluminium: results and discussions. Journal of 2002PA000791.
Geophysical Research Atmospheres 105, 24731–24738. Raymo, M., Ruddiman, Froelich, P.M. 1992. Influence of late
Ramillien, G., Lombard, A., Cazenave, A., et al. 2006. Interannual Cenozoic mountain building on ocean geochemical cycles.
variations of the mass balance of the Antarctica and Greenland Geology 16, 649–653.
ice sheets from GRACE. Global and Planetary Change 53, Raymo, M.E., Mitrovica, J.X., O’Leary, R.M., et al. 2011. Departures
198–208. from eustasy in Pliocene sea-level records. Nature Geoscience 4,
Rampino, M.R., Self, S., Fairbridge, R.W. 1979. Can rapid climate 328–332.
change cause volcanic eruptions? Science 206, 826–829. Rayner, D., Hirschi, J.J.-M., Kanzow, T., et al. 2011. Monitoring the
Ran, Y., Li, X., Cheng, G., et al. 2012. Distribution of permafrost in Atlantic meridional overturning circulation. Deep-Sea Research
China: an overview of existing permafrost maps. Permafrost and II 58, 1744–1753.
Periglacial Processes 23, 322–333. Rea, B.R. 2009. Defining modern day Area-Altitude Balance ratios
Randall, D.A., Wood, R.A., Bony, S., et al. 2007. Climate models and (AABRs) and their use in glacier–climate reconstructions.
their evaluation. In Climate Change 2007: Working Group I: Quaternary Science Reviews 28, 237–248.
The Physical Science Basis (edited by S. Solomon, D. Quin, Rea, D.K., Snoeck, H., Joseph, L.H. 1998. Late Cenozoic eolian
M. Manning, et al.). Cambridge University Press, Cambridge, deposition in the North Pacific: Asian drying, Tibetan uplift, and
591–662. cooling of the Northern Hemisphere. Paleoceanography 13,
Rankin, A.M., Wolff, E.W., Mulvaney, R. 2004. A reinterpretation 215–224.
of sea-salt records in Greenland and Antarctic ice cores. Annals Reeves, J.M., Chivas, A.R., Garcia, A., et al. 2007. Palaeo-
of Glaciology 39, 276–282. environmental change in the Gulf of Carpentaria (Australia) since
Räsenen, M.E., Auri, J.M., Huitti, J.V., et al. 2009. A shift from
the last interglacial based on Ostracoda. Palaeogeography,
lithostratigraphic to allostratigraphic classification of Quaternary
glacial deposits. GSA Today 19, 4–11.
Refsnider, K.A., Brugger, K.A. 2007. Rock glaciers in Central
Rasmussen, S.O., Andersen, K.K., Johnsen, S.J., et al. 2005.
Colorado, U.S.A., as indicators of Holocene climate change.
Deconvolution-based resolution enhancement of chemical ice
Arctic, Antarctic, and Alpine Research 39, 127–136.
core records obtained by continuous flow analysis. Journal of
Refsnider, K.A., Laabs, B.J.C., Plummer, M.A., et al. 2008. Last
Geophysical Research doi: 10.1029/2004JD005717.
glacial maximum climate inferences from cosmogenic dating
Rasmussen S.O., Andersen K.K., Svensson A.M., et al. 2006. A new
and glacier modeling of the western Uinta ice field, Uinta
Greenland ice core chronology for the last glacial termina-
Mountains, Utah. Quaternary Research 69, 130–144.
tion. Journal of Geophysical Research, 111: D06102, doi: 10.1029/
2005JD006079. Refsnider, K.A., Brugger, KA., Leonard, E.M., et al. 2009. Last
Rasmussen, S.O., Vinther, B., Clausen, H.B., et al. 2007. Early Glacial Maximum equilibrium-line altitude trends and
Holocene climate oscillations recorded in three Greenland ice precipitation patterns in the Sangre de Cristo Mountains,
cores. Quaternary Science Reviews 26, 1907–1914. southern Colorado, USA. Boreas 38, 663–678.
Rasmussen, S.O., Seierstad, I.K., Andersen, K.K., et al. 2008. Reichert, K.L., Licciardi, J.M., Kaudfmnan, D.S. 2011. Amino acid
Synchronisation of the NGRIP, GRIP and GISP2 ice cores across racemization in lacustrine ostracodes, Part II: paleother-
MIS 2 and palaeoclimatic implications. Quaternary Science mometry in Pleistocene sediments at Summer Lake, Oregon.
Reviews 27, 18–28. Quaternary Geochronology 6, 174–185.
Rasmussen, T.L., Thomsen, E. 2004. The role of the North Atlantic Reille, M. 1995. Pollen et Spores d’Europe et d’Afrique du Nord.
Drift in the millennial timescale glacial climate fluctuations. Supplement 1. Laboratoire de Botanique Historique et de
Palaeogeography, Palaeoclimatology, Palaeoecology 210, 101–116. Palynologie, Université d’Aix-Marseille.
Rathore, V.S., Nathawat, M.S., Champatiray, P.K. 2010. Reille, M., de Beaulieu, J.-L., Svobodova, H., et al. 2000. Pollen
Palaeochannel detection and aquifer performance assessment in analytical biostratigraphy of the last five climatic cycles from a
Mendha River catchment, Western India. Journal of Hydrology long continental sequence from the Velay region (Massif Central,
395, 216–225. France). Journal of Quaternary Science 15, 665–685.
Rawson, P.F., Allen, P.M., Brenchley, P.J., et al. 2002. Stratigraphical Reimer, P.J., Reimer, R.W. 2001. A marine reservoir correction
Procedure (‘Professional Handbook’ Series). Geological Society, database and on-line interface. Radiocarbon 43, 461–463.
Bath. Reimer, P.J., Bard, E., Bayliss, A., et al. 2013. IntCal13 and Marine13
Rayburn, J.A., Cronin, T.M., Franzi, D.A., et al. 2011. Timing and radiocarbon age calibration curves 0–50,000 years cal BP.
duration of North American glacial lake discharges and the Radiocarbon 55, 1869–1887.
Younger Dryas climate reversal. Quaternary Research 75, Reiss, R.A. 2006. Ancient DNA from ice age insects: proceed with
541–551. caution. Quaternary Science Reviews 25, 1877–1893.
REFERENCES 503
Rember, W.C., Erdman, T.W., Hoffman, M.L., et al. 1993. Dating Rial, J.A., Oh, J., Reischmann, E. 2013. Synchronization of the
of mine waste in lacustrine sediments using cesium-137. Environ- climate system to eccentricity forcing and the 100,000-year
mental Earth Sciences 22, 242–245. problem. Nature Geoscience 6, 289–293.
Rémy, F., Parouty, S. 2009. Antarctic ice sheet and radar altimetry: Ribera, I., Vogler, A.P. 2004. Speciation of Iberian diving beetles in
a review. Remote Sensing 1, 1212–1239. Pleistocene refugia (Coleoptera, Dytiscidae). Molecular Ecology
Renberg, I., Braännvall, M.-L., Bindler, R., et al. 2002. Stable lead 13, 179–193.
isotopes and lake sediments – a useful combination for the Richards, D.A., Smart, P.L. 1991. Potassium–argon and argon–argon
study of atmospheric lead pollution history. The Science of the dating. In Quaternary Dating Methods – A User’s Guide (edited
Total Envirnment 292, 45–54. by P.L. Smart, P.D. Frances). Technical Guide 4. Quaternary
Renberg, I., Bigler, C., Bindler, R., et al. 2009. Environmental history: Research Association, Cambridge, 37–44.
a piece in the puzzle for establishing plans for environmental Richards, D.A., Dorale, J.A. 2003. U-series chronology and
management. Journal of Environmental Management 90, environmental applications of speleothems. In Uranium Series
Geochemistry (edited by B. Bourdon, G.M. Henderson, C.C.
2794–2800.
Lundstrom, et al.). Reviews in Mineralogy and Geochemistry 52,
Renfrew, C. 1973. Before Civilisation: The Radiocarbon Revolution
407–460.
and Prehistoric Europe. Jonathan Cape, London.
Richter, D. 2007. Advantages and limitations of thermolumin-
Renne, P.R., Sharp, W.D., Deino, A.L., et al. 1997. 40Ar/39Ar dating
escence dating of heated flint from Palaeolithic sites. Geo-
into the historical realm: calibration against Pliny the Younger. archaeology 22, 671–683.
Science 277, 1279–1280. Rickaby, R.E.M., Elderfield, H. 2005. Evidence from the high-latitude
Renssen, H., Vandenberghe, J. 2003. Investigation of the relationship North Atlantic for variations in Antarctic Intermediate water
between permafrost distribution in NW Europe and extensive flow during the last deglaciation. Geochemistry, Geophysics,
winter sea-ice cover in the North Atlantic Ocean during the Geosystems 6, Q05001, doi: 10.1029/2004GC000858.
cold phases of the Last Glaciation. Quaternary Science Reviews Ridefelt, H., Boelhouwers, J., Eiken, T. 2009. Measurement of
22, 209–223. solifluction rates using multi-temporal aerial photography. Earth
Renssen, H., Isarin, R.F.B., Vandenberghe, J. 2002. Thermal gradients Surface Processes and Landforms 34, 725–737.
in Europe during the last glacial–interglacial transition. Geologie Ridente, D., Trincardi, F. 2002. Eustatic and tectonic control on
en Mijnbouw (Netherlands Journal of Geosciences) 81, 113–122. deposition and lateral variabilty of Quaternary regressive
Renssen, H., Goosse, H., Fichefet, T., et al. 2005. Holocene climate sequences in the Adriatic basin (Italy). Marine Geology 184,
evolution in the high-latitude Southern Hemisphere simulated 273–293.
by a coupled atmosphere–sea ice–ocean–vegetation model. The Ridge, J.C., Balco, G., Bayless, R.L. 2012. The new North American
Holocene 15, 951–964. Varve Chronology: a precise record of southeastern Lauren-
Renssen, H., Kasse, K., Vendenberghe, J., et al. 2007a. Weichselian tide Ice Sheet deglaciation and climate, 18.2–12.5 kyr BP, and
Late Pleniglacial surface winds over northwest and central correlations with Greenland ice core records. American Journal
Europe: a model-data comparison. Journal of Quaternary Science of Science 312, 685–722.
22, 281–293. Ridgwell, A.J. 2002. Dust in the Earth system: the biogeochemical
Renssen, H., Goose, H., Fichefet, T. 2007b. Simulation of Holocene linking of land, air and sea. Philosophical Transactions of the
cooling events in a coupled climate model. Quaternary Science Royal Society of London A 360, 2905–2924.
Reviews 26, 2019–2029. Ridings, R. 1996. Where in the world does obsidian hydration work?
Retallack, G.J. 2001. Soils of the Past, 2nd edn. Blackwell, New York. Antiquity 61, 136–148.
Riera, S., López-Sáez, J.A., Julià, R. 2006. Lake responses to histor-
Reuther, A.U., Urdea, P., Geoger, C., et al. 2007. Late Pleistocene
ical land use changes in northern Spain: the contribution of
glacial chronology of the Pietrele Valley, Retezat Mountains,
non-pollen palynomorphs in a multiproxy study. Review of
southern Carpathians, constrained by 10Be exposure dating and
Palaeobotany and Palynology 141, 127–137.
pedological investigations. Quaternary International 164/165,
Rignot, E., Hallet, B., Fountain, A. 2002. Rock glacier surface motion
151–169.
in Beacon Valley, Antarctica, from synthetic-aperture radar
Reyes, A.V., Wiles, G.C., Smith, D.J., et al. 2006. Expansion of alpine interferometry. Geophysical Research Letters 29, 1607, doi:
glaciers in Pacific North America in the first millennium AD. 10.1029/2001GL013494.
Geology 34, 57–60. Rind, D., Peteet, D., Broecker, W.S., et al. 1986. The impact of
Rhodes, R.H., Bertler, N.A.N., Baker, J.A., et al. 2012. Little Ice Age cold North Atlantic sea surface temperatures on climate. Implica-
climate and oceanic conditions of the Ross Sea, Antarctica from tions for the Younger Dryas cooling. Climate Dynamics 1, 3–33.
a coastal ice core record. Climate of the Past 8, 1223–1238. Ringrose, S., Harris, C., Huntsman-Mapila, P., et al. 2009. Origins
Rial, J.A. 2012. Synchronization of polar climate variability over the of strandline duricrusts around the Makgadikgadi Pans
last ice age: in search of simple rules at the heart of climate’s (Botswana Kalahari) as deduced from their chemical and isotope
complexity. American Journal of Science 312, 417–448. composition. Sedimentary Geology 219, 262–279.
Rial, J.A., Anaclerio, C.A. 2000. Understanding non-linear response Rink, W.J. 1997. Electron spin resonance (ESR) dating and ESR
of the climate system to orbital forcing. Quaternary Science applications in Quaternary science and archaeometry Radiation
Reviews 19, 1709–172. Measurements 27, 975–1025.
504 REFERENCES
Rinterknecht, V., Clark, P.U., Raisbeck, G.M., et al. 2004. Cosmogenic Oasis in the eastern Sahara. Journal of African Earth Sciences 44,
10
Be dating of the Salpausselkä I Moraine in southwestern 229–240.
Finland. Quaternary Science Reviews 23, 2283–2289. Robinson, S.G., Maslin, M.A., McCave, I.N. 1995. Magnetic suscepti-
Rinterknecht, V.R., Marks, L., Piotrowski, J. A., et al. 2005. Cosmo- bility variations in Upper Pleistocene deep-sea sediments on
genic 10Be ages on the Pomeranian Moraine, Poland. Boreas 34, the NE Atlantic: implications for ice-rafting and palaeocircu-
186–191. lation at the last glacial maximum. Paleoceanography 10,
Rinterknecht, V.R., Clark, P.U., Raisbeck, G.M., et al. 2006. The last 221–250.
deglaciation of the southeastern sector of the Scandinavian Ice Robinson, S., Beaudoin, A.B., Froese, D.G., et al. 2007. Plant
Sheet. Science 311, 1449–1452. macrofossils associated with an early Holocene beaver dam in
Rinterknecht, V.R., Pavlovskaya, I.E., Clark, P.U. 2007. Timing of interior Alaska. Arctic 60, 430–438.
the last deglaciation in Belarus. Boreas 36, 307–313. Robson, J., Sutton, R., Lohmann, K., et al. 2012. Causes of the rapid
Rinterknecht, V., Braucher, R., Böse, M., et al. 2012. Late Quaternary warming of the North Atlantic Ocean in the mid-1990s. Journal
ice sheet extents in northeastern Germany inferred from surface of Climate 25, 4116–4134.
exposure dating. Quaternary Science Reviews 44, 89–95. Rodbell, D.T. 1990. Soil-age relationships on Late Quaternary
Rio, D., Sproviere, R., Castradori, D., et al. 1998:. The Gelasian moraines, Arrowsmith Range, Southern Alps, New Zealand.
Stage (Upper Pliocene): a new unit of the global standard Arctic & Alpine Research 22, 355–365.
chronostratigraphic scale. Episodes 21, 82–87. Rodriguez-Lazaro, J., Ruiz-Muñoz, F. 2012. A general introduction
Ritchie, J.C., Cwynar, L.C., Spear, R.W. 1983. Evidence from north- to ostracods: morphology, distribution, fossil record and
west Canada for an early Holocene Milankovitch thermal applications. In Developments in Quaternary Science. Volume 17:
maximum. Nature 305, 126–228. Ostracoda as Proxies for Quaternary Climate Change (edited by
Rixhon, G., Braucher, R., Bourlès, D., et al. 2011. Quaternary river D.J. Horne, J.A. Holmes, J. Rodriguez-Lazaro, et al.). Elsevier,
incision in NE Ardennes (Belgium) – insights from 10Be/26Al Amsterdam, 1–14.
dating of river terraces. Quaternary Geochronology 6, 273–284. Roe, G. 2006. In defense of Milankovitch. Geophysical Research
Robert, C.M. 2009. Developments in Marine Geology. Volume 3: Letters 33, L24703, doi: 10.1029/2006GL027817.
Global Sedimentology of the Ocean: An Interplay Between Geo- Roe, H., van de Plassche, O. 2005. Modern pollen distribution in a
dynamics and Paleoenvironment. Elsevier, Oxford. Connecticut saltmarsh: implications for studies of sea-level
Roberts, A.P., Tauxe, L., Hesliop D. 2013. Magnetic paleointensity change. Quaternary Science Reviews 24, 2030–2049.
stratigraphy and high-resolution Quaternary geochronology: Roe, H.M., Coope, G.R., Devoy, R.J.N., et al. 2009. Differentiation
successes and future challenges. Quaternary Science Reviews 61, of MIS 9 and MIS 11 in the continental record: vegetational,
1–16. faunal, aminostratigraphic and sea-level evidence from coastal
Roberts, D.H., Hart, J.K. 2005. The deforming bed characteristics sites in Essex, UK. Quaternary Science Reviews 28, 2342–2373.
of a stratified till assemblage in north East Anglia, UK: investi- Rogalla, U., Andruleit, H. 2005. Precessional forcing of cocco-
gating controls on sediment rheology and strain signatures. lithophore assemblages in the northern Arabian Sea: implica-
Quaternary Science Reviews 24, 123–140. tions for monsoonal dynamics during the last 200,000 years.
Roberts, H.M., Plater A.J. 2007. Reconstruction of Holocene foreland Marine Geology 217, 31–48.
progradation using optically stimulated luminescence (OSL) Rogers, A.K. 2007. Effective hydration temperature of obsidian: a
dating: an example from Dungeness, UK. The Holocene 17, diffusion theory analysis of time-dependant hydration rates.
495–505. Journal of Archaeological Science 34, 656–665.
Roberts, H.M., Muhs, D.R., Wintle, A.G., et al. 2003. Unprecedented Rogers, R.R., Eberth, D.A., Fiorillo, R. (eds). 2007. Bonebeds: Genesis,
last-glacial mass accumulation rates determined by luminescence Analysis and Palebiological Significance. University of Chicago
dating of loess from western Nebraka. Quaternary Research 59, Press, Chicago.
411–419. Rogerson, M., Bigg, G.R., Rohling, E.J. 2011. Vertical density gradient
Roberts, N., Jones, M.D., Benkaddour, A., et al. 2008. Stable isotope in the eastern North Atlantic during the last 30,000 years. Climate
records of Late Quaternary climate and hydrology from Dynamics 39, 589–598.
Mediterranean lakes: the ISOMED synthesis. Quaternary Science Rogerson, M., Rohling, E.J., Bigg, G.R., et al. 2012. Paleoceanography
Reviews 27, 2426–2441. of the Atlantic–Mediterranean exchange: overview and first
Roberts, N.L., Piotrowski, A.M., McManus, J.F., et al. 2010. quantitative assessment of climatic forcing. Reviews of Geophysics
Synchronous deglacial overturning and water mass source 50, RG2003, doi: 10.1029/2011RG000376.
changes. Science 327, 75–78. Rohling, E., Pälike, H. 2005. Centennial-scale climate cooling with
Robinson C., El-Baz, F., Ozdogan, M., et al. 2000. Use of radar data a sudden cold event around 8,200 years ago. Nature 434, 975–979.
to delineate palaeodrainage flow directions in the Selima Sand Rohling, E., Marsh, R., Wells, N.C., et al. 2004. Similar meltwater
Sheet, Eastern Sahara. Photogrammetric Engineering and Remote contributions to glacial sea level changes from Antarctic and
Sensing 66, 645–753. northern ice sheets. Nature 430, 1016–1021.
Robinson, C.A., El-Baz, F., Al-Saud, T.S.M., et al. 2006. Use of Rohling, E.J., Grant, K., Hemleben, C., et al. 2008. New constraints
radar data to delineate palaeodrainage leading to the Kufra on the timing of sea-level fluctuations during early to middle
REFERENCES 505
marine isotope stage 3. Paleoceanography 23, PA3219, doi: Ross, M., Campbell, J.E., Parent, M., et al. 2009. Palaeo-ice streams
10.1029/2008PA001617. and the subglacial landscape mosaic of the North American
Rolland, N., Larocque, I. 2007. The efficiency of kerosene flotation mid-continental prairies. Boreas 38, 421–439.
for extraction of chironomid head capsules from lake sediments Ross, N., Harris, C., Christiansen, H.H., Brabham, P.J. 2005. Ground
samples. Journal of Paleolimnology 37, 565–572. penetrating radar investigations of open system pingos,
Romero, O., Schmieder, F. 2006. Occurrence of thick Ethmodiscus Adventdalen, Svalbard. Norsk Geografisk Tidsskrift 59, 129–138.
oozes associated with a terminal Mid-Pleistocene Transition Rossignol, J., Moine, O., Rousseau, D.-D. 2004. The Buzzard’s Roost
event in the oligotrophic subtropical South Atlantic. Palaeo- and Eustis mollusc sequences: comparison between the paleo-
geography, Palaeoclimatology, Palaeoecology 235, 321–329. environments of two sites in the Wisconsinan loess of Nebraska,
Romero, O.E., Dupont, L., Wyputta, U., et al. 2003. Temporal USA. Boreas 33, 145–154.
variability of fluxes of eolian-transported freshwater diatoms, Rothwell, R.G. 2006. New Techniques in Sediment Core Analysis.
phytoliths and pollen grains off Cape Blanc as reflection of Geological Society of London, Special Publication 267. Geological
land–atmosphere–ocean interactions in northwest Africa. Journal Society, London.
of Geophysical Research 108, doi: 10.1029/2000JC000375. Roucoux, K.H., de Abreu, L., Shackleton, N.J., et al. 2005. The
Romero-Viana, L., Kienel, U., Sachse, D. 2012. Lipid biomarker response of NW Iberian vegetation to North Atlantic climate
signatures in a hypersaline lake on Isabel Island (Eastern Pacific) oscillations during the last 65 kyr. Quaternary Science Reviews 24,
as a proxy for past rainfall anomaly (1942–2006 AD). Palaeo- 1637–1653.
geography, Palaeoclimatology, Palaeoecology 350–352, 49–61. Round, F. E., Crawford, R.M., Mann, D.G. 1990. The Diatoms:
Roos-Barraclough, F., van der Knaap, W.O., van Leeuwen, J.F.N., Biology and Morphology of the Genera. Cambridge University
et al. 2004. A Late-glacial and Holocene record of climatic Press, Cambridge.
change from a Swiss peat humification profile. The Holocene Rousseau, D.-D. 2001. Loess biostratigraphy: new advances and
14, 21–34. approaches in mollusk studies. Earth-Science Reviews 54,
Rose, J. 1985. The Dimlington Stadial/Dimlington Chronozone: 157–171.
a proposal for naming the main glacial episode of the Late Rousseau, D.-D., Limondin, N., Puissegur, J.-J. 1993. Holocene
Devensian in Britain. Boreas 14, 225–230. environmental signals from mollusc assemblages in Burgundy
Rose, J., Boardman, J., Kemp, R.A., et al. 1985. Palaeosols and the (France). Quaternary Research 40, 237–253.
interpretation of the British Quaternary stratigraphy. In Rousseau, D.-D., Limondin, N., Magnin, F., et al. 1994. Temperature
Geomorphology and Soils (edited by K.S. Richards, R.R. Arnett, oscillations over the last 10,000 years in western Europe estimated
S. Ellis). George Allen and Unwin, London, 348–375. from terrestrial mollusc assemblages. Boreas 23, 66–73.
Rose, J., Whiteman, C.A., Allen, P., et al. 1999. The Kesgrave Sands Rousseau, D.-D., Schevin, P., Duzer, D., et al. 2006. New evidence
and Gravels: ‘preglacial’ Quaternary deposits of the River Thames of long distance pollen transport to southern Greenland in late
in East Anglia and the Thames valley. Proceedings of the Geologists’ spring. Review of Palaeobotany and Palynology 141, 277–286
Association 110, 93–116. Rowe, P.J., Maher, B.A. 2000. ‘Cold’ stage formation of calcrete
Rose, J., Lee, J.A., Kemp, R.A., et al. 2000. Palaeoclimate, sedi- nodules in the Chinese Loess Plateau: evidence from U-series
mentation and soil development during the Last Glacial Stage dating and stable isotope analysis. Palaeogeography, Palaeo-
(Devensian), Heathrow Airport, London, UK. Quaternary Science climatology, Palaeoecology 157, 109–125.
Reviews 19, 827–847. Roy, K., Jablonsk, D., Valentine, J.W. 1995. Thermally anomalous
Rose, J., Moorlock, B.S.P., Hamblin, R.J.O. 2001. Pre-Anglian fluvial assemblages revisited: patterns in the extraprovincial latitudinal
and coastal deposits in eastern England: lithostratigraphy and range shifts of Pleistocene marine mollusks. Geology 23,
palaeoenvironments. Quaternary International 79, 5–22. 1071–1074.
Rose, K.A., Sikes, E.L., Guilderson, T., et al. 2010. Upper-ocean-to- Roy, M., Clark, P.U., Duncan, R.A., et al. 2007. Insights into the late
atmosphere radiocarbon offsets imply fast deglacial carbon Cenozoic configuration of the Laurentide Ice Sheet from
40
dioxide release. Nature 466, 1093–1097. Ar/39Ar dating of glacially transported minerals in midcontinent
Rosell-Melé, A., McClymont, E.L. 2007. Biomarkers as paleoceano- tills. Geochemistry, Geophysics, Geosystems 8, doi: 10.1029/2006
graphic proxies. In Developments in Marine Geology. Volume 1: GC001572.
Proxies in Late Cenozoic Paleoceanography (edited by C. Hillaire- Ruddiman, W.F. 2003a. Orbital insolation, ice volume and green-
Marcel, A. de Vernal). Elsevier, Amsterdam, 441–509. house gases. Quaternary Science Reviews 22, 1597–1629.
Rosen, D., Dumayne-Peaty, L. 2001. Human impact on the vegetation Ruddiman, W.F. 2003b. The anthropogenic greenhouse era began
of South Wales during late historical times: palynological; and thousands of years ago. Climate Change 61, 261–293.
palaeoenvironmental results from Crymlyn Bog NNR, West Ruddiman, W.F. 2005a. Plows, Plagues and Petroleum. Princeton
Glamorgan, Wales, UK. The Holocene 11, 11–23. University Press, Princeton, NJ.
Rosendahl, D., Ulm, S., Weisler, M.I. 2007. Using foraminifera Ruddiman, W.F. 2005b. The early anthropogenic hypothesis a year
to distinguish between natural and cultural shell deposits in later. Climatic Change 69, 427–434.
coastal eastern Australia. Journal of Archaeological Science 34, Ruddiman, W.F. 2006. Ice-driven CO2 feedback on ice volume.
1584–1593. Climate of the Past 2, 43–55.
506 REFERENCES
Ruddiman, W.F. 2013. The Anthropocene. Annual Review of Earth Sacchetti, F., Benetti, S., Quinn, R., et al. 2013. Glacial and post-
and Planetary Sciences 41, 1–24. glacial sedimentary processes in the Irish Rockall Trough from
Ruddiman, W.F., McIntyre, A. 1976. Northeast Atlantic palaeo- an integrated acoustic analysis of near-seabed sediments. Geo-
climatic changes over the past 600,000 years. Geological Society Marine Letters 33, 49–66.
of America Memoir 145, 111–146. Sachs, J.P., Anderson, R.F. 2005. Increased productivity in the sub-
Ruddiman, W.F., Kutzbach, J.E. 1990. Late Cenozoic plateau uplift antarctic ocean during Heinrich events. Nature 434, 1118–1121.
and climate change. Transactions of the Royal Society of Edinburgh: Sack, D. 2009. Evidence for climate change from desert basin
Earth Sciences 81, 301–314. palaeolakes. In Geomorphology of Desert Environments, 2nd edn
Ruddiman, W.F., Thomson, J.F. 2001. The case for human causes (edited by A.J. Parsons, A. Abrahams). Springer, Dordrecht,
of increased CH4 over the last 5000 years. Quaternary Science 743–756.
Reviews 20, 1769–1777. Saher, M., Kristensen, D.K., Hald, M., et al. 2012. Changes in
Ruddiman, W.F., Raymo, M.E. 2003. A methane-based time scale distribution of calcareous benthic foraminifera in the central
for Vostok ice. Quaternary Science Reviews 22, 141–155. Barents Sea between the periods 1965–1992 and 2005–2006.
Ruddiman, W.F., Ellis, E.C. 2009. Effect of per-capita land use Global and Planetary Change 98–99, 81–96.
changes on Holocene forest clearance. Quaternary Science Reviews Saikku, R., Stott, L., Thunell, R. 2009. A bi-polar signal recorded in
27/28, 3011–3015. the western tropical Pacific: Northern and Southern Hemisphere
Ruddiman, W.F., Vavrus, S.J., Kutzbach, J.E. 2005. A test of the climate records from the Pacific warm pool during the last Ice
overdue-glaciation hypothesis. Quaternary Science Reviews 24, Age. Quaternary Science Reviews 28, 2374–2385.
1–10. St. Jacques, J.-M., Cumming, B.F., Smol, J.P. 2008. A pre-European
Ruddiman, W.F., Kutzbach, J.E., Vavrus, S.J. 2011. Can natural or settlement pollen–climate calibration set for Minnesota, USA:
anthropogenic explanations of Late Holocene CO2 and CH4 developing tools for palaeoclimatic reconstructions. Journal of
increases be falsified? The Holocene 21, 865–879. Biogeography 35, 306–324.
Ruiz, F., Abad, M., Cáceres, L.M., et al. 2010. Ostracods as tsunami St. Jacques, J.-M., Cumming, B.F., Smol, J.P. 2009. A 900-yr diatom
tracers in Holocene sequences. Quaternary Research 73, 130–135. and chrysophyte record of spring mixing and summer
Ruiz, Z., Brown, A.G., Langdon, P.G. 2006. The potential of chiro- stratification from varved Lake Mina, west-central Minnesota,
nomid (Insecta: Diptera) larvae in archaeological investigations USA. The Holocene 19, 537–547.
of floodplain and lake settlements. Journal of Archaeological Salvador, A. (ed.). 1994. International Stratigraphic Guide: A Guide
Science 33, 14–33. to Stratigraphic Classification, Terminology and Procedure, 2nd
Rundgren, M., Björck, S. 2003. Late-glacial and early Holocene edn. International Commission on Stratigraphy, Geological
variations in atmospheric CO2 concentration indicated by Society of America, Boulder, CO.
high-resolution stomatal index data. Earth and Planetary Science Salvigsen, O., Forman, S.L., Miller, G.H. 1992. Thermophilous
Letters 213, 191–204. molluscs on Svalbard during the Holocene and their paleo-
Ruszkiczay-Rüdiger, Z., Fodor, L., Bada, G., et al. 2005. Quantifi- climatic implications. Polar Research 11, 1–10.
cation of Quaternary vertical movements in the central Pan- Sambridge, M., Grün, R., Eggins, S. 2012. U-series dating of bone
nonian Basin: A review of chronologic data along the Danube in an open system: the diffusion-adsorption-decay model.
River, Hungary. Tectonophysics 410, 157–172. Quaternary Geochronology 9, 42–53.
Rutherford, S., D’Hoydt, S. 2000. Early onset and tropical forcing of Sampson, K.M., Smith, L.C 2006. Relative ages of Pleistocene
100,000-year Pleistocene glacial cycles. Nature 408, 72–75. moraines discerned from pebble counts: Eastern Sierra Nevada,
Rutter, N.W., Ding, Z. 1993. Palaeoclimates and monsoon variations California. Physical Geography 27, 223–235.
interpreted from micromorphogenic features of the Baoji Sánchez Goñi, M.F., Landais, A., et al. 2008. Contrasting impacts of
palaeosols, China. Quaternary Science Reviews 12, 853–862. Dansgaard–Oeschger events over a western European latitudinal
Rutter, N.W., Ding, Z., Liu T. 1996. Long paleoclimate records from transect modulated by orbital parameters. Quaternary Science
China. Geophysica 32, 7–34. Reviews 27, 1136–1151,
Rutter, N.W., Velichko, A.A., Dlussky, K.G., et al. 2006. New insights Santacreu, D.A., Vicens, G.M. 2012. Raw materials and pottery
on the loess/paleosol Quaternary stratigraphy from key sections production at the Late Bronze and Iron Age Site of Puig de Sa
in the US Midwest. Catena 67, 15–34 Morisca, Mallorca, Spain. Geoarchaeology 27, 285–299.
Ryder, J.M. 1971. The stratigraphy and morphology of paraglacial Sapart, C.J., Monteil, G., Prokopiou, M., et al. 2012. Natural and
alluvial fans in south-central British Columbia. Canadian Journal anthropogenic variations in methane sources during the past
of Earth Sciences 8, 1252–1264. two millennia. Nature 490, 85–88.
Ryves, D.B., Jewson, D.H., Sturm, M., et al. 2003. Quantitative and Sarala, P., Peuraniemi, V. 2007. Exploration using till geochemistry
qualitative relationships between planktonic diatom commun- and heavy minerals in the ribbed moraine area of southern
ities and diatom assemblages in sedimenting material and surface Finnish Lapland. Geochemistry: Exploration, Environment,
sediments in Lake Baikal, Siberia. Limnology and Oceanography Analysis 7, 195–205.
48, 1643–1661. Sarnthein, M., Kiefer, T., Grootes, P.M., et al. 2006. Warmings in
Saarinen, T. 1999. Palaeomagnetic dating of Late Holocene sediments the far northwestern Pacific promoted pre-Clovis immigration
in Fennoscandia. Quaternary Science Reviews 18, 889–897. to America during Heinrich event 1. Geology 34, 141–144.
REFERENCES 507
Sawada, M., Viau, A.E., Vettoretti, G., et al. 2004. Comparison of Scholz, D., Hoffman, D.L., Hellstrom, J., et al. 2012. A comparison
North-American pollen-based temperature and global lake- of different methods for speleothem age modelling. Quaternary
status with CCCma AGCM2 output at 6 ka. Quaternary Science Geochronology 14, 94–104.
Reviews 23, 225–244. Schöne, B.R. 2003. A ‘clam-ring’master-chronology constructed
Saxena, A. 2005. Text Book of Mollusca. Discovery Publishing House, from a short-lived bivalve mollusc from the northern Gulf of
New Delhi. California, USA. The Holocene 13, 39–50.
Scarciglia, F., Pulice, I., Robustelli, G., et al. 2006. Soil chrono- Schöne, B.R., Feibig, J. 2009. Seasonality in the North Sea during
sequences on Quaternary marine terraces along the north- the Allerød and Late Medieval Climate Optimum using
western coast of Calabria (southern Italy). Quaternary Inter- bivalve sclerochronology. International Journal of Earth Science
national 156/157, 133–155. (Geologische Rundschau) 98, 83–98.
Schaefer, J.M., Denton, G.H., Barrell, D.J.A., et al. 2006. Near- Schöne, B.R., Oschmann, W., Rössler, J., et al. 2003. North Atlantic
synchronous interhemispheric termination of the Last Glacial Oscillation dynamics recorded in shells of a long-lived bivalve
Maximum in mid-latitudes. Science 312, 1510–1513. mollusk. Geology 31, 1037–1040.
Scharpenseel, H.W., Becker-Heidemann, P. 1992. Twenty-five years Schöne, B.R., Dunca, E., Mutvei, H., et al. 2004. A 217-year record
of radiocarbon dating soils: paradigm erring and learning. of summer air temperature reconstructed from freshwater
Radiocarbon 34, 541–549. pearl mussels (M. margarifitera, Sweden). Quaternary Science
Schellenberger, A., Heller, F., Veit, H. 2003. Magnetostratigraphy Reviews 23, 1803–1816.
and magnetic susceptibility of the Las Carreras loess–paleosol Schönfeld, J. 2012. History and development of methods in Recent
sequence in Valle de Tafí, Tucumán, NW-Argentina. Quaternary benthic foraminiferal studies. Journal of Micropalaeontology
International 106/107, 159–167. 31, 53–72.
Schellmann, G., Radtke, U. 2004. A revised morpho- and chrono- Schrag, D.P., Hampt, G., Murray, D.W. 1996. Pore fluid constraints
stratigraphy of the Late and Middle Pleistocene coral reef terraces on the temperature and oxygen isotopic composition of the
on Southern Barbados (West Indies). Earth-Science Reviews 64, glacial ocean. Science 272, 1930–1932.
157–187. Schramm, A., Stein, M., Goldstein, S.L. 2000. Calibration of the
14
Schellmann, G., Beerten, K., Radtke, U. 2008. Electron spin resonance C timescale to >40 ka by 234U-230Th dating of Lake Lisan
(ESR) dating of Quaternary materials. Eiszeitalter und Gegenwart sediments (last glacial Dead Sea). Earth and Planetary Science
(Quaternary Science Journal) 57, 150–178. Letters 175, 27–40.
Schellnhuber, H.J. 1999. ‘Earth System’ analysis and the second Schreve, D. 2007. Vertebrate overview. In Encyclopedia of Quaternary
Copernican revolution. Nature 402 (Suppl.), C19–C23. Science (edited by S.A. Elias). Elsevier, Amsterdam, 3123–3132.
Schettler, G., Liu, Q., Mingram, J., et al. 2006. Palaeovariations in Schreve, D.C., Thomas, G.N. (eds). 2001. European Quaternary
the East-Asian monsoon regime geochemically recorded in biostratigraphy. Quaternary Science Reviews 20, 1577–1764.
varved sediments of Lake Sihailongwan (Northeast China, Schreve, D., Candy, I. 2010. Interglacial climates: advances in our
Jilin Province). Part 1. Hydrological conditions and dust flux. understanding of warm climate episodes. Progress in Physical
Journal of Paleolimnology 35, 239–270. Geography 34, 845–856.
Schlitzer, R. 2003. eWoce – Electronic Atlas of WOCE Hydrographic Schreve, D.C., Keen, D.H., Limondon-Lozouet, N., et al. 2007.
and Tracer Data. Alfred Wegener Institute for Polar and Marine Progress in faunal biostratigraphy of Late Cenozoic fluvial
Research, Bremerhaven, http://www.ewoce.org. sequences during IGCP 449. Quaternary Science Reviews 26,
Schmidt, G.A. 2010. Enhancing the relevance of palaeoclimate 2970–2995.
model/data comparisons for assessments of future climate Schulz, K.G., Zeebe, R.E. 2006. Pleistocene glacial terminations
change. Journal of Quaternary Science 25, 79–87. triggered by synchronous changes in Southern and Northern
Schmieder, J., Fritz, S.C., Swinehart, J.B., et al. 2011. A regional-scale Hemisphere insolation: the insolation canon hypothesis. Earth
climate reconstruction of the last 4000 years from lakes in the and Planetary Science Lettters 249, 326–336.
Nebraska Sand Hills, USA. Quaternary Science Reviews 30, Schulz, M. 2002. On the 1470-year pacing of Dansgaard–Oeschger
1797–1812. warm events. Paleoceanography 17, doi: 10.1029/2000PA000571.
Schmittner, A., Saenko, O.A., Weaver, A.J. 2003. Coupling of the Schurr, M.R., Gregory, D.A. 2002. Fluoride dating of faunal materials
hemispheres in observations and simulations of glacial climate by ion-selective electrode: high resolution relative dating at an
change. Quaternary Science Reviews 22, 659–671. early agricultural period site in the Tucson basin. American
Schneider, B., Schmittner, A. 2006. Simulating the impact of the Antiquity 67, 281–299.
Panamanian seaway closure on ocean circulation, marine pro- Schuster, M., Roquin, C., Duringer, P., et al. 2005. Holocene Lake
ductivity and nutrient cycling. Earth and Planetary Science Letters Mega-Chad palaeoshorelines from space. Quaternary Science
246, 367–380. Reviews 24, 1821–1827.
Schnurrenberger, D., Russell, J., Kelts, K. 2003. Classification of Schüttenhelm, R.T.E., Laban, C. 2005. Heavy minerals, provenance
lacustrine sediments based on sedimentary components. Journal and large scale dynamics of seabed sands in the Southern North
of Paleolimnology 29, 141–154. Sea: Baak’s (1936) heavy mineral study revisited. Quaternary
Scholz, D., Magini, A., Felis, T. 2004. U-series dating of diagenetic- International 133/134, 170–193.
ally altered fossil reef corals. Earth and Planetary Science Letters Schweingruber, F.H. 1996. Tree Rings and Environment: Dendro-
218, 163–178 ecology. Paul Haupt, Bern.
508 REFERENCES
Scott, D.B., Collins, E.S., Gayes, PT., et al. 2003. Records of prehistoric Seong Y.B., Owen, L.A., Yi, C., et al. 2009. Geomorphology of
hurricanes on the South Carolina coast records. Geological Society anomalously high glaciated mountains at the northwestern end
of America Bulletin 115, 1027–1039. of Tibet: Muztag Ata and Kongur Shan. Geomorphology 103,
Scott, D.B., Schell, T., Rochon, A., et al. 2008. Benthic foraminifera 227–250.
in the surface sediments of the Beaufort Shelf and slope, Beaufort Seppä, H., Bennett, K.D. 2003. Quaternary pollen analysis: recent
Sea, Canada: applications and implications for past sea-ice progress in palaeoecology and palaeoclimatology. Progress in
conditions. Journal of Marine Systems 74, 840–863. Physical Geography 27, 548–579.
Scott, E.M., Cook, G.T., Naysmith, P. 2010a. A Report on Phase 2 Seppä, H., Birks, H.J.B., Odland, A., et al. 2004. A modern
of the Fifth International Radiocarbon Intercomparison (VIRI). pollen–climate calibration set from northern Europe: developing
Radiocarbon 52, 846–858. and testing a tool for palaeoclimatological reconstructions.
Scott, E.M., Cook, G.T., Naysmith, P. 2010b. The Fifth International Journal of Biogeography 31, 251–267.
Radiocarbon Intercomparison (VIRI): an assessment of labora- Serefiddin, F., Schwarcz, H.P., Ford, D.C., et al. 2004. Late Pleistocene
tory performance in Stage 3. Radiocarbon 52, 859–865. paleoclimate in the Black Hills of South Dakota from isotope
Scourse, J., Richardson, C., Forsythe, G., et al. 2006. First cross- records in speleothems. Palaeogeography, Palaeoclimatology,
matched floating chronology from the marine fossil record: Palaeoecology 203, 1–17.
data from growth lines of the long-lived bivalve mollusc Arctica Serieyssol, K., Chatelard, S., Cubizolle, H. 2011. Diatom fossils in
islandica. The Holocene 16, 967–974. mires: a protocol for extraction, preparation and analysis in
Scourse, J.D., Haapaniemi, A.I., Colmenero-Hidalgo, E., et al. palaeoenvironmental studies. Mires and Peat 7(art. 12), 1–11,
2009. Growth, dynamics and deglaciation of the last British–Irish http://www.mires-and-peat.net/.
ice sheet: the deep-sea ice-rafted detritus record. Quaternary Severinghaus, J.P. 2009. Southern see-saw seen. Nature 457,
Science Reviews 28, 3066–3084. 1093–1094.
Scrivner, A.E., Vance, D., Rohling, E.J. 2004. New neodymium Shackleton, N.J. 1987. Oxygen isotopes, ice volume and sea level.
isotope data quantify Nile involvement in Mediterranean anoxic Quaternary Science Reviews 6, 183–190.
episodes. Geology 32, 565–568. Shackleton, N.J. 2000. The 100,000-year ice-age cycle identified and
Sedov, S., Sycheva, S., Targulian, V., et al. 2013. Last Interglacial found to lag temperature, carbon dioxide and orbital eccentricity.
paleosols with Argic horizons in Upper Austria and Central
Science 289, 1897–1902.
Russia – pedogenetic and palaeoenvironmental inferences from
Shackleton, N.J. 2006. Formal Quaternary stratigraphy – what do we
comparison with the Holocene analogues. Eiszeitalter und
expect and need? Quaternary Science Reviews 25, 3458–3462.
Gegenwart (Quaternary Science Journal) 62, 44–58.
Shackleton, N.J., Opdyke, N.D. 1973. Oxygen isotope and palaeo-
Seidov, D., Maslin, M. 2001. Atlantic Ocean heat piracy and the bipolar
magnetic stratigraphy of equatorial Pacific core V28-238: oxygen
climate see-saw during Heinrich and Dansgaard–Oeschger
isotope temperatures and ice volume on a 105 and 106 year scale.
events. Journal of Quaternary Science 16, 321–328.
Seidov, D., Haupt, B.J., Barron, E.J., et al. 2001. Ocean bi-polar seesaw:
Shackleton, N.J., Berger, A., Peltier, W.R. 1990. An alternative
southern versus northern meltwater impacts. In The Oceans
astronomical calibration of the lower Pleistocene timescale
and Rapid Climate Change: Past, Present & Future (edited by
D. Seidov, B.J. Haupt, M. Maslin). American Geophysical Union, based on ODP Site 677. Transactions of the Royal Society of
Washington DC, 169–197. Edinburgh: Earth Science 81, 251–261.
Seidov, D., Stouffer, R.J., Haupt, B.J. 2005. Is there a simple bi-polar Shackleton, N.J., Fairbanks, R.G., Chiu, T.-C., et al. 2004. Absolute
ocean seesaw? Global and Planetary Change 49, 19–27. calibration of the Greenland time scale: implications for Ant-
Sejrup, H.P., Hjelstuen, B.O., Dahlgren, K.I.T., et al. 2005. Pleisto- arctic time scales and for Δ14C. Quaternary Science Reviews 23,
cene glacial history of the NW European continental margin. 1513–1522
Marine and Petroleum Geology 22, 1111–1129. Shakesby, R.A. 1997. Pronival (protalus) ramparts: a review of
Sejrup, H.P., Nygård, A., Hall., A.M., et al. 2009. Middle and Late forms, processes, diagnostic criteria and palaeoenvironmental
Weichselian (Devensian) glaciation history of south-western implications. Progress in Physical Geography 21, 394–418.
Norway, North Sea and eastern UK. Quaternary Science Reviews Shakesby, R.A., Matthews, J.A., Winkler, S. 2004. Glacier vari-
28, 370–380. ations in Breihemen, southern Norway: relative-age dating of
Selby, K., Smith, D.E. 2007. Late Devensian and Holocene relative Holocene moraine complexes at six high-altitude glaciers. The
sea-level changes on the Isle of Skye, Scotland, UK. Journal of Holocene 14, 899–910.
Quaternary Science 22, 119–139. Shakesby, R.A., Matthews, J.A., Owen, G. 2006. The Schmidt hammer
Sella, G.F., Stein, S., Dixon, T.H., et al. 2007. Observation of glacial as a relative-age tool and its potential for calibrated-age dating
isostatic adjustment in ‘stable’ North America with GPS. in Holocene glaciated environments. Quaternary Science Reviews
Geophysical Research Letters 34, doi: 10.1029/2006GL027081. 25, 2846–2867.
Sémah, F., Salekei, H., Falguères, C. 2000. Did early man reach Java Shakun, J.D., Clark, P.U., Feng, H., et al. 2012. Global warming
during the Late Pliocene? Journal of Archaeological Science preceded by increasing carbon dioxide concentrations during
27, 763–769. the last deglaciation. Nature 484, 49–55.
Sen Gupta, B.K. (ed.). 2002. Modern Foraminifera. Kluwer Academic Shanahan, T.M., Overpeck, J.T., Anchukaitis, K.J., et al. 2009. Atlantic
Publishers, Dordrecht. forcing of persistent drought in West Africa. Science 324, 377–380.
REFERENCES 509
Shane, P. 2005. Towards a comprehensive distal andesitic tephro- Siegert, M.J. 2001. Ice Sheets and Late Quaternary Environmental
stratigraphic framework for New Zealand based on eruptions Change. John Wiley, Chichester and New York.
from Egmont volcano. Journal of Quaternary Science 20, 45–57. Siegert, M.J. 2005. Lakes beneath the ice sheet: the occurrence,
Shane, P.A., Narin, I.A., Martin, S.B., et al. 2008. Compositional analysis, and future exploration of Lake Vostok and other
heterogeneity in tephra deposits resulting from the eruption of Antarctic subglacial lakes. Annual Review of Earth and Planetary
multiple magma bodies: implications for tephrochronology. Sciences 33, 215–245.
Quaternary International 178, 44–53. Siegert, M.J., Dowdeswell, J.A. 2004. Numerical reconstructions of
Sharp, M., Dowdeswell, J.A., Gemmell, J.C. 1989. Reconstructing past the Eurasian Ice Sheet and climate during the Late Weichselian.
glacier dynamics and erosion from glacial geomorphic evidence: Quaternary Science Reviews 23, 1273–1284.
Snowdon, North Wales. Journal of Quaternary Science 4, 115–130. Sierro, F.J., Andersen, N., Bassetti, M.A., et al. 2009. Phase
Sharp, W.D., Ludwig, K.R. Chadwick, O.A., et al. 2003. Dating relationship between sea level and abrupt climate change.
fluvial terraces by 230Th/U on pedogenic carbonate, Wind River Quaternary Science Reviews 28, 2867–2881.
Basin, Wyoming. Quaternary Research 59, 139–150. Sigman, D.M., Boyle, E.A. 2000. Glacial/interglacial variations in
Shaw, J., Gareau, P., Courtney, R.C. 2002. Paleogeography of Atlantic atmospheric carbon dioxide. Nature 407, 859–869.
Canada 13–0 kyr. Quaternary Science Reviews 21, 1861–1878. Sigman, D.M., Hain, M.P., Haug, G.H. 2010. The polar and glacial
Shennan, I., Horton, B. 2002. Holocene land- and sea-level changes cycles in atmospheric CO2 concentration. Nature 466, 47–55.
in Great Britain. Journal of Quaternary Science 17, 511–526. Sime, L.C., Ferguson, R.I. 2003. Information on grain sizes in
Shennan, I., Innes, J., Long, A.J., et al. 1994. Late Devensian and gravel-bed rivers by automated image analysis. Journal of
Holocene sea-level changes at Loch an Eala near Arisaig, Sedimentary Research, Section B: Stratigraphy and Global Studies
northwest Scotland. Journal of Quaternary Science 9, 261–284. 73, 630–636.
Shennan, I., Bradley, S., Milne, G., et al. 2006a. Relative sea-level Sime, L.C., Wolff, E.W., Oliver, K.I.C., et al. 2009. Evidence for warmer
changes, glacial isostatic modelling and ice-sheet reconstructions interglacials in East Antarctic ice cores. Nature 462, 342–346.
from the British Isles since the Last Glacial Maximum. Journal Simms, A.R., Lambeck, K., Purcell, A., et al. 2007. Sea-level history
of Quaternary Science 21, 585–599. of the Gulf of Mexico since the Last Glacial Maximum with
Shennan, I., Hamilton, S., Hiller, C., et al. 2006b. Relative sea-level implications for the melting history of the Laurentide Ice Sheet.
observations in western Scotland since the Last Glacial maximum
Simpson, G.L., Shilland, E.M., Winterbottom, J.M., et al. 2005.
for testing models of glacial isostatic land movements and ice-
Defining reference conditions for acidified waters using a modern
sheet reconstructions. Journal of Quaternary Science 21, 601–613.
analogue approach. Environmental Pollution 137, 119–133.
Sherwood, S.C., Huber, M. 2010. An adaptability limit to climate
Simpson, M.J.R., Milne, G.A., Huybrechts, P., et al. 2009. Calibrating
change due to heat stress. Proceedings of the National Academy
a glaciological model of the Greenland ice sheet from the Last
of Sciences of the United States of America 107, 9552–9555.
Glacial Maximum to present-day using field observations of
Shi, C., Zhu, R., Glas, B.P., et al. 2003. Climate variations since the
relative sea level and ice extent. Quaternary Science Reviews 28,
last interglacial recorded in Czech loess. Geophysical Research
1631–1657.
Letters 30, doi: 10.1029/2003GL017251.
Singer, B.S., Ackert, R.P., Guillou, H. 2004. 40Ar/39Ar and K–Ar
Shimada, C., Sato, T., Toyoshima, S., et al. 2008. Paleoecological chronology of Pleistocene glaciations in Patagonia. Geological
significance of laminated diatomaceous oozes during the middle- Society of America Bulletin, 434–450.
to-late Pleistocene, North Atlantic Ocean (IODP Site U1304). Singhvi, A.K., Porat, N. 2008. Impact of luminescence dating on
Marine Micropaleontology 69, 139–150. geomorphological and palaeoclimate research in drylands. Boreas
Shindell, D.T., Schmidt, G.A., Miller, R.L., et al. 2003. Volcanic and 37, 536–558.
solar forcing of climate change during the Preindustrial Era. Singhvi, A.K., Williams, M.A.J., Rajaguru, S.N., et al. 2010. A c. 200
Journal of Climate 16, 4094–4107. ka record of climatic change and dune activity in the Thar Desert,
Shuman, B., Newby, P., Huang, Y., et al. 2004. Evidence for the close India. Quaternary Science Reviews 29, 3095–3105.
climatic control of New England vegetation history. Ecology 85, Sinka, H.J., Atkinson, T.C. 1999. A mutual climatic range method
1297–1310. for reconstructing palaeoclimate from plant remains. Journal of
Siddall, M., Rohling, E.J., Almogi-Labin, A., et al. 2003. Sea-level the Geological Society of London 156, 381–396
fluctuations during the last glacial cycle. Nature 423, 853–858. Sissons, J.B. 1967. The Evolution of Scotland’s Scenery. Oliver and Boyd,
Siddall, M., Rohling, E.J., Thompson, W.G., et al. 2008. Marine isotope Edinburgh.
stage 3 sea level fluctuations: data synthesis and new outlook. Sissons, J.B., 1974. A Late-glacial ice cap on the central Grampians,
Review of Geophysics 46, RG4003, doi: 10.1029/2007RG000226. Scotland. Transactions of the Institute of British Geographers 62,
Siddall, M., Hönisch, B., Waelbroeck, C., et al. 2010a. Changes in deep 95–114.
Pacific temperature during the mid-Pleistocene Transition and Sissons, J.B. 1979. The Loch Lomond Stadial in the British Isles. Nature
Quaternary. Quaternary Science Reviews 29,170–181. 280, 1990–202.
Siddall, M., Kaplan, M.R., Schaefer, J.M., et al. 2010b. Changing Sissons, J.B., 1980. Palaeoclimatic inferences from Loch Lomond
influence of Antarctic and Greenlandic temperature records on Advance glaciers. In Studies in the Lateglacial of North-West
sea-level over the last glacial cycle. Quaternary Science Reviews Europe (edited by J.J. Lowe, J.M. Gray, J.E. Robinson). Pergamon
29, 410–423. Press, Oxford, 31–44.
510 REFERENCES
Sivakumar, B. 2004. Chaos theory in geophysics: past, present and Smol, J.P., Stoermer, E.F. (eds). 2010. The Diatoms: Applications for
future. Chaos, Solitons & Fractals 19, 441–462. the Environmental and Earth Sciences, 2nd edn. Cambridge
Sjunneskog, C., Scherer, R.P. 2005. Mixed diatom assemblages in University Press, Cambridge.
glacigenic sediment from the central Ross Sea, Antarctica. Snowball, I., Sandgren, P. 2002. Geomagnetic field variations in
Palaeogeography, Palaeoclimatology, Palaeoecology 218, 287–300. northern Sweden during the Holocene quantified from varved
Skinner, L.C., Fallon, S., Waelbroeck, C., et al. 2010. Ventilation of lake sediments and their implications for cosmogenic nuclide
the deep Southern Ocean and deglacial CO2 rise. Science 328, production rates. The Holocene 12, 517–530.
1147–1151. Snowball, I., Zillen, L., Ojala, A., Saarinen, T., et al. 2007. FENNO-
Smart, D. 2003. Sub-sea level speleothems from the Andaman coast STACK and FENNORPIS: varve dated Holocene palaeomagnetic
of southern Thailand, and sea-level change in Southeast Asia. Cave secular variation and relative palaeointensity stacks for
and Karst Science 30, 39–42. Fennoscandia. Earth and Planetary Science Letters 255, 106–115
Smart, P.L. 1991a. Uranium series dating. In Quaternary Dating Snowball, T., Thompson, R. 1992. a mineral magnetic study of
Methods – A User’s Guide (edited by P.L. Smart, P.D. Frances). Holocene sediment yields and deposition patterns in the Llyn
Technical Guide 4. Quaternary Research Association, Cam- Geirionydd catchment, north Wales. The Holocene 2, 238–248.
bridge, 45–83. Solomina, O., Haeberli, W., Kull, C., et al. 2008. Historical and
Smart, P.L. 1991b. Electron Spin Resonance (ESR) dating. In Holocene glacier–climate variations: general concepts and
Quaternary Dating Methods – A User’s Guide (edited by P.L. overview. Global and Planetary Change 60, 1–9.
Smart, P.D. Frances). Technical Guide 4. Quaternary Research Solomon, S., Qin, D., Manning, M et al. (eds). 2007. Climate Change
Association, Cambridge, 128–160. 2007: The Physical Science Basis. Contribution of Working
Smith, A.M., Murray, T., Nicholls, K.W., et al. 2007. Rapid erosion, Group I to the Fourth Assessment Report of the Intergovern-
drumlin formation, and changing hydrology beneath an mental Panel on Climate Change. Cambridge University Press,
Antarctic ice stream. Geology 35, 127–130. Cambridge.
Smith, C.W. 2003. Archaeological Conservation Using Polymers. Texas Sonnett, C.P., Finney, S.A. 1990. The spectrum of radiocarbon.
A & M University, Anthropology Series 6. Texas A & M University Philosophical Transactions of the Royal Society, London A330,
Press, College Station. 238–248.
Smith, D., Kenward, H. 2011. Roman grain pests in Britain: Spalding, M.D., Fox, H.E., Allen, G.R., et al. 2007. Marine ecoregions
implications for grain supply and agricultural production. of the world: a bioregionalization of coastal and shelf areas.
Britannia 42, 243–262. BioScience 57, 573–583.
Smith, D.E., Shi, S., Cullingford, R.A., et al. 2004. The Holocene Speer, J.H. 2010. Fundamentals of Tree Ring Research. University of
Storegga Slide tsunami in the United Kingdom. Quaternary Arizona Press, Tucson.
Science Reviews 23, 2291–2321. Speirs, J.C., McGowan, H.A., Neil, D.T. 2008. Meteorological
Smith, D.E., Fretwell, P.T., Cullingford, R.A., et al. 2006. Towards controls on sand transport and dune morphology in a polar-
improved empirical isobase models of Holocene land uplift for desert: Victoria Valley, Antarctica. Earth Surface Processes and
mainland Scotland, UK. Philosophical Transactions of the Royal Landforms 33, 1875–1891.
Society of London A 364, 949–972. Spencer, J., Frizzelle, B.G., Page, P.H., et al. 2003a. Global Positioning
Smith, D.E., Davies, M.H., Brooks, C.L., et al. 2010. Holocene relative System: a Field Guide for the Social Sciences. Blackwell, Malden,
sea levels and related prehistoric activity in the Forth lowland, MA and Oxford.
Scotland, United Kingdom. Quaternary Science Reviews 29, Spencer, L.M., Van Valkenburgh, B., Harris, J.M. 2003b.
2382–2410. Taphonomic analysis of large mammals recovered from the
Smith, D.E., Hunt, N., Firth, C.R., et al. 2012. Patterns of Holocene Pleistocene Rancho La Brea tar seeps. Paleobiology 29, 561–575.
relative sea level change in the north of Britain and Ireland. Speranza, A., van Geel, B., van der Plicht, J. 2002. Evidence for solar
Quaternary Science Reviews 54, 58–76. forcing of climate change at ca. 850 cal. BC from a Czech peat
Smith, G.I. 2009. Late Cenozoic geology and lacustrine history of sequence. Global and Planetary Change 35, 51–65.
Searles Valley, Inyo and San Bernardino Counties, California. US Spötl, C., Mangini, A. 2002. Stalagmite from the Austrian Alps
Geological Survey Professional Paper 1727. reveals Dansgaard–Oeschger events during isotope stage 3:
Smith, G.I., Street-Perrott, A.F. 1983. Pluvial lakes of the western implications for the absolute chronology of Greenland ice cores.
United States. In Late Quaternary Environments of the United Earth and Planetary Sciences Letters 203, 507–518.
States. 1. The Late Pleistocene (edited by S.C. Porter). Longman, Spötl, C., Magnini, A. 2007. Speleothems and palaeoglaciers. Earth
London, 190–212. and Planetary Science Letters 254, 323–331.
Smith, M.A., Watchman, A., Ross, J. 2009. Direct dating indicates a Spötl, C., Scholz, D., Mangini, A. 2008. A terrestrial U/Th-dated stable
Mid Holocene age for archaic rock engravings in arid Central isotope record of the Penultimate Interglacial. Earth And
Australia. Geoarchaeology 24, 191–203. Planetary Science Letters 276, 283–292.
Smith, M.W., Riseborough, D.W. 2002. Climate and the limits of Staff, R.A., Ramsey, C.B., Bryant, C.L., et al. 2011. New 14C
permafrost: a zonal analysis. Permafrost and Periglacial Processes determinations from Lake Suigetsu, Japan: 12,000 to 0 cal BP.
13, 1–15 Radiocarbon 53, 3, 511–528.
REFERENCES 511
Stager, J.C., Ruzmaikin, A., Conway, D., et al. 2007. Sunspots, El Niño, Sternberg, R.S. 2001. Magnetic properties and archaeomagnetism.
and the levels of Lake Victoria, East Africa. Journal of Geophysical In Handbook of Archaeological Sciences (edited by D.R. Brothwell,
Research 112, D15, 16: doi: 10.1029/2006/JD008362. A.M. Pollard). John Wiley, Chichester and New York, 73–79.
Stahl, P.W. 1996. The recovery and interpretation of microvertebrate Stevens, R.E., Metcalfe, S.E., Leng, M.J., et al. 2012. Reconstruction
bone assemblages from archaeological contexts. Journal of of late Pleistocene climate in the Valsequillo Basin (Central
Archaeological Method and Theory 3, 31–75. Mexico) through isotopic analysis of terrestrial and fresh-
Stahle, D.W., Fyfe, F.K., Cook, E.R., et al. 2007. Tree-ring recon- water snails. Palaeogeography, Palaeoclimatology, Palaeoecology
structed megadroughts over North America since AD 1300. 319–320, 16–27.
Climate Change 8, 133–149. Stewart, J.R. 2009. The evolutionary consequence of the indi-
Staiger, J.W., Gosse, J., Little, E.C., et al. 2006. Glacial erosion and vidualistic response to climate change. Journal of Evolutionary
sediment dispersion from detrital cosmogenic nuclide analyses Biology 22, 2363–2375.
of till. Quaternary Geochronology 1, 29–42. Stewart, J.R., Lister, A.M., Barnes, I., et al. 2010. Refugia revisited:
Stambaugh, M.C., Guyette, R.P. 2009. Progress in constructing a long individualistic responses of species in space and time. Proceedings
oak chronology from the central United States. Tree-Ring of the Royal Society B 277, 661–671.
Research 65, 147–156. Stewart, T.G. 1991. Glacial marine sedimentation from tidewater
Stanford, J.D., Rohling, E.J., Bacon, S., et al. 2011. A new concept for glaciers in the Canadian High Arctic. In Glacial Marine Sedi-
the paleoceanographic evolution of Heinrich event 1 in the mentation; Paleoclimatic Significance (edited by J.B. Anderson,
North Atlantic. Quaternary Science Reviews 30, 1047–1066. G.M. Ashley), Geological Society of America, Special Paper 261,
Stanton, T., Snowball, I., Zillén, L., et al. 2010. Validating a Swedish 95–105.
varve chronology using radiocarbon, palaeomagnetic secular Stickley, C.E., Pike, J., Leventer, A., et al. 2005. Deglacial ocean and
variation, lead pollution history and statistical correlation. climate seasonality in laminated diatom sediments, MacRobert-
Quaternary Geochronology 5, 611–624. son Shelf, Antarctica. Paleogeography, Paleoclimatology, Palaeo-
Starkel, L. 2003. Climatically controlled terraces in uplifting ecology 227, 290–310.
mountain areas. Quaternary Science Reviews 22, 2189–2198. Stiner, M.C. 1999. Palaeolithic mollusc exploitation at Riparo Mochi
Steadman, D.W., Franz, R.S., Morgan, G.S., et al. 2007. Exceptionally (Balzi Rossi, Italy): food and ornaments from the Aurignacian
well preserved late Quaternary plant and vertebrate fossils from through Epigravettian. Antiquity 73, 735–754.
a blue hole on Abaco, The Bahamas. Proceedings of the National Stirling, C.H., Andersen, M.B. 2009. Uranium series dating of fossil
Academy of Sciences of the United States of America 104, coral reefs: extending the sea-level record beyond the last glacial
19897–19902. cycle. Earth and Planetary Science Letters 284, 269–283.
Stearta, D.C., Greenwood, D.R., Boon, P.I. 2009. The chemical Stocchi, P., Spada, G. 2007. Glacio and hydro-isostasy in the
constraints upon leaf decay rates: taphonomic implications Mediterranean Sea: Clark’s zones and role of remote ice sheets.
among leaf species in Australian terrestrial and aquatic environ- Annals of Geophysics 50, 741–761.
ments. Review of Palaeobotany and Palynology 157, 358–374. Stocker, T.F. 1998. The see-saw effect. Science 282, 61–62.
Steffen, W., Grinevald, J., Crutzen, P.J., et al. 2011. The Anthropo- Stocker, T.F., Johnsen, S.J. 2003. A minimum thermodynamic model
cene: conceptual and historical perspectives. Philosophical for the bipolar seesaw. Paleoceanography 18, doi: 10.1029/
Transactions of the Royal Society of London A 369, 842–867. 2003PA000920.
Steffensen, J.P., Andsersen, K.K., Bigler, M., et al. 2008. High- Stoddard, J.L., Jeffries, D.S., Lükewille, A., et al. 1999. Regional
resolution Greenland ice core data show abrupt climate change trends in aquatic recovery from acidification in North America
happens in few years. Science 321, 680–684. and Europe. Nature 401, 575–578.
Stein, J.K., Deo, J.N. 2003. Big sites – short time. Accumulation Stoermer, E.F. 2001. Diatom taxonomy for paleolimnologists. Journal
rates in archaeological sites. Journal of Archaeological Science of Paleolimnology 25, 393–398.
30, 297–316. Stokes, C.R., Tarasov, L. 2010. Ice streaming in the Laurentide Ice
Stein, M. 2001. The sedimentary and geochemical record of Sheet: a first comparison between data-calibrated numerical
Neogene–Quaternary water bodies in the Dead Sea Basin: infer- model output and geological evidence. Geophysical Research
ences for the regional paleoclimatic history. Journal of Paleo- Letters 37, doi: 10.1029/2009GL040990.
limnology 26, 271–282. Stokes, C.R., Clark, C.D., Darby, D.A., et al. 2005. Late Pleistocene
Stein, R., Fahl, K. 2013. Biomarker proxy shows potential for study- ice export events into the Arctic Ocean from the M’Clure Strait
ing the entire Quaternary Arctic sea ice history. Organic Ice Stream, Canadian Arctic Archipelago. Global and Planetary
Geochemistry 55, 98–102. Change 49, 139–162.
Steinhilber, F., Abreu, J.A., Beer, J., et al. 2012. 9,400 years of cosmic Stokes, C.R., Tarasov, L., Dyke, A.S. 2012. Dynamics of the North
radiation and solar activity from ice cores and tree rings. American Ice Sheet Complex during its inception and build-up
Proceedings of the National Academy of Sciences of the United States to the Last Glacial Maximum. Quaternary Science Reviews 50,
of America, doi: 10.1073/pnas.1118965109. 86–104.
Stenni, B., Buiron, D., Frezzotti, M., et al. 2011. Expression of the Stone, E.J., Lunt, D.J., Annan, J.D. 2013. Quantification of the
bipolar see-saw in Antarctic climate records during the last Greenland ice sheet contribution to Last Interglacial sea level rise.
deglaciation. Nature Geoscience 4, 46–49. Climate of the Past Discussions 9, 621–639.
512 REFERENCES
Stone, J.O. 2000. Air pressure and cosmogenic isotope production. Sugden, D.E., McCulloch, R.D., Bory, A.J.-M., et al. 2009. Influence
Journal of Geophysical Research 105, 23753–23759. of Patagonian glaciers on Antarctic dust deposition during the
Stoops, G. (ed.). 2003. Achievements in micromorphology. Catena Last glacial period. Nature Geoscience 2, 281–285.
54, 317–678. Sugita, S. 1994. Pollen representation of vegetation in Quaternary
Stott, A.W., Berstan, R., Evershed, P., et al. 2001. Radiocarbon dating sediments: theory and method in patchy vegetation. Journal of
of single compounds isolated from pottery cooking vessel Ecology 82, 881–897.
residues. Radiocarbon 43, 191–197. Sugita, S. 2007. Theory of quantitative reconstruction of vegeta-
Stott, K.J., Austin, W.E.N., Sayer, M.D.J., et al. 2010. The potential tion I: pollen from large sites REVEALS regional vegetation
of Arctica islandica growth records to reconstruct coastal climate composition. The Holocene 17, 229–241.
in north west Scotland, UK. Quaternary Science Reviews Sulpizio, R., Giaccio, B., Isaia, R., et al. (eds). 2008. Explosive
29,1602–1633. volcanism in the central Mediterranean area during the late
Stott, L., Timmermann, A., Thunell, R. 2007. Southern hemisphere Quaternary – linking sources and distal archives. Journal of
and deep-sea warming led deglacial atmospheric CO2 rise and Volcanology and Geothermal Research 177, 1–300.
tropical warming. Science 318, 435–438. Summerhayes, C.P., Thorpe, S.A. 2002. Oceanography: An Illustrated
Street, F.A., Grove, A.T. 1979. Global maps of lake-level fluctuations Guide, 3rd impr. Manson Publishing, London.
in Africa. Quaternary Research 12, 83–118. Sun, D., Gagan, M.K., Cheng, H., et al. 2005. Seasonal and interannual
Street-Perrott, F.A., Holmes, J.A., Waller, M.P., et al. 2000. Drought variability of the Mid-Holocene East Asian monsoon in coral
and dust deposition in the West African Sahel: a 5500-year δ18O records from the South China Sea. Earth and Planetary
record from Kajemarun Oasis, northeastern Nigeria. The Science Letters 237, 69–84.
Holocene 10, 293–302. Sun, J., Huang, X. 2006. Half-precessional cycles recorded in Chinese
Stuart, A.J. 1979. Pleistocene occurrences of the European pond loess: response to low-latitude insolation forcing during the
tortoise (Emys orbicularis L.) in Britain. Boreas 8, 359–371. Last Interglaciation. Quaternary Science Reviews 25,1065–1072.
Stuart, A.J. 2005. The extinction of woolly mammoth (Mammuthus Sun, Q., Chu, G., Liu, J., et al. 2006a. A 150-year record of heavy metals
primigenius) and straight-tusked elephant (Palaeoloxodon in the varved sediments of Lake Bolterskardet, Svalbard. Arctic
antiquus) in Europe. Quaternary International 126–12, 171–177. and Alpine Research 38, 436–445.
Stuart, A.J., Lister, A.M. 2001. The mammalian faunas of Pakefield/ Sun, Y., Clemens, S.C., An, Z., et al. 2006b. Astronomical timescale
Kessingland and Corton, Suffolk, UK: evidence for a new and palaeoclimatic implication of stacked 3.6-Myr monsoon
temperate episode in the British Middle Pleistocene. Quaternary records from the Chinese Loess Plateau. Quaternary Science
Science Reviews 20, 1677–1692. Reviews 25, 33–48.
Stuart, A.J., Lister, A.M. 2011. Extinction chronology of the cave lion Sun, Y.J., Lai, Z.P., Madsen D., et al. 2012. Luminescence dating of
Panthera spelaean. Quaternary Science Reviews 30, 2329–2340. a hearth from the archaeological site of Jiangxigou in the Qinghai
Stuart, A.J., Lister, A.M. 2012. Extinction chronology of the woolly Lake area of northeastern Qinghai-Tibetan Plateau. Quaternary
rhinoceros Coelodonta antiquitatis in the context of late Geochronology 12, 107–110.
Quaternary megafaunal extinctions in northern Eurasia. Surić, M., Richards, D., Hoffman, D.L., et al. 2009. Sea-level change
Quaternary Science Reviews 30, 2329–2340. during MIS 5e based on submerged speleothems from eastern
Stuiver, M., Brazunias, T.F., Becker, B., et al. 1991. Climatic solar, Adriatic Sea (Croatia). Marine Geology 262, 62–67.
oceanic and geomagnetic influences on Late-Glacial and Sutherland, D.G. 1984. Modern glacier characteristics as a basis
Holocene atmospheric 14C/12C change. Quaternary Research 35, for inferring former climates, with particular reference to the Loch
1–24. Lomond Stadial. Quaternary Science Reviews 3, 291–309.
Stuiver, M., Reimer, P.J., Bard, E., et al. 1998. INTCAL98 radiocarbon Svendsen, J.I., Alexanderson, H., Astakhov, V.I., et al. 2004. Late
age calibration, 24,000–0 cal BP. Radiocarbon 40, 1041–1084. Quaternary ice sheet history of northern Eurasia. Quaternary
Stuiver, M., Reimer, P.J., Reimer, R. 2013. CALIB Radiocarbon Science Reviews 23, 1229–1271.
Calibration Version 7.0, http://calib.qub.ac.uk/calib/. Svensson, A., Nielsen, S.W., Kipfstuhl, S., et al. 2005. Visual
Sturm, C.F., Pearce, T.A., Valdés, A. 2006. The Mollusks: A Guide to stratigraphy of the NorthGRIP ice core during the last glacial
their Study, Collection and Preservation. American Malacological period. Journal of Geophysical Research 110, D02108, doi: 10.1029/
Society, Pittsburgh, PA. 2004JD005134.
Sturm, M., Lotter A. 1995. Lake sediments as environmental archives. Svensson, A., Andersen, K.K., Bigler, M., et al. 2008. A 60 000 year
EAWAG News 38E, 6–9. Greenland stratigraphic ice core chronology. Climate of the Past
Suckow, A., Morgenstern, U., Kudrass, H.-R. 2001. Absolute dating 4, 47–57.
of recent sediments in the cyclone-influenced shelf area off Svensson, A., Bigler, M., Blunier, T., et al. 2012. Direct linking of
Bangladesh: comparison of gamma spectrometric (137Cs, 210Pb, Greenland and Antarctic ice cores at the Toba eruption (74 kyr
228
Ra), radiocarbon and 32Si ages. Radiocarbon 43, 917–927. BP). Climate of the Past 8, 5389–5427.
Sucre, E.B., Tuttle, J.W., Fox, T.R. 2011. The use of ground- Svensson, H. 2005. From relict polygon patterns in South Scandinavia
penetrating radar to accurately estimate soil depth in rocky to active ice wedge polygons in Svalbard: A review of former
forest soils. Forest Science 57, 59–66. studies. Norsk Geografisk Tidsskrift 59, 164–170.
REFERENCES 513
Sverdrup, K.A., Kudela, R.M. 2011. Investigating Oceanography. Taylor, K.C., Alley, R.B.V., Lamorey, G.W., et al. 1997. Electrical
McGraw-Hill Companies, Higher Education e-book. measurements on the Greenland Ice Sheet Project 2 core. Journal
Sweetman, J.N., LaFace, E., Rühland, K.M., et al. 2008. Evaluating of Geophysical Research 102, 26511–26517.
the response of Cladocera to recent environmental changes in Taylor, K.C., Alley, R.B., Meese, D.A., et al. 2004. Dating the Suple
lakes from the central Canadian Arctic treeline region. Arctic, Dome (Antarctica) ice core by manual and computer inter-
Antarctic, and Alpine Research 40, 584–591. pretation of annual layering. Journal of Glaciology 50, 453–461.
Swindles, G.T., Plunkett, G., Roe, H. 2007. A delayed climatic Tchakerian, V.P. 2009. Palaeoclimatic interpretations from desert
response to solar forcing at 2800 cal. BP: multiproxy evidence from dunes and sediments. In Geomorphology of Desert Environments,
three Irish peatlands. The Holocene 17, 177–182. 2nd edn (edited by A.J. Parsons, A. Abrahams). Springer,
Swingedouw, D., Fichefet, T., Goosse, H., et al. 2009. Impact of Dordrecht, 757–772.
transient freshwater releases in the Southern Ocean on the Teixell, A., Bertotti, G., de Lamotte, D.F., et al. 2009. The geology
AMOC and climate. Climate Dynamics 33, 365–381. of vertical movements of the lithosphere: an overview. Tectono-
Sykes, G. 1991. Amino-acid dating. In Quaternary Dating Methods: physics 475, 1–8.
A User’s Guide (edited by P.L. Smart, P.D. Francis). Technical Telford, R.J. 2006. Limitations of dinoflagellate cyst transfer
Guide 4. Quaternary Research Association, Cambridge, 161–176. functions. Quaternary Science Reviews 25, 1375–1382.
Sykes, G.A., Collins, M.J., Walton, D.I. 1995. The significance of a Telford, R.J., Heegaard, E., Birks, H.J.B. 2004. All age-depth models
geochemically isolated intra-crystalline organic fraction within are wrong: but how badly? Quaternary Science Reviews 23, 1–5.
biominerals. Organic Geochemistry 23, 1059–1065. Thomas, G.N. 2001. Late Middle Pleistocene pollen bisotratigraphy
Syvitski, J.P.M. 2002. Sediment discharge variability in Arctic rivers: in Britain: pitfalls and possibilities in the separation of interglacial
implications for a warmer future. Polar Research 21, 323–330. sequences. Quaternary Science Reviews 20, 1621–1630.
Syvitski, J.P.M., Vörösmarty, V.J., Kettner, A.J., et al. 2005. Impact Thomas, G.S.P., Chiverrell, R.C. 2007. Structural and depositional
of humans on the flux of terrestrial sediment to the global coastal
evidence for repeated ice-marginal oscillation along the eastern
ocean. Science 308, 376–380.
margin of the Late Devensian Irish Sea Ice Stream. Quaternary
Szabo, B.J. 1990. Ages of travertine deposits in eastern Grand Canyon
National Park, Arizona. Quaternary Research 34, 24–32.
Thomas, G.S.P., Chiverrell, R.C., Huddart, D. 2004. Ice-marginal
Szostek, K. 2009. Chemical signals and reconstruction of life strat-
depositional responses to readvance episodes in the Late
egies from ancient human bones and teeth – problems and
Devensian deglaciation of the Isle of Man. Quaternary Science
perspectives. Anthropological Review 72, 3–30.
Reviews 23, 85–106.
Takano, Y., Kobayashi, K., Ishikawa, Y., et al. 2005. Emergence of the
Thomas, P.A., Easterbrook, D.J., Clark, P.U. 2000. Early Holocene
inflection point on racemization rate constants for d- and l-amino
glaciation on Mount Baker, Washington State, USA. Quaternary
acids in the early stages of terrestrial diagenesis. Organic
Geochemistry 37, 334–341. Science Reviews 19, 1043–1046.
Támas, T., Onac, B., Bojar, A.-V. 2005. Lateglacial–Middle Holocene Thomas, P.J. 2009. Luminescence dating of beachrock in the
stable isotope records in two coeval stalagmites from the Bihor southeast coast of India – potential for Holocene shoreline
Mountains. NW Romania. Geological Quarterly 49, 185–194. reconstruction. Journal of Coastal Research 25, 1–7.
Tanabe, S., Tateishi, M., Shibata, Y. 2009. The sea-level record of the Thompson, D.W.J., Lee, S., Baldwin, M.P. 2003. Atmospheric
last deglacial in the Shinano River incised-valley fill, Echigo processes governing the Northern Hemisphere annular mode/
Plain, central Japan. Marine Geology 266, 223–231. North Atlantic Oscillation. In The North Atlantic Oscillation:
Tang, K., Feng, X., Ettl., G.J. 2000. The variations in δD of tree- Climatic Significance and Environmental Impact (edited by J.
rings and the implications for climatic reconstructions. Hurrell, Y. Kushnir, G. Ottersen, et al.). Geophysical Monograph
Geochimica et Cosmochimica Acta 64, 1663–1673. 134. American Geophysical Union, Washington DC, 81–112
Tankersley, K.B., Schlecht, K.D., Laub, R.S. 1998. Fluoride dating of Thompson, L.G. 2000. Ice core evidence for climate change in the
mastodon bone from an early palaeoindian spring site. Journal Tropics: implications for our future. Quaternary Science Reviews
of Archaeological Science 25, 805–811. 19, 19–35.
Tapley, B.D., Bettadpur, S., Ries, J.C., et al. 2004. GRACE measure- Thompson, L.G., Yao, T. Davis, M.E., et al. 1997. Tropical climate
ments of mass variability in the earth system. Science 305, instability: the last glacial cycle from a Qinghai-Tibetan ice core.
503–505. Science 276, 1821–1825.
Tarasov, L., Peltier, W.R. 2004. A geophysically constrained large Thompson, L.G., Mosley-Thompson, E., Davis, M.E., et al. 2013.
ensemble analysis of the deglacial history of the North American Annually-resolved ice core records of tropical climate variability
ice-sheet complex. Quaternary Science Reviews 23, 359–388. over the past ~1800 years. Science 340, 945–950.
Tattersfield, P., Warui, C.M., Seddon, M.B., et al. 2001. Land-snail Thompson, R., Oldfield, F. 1986. Environmental Magnetism. Allen
faunas of afro-montane forests of Mount Kenya, Kenya: ecology, and Unwin, London.
diversity and distribution patterns. Journal of Biogeography 28, Thompson, R., Price, D., Cameron, N., et al. 2005. Quantitative
843–861. calibration of remote mountain-lake sediments as climatic
Tauber, H. 1965. Differential pollen dispersal and the interpretation recorders of air temperature and ice-cover duration. Arctic,
of pollen diagrams. Danmarks Geologiske Undersøgelse II, 89, 1–69. Antarctic and Alpine Research 37, 626–635.
514 REFERENCES
Thompson, W.G., Goldstein, S.L. 2005. Open-system coral ages Tonni, E.P., Cione, A.L., Figini, A.J. 1999. Predominance of arid
reveal persistent suborbital sea-level cycles. Science 308, 401–404. climates indicated by mammals in the pampas of Argentina
Thompson, W.G., Goldstein, S.L. 2006. A radiometric calibration during the Late Pleistocene and Holocene. Palaeogeography,
of the SPECMAP timescale. Quaternary Science Reviews 25, Palaeoclimatology, Palaeoecology 147, 257–281.
3207–3215. Tooth, S. 2000. Process, form and change in dryland rivers: a review
Thompson, W.G., Cirran, H.A., Wilson, M.A., et al. 2011. Sea-level of recent research. Earth-Science Reviews 51, 67–107.
oscillations during the last interglacial highstand recorded by Torfstein, A., Haase-Schramm, A., Waldmann, N., et al. 2009. U-series
Bahamas corals. Nature Geoscience 4, 684–687. and oxygen isotope chronology of the mid-Pleistocene Lake
Thompson Davis, P., Menounos, B., Osborn, G. 2009. Holocene and Amora (Dead Sea basin). Geochimica et Cosmochimica Acta 73,
latest Pleistocene alpine glacier fluctuations: a global perspective. 2603–2630.
Quaternary Science Reviews 28, 2021–2033. Torfstein A., Goldsteain, S.L., Stein, M.I. 2013. Impacts of abrupt
Thorn, C.E., Hall, K. 2002. Nivation and cryoplanation: the case climate changes in the Levant from Last Glacial Dead Sea levels.
for scrutiny and integration. Progress in Physical Geography 26, Quaternary Science Reviews 69, 1–7.
533–550 Törnqvist, T.E. 2007. Fluvial environments: responses to rapid
Thornalley, D.J.R., Elderfield, H., McCave, I.N. 2010. Intermediate environmental change. In Encyclopedia of Quaternary Science
and deep water paleoceanography of the northern North (edited by S.A. Elias). Elsevier, Amsterdam, 686–694.
Atlantic over the past 21,000 years. Paleoceanography 25, doi: Törnqvist, T.E., González, J.L., Newsom, L.A., et al. 2004.
10.1029/2009PA001833. Deciphering Holocene sea-level history on the US Gulf Coast:
Thornalley, D.J.R., McCave, I.N., Elderfield, H. 2011. Tephra in a high-resolution record from the Mississippi Delta. Geological
deglacial ocean sediments south of Iceland: stratigraphy, geo- Society of America Bulletin 116, 1026–1039.
chemistry and oceanic reservoir ages. Journal of Quaternary Torres, V., Hooghiemstra, H., Lourens, L., et al. 2013. Astronomical
Science 26, 190–198. tuning of long pollen records reveals the dynamic history of
Thorne, P.D., Hanes, D.M. 2002. A review of acoustic measurement montane biomes and lake levels in the tropical high Andes
of small-scale sediment processes. Continental Shelf Research during the Quaternary. Quaternary Scence Reviews 63, 59–72.
22, 603–632. Toyoda, S., Tsukamoto, S., Hameau, S., et al. 2006. Dating of Japanese
Thorp, P.W. 1981. A trimline method for defining the upper limit tephras by ESR and luminescence methods. Quaternary
of Loch Lomond Advance glaciers: examples from the Loch Geochronology 1, 320–326.
Leven and Glen Coe areas. Scottish Journal of Geology 17, 49–64. Trauth, M.H., Deino, A., Strecker, M.R. 2001. Response of the East
Thorson, R.M. 2000. Glacial tectonics: a deeper perspective. African climate to orbital forcing during the last interglacial
Quaternary Science Reviews 14–15, 1391–1398. (130–117 ka) and the early last glacial (117–60 ka). Geology 29,
Thunell, R.C., Varela, R., Llano, M., et al. 2000. Organic carbon fluxes, 499–503.
degradation, and accumulation in an anoxic basin: sediment trap Trauth, M.H., Maslin, M.A., Deino, A., et al. 2005. Late Cenozoic
results from the Cariaco Basin. Limnology and Oceanography 45, moisture history of East Africa. Science 23, 2051–2053.
300–308. Trimble Navigation. GPS Tutorial, http://www.trimble.com/
Tietze, E., De Francesco, C.G. 2012. Compositional fidelity of gps_tutorial/.
subfossil mollusk assemblages in streams and lakes of the south- Tripaldi, A., Forman, S.L. 2007. Geomorphology and chronology
eastern Pampas, Argentina. Palaios 27, 401–413. of Late Quaternary dune fields of western Argentina. Palaeo-
Timm, O., Timmermann, A. 2007. Simulation of the last 21 000 years geography, Palaeoclimatology, Palaeoecology 251, 300–320.
using accelerated transient boundary conditions. Journal of Tripcevich, N., Eerkens, J.W., Carpenter, T.R. 2012. Obsidian
Climate 20, 4377–4401. hydration at high elevation: archaic quarrying at the Chivay
Timmermann, A., An, S.-I., Krebs, U., et al. 2005. ENSO Suppression source, southern Peru. Journal of Archaeological Science 39,
due to weakening of the North Atlantic Thermohaline Circu- 1360–1367.
lation. Journal of Climate 18, 3122–3139. Tröften, P.-E. 2000. The use of varved clay chronology for dating
Timmermann, A., Timm, O., Stott, L., et al. 2009. The roles of CO2 palaeoseismic events: the Erstavik record in the Stockholm area,
and orbital forcing in driving Southern Hemispheric temperature south Sweden. Sedimentary Geology 130, 167–181.
variations during the last 21 000 yr. Journal of Climate 22, Tsai, H., Hseu, Z.-Y., Huang, W.-S., et al. 2007. Pedogenic approach
1626–1640. to resolving the geomorphic evolution of the Pakua river terraces
Tinner, W., Hubschmid, P., Wehrli, M., et al. 1999. Long-term in central Taiwan. Geomorphology 83, 14–28.
forest fire ecology and dynamics in southern Switzerland. Journal Tsatskin, A., Ronen, A. 1999. Micromorphology of a Mousterian
of Ecology 87, 273–289. paleosol in aeolianites at the site Habonim, Israel. Catena 34,
Tobolski, K. Ammann, B. 2000. Macrofossils as records of plant 365–384.
responses to rapid Late Glacial climatic changes at three sites in Tschumi, T., Joos, F., Gehlen M., et al. 2010. Deep ocean ventilation,
the Swiss Alps. Palaeogeography, Palaeoclimatology, Palaeoecology carbon isotopes, marine sedimentation and the deglacial CO2 rise.
159, 251–259. Climate of the Past 7, 771–800.
Tomescu, A.M.F. 2000. Evaluation of Holocene pollen records from Tudhope, A.W., Chilcott, C.P., McCulloch, M.T., et al. 2001.
the Romanian Plain. Review of Palaeobotany and Palynology 109, Variability in the El Niño–Southern Oscillation through a
219–233. glacial–interglacial cycle. Science 291, 1511–1517.
REFERENCES 515
Tuovinen, N., Virtasalo, J.J., Kotilainen, A.T. 2008. Holocene diatom Valero-Garcés, B.L., Arenas, C., Delgada-Huertas, A. 2001.
stratigraphy in the Archipelago Sea, northern Baltic Sea. Journal Depositional environments of Quaternary lacustrine travertine
of Paleolimnology 40, 793–807. and stromatolites from high-altitude Andean lakes, northwestern
Turner, M.B., Cronin, S.J., Bebbington, M.S., et al. 2008. Developing Argentina. Canadian Journal of Earth Sciences 38, 1263–1283.
probabilistic eruption forecasts for dormant volcanoes: a case Van Calsteren, P., Thomas, L. 2006. Uranium-series dating applica-
study from Mt Taranaki, New Zealand. Bulletin of Volcanology tions in natural environmental science. Earth-Science Reviews 75,
70, 507–515. 155–175.
Turner-Walker, G. 2008. The chemical and microbial degradation Van den Bergh, G.D., Meijer, H.J.M., Awe, R.D., et al. 2009. The Liang
of bones and teeth. In Advances in Human Palaeopathology Bua faunal remains: a 95 k.yr. sequence from Flores, East
(edited by R. Pinhasi, S. Mays). John Wiley, Chichester and New Indonesia. Journal of Human Evolution 57, 527–537.
York, 3–30. Vandenberghe, J. 1988. Cryoturbations. In Advances in Periglacial
Turney, C.S.M., Lowe, J.J. 2001. Tephrochronology. In Tracking Geomorphology (edited by M.J. Clark). John Wiley, Chichester
Environmental Change Using Lake Sediments. Volume 1: Basin and New York, 179–198.
Analysis, Coring and Chronological Techniques (edited by J.P. Last, Vandenberghe, J. 2003. Climate forcing of fluvial system develop-
W.M. Smol). Kluwer, Dordrecht, 451–471. ment: an evolution of ideas. Quaternary Science Reviews 22,
Turney, C.S.M., Brown, H. 2007. Catastrophic early Holocene sea 2053–2060.
level rise, human migration and the Neolithic transition in Vandenberghe, J., Woo, M. 2002. Modern and ancient periglacial river
Europe. Quaternary Science Reviews 26, 2036–2041. types. Progress in Physical Geography 26, 479–506.
Turney, C.S.M., Coope, G.R., Harkness, D.D., et al. 2000. Vandenberghe, J., Zhijiu, C., Liang, Z., et al. 2004. Thermal-
Implications for the dating of Wisconsinan (Weichselian) contraction-crack networks as evidence for late-Pleistocene
Lateglacial events of radiocarbon age differences between permafrost in Inner Mongolia, China. Permafrost and Periglacial
Processes 15, 21–29.
terrestrial macrofossils from a site in SW Ireland. Quaternary
Van den Bogaard, P. 1995. 40Ar/39Ar ages of sanidine phenocrysts
Research 53, 114–121.
from Laacher See Tephra (12,900 yrs BP): chronostratigraphic
Tweddle, J.C., Edwards, K.J. 2010. Pollen preservation zones as an
and petrologic significance. Earth and Planetary Science Letters
interpretative tool in Holocene palynology. Review of Palaeo-
133, 163–174.
botany and Palynology 161, 59–76.
Van den Haute, P., Corte, F. (eds). 1998. Advances in Fission Track
Twiddle, C.L., Bunting, M.J. 2010. Experimental investigations
Geochronology. Kluwer, Dordrecht.
into the preservation of pollen grains: a pilot study of four
Van der Geer, A., Lyras, G., de Vos, J., et al. 2010. Evolution of Island
pollen types. Review of Palaeobotany and Palynology 162,
Mammals: Adaptation and Extinction of Placental Mammals on
621–630.
Islands. Wiley-Blackwell, Hoboken, NJ.
Tyrácek, J., Havlícek, P. 2009. The fluvial record in the Czech Vandergoes, M.C., Dieffenbacher-Krall, A.C., Newnham, R.M., et al.
Republic: a review in the context of IGCP 518. Global and 2008. Cooling and changing seasonality in the Southern Alps, New
Planetary Change 68, 311–325. Zealand during the Antarctic Cold Reversal. Quaternary Science
Tzedakis, P.C., McManbus, J.F., Hooghiemstra, H., et al. 2003. Reviews 27, 589–601.
Comparison of changes in vegetation in northeast Greece Vandergroes, M.J., Prior, C.A. 2003. The AMS dating of pollen
with records of climatic variability on orbital and suborbital concentrates: a methodological study of Late Quaternary
frequencies over the last 450,000 years. Earth and Planetary sediments from South Westland, New Zealand. Radiocarbon
Science Letters 212, 197–212. 45, 479–491.
Tzedakis, P.C., Roucoux, K.H., de Abreu, L., et al. 2004. The duration Van der Knaap, W.O., Lamentowicz, M., van Leeuwen, J.F.N., et al.
of forest stages in Southern Europe and interglacial climate 2011. A multi-proxy, high-resolution record of peatland
variability. Science 306, 2231–2235. development and its drivers during the last millennium from the
Tzedakis, P.C., Channell, J.E.T., Hodell, D.A., et al. 2012. Determining subalpine Swiss Alps. Quaternary Science Reviews 30, 3467–3480.
the natural length of the current interglacial. Nature Geoscience Van der Knaap, W.O., van Leeuwen, F.F.N., Finsinger, W., et al. 2005.
5, 138–141. Migration and population expansion of Abies, Fagus, Picea and
Tziperman, E., Gildor, H. 2003. On the mid-Pleistocene transition Quercus since 15000 years in and across the Alps, based on
to 100-kyr glacial cycles and the asymmetry between glaciation pollen-percentage threshold values. Quaternary Science Reviews
and deglaciation times. Paleoceanography 18, doi: 10.1029/ 24, 645–680.
2001PA000627. Van der Meer, J.J.M., Menzies, J., Rose, J. 2003. Subglacial till: the
Ubelaker, D.H., Zarenko, K.M. 2011. Adipocere: what is known after deforming glacier bed. Quaternary Science Reviews 22, 1659–1685.
over two centuries of research? Forensic Science International Van der Plicht, J., van der Sanden, W.A.B., Aerts, A.T., et al. 2004.
208, 167–172. Dating of bog bodies by means of 14C-AMS. Journal of Archaeo-
Valentine, K.W.G., Dalrymple, J.B. 1975. The identification, lateral logical Science 31, 471–491.
variation and chronology of two buried palaeocatenas at Van der Veen, M. 2007. Formation processes of desiccated and
Woodhall Spa and West Runton, England. Quaternary Research carbonized plant remains – the identification of routine practice.
4, 551–590. Journal of Archaeological Science 34, 968–990.
516 REFERENCES
Van der Zwaan, G.J., Duijnstee, I.A.P., den Dilk, M., et al. 1999. for preservation of Irish last interglacial deposits. Journal of
Benthic foraminifers: proxies or problems?: A review of Quaternary Science 19, 577–590.
paleocological concepts. Earth-Science Reviews 46, 213–236. Velichkevich, F.Y., Zastawniak, E. 2008. Atlas of the Pleistocene
Van Geel, B., Aptroot, A. 2006. Fossil ascomycetes in Quaternary Vascular Plant Macrofossils of Central and Eastern Europe – Part
deposits. Nova Hedwigia 82, 313–329. 2: Herbaceous Dicotyledones. Polish Academy of Sciences, W.
Van Geel, B., Coope, G.R., van der Hammen, T. 1989. Palaeoecology Szafer Institute of Botany, Krakow.
and stratigraphy of the Lateglacial type section at Usselo (the Veres, D., Bazin, L., Landais, A., et al. 2012. The Antarctic ice core
Netherlands). Review of Palaeobotany and Palynology 60, 25–129. chronology (AICC2012): an optimised multi-parameter and
Van Geel, B., Buurman, J., Waterbolk, H.T. 1996. Archaeological multi-site dating approach for the last 120 thousand years.
and palaeoecological indications for an abrupt climate change Climate of the Past Discussions 8, 6011–1049.
in the Netherlands and evidence for climatological telecon- Verleye, T.J., Louwye, S. 2010. Late Quaternary environmental
nections around 2650 BP. Journal of Quaternary Science 11, changes and latitudinal shifts of the Antarctic Circumpolar
451–460. Current as recorded by dinoflagellate cysts from offshore Chile
Van Geel, B., van der Plicht, J., Renssen, H. 2003. Major Δ14C (40°S). Quaternary Science Reviews 29, 1025–1039.
excursions during the late glacial and early Holocene: changes Verschuren, D., Eggermont, H. 2006. Quaternary paleoecology of
in ocean ventilation or solar forcing of climate change. aquatic Diptera in tropical and Southern Hemisphere regions,
Quaternary International 105, 71–76. with special reference to the Chironomidae. Quaternary Science
Van Geel, B., Bokovenko, N.A., Burova, N.D., et al. 2004. Climate Reviews 25, 1926–1947.
change and the expansion of the Scythian culture after 850 BC, Veski, S., Heinsalu, A., Klassen, V., et al. 2005a. Early Holocene coastal
a hypothesis. Journal of Archaeological Science 31, 1735–1742. sediments and palaeoenvironment on the shore of the Baltic Sea
Van Geel, B., Zazula, G.D., Schweger, C.E. 2007. Spores of coat Pärnu, southwestern Estonia. Quaternary International 130,
prophilous fungi from under the Dawson tephra (25,300 14C years 75–85.
BP), Yukon Territory, northwestern Canada. Palaeogeography,
Veski, S., Koppel, K., Poska, A. 2005b. Integrated palaeoecological
and historical data in the service of fine-resolution land use
Van Geel, B., Aptroot, A., Baittinger, C., et al. 2008. The ecological
and ecological change assessment during the last 1000 years in
implications of a Yakutian mammoth’s last meal. Quaternary
Rõuge, southern Estonia. Journal of Biogeography 32, 1473–1488.
Research 69, 361–376.
Veski, S., Amon, L., Heinsalu, A., et al. 2012. Lateglacial vegetation
Van Huissteden, K., Vandenberghe, J., Pollard, D. 2003.
dynamics in the eastern Baltic region between 14,500 and 11,400
Palaeotemperature reconstructions of the European permafrost
cal yr BP: a complete record since the Bølling (GI-1e) to the
zone during marine oxygen isotope stage 3 compared with
Holocene. Quaternary Science Reviews 40, 39–53.
climate model results. Journal of Quaternary Science 18, 453–464.
Viau, A.E., Gajewski, K. 2001. Holocene variations in the global
Van Koenigswald, W., van Kolfschoten, T. 1996. The Mimomys–
Arvicola boundary and the enamel thickness quotient (SDQ) hydrological cycle quantified by objective gridding of lake level
of Arvicola as stratigraphic markers in the Middle Pleistocene. databases. Journal of Geophysical Research 106, 31703–31716.
In The Early–Middle Pleistocene in Europe (edited by C. Turner). Viehberg, F.A., Mesquita-Joanes, F. 2012. Quantitative transfer
Balkema, Rotterdam, 211–226. function approaches in palaeoclimatic reconstruction using
Van Leeuwen, R.J.W., Beets, D.J., et al. 2000. Stratigraphy and Quaternary ostracods. In Developments in Quaternary Science.
integrated facies analysis of the Saalian and Eemian sediments Volume 17: Ostracoda as Proxies for Quaternary Climate Change
in the Amsterdam-Terminal borehole. Geologie en Mijnbouw (edited by D.J. Horne, J.A. Holmes, J. Rodriguez-Lazaro, et al.).
(Netherlands Journal of Geosciences) 79, 161–196. Elsevier, Amsterdam, 47–64.
Van Raden, U.J., Colombaroli, D., Gilli, A., et al. 2013. High- Viklund, K. 2011. Flax in Sweden: the archaeobotanical,
resolution late-glacial chronology for the Gerzensee lake record archaeological and historical evidence. Vegetation History and
(Switzerland): δ18O correlation between a Gerzensee-stack and Archaeobotany 20, 509–515.
NGRIP. Palaeogeography, Palaeoclimatology, Palaeoecology 391, Viles, H., Goudie, A., Grab, S., et al. 2011. The use of the Schmidt
13–24. hammer and Equotip for rock hardness assessment in
Vardy, M.E., Pinson, J.J.W., Bull, J.M., et al. 2010. 3D seismic geomorphology and heritage science: a comparative analysis.
imaging of buried Younger Dryas mass movement flows: Lake Earth and Planetary Science Letters 36, 320–333.
Windermere, UK. Geomorphology 118, 176–187. Vimeux, F., Ginot, P., Schwikowski, M., et al. 2009. Climate variability
Varela, S., Lobo, J.M., Rodriguez, J., et al. 2010. Were the Late during the last 1000 years inferred from Andean ice core: a
Pleistocene climatic changes responsible for the disappearance review of methodology and recent results. Palaeogeography,
of the European spotted hyena populations? Hindcasting a Palaeoclimatology, Palaeoecology 281, 229–241.
species geographic distribution across time. Quaternary Science Vink, A., Steffen, H., Reinhardt, L., et al. 2007. Holocene relative
Reviews 29, 2027–2035. sea-level change, isostatic subsidence and the radial viscosity
Vaughan, A.P.M., Dowling, L.A., Mitchell, F.J., et al. 2004. structure of the mantle of northwest Europe (Belgium, the
Depositional and post-depositional history of warm stage Netherlands, Germany, southern North Sea. Quaternary Science
deposits at Knocknacran, Co. Monaghan, Ireland: implications Reviews 26, 3249–3275.
REFERENCES 517
Vinther, B.M., Clausen, H.B., Johnsen, S.J., et al. 2006. A synchronised Walker, M. 2005. Quaternary Dating Methods. John Wiley, Chichester
dating of three Greenland ice cores throughout the Holocene. and New York.
Journal of Geophysical Research 111, D13201, doi: 13110.11029/ Walker, M.J.C., Lowe, J.J. 1990. Reconstructing the environmental
12005JD006921. history of the last glacial–interglacial transition: evidence from
Voelker, A.H.L. 2002. Global distribution of centennial-scale records the Isle of Skye, Inner Hebrides, Scotland. Quaternary Science
for marine isotope stage (MIS)3: a database. Quaternary Science Reviews 9, 15–49.
Reviews 21, 1185–1212. Walker, M.J.C., Merritt, J.W., Auton, C.A., et al. 1992. Allt Odhar
Voelker, A.H.L., Grootes, P.M., Nadeau, M.-J., et al. 2000. Radio- and Dalcharn: two pre-Late Devensian sites in northern Scotland.
carbon levels in the Iceland Sea from 25–53 kyr and their link to Journal of Quaternary Science 7, 69–86.
the earth’s magnetic field. Radiocarbon 42, 437–452. Walker, M.J.C., Coope, G.R., Lowe, J.J. 1993. The Devensian
Voinchet, P., Despriée, J. Tissoux, H., et al. 2010. ESR chronology (Weichselian) Lateglacial palaeoenvironmental record from
of alluvial deposits and first human settlement of the Middle Gransmoor, East Yorkshire, England. Quaternary Science Reviews
Loire Basin (Region Centre, France). Quaternary Geochronology 12, 659–680.
5, 381–384. Walker, M.J.C., Bjorck, S., Lowe, J.J., et al. 1999. Isotopic ‘events’ in
Völkel, J., Grunert, J. 1990. To the problem of dune formation and the GRIP ice core: a stratotype for the Late Pleistocene.
dune weathering during the Late Pleistocene and Holocene Quaternary Science Reviews 18, 1143–1151.
in the southern Sahara and the Sahel. Zeitschrift für Geomorpho- Walker, M.J.C., Bryant, C., Coope, G.R., et al. 2001. Towards a
logie 34, 1–17. radiocarbon chronology for the Lateglacial: sample selection
Vonmoos, M., Beer, J., Muscheler, R. 2006. Large variations in strategies. Radiocarbon 43, 1007–1020.
Holocene solar activity: constraints from 10Be in the Green- Walker, M.J.C., Coope, G.R., Sheldrick, C., et al. 2003. Devensian
land Ice Core project ice core. Journal of Geophysical Research 111, Lateglacial environmental changes in Britain: a multi-proxy
A10105, doi: 10.1029/2005JA011500. environmental record from Llanilid, South Wales, UK.
Vork, K.A., Thomsen, E. 1996. Lusitanian/Mediterranean ostracods Quaternary Science Reviews 22, 475–520.
in the Holocene of Denmark: implications for the interpretation Walker, M., Johnson S., Rasmussen S.O., et al. 2009. Formal
of winter temperatures during the postglacial temperature definition and dating of the GSSP (Global Stratotype Section and
maximum. The Holocene 6, 423–432.
Point) for the base of the Holocene using the Greenland NGRIP
Vorren T.O., Plassen, L. 2008. Deglaciation and palaeoclimate of the
ice core, and selected auxiliary records. Journal of Quaternary
Andfjord Vågsfjord area, North Norway. Boreas 31, 97–125.
Science 24, 3–17.
Vött, A. 2007. Relative sea level changes and regional tectonic
Walker, M.J.C., Berkelhammer, M., Bjorck, S., et al. 2012. Formal
evolution of seven coastal areas in NW Greece since the mid-
subdivision of the Holocene Series/Epoch: a Discussion Paper
Holocene. Quaternary Science Reviews 26, 894–919.
by a Working Group of INTIMATE (Integration of ice-core,
Waelbroeck, C., Labeyrie L., Michel E., et al. 2002. Sea-level and
marine and terrestrial records) and the Subcommission on
deepwater temperature changes derived from benthic foramini-
Quaternary Stratigraphy (International Commission on Strati-
fera isotopic records. Quaternary Science Reviews 21, 295–305.
graphy). Journal of Quaternary Science 27, 649–659.
Waelbroeck, C., Mulitza, S., Spero, H., et al. 2005. A global com-
pilation of late Holocene planktonic foraminiferal δ18O: Walker, S.E., Goldstein, S.T. 1999. Taphonomic tiering: experimental
relationship between surface water temperature and δ18O. field taphonomy of molluscs and foraminifera above and
Quaternary Science Reviews 24, 853–868. below the sediment–water interface. Palaeogeography, Palaeo-
Wagner, G.A. 1998. Age Determination of Young Rocks and Artefacts. climatology, Palaeoecology 149, 227–244.
Springer, Heidelberg, Berlin and New York. Walkington, H. 2010. Soil science applications in archaeological
Wagner, G.A., van den Haute, P. 1992. Fission Track Dating. Kluwer, contexts: a review of key challenges. Earth-Science Reviews 103,
Dordrecht. 122–134.
Wagner, G., Masarik, J., Beer, J., et al. 2000. Reconstruction of the Wallin, J.-E. 1996. History of sedentary farming in Ångermanland,
geomagnetic field between 20 and 60 kyr BP from cosmogenic northern Sweden, during the Iron Age and Medieval period
radionuclides in the GRIP ice core. Nuclear Instruments and based on pollen analytical investigations. Vegetation History and
Methods in Physics Research Section B: Beam Interactions with Archaeobotany 5, 301–312.
Materials and Atoms 172, 597–604. Wallis, L.A. 2001. Environmental history of northwest Australia
Walden, J., Oldfield, F., Smith, J. (eds). 1999. Environmental based on phytolith analysis at Carpenter’s Gap 1. Quaternary
Magnetism: a Practical Guide. Technical Guide 6. Quaternary International 83–85, 103–117.
Research Association, London. Walsh, S. 2005. The role of stratotypes in stratigraphy. Part 1.
Walker, M.J.C. 1995. Climatic changes in Europe during the Last Stratotype functions. Earth-Science Reviews 69, 307–332.
Glacial/Interglacial Transition. Quaternary International 28, Walsh, S.J., Butler, D.R., Malanson, J.P., et al. 2003. Mapping,
63–76. modeling, and visualization of the influences of geomorphic
Walker, M.J.C. 2001. Rapid climate change during the last processes on the alpine treeline ecotone, Glacier National Park,
glacial–interglacial transition; implications for stratigraphic MT, USA. Geomorphology 53, 129–145.
subdivision, correlation and dating. Global and Planetary Change Wanamaker Jr., A.D., Hetzinger, S., Halfar, J. 2011. Reconstructing
30, 59–72. mid- to high-latitude marine climate and ocean variability using
518 REFERENCES
bivalves, coralline algae, and marine sediment cores from the Webb, S.D., Graham, R.W., Barnosky, A.D., et al. 2003b. Vertebrate
Northern Hemisphere. Palaeogeography, Palaeoclimatology, paleontology. In Developments in Quaternary Science. Volume 1:
Palaeoecology 302, 1–9 The Quaternary Period in the United States (edited by A.R.
Wang, H., Hughes, R.E., Steele, J.D., et al. 2003. Correlation of Gillespie, S.C. Porter, B.F. Atwater). Elsevier, Amsterdam,
climatic cycles in middle Mississippi Valley loess and Greenland 519–538.
ice. Geology 31, 179–182. Webb III, T., Ruddiman, W.F., Street-Perrott, F.A., et al. 1993.
Wang, L., Sarnthein, M., Grootes, P.M., et al. 1999. Millennial Climatic changes during the past 18,000 years: regional syn-
reoccurrence of century-scale abrupt events of East Asian theses, mechanisms and causes. In Global Climates Since the
monsoon: a possible heat conveyor for the global deglaciation. Last Glacial Maximum (edited by H.E.Wright Jr., J.E. Kutzbach,
Paleoceanography 14, 725–731. T. Webb III, et al.). University of Minnesota Press, Minneapolis
Wang, L., Duan, J., Chen, J., et al. 2010. Temperature reconstruction and London, 514–535.
from tree-ring maximum density of Balfour spruce in eastern Webb III, T., Shuman, B., Williams, J.W. 2003a. Climatically
Tibet, China. International Journal of Climatology 30, 972–979. forced vegetation dynamics in eastern North America during
Wang, X., Dong, Z., Zhang, J., et al. 2004. Formation of the complex the Late Quaternary Period. In Developments in Quaternary
linear dunes in the central Taklimakan Sand Sea, China. Earth Science. Volume 1: The Quaternary Period in the United States
Surface Processes and Landforms 29, 677–686. (edited by A.R. Gillespie, S.C. Porter, B.F. Atwater). Elsevier,
Wang, X., Auler, A.S., Edwards, R.L., et al. 2006. Interhemispheric Amsterdam, 459–478.
anti-phasing of rainfall during the last glacial period. Quaternary Weber, S.L. 2010. The utility of Earth system Models of Intermediate
Science Reviews 25, 3391–3403. Complexity (EMICs). Wiley Interdisciplinary Reviews: Climate
Wang, Y., Amundson, R., Trumbore, S. 1996. Radiocarbon dating Change 1, 243–252.
of soil organic matter. Quaternary Research 45, 282–288. Weber, S.L., Drijfhout, S.S., Abe-Ouchi, A., et al. 2007. The modern
Wang, Y.J., Cheng, H., Edwards, R.L., et al. 2001. A high-resolution and glacial overturning circulation in the Atlantic ocean in PMIP
absolute-dated Late Pleistocene monsoon record from Hulu coupled model simulations, Climate of the Past 3, 51–64.
Cave, China. Science 294, 2345–2348. Webster, R., Oliver, M.A. 2000. Geostatistics for Environmental
Wang, Y., Cheng, H., Edwards, R.L., et al. 2008. Millennial- and Scientists. Wiley, Chichester.
orbital-scale changes in the East Asian monsoon over the past Wehmiller, J.F., Miller, G.H. 2000. Aminostratigraphic dating
224,000 years. Nature 451, 1090–1093. methods in Quaternary geology. In Quaternary Geochronology,
Wanner, H., Solomina, O., Grosjean, M., et al. 2011. Structure and Methods and Applications (edited by J.S. Noller, J.M. Sowers,
origin of Holocene cold events. Quaternary Science Reviews 30, W.R. Letts). American Geophysical Union, Washington, DC,
3109–3123. 187–222.
Washburn A.L. 1980. Permafrost features as evidence of climatic Weidman, C.R., Jones, G.A., Lohmann, K.C. 1994. The long-lived
change. Earth-Science Reviews 15, 327–402. mollusc Arctica islandica: a new paleoceanographic tool for
Wastegård, S., Rasmussen, T. 2001. New tephra horizons from the reconstruction of bottom temperatures for the continental
Oxygen Isotope Stage 5 in the North Atlantic: correlation shelves of the northern North Atlantic Ocean. Journal of
potential for terrestrial, marine and ice-core archives. Quaternary Geophysical Research 99, 18305–18314.
Science Reviews 20, 1587–1593. Weiss, D., Shotyk, W., Kempf, O. 1999. Archives of atmospheric lead
Watanabe, O., Jouzel, J., Johnsen, S., et al. 2003b. Homogeneous pollution. Naturwissenschaften 86, 262–275.
climate variability across East Antarctica over the past three Weldeab, S. 2012. Bipolar modulation of millennial-scale West
glacial cycles. Nature 422, 509–512. African monsoon variability during the last glacial (75,000–
Watanabe, S., Ayta, W.E.F., Hamaguci, H., et al. 2003a. Some 25,000 years ago). Quaternary Science Reviews 40, 21–29.
evidence of a date of first humans to arrive in Brazil. Journal of Weng, C., Hooghiemstra, H. and Duivenvoorden, J.F. 2006.
Archaeological Science 30, 351–354. Challenges in estimating past plant diversity from fossil pollen
Watanabe, T., Nakamura, T., Watanabe Nara, F., et al. 2009. data: statistical assessment, problems and possible solutions.
High-time resolution AMS 14C data sets for Lake Baikal and Biodiversity and Distributions 12, 310–318.
Lake Hovsgol sediment cores: changes in radiocarbon age and Weninger, B., Jöris, O. 2008. A 14C age calibration curve for the last
sedimentation rates during the transition from the last glacial to 60 ka: the Greenland-Hulu U/Th timescale and its impact on
the Holocene. Quaternary International 205, 12–20. understanding the Middle to Upper Paleolithic transition in
Watchman, A. 2000. A review of the history of dating rock varnishes. Western Eurasia. Journal of Human Evolution 55, 772–781.
Earth-Science Reviews 49, 261–277. Werdelin, L., Lewis, M.E. 2005. Plio–Pleistocene Carnivora of eastern
Watson, A.J., Bakker, D.C.E., Ridgwell, A.J., et al. 2000. Effect of iron Africa: species richness and turnover patterns. Zoological Journal
supply on Southern Ocean CO2 uptake and implications for of the Linnean Society 144, 121–144.
atmospheric CO2. Nature 407, 730–733. Werner, C.M., Mason, J.A., Hanson, P.R. 2011. Non-linear
Watson, E., Luckman, B.H. 2001. Dendroclimatic reconstruction of connections between dune activity and climate in the High
precipitation for sites in the southern Canadian Rockies. The Plains, Kansas and Oklahoma, USA. Quaternary Research 75,
Holocene 11, 203–213. 267–277.
REFERENCES 519
Werner, M., Mikolajewicz, U., Heimann, M., et al. 2000. Borehole Wiles, G.C., Barclay, D.J., Young, N.E. 2010. A review of lichen-
versus isotope temperatures on Greenland: seasonality does ometric dating of glacial moraines in Alaska. Geografiska Annaler
matter. Geophysical Research Letters 27723–27726. 92, 101–109.
West, R.G. 1977. Pleistocene Geology and Biology, 2nd edn. Longman, Wilkins, D., De Deckker, P., Fifield, L.K., et al. 2012. Comparative
London. optical and radiocarbon dating of laminated Holocene sedi-
West, R.G. 2000. Plant Life of the Quaternary Cold Stages. Cambridge ments in two maar lakes: Lake Kellambete and Lake Gnotuk,
University Press, Cambridge. south-western Victoria, Australia. Quaternary Geochronology 9,
Westaway, R. 2002. Geomorphological consequences of weak lower 3–15.
continental crust, and its significance for studies of uplift, Willerslev, E., Cooper, A. 2005. Ancient DNA. Proceedings of the Royal
landscape evolution, and the interpretation of river terrace Society of London B 272, 3–16.
sequences. Geologie en Mijnbouw (Netherlands Journal of Willerslev, E., Cappellini, E., Boomsma, W., et al. 2007. Ancient
Geosciences) 81, 283–303. biomolecules from deep ice cores reveal a forested southern
Westaway, R. 2009. Calibration of decomposition of serine to alanine Greenland. Science 317, 111–114.
in Bithynia opercula as a quantitative dating technique for Middle Williams, D.F., Kuzmin, M.I., Prokopenko, A.A., et al. 2001. The
and Late Pleistocene sites in Britain. Quaternary Geochronology Lake Baikal drilling project in the context of a global drilling
4, 241–259. initiative. Quaternary International 80–81, 3–18.
Westaway, R., Bridgland, D.R., Sinha, R., Demir, T. 2009. Fluvial Williams, J.W. 2002. Variations in tree cover in North America
sequences as evidence for landscape and climatic evolution in since the last glacial maximum. Global and Planetary Change
the Late Cenozoic: a synthesis of data from IGCP 518. Global and 35, 1–23.
Planetary Change 68, 237–253. Williams, J.W., Shuman, B.N., Webb III, T., et al. 2004b. Late
Westgate, J. 1989. Isothermal plateau fission track ages of hydrated Quaternary vegetation dynamics in North America: scaling from
glass shards from silicic tephra beds. Earth and Planetary Science taxa to biomes. Ecological Monographs 74, 309–334.
Letters 95, 226–234. Williams, M. 2012. The ~73 ka Toba super-eruption and its impact:
Westgate, J.A., Preece, S.J., Froese, D.G., et al. 2001. Dating Early and history of a debate. Quaternary International 258, 19–29.
Middle (Reid) Pleistocene glaciations in central Yukon by Williams, M.A.J., Dunkerley, D.L., De Deckker, P., et al. 1998.
Quaternary Environments, 2nd edn. Edward Arnold, London.
tephrochronology. Quaternary Research 56, 335–348.
Williams, M.A.J., Ambrose, S.H. van der Kaars, S., et al. 2009.
White, W.B. 2007. Cave sediments and paleoclimate. Journal of
Environmental impact of the 73 ka Toba super-eruption in
Cave and Karst Studies 69, 76–93.
South Asia. Palaeogeography, Palaeoclimatology, Palaeoecology
Whitehouse, N.J. 2006. The Holocene British and Irish ancient
284, 295–314.
forest fossil beetle fauna: implications for forest history,
Williams, P.W., King, D.N.T., Zhao, J.-X., et al. 2004a. Speleothem
biodiversity and faunal colonisation. Quaternary Science Reviews
master chronologies: combined Holocene 18O and 13C records
25, 1755–1789.
from the North Island of New Zealand and their palaeo-
Whitehouse, N.J., Smith, D. 2010. How fragmented was the British
environmental implications. The Holocene 14, 194–208.
Holocene wildwood? Perspectives on the ‘Vera’ grazing debate
Williams, P.W., Neil, H.L., Zhao, J.-X. 2010. Age frequency
from the fossil beetle record. Quaternary Science Reviews 29, distribution and revised stable isotope curves for New Zealand
539–553. speleothems: palaeoclimatic implications. International Journal
Whiteman, C.A., Kemp, R.A. 1990. Pleistocene sediments, soils and of Speleology 39, 99–112.
landscape evolution at Stebbing, Essex. Journal of Quaternary Willis, K.J., van Andel, T.H. 2004. Trees or no trees? The environ-
Science 5, 145–161. ments of central and eastern Europe during the Last Glaciation.
Whitlock, C., Bianchi, M.M., Bartlein, P.J., et al. 2006. Postglacial Quaternary Science Reviews 23, 2369–2387.
vegetation, climate, and fire history along the east side of the Willis, K.J., Araújo, M.B., Bennett, K.D., et al. 2007. How can a
Andes (lat 41–42.5°S), Argentina. Quaternary Research 66, knowledge of the past help to conserve the future? Biodiversity
187–201. conservation and the relevance of long-term ecological studies.
Whitmore, J., Gajewski, K., Sawada, M., et al. 2005. Modern pollen Philosophical Transactions of the Royal Society of London B 362,
data from North America and Greenland for multi-scale paleo- 175–187.
environmental applications. Quaternary Science Reviews 24, Willman, H.B., Frye, J.C. 1970. Pleistocene stratigraphy of Illinois.
1828–1848. Illinois Geological Survey Bulletin 94.
Widdowson, M. 2007. Laterite and Ferricrete. In Geochemical Wilmshurst, J.M., McGlone, M.S. 2005. Corroded pollen and
Sediments and Landscapes (edited by D.J. Nash, S.J. McLaren). spores as indicators of changing lake sediment sources and
Blackwell, Oxford. catchment disturbance. Journal of Paleolimnology 34, 503–517.
Wiersma, A.P., Riche, D.M., Renssen, H. 2011. Fingerprinting the Wilmshurst, J.M., Higham, T.F.G., Allen, H., et al. 2004. Early
8.2 ka event climate response in a coupled climate model. Journal Maori settlement impacts in northern coastal Taranaki, New
of Quaternary Science 26, 118–127. Zealand. New Zealand Journal of Ecology 28, 167–179.
Wild, M.T., Tabner, B.J., Macdonald, R. 1999. ESR dating of quartz Wilson, R.S.J., Luckman, B.H. 2003. Dendroclimatic reconstruc-
phenocrysts in some rhyolitic extrusive rocks using Ti and Al tion of maximum summer temperatures from upper treeline
impurity centres. Quaternary Science Reviews 18, 1507–1514. sites in Interior British Columbia. The Holocene 3, 851–861.
520 REFERENCES
Winchester, V. Harrison S. 2000. Dendrochronology and lichen- Wolin, J.A., Stone, J.R. 2010. Diatoms as indicators of water-level
ometry: colonisation, growth rates and dating of geomorpho- change in freshwater lakes. In The Diatoms: Applications for
logical events on the east side of the North Patagonian Icefield, the Environmental and Earth Sciences, 2nd edn (edited by J.P.
Chile. Geomorphology 34, 181–194. Smol, E.F. Stoermer). Cambridge University Press, Cambridge,
Winkler, S. 2000. The ‘Little Ice Age’ maximum in the Southern 174–185.
Alps, New Zealand: preliminary results at Mueller Glacier. The Woodhead, J., Hellstrom, J., Maas, R., et al. 2006. U–Pb geochron-
Holocene 10, 643–647. ology of speleothems by MC-ICPMS. Quaternary Geochronology
Winkler, S. 2004. Lichenometric dating of the ‘Little Ice Age’ 1, 208–221.
maximum in Mt Cook National Park, Southern Alps, New Woodroffe, C.D., Murray-Wallace, C.V. 2012. Sea-level; rise and
Zealand. The Holocene 14, 911–920. coastal change: the past as a guide to the future. Quaternary Science
Winkler, S., Matthews, J.A., Shakesby, R.A., et al. 2003. Glacier Reviews 54, 4–11.
variations in Breheimen, southern Norway: dating Little Ice Woodroffe, S.A. 2009. Recognising subtidal foraminiferal assem-
Age moraine sequences at seven low-altitude glaciers. Journal of blages: implications for quantitative sea-level reconstruc-
Quaternary Science 18, 395–413. tions using a foraminfera-based transfer function. Journal of
Winsborrow, M.C.M., Clark, C.D., Stokes, C.R. 2010. What controls Quaternary Science 24, 215–223.
the location of ice streams? Earth-Science Reviews 103, 45–59. Woodroffe, S.A., Horton, B.P. Larcombe, P., et al. 2005. Con-
Winsemann, J., Asprion, U., Meyer, T., et al. 2007. Facies charac- temporary intertidal foraminifera distributions of mangrove
teristics of Middle Pleistocene (Saalian) ice-margin subaqueous environments from the central Great Barrier Reef shelf, Aus-
fan and delta deposits, glacial Lake Leine, NW Germany. tralia: implications for sea-level reconstructions. Journal of
Sedimentary Geology 193, 105–129. Foraminiferal Research 35, 259–270.
Wintle, A.G. 1996. Archaeologically-relevant dating techniques Woodward, J.C. 2014. The Ice Age: A Very Short Introduction. Oxford
for the next century. Journal of Archaeological Science 23,
University Press, Oxford.
395–413.
Woodward, J.C., Goldberg, P. 2001. The sedimentary record in
Wintle, A.G. 2008. Luminescence dating; where it has been and
Mediterranean rockshelters and caves: archives of environ-
where it is going. Boreas 37, 471–482.
mental change. Geoarchaeology 16, 327–354.
Wintle, A.G. 2010. Future directions of luminescence dating of
Wooldridge, S.W., Linton, D.L. 1955. Structure, Surface and
quartz. Geochronometrica 37, 1–7.
Drainage in South-east England. George Philip, London.
Wintle, A.G., Murray, A.S. 2000. Quartz OSL: effects of thermal
Woolfenden, W.B. 2003. A 180,000-year pollen record from
treatment and their relevance to laboratory dating procedures.
Owens Lake, CA: terrestrial vegetation change on orbital scales.
Radiation Measurements 32, 387–400.
Wiseman, J., El-Baz, F. 2007. Remote Sensing in Archaeology. Springer,
Heidelberg and New York. Wooller, M.J., Francis, D., Fogel, M.L., et al. 2004. Quantitative
Wittkop, C.A., Teranes, J.L., Dean, W.E., et al. 2009. A lacustrine paleotemperature estimates from δ18O of chironomid head
carbonate record of Holocene seasonality and climate. Geology capsules preserved in arctic lake sediments. Journal of Paleo-
37, 695–698. limnology 31, 267–274.
Wohlfarth, B., Possnert, G. 2000. AMS radiocarbon measurements Worni, R., Stioffel, M., Huggel, C., et al. 2012. Analysis and dynamic
from the Swedish varved clays. Radiocarbon 42, 323–333. modelling of a moraine failure and glacier lake outburst flood
Wohlfarth, B., Blaauw, M., Davies, S.M., et al. 2006. Constraining at Ventisquero Negro, Patagonian Andes (Argentina). Journal
the age of Lateglacial and early Holocene pollen zones and of Hydrology 444/445, 134–145.
tephra horizons in southern Sweden with Bayesian probability Wright Jr., H.E., Kutzbach, J.E., Webb III, T., et al. (eds). 1993. Global
methods. Journal of Quaternary Science 21, 321–334. Climates Since the Last Glacial Maximum. University of
Woillard, G.M., Mook, W.G. 1982. Carbon-14 dates at Grande Pile: Minneapolis Press, Minnesota.
correlation of land and sea chronologies. Science 215, 159–161. Wright Jr., H.E., Stefanova, I., Tian, J., et al. 2004. A chronological
Wolfe, A.P. 1997. On diatom concentrations in lake sediments: framework for the Holocene vegetational history of central
results from an inter-laboratory comparison and other tests Minnesota: the Steel Lake pollen record. Quaternary Science
performed on a uniform sample. Journal of Paleolimnology 18, Reviews 23, 611–626.
261–268. Wright, J.S. 2001. ‘Desert’ loess versus ‘glacial’ loess: quartz silt
Wolfe, A.P. 2002. Climate modulates the acidity of Arctic lakes on formation, source areas and sediment pathways in the formation
millennial time scales. Geology 30, 215–218. of loess deposits. Geomorphology 36, 231–256.
Wolff, E.W., Fischer, H., Röthlisberger, R. 2009. Glacial terminations Wroe, S., Field, J. 2006. A review of the evidence for a human role
as southern warmings without northern control. Nature in the extinction of Australian megafauna and an alternative
Geoscience 2, 206–209. interpretation. Quaternary Science Reviews 25, 2692–2703.
Wolff, E., Barbante, C., Becagli, S., et al. 2010. Changes in Wu, H., Guiot, J., Brewer, S., et al. 2007. Climate changes in Eurasia
environment over the last 800,000 years from chemical analysis and Africa at the last glacial maximum and mid-Holocene:
of the EPICA Dome C ice core. Quaternary Science Reviews 29, reconstruction from pollen data using inverse vegetation
285–295. modelling. Climate Dynamics 29, 211–229.
REFERENCES 521
Wu, N., Liu, T., Liu, X.P., et al. 2002. Mollusk record of millennial Yasuhara, M., Hunt, G., Cronin, T.M., et al. 2012. Climatic forcing
climate variability in the Loess Plateau during the Last Glacial of Quaternary deep-sea benthic communities in the North
Maximum. Boreas 31, 20–27. Pacific Ocean. Paleobiology 38, 162–179.
Wu, Y., Shijie, L., Lücke, A., et al. 2010. Lacustrine radiocarbon Yeloff, D., Charman, D., van Geel, B., et al. 2007. Reconstruction of
reservoir ages in Co Ngoin and Zigê Tangco, central Tibetan hydrology, vegetation and past climate change in bogs using fungal
Plateau. Quaternary International 212, 21–25. microfossils. Review of Palaeobotany and Palynology 146,
Wulf, S., Kraml, M., Brauer, A., et al. 2004. Tephrochronology 102–145.
of the last 100 ka lacustrine sediment record of Lago Grande di Yokoyama, Y., Lambeck, K., De Deckker, P., et al. 2000. Timing of
Monticchio (southern Italy). Quaternary International 122, 7–30. the Last Glacial Maximum from observed sea-level data. Nature
Wünnemann, B., Hartmann, K., Janssen, M., et al. 2007. Responses 406, 713–716.
of Chinese desert lakes to climate instability during the past Young, G.H.F., Bale, R.J., Loader, N.J., et al. 2012. Central England
45,000 years. In Developments in Quaternary Sciences, Volume 9: temperature since AD 1850: the potential of stable carbon isotopes
Late Quaternary Climate Change and Human Adaptation in Arid in British oak trees to reconstruct past summer temperatures.
China (edited by D.B. Madsen, F.H. Chen, X. Gao). Elsevier, Journal of Quaternary Science 27, 606–614.
Amsterdam, 11–24. Young, N.E., Schaeffer, J.M., Briner, J.P., et al. 2013. A 10Be prduction-
Wunsch, C. 2004. Quantitative estimate of the Milankovitch-forced rate calibration for the Arctic. Journal of Quaternary Science 28,
contribution to observed Quaternary climate change. Quaternary 515–526.
Science Reviews 23, 1001–1012. Yu, G., Harrison, S.P. 1995. Holocene changes in atmospheric
Wunsch, C. 2010. Towards understanding the Paleocean. Quaternary circulation patterns as shown by lake status change in northern
Science Reviews 29, 1960–1967. Europe. Boreas 24, 260–268.
Wurth, G., Niggeman, S., Richter, D.K., et al. 2004. The Younger Yu, Z., Wright Jr., H.E. 2001. Response of interior North America
Dryas and Holocene climate record of a stalagmite from Hölloch to abrupt climate oscillations in the North Atlantic region during
Cave (Bavarian Alps, Germany). Journal of Quaternary Science the last deglaciation. Earth-Science Reviews 52, 333–369.
19, 291–298. Zachos, J.C., Pagani, M., Sloan, L., et al. 2001. Trends, rhythms and
Wyrwoll, K.-H., Miller, G.H. 2001. Initiation of the Australian aberrations in global climate 65 Ma to present. Science 292,
summer monsoon, 14,000 years ago. Quaternary International 686–693.
83–85, 119–128. Zagwijn, W.H. 1994. Reconstruction of climate change during the
Xiong,S., Ding, Z., Liu, T. 2001. Climatic implications of loess Holocene in western and central Europe based on pollen records
deposits from the Beijing area. Journal of Quaternary Science of indicator species. Vegetation History and Archaeobotany 3,
16, 575–582. 65–88.
Yadava, M.G., Ramesh, R. 2005. Monsoon reconstruction from Zalasiewicz, J., Smith, A., Brenchley, P., et al. 2004. Simplifying the
radiocarbon dated tropical Indian speleothems. The Holocene stratigraphy of time. Geology 32, 1–4.
15, 48–59. Zalasiewicz, J., Smith, A., Barry, T.L. 2008. Are we now living in the
Yamazaki, T., Oda, H. 2002. Orbital influence on Earth’s magnetic Anthropocene? GSA Today 18, 4–8.
field: 100,000-year periodicity in inclination. Science 295, Zalasiewicz, J., Williams, M., Haywood, A., et al. 2011. The
2435–2438. Anthropocene: a new epoch of geological time. Philosophical
Yanes , Y., Aguirre, J., Alonso, M.R., et al. 2011. Ecological fidelity Transactions of the Royal Society of London A 369, 1036–1055.
of Pleistocene–Holocene land snail shell assemblages preserved Zárate, M.A. 2003. Loess of southern South America. Quaternary
in carbonate-rich paleosols. Palaios 26, 406–419. Science Reviews 22, 1987–2006.
Yang, G., Zhang, X., Tian, M., et al. 2011. Alluvial terrace systems in Zazo, C., Goy, J.L., Dabrio, C.J., et al. 2003. Pleistocene raised marine
Zhangjiajie of northwest Hunan, China: Implications for climatic terraces of the Spanish Mediterranean and Atlantic coasts:
change, tectonic uplift and geomorphic evolution. Quaternary records of coastal uplift, sea-level highstands and climate changes.
International 233, 27–39. Marine Geology 194, 103–133.
Yang, J.-D., Chen, J., Zhang, J.-X. 2005. Variations in 87Sr/86Sr of Zazula, G.D., Telka, A.M., Harington, C.R., et al. 2006. Macrofossils
the Huanxian profile, Loess Plateau of China from 40 ka B.P. from Yukon Territory: implications for Late Pleistocene refugia
to 10 ka B.P. and Heinrich events. Geochemical Journal 39, in Eastern Beringia. Arctic 59, 391–400.
165–171. Zech, W., Glaser, B., Abrammowski, U., et al. 2003. Reconstruction
Yang, X., Rost, K.T., Lehmkuhl, F., et al. 2004. The evolution of dry of the Late Quaternary glaciation of the Macha Khola Valley
lands in northern China and in the Republic of Mongolia since (Gorkha Himal, Nepal) using relative and absolute (14C, 10Be,
the Last Glacial Maximum. Quaternary International 118–119, dendrochronology) dating techniques. Quaternary Science
69–85 Reviews 22, 2253–2265.
Yasuhara, M., Yamazaki, H., Irizuki, T., et al. 2003. Temporal changes Zeeb, B.A., Smol, J.P. 2001. Chrysophyte scales and cysts. In Track-
of ostracode assemblages and anthropogenic pollution during ing Environmental Change Using Lake Sediments. Volume 3:
the last 100 years, in sediment cores from Hiroshima Bay, Japan. Terrestrial, Algal and Silicious Indicators (edited by J.P.Smol,
The Holocene 13, 527–536. H.J.B. Birks, W.M. Last). Kluwer, Dordrecht, 203–223.
522 REFERENCES
Zerboni, A. 2008. Holocene rock varnish on the Messak plateau Zhou, EW., Zheng, Y., Meyers, P.A., et al. 2010. Postglacial climate-
(Libyan Sahara): chronology of weathering processes. change record in biomarker lipid compositions of the Hani peat
Geomorphology 102, 640–651. sequence, Northeastern China. Earth and Planetary Science
Zhai, D., Xiao, J., Zhou, L., et al. 2011. Holocene East Asian monsoon Letters 294, 37–46.
variation inferred from species assemblage and shell chemistry Zhou, S., Li, J., Zhang, S. 2002. Quaternary glaciation of the Bailang
of the ostracodes from Hulun Lake, Inner Mongolia. Quaternary River Valley, Qilian Shan. Quaternary International 97–98,
Research 75, 512–522. 103–110.
Zhand, M., DaoXian, Y., Lin, Y., et al. 2004. A 6000-year high- Zic, M., Negrini, R.M., Wigand, P.E. 2002. Evidence of synchronous
resolution stalagmite in Xiangshui Cave, Guinin, China. The climatic change across the Northern Hemisphere between the
Holocene 14, 697–702. North Atlantic and the northwestern Great Basin, United States.
Zhang, J., Chen, F., Holmes, J.A., et al. 2011. Holocene monsoon Geology 30, 635–638.
climate documented by oxygen and carbon isotopes from lake Zillén, L., Lenz, C., Julbert, T. 2012. Stable lead (Pb) isotopes and
sediments and peat bogs in China: a review and synthesis. concentrations – a useful independent dating tool for Baltic Sea
Quaternary Science Reviews 30, 1973–1987. sediments. Quaternary Geochronology 9, 41–45.
Zhang, R., Delworth, T.L. 2006. Impact of Atlantic Multidecadal Zink, A.J.C., Susino, G.J., Porto, E., et al. 2012. Direct OSL dating of
Oscillations on India/Sahel rainfall and Atlantic hurricanes.
Iron Age pottery from South Africa – preliminary dosimetry
Geophysical Research Letters 33, L17712, doi: 10.1029/2006
investigations. Quaternary Geochronology 8, 1–9.
GL026267.
Zinovjev, E.V. 2006. Problems of ecological interpretation of
Zhang, Y., Fountain, D.W., Hodgson, R.M., et al. 2004a. Towards
Quaternary insect faunas from the central part of Northern
automation of palynology 3: pollen pattern recognition using
Eurasia. Quaternary Science Reviews 25, 1821–1840.
Gabor transforms and digital moments. Journal of Quaternary
Zolitschka, B. 2007. Varved lake sediments. In Encyclopedia of
Science 19, 763–768.
Zhang, Z., Mann, M.E., Cook, E.R. 2004b. Alternative methods of Quaternary Science (edited by S.A. Elias). Elsevier, Amsterdam,
proxy-based climate field reconstruction: applications to summer 3105–3114.
drought over the coterminous United States back to AD 1700 from Zolitschka, B., Brauer, A., Negendank, J.F.W., et al. 2000. Annually
tree-ring data. The Holocene 14, 502–516. dated late Weichselian continental paleoclimate record from the
Zhao, J., Xia, Q., Collerson, K. 2001. Timing and duration of the last Eifel, Germany. Geology 28, 783–786.
interglacial inferred from high resolution U-series chronology Zolitschka, B., Behre, K.-E., Schneider, J. 2003. Human and climatic
of stalagmite growth in Southern Hemisphere. Earth and impact on the environment as derived from colluvial, fluvial
Planetary Science Letters 184, 635–644. and lacustrine archives – examples from the Bronze Age to the
Zhao, Y., Sayer, C. D., Birks, H. H., et al. 2006. Spatial representation Migration period, Germany. Quaternary Science Reviews 22,
of aquatic vegetation by macrofossils and pollen in a small and 81–100.
shallow lake. Journal of Paleolimnology 35, 335–350. Zwally, H.J., Giovinetto, M.B., Li, J., et al. 2005. Mass changes of
Zhou, C., Liu, Z., Wang, Y., et al. 2001. Climatic cycles investigated the Greenland and Antarctic ice sheets and shelves and
by sediment analysis in Peking Man’s Cave, Zhoukoudian, contributions to sea-level rise: 1992–2002. Journal of Glaciology
China. Journal of Archaeological Science 27, 101–109. 51, 509–527.
Index
Page numbers in bold refer to figures, page numbers in italic refer to tables, n refers to notes
2.6 Ma event 5 airborne laser swath mapping (ALSM) 26 ice-accumulation years 174; ice-core
2.8 ka event 433 Alaska, United States of America 214 stratigraphy 8, 172, 173, 177–8, 394–5,
4.2 ka event 431, 432, 433 algal blooms 246 395, 396, 397, 397, 399, 405–6, 405; ice-
8.2 ka event 429, 430, 431 Allerød Interstadial 12, 413 sheet modelling 46–7; isotope ratios
allostratigraphy 352 411, 411; Last Termination 414, 414,
AABR (area-altitude balance ratio) 49 alpha-recoil track (ART) dating 290 415; Lateglacial climate change 422–3;
AAR (accumulation area ratio) 48–50 altiplanation 54 methane signal 406, 406; Milankovitch
AAR chronology 334, 335 American Bureau of Standards 272 cycle correlation 371; teleconnection
ablation area (glacier) 48 American Stratigraphic Commissions 405–7, 405, 406; tephra marker horizons
accumulation area (glacier) 48 359 329
abandoned lake shorelines 83–5, 83 amino-acid geochronology 332; AAR Antevs, Ernst 307
absolute dating 343 chronology 334, 335; aminostratigraphy Anthropocene, the 439–40
absolute sea-level change 59 334; applications 336, 337, 338–9; anthropogenic signature 196, 197, 427,
accelerator mass spectrometry 271, 272–5, diagenesis 334, 336; epimerization 334; 436–40, 437, 438
273, 274, 278–9 fossils 339; intra-crystalline fraction 335, anthropogenic warming 437–9, 437,
accretionary soils 123 336; isomeric forms 333; molluscan 438
acidity profiles 178 remains 336, 338, 338, 339; Aqua (EOS PM-1) 23
acoustic profiling 24 palaeothermometry 339; problems 334, Arabian Sea 246
aerial photography 20, 22 336; protein chemistry 332–3, 333; Aral Sea 83–4
Africa: dunefields 87; early hominid sites protein degradation 336; racemization Aran Islands 196, 197
285; ice-core stratigraphy 172; lake- 334, 336, 339; recent developments 336; archaeobotanical evidence 215
levels 135, 137, 138; pluvial lake sequence correlation 335, 336, 337, 338; archaeoentomology 225
sequences 133; pollen analysis 192; source material 334 archaeology 24, 205–6, 225, 234–5, 241,
Quaternary environments 260; amino-acid ratios 334 243, 265, 293, 339, 340, 342
vertebrate remains 260 aminostratigraphy 334 archaeomagnetic measurements 320
African Humid Period 137 aminozones 336 Arduino, Giovanni 5
Agassiz, Louis 6 Anglian Glacial Stage 351 argon–argon dating 285
age estimate techniques 267 Annual Greenhouse Gas Index 396 argon-isotope dating 284–5, 284, 322–3,
age modelling: Lateglacial 420–1; annual ice increments 173–5, 174 325, 327
radiocarbon dating 282–3, 283 Antarctic Bottom Water current 404 Arrhenius, Svante 437
age-equivalence techniques 267 Antarctic Circumpolar Current 246 assemblage biozones 353, 354
age-equivalent stratigraphic markers Antarctic Ice Core Chronology 315 astronomical forcing 4, 390–2, 393
319; biostratigraphy 319; oxygen Antarctica: 4.2 ka event 431; annual layers astronomical parameters 4
isotope chronostratigraphy 319; 312, 313, 315; anti-phase behaviour Astronomical Theory 13–15, 14, 389,
palaeomagnetism 319, 319–24, 321, 406–7; correlation 377; deglaciation 389–90, 390–2, 398
322, 323, 324, 325; tephrochronology 404; dust record 132; glacial–interglacial Atlantic Bottom Water (ABW) current
319, 325–31, 326, 328, 329, 331 cycles 394–5; glaciation, 35 Ma 5; 401, 402, 403, 408, 408
agriculture, human impacts 437 ice streams 105; ice volume 170, 392; Atlantic Cold Reversal (ACR), the 414
524 INDEX
Atlantic Meridional Overturning biomolecular products 181 Cader Idris, Wales 57

Circulation (AMOC) 401, 402, 403–4, biostratigraphic correlation 233, 237, 259 caesium-137 296, 297–8
403, 406–7, 408, 408, 417, 422, 424 biostratigraphy 181, 319, 331–2, 348, Cairngorm Mountains, Scotland 54
Atlantic Multidecadal Oscillation (AMO) 353–4, 353, 363 calcareous nannoplankton 245
436 biozones 348, 353–4, 353; boundaries 354, calcrete 91
atmospheric circulation 140; general 419 CALIB 280–1
circulation models 8–9; and lake-level bipolar seesaw 406–7, 409, 411, 412, 418, California, Gulf of 306
changes 137, 160 421–2, 440 Callendar, Guy 437
atmospheric forcing 395–7, 395, 396, 397 Black Sea 237, 240 Canada 68, 68; isobase patterns 71;
atmospheric variations, CO2 172 block-fields 54, 54, 58 Keewatin 41; lake sedimentation rates
atomic absorption spectrophotometer Blytt–Sernander climatic scheme 161 158; sea-level change 72
(AAS) 98 bogs: see lake, mire and bog sediments carbon: global reservoirs 270, 276; isotopic
atomic force microscopy (AFM) 97 Bølling 12 fractionation 276, 277; marine
atomic numbers 269 Bond, Gerard 400 circulation 276–7
Aura (EOS CH-1) 23 Bond cycles 400, 401, 404 carbon cycle 195
Australia 66, 69, 73; amino-acid bone, vertebrate 254, 255, 256; amino-acid carbon dioxide 15, 15, 16, 270–1;
geochronology 336, 337; geochronology 334; dating 258, 278, atmospheric variations 172; exchange
dendrochronology 302; dunefields 86, 289; excavation 257; fluorine content 408–9; greenhouse effect 437–8; human
87, 87–8, 87, 88; dust circulation 88, 89; 339–40; fossilization 256; identification impacts 437–9, 438, 442
fire histories 214–15; fluvial landforms 257; micro-vertebrates 257; uranium carbon isotopes: deep-sea sediment 171–2;
89; lake-level changes 137; Little Ice Age content 339–40 molluscan remains, non-marine 234
434; mass extinctions 261–2; mutual borehole records 94–5 carbon modelling 284
climatic range (MCR) 224; ostracod borehole thermometry 177 carbon sinks 384
analysis 240; plant macrofossils 212; boundary conditions 251, 266n9 carbon storage and release 195, 407–9, 408,
Quaternary environments 260; sand seas boundary stratotypes 3, 4–5, 4, 349 416–17
131; sea-level change 169; speleothem box models 380–1, 381, 393 carbonate cave deposits 149–50, 150, 289;
growth and climatic change 145; Braldu Valley, Pakistan 108 see also speleothem
vertebrate remains 260 Brandenburg moraines 30–1 carbonate dissolution, deep-sea sediment
Austria 143 Brazil 62, 150, 150 171
autochthonous sediments 142 brickearths 128 carbonate lake deposits 138, 140
Britain and Ireland: biostratigraphy 354; carbonate production 91
Baffin Island 115 Coleoptera 218, 219, 220, 220, 222, 224, carbonate varves 306
bald-earth elevation mapping 25 224; deglaciation 34–5, 307; episodes of Cariaco Basin, Venezuela 280, 306, 309,
Baltic Sea 160 climatic deterioration 213, 213; eustatic 373, 373, 423
Barbados, oxygen isotope record 61 sea-level change 64–5, 65, 66; extent of Caspian Sea 83–4
bathymetry 24, 25, 33 ice cover 32, 33–5, 34; fire histories 214; Catastrophist philosophy 5
Bayesian-based statistical modelling 282–3, Foraminifera 242; ice-sheet modelling cation-ratio dating 341
282: see also radiocarbon dating 45, 46; indicator erratics 113; lake causality 380
Beaver Lake, United States of America 201 sedimentation rates 157–8, 158; mean cave sediments 140–1; autochthonous 142;
Belgium 56, 119, 120 annual air temperature (MAAT) carbonate 149–50, 150; cave earth 142;
benthic foraminifera 169, 170, 171, 172 119–20, 121; molluscan remains 228, dating 141; detrital 141–3; detritus types
Berger–Parker index (BPI) 233 231, 232, 233, 234, 336, 338, 338; North 140–1, 141; dripstone 141, 143;
Bering Sea 221 Atlantic Oscillation (NAO) 436; flowstone 141, 143; fossil content 142;
Bering Strait 73 ostracod analysis 239; pingos 56; plant human coprolites 143; molluscan
Bermuda 62 macrofossils 207, 213; pollen assemblage remains 231–2; organic components
Bernhardi, Albrecht 5 zones 332; Quaternary environments 142; plant macroremains 142; pollen
Beschel, Roland 315 261; readvances 34–5, 47; soil 142, 193; sedimentation 142; sequences
beta decay 271–2 stratigraphy 355; surface wetness change 142–3; speleothem 141, 143, 149, 334,
Bhutan-Himalaya 28 163, 164; vertebrate biostratigraphy 259; 371; speleothem formation and sea-level
binge-purge episodes 404, 405 vertebrate remains 261 change 149; speleothem formation and
biogenic sediments 93 BRITICE project 33, 91n6 tectonic activity 149; speleothem growth
biogeochemical cycle 384 Broecker, Wally 403 and climatic change 143–4, 144;
biological evidence 181, 266 Brückner, Eduard 6, 13, 33 speleothem isotope ratios 145–8, 147,
biomarkers 253, 265 Buckland, William 7 148; vertebrate remains 257–8; water-
biomass 195 Bugs-CEP website 218 filled passages 179n7
BIOME-6000 388 buried palaeosols 122, 126 cave sequences 126–7
INDEX 525
Cenozoic era 1, 2, 3, 3 climate GCMs 383 370, 371, 378; oxygen isotope ratios 399;
Challenger, HMS 8, 241 climate rhythms 4–5 palaeosols 364; pollen and pollen
Chapman Glacier, Canada 49 Climate/Long Range Investigation analysis 368; principles 362–3;
Chappell, J. 62 Mapping and Prediction (CLIMAP) 9, radiometric dating 365; record
charcoal 127, 142, 196, 196, 210, 214–15 250–1, 265–6, 379 synchronization 374–8, 375, 376, 377;
chemical alteration processes chronology climatic events, synchronization 420, 421 sea-level change 365; shoreline 364–5;
332; amino-acid geochronology 332–9, climatic signal, isotopic ratios 145 sub-Milankovitch timescales 371, 371,
333, 335, 337, 338; bone fluorine climatic terraces 77 372, 373–4, 373, 374; tephra 364
content 339–40; bone uranium content climatostratigraphy 9, 348, 358–61, 359, cosmic ray flux 433
339–40; obsidian hydration dating 360, 361, 378 cosmogenic nuclides 115, 294, 295, 295
(OHD) 340; pedogenesis 342–3; rock- Clogher Head Readvance, Ireland 35 cosmogenic radionuclide (CRN) dating
weathering 340–2, 341 closed basin (endoreic) lakes 133–4, 203 294–6, 295, 297
chemical fossils 265 Coccolithophores 245–6, 245, 246, 248, coversand 127, 130, 130
chemical remanent magnetism (CRM) 320 250 Craford Lake, Canada 158
chemical stratigraphy 348 Cockburn Event, the 38 crag-and-tail features 41
chemical tracers 254 coercivity curves 324 crescentic fractures 40, 42
chemical varves 152, 304, 306 COHMAP (Co-operative Holocene crescentic gouges 40
China: 4.2 ka event 431; biomarkers 265; Mapping Project) 9, 251, 379, 384, Croatia 149
dust flux 309; Loess Plateau 127, 128, 384 Croll, James 13
128, 129, 131–2, 132, 234, 325, 355, 364, cold stage 9 crossing striations 40, 41
370, 371, 374; molluscan remains 234; Coleoptera 119, 215, 215–16, 216, 217; crustal flexuring 68, 68; isostatic recovery
pollen analysis 192; sand seas 131 analysis 218–25, 220, 221, 223, 224; 69, 69
Chironomidae 225–8, 226, 227 archaeoentomology 225; climate change crustal warping 6
chronohorizons 361–2 boundaries 360; distribution 221–2, 221; cryoplanation terraces 54
chronostratigraphic units 2, 361–2, 362 evolutionary stasis 218–19; habitat cryostatic facies 121
chronostratigraphy 348, 361–2, 362, 378 preferences 219–21, 220; human cryostructures 117
chronozones 362 impacts 225; identification 216; isotope cryoturbation structures 117–18, 117, 118,
chrons 2 ratios 224; palaeoenvironmental 119
Chrysophytes 263 reconstruction 221–4, 221, 223, 224; cultural eutrophication 205
Cladocera 263 record 216, 218; sensitivity to cultural landscape 196
clastic varves 304, 304–5, 305 temperature variations 222; speciation cumulative soils 123
clasts 93 219
clay mineral ratios 115 computer-based models 8–9 Dansgaard–Oeschger (DO) events 78, 176,
clay mineralogy 98 continental levering 69 243, 261, 371, 399, 400, 400, 401, 403–5,
CLIMAP (Climate/Long Range continuous flow analysis (CFA) 175 440
Investigation Mapping and Prediction) Coope, Russell 215, 218–19, 222 Danube, River 75, 77, 78, 128, 338
9, 250–1, 265–6, 379 Co-operative Holocene Mapping Project dating 2, 7–8, 33; accuracy 267, 269, 345;
climate change 309; anthropogenically (COHMAP) 9, 251, 379, 384, 384 bias 267; carbonate cave deposits 289;
induced 388–9; Astronomical Theory coral polyps 263–4 cave sediments 141; corals 288–9, 294;
13–15, 14; astronomically driven 389; coral reef terraces 67, 67 deep-sea sediment 287; dunefields 87,
atmospheric influences 15–16, 15; coral reefs 61–2, 263–4, 331; dating 288–9, 88; early hominid sites 285; glacial
boundaries 358–60, 359, 360; causes 294 sediments 292, 294; glacier ice 298,
13–16, 14, 15; and coccolith productivity coral stratigraphy 264, 316, 317, 317, 318, 311–12; Greenland ice sheet 268;
246; dendroclimatology records 303; 319 Heinrich events 373; ice-core
feedback loops 15–16; geologic-climatic Cordilleran ice sheet, North America 35, stratigraphy 175, 297, 313; lake
units 358–61; and lake sediments 36, 38–9, 68, 73 sediment 278, 279, 287, 297; lake-level
155–61, 156, 157, 158, 159; lake-level coring 94–6, 95, 151; see also ice-core changes 135; Lateglacial 419–20; marine
changes and 135, 159–60, 159; stratigraphy oxygen isotope record 268; molluscan
Lateglacial 421–6, 423, 424, 425, 426; coring equipment 94, 95, 172 remains 235, 277, 278, 289, 294, 339;
Milankovitch cycles 389–99, 390, 391, correlation 2; definition 363; difficulties moraines 296, 297, 316, 342; palaeosols
394, 395, 396, 397, 398; the Milankovitch 359, 363; event stratigraphy 365; 126; peat deposits 161–2, 289; periglacial
hypothesis 14–15; and river terraces homotaxial errors 363; ice-core structures 122; plant macrofossils 210;
76–7, 81–2; and speleothem growth stratigraphy 366–7; land–sea 366–7, 371, polarity events 322–3; pollen 189, 362;
143–4, 144; sub-Milankovitch events 16; 373, 377–8, 377; Lateglacial 420–1, 421; precision 267, 269, 274–5, 345, 426, 440;
time-transgression 367; and volcanic marine oxygen isotope record 366, 399; Quaternary events 308–9; ratite eggshell
activity 330 Milankovitch cycles 367–9, 368, 369, 338; reliability 267; river terraces 77;
526 INDEX
sea-level change 59, 61–2; sea-waters States of America 301, 301; wood earth hummocks 58
276–7; soils 279; speleothem 143, 144, density 303 Earth Observing System (EOS) 23
149, 286–7, 288, 293, 298, 316–17; dendroclimatology 299, 302–4 Earth Resource Technology Satellites
stratigraphic resolution 343–4, 344; denivation features 130 (ERTS) 22–3
temporal resolution 343–4, 344; tephra Denmark 32, 157–8, 207, 293 earth system models of intermediate
325, 327; trapped gas 312; uncertainty depositional landforms 42–3 complexity (EMICs) 383–4, 384, 385,
267, 272, 362; vertebrate remains 258, Desnoyers, Jules 5 386, 386, 393
278, 289, 294; volcanic rocks 284–5, 289; detrital remanent magnetism (DRM) 320 East African Rift Valley 330, 369
weathering 294 Devensian Lateglacial 10, 12, 33–5, 34, East Asian monsoon system 131–2, 240,
dating methods 267, 268, 344–5; see also 44–5, 45 245, 309, 371, 374, 409
individual methods diagenesis 123–4, 179n5 East Jutland Line 31
daughter deficient (DD) U-series dating diamicton 99, 179n2 echo sounders 23–4
287 diatom analysis 197–8, 246, 306; ecological changes, early evidence 7
daughter excess (DE) U-series dating 287 applications 202–3, 204, 205–6, 205, ecological niche models 261
debris flow deposit 102, 103 206; concentrations 200, 201; diatom ecosystems 182
deep-sea sediment 2, 8; accumulation 165; range 199; diatom types 198–200, 199; Eemian 10, 126, 307–8
carbon isotopes 171–2; carbonate halobian system 202–3; human impacts eggshell, ratite, dating 338
compensation depth (CCD) 171; 205–6; identification 200; interpretation Egypt 88, 89, 131, 133
carbonate dissolution 171; correlation 202; and lake-level changes 203; and pH eigenvalues 114
365, 366, 366–7, 377–8, 377; dating 203, 205, 206; procedures 198, 200, 201, El Niño–Southern Oscillation (ENSO) 214,
287; event stratigraphy 365; fossil 202; salinity indicators 202–3, 205; 431, 435, 440
record 166; Heinrich layers 166, 166; samples 200; and sea-level change electrical conductivity measurements
ice-rafted debris (IRD) 165–6; isotopic 202–3, 202, 204; species distribution 200 (ECM) 175, 348
equilibrium 171; magnetostratigraphy diatom mats 248 electron paramagnetic resonance 293
323–4; micropalaeontology 243–8, 245, diatom stratigraphy 206 electron probe micro-analysis (EPMA)
247, 249, 250–4, 250, 251, 252, 253; differential frost heave 118 325
mixing 171; nature of 165–6, 166; differential global positioning system electron spin resonance (ESR) 293–4
oxygen isotope ratios 166–71, 168, 169, (DGPS) 91n2 electron traps 291, 294
390, 390; oxygen isotope record 9; differential receivers 22 electronic distance measures (EDMs) 20
sedimentation rates 170–1; stratigraphic digital elevation models (DEM) 25–6, 26 Ellesmere Island 39
resolution 170–1 digital images 23 Emiliani, Cesare 170
deep-sea temperature 169 digital surface models (DSM) 26 Emiliani chron 391, 391
deformation 179n3 digital terrain models (DTM) 25–6, 26 end moraines 27, 29, 29–30
deformation tills 102–5, 104, 105, 106 diluvium 99 englacial subenvironments 100, 100
deforming bed 103, 104–5 Dimlington Stadial 33 ENSO (El Niño–Southern Oscillation) 214,
deglaciation 393; Antarctica 404; Britain Dinoflagellates (dinocysts) 246, 247, 248 431, 435, 440
and Ireland 34–5; freshwater release 404; DIRTMAP 387–8 environmental change: record
Last Termination 418; North America DiTe index 265 synchronization 374–8, 375, 376, 377;
37, 39, 404; Northern Europe 32; DNA analysis 8, 262, 265, 332 time-stratigraphic framework 347; see
paraglacial deposits 105–7, 108; varve drift 6 also global environmental change
chronology 307, 308 drumlins 40, 42–3, 42 environmental simulation models (ESMs)
dendrochronology 7, 298; 2.8 ka event dump moraines 27, 28 380; box models 380–1, 381, 393; earth
433; Australia 302; cross-dating Dundalk Bay 35 system models of intermediate
299–300, 300; dendroclimatology 299, dune patterns 88–9 complexity (EMICs) 383–4, 384, 385,
302–4; Europe 301–2, 302, 303; general dunefields 86–9, 87, 88 386–6, 386, 393; general circulation
principles 298, 299; Hohenheim master duricrusts 90–1 models (GCMs) 381–3, 382, 383;
chronology 301–2, 302; isotope dust: air-transported 127, 130; annual hindcasting 388; importance of 388–9;
dendroclimatology 303–4; Lateglacial cycles 175; circulation 410–11, 410; Integrated Global System Model 386,
420; Little Ice Age 434; markers 299; and global cooling 410–11; ice-core 386; limitations of 388; Milankovitch
master chronology 299, 300; New stratigraphy 175, 178; Last Glacial cycles 393; oversimplification 388;
Zealand 302; North America 301, 301; Maximum (LGM) 131; palaeodata-model comparisons 387–8;
Pacific Decadal Oscillation (PDO) 436; Pleistocene–Holocene transition 428; testing 387; transient simulations 386–7
and radiocarbon dating calibration role of 412–13, 412; and sea-level Envisat 23
279, 280, 283; records 300; ring change 411, 411; transport fluxes 132; EPICA (European Project for Ice Coring in
measurement 298–9; ring-width indices varve chronology 309 Antarctica) 8, 172, 177–8
299, 301; tree-ring width 299; United dust bowl events 309 epochs 2
INDEX 527
equifinality 126 Flandrian 11 geochemical fingerprinting 325

equilibrium line altitude (ELA) 48–50, 57; Flint, Richard Foster 14 geochronological potential 343
palaeoclimatic inferences 50–3, 51, 52 flint artefacts 151 geochronological resolution 343, 344
eras 1, 2 floating chronologies 304, 307, 308, 317, geochronological units 2, 348, 361–2,
erathem 2 319, 420 362
Erdalen Event, the 32–3 Florida, United States of America 212 geochronology 361–2, 362
erosion rates 294 flow till 102, 103, 103 geodetic datum 20
erratics 111, 113, 113 fluted moraines 40, 42 geoidal eustasy 66–7
Esmark, Jens 5 fluvial landforms, low latitude 89–90, 90 geological ages, abbreviating conventions
estuarine sediments, and sea-level change Foraminifera 241–4, 241, 243, 246, 248, 17n1
63–4, 64 249, 250, 309 geological timescale 1, 2–3
Ethiopia 234 Forbes, Sir Edward 1, 7 geologic-climatic units 358–61, 363, 366,
Europe: 2.8 ka event 433; amino-acid forcing factors 379 378
geochronology 338; Coleoptera 221, forest history 214 geomagnetic field: components 320;
221; coversand 130, 130; deglaciation fossils and the fossil record: amino-acid polarity changes 319–20, 321–4, 322,
404; dendrochronology 301–2, 303; geochronology 339; analysis 181; 323, 324; polarity epochs 321–3, 323;
dendroclimatology 299; human biomolecular products 181; secular variations 320–1, 321
evolution 330; Little Ice Age 433–4; loess biostratigraphy 348; deep-sea sediment geomorphological evidence 19, 91;
deposits 128, 131; molluscan remains 166; formation 182; identification 182; direction of ice movement 39–43, 41,
233; phylogeography 262; textural B indices of environmental change 8; 42, 43; extent of ice cover 27–30, 28, 29;
horizon 126; Ulmus (elm) decline 194; interpretation 182–3; lake sediment 155, extent of ice cover, Britain and Ireland
vertebrate biostratigraphy 259; wind 305–6; macrofossils 181; microfossils 32, 33–5, 34; extent of ice cover, North
directions 131; Younger Dryas Stadial 181; modern analogue approach 182, America 35, 36, 37–9, 37, 39; extent of
424 183; molecular clock analysis 332; ice cover, Northern Europe 30–3, 30, 31;
European Greenland Ice-core Project 8 palaeobiological context 182; field methods 19–26, 21, 25, 26; low
European North Greenland Eemian Project preservation 182; process 256; proxy latitude landforms 82–91, 83, 84, 85, 86,
8 records 182; range 181; screening 339; in 87, 88, 89, 90; river terraces 73–82, 74,
European Project for Ice Coring in situ fossils 182; stromatolites 150; 75, 76, 78, 79, 80; sea-level change
Antarctica (EPICA) 8, 172, 177–8 taphonomy 182 58–73; see also glacial landforms
European Space Agency (ESA) 23 France: amino-acid geochronology 338; geomorphological mapping 20, 21
eustatic sea-level change 59, 59–60, 60–7, Coleoptera 224, 224; Grotte de geophones 24
60, 61, 62, 63, 64, 65, 66, 76 Clamouse 145, 148; Lac d’Annecy 306; GEOSAT 23
evaporitic varves 306 molluscan remains 233; permafrost geosols 355
event alignment 367 boundary 119; Sevrier-Les-Charretières Germany 32, 56, 68, 119, 120, 161, 190,
event stratigraphy 348, 365 159; Vosges Mountains 212 207, 212–13, 285, 301, 302, 309, 376,
exemplary stratotypes 349 freeze-thaw activity 115, 121 424, 433
exhumed palaeosols 123 freshwater hosing experiments 380, 424 GICC05 see Greenland Ice-core
Eyjafjallajökull 327 freshwater release 404, 422, 423–4, 431 Chronology
friction cracks 40 Gilbert, Grove Karl 6–7
facies 100, 102, 109–11, 111, 112, 351 frost action 27, 29 GISP2 see Greenland Ice Sheet Project
facies logs 109–10, 111, 112 frost-shattering 54 Glacial Epoch 1
far field locations 62–3 fungal remains 264 glacial ice storage, and marine oxygen
faunal evidence 5 isotope ratios 167–9, 168
feedback loops: climate change 15–16; gas, trapped 312 glacial landforms: distribution 27; and
global environmental change 381, 384, GCMs (general circulation models) 8–9, extent of ice cover 27, 28, 29;
392, 395–6, 395, 396, 396–7, 396, 404; 381–3, 382, 383 geomorphological evidence 26–7;
Holocene 435; interdependency 412–13, Geer, Gerard de 7, 304–5, 307 modification 27, 29; preservation 27
413; Last Termination 417–18; Geikie, Archibald 99 Glacial North Atlantic Intermediate Water
teleconnection 407, 409, 410–11 Geikie, James 6, 27 (GNAIW) 409
fidelity 232 Gelasian 5 glacial sediments 99; dating 292, 294;
field mapping 19–20, 21, 40 gelifluction 54, 115 deformation 102; deformation tills
Finland 32, 307, 322 general circulation models (GCMs) 8–9, 102–5, 104, 105, 106; erratics 111, 113,
Finse Event, the 32–3 381–3, 382, 383 113; extent 99; facies 100, 102, 109–11,
fire histories 214–15 genetic facies 102 111, 112, 351; factors affecting 100, 100,
fission track dating 289–91, 325, 327 genetic modification 262 102; flow till 102, 103, 103; glaciofluvial
flame photometer 98 GENIE initiative 386 99–100; glaciolacustrine deposits 100;
528 INDEX
and ice thermal regime 107–9; ice- global circulation models 251 Greenland Ice-Core Chronology (GICC05)
contact deposits 99; ice-directional global circulation processes 385 312–13, 313, 314, 355, 326, 375, 406,
indicators 111, 113–15, 113, 114; global conveyor 401, 402 419, 420, 421, 423
lithofacies interpretations 109–11, 111; global environmental change 379–80, 408, Greenland ice sheet 73, 160; 8.2 ka event
lodgement till 102, 103, 104; 440, 441, 442–3, 442; amplification 429; annual layers 312; anti-phase
luminescence dating 292; melt-out till agents 392–4; the Anthropocene behaviour 406–7; Chironomidae 227,
102, 103, 104; micromorphology 103–4; 439–40; anthropogenic signature 228; coring 8; correlation 371, 371, 372,
modes of deposition 101, 102; 436–40, 437, 438; astronomically driven 373–4, 373, 374, 377; crustal flexuring
nomenclature 99; outwash deposits 389; bipolar teleconnection 405–7, 405, 68; dating 268; dust record 132, 410,
99–100; paraglacial deposits 105–7, 108; 406; boundary conditions 381–2, 383–4, 410; ice core chronology 312–13, 313,
particle size and shape analysis 109, 110, 384, 387; cause-and-effect pathways 314; ice-accumulation years 174; ice-
112; plastering on process 103; involved 388; complexity 384; dynamic core stratigraphy 172, 173, 176–7, 176,
proglacial 99–100; sequence analysis conditions 382; environmental 178, 399, 405–6, 405, 419–20, 422; ice-
109–11, 110, 111; stratified 99; simulation models (ESMs) 380–9; sheet modelling 45–6, 46–7; isotope
subglacial traction till 104; sublimation feedback loops 381, 384, 392, 395–6, ratios 176, 176; Last Termination 414,
till 102, 103; till classifications 102–7, 395, 396, 396–7, 396, 397–8, 404; 415–16; Lateglacial 419–20; Lateglacial
103, 104, 105, 106, 107; till fabric glacial–interglacial cycles of the last 800 climate change 422, 423–4; methane
113–15, 114; till matrix properties 115; ka 394–7, 395, 396, 397, 398, 399; global signal 406, 406; oxygen isotope ratios
till provenance 113; till sequences 109; teleconnections 407–13, 408, 410, 411, 427; Pleistocene–Holocene transition
unstratified 99 412; the Great Acceleration 440; 428; principal sites 172, 173; retreat 39,
glacial stages 1, 9, 10 greenhouse effect 437–8, 437, 438; 203, 204; teleconnection 405, 405–7,
Glacial Theory 5–6, 26 Holocene. see Holocene; human 406; temperature change 177
glacial troughs 41–2 impacts 436–40, 437, 438, 442–3; Greenland Ice-core Project (GRIP) 8, 12,
glacial–interglacial cycles 2, 359–60, 360; interdependency 412–13, 412; lags 380; 12, 172, 176–7, 176, 312–13, 314, 349,
last 800 ka 394–7, 395, 396, 397, 398, Last Termination 380, 413–19, 414, 415, 371, 371, 375, 377, 401, 415, 423, 433
399 416, 417; Lateglacial 419–26, 421, 423, Greenland Ice Sheet Project (GISP2) 8,
glacially moulded/sculpted terrain 41, 42 424, 425, 426; leads 380; Middle 146, 148, 172, 177, 309, 312, 373, 373,
glaciations: correlation 359, 363; early Pleistocene Transition (MPT) 389, 374, 411, 411
mapping 6; maximum extent 6; outwash 390–4, 390, 391, 394, 398–9; Greenland Interstadials (GI) 12, 12, 371
terraces 77; and river terraces 77 Milankovitch cycles 380, 389–99, 390, Grimsel Pass, Switzerland 29
glacier ice: annual layers 298, 310–13, 313, 391, 394, 395, 396, 397, 398; ocean– GRIP see Greenland Ice-core Project
314, 314, 315; dating 298, 311–12; ice interactions, North Atlantic 400–1, Gros Morne Mountain, Newfoundland 54,
thermal regime 107–9 401, 402, 403–5, 403; proxy records 55
glaciers 48; ablation area 48; accumulation 379–80; sub-Milankovitch timescales Grotte de Clamouse, France 145, 148
area 48; accumulation area ratio (AAR) 380; sub-orbital (millennial) timescales ground ice 54
48–50; cirque floor altitude (CFA) 49, 399–413, 400, 401, 402, 403, 405, 406, ground moraine 99
50; equilibrium line 48, 48–50, 57; firn 408, 410, 411, 412; timescale 440; ground penetrating radar (GPR) 24
lines 48, 48–50, 49; ice surface contours tipping points 386, 404, 438, 440; ground wedges 117
47–8, 48; lapse-rate calculations 50; transitions 387; variability 440, 441, 442 grounding lines 47
palaeoclimatic inferences from ELA Global Lake Status Data Base 135, 136, Gulf Stream 403, 406–7
50–3, 51, 52; as palaeoenvironmental 388 gypcrete 91
archives 173; readvance 30; Global Positioning System (GPS). 20, 22
reconstruction 47–50, 49; retreat, rate of global stratotypes 349 hard-water error 277–8
27; surges 30, 91n5; toe-to-headwall global tephra database 327 heat piracy 407
(THAR) 50; transition zones 92n9 Grand Canyon, Arizona 149 heavy mineral assemblages 98
glacigenic sequences 100, 109–11, 110, 111 Gravity Recovery and Climate Experiment Heer, Oswald 7
glacigenic varves 304, 304–5 (GRACE) satellite 47 Heinrich events 62, 105, 138, 246, 365, 371,
glacimarine deposits 100 Great Acceleration, the 440 371, 401, 436; dating 373; deep-sea
Glacio-eustatic Theory 6 Great Barrier Reef 242 sediment 166, 166; Lateglacial 423;
glacio-eustatic sea-level change 60–1 Great Ice Age 27 North Atlantic 372, 373, 374, 404, 411;
glacio-isostasy 6, 68–9, 69–73, 70, 71, 72 Greatlakean Advance 35, 37 proxy records 373
glacitectonite 104 greenhouse effect 397, 437–8, 437, 438 Heinrich layers 166, 166
Glen Roy, Scotland 75 greenhouse gases 16, 73, 395–6, 409, 437–8, Hölloch Cave, Bavarian Alps 146, 147
Global Boundary Stratotype Sections and 438, 442 Holocene 2–3, 10, 32, 33, 39, 413, 427;
Points (GSSPs) 4, 349 Greenland: lichen growth curve 315, 315; 2.8 ka event 433; 4.2 ka event 431,
global carbon reservoirs 270, 276 ostracod analysis 240 432, 433; 8.2 ka event 429, 430, 431;
INDEX 529
anthropogenic signature 427; Atlantic hydrographs 84, 86 Greenland ice sheet 172, 173, 176–7,
Multidecadal Oscillation (AMO) 436; hydro-isostasy 60, 68, 69 176, 178, 399, 405–6, 405, 419–20, 422;
atmospheric circulation regime 428; hydrophones 24 human impacts 178; isotope ratios 175,
climate change episodes 427; climate hydroseral successions 153, 154, 155 175–6, 176; Lateglacial 419; the
trends 427, 427; climatic cycles 434–6; palaeoenvironmental archive 173;
definition 427; dunefields 87; El Iberian Penisula 219 palaeoenvironmental reconstruction
Niño–Southern Oscillation (ENSO) ice, dielectrical properties (DEP) 175 175–8, 176; proxy records 176; statistical
431, 435; eustatic sea-level change 65–6; ice cover, extent of 27–30, 28, 29, 342; uncertainties 177; and temperature
feedback loops 435; global Britain and Ireland 32, 33–5, 34; change 177; tephra 327; trace substances
environmental change 380; human Foraminifera and 243; Last Glacial 175, 178; upper levels 8
impacts 436–40, 437, 438; ice- Maximum (LGM) 33; North America Iceland 326, 327, 329
accumulation years 174; lake-level 35, 36, 37–9, 37, 39; Northern Europe Icelandic Sea 277, 330
changes 138, 160; Little Ice Age 433–4, 30–3, 30, 31; readvances, Britain and ice-moulded (streamlined) bedrock 40–2,
433; methane levels 431; North Atlantic Ireland 35; readvances, North America 42
Oscillation (NAO) 435–6; onset 413, 35, 37, 38, 38–9 ice-rafted debris (IRD) 165–6, 327, 373,
414; Pacific Decadal Oscillation (PDO) ice divides 46 404
436; Pleistocene–Holocene transition ice lobes 37, 44, 45, 46 ICESat 23
428–9, 428, 429; proxy records 427; ice masses, as palaeoenvironmental ice-sheet modelling: and ice divides 46;
sea-level change 66, 69; solar output archives 173 inverse models 44–5, 45; numerical
variations 431, 433, 433, 434, 434–5; ice masses, reconstruction: flowlines 43, 43; simulations 44–7, 44; parameters 43–4,
speleothem growth 146; sub-epochs glaciers 47–50, 49; ice caps 47, 48; ice- 44; present Antarctic and Greenland ice
431; temperature anomaly record 398; sheet modelling 43–7, 44, 45, 46, 47; sheets 46–7; steady-state models 44; and
temperatures 427, 427, 434 palaeoclimatic inferences 50–3, 51, 52; wind patterns 45–6
Holocene Thermal Maximum 427 parameters 43; steady-state models 44 ice-wedges 116–17, 116, 118–19, 122
homotaxial errors 363 ice movement, direction 39–40; crescentic illuviation 109
Hudson Bay, Canada 63 fractures 42; crossing striations 40; imaging radar 24
Hulu Cave, China 146, 148 flowlines 43; friction cracks 40; glacial incremental dating methods 267;
human evolution, and volcanic activity sediments and 113–15, 113, 114; ice biostratigraphy 331–2;
330 streams 40, 41; ice-moulded dendrochronology 298, 298–304, 299,
human impacts: agriculture 437; the (streamlined) bedrock 40–2, 42; macro- 300, 301, 302, 303; glacier ice annual
Anthropocene 439–40; anthropogenic scale patterns 43, 43; streamlined glacial layers 298, 310–13, 313, 314, 314, 315;
indicators 196, 197; anthropogenic deposits 42–3; striations 40, 41, 42 lichenometry 298, 315–16, 315;
warming 437–9, 437, 438; Coleoptera ice sheets 47; annual ice increments 173–5, molecular clock analysis 332, 345;
225; diatom analysis 205–6; fire histories 174; annual layering 311; collapse 404, oxygen isotope chronology 330–1, 331;
214; global environmental change 422; freshwater release 404, 422, 423–4, sclerochronology 298, 316, 317, 318,
442–3; the Great Acceleration 440; 431; steady-state 29, 44, 49, 91n4 319; speleothem 298, 316–17; see also
Holocene 427, 436–40, 437, 438; ice- ice streams 40, 44, 46 varve chronology
core stratigraphy 178; industrial activity ice volume 27; environmental simulation India 62, 89, 90
178; lake sediment 158–9; molluscan models (ESMs) 393; global 26–7; last indicator erratics 111
remains, non-marine 234–5; on peat cold stage 9; Middle Pleistocene Indonesian Islands 73
deposits 165; plant macrofossils 212, Transition (MPT) 392; and oxygen inductively coupled plasma atomic
215; and pollen analysis 183, 195–6, 196, isotope ratios 169; emission spectrometry (ICP-AES) 98
197; and population numbers 442, 442; Pleistocene–Holocene transition 429; industrial activity 8, 178, 437, 437–8, 438
on radiocarbon dating 276; and river and sea-level change 169–70 Industrial Revolution 438, 439
terraces 77–8, 78; synanthropic species ice-accumulation years 174 infrared stimulated luminescence 292
225; temperatures 442; time- ice-contact deposits 99 Innuitian ice sheet, North America 35, 36,
transgression 440; tipping point 438, ice-core stratigraphy 8, 172–3; 4.2 ka event 39
440; varve chronology 309; vertebrate 431; 8.2 ka event 431; acidity profiles inorganic deposits 93
remains 261–2 178; analysis 173–5, 174; annual insect remains, fossil 215; analysis 218–25,
human occupation sites 258 increments 173–5, 174; Antarctica 8, 220, 221, 223, 224; archaeoentomology
human population numbers 442, 442 172, 173, 177–8, 394–5, 395, 396, 397, 225; Chironomidae 215, 225–8, 226,
human skeletal remains 256–7 399, 405–6, 405; borehole thermometry 227; Coleoptera 215, 215–16, 216, 217,
hummocky moraines 27, 28 177; chemical content 175; chronology 218–25, 218, 220, 221, 223, 224;
Huon Peninsula, Papua New Guinea 67, 67 errors 312; correlation 366–7, 377; distribution 221–2, 221; evolutionary
Hutton, James 5 dating 175, 297, 313; dust content 175, stasis 218–19; human impacts 225;
hydraulic sorting 258 178; glacier ice annual layers 311, 311; identification 216; isotope ratios 224;
530 INDEX
laboratory methods 216, 218, 218; sediment 160–1; micropalaeontology, evidence from lake sediments 154–61,
palaeoenvironmental reconstruction deep-sea sediment 253–4; molluscan 156, 157, 158, 159; palaeoenvironmental
and 221–4, 221, 223, 224; types 215 remains 234, 237; peat deposits 163, evidence from mire and bog sediments
instrumental levelling 20, 22 165; speleothem 145–8, 147, 148 161–5, 162, 163, 164; palaeolimnology
Inter-Allerød Cold Phase: IACP 413, 422 isotopes 17n5, 269 155; and palaeotemperatures 160–1;
interdisciplinary approach 2, 9 isotopic enrichment 271–2 peat formation 153–4; plant
interferometric synthetic aperture radar isotopic equilibrium 145, 171 macrofossils 210, 211; pollen content
(IFSAR) 25 isotopic fractionation 145, 167, 179n9, 155, 157, 192; pollen diagrams 185, 186,
interglacials 1; definition 9; duration 271–2, 276, 277 188–9, 188; radiocarbon dating 278,
307–8; number 10 isotopic records 8 279; sedimentation patterns 154–5;
Intergovernmental Panel on Climate isotopic signals, oxygen isotope record 9, sedimentation rates 157–8, 158;
Change (IPCC) 388, 437, 443 10 sequences 151, 151, 155, 155–6, 157,
International Atomic Energy Authority isotopic stages 9–10, 10, 169, 330, 358 159; shoreline correlation 365; shoreline
(IAEA) 167, 272 Israel 67, 131 features 150; stratification 152;
International Geological Congress: 1885 3; Italy 78, 149, 326, 327 vegetation patterns 158–9; vertebrate
1948 4 remains 258; see also pluvial lakes
International Union of Geological Sciences James Ross Island, Antarctica 28 Lake Baikal 369, 369
(IUGS) 3, 17n3 Jamieson, Thomas 6 Lake Bonneville 82, 83–6, 84, 86, 133,
International Union of Quaternary Japan 283, 309–10, 310, 365, 422, 423 339
Research (INQUA) 3, 17n4; Congress, Jason-1 23 Lake Chad 82, 85, 89
1982 4 jet stream 137 Lake Kinneret 133
interpluvial episodes 6–7, 82 jökulhlaups 77 Lake Lahontan 82, 83, 84
interstadial episodes, definition 9 Jutland 31, 31 Lake Lisan 133
Inter-Tropical Convergence Zone (ITCZ) Lake Malawi 135
246, 407, 409–10, 418, 431 kame terraces 27 Lake Neuchatel 215
INTIMATE (INTegration of Ice-core, Keewatin, Canada 41 Lake Tana 135
MArine and TErrestrial records) 375–6, Kennard, Alfred 7, 228 Lake Torreberga 160
419, 421 Kenya 133 lake-level changes 132–3, 154–5, 409; and
involution structures 117–18, 117, 118 Killard Point Readvance, Ireland 35 atmospheric circulation patterns 160;
ionization 269 Kråkenes Lake, Norway 424–6, 425, 426 causes 134–5, 137–8, 140;
IP25 index 265 Kullenberg, Börje 8 Chironomidae 227; and climatic change
Ipswichian 10 135, 159–60, 159; dating 135; diatoms
Ireland 33, 35, 43, 72–3 laboratory methods 96–9 and 203; geochemical variations 134,
Irish Sea 33, 34, 35; ice wastage record Lac d’Annecy, France 306 134, 135; Global Lake Status Data Base
110–11 lake, mire and bog sediments: 135, 136; hydrological balance 134;
island rule, the 260 accumulation 151; allochthonous 151; isotope variations 138, 140; Last Glacial
Islay, Hebrides 58 autochthonous 151; biogeochemistry Maximum (LGM) 135, 137; ostracod
isobases 71, 71 279; bog formation 155; bog types 154; analysis 240; palaeoclimatic
isolation basins 64, 65 chemical composition 155–7, 157; and reconstruction 135, 137–9, 140; plant
isopollen maps 195, 195, 332 climatic change 155–61, 156, 157, 158, macrofossils and 211; pluvial lake
isostatic equilibrium 59 159; coring 151; dating 287, 297; sequences 133–5, 133, 134; proxy
isostatic movements 59 density/thermal stratification 152; records 138, 139; reconstruction data
isostatic recovery 69, 69, 71, 72–3 deposits 151–2; eutrophic 151; fossil 159–60; sedimentary sequences 133–4;
isostatic sea-level change 66, 68–73, 68, 69, record 155, 305–6; human impacts status data 136; and the sunspot cycle
70, 71, 72 158–9; importance 150–1; insect 138; water balance models 135
isothermal plateau fission track technique remains 215; isotope ratios 160–1; lake lakes: age 150–1; catchment histories
(ITPFT) 290 age 150–1; and lake chemistry 152; lake 155–6; classification 150; closed-basin
isothermal remanent magnetization (IRM) classification 150, 150; Lateglacial 425, 203; density structure 152; high-altitude
324 425; limnic accumulation 151, 151; 152; pH 205; thermocline 152; trophic
isotope dendroclimatology 303–4 magnetic data 158–9; Milankovitch status 153, 205; water temperature 152;
isotope ratio mass spectrometry 273–4 cycle correlation 368–9, 369, 370, 371; see also lake, mire and bog sediments;
isotope ratios 98–9; Antarctica 411, 411; mire formation 153, 155; mire types lake-level changes; pluvial lakes
climatic signal 145; Coleoptera 224; 154; molluscan remains 232; nature of land bridges 73
deep-sea sediment 166–72, 168, 169; 151–4, 154; oligotrophic 151; organic Landsat 22–3
dendrochronology 303–4; ice-core varves 305–6; and oxygenation of lake land-use changes 309
stratigraphy 175, 175–6, 176; lake waters 152; palaeoenvironmental Lapland 300, 302, 303
INDEX 531
laser altimetry 26 Loch Lomond Stadial 12, 35, 40, 155, 156; marine palaeoclimatology 248, 249, 250–3,
laser diffraction 96–7 ELAs 52, 52; palaeoclimatic inferences 250, 251, 252, 253
laser technology 22 51–2, 51, 52; pronival (protalus) marine palaeoproductivity 253–4
Last Glacial Maximum (LGM) 33, 36, 260, ramparts 57, 57; reconstruction 47–8, marine regressions 64, 64
383–4, 413; dust flux 131; forest refugia 48, 50 marine reservoir age 276–7
214; and ice volume 169; ice-sheet lodgement till 102, 103, 104 marine reservoir errors 276, 329–30
modelling 45–6, 47; lake-levels 135, 137; loess: cooling onset evidence 5; correlation marine snow 248
marine oxygen isotope ratios 167–8; 374; extent 127; fabric 127; magnetic Marine Species Identification Portal 242
megadroughts 86; multi-proxy susceptibility 325; Milankovitch cycle marine terraces 62
investigation 266; onset 415; River correlation 369, 370; particle size 127, Marquette Advance, North America 37
Thames 81; sea-level change 62–3; 131–2, 132; sequences 122, 127, 128, mass balance 43–4
sea-surface temperatures (SSTs) 251, 129; sources 128; stratigraphy 127–8, mass extinctions 261–2
252; temperatures 148, 339 128, 129, 130, 130, 355; thickness 127 mass flow diamicton 102, 103
Last Glacial–Interglacial Transition 413 low latitude landforms 82; dunefields 86–9, mass flow tills 114–15
Last Termination 413; bipolar seesaw 87, 88; fluvial landforms 89–90, 90; mass spectrometric analyses 167
418–19; carbon storage and release pluvial lakes 82–6, 83, 84, 85, 86; Massachusetts Institute of Technology 386,
416–17; definition 413–15, 414, 415; weathering crusts 90–1 386
feedback loops 417–18; global Lozinski, Walery von 7, 55 maximum counting error (MCE) 312
environmental change 380; global luminescence dating 61–2, 291–3, 291, 327 mean annual air temperature (MAAT) 53,
temperature records 417–18, 417; Lyell, Charles 5, 6, 99 118–19, 119, 121, 121, 122
isotope records 414–15, 415; onset Lytton, British Columbia 56 mean annual ground temperature (MAGT)
415–18, 416, 417, 418; record 122
synchronization 375; sea-surface Mackereth, F.J.H. 155, 156–7 Mediterranean 62, 69, 77–8, 138, 149, 237,
temperatures (SSTs) 415, 416; sequence MacLaren, Charles 6 240, 242, 243, 277, 411
417–18, 418; teleconnections 418–19; McMurdo Ice Shelf 115 Mediterranean precession-related
terminology 413–14; trigger 415–17 macrofossils 181 sapropels 377–8
Late Weichselian ice sheet 30–1, 30–3, 30, magnetic field variation 283 megadroughts 86
31, 73, 117 magnetic susceptibility 324, 325 megafaunal extinctions 261–2
Late Wisconsinan events, correlation 364 magnetostratigraphy 320–4, 321, 322, 324, mega-grooved terrain 41, 42
Lateglacial Interstadial 12, 413, 419–26, 325, 348, 363–4 melt-out till 102, 103, 104
421, 422, 423, 423, 424, 425, 426 magnetozones 363–4 metallic ions, sediments 98
laterites 90–1 Main Stationary Line 31 methane 15, 15, 405–6, 406, 431
Latvia 209 mammoths 254, 255, 258, 260, 260 Methuselah Tree 301, 301
Laurentide ice sheet, North America 35, MARGO 253, 388 Mexico, Gulf of 68, 69
36, 37–8, 37, 137, 384, 385; 8.2 ka event marine carbon circulation 276–7 Micromorphological analysis 97
431; collapse 404; discharges 105; ice- marine deposits: dust record 132; and sea- microfloral assemblages, deep-sea
sheet modelling 44, 44, 45, 46; retreat level change 58 sediment: chemical tracers 254;
243 marine molluscan remains 235–8, 235, Coccolithophores 244, 245, 246, 248,
lead-210 296, 297 236, 237 250; Dinoflagellates (dinocysts) 246,
Leszno moraine 31 marine oozes 166, 248 247, 248; exhumation and reburial 248;
Levene Moraine 32 marine oxygen isotope record 4, 355–7, Foraminifera 246, 248, 249, 250;
Libby, Willard 7, 270 357, 358, 400; boundaries 356; identifications 248; isotope ratios 253–4;
lichen growth curve 315, 315 correlation 366, 399; dating 268; deep- laboratory methods 248; and marine
lichenometry 298, 315–16, 315 sea sediment 166–71, 168, 169; and palaeoclimatology 248, 249, 250–3, 250,
light detection and ranging (LIDAR) 26 glacial ice storage 167–9, 168; isotopic 251, 252, 253; planktonic foram tests
lithalsas 56 stages 356–7, 357; Last Glacial 252–3, 253; Radiolaria 244, 244–5, 245,
lithofacies 102, 112 Maximum (LGM) 167–8; marker 248
lithological changes 93 horizons 366; Middle Pleistocene micromorphology 97, 103–4, 105, 106
lithostratigraphy 348, 349–53, 350, 351, Transition (MPT) 390, 390, 391; micropalaeontology, deep-sea sediment
352, 378 Milankovitch cycles 367–8, 368, 369, 244; chemical tracers 254;
Little Ice Age 33, 39, 78, 243, 309, 433–4, 369, 389; and sea-level change 169–70, Coccolithophores 245–6, 245, 246, 248,
433 169; and sea-surface temperatures 170; 250; Dinoflagellates (dinocysts) 246,
load structures 117–18, 117, 118, 179n4 signal noise 366; stage boundaries 366; 247, 248; Foraminifera 246, 248, 249,
local vegetation reconstructions 194 terminations 357, 358, 366, 413; time- 250; identifications 248; isotope ratios
locations, pinpointing 20, 22 transgression 366; tuning 367; warm 253–4; laboratory methods 248; and
Loch Lomond Readvance 35, 47 phases 356, 357 marine palaeoclimatology 248, 249,
532 INDEX
250–3, 250, 251, 252, 253; planktonic monolith tins 94, 94 the Cockburn Event 38; Coleoptera 219,
foram tests 252–3, 253; Radiolaria monsoon belts 409–10 220, 223; Cordilleran ice sheet 35, 36,
244–5, 245, 248 monsoonal circulation 15–16 38–9, 68, 73; coversand 130;
microscopy 248 Monte San Nicola section 4, 5, 10 deglaciation 37, 39, 307, 404;
micro-vertebrates 257 moraines 27, 28, 29, 29–30, 40, 42, 99; dendrochronology 301, 301; dunefields
Middle Pleistocene Transition (MPT) 15, dating 296, 297, 316, 342; Late 87; ELAs 52; extent of ice cover 35, 36,
389, 390–4, 390, 391, 394, 398–9 Weichselian ice sheet 30–1, 31; 37–9, 37, 39; fission track dating 290–1;
Mid-Pleistocene Revolution 15, 390 morphostratigraphy 354–5, 354 Foothills Erratics Train 113; fungal
Milankovitch chron, the 390–1, 393 morainic rock glaciers 55 remains 264; Greatlakean Advance 35,
Milankovitch cycles 15, 78, 81, 167, 169, morphological mapping 19–20 37; ice-sheet modelling 45; Innuitian ice
262–3, 330, 367, 379, 440; the 100 kyr morphostratigraphic units 354 sheet 35, 36, 39; lake sedimentation rates
problem 389, 392, 393, 394, 398–9; morphostratigraphy 348, 354–5, 354, 378 158; lake-level changes 135, 137; Last
astronomical forcing 390–2, 393; Mt Edziza, British Columbia 55 Glacial Maximum (LGM) 36; Lateglacial
astronomically driven 389, 398; climate Mt St Helens, United States of America climate change 422; Little Ice Age 434;
change over 389–99, 390, 391, 394, 395, 327, 328 loess deposits 128, 130; Marquette
396, 397, 398; correlation 367–9, 368, multidisciplinary approach 2, 9 Advance 37; mass extinctions 261;
369, 370, 371, 378; deaf zones 390–2; multi-parameter approach 174, 311 molluscan remains 233, 234; ostracod
environmental simulation models multi-proxy palaeoecological studies 265–6 analysis 240; palaeoclimatic inferences
(ESMs) 393; glacial–interglacial cycles multispectral (multi-wavelength) scanning 52; palaeosol correlation 364;
of the last 800 ka 394–7, 395, 396, 397, systems 22, 23 palaeovegetation maps 195, 195;
398, 399; marine oxygen isotope record Munsell colour charts 93 permafrost 118; phylogeography 262;
389; Middle Pleistocene Transition mutual climatic range (MCR) 212–13, pluvial lakes 82, 83–6, 83, 84, 86, 133;
(MPT) 389, 390–4, 390, 391, 394, 398–9; 222–4, 223, 224, 234, 261 Port Huron Advance 35; readvances 35,
precessional cycles 393; rhythms 389, mutual temperature range 240 37, 38, 38–9; record synchronization
392–3; signal clarity 389 375; retreat pattern 37; soil stratigraphy
Milankovitch hypothesis, the 14–15, 178 National Aeronautics and Space 355; Tahoe Moraines 38, 39;
mineral carbon error 277–8 Administration (NASA) 22–3 Tiedemann–Peyto Advance 39; Tsuga
mineral magnetic analysis 98, 125, 126 National Climatic Data Center 438 (hemlock) decline 194; vertebrate
mineral magnetic susceptibility 324, 325 National Oceanic and Atmospheric biostratigraphy 259; vertebrate remains
mineral palsas 56 Administration 22 261; Younger Dryas Stadial 424; see also
mires: see lake, mire and bog sediments natural luminescence signal 291 Laurentide ice sheet, North America
MIS (marine isotope stage) 9–10, 11 natural remanent magnetism (NRM) North Atlantic 250, 379; 2.8 ka event 433;
molecular clock analysis 8, 332, 345 320 8.2 ka event 431; circulation system
molluscan remains: absolute abundance NEEM ice core 172, 313 400–1, 402, 403–4, 403, 406–7, 408–9,
231, 231; amino-acid geochronology negative sea-level tendency 64, 64 417, 431; CO2 content 408–9, 408; and
336, 338, 338, 339; archaeological Neogene 1, 3 freshwater release 404, 422, 423–4, 431;
relevance 234–5; assemblage zones 230, Neoglacial period 33 Heinrich events 372, 373, 374, 404, 411;
231; and biostratigraphic correlation Neolithic 73 ice-rafted debris (IRD) 404; Last
233, 237; climate change boundaries Netherlands 56, 119, 160, 368, 368 Termination 417; Lateglacial 422; Little
360, 361; dating 277, 278, 289, 294, 339; New England Varve Chronology (NEVC) Ice Age 434; magnetostratigraphy 323;
distribution 229, 234, 235; fidelity 232; 307 marine molluscs 235, 236, 236, 319;
field work 229; growth banding 319; New Guinea 61, 73, 169, 169, 240 ocean–ice interactions 400–1, 401, 402,
habitat groups, non-marine 232–3, 232; New Zealand 77, 145, 237, 283, 302, 329, 403–5, 403; sea-surface temperatures
identification 230–1; isotope ratios 234, 375–6, 434 (SSTs) 251–2, 252; tephra marker
237; laboratory methods 229–31; marine NEXTMap 25–6, 26 horizons 329; thermal stratification 401,
235–8, 235, 236, 237, 319; marker bands NGRIP ice core see North Greenland Ice 402, 403; thermohaline circulation:
319; mutual climatic range (MCR) 234; Core Project THC 401, 402, 436
nature of 229; non-marine 227–35, 228, nivation ridges 57 North Atlantic Ash Zone 326
230, 231, 232; palaeoenvironmental niveo-aeolian deposition 130 North Atlantic Deep Water (NADW)
reconstruction 234, 236–7, 237; in nominal stratotypes 349 current 401, 402, 403, 404, 408–9, 408,
palaeosols 122; relative abundance 230, non-marine molluscan remains 227–35, 422, 424, 431, 434
231; species bias 231–2; species diversity, 228, 230, 231, 232, 234 North Atlantic Event Stratigraphy 375
non-marine 233; taphonomy, marine Norden Model 419 North Atlantic Oscillation (NAO) 303, 309,
236; taphonomy, non-marine 231–2; Nordic Sea 251–2, 251 404, 435–6
types 229 North America: 2.8 ka event 433; 4.2 ka North Atlantic Polar Front 236, 243, 250,
Mono Lake, United States of America 150 event 431; arrivals of humans in 436; 266n6
INDEX 533
North Cascade Range, United States of mass spectrometric analyses 167; palaeothermometry 339
America 210, 214 micropalaeontology, deep-sea sediment palaeovegetation maps 195
North Greenland Ice Core Project 253–4; Middle Pleistocene Transition palsas 56–7
(NGRIP) 8, 172, 174, 176, 176, 177, 227, (MPT) 390, 390, 391; the Milankovitch palynology 183
311, 312–14, 313, 314, 315, 326, 349, hypothesis 15; profiles 168–9, 168; sea- paraglacial deposits 105–7, 108
375, 377, 399, 405, 406, 413–15, 415, level change 60–1, 61; and sea-level paraglacial landscape 106, 108
421, 425, 426, 428, 428, 429, 429, 430, change 169–70, 169; seasonal variations paraglacial period 106–7, 116
442 174; and sea-surface temperatures 170; particle settling velocities 97
North Icelandic Shelf 243 signal 9, 10; speleothem 373–4; particle shape: glacial sediments 109;
North Sea 33, 319 standards 167, 179n10; stratigraphic sediments 97
Northern Europe: deglaciation 32–3; application 170; terminations 357, 358; particle size: distribution 96–7, 97; glacial
extent of ice cover 30–3, 30, 31; ice-sheet trace timescale 1; tuning 367 sediments 109, 110, 112; loess 127;
modelling 44–5, 45; indicator erratics oxygen isotope stratigraphy 170, 348, wind-blown sediments 131–2, 132
113; mean annual temperatures 119–20, 355–7, 357, 358 peat deposits 151; 2.8 ka event 433;
120; permafrost limits 119, 121, 121 biomarkers 265; chronologies 161;
Norway 32, 51, 52, 73, 118, 143, 145, 224, Pacific Decadal Oscillation (PDO) 303, dating 161–2, 289; down-profile
224, 236–7, 315, 424–6, 425, 426, 436 436 variations 162; formation 153–4;
Norwegian Sea 277 Pakefield, Suffolk 151 formation rates 161; human impacts on
Nova Scotia 165 palaeocatenas 124 165; isotope ratios 163, 165;
nuclear bomb testing 440 palaeooceanography 265 palaeoenvironmental evidence from
palaeoclimatic reconstruction 358–9, 360; mire and bog sediments 161, 161–3,
obsidian diffusion dating 340 and ELA 50–3, 51, 52; lake-level changes 162, 163; plant macrofossils 162, 213;
obsidian hydration dating (OHD) 340 135, 137–9, 140 pollen content 161, 165; profiles 162;
ocean circulation 254, 401, 402 palaeodata-model comparisons 387–8 radiocarbon dating 161–2; radiocarbon
ocean siphoning 69 palaeo-drainage channels 90, 90, 131 wiggle-matching 282, 283; recurrence
ocean–ice interactions, North Atlantic palaeoecology 7, 181, 265–6 surfaces 161; Sphagnum 161, 162; tephra
400–1, 401, 402, 403–5, 403 palaeoenvironmental reconstruction 327; testate amoebae analysis 162–3
oceans, salt-density variations 16 360–1, 361; Coleoptera and 221–4, 221, pedogenesis 342–3, 355
Okotoks erratic, The Big Rock 111, 113, 223, 224; dendroclimatology 299, pedogenic carbonate 150
113 302–4; Foraminifera 241, 242; glaciers Penck, Albrecht 6, 13
Older Dryas (OD) cold oscillation 413, and 50–3, 51, 52; ice-core stratigraphy Pennington, Winifred 155
422 175–8, 176; and lake sediments 154–61, periglacial domain, the 53–4
optically stimulated luminescence 291 156, 157, 158, 159; marine 248, 249, periglacial landforms 53–4, 53; air
orbital forcing 15, 138, 390–2 250–3, 250, 251, 252, 253; and mire and temperature thresholds 119; features 54,
orbital tuning 4, 330–1, 331, 366 bog sediments 161–5, 162, 163, 164; 55; palaeoclimatic inferences 55–8, 56,
Ordnance Survey Bench Marks 20 molluscan remains, marine 236–7, 237; 57, 58; patterned ground 54–5, 55
organic varves 152, 304, 305–6, 305 mutual climatic range (MCR) method periglacial loading 117–18
Orkney Islands 33 222–4, 223, 224; ostracod analysis 240; periglacial sediments 115–22, 116, 117,
Oro Lake, Canada 203, 205 palaeosols and 125–7; periglacial 118, 119, 120, 121
orthorectification 26 structures and 55–8, 56, 57, 58, 118–22, perlite 340
ostracod analysis 238–41, 239 119, 120, 121; plant macrofossils permafrost 53, 53, 56; climate 118–21;
outwash terraces 77 212–14, 213; pollen data 197, 212; range continuous 116, 119; discontinuous 116,
overkill hypothesis, the 262 overlap method 222; wind-blown 119–20, 121; extent 118, 119, 121, 121;
Owen, Richard 7 sediments and 131–2, 132 and MAAT 118–19, 121, 121;
OxCal 281, 282 palaeoenvironmental synthesis 2 palaeoclimatic inferences 118–22, 119,
oxygen isotope chronology 330–1, 331 Palaeogene 1, 3 120, 121; periglacial sediments 115–22,
oxygen isotope chronostratigraphy 319 palaeo-ice streams 41 116, 117, 118, 119, 120, 121; structures
oxygen isotope ratios: and cave sediments palaeolimnology 155 116–18, 116, 117, 118
142; climate change boundaries 360; palaeomagnetic correlation 363–4 pH, diatoms and 203, 205, 206
correlation 366, 399; dating 268; deep- palaeomagnetism 4, 7, 319, 319–24, 321, Phanerozoic 59, 92n10
sea sediment 166–71, 168, 169, 390; and 322, 323, 324, 325 phenotypic plasticity 262
fractionation 167; and glacial ice storage palaeoseismicity 309 photography 20, 22
167–9, 168; Holocene 427; and ice palaeosols 122–7, 123, 125, 127, 128, 355, photo-imaging 95
volume 169; ice-core stratigraphy 175; 364 phreatic conditions 143
isotopic forms 167; isotopic stages 9–10, palaeotemperatures, and lake sediments phylogeography 262
10, 169, 330, 358; limitations 170–1; 160–1 phytoliths 142, 179n6
534 INDEX
pingos 56–7 194; regional vegetation reconstructions 2–3; geological timescale 1, 2–3; glacial
pixels 23 194; R-values 190–1; sample sources stages 1, 9; interglacial stages 9;
planetary harmonics 389 184–5; space–time reconstructions 195, interstadial episodes 9; onset 3; sequence
plankton blooms 304 195; taphonomy 197; transport correlation 336, 337, 338; stadial
planktonic assemblage zones 332 mechanisms 191; Tsuga (hemlock) episodes 9; system 2; time intervals
planktonic foram tests 252–3, 253 decline 194; Ulmus (elm) decline 194; 307–8
plant function types (PFTs) 266n2 zone boundaries 376, 419 Quaternary studies 5–9
plant macrofossils 142, 162, 207–15, 207, pollen diagrams 185, 186, 187–9, 187, 188, Queen Charlotte Island 68
209, 210, 211, 213, 332 189, 197; interpretation 190–4, 190, 191, Queen Maud Mountains, Antarctica 29
plant refugia 195 193 Quelcayya ice cap, Peru 174
plasma 103 pollen response functions 197
plateau correction procedure 290 pollen stratigraphy, applications 194–7, radar sensing 22, 23–4, 90, 90
Playfair, John 5, 6 194, 195, 196, 197 radiative forcing 395–6, 396
Pleistocene 2, 5, 104, 126 pollen zones 189 radioactive decay 268, 269–70, 286, 286
Pleistocene–Holocene transition 428–9, pollution levels 243, 297, 307, 309, 376, radioactive nuclides 269, 440
428, 429 419 radioactivity 268–70
Pliocene 4, 5, 61 Port Huron Advance 35 Radiocarbon (journal) 270
pluvial episodes 6–7, 82 Post, Lennart von 7 radiocarbon dating 7; accelerator mass
pluvial lakes 82, 82–6, 84, 85, 86, 140 postglacial erosion 29 spectrometry 271, 272–5, 273, 274,
PMIP (the Paleoclimate Model postglacial rebound 69 278–9; age estimates 271; age modelling
Intercomparison Project) 388 potassium–argon dating 284–5, 284 282–3, 283; applicability 268, 270, 274;
Poland 31, 32, 56, 77, 119, 120, 224, 224 Poznan moraine 31 Bayesian-based statistical modelling
polar bears 262 precipitation 120, 140; and ELA 50, 52; 282–3, 282; beta decay 271; bone 278;
polar deserts 55 isotope dendroclimatology 304; calendar years 275, 275, 279; CALIB
pollen and pollen analysis 183, 209, 214, monsoonal circulation 15–16; and peat 280–1; calibration 279–84, 280, 281,
266n1, 308; anthropogenic indicators deposits 162, 163; Pleistocene–Holocene 282, 283, 309–10; calibration curve
196, 197; applications 194–7, 194, 195, transition 429; and river landforms 74 280–1, 281, 282, 283–4; CalPal 281;
196, 197; assemblage zones 331–2; pronival (protalus) ramparts 57–8, 57 chronological order reversal 281;
calibration 197; cave sediments 142, protalus lobes 56 contamination 277–9, 278; counting
193; climate change boundaries 360; protalus rock glaciers 56 statistics 274–5; date calculation 272;
compositional variation 192; protein chemistry 332, 333, 334, 336 decay curve 270–1, 271, 271; Fairbanks
concentrations 188–9; cooling onset proxy records 2, 379; 2.8 ka event 433; 8.2 calibration model 281; fluctuations 280;
evidence 5; correlation 368; corrosion ka event 430, 431; databases 387–8; gas proportional counting 271; Gaussian
193, 193; dating 189, 362; deposition definition 17n2; dendroclimatology 302; errors 281, 282; general principles
192; deterioration 193, 193; dispersal fossils and the fossil record 182; global 270–1; hard-water error 277–8; human
191–2; dissemination 183–4; exotic 188; environmental change 379–80; Heinrich impacts on 276; IntCal model 280, 281,
field work 184–5; human impacts 165, events 373; Holocene 427; ice-core 281; isotopic enrichment 271–2;
183, 195–6, 196, 197; identification 185, stratigraphy 176; lake-level changes 138; isotopic fractionation 276, 277; lake
194; indicator species 197; laboratory multi 182, 265–6, 304; North Atlantic sediment 278, 279; lake-level changes
methods 185; lake, mire and bog Oscillation (NAO) 436; palaeodata- 135; Late Weichselian ice sheet 30–1;
sediments 155, 157, 185, 186, 188–9, model comparisons 387–8; liquid scintillation counting 271; marine
188, 192; local vegetation Pleistocene–Holocene transition 428, carbon circulation 276–7; marine
reconstructions 194; mixing 192–3; 429; quality and quality control 380, reservoir errors 276, 329–30;
nature of 183–4, 184; oxidation 193, 387–8; record synchronization 375; measurement of 14C 271–5, 273, 274;
193; palaeoenvironmental single 182; surface wetness change mineral carbon error 277–8; modern
reconstruction 197, 212; peat deposits 162–3, 164; tuning 312 reference standard 272; molluscan
161, 165; pollen assemblage 185; pollen pseudosoils 124 remains 277, 278; OxCal 281, 282;
diagrams 185, 186, 187–9, 187, 188, 189, pulse counters 96 palaeoenvironmental applications
197; pollen diagrams, interpretation Pyramid Lake, United States of America 283–4; peat deposits 161–2; plant
190–4, 190, 191, 193; pollen grains 183, 228 macrofossils 210; pollen 189; precision
184; pollen influx 189, 192; pollen 274–5; quality assurance 275;
sources 191–2; pollen spectrum 185; quartz particles 97 radiocarbon wiggle-matching 280, 282,
pollen structure 183; the pollen sum Quaternary: character 1–2; definition 1, 5; 283; radiocarbon years 275, 275;
185, 187; pollen zones 189; preservation duration 3–5; event dating 308–9; radiometric dating 271, 271–2; reference
184; production 190–1, 191; redeposited framework 9–10, 10, 11, 12–13, 12; first year 271; sea-waters 276–7; sediments
pollen 193; regional pollen databases use of term 5; geochronological status 282; soils 279; standards 272; temporal
INDEX 535
variations, 14C production 275–6, 275; outwash 77; paired 75; and 66, 76; Foraminifera and 242, 243;
timescale calibration 275; uncertainty palaeoenvironmental reconstruction 78; geoidal irregularity 66–7;
272, 274–5 preservation 74–5; River Thames 78–82, geomorphological evidence 58–73;
radiocarbon wiggle-matching 280, 282, 79, 80; and sea-level change 76, 81; geomorphological features 58, 58, 62;
283 sequences 75–6, 77, 78, 81; staircases Glacio-eustatic Theory 6; glacio-eustasy
Radiolaria 248 74–5, 76, 79, 82; strath 75, 76; tectonic 60–1; glacio-isostasy 68–9, 68, 69–73,
radiometric clocks 270 influences 77; thalassostatic 76; 70, 71, 72; Holocene 65–6, 66, 69;
radiometric dating 267, 268–70; alpha- unpaired 75 hydro-isostasy 68, 69; and ice volume
recoil track (ART) dating 290; argon- roches moutonnées 40, 92n8 169–70; isostatic 59, 66, 68–73, 68, 69,
isotope dating 284–5, 284; correlation rock glaciers 55–6, 56 70, 71, 72; isostatic recovery 69, 71,
365; cosmogenic radionuclide (CRN) rock platforms 58, 58 72–3; Last Glacial Maximum (LGM)
dating 294–6, 295, 297; electron spin rock-hardness 341 62–3; Lateglacial 422; and marine
resonance (ESR) 293–4; fission track rock-weathering 340–2, 341 deposits 58; and marine oxygen isotope
dating 289–91; luminescence dating Roman period 78 record 169–70, 169; marine regressions
291–3, 291; short-lived radioactive Russell, Israel 6–7 64, 64; meltwater pulse (MWP) 1a 63;
isotopes 296–8; uranium-series Russia 31, 259, 304 Middle Pleistocene Transition (MPT)
(U-series) dating 286–9, 286, 288; 390; the Milankovitch hypothesis 14–15;
see also radiocarbon dating Sahara, the 86 modelling 66; negative sea-level
raised shorelines 70, 71, 72–3 salinity indicators: Coccolithophores 246; tendency 64, 64, 202, 203; ocean
range charts 242 diatoms as 202–3, 205; Foraminifera siphoning 69; ostracod analysis 240;
range fossils 233 241, 243; ostracod analysis 240 oxygen isotope record 60–1, 61;
range overlap method 222 Salpausselkä Moraines 32 palaeoenvironmental significance of 73;
ratite eggshell, dating 338 salt lakes 134, 203 Pliocene 60; positive sea-level tendency
Rayleigh test 443n1 salt oscillator hypothesis 403–4 64, 64, 202, 203; Pre-Quaternary
Reboul, Henri 5 salt-density variations 16 eustatic 59–60; reconstruction 59;
recessional stages 27 San Francisco Bay 212 relative curves 243; and river terraces 76,
record synchronization 374–8, 375, 376, sand seas 127, 131 81; and sea-floor spreading 60; and
377 sand wedges 116–17, 116, 119 speleothem formation 149; tectonic
recurrence surfaces 161 Sandar 91n3 influences 67–8, 67; teleconnection 411,
Red Sea 61, 411, 411 Sangamonian 10 411; worldwide 59, 60
regional pollen databases 194 sapropels 17n6, 377–8, 377 sea-level index point 63, 64
regional stratotypes 349 satellite images 20, 22–3, 33 Searles Lake basin, United States of
regional vegetation reconstructions 194 Saussure, Horace-Bénédict de 111 America 133
Reid, Clement, The Origin of the British Scandinavia 71, 307, 308, 433 sea-surface temperatures (SSTs) 16, 373;
Flora 207 scanning electron microscopy (SEM) 97, cooling trend 394; and coral 319;
relative age methods 267 107, 127, 179n1, 207, 208 Dinoflagellates (dinocysts) and 246; Last
relict palaeosols 122 Schmidt hammer 341, 341–2 Termination 415, 416; and marine
remanent magnetism 320 Schott, Wolfgang 8 oxygen isotope record 170;
remote sensing 7, 22–6, 25, 26, 31–2, 40, Scilly Isles 33 micropalaeontology, deep-sea sediment
95 sclerochronology 298, 316, 317, 318, 319 249, 250–1, 250, 251, 252; Middle
reservoir effect, the 278 Scotland 33, 35; equilibrium line altitude Pleistocene Transition (MPT) 390;
residual rebound 69 (ELA) 57, 57; isobase patterns 71; molluscan remains and 237; ocean–ice
restrained rebound 69 marine molluscs 319; palaeoclimatic interactions, North Atlantic 400;
Rhine, River 75, 76, 77 inferences 51–2, 51, 52; sea-level change ostracod analysis 240;
Rhine–Meuse system 77 70 Pleistocene–Holocene transition 429;
Rhône, River 212 seabed, hydro-isostatic load 60 Radiolaria and 245
rice farming 438 sea-floor spreading, and sea-level change sea-waters, age estimates 276–7
Riss–Würmian 10 60 secular magnetic database 345n2
river sediments, vertebrate remains 258 sea-ice formation 178 secular magnetic variations 320–1, 321,
river terraces 74–5, 75; aggradation 76, sea-level change 19; absolute 59; cause 363–4
76–7; alluviation 78, 78; backswamp differentiation 66; chronologies 61–2; sedigraph 97
deposits 78; climatic 76–7; and climatic and coral 319; and coral polyps 263–4; sedimentary record 93
change 76–7, 81–2; cut and fill 75, 76; correlation 365; cycles 62, 62; dating 59, sedimentary sequences 93, 94–5
dating 77; evolution 78–82; and 61–2; diatoms and 203, 204; and dust sedimentation rates: deep-sea sediment
glaciation 77; gradients 77; and human 411, 411; and estuarine sediments 63–4, 170–1; lakes 157–8, 158
activity 77–8, 78; origins 76–8, 78; 64; eustatic 59–77, 60, 61, 62, 63, 64, 65, sedimentological analysis 7
536 INDEX
sediments 97–8, 178–9; beds 349; clay soils: dating 126, 279; diagenesis 123–4; 439; approaches 348; beds 350, 351;
mineralogy 98; dating 282, 292; deep-sea fabric 124–5; formation 122–3, 355; biostratigraphy 348, 353–4, 353, 363;
165–71, 166, 168, 169; deformation 352, horizon boundaries 124; mean residence biozone correlation 353–4; chemical
353; fabrics 93; Foraminifera 241, 242; time 279; micromorphology 124–5, 125; 348; chronostratigraphy 348, 361–2,
heavy mineral assemblages 98; insect mineral magnetic analysis 125, 126; 362, 378; climatostratigraphy 348,
remains 215, 226; lake-level records plant macrofossils 207; profiles 123, 124; 358–61, 359, 360, 361, 378; codes of
132–6, 133, 134, 136, 137–8, 137, 139, properties 124; radiocarbon dating 279; practice 348; correlation 347, 359,
140–1, 140; lithostratigraphic units relict features 122, 123; stratigraphy 355; 362–3; definition 347; diatom 206;
349–51, 350, 351; luminescence dating see also palaeosols electrical conductivity measurements
292; metallic ions 98; Milankovitch cycle solar output variations 16, 160, 275–6, 283, 348; equifinality 352–3; event 348, 365;
correlation 368–9, 369, 370, 371; 296, 404, 415, 431, 433, 433, 434, 434–5, geochronological units 348;
mineral magnetic analysis 98; molluscan 438, 440 geochronology 361–2, 362; Lateglacial
remains 231–2; palaeosols 122–7, 123, Somme, River 75, 76, 77 synchronization 419–20; lateral 363;
125; particle shape 97; particle size sonar signals 22, 24, 25, 95 lithostratigraphy 348, 349–53, 350, 351,
distribution 96–7, 97; periglacial Soreq Cave, Israel 146, 148 352, 378; loess 127–8, 128, 129, 130,
115–22, 116, 117, 118, 119, 120, 121; South America 128; Quaternary 130; magnetostratigraphy 348;
plant macrofossils 207, 209, 210, 211; environments 260; sand seas 131; morphostratigraphy 348, 354–5, 354,
radiocarbon dating 282; sections 93–4, vertebrate biostratigraphy 259; 378; oxygen isotope 348, 355–7, 357,
94; sequence stratigraphy 352, 352; shelf vertebrate remains 260 358; palaeomagnetic correlation 363–4;
seas 242; stable isotope analysis 98–9; Southern Oscillation 435 palaeosol correlation 364; pollen 194–7,
thin sections 97, 106; types 93; varved Space Imaging, Inc 22 194, 195, 196, 197; principles 347–8,
304; vertebrate remains 257–8; wind- Spain 91, 240, 259 349; record synchronization 374–8, 375,
blown 127–8, 128, 129, 130–2, 130, 132; spallation 294 376, 377; seismic 348; sequence 351–2,
see also cave sediments; glacial species diversity 233 352; shoreline correlation 364–5; soil
sediments; lake, mire and bog sediments species richness 233 stratigraphy 355; stages 362, 378;
seismic events 309 SPECMAP timescale 330–1, 331, 356 stratotypes 349; synchronization 367;
seismic sensing 24 spectral analysis 15, 17n7, 393 tephra correlation 364; time-rock units
seismic stratigraphy 348 speleothem 141, 409; amino-acid 362; time-stratigraphic markers 362;
seismic surveys 24 geochronology 334; dating 143, 144, time-transgression 348, 349, 362, 363;
sequence stratigraphy 351–2, 352 149, 286–7, 288, 293, 298, 316–17; units 348, 349; vertebrate 259; see also
series 2 formation 141, 141, 143; formation ice-core stratigraphy
Sernander, Rutger 7 and sea-level change 149; formation stratotypes 4
Sevrier-Les-Charretières, France 159 and tectonic activity 149; growth and streamlined glacial deposits 42–3
shallow marine water mass, temperature climatic change 143–4, 144; growth striations 40, 41, 42
change 243–4 layers 143, 144, 316–17, 371; isotope stromatolites 134, 150
Shannon diversity index (SDI) 233 ratios 145–8, 147, 148; Milankovitch Subcommission on Quaternary
Shetland Isles 33 cycle correlation 371; oxygen isotope Stratigraphy (SQS) 3, 17n3, 349
shoreline correlation 364–5 ratios 373–4; trace elements 148 subenvironments 100, 100, 101, 102
shoreline displacement 67–8, 67 Sphagnum mosses 154, 155, 210, 213 subglacial deformation 102–3
side-looking airborne radar (SLAR) 24 Spitsbergen 102 subglacial subenvironments 100, 100
silcrete 91 spontaneous fission 289, 290 subglacial traction till 104
silicon-32 297, 298 spores: see pollen and pollen analysis sublimation 102
SIMMAX 251–2 SPOT 23 sublimation till 102, 103
single aliquot regeneration method (SAR) stable isotope analysis 98–9 submarine landforms 33
292 stadial episodes, definition 9 sub-Milankovitch timescales 16, 367, 371,
single laser crystal fusion (SLCF) 40Ar/39Ar stages 2 371, 372–4, 372, 373, 374, 378, 379
dating 285 Steel Lake, United States of America 188 Sudan 90, 234
skeleton grains 103 stereonets 114, 114 Sumatra 327
Skye, Isle of 157 stochastic resonance 413, 443n2 sunspot cycle 138, 434
S-matrix structures 104 Stockholm 305 superposition, 267, 336, 351, 361, 363, 365,
smear slides 248 stoss-and-lee landforms 40 375; principle of 345n1
Smith, James 7 stratigraphic boundaries 349, 349 supraglacial subenvironments 100, 100
snow shadow 45–6 stratigraphic markers 127, 267 surface exposure dating 294
Snowdonia, Wales 40 stratigraphic resolution 343–4, 344 surface wetness change: and peat deposits
soil catena chronosequences 343 stratigraphy 264, 347; age modelling 420–1; 162, 163; and plant macrofossils 162;
soil wedges 117, 119 allostratigraphy 352; the Anthropocene proxy records 162–3, 163, 164
INDEX 537
surging 30, 37 tephrochronology 319, 325–31, 326, 328, transient accelerated palaeo-simulations
Svalbard-Barents Sea 31 329, 331, 364, 420 387
swathe bathymetry 24 tephrostratigraphy 325 transient simulations 386–7
Sweden 63, 161, 224, 224, 305, 307, 309 terminations 246, 357, 358, 365, 366, 367, travertine 149–50, 150
Switzerland 42, 160, 196, 196 396, 413 tree rings: see dendrochronology
syngenetic ice wedges 116 terminology (stratigraphic) 9 trimlines 27, 28, 29, 33
synthetic aperture radar (SAR) 24, 90, 90 Terra (EOS AM-1) 23 trophic status 153, 205
TerraSAR-X satelite 24 tsunami deposits 240, 364–5
Tahoe Moraines, North America 38, 39 terrestrial in situ cosmogenic nuclide tufa 150, 150
Tajikistan 128 (TCN) dating 294 tunnel valleys 33, 91n7
TALDICE consortium 172 Tertiary 3 Turkey 78, 78
technological advances 7 testate amoebae analysis 162–3, 264–5 turnover rate 262
tectomicts 104 TEX86 index 265 Tuscany, Italy 233
tectonic basins 368 textural B horizon, Europe 126 Tyndall, John 437
tectonic influences 15: Middle Pleistocene thalassostatic terraces 76 type sites 349
Transition (MPT) 392; river terraces 77; Thames, River 1; amino-acid
sea-level change 67–8, 67; speleothem geochronology 336, 338, 338; drainage uniformitarian principle, the 125–6, 181,
formation 149; uplift 77 system 80, 81; Last Glacial Maximum 182
teeth, vertebrate 254, 255, 256, 294 (LGM) 81; luminescence dating 293; Uniformitarianism 5
teleconnection 399–400; anti-phase river terrace sequences 78–82, 79, 80; United States of America: Alaska 214;
behaviour 406–7; bipolar 405–7, 405, vertebrate biostratigraphy 259 Beaver Lake 201; Bishop Tuff 290–1;
406; bipolar seesaw 406–7, 409, 411, thermal ionization mass spectrometry 287 dendrochronology 301, 301; dust flux
412, 418, 440; dust circulation 410–11, thermally transferred optically stimulated 309; Florida 212; Lake Bonneville 82,
410, 412; feedback loops 407, 409, luminescence 292–3 83–6, 84, 86, 133, 339; Lake Lahontan
410–11; global mechanisms 407–13, thermistors 177, 179n111 82, 83, 84; Mono Lake 150; Mt St Helens
408, 410, 411, 412; heat piracy 407; thermoclastic scree 142 327, 328; North Cascade Range 210,
interdependency 412–13, 412; Last thermohaline circulation 16, 401, 402 214; Pyramid Lake 228; record
Termination 418–19; ocean–ice thermokarst 55 synchronization 375; Searles Lake basin
interactions, North Atlantic 400, 400–1, thermoluminescence 291 133; soil stratigraphy 355; Steel Lake
401, 402, 403–5, 403; problems 412–13; thermoremanent magnetization (TRM) 188; tephra 327, 328; Yellowstone 285,
sea-level change 411, 411; tunnel linking 320 327; see also North America
400 thin sections 97, 106 uranium-series (U-series) dating 286–9,
temperate stage 9 thúfur 58 286, 288
temperatures 1–2; anthropogenic Tibet 134, 138, 172 uranium-series disequilibrium dating
warming 437–9, 437, 438; bipolar Tiedemann–Peyto Advance, North method 286–7
seesaw 406–7; Chironomidae 227–8, America 39 urbanization 437
227; deep-sea 169; and ELA 50, 51; tills 99; classification of 102–7, 103, 104, Urströmtaler 30
environmental simulation models 105, 106, 107; clay mineral ratios 115; U.S. Geological Survey (USGS) 22,
(ESMs) 393, 394; GCMs 382, 383; cosmogenic nuclides 115; fabric 113–15, 200
glacial–interglacial cycles of the last 114; indicator minerals 115; matrix Ussher, James, Archbishop of
800 ka 398, 399; greenhouse effect properties 115; provenance 113 Armagh 5
437–8, 437, 438; Holocene 427, 427, time-rock units 362
434; human impacts 436, 437–8, 437, time-stratigraphic correlation 374–8, 375, vadose 143
438, 442; ice-core stratigraphy 177; 376, 377 varnish micro-lamination (VML) dating
Last Glacial Maximum (LGM) 148, time-stratigraphic markers 362 341
339; Last Termination 417–18, 417; time-transgression 354; climate change varves and varve chronology 100, 298,
Little Ice Age 434; mean annual 367; human impacts 440; marine oxygen 304; applications 307–10, 308, 310;
temperatures 119–20; mean global 417; isotope record 366; stratigraphy 348, chemical 306; chronological errors 306;
palaeothermometry 339; periglacial 349, 362, 363 clastic 304, 304–5, 305; complex 306;
landforms and 56; shallow marine water tipping points 386, 404, 438, 440 counting error sources 306–7; cross
mass 243–4; surface water 227–8, 227; TOPEX/Poseidon satellite 23 dating 306; glacigenic 304, 304–5;
summer 51–2; see also sea-surface toposequences 124 human impacts 309; Lateglacial 420;
temperatures (SSTs) tors 54, 54 nature of 304; organic 304, 305–6,
temporal resolution 343–4, 344 total stations 20 305; radiocarbon dating calibration
tephra 364, 375; correlation 420; marker transfer functions 192, 205, 212, 234, 240, 309–10; sublaminations 306; Swedish
horizons 329, 329 242, 243, 246, 250–1, 263–4, 266n4 Timescale (STS) 307; year zero 307
538 INDEX
vegetation reconstructions 194, 195, 195 volcanic activity: and climate change 330; Working Group of the Subcommission
vegetational change, early evidence 7 and human evolution 330; record on Quaternary Stratigraphy 439
vegetational productivity 211 synchronization 377 World Modern Foraminifera Database
vegetational signatures 9 volcanic ash: see tephrochronology 242
Venetz, Ignaz 5 volcanic histories 330 Würmian 10
vertebrate biogeography 259–60, 259, volcanic rocks: dating 284–5, 289;
260 luminescence dating 291; X-radiography 95, 96, 96
vertebrate biostratigraphy 259 recrystallization 285 X-ray densitometry 299
vertebrate remains 254; bioclimatic limits Vosges Mountains, France 212 X-ray diffraction (XRD) 98
261; biogeography 259–60, 259, 260; Vostok core, the 8 X-ray fluorescence (XRF), 95–6
bone 254, 255, 289, 334; bone structure Vrica section, Calabria 4
256; burial effects 256; cave and fissure Yellowstone, United States of America
deposits 257–8; changes in animal Wales 33, 40, 107, 111, 218 285, 327
morphology 260, 260; dating 258, 278, Walton Moss 163, 163 Younger Dryas Stadial 12, 52, 56, 138, 155,
289, 294; decalcification 256; DNA water volume 59–60 277, 413, 414, 419; dendrochronology
analysis 262; ecological niche models weathering: crusts 90–1; dating 294; 301–2; duration of 308; lake-level
261; excavation 257; and faunal profiles 124; rinds 341, 341; rocks changes 160; mean annual air
evolution 262–3; field methods 256–7; 340–2, 341 temperature (MAAT) 121, 121;
fluorine content 339–40; fluvial Weichselian Lateglacial 10, 12 molluscan remains 237; moraines 32;
sediments 258; fossilization process welded soils 122 mutual climatic range (MCR) 224;
256; human impacts 261–2; hydraulic West Runton, Norfolk 104 onset 423–5, 424, 425; permafrost
sorting 258; identification 257; indicator whalebacks 40 limits 121, 121; plant macrofossils 212;
species 260–1; lacustrine sediments wiggle matching 280, 282, 283, 367 Polar Front position 424–5, 424, 426;
258; micro-vertebrates 257; molecular wind erosion 54 speleothem growth 146; transitions 426,
clock analysis 332; phylogeography 262; wind patterns: 2.8 ka event 433; and ice- 426
and Quaternary environments 260–2; sheet modelling 45–6; reconstruction Yukon Territory, Canada 214
range 254, 259–60, 259; records 258–63, 131
259, 260; recovery sites 256; wind-blown sediments 127–8, 128, 129, Zeuner, Frederick 261
stratigraphic ranges 259; taphonomy 130–2, 130, 132 zonal soils 122
257–8; teeth 254, 255, 256, 294; Windermere Interstadial 12 zoogeographical provinces: marine fauna
turnover rate 262; uranium content Wisconsinan 10, 35, 36, 39 236, 236
339–40; weathering 256 Woodward, Bernard 7, 228

Lowe, Joseph John - Walker, Mike - Reconstructing Quaternary Environments-Routledge (2015)

Uploaded by

Copyright:

Available Formats

Lowe, Joseph John - Walker, Mike - Reconstructing Quaternary Environments-Routledge (2015)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lowe, Joseph John - Walker, Mike - Reconstructing Quaternary Environments-Routledge (2015)

Uploaded by

Copyright:

Available Formats

Reconstructing

0 200 400 600 800

John Lowe and

ISBN: 978-0-415-74075-3 (hbk)

Typeset in Minion and Univers

List of ﬁgures and tables xv

1 The Quaternary record 1

2.3.2.3 North America 35

3.2.3.1 Particle size measurements 96

3.8.4.2 Stable isotope ratios in cave speleothem 145

4 Biological evidence 181

4.2.2 The nature of pollen and spores 183

4.6.6.3 Archaeological relevance 234

5 Dating methods 267

5.4.2.2 Clastic varves 304

6 Approaches to Quaternary stratigraphy and

7 Global environmental change during the Quaternary 379

7.4.4 Global teleconnections: linking mechanisms 407

6.1 Time-transgression in lithostratigraphic and biostratigraphic boundaries 349

3.3 Classiﬁcation of lakes by mode of formation 151

1.1 INTRODUCTION were interspersed with warm episodes (interglacials) during

Periods 0 Periods E S S E evidence evidence

Tertiary 2 ILasI temperate stage

UIO Cretaceous 10 0.4

Epoch Age Ma However, following a protracted campaign by the Sub-

Louis Agassiz to expound, in 1837, the ﬁrst coherent theory

MARINE ISOTOPE R E C O R D NORTHERN HEMISPHERE STAGES

1.2 3-3 Menapian Pre-lllinoian G

Holocene Holocene Flandrian

11,500 -10,050 -11,500

12,000 Greenland Younger Loch -10,255 -12,000

12,500 -10,475 -12,500

13,000 Gl-ib -11,000 -13,000

15,000 -12,700 -15,000

a) Almost elliptical Almost circular

Period icily 9$,00Qyr

Period icily 42.DO.0yr ETP

Equator Wiriter 0.0

ecliptic; c) precession of the equinoxes.

cycle (Mesolella et al., 1969). More signiﬁcant, however,

fossils which provided a long-term proxy record of environ- 0

geomorphological mapping. The spatial resolution and

expansion, and therefore reﬂect a glacier readvance. Where MKt-

(Lingle & Fatland, 2003). Despite these difﬁculties, however, ta

there is now a broad consensus about the positions of the tH a m b u r g l

2.3.2 Geomorphological evidence and the

112 W 100" 80-

2.3.3.1 Striations 2.3.3.2 Friction cracks

Whale ESE esker pulling

enable comparisons to be made between drumlins in

2.3.4 Reconstruction of former ice masses

2.3.4.1 Ice-sheet modelling

Subgrid Topographic Attributes: Ice T h i c k n e s s H (A.Q, i)

Initial Ice T h i c k n e s s H (A,S, trf

-50 -40-35-30-25 -20-15-10 -5 -1 0 1 0 10 20 501001502002505007501000

a! left shfral surface (m) a!

Forms glacier rnargn

Fame* ice-dammed bfce

(the exact relationship can be measured for any particular