Fish Cognition and Behavior (Fish and Aquatic Resources)

BLBK374-FM BLBK374-Brown June 7, 2011 11:29 Copyeditor’s Name: Trim: 244mm X 172mm Fish Cognition and Behavior i ...

Author: Culum Brown | Kevin Laland | Jens Krause

59 downloads 2471 Views 8MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

BLBK374-FM

BLBK374-Brown

June 7, 2011

11:29

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Cognition and Behavior

i

Char Count=

BLBK374-FM

BLBK374-Brown

June 7, 2011

11:29

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish and Aquatic Resources Series Series Editor: Tony J. Pitcher Professor of Fisheries Policy & Ecosystem Restoration in Fisheries, Fisheries Centre, Aquatic Ecosystems Research Laboratory, University of British Columbia, Canada The Wiley-Blackwell Fish and Aquatic Resources Series is an initiative aimed at providing key books in this fast-moving field, published to a high international standard. The Series includes books that review major themes and issues in the science of fishes and the interdisciplinary study of their exploitation in human fisheries. Volumes in the Series combine a broad geographical scope with in-depth focus on concepts, research frontiers, and analytical frameworks. These books will be of interest to research workers in the biology, zoology, ichthyology, ecology, and physiology of fish and the economics, anthropology, sociology, and all aspects of fisheries. They will also appeal to non-specialists such as those with a commercial or industrial stake in fisheries. It is the aim of the editorial team that books in the Wiley-Blackwell Fish and Aquatic Resources Series should adhere to the highest academic standards through being fully peer reviewed and edited by specialists in the field. The Series books are produced by Wiley-Blackwell in a prestigious and distinctive format. The Series Editor, Professor Tony J. Pitcher, is an experienced international author, and founding editor of the leading journal in the field, Fish and Fisheries. The Series Editor, and Publisher at Wiley-Blackwell, Nigel Balmforth, will be pleased to discuss suggestions, advise on scope, and provide evaluations of proposals for books intended for the Series. Please see contact details listed below. Titles currently included in the Series 1. Effects of Fishing on Marine Ecosystems and Communities (S. Hall) 1999 2. Salmonid Fishes (Edited by Y. Altukhov et al.) 2000 3. Percid Fishes (J. Craig) 2000 4. Fisheries Oceanography (Edited by P. Harrison and T. Parsons) 2000 5. Sustainable Fishery Systems (A. Charles) 2000 6. Krill (Edited by I. Everson) 2000 7. Tropical Estuarine Fishes (S. Blaber) 2000 8. Recreational Fisheries (Edited by T. J. Pitcher & C. E. Hollingworth) 2002 9. Flatfishes (Edited by R. Gibson) 2005 10. Fisheries Acoustics (J. Simmonds & D. N. MacLennan) 2005 11. Fish Cognition and Behavior (Edited by C. Brown, K. Laland & J. Krause) 2006 12. Seamounts (Edited by T. J. Pitcher, T. Morato, P. J. B. Hart, M. R. Clark, N. Haggan & R. S. Santos) 2007 13. Sharks of the Open Ocean (Edited by M. D. Camhi, E. K. Pikitch and E. A. Babcock) 2008 14. World Fisheries (Edited by R. E. Ommer, R. I. Perry, K. Cochrane & P Cury) 2011 15. Fish Cognition and Behavior, Second Edition (Edited by C. Brown, K. N. Laland & J. Krause) 2011 For further information concerning existing books in the series, please visit: www.wiley.com To discuss an idea for a new book, please contact: Nigel Balmforth, Life Sciences, Wiley-Blackwell, 9600 Garsington Road, Oxford OX4 2DQ, UK Tel: +44 (0) 1865 476501 Email: [email protected]

ii

BLBK374-FM

BLBK374-Brown

June 7, 2011

11:29

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior Edited by

Culum Brown Department of Biological Sciences, Macquarie University, Sydney, Australia

Kevin Laland Centre for Social Learning and Cognitive Evolution, School of Biology, University of St Andrews, UK

Jens Krause Department of Biology and Ecology of Fishes, Leibniz-Institute of Freshwater Ecology and Inland Fisheries, Berlin; Humboldt University, Berlin

A John Wiley & Sons, Ltd., Publication

iii

BLBK374-FM

BLBK374-Brown

June 7, 2011

11:29

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

First edition published 2006 C 2011, 2006 by Blackwell Publishing Ltd. This edition first published 2011 Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing program has been merged with Wiley’s global Scientific, Technical and Medical business to form Wiley-Blackwell. Registered office John Wiley & Sons, Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial offices 9600 Garsington Road, Oxford, OX4 2DQ, UK The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK 2121 State Avenue, Ames, Iowa 50014-8300, USA For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell. The right of the author to be identified as the author of this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data Fish cognition and behavior / edited by Culum Brown. – 2nd ed. p. cm. – (Fish and aquatic resources series) Includes bibliographical references and index. ISBN 978-1-4443-3221-6 (hardcover : alk. paper) 1. Fishes–Behavior. 2. Fishes–Psychology. 3. Cognition in animals. I. Brown, Culum. QL639.3.F575 2011 597–dc22 2011002188 A catalogue record for this book is available from the British Library. This book is published in the following electronic formats: ePDF 9781444342505; Wiley Online Library 9781444342536; ePub 9781444342512; Mobi 9781444342529 R Inc., New Delhi, India Set in 10/13pt Times New Roman by Aptara

1

2011

iv

BLBK374-FM

BLBK374-Brown

June 8, 2011

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Contents

Preface and Acknowledgements Series Foreword List of Contributors

1

2

xv xvi xix

Fish Cognition and Behaviour Brown, Laland and Krause

1

1.1 Introduction 1.2 Contents of this book References

1 3 9

Learning of Foraging Skills by Fish Warburton and Hughes

10

2.1 Introduction 2.2 Some factors affecting the learning process 2.2.1 Reinforcement 2.2.2 Drive 2.2.3 Stimulus attractiveness 2.2.4 Exploration and sampling 2.2.5 Attention and simple association 2.2.6 Cognition 2.2.7 Memory systems and skill transfer 2.3 Patch use and probability matching 2.4 Performance 2.5 Tracking environmental variation 2.6 Competition 2.7 Learning and fish feeding: some applications 2.8 Conclusions Acknowledgements References

10 12 12 12 12 14 14 15 18 19 21 23 26 27 27 28 29

BLBK374-FM

BLBK374-Brown

June 8, 2011

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

vi

Contents

3

Learned Defences and Counterdefences in Predator–Prey Interactions Kelley and Magurran

36

3.1 3.2

36 38 39 39 40 41 42 42 43 43 44 45 46 47 47 47 47 49 49 50 50 51 52 53

3.3

4

Introduction The predator–prey sequence 3.2.1 Encounter 3.2.1.1 Avoiding dangerous habitats 3.2.1.2 Changing activity patterns 3.2.2 Detection 3.2.2.1 Crypsis 3.2.2.2 Sensory perception 3.2.3 Recognition 3.2.3.1 Associative learning 3.2.3.2 Learning specificity 3.2.3.3 Search images 3.2.3.4 Aposematism and mimicry 3.2.4 Approach 3.2.4.1 Pursuit deterrence 3.2.4.2 Gaining information about the predator 3.2.4.3 Social learning 3.2.4.4 Habituation 3.2.5 Evasion 3.2.5.1 Reactive distance and escape speed and trajectory 3.2.5.2 Survival benefits/capture success Summary and discussion Acknowledgements References

Learning about Danger: Chemical Alarm Cues and Threat-Sensitive Assessment of Predation Risk by Fishes Brown, Ferrari and Chivers 4.1 4.2

4.3

4.4

4.5

Introduction Chemosensory cues as sources of information 4.2.1 Learning, innate responses and neophobia 4.2.2 Learned predator recognition through conditioning with alarm cues Variable predation risk and flexible learning 4.3.1 Assessing risk in time 4.3.2 Sensory complementation and threat-sensitive learning Generalisation of risk 4.4.1 Generalising of predator cues 4.4.2 Generalisation of non-predator cues Predator recognition continuum hypothesis 4.5.1 Ecological selection for innate versus learned recognition of predators 4.5.2 Ecological selection for generalised learning

59 59 60 60 62 62 64 65 66 66 67 68 69 69

BLBK374-FM

BLBK374-Brown

June 8, 2011

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Contents

4.6 4.7

4.8

5

Retention: the forgotten component of learning Conservation, management and learning 4.7.1 Conditioning predator recognition skills 4.7.2 Anthropogenic constraints 4.7.3 Field-based studies Conclusions Acknowledgements References

vii

70 72 72 73 73 74 74 74

Learning and Mate Choice Witte and N¨obel

81

5.1 5.2

81 82

5.3 5.4

5.5

5.6

5.7 5.8

5.9

5.10 5.11

Introduction Sexual imprinting 5.2.1 Does sexual imprinting promote sympatric speciation in fishes? Learning after reaching maturity Eavesdropping 5.4.1 Eavesdropping and mate choice 5.4.2 Benefits of eavesdropping 5.4.3 The audience effect Mate-choice copying 5.5.1 Mate-choice copying – first experimental evidence and consequence 5.5.2 Mate-choice copying – evidence from the wild 5.5.3 Mate-choice copying when living in sympatry or allopatry 5.5.4 Mate-choice copying – the role of the early environment 5.5.5 Quality of the model fish Social mate preferences overriding genetic preferences 5.6.1 Indications from guppies 5.6.2 Indications from sailfin mollies Cultural evolution through mate-choice copying Does mate-choice copying support the evolution of a novel male trait? 5.8.1 Theoretical approaches 5.8.2 Experimental approaches Is mate-choice copying an adaptive mate-choice strategy? 5.9.1 Benefits of mate-choice copying 5.9.2 Costs of mate-choice copying Outlook Conclusions References

82 83 84 84 84 85 87 88 89 91 92 93 94 94 95 96 96 97 98 99 99 100 101 102 102

BLBK374-FM

BLBK374-Brown

viii

6

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Contents

Aggressive Behaviour in Fish: Integrating Information about Contest Costs Hsu, Earley and Wolf 6.1 6.2 6.3

6.4 6.5

7

June 8, 2011

Introduction Information about resource value Information about contest costs 6.3.1 Assessing fighting ability 6.3.2 Information from past contests 6.3.2.1 Winner and loser effects 6.3.2.2 Individual recognition 6.3.2.3 Social eavesdropping 6.3.3 Integrating different types of cost-related information Physiological mechanisms Conclusions and future directions Acknowledgements References

108 108 110 110 111 113 113 117 117 118 119 126 128 128

Personality Traits and Behaviour Budaev and Brown

135

7.1 7.2

135 137 137 138 140 140 140 142

7.3 7.4 7.5

7.6 7.7

7.8

Introduction Observation and description of personality 7.2.1 Current terminology 7.2.1.1 Shyness–boldness 7.2.1.2 Coping styles 7.2.1.3 Behavioural syndromes 7.2.2 Objectivity 7.2.3 Labelling personality traits; construct validity 7.2.4 Objective and subjective measurements of personality 7.2.5 Modern terminology and statistical approaches Proximate causation Ontogeny and experience Is personality adaptive? 7.5.1 Frequency- and density-dependent selection 7.5.2 State-dependent models Evolution Wider implications 7.7.1 Fish production and reproduction 7.7.2 Personality and population dynamics Conclusions Acknowledgements References

142 145 146 149 150 150 151 153 155 155 155 156 157 157

BLBK374-FM

BLBK374-Brown

June 8, 2011

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Contents

8

The Role of Learning in Fish Orientation Odling-Smee, Simpson and Braithwaite

166

8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9

166 166 167 168 171 172 173 174 174 174 175 176 177 179 179 180

8.10 8.11

9

Introduction Why keep track of location? The use of learning and memory in orientation Learning about landmarks Compass orientation Water movements Inertial guidance and internal ‘clocks’ Social cues How flexible is orientation behaviour? 8.9.1 When to learn? 8.9.2 What to learn? 8.9.3 Spatial learning capacity Salmon homing – a case study Conclusion Acknowledgements References

Social Recognition of Conspecifics Griffiths and Ward

186

9.1 9.2

186 186 187 187 191 194 195 196 196 200 201 201

9.3

9.4

10

ix

Introduction Recognition of familiars 9.2.1 Laboratory studies of familiarity 9.2.2 Mechanisms of familiarity recognition 9.2.3 Functions of associating with familiar fish 9.2.4 Familiarity in free-ranging fishes 9.2.5 Determinants of familiarity Familiarity or kin recognition? 9.3.1 Kin recognition theory 9.3.2 Evidence for kin recognition from laboratory studies 9.3.3 Advantages of kin discrimination 9.3.4 Kin association in the wild 9.3.5 Explaining the discrepancies between laboratory and field 9.3.6 Kin avoidance Conclusion References

203 205 206 207

Social Organisation and Information Transfer in Schooling Fish Ioannou, Couzin, James, Croft and Krause

217

10.1 10.2 10.3

217 218 219

Introduction Collective motion Emergent collective motion in the absence of external stimuli

BLBK374-FM

BLBK374-Brown

x

10.5 10.6 10.7 10.8 10.9

12

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Contents

10.4

11

June 8, 2011

Response to internal state and external stimuli: Information processing within schools 10.4.1 Collective response to predators 10.4.2 Mechanisms and feedback in information transfer 10.4.3 Information transfer during group foraging and migration Informational status, leadership and collective decision-making in fish schools The structure of fish schools and populations Social networks and individual identities Community structure in social networks Conclusions and future directions Acknowledgements References

220 220 222 225 225 227 229 232 233 234 234

Social Learning in Fishes Brown and Laland

240

11.1 Introduction 11.2 Antipredator behaviour 11.3 Migration and orientation 11.4 Foraging 11.5 Mate choice 11.6 Aggression 11.7 Trade-offs in reliance on social and asocial sources of information 11.8 Concluding remarks Acknowledgements References

240 241 244 247 248 249 250 252 252 252

Cooperation and Cognition in Fishes Alfieri and Dugatkin

258

12.1 12.2 12.3

258 259 261 261 261

Introduction Why study cooperation in fishes? Cooperation and its categories 12.3.1 Category 1 – kin selection 12.3.1.1 Cognition and kin selection 12.3.1.2 Example of kin selected cooperation: Cooperative breeding 12.3.1.3 Example of kin selected cooperation: Conditional territory defence 12.3.2 Category 2 – reciprocity 12.3.2.1 Cognition and reciprocity 12.3.2.2 Example of reciprocity: Egg trading 12.3.2.3 Example of reciprocity: Predator inspection 12.3.2.4 Example of reciprocity: Interspecific cleaning behaviour

262 262 263 264 265 266 267

BLBK374-FM

BLBK374-Brown

June 8, 2011

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Contents

12.3.3

12.4

13

Category 3 – by-product mutualism 12.3.3.1 Cognition and by-product mutualism 12.3.3.2 Example of by-product mutualism: Cooperative foraging 12.3.4 Category 4 – trait group selection 12.3.4.1 Cognition and trait group selection 12.3.4.2 Example of trait group selected cooperation: Predator inspection Conclusion Acknowledgements References

xi

268 268 269 270 270 270 271 272 272

Machiavellian Intelligence in Fishes Bshary

277

13.1 13.2

277 279

13.3

13.4

Introduction Evidence for functional aspects of Machiavellian intelligence 13.2.1 Information gathering about relationships between other group members 13.2.2 Predator inspection 13.2.3 Group-living cichlids 13.2.4 Machiavellian intelligence in cleaning mutualisms 13.2.4.1 Categorisation and individual recognition of clients 13.2.4.2 Building up relationships between cleaners and resident clients 13.2.4.3 Use of tactile stimulation by cleaners to manipulate client decisions and reconcile after conflicts 13.2.4.4 Audience effects in response to image scoring and tactical deception 13.2.4.5 Punishment by males during pair inspections Evidence for cognitive mechanisms in fishes 13.3.1 What cognitive abilities might cleaners need to deal with their clients? 13.3.2 Other cognitive mechanisms Discussion 13.4.1 Future avenues I: How Machiavellian is fish behaviour? 13.4.2 Future avenues II: Relating Machiavellian-type behaviour to brain size evolution 13.4.3 Extending the Machiavellian intelligence hypothesis to general social intelligence Acknowledgements References

279 280 281 283 283 284

284 285 285 286 286 287 288 289 290 291 291 291

BLBK374-FM

BLBK374-Brown

xii

14

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Contents

Lateralization of Cognitive Functions in Fish Bisazza and Brown

298

14.1 14.2

298 300 300 301 302 303 304 304 304 306 306 307 307 308 308 309

14.3

14.4

14.5

15

June 8, 2011

Introduction Lateralized functions in fish 14.2.1 Antipredator behavior 14.2.1.1 Predator inspection 14.2.1.2 Predator evasion 14.2.1.3 Fast escape response 14.2.2 Mating behavior 14.2.3 Aggression 14.2.4 Shoaling and social recognition 14.2.5 Foraging behavior 14.2.6 Exploration and response to novelty 14.2.7 Homing and spatial abilities 14.2.8 Communication Individual differences in lateralization 14.3.1 Hereditary basis of lateralization 14.3.2 Sex differences in lateralization 14.3.3 Environmental factors influencing development of lateralization 14.3.4 Lateralization and personality Ecological consequences of lateralization of cognitive functions 14.4.1 Selective advantages of cerebral lateralization 14.4.2 Costs of cerebral lateralization 14.4.3 Maintenance of intraspecific variability in the degree of lateralization 14.4.4 Evolutionary significance of population biases in laterality Summary and future research Acknowledgements References

310 311 312 312 314 316 316 317 318 319

Brain and Cognition in Teleost Fish Broglio, G´omez, Dur´an, Salas and Rodr´ıguez

325

15.1 15.2

325 327

15.3

Introduction Classical conditioning 15.2.1 Delay motor classical conditioning and teleost fish cerebellum 15.2.2 Role of the teleost cerebellum and telencephalic pallium in trace motor classical conditioning Emotional learning 15.3.1 Role of the medial pallium in avoidance conditioning and taste aversion learning 15.3.2 Teleost cerebellum and fear conditioning

328 330 331 332 334

BLBK374-FM

BLBK374-Brown

June 8, 2011

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Contents

15.4

15.5

16

Spatial cognition 15.4.1 Allocentric spatial memory representations in teleost fishes 15.4.2 Role of the teleost telencephalon in egocentric and allocentric spatial navigation 15.4.3 Map-like memories and hippocampal pallium in teleost fishes 15.4.4 Neural mechanisms for egocentric spatial orientation Concluding remarks Acknowledgements References

Char Count=

xiii

336 337 340 345 347 349 350 350

Fish Behaviour, Learning, Aquaculture and Fisheries Fern¨o, Huse, Jakobsen, Kristiansen and Nilsson

359

16.1 16.2

359 362 362 362 363

16.3

16.4 16.5 16.6 16.7

Fish learning skills in the human world Fisheries 16.2.1 Spatial dynamics 16.2.1.1 Learning skills and movement 16.2.1.2 Social learning of migration pattern 16.2.1.3 Implications of learning for fisheries management 16.2.2 Fish capture 16.2.2.1 Natural variations in spatial distribution and behaviour 16.2.2.2 Avoidance and attraction before fishing 16.2.2.3 Before physical contact with the gear 16.2.2.4 After physical contact with the gear 16.2.2.5 Behaviour after escaping the gear and long-term consequences 16.2.3 Abundance estimation Aquaculture 16.3.1 Ontogeny 16.3.2 Habituation, conditioning and anticipation 16.3.3 Pavlovian learning – delay and trace conditioning 16.3.4 Potential use of reward conditioning in aquaculture 16.3.5 Operant learning 16.3.6 Individual decisions and collective behaviour Stock enhancement and sea-ranching Escapees from aquaculture Capture-based aquaculture Conclusions and perspectives Acknowledgements References

366 367 369 369 369 371 372 374 375 375 376 378 379 382 383 384 388 389 389 391 391

BLBK374-FM

BLBK374-Brown

xiv

17

June 8, 2011

18:8

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Contents

Cognition and Welfare Sneddon

405

17.1

405 406 407 408 408 409 410 410 411 411 413 413 414 416 417 420 420 421 425 425 426 427 427 429 429

Introduction 17.1.1 Fish welfare 17.1.2 Preference and avoidance testing 17.1.3 Behavioural flexibility and intraspecific variation 17.2 What is welfare? 17.2.1 Sentience and consciousness 17.2.2 Cognition and welfare 17.3 What fishes want 17.3.1 Preference tests 17.3.1.1 Physical habitat 17.3.1.2 Breeding 17.3.1.3 Diet 17.3.1.4 Social interactions 17.4 What fishes do not want 17.5 Pain and fear in fish 17.6 Personality in fish 17.7 Wider implications for the use of fish 17.7.1 Aquaculture 17.7.2 Fisheries 17.7.3 Recreational fishing 17.7.4 Research 17.7.5 Companion fish 17.8 Conclusion Acknowledgements References Species List Index

435 443

BLBK374-FM

BLBK374-Brown

June 7, 2011

11:29

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Preface and Acknowledgements

This is now the second edition of this book, which is a follow-up from our successful volume of Fish and Fisheries dedicated to learning in fishes. All of the contributors to that volume and our previous edition have updated their work in this second edition, and we have added several more contributions covering a broad range of fish behaviour. It is encouraging to see a range of contributions from both established and emerging experts in fish behaviour. The editors would like to thank all of the contributors for their hard work and enthusiasm whilst producing this volume. Such an undertaking would be far too big a task for one person alone, given the increasing volume of behavioural research conducted on fishes. There is also a long list of reviewers whose comments have made valuable contributions to each of the chapters. We would like to thank Nigel Balmforth and his colleagues at Wiley-Blackwell for their valuable support and Tony J. Pitcher for writing the Series Foreword.

BLBK374-FM

BLBK374-Brown

June 8, 2011

10:27

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Series Foreword

Many years ago, teaching a series of practical animal behaviour classes to undergraduates, I asked my students to try to train goldfish to feed on food pellets delivered from colored tubes. It all worked well enough, the goldfish learned to feed from tubes painted with, to our eyes, subtly different colors, and after running the usual controls for light intensity, the students discovered that the fish had very effective color vision. After the classes, the goldfish were returned to a stock aquarium and were left alone for a year, although some of them may have taken part in other experiments. The following year, it was evident right at the start of the student practical that each goldfish remembered the exact color and location of its feeding tube from 1 year before, a remarkable cognitive feat from an animal that is supposed to have only a 3-second memory, as satirized in the recent Pixar Finding Nemo film. Fishermen, anglers, and most of the general public encounter live fish only when they are flopping helplessly, and apparently dumbly, on the boat deck or seashore. In such circumstances, it is hard to believe that fish are intelligent sentient beings: Even in the least speciesist1 science fiction, in Douglas Adams’ (1979) otherwise splendid Hitchhiker’s Guide to the Galaxy for example, fish are merely food for whales (except for one smart automaton, the universal translator known as Babelfish). On the other hand, watching fish hunting for food, engaging with mates, or raising young, aquarists and divers gain a very different view of the behavioural complexities, elegant adaptations and cognitive abilities that lie behind the actions of fish. Fish are endowed with a complex evolved neural and cognitive capacity that reflects the challenges faced by their ancestors, rather than any phylogenetic proximity to humans. This is the scientific reality, and it is a subject of this volume in the Fish and Aquatic Resources series. This is the second edition of the book Fish Cognition and Behavior, which grew out of a 2003 special issue of the Wiley-Blackwell journal Fish and Fisheries (Brown et al. 2003), itself was built upon a pioneering review paper published in the early 1990s (Kieffer & Colgan 1992). In the second edition of this book, we have a set of 17 expanded and updated chapters written by internationally renowned authors that review this important area. The book has been put together and edited by three of the world leaders in this field: Culum Brown from Macquarie University, Australia; Kevin Laland from St Andrews, Scotland; and Jens Krause from Humboldt University, Berlin.

1

“Speciesism” involves assigning different values or rights to beings on the basis of their species. The term was coined by Richard D. Ryder in 1970 and is used to denote prejudice similar in kind to sexism and racism.

BLBK374-FM

BLBK374-Brown

June 8, 2011

10:27

Copyeditor’s Name:

Trim: 244mm X 172mm

Series Foreword

Char Count=

xvii

The editors point out that cognition includes perception, attention, memory formation, and executive functions related to information processing such as learning and problem solving. Fish, it turns out, are not primitive in these respects. Bony fish have had, after all, over 60 million years for their genes to evolve the capacity to build and run fish brains that can deal flexibility with the diverse but volatile underwater environment—a time span ten times longer than our human line. Indeed, the editors of this book show that, far from being primitive automatons as had once been thought, fish “have evolved complex cultural traditions, pursue Machiavellian strategies of manipulation, deception and reconciliation, can monitor the social prestige of others, and can cooperate during foraging, navigation, reproduction and predator avoidance.” This volume presents fascinating, timely, and comprehensive “state of the art” reviews of the cognitive abilities of fish, and readers will find the elements of a fresh synthesis in this field. Therefore, it should find a home on the bookshelves and in the libraries of a broad set of practitioners and students concerned with fish evolution, behaviour, and ecology, including those, like myself, who might still wish to call themselves ichthyologists. Professor Tony J. Pitcher Series Editor: Wiley-Blackwell Fish and Aquatic Resources Series Fisheries Centre, University of British Columbia, Vancouver, Canada

References Adams, D. (1979) The Hitchhiker’s Guide to the Galaxy. Heinemann, London, UK. Brown, C., Laland, K. & Krause, J. (2003) Special issue on learning in fishes: why they are smarter than you think. Fish and Fisheries, 4(3), 197–288. Kieffer, J.D. & Colgan, P.W. (1992) The role of learning in fish behaviour. Reviews in Fish Biology and Fisheries, 2, 125–143.

BLBK374-FM

BLBK374-Brown

June 8, 2011

10:27

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

List of Contributors

Michael S. Alfieri Biology Department Viterbo University 900 Viterbo Drive La Crosse, WI 54601, USA Email: [email protected]

Grant E. Brown Department of Biology Concordia University 7141 Sherbrooke St., W. Montreal Quebec, H4B 1R6, Canada Email: [email protected]

Angelo Bisazza Comparative Psychology Research Group University of Padova Padova, Italy Email: [email protected]

Redouan Bshary Universit´e de Neuchˆatel, Rue Emile-Argand 11 CH-2007 Neuchˆatel, Switzerland Email: [email protected]

Cristina Broglio Laboratorio de Psicobiologia Universidad de Sevilla c/ Camilo Jose Cela s/n, 41018 Sevilla, Spain Email: [email protected]

Sergey Budaev Severtsov Institute of Ecology and Evolution Russian Academy of Sciences Leninsky prospect 33 Moscow 119071, Russia Email: [email protected]

Victoria A. Braithwaite School of Forest Resources and Department of Biology Pennsylvania State University University Park PA 16802, USA Email: [email protected]

Douglas P. Chivers Department of Biology University of Saskatchewan Saskatoon, Saskatchewan SK S7N 5E2, Canada Email: [email protected]

Culum Brown Department of Biological Sciences Macquarie University Sydney 2109, Australia Email: [email protected]

Iain D. Couzin Department of Ecology and Evolution Princeton University Princeton, NJ 08544-2016, USA Email: [email protected]

BLBK374-FM

BLBK374-Brown

xx

June 8, 2011

10:27

Copyeditor’s Name:

Trim: 244mm X 172mm

List of Contributors

Darren P. Croft School of Psychology Exeter University Perry Road Exeter, EX4 4QG, UK Email: [email protected] Lee A. Dugatkin Department of Biology University of Louisville Louisville, KY 40208, USA Email: [email protected] Emilio Dur´an Laboratorio de Psicobiologia Universidad de Sevilla c/ Camilo Jose Cela s/n, 41018 Sevilla, Spain Email: [email protected] Ryan L. Earley Department of Biological Sciences University of Alabama Box 870344 Tuscaloosa, Alabama 35487, USA Email: [email protected]

Siˆan W. Griffiths Cardiff School of Biosciences PO Box 915, Cardiff Wales, CF10 3TL, UK Email: [email protected] Yuying Hsu Department of Life Science National Taiwan Normal University No. 88, Section 4, Ting-Chou Road Taipei 116, Taiwan Email: [email protected] Roger Hughes School of Biological Sciences Environment Centre University of Wales, Bangor Gwynedd, LL57 2UW, UK Email: [email protected] Geir Huse Institute of Marine Research PO Box 1870-Nordnes N-5817 Bergen, Norway Email: [email protected]

Anders Fern¨o Department of Biology University of Bergen PO Box 7800 N-5020 Bergen, Norway Email: [email protected]

Christos C. Ioannou Department of Ecology and Evolution Princeton University Princeton, NJ 08544-2016, USA Email: [email protected]

Maud C.O. Ferrari Department of Biomedical Sciences WCVM, University of Saskatchewan Saskatoon, SK, Canada Email: [email protected]

Per Johan Jakobsen Department of Biology University of Bergen PO Box 7800 N-5020 Bergen, Norway Email: [email protected]

Antonia G´omez Laboratorio de Psicobiologia Universidad de Sevilla c/ Camilo Jose Cela s/n, 41018 Sevilla, Spain Email: [email protected]

Richard James Department of Physics University of Bath Bath, BA2 7AY, UK Email: [email protected]

Char Count=

BLBK374-FM

BLBK374-Brown

June 8, 2011

10:27

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

List of Contributors

Jens Krause Department of Biology and Ecology of Fishes Leibniz-Institute of Freshwater Ecology and Inland Fisheries Berlin, Germany Email: [email protected] Jennifer L. Kelley Centre for Evolutionary Biology School of Animal Biology The University of Western Australia Nedlands, WA 6009, Australia Email: [email protected] Tore S. Kristiansen Institute of Marine Research PO Box 1870-Nordnes N-5817 Bergen, Norway Email: [email protected] Kevin Laland Centre for Social Learning and Cognitive Evolution School of Biology University of St Andrews St Andrews, Fife, KY16 9TS, Scotland Email: [email protected] Anne E. Magurran Gatty Marine Laboratory University of St Andrews St Andrews, Fife, KY16 8LB, Scotland Email: [email protected] Jonatan Nilsson Institute of Marine Research PO Box 1870-Nordnes N-5817 Bergen, Norway Email: [email protected] Sabine N¨obel Department of Biology University of Siegen Adolf-Reichwein-Str. 2 D-57068 Siegen, Germany Email: [email protected]

Lucy Odling-Smee Nature Publishing Group 4 Crinan Street, London N1 9XW, UK Email: [email protected] Tony Pitcher Fisheries Centre The University of British Columbia Vancouver, BC, V6T 1Z4, Canada Email: [email protected] Fernando Rodr´ıguez Laboratorio de Psicobiologia Universidad de Sevilla c/ Camilo Jose Cela s/n, 41018 Sevilla, Spain Email: [email protected] Cosme Salas Laboratorio de Psicobiologia Universidad de Sevilla c/ Camilo Jose Cela s/n, 41018 Sevilla, Spain Email: [email protected] Lynne U. Sneddon Integrative Biology University of Liverpool Crown Street Liverpool, L69 7ZB, UK Email: [email protected] Kevin Warburton School of Environmental Sciences Faculty of Science Charles Sturt University Thurgoona, New South Wales, Australia Email: [email protected]. Ashley Ward School of Biological Sciences University of Sydney New South Wales, Australia Email: [email protected]

xxi

BLBK374-FM

BLBK374-Brown

xxii

June 8, 2011

10:27

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

List of Contributors

Klaudia Witte Department of Biology University of Siegen Adolf-Reichwein-Str. 2 D-57068 Siegen, Germany Email: [email protected]

Larry L. Wolf Department of Biology Syracuse University 107 College Place, Life Sciences Complex Syracuse, New York, 13244, USA Email: [email protected]

BLBK374-01

BLBK374-Brown

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 1

Fish Cognition and Behaviour Culum Brown, Kevin Laland and Jens Krause

1.1

Introduction

The field of animal cognition is the modern approach to understanding the mental capabilities of animals. The theories are largely an extension of early comparative psychology with a strong influence of behavioural ecology and ethology. Cognition has been variously defined in the literature. Some researchers confine cognition to higher order mental functions including awareness, reasoning and consciousness. However, a more general definition of cognition also includes perception, attention, memory formation and executive functions related to information processing such as learning and problem solving. The study of animal cognition has been largely confined to birds and mammals, particularly non-human primates. This bias in the literature is in part due to the approach taken in the 1950s when cognitive psychologists began to compare known human mental processes with other closely related species. This bias was reinforced by an underlying misconception that learning played little or no role in the development of behaviour in reptiles and fishes. Throughout scientific history fishes have largely been viewed as automatons. Their behaviour was thought to be almost exclusively controlled by unlearned predispositions. Ethologists characterised their behaviour as a series of fixed action patterns released on exposure to appropriate environmental cues (sign stimuli). Whilst there is no doubt that fishes are the most ancient form of vertebrates, they are only ‘primitive’ in the sense that they have been on earth for in excess of 500 million years and that all other vertebrates evolved from some common fish-like ancestor (around 360 million years ago). However, it is important to note that fishes have not been stuck in an evolutionary quagmire during this time. Their form and function have not remained stagnant over the ages. On the contrary, within this time frame they have diversified immensely to the point where there are more species of fish than all other vertebrates combined (currently over 32,000 described species) occupying nearly every imaginable aquatic niche. The erroneous view that both behavioural and neural sophistications are associated in a linear progression from fishes through reptiles and birds to mammals is largely due to a heady mix of outdated and unscientific thinking. Aristotle’s concept of Scala naturae

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-01

BLBK374-Brown

2

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(the scale of nature) and a Christian fundamentalist view that man is the pinnacle of the natural world have dominated conceptions of animal intelligence for millennia. However, Darwin’s theory of evolution is fundamentally inconsistent with a gradual progression of behavioural flexibility and cognitive complexity from ‘primitive’ to ‘advanced’ life forms, leading inevitably to humans at the peak (i.e. the wrong-headed notion of an evolutionary ladder). There is nothing progressive about Darwinian evolution, and any semblance of progression merely reflects our anthropocentric bias to track evolutionary lineages that culminate in our species, and to evaluate other species by their similarity to ourselves. The cognitive capabilities of a species will reflect the history of selection amongst its ancestors, rather than phylogenetic proximity to humanity. Amongst the vertebrates, fishes have suffered the most from the common misconception of the evolutionary ladder. However, over the last few decades this fallacy has begun to be redressed. Researchers now realised that, like the rest of the vertebrate kingdom, fishes exhibit a rich array of sophisticated behaviour and that learning plays a pivotal role in behavioural development of fishes. Gone, or at least redundant, are the days where fishes were looked down upon as pea-brained machines whose only behavioural flexibility was severely curtailed by their infamous 3-second memory (`a la Dory in Disney’s Finding Nemo). As this book will reveal, many fishes in fact have impressive long-term memories comparable to most other vertebrates (Brown 2001; Warburton 2003). Their neural architecture has both analogous and homologous components with mammals, and is capable of much the same processing power (Broglio et al. 2003). Their cognitive capacity in many domains is comparable with that of non-human primates (Bshary et al. 2002; Laland & Hopitt 2003; Odling-Smee & Braithwaite 2003). Fishes have evolved complex cultural traditions and pursue Machiavallian strategies of manipulation, deception and reconciliation (Bshary et al. 2002; Brown & Laland 2003). They not only recognise one another, but can monitor the social prestige of and dominance relations amongst others (McGregor 1993; Griffiths 2003; Grosenick et al. 2007) and cooperate in a variety of ways during foraging, navigation, reproduction and predator avoidance (Huntingford et al. 1994; Johnstone & Bshary 2004; Fitzpatrick et al. 2006). It is clear that the recent developments in our understanding of fish behaviour require a substantial reappraisal of their behavioural flexibility that warrants further investigation. Since the 1960s there has been a rapid increase in the number of papers published on learning in fishes and those published since 1991 has risen dramatically (Fig 1.1). In the early 1990s James Kieffer and Patrick Coglan published the first comprehensive review of the role of learning in the development of fish behaviour (Kieffer & Colgan 1992). In their review, they were able to draw on some 70 published papers on learning in fishes, a vast improvement over previous works (Thorpe 1963; Gleitman & Rozin 1971). In 2003, we published a collection of reviews on the topic in a special issue of the journal Fish and Fisheries. The special issue contained eight reviews on various aspects of learning in fishes referring to over 500 papers. In the first edition of this book, which contained 14 chapters, many of these reviews were revised and extended. The second edition has been significantly expanded again, with revision of most chapters and the inclusion of three more chapters on laterality, personality and welfare consequences of cognition. This new edition now examines the role of cognition in every major aspect of fish biology, from foraging and predator avoidance to fighting and social relationships.

BLBK374-Brown

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behaviour

3

60 50 Number of publications

40

30 20 10 0 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10

BLBK374-01

Year

Fig. 1.1 The number of publications on fish learning and cognition since 1991 has increased substantially. Data based on key word search (fish, fishes, learn, learning and cognition) in Web of Science.

1.2

Contents of this book

Apart from this opening introduction, Chapter 2, by Kevin Warburton and Roger Hughes, investigates the role of learning in foraging behaviour, drawing on both psychological and behavioural ecology literature. They suggest that learning and memory play significant roles in the foraging activities in fish and that memory, like many traits, seems to be highly adapted to the specific requirements of each species. Interestingly, they suggest that in some circumstances forgetting might be just as important as remembering. The chapter highlights that the similarities between vertebrate learning systems are far more striking than the differences and fishes rely on a wide array of learning mechanisms in their daily lives. The literature shows that learning is vital in many aspects of fish foraging behaviour, from the formation of foraging search images, to prey capture and handling. Warburton and Hughes also outline various experiments that explore foraging theory and point out that fishes are frequently ideal candidates for such research. It is often assumed that anti-predator behaviour should have a significant unlearned component to it because fishes need to be able to escape predators from the moment they hatch. The penalty for failure in this instance is death, so there is an expectation that natural selection will exert significant evolutionary pressure in this respect. Jennifer Kelley and Anne Magurran point out in Chapter 3 that while this is the case to some degree, learning still plays a key role in the fine-tuning of predator recognition and response systems. In environments that are unpredictable from moment to moment and, perhaps more importantly, from generation to generation, it is essential that prey species have some general template for predator recognition, but that this template be flexible enough to enable fine-tuning to match the prevailing predatory threats. Kelly and Magurran discuss the various ways in which fishes learn about predators and the need for prey species to be

BLBK374-01

BLBK374-Brown

4

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

able to accurately assess potential risks and act accordingly. They cover the evolutionary arms race between predators and prey highlighting the role learning plays in this race from both perspectives. There are many ways in which prey can learn about predators without high-risk exposure, including the observation of conspecifics as they interact with, or detect, predators. One such method is the reliance on predator odours and prey alarm cues that may be detected from some distance and this is the focus of Chapter 4. Here, Grant Brown, Maud C.O. Ferrari and Douglas P. Chivers explore how fishes use chemical cues both to assess risk and to learn about predators. There are obviously great fitness advantages to be had by the accurate assessment of risk, primarily because it frees the individual time budget from unnecessary anti-predator behaviour (Lima & Dill 1990). Fishes not only learn from conspecifics but may also respond to the alarm signals generated by heterospecifics that are part of the same prey guild, thus enabling the recognition of predators and dangerous habitats alike. It is interesting to note that fishes often undergo massive growth from larval to adult stages and in doing so pass through a series of predatory guilds each with its own specific threats. In this scenario, Brown, Ferrari and Chivers point out that learning may play a larger role in the development of anti-predator behaviour than previously suspected. In Chapter 5, Klaudia Witte and Sabine N¨obel explore the role of learning in matechoice decisions. In their review, Witte and N¨obel examine the evidence for the influence of imprinting during the critical period of early life-history stages on later mate-choice decisions. They reveal that imprinting is most likely to occur in those species that show some kind of extended parental care, such as the cichlids. However, it is also evident that other social influences can also affect mate-choice decisions later in life. For example, naive male guppies can learn to discriminate between conspecifics and heterospecifics and alter their mating strategy to concentrate on courting conspecifics. Part of this alteration in behaviour may be mediated by their mating success and feedback from the females they are attempting to court. Species recognition may be reinforced by learning in those areas where multiple closely related species coexist. Whilst mate choice often relies on some predetermined innate recognition and preference system, Witte and N¨obel reveal that these unlearned preferences can be overcome by learning and especially by copying the mate-choice decisions of others. As discussed in many of the chapters, fishes are capable of relying on a mixture of eavesdropping and social information to help them make important decisions, and mate choice is no exception. Reliance on public information may enable females to gauge the quality and aggression levels of a potential mate without having to suffer any negative consequences associated with the early stages of courtship. Yuying Hsu, Ryan L. Earley and Larry L. Wolf examine the modulation of aggression through prior experience in Chapter 6. Many factors combine to influence the outcome of aggressive encounters, including size, motivation, prior residency and, as Hsu and his colleagues highlight, prior experience with fights can also play a large role. The outcome of fights can have considerable consequences including access to food, mates or territories, so it is important to understand how experience can influence the outcomes of fights. Recent literature suggests that fishes that have recently lost a fight are more likely to lose a second encounter compared to winners, all else being equal. Therefore, an individual’s history must be considered when predicting the outcome of a fight at the present time. All of us know that confidence can influence our behaviour considerably and this is likely to be mediated both

BLBK374-01

BLBK374-Brown

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behaviour

5

through physiological as well as psychological mechanisms. Relying on both modelling and empirical data, Hsu et al. explore how previous experience combines or interacts to shape an individual’s present fighting capability. Whilst Darwin and his immediate successors described animals as having personalities, this was characterised as anthropomorphic and fell out of favour, as a result of which, until fairly recently, discussions on animal personality have been something of a taboo. Perhaps there was a superficial acceptance that domestic animals such as dogs could have personality traits, but fishes? In Chapter 7, Sergey V. Budaev and Culum Brown explore the recent explosion in animal personality literature in which fishes have played a leading role. Owing in part to this fear of anthropomorphism, the literature relating to fish personality has been heavily fragmented with the adoption of alternative synonyms such as ‘coping style’, ‘behavioural syndrome’ and ‘boldness–shyness continuum’. This chapter represents the first attempt to bring these streams of research together. The authors examine both proximate and ultimate explanations of fish personality. Budaev and Brown conclude that personality traits play a neglected role in evolution since individual variation is the bread and butter of natural selection. Personality traits not only are heritable but also have fitness consequences. The authors claim that examination of personality traits in fishes requires a holistic view of behaviour in which multiple traits may be correlated with one another across a range of contexts, and warn against too narrow a view that misses important relationships that constrain behavioural evolution. In Chapter 8, Victoria Braithwaite and Lucy Odling-Smee explore the role of cognition in spatial orientation, navigation and migration. The authors point out that, like most animals, the resources fish utilise are often widely separated in space. Many of these biologically important locations are relatively temporally and spatially stable and as such can be reliably found by learning and memory retrieval. As Warburton and Hughes pointed out in Chapter 2, here it is also the case that natural selection has favoured learning strategies to closely match the needs of the species under consideration. Like in all animals, cue reliance is constrained by the species’ perception, and fishes display a huge array of perceptual capabilities, many of which are only just beginning to be understood, such as electroreception and UV vision. It is evident that fishes rely on a wide array of navigation cues and mechanisms, ranging from egocentric turns to the formation of cognitive maps, to move accurately around their environments. Natural selection would favour the ability to select the most efficient movement pathways possible so as to reduce any potential waste of time and energy. Thus, accurate navigation is a key component to an individual’s fitness landscape. In the final part of their chapter, Braithwaite and Odling-Smee concentrate on large-scale migration in salmon as a case study, highlighting both the recall of long-term memory and initial imprinting processes. Sian Griffiths and Ashley Ward review the evidence for individual recognition in Chapter 9. When closely examining social interactions, it is apparent that not all individuals are treated equally by a given fish. For example, as discussed by Hsu et al. in Chapter 9, closely related fishes often receive less aggression than non-relatives. Individual recognition has several implications on multiple levels, including predicting species dispersal patterns, which has conservation and fisheries management outcomes. But how do fishes recognise one another? Griffiths and Ward review the ever-increasing body of publications that fishes not only recognise kin, but they can also distinguish between familiar

BLBK374-01

BLBK374-Brown

6

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

and unfamiliar individuals. This process seems to build up over 10–14 days although it may vary from species to species. Being able to recognise, and preferentially associate with kin or familiar individuals, potentially has substantial direct and indirect fitness benefits. For example, there is evidence that shoals comprised of familiar individuals show more efficient schooling behaviour than those comprised of strangers. Such benefits may accrue due to an increase in an individual fish’s ability to predict the response of familiar individuals across a variety of contexts. Individual recognition is germane to other aspects of fish behaviour, including cooperation (Chapter 12), exploitation of social cues and signals (Machiavellian Intelligence in Fishes; Chapter 13) and social learning (Chapter 11). In Chapter 10, Christos Ioannou, Iain Couzin, Richard James, Darren Croft and Jens Krause develop mathematical approaches and review current literature that links the behaviour of individuals to the higher order properties at the group and population levels. It is evident that the behaviour of individuals within a social group is largely influenced by their fellow group members. Through the rapid transfer of information between group members, shoals of fish often seem to behave as a single collective. However, a few individuals within a group can assert undue influence on the behaviour of the majority, particularly if these ‘leaders’ are more motivated to perform some behaviour than the remainder of the shoal (i.e. they are more directed than the average). Such processes may have significant impact on the three-dimensional structure and movement of shoals. Moreover, because information is shared between group members, a shoal as a whole may be able to solve problems more efficiently than singletons (e.g. navigation), for example, by filtering environmental noise or collective detection and processing of external cues. In addition, examination of association networks by Ioannou et al. can be utilised to predict the path through which information is likely to be transferred within the group. The transfer of information between individuals is reliant on social learning processes. Social learning refers to those situations where individuals acquire new information or behaviour by observation of, or interaction with, others. Social learning can occur across a wide variety of contexts and appears to be a ubiquitous form of learning within fishes. Social learning often enables individuals to acquire information more rapidly and efficiently than would be the case if they themselves had to explore their environment fully and learn via trial and error. Traditionally, social learning was thought to be restricted to mammals and birds, but in Chapter 11, Culum Brown and Kevin Laland explore the substantive body of evidence showing the widespread existence of social learning in fishes. Social learning that occurs across generations (vertical or oblique transmission) can lead to the establishment of localised, stable behavioural traditions that form the very roots of animal culture. Such cultural evolution can operate in tandem with biological evolution and these processes interact in many interesting ways. Brown and Laland argue that social learning is likely to play a key role in the development of fish behaviour and point out that exploitation of such processes could be utilised in training regimes for fisheries and in conservation management programmes such as restocking. Cooperation between individuals has long been considered something of an enigma within evolutionary biology. If Darwinian fitness is all about out-competing others then one might think all individuals ought to behave selfishly. This notion is central to many existing theories such as the selfish herd hypothesis which is particularly pertinent to group-living animals such as fishes. However, it became clear that the evolution of cooperation could

BLBK374-01

BLBK374-Brown

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behaviour

7

be explained through a number of alternative hypothesis, namely kin selection, reciprocity, by-product mutualism and group selection, in which individuals gain long-term fitness benefits in spite of short-term costs. In Chapter 12, Michael Alfieri and Lee A. Dugatkin suggest that cooperation not only occurs in fishes but may also be widespread in a number of contexts. Redouan Bshary, in Chapter 13, continues the social theme by examining the evidence for social or Machiavellian intelligence in fishes, largely stemming from his early observations of the behaviour of cleaning wrasse. Here, he extends his earlier review (Bshary et al. 2002) on the topic and presents an overview on the social strategic cognitive abilities of fishes. The primary thesis of the Machiavellian intelligence hypothesis (Whiten & Byrne 1997) is that one of the principal driving forces for the evolution of cognition was the challenge to cope with and exploit the complexity of an individual’s social environment. For years the hypothesis was used almost exclusively to ‘prop-up’ the apparent existence of the higher cognitive capacity of primates including humanity. However, it soon became apparent that the theory, if true, should apply equally to other vertebrate groups. Bshary provides evidence for individual recognition, individualised group living, cooperation, manipulation, reconciliation and deception in various fishes. The second new addition to this edition examines the role of laterality on fish behaviour. Like all vertebrates, fishes show very strong left–right biases in a range of behaviour patterns which are generated by the preferential processing of information in either hemisphere of the brain. In fishes, lateralization of cognitive function is overtly displayed by such things as eye preferences whilst viewing particular scenes, objects or turn biases during startle responses. For example, many species prefer to view predators with one eye and familiar conspecifics with the other. In Chapter 14, Angelo Bisazza and Culum Brown summarise the extensive literature on fish laterality. The authors first discuss proximate causes of laterality, ranging from brain formation to genetic heritability, and then address the ultimate consequences by examining the costs and benefits of laterality in the context of evolutionary ecology. The fact that many species of fishes are lateralized at the population level (much like 90% of humans are right-handed) begs an intriguing question regarding the evolution of laterality in group-living species. In schooling species, for example, we often find that fish that are strongly left-eye-biased in social contexts will take up positions on the right side of the school so they can monitor the behaviour of conspecifics with their preferred eye and vice versa for right-biased fishes. It seems likely that laterality is under frequency-dependent selection in some species whilst in other species key environmental variables, such as the level of predation, likely shape the trait. For decades the cognitive ability of fishes was highly underrated, largely due to a lack of direct experimentation. However, an additional factor here was a reliance on direct comparisons of the fish brain with that of mammals, in which the majority of studies on cognition had occurred (particularly primates and rodents), and for which a great deal was known about the function and connectivity of brain structures. Such comparison suggested that the brains of fishes and mammals differ in many ways, with fish brains typically smaller and less structured than those of mammals. Because of this it was often indirectly inferred that fishes must lack certain cognitive abilities observed in mammals because their brain structure was not the same as mammals. Not until very recently have scientists begun to study the brains of fishes and their function in any detail. It should be pointed out that

BLBK374-01

BLBK374-Brown

8

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

virtually nothing is known about the vast majority of fish species, let alone anything about their brain structure and function. The results of these pioneering studies, as Fernando Rodriguez and his laboratory members (Christina Broglio, Emilio Duran, Antonio Gomez and Cosme Salas) highlight in Chapter 15, are startling. They reveal many similarities between the mammalian and fish brains in terms of their functions. In the light of recent developmental, neuroanatomical and functional data, it appears that many functions are highly conserved right across vertebrates, despite the fact that morphology can often differ substantially. Rodriguez et al. point out that these morphological differences stem from an entirely different developmental pathway. For example, the fish telencephalon goes through a process of eversion (bending out) during embryonic development whereas the brain of the rest of the vertebrates develops by evagination (bending in). Through the results of their extensive research and through the review of related literature, Rodriguez et al. challenge the prevailing notion that fishes lack most of the brain centres and neural circuits that support cognition capabilities in the other vertebrate groups. Chapter 16 examines just one of the many practical applications that can stem from a greater understanding of fish cognition and behaviour. Here, Anders Fern¨o and colleagues explore the role of fish learning in aquaculture and fisheries. For thousands of years humans have relied on a steady harvest of fish from the rivers and oceans as an important source of protein. Today fishes remain the only wild food source humans harvest and through a greater understanding of their behaviour we have begun to farm them and exploit natural populations at an ever-growing rate. However, as Fern¨o et al. point out, human fishing methods have evolved at a far greater rate than the fishes’ response to this selection pressure and there is now a huge gap between fishes’ natural responses to predators and our modern fishing techniques. However, fish can respond to the threat of fishing through learning. There is now some evidence that fishes learn to respond to fishing gear, largely by avoidance of vessels, and such responses may interfere with our estimates of stock sizes. Fishing may also affect fish learning. For instance, removal of larger, more knowledgeable individuals from stocks may disrupt social transmission chains, thus breaking long-standing cultural traditions in some of the economically most important fish species (e.g. the location of feeding, migration routes or breeding grounds). Following the crash of the Northern cod stocks, for example, an abrupt change was realised in the stocks distribution. Fern¨o et al. also investigate the ways in which behavioural flexibility can be utilised in aquaculture scenarios. They highlight the fact that due consideration must be given to the large influence of early experience in the development of fish behaviour when managing hatchery stocks, particularly in those instances where the stocks are used for conservation reintroductions or to buffer existing natural stocks from the pressures of commercial and recreational fisheries. The final chapter, Fish Cognition: Implications for Fish Welfare, in this collection of reviews examines the implications of fish cognition for fish welfare. With our ever-growing appreciation of fish cognition we undoubtedly have a moral duty to address the potential welfare considerations of fish in a variety of contexts. The most obvious consequences will be felt in the fisheries and aquaculture industries, but think also of the very large aquarium trade. Fishes are third only to cats and dogs as the most popular pets in the world. In 2005, two Italian cities banned the use of small fish bowls for keeping fish on welfare ground. Moreover, fishes are widely used as experimental animals in scientific experiments. In Chapter 17, Lynne Sneddon summarises the key facts about cognition in fishes that until

BLBK374-01

BLBK374-Brown

May 11, 2011

13:48

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behaviour

9

recently have largely clouded the issue of whether or not fish deserved to be treated in the same way as other vertebrates. Two of the most important issues are ‘do fish feel pain?’ and ‘how do we measure the needs of fish and assess their welfare?’ Finally, she draws conclusions regarding the implications of our knowledge about fish cognition for industry and society as a whole. Two themes emerge in this book. The first is that the learning abilities and complexity of behaviour of fishes are, in many respects, comparable to land vertebrates. The second is that fish provide a flexible and pragmatic biological model system for studying many aspects of animal learning and cognition. These observations lead us to the view that interest in the topic of fish learning, cognition and behaviour is likely to continue to grow for the foreseeable future.

References Broglio, C., Rodriguez, F. & Salas, C. (2003) Spatial cognition and its neural basis in teleost fishes. Fish and Fisheries, 4, 247–255. Brown, C. (2001) Familiarity with the test environment improves escape responses in the crimson spotted rainbowfish, Melanotaenia duboulayi. Animal Cognition, 4, 109–113. Brown, C. & Laland, K. (2003) Social learning in fishes: a review. Fish and Fisheries, 4, 280–288. Bshary, R., Wickler, W. & Fricke, H. (2002) Fish cognition: a primate’s eye view. Animal Cognition, 5, 1–13. Fitzpatrick, J.L., Desjardins, J.K., Stiver, K.A., Montgomerie, R. & Balshine, S. (2006) Male reproductive suppression in the cooperatively breeding fish Neolamprologus pulcher. Behavioral Ecology, 17, 25–33. Gleitman, H. & Rozin, P. (1971) Learning and memory. In: W.S Hoar & D.J. Randall (eds) Fish Physiology, 6, pp. 191–278. Academic Press, New York. Griffiths, S.W. (2003) Learned recognition of conspecifics by fishes. Fish and Fisheries, 4, 256–268. Grosenick, L., Clement, T.S. & Fernald, R.D. (2007) Fish can infer social rank by observation alone. Nature, 445, 429–432. Huntingford, F.A., Lazarus, J., Barrie, B.D. & Webb, S. (1994) A dynamic analysis of cooperative predator inspection in sticklebacks. Animal Behaviour, 47, 413–423. Johnstone, R.A. & Bshary, R. (2004) Evolution of spite through indirect reciprocity. Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 1917–1922. Kieffer, J.D. & Colgan, P.W. (1992) The role of learning in fish behaviour. Reviews in Fish Biology and Fisheries, 2, 125–143. Laland, K.N. & Hoppitt, W. (2003) Do animals have culture? Evolutionary Anthropology, 12, 150–159. Lima, S.L. & Dill, L.M. (1990) Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology, 68, 619–640. McGregor, P.K. (1993) Signaling in territorial systems – a context for individual identification, ranging and eavesdropping. Philosophical Transactions of the Royal Society of London Series B – Biological Sciences, 340, 237–244. Odling-Smee, L. & Braithwaite, V.A. (2003) The role of learning in fish orientation. Fish and Fisheries, 4, 235–246. Thorpe, W.H. (1963) Learning and Instinct in Animals. Methuen, London. Warburton, K. (2003) Learning of foraging skills by fish. Fish and Fisheries, 4, 203–215. Wilson, E.O. (1975) Sociobiology: The New Synthesis. Harvard University Press, Cambridge, MA. Whiten, A. & Byrne, R.W. (1997) Machiavellian Intelligence II: Extensions and Evaluations. Cambridge University Press, Cambridge.

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 2

Learning of Foraging Skills by Fish Kevin Warburton and Roger Hughes

2.1

Introduction

Investigations of the role of learning and memory in the foraging behaviour of fishes are at an exciting stage. Although less thoroughly studied than traditional laboratory favourites such as rats, pigeons or bees, fish species are no longer to be consigned to the ‘poorly understood’ category. It is now possible to place the capacities of fish in the context of learning and memory as a whole, as evidenced by previous reviews such as those by Hart (1986, 1993), Hughes et al. (1992) and Kieffer & Colgan (1992). The present chapter attempts to integrate perspectives and findings from the fields of behavioural ecology and comparative psychology. This approach has been adopted for the following reasons:

1. Comparative psychology has revealed broad regularities in the general principles of learning across invertebrate and vertebrate taxa (Logue 1988; Domjan 1998) and across spatial and temporal domains (Cheng & Spetch 2001). The general principles that apply to learning in bees, pigeons and rats are likely to apply to fishes also. Psychology can clarify the mechanisms that underlie observed behaviour, while behavioural ecology can evaluate the adaptive significance of behavioural capacities demonstrated by psychology. 2. In several cases, static first-generation models based on optimal foraging theory (OFT) that do not represent temporal changes in internal state fail to predict observed behaviour (Hart 1993). More recent studies using more flexible (e.g. dynamic-programming) models have been more successful because of their ability to represent changes in internal state (e.g. physiology and learning), interactions between intrinsic and extrinsic variables and continuous behavioural adjustment in response to these factors (Ehlinger 1989; Kieffer & Colgan 1991; Hart 1993; Dall et al. 1999). Although standard OFT models predict that animals should exhibit all-or-nothing choice, experiments usually reveal partial preferences. It is likely that future models will draw increasingly on psychological effects (e.g. discrimination, memory, cue competition, interference and attention) to explain such divergences from the predictions of simple or idealised models and to test models based on risk and information (Shettleworth 1988). Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

FORAGING PERFORMANCE

• • • •

11

Spatial environment Patch profitability Prey availability Predation risk

LEARNING

Search speed Patch location Prey capture Handling skill

Simple association (stimulus substitution)

Cognition (active anticipation)

Attention

Holistic awareness

Exploration/ sampling

Motivation

Preparedness Drive/ deprivation

Stimulus attractiveness/ incentive

Cue competition

Reinforcement Hunger Stress Threat Competition

Contiguity Frequency Intensity

Fig. 2.1 Fish foraging: Suggested relationships between foraging performance, learning and contributory factors. Bold arrows indicate main influences.

A suggested conceptual framework for fish foraging is outlined in Fig. 2.1. The main intention of this framework is to highlight how different contributory factors combine to affect foraging performance. Intrinsic and extrinsic (stimulus-related) factors affect motivation, which in turn influences the attention that is directed to relevant stimuli and the willingness to explore the general environment in which such stimuli occur. Holistic environmental awareness is a key prerequisite for cognitive appraisal, while attention plays an important role in the formation of simple stimulus associations and in the development of foraging skills. Both association-formation and cognition contribute to learning. Such associative and cognitive information processing enhances the development of physical skills and thus improves foraging performance. This logic forms the basis for the organisation of

BLBK374-02

BLBK374-Brown

12

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

the first part of the present chapter. In later sections, the adaptive significance of learning and memory is considered in terms of the resulting ability of fish to track environmental variation and improve competitive foraging success.

2.2 2.2.1

Some factors affecting the learning process Reinforcement

Reinforcement is the increase in response probability following a stimulus event. Reinforcement appears to affect learning mainly by influencing what is learned (rather than how or how well it is learned) and appears to have a greater effect on motivation than on learning (Crespi 1942; Lieberman 1990). As in other animals, associative learning by fish is strongly influenced by the frequency and intensity of reinforcement and the spatial and temporal contiguity of events. In goldfish, Carassius auratus (Cyprinidae), the formation of learned associations between new stimuli (e.g. visual cues) and rewards (e.g. food) occurs more efficiently when the delay between the stimulus and the reward is short (Breuning et al. 1981).

2.2.2

Drive

While the explanatory value of the term ‘drive’ is the subject of continuing debate, in the present context it seems useful to recognise drive (deprivation level) and stimulus attractiveness (incentive value) as two distinct components of motivation (Lieberman 1990). When fishes are hungry they are less distracted by other pressures (Milinski 1993), which enhances foraging learning. In social learning by guppies, Poecilia reticulata (Poeciliidae), food deprivation has a stronger effect on male than on female performance (Reader & Laland 2000). Isolation stress (as caused, for example, by predatory attacks) interferes with memory formation (goldfish; Laudien et al. 1986). Fish may learn feeding skills more slowly when isolated than when in shoals (Jain & Sahai 1989), presumably as a result of enhanced perceived threat and/or a lack of opportunities for social learning. Similarly, foraging motivation is seriously reduced or abolished entirely by predator threat. There is evidence that fish continually adjust their behaviour in accordance with a risk-balancing forage-refuge trade-off (Mittelbach 1981; Pitcher et al. 1988; Gotceitas & Colgan 1990). Thus, three-spined sticklebacks, Gasterosteus aculeatus (Gasterosteidae), learn to feed in profitable patches but abandon them when they are threatened (Huntingford & Wright 1989). Other factors also influence risk assessment; for example, after being alarmed, parasitised sticklebacks (G. aculeatus) resume feeding more quickly than non-parasitised fish (Giles 1983), and brown trout (Salmo trutta) in familiar groups resume feeding more quickly after a simulated predator attack than do fish in unfamiliar groups (Griffiths et al. 2003). These examples demonstrate that feeding drive is affected by a complex interaction between hunger, social context, threat and physiological burden.

2.2.3

Stimulus attractiveness

Preparedness, which is the tendency to associate some stimulus combinations more readily than others, varies across taxa as a consequence of evolutionary selection for specialised

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

13

Fig. 2.2 Goldfish can be readily trained to forage for flake food hidden under surface gravel.

sensory, receptor and associative apparatus (Seligman 1970). Responses must be chosen from the adaptive, species-specific repertoire of the organism (Seligman 1970), which explains why rats can be readily trained to jump or run, but not press a bar, to escape from a dangerous place (D´amato & Schiff 1964). Populations of cells in the primate amygdala respond selectively to faces (Desimone 1991; Dahl et al. 2009). In fish, visual landmarks can vary in salience; for example, goldfish (Fig. 2.2) form feeding associations more readily with tall, coloured columns than with weed/plant combinations, and with vertically striped screens more readily than with horizontally striped screens (Warburton 1990; K. Warburton’s unpublished observations). Patterns of retinal topography in fish have been proposed to reflect habitat variation (Collin & Pettigrew 1988a, 1988b), and in this connection, it is interesting that goldfish have vertically oriented retinal cell patterns (Mednick & Springer 1988) and favour ‘vertical’ vegetated habitats (McDowall 1996). Also noteworthy here is the fact that, in mammals, rearing in a striped environment affects the development of the visual cortex such that more surface area becomes devoted to the experienced orientation than to the orthogonal one (Sengpiel et al. 1999). Generalisation occurs when a conditioned response is elicited to stimuli that are similar to a conditioned stimulus. Learning in fish is expected to conform to Shepard’s law of generalisation (Shepard 1987, 1988), in that fish should respond to stimuli that are similar to those involved in pre-existing associations. Generalisation gradients are typically negatively exponential (i.e. the response falls off rapidly as the similarity to the original stimulus declines) if the relevant stimuli are clearly distinguishable, but are more likely to be Gaussian if the stimuli are not distinguishable (Shepard 1988). In the spatial domain, area shift occurs when animals position themselves close to a rewarded location but on the side away from a nearby, unrewarded location. Cue competition is observed when the conditioning associated with a stimulus is reduced or overshadowed when it is reinforced in compound with another stimulus. Further, prior conditioning to one element can block (i.e. prevent conditioning to) other elements of a compound stimulus. Thus, nearer landmarks may dominate over more distant landmarks (overshadowing), or

BLBK374-02

BLBK374-Brown

14

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

earlier encountered landmarks may dominate (blocking). Blocking may also occur between different types of stimuli (e.g. colours and landmarks; Couvillon et al. 1997). Salient (e.g. large) cues, or those with intrinsic biological significance, are less susceptible to blocking (Denniston et al. 1996). Although relatively poorly documented in the case of fish, psychological principles such as those described in this section may be used to explain ‘suboptimal’ patterns of fish foraging, such as the consumption of non-preferred prey types and spatial biases due to the visual environment. Cue competition may help explain why fifteen-spined sticklebacks, Spinachia spinachia (Gasterosteidae), and corkwing wrasse, Crenilabrus melops (Labridae), in a radial maze make associations between visual cues (coloured tiles) and food less efficiently as the diversity of cues increases (Hughes & Blight 2000).

2.2.4

Exploration and sampling

Fishes discover food through individual sampling and by observing other foragers (Pitcher & Magurran 1983; Pitcher & House 1987). There are good reasons to suppose that exploration and sampling are fundamental, integral aspects of foraging. This idea is captured in the information primacy hypothesis of Inglis et al. (2001), according to which a major determinant of behaviour is the need to obtain information continually in order to deal with environmental variability. The information primacy hypothesis helps to account for behavioural observations that cannot be explained by conventional reward theory, such as spontaneous alternation, patrolling, effects of hunger on the variability of learned results, latent learning, contrafreeloading and behaviour following changes in food availability (see Inglis et al. (2001) for details). Such phenomena have not attracted much study in fish, but there is enough evidence to suggest that the information primacy hypothesis is valid for fish. For example, goldfish continue to sample widely even in the presence of stable, high yield food patches, and fishes that spend more time in food patch sampling are able to switch patches faster when patch profitabilities change (Pitcher & Magurran 1983; Warburton 1990). Such results indicate the existence of latent learning, where learning occurs on non-reinforced trials but remains unused until the introduction of a reinforcer provides an incentive for using it (Lieberman 1990). A contrasting effect, latent inhibition, occurs when exposure to an isolated stimulus reduces subsequent conditioning to that stimulus. When visual cues are consistently associated with food, cue fixation may occur and exploratory activity in other areas may decline markedly (Warburton 1990). While the value of sampling is in little doubt, it is important to separate purposeful sampling from inefficient patch use: At least some ‘sampling’ can be explained in terms of phenomena such as delay reduction and scalar expectancy rather than as a special type of information-gathering behaviour (Shettleworth 1988).

2.2.5

Attention and simple association

Only certain stimuli influence behaviour. Increases in foraging motivation tend to improve attention to relevant cues, such as features of food patches and prey. Fishes in larger groups are better able to sample their environment and overcome the confusion effect

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

15

caused by random movements of swarming prey, presumably because increased group size allows individuals to benefit from shared vigilance and attend more to foraging (Pitcher & Magurran 1983; Smith & Warburton 1992). The mechanism underlying attention switching is a central concern in sensory physiology (Rice 1989). By restricting an animal’s focus, attention helps to form associations when problems are simple: As a result, in conditioning experiments the conditioned stimulus effectively replaces the unconditioned stimulus in an animal’s brain. Stimulus substitution is the formation of simple associations. However, if problems are more complex, increased focus may be counterproductive because subtle, though relevant, cues may be missed (Lieberman 1990). Predators that take a wide range of prey may be expected to be relatively inefficient foragers and to suffer from divided attention (Bernays et al. 2004; see also Griffiths et al. 2004). The foraging behaviour of young lake trout (Salvelinus namaycush) is disrupted by the caudal spine possessed by the cladoceran Bythotrephes, which coexists with edible prey such as Daphnia (Barnhisel & Kerfoot 2004). However, when the rate of encounter with Bythotrephes is high enough, trout exhibit improved prey discrimination and concentrate on Daphnia, which leads to improved foraging efficiency. Working with Bidyanus bidyanus (silver perch) fingerlings foraging on brine shrimp and chironomid larvae, Warburton & Thomson (2006) found that when fishes that were familiar with one of the two prey types were offered both prey types simultaneously, the rate at which they captured both familiar and unfamiliar prey dropped progressively over succeeding trials. This result was not predicted by simple learning models, according to which a steady improvement in predation efficiency would be expected, but it could be explained in terms of an interaction between learning and attention. The authors postulated that when fish were faced with mixed prey populations, cognitive constraints associated with divided attention impaired foraging efficiency. This effect increased over time because experience increased awareness of both prey types, which then competed for attention. The presence of two alternative prey types led to substantial fluctuations in reward rate over extended periods (20 days), even when prey densities at the start of each trial were kept constant. Warburton & Thomson (2006) interpreted these effects as examples of costs of learning.

2.2.6

Cognition

Fishes have extensive and diverse abilities for pattern discrimination and categorisation (Douglas & Hawryshyn 1990; Chase 2001). Cognitive processing typically allows the subject to select from a wide range of preparatory responses, not just from innate ones. This permits flexibility of response, and goal-directed action represents the most basic behavioural marker of cognition (Dickinson 1994). Classical conditioning involves two systems: (1) stimulus substitution and (2) a cognitive/awareness system. The latter system involves active anticipation, or expectation, of an unconditioned stimulus. Fish can be trained to feed at a particular location and at a particular time of day. For example, within an experimental period of 10 days, Arctic charr, S. alpinus, learned to concentrate their feeding activity within a restricted time window (2 hours) (Brannas et al. 2005). Golden shiners, Notemigonus crysoleucas (Cyprinidae), learned to expect food at midday in one of the brightly lit corners of their tank (Reebs 2000). They displayed daily food-anticipatory activity by leaving the shady area of the tank and spending more and more time in the

BLBK374-02

BLBK374-Brown

16

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

food corner up to the normal feeding time. Time–place learning need not depend on social influences, since individually reared cichlid angelfish (Pterophyllum scalare) show a capacity for such learning. However, this may not be true of all species, since time–place learning is not exhibited by other cichlids (e.g. pearl cichlid, Geophagus brasiliensis) that have been raised in isolation (Barreto et al. 2006). As with the length of memory windows (Mackney & Hughes 1995; see Section 2.5), such differences may be related to ecological variation between species. The results of some studies (e.g. Roche et al. 1998) suggest that foraging fishes prefer sites with maximum temporal stability in terms of food abundance. However, in experiments of relatively long duration (1–1.5 months), in which the pattern of food distribution remained constant, Gerasimov (2008) found that the behaviour of three cyprinids (Abramis brama, Rutilus rutilus, C. auratus) changed. After an initial period during which rates of food intake and swimming activity increased and then stabilised, visits to food patches became shorter but more frequent, and more time was spent on non-search activities. At the same time, rates of intake and growth fell. Such changes were not exhibited by fishes in other trials, where the food distribution was continually varied. The author speculated that under predictable feeding conditions, fishes could afford to reduce their intake rate because they had a lower need to store excess energy, but more research is necessary to establish if fishes adopt such a potentially risky strategy, or whether there are other explanations for the observed behaviour. Mobile aggregations of Chromis chrysurus (Pomacentridae) feeding on zooplankton search specific foraging locations slowly and tortuously, but swim rapidly between foraging locations, so that local search involves spatial memory and expectation of resource use (Noda et al. 1994). Dugatkin & Wilson (1992) found that individual bluegill sunfish, Lepomis macrochirus (Centrarchidae), could remember their feeding success with particular conspecifics and used that information to prefer or avoid those associates over a period of several weeks. Therefore, the fish displayed cognitive abilities for strategic behaviour, a topic that is explored in depth in Chapters 12 and 13. Psychological experiments indicate that retrieval of memories may be affected by interference by other memories. Interference may be proactive (due to interference from experiences that preceded the event to be remembered) or retroactive (due to interference from events that followed the event; Lieberman 1990). Proactive interference tends to increase and retroactive to decrease as the delay between learning and recall lengthens (Briggs 1954; Baddeley 1976). Fishes have difficulties in exploiting different feeding strategies (such as those required for different types of prey) simultaneously (Persson 1985). Fifteen-spined sticklebacks learn less efficiently and have shorter memories when fed alternating rather than pure diets: The mean handling time with amphipod (Gammarus locusta) prey was reduced to a greater extent (by 68%) in monospecific trials than in trials where amphipods and brine shrimp (Artemia spp.) were presented alternately (59%; Croy & Hughes 1991a). A short ‘reorientation’ lag to consumption occurred at the beginning of experimental sessions where silver perch, B. bidyanus (Terapontidae) were offered Chironomus spp. larvae and brine shrimp in alternating trials (Warburton & Thomson, 2006). However, different components of foraging behaviour do not necessarily involve the same cognitive processes. Learning and memory of times and places where prey have

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

17

been found, or of prey-recognition stimuli, may follow different operational rules from ‘procedural’ learning and memory of prey-subjugation skills (Sherry & Schacter 1987). However, different systems may interact as shown by the effects of environmental context on the learning and memory of prey-subjugation skills. In a classical conditioning experiment, cod, Gadus morhua, learned initially to approach blinking lights (CS) before swimming to a feeder (US) placed on the opposite side of the aquarium once delivery of fish pellets had begun. In later trials, the cod moved from the CS and congregated below the feeder before food delivery had actually begun (Nilsson et al. 2008b). The sequential responses may reflect a rapid, reflexive, unconsciously elicited response to the CS (sign-tracking or autoshaping) and an operationally slower, more flexible cognitive response based on expected location and timing of the US (goal-tracking) (Lieberman 1990). Reflexive signtracking may promote more rapid, though less flexible, response to prey-recognition stimuli, thereby enabling faster orientation and approach to mobile prey (see Fig. 2.3). Additionally, sign-tracking may enable foragers to learn associations between environmental cues such as stones or weed and the likely reward from probing for prey hiding beneath or within them. Slower, but more flexible, cognitively guided responses such as goal-tracking may enable foragers to update behaviour in temporally variable habitats and food spectra.

Orientate

Fixate

Turn away

Approach S-bend Attack

Backward

Miss

Grasp Hold Spit

Ingest Turn prey

Chew

Swallow Fig. 2.3 To what extent does skill transfer enhance an individual’s ability to handle novel prey? Is there a relationship between skill transfer, memory retention and habitat-related variability of prey spectra? Behaviours are defined as follows. Orientate: Alignment of the body towards the prey. Fixate: Focus with both eyes on the prey. Approach: Swimming towards the prey. S-bend: Priming posture in readiness for striking the prey. Attack: Striking at the prey. Grasp: Gripping the prey after a successful strike. Miss: A strike that misses the prey. Hold: Prolonged grasping in readiness for manipulation of the prey. Backward: Reversed swimming while still facing or holding the prey. Spit: Forceful ejection of the prey. Ingest: Sucking or manipulating the prey into the buccal cavity. Turn prey: Reversing the orientation of the prey within the mouth. Chew: Fragmentation of the prey by chewing motion of the jaws and gill rakers. Swallow: Prey is swallowed into the oesophagus. Turn away: A half-turn of the body away from the prey in readiness for renewed subjugation cycle.

BLBK374-02

BLBK374-Brown

18

2.2.7

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Memory systems and skill transfer

Learning and memorising how to accomplish one task may impair or enhance learning and memorising another, depending on the nature and similarity of the tasks. Impairment could result from proactive or retroactive interference, discussed in Section 2.2.6, or from ‘negative transfer’ of inappropriate skills between tasks. Enhancement could result from ‘positive transfer’ of skills appropriate to both tasks. Learning and memorising recognition stimuli of different prey are likely to suffer increasing interference due to a greater risk of cue-misidentification during recall as the tasks become more similar. Three-spined sticklebacks trained consecutively on three differently coloured, but otherwise physically identical, ‘prey’ models and tested after a retention period of 8 days, were readily able to recognise the third model, less able to recognise the second and unable to recognise the first (Gibbons 1998). Short-term forgetting of recognition stimuli questions an assumption of OFT that foragers are able to recognise and rank prey during a foraging bout and begs an alternative explanation of the prey selection mechanism, perhaps involving ‘rules of thumb’ (Bergelson 1985). As discussed in Section 2.2.6, learning and remembering associations between physical attributes and dietary reward follows the principles of classical conditioning. Accordingly, the Rescorla–Wagner (1972) theory may be invoked to predict that learning to associate physical attributes of prey (CS) with food reward (US) is proportional to the difference between the current strength of the association and the maximum strength that the US will allow, and that multiple CS will compete in their contribution to strength of the association (see Section 2.2.3). The outcome of such cognitive processes will depend on the variety of prey and the rate at which different types are encountered and could provide a useful framework for predicting memory retention and diet selection in different habitats. Learning and memorising procedural tasks, for example prey-subjugation skills (Fig. 2.3), are more appropriately represented by the operant conditioning paradigm. As tasks become more similar, common elements will begin to allow generalisation through skill transfer from familiar to novel tasks (Osgood 1949; Holding 1976). However, common procedural elements may become fewer and transfer less likely as skills become more highly specialised (Henry 1968). Study of skill transfer has many applications in applied psychology and sports science (Taylor et al. 2005; Figueredo 2006; Osman 2008). For example, students trained to juggle a football with the feet performed better than controls at juggling with the knees after subsequent training (Weigelt et al. 2000). Skill transfer is also pertinent to foraging behaviour. Crabs, Carcinus maenas, showed positive skill transfer when learning to break open mussels after being trained to break open similarly shaped snail shells, but did not show negative skill transfer when trained on dissimilar prey (Hughes & O’Brien 2001). Similarly, three-spined sticklebacks trained to catch and ingest freeswimming brine shrimp, Artemia salina, showed enhanced learning and memory compared with controls when subsequently trained to subjugate amphipods, G. pulex, implying that motor skills needed to subjugate mobile prey may include basic, transferable elements (Gibbons 1998; see also Brown et al. 2003). Learning and memory may depend not only on characteristics of the tasks involved, but also on the environmental context. Three-spined sticklebacks trained to feed on amphipods in aquaria with either plain or chequered walls showed impaired performance on retraining

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

2.0 Log handling time (seconds)

BLBK374-02

19

Soft food Hard food

1.5

1.0

0

20 10 Time (hours)

30

Fig. 2.4 Mean log handling time and standard errors for red drum reared on hard (black) and soft (grey) foods. Four separate trials were conducted in 2 days. Univariate tests showed that fish reared on soft food took significantly longer to consume hard food in the first trial (** P < 0.01), but not in successive trials. (Reproduced from Ruehl & DeWitt (2007). Copyright 2007, with permission of Elsevier.)

after a 4-day retention period when wall patterns were transposed, but not when they were kept the same (Gibbons 1998). In a two-way factorial experiment, Brown et al. (2003) tested the ability of salmon parr, S. salar (Salmonidae), to forage on live Artemia, after having reared the parr in tanks containing rocks, plants and other objects or in tanks lacking such features and having trained them to feed on live bloodworm or fish flakes. Transfer, or generalisation, of subjugation skills from live bloodworm to live Artemia was shown only by parr reared in the complex environment. Thus, unfamiliarity of environmental context might contribute to reduced performance of hatchery-reared fish when first released into the wild, or indeed of any fish displaced from its usual microhabitat. Such a reduction in foraging ability could put displaced subjects at a competitive disadvantage to residents. When paired, pre-trained, three-spined sticklebacks were allowed after a 4-day retention period to compete for amphipods in a plain or chequered aquarium, foraging success was greater in subjects experiencing the same visual context as in pre-training (Gibbons 1998). Therefore, context-dependent learning and memory may reinforce habitat selection and ultimately genetic and phenotypic differentiation among populations (McPhail 1994).

2.3

Patch use and probability matching

One of the central issues in foraging theory concerns the ability of animals to assess patch quality and adjust their behaviour accordingly. Do fishes have a memory for patch profitabilities that enables them to spend longer in patches with higher prey densities?

BLBK374-02

BLBK374-Brown

20

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

If so, what decision rules do they use? To address these questions, Marschall et al. (1989) and DeVries et al. (1989) examined the behaviour of bluegill sunfish foraging for chironomid (Chironomus riparius) larvae among patches of artificial macrophytes. Both studies found that bluegills searched randomly within a patch, but came to different conclusions with respect to the patch departure rules that the fish used. Marschall et al. (1989) concluded that, of three different types of decision rules based on time, total prey caught or capture rate in a patch, a constant residence time rule explained the observed data best. According to this rule, a forager is expected to stay in each patch for a constant optimal amount of time. Giving-up time (GUT) is the time spent in a patch after the final prey capture event. Although the optimal strategy is to persist in highly profitable patches when quality differs among patches within a foraging area, bluegills did just the opposite, displaying shorter GUTs in highly profitable patches. This result may have occurred because fishes did not have perfect environmental knowledge and were unable to assess patches without sampling them. In contrast, the study by DeVries et al. (1989) suggested that bluegills used a patch departure rule based on capture rate. Observed GUTs were longer than those predicted by optimality theory. However, this bias was in a direction that minimised the cost of poorly approximating the optimal solution. The different outcomes of the two studies were attributed to variations in experimental design: Marschall et al. (1989) varied patch quality within habitats but kept habitat quality (i.e. total prey density) constant across treatments, while DeVries et al. (1989) varied habitat quality but kept patch quality constant within a habitat. The main conclusion was that bluegills could assess environmental characteristics such as patch depletion and adjust their patch departure rules accordingly. Further work by Wildhaber et al. (1994), where bluegills were offered food pellets in a partitioned, two-patch shuttle tank, indicated that patch GUTs were based on foraging experience. Bluegills had longer residence times and GUTs in relatively poor environments. However, within an environment, residence times and GUTs increased linearly with prey encounter and a variable GUT rule was the best predictor of departure. Individuals varied in terms of mean GUTs but there was little variation in overall patterns of patch use. Taken together, these studies suggest that bluegills can adjust their foraging strategy as patterns of patch profitability change. Other generalist foragers may show similar responses. Flexible decision-making permits matching between patch exploitation and patch profitability. An example would be the use of a linear GUT rule coupled with a memory for patch quality. Since tests of OFT are essentially similar to studies of reinforcement schedules (Shettleworth 1988), it is relevant to consider whether fishes are capable of probability matching, i.e. whether they are able to choose alternative food sources in proportion to their associated likelihood of reward. Mackintosh et al. (1971) trained goldfish either on visual probability discrimination with irrelevant spatial cues or on spatial problems with irrelevant visual cues. In the test tank, fishes were presented with a screen containing two holes behind which were paddles onto which red, green or white lights could be projected. Food rewards were delivered via a magazine opening midway between the screen holes. When subjected to a 70:30 probability schedule (i.e. two patches that were rewarded on 70% and 30% of occasions, respectively), fishes showed probability matching for both

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

21

visual and spatial tests, in that they chose the majority stimulus on approximately 70% of trials. However, their responses were not random. Most subjects showed significant biases to one or other value of the irrelevant dimension, presumably by failing to maintain consistent attention to the inconsistently reinforced relevant cue. Birds and mammals that have been tested in the same way also show non-random responses but a greater tendency for maximising (i.e. concentrating only on the most-reinforced stimulus) than probability matching (Mackintosh et al. 1971). While these findings show that fishes are capable of probability matching, the reinforcement context appears to be critical, since probability matching was replaced by maximising unless a ‘guidance’ procedure was used: in such a procedure, if on any trial the unreinforced stimulus was chosen initially, it was removed and the fish was allowed to earn a reward for response to the other stimulus (Behrend & Bitterman 1961). Despite intensive research on a wide range of animal taxa, none of the major theories about the psychological processes that underlie matching satisfactorily explains all observed data, and there is still uncertainty as to whether matching or maximising is the basic choice rule (Williams 1994). However, evidence from both psychology and behavioural ecology indicates that fishes are capable of flexible exploitation strategies in response to reinforcement variation. Detailed cross-interpretation of findings from the two fields is complicated by the use of different protocols and performance currencies, but valid analogies may be drawn in terms of the use of information by fishes. In a reinforcement study, the more frequently reinforced stimulus will be associated with both a higher mean reward rate and a higher degree of certainty, since it will be better sampled by the subject. Both these factors will encourage maximising with respect to that stimulus. Guidance procedures increase return rates and certainty with respect to the less frequently reinforced stimulus, thus encouraging probability matching. In larger foraging areas and natural habitats, with more widely spaced food patches and higher travel costs, residence times per visit are likely to be higher and animals will acquire more accurate information on mean profitability – a process akin to guidance. At the same time, attentional deficits (Mackintosh et al. 1971) and other factors will contribute to uncertainty. Working with juvenile Atlantic salmon (S. salar) at natural densities in an artificial stream system, MacLean et al. (2003) found that patch switching was largely unrelated to food availability. Thus, the fact that fishes are capable of following simple foraging rules under laboratory conditions is no guarantee that the same rules are equally important in complex situations in the wild. There is a need for more research on the impact of intrinsic and extrinsic factors on patch use strategies by fishes in natural environments.

2.4

Performance

Learning and motivation can both enhance performance. Much of the observed improvement in skill development over a series of experimental trials is due to increased speed, and the period when associations are actually formed is likely to be much shorter (Lieberman 1990). As a necessary preliminary to successful prey ingestion, several distinct behaviours must be combined in sequence (Fig. 2.3; Croy & Hughes 1990). Repeated experience can improve the efficiency of prey recognition, attack, manipulation and ingestion by fish

BLBK374-02

BLBK374-Brown

22

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(e.g. Ware 1971; Colgan et al. 1986; Ranta & Nuutinen 1986; Mills et al. 1987; Croy & Hughes 1991a; Hughes & Croy 1993). Planktivores, such as bluegill sunfish, use past encounter rates to decide when to pursue prey (Werner et al. 1981). The speed with which discrimination is learned may vary between prey types (Croy & Hughes 1991b; Kaiser et al. 1992). Fifteen-spined sticklebacks (S. spinachia) find oligochaetes (whiteworms, Enchytraeus albidus) easier to catch and handle than amphipods, for which the capacity for learned improvements in attack efficiency, handling efficiency and handling time is greater (Mackney & Hughes 1995). Croy & Hughes (1991a) found that S. spinachia required five to eight trials to learn to handle new prey. Similarly, bluegill sunfish required six to eight exposures to reach a peak handling efficiency on Daphnia (Werner et al. 1981). During the course of a 9-day experiment, na¨ıve reared turbot, Scophthalmus maximus, that were offered shrimp, Crangon crangon, achieved the consumption rates of wild fish (Ellis et al. 2002). Silver perch fingerlings reached peak efficiency on chironomid larvae and brine shrimp after five trials (Warburton & Thomson 2006). Reiriz et al. (1998) found that na¨ıve Atlantic salmon increased their consumption of Hydropsyche drastically and decreased that of Gammarus after only three attempts, thus matching the selection pattern of wild fish. The preference of pellet-reared juvenile Atlantic salmon for live prey increased within 16 attempts (Stradmeyer & Thorpe 1987). The ability of na¨ıve coho salmon, Oncorhynchus kisutch (Salmonidae), to capture shrimps (Crangon spp.) also increased with experience (Paszkowski & Olla 1985). Johnsen & Ugedal (1986, 1989, 1990) reported that hatchery-reared brown trout, S. trutta (Salmonidae), had broadened their diets to coincide with those of wild fish after a few weeks at liberty. Individual juvenile Atlantic salmon that were offered prey in a sequential encounter context increased their consumption of all prey types as they gained experience, but the improvement was higher for prey types that were less consumed initially (Reiriz et al. 1998). The fast learning response towards novel prey was interpreted as a way of maintaining high foraging efficiency in the face of frequent changes in prey availability. Na¨ıve reared turbot, S. maximus, attacked stones in preference to newly offered shrimp (C. crangon), evidently because of a persistent memory of pellet-like characteristics. This behaviour persisted in some fishes for at least 6 weeks, which Ellis et al. (2002) interpreted in terms of costs of memory. Juvenile red drum, Sciaenops ocellatus, reared on hard foods (crayfish limb segments) developed morphological features (e.g. heavier feeding muscles) that differed from those fish reared on soft foods (crayfish meat without exoskeleton), and consumed hard foods 2.6 times faster (Ruehl & DeWitt 2007). However, mean handling times of the two groups converged within four trials (Fig. 2.4) demonstrating that learning can compensate for small physical variations (Ruehl & DeWitt 2007).) Day (1999) obtained evidence for context-dependent familiarity in rainbowfish (Melanotaenia duboulayi). Individuals became familiar with different shoals in various contexts, namely feeding, threat or neutral (non-feeding, non-threat). In later trials, when given a choice between shoals, the same individuals showed different preferences depending on the context of the test situation. After being exposed to threat during the testing period, but not before, fishes preferred to associate with the shoal with which they had been threat-exposed during training. In the context of feeding, the opposite trend was observed, with individuals spending more time with familiar foraging shoalmates before, but

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

23

not after, food-exposure during the test period. The context also affected the frequency of ‘S-wiggles’ (sigmoid movements that appeared to be non-aggressive social signals), with more being directed to fishes outside the familiar foraging group. While further work is required to clarify the complex dynamics of context-dependency, these results suggest that fishes were responding to both their external environment (e.g. threat level) and their internal motivational state (e.g. hunger state and curiosity). The larvae of the cane toad, Bufo marinus, which has been introduced to Australia, are toxic to predatory native Australian fish species. On exposure to B. marinus, na¨ıve native fish species show different patterns of learned behaviour. In experiments most barramundi (Lates calcarifer) rapidly learned to avoid toad tadpoles, while sooty grunter (Hephaestus fuliginosus) showed wide intraspecific variation in behaviour: Some individuals learned to avoid tadpoles after only a few attacks while others continued to attack and reject tadpoles throughout the series of trials. Differences in fish learning ability, hunger levels and tadpole palatability may have contributed to the observed behavioural variation (Crossland 2001). Learning can magnify the effect of small behavioural variations into pronounced individual foraging differences (Ehlinger 1989; Kieffer & Colgan 1991; Kohda 1994) and thus help to explain individual variation in the diets of fishes in the wild (Warburton et al. 1998). Kieffer & Colgan (1991) observed pumpkinseed sunfish, L. gibbosus (Centrarchidae), feeding on whiteworms in structured and open habitats and noted significant individual variation in learning rates. Moreover, habitat structure had a significant effect on the efficiency of individuals with respect to the inter-capture interval and total time feeding. Such individual variation may represent adaptive flexibility in foraging behaviour (Magurran 1993). The finding that pumpkinseed exhibited a positive successive contrast effect, where fishes that were familiar with a low yield, vegetated habitat showed enhanced performance when exposed to an open habitat, indicates that differing background experience may affect the foraging efficiency and competitive success of individuals (Kieffer & Colgan 1991). By changing the efficiency of prey exploitation and thus relative prey profitabilities, learning can also help to account for prey switching and frequency-dependent predation (Murdoch 1969; Ringler 1985). Experiments with fifteen-spined sticklebacks by Croy & Hughes (1991a) indicated that Artemia became more profitable than Gammarus as trials proceeded. Shoals comprised of individuals that are familiar with one another are able to locate food patches more quickly and consume more prey items than unfamiliar shoals (Swaney et al. 2001; Ward & Hart 2005). Members of natural guppy (P. reticulata) shoals enjoy better foraging success than those of artificially constructed shoals, presumably due to enhanced social learning and/or a reduced perception of risk among familiarised individuals (Morrell et al. 2008).

2.5

Tracking environmental variation

Food patch discrimination can be improved by remembering the spatial position of previously exploited patches, and predators must continually compare present information on patches and prey types with search images held in memory (Hart 1986, 1993;

BLBK374-02

BLBK374-Brown

24

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Odling-Smee & Braithwaite 2003). There appear to be individual differences in food patch learning: some fishes seem to identify food patches by reference to local visual cues while others rely more on global cues (Huntingford & Wright 1989). When tested in an eight-arm radial maze in the absence of spatial cues, fifteen-spined sticklebacks and corkwing wrasse adopted an algorithmic strategy of visiting every third arm, but in the presence of spatial cues (coloured tiles) algorithmic behaviour was largely subsumed by the use of spatial memory (Hughes & Blight 1999). The results suggested that fishes were able to memorise the spatial configuration of the cues and indicate that when spatial information is limited, reliance on search algorithms that reduce the likelihood of revisiting depleted patches tends to increase (see Chapter 8 for more detail). Fauchald (1999) modelled prey search in a stochastic hierarchical patch system where high-density patches at small scales were nested within low-density patches at larger scales. The scaling of the prey system was adapted for schooling marine fish and krill. Simulations suggested that information flow and tracking efficiency were maximal at high levels of prey abundance but at intermediate levels of prey aggregation. These conditions describe patch systems that are well structured and hierarchical, where clear structural cues are available to guide foraging behaviour. The allocation of priorities given to patches with changing profitabilities is made possible by patch sampling, which permits fish to switch rapidly between win-shift and win-stay behaviour when revisiting preferred locations. This is especially valuable in habitats where food patch distributions are subject to frequent change as found, for example, in intertidal environments (Hughes & Blight 2000). Sticklebacks (G. aculeatus) can develop foraging preferences for certain types of subhabitat as well as different spatial locations; Webster & Hart (2006) found that preferences for sand or gravel sub-habitats based on profitability took about 14 days to develop, and then persisted regardless of prey density. However, in the presence of conspecifics, fish used social foraging cues in preference to private information. Braithwaite et al. (1996) showed that juvenile Atlantic salmon had the ability to distinguish between two identical visual landmarks and to learn to track the movements of one of them to predict the location of food. They concluded that chemosensory cues, perhaps originating from substrate marking by the fish themselves, could be used in conjunction with visual cues. Salmonids are known to mark substrates (Halvorsen & Stabell 1990) and switch to nocturnal foraging in winter (Fraser et al. 1993), possibly in response to changes in prey availability or predator threat. Context switching and the retention interval appear to have additive effects on memory (Bouton et al. 1999). A strong case can be made in support of the idea that some forgetting may be adaptive rather than a negative outcome of process failure (Kraemer & Golding 1997). Foraging covers a diverse set of activities in which the animal may draw on several learning and memory systems; for example, spatial memory to construct a cognitive map of its environment (Broglio et al. 2003; Odling-Smee & Braithwaite 2003; see also Chapters 8 and 15), a memory for cues associated with particular food patches and food types, an ability to recognise individual shoalmates (Griffiths 2003; Chapter 9) and to benefit from social transmitted information (Brown & Laland 2003; Chapter 11) and a memory for predator-related cues that indicate when the risks of foraging are unacceptably high (Brown 2003 Chapter 4; Kelley & Magurran 2003 Chapter 3). The typical memory windows for

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

25

these different systems are likely to reflect corresponding levels of environmental variation. For example, in a structured, relatively undisturbed environment spatial memory should be persistent, while associations involving varying patch profitabilities and ephemeral prey types should be relatively short. Curves of forgetting against time are typically exponential in form (White 2001) and foraging theory has tended to assume that recently obtained information should be the most valued (e.g. Cuthill et al. 1990). However, little is known about rates of forgetting and the factors that lead to variations in such rates, including the effects of intervening experience. Fish species appear to vary widely in terms of their memory window, defined as the duration of learned foraging skills. In the absence of reinforcement, fifteen-spined sticklebacks start to forget foraging skills after 2 days and return to na¨ıve levels by 8 days (Croy & Hughes 1991a), while corresponding times for rainbow trout, O. mykiss (Salmonidae), are 14 days (onset of forgetting) and 3 months (return to na¨ıve level) (Ware 1971). In contrast, silver perch can retain learned foraging skills for at least 5 weeks (Warburton & Thomson 2006). It seems likely that the skills to recognise and capture a number of prey types are retained simultaneously, but the decisions to draw on those skills are based on recent experience. These results are consistent with proposals that extinction is not equivalent to unlearning, and that in an appropriate context previous learning can be reinstated (Pearce & Bouton 2001). Feeding habits, which are easily learned in early life, can be retained for a significant length of time, and early training on natural foods may develop preferences that persist through long periods of feeding on artificial foods (Suboski & Templeton 1989). Norris (2003) found that two groups of juvenile whiting, Sillago maculata (Sillaginidea), one grown on a diet of live food and the other on artificial pellets over a period of 4 months, diverged in terms of gross morphology, taste bud distribution and learned feeding responses. The observed differences in development were attributed to the different stimuli required to locate food on the two diets, pellets being approached primarily using chemical cues and live prey mainly via visual stimuli. The development of an individual’s ability to use social cues in foraging also depends critically on experience (Huntingford 1993). Evidence suggests that the memory window for prey is related to the predictability of the feeding environment, and there appear to be significant interspecific and inter-population differences correlated with ecology. Although freshwater and anadromous three-spined sticklebacks and marine fifteen-spined sticklebacks exhibit similar rates of prey learning, their memory windows differ considerably, being >25 days, 10 days and 8 days, respectively. For residential or anadromous marine populations that move with the tide and where older information is rapidly devalued, a long memory window could be maladaptive by retarding behavioural adjustment to changing conditions (Mackney & Hughes 1995). In silver perch, a freshwater species, the relatively long memory window (>5 weeks) is consistent with evidence that individuals respond to slower cycles of food availability and unpredictable, intermittent peaks in particular prey types (Warburton et al. 1998). Using an appetitive conditioning regime, Nilsson et al. (2008a) showed that Atlantic cod, G. morhua are able to associate two events (blinking light, food presentation) separated by up to 60 seconds, and to remember the association for at least 3 months. In radial maze experiments, imposition of a delay within trials had no effect on foraging efficiency of wrasse and sticklebacks when memory for spatial cues could be used to guide foraging, but

BLBK374-02

BLBK374-Brown

26

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

in the absence of cues the behavioural algorithm was reset, leading to reduced efficiency (Hughes & Blight 1999). Memory retention for previous choices lay within the range of 0.5–5.0 minutes, consistent with the timescale of rapidly changing prey distributions during the tidal cycle. Phylogenetic influences on the length of the memory window cannot be ruled out. However, systematic studies of a range of fish taxa tested with the same experimental protocol are needed before the importance of such influences can be properly assessed. Traditional assumptions that animal learning and memory are influenced mainly by phylogenetic relationships have been significantly revised as evidence for the importance of lifestyle and ecological context has emerged (Healy 1992). Though very incomplete, our present knowledge of the role of learning and memory in the foraging behaviour of fishes is consistent with this revised interpretation; for example, the difference in memory window between two forms of the three-spined stickleback is as great as that between one form of three-spined stickleback and the fifteen-spined stickleback, a species from a different genus (Mackney & Hughes 1995).

2.6

Competition

Learning and memory can influence the distributional patterns of competing individuals. The relative pay-off sum (RPS) learning rule predicts that good competitors will decide where to feed earlier and switch less between patches than poor competitors (Regelmann 1984). Within-shoal competition can cause subordinate fish to abandon patch sampling (Croy & Hughes 1991b). Hakayama & Iguchi (2001) recorded patterns in distribution, aggression, food intake and growth of the salmonid Onchorhynchus masou ishika that had free access to two patches and were able to use long-term memory to assess patch quality. The within-group variation in body weight increased with time, and over the 4-week experimental period, the pattern of resource use changed from a random distribution to an ideal free distribution and finally to an ideal despotic distribution. On average, the better patch was used by more individuals than predicted by a random distribution but by fewer than predicted by an ideal free distribution. Therefore, competition was a contributory reason for the non-occurrence of an ideal free distribution. A capacity for individual recognition allows foraging fish to identify potential competitors (see Griffiths 2003; Chapter 9). In rainbow trout, aggression, food intake and growth rate are positively correlated and aggressive dominants deny subordinates access to food (Johnsson 1997). Levels of aggression in contests between familiar individuals are lower than in contests between strangers, so that familiarity appears to reduce aggression and increase the foraging rate (Griffiths et al. 2004). This is consistent with the hypothesis that individual recognition is used to reduce the costs of contesting resources: After an initial contest, encounters between familiar individuals should be settled with less aggression and a lower probability of status reversal than encounters between unfamiliar fish (Switzer et al. 2001). However, this effect decays rapidly with time since previously familiar individuals are treated as strangers after 3 days of separation (Johnsson 1997).

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

2.7

27

Learning and fish feeding: some applications

Stimuli that elicit feeding responses such as approach and oral contact serve to initiate individual learning mechanisms based on the palatability of the novel food (Suboski & Templeton 1989). In aquaculture, self-feeding devices rely on the fact that fish can be trained to press a trigger to obtain food (Boujard & Leatherland 1992). The key stimuli (presence of potential food particles, the sight of other fish feeding) are the same as those involved in the wild. Na¨ıve fish appear to mistake the self-feeding triggers for food particles but later learn to distinguish between the trigger and food and reduce the force they apply to the trigger (Alanara 1996). Trout reared in tanks in relatively small groups of 100–300 individuals reach a stable level of self-feeding in approximately 25 days, but under large-scale rearing conditions where 1000–2000 individuals are kept in cages, learned associations between trigger and food seem less prevalent since the frequency of trigger actuations stays high even when food is not provided (Alanara 1996). Dominant fish may guard access to the trigger and this in turn influences access and learning capability of all fish in the system. Also relevant to aquaculture is the ability of fish to learn the time of day when food is available. This may be evidenced by food-anticipatory behaviour in the form of a gradual increase in locomotory activity prior to feeding time (Sanchez-Vazquez et al. 1997). Amano et al. (2005) found that goldfish (C. auratus) fed once a day showed approach behaviour to food odour only at the time when they were normally fed, but fishes fed three times a day showed no significant approach behaviour at any time of day. Therefore, anticipatory behaviour may be most effective at inducing fish to aggregate for feeding when the rewards are relatively high (i.e. food is concentrated in one meal) and clearly associated with a single time of day. There is growing interest in the possibility of life-skills training of hatchery-reared fish before release into the wild. Compared to wild fishes, hatchery-reared individuals tend to suffer behavioural deficits that may significantly reduce post-release survivorship (Brown & Laland 2001; Brown & Day 2002). Such deficits include reduced food consumption, lower diversity of prey types, consumption of non-prey objects, longer delays in attacking and consuming prey, slower prey switching and atypical microhabitat choice (Sosiak et al. 1979; Ersbak & Haase 1983; Olla et al. 1998; Neveu 1999; Sundstrom & Johnsson 2001), all of which may be redressed if steps are taken to provide appropriate learning opportunities prior to release into the wild.

2.8

Conclusions

Learning and memory play a decisive role in the foraging activities of fish. Learning and associated improvements in prey search, capture and handling efficiency can lead to significant enhancements in foraging performance after only a few exposures. Fishes are capable of adjusting their foraging strategy as patterns of patch profitability change. There is also evidence that forgetting seems to have adaptive significance, since the length of the memory window appears to be related to environmental variability in at least some species.

BLBK374-02

BLBK374-Brown

28

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Thus, learning and memory permit rapid and flexible adaptation to changing prey densities, identity and location. Fishes have been used to test the predictions of classic foraging theory. These studies show that fish can match exploitation rates to patch profitabilities. To do this, they draw on memories for general patch quality (based on previous sampling and exploitation) and on departure rules for leaving the current patch. However, it appears that such patch departure rules can be varied according to ecological circumstances. In this respect fishes resemble higher vertebrates (e.g. jays, Garrulus glandarius (Corvidae); Shettleworth 1988). Evidence from the fish foraging literature is consistent with the proposal that in terms of basic learning mechanisms, the similarities across animal taxa are more striking than the differences (Lieberman 1990). If cognition is viewed as a collection of adaptively specialised modules, then all extant species are equally intelligent in their own ways and it makes no sense to propose a linear evolutionary hierarchy of cognitive modules (Shettleworth 1998). Like other vertebrates, fish draw on an array of learning and memory systems as part of a broad cognitive repertoire. There is much to be discovered about the ways in which individually and socially learned information about the feeding habitat, prey types, conspecifics and predator risk are combined to make foraging decisions. Convergence between behavioural ecology and comparative psychology offers promise in terms of developing more mechanistically realistic foraging models and explaining apparently ‘suboptimal’ patterns of behaviour. Foraging decisions involve the interplay between several distinct systems of learning and memory, including those that relate to habitat, food patches, prey types, conspecifics and predators. The expression of such behaviours will vary with the ‘personalities’ of individuals (see Chapter 7). Therefore, fish biologists face an interesting challenge in developing integrated accounts of fish foraging that explain how cognitive sophistication can help individual animals deal with the complexity of the ecological context. Although progress has been made in identifying general psychological principles underlying animal behaviour, in most cases their impacts on the foraging efficiency of fish remain to be explored. For example, to what extent does skill transfer enhance an individual’s ability to handle novel prey? Is there a relationship between skill transfer, memory retention and habitat-related variability of prey spectra? How do visual stimuli bias movement patterns, independent of patch profitabilities? How is learning constrained by attentional limitations? Which factors encourage latent learning and latent inhibition? How do order effects, such as successive contrast, affect the competitive foraging success of individuals in shoals? There is scope for imaginative experiments to address such questions. Almost all previous work has been carried out under laboratory conditions and there is a pressing need to assess the role of learning in natural, complex environments.

Acknowledgements We would like to thank Paul Cunningham, Craig Hull, Ottmar Lipp and Culum Brown for their helpful comments on the chapter, Andrew Norris for access to unpublished information on the sensory development of whiting and Viviana Gamboa-Pickering for assistance with referencing.

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

29

References Alanara, A. (1996) The use of self-feeders in rainbow trout (Oncorhynchus mykiss) production. Aquaculture, 145, 1–20. Amano, M., Iigo, M. & Yamamori, K. (2005) Effects of feeding time on approaching behaviour to food odor in goldfish. Fisheries Science, 71, 183–186. Baddeley, A.D. (1976) The Psychology of Memory. Harper and Row, New York. 430 p. Barnhisel, D.R. & Kerfoot, W.C. (2004) Fitting into food webs: behavioural and functional response of young lake trout (Salvelinus namaycush) to an introduced prey, the spiny cladoceran (Bythotrephes cederstroemi). Journal of Great Lakes Research, 30 (Suppl. 1), 300–314. Barreto, R.E., Rodrigues, P., Luchiari, A.C. & Delicio, H.C. (2006) Time-place learning in individually reared angelfish, but not in pearl cichlid. Behavioural Processes, 73, 367–372. Behrend, E.R. & Bitterman, M.E. (1961) Probability-matching in the fish. American Journal of Psychology, 74, 542–551. Bergelson, J.M. (1985) A mechanistic interpretation of prey selection by Anax junius larvae (Odonata: Aeschnidae). Ecology, 66, 1699–1705. Bernays, E.A., Singer, M.S. & Rodrigues, D. (2004) Foraging in nature: foraging efficiency and attentiveness in caterpillars with different diet breadths. Ecological Entomology, 29, 389–397. Boujard, T. & Leatherland, J.F. (1992) Demand-feeding behaviour and diel pattern of feeding activity in Oncorhynchus mykiss held under different photoperiod regimes. Journal of Fish Biology, 40, 535–544. Bouton, M.E., Nelson, J.B. & Rojas, J.M. (1999) Stimulus generalization, context change, and forgetting. Psychological Bulletin, 125, 171–186. Braithwaite, V.A., Armstrong, J.D., McAdam, H.M. & Huntingford, F.A. (1996) Can juvenile Atlantic salmon use multiple cue systems in spatial learning? Animal Behaviour, 51, 1409–1415. Brannas, E., Berglund, U. & Eriksson, L.O. (2005) Time learning and anticipatory activity in groups of Arctic charr. Ethology, 111, 681–692. Breuning, S.E., Ferguson, D.G. & Poling, A.D. (1981) Second-order schedule effects with goldfish: a comparison of brief-stimulus, chained, and tandem schedules. Psychological Record, 31, 437–445. Briggs, G.E. (1954) Acquisition, extinction and recovery functions in retroactive inhibition. Journal of Experimental Psychology, 47, 285–293. Broglio, C., Rodr´ıguez, F. & Salas, C. (2003) Spatial cognition and its neural basis in teleost fish. Fish and Fisheries, 4, 247–255. Brown, C. & Day, R. (2002) The future of stock enhancements: bridging the gap between hatchery practice and conservation biology. Fish and Fisheries, 3, 79–94. Brown, C. & Laland, K.N. (2001) Social learning and life skills training for hatchery reared fish. Journal of Fish Biology, 59, 471–493. Brown, C. & Laland, K.N. (2003) Social learning in fishes: a review. Fish and Fisheries, 4, 280–288. Brown, C., Davidson, T. & Laland, K. (2003) Environmental enrichment and prior experience of live prey improve foraging behaviour in hatchery-reared Atlantic salmon. Journal of Fish Biology, 63, 187–196. Brown, G.E. (2003) Learning about danger: chemical alarm cues and local risk assessment in prey fishes. Fish and Fisheries, 4, 227–234. Chase, A.R. (2001) Music discriminations by carp (Cyprinus carpio). Animal Learning and Behavior, 29, 336–353. Cheng, K. & Spetch, M.L. (2001) Blocking in landmark-based search in honeybees. Animal Learning and Behavior, 29, 1–9. Colgan, P.W., Brown, J.A. & Orsatti, S.D. (1986) Role of diet and experience in the development of feeding behaviour in large-mouth bass (Micropterus salmoides). Journal of Fish Biology, 28, 161–170. Collin, S.P. & Pettigrew, J.D. (1988a) Retinal topography in reef teleosts. I. Some species with well-developed areae but poorly-developed streaks. Brain Behavior and Evolution, 31, 269–282.

BLBK374-02

BLBK374-Brown

30

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Collin, S.P. & Pettigrew, J.D. (1988b) Retinal topography in reef teleosts. I. Some species with prominent horizontal streaks and high-density areae. Brain Behavior and Evolution, 31, 283–295. Couvillon, P.A., Arakaki, L. & Bitterman, M.E. (1997) Intramodal blocking in honeybees. Animal Learning and Behavior, 25, 277–282. Crespi, L.P. (1942) Quantitative variation in incentive and performance in the white rat. American Journal of Psychology, 55, 467–517. Crossland, M.R. (2001) Ability of predatory native Australian fishes to learn to avoid toxic larvae of the introduced toad Bufo marinus. Journal of Fish Biology, 59, 319–329. Croy, I.M. & Hughes, R.N. (1990) The combined effects of learning and hunger in the feeding behaviour of the fifteen-spined stickleback (Spinachia spinachia). In: R.N. Hughes (ed) Behavioural Mechansisms of Food Selection. NATO ASI Series, Vol. G20, pp. 215–233. Springer, Berlin. Croy, M.I. & Hughes, R.N. (1991a) The role of learning and memory in the feeding behaviour of the fifteen-spined stickleback, Spinachia spinachia L. Animal Behaviour, 41, 149–159. Croy, M.I. & Hughes, R.N. (1991b) The influence of hunger on feeding behaviour and on the acquisition of learned foraging skills by the fifteen-spined stickleback, Spinachia spinachia L. Animal Behaviour, 41, 161–170. Cuthill, I.C., Kacelnik, A., Krebs, J.R., Haccou, P. & Iwasa, Y. (1990) Starlings exploiting patches: the effect of recent experience on foraging decisions. Animal Behaviour, 40, 625–640. Dahl, C.D., Wallraven, C., Bulthoff, H.H. & Logothetis, N.K. (2009) Humans and macaques employ similar face-processing strategies. Current Biology, 19, 509–513. Dall, S.R.X., McNamara, J.M. & Cuthill, I.C. (1999) Interruptions to foraging and learning in a changing environment. Animal Behaviour, 57, 233–241. D´amato, M.R. & Schiff, J. (1964). Long-term discriminated avoidance performance in the rat. Journal of Comparative and Physiological Psychology, 57, 123–126. Day, J. (1999) Context-dependent familiarity in rainbowfish. Honours thesis, Department of Zoology and Entomology, University of Queensland, 88 p. Denniston, J.C., Miller, R.R. & Matute, H. (1996) Biological significance as a determinant of cue competition. Psychological Science, 7, 325–331. Desimone, R. (1991). Face-selective cells in the temporal cortex of monkeys. Journal of Cognitive Neuroscience, 3, 1–8. DeVries, D.R., Stein, R.A. & Chesson, P.L. (1989) Sunfish foraging among patches: the patchdeparture decision. Animal Behaviour, 37, 455–464. Dickinson, A. (1994) Instrumental conditioning. In: N.J. Mackintosh (ed) Animal Learning and Cognition, pp. 45–79. Academic Press, San Diego, CA. Domjan, M. (1998) The Principles of Learning and Behavior, 435 p. Brooks/Cole, Pacific Grove, CA. Douglas, R.H. & Hawryshyn, C.W. (1990) Behavioural studies of fish vision: an analysis of visual capabilities. In: R.H. Douglas & M.B.A. Djamgoz (eds) The Visual System of Fish, pp. 373–418. Chapman and Hall, London. Dugatkin, L.A. & Wilson, D.S. (1992) The prerequisites for strategic behaviour in bluegill sunfish, Lepomis macrochirus. Animal Behaviour, 44, 223–230 Ehlinger, T.J. (1989) Learning and individual variation in bluegill foraging: habitat-specific techniques. Animal Behaviour, 38, 643–658. Ellis, T., Hughes, R.N. & Howell, B.R. (2002) Artificial dietary regime may impair subsequent foraging behaviour of hatchery-reared turbot released into the natural environment. Journal of Fish Biology, 61, 252–264. Ersbak, K. & Haase, B.L. (1983) Nutritional deprivation after stocking as a possible mechanism leading to mortality in stream-stocked brook trout. North American Journal of Fisheries Management, 3, 142–151. Fauchald, P. (1999) Foraging in a hierarchical patch system. American Naturalist, 153, 603–613. Figueredo, L. (2006) Using the known to chart the unknown: a review of first-language influence on the development of English-as-a-second-language spelling skill. Reading and Writing, 19, 873–905.

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

31

Fraser, N.H.C., Metcalfe, N.B. & Thorpe, J.E. (1993) Temperature-dependent switch between diurnal and nocturnal foraging in salmon. Proceedings of the Royal Society of London Series B – Biological Sciences, 252, 135–139. Gerasimov, Y.V. (2008) Fish foraging behaviour at different stability levels of food distribution. Inland Water Biology, 1, 175–181. Gibbons, M.J. (1998) The Evolutionary Ecology of Memory in Sticklebacks. Ph.D. thesis, Bangor University, 296 p. Giles, N. 1983. Behavioural effects of the parasite Schistocephalus solidus (Cestoda) on an intermediate host, the three-spined stickleback, Gasterosteus aculeatus L. Animal Behaviour, 31, 1192–1194. Gotceitas, V. & Colgan, P.W. (1990) The effects of prey availability and predation risk on habitat selection by juvenile bluegill sunfish. Copeia, 1990, 409–417. Griffiths, S.W. (2003) Learned recognition of conspecifics by fishes. Fish and Fisheries, 4, 256–268. Griffiths, S.W., Hoejesjoe, J. & Johnsson, J.I. (2003) Familiarity confers anti-predator and foraging advantages on juvenile brown trout. Journal of Fish Biology, 63 (Suppl. A), 226–245. Griffiths, S.W., Brockmark, S., Hoejesjoe, J. & Johnsson, J.I. (2004) Coping with divided attention: the advantage of familiarity. Proceedings of the Royal Society of London Series B – Biological Sciences, 1540, 695–699. Hakayama, H. & Iguchi, K. (2001) Transition from a random to an ideal free to an ideal despotic distribution: the effect of individual difference in growth. Journal of Ethology, 19, 129–137. Halvorsen, M. & Stabell, O.B. (1990) Homing behaviour of displaced stream-dwelling trout. Animal Behaviour, 39, 1089–1097. Hart, P.J.B. (1986) Foraging in teleost fishes. In: T.J. Pitcher (ed) The Behaviour of Teleost Fishes, pp. 211–235. Croom Helm, London. Hart, P.J.B. (1993) Teleost foraging: facts and theories. In: T.J. Pitcher (ed) Behaviour of Teleost Fishes, 2nd edition, pp. 253–284. Chapman and Hall, London. Healy S, 1992. Optimal memory: towards an evolutionary ecology of animal cognition. Trends in Ecology and Evolution, 7, 399–400. Henry, F.M. (1968) Specificity vs. generality in learning motor skill. In: R.C. Brown & G.S. Kenyon (eds), Classical Studies on Physical Activity, pp. 331–340. Prentice Hall, Englewood Cliffs, NJ. Holding, D.H. (1976) An approximate transfer. Journal of Motor Behaviour, 15, 230–296. Hughes, R.N. & Blight, C.M. (1999) Algorithmic behaviour and spatial memory are used by two intertidal fish species to solve the radial maze. Animal Behaviour, 58, 601–613. Hughes, R.N. & Blight, C.M. (2000) Two intertidal fish species use visual association learning to track the status of food patches in a radial maze. Animal Behaviour, 59, 613–621. Hughes, R.N. & Croy, M.I. (1993) An experimental analysis of frequency-dependent predation (switching) in the 15-spined stickleback, Spinachia spinachia. Journal of Animal Ecology, 62, 341–352. Hughes, R.N., Kaiser, M.J., Mackney, P.A. & Warburton, K. (1992) Optimizing foraging behaviour through learning. Journal of Fish Biology, 41 (Suppl. B), 77–91. Hughes, R.N. & O’Brien, N. (2001) Shore crabs are able to transfer learned handling skills to novel prey. Animal Behaviour, 61, 711–714. Huntingford, F.A. (1993) Development of behaviour in fish. In: T.J. Pitcher (ed) Behaviour of Teleost Fishes, pp. 57–83. Chapman and Hall, London. Huntingford, F.A. & Wright, P.J. (1989) How sticklebacks learn to avoid dangerous feeding patches. Behavioural Processes, 19, 181–189. Inglis, I.R., Langton, S., Forkman, B. & Lazarus, J. (2001) An information primacy model of exploratory and foraging behaviour. Animal Behaviour, 62, 543–557. Jain, V.K. & Sahai, S. (1989) Learning behaviour of the black molly, Molliensia sphenops. Environmental Ecology, 7, 337–344.

BLBK374-02

BLBK374-Brown

32

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Johnsen, B.O. & Ugedal, O. (1986) Feeding by hatchery-reared and wild brown trout, Salmo trutta L., in a Norwegian stream. Aquaculture and Fisheries Management, 17, 281–287. Johnsen, B.O. & Ugedal, O. (1989) Feeding by hatchery-reared brown trout, Salmo trutta L. released in lakes. Aquaculture and Fisheries Management, 20, 97–104. Johnsen, B.O. & Ugedal, O. (1990) Feeding by hatchery- and pond-reared brown trout, Salmo trutta L., fingerlings released in a lake and in a small stream. Aquaculture and Fisheries Management, 21, 253–258. Johnsson, J.I. (1997) Individual recognition affects aggression and dominance relations in rainbow trout, Oncorhynchus mykiss. Ethology, 103, 267–282. Kaiser, M.J., Gibson, R.N. & Hughes, R.N. (1992) The effects of prey type on the predatory behaviour of the fifteen-spined stickleback Spinachia spinachia L. Animal Behaviour, 43, 147–156. Kelley, J.L. & Magurran, A.E. (2003) Learning of predator recognition and anti-predator responses in fishes. Fish and Fisheries, 4, 216–226. Kieffer, J.D. & Colgan, P.W. (1991) Individual variation in learning by foraging pumpkinseed sunfish, Lepomis gibbosus: the influence of habitat. Animal Behaviour, 41, 603–611. Kieffer, J.D. & Colgan, P.W. (1992) The role of learning in fish behaviour. Reviews in Fish Biology and Fisheries, 2, 125–143. Kohda, M. (1994) Individual specialized foraging repertoires in the piscivorous cichlid fish Lepidolamprologus profundicola. Animal Behaviour, 48, 1123–1131. Kraemer, P.J. & Golding, J.M. (1997) Adaptive forgetting in animals. Psychonomic Bulletin and Review, 4, 480–491. Laudien, H., Frever, J., Erb, R. & Denzer, D. (1986) Influence of isolation stress and inhibited protein biosynthesis on learning and memory in goldfish. Physiology and Behaviour, 38, 621–628. Lieberman, D.A. (1990) Learning: Behavior and Cognition, 500 p. Wadsworth, Belmont, CA. Logue, A.W. (1988) A comparison of taste aversion learning in humans and other vertebrates: evolutionary pressures in common. In: R.C. Bolles & M.D. Beecher (eds) Evolution and Learning, pp. 97–116. Erlbaum, Hillsdale, NJ. Mackintosh, N.J., Lord, J. & Little, L. (1971) Visual and spatial probability learning in pigeons and goldfish. Psychonomic Science, 24, 221–223. Mackney, P.A. & Hughes, R.N. (1995) Foraging behaviour and memory window in sticklebacks. Behaviour, 132, 1241–1253. MacLean, A., Huntingford, F.A., Armstrong, J.D. & Ruxton, G.D. (2003) Fish don’t read textbooks: juvenile salmon prove ignorant of foraging theory. Journal of Fish Biology, 63 (Suppl. A), 226–245. Magurran, A.E. (1993) Individual differences and alternative behaviours. In: T.J. Pitcher (ed) Behaviour of Teleost Fishes, pp. 441–477. Chapman and Hall, London. Marschall, E.A., Chesson, P.L. & Stein, R.A. (1989) Foraging in a patchy environment: prey-encounter rate and residence time distributions. Animal Behaviour, 37, 444–454. McDowall, R.M. (1996) Freshwater Fishes of Southeastern Australia. Reed, Sydney. 247 p. McPhail, J.D. (1994) Speciation and the evolution of reproductive isolation in the sticklebacks (Gasterosteus) of South-Western British Columbia. In: M.A. Bell & S.A Foster (eds), The Evolutionary Biology of the Three-Spine Stickleback, pp. 399–426. Oxford University Press, Oxford. Mednick, A.S. & Springer, A.D. (1988) Asymmetric distribution of retinal ganglion cells in goldfish. Journal of Comparative Neurology, 268, 49–59. Milinski, M. (1993) Predation risk and feeding behaviour. In: T.J. Pitcher (ed) Behaviour of Teleost Fishes, pp. 285–305. Chapman and Hall, London. Mills, E.L., Widzowski, D.V. & Jones, S.R. (1987) Food conditioning and prey selection by young yellow perch (Perca flavescens). Canadian Journal of Fisheries and Aquatic Sciences, 44, 549–555. Mittelbach, G.G. 1981. Foraging efficiency and body size – a study of optimal diet and habitat use by bluegills. Ecology, 62, 1370–1386. Morrell, L.J., Croft, D.P., Dyer, J.R.G., Chapman, B.B., Kelley, J.L., Laland, K.N. & Krause, J. (2008) Association patterns and foraging behaviour in natural and artificial guppy shoals. Animal Behaviour, 76, 855–864.

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

33

Murdoch, W.W. (1969) Switching in general predators: experiments on predator specificity and stability of prey populations. Ecological Monographs, 39, 335–354. Neveu, A. (1999) Feeding strategy of the brown trout (Salmo trutta L.) in running water. In: J.L. Bagliniere & G. Maisee (eds) Biology and Ecology of the Brown and Sea Trout, pp. 91–113. Praxis, Chichester. Nilsson, J., Kristiansen, T.S., Fosseidengen, J.E., Ferno, A. & van den Bos, R. (2008a) Learning in cod (Gadus morhua): long trace interval retention. Animal Cognition, 11, 215–222. Nilsson, J., Kristiansen, T.S., Fosseidengen, J.E., Ferno, A. & van den Bos, R. (2008b) Sign- and goal-tracking in Atlantic cod (Gadus morhua). Animal Cognition, 11, 651–659. Noda, M., Gushima, K. & Kakuda, S. (1994) Local prey search based on spatial memory and expectation in the planktiovorous reef fish, Chromis chrysurus (Pomacentridae). Animal Behaviour, 47, 1413–1422. Norris, A.J. (2003) Sensory Modalities, Plasticity and Prey Choice in Three Sympatric Species of Whiting (Pisces: Sillaginidae). Ph.D. Thesis, University of Queensland. 233 p. Odling-Smee, L. & Braithwaite, V.A. (2003) The role of learning in fish orientation. Fish and Fisheries, 4, 235–246. Olla, B.L., Davis, M.W. & Ryer, C.H. (1998) Understanding how the hatchery environment represses or promotes the development of behavioural survival skills. Bulletin of Marine Science, 62, 531–550. Osgood, C.E. (1949) The similarity paradox in human learning: a resolution. Psychological Review, 56, 132–143. Osman, M. (2008) Positive transfer and negative transfer/antilearning of problem-solving skills. Journal of Experimental Psychology: General, 137, 97–115. Paszkowski, C.A. & Olla, B.L. (1985) Foraging behaviour of hatchery-produced coho salmon (Oncorhynchus kisutch) smolts on live prey. Canadian Journal of Fisheries and Aquatic Sciences, 42, 1915–1921. Pearce, J.M. & Bouton, M.E. (2001) Theories of associative learning in animals. Annual Reviews Psychology, 52, 111–139. Persson, L. (1985) Optimal foraging: the difficulty of exploiting different feeding strategies simultaneously. Oecologia, (Berlin) 67, 338–341. Pitcher, T.J., Lang, S.H. & Turner, J.R. (1988) A risk-balancing trade-off between foraging rewards and predation risk in shoaling fish. Behavioral Ecology and Sociobiology, 22, 225–228. Pitcher, T.J. & House, A.C. (1987) Foraging rules for group feeders: area copying depends upon food density in shoaling goldfish. Ethology, 76, 161–167. Pitcher, T.J. & Magurran, A.E. (1983) Shoal size, patch profitability and information exchange in foraging goldfish. Animal Behaviour, 31, 546–555. Ranta, E. & Nuutinen, V. (1986) Experience affects performance of ten-spined sticklebacks foraging on zooplankton. Hydrobiologia, 140, 161–166. Reader, S.M. & Laland, K.N. (2000) Diffusion of foraging innovations in the guppy. Animal Behaviour, 60, 175–180. Reebs, S.G. (2000) Can a minority of informed leaders determine the foraging movements of a fish shoal? Animal Behaviour, 59, 403–409. Regelmann, K. (1984) Competitive resource sharing: a simulation model. Animal Behaviour, 32, 226–232. Reiriz, L., Nicieza, A.G. & Brana, F. (1998) Prey selection by experienced and naive juvenile Atlantic salmon. Journal of Fish Biology, 53, 100–114. Rescorla, R.A. & Wagner, A.R. (1972) A theory of Pavlovian conditioning: variations on the effectiveness of reinforcement and non-reinforcement. In: A.H. Black & W.F. Prokasy (eds) Classical Conditioning: II. Current Research and Theory, pp. 64–99. Appleton-Century-Crofts, New York. Rice, M.J. (1989) The sensory physiology of pest fruit flies: conspectus and prospectus. In: A.S. Robinson & G.H.S. Hooper (eds) Fruit Flies. Their Biology, Natural Enemies and Control, pp. 249–272. Elsevier, Amsterdam.

BLBK374-02

BLBK374-Brown

34

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Ringler, N.H. (1985) Individual and temporal variation in prey switching by brown trout, Salmo trutta. Copeia, 1985, 918–926. Roche, J.P., Dravet, S.M., Bolyard, K., & Rowland, W. (1998) Risk sensitivity in foraging Jack Dempsey cichlids (Cichlasoma octofasciatum), Ethology, 104, 593–602. Ruehl, C.B. & DeWitt, T.J. (2007) Trophic plasticity and foraging performance in red drum, Sciaenops ocellatus (Linnaeus). Journal of Experimental Marine Biology and Ecology, 349, 284–294. Sanchez-Vazquez, F.J., Madrid, J.A., Zamora, S. & Tabata, M. (1997) Feeding entrainment of locomotor activity rhythms in the goldfish is mediated by a feeding-entrainable circadian oscillator. Journal of Comparative Physiology, 181, 121–132. Seligman, M.E.P. (1970) On the generality of the laws of learning. Psychological Review, 77, 406–418. Sengpiel, F., Stawinski, P. & Bonhoeffer, T. (1999) Influence of experience on orientation maps in cat visual cortex. Nature Neuroscience, 2, 727–732. Shepard, R.N. (1987) Toward a universal law of generalisation for psychological science. Science, 237, 1317–1323. Shepard, R.N. (1988) Time and distance in generalization and discrimination. Reply to Evans (1988). Journal of Experimental Psychology: General, 117, 415–416. Sherry, D.F. & Schacter, D.L. (1987) The evolution of multiple memory systems. Psychological Review, 94, 439–454. Shettleworth, S.J. (1988) Foraging as operant behavior and operant behavior as foraging: what have we learned? Psychology of Learning and Motivation, 22, 1–49. Shettleworth, S.J. (1998) Cognition, Evolution, and Behavior, 688 p. Oxford University Press, Oxford. Smith, M.F.L. & Warburton, K. (1992) Predator shoaling moderates the confusion effect in blue-green chromis, Chromis viridis. Behavioral Ecology and Sociobiology, 30, 103–107. Sosiak, A.J., Randall, R.G. & McKenzie, J.A. (1979) Feeding by hatchery-reared and wild Atlantic salmon (Salmo salar) parr in streams. Journal of the Fisheries Research Board of Canada, 36, 1408–1412. Stradmeyer, L. & Thorpe, J.E. (1987) The response of hatcher-reared Atlantic salmon, Salmo salar L., parr to pelletted and wild prey. Aquaculture and Fisheries Management, 18, 51–62. Suboski, M.D. & Templeton, J.J. (1989) Life skills training for hatchery fish: social learning and survival. Fisheries Research, 7, 343–352. Sundstrom, L.F. & Johnsson, J.I. (2001) Experience and social environment influence the ability of young brown trout to forage on live novel prey. Animal Behaviour, 61, 249–255. Swaney, W., Kendall, J., Capon, H., Brown, C. & Laland, K.N. 2001. Familiarity facilitates social learning of foraging behaviour in the guppy. Animal Behaviour, 62, 591–598. Switzer, P.V., Stamps, J.A. & Mangel, M. (2001) When should a territory resident attack? Animal Behaviour, 62, 749–759. Taylor, P.J, Russ-Eft, D.F. & Chan, D.W.L. (2005) A meat-analytic review of behaviour modelling training. Journal of Applied Psychology, 90, 692–709. Underwood, G. (1979) Memory systems and conscious process. In: G. Underwood & R. Stevens (eds) Aspects of Consciousness Vol 1: Psychological Issues, pp. 91–121. Academic Press, London. Warburton, K. (1990) The use of local landmarks by foraging goldfish. Animal Behaviour, 40, 500–505. Warburton, K. & Thomson, T. (2006) Costs of learning: the dynamics of mixed-prey exploitation by silver perch, Bidyanus bidyanus. Animal Behaviour, 71, 361–370. Warburton, K., Retif, S. & Hume, D. (1998) Generalists as sequential specialists: diets and prey switching in juvenile silver perch. Environmental Biology of Fishes, 51, 445–454. Ward, A.J.W. & Hart, P.J.B. (2005) Foraging benefits of shoaling with familiars may be exploited by outsiders. Animal Behaviour, 69, 329–335. Ware, D.M. (1971) Predation by rainbow trout (Salmo gairdneri): the effect of experience. Journal of the Fisheries Research Board of Canada, 28, 1847–1852.

BLBK374-02

BLBK374-Brown

May 13, 2011

17:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning of Foraging Skills by Fish

35

Webster, M.M. & Hart, P.J.B. (2006) Subhabitat selection by foraging threespine stickleback (Gasterosteus aculeatus): previous experience and social conformity. Behavioral Ecology and Sociobiology, 60, 77–86. Weigelt, C., Williams, A.M., Wingrove, T. & Scott, M.A. (2000) Transfer and motor skill learning in association football. Ergonomics, 43, 1698–1707. Werner, E.E., Mittelbach, G.G. & Hall, D.J. (1981) The role of foraging profitability and experience in habitat use by the bluegill sunfish. Ecology, 62, 116–125. White, K.G. (2001) Forgetting functions. Animal Learning and Behavior, 29, 193–207. Wildhaber, M.L., Green, R. & Crowder, L.B. (1994) Bluegills continuously update patch giving-up times based on foraging experience. Animal Behaviour, 47, 501–513. Williams, B.A. (1994) Reinforcement and choice. In: N.J. Mackintosh (ed) Animal Learning and Cognition, pp. 81–108. Academic Press, San Diego, CA.

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 3

Learned Defences and Counterdefences in Predator–Prey Interactions Jennifer L. Kelley and Anne E. Magurran

3.1

Introduction

Living with predators is an unavoidable aspect of life for almost all fishes. The activity patterns of predators are highly variable over space and time and consequently prey are faced with the continual need to balance their habitat use (Chapter 8), foraging decisions (Chapter 2) and reproduction (Chapter 5) with the risk of predation (Lima & Dill 1990; Kats & Dill 1998; Lima 1998). Learning is the mechanism by which prey can achieve this important outcome because it allows prey to fine-tune their antipredator responses to variations in predation risk that can occur seasonally, across lunar cycles and on a momentto-moment basis (Lima & Bednekoff 1999). Through learned predator recognition, prey can respond to novel introduced predators (Kristensen & Closs 2004), changes in community structure (Pollock & Chivers 2004), or learn a response to predators that were previously extinct in the local area (Berger et al. 2001). It was once thought that there could be little role for learning in the defence against predators because a failure to respond appropriately during the first encounter would result in death. However, predator–prey interactions are far more complex than this scenario implies and there are many ways in which prey can learn about predators while being exposed to relatively low levels of risk. This is particularly the case for remote cues such as odour because prey can obtain information about the predator even when it is not in the vicinity (Smith 1997; Chivers & Smith 1998; Kats & Dill 1998; see Chapter 4). There is also the potential for prey to acquire information at low risk by observing conspecifics or heterospecifics being attacked, or by observing other prey responding to predator-related cues (i.e. through social learning; see Chapter 11). Predation risk is highly variable during the predator–prey interaction (see below) and prey must display an antipredator response that reflects the magnitude of the risk posed (‘risksensitive predator avoidance’; Helfman 1989). This is because prey need to balance their risk of predation against other activities that influence fitness, such as feeding, courtship and habitat use (e.g. Sih 1980, 1988). This balance is achieved when prey display an Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learned Defences and Counterdefences in Predator–Prey Interactions

37

antipredator response that is appropriate to the magnitude of the threat, such that greater threats produce stronger avoidance responses (Helfman 1989). Prey that show a minimal avoidance response may have an increased risk of mortality, whereas displaying an overly cautious antipredator response results in a loss of time and energy available for other important activities. Most studies of predator–prey interactions have addressed the adaptations of prey in their defence against predators. This follows a general trend in behavioural ecology in which the behaviour of predators in the predator–prey interaction appears to have been largely ignored (Lima 2002). In line with this tendency, the majority of studies that have addressed learning in the context of predator–prey interactions have examined the ways in which prey modify their antipredator behaviour as a result of experience with predators. In contrast, few investigators have examined the importance of learning in shaping predatorhunting behaviour. Predator learning is a key assumption behind mathematical models of the evolution of aposematism (Speed 2001), but surprisingly few studies have addressed the cognitive abilities of predators that are predicted by these models. Finally, predator learning is an important means by which predators can counteract the behavioural plasticity of their prey and may, therefore, be a crucial weapon in the predator–prey arms race. In this chapter, we evaluate the evidence for learning by both predators and prey at each of the five stages of the predator–prey interaction (Fig. 3.1). At each stage of the predator–prey interaction, predator attack and prey defence may be initiated in response to behavioural, chemical and morphological cues that are associated with the presence of predators and prey. Many experimental studies of predator–prey interactions have examined these cues in isolation and, as a result, many of the examples given in this chapter are based on a particular Prey defences

Predator adaptations

Time–place learning Rarity and activity

Encounter

Time–place learning

Immobility, crypsis, confusion Sensory perception

Detection

Development of sensory perception

Associative learning (visual or chemical) Masquerade, confusion, aposematism, mimicry, honest signalling of unprofitability Fleeing, refuging, schooling Social learning, habituation/reinforcement

Fleeing, hiding, schooling, crypsis, immobility Negative reinforcement

Recognition

Approach

Evasion/attack

Learned prey recognition Search images and memory for particular prey shapes and colour patterns

Information acquired about awarenessand condition of inspector?

Previous experience with fleeing prey

Fig. 3.1 Stages of the predation sequence and opportunities for learning by both predators and prey. (Adapted from Endler 1986, 1991; Lima & Dill 1990.)

BLBK374-03

BLBK374-Brown

38

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

cue, usually those that are visual or chemical. We give examples of learning processes that use both these stimuli, but the reader is referred to Chapter 4 for a detailed discussion of chemically mediated learning. This chapter largely focuses on behavioural interactions between predators and prey; however, because it is often the interaction between behaviour and morphology that determines the success of antipredator defences or predator attack strategies, morphological adaptations are included in cases where learning is considered to be particularly relevant.

3.2

The predator–prey sequence

The interaction between a predator and its prey can be considered as a sequence of events that begins when predator and prey encounter one another and ends when the prey escapes or is consumed by the predator. As the sequence progresses, individual prey are exposed to increasing levels of predation risk (Endler 1986, 1991). The energetic costs of avoiding predation increase as the interaction progresses and the antipredator behaviours that are performed become less frequent and more specialised (Endler 1986, 1991); for example, evasive manoeuvres such as short bursts of high-speed swimming are far more specialised and energetically demanding than avoidance behaviours that reduce the probability of predator encounter (Fuiman & Magurran 1994). However, avoiding risky habitats could be costly in the long term, particularly if alternative habitats are suboptimal or unsuitable. As in the case of prey, the counterdefences of predators tend to be less frequent, increasingly specialised and more costly to perform as the sequence progresses (Endler 1991); for example, predators tend to use the same sensory and cognitive processes for detecting and recognising many different prey species (Endler 1991). However, to overcome prey defences at the later stages of the sequence, predators often have costly and specialised adaptations, such as fast attack speeds for capturing their prey and teeth and venom for subduing them (Endler 1991). At every stage of the interaction, the behavioural defences of prey and the counteradaptations of predators are in conflict. Prey defences are deployed to allow the prey to escape the interaction as early as possible, whereas predator counterdefences aim to continue the interaction until it ends in prey capture (Endler 1991). If attack/defence strategies can be modified through experience, then we predict that learning by both predators and their prey is a crucial weapon in the predator–prey arms race. Opportunities for learning can arise at any one of the five stages of the predation sequence. However, we predict that learning is likely to play a greater role during the earlier stages of the sequence than during the later stages. This is based on the assumption that behaviours initiated at these early stages are more generalised and less energetically costly to perform than those used at the later stages (see Section 3.2, Endler 1991). The frequency at which predators and prey experience each stage of the sequence is also important. Prey experience the earlier stages of the interaction more often, for example when encountering a cue from a predator, than at the later stages of the sequence, such as being chased by a predator. Learning in the latter stages will also incur considerable risks to individual prey because these stages of the interaction are associated with higher levels of predation risk; for example, the predation risk of learning to associate a particular habitat with danger is less than learning an appropriate escape response to a pursuing predator. We now

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

Char Count=

39

consider examples of learning by both predators and prey at each of the five stages of the predator–prey sequence.

3.2.1

Encounter

One of the most effective ways for prey to reduce their risk of predation is to adopt behaviours that reduce their probability of encountering a predator. However, given that many prey are constantly exposed to some level of predation risk, avoidance behaviours must function to manage the level and duration of risk, rather than avoid it altogether. Consequently, predation risk should be considered in terms of the relative level, duration and frequency of risky periods rather than a high-risk/no-risk situation, which is typical of laboratory studies of antipredator behaviour (Sih et al. 2000). The risk allocation hypothesis (RAH) predicts that as the duration or frequency of relatively high-risk periods increases, investment in antipredator behaviour should decrease or an overall reduction in fitness will result (Lima & Bednekoff 1999; see Ferrari et al. 2009 for a review of studies testing the RAH). Risk allocation can be achieved in a variety of ways such as minimising the time spent in habitats that are associated with elevated risk and reducing activity levels during the periods in which predators are most active (Lima & Dill 1990). The comprehensive reviews provided by Lima & Dill (1990) and Lima (1998) provide many examples of behavioural decisions that reduce a prey’s chance of encountering a predator; for example, the decision of where to feed and for how long is an important one when there is spatial and temporal variability in predator activity (Lima & Bednekoff 1999). We now consider studies that have examined whether prey can learn to avoid dangerous habitats and whether predators and prey adjust their activity patterns in order to increase their hunting efficiency or reduce their risk of predation. 3.2.1.1

Avoiding dangerous habitats

Several studies have demonstrated that fishes can learn to avoid dangerous foraging patches following an encounter with a predator; for example, Huntingford & Wright (1992) found that three-spined sticklebacks (Gasterosteus aculeatus, Gasterosteidae) learned to avoid a feeding patch that they previously favoured following a simulated predatory attack. Utne-Palm (2001) found that na¨ıve two-spotted gobies (Gobiusculus flavescens, Gobiidae) subsequently avoided the habitat where they last saw a live cod predator (Gadus morhua, Gadidae), but avoided cod odour only after observing cod on three successive occasions. Brown (Brown C. 2003) similarly found that rainbowfish (Melanotaenia spp., Melanotaeniidae) avoided habitats where they had previously encountered a model of their natural predator, the mouth almighty (Glossamia aprion, Apogonidae). In Brown’s experiment (Brown C. 2003), the observation arena was rotated through 90◦ following predator exposure, suggesting that the fish remembered the location where the predator model appeared using features of the habitat as well as global cues (outside the test arena). Gobies (Bathygobius soporator, Gobiidae) living in tidal rock pools also use spatial learning to jump between pools at low tide and avoid contact with fish predators (Aronson 1971). Studies with fathead minnows (Pimephales promelas, Cyprinidae) demonstrate that fish can also learn to recognise dangerous habitats by associating chemical odours from that

BLBK374-03

BLBK374-Brown

40

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

habitat with ‘damage-released alarm cues’ (Chivers & Smith 1995a, 1995b). Damagereleased alarm cues (hereafter referred to as alarm cues) are chemical cues in the epidermis of the skin that are released if the fish is injured or captured by a predator (von Frisch, cited in Pfeiffer 1974; Smith 1992, 1997; Brown G.E. 2003; Chapter 4). Detection of alarm cues by conspecifics (or heterospecifics) elicits an unlearned ‘fright reaction’, characterised by freezing, dashing, hiding, shoaling and/or reduced foraging/mating activity (Chivers & Smith 1998). A large number of studies, particularly those associated with learned predator recognition (see Subsection 3.2.3), have provided evidence for associative learning through the association of alarm cues with previously novel stimuli. Fathead minnows were presented with water from either an open habitat or water from an area with vegetated cover (in the same stream) paired with either alarm cues or distilled water (control). The fathead minnows displayed a learned response to both types of water that were previously presented in conjunction with alarm cues, but not to water that was not paired with alarm cues (Chivers & Smith 1995a, 1995b). Chivers & Smith (1995a) also showed that a learned response to water from a particular habitat can be socially transmitted to na¨ıve observers. Collectively, these studies demonstrate that spatial memory allows prey to associate visual features of a habitat with increased risk, whereas associative learning is a mechanism by which fish can learn to respond to chemical cues from dangerous habitats. In line with risk allocation, it should be noted that prey may not always choose habitats that limit their likelihood of encountering a predator. Furthermore, habitat use under predation risk can be influenced by interactions between the prey’s escape mode, the hunting tactic of the predator and features of the landscape (see Wirsing et al. 2010 for a review). Thus, prey may sometimes select habitats where predators are relatively common because the landscape increases their chance of a successful escape or affords greater foraging opportunity. It would be very interesting to understand whether these decisions are contingent on a prey’s previous experience with the habitat and/or predators with particular attack strategies. Similarly, we might expect the attack success of predators to be related to features of the environment with predators selecting habitats that are more conducive to their particular attack strategies.

3.2.1.2

Changing activity patterns

The activity patterns of both marine (reviewed by Neilson & Perry 1990) and freshwater fishes (reviewed by Reebs 2002) are highly plastic and migrations of reef fishes are almost certainly driven by trade-offs between foraging opportunities and risk of predation (Smith 1997); for example, Helfman (1986) found that the timing of migration of juvenile grunts (Haemulon flavolineatum, Haemulidae) was delayed following a simulated increase in attack rate by a model predator, a lizardfish (Synodus intermedius, Synodontidae). Subsequent studies have demonstrated that social learning allows French grunts (H. flavolineatum, Haemulidae) and bluehead wrasse (Thalassoma bifasciatum, Labridae) to maintain consistent migration routes over several generations (Helfman & Schultz 1984; Warner 1988; Brown & Laland 2003), but the timing of the migration may vary depending on the perceived level of threat. Odling-Smee et al. provide more information on learned migratory patterns in Chapter 7.

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

Char Count=

41

Pettersson et al. (2000) showed that crucian carp (Carassius carassius, Cyprinidae) that were exposed to water containing a predatory pike exhibited low levels of night-time activity, whereas carp exposed to untreated water were nocturnal. The guppy (Poecilia reticulata, Poeciliidae), a species that was previously considered to be diurnal (Houde 1997), readily forages at night under release from its nocturnal predator the trahira (Hoplias malabaricus, Erythrinidae; Fraser et al. 2004). Guppies denied the opportunity of night feeding exhibit low growth rates and reduced daytime courtship intensities relative to their nocturnally foraging counterparts (Fraser et al. 2004), illustrating the significance of predator activity on guppy behaviour and life history. Reebs (1999) examined whether prey fishes learn to adjust the timing of their activities in response to predation risk. In this study, shoals of inangas (Galaxias maculatus, Galaxiidae) were presented with either food, a simulated predation threat (a model of a heron’s bill), or both treatments. The stimuli were presented twice in the morning in one half of the tank and twice in the afternoon in the other half for a period of 14 days. After this time the stimuli were removed and the activity patterns of the shoals measured. Fish that had received the food treatment continued to anticipate the time and place at which food was delivered, but fish in the predation treatments and in the combined treatments did not display a time–place association (Reebs 1999). In contrast, a large number of studies have shown that fish predators adjust their activity patterns according to changes in the behaviour of their prey. A well-known example of this is the migration of fishes in response to diurnal migrations of their plankton prey. Zooplankton descend into the water column during the day in order to avoid visual-hunting predators. When the risk of predation is lower at night, the zooplankton migrate to the nutrient-rich surface waters. Predator odour has a direct effect on the swimming behaviour of daphnia by increasing the proportion of individuals that perform the migration pattern (Dodson et al. 1997). It is most likely that diel patterns of activity are a result of natural selection rather than arising through experience with migrating predators and prey. Although the rate at which predators encounter prey is influenced by prey abundance, predator learning and experience also play an important part in this process (Endler 1991). Predator hunting tactics are based on the optimal search rate hypothesis, which states that predators should spend more time searching for prey in patches that have high prey densities and less time in patches in which prey are rare (Gendron & Staddon 1983; Chapter 2). The ability of predators to adapt their foraging behaviour according to the availability of different prey types and features of the habitat is discussed under ‘Recognition’ (Subsection 3.2.3).

3.2.2

Detection

One of the most important ways in which prey can avoid being detected by a predator is through being cryptic and matching their colouration with that of their background environment (Edmunds 1974). Although crypsis is usually considered in terms of colouration, prey chemical, auditory and electrical signals can also be difficult for predators to detect at a level above background ‘noise’ (Smith 1997). Most research on predator–prey interactions in fishes has focused on visual crypsis. The diverse body shapes of species such as pipefishes (Syngnathus spp., Syngnathidae), the barbeled leaf fish (Monocirrhus polyacanthus, Polycentridae), the seadragon (Phyllopteryx eques, Syngnathidae) and the frogfish (Antennarius marmoratus, Antennariidae) allow them to be virtually indistinguishable from their

BLBK374-03

BLBK374-Brown

42

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

surroundings (Keenleyside 1979). Behaviour is an essential component of crypsis and fishes such as the barbeled leaf fish (M. polyacanthus) hold themselves vertical and sway back and forth in eelgrass (Keenleyside 1979). More commonly, prey must be motionless in order to appear cryptic (Cott 1940); these tactics (inactivity and crypsis) function synergistically to promote concealment (Ioannou & Krause 2009). Because crypsis is a function of the environment as well as the prey’s movement, factors that influence habitat selection and the time spent in that habitat will directly affect how conspicuous an animal is in a given environment. 3.2.2.1

Crypsis

Flatfishes (Pleuronectiformes) provide a particularly good example of crypsis as they are able to change their dorsal colouration so that it matches that of the substratum. Burying behaviour also enhances crypsis: sole (Solea solea, Soleidae) that are buried in sand react to a predation stimulus at a shorter distance (hence relying on crypsis) than those that are not buried (Ellis et al. 1997). The low post-release survival of hatchery-reared flatfishes has been attributed to high predation mortality through poor crypsis, suggesting that some aspect of substratum matching requires experience (Howell 1994). Although flatfish with no experience of burying quickly dig themselves into the substratum when given the opportunity, sole (Ellis et al. 1997) and winter flounder (Pseudopleuronectes americanus, Pleuronectidae; Fairchild & Howell 2004) with previous experience in a sandy substratum are more efficient buriers (measured as the proportion of sand covering their dorsal surface) than those reared in hard-bottomed tanks. In the case of flounder, 5 days of contact with a sandy substratum were required for the burying behaviour of fish reared in hard-bottomed tanks to resemble that of fish reared in sandy-bottomed tanks (Fairchild & Howell 2004). The ability to change colour in both these species is also influenced by the environment in which fishes are reared; the colour (lightness, intensity and hue) of hatchery flatfish reared in hard-bottomed tanks took up to 69 days, in the case of sole, to resemble that of wild fish (Ellis et al. 1997), and over 90 days for winter flounder (Fairchild & Howell 2004). Although colour change is largely a physiological process rather than a behavioural one, it is the interaction between behaviour and morphology that determines the effectiveness of crypsis (Cott 1940; Edmunds 1974). Few studies with fishes have examined the relationship between prey conspicuousness and the effect of predator foraging experience. However, Johnsson & Kjallman-Eriksson (2008) examined the ability of brown trout to improve their foraging performance when presented with prey (live brown maggots) that appeared either conspicuous (presented on a green background) or cryptic (presented on a brown background). Search times were longer for cryptic prey than for conspicuous prey and foraging performance increased over successive trials for both prey types, indicating that learning was occurring. However, there was no difference in the rate of learning between cryptic and conspicuous prey. It would be interesting to further these experiments and investigate how the level of prey crypsis influences the rate of predator learning. 3.2.2.2

Sensory perception

Predators are far more successful at capturing prey when the prey animal is unaware of the predator’s presence (Dugatkin & Godin 1992). As a result, there is strong selection

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

Char Count=

43

on predators for rapid detection of prey and vice versa. Movement is the primary means through which predators and prey detect one another (Kislalioglu & Gibson 1976). When detecting the cues from a predator, prey fishes commonly increase schooling, freeze, sink lower into the water column and/or hide in an attempt to avoid being detected. The type of response adopted may depend on the habitat the fishes originate from; for example, rainbowfish (Melanotaenia eachamensis, Melanotaeniidae) that coexist with predators in an open lake habitat increase their schooling but do not seek refuge when exposed to visual predatory cues (Brown & Warburton 1997). In contrast, rainbowfish from a more structured but predator-free environment were more likely to hide under vegetation when exposed to a predator model (Brown & Warburton 1997). Exposure of prey to alarm cues results in freezing and/or hiding even in the absence of direct predator cues (Suboski et al. 1990; Hall & Suboski 1995; Yunker et al. 1999). While also functioning in learned predator recognition and avoidance of dangerous habitats (see ‘Avoidance’ and ‘Recognition’, Subsections 3.2.1 and 3.2.3, respectively), alarm cues allow prey to respond rapidly to cues that indicate that predators are in the vicinity, allowing them to adopt behaviours that reduce their chance of being detected at an early stage of the encounter. For both predators and prey, detection is a function of the distance over which sensory perception operates and is dependent on the environment. Hartman & Abrahams’ (2000) sensory compensation model predicts that fishes rely more on alternative cues when the primary source of information is reduced. The relative importance of different sensory cues will fluctuate with both spatial and temporal changes in the environment; in turbid habitats, for instance, both predators and prey are likely to rely disproportionately on nonvisual sensory systems such as the lateral line, electroreception and olfaction (Hartman & Abrahams 2000). Interestingly, recent work with guppies suggests that the environment encountered during rearing affects the responsiveness of the sensory system, with individuals compensating for deficiencies in one sense by enhancing acuity in another (‘sensory plasticity’). Guppies raised in low-light conditions showed increased sensitivity to chemical food cues compared with those reared at high light levels, irrespective of the light environment that they were tested in (Chapman et al. 2010). The observation that guppies reared in low-light conditions continued to show a reduced response to visual only food cues even when tested in a highlight environment suggests that developmental constraints on the sensory system may not easily be reversed (Chapman et al. 2010). This is an interesting area for further investigation as such developmental effects could constrain the use of sensory information for mediating a variety of behaviours, including those that are learned. Therefore, human impacts that lead to changes in the sensory environment could have important effects on predator–prey interactions and overall community dynamics.

3.2.3 3.2.3.1

Recognition Associative learning

The best-known mechanism through which fishes learn to recognise predators is through associative learning or releaser-induced recognition learning (Suboski 1990). This learning process is comparable to Pavlovian conditioning and occurs when na¨ıve individuals acquire

BLBK374-03

BLBK374-Brown

44

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

a response to a predator cue (the conditional stimulus) by associating it with an alarm cue (the unconditioned stimulus) (Suboski 1990). It was previously thought that only fishes belonging to the Ostariophysian family display an unlearned response to alarm cues; however, the phenomenon is now known to be widespread among fishes (Chivers & Smith 1998). Although most examples of associative learning have paired alarm cues with predator odour, prey can also be conditioned to respond to visual cues from predators (Chivers & Smith 1994). Furthermore, conditioning cues need not be biologically relevant, as zebra danios (Brachydanio rerio, Cyprinidae), fathead minnows and glowlight tetras can learn a response to unnatural odours (Suboski et al. 1990), flashing lights (Hall & Suboski 1995; Yunker, et al. 1999) and novel auditory cues (Wisenden et al. 2008). In most cases, associative learning occurs after just one simultaneous presentation of the cue and the stimulus (Magurran 1989) and the response can be retained for up to 2 months (Chivers & Smith 1994). If a single exposure to a chemical cue can lead to marked and long-lasting changes in antipredator behaviour, what prevents fishes from learning a response to irrelevant stimuli? Responding to non-biological or irrelevant stimuli would entail significant fitness costs and reduce time available for other activities (Lima & Dill 1990). There is some evidence that fishes are predisposed to acquire responses to moving objects (Brown & Warburton 1997; Vogel & Bleckmann 2001; Wisenden & Harter 2001) and to particular predator cues (i.e. ‘learning specificity’; Griffin et al. 2001); for example, fathead minnows learn to fear the sight of both pike and goldfish, but when retested 2 months later the response to the pike had remained unchanged, whereas the response to the goldfish had diminished (Chivers & Smith 1994). European minnows (Phoxinus phoxinus, Cyprinidae) that had acquired a response to the odour of both their natural predator (pike) and a non-piscivorous cichlid showed a stronger response towards the pike (spent more time schooling and less time foraging; Magurran 1989). Prey may be predisposed to learn a response to generalised cues that indicate a fish is likely to be predatory. If prey fishes encounter varying levels of predation risk at different life stages, then both their response to predators and their ability to learn from predator encounters may be variable. For example, Hawkins et al. (2008) investigated the learning ability of different age groups of predator-na¨ıve salmon by exposing them to predator (pike) odour paired with either alarm cues or distilled water (control). Salmon aged 16–20 weeks showed a learned response when they were subsequently exposed to pike odour but there was no effect of treatment (alarm cues or distilled water) for fishes in the younger age group (3 weeks old). These findings suggest that age-dependent learning allows prey to respond to changes in risk at particular life stages, in this case an increase in risk upon emergence from the substrate at around 7 weeks of age (Hawkins et al. 2008).

3.2.3.2

Learning specificity

Karplus & Algom (1981) tested the idea of predisposed antipredator responses towards visual cues by conducting a detailed morphometric analysis of the facial features and feeding habitats of 105 species of reef fishes. Although a large number of facial features were associated with a piscivorous habit, fishes with large mouths and eyes that are widely spaced

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

Char Count=

45

tended to be predatory. Karplus et al. (1982) then tested this experimentally, confirming that chromis (Chromis caeruleus, Pomacentridae) showed a stronger response to models with these facial characteristics than they did to models with non-predatory features. Similar generalised visual features of predators may allow prey to learn a response to piscivorous species more rapidly than non-predatory species. Csanyi & Doka (1993) tested this idea by presenting paradise fish (Macropodus opercularis, Anabantidae) with either a live goldfish or fish models of varying realism in conjunction with an electric shock. Paradise fish showed the strongest learned response towards the goldfish and were more likely to learn a response towards fish models with lateral eye-like spots than models with only one eye or no eyes (Csanyi & Doka 1993). Learning specificity allows prey fishes to readily associate particular predatory cues with increased risk. However, the same mechanism may also allow prey fishes to generalise their learned antipredator responses to other (e.g. taxonomically related) predators with similar features. This concept has been termed ‘generalisation of predator recognition’ and suggests that prey are more likely to display an antipredator response to a novel predator if it is closely related to a predator they have recently learned to recognise (Ferrari et al. 2007). A similar process may allow predators to generalise their experience with prey to other species with similar morphologies. Generalisation of predator recognition by prey is developed further by Brown and colleagues in Chapter 3.

3.2.3.3

Search images

An animal’s brain can only process a limited amount of information; therefore, predators are predicted to search for only one prey type at any given time, a ‘search image’ (Tinbergen 1960; Dukas 1998). The development of a search image depends on the predator’s experience with particular prey types. Predators develop a search image based on the prey that is encountered most frequently and once this prey type becomes less abundant or less profitable, a new search image will be developed (Dukas 1998). However, switching between search images is potentially costly because the predator suffers from divided attention whilst attempting to learn the new prey image resulting in reduced foraging efficiency (Croy & Hughes 1991). Animals that are able to rapidly adjust their foraging behaviour in response to changes in prey availability will have high fitness. Ehlinger (1989) investigated the influence of experience on the foraging behaviour of bluegill sunfish (Lepomis macrochirus, Centrarchidae) in both open water and vegetated habitats. In both habitat types, sunfish adapted their foraging behaviour according to the conspicuousness of the prey. Sunfish that foraged in the open water habitat learned to increase their searching speed, whereas those in the vegetated habitat where prey were more cryptic learned to search more slowly (Ehlinger 1989). If search images for a given prey morphology are acquired more easily and retained in the memory for longer, then selection will be stronger on memorable prey types than on those that are readily forgotten (Endler 1991). In this way, we can envisage how experience with particular prey morphs (which is dependent on the environment, as illustrated by sunfish in the above example) creates variation in predation pressure that is largely influenced by the cognitive ability of predators (see Chapter 2 for further discussion).

BLBK374-03

BLBK374-Brown

46

3.2.3.4

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Aposematism and mimicry

Prey that are aposematic rely on their conspicuousness to warn potential predators of their unpalatability. Theoretical models suggest that in order for aposematism to evolve, predators must rapidly learn to associate the noxiousness of the prey with its conspicuous colour patterns (Speed 2001; Ruxton et al. 2004). If conspicuous colouration is to act as a warning signal, it is predicted that predators learn to associate noxiousness with a conspicuous colour pattern more rapidly than with a cryptic colour pattern (Guildford 1990). There is some evidence for this in birds; chicks learn to avoid conspicuous baits more rapidly than they learn to avoid cryptic distasteful baits (Roper & Wistow 1986; Roper & Redston 1987). Interestingly, aposematism appears to be uncommon in fishes. This is perhaps because less than 3% of teleosts (the majority of which are marine) possess toxic or noxious chemicals in the skin, spines or viscera (Godin 1997). There is evidence for aposematism in marine invertebrate prey such as nudibranchs, which are brightly coloured and often possess chemical and/or physical defences (e.g. nematocysts) that make them distasteful to fish predators (Edmunds 1974). Aguado & Marin (2007) investigated whether the colour and shape of the aeolid nudibranch (Cratena peregrina) might function as a warning signal to one of its fish predators, the ornate wrasse (T. pavo). Wrasse were presented with nudibranch models that were either matched for colour but differed in shape (with or without dorsal protuberances) or matched for shape but differed in colour (all blue or white body with orange protuberances). At first presentation, wrasse preferred to attack the models with the dorsal protuberances over those without these projections but did not discriminate between the different coloured models. When the models were presented in combination with nematocysts containing distasteful physical/chemical cues, attack rates towards these models were diminished and this occurred most rapidly in models that differed in colour versus those that differed in shape. The post-training test (no distasteful cues present) revealed that fish had learned to avoid the noxious models on the basis of both model shape and colour. Batesian mimicry, in which a palatable prey imitates an unpalatable species, relies on predators associating both the mimic and the model species with unpalatability. In fishes, the colour patterns of the leatherjacket (Paraluteres prionurus, Monacanthidae) closely resemble those of its putative Batesian mimic, a toxic pufferfish species (Canthigaster valentini, Tetraodontidae). In a study designed to investigate the effectiveness of mimicry in relation to the degree of resemblance in colour pattern between the pufferfish species and its mimic, Caley & Schluter (2003) showed that painted model replicas that most resembled the pufferfish received fewer visits from piscivorous fish. In another study with reef fish, Cheney (2008) found that staghorn damselfish (Plagiotremus rhinorhynchos) avoided bluestreak cleaner fish (Labroides dimidiatus) and their aggressive mimics (blue striped fangblennys, P. rhinorhynchos) following aggressive interactions with the latter. However, avoidance learning by damselfish was more pronounced when fangblennys displayed nonmimic colouration rather than mimic colouration (the species can rapidly switch between the different colour forms). Here, both model–mimic colour pattern similarity and the damselfish’s (‘receiver’) learning ability appear to play a role in maintaining this system of aggressive mimicry. Considering the recent interest in predator learning and memory, it is surprising that few studies with fishes have addressed the role of learning in this context.

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

3.2.4 3.2.4.1

Char Count=

47

Approach Pursuit deterrence

Fishes that detect and recognise a predator frequently perform what is referred to as inspection behaviour, where a small group of fish leaves the shoal and slowly approaches the predator, often swimming slowly along the length of its body, before returning to the shoal (Pitcher et al. 1986). One function of this apparently paradoxical behaviour may be pursuit deterrence (also known as attack inhibition) in which the approach ‘informs’ the predator that it has been detected (Magurran 1990; Dugatkin & Godin 1992; Godin & Davis 1995; Brown et al. 1999). Fin flicking in glowlight tetras (Hemigrammus erythrozonus, Characidae; Brown et al. 1999) and head bobbing in gobies (Asterropteryx semipunctatus and Gnatholepsis anjerensis, Gobiidae; Smith & Smith 1989) have been proposed to serve a similar function. Predators are more successful at attacking prey that may be less vigilant (e.g. when foraging; Krause & Godin 1996) and consequently the predator may redirect its attention towards fishes that have not signalled their awareness (Pitcher & Parrish 1993). Magurran (1990) and Dugatkin & Godin (1992) provide evidence that predators are less likely to attack fishes that are inspecting than those that are not inspecting. It has previously been suggested that approach behaviour (and behaviours that are associated with predator detection, e.g. head bobbing and fin flicking) might serve as warning or ‘alarm signals’, which alert nearby conspecifics of the location of the predator and its potential danger (Smith & Smith 1989). A further proposed function of predator inspection behaviour is the preferential selection of bolder males by females (Godin & Dugatkin 1996). 3.2.4.2

Gaining information about the predator

Another function of inspection behaviour is that it allows the inspectors to gain information about the predator, such as its condition and motivation to attack. This information may be more reliable than that gained ‘second-hand’ by observing the inspection behaviour of other individuals (see Subsection 3.2.4.3). Several studies have suggested that the behaviour of inspectors is changed as a result of inspection, suggesting that information has been acquired about the predator’s attack motivation (Pitcher et al. 1986; Magurran & Higham 1988; Pitcher 1992). The predator’s posture affects reaction distance in three-spot damselfish (Stegastes planifrons, Pomacentridae; Helfman 1989) and allows inspecting guppies to differentiate between hungry and satiated predators (Licht 1989). Glowlight tetras and finescale dace (P. neogaeus, Cyprinidae) are more reluctant to inspect and do so from further away and in smaller groups if they detect alarm cues in the odour of the predator (Brown & Godin 1999; Brown et al. 2000; Brown et al. 2001; Brown & Schwarzbauer 2001; Brown & Dreier 2002). When tetras were subsequently presented with only visual cues from a cichlid, only those tetras that had previously been exposed to the predator and its dietary cues displayed a response (Brown & Godin 1999). 3.2.4.3

Social learning

Fishes that do not take part in the inspection change their behaviour when the inspector(s) returns, suggesting that information is socially transmitted to the shoal (Pitcher et al. 1986;

BLBK374-03

BLBK374-Brown

48

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Chapter 11). Minnows that could not see a pike model reduced their level of activity after observing the ‘skittering’ behaviour (a startle response) of inspecting fish (Magurran & Higham 1988). Social learning occurs when na¨ıve individuals modify their behaviour after observing conspecifics responding fearfully to a particular stimulus, and allows individuals to acquire information about the predator without incurring the potential costs (increased predation risk) associated with gaining the information independently (Box 1984). Note that opportunities for learning from conspecifics (or heterospecifics) could arise at other stages of the interaction, such as during predator detection. However, we discuss it here because of its demonstrated relevance to inspection behaviour. Changes in the behaviour of the inspectors (including signals that may act as pursuit deterrents) could provide clues to the non-inspectors regarding the level of threat posed by the predator. This would parallel the situation in which successful foragers inadvertently disclose information about good foraging sites through changes in their behaviour (Pitcher et al. 1982; Pitcher & Parrish 1993). This assumes that the behaviour of the inspector (either during inspection or when returning to the shoal) is a reliable indicator of the behaviour and motivation of the predator. Because inspection behaviour is a risk-sensitive behaviour, balancing the potential risk of being captured against the benefits gained through information acquisition (Murphy & Pitcher 1991), this seems quite likely. Suboski et al. (1990) were the first to demonstrate that conspecific alarm cues can be socially transmitted among na¨ıve conspecifics (Brown & Laland 2003; Chapter 11). In their study, zebrafish were conditioned to respond to morpholine (an artificial odour) by presenting it in conjunction with conspecific alarm cues. Na¨ıve observers displayed an alarm response after observing (through a clear barrier) conditioned fish responding to morpholine and subsequently retained this response when later tested alone (Suboski et al. 1990). Mathis et al. (1996) similarly showed that na¨ıve fathead minnows can learn to recognise a novel predator odour by observing conspecifics displaying a conditioned fright response. Alarm reactions can also be transmitted among heterospecifics. In Mathis et al.’s study (1996), brook stickleback, a species that forms mixed shoals with fathead minnows, acquired a fright reaction after observing minnows displaying a conditioned response. Krause (1993) also reported transmission of information among heterospecifics by demonstrating that sticklebacks exhibited a fright response after observing chub (Leuciscus cephalus, Cyprinidae) responding to alarm cues. In all of these experiments, fishes that had never been exposed to the alarm substance were able to learn a fright reaction to a novel odour. This suggests that the response acquired from detecting an alarm cue is similar to that learned by observing the conditioned fright response of conspecifics. It would be interesting to know whether these two sources of information (individual and social) are equivalent, and, specifically, which behaviours are involved in this process. It has been suggested that dashing movements (Chivers & Smith 1994) or the position of the fish relative to the substratum might be important (Griffin 2004). Social learning may partly account for the rapid acquisition of learned predator recognition when novel predators are introduced into a previously predator-free population. Following the introduction of pike into two previously pike-free populations (of approximately 20,000 and 78,000 fish), minnows learned to recognise pike odour within just 14 days (Chivers & Smith 1995c). For a full discussion of the role of social learning in fishes, see Chapter 11.

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

3.2.4.4

Char Count=

49

Habituation

Importantly, inspection behaviour may play an important role in habituation, a type of learning in which there is a reduced response after repeated exposure to a stimulus (Shettleworth 1988). Inspectors often repeatedly approach the predator, providing an ongoing assessment of the motivation and likely risk posed by it (Pitcher & Parrish 1993). Inspecting fishes on acquiring information suggesting that a predator is not actively hunting may discontinue inspection if they perceive that there is little threat, or they may continue to inspect the predator but show a less cautious response during subsequent approaches. In this manner, habituation may play an important part in allowing prey to display threat-sensitive antipredator responses. Habituation is less likely to occur towards stimuli that prey fishes are predisposed to show a fear response to, but experience also mediates the response. Csanyi (1985) found that paradise fish became habituated to both a goldfish and a satiated pike, but whereas inspection rate rapidly diminished towards the goldfish, paradise fish continued to approach the pike. Magurran & Girling (1986) presented European minnows with pike models that differed in realism. Their results confirmed that minnows showed a stronger response towards the most realistic model and that they habituated most rapidly to the least realistic model. If predator inspection does indeed function as a pursuit deterrence signal, then predators are also acquiring information during this interaction. This is an intriguing possibility that has so far been little considered. Krause & Godin (1996) reported that subtle differences in prey foraging posture influence the attack success of predators. The predatory cichlid – the blue acara (Aequidens pulcher, Cichlidae) – preferred to attack foraging guppies rather than non-foraging ones, and guppies that foraged in a ‘nose-down’ position rather than horizontally (Krause & Godin 1996). This is probably because of the reduced vigilance of foraging prey in a head-down posture. Prey that are foraging or in nose-down postures may also have difficulty performing a fast start or C-start escape response, although this explanation is unlikely in this case as the experimental design controlled for different positional effects (Krause & Godin 1996). It would be very interesting to investigate whether the recognition of these and other vigilance cues is based on previous hunting experience.

3.2.5

Evasion

Prey fishes that are under imminent attack from a predator may either flee, freeze or hide in an attempt to avoid being captured (Edmunds 1974). Acquisition of the latter two responses has been discussed in Subsections 3.2.2 and 3.2.3 in the context of avoiding detection and learned predator recognition. We know of no studies that demonstrate that previous experience with an attacking predator enhances the freezing or hiding responses of prey. This is probably because these responses are used in combination with morphological defences or when the predator is very close and escape seems unlikely (Godin 1997). Fleeing is the more common response to an approaching predator (Edmunds 1974) and several studies demonstrate that the timing, swimming speed and trajectory of the flight response can be improved through experience.

BLBK374-03

BLBK374-Brown

50

3.2.5.1

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Reactive distance and escape speed and trajectory

Dill (1974) found that the reactive distance of zebra danios, the distance between the prey and the moving stimulus when the prey displays an escape response, is increased through repeated exposure to a film of an approaching object or a model predator. Escape velocity also increased with experience of the filmed predator, but not with experience of the predator model (Dill 1974). Huntingford et al. (1994) demonstrated that the escape response of three-spined stickleback fry is contingent on both their population of origin and their previous experience. If stickleback fry stray from their nest, the father chases them and carries them back to the nest in his mouth. Huntingford et al. (1994) found that the escape speed of fry and the retrieval speed of fathers were highly correlated, but that the speeds were significantly greater for fry from high-risk populations than low-risk ones. Fleeing prey often display a zigzag swimming trajectory (Edmunds 1974), and the angle at which they initially flee and the number of turns executed can affect the success of evading capture (Godin 1997). For example, insects such as cockroaches display a fleeing trajectory that is non-random (as that would occasionally entail movement toward the predator) but encompasses a variety of escape routes at fixed angles from the threatening stimulus (Domencini et al. 2008). In fishes, stickleback fry from a high-risk population are more likely to flee at an angle from a model predator than fishes from a low-risk population (Huntingford et al. 1994). These unpredictable or ‘protean’ movements commonly displayed by fleeing prey are thought to make it difficult for predators to learn fixed escape patterns (Domencini & Blake 1993; Godin 1997). It would be very interesting to explore this idea and incorporate models of predator learning as well as prey fleeing trajectories. A prey’s escape trajectory may also relate to its lateralization, which refers to the selective partitioning of cognitive function in either the left or right hemisphere of the brain (see Chapter 14 for more on lateralization and antipredator behaviour). Although fishes from high predation populations are strongly lateralized relative to low predation fishes (Brown et al. 2004), a study with golden topminnows (Girardinus falcatus) found that escape responses of lateralized and non-lateralized fishes were similar, irrespective of whether a visual predatory stimulus was presented in the left or right visual field of the test fish (Agrillo et al. 2009).

3.2.5.2

Survival benefits/capture success

A large number of studies have demonstrated that prior experience with live predators increases their subsequent chance of survival in a predator encounter. Patten (1977) found that the predation mortality of coho salmon (Oncorhynchus kisutch) was lower when fish had previous experience with predatory sculpin or when they had contact with predatorexperienced salmon than that for predator-na¨ıve salmon. Similar findings were reported by Olla & Davis (1989). Mirza & Chivers (2000) were the first to demonstrate that learned recognition of predators through conditioning with alarm cues confers a survival advantage. Fathead minnows that were conditioned with pike odour and alarm cues survived significantly longer in subsequent trials with live pike. In their study, Mirza and Chivers found positive correlations between shelter use and survival time and between shoaling behaviour and survival time, suggesting that these behaviours were particularly important

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

Char Count=

51

(Mathis & Smith 1993). Parallel results have been presented for brook trout (Salvelinus fontinalis, Salmonidae) exposed to live predatory chain pickerel (Esox niger, Esocidae); trout conditioned with alarm cues had a higher survival rate than those conditioned with odour not containing alarm cues (Mirza & Chivers 2000). Although many fishes can be trained to respond to novel predator odours under experimental conditions, this ability does not necessarily increase their survival in the wild. Hawkins et al. (2007) found that hatchery reared salmon smolts that had been conditioned to respond to pike odour were no more likely than control fishes to survive following release into a Scottish river system in which pike are the main predator. The high mortality of both conditioned and control fishes was probably due to the tendency of the hatchery-reared smolts to remain in lochs for much longer than their wild conspecifics. Thus, learned antipredator responses appear to be overridden by the non-adaptive migratory behaviour of hatchery-reared smolts. Visual cues can also contribute to learned avoidance behaviour. Berejikian (1995) showed that fry of hatchery-reared steelhead trout (O. mykiss, Salmonidae) suffered higher predation mortality than their wild counterparts. After observing sculpin prey on trout through a clear barrier, both wild and hatchery fishes showed improved survival skills but wild fishes survived for longer. The experienced of being chased when young has been shown to confer survival benefits in adult guppies and sticklebacks (Goodey & Liley 1986; Tulley & Huntingford 1987). Few studies with piscivorous fishes have investigated how predators learn to capture their prey, probably because of the ethical issues associated with such experiments. However, a study with hatchery-reared jade perch (Scortum barcoo) found that previous experience with live mosquitofish (Gambusia holbrooki) enhanced capture success compared with perch that had previously been fed a diet of freshly killed mosquitofish or fish food pellets (Reid et al. 2010). There are also suggestions in the literature that the hunting behaviour of predators is enhanced through social mechanisms; perch were more successful foraging as a shoal than when foraging alone (Eklov 1992). Again, these ideas remain to be tested. Interestingly, mathematical models that incorporate predator learning (both the individual experience of prey handling and social transmission of learned behaviour) predict increased foraging specialisations by predators (Tinker et al. 2009).

3.3

Summary and discussion

Predator–prey interactions are complex and learning about predators does not necessarily have to be a life-or-death experience. There are a variety of mechanisms that allow prey to learn about their predators, some of which are surprisingly subtle. Our understanding of the role of chemical cues, particularly alarm cues, is far ahead of that in other sensory domains and in many cases, the mechanisms by which learning is achieved remain unresolved. The study of animal cognition is still relatively new and has developed through two distinct research disciplines: psychology and ecology (Kamil 1998). These two perspectives still require integration and we know little of the importance of learning specificity, memory constraints, habituation and reinforcement in shaping the behaviour of predators and prey in the wild.

BLBK374-03

BLBK374-Brown

52

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

One of the most striking things to arise from our review of the literature is the lack of research on the learning abilities of predators. In particular, the idea that many of the bright, conspicuous colour patterns of prey have evolved as a result of the memory and learning capabilities of predators is a very exciting notion that requires further investigation. Although predator psychology has been considered in the context of optimal foraging, these investigations often overlook factors affecting prey conspicuousness such as movement and colouration. Experiments are needed to test the learning and memory capability of predators for particular prey colour patterns and movements against a variety of backgrounds (Endler 1986). The use of computer animations is becoming increasingly popular in behavioural ecology and would be one way to disentangle the multiple effects of behaviour, colouration, background and movement. We have suggested that opportunities for learning are likely to be greater at earlier stages of the predator–prey interaction when the predation risk to individual prey is relatively low and antipredator behaviours are less costly of energy than at the later stages. This review of the literature has revealed examples of learning at every stage of the sequence with no obvious bias of studies towards any particular stage. However, it would be relatively simple to design a set of experiments to test the learning ability of prey at each stage of the sequence. Although there is some evidence for learned avoidance of dangerous habitats, we know little about the influence of predator activity patterns on prey behaviour. Prey fishes can learn to recognise and respond to a novel predator but do they alter the timing of their activities as a result of previous encounters and can predators learn to ‘track’ the behaviour of their prey? Interactions between predators and prey often involve a diversity of species and more than just one predator and prey. Importantly, the combined effects of multiple predators can be very different to those arising from pairwise predator–prey interactions (reviewed by Sih et al. 1998), particularly when a prey’s response to one predator increases its predation risk to another (Charnov et al. 1976). Studies of generalisation of predator recognition have an important role to play in determining the learned response of prey to multiple predators and it would be interesting to know if the same mechanisms apply to predator learning of prey phenotypes. Such experiments would allow us to predict the response of populations to changes in species assemblages, such as those occurring through the introduction or invasion of exotic species. An exciting area of future research is to investigate how predator–prey interactions proceed when both predators and prey learn. A comparative analysis of predator–prey pairs comprising 277 species of fishes found a positive correlation between predator and prey brain sizes (Kondoh 2010). If brain size can be considered a proxy for cognitive ability, then there is some support for the role of cognition in the predator–prey arms race – evolution of larger brained prey results in selection for larger brained predators. Selection experiments are necessary to disentangle these evolutionary processes and understand the consequences of predator–prey learning on a larger scale (Kondoh 2010).

Acknowledgements We wish to thank Peter Banks, Culum Brown and John Endler for their valuable comments on an earlier version of this chapter. Jennifer L. Kelley is funded by a University of Western Australia Postdoctoral Research Fellowship.

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

Char Count=

53

References Agrillo, C., Dadda, M. & Bisazza, A. (2009) Escape behaviour elicited by a visual stimulus. A comparison between lateralised and non-lateralised female topminnows. Laterality: Assymetries of Body, Brain and Cognition, 14, 300–314. Aguado, F. & Marin, A. (2007) Warning coloration associated with nematocyst-based defences in aeolidiodean nudibranchs. Journal of Molluscan Studies, 73, 23–28. Aronson, L.R. (1971) Further studies on orientation and jumping behaviour in the gobiid fish, Bathygobius soporator. Annals of the New York Academy of Sciences, 188, 378–407. Berejikian, B.A. (1995) The effects of hatchery and wild ancestry and experience on the relative ability of steelhead trout fry (Oncorhynchus mykiss) to avoid a benthic predator. Canadian Journal of Fisheries and Aquatic Sciences, 52, 2476–2482. Berger, J., Swenson, J.E. & Persson, I.-L. (2001) Recolonizing carnivores and na¨ıve prey: conservation lessons from Pleistocene extinctions. Science, 291, 1036–1039. Box, H.O. (1984) Primate Behavior and Social Ecology. Chapman & Hall, London. Brown, C. (2003) Habitat-predator association and avoidance in rainbowfish (Melanotaenia spp.). Ecology of Freshwater Fish, 12, 118–126. Brown, C., Gardner, C. & Braithwaite, V.A. (2004) Population variation in lateralized eye use in the Poeciliid Brachyraphis episcopi. Proceedings of the Royal Society of London, 271, S455–S457. Brown, C. & Laland, K.N. (2003) Social learning in fishes: a review. Fish and Fisheries, 4, 280–288. Brown, C. & Warburton, K. (1997) Predator recognition and antipredator responses in the rainbowfish, Melanotaenia eachamensis. Behavioural Ecology and Sociobiology, 41, 61–68. Brown, G.E. (2003) Learning about danger: chemical alarm cues and local risk assessment in prey fishes. Fish and Fisheries, 4, 227–234. Brown, G.E. & Dreier, V.M. (2002) Predator inspection behaviour and attack cone avoidance in a characin fish: the effects of predator diet and prey experience. Animal Behaviour, 63, 1175–1181. Brown, G.E. & Godin, J.-G.J. (1999) Who dares, learns: chemical inspection behaviour and acquired predator recognition in a characin fish. Animal Behaviour, 57, 475–481. Brown, G.E., Godin, J.-G.J. & Pederson, J. (1999) Fin-flicking behaviour: a visual anti-predator alarm signal in a characin fish, Hemigrammus erythrozonus. Animal Behaviour, 58, 469–475. Brown, G.E., Golub, J.L. & Plata, D.L. (2001) Attack cone avoidance during predator inspection visits by wild finescale dace (Phoxinus neogaeus): the effects of predator diet. Journal of Chemical Ecology, 27, 1657–1666. Brown, G.E., Paige, J.A. & Godin, J.-G.J. (2000) Chemically mediated predator inspection behaviour in the absence of predator visual cues by a characin fish. Animal Behaviour, 60, 315–321. Brown, G.E. & Schwarzbauer, E.M. (2001) Chemical predator inspection and attack cone avoidance in a characin fish: the effects of predator diet. Behaviour, 138, 727–739. Caley, M.J. & Schluter, D. (2003) Predators favour mimicry in a tropical reef fish. Proceedings of the Royal Society of London B, 270, 667–672. Chapman, B.B., Morrell, L.J., Tosh, C.R. & Krause, J. (2010) Behavioural consequences of sensory plasticity in guppies. Proceedings of the Royal Society of London, B. doi: 10.1098/rspb.2009.2055. Charnov, E.L., Orians, G.H. & Hyatt, K. (1976) Ecological implications of resource depression. American Naturalist, 110, 247–259. Cheney, K.L. The role of avoidance learning in an aggressive mimicry system. Behavioral Ecology, 19, 583–588. Chivers, D.P. & Smith, R.J.F. (1994) Fathead minnows, Pimephales promelas, acquire predator recognition when alarm substance is paired with the sight of an unfamiliar fish. Animal Behaviour, 48, 597–605. Chivers, D.P. & Smith, R.J.F. (1995a) Chemical recognition of risky habitats is culturally transmitted among fathead minnows, Pimephales promelas (Osteichthyes, Cyprinidae). Ethology, 99, 286–296. Chivers, D.P. & Smith, R.J.F. (1995b) Fathead minnows, Pimephales promelas, learn to recognise chemical stimuli from high risk habitats by the presence of alarm substance. Behavioural Ecology, 6, 155–158.

BLBK374-03

BLBK374-Brown

54

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Chivers, D.P. & Smith, R.J.F. (1995c) Free-living minnows rapidly learn to recognize pike as predators. Journal of Fish Biology, 46, 949–954. Chivers, D.P. & Smith, R.J.F. (1998) Chemical alarm signalling in aquatic predator–prey systems: a review and prospectus. Ecoscience, 5, 338–352. Cott, H.B. (1940) Adaptive Coloration in Animals. Methuen, London. Croy, M.I. & Hughes, R.N. (1991) The role of learning and memory in the feeding behavior of the fifteen-spines stickleback, Spinachia spinachia L. Animal Behaviour, 41, 149–159. Csanyi, V. (1985) Ethological analysis of predator avoidance by the paradise fish (Macropodus opercularis L.) 1. Recognition and learning of predators. Behaviour, 92, 227–240. Csanyi, V. & Doka, A. (1993) Learning interactions between prey and predator fish. Marine Behaviour and Physiology, 23, 63–78. Dill, L.M. (1974) The escape response of the zebra danio (Brachydanio rerio) II. The effect of experience. Animal Behaviour, 22, 723–730. Dodson, S.I., Tollrian, R. & Lampert, W. (1997) Daphnia swimming behaviour and vertical migration. Journal of Plankton Research, 19, 969–978. Domencini, P. & Blake, R.W. (1993) Escape trajectories in angelfish. Journal of Experimental Biology, 177, 253–272. Domencini, P., Booth, D., Blagburn, J.M. & Bacon, J.P. (2008) Cockroaches keep predators guessing by using preferred escape trajectories. Current Biology, 15, 1792–1796. Dugatkin, L.A. & Godin, J.-G.J. (1992) Prey approaching predators: a cost-benefit perspective. Annals Zoologica Fennici, 29, 233–252. Dukas, R. (1998) Constraints on information processing and the effects on behaviour. In: R. Dukas (ed) Cognitive Ecology, pp. 89–127. Chicago University Press, Chicago. Edmunds, M. (1974) Defense in Animals: A Survey of Anti-predator Defenses. Longmans, London. Ehlinger, T.J. (1989) Learning and individual variation in bluegill foraging: habitat-specific techniques. Animal Behaviour, 38, 643–658. Eklov, P. (1992) Group foraging versus solitary foraging efficiency in piscivorous predators: the perch, Perca fluviatilis, and pike, Esox lucius, patterns. Animal Behaviour, 44, 313–326. Ellis, T., Howell, B.R. & Hughes, R.N. (1997) The cryptic responses of hatchery-reared sole to a natural sand substratum. Journal of Fish Biology, 51, 389–401. Endler, J.A. (1986) Defense against predators. In: J.A. Endler (ed) Predator–Prey Relationships: Perspectives and Approaches from the Study of Lower Vertebrates, pp. 109–134. University of Chicago Press, Chicago. Endler, J.A. (1991) Interactions between predators and prey. In: J.A. Endler (ed) Behavioural Ecology: An Evolutionary Approach, p. 482. Blackwell Publishing Ltd., Oxford. Fairchild, E.A. & Howell, W.H. (2004) Factors affecting the post-release survival of cultured juvenile Pseudopleuronectes americanus. Journal of Fish Biology, 65, 69–87. Ferrari, M.C.O., Gonzalo, A., Messier, F. & Chivers, D.P. (2007) Generalization of learned predator recognition: an experimental test and framework for future studies. Proceedings of the Royal Society of London Series B, 274, 1853–1859. Ferrari, M.C.O., Sih, A. & Chivers, D.P. (2009) The paradox of risk allocation: a review and prospectus. Animal Behaviour, 78, 579–585. Fraser, D.F., Gilliam, J.F., Akkara, J.T., Albanese, B.W. & Snider, S.B. (2004) Night feeding by guppies under predator release: effects on growth and daytime courtship. Ecology, 85, 312–319. Fuiman, L.A. & Magurran, A.E. (1994) Development of predator defences in fishes. Reviews in Fish Biology and Fisheries, 4, 145–183. Gendron, R.P. & Staddon, J.E.R. (1983) Searching for cryptic prey: the effect of search rate. American Naturalist, 121, 172–186. Godin, J.-G.J. (1997) Evading predators. In: J.-G.J. Godin (ed) Behavioural Ecology of Teleost Fishes, pp. 191–236. Oxford University Press, Oxford. Godin, J.-G.J. & Davis, S.A. (1995) Who dares, benefits: predator approach behaviour in the guppy (Poecilia reticulata) deters predator pursuit. Proceedings of the Royal Society of London Series B, 259, 193–200.

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

Char Count=

55

Godin, J.-G.J. & Dugatkin, L.A. (1996) Female mating preferences for bold males in the guppy (Poecilia reticulata). Proceedings of the National Academy of Sciences USA, 93, 10262– 10267. Goodey, W. & Liley, N.R. (1986) The influence of early experience in escape behaviour in the guppy (Poecilia reticulata). Canadian Journal of Zoology, 64, 885–888. Griffin, A.S. (2004) Social learning by predators: a review and prospectus. Learning and Behavior, 32, 131–140. Griffin, A.S., Evans, C.S. & Blumstein, D.T. (2001) Learning specificity in acquired predator recognition. Animal Behaviour, 62, 577–589. Guildford, T. (1990) The evolution of aposematism. In: T. Guildford (ed) Insect Defense: Adaptive Mechanisms and Strategies of Prey and Predators, pp. 23–61. State University of New York Press, New York. Hall, D. & Suboski, M.D. (1995) Visual and olfactory stimuli in learned release of alarm reactions by zebra danio fish (Brachydanio rerio). Neurobiology of Learning and Memory, 63, 229– 240. Hartman, E.J. & Abrahams, M.V. (2000) Sensory compensation and the detection of predators: the interaction between chemical and visual information. Proceedings of the Royal Society of London B, 267, 571–575. Hawkins, L.A., Armstrong, J.A. & Magurran, A.E. (2007) A test of how predator conditioning influences survival of hatchery-reared Atlantic salmon, Salmo salar, in restocking programmes. Fisheries Management and Ecology, 14, 291–293. Hawkins, L.A., Magurran, A.E. & Armstrong, J.D. (2008) Ontogenetic learning of predator recognition in hatchery-reared Atlantic salmon, Salmo salar. Animal Behaviour, 75, 1663–1671. Helfman, G.S. (1986) Behavioral response of prey fishes during predator–prey interactions. In: G.S. Helfman (ed) Predator–Prey Relationships, pp. 135–156. Chicago University Press, Chicago. Helfman, G.S. (1989) Threat-sensitive predator avoidance in damselfish–trumpetfish interactions. Behavioral Ecology and Sociobiology, 24, 47–58. Helfman, G.S. & Schultz, E.T. (1984) Social transmission of behavioural traditions in a coral reef fish. Animal Behaviour, 32, 379–384. Houde, A.E. (1997) Sex, Colour, and Mate Choice in Guppies. Princeton University Press, Princeton, N.J. Howell, B.R. (1994) Fitness of hatchery-reared fish for survival in the sea. Aquaculture and Fisheries Management, 25, 3–17. Huntingford, F.A. & Wright, P.J. (1992) Inherited population differences in avoidance conditioning in three-spined sticklebacks, Gasterosteus aculeatus. Behaviour, 122, 264–273. Huntingford, F.A., Wright, P.J. & Tierney, J.F. (1994) Adaptive variation in antipredator behaviour in threespine stickleback. In: F.A. Huntingford, P.J. Wright & J.F. Tierney (eds) The Evolutionary Biology of the Threespine Stickleback, pp. 277–296. Oxford University Press, Oxford. Ioannou, C.C. & Krause, J. (2009) Interactions between background matching and motion during visual detection can explain why cryptic animals keep still. Biology Letters, 5, 191–193. Johnsson, J.K. & Kj¨allman-Eriksson, K. (2008) Cryptic prey colouration increases search time in brown trout (Salmo trutta): effects of learning and body size. Behavioral Ecology and Sociobiology, 62, 1613–1620. Kamil, A.C. (1998) On the proper definition of cognitive ethology. In: A.C. Kamil (ed) Animal Cognition in Nature, pp. 1–28. Academic Press, San Francisco, CA. Karplus, I. & Algom, D. (1981) Visual cues for predator face recognition by reef fishes. Zeitschrift f¨ur Tierpsychologie, 55, 343–364. Karplus, I., Goren, M. & Algom, D. (1982) A preliminary experimental analysis of predator face recognition by Chromis caerulus (Pisces, Pomacentridae). Zeitschrift f¨ur Tierpsychologie, 58, 53–65. Kats, L.B. & Dill, L.M. (1998) The scent of death: chemosensory assessment of predation risk by prey animals. Ecoscience, 5, 361–394. Keenleyside, M.H.A. (1979) Diversity and Adaptation in Fish Behaviour. Springer Verlag, New York.

BLBK374-03

BLBK374-Brown

56

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Kislalioglu, M. & Gibson, R.N. (1976) Some factors governing prey selection by the 15-spined stickleback, Spinachia spinachia (L.). Journal of Experimental Marine Biology and Ecology, 25, 159–169. Kondoh, M. (2010) Linking learning adaptation to trophic interactions: a brain size-based approach. Functional Ecology, 24, 35–43. Krause, J. (1993) Transmission of fright reaction between different species of fish. Behaviour, 127, 35–48. Krause, J. & Godin, J.-G.J. (1996) Influence of prey foraging posture on predator detectability and predation risk: predators take advantage of unwary prey. Behavioral Ecology, 7, 264–271. Kristensen, E.A. & Closs, G.P. (2004) Anti-predator response of na¨ıve and experienced common bully to chemical alarm cues. Journal of Fish Biology, 64, 643–652. Licht, T. (1989) Discrimination between hungry and satiated predators: the response of guppies (Poecilia reticulata) from high and low predation sites. Ethology, 82, 238–243. Lima, S.L. (1998) Stress and decision making under the risk of predation: recent developments from behavioral, reproductive, and ecological perspectives. Advances in the Study of Animal Behaviour, 27, 215–290. Lima, S.L. (2002) Putting predators back into behavioral predator–prey interactions. Trends in Ecology and Evolution, 17, 70–75. Lima, S.L. & Bednekoff, P.A. (1999) Temporal variation in danger drives antipredator behaviour: the predation risk allocation hypothesis. American Naturalist, 153, 649–659. Lima, S.L. & Dill, L.M. (1990) Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology, 68, 619–640. Magurran, A.E. (1989) Acquired recognition of predator odour in the European minnow (Phoxinus phoxinus). Ethology, 82, 216–223. Magurran, A.E. (1990) The inheritance and development of minnow anti-predator behaviour. Animal Behaviour, 39, 834–842. Magurran, A.E. & Girling, S.L. (1986) Predator model recognition and response habituation in shoaling minnows. Animal Behaviour, 34, 510–518. Magurran, A.E. & Higham, A. (1988) Information transfer across fish shoals under threat. Ethology, 78, 153–158. Mathis, A., Chivers, D.P. & Smith, J.F. (1996) Cultural transmission of predator recognition in fishes: intraspecific and interspecific learning. Animal Behaviour, 51, 185–201. Mathis, A. & Smith, R.J.F. (1993) Chemical alarm signals increase the survival time of fathead minnows (Pimaphales promelas) during encounters with northern pike (Esox lucius). Behavioral Ecology, 4, 260–265. Mathis, A. & Smith, R.J.F. (1993b) Chemical labelling of northern pike, Esox lucius, by the alarm pheromone of fathead minnows, Pimephales promelas. Journal of Chemical Ecology, 19, 1967–1979. Mirza, R.S. & Chivers, D.P. (2000) Predator-recognition training enhances the survival of brook trout: evidence from laboratory and field-enclosure studies. Canadian Journal of Zoology, 78, 2198–2208. Murphy, K.E. & Pitcher, T.J. (1991) Individual behavioural strategies associated with predator inspection in minnow shoals. Ethology, 88, 307–319. Neilson, J.D. & Perry, R.I. (1990) Diel vertical migrations of marine fishes – an obligate or facultative process. Advances in Marine Biology, 26, 115–168. Olla, B.L. & Davis, M.W. (1989) The role of learning and stress in predator avoidance of hatchery reared coho salmon (Oncorhynchus kisutch) juveniles. Aquaculture, 76, 209–214. Patten, B.G. (1977) Body size and learned avoidance as factors affecting predation on coho salmon fry, Oncorhynchus kisutch, by torrent sculpin, Cottus rhotheus. Fisheries Bulletin, 75, 457–459. Pettersson, L.B., Nilsson, P.A. & Bronmark, C. (2000) Predator recognition and defence strategies in crucian carp, Carassius carassius. Oikos, 88, 200–212. Pfeiffer, W. (1974) Pheromones in fish and amphibia. In: W. Pfeiffer (ed) Pheromones, pp. 269–296. North-Holland, Amsterdam.

BLBK374-03

BLBK374-Brown

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Learned Defences and Counterdefences in Predator–Prey Interactions

Char Count=

57

Pitcher, T.J. (1992) Who dares wins: the function and evolution of predator inspection behaviour in fish shoals. Netherlands Journal of Zoology, 42, 371–391. Pitcher, T.J., Green, D.A. & Magurran, A.E. (1986) Dicing with death: predator inspection behaviour in minnow shoals. Journal of Fish Biology, 28, 439–448. Pitcher, T.J., Magurran, A.E. & Winfield, I. (1982) Fish in large shoals find food faster. Behavioral Ecology and Sociobiology, 10, 149–151. Pitcher, T.J. & Parrish, J.K. (1993) Functions of shoaling behaviour in teleosts. In: T.J. Pitcher & J.K. Parrish (eds) Behaviour of Teleost Fishes, pp. 363–439. Chapman & Hall, London. Pollock, M.S. & Chivers, D.P. (2004) The effects of density on the learned recognition of heterospecific alarm cues. Ethology, 110, 341–349. Reebs, S.G. (1999) Time-place learning based on food but not on predation risk in a fish, the inanga (Galaxias maculatus). Ethology, 105, 361–371. Reebs, S.G. (2002) Plasticity of diel and circadian activity rhythms in fishes. Reviews in Fish Biology and Fisheries, 12, 349–371. Reid, A.L., Seebacher, F. & Ward, A.J.W. (2010) Learning to hunt: the role of experience in predator success. Behaviour, 147, 223–233. Roper, T.J. & Redston, S. (1987) Conspicuousness of distasteful prey affects the strength and durability of one-trial avoidance learning. Animal Behaviour, 35, 739–747. Roper, T.J. & Wistow, R. (1986) Aposematic coloration and avoidance learning in chicks. Quarterly Journal of Experimental Psychology, 38B, 141–149. Ruxton, G.D., Sherratt, T. & Speed, M.P. (2004) The Evolutionary Ecology of Crypsis, Warning Signals and Mimicry. Oxford University Press, Oxford. Shettleworth, S.J. (1988) Cognition, Evolution and Behavior. Oxford University Press, New York. Sih, A. (1980) Optimal behaviour: can foragers balance two conflicting demands? Science, 210, 1041–1043. Sih, A. (1988) The effects of predators on habitat use, activity and mating behaviour in a semi-aquatic bug. Animal Behaviour, 36, 1846–1848. Sih, A., Englund, G. & Wooster, D. (1998) Emergent impacts of multiple predators on prey. Trends in Ecology and Evolution, 13, 350–355. Sih, A., Ziemba, R. & Harding, K.C. (2000) New insights on how temporal variation in predation risk shapes prey behaviour. Trends in Ecology and Evolution, 15, 3–4. Smith, R.J.F. (1992) Alarm signals in fishes. Reviews in Fish Biology and Fisheries, 2, 33–63. Smith, R.J.F. (1997) Avoiding and deterring predators. In: R.J.F. Smith (ed) Behavioural Ecology of Teleost Fishes, pp. 163–190. Oxford University Press, Oxford. Smith, R.J.F. & Smith, M.J. (1989) Predator-recognition behaviour in two species of Gobiid fishes, Asterropteryx semipunctatus and Gnatholepis anjerensis. Ethology, 83, 19–30. Speed, M.P. (2001) Can receiver psychology explain the evolution of aposematism? Animal Behaviour, 61, 205–216. Suboski, M.D. (1990) Releaser-induced recognition learning. Psychological Reviews, 97, 271– 284. Suboski, M.D., Bain, S., Carty, A.E., McQuoid, L.M., Seelen, M.I. & Seifert, M. (1990) Alarm reaction in acquisition and social transmission of simulated-predator recognition by zebra danio fish (Brachydanio rerio). Journal of Comparative Psychology, 104, 101–112. Tinbergen, L. (1960) The natural control of insects on pinewoods. I. Factors influencing the intensity of predation by songbirds. Archives Neerlandaises de Zoologie, 13, 265–343. Tinker, M.T., Mangel, M. & Estes, J.A. (2009) Learning to be different: acquired skills, social learning, frequency dependence, and environmental variation can cause behaviourally mediated foraging specializations. Evolutionary Ecology Research, 11, 841–869. Tulley, J.J. & Huntingford, F.A. (1987) Parental care and the development of adaptive radiation in antipredator responses in sticklebacks. Animal Behaviour, 35, 1570–1572. Utne-Palm, A.C. (2001) Response of na¨ıve two-spotted gobies Gobiusculus flavescens to visual and chemical stimuli of their natural predator, cod Gadus morhua. Marine Ecology Progress Series, 218, 267–274.

BLBK374-03

BLBK374-Brown

58

May 27, 2011

10:35

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Vilhunen, S. & Hirvonen, H. (2003) Innate antipredator responses of Arctic charr depend on predator species and their diet. Behavioral Ecology and Sociobiology, 55, 1–10. Vogel, D. & Bleckmann, H. (2001) Behavioral discrimination of water motions caused by moving objects. Journal of Comparative Physiology, 186, 1107–1117. Warner, R.R. (1988) Traditionality of mating site preferences in a coral reef fish. Nature, 335, 719–721. Wirsing, A.J., Cameron, K.E. & Heithaus, M.R. (2010) Spatial responses to predators vary with prey escape mode. Animal Behaviour, 79, 531–537. Wisenden, B.D. & Harter, K.R. (2001) Motion, not shape, facilitates association of predation risk with novel objects by fathead minnows (Pimephales promelas). Ethology, 107, 357–364. Wisenden, B.D., Pogatshnik, J., Gibson, D., Bonacci, L., Schumacher, A. & Willett, A. (2008) Sound the alarm: learned association of predation risk with novel auditory stimuli by fathead minnows (Pimephales promelas) and glowlight tetras (Hemigrammus erythrozonus) after single simultaneous pairings with conspecific chemical alarm cues. Environmental Biology of Fishes, 81, 141–147. Yunker, W.K., Wein, D.E. & Wisenden, B.D. (1999) Conditioned alarm behavior in fathead minnows (Pimephales promelas) resulting from association of chemical alarm pheromone with a nonbiological visual stimulus. Journal of Chemical Ecology, 25, 2677–2686.

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 4

Learning about Danger: Chemical Alarm Cues and Threat-Sensitive Assessment of Predation Risk by Fishes Grant E. Brown, Maud C.O. Ferrari and Douglas P. Chivers

4.1

Introduction

Most species are at risk of predation during some, if not all, phases of their lives. As a result, there exists strong selection pressure for early detection and avoidance of potential predation threats. However, predator avoidance has the potential to be very costly, as it reduces time and energy available for other activities such as foraging, mating and territorial defence (Godin & Smith 1988; Sih 1992) or forces prey to utilise suboptimal habitats (Gotceitas & Brown 1993), leading to a reduction in energy intake and reproductive output (Lima & Dill 1990). Presumably, prey that can adjust the intensity of their predator avoidance response according to the level of perceived risk should be at a selective advantage (Helfman 1989; Lima & Dill 1990). This supposition is known as threat-sensitive predator avoidance (Helfman 1989; Chivers et al. 2001) Threat-sensitive assessment of predation risk is complicated by the fact that predation itself is highly variable in space and time (Sih et al. 2000; Griffin 2004; Lima & Steury 2005; Ferrari et al. 2009a). Likewise, the form of predation and the degree of risk may change dramatically as prey individuals grow (size-dependent predation risk; Br¨onmark & Miner 1992), shift habitat preferences with ontogeny (Werner & Gilliam 1984) or move within heterogeneous habitats (Golub et al. 2005). Prey may also move between prey guilds (Olson et al. 1995; Olson 1996; Brown et al. 2001) and are subject to seasonal changes in biotic and/or abiotic conditions (Gilliam & Fraser 2001). This high degree of variability can result in unpredictable and variable predation risk. Prey that can reliably assess local predation risk based on recent experience (i.e. learning) should be better able to deal with variability in predation pressure. Our recent reviews (Brown & Chivers 2005, 2006) have highlighted the mechanisms of chemically mediated learning in prey fishes. Our goal here is to review some of the new directions in the field of chemically mediated predator recognition. Specifically, we examine the role-learning plays in the threat-sensitive mediation of predation risk.

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

60

Fish Cognition and Behavior

4.2

Chemosensory cues as sources of information

Char Count=

An underlying assumption of any approach to threat-sensitive behavioural decision models is that prey can reliably assess local predation threats in real time. Within aquatic ecosystems, this risk assessment is often mediated via chemosensory cues (Kats & Dill 1998; Wisenden & Chivers 2006). Public information regarding local predation risk is available from a suite of cues, including damage-released chemical alarm cues, disturbance cues and predator odours (Chivers & Smith 1998). Damage-released cues are found in a wide variety of taxonomically diverse prey fishes and are produced and/or stored in the epidermis. When released, following mechanical damage to the skin, these cues can elicit dramatic short-term increases in species typical antipredator behaviours (Chivers & Smith 1998; Wisenden & Chivers 2006). Recent studies suggest that there is often a strong relationship between the relative concentration of alarm cues prey detect and the intensity of the antipredator responses displayed by the prey (Jachner & Rydz 2002; Dupuch et al. 2004; Brown et al. 2006a, 2009). Moreover, prey may attend to alarm cues at very low concentrations by increasing vigilance towards secondary risk assessment cues (Brown et al. 2004) or may show subtle adjustments in their foraging tactics (Foam et al. 2005a). A second class of chemosensory cues are the so called ‘disturbance cues’. These are metabolic by-products released by stressed or disturbed prey prior to an attack by a predator (Wisenden et al. 1995; Jord˜ao & Volpato 2000; Vavrek et al. 2008). They are released in the absence of skin damage to the prey. While behavioural response to disturbance is typically of a lower intensity than to the more often studied damage-released cues, there still exists a strong threat-sensitive response to varying concentrations of disturbance cues (Vavrek & Brown 2009). Finally, the chemosensory cues originating from predators themselves can provide information regarding the intensity of local threats (Kusch et al. 2004; Ferrari et al. 2006a). In some cases, the degree of sophistication of chemosensory risk assessment is remarkable. Fathead minnows (Pimephales promelas, Cyprinidae), for example, are known to distinguish predator size, predator proximity and predator density – all based on predator odours (Kusch et al. 2004; Ferrari et al. 2006b).

4.2.1

Learning, innate responses and neophobia

Researchers studying predator recognition have spent considerable time attempting to identify the relative importance of experience versus genetic factors in the acquisition of responses to predators. We have some good examples of prey fishes that do not appear to respond to predators unless they have experience, cases where prey seem to respond to predators without experience and examples where experience modifies what appears to be ‘innate’ responses. Chivers & Smith (1994a,1994b) demonstrated the importance of experience in responses of fathead minnows to pike (Esox lucius, Esocidae) cues. They found that fathead minnow eggs collected from pike sympatric populations and reared under laboratory conditions exhibited no apparent recognition of either the chemical or visual

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

Char Count=

61

cues of a predatory pike, while wild caught individuals (from the same population) of the same size/age did (Chivers & Smith 1994b). Subsequent stocking experiments demonstrated that naive populations of minnows would quickly acquire recognition of introduced predators (Chivers & Smith 1995). Perhaps the most striking example of this rapid recognition was demonstrated by Brown et al. (1997). They introduced 39 juvenile pike into a 4-ha lake containing approximately 78,000 minnows and found recognition of the pike odour by minnows within 4 days. Several other species, including brook stickleback (Culaea inconstans, Gasterosteidae; Chivers et al. 1995), brook trout (Salvelinus namaycush, Salmonidae; Mirza & Chivers 2000, 2001), brown trout (Salmo trutta, Salmonidae; Alvarez & Nicieza 2003), common bully (Gobiomorphus cotidianus, Eleotridae; Kristensen & Closs 2004) and zebra danio (Danio rerio, Cyprinidae; Bass & Gerlai 2008), fail to exhibit responses to predators without prior experience. Conversely, Scheurer et al. (2007) provided strong support that fishes can respond to predators even though they have no prior experience with them. They tested F2 hatcheryreared offspring of steelhead trout originally collected from a stream population that contained Dolly Varden (S. malma, Salmonidae, a common predator of juvenile steelhead) and a lake population that was devoid of Dolly Varden. The lake population was originally stocked with steelhead collected from the stream population at least 15 generations previously. Despite at least 15 generations of isolation, the lake population showed responses to the odour of Dolly Varden predators, as did the stream population, when tested under common laboratory conditions. These results might indicate genetically fixed, i.e. innate, predator recognition. There are several other examples of what may appear to be ‘innate’ predator recognition in a variety of prey fishes. Juvenile Chinook salmon (Berejikian et al. 2003) and arctic charr (S. alpinus, Salmonidae; Vilhunen & Hirvonen 2003) exhibit increased antipredator responses when exposed to novel predator odours. Likewise, Hawkins et al. (2004) have shown that juvenile Atlantic salmon (S. salar, Salmonidae) significantly increased opercular flap rates upon detection of novel cues, suggesting increased vigilance or olfactory sampling as opposed to true recognition (Gibson & Mathis 2006). Predator naive Nile tilapia (Oreochromis niloticus, Cichlidae) show a similar increase in opercular movements during exposure to visual predator cues (Barreto et al. 2003). Can such apparent innate recognition occur in the absence of genetic fixation? The answer is likely yes. Prey may exhibit strong avoidance responses to any novel cue, a phenomenon known as neophobia (Sneddon et al. 2003). Responding with a fright response to any novel cue is much different than having genetically fixed responses to specific predator cues. In either case, being able to respond to predators upon a first encounter should eliminate the cost of learning (Blumstein 2006; Ferrari et al. 2007). As prey grow or develop sufficient behavioural plasticity, learning should replace neophobic responses, allowing prey to ‘fine-tune’ their recognition and avoidance of predators. This is likely very important in populations characterised by variable predation pressure, where true innate recognition would prove limiting. We will return to the importance of learning versus innate responses later when we discuss the predator recognition continuum hypothesis in Section 4.5.

BLBK374-04

BLBK374-Brown

62

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

4.2.2 Learned predator recognition through conditioning with alarm cues It is well established that many fishes can learn to recognise unknown predators through conditioning with alarm cues. In this well-studied mechanism, the pairing of a damagereleased alarm cue with either the visual or chemical cues of a novel predator results in learned recognition of that predator (Chivers & Smith 1998; Brown & Chivers 2005, 2006). For example, Magurran (1989) showed that European minnows (Phoxinus phoxinus, Cyprinidae) acquire the recognition of the chemical cues of predatory pike after a single exposure to pike odour paired with conspecific alarm cues. Chivers & Smith (1994a,1994b) likewise reported that fathead minnows could learn the sight or the odour of a pike in the same way. Similar conditioning results have been shown for brook stickleback (Chivers et al. 1995), brown trout (Alvarez & Nicieza 2003), common bully (Kristensen & Closs 2004) and zebra danios (Bass & Gerlai 2008). Learned recognition through conditioning with alarm cues enhances survival during staged encounters with live predators (Mirza & Chivers 2000, 2001; Darwish et al. 2005). There is growing evidence that chemically mediated learning allows prey to fine-tune their predator recognition. For example, Berejikian et al. (2003) found that the strength of an ‘innate’ response to novel predators could be enhanced through conditioning with alarm cues. Recent studies have also shown a significant ontogenetic effect on both innate predator avoidance and chemosensory learning. Juvenile Atlantic salmon exhibited significant innate responses to predator odours when tested 10–15 weeks post-hatching; however, younger (<3 weeks post-hatching) or older (28–36 weeks post-hatching) salmon did not (Hawkins et al. 2008). The same study showed that young salmon did not acquire the recognition of a novel predator odour paired with conspecific alarm cues, while older salmon did show significant learned responses (Hawkins et al. 2008). These data strongly suggest a complementary relationship between ‘innate’ and ‘learned’ responses.

4.3

Variable predation risk and flexible learning

It is now abundantly clear that prey fishes can show dramatic adjustments in the intensity of their antipredator response according to the level of perceived risk and that they are capable of learning to recognise novel predator cues. A growing body of research has built upon these two findings, examining the role of chemically mediated learning in the development of threat-sensitive response patterns. Ferrari et al. (2005) asked if the strength of the learned response to a novel predator odour is related to the intensity of the initial conditioning experience. They exposed predator-naive fathead minnows to low, intermediate or high relative concentrations of conspecific alarm cues, paired with the odour of a novel predator (brook charr, S. fontinalis, Salmonidae). During the initial conditioning phase, minnows exhibited stronger antipredator responses to the high concentration cues than the low concentration cues. When retested 24 hours later, Ferrari et al. (2005) found that the learned response to charr odour matched the intensity of response during the initial conditioning phase. Likewise, they found a similar correlation between the intensity of response to charr odour between experienced tutors

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

Char Count=

63

and naive observers in a social learning study (Ferrari et al. 2005). Thus, acquired predator recognition through a single conditioning with alarm cues or a single observational learning opportunity appears to allow for the acquisition of threat-sensitive response patterns. Zhao et al. (2006) extended these findings, demonstrating that goldfish (Carassius auratus, Cyprinidae) were able to make similar threat-sensitive associations based on predator diet cues. They fed northern pike diets of either goldfish (with a recognisable alarm cue) or swordtails (Xiphophorus helleri, Poceiliidae; lacking an alarm cue recognised by minnows) and then exposed goldfish to varying concentrations of the predator odour. Goldfish exhibited increased antipredator responses proportional to the concentration of predator odour presented. When tested to a standard predator odour (pike-fed swordtails) during recognition trials, goldfish exhibited predator avoidance responses matching the intensity the fish exhibited during the initial conditioning trials. Moreover, when exposed to a live predator during staged encounters, goldfish initially conditioned to the highest concentration of predator odour (i.e. strongest learned recognition) were more likely to survive than those exposed to lower concentrations (i.e. weaker learned recognition) or the non-conditioned controls (i.e. no recognition of pike). Together, these data suggest that there is indeed a functional link between learning and the acquisition of context appropriate threat-sensitive response patterns. However, under ecologically realistic conditions, we should expect prey to be exposed to multiple learning opportunities. This raises the issue of ‘conflicting’ information, where prey may experience relatively high- and low-risk situations within short time frames. How might this influence threat-sensitive learning? Ferrari & Chivers (2006a) tested whether recent experience shapes threat-sensitive learning. Prey fishes may continually update their learned recognition of potential predators, with the intensity of antipredator behaviour mimicking the most recent learning experience. Ferrari & Chivers (2006a) tested this by exposing fathead minnows to either a high or a low concentration of conspecific alarm cue plus the odour of brook charr. The paired cues were given daily, for 6 consecutive days, one of four combinations: 6 low, 5 low 1 high, 1 low 5 high or 6 high concentrations of conspecific alarm cue plus the odour of brook charr. The intensity of antipredator behaviour when exposed to charr odour alone appeared to match the most recent ‘conditioning’ experience. This suggests that prey do not simply ‘average’ the intensity of learning opportunities, but do indeed adjust their level of antipredator response to the most recent experience. The concentration of predator odour detected during conditioning events also provides valuable information about the threat of the predator. Ferrari et al. (2006c) conditioned fathead minnows with conspecific alarm cue paired with either high or low concentrations of northern pike odour. Regardless of the concentration, there was no difference in response intensity when minnows were tested for recognition to the same concentration of predator odour used during the conditioning event (i.e. low–low = high–high). Perhaps this is not surprising, as the initial intensity of the antipredator response is governed by the concentration of alarm cue detected (which was the same for this study). However, when conditioned with a low concentration predator odour and then tested against a high concentration cue, minnows increased the intensity of the predator avoidance response. Conversely, when conditioned to a high concentration and then tested against a low concentration, minnows reduced the intensity of their response. Taken together with previous studies, these data

BLBK374-04

BLBK374-Brown

64

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

highlight the sophistication of the chemically mediated learning system, allowing prey to associate context-dependent risk with specific learned cues.

4.3.1

Assessing risk in time

Predation risk is known to vary from moment to moment and over daily, seasonal, yearly and generational scales (Sih et al. 2000; Magurran 2005). Until recently there was a surprising lack of information on how prey fishes dealt with such variation in risk. This stands in sharp contrast to numerous studies showing that prey animals, including fishes, can learn to associate specific foraging location with times of day or specific seasons (Foote & Brown 1998; Reebs 1999). Reebs (1999) demonstrated that inangas (Galaxias maculatus, Galaxiidae) were able to learn to associate a specific location with the temporal availability of food. However, inangas were unable to make the same association with a predictable predation risk (Reebs 1999). Yl¨onen et al. (2006) found partial support for temporally dependent learning in juvenile yellow perch (Perca fluviatilis, Percidae) and ruffe (Gymnocephalus cernuus, Percidae). Both perch and ruffe were able to acquire the recognition of the chemical cues of pike and burbot (Lota lota, Lotidae). However, while neither prey species showed clear evidence of diel patterns in avoidance, juvenile perch did respond with a stronger response to burbot at night. Yl¨onen et al. (2006) suggest that this may be due to the primarily nocturnal foraging habits of burbot. However, recent studies with larval amphibians demonstrate a clear ability to associate acquired predator recognition cues with time of day. Ferrari et al. (2008a) exposed larval wood frogs (Rana sylvatica) to predatory tiger salamanders (Ambystoma tigrinum) in the presence of wood frog alarm cues in the evening and tiger salamanders alone in the morning (vs. salamander only morning and evening controls) over a 9-day period. Wood frog tadpoles conditioned to recognise the salamander cue exhibited a significant recognition of the predator cue when tested in the morning, but showed a significantly stronger response when tested in the evening. These results suggest that wood frogs not only are able to learn to recognise the chemical cue of a common predator but can learn to associate the level of risk with time of day. The ability to learn threat-sensitive recognition based on predictable temporal patterns presumably would provide considerable benefits to prey fishes. Clearly, this hypothesis requires additional examination. Recent studies have examined the way in which prey fishes adjust the intensity of their antipredator responses within the context of the risk allocation model proposed by Lima & Bednekoff (1999). This model predicts that as predation risk fluctuates over time, the intensity of prey vigilance and foraging should depend on both the level of risk and the proportion of time that predators are present. If predators are usually absent, prey can meet their energy demands during safe periods, and thus respond strongly during the rare times when predators are present. In contrast, if predators are omnipresent, prey might need to forage actively even though predators are present. Studies with rainbow trout (Mirza et al. 2006), convict cichlids (Foam et al. 2005b; Brown et al. 2006b; Ferrari et al. 2010a), guppies (Poecilia reticulata, Poeciliidae; Brown et al. 2009b) and several flatfishes (Boersma et al. 2008) provide at least partial support for the risk allocation model.

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

4.3.2

Char Count=

65

Sensory complementation and threat-sensitive learning

Under natural conditions, we might expect that prey would be exposed to multiple sources of information simultaneously. How these multiple information sources interact to allow for threat-sensitive behavioural decisions has only recently been examined. The sensory complementation hypothesis suggests that prey capable of accessing multiple risk assessment cues should be better able to gain reliable information regarding local threats (Smith & Belk 2001; Brown & Magnavacca 2003; Blanchet et al. 2007) and that multiple cues should interact in an additive or synergistic fashion (Lima & Steury 2005). There is a growing body of evidence that supports this hypothesis. For example, Blanchet et al. (2007) found that under laboratory conditions, stream dwelling young-of-the-year (YOY) Atlantic salmon responded in an additive fashion when presented with predator odour (adult rainbow trout-fed YOY Atlantic salmon) combined with a visual predator model in comparison to when the fishes were exposed to either visual or chemical cues alone. They found a similar response patterns when quantifying different response variables. Smith & Belk (2001) found similar chemical plus visual interactions in their study of the threat-sensitive predator inspection behaviour of western mosquitofish (Gambusia affinis, Poeciliidae). Recent studies have shown that sensory complementary effects may differ depending on age and/or experience. Kim et al. (2009) have shown an interaction between the additive value of combined chemical and visual cues with age class in wild Atlantic salmon. YOY and parr salmon were exposed to conspecific alarm cues (or a stream water control) followed by the presentation of a visual threat. Both YOY and parr exhibited significant increases in their latency to resume foraging following exposure to the chemical cue (relative to the control), but only YOY salmon exhibited a significant increase in latency to resume foraging following the subsequent exposure to the visual cue. Parr initially exposed to the control or conspecific alarm cue showed similar latencies to resume foraging after exposure to the visual cue. However, parr exhibited an additive response to chemical plus visual cues when looking at the flight initiation distance. Parr initially exposed to the chemical alarm cue exhibited a significantly greater reactive distance to the visual cue. YOY salmon showed no difference in the reactive distance. Thus, additive effects may be present, but differ depending on experience or age (Kim et al. 2009). Chris Elvidge (unpublished data) suggests that sensory complementation of information might depend on learning or experience. Elvidge presented wild juvenile Atlantic salmon with realistic model predators (visual cues) in a series of streams varying in ambient pH. Previous studies have shown that weakly acidic conditions inhibit the detection and response to conspecific alarm cues by juvenile salmon (Leduc et al. 2006). Presumably, salmon in neutral streams have had previous opportunity to assess risk based on the combined visual plus chemical cues, whereas salmon in the acidic streams would not have prior experience with chemical information. Elvidge’s results suggest that under weakly acidic conditions, juvenile salmon treat a standardised visual threat as a higher risk than under neutral conditions. The sensory complement model would argue that if prey have had previous experience with combined cues (as in the neutral streams), detection of a potential threat via a single sensory modality would be perceived as a lower risk than the same cue accompanied by additional sensory inputs. Likewise, Jachner (2001) demonstrated an interesting interaction between experience and the response of juvenile roach

BLBK374-04

BLBK374-Brown

66

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(Rutilus rutilus, Cyprinidae) to conspecific alarm cues. Hatchery-reared roach exhibited threat-sensitive responses to conspecific alarm cues. Moreover, when repeatedly exposed to alarm cues, roach increased time on a foraging patch, suggesting a reduced antipredator response. However, wild caught, presumably experienced, roach exhibited only a weak response to alarm cues. Jachner (2001) argued that upon detection of the alarm cues experienced roach would increase vigilance towards either the smell or sight of a predator. In the absence of these secondary cues, roach did not show a strong antipredator response. Together, these studies provide indirect evidence of the role of sensory complementation in predator recognition learning. To date, we know of only a single study examining the direct effects of additive sensory inputs on chemosensory learning in fishes. Ferrari et al. (2008c) tested juvenile rainbow trout for additive responses to two different chemosensory cues. Trout exposed conspecific disturbance and chemical alarm cues combined responded in an additive fashion, supporting the sensory complement hypothesis. They then tested for additive learning effects by conditioning trout to recognise a novel predator odour paired with disturbance cue, alarm cue or the combined disturbance plus alarm cue. While they report no learning in trout initially exposed to the disturbance cue plus predator odour treatment, they did find stronger learning among trout initially conditioned with disturbance plus alarm cues. Thus, their results suggest that complementary interactions among multiple sensory modalities result in enhanced learning opportunities.

4.4

Generalisation of risk

Though learning has the obvious benefit of allowing prey to respond to temporal and spatial variations in risk, it is not without costs. In order to learn to recognise a novel predator, prey must have an initial encounter with the predator, representing at least a momentary increase in risk. Presumably, if prey can generalise what they actually learn, this would reduce the ‘direct learning’ costs associated with learning specific predators. For both visual and chemical predator cues, there should exist considerable variability among individual conspecific predators. Yet prey conditioned to recognise an individual predator are able to generalise other predators of the same species. But, can prey actually generalise learned information across different predator species that are similar in appearance or smell to the ones the prey already recognises?

4.4.1

Generalising of predator cues

A series of elegant studies with mammals (Griffin et al. 2001; Stankowich & Coss 2007), amphibians (Ferrari et al. 2009b) and fishes (Ferrari et al. 2007, 2010b; Brown et al. 2011a) have shown that prey conditioned to a reference predator can exhibit learned responses to chemical and visual cues of predators with which the prey have no experience. Moreover, predators that are more distantly related to the reference predator are not recognised. For example, Ferrari et al. (2007) conditioned predator-naive fathead minnows to recognise lake trout (S. namaycush, Salmonidae) and then tested for the recognition of lake trout, brook charr, rainbow trout (Oncorhynchus mykiss, Salmonidae), pike or white sucker (Catostomus

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

Char Count=

67

commersoni, Catostomidae). Minnows exhibited a learned response to the three salmonid odours, and no significant response towards the pike or sucker odours. More interesting is that the intensity of the learned response was strongest for the reference predator and weakest (though still significant) towards the confamilial predator (rainbow trout), with the response to the congeneric predator (brook charr) being intermediate. By generalising learned predator recognition, prey should be able to gain the benefits of learning specific predator cues, without having to directly assess multiple predators independently (Griffin et al. 2001; Darwish et al. 2005; Ferrari et al. 2007). What specifically is being generalised? The results of Ferrari et al. (2007) suggest that generalisation is based on chemical similarities among phylogenetically related predators, but was independent of the predator’s diet. Presumably, related predator species would produce chemosensory cues more similar to each other than would more distantly related predators. We might also expect predators sharing similar diets, regardless of phylogenetic relatedness, to produce cues that are readily generalised by prey. Dietary cues are known to allow for the recognition of novel predators (Mathis & Smith 1993a, 1993b; Chivers & Mirza 2001). It is likely that a combination of chemical similarities among related predators plus common diets (especially among sympatric predators) would provide sufficient information to allow for generalisation of learned predator recognition. Ferrari et al. (2010c) show that generalisation can also be based on visual cues. Fathead minnows conditioned to recognise the visual cues of brown trout exhibited a similar response to the visual cues of rainbow trout, but not yellow perch. They argue that this is due to learning of similar body shape or some other visual feature. Intuitively, this makes sense. Magurran (1989) found that European minnows (P. phoxinus, Cyprinidae) were less likely to learn to recognise the visual cues of the non-predatory tilapia, compared to those of the predatory pike. Chivers & Smith (1994b) showed similar effects with fathead minnows. These results suggest that some specific visual cues may be more easily generalised in the context of predator recognition (Smith 1997). Studies looking at chemosensory cues typically restrict predators to a common diet in order to minimise possible dietary effects. An intriguing study would be to vary diet and predator taxa to directly assess this. If, as argued by Ferrari et al. (2007), prey benefit from generalised predator recognition via a reduction in the costs associated with learning specific predators, then we would expect higher risk cues to be more readily generalised than lower risk cues. Indeed, Ferrari et al. (2008c) found that fathead minnows conditioned with a high concentration of alarm cue (high perceived risk) exhibited generalised learning of trout cues, while minnows conditioned with low concentrations of alarm cue (low perceived risk) did not. These results suggest that prey may generalise the recognition only to highly threatening species but not to those of lower perceived risks.

4.4.2

Generalisation of non-predator cues

Pre-exposure to a predator cue, in the absence of any conditioning stimulus (latent inhibition), can result in the inability of prey to acquire recognition of a novel predator (Acquistapace et al. 2003; Ferrari & Chivers 2006b, 2009). For example, Ferrari & Chivers (2006b) exposed fathead minnows to brook charr odour or distilled water once per day for 5 consecutive days and then conditioned them to recognise brook charr as a predator.

BLBK374-Brown

68

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

80 A. Generalised learning of predator odours

B. Generalised learning of non-predator odours

60 Change in time moving (seconds)

BLBK374-04

Pre-exposed to pumpkinseed odour Pre-exposed to distilled water

40 20 0 –20 –40 –60 –80

PS

LE

RB

YP

DW

PS

LE

RB

YP

Fig. 4.1 Mean (SE) change in time spent moving for juvenile rainbow trout. Panel A: Juvenile rainbow trout were initially conditioned with trout alarm cue paired with the odour of a novel predator (pumpkinseed) and subsequently tested for the learned avoidance of pumpkinseed (PS), longear sunfish (LE), rock bass (RB), yellow perch (YP) or a distilled water (DW) control. Trout were able to generalise the learned recognition of pumpkinseed to the congeneric longear sunfish and the confamilial rock bass, but not to yellow perch (no different from the distilled water control). Panel B: Juvenile rainbow trout were pre-exposed to distilled water twice per day for 3 days and conditioned with alarm cue paired with the odour of one of four predators. When tested for recognition, trout were equally able to acquire the recognition of the predator odours (open circles). However, when pre-exposed to pumkinseed odour twice per day for 3 days (solid circles) and then conditioned to recognise the predator odours, trout were inhibited from learning pumpkinseed odour. In addition, trout appeared to generalise this inhibition and were unable to learn to recognise the odour of the congeneric longear sunfish. There was no inhibition of learning of rock bass or yellow perch odours. Modified from Brown et al. (2011a).

Minnows pre-exposed to distilled water were able to learn to recognise charr as a predator, but those pre-exposed to charr odour were not. Brown et al. (2011a) used this latent inhibition mechanism to test the hypothesis that juvenile rainbow trout can generalise the recognition of both predator and non-predator cues. In their first experiment, juvenile trout were conditioned to recognise pumpkinseed odour as a reference predator and then tested for the recognition of pumpkinseed, longear sunfish, rock bass or yellow perch odours. Trout conditioned to recognise pumpkinseed exhibited strong learned antipredator responses to the pumpkinseed and the odour of the congeneric longear sunfish. They observed a weaker (but still significant) response to the confamiliar rock bass and no response to the more distantly related yellow perch (Fig. 4.1A). In the second experiment, trout that had been pre-exposed to pumpkinseed odour did not learn to recognise pumpkinseed as a predation threat. Moreover, pre-exposure to pumpkinseed odour also inhibited learning of the odour of longear sunfish cues (Fig. 4.1B).

4.5

Predator recognition continuum hypothesis

Ferrari et al. (2007) argued that learning to recognise novel predators is merely one point along a continuum of predator recognition. At one extreme, prey would be genetically fixed to exhibit true innate recognition of predators, independent of any experience. At the

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

Char Count=

69

other extreme, recognition would be learned, but any acquired information could be fully generalised to appropriate stimuli. Ferrari et al. (2007, 2008c) argue that such a continuum would arise primarily through variability in the spatial and/or temporal predictability of predation threats. This model consists of two related issues: (1) selection to learn and (2) selection to generalise.

4.5.1 Ecological selection for innate versus learned recognition of predators Given the demonstrated range of responses from innate to learned in how prey fishes recognise potential predators (see Subsection 4.2.1), we can make some predictions regarding the ecological conditions favouring either extreme form of recognition (Brown & Chivers 2005; Ferrari et al. 2007). Though poorly understood, the key factor is likely variability in predation pressure, in both overall intensity and predator type. Under conditions of intense predation pressure or conditions of a small and stable predator guild, selection should favour an experience independent response to novel predators (Breden et al. 1987; Vilhunen & Hirvonen 2003; Riechert 2005). Under such conditions, exposure to novel situations, including a novel predator, may elicit a strong predator avoidance response simply because the costs of not doing so would be disproportionately high (Hirvonen et al. 2000). Such ‘experience independent’ responses could be in the form of true innate responses (i.e. genetically fixed) or neophobia. Neophobic responses (see Subsection 4.2.1) would be particularly beneficial for species lacking sufficient behavioural plasticity in early life history phases to allow for threat-sensitive responses (Killen & Brown 2006). Moreover, neophobic responses would not require genetic fixation to a specific predator type, and thus might represent a ‘midpoint’ between innate and learned responses. Conversely, populations exposed to variable predation risk may be selected to learn context-appropriate responses due to the relatively high costs associated with ‘false responses’ to non-risky stimuli. Such variability in predation risk may arise from a number of ecological variables including spatial and temporal variabilities in predator encounters (Sih et al. 2000; Lima & Steury 2005), ontogenetic shifts in microhabitat use (Olson et al. 1995; Olson 1996) or size-dependent risk (Br¨onmark & Miner 1992; Nilsson & Br¨onmark, 2000). While clearly an oversimplification, this theoretical starting point could potentially lead to testable predictions.

4.5.2

Ecological selection for generalised learning

Ferrari et al. (2007) argued that selection towards generalised learning may be linked to the ratio of predators to non-predators. Under conditions where heterospecifics are predominately potential predation threats (high predator to non-predator ratio), prey may benefit from generalising the visual and/or chemical cues of learned predators to all heterospecifics with similar traits. This would allow prey to reduce the costs associated with direct interactions (learning) while still responding to variable predation risks. However, under conditions where fewer heterospecifics represent actual predation threats (low predator to non-predator ratio), generalising to all heterospecifics might represent a reduction in time and energy available for other fitness-related activities. As such, depending on the predator and prey

BLBK374-04

BLBK374-Brown

70

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

guild structure, the use of generalised or non-generalised predator recognition may allow prey to optimise threat-sensitive trade-offs (Ferrari et al. 2007, 2008c).

4.6

Retention: the forgotten component of learning

Despite the large volume of literature focusing on the acquisition of predator recognition in fishes and other taxa, surprisingly little is known regarding the retention of acquired information. Following a single pairing of an alarm cue and a novel predator odour, hatcheryreared rainbow trout retain a learned response for up to 21 days (Brown & Smith 1998) though the response begins to decline rapidly after 10 days (Mirza & Chivers 2000). Fathead minnows conditioned to recognise visual cues of either northern pike or goldfish as potential predators showed similar intensities of learned predator recognition when tested 2 days post-conditioning. However, when re-tested for visual recognition nearly 2 months post-conditioning, those initially conditioned to pike show a stronger response than did those initially conditioned to goldfish (Chivers & Smith, 1994a), suggesting differential retention. Presumably, prey should only respond to learned predator cues as long as they represent an actual threat (Gonzalo et al. 2009). So far, we have argued that learning to recognise novel predators allows prey to balance the conflicting pressures associated with successful predator avoidance and a suite of other behavioural activities. To date, the majority of studies have focused on the presence of learning abilities and on the mechanisms and selection that shape learning abilities. A poorly understood aspect of the learning equation is ‘how long should prey retain’ acquired information. Several models have addressed the issue of retention of learned responses within the context of foraging decisions (McNamara & Houston 1987; Mangel 1990; Hirvonen et al. 1999). These models generally predict that there should exist a ‘memory retrieval’ window that allows for a flexible response pattern. Under relatively constant environmental conditions, information regarding foraging decisions should be retained longer (i.e. remain within the memory window), while under highly variable environmental conditions, older learned foraging information would be of lower value, hence fall outside this window (i.e. be ‘forgotten’). Such models predict that learned information should only be retained as long as it is relevant (Pravosudov & Clayton 2002; Brydges et al. 2008) and that acquired information that is no longer relevant is forgotten (i.e. no longer capable of eliciting an adaptive behavioural response; Mackney & Hughes 1995). Using this framework, Ferrari et al. (2010b) have proposed an analogous model dealing with learned predator recognition. They argue that the question of how long prey should retain information is equally important in the context of adaptive value as the ability to acquire recognition of novel predators. The retention of information or ‘memory window’ should only exist as long as the information is relevant. Responding to outdated or suboptimal information is costly, as it takes away from time and energy available for other activities such as foraging. Ferrari et al. (2010b) proposed that a suite of extrinsic and intrinsic factors should interact to shape the ‘memory window’ of prey fishes. Extrinsic factors such as high turnover rate of predator communities and high frequency of diet shifts of predators should have the effect of reducing the length of the plastic or flexible ‘memory window’ as individual prey would likely only be at risk of particular

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

71

predators for a limited time frame. Conversely, increased predator encounter rates, predictability of predators and/or probability of predator attacks are extrinsic factors, which would be predicted to extend the memory window. Likewise, intrinsic factors such as increased growth rates, impending life history shifts or morphological antipredator defences should result in a shortened memory window. Slower growth rates, simple life histories or a lack of morphological defences should increase the value of learned information, and hence extend the length of the memory window. Brown et al. (2011b) directly tested this model by manipulating the growth rate of juvenile rainbow trout and testing for the retention of the learned recognition of pumpkinseed odour. Trout were fed either high or low food rations for a period of 7 days (sufficient to induce different growth rates) and then conditioned to recognise the odour of pumpkinseed as a threat. Trout tested for the recognition of pumpkinseed odour 24 hours post-conditioning exhibited similar learned responses, regardless of growth rate (Fig. 4.2A). However, when tested for a learned response 8 days post-conditioning, only those on the low food ration exhibited an antipredator response. Trout fed on the high food ration were no different from distilled water controls (Fig. 4.2A). As a companion study, Brown et al. (in press b) conditioned trout of two size classes (similar to the final size of high vs. low food ration groups from the first study) that had been fed proportionally similar food rations and tested for the retention of pumpkinseed odour 24 hours and 8 days post-conditioning. They reported no difference in the retention of learned information (Fig. 4.2B). These data

20 A. High vs. low-growth rates Change in time moving (seconds)

BLBK374-04

B. Large vs. small initial mass

0

–20

–40

–60

–80

High Low growth rate growth rate

High Low growth rate growth rate

24 hours post-conditioning 8 days post-conditioning

Large initial size

Small initial size

24 hours post-conditioning

Small Large initial size initial size

8 days post-conditioning

Fig. 4.2 Mean (SE) change in time moving for juvenile rainbow trout initially conditioned with trout alarm cue paired with a novel predator (pumpkinseed) odour and subsequently exposed to pumkinseed odour alone. Panel A: Trout on a high versus low growth (5% vs. 1% mean body weight day per day) diet exhibited similar reductions in time spent moving when tested for recognition of pumpkinseed odour 24 hours post-conditioning. However, when tested for recognition 8 days post-conditioning, only trout in the low-growth treatment exhibited recognition of pumpkinseed odour; trout on the high-growth diet were no different from controls (not shown). Panel B: In order to determine if these trends were due to growth rate or absolute size, trout of two size ranges (fed 1% mean body weight day per day) were conditioned and tested as above. Small (∼0.6 g) and large (∼1.8 g) trout exhibited similar responses when tested for recognition on both testing days, suggesting that retention of learned recognition is not influenced by absolute size. Modified from Brown et al. (2011b).

BLBK374-04

BLBK374-Brown

72

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

suggest that growth rate does indeed influence the length of the memory window. A likely proximate mechanism in this case might be that increased growth rates result in increased energy demands, resulting in a shift in threat-sensitive trade-offs in favour of risk-prone behavioural tactics in the case of high growth rate trout. This raises the intriguing possibility of information renewal (Brown & Chivers 2005). Given that predation risk is variable, we might expect that prey should indeed ‘forget’ acquired information that is no longer relevant. But under high-risk conditions (high rates of encounter), we would expect this information to be continually reinforced. The question remains if repeated exposure over time results in enhanced learning opportunities in which the learned response is greater (stronger) than that resulting from a single conditioning event or if the retention is extended, or both. Vilhunen (2006) conditioned hatchery-reared Arctic charr juveniles to the odour of pikeperch (Sander lucioperca, Percidae) which had been fed a diet of Arctic charr once or four times (with 4 days between each conditioning). Vilhunen (2006) reports stronger antipredator responses (relative to unconditioned controls) by charr exposed four times to the predator. In addition, when exposed to a live predator during staged encounters, those that received multiple exposures had a higher probability of survival than did those conditioned a single time. While not a direct test of the hypothesis that repeated exposure enhances intensity and retention of learning, it does suggest that multiple conditionings increase the strength (and possibly retention) of learned information.

4.7

Conservation, management and learning

Here, we have reviewed the latest developments in the field of chemically mediated predator recognition learning, highlighting the incredible degree of sophistication in this learning process. Moreover, we have highlighted the dynamic interplay between acquired predator recognition and threat-sensitive behavioural decision-making in prey fishes. In this final section, we will briefly touch on some conservation and management implications of this work.

4.7.1

Conditioning predator recognition skills

A commonly stated goal of predator training is that it may allow us to ‘teach’ threatened or commercially important species raised in hatcheries to recognise potential predation threats prior to stocking (Brown & Day 2002; Brown & Laland 2003; Bischof & Zedrosser 2009). Hatchery-reared fishes stocked into natural waterways often suffer intense predation pressure (Shively et al. 1996; Henderson & Letcher 2003), particularly within the first days to weeks post-stocking. However, to date, such an approach has met with limited success (Wisenden et al. 2004; Hawkins et al. 2007). For example, Hawkins et al. (2007) conditioned hatchery-reared juvenile Atlantic salmon to recognise pike and released conditioned and unconditioned salmon into a loch in which pike was the major predator of juvenile salmon. They report no difference in survival between the conditioned and unconditioned prey. These results differ from those of Berejikian et al. (1999) that found enhanced survival of Chinook salmon smolts following predator recognition training. One reason for the failure to find increased survival of trained prey may be the issue of retention. Hatchery

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

Char Count=

73

fishes are typically fed high-growth diets, possibly leading to reduced retention rates of acquired information (see Section 4.6). Another possibility might be the effects of hatchery selection itself. A growing body of literature shows that hatchery-breeding programmes may actually be selecting for high growth and risk-taking behavioural tactics (Malavasi et al. 2008; Houde et al. 2010). Selection acting on threat-sensitive tactics may reduce the likelihood of retention of responses to learned information. Studies comparing hatchery reared to wild stock conspecifics are needed to address this question.

4.7.2

Anthropogenic constraints

A growing volume of literature demonstrates significant impairment of alarm cue detection and function by prey exposed to sublethal concentrations of anthropogenic pollutants (Schotz et al. 2000; Scott et al. 2003; McPherson et al. 2004; Leduc et al. 2006; Kusch et al. 2008; Tierney et al. 2008). Presumably, an inability to detect alarm cues would result in impaired learning opportunities. Recent studies by Leduc et al. (2004a, 2006, 2010) demonstrate that juvenile salmonids are unable to detect conspecific alarm cues under weakly acidic conditions in laboratory and field trials. Moreover, the ability to learn (Leduc et al. 2004b, 2007a, 2007b) and retain (Smith et al. 2008) predator recognition is likewise impaired under weakly acidic conditions. Moreover, what remains unknown is whether such a sublethal effect on risk assessment and learning has real fitness consequences for prey fishes. Presumably, an inability to assess chemosensory predator cues might lead to increased risk of predation. Such an increase in predation risk may have direct effects (i.e. reduced survival) or indirect effects (reduced foraging opportunity, exclusion to suboptimal habitats or reduced recruitment). Clearly, additional laboratory and (perhaps most importantly) field studies are required to assess the potential impacts of anthropogenic stressors on prey fish populations.

4.7.3

Field-based studies

To date, only a handful of studies have examined learning under fully natural conditions. This represents a critical gap in our understanding of the functional importance of predator recognition learning. Clearly, laboratory studies are important, but they often lack ecological realism and are typically conducted under ‘ideal’ conditions (i.e. well-fed prey, absence of background predation risks). In order to assess the full ecological relevance of chemically mediated learning, additional studies are required to test the degree to which prey fishes rely on learned information under fully natural conditions. For example, one factor that deserves attention is the potential impact of abiotic conditions on chemically mediated learning in wild populations. While Leduc et al. (2007a, 2007b) have shown that juvenile Atlantic salmon can acquire the recognition of novel odours in natural streams, C.J. Macnaughton & G.E. Brown (unpublished data), working with the same population, failed to replicate these findings. One logical difference between these studies is that while Leduc conducted his trials during periods of low water depth and current speeds, Macnaughton & Brown (unpublished data) conducted their studies during 2 years of abnormally high rainfall (leading to increased stream depth and current speeds). This suggests that naturally occurring abiotic cycles may seasonally limit learning opportunities.

BLBK374-04

BLBK374-Brown

May 27, 2011

74

Fish Cognition and Behavior

4.8

Conclusions

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Carefully controlled laboratory studies have shown an exceptionally high degree of sophistication in the learning abilities of prey fishes. Prey fishes can be conditioned to recognise predators based on a single learning event. The prey learns to recognise not only the predator as a threat, but also the level of threat posed by the predator, making it possible for the prey to match the intensity of their antipredator response to the risk posed by the predator. Prey fishes continually update information regarding the risk level of predators and learn the temporal foraging patterns of their predators. Prey fishes can generalise recognition of one predator to other similar species and hence avoid some of the costs associated with learning. We are just beginning to understand how long prey should retain information about predators and how learning operates under natural conditions. In order to assess the full ecological relevance of chemically mediated learning, additional studies are required under fully natural conditions. Ideally, this ecologically relevant information should allow for the planning and implementation of realistic management and conservation programmes.

Acknowledgements We thank Chris Elvidge, Christopher Jackson and James Grant for their helpful comments. Financial support for our research on predator learning was provided by Concordia University, the University of Saskatchewan and NSERC of Canada.

References Alvarez, D. & Nicieza, A.G. (2003) Predator avoidance behaviour in wild and hatchery-reared brown trout: the role of experience and domestication. Journal of Fish Biology, 63, 1565–1577. Acquistapace, P., Hazlett, B.A. & Gherardi, F. (2003) Unsuccessful predation and learning of predator cues by crayfish. Journal of Crustacean Biology, 23, 364–370. Barreto, R.E., Luchiari, A.C. & Marcondes, A.L. (2003) Ventilatory frequency indicates visual recognition of an allopatric predator in Nile tilapia. Behavioural Processes, 60, 235–239. Bass, S.L.S. & Gerlai, R. (2008) Zebrafish (Danio rerio) responds differentially to stimulus fish: the effects of sympatric and allopatric predators and harmless fish. Behavioural Brain Research, 186, 107–117. Berejikian, B.A., Tezak, E.P. & LaRae, A.L. (2003) Innate and enhanced predator recognition in hatchery-reared Chinook salmon. Environmental Biology of Fishes, 67, 241–251. Berejikian, B.A., Smith, R.J.F., Tezak, E.P. et al. (1999) Chemical alarm signals and complex hatchery rearing habitats affect antipredator behaviour and survival of Chinook salmon (Oncorhynchus tshawytscha) juveniles. Canadian Journal of Fisheries and Aquatic Sciences, 56, 830–838. Bischof, R. & Zedrosser, A. (2009) The educated prey: consequences for exploitation and control. Behavioral Ecology, 20, 1228–1235. Blanchet, S., Bernatchez, L. & Dodson, J.J. (2007) Behavioural and growth responses of a territorial fish (Atlantic salmon, Salmo salar, L.) to multiple predatory cues. Ethology, 113, 1061–1072. Blumstein, D. (2006) The multipredator hypothesis and the evolutionary persistence of antipredator behavior. Ethology, 112, 209–217. Boersma, K.S., Ryer, C.H., Hurst, T.P. et al. (2008) Influences of divergent behavioural strategies upon risk allocation in juvenile flatfishes. Behavioral Ecology and Sociobiology, 62, 1959–1968.

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

Char Count=

75

Breden, F., Scott, M. & Michel, E. (1987) Genetic differentiation for anti-predator behavior in the Trinidadian guppy, Poecilia reticulata. Animal Behaviour, 35, 618–620. Br¨onmark, C. & Miner, J.G. (1992) Predator-induced phenotypic change in body morphology in crucian carp. Science, 258, 1348–1350. Brown, C. & Day, R. (2002) The future of stock enhancements: bridging the gap between hatchery practice and conservation biology. Fish and Fisheries, 3, 79–94. Brown, C. & Laland, K.N. (2003) Social learning in fishes: a review. Fish and Fisheries, 4, 280– 288. Brown, G.E. & Chivers, D.P. (2005) Learning as an adaptive response to predation. In: P. Barbosa & I. Castellanos (eds) Ecology of Predator–Prey Interactions, pp. 34–54. Oxford University Press, Oxford. Brown, G.E. & Chivers, D.P. (2006) Learning about danger: chemical alarm cues and the assessment of predation risk by fishes. In: C. Brown, K. Laland & J. Krause (eds) Fish Cognition and Behavior, pp 49–69. Blackwell Publishing Ltd., Oxford. Brown, G.E. & Magnavacca, G. (2003) Predator inspection behaviour in a characin fish: an interaction between chemical and visual information? Ethology, 109, 739–750. Brown, G.E. & Smith, R.J.F. (1998) Acquired predator recognition in juvenile rainbow trout (Oncorhynchus mykiss): conditioning hatchery-reared fish to recognize chemical cues of a predator. Canadian Journal of Fisheries and Aquatic Sciences, 55, 611–617. Brown, G.E., Poirier, J.-F. & Adrian, J.C., Jr. (2004) Assessment of local predation risk: the role of subthreshold concentrations of chemical alarm cues. Behavioral Ecology, 15, 810–815. Brown, G.E., Chivers, D.P. & Smith, R.J.F. (1997) Differential learning rates of chemical versus visual cues of a northern pike by fathead minnows in a natural habitat. Environmental Biology of Fishes, 49, 89–96. Brown, G.E., LeBlanc, V.J. & Porter, L.E. (2001) Ontogenetic changes in the response of largemouth bass (Micropterus salmoides, Centrarchidae, Perciformes) to heterospecific alarm pheromones. Ethology, 107, 401–414. Brown, G.E., Bongiorno, T., DiCapua, D.M. et al. (2006a) Effects of group size on the threat-sensitive response to varying concentrations of chemical alarm cues by juvenile convict cichlids. Canadian Journal of Zoology, 84, 1–8. Brown, G.E., Rive, A.C., Ferrari, M.C.O. et al. (2006b) The dynamic nature of antipredator behavior: prey fish integrate threat-sensitive antipredator responses within background levels of predation risk. Behavioral Ecology and Sociobiology, 61, 9–16. Brown, G.E., Ferrari, M.C.O., Malka, P.H. et al. (2011a) Generalization of predators and nonpredators by juvenile rainbow trout: learning what is and what is not a threat. Animal Behaviour. doi: 10.1016/j.anbehav.2011.03.013. Brown, G.E., Ferrari, M.C.O., Malka, P.H., et al. (2011b) Growth rate and retention of learned predator cues by juvenile rainbow trout: faster-growing fish forget sooner. Behavioral Ecology and Sociobiology. doi: 10-1007/s00265-011-1140-3. Brown, G.E., Macnaughton, C.J., Elvidge, C.K. et al. (2009) Provenance and threat-sensitive predator avoidance patterns in wild-caught Trinidadian guppies. Behavioral Ecology and Sociobiology, 63, 699–706. Brydges, N.M., Heathcote, R.J.P. & Braithwaite, V.A. (2008) Habitat stability and predation pressure influence learning and memory in populations of three-spined sticklebacks. Animal Behaviour, 75, 935–942. Chivers, D.P. & Mirza, R.S. (2001) Predator diet cues and the assessment of predation risk by aquatic vertebrates: a review and prospectus. In: A. Marchlewska-Koj, J.J. Lepri & D. M¨uller-Schwarze (eds) Chemical Signals in Vertebrates, Vol. 9, pp. 227–284. Kluwer Academic, New York. Chivers, D.P. & Smith, R.J.F. (1994a) Fathead minnows, Pimephales promelas, acquire predator recognition when alarm substance is associated with the sight of unfamiliar fish. Animal Behaviour, 48, 597–605. Chivers, D.P. & Smith, R.J.F. (1994b) The role of experience and chemical alarm signaling in predator recognition by fathead minnows, Pimephales promelas. Journal of Fish Biology, 44, 273–285.

BLBK374-04

BLBK374-Brown

76

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Chivers, D.P. & Smith, R.J.F. (1995) Free-living fathead minnows rapidly learn to recognize pike as predators. Journal of Fish Biology, 46, 949–954. Chivers, D.P. & Smith, R.J.F. (1998) Chemical alarm signalling in aquatic predator–prey systems: a review and prospectus. Ecoscience, 5, 338–352. Chivers, D.P., Brown, G.E. & Smith, R.J.F. (1995) Acquired recognition of chemical stimuli from pike, Esox lucius, by brook sticklebacks, Culaea inconstans (Osteichthyes, Gasterosteidae). Ethology, 99, 224–234. Chivers, D.P., Mirza, R.S., Bryer, P.J. et al. (2001) Threat-sensitive predator avoidance by slimy sculpins: understanding the role of visual versus chemical information. Canadian Journal of Zoology, 79, 867–873. Darwish, T.L., Mirza, R.S., Leduc, A.O.H.C. et al. (2005) Acquired recognition of novel predator odour cocktails by juvenile glowlight tetras. Animal Behaviour, 70, 83–89. Dupuch, A., Magnan, P. & Dill, L.M. (2004) Sensitivity of northern redbelly dace, Phoxinus eos, to chemical alarm cues. Canadian Journal of Zoology, 82, 407–415. Ferrari, M.C.O. & Chivers, D.P. (2006a) The role of learning in the development of threat-sensitive predator avoidance: how do fathead minnows incorporate conflicting information? Animal Behaviour, 71, 19–26. Ferrari, M.C.O. & Chivers, D.P. (2006b) The role of latent inhibition in acquired predator recognition by fathead minnows. Canadian Journal of Zoology, 84, 505–509. Ferrari, M.C.O. & Chivers, D.P. (2009) Latent inhibition of predator recognition by embryonic amphibians. Biology Letters, 5, 160–162. Ferrari, M.C.O., Lysak, K. & Chivers, D.P. (2010c) Turbidity as an ecological constraint on learned predator recognition and generalization in a prey fish. Animal Behaviour, 79, 515–519. Ferrari, M.C.O., Messier, F. & Chivers, D.P. (2006b) The nose knows: minnows determine predator proximity and density through detection of predator odours. Animal Behaviour, 72, 927–932. Ferrari, M.C.O., Messier, F. & Chivers, D.P. (2008a) Larval amphibians learn to match antipredator response intensity to temporal patterns of risk. Behavioral Ecology, 19, 980–983. Ferrari, M.C.O., Messier, F. & Chivers, D.P. (2008c) Can prey exhibit threat-sensitive generalization of predator recognition? Extending the predator recognition continuum hypothesis. Proceedings of the Royal Society of London, Series B, 275, 1811–1816. Ferrari, M.C.O., Sih, A. & Chivers, D.P. (2009a) The paradox of risk allocation: a review and prospectus. Animal Behaviour, 78, 579–585. Ferrari, M.C.O., Brown, G.E., Bortolotti, G.R. et al. (2010b) Linking predator risk and uncertainty to adaptive forgetting: a theoretical framework and empirical test using tadpoles. Proceedings of the Royal Society of London, Series B, 277, 2205–2210. Ferrari, M.C.O., Brown, G.E., Messier, F. et al. (2009b) Threat-sensitive generalization of predator recognition by amphibians. Behavioral Ecology and Sociobiology, 63, 1369–1375. Ferrari, M.C.O., Gonzalo, A., Messier, F. et al. (2007) Generalization of learned predator recognition: an experimental test and framework for future studies. Proceedings of the Royal Society of London, Series B, 274, 1853–1859. Ferrari, M.C.O., Trowell, J.J., Brown, G.E. et al. (2005) The role of learning in the development of threat-sensitive predator avoidance by fathead minnows. Animal Behaviour, 70, 777–784. Ferrari, M.C.O., Capitania-Kowk, T. & Chivers, D.P. (2006c) The role of learning in the acquisition of threat-sensitive responses to predator odours. Behavioral Ecology and Sociobiology, 60, 522– 527. Ferrari, M.C.O., Elvidge, C.K., Jackson, C.D. et al. (2010a) The responses of prey fish to temporal variation in risk: sensory habituation or risk assessment? Behavioral Ecology, 21, 532– 536. Ferrari, M.C.O., Vavrek, M.A., Elvidge, C.K. et al. (2008b) Sensory complementation and the acquisition of predator recognition by salmonid fishes. Behavioral Ecology and Sociobiology, 63, 113–121. Foam, P.E., Harvey, M.C., Mirza, R.S. et al. (2005a) Heads up: juvenile convict cichlids rely on chemosensory information to make threat-sensitive foraging decisions. Animal Behaviour, 70, 601–607.

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

Char Count=

77

Foam, P.E., Mirza, R.S., Chivers, D.P. et al. (2005b) Juvenile convict cichlids (Archocentrus nigrofasciatus) allocate foraging in response to temporal variation in predation risk. Behaviour, 142, 129–144. Foote, C.J. & Brown, G.S. (1998) Ecological relationship between freshwater sculpins (genus Cottus) and beach-spawning sockeye salmon (Oncorhynchus nerka) in Iliamna Lake, Alaska. Canadian Journal of Fisheries and Aquatic Sciences, 55, 1524–1533. Gibson, A.K. & Mathis, A. (2006) Opercular beat rate for rainbow darters Etheostoma caeruleum exposed to chemical stimuli from conspecific and heterospecific fishes. Journal of Fish Biology, 69, 224–232. Gilliam, J.F. & Fraser, D.F. (2001) Movement in corridors: enhancement by predation threat, disturbance, and habitat structure. Ecology, 82, 124–135. Godin, J.-G.J. & Smith, S.A. (1988) A fitness cost of foraging in the guppy. Nature, 333, 69–71. Golub, J.L., Vermette, V. & Brown, G.E. (2005) The response of pumpkinseed sunfish to conspecific and heterospecific chemical alarm cues under natural conditions: the effects of stimulus type, habitat and ontogeny. Journal of Fish Biology, 66, 1073–1081. Gonzalo, A., L´opez, P. & Mart´ın, J. (2009) Learning, memory and apparent forgetting of chemical cues from new predators by Iberian green frog tadpoles. Animal Cognition, 12, 745–750. Gotceitas, V. & Brown, J.A. (1993) Substrate selection by juvenile Atlantic cod (Gadus morhua): effects of predation risk. Oecologia, 93, 31–37. Griffin, A.S. (2004). Social learning about predators: a review and prospectus. Learning and Behavior, 32, 131–140. Griffin, A.S., Evans, C.S. & Blumstein, D.T. (2001) Learning specificity in acquired predator recognition. Animal Behaviour, 62, 577–589. Hawkins, L.A., Armstrong, J.D., & Magurran, A.E. (2007) A test of how predator conditioning influence survival of hatchery-reared Atlantic salmon, Salmo salar, in restocking programmes. Fisheries Management and Ecology, 14, 291–293. Hawkins, L.A., Magurran, A.E. & Armstrong, J.D. (2004) Innate predator recognition in newlyhatched Atlantic salmon. Behaviour, 141, 1249–1262. Hawkins, L.A., Magurran, A.E. & Armstrong, J.D. (2008) Ontogenetic learning of predator recognition in hatchery-reared Atlantic salmon, Salmo salar. Animal Behaviour, 75, 1663–1671. Helfman, G.S. (1989) Threat-sensitive predator avoidance in damselfish–trumpetfish interactions. Behavioral Ecology and Sociobiology, 24, 47–58. Henderson, J.N. & Letcher, B.H. (2003) Predation on stocked Atlantic salmon (Salmo salar) fry. Canadian Journal of Fisheries and Aquatic Sciences, 60, 32–42. Hirvonen, H., Ranta, E., Rita, H. et al. (1999) Significance of memory properties in prey choice decisions. Ecological Modelling, 115, 177–189. Hirvonen, H., Ranta, E., Piironen, J. et al. (2000) Behavioural responses of na¨ıve Arctic charr young to chemical cues from salmonid and non-salmonid fish. Oikos, 88, 191–199. Houde, A.L.S., Fraser, D.J. & Hutchings, J.A. (2010) Reduced anti-predator responses in multigenerational hybrids of farmed and wild Atlantic salmon (Salmo salar L.). Conservation Genetics, doi: 10.1007/s10592-009-9892-2. Jachner, A. (2001) Anti-predator behaviour of na¨ıve compared with experience juvenile roach. Journal of Fish Biology, 59, 1313–1322. Jachner, A. & Rydz, M.A. (2002) Behavioural response of roach (Cyprinidae) to different doses of chemical alarm cues (Schreckstoff). Archieves Hydrobiologica, 155, 369–381. Jord˜ao, L.C. & Volpato, G.L. (2000) Chemical transfer of warning information in non-injured fish. Behaviour, 137, 681–690. Kats, L.B. & Dill, L.M. (1998) The scent of death: chemosensory assessment of predation risk by prey animals. Ecoscience, 5, 361–394. Killen, S.S. & Brown, J.A. (2006) Energetic costs of reduced foraging under predation threat in newly hatched ocean pout. Marine Ecology Progress Series, 321, 255–266. Kim, J.-W., Brown, G.E., Dolinsek, I.J. et al. (2009) Combined effects of chemical and visual information in eliciting antipredator behaviour in juvenile Atlantic salmon Salmo salar. Journal of Fish Biology, 74, 1280–1290.

BLBK374-04

BLBK374-Brown

78

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Kristensen, E.A. & Closs, G.P. (2004) Anti-predator response of naive and experienced common bully to chemical alarm cues. Journal of Fish Biology, 64, 643–652. Kusch, R.C., Mirza, R.S. & Chivers, D.P. (2004) Making sense of predator scents: investigating the sophistication of predator assessment abilities of fathead minnows. Behavioral Ecology and Sociobiology, 55, 551–555. Kusch, R.C., Krone, P.H. & Chivers, D.P. (2008) Chronic exposure to low concentrations of waterborne cadmium during embryonic and larval development of zebrafish results in long-term hindrance of antipredator responses to alarm cues. Environmental Toxicology and Chemistry, 27, 705–710. Leduc, A.O.H.C., Kelly, J.M. & Brown, G.E. (2004a) Detection of conspecific chemical alarm cues by juvenile salmonids under neutral and weakly acidic conditions: laboratory and field tests. Oecologia, 139, 318–324. Leduc, A.O.H.C., Ferrari, M.C.O., Kelly, J.M. et al. (2004b) Learning to recognize novel predators under weakly acidic conditions: the effects of reduced pH on acquired predator recognition by juvenile rainbow trout (Oncorhynchus mykiss). Chemoecology, 14, 107–112. Leduc, A.O.H.C., Roh, E., Harvey, M.C. et al. (2006) Impaired detection of chemical alarm cues by juvenile Atlantic salmon (Salmo salar) in a weakly acidic environment. Canadian Journal of Fisheries and Aquatic Sciences, 63, 2356–2363. Leduc, A.O.H.C., Roh, E., Breau, C. et al. (2007a) Learned recognition of a novel odour by wild juvenile Atlantic salmon, Salmo salar, under fully natural conditions. Animal Behaviour, 73, 471–477. Leduc, A.O.H.C., Roh, E., Breau, C. et al. (2007b) Effects of ambient acidity on chemosensory learning: an example of an environmental constraint on acquired predator recognition in wild juvenile Atlantic salmon (Salmo salar). Ecology of Freshwater Fishes, 16, 385–394. Leduc, A.O.H.C., Roh, E., Macnaughton, C.J. et al. (2010) Ambient pH and the response to chemical alarm cues in juvenile Atlantic salmon: mechanisms of reduced behavioral responses. Transactions of the American Fisheries Society, 139, 117–128. Lima, S.L. & Bednekoff, P.A. (1999) Temporal variation in danger drives anti-predator behavior: the predator risk allocation hypothesis. American Naturalist, 153, 649–659. Lima, S.L. & Dill, L.M. (1990) Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology, 68, 619–640. Lima, S.L. & Steury, T.D. (2005) Perception of predation risk: the foundation of nonlethal predatorprey interactions. In: P. Barbosa & I. Castellanos (eds) Ecology of Predator–Prey Interactions, pp. 166–188, Oxford University Press, Oxford. Mackney, P.A. & Hughes, R.N. (1995) Forgaging behaviour and memory window in sticklebacks. Behaviour, 132, 1241–1253. Malavasi, S., Georgalas, V., Mainardi, D. et al. (2008) Antipredator responses to overhead fright stimuli in hatchery-reared and wild European sea bass (Dicentrarchus labrax L.) juveniles. Aquaculture Research, 39, 276–282. Mangel, M. (1990) Dynamic information in uncertain and changing worlds. Journal of Theoretical Biology, 146, 317–332. McNamara, J.M. & Houston, A.I. (1987) Memory and the efficient use of information. Journal of Theoretical Biology, 125, 385–395. Magurran, A.E. (1989) Acquired recognition of predator odour in the European minnow (Phoxinus phoxinus). Ethology, 82, 216–233. Magurran, A.E. (2005) Evolutionary Ecology: The Trinidadian Guppy. Oxford Series in Ecology and Evolution. Oxford University Press, Oxford. Mathis, A. & Smith, R.J.F. (1993a) Fathead minnows (Pimephales promelas) learn to recognize pike (Esox lucius) as predators on the basis of chemical stimuli in the pike’s diet. Animal Behaviour, 46, 645–656. Mathis, A. & Smith, R.J.F. (1993b) Chemical labelling of northern pike, Esox lucius, by the alarm pheromone of fathead minnows, Pimephales promelas. Journal of Chemical Ecology, 19, 1967–1979.

BLBK374-04

BLBK374-Brown

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning about Danger: Chemical Alarm Cues and Threat Sensitivity

Char Count=

79

McPherson, T.D., Mirza, R.S. & Pyle, G.G. (2004) Responses of wild fishes to alarm chemicals in pristine and metal-contaminated lakes. Canadian Journal of Zoology, 82, 694–700. Mirza, R.S. & Chivers, D.P. (2000) Predator-recognition training enhances survival of brook trout: evidence from laboratory and field-enclosure studies. Canadian Journal of Zoology, 78, 2198–2208. Mirza, R.S. & Chivers, D.P. (2001) Chemical alarm signals enhance survival of brook charr, Salvelinus fontinalis, during encounters with predatory chain pickerel, Esox niger. Ethology, 107, 989–1006. Mirza, R.S., Mathis, A. & Chivers, D.P. (2006) Does temporal variation in predation risk influence the intensity of anti-predator responses? A test of the risk allocation hypothesis. Ethology, 112, 44–51. Nilsson, P.A. & Br¨onmark, C. (2000) Prey vulnerability to a gape-limited predator: behavioural and morphological impacts on northern pike piscivory. Oikos, 88, 539–546. Olson, M.H. (1996) Ontogenetic niche shifts in largemouth bass: variability and consequences for first-year growth. Ecology, 77, 179–190. Olson, M.H., Mittelback, G.G. & Osenberg, C.W. (1995) Competition between predator and prey: resource-based mechanisms and implication for stage-structured dynamics. Ecology, 76, 1758–1771. Pravosudov, V.V. & Clayton, N.S. (2002) A test of the adaptive specialization hypothesis: population differences in caching, memory and the hippocampus in black-capped chickadees (Poecile atricapilla). Behavioral Neuroscience, 116, 515–522. Reebs, S. (1999) Time-place learning based on food but not on predation risk in a fish, the inanga (Galaxias maculates). Ethology, 105, 361–371. Reiechert, S.E. (2005) Patterns of inheritance of traits associated with predator foraging behavior. In: P. Barbosa & I. Castellanos (eds) Ecology of Predator–Prey Interactions, pp. 55–76. Oxford University Press, Oxford. Schotz, N.L., Turelove, N.K., French, B.L. et al. (2000) Diazinon disrupts antipredator and homing behaviors in Chinook salmon (Oncorhynchus tshawytscha). Canadian Journal of Fisheries and Aquatic Sciences, 57, 1911–1918. Scheurer, J.A., Berejikian, B.A., Thrower, F.P. et al. (2007) Innate predator recognition and freight response in related populations of Oncorhynchus mykiss under different predation pressure. Journal of Fish Biology, 70, 1057–1069. Scott, G.R., Sloman, K.A., Rouleau, C. et al. (2003) Cadmium disrupts behavioural and physiological responses to alarm substance in juvenile rainbow trout (Oncorhynchus mykiss). Journal of Experimental Biology, 206, 1779–1790. Shively, R.S., Poe, T.P. & Sauter, S.T. (1996) Feeding response by northern squawfish to a hatchery release of juvenile salmonids in the Clearwater River, Idaho. Transactions of the American Fisheries Society, 125, 230–236. Sih, A. (1992) Prey uncertainty and the balancing of antipredator and foraging needs. American Naturalist, 139, 1052–1069. Sih, A., Ziemba, R. & Harding, K.C. (2000) New insights on how temporal variation in predation risk shapes prey behavior. Trends in Ecology and Evolution, 15, 3–4. Smith, J.J., Leduc, A.O.H.C. & Brown, G.E. (2008) Chemically mediated learning in juvenile rainbow trout. Does predator odour pH influence intensity and retention of acquired predator recognition? Journal of Fish Biology, 72, 1750–1760. Smith, M.E. & Belk, M.C. (2001) Risk assessment in western mosquitofish (Gambusia affinis): do multiple cues have additive effects? Behavioral Ecology and Sociobiology, 51, 101–107. Smith, R.J.F. (1997) Avoiding and deterring predators. In: J.-G.J. Godin (ed) Behavioural Ecology of Teleost Fishes, pp. 163–190. Oxford University Press, Oxford. Sneddon, L.U., Braithwaite, V.A. & Gentle, M.J. (2003) Novel object test: examining nociception and fear in the rainbow trout. The Journal of Pain, 4, 431–440. Stankowich, T. & Coss, R.G. (2007) The re-emergence of felid camouflage with the decay of predator recognition in deer under relaxed selection. Proceedings of the Royal Society of London, Series B, 274, 175–182.

BLBK374-04

BLBK374-Brown

80

May 27, 2011

10:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Tierney, K.B., Sampson, J.L., Ross, P.S. et al. (2008) Salmon olfaction is impaired by an environmentally realistic pesticide mixture. Environmental Science and Technology, 42, 4996–5001. Vavrek, M.A. & Brown, G.E. (2009) Threat-sensitive responses to disturbance cues in juvenile convict cichlids and rainbow trout. Annales Zoologica Fennici, 46, 171–180. Vavrek, M.A., Elvidge, C.K., DeCaire, R. et al. (2008) Disturbance cues in freshwater prey fishes: do juvenile convict cichlids and rainbow trout respond to ammonium as an ‘early warning’ signal? Chemoecology, 18, 255–261. Vilhunen, S. (2006). Repeated antipredator conditioning: a pathway to habituation or to better avoidance? Journal of Fish Biology, 68, 25–43. Vilhunen, S. & Hirvonen, H. (2003) Innate antipredator response of Arctic charr (Salvelinus alpinus) depend on predator species and their diet. Behavioral Ecology and Sociobiology, 55, 1–10. Werner, E.E. & Gilliam, J.F. (1984) The ontogenetic niche and species interactions in size-structured populations. Annual Reviews in Ecology and Systematics, 15, 393–425. Wisenden, B.D. & Chivers, D.P. (2006) The role of public chemical information in antipredator behaviour. In: F. Ladich, S.P. Collins, P. Moller, & B.G. Kapoor (eds) Fish Chemoreception, pp 259–278. Science Publisher, Enfield, NH. Wisenden, B.D., Chivers, D.P. & Smith, R.J.F. (1995) Early warning in the predation sequence: a disturbance pheromone in Iowa darters (Etheostoma exile). Journal of Chemical Ecology, 21, 1469–1480. Wisenden, B.D., Klitzke, J., Nelson, R. et al. (2004) Predator-recognition training of hatchery-reared walleye (Stizostedion vitreum) and a field test of a training method using yellow perch (Perca favescens). Canadian Journal of Fisheries and Aquatic Sciences, 61, 2144–2150. Yl¨onen, H., Kortet, R., Myntti, J. et al. (2006) Predator odor recognition and antipredator response in fish: does the prey know the predator diel rhythm? Acta Oecologica, 31, 1–7. Zhao, X., Ferrari, M.C.O. & Chivers, D.P. (2006) Threat-sensitive learning of predator odours by a prey fish. Behaviour, 143, 1103–1121.

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 5

Learning and Mate Choice Klaudia Witte and Sabine N¨obel

5.1

Introduction

One of the most fascinating questions within the field of sexual selection is why and how females and males choose particular conspecifics as mates. Sexual selection theories provide different explanations for the origin and evolution of ornamental traits and mate preferences for such traits (overview in Andersson 1994; Jennions et al. 2001; Kokko et al. 2003; Kokko & Jennions 2008; Clutton-Brock 2009). Genes for the ornamental trait and those for the concomitant preference can co-evolve as a result of a genetic linkage between those genes as a result of the Fisherian runaway process (Fisher 1930; Lande 1981; Kirkpatrick 1982; Brooks 2000) or as a result of selection for mates advertising ‘good genes’ (indicator models; Zahavi 1975; Møller & Alatalo 1999). These models generally assume that mate preferences are genetically based (Bakker 1999; Iwasa & Pomiankowski 1999; Andersson & Simmons 2006). Mate preferences can be highly variable within populations. Part of this variation is owing to flexibility in mating preferences expressed by individuals during their lifetime (Jennions & Petrie 1997). Of increasing interest is the potential influence of the social environment on mate-choice decisions (e.g. Gibson & H¨oglund 1992; Dugatkin 1996a, 1996b; Witte 2006). Forming mate preferences is a complex process involving not only genetic factors but also non-genetic factors. Increasing evidence suggests that the social environment (Dugatkin 1996a; Westneat et al. 2000) and learning are important factors in forming mate preferences. The mate choice of conspecifics influences the mate-choice decisions of an individual, who can alter mate preferences through learning processes. Therefore, social learning and using public information (Danchin et al. 2004; Dall et al. 2005; Bonnie & Earley 2007; and see Chapter 11) and other kinds of learning significantly influence the process of sexual selection. Forms of social learning have now been recognised as meaningful mechanisms for the non-genetic inheritance (i.e. cultural transmission) of mate preferences (Brooks 1998), leading to cultural evolution of mate preferences. This chapter illustrates how learning is involved in the mate choices of fishes and emphasises the important roles that different kinds of learning, particularly social learning, play in sexual selection. It focuses on four different kinds of learning: (1) sexual imprinting; (2) learning

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-05

BLBK374-Brown

82

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

after reaching maturity; (3) eavesdropping including audience-effect; (4) and mate-choice copying.

5.2

Sexual imprinting

Sexual imprinting is a learning process restricted to a specific period during early development (the sensitive phase), which influences mate preferences later on in life (Immelmann 1972). A prerequisite for sexual imprinting is that at least one parent, a genetic or social parent, cares for the young to ensure that young have intense contact with the parent(s) and get the opportunity to learn specific traits of the parent(s). Mate preferences learned by sexual imprinting can be transmitted from one generation to the next generation in a socially inherited way. Thus, sexual imprinting is assumed to be a powerful mechanism for the cultural evolution of mate preferences. Theoretical models show that sexual imprinting potentially plays an important role in sexual selection (Aoki et al. 2001) and in sympatric speciation (Laland 1994a).

5.2.1

Does sexual imprinting promote sympatric speciation in fishes?

Sexual imprinting in cichlids, which was studied intensively in the 1970s and 1980s (Fern¨o & Sj¨olander 1976; Siepen & Capron de Caprona 1986; more details in Witte 2006), has begun to receive attention within behavioural ecology again, largely because sexual imprinting is thought to be a meaningful mechanism for promoting sympatric speciation. Verzijden & ten Cate (2007) investigated the effect of sexual imprinting on the mother’s phenotype in two sister species of African mouth-breeding cichlids (Pundamilia nyererei, P. pundamilia, Cichlidae) living in sympatry in the Lake Victoria, Africa. In both species, only females provide brood care to young. In a cross-fostering experiment, Verzijden & ten Cate (2007) exchanged eggs between breeding females of the two different species. Results show that cross-fostering influenced the mate-choice decisions of females in these species. Thus, female offspring sexually imprinted on the mother’s phenotype and preferred heterospecific males over conspecific males when reared by a heterospecific foster mother. In these species sexual imprinting on the mother’s phenotype forms mate preferences of young females and can promote reproductive isolation in sympatry. In 2009, Verzijden et al. repeated the cross-fostering experiment mentioned above with males of P. pundamilia and P. nyererei and investigated male–male aggression behaviour and male mate preferences. Males raised by conspecific and heterospecific foster mothers showed no differences in mate preferences and preferred conspecific females as mates. Males of both species directed aggression in territorial defence primarily towards conspecific intruders. Thus, there is no evidence for sexual imprinting in males in these two species. Both studies together show that there is a sex difference in the potential of sexual imprinting in these cichlid species as has been shown in bird species (Witte & Sawka 2003; Witte & Caspers 2006). Sexual imprinting might also play a role in forming mate preferences in the three-spined stickleback (Gasterosteus aculeatus, Gasteroidae), a species in which males care intensively for the young. In several lakes in British Columbia, Canada, three-spined sticklebacks occur

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

83

in sympatric species pairs. These pairs consist of a large-bodied invertebrate-feeding benthic species and a small-bodied zooplankton-feeding limnetic species (Schluter & McPhail 1992). In the same species-pair of sticklebacks, Kozak & Boughman (2009) showed that both sexes enhance conspecific mate preferences through experience with them, but in opposite directions. Female limnetic sticklebacks learned to prefer conspecific males if they were raised with them. However, limnetic and benthic males learned to court conspecifc females less when reared with conspecifics. It was the experience with heterospecific females that enhances discrimination against heterospecifics. Thus, males learn to discriminate against heterospecific females via sexual imprinting, whereas limnetic females learn to prefer conspecifics by imprinting on siblings. Sexual imprinting can potentially be an important learning process in fishes for forming mate preferences. It may be worth investigating which factors facilitate the occurrence of sexual imprinting in a species and its potential role in sympatric speciation.

5.3

Learning after reaching maturity

Whereas sexual imprinting is a learning process that is restricted to a specific sensitive period during early development, there are other learning processes that are not restricted to a specific sensitive period and occur in sexually mature individuals. Learning processes exhibited later on in life that affect mate choice are observed in a wide range of fish species, including those without any parental care, like livebearing fishes. In an early study, Haskins & Haskins (1950) showed that when guppy (Poecilia reticulata, Poeciliidae) males were reared in isolation until sexual maturity and then exposed to females of a specific colour variant, which differed from their own colour variant, for a month or longer, males preferred females of the other colour variant with which males were reared after reaching sexual maturity. Liley (1966) investigated species recognition in four sympatric species within the family Poeciliidae. Guppy males with female experience restricted to conspecifics did not show a preference for conspecific females when females of three other species were present. He concluded that males must require experience with females of their own species, as well as with females of other species, to learn to discriminate between conspecific and heterospecific females. Haskins & Haskins (1949) investigated whether male guppies learn to discriminate between conspecific and heterospecific females by experience. They presented guppy males with females of three related species (P. reticulata, P. picta and P. vivipara). Male guppies that had had no experience with heterospecifics, initially directed most of their courtship displays towards swamp guppy females (P. picta). However, after about a week males courted mostly conspecific females (P. reticulata). Magurran & Ramnarine (2004) investigated sexually mature Trinidadian guppy (P. reticulata) males and swamp guppy (P. picta) males living in either sympatry or allopatry. In a baseline mate-choice test, two P. reticulata males, collected from the same locality, or two P. picta males, could physically interact with one P. reticulata female and one P. picta female, matched for size. One of the two males was the focal male and the authors recorded the number of sneaky matings of that male. Males of both species living in allopatry attempted matings with heterospecific females and conspecific females at random. However, males living in sympatry with the other species preferred to attempt matings with females of their own species. In a test for

BLBK374-05

BLBK374-Brown

84

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

learned preferences, two males, both from the same localities, were housed together with two size-matched females, one of each species. Trinidadian guppy males with no experience with P. picta females learned to discriminate between heterospecific and conspecific females and preferred conspecific females within a few days. Thus, learning may modify mate preferences in guppy males and may help to prevent them from mating with the wrong species. Such modification of choice via learning may be mediated via feedback emanating from potential mates. Several other studies have shown that male preferences in guppy females are also altered by experience (Breden et al. 1995; Rosenqvist & Houde 1997; Jirotkul 1999).

5.4

Eavesdropping

Eavesdropping is defined as the act of extracting information from signalling interactions between conspecifics (McGregor & Dabelsteen 1996). Eavesdropping occurs when information from an animal transmitting a signal to another individual is picked up by one or more bystanders towards whom the signal was not directed (McGregor 2005). Eavesdropping is now recognised as representing an important component of animal communication, particularly communication in a network, and has been studied intensively in songbirds and fishes (Peake 2005). Females can gain information about potential males by assessing their quality on the basis of morphological cues (Endler & Houde 1995; Houde 1997). In addition, by observing two males interacting (e.g. fighting) with each other, females gather further reliable information about these males that they can then use to guide mate-choice decisions. Eavesdropping can be an effective way for females to evaluate potential males.

5.4.1

Eavesdropping and mate choice

Doutrelant & McGregor (2000) investigated whether female Siamese fighting fish (Betta splendens, Osphronemidae) monitor aggressive interactions between two males and whether the information gained by eavesdropping is used to guide mate-choice decisions. In a wellcontrolled experimental set-up, they found that females that had the opportunity to watch two displaying males subsequently first visited the winner significantly more often, spent significantly more time with the winner and displaying to the winner, than to other male. Females that had not seen the interaction between two males visited the loser first more often than females under other conditions, and did not behave differently to winner and loser. This experiment shows that females use the information gained from an aggressive interaction between two males in their mate-choice decision. In the sexually role-reversed pipefish Syngnathus typhle males that have observed fighting females preferred to associate with the more dominant female than with the more attractive female (Berglund & Rosenqvist 2001).

5.4.2

Benefits of eavesdropping

What are the benefits of eavesdropping? As eavesdroppers are not engaged in conspicuous, probably risky communication, eavesdropping could constitute an easy, rapid and safe

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

85

way to acquire information on conspecifics and should be rather common (McGregor 1993). In general, mate choice is costly for females because it requires them to devote to evaluating males and may expose them to enhanced predation risk (Andersson 1994). Additionally, mate-sampling females may be injured in aggressive courtship displays by males or even suffer from harassment by males (Schlupp et al. 2001). Plath et al. (2007) performed a comparative approach with nine species of livebearing fish and examined costs of male sexual harassment for females as reduced feeding time. In all species females spent significantly less time feeding in the presence of a male. Ojanguren & Magurran (2007) could show that in guppies (P. reticulata) male harassment causes a direct reduction in female short-term fitness by significantly reducing the number of offspring produced. Eavesdropping females can avoid some of these costs, gaining information about male quality without being directly involved in an interaction with conspecifics. Moreover, they may be able to watch interactions between several conspecifics at the same time, thus providing direct comparisons. Females can then use this information about male quality gained from eavesdropping to supplement direct information gained on the basis of male morphological cues. Eavesdropping females gain information on the relative quality of males at little cost and/or risk (McGregor & Peake 2000). Information gained by observing an aggressive interaction between two individuals is assumed to be reliable and not subject to cheating. From this perspective, eavesdropping may be more reliable than mate-choice copying (see below), where a female may copy a ‘wrong’ choice of the model female. While eavesdropping seems to be a good strategy for mate choice, as yet there is no quantitative evidence for any fitness advantages based on this strategy.

5.4.3

The audience effect

The individual that eavesdrops not only gains information about the two interacting individuals, but the presence of the eavesdropper may also influence the nature of the interaction. This so-called ‘audience effect’ or ‘bystander effect’ has been intensively investigated in Siamese fighting fish, but recently also in the livebearers (Poeciliidae). Doutrelant et al. (2001) tested whether the presence of a female or male changed the intrasexual interaction between two fighting males. In the experiments, the two fighting males could interact with each other through clear partitions, over two trials. In one trial, both males saw a female prior to interacting with the other male. In the other trial, the same males did not see a female before the interaction started. A similar experiment was performed with a male as an audience. Results clearly show that a female audience changes the male–male interactions. With a female audience males performed more tail beats, spent more time with gill cover erected, interacted farther away from the other male and performed aggressive displays that are used only in male–male interactions and more of the displays that are considered more conspicuous used in the presence of both sexes. Conversely, whether a male audience was previously present or not did not significantly change the characteristics of the male–male interaction. A similar result was found by Matos & McGregor (2002) in the same species. When a male audience was present prior to encounter, males attempted significantly more bites and spent less time near the opponent than when a female audience was observed prior to the encounter.

BLBK374-05

BLBK374-Brown

86

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Due to the fact that female B. splendens eavesdrop on male contests, Herb et al. (2003) tested mate-choice decisions of male fighting fish after losing or winning a contest. They conducted two sets of trials: in one set-up they tested the losers’ courting preference for eavesdropping females or na¨ıve females (females which had no opportunity to observe the contestants). In the other set-up they tested the winners’ courting preference for eavesdropping females or na¨ıve females. The loser male preferentially courted a na¨ıve female, while the winner showed no preference for either female type. This suggests that loser males are able to assess the effect of their subordinate status on their attractiveness to females. By courting na¨ıve females, loser males may increase their chance of being accepted by that female, and reducing the costs of courting a female with a high probability of being rejected by her. These results suggest that B. splendens males moderate their behaviour in response to an audience in ways more complex than previously thought. Similar to Siamese fighting fish, males of the northern swordtail (Xiphophorus birchmanni, Poeciliidae) perform courtship display with an audience ‘in mind’ (Fisher & Rosenthal 2007). Male northern swordtails display large, sail-like dorsal fins. Dorsal fin raising is a dynamic component in both courtship displays and male–male aggression displays. In this species typically males with large dorsal fins are more dominant and more aggressive than other males. Fisher & Rosenthal (2007) showed that males, together with a mate, raise their dorsal fin more frequently when an audience male is present than with an audience female or no audience. They assume that male swordtails have not evolved courtship behaviour involving dorsal fin raising in response to exploit a pre-existing preference in females, but rather as a competitive signal to male bystanders. This assumption is supported by the fact that females reject males with large fins, possibly due to their general high aggressiveness. Recently, the audience effect has been investigated in the Atlantic molly (P. mexicana) (Plath et al. 2008a). The authors tested males in two mate-choice situations. Males could choose between a conspecific P. mexicana female and a heterospecific P. formosa female (experiment 1) or between a large conspecific female and a small conspecific female (experiment 2). Each experiment had two trials: first trials without an audience male and the second trial with an audience male. In both experiments, males spent significantly less time near the initially preferred female (a conspecific female or the larger of the two females) and spent significantly more time near the initially non-preferred female when a conspecific audience male was present. In a control Plath et al. (2008a) showed that males choose consistently without the presence of an audience male. This study highlights that the social environment has an important influence on male mate-choice decisions, and that visual presence of a conspecific competitor alone can affect mate-choice decisions. In another study (Plath et al. 2008b) P. mexicana males could physically interact both with a large conspecific female and a small conspecific female or with two size-matched females, one a conspecific P. mexicana female and the other a heterospecific P. formosa female. After the initial preference test, an audience male was presented to the focal male. In the presence of the audience male, focal males reduced their mating activity and preferred no longer one of the two females. Rather focal males directed their first sexual interaction toward the initially non-preferred female when an audience male was present. Plath et al. (2008b) suggest that focal males tried to deceive the audience male about their mating preferences to avoid sperm competition because surrounding males may use public information and copy the focal male’s mate choice (see Section 5.5).

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

87

In 2009, Plath et al. performed a similar study with P. mexicana females. However, females interacted with the audience female and their mate-choice decisions were not affected by the audience female. We have now evidence that P. mexicana females also respond to an audience female and change their mate preferences in the presence of an audience female (N¨obel et al. submitted). But why does a male audience affect male mate choice? One potential benefit of changing mate-choice decisions in the presence of a male audience is reducing sperm competition risk. Sperm competition can be reduced if the audience male copy the ‘deceived’ choice of the focal male. Ziege et al. (2009) investigated this aspect in P. mexicana males. They gave P. mexicana males the opportunity to choose between two different females (large vs. small) under five conditions. Results indicated that focal males showed weaker expression of mating preferences when being observed by a rival. This suggests that focal males tried to deceive surrounding males so that they will copy the false mate-choice decision. Focal males seem to gain benefits from concealing their original mating preferences. However, to understand the complex social information network in P. mexicana males in the context of mate choice, it is necessary to investigate whether P. mexicana males copy the mate choice of other males. There is now evidence that this is the case (N¨obel et al. submitted). Whether an individual acts as an eavesdropper or an audience makes a big difference for the information he or she can gather. In the case of an eavesdropper this individual is not recognised by the two interacting individuals and can get reliable information on the basis of signals transmitted back and forth between the two interacting individuals, i.e. on signals that are not directed to the eavesdropper. However, when two interacting individuals are aware of an audience, these individuals will change their behaviour in reaction to the audience. Fighting males might change their behaviour to ‘impress’ a watching female. However, males during mate choice might conceal their real mate preference and the audience male will receive false information. This shows that these small social networks are already highly complex. Further research might focus on benefits and costs for individuals that are eavesdropped and those individuals that make up the audience.

5.5

Mate-choice copying

Models of sexual selection assume that females and males choose among potential mates independently of other conspecifics and on the basis of their genetically determined preferences. However, there is strong evidence that females and males choose a mate nonindependently of other conspecifics and on the basis of social information. A sophisticated form of inadvertent social information (ISI) arises as a by-product of performance activities of individuals. When this by-product provides information about the quality of any environmental parameter, including conspecifics and heterospecifics, it is called public information (Valone 1989). Public information is gained by an individual observing the performance of others and used to estimate the quality of environmental parameters or by noting the behavioural decisions of other individuals and gain information about the quality of conspecifics (Valone 2007). Performance-based public information has been observed in the context of assessing the quality of potential mates (Danchin et al. 2004; Valone 2007). Mate-choice copying occurs when an observation of a sexual interaction between a male

BLBK374-05

BLBK374-Brown

88

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

and a female influences the subsequent mate-choice decision of the observing individual, biasing his or her decision to favour the observed mating individual. Mate-choice copying is a form of social learning in which individuals gain information and learn about potential mates by observing conspecifics (see Chapter 11 for a review of social learning in fishes). Mate-choice copying is an important mate-choice strategy demonstrating that individuals gather and use social and public information (Danchin et al. 2004; Dall et al. 2005). The first prerequisite and necessary condition for mate-choice copying to occur is that individuals must be able to observe the mate choices of others (Losey et al. 1986). To qualify as matechoice copying, it must be the sexual interaction, and not the consequence of the choice of a female or a male, that influences the mating decision of another (Pruett-Jones 1992). Thus, it is not mate-choice copying when females prefer to lay their eggs in nests that already contain eggs, as in the Bullhead goby (Cottus gobio, Cottidae; Marconato & Bisazza 1986), fathead minnow (Pimephales promelas, Cyprinidae; Unger & Sargent 1988), fantail darter (Ethiostoma flabellare, Percidae; Knapp & Sargent 1989), and three-spined sticklebacks (Ridley & Rechten 1981; Goldschmidt et al. 1993). This behaviour can be explained by dilution of the risk of egg predation or egg cannibalism (Rohwer 1978), or as resulting when male sticklebacks that have eggs in their nests court more vigorously and are, therefore, preferred by females (Jamieson & Colgan 1989). Mate-choice copying is most likely to occur in polygynous and promiscuous mating systems with no parental care or with maternal care only. Several theoretical models have investigated how mate-choice copying could evolve and be maintained in a population (Losey et al. 1986; Gibson & H¨oglund 1992; Pruett-Jones 1992; Laland 1994b; Nordell & Valone 1998; St¨ohr 1998; Sirot 2001; Uehara et al. 2005; Brennan et al. 2008). Wade & Pruett-Jones (1990) showed that copying is likely to increase the variance in mating success among males, and thus intensify sexual selection. Servedio & Kirkpatrick (1996) showed theoretically that an allele for copying can spread through a population even when there is mild selection against it.

5.5.1 Mate-choice copying – first experimental evidence and consequence The first experimental evidence for mate-choice copying came from Dugatkin et al.’s (1992) study with Trinidadian guppies (P. reticulata, Poeciliidae). Following his study, several other studies regarding mate-choice copying in fishes have been published (see Subsection 5.5.2) and research on this topic is continuing (Amlacher & Dugatkin 2005; Godin et al. 2005; Goulet & Goulet 2006; Hill & Ryan 2006; Widemo 2006; Witte 2006; Vukomanovic & Rodd 2007; Heubel et al. 2008; Frommen et al. 2009; Godin & Hair 2009; Witte & Godin 2010). At present there is good evidence for mate-choice copying in guppies (Dugatkin & Godin 1992, 1993; Dugatkin 1996a, 1996b, 1998a, 1998b; Amlacher & Dugatkin 2005; Godin et al. 2005; Vukomanovic & Rodd 2007; Godin & Hair 2009), and in other fish species like the sailfin molly (P. latipinna, Poeciliidae; Schlupp et al. 1994; Schlupp & Ryan 1997; Witte & Ryan 1998, 2002; Witte & Noltemeier 2002; Witte & Massmann 2003; Witte & Ueding 2003; Hill & Ryan 2006; Witte 2006; Heubel et al. 2008), the humpback limia (Limia nigrofasciata, Poeciliidae; Munger et al. 2004), the Japanese medaka

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

89

(Oryzias latipes, Adrianichthyidae; Grant & Green 1996, but see Howard et al. 1998), three-spined sticklebacks (G. aculeatus, Gasterosteidae; Frommen et al. 2009), white belly damselfish (Amblyglyphidodon leucogaster, Pomacentridae; Goulet & Goulet 2006), ocellated wrasse (Symphodus ocellatus, Labridae; Alonzo 2008) and deep-snouted pipefish (S. typhle, Syngnathidae; Widemo 2006). However, there are a few studies that failed to detect mate-choice copying in fish species. Brooks (1996) could not detect mate-choice copying in guppies from a feral South African population; Lafleur et al. (1997) found no indication for mate-choice copying in pet store guppies (but see Dugatkin 1998a for a comment on this study); and Patriquin-Meldrum & Godin (1998) found no evidence for mate-choice copying in the three-spined stickleback (but see Frommen et al. 2009). Ambiguous results for the Perugia’s limia (L. perugiae, Poeciliidae) were reported by Applebaum & Cruz (2000). It is currently unclear why mate-choice copying should be observed in some populations and not in others. According to the definition of Pruett-Jones (1992), female mate-choice copying occurs when ‘the conditional probability of choice of a given male by a female is either greater or less than the absolute probability of choice depending whether that male mated previously or was avoided, respectively. The outcome of the female is that if one female mates with or avoids a specific male successively choosing females will be accordingly more or less likely to mate with that male than they would otherwise have been’ (Pruett-Jones 1992, pp. 1001–1002). Thus, mate-choice copying may decrease the probability that a female mates with a particular male when the female has observed that another female has rejected that male. This idea was investigated experimentally by Witte & Ueding (2003). They showed that female sailfin mollies rejected a previously preferred male, after observing another female escaping from that male. So far this is the only study for copying mate rejection.

5.5.2

Mate-choice copying – evidence from the wild

There is convincing evidence that females and males copy the mate choice of other conspecifics in several fish species. Although most studies have been performed in the laboratory, there are an increasing number of studies demonstrating the occurrence of mate-choice copying in fish species in their natural environment. These studies emphasise the biological relevance of this fascinating mate-choice strategy and its prevalence and generality for using social information in the context of mate choice. The first experimental evidence that mate-choice copying exists in wild populations of a fish species comes from the sailfin molly P. latipinna (Witte & Ryan 2002). They performed experiments for male mate-choice copying, female mate-choice copying (Fig. 5.1a) and a control for schooling behaviour in each sex (Fig. 5.1b). In the case of female mate-choice copying, they presented focal females with two stimulus males in a river. Next to a stimulus male was a jar containing a model female or no fish for a symmetrical set-up. The two pairs of jars formed a corridor. Witte & Ryan (2002) counted the number of females swimming into this corridor and interacting with the stimulus male next to the model female or with the lone stimulus male. In the case of male mate-choice, they presented stimulus females and a model male in the same manner and counted the number of males swimming into this corridor and interacting with the stimulus females. These experiments indicated that males and females copy the choices of others, i.e. females preferred to associate with the male next to a model female

BLBK374-05

BLBK374-Brown

90

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(a) Fig. 5.1(a) Top view of the set-up of the female mate-choice test in the sailfin molly in a river. Two jars stood on two upside-down plastic tanks. Each jar had a net on top and was filled with water from the river. In the female mate-choice tests, Witte & Ryan (2002) presented stimulus males (SM) in two jars, in one jar next to a stimulus male was the model female (MF), the fourth jar had no fish (NF). Only females were counted when they entered the set-up from the side with the empty jar and the model female (indicated by the arrows) came through the ‘corridor’ and stopped within body length of the jar containing the stimulus males. For the male mate-choice test, stimulus females replaced stimulus males and a model male the model female. Only males were counted, when they entered the set-up from the side with the empty jar and the model male (indicated by the arrows) came through the ‘corridor’ and stopped at the jar within one body length of the jar with the stimulus females.

rather than the lone male, and males preferred to associate with a female next to a model male rather than next to a lone female. Neither sex showed shoaling behaviour in this experimental set-up. This field experiment is a convincing indication that mate-choice copying is a biologically relevant mate-choice strategy in sailfin mollies. Witte & Ryan (2002) provide a practicable design for mate-choice copying studies in the natural habitat of a fish species. This design was recently used and significantly advanced by Godin & Hair (2009) to investigate whether female Trinidadian guppies copy the mate choice of others in the natural riverine environment in the Quar´e River. The authors showed convincingly that female guppies copy the mate choice of others in the natural environment. To date there exist two other studies on mate-choice copying in the natural environment. Goulet & Goulet (2006) showed that females of the coral reef white belly damselfish, A. leucogaster, copy the mate choice and prefer to spawn with males that have recently mated. With an egg-switching experiment, they could exclude the hypothesis that females base their mate choice on the presence of eggs in the nest of a male. Alonzo (2008) demonstrated that female ocellated

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

91

(b) Fig. 5.1 (Continued) (b) Top view of the set-up of the female control test for shoaling in the river. In the set-up of the female control for shoaling there was a stimulus female (SF) and a jar with a model female (MF) on one tank and a stimulus female next to a jar with no fish (NF) on the other tank. Females were counted only when they entered the set-up from the side with the empty jar and the extra female (indicated by the arrows) came through the ‘corridor’ and stopped within one body length of the jar with a stimulus female. In the male control for shoaling, Witte & Ryan (2002) presented two stimulus males and an extra male.

wrasse S. ocellatus copy the mate choice of others in the wild. Female ocellated wrasses are more likely to spawn with a male when another spawning female is present than with a male that is alone. She could show that females copy the mating behaviour of other females directly. These studies give strong evidence that mate-choice copying is a biologically relevant mate-choice strategy, which can be demonstrated in the natural environment as well. The next step would be now to investigate the ecological factors in the natural environment, which facilitate or complicate the evolution of mate-choice copying in different model systems.

5.5.3

Mate-choice copying when living in sympatry or allopatry

The Amazon molly (P. formosa) is an all female gynogen hybrid species originated by natural hybridisation between the sailfin molly P. latipinna and the Atlantic molly P. mexicana. It occurs in diverse freshwater habitats from the lower Rio Grande Valley (USA) to Tuxpan in Northeast Mexico. Most of these populations live in sympatry with one of their parental species. Amazon mollies reproduce through a sperm-dependent parthenogenesis (gynogenesis). The sperm of the host species only triggers embryogenesis and do not

BLBK374-05

BLBK374-Brown

92

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

contribute any genetic information to offspring (Schlupp et al. 2007). Gabor & Ryan (2001) have demonstrated that sailfin molly males from populations sympatric with Amazon mollies have significantly stronger mating preferences for conspecific females than have males from populations allopatric with Amazon mollies. Male sailfin mollies cannot increase their fitness by mating with Amazon mollies because male genes are not transferred to offspring. However, males gain some benefits when conspecific females see them courting and copulating with heterospecific Amazon molly females. These males become more attractive to conspecific females because female sailfin mollies copy the mate choice of Amazon mollies (Schlupp et al. 1994). This study provides first evidence that it might be adaptive for males to serve as sperm donors for heterospecific females. So far this is the only species system for heterospecific mate-choice copying. Recently, Heubel et al. (2008) demonstrated that both P. latipinna and P. mexicana females and also the asexual Amazon molly females copy each other’s mate-choice decisions when they live in sympatry, but females from allopatric populations do not show heterospecific mate-choice copying. This supports the hypothesis that heterospecific mate-choice copying is beneficial for sailfin molly males in sympatric populations. The absence of heterospecific mate-choice copying in females from allopatric populations suggests that mate-choice copying is a population-specific response (Heubel et al. 2008). In sympatric populations of P. formosa and P. latipinna, males show high seasonal variation in association patterns. They spent less time with the asexual P. formosa females at times that may coincide with reproductive peaks of their own species (Heubel & Schlupp 2008). So males may gain some benefits from mating with the heterospecific P. formosa females, but prefer conspecific females in times of reproductive peaks in spring. By this strategy males maximise their own fitness and stabilise this sexual/asexual species complex.

5.5.4

Mate-choice copying – the role of the early environment

Mate-choice copying has been successfully investigated in many fish species. So far, little is known about the influence of the early environment on the decision of an observer fish to copy the mate choice of others when mature. Dugatkin (2007) showed that early environmental conditions have indeed an influence on copying behaviour later on in life. He tested whether certain social environments favour the use of social information in the context of mate choice later on in life. Groups of fifteen 1- to 2-day-old guppies were raised in individual pools for 35 days under five different ‘social environments’. After 35 days fishes were removed from pools and kept individually for 14 days. During that period they underwent sexual maturity. Females were then tested for their tendency to copy the mate choice of others in a standardised mate-choice copying experiment (Dugatkin 2007). Dugatkin showed that guppy females copy the choice of other females when they were raised without adults and when they were raised with adult males and sexually receptive females. These two conditions correspond to the natural social environment of young guppies. During their first week of life, young guppies swim together with other young guppies. Later on they join shoals with adult males and sexually receptive females (simulated by exchanging virgin females every 14 days). Dugatkin assumed that the natural developmental environment primes the guppies for using social information. His study is a first stimulation to further investigate the role of early environment on mate-choice copying

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

93

to understand developmental processes potentially involved in mate-choice copying which might explain variation in mate-choice copying between populations and species.

5.5.5

Quality of the model fish

In a typical mate-choice copying situation the observer (copier) female can potentially assess the quality of the male, the quality of the model female or both, and use this information subsequently in her mate-choice decisions (Westneat et al. 2000; Witte 2006; Witte & Godin 2010). Thus, the observer (copier) female can obtain social information about the quality of a prospective mate by observing a sexual interaction between that male and another (model) female. In this chapter, we focus on the essential role of the model female within the social information network in mate-choice copying and take into account the quality of the model, i.e. different cues of the model female like size, colour, age, experience, condition, behaviour, etc. Dugatkin & Godin (1993) have investigated the role of the model female in mate-choice copying in guppies. They presented (copier) female guppies with two males consorting with model females of different quality – either a small (young) female that was probably inexperienced in mate choice or a large (older) female that was likely experienced in mate choice. Observer females copied only the choice of larger (presumably higher quality) model females. Amlacher & Dugatkin (2005) investigated the role of the quality of the model female in guppies. They presented an observer young guppy female two males matched for colour and body size. But this time each male was together with a model female, one male with a smaller (younger) model female, and the other interacted with a larger (older) model female. Test females preferred to associate with the male they have seen together with an older female. Similarly, Vukomanovic & Rodd (2007) repeated and extended the study by Dugatkin & Godin (1993) by testing four model/copier female combinations (i.e. large/large, small/small, large/small and small/large) in the guppy. Their results corroborate those of Dugatkin & Godin (1993) and, additionally, they found that large female guppies copy the choice of large model females, but small guppy females did not copy the choice of small model females. Vukomanovic & Rodd (2007) concluded that females are more likely to copy when they perceive that there is an imbalance between their assessment ability and that of another female. Thus, females used the quality of the model female as a cue in mate-choice copying. In the sailfin molly, Hill & Ryan (2006) showed that the quality of the model female indeed matters in mate-choice copying. In their study, Hill and Ryan gave (observer) sailfin molly females (P. latipinna) a binary choice between two conspecific males as potential mates. Thereafter, females could observe the previously preferred male interacting with a heterospecific female (a low-quality model), the gynogenetic hybrid species, the Amazon molly (P. formosa). The previously non-preferred male was together with a conspecific female (a high-quality model). Subsequently, sailfin molly females significantly increased the time spent with the males they have previously seen interacting with a conspecific model female and they significantly decreased the association time with males they have seen interacting with a heterospecific model female. Thus, again the phenotypic quality of the model female influenced the mate-choice copying behaviour in females. Age and correlated experience in mate choice and the species type of the model female are only two possible qualities for model females. To get a better understanding how information and what information of the model females influences

BLBK374-05

BLBK374-Brown

94

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

mate-choice copying behaviour in the observer female, one should manipulate other modes of the model female qualities in future studies. Not only the quality of a model female influences the mate-choice copying behaviour of observing individuals, but also the number of model females has an impact on the matechoice decision. Dugatkin (1998b) showed that there is a significant stronger copying effect in female guppies when the copier female observes two model females interacting one after another with a drab coloured male for 5 minutes each. In this situation 13 of 20 females copied the choice for a drabber male, which is in contrast to their genetically based mate preference (see Section 5.6). A similar result was found in the sailfin molly. When females have a choice between a larger male and a smaller male they prefer the larger male. In mate-choice copying experiments females do not copy the choice for a smaller male if only one model female is presented next to the smaller male for 10 minutes. However, when two model females are interacting with a smaller male, each for 5 minutes, females copy the choice for the smaller male (Witte & Noltemeier 2002, see also Chapter 6). These studies emphasise the role of the model female in the context of mate-choice copying.

5.6 5.6.1

Social mate preferences overriding genetic preferences Indications from guppies

The evolution of mate preference is a complex process in which genetic and non-genetic factors are involved. Several models indicate how genetic factors influence mate choice, and we know how social cues and environment can influence the mate-choice decision. However, it is less clear how genetic and social factors interact and how this interaction can influence a female’s mate-choice decision. Two studies show how a genetically based mate preference is influenced by social learning, i.e. by mate-choice copying. Guppy females generally exhibited a genetically based preference for males with a higher amount of orange-coloured body surface (Houde 1988, 1992, 1997; Endler & Houde 1995). Guppy females also copy the mate choice of another female when both males presented in a test are matched for size and body colouration, and this is true for guppies of different populations (Dugatkin 1992, 1996a; Dugatkin & Godin 1992; Briggs et al. 1996). How do guppy females respond when they are challenged with a conflict between their genetically based mate preference and a socially based mate preference? Dugatkin (1996b) presented guppy females with this conflict in a mate-choice copying experiment. He varied the difference in male body colouration between males presented in a test. The two males presented simultaneously in a binary choice situation differed by 10%, 25% or 40% in total orange body colouration. In all cases, test females observed model females next to and interacting with the less colourful male. Afterwards, the test female was allowed to choose between the two males. When males differed in only 10% or 25% of the amount of orange body colouration, females copied the choice of the model female and preferred to associate with the paler of two males, despite a strong genetic preference for more colourful males. However, when males differed by 40% in orange body colour, test females always preferred the more colourful male, although they observed an interaction between the model female and the paler male. Thus, in this case, the genetic preferences seem to have a stronger influence on the mate-choice decision than the social cues.

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

95

In a later study, Dugatkin (1998b) further explored the interaction between genetic and social factors with regard to a preference for orange-coloured males in guppies. In this study, he presented two males simultaneously to females and these males always differed in the amount of orange by an average of 40%. In different experiments, test females could observe either no model female, one model female interacting with the drabber male for 5 minutes, two different model females interacting with the drabber male each for 5 minutes or one model female interacting with the drabber male for 10 minutes. When females observed no model female or one model female next to the drabber male, they did not copy the choice and preferred the more colourful male, thus females followed their genetic preferences. These results were consistent with the previous findings (Dugatkin 1996b). When females observed two different model females next to the drabber male, 12 of 20 females preferred the drabber male and 13 of 20 females that had observed one model female next to the drabber male for 10 minutes preferred the drabber male. Thus, in these two experiments, social cues were shown to override the genetic predisposition and had a stronger influence on mate-choice decision than genetic factors. Therefore, it seems that the amount of information a female can gain by observing the sexual interaction between a male and the model female lowers the threshold in favour of social cues having a stronger influence on the mate-choice decision than genetic factors. In these cases, social preference overrides the genetic preference in guppy females.

5.6.2

Indications from sailfin mollies

Sailfin mollies also provide evidence for an interaction between genetic and social factors influencing mate-choice decisions. Sailfin molly females show a strong preference for larger over smaller males, which had been documented in different populations of sailfin mollies (Schlupp et al. 1994; Marler & Ryan 1997; Ptacek & Travis 1997; Witte & Ryan 1998; Gabor 1999; Witte & Noltemeier 2002; MacLaren et al. 2004). Marler & Ryan (1997) provide strong evidence that the preference for larger males in sailfin molly females is genetically based. Witte & Noltemeier (2002) investigated the relative importance of genetic and social cues regarding the female preference for larger males. In a standard mate-choice copying experiment, females could first independently choose between a smaller and a larger male. In the first mate-choice test, all females preferred the larger male over the smaller male. After this independent mate choice, females had the opportunity to observe a sexual interaction between a model female and the smaller male. Afterwards, females were allowed to choose between the same larger and smaller males again. The authors varied the situation during the observing period in three experiments. In the first experiment, females could observe one model female next to the smaller male for 10 minutes. Afterwards, as predicted by previous experiments (Witte & Ryan 1998), females did not copy and still preferred the larger male over the smaller male. In a second experiment, females were allowed to observe two different females interacting with the smaller male each for 5 minutes. Thereafter, seven of 15 females reversed their initial preference for the larger of two males and spent significantly more time with the smaller male. The strongest effect was in the experiment in which females could observe one model female interacting with the smaller of two males for 20 minutes. Here, 13 of 15 females reversed their mate choice in favour of smaller males. Thus, social preference overrides the genetic preference in favour of smaller males. In several control

BLBK374-05

BLBK374-Brown

96

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

conditions where there was no opportunity to copy, Witte & Noltemeier (2002) found that females consistently preferred larger males over smaller males, while females in other control conditions exposed solely to stimulus females did not show shoaling behaviour that might explain the experimental findings. This study suggests that genetic factors interact with social cues during mate choice. Depending on the amount of social information received, females may be more influenced by their genetically determined mate preference or social cues. These experiments demonstrate the significance of social learning for mate choice and emphasise the potential of mate-choice copying to precipitate sexual selection.

5.7

Cultural evolution through mate-choice copying

Several studies have demonstrated that females change their initial mate preferences as a result of mate-choice copying. However, for mate-choice copying to be a meaningful mechanism for the cultural inheritance of mate preferences, it is necessary to show that females do not only copy the choice of a particular male, but also acquire and maintain a preference for a particular male phenotype (Brooks 1998). We now have evidence from two studies in fishes that mate-choice copying achieves these criteria for cultural inheritance of female mate preferences. Female sailfin mollies that had previously copied the mate choice of a smaller male after observing one model female interacting with the smaller of two males for 20 minutes were retested by Witte & Noltemeier (2002) for a preference for smaller males up to 36 days after copying. Females that had previously reversed their mate preference in favour of smaller males through mate-choice copying maintained this preference for smaller males in binary mate-choice tests. This was the first evidence in fishes that females copied a mate choice for a male phenotype and that females maintain a mate preference learned by mate-choice copying for a considerable period of time. These females may serve as models for other females and may induce a new mate preference in favour of smaller males within a population. Thus, the prerequisites for mate-choice copying as a mechanism for the cultural inheritance of mate preferences were fulfilled. Godin et al. (2005) have presented further evidence that guppy females copy the choice for a male phenotype and not just a choice for an individual male and that females maintained their copied mate preference. The authors showed that guppy females copied the choice of other females for less colourful males and that these females still preferred less colourful males the next day when different males were presented in a mate-choice experiment. These two studies provide good evidence that cultural transmission of mate preferences via mate-choice copying is possible in fishes even when the socially induced mate choice conflicts with the genetically based mate preference. Therefore, these studies emphasise mate-choice copying as a powerful mechanism in sexual selection.

5.8 Does mate-choice copying support the evolution of a novel male trait? How secondary sexual traits have evolved in males is one of the most fascinating questions in sexual selection. The sensory exploitation hypothesis (Ryan & Keddy-Hector 1992;

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

97

Ryan 1998) states that females have latent preferences for particular male traits before the evolutionary appearances of these traits in males. These latent mate preferences are shaped by natural selection, mostly acting in the context of foraging behaviour, through pleiotropic effects of genes expressed in both foraging and mate choice. Guppy females prefer males with a higher amount of orange colouration on the body surface (Houde 1988, 1992, 1997; Endler & Houde 1995). Rodd et al. (2002) showed that this mate preference probably originated as an innate preference for orange as a cue for rare and high quality food sources in both sexes. Therefore, males that developed orange spots exploited the preexisting preference for orange in females and are preferred as mates by females. However, in a later study Grether et al. (2005) interpreted the female preference for orange in males as a correlated effect of selection on mate preferences.

5.8.1

Theoretical approaches

Although the sensory exploitation hypothesis explains how a female preference for male traits has originated, it does not explain how a novel male trait can spread within a population. Therefore, an interesting question is whether mate-choice copying can support the spread of a novel male trait within a population. This fascinating question has been examined theoretically by Kirkpatrick & Dugatkin (1994). They assumed that female mate preferences evolve only through cultural evolution, whereas the male trait on which they act is inherited via a haploid autosomal or a Y-linked locus. In their model, they simulated two different copying situations: ‘single mate copying’, in which younger females copy the choice of only one older female, and ‘mass copying’, in which younger females have the opportunity to copy the choices of a large number of older females. Thus, copying females strengthen their mate preference towards the male type they have observed mating. As a result of frequency dependence, females in the ‘mass copying’ scenario have a stronger preference for the male type they have seen mating most frequently. On the one hand, Kirkpatrick and Dugatkin’s model shows that copying can lead to a rapid exaggeration of the male trait and female preference for it. On the other hand, copying seems to make it more difficult for a rare male trait to become established and does not maintain a polymorphism for that trait. Only under specific conditions can copying lead to two alternative evolutionary equilibria for the male trait. Female preference and the male trait can rapidly co-evolve, with a positive frequency-dependent advantage to the more common male trait allele. This is true even for a male trait that lowers male viability, when it has reached a certain threshold in frequency. Both scenarios lead to a positive frequency-dependent advantage to males: the more common a male type, the stronger is the female preference for it. This effect of frequency dependence is stronger in the ‘mass copying’ scenario than in the ‘single copying’ scenario. Thus, according to the model of (Kirkpatrick & Dugatkin 1994), matechoice copying does not favour the spread of a novel male trait within a population. Similar conclusions were reached by Laland (1994b). Agrawal (2001) has developed another model on evolutionary consequences of matechoice copying on male traits. In contrast to Kirkpatrick & Dugatkin’s model (1994), Agrawal’s model shows that mate-choice copying can cause positive or negative directional selection on male traits, or positive or negative frequency-dependent selection on male traits. Whereas Kirkpatrick & Dugatkin (1994) assume that each copying event influenced the

BLBK374-05

BLBK374-Brown

98

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

mate-choice decision of the observing female equally, Agrawal (2001) assumed in his model that different observations have differing degrees of influence on the mate-choice decision. Agrawal assumed that females are influenced by the extent to which the male-observed mating successfully differs from the population mean regarding the focal male trait. He concluded from his model that mate-choice copying can, first, facilitate the spread of a novel male type through a population, even if there is no inherent preference for the novel male trait, and second, that mate-choice copying can maintain genetic variation for sexually selected male traits. When a female observes males of different phenotypes successfully mating in proportion to their frequency in the population, her mating preferences are not altered by social cues like mate-choice copying. When a female observes a particular male phenotype mating disproportionately more often than other male phenotypes, her preference is biased towards this type of male. Thus, a female that observes a rare male type mating is more strongly biased towards this rare male phenotype than a female that observes a common male phenotype successfully mating. This assumption is based on the notion that unusual or unexpected stimuli affect individuals more strongly than common stimuli (Cohen 1984). The legitimacy of this assumption has yet to be evaluated within the context of mate-choice behaviour, although see Subsection 5.8.2. Nonetheless, the model of Agrawal (2001) provides some indications that mate-choice copying may favour the spread of a novel rare male trait within a population.

5.8.2

Experimental approaches

Is there any experimental evidence for mate-choice copying supporting or preventing the spread of a novel male trait? Sailfin molly females from the Comal River, Texas, USA, have no pre-existing preference for males with an artificial sword imitating the natural sword of male green swordtails (Witte & Klink 2005), and several previous experiments have demonstrated that sailfin molly females copy the choice of other females (Witte & Ryan 1998, 2002; Witte & Noltemeier 2002; Witte & Massmann 2003). To investigate this question, Witte (2006) attached an artificial yellow plastic sword with a black border or a transparent plastic sword to the base of the male tail fin and created video playbacks of courting males bearing the yellow sword or the transparent sword. In copying experiments, females could first choose between the two male videos presented on television monitors at each end of the female test tank. They quantified the time the female spent within a preference zone at each end of the test tank. After this first preference test, females had the opportunity to observe the male with the yellow sword courting another female on a video, whereas the male with the transparent sword was alone. In the second preference test, females were allowed to choose between the two males, one with the coloured sword and the male with the transparent sword, a second time. Fourteen of 23 females that had rejected the male with the coloured sword in the first preference test preferred that male in the second preference test after having observed the male courting another female (McNemar’s test, P = 0.001). This result seems to indicate that mate-choice copying can support the spread of a novel male trait, because females copy the choice for that novel male type. However, ten of 17 females, which had preferred the male with the coloured sword in the first preference test, changed their preference and preferred the male with the transparent sword in the second preference test (McNemar, P = 0.031). Thus, in this situation

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Learning and Mate Choice

Char Count=

99

mate-choice copying prevented the spread of a novel male trait. These experiments at least support Agrawal’s (2001) assumption that different observations have differing degrees of influence on the female mate-choice decision. The observation of an unattractive male interacting with a female seems to increase the probability that females copy the mate choice, whereas the observation of an attractive male interacting with another female decreases the probability that females copy the mate choice of others. An alternative explanation might be that fishes adopt a ‘copy when uncertain’ strategy (Laland 2004). An attractive male with familiar characteristics evokes a clear preference and leaves little uncertainty as to courtship behaviour. Conversely, a strange-looking male generates uncertainly as to whether he is an appropriate mate, so fishes look to the behaviour of others for guidance. There is good evidence for this strategy being utilised by fishes (van Bergen et al. 2004). Further experiments are necessary to estimate the evolutionary consequences of mate-choice copying for the evolution of novel traits in males.

5.9

Is mate-choice copying an adaptive mate-choice strategy?

So far we have provided strong evidence that mate-choice copying exists in several fish species and that mate-choice copying is a biologically meaningful mate-choice strategy. Unfortunately, it is still not clear what the adaptive value of this mate-choice strategy is. What are the benefits and what are the costs of this socially driven mate preference?

5.9.1

Benefits of mate-choice copying

Pruett-Jones (1992) demonstrated in a game-theoretical model that the adaptive significance of mate-choice copying depends on the ratio of costs to benefits of independent mate choice. Gibson & H¨oglund (1992) proposed two important benefits resulting from copying. Copying can serve to increase the accuracy of mate assessment and reduce the costs of mate choice. Increasing the accuracy of mate assessment through mate-choice copying (Losey et al. 1986) is especially valid for females inexperienced in mate choice. Dugatkin & Godin (1993) found in guppies that young females, which are assumed to be relatively inexperienced in mate choice, copy the choice of older, presumably more experienced, females in mate choice, but not the reverse. Inexperienced females can learn to recognise a male or male phenotype of good quality by copying the choice of experienced females. Another example of copying facilitating the learning of mate assessment is provided by the sailfin molly. Sailfin molly females copy the choice of others when both males presented in a test are similar in colour and body size. Females do not copy the choice for a smaller male when both males presented in a test differed obviously in size. In the latter case, females prefer the larger of two males, even though a model female is placed next to the smaller male (Witte & Ryan 1998). Thus, when it is difficult to distinguish between males, females are more likely to copy than in a situation when males clearly differ in quality, i.e. females are more likely to copy when they are uncertain in their mating strategy. Another benefit of mate-choice copying is that the observing female is not physically involved in courtship displays with a prospective mate. In some species, courting males behave aggressively towards females, or even harass females, during courtship displays

BLBK374-05

BLBK374-Brown

100

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(Schlupp et al. 2001; Ojanguren & Magurran 2007; Plath et al. 2007). A female that observes how a male courts another female gains information about this male, and may reject an aggressive male, without being physically involved (Witte & Ueding 2003). Mate-choice copying is assumed to reduce the time costs of searching for a mate. Females often visit several males before they choose one (Forsgren 1997). By observing others’ mate-choice decisions, copying females may save time for their own mate choice. Copying females can decrease the time spent on directly assessing potential mates by copying the mate choice of others, and thus can minimise the ‘opportunity costs’ associated with the assessment of males. Briggs et al. (1996) tested this hypothesis. They investigated whether female guppies show a higher tendency for mate-choice copying when a predator is present than with no predator around. Mate-choice copying should reduce the time for mate inspection, and thus should increase the time available for predator vigilance and, therefore, reduce the risk of predation. In both situations, with predator present and predator absent, females copied the mate choice of others, but the authors found no indication of a higher mate-choice copying tendency in females when a predator was present. Copying females may be able to reduce the time for mate assessment and, therefore, increase the time for foraging activities. Hungry females should show a higher tendency to copy the mate choice of others than satiated females. Dugatkin & Godin (1998) tested this assumption in guppy females. However, their results contradicted the expectation; only the well-fed females copied the mate choice of others significantly more often than expected by chance.

5.9.2

Costs of mate-choice copying

Mate-choice copying is also likely to entail some costs. Copying females might risk a reduced fertility as a result of sperm depletion in males when the copying female immediately copulates with that male after he has already copulated with several females. Male courtship display is a highly conspicuous behaviour not only to conspecifics but also to predators (Houde 1997); therefore, it might be risky for a copying female to mate with a male immediately after that male has courted another female and might have attracted the attention of a predator. However, both disadvantages would be reduced when copying females do not have to copulate with a particular male immediately after observing a sexual interaction between another female and that male. Witte & Massmann (2003) showed that sailfin molly females are able to memorise an observed interaction between a male and a female for at least 1 day. Thus, copying females may copy the choice of others not immediately but rather later, when the male has replenished his sperm supply and at a safer time. Godin et al. (2005) showed that guppy females copy the choice for a male phenotype 1 day after they have seen another female interacting with such a male phenotype. It is also possible that the female may acquire outdated, inappropriate or inaccurate information about mate quality through mate-choice copying. Thus, females always have to decide whether to base their mate choice on social or private information. Nordell & Valone (1998) showed that females do better when copying the choice of another female. Even if the other female has no good discrimination abilities regarding the quality of males, the copying female will not be worse than choosing independently. Due to the fact that females can assess the quality

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning and Mate Choice

101

of other females and use this information for mate-choice decision, females that copy the choice of a ‘good’ model will be on the safer side.

5.10

Outlook

Learning in mate choice is a highly dynamic research field in behavioural ecology with an enormous increase in experimental and theoretical work in recent years. Different types of learning have a specific potential as a driving force in sexual selection. Cultural evolution, i.e. inheritance of mate preferences based on and inherited through social learning provides a faster and a more dynamic potential for evolution of mate preferences than genetically determined mechanisms and can even overwrite genetically based mate preferences. In future researchers should focus on how widespread cultural evolution via social learning is and what factors facilitate or complicate cultural evolution, specifically the evolution of novel mate preferences. Sexual imprinting is an efficient mechanism to promote sympatric speciation. It may be worth investigating which specific factors facilitate the occurrence of sexual imprinting in a species and its potential role in sympatric speciation. The social environment provides additional information for ‘networkers’ using such public information about potential mates. Studies on eavesdropping, the audience effect and mate-choice copying showed that individuals use this public information, but they are also a part of the complex information network and they can be manipulated by others as well. Future studies should investigate costs and benefits for individuals taking part in the social network. The social environment plays an interesting role in forming mate-choice decisions in all stages of life. To get a better understanding on mate-choice decisions in adult fish, one should focus on the social environment individuals have experienced early in life. This aspect might help to understand variation in mate-choice decisions between populations and species. In spite of good progress in understanding mate-choice copying, there is still little indication that mate-choice copying is an adaptive mate-choice strategy. It would be valuable if future experiments focus on this question to aid understanding of the function of mate-choice copying. It is also important for future studies to investigate the relative reproductive success of copying and non-copying females. Answers to these questions would strengthen the claim that mate-choice copying plays an important role in sexual selection. Although the typical mate-choice copying situation involves only three players, the male, the model female and the focal female, these players are connected in a complex social information network. The focal female can gain information about the male via the male itself, the interaction of the male with the model female and the quality of the model female itself. The question what is the portion of each of the components that results in the mate-choice decision of the focal female is still open. Future research might focus on these interesting questions. Although mate-choice copying has been studied experimentally in many fish species, there is no clear experimental evidence that mate-choice copying increases the fitness of a copying individual. In other words, there is no experimental indication that mate-choice copying is an evolutionary adaptive strategy. In theory, there are several benefits but also some costs associated with mate-choice copying. It should be the focus for experimental studies to investigate the adaptive value of mate-choice copying in the next years.

BLBK374-05

BLBK374-Brown

May 12, 2011

102

Fish Cognition and Behavior

5.11

Conclusions

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning has an enormous influence on mate choice in fishes and is, therefore, potentially an important influence on sexual selection. Learning during an early phase of development (sexual imprinting) can shape mate preferences later on in life when the individual has reached sexual maturity. Other forms of learning, which involve experience with conspecifics, occur during all phases of life and can form and change mate preferences in adults. Social learning, which includes observing conspecifics, is arguably the most fascinating kind of learning. Individuals can gather ISI from conspecifics about the quality of potential mates and use this information for their own mate-choice decision. The evolutionary consequences for social learning by using ISI is a new expanding research field in evolutionary biology and will provide novel aspects for the intriguing role of socially induced mate preferences in sexual selection.

References Agrawal, A.F. (2001) The evolutionary consequences of mate choice copying on male traits. Behavioral Ecology and Sociobiology, 51, 927–931. Alonzo, S.H. (2008) Female mate choice copying affects sexual selection in wild populations of the ocellated wrasse. Animal Behaviour, 75, 1715–1723. Amlacher, J. & Dugatkin, L.A. (2005) Preference for older over younger models during mate-choice copying in young guppies. Ethology Ecology & Evolution, 17, 161–169. Andersson, M. (1994) Sexual Selection. Princeton University Press, Princeton, NJ. Andersson, M. & Simmons, L.W. (2006) Sexual selection and mate choice. Trends in Ecology and Evolution, 21, 296–302. Aoki, K., Feldman, M.W. & Kerr, B. (2001) Models of sexual selection on a quantitative genetic trait when preference is acquired by sexual imprinting. Evolution, 55, 25–32. Applebaum, S.L. & Cruz, A. (2000). The role of mate-copying and disruption effects in mate preference determination of Limia perugiae (Cyprinodontiformes, Poecilidae). Ethology, 106, 933–944. Bakker, T.C.M. (1999) The study of intersexual selection using quantitative genetics. Behaviour, 136, 1237–1265. Berglund, A. & Rosenqvist, G. (2001) Male pipefish prefer dominant over attractive females. Behavioral Ecology, 12, 402–406. Bonnie, K.E. & Earley, R.L. (2007) Expanding the scope for social information use. Animal Behaviour, 74, 171–181. Breden, F., Novinger, D. & Schubert, A. (1995) The effect of experience on mate choice in the Trinidadian guppy, Poecilia reticulata. Environmental Biology of Fishes, 7, 323–328. Brennan, B.J., Flaxman, S.M. & Alonzo, S.H. (2008) Female alternative reproductive behaviors: the effect of female group size on mate assessment and copying. Journal of Theoretical Biology, 253, 561– 569. Briggs, S.E., Godin, J.-G.J. & Dugatkin, L.A. (1996) Mate choice copying under predation risk in the Trinidadian guppy (Poecilia reticulata). Behavioral Ecology, 7, 151–157. Brooks, R. (1996) Copying and the repeatability of mate choice. Behavioral Ecology and Sociobiology, 39, 323–329. Brooks, R. (1998) The importance of mate copying and cultural inheritance of mating preferences. Trends in Ecology and Evolution, 13, 45–46. Brooks, R. (2000) Negative genetic correlation between male sexual attractiveness and survival. Nature, 406, 67–70.

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning and Mate Choice

103

Clutton-Brock, T. (2009) Sexual selection in females. Animal Behaviour, 77, 3–11. Cohen, J.A. (1984) Sexual selection and the psychophysics of female choice. Zeitschrift f¨ur Tierphysiologie, 64, 1–8. Dall, S.R.X., Giraldaeu, L.-A., Olsson, O., McNamara, J.M. & Stephens, D.W. (2005) Information and its use by animals in evolutionary ecology. Trends in Ecology and Evolution, 20, 187–193. Danchin, E., Giraldeau, L.-A., Valone, T.J. & Wagner, R.H. (2004) Public information: from noisy neighbors to cultural evolution. Science, 305, 487–491. Doutrelant, C. & McGregor, P.K. (2000) Eavesdropping and mate choice in female fighting fish. Behaviour, 137, 1655–1669. Doutrelant, C., McGregor, P.K. & Oliveira, R.F. (2001) The effect of an audience on intrasexual communication in male Siamese fighting fish, Betta splendens. Behavioral Ecology, 12, 283–286. Dugatkin, L.A. (1992) Sexual selection and imitation: females copy the mate choice of others. The American Naturalist, 139, 1384–1389. Dugatkin, L.A. (1996a) Copying and mate choice. In: C.M. Heyes & B.G. Galef Jr. (eds) Social Learning in Animals: The Roots of Culture, pp. 85–105. Academic Press, New York. Dugatkin, L.A. (1996b) The interface between culturally-based preferences and genetic preferences: female mate choice in Poecilia reticulata. Proceedings of the National Academy of Sciences USA, 93, 2770–2773. Dugatkin, L.A. (1998a) A comment on Lafleur et al’s re-evaluation on mate-choice copying in guppies. Animal Behaviour, 56, 513–514. Dugatkin, L.A. (1998b) Genes, copying, and female mate choice; shifting thresholds. Behavioral Ecology, 9, 323–327. Dugatkin, L.A. (2007) Developmental environment, cultural transmission, and mate choice copying. Naturwissenschaften, 94, 651–656. Dugatkin, L.A. & Godin, J.-G.J. (1992) Reversal of female mate choice by copying in the guppy (Poecilia reticulata). Proceedings of the Royal Society of London Series B, 249, 179–184. Dugatkin, L.A. & Godin, J.-G.J. (1993) Female mate copying in the guppy (Poecilia reticulata): age-dependent effects. Behavioral Ecology, 4, 289–292. Dugatkin, L.A. & Godin, J.-G.J. (1998) Effects on hunger on mate-choice copying in the guppy. Ethology, 104, 194–202. Endler, J.A. & Houde, A.E. (1995) Geographic variation in female preferences for male traits in Poecilia reticulata. Evolution, 49, 456–468. Fern¨o, A. & Sj¨olander, S. (1976) Influence of previous experience on the mate selection of two colour morphs of the convict cichlid, Cichlasoma nigrofasciatum (Pisces, Cichlidae). Behavioural Processes, 1, 3–14. Fisher, H.S. & Rosenthal, G.G. (2007) Male swordtails court with an audience in mind. Biology Letters, 3, 5–7. Fisher, R.A. (1930) The Genetical Theory of Natural Selection. Clarendon Press, Oxford. Forsgren, E. (1997) Mate sampling in a population of sand gobies. Animal Behaviour, 53, 267–276. Frommen, J.G., Rahn, A.K., Schroth, S.H., Waltschyk, N. & Bakker, T.C.M. (2009) Mate-choice copying when both sexes face high costs of reproduction. Evolutionary Ecology, 23, 435–446. Gabor, C. (1999) Association patterns of sailfin mollies (Poecilia latipinna). Alternative hypotheses. Behavioral Ecology and Sociobiology, 46, 333–340. Gabor, C.R. & Ryan, M.J. (2001) Geographic variation in reproductive character displacement in mate choice by male sailfin mollies. Proceedings of the Royal Society of London Series B, 268, 1063–1070. Gibson, R.M. & H¨oglund, J. (1992) Copying and sexual selection. Trends in Ecology and Evolution, 7, 229–232. Godin, J.-G.J. & Hair, K.P.E. (2009) Mate choice copying in free-ranging Trinidadian guppies (Poecilia reticulata). Behaviour, 146, 1443–1461. Godin, J.-G.J., Herdmann, E.J.E. & Dugatkin, L.A. (2005) Social influences on female mate choice in the guppy, Poecilia reticulata: generalized and repeatable trait-copying behaviour. Animal Behaviour, 69, 999–1005.

BLBK374-05

BLBK374-Brown

104

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Goldschmidt, T., Bakker, T.C.M. & Feuth-De Bruijn, E. (1993) Selective copying in mate choice of female sticklebacks. Animal Behaviour, 45, 541–547. Goulet, D. & Goulet, T.L. (2006) Nonindependent mating in a coral reef damselfish: evidence of mate choice copying in the wild. Behavioral Ecology, 17, 998–1003. Grant, J.W.A. & Green, L.D. (1996) Mate copying versus preference for actively courting males by female Japanese medaka (Oryzias latipes). Behavioral Ecology, 7, 165–167. Grether, G.F., Kolluru, G.R., Rodd, F.H., de la Cerda, J. & Shimazaki, K. (2005) Carotenoid availability affects the development of a colour-based mate preference and the sensory bias to which it is genetically linked. Proceedings of the Royal Society of London Series B, 272, 2181–2188. Haskins, C.P. & Haskins, E.F. (1949) The role of sexual selection as an isolating mechanism in three species of poeciliids fishes. Evolution, 3, 160–169. Haskins, C.P. & Haskins, E.F. (1950) Factors governing sexual selection as an isolating mechanism in the poeciliids fish Lebistes reticulatus. Proceedings of the National Academy of Sciences USA, 36, 464–476. Herb, B.M., Biron, S.A. & Kidd, M.R. (2003) Courtship by subordinate male Siamese fighting fish, Betta splendens: their response to eavesdropping and na¨ıve females. Behaviour, 140, 71–78. Heubel, K.U. & Schlupp, I. (2008) Seasonal plasticity in male mating preferences in sailfin mollies. Behavioral Ecology, 19, 1080–1086. Heubel, K.U., Hornhardt, K., Ollmann, T., Parzefall, J., Ryan, M.J. & Schlupp, I. (2008) Geographic variation in female mate-copying in the species complex of a unisexual fish, Poecilia formosa. Behaviour, 145, 1041–1064. Hill, S.E. & Ryan, M.J. (2006) The role of model female quality in the mate choice copying behaviour of sailfin mollies. Biology Letters, 2, 203–205. Houde, A.E. (1988) Genetic difference in female choice between two guppy populations. Animal Behaviour, 36, 510–516. Houde, A.E. (1992) Sex-linked heritability of sexually selected character in a natural population of Poecilia reticulata (Pisces: Poeciliidae) (guppies). Heredity, 69, 229–235. Houde, A.E. (1997) Sex, Color, and Mate Choice in Guppies. Princeton University Press, Princeton, NJ. Howard, R.D., Martens, R.S., Innis, S.A., Drenvich, J.M. & Hale, J. (1998) Mate choice and mate competition influence male body size in Japanese medaka. Animal Behaviour, 55, 1151–1163. Immelmann, K. (1972) Sexual and other long-term aspects of imprinting in birds and other species. Advances in the Study of Behaviour, 4, 147–174. Iwasa, Y. & Pomiankowski, A. (1999) Good parent and good genes models of handicap evolution. Journal of Theoretical Biology, 200, 97–109. Jamieson, I.G. & Colgan, P.W. (1989) Eggs in the nest of males and their effect on mate choice in the three-spined stickleback. Animal Behaviour, 38, 859–865. Jennions, M.D. & Petrie, M. (1997) Variation in mate choice and mating preference – a review of causes and consequences. Biological Reviews, 12, 283–327. Jennions, M.D., Møller, A.P. & Petrie, M. (2001) Sexually selected traits and adult survival: a metaanalysis. Quarterly Review in Biology, 76, 287–294. Jirotkul, M. (1999) Operational sex ratio influences female preferences and male–male competition in guppies. Animal Behaviour, 58, 287–294. Kirkpatrick, M. (1982) Sexual selection and the evolution of female preferences. Evolution, 36, 1–12. Kirkpatrick, M. & Dugatkin, L.A. 1994. Sexual selection and the evolutionary effects of dominance, coloration and courtship. Behavioral Ecology and Sociobiology, 34, 443–449. Knapp, R.A. & Sargent, R.C. (1989) Egg mimicry as a mating strategy in the fantail darter, Ethiostoma flabellare: females prefer males with eggs. Behavioral Ecology and Sociobiology, 25, 321– 326. Kokko, H., Brooks, R., Jennions, M.D. & Morley, J. (2003) The evolution of mate choice and mating biases. Proceedings of the Royal Society of London Series B, 270, 653–664.

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning and Mate Choice

105

Kokko, H. & Jennions, M.D. (2008) Parental investment, sexual selection and sex ratios. Journal of Evolutionary Biology, 21, 919–948. Kozak, G.M. & Boughman, J.W. (2009) Learned consepcific mate preference in a species pair of sticklebacks. Behavioral Ecology, 20, 1282–1288. Lafleur, D.L., Lozano, G.A. & Scalfini, M. (1997) Female mate-choice copying in guppies, Poecilia reticulata: a re-evaluation. Animal Behaviour, 54, 579–586. Laland, K. (1994a) On the evolutionary consequences of sexual imprinting. Evolution, 48, 477–489. Laland, K.N. (1994b) Sexual selection with a culturally transmitted mating preference. Theoretical Population Biology, 45, 1–15. Laland, K.N. (2004) Social learning strategies. Learning and Behavior, 32, 4–14. Lande, R. (1981) Models of speciation by sexual selection on polygenic traits. Proceedings of the National Academy of Sciences USA, 78, 3721–3725. Liley, N.R. (1966) Ethological isolating mechanisms in four sympatric species of poeciliid fishes. Behaviour, (Suppl. 13), 1–197. Losey, G.S. Jr., Stanton, F.G., Telecky, T.M., Tyler, W.A. III & Zoology 691 Graduate seminar class 1986 (1986) Copying others, an evolutionary stable strategy for mate choice: a model. American Naturalist, 128, 653–664. MacLaren, R.D., Rowland, W.J. & Morgan, N. (2004) Female preference for sailfin and body size in sailfin molly, Poecilia latipinna. Ethology, 110, 363–379. Magurran, A.E. & Ramnarine, I. (2004) Learned mate recognition and reproductive isolation in guppies. Animal Behaviour, 67, 1077–1082. Marconato, A. & Bisazza, A. (1986) Males whose nest contains eggs are preferred by female Gottus gobio L. (Pisces, Cottidae). Animal Behaviour, 34, 1580–1582. Marler, C.A. & Ryan, M.J. (1997) Origin and maintenance of a female mating preference. Evolution, 51, 1244–1248. Matos, R.J. & McGregor, P.K. (2002) The effect of the sex of an audience on male-male displays of Siamese fighting fish (Betta splendens). Behaviour, 139, 1211–1221. McGregor, P.K. (1993) Signalling in territorial systems: a context for individual identification, ranging and eavesdropping. Philosophical Transactions of the Royal Society B, 340, 237–244. McGregor, P.K. (2005) Animal Communication Networks. Cambridge University Press, Cambridge. McGregor, P.K. & Dabelsteen, T. (1996) Communication networks. In: D.E. Kroodsma & E.H. Miller (eds) Ecology and Evolution of Acoustic Communication in Birds, pp. 409–425. Cornell University Press, Ithaca, NY. McGregor, P.K. & Peake, T.M. (2000) Communication networks: social environments for receiving and signalling behaviour. Acta Ethologica, 2, 71–81. Møller, A.P. & Alatalo, R.V. (1999) Good-genes effects in sexual selection. Proceedings of the Royal Society of London Series B, 266, 85–91. Munger, L., Cruz, A. & Applebaum, S. (2004) Mate choice copying in female humpback limia (Limia nigrofasciata, Family Poeciliidae). Ethology, 110, 563–573. Nordell, S.E. & Valone, T.J. (1998) Mate choice copying as public information. Ecology Letters, 1, 74–76. Ojanguren, A.F. & Magurran, A.E. (2007) Male harassment reduces short-term female fitness in guppies. Behaviour, 144, 503–514. Patriquin-Meldrum, K.J. & Godin, J.-G.J. (1998) Do female three-spined sticklebacks copy the mate choice of others? American Naturalist, 151, 570–577. Peake, T.M. 2005. Eavesdropping in communication networks. In: P.K. McGregor (ed) Animal Communication Networks, pp. 13–37. Cambridge University Press, Cambridge. Plath, M., Makowicz, A.M., Schlupp, I. & Tobler, M. (2007) Sexual harassment in live-bearing fishes (Poeciliidae): comparing courting and noncourting species. Behavioral Ecology, 18, 680–688. Plath, M., Blum, D., Schlupp, I. & Tiedemann, R. (2008a) Audience effect alters mating preferences in a livebearing fish, the Atlantic molly, Poecilia mexicana. Animal Behaviour, 75, 21–29. Plath, M., Blum, D., Tiedemann, R. & Schlupp, I. (2008b) A visual audience effect in a cavefish. Behaviour, 145, 931–947.

BLBK374-05

BLBK374-Brown

106

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Plath, M., Kromuszczynski, K. & Tiedemann, R. (2009) Audience effect alters male but not female mating preferences. Behavioral Ecology and Sociobiology, 63, 381–390. Pruett-Jones, S. (1992) Independent versus non-independent mate-choice: do females copy each other? American Naturalist, 140, 1000–1009. Ptacek, M.B. & Travis, J. (1997) Mate choice in the sailfin molly, Poecilia latipinna. Evolution, 51, 1217–1231. Ridley, M. & Rechten, C. (1981) Female sticklebacks prefer to spawn with males whose nests contain eggs. Behaviour, 76, 152–161. Rodd, F.H., Hughes, K.A., Grether, G.F. & Baril, T.C. (2002) A possible non-sexual origin of mate preference: are male guppies mimicking fruit? Proceedings of the Royal Society of London Series B, 269, 467–481. Rohwer, S. (1978) Parent cannibalism of offspring and egg raiding as a courtship strategy. American Naturalist, 112, 429–440. Rosenqvist, G. & Houde, A. (1997) Prior exposure to male phenotypes influences mate choice in the guppy, Poecilia reticulata. Behavioral Ecology, 8, 194–198. Ryan, M.J. & Keddy-Hector, A. (1992) Directional patterns of female mate choice and the role of sensory biases. American Naturalist, 139, S4–35. Ryan, M.J. (1998) Sexual selection receiver bias, and the evolution of sex differences. Science, 281, 1999–2003. Schlupp, I., Riesch, R. & Tobler, M. (2007) Amazon mollies. Current Biology, 17, R536–R537. Schlupp, I. & Ryan, M.J. (1997) Male sailfin mollies (Poecilia latipinna) copy the mate choice of other males. Behavioral Ecology, 8, 104–107. Schlupp, I., Marler, C. & Ryan, M.J. (1994) Benefit to male sailfin mollies of mating with heterospecific females. Science, 263, 373–374. Schlupp, I., McKnab, R. & Ryan, M.J. (2001) Sexual harassment as a cost for molly females: bigger males cost less. Behaviour, 138, 277–286. Schluter, D. & McPhail, J.D. (1992) Ecological character displacement and speciation in sticklebacks. American Naturalist, 140, 85–108. Servedio, M.R. & Kirkpatrick, M. (1996) The evolution of mate choice copying by indirect selection. American Naturalist, 148, 848–867. Siepen, G. & Capron de Caprona, M.-D. (1986) The influence of parental color morph on mate choice in the cichlid fish Cichlasoma nigrofasciatum. Ethology, 71, 187–200. Sirot, E. (2001) Mate choice copying by females: the advantages of a prudent strategy. Journal of Evolutionary Biology, 14, 418–423. St¨ohr, S. (1998) Evolution of mate choice-copying: a dynamic model. Animal Behaviour, 55, 893–903. Uehara, T., Yokomizo, H. & Iwasa, Y. (2005) Mate choice copying as Bayesian decision making. American Naturalist, 165, 403–410. Unger, L.M. & Sargent, R.C. (1988) Alloparental care in the fathead minnow, Pimephales promelas: females prefer males with eggs. Behavioral Ecology and Sociobiology, 23, 27–32. Valone, T.J. (1989) Group foraging, public information, and patch estimation. Oikos, 56, 357–363. Valone, T.J. (2007) From eavesdropping on performance to copying the behaviour of others: a review of public information use. Behavioral Ecology and Sociobiology, 62, 1–14. Van Bergen, Y., Coolen, I. & Laland, K.N. (2004) Nine-spined sticklebacks exploit the most reliable source when public and private information conflict. Proceedings of the Royal Society of London Series B, 271, 957–962. Verzijden, M.N. & ten Cate, C. (2007) Early learning influences species assortative mating preferences in Lake Victoria cichlid fish. Biology Letters, 3, 134–136 Verzijden, M.N., Zwinkels, J. & ten Cate, C. (2009) Cross-fostering does not influence the mate preferences and territorial behaviour of males in Lake Victoria cichlids. Ethology, 115, 39–48. Vukomanovic, J. & Rodd, F.H. (2007) Size-dependent female mate copying in the guppy (Poecilia reticulata): large females are role models but small ones are not. Ethology, 113, 579–586. Wade, M.J. & Pruett-Jones, S.G. (1990) Female copying increases the variance in male mating success. Proceedings of the National Academy of Sciences USA, 87, 5749–5733.

BLBK374-05

BLBK374-Brown

May 12, 2011

13:40

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Learning and Mate Choice

107

Westneat, D.F., Walters, A., McCarthy, T.M., Hatch, M.I. & Hein, W.K. (2000) Alternative mechanisms of nonindependent mate choice. Animal Behaviour, 59, 467–476. Widemo, M.S. (2006) Male but not female pipefish copy mate choice. Behavioral Ecology, 17, 255–259. Witte, K. (2006) Mate choice in fish. In: Brown, C., Laland, K. N. & Krause, J. (eds) Fish Cognition and Behaviour, pp. 70–95. Blackwell Publishing Ltd., Oxford. Witte, K. & Caspers, B. (2006) Sexual imprinting on a novel blue ornament in zebra finches. Behaviour, 143, 969–991. Witte, K. & Godin J.-G.J. (2010) Mate choice copying and mate quality bias: are they different processes? A reply to A. Vakirtzis and S.C. Roberts. Behavioral Ecology, 21, 193–194 Witte, K. & Klink, K. (2005) No pre-existing bias in sailfin molly females (Poecilia latipinna) for a sword in males. Behaviour, 142, 283–303. Witte, K. & Massmann, R. (2003) Females remember males and copy the choice of others after one day in sailfin mollies (Poecilia latipinna). Animal Behaviour, 65, 1151–1159. Witte, K. & Noltemeier, B. (2002) The role of information in mate-choice copying in female sailfin mollies (Poecilia latipinna). Behavioral Ecology and Sociobiology, 52, 194–202. Witte, K. & Ryan, M.J. (1998) Male body length influences mate-choice copying in the sailfin molly Poecilia latipinna. Behavioral Ecology, 9, 534–539. Witte, K. & Ryan, M.J. (2002) Mate copying in the sailfin molly, Poecilia latipinna, in the wild. Animal Behaviour, 63, 943–949. Witte, K. & Ueding, K. (2003) Sailfin molly females (Poecilia latipinna) copy the rejection of a male. Behavioral Ecology, 14, 389–395. Witte, K. & Sawka, K. (2003) Sexual imprinting on a novel trait in the dimorphic zebra finch: sexes differ. Animal Behaviour, 65, 195–203. Zahavi, A. (1975) Mate selection – a selection for a handicap. Journal of Theoretical Biology, 53, 205–214. Ziege, M., Mahlow, K., Hennige-Schulz, C., Kronmarck, C., Tiedemann, R., Streit, B. & Plath, M. (2009) Audience effects in the Atlantic molly (Poecilia mexicana) – prudent male mate choice in response to perceived sperm competition risk? Frontiers in Zoology, 6, 17.

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 6

Aggressive Behaviour in Fish: Integrating Information about Contest Costs Yuying Hsu, Ryan L. Earley and Larry L. Wolf

6.1

Introduction

As the introductory chapter to the first edition of this book (Brown et al. 2006) noted, fish behaviour has long been viewed as stereotyped and not strongly influenced by context or experience. The recognition that experience and learning can influence fish behaviour coupled with observations of considerable variation in behaviour within and between individuals has led to an increasing awareness of the complexity of fish behaviour. Behavioural ecologists are interested in understanding the causes and effects of behavioural variation within individuals (e.g. over time), among individuals of a population and among populations. These causes and effects are related to the benefits and costs of the behaviour that influence the evolutionary success (fitness) of the individual. Over 40 years ago Tinbergen (1963) noted that behavioural variation could be examined from multiple perspectives. This variation could result, for example, from differences within and among individuals in genetics, age, size, sex, developmental history and morphological and physiological constraints. The variation could also be related to environmental situations encountered by an individual in the present or past and could result from the individual’s past experiences. Behavioural variation presumably results from how each of these factors influences the benefits and costs of behaviour at a particular time and place. However, a more comprehensive understanding of behavioural variation will emerge only by integrating these different perspectives. Experiences could influence variables such as size, developmental history and physiological state that are predicted or observed to lead to behavioural differences. Experiences could also influence the individual’s perception of its current environmental situation. For example, a fish that recently encountered a predator in a particular location may be less likely to engage in intense aggressive behaviour (Brick 1998, 1999) and spend more time scanning for potential predators than another fish at the same place that has never encountered a predator there. Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

109

Early ecological models of behavioural variation assumed that organisms had perfect information about the benefits and costs of behaviour in the situation being investigated (e.g. Charnov 1976). Thus, a forager supposedly knew about size distributions or abundances of prey and the costs of foraging as it selected its diet, but that is unlikely to be the case if those variables change rapidly through time (Dunlap et al. 2009). Most organisms undoubtedly do not have perfect information about any situation and some of the variation in behaviour reflects differences in individuals’ perceptions of the situation. Experience can reduce future uncertainty by modifying perceptions or expectations of benefits and/or costs in the future (Dall et al. 2005). While this variation in uncertainty applies to all types of behaviour, including feeding, antipredator strategies, mating and fighting, we concentrate in this chapter on how experience provides information about benefits and costs of future contests between individuals. In some contexts involving aggressive behaviour, information might be obtained about potential benefits. For example, how hard a resident male bowl and doily spider fights to maintain access to a female may give a male intruder information about the egg-laying status of the female (Austad 1983). Experience in prior contests presumably may provide information about potential costs of future contests. A prior contest could provide information about an individual’s fighting ability and the rate at which it would accumulate costs in a subsequent contest. This reduction in uncertainty of an individual’s own fighting ability could then influence how an individual behaves during the subsequent contest. Considerable evidence now indicates that fishes that lose a contest are more likely to lose a subsequent contest (loser effect) and fishes that win are more likely to win a subsequent contest (winner effect) (review in Hsu et al. 2006a). While an individual accumulates information about fighting ability from prior contests, the effect of that information on the outcome of a subsequent contest depends on information accumulated by the other contestant (i.e. its prior experience). The outcome of the later contest depends on information gained by all contestants and how that information influences the ongoing behaviour of each. An important area of current research is how much an individual’s behaviour in a contest is influenced by its own information and how much is influenced by the accumulated information of the opponent. These experience effects on fighting raise numerous questions that are the subject of other current research into the mechanisms and theory of information accumulation. Information that helps an organism predict benefits and costs of a situation should be retained while information that is no longer useful should be jettisoned. But when and for how long is information from a contest useful to the individual? In a highly variable environment, or when fighting ability changes quickly (e.g. high growth rates), information may become outdated very rapidly; in these conditions, we might predict that the effect of experience from a previous contest disappears quite rapidly, especially if maintaining the information has significant costs. The maintenance costs of information could depend on how the information is accumulated. Carriers of short-term information regarding social context, such as hormone titers, may carry significant costs; for instance, persistent elevations in stress hormones (e.g. adrenocorticotropins) or sex hormones (e.g. 11-ketotestosterone) can negatively impact reproduction or immune function, respectively (Kurtz et al. 2007; Alsop et al. 2009). On the other hand, learning as a long-term mechanism of information accumulation carries costs (Domjan et al. 2000) of, for instance, maintaining the new neural

BLBK374-06

BLBK374-Brown

110

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

connections, competing uses of those connections and changes in synaptic characteristics (e.g. alterations in receptor expression; Meyer et al. 2004). This chapter explores the state of research on cost-related information in animal contests with an emphasis on the role of experience in contest outcomes and points to important areas of future work.

6.2

Information about resource value

Before examining the importance of cost-related information to contest behaviour and outcomes, we first briefly discuss information about resource value. Fishes have been shown to respond to differences in reproduction or survival-related resources and adjust contest decisions accordingly (Arnott & Elwood 2008). Territorial male beaugregory damselfish (Stegastes leucostictus), for instance, were more aggressive and delivered more bites to stimulus males when an artificial breeding site was present (Snekser et al. 2009). Juvenile convict cichlids (Archocentrus nigrofasciatus) adjust their rates of aggression with resource availability, being most aggressive when supplied with intermediate levels of food (Grant et al. 2002). Resource expectations can also influence motivation to attack. Dugatkin & Ohlsen (1991) showed that, despite large size asymmetries, smaller pumpkinseed sunfish (Lepomis gibbosus) attack first and win more often when trained to expect greater food returns after presentation of a cue. An individual’s internal state can also influence its assessment of resource value; the longer the residence of replacement-owner brown trout (Salmo trutta), the more intense their contests and the higher the probability of their winning, suggesting that the motivation to defend a territory increases with residence time (Johnsson & Forser 2002). Moreover, the development of reproductive tissue in male cichlid fish, Tilapia zillii, better predicted contest outcome than body size (Neat et al. 1998a), leading the authors to suggest that differences in gonad development created asymmetries in perceived territory value and caused the males with larger testes to behave more aggressively, persist longer and win more. To strengthen this hypothesis, the authors subsequently showed that winners, losers and controls did not differ in plasma concentrations of testosterone or 11-ketotestosterone, ruling out a difference in androgen levels as the cause (Neat & Mayer 1999). Overall, fishes appear to monitor their environment closely and adjust their contest behaviour readily with changes in the actual or perceived value of the contested resource. Furthermore, their assessment of a resource’s value depends not only on its quality or quantity but also on their internal state.

6.3

Information about contest costs

In contests, animals expend energy and time, risk physical injuries and predation and forgo other opportunities (Neat et al. 1998b; Brick 1999). An individual’s potential contest cost should decrease with its fighting ability and increase with its opponent’s, since the more able contestant has the better chance of winning, resolving a contest quickly and avoiding injury (Enquist et al. 1990). As expected, contest duration, intensity and outcome vary with competitors’ fighting abilities (Enquist et al. 1990; Leiser et al. 2004; Hsu et al. 2008). These studies typically used body or weapon size to index fighting ability because

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

111

• Transitive inference • Eavesdropping • Individual recognition • Recent winning/losing experiences

• Recent winning/losing experiences

Self-fighting ability assessment

Benefit assessment

Opponent’s fighting ability assessment

Cost assessment

Contest decision Fig. 6.1 Individuals could use information from various sources to modify their cost assessment for a contest, which subsequently influences their contest decision.

of their correlation with contest success (see Table 2 in Hsu et al. 2006a). For instance, in Kryptolebias marmoratus, the probability of the larger opponent winning is positively correlated with the difference in the two opponents’ sizes while contest duration and intensity are negatively correlated with it (Hsu et al. 2008). Individuals could acquire information about fighting abilities from various sources (Fig. 6.1). An individual could adjust its contest behaviour based on an assessment of its own fighting ability (‘self-assessment’; Arnott & Elwood 2009) and/or its opponent’s (‘mutual assessment’; Arnott & Elwood 2009). Individuals could modify these assessments based on information from previous contests in which they participated or which they witnessed (Hsu et al. 2006a).

6.3.1

Assessing fighting ability

Differentiating experimentally between whether fishes use self-assessment (where animals make contest decisions based on their own fighting ability alone) or mutual assessment (where they also appraise their opponents’ strength) has remained difficult. A significant negative relationship between contest duration/intensity and size disparity was historically regarded as evidence for mutual assessment, on the assumption that the smaller animal retreated more quickly in the face of a much larger opponent (reviewed in Arnott & Elwood 2009; and for fish Enquist et al. 1990; Neat et al. 1998a). However, this relationship could arise simply because the smaller opponent’s size determines contest duration/intensity (the smaller opponent persists or retreats in a contest based on its own energy reserve without assessing the larger opponent’s ability) (Taylor & Elwood 2003). To detect mutual assessment, it is necessary, but not sufficient, (see Taylor & Elwood 2003 for discussion of ‘cumulative assessment’) to show that contest duration/intensity relates positively with the smaller opponent’s size and negatively with the larger opponent’s size. These conditions,

BLBK374-Brown

112

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Larger

26

Smaller

25 Body length (mm)

BLBK374-06

24 23 22 21 20 19 Display

Attack

Escalation

Unresolved

Fig. 6.2 The intensity of contests staged between different-sized individuals of Kryptolebias marmoratus depends on the sizes of both opponents; contests between a large larger and a small smaller opponent tend to end after bouts of mutual displays or an attack from the larger opponent while contests between a small larger opponent and a large smaller opponent are likely to progress to escalation (mutual attacks) or remain unresolved after 1 hour. (Data adopted from Hsu et al. 2008.)

although they still do not guarantee mutual assessment, provide evidence for the larger opponent influencing the contest. Following these guidelines, individuals of mangrove killifish (K. marmoratus) were found to switch assessment strategy during a contest (Hsu et al. 2008). The killifish adopt mutual assessment at earlier stages when deciding whether to escalate a contest from mutual displays to physical interactions (Fig. 6.2); contest intensity is positively related to the smaller opponent’s size and negatively to the larger opponent’s size. However, the duration of the escalated portion of the contest (involving mutual attacks) is positively related with the loser’s size but has no relation with the winner’s size. This indicates that the fishes switch to self-assessment once a contest is escalated to physical contact. The killifish has the capability to evaluate an opponent’s ability and adjust its contest behaviour accordingly but it does not (or cannot) exercise this ability throughout the entire contest. However, the assessment strategy of other fish is not as straightforward. In swordtail fish (Xiphophorus helleri), although the difference in body size was important for the likelihood of the larger fish winning, the duration of contests between similar-sized swordtails was not associated with the size of the larger or the smaller contestants (Prenter et al. 2008). Contest duration correlated negatively with the winner’s sword length but had no relationship with the loser’s. Therefore, it is not clear whether the fishes assess either their own or their opponents’ strength using either body size or sword length. In male convict cichlids (A. nigrofasciatus), contests between size-matched opponents showed that large- and smallsized pairs fought for similar durations (Leiser et al. 2004) as predicted by the mutual assessment but not the self-assessment hypothesis. However, small pairs adopted different fighting strategies and escalated to biting more quickly than large pairs, showing the importance of opponents’ absolute size (not just relative size) to contest strategy; this is not predicted by the mutual assessment hypothesis. Thus, no firm conclusion about the fish’s assessment strategy can be drawn from these results.

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

113

Based on the few studies summarised above, it appears that different species of fish may adopt different assessment strategies, but the data are not yet sufficient to gain a good sense of the diversity of assessment strategies employed by different fishes, or by the same species in different contexts or during different contest stages. One area suggested by the studies as worthy of further research is whether assessment strategies vary with underlying ecological or evolutionary factors such as social system (e.g. killifish are solitary, swordtails are highly social and cichlids are territorial).

6.3.2

Information from past contests

Contestants may derive information about their fighting ability from contest(s) in the recent past. For instance, outcomes of previous contests could give individuals information about how their fighting ability compares with the overall population (winner and loser effects; Whitehouse 1997). Contestants that have fought before could recognise each other and avoid unnecessary costs by using the outcome of their previous interaction to settle their future conflicts (individual recognition; Tibbetts & Dale 2007). Furthermore, an individual could extract information about the participants’ fighting abilities from the interaction in or outcome of contests it has previously witnessed but not taken part in (social eavesdropping; McGregor & Dabelsteen 1996). 6.3.2.1

Winner and loser effects

Many fishes adjust their fighting decisions as a result of the outcomes of their previous contests (Table 1 in Hsu et al. 2006a). After a recent win, individuals often become more aggressive (e.g. more likely to initiate contests, retaliate when provoked and persist longer before retreating) and, as a consequence, have a higher probability of winning the next contest. In contrast, after a recent loss, individuals become more passive and likely to retreat sooner when challenged. Winner and loser effects are usually hypothesised to be the result of prior winning and losing experience influencing an individual’s assessment of its own fighting ability and its estimated fighting costs in later contests (Whitehouse 1997). Winning experiences appear to raise and losing experiences to lower an individual’s perception of its fighting ability. However, not all species display experience effects and not all species whose contest decisions are influenced by past contest experiences display both winner and loser effects. The general trend is that loser effects are detected more frequently, have a stronger impact and last longer than winner effects. In fishes, winner effects have been observed in sticklebacks (Gasterosteus aculeatus; Bakker & Sevenster 1983; Bakker et al. 1989), mangrove killifish (K. marmoratus; Hsu & Wolf 1999; Hsu et al. 2009), pumpkinseed sunfish (L. gibbosus; Chase et al. 1994), Mozambique tilapia (Oreochromis mossambicus; Oliveira et al. 2009) and blue gourami (Trichogaster trichopterus; Frey & Miller 1972). Loser effects, but not winner effects, have also been detected in green sunfish (Lepomis cyanellus; McDonald et al. 1968), paradise fish (Macropodus opercularis; Francis 1983) and African cichlid (Melanochromis auratus; Chase et al. 2003). Siamese fighting fish (Betta splendens; Wallen & Wojciechowski-Metzlar 1985) and steelhead trout (Salmo gairdneri; Abbott et al. 1985) are reported to exhibit experience effects; however, the experimental procedures used (pitting prior winners against prior losers) did not allow

BLBK374-06

BLBK374-Brown

114

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

winner or loser effects to be distinguished. Several studies of green swordtail fish (X. helleri) produced mixed conclusions regarding the presence of effects (winner and/or loser effects: Franck & Ribowski 1987; Beaugrand et al. 1991, 1996; Beaugrand & Goulet 2000; no experience effects: Earley & Dugatkin 2002). The magnitude and permanence of winner and loser effects vary considerably among species (Hsu et al. 2006a). For instance, the probability of prior winners winning against size-matched, naive opponents ranges from 0.5 (no effect; paradise fish, Francis 1983) to 0.78 (pumpkinseed sunfish, Chase et al. 1994) and the effect can decay completely in an hour (pumpkinseed sunfish, Chase et al. 1994) or last more than 2 days (mangrove killifish, Hsu & Wolf 1999). The probability of prior losers winning against size-matched naive opponents varies from 0.5 (no effect; green swordtail fish, Earley & Dugatkin 2002) to 0 (sticklebacks, Bakker et al. 1989) and the effect can disappear in 24 hours (as suggested by unanalysed preliminary data, Bakker et al. 1989) or last more than 3 days (paradise fish, Francis 1983). Some part of the differences in the magnitude and permanence of experience effects between species and the asymmetrical winner and loser effects within species could result from differences in methodology (see Hsu et al. 2006a, 2006b for a more detailed discussion). These differences include the protocol for training animals with winning and losing experiences, the frequency and duration of experience training, the time interval between the completion of experience training and the subsequent contest and whether (and for how long) study animals were isolated before experience training. The procedures used to train animals can be roughly grouped into self- or random-selection. For self-selection, the winner and the loser of a size-matched fight are treated as having a winning and a losing experience, respectively, a procedure which could confound experience effect with intrinsic fighting ability (Chase et al. 1994). B´egin et al. (1996) concluded that a selfselected winner has a 0.67 probability of having higher intrinsic fighting ability than a size-matched, naive opponent (0.83 if compared with self-selected losers), as opposed to a 0.5 probability as usually assumed and tested. On the other hand, random selection gives predetermined winning or losing experiences to individuals chosen at random by pitting the study animals against smaller, habitual losers or larger, habitual winners, respectively. Studies that employ self-selection procedures and test against a null of 0.5 may, therefore, find statistically significant but bogus winner and loser effects that would not be detected by studies employing random-selection procedures. The difference in the significance of an experience effect detected from these two procedures could be further complicated by the possibility that the ‘quality’ of an experience depends on the opponent and the interaction with the opponent. A low-quality opponent that is easy to beat may give an individual less information about its fighting ability than a high-quality opponent. Thus, a win/loss against a similar-sized or a much smaller/larger trainer may have different influences on an individual’s perceived fighting ability. Because experience training can cause energy depletion, bodily injury and physical exhaustion, prolonged experience training can compromise the physical condition of trained winners and losers and cause winner and loser effects to appear weaker and stronger, respectively, than they really are. Researchers do not usually test for experience effects immediately after completion of the experience training. A long time interval between the completion of experience training and the subsequent contest provides study animals a chance to recover from the physical exhaustion/injury of

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

115

experience training. However, because experience effects decay with time (Bakker et al. 1989; Chase et al. 1994; Hsu & Wolf 1999), the length of this interval will influence the likelihood of detecting any experience effects and the magnitude of the effects detected. Finally, isolating study animals before experience training could further complicate the interpretation of the observed winner and loser effects. The purpose of this isolation is to allow focal individuals sufficient time for the effects of previous agnostic experience to dissipate. However, one must also consider the varied effects of isolation itself on agonistic behaviour (Gomez-Laplaza & Morgan 2000); depending on species and age, social isolation can lead to increases or decreases in aggressive behaviour. Asymmetrical winner and loser effects could be adaptive. Engaging in contests but losing may incur more costs (time, energy, injuries) than retreating without confrontation (Neat et al. 1998b), which could select individuals that refrain from engaging in contests after a recent loss (i.e. for a stronger loser than winner effect). However, it is also possible that winner effects are simply harder to detect experimentally because of differences in the behaviour of losers and winners when each is faced with a naive opponent (MestertonGibbons 1999). Individuals with prior losing experience often voluntarily retreat from a subsequent contest and lose to their naive opponents (e.g. Bakker & Stevenster 1983). On the other hand, contests between prior winners and naive opponents are more likely to escalate into physical fights, in which prior winners and their size-matched na¨ıve opponents are expected to have an equal chance of winning – if prior contest experience alters only an individual’s perceived but not actual fighting ability (Hsu & Wolf 1999). In this case, contest-related behaviours (e.g. escalation rate) that are sensitive to both winning and losing experiences might be more appropriate for measuring experience effects than the probability of winning. Differences among species in the importance and permanence of experience effects may reflect differences in the usefulness of prior contest information, which should depend on its reliability for use in future contests. Therefore, factors (physiology, ecology, etc.) that influence information reliability may influence the magnitude and permanence of the effect. For example, where size is important to fighting costs, slow-growing species or age groups may retain information from a prior contest longer than those that grow more quickly. For species with indeterminate growth (e.g. fish; Patnaik et al. 1994), experience effects may remain transitory for life. The frequency of social encounters might also have an effect (Schuett 1997; Hsu et al. 2006a). If outcomes of previous contests offer an individual information about how its fighting ability compares with the overall population (Whitehouse 1997), then individuals in populations with more social encounters will more frequently obtain recent and more reliable information and thus do not need to preserve information from past interactions for as long as individuals in populations with fewer social encounters. 6.3.2.1.1 Testing the behavioural mechanisms of experience effects Winner/loser effects are usually thought to be a consequence of individuals re-estimating their fighting ability after one or more winning/losing experiences and changing fighting behaviour as a result (Whitehouse 1997; Hsu et al. 2006a). However, it is possible that prior winners and losers also release status-related cues which opponents can detect and use to adjust their contest strategy. Opponent’s use of these status-related cues could enhance the

BLBK374-06

BLBK374-Brown

116

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

winner/loser effects (social-cue hypothesis, Rutte et al. 2006). Dominant and subordinate individuals of some fishes have been observed to release different chemicals or different quantities of the same chemicals into the water (Oliveira et al. 1996; Barata et al. 2007). For instance, in the presence of females that are ready to spawn, dominant Mozambique tilapia (O. mossambicus) males urinate more frequently and produce more urine than subordinates and their urine has greater olfactory potency (Barata et al. 2007, 2008). Furthermore, female tilapia’s olfactory system is more sensitive to dominants’ urine, which leads the authors to conclude that dominant tilapia males use urine odour to signal dominance to females (Barata et al. 2008). Although fishes are clearly capable of detecting and responding to chemical cues released by conspecifics, Hsu et al. (2009) did not find evidence to support the socialcue hypothesis; individuals of K. marmoratus adjusted their contest strategy based on their own recent contest experiences but not their opponents’ experiences. Although not directly tested, the social-cue mechanism probably also does not play an important role in Siamese fighting fish’s (B. splendens) contest decisions. The fish appears to exhibit experience effects such that individuals with recent winning experiences behave more aggressively and win more contests fighting against individuals with recent losing experiences (Wallen & Wojciechowski-Metzlar 1985). However, bystanders do not respond differently to prior winners and losers that they did not observe fighting (Oliveira et al. 1998), suggesting that behavioural cues or other types of cue indicative of status are either unavailable to or unused by bystanders. Thus, the winner/loser effect in Siamese fighting fish appears to operate through individuals changing contest decisions based on their own contest experiences, as in the mangrove killifish. Because this mechanism has not been examined in the other species, it is not clear whether status-related cues contribute to the winner/loser effect in any species. More studies are needed to determine this.

6.3.2.1.2 Integrating information from multiple contest experiences Only a few studies have examined the effect of multiple contest experiences on future contest behaviour (Hsu & Wolf 1999; Bekoff & Dugatkin 2000; Oyegbile & Marler 2005, 2006). Evidence so far indicates that the effects of different prior contest experiences on behaviour and physiology are cumulative but that they decay with time. In the mangrove killifish, a more recent experience (24 hours earlier) had a stronger impact on contest behaviour than an older one (48 hours earlier) (Hsu & Wolf 1999). However, we know very little about how information from different experiences is combined. For instance, are the effects from different contest experiences additive or multiplicative? If effects are multiplicative, their magnitude is a function of an individual’s prior perceived fighting ability and recent contest experiences; if additive, their magnitude is fixed (see Hsu et al. 2006a for a discussion). Another complicating factor is whether the value of an experience is influenced by other experiences (Bouton & Moody 2004; Jonides et al. 2008) as well as deteriorating with time (Bakker et al. 1989; Chase et al. 1994). The memory of an experience may be influenced by older (proactive interference) or newer (retroactive interference) experiences. These are frequently discussed in the learning literature (Bouton & Moody 2004; Jonides et al. 2008), but have not been explored in the context of animal contests. If different contest experiences interfere with each other, the combined effect will be different from the sum of the individual effects adjusted for time decay.

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

6.3.2.2

Char Count=

117

Individual recognition

The behavioural decisions of shoaling fish and the dynamics of competitive interactions in territorial species are mediated, in part, by the ability of individuals to distinguish among familiar conspecifics (Mikl´osi et al. 1995; Mikl´osi et al. 1997; Griffiths 2003). Individual recognition can reduce fighting costs if the probability of encountering the same individual on a regular basis is high (Pagel & Dawkins 1997; Tibbetts & Dale 2007). In fishes, memory of past opponents is ascertained by comparing the behavioural response of losers when encountering (a) their former dominant and (b) an unfamiliar opponent. Losers generally exhibit more pronounced behavioural and physiological (e.g. skin darkening) avoidance responses when faced with familiar opponents (Mikl´osi et al. 1995, 1997; Morris et al. 1995; Johnsson 1997; O’Connor et al. 2000; Utne-Palm & Hart 2000), suggesting that individual recognition amplifies the loser effect. Contests between familiar opponents seldom escalate (Keeley & Grant 1993; Earley et al. 2003). It is unclear whether winners fail to escalate because they recognise a former subordinate and/or because the opponent behaves submissively. Individual recognition and winner/loser effects may contribute to the formation and stability of dominance hierarchies (e.g. Dugatkin 1997; Dugatkin & Earley 2004; Hock & Huber 2009). The importance of recent contest experience on dominance ranks has been demonstrated in green swordtail fish (X. Helleri; Dugatkin & Druen 2004). When sizematched males were given winning, losing or no experience and then placed together to form a dominance hierarchy, previous winners were more likely to emerge at the top and prior losers at the bottom. Individual recognition might stabilise dominance hierarchies by reducing aggression among group members (Morris et al. 1995; Johnsson 1997; Hojesjo et al. 1998). Small hierarchies are generally more linear than large hierarchies (e.g. see Chase 1974), which suggests that where recognition of all group members is possible or where the benefits of recognition strategies exceed the costs (Pagel & Dawkins 1997) distinguishing among individual opponents stabilises the hierarchy.

6.3.2.3

Social eavesdropping

In some fish species, individuals appear to obtain a relatively accurate estimate of possible costs in future contests by social eavesdropping, the act of extracting information from contest interactions between others (Peake & McGregor 2004; Peake 2005; Bonnie & Earley 2007; Valone 2007). This might be particularly advantageous when the costs of physical combat are high (Johnstone 2001). In Siamese fighting fish (B. splendens) and green swordtail fish (X. helleri), observers appear to update their perception of the watched individuals’ fighting abilities based on the dynamics and/or outcome of the witnessed contest (Oliveira et al. 1998; McGregor et al. 2001; Earley & Dugatkin 2002; Brown & Laland 2003). An important consideration for studies on eavesdropping in fishes is whether the observer’s response is specific to the watched individuals or more general. Observing an aggressive interaction elevates urinary 11-ketotestosterone levels (O. mossambicus: Oliveira et al. 2001) and increases the aggressive behaviour of male B. splendens towards unobserved opponents (Clotfelter & Paolino 2003). It is possible that watching fights elicits behavioural

BLBK374-06

BLBK374-Brown

118

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

and physiological ‘priming’ responses (Hollis et al. 1995), which cause post-observational changes in agonistic behaviour regardless of the future opponent. However, studies on green swordtail fish indicate that eavesdroppers predictably modify their response towards individuals that they had observed to win or lose, but not towards unobserved animals (Earley & Dugatkin 2002; Earley et al. 2005). The response of swordtail bystanders towards winners and losers also indicates that fishes might be capable of rather sophisticated, indirect assessment of fighting ability. Although bystanders avoided confrontation with observed winners, their response towards observed losers depended on how long the loser persisted in the watched contest (bystanders avoid fights with persistent losers; Earley & Dugatkin 2002). This suggests that fish bystanders can cue in on both contests dynamics and outcome and modify their behaviour in future encounters accordingly. Recent simulation models have addressed whether eavesdropping can combine with winner and loser effects to promote linear dominance hierarchies, and there is some indication that they can (Dugatkin 2001). However, the verdict is uncertain, due in part to unnecessarily strict assumptions including how these experience effects accumulate (e.g. winners can increase perceptions of fighting ability without bound; losers are bound at zero; Earley & Dugatkin 2005). Transitive inference allows an animal to respond more appropriately to its social environment by combining individual experience with a particular opponent and information obtained through eavesdropping. For instance, if individual A loses to B and then witnesses C defeat B, A may avoid the costs of fighting with C and losing again. Altmann (1981) proposed that non-primate animals were incapable of transitive inference, but pinyon jays (Paz-y-Mi˜no et al. 2004), chickens (Hogue et al. 1996) and hyenas (Engh et al. 2005) show the capacity for transitive inference, or at least assessment of third-party relationships in a social context (see also Peake et al. 2002 for support in great tits). A fish species (Astatotilapia burtoni) has also been demonstrated to be capable of inferring hierarchical relationships from fights that occur around them (Grosenick et al. 2007), showing its potential to synthesise social information to guide future contest behaviour.

6.3.3

Integrating different types of cost-related information

As shown above, fishes can adjust fighting strategies based on information that facilitates assessment of their own and their opponents’ abilities, previous winning/losing experiences, the identities of the opponents and the opponents’ past performances. However, most studies examine the use of only one type of cost-related information. Therefore, it is difficult to find data on whether and how they integrate cost-related information from different sources to arrive at fighting decisions. If prior winning/losing experiences, for instance, provide an individual with information about how its fighting ability compares with the population at large, they would be of limited use to an individual that persists in a contest based solely on its own endurance or energy reserve (self-assessment: energetic war of attrition, Payne & Pagel 1996; war of attrition without assessment, Mesterton-Gibbons et al. 1996). Prior contest experiences in this case could still provide the individual useful information such as its energy consumption rate that enables it to decide how long to persist in a contest, although this is not a winner/loser effect. Previous fighting experience might also train individuals to do better in future contests regardless of the outcome of their prior contests

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

119

(e.g. Kim & Zuk 2000). Similarly, individual recognition, eavesdropping and transitive inference should be more useful to individuals that adopt the mutual assessment strategy, but have limited value to individuals that adopt the self-assessment strategy. The magnitude of winner/loser effects on contest decisions is negatively influenced by the discrepancy between the contestants’ sizes (Beacham 1988; Beaugrand et al. 1991, 1996), indicating that cost-related information from different sources competes to guide contest decisions. It is conceivable that the importance of the information acquired from witnessing contests between other individuals is affected by whether or not the observer has a recent winning or losing experience and by its body/weapon size relative to the previously witnessed opponents. The acquisition and maintenance of the information from different sources may impose different costs on a contestant, and influence which information is used and how long it is retained. For an individual to assess its opponent, for instance, requires it to recognise and respond to characteristics that give a reliable indication of an individual’s fighting ability which could divert the individual’s attention and compromise its performance in a contest. Eavesdropping and transitive inference require individual recognition, but winner/loser effects do not require such abilities and thus have a lower cognitive requirement, although they do require some physiological mechanisms for the individual to ‘remember’ the experience (Section 6.4). Activating and maintaining such mechanisms (e.g. elevated testosterone or corticosteroid levels) could involve costs to an individual (e.g. decreased immune system responses; Buchanan 2000; Casto et al. 2001). However, they might provide important cost-related information for individuals that do not live in groups and do not encounter the same competitors regularly. As mentioned in Subsection 6.3.2.3, Siamese fighting fish (B. splendens) and green swordtail fish (X. helleri) appear to update their estimate of the fighting ability of individuals they have observed (Oliveira et al. 1998; McGregor et al. 2001; Earley & Dugatkin 2002; Brown & Laland 2003). Siamese fighting fish seem to exhibit winner/loser effects (Wallen & Wojciechowski-Metzlar 1985), while evidence for the presence of experience effects in green swordtail fish is equivocal (Earley & Dugatkin 2002). Therefore, it is not obvious whether individuals that are capable of recognising individuals and extracting information from observing contests between others also display winner/loser effects. Because contestants are likely to obtain and integrate cost-related information from multiple sources, understanding how the information from difference sources interact with each other and are combined to determine contest behaviour should enable us to have a better overview of decision making in animal contests.

6.4

Physiological mechanisms

Decision making during a contest and changes in behaviour that result from information gained in prior wins and losses are modulated through an individual’s physiology. During mutual assessment, aspects of an opponent’s size, strength and persistence are filtered through the sensory machinery and interpreted through changes in neurochemical processes (e.g. brain serotonergic or dopaminergic activity; Winberg & Lepage 1998). Self-assessment probably entails some way to relay information between gauges of energy reserve (e.g. glycogen stores) or exhaustion (e.g. lactate accumulation) and the neural machinery, which

BLBK374-06

BLBK374-Brown

120

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

B

A

Contest

Δ Physiology

Δ Physiology

?

Δ Aggression or contest success

Fig. 6.3 A rich literature has demonstrated that fighting alters an array of physiology parameters (A), and that manipulations of these parameters can cause changes in aggression and/or contest success (B). However, under natural circumstances, animals can engage in multiple interactions over a relatively short time course. The black box with the question mark indicates a void in our understanding of how physiological responses to an initial contest change during an inter-contest interval (three different hypothetical patterns are illustrated), and how the dynamics of this change might impact behaviour and/or success in future contests.

dictates whether an animal persists or gives up in a contest. This basic physiological architecture for decision making in contests could also be modulated in significant ways by changes in endocrine profiles (e.g. androgens, estrogens, glucocorticoids) both during and after a fight. Indeed, post-fight increases in steroid hormones are thought to play an important role in preparing animals to engage in (or avoid) future social interaction (Oliveira 2004). This role for hormones has been supported by an abundance of studies demonstrating that administering hormones prior to a contest predictably modifies behaviour (e.g. androgens elevate aggressive responsiveness: Trainor et al. 2004). At this point, we know quite a bit about how the physiology of animals changes after a fight and how altering physiology prior to a fight impacts contest performance. Both during and after fights, an individual’s metabolism, endocrine profiles, neurochemistry and neural connectivity/ excitability are modified in significant ways. These physiological changes are a source of information about past encounters that animals can use to guide future behavioural decisions. For instance, these changes may facilitate continued exertion/aggression or orchestrate behavioural changes commonly associated with being a winner or loser (e.g. Oliveira 2009). However, we understand very little of how these physiological parameters change temporally between contests (Fig. 6.3; see Summers & Winberg 2006) and how the physiological changes that accompany an initial win or loss influence behaviour and perhaps success (i.e. winner and loser effects) during subsequent contests. To fully understand this requires time course data for both the winner and loser effects and physiological parameters that could underlie changes in behaviour that are characteristic of these effects. In the vast majority of fish species that have been studied, winner and loser effects disappear quickly (within several days; see Subsection 6.3.2.1). However, few studies have examined winner and loser effects over an extended time course (Hsu et al. 2006a) and there is some evidence that prior experiences may have a lasting, background impact on escalation and winning probabilities for up to 1 month (Earley & Hsu 2008). Experience is translated into behavioural change through alterations in physiology and it is likely that a different combination of mechanisms is responsible for short-term (hours to days) versus long-term (weeks to months) changes in aggressive motivation. We forward two very preliminary, non-exclusive hypotheses of how physiological parameters fluctuate between contest experiences, and how these fluctuations might drive the behavioural changes characteristic of winner and loser effects. Each hypothesis maintains that

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

121

contests elicit pronounced physiological changes that precipitate modifications to subsequent behaviour. The difference lies in how physiological change translates into behavioural change, and whether learning processes are involved. In the by-product hypothesis, physiological responses during and/or immediately after a contest directly elicit behaviour that increases/decreases the likelihood of winning future contests (non-learning). In the organisational hypothesis, early physiological change(s) may additionally trigger a cascade of events that reorganises neural circuits to maintain behavioural differentiation between winners and losers (learning). It will be important also to integrate these hypotheses with an understanding of the natural history and life history of the organisms being studied; growth rates, type of social system (e.g. shoaling vs. territorial), interaction frequencies in the field and reproductive seasonality could all be superimposed on to these basic models to provide a richer understanding of variation in the existence and permanence of winner–loser effects among species. Winner and loser effects could arise as a by-product of physiological deviations from baseline in response to fighting experience, and could dissipate rather quickly with subsequent recovery of baseline physiology (by-product hypothesis; Fig. 6.4a). Contestants must mobilise energy stores to maintain high activity levels during a fight, and this often is reflected as depleted glycogen reserves and increased blood glucose levels (e.g. Haller et al. 1996; Campbell et al. 2005). To fuel metabolic processes and generate energy, fighters also increase rates of oxygen consumption (Ros et al. 2006) and, if the contest escalates persistently, they turn to anaerobic metabolism. This switch to anaerobic metabolism results in elevated tissue lactate levels, typically in both contestants (Briffa & Sneddon 2007). Depletion of energy reserves and the accumulation of lactate can constrain the ability of animals to persist in a current contest (Abraham et al. 2005) and, depending on recovery times and interaction frequencies, could also limit an animal’s ability to engage successfully in future contests. At first glance, changes in metabolic physiology would seem to impact both winners’ and losers’ future performance negatively, perhaps equally so, and thus fail to explain behavioural differences characteristic of the experience effects. However, if eventual losers accrue metabolic costs at a faster rate than eventual winners (Briffa & Sneddon 2007), and if recovery times are independent of status, then losers might be expected to take longer to restore homeostasis than winners. Thus, we might expect prior losers to refrain from aggressive contests or to give up when challenged over a considerably longer time interval than prior winners, who may fully recover in time for a second interaction shortly after the first. This might explain why, in some species, loser effects exist without a corresponding winner effect, and perhaps why loser effects generally last longer (Hsu et al. 2006a). Another possibility is that eventual winners and losers accrue metabolic costs at the same rate but that some other status-dependent variable intervenes in ways that slow the recovery of losers, thereby leading to pronounced loser effects (and no winner effects). In many fish species, losers exhibit significantly higher plasma concentrations of the stress hormone cortisol than winners (e.g. Øverli et al. 1999; Hoglund et al. 2000; Sloman et al. 2001; but see Earley et al. 2006; Earley & Hsu 2008). Following physical exertion (e.g. intense contests), high levels of cortisol can impair lactate recovery and the ability of fish to replenish muscle glycogen stores (Milligan 2003). In this case, the interaction between endocrine parameters and metabolic physiology might constrain the ability of losers to engage in, or win, future contests and thus explain the prevalence of loser effects.

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

0.5

Baseline

Tissue lactate levels

Probability of winning second contest

122

Significant loser effect 24 h Contest settles 0h Contest begins

0.5

Baseline

Circulating testosterone concentrations

Significant winner effect Probability of winning second contest

BLBK374-06

24 h Contest settles 0h Contest begins (a) Fig. 6.4 Schematic diagrams representing two basic, non-independent hypotheses for mechanisms driving the winner and loser effects. (a) Two representations of the by-product hypothesis, wherein probabilities of winning a second contest (dotted curve) track changes in physiological parameters that occur as a consequence of winning or losing a first contest (solid curve). These physiological parameters include metabolic constraints on contest performance, such as tissue lactate (top panel), or hormones such as testosterone that activate behavioural responses (e.g. attack initiation) relevant to contest success (bottom panel). Note that in both representations, changes in winning probabilities are short-lived (<24 hours). Winning probabilities refer to the success of an experienced animal (prior win or loss) against a size-matched conspecific opponent encountered at some point following resolution of the initial contest (dotted line). In both panels, the slight lag between hormone/lactate elevations and behavioural change results from the time it takes hormones to bind cytosolic, nuclear, or cell surface receptors and initiate their direct (perhaps non-genomic) actions on behaviour or for lactate to exert its metabolic effects.

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

123

Δ neural connectivity, synaptic properties

Δ hormones

Δ gene expression Δ protein expression

0.5

Baseline

Significant loser effect

24 h

48 h

72 h

Physiological parameters

Probability of winning second contest

BLBK374-06

96 h

Contest settles 0h Contest begins

(b) Fig. 6.4 (Continued) (b) A representation of the organisational hypothesis where, for instance, hormone elevations that result from an initial losing experience trigger a cascade of molecular and cellular events that crystallise the effects of experience on future winning probabilities, perhaps for extended periods of time (e.g. >96 hours). The assumption here is that the different physiological parameters (e.g. hormones, gene expression, neural connectivity) are causally and sequentially linked. In this model, the probability of winning may not correlate well with any one parameter (e.g. hormone concentrations, gene expression) because changes in contest behaviour hinge on an integration of various molecular neuroendocrine reorganisations. The temporal trajectories, amplitude of the peaks and order of events shown in these figures (a and b) are not meant to be precise representations of what occurs physiologically; rather, they should be seen as a heuristic for understanding potential relationships between physiology and behaviour.

These ideas on the metabolic correlates of fighting, and their relevance to understanding winner and loser effects, produce at least two important considerations. First, if loser effects are driven largely by changes in metabolic physiology, then we might expect to see variation within species in the magnitude and/or duration of the loser effect. Individuals could retreat immediately or engage in various levels of escalation en route to a loss, and the intensity of the contest will likely impact their physiology in different ways (e.g. individuals who retreat immediately suffer little metabolic cost). Thus, we might expect variation in the magnitude of loser effects to scale in some fashion to the intensity of the initial contest, with losers of metabolically costly contests exhibiting a more pronounced experience effect than losers that incurred little cost. To date, no study has investigated the effects of initial contest dynamics (e.g. whether and for how long the eventual loser was willing to escalate) on the existence or persistence of experience effects. Second, it is often assumed that winner and

BLBK374-06

BLBK374-Brown

124

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

loser effects are driven by similar mechanisms acting in opposing fashions. As should be evident from the discussion above, this need not be the case. The accumulation of metabolic costs can explain why prior losers become behaviourally inhibited but cannot explain why prior winners often exhibit aggressive motivation that exceeds that of naive animals. Winner effects could also arise as a by-product of status-dependent changes in physiology. Fishes in positions of dominance often have higher levels of androgens (e.g. testosterone and 11-ketotestosterone) than subordinates (e.g. Neat et al. 1998; Elofsson et al. 2000), and aggressive challenges tend to stimulate androgen production (Hirschenhauser et al. 2004), although the responses vary among species and contexts. Post-fight elevations in androgen concentrations are thought to prepare individuals to respond more quickly or with greater vigour to future aggressive challenges (Oliveira 2004). This is supported by studies showing that administration of androgens enhances the expression of aggressive behaviour in fishes (Fernald 1976; Trainor et al. 2004; Remage-Healey & Bass 2006), and that individuals with higher baseline androgen levels tend to initiate attacks in a contest (Earley & Hsu 2008). Cichlids trained in a classical conditioning paradigm pairing light with territorial intrusion show an anticipatory increase in androgens when presented with light alone, suggesting a critical role for androgens in priming aggression (Antunes & Oliveira 2009). One of the most robust findings related to experience effects is the propensity of prior winners to attack quickly and therefore win subsequent contests. If aggressive motivation is increased only when circulating androgens exceed some threshold, and if the by-product hypothesis holds true, we might expect the winner effect to last only as long as androgen levels remain high. Although significant evidence points to a reciprocal relationship between androgens and contest behaviour (Oliveira 2009), no study has tracked the decay of androgens following an initial contest and mapped this decay onto changing winning probabilities. An alternative to the by-product hypothesis is that winner and loser effects result from coordinated changes in physiology at multiple organisational levels (e.g. from molecules to cells and tissues) that reorganise neural pathways in the brain in ways that promote adaptive behavioural flexibility in the face of changing social conditions (Fig. 6.4b). These changes constitute potential mechanisms of learning – persistent changes in behaviour that result from past experiences (Shettleworth 1998). Arguably, the most well-documented physiological response to fighting is a change in steroid hormone concentrations (Hsu et al. 2006a); levels of stress hormones (e.g. cortisol, corticosterone) and sex hormones (e.g. testosterone, 11-ketotestosterone, estrogens, progesterone) can be altered markedly following a contest, often in a status-dependent manner. These hormones serve purposes other than directly activating behavioural responses (as assumed by the by-product hypothesis) many of which are particularly relevant to winner and loser effects. For instance, elevated stress hormones and aggressive interaction both upregulate NMDA receptor expression in the hippocampus of Anolis lizards (Meyer et al. 2004); androgens and estrogens can also modulate NMDA receptor activity in the brain (e.g. White et al. 1999; Srivastava et al. 2008). NMDA receptors bind the neurotransmitter glutamate. The availability of NMDA receptors at a synapse, coupled with the trafficking of AMPA receptors (also glutamate-binding) to the cell surface and the morphology and number of dendritic spines, can mediate synaptic plasticity, learning and memory (Kasai et al. 2003). In addition, social interaction and steroid hormones, particularly stress hormones, are known to modulate cell proliferation in the brains of fish, which could also impact learning and memory. For instance, Dunlap et al. (2006) paired

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

125

brown ghost knife fish (Apteronotus leptorhynchus) with a conspecific or treated them with cortisol and compared neurogenesis between these animals and isolates or untreated fishes, respectively. The treated fish showed increased neurogenesis and higher densities of radial glial fibres, along which new cells migrate to their destinations, in behaviourally relevant brain areas (e.g. periventricular zone) compared to the isolates and untreated fish. Sørensen et al. (2007) also demonstrated a reduction in cell proliferation in both dominant and subordinate rainbow trout (Oncorhynchus mykiss) relative to controls, suggesting that social interaction can affect neurogenesis in diverse ways depending on social circumstance. Thus, steroid hormones may organise the brain in ways that enhance synaptic strengths or modify cellular constituents in relevant brain areas to modulate learning and memory processes that might be necessary to alter perceived fighting ability following wins and losses. NMDA receptors, AMPA receptors and dendritic spines have been characterised in fishes (e.g. mormyrid and gymnotiform electric fish: O’Brien & Unwin 2006; Harvey-Girard et al. 2007; Fortune & Chacron 2009), and there are established techniques for examining neurogenesis. However, the relevance of these factors to complex social behaviour has received little attention in fishes. Examining steroid- or status-dependent changes in the expression and/or activity of glutamate receptors, the morphology and turnover of dendritic spines or adult neurogenesis, and the time course of these changes following experience, could be a lucrative area for future study to determine mechanisms underlying winner and loser effects. Goodson (2005) proposed a cross-taxonomic vision of a ‘social behaviour network’ that includes interconnected brain regions such as the preoptic area, medial amygdala and lateral septum (fish’s ventral telencephalic nuclei) and periaqueductal gray, all of which are responsive to steroids and are implicated in the control of many forms of social behaviour. This social behaviour network could serve as an excellent starting point for isolating regionspecific, experience-induced changes in brain neurochemistry or morphology that might impact future contest decisions. Lastly, engaging in aggressive interactions could change gene expression patterns, alter the secretion of neuromodulators, or modify steroid receptor densities to make animals more or less sensitive to their endocrine milieu or prone to respond to social stimuli in an excitatory manner. In the African cichlid, A. burtoni, dominant males showed elevated androgen receptor mRNA expression in the anterior portion of the brain (which contains several regions associated with the social behaviour network) relative to subordinates (Burmeister et al. 2007). Given that steroid receptors are necessary for hormones to exert their actions on behaviour, it is possible that changes in receptor expression could result in prior winners being more sensitive to androgens than prior losers. Thus, even in the absence of marked changes in hormone production between winners and losers following a contest (e.g. Earley & Hsu 2008), differences in receptor availability could contribute to the ‘winner phenotype’ (e.g. increased attack motivation). Indeed, prior winners whose androgen receptors were blocked failed to initiate and win future contests although untreated winners still did (Oliveira et al. 2009). Fighting experience may also induce changes in the secretion of peptide neuromodulators such as corticotropin-releasing factor (CRF) and arginine vasotocin (AVT) and monoamines such as serotonin (e.g. Winberg & Lepage 1998), which can alter neuronal sensitivity, neurotransmission and future behaviour. For instance, rainbow trout administered CRF shows higher aggressive motivation during contests (e.g. reduced attack latency; Carpenter et al. 2009) and damselfish (S. leucostictus) treated with intermediate

BLBK374-06

BLBK374-Brown

126

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

doses of AVT showed significant elevations in aggressive behaviour (Santangelo & Bass 2006). The precise mechanisms driving changes in behaviour and contest outcome that characterise winner and loser effects probably represent some combination of the factors described above. Although complex interactions may occur among steroids, neuromodulators, neurotransmitters, receptors and metabolism, this should not dissuade us from investigating the mechanistic bases of winner and loser effects. These effects may arise as a by-product of physiological processes, as an adaptive response coordinated by a multitude of temporally associated endocrine and neural processes or perhaps both, depending on the species, stage of development and social and/or environmental context. Based on the current evidence, it is quite possible that early post-contest changes in physiology (e.g. hormone titres) trigger both immediate behavioural change and organisational processes that crystallise experience leading to long-term alterations in contest behaviour. It is also conceivable that long-term shifts in contest behaviour occur only after an animal has integrated many consistent experiences (e.g. several losses in a row). One could expand upon Fig. 6.4b to incorporate how individual physiology and behaviour are modulated by successive experiences, perhaps at different intervals. Exploring these possibilities will probably yield fascinating insights into the neurobiology of learning and aggression and the interconnectedness of physiological systems that drive behaviour. Comparative approaches might also provide clues into the diverse ways that selection shapes individual responses to fighting experience.

6.5

Conclusions and future directions

More information is accumulating about how experience effects influence subsequent contests. Most studies confirm that prior contest experiences, whether as a participant or an observer, have some degree of influence on the course of future contests, but also show that these effects vary in magnitude and persistence. The obvious conclusion from the accumulating data is that the relationship between experience and contest behaviour is complex and appears to vary at all levels, not only among species but also within individuals over time. The recognition of complexity in experience effects is hardly surprising, but it does focus attention on the need for more diverse studies that will permit attempts to test possible generalisations. Hypothesised generalisations are most likely to derive from carefully chosen experiments using a variety of fish species reflecting differences in such characteristics as population sizes, group sizes, growth rates and the types of resources for which individuals are competing. Comparison among these studies also will be facilitated by standardisation of methods. Currently, variation in such things as how experiences are given, when subsequent contests are staged and what variables are measured makes it difficult to compare results. Variables used to measure experimental outcomes should be carefully chosen for their value in discriminating between hypotheses. This means that it is important to define or understand what each variable measures. Items such as contest outcome and the duration of a total contest or of specific stages such as escalation are often analysed. Contest duration is controlled by the loser, so what does it tell us about the winner? And what determines when the loser will retreat? Some models assume that duration reflects depletion of the loser’s

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

127

fighting resources while others assume that the loser assesses its probability of winning before deciding to quit. The physical interaction in an escalated contest presumably gives the contestants accurate and up-to-date information about their opponents’ real fighting ability. Does this mean that their prior experience will have less influence on the outcome of escalated contests? Prior experience is only one of many possible factors influencing contest outcome. So far most investigations of experience effects have matched opponents for size, although size differences are a very significant predictor of contest outcomes. A few recent studies have also begun to examine the important question of whether fishes follow self- and/or mutual assessment in determining contest strategies. However, very little research has been carried out on how these multiple influences combine to understand how much and when each factor impacts the outcome. Presumably individuals’ ability to integrate information from multiple sources varies with the marginal value and reliability of the additional information and the cost of obtaining and storing it. Following on the work of Beaugrand et al. (1991, 1996), who integrated size difference and prior experience, we need more experiments that manipulate more than one possible source of information. This chapter has not mentioned sex or age differences as variables associated with experience effects, primarily because they have not been examined. Differences in benefits of fighting, possible hormonal correlates and even fitness costs of fighting suggest more attention should be directed to female fish. The fact that some fishes are hermaphrodite or change sex might facilitate experiments that provide significant insights into various physiological ways in which sex influences contest behaviour. Fishes of different ages may also respond differently to experience, perhaps due to variations in benefits and costs of fighting or to physiological differences. Understanding how information is integrated from various sources will depend in part on shedding light on the underlying physiological mechanisms. So too will our understanding of the persistence of experience effects and their relative importance in determining the outcome of later contests. To date, the preponderance of research on physiological mechanisms has focused on steroid hormones. As the research moves forward, it is essential that we study a greater variety of potential mechanisms, including metabolic parameters, neuromodulators, neurogenesis and patterns of gene and protein expression specific to particular brain regions. These mechanisms should ideally be examined not just individually, but in combination with each other and to see how they vary with other components of the phenotype such as morphology. Furthermore, since most such mechanisms affect a wide variety of behaviours, variations in contest behaviour among species, or among individuals in different circumstances, may have evolved because either different responses confer advantages on animals with, for example, different social systems, or because the same mechanisms are affected by selection acting on other traits, either constraining or relaxing constraints on responses to aggression (Ketterson et al. 2009). Therefore, future studies that address contests from an integrative perspective will help to shed light on individualand species-level variations in contest behaviour. Fishes generally are considered to be good subjects for contest research because they regularly fight, have been shown to respond to complex information from many different sources and are easy to work with experimentally. We still know relatively little about how information from multiple sources is assimilated, how contest behaviour varies with species’

BLBK374-06

BLBK374-Brown

128

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

ecology, about the physiological mechanisms underlying experience effects or about the relationship between these questions. Nevertheless, fishes are excellent organisms for these investigations and the field seems poised to make important advances in the next few years. Carefully focused studies should lead to a much better understanding of fighting behaviour, physiological mechanisms and contest outcomes.

Acknowledgements This research was supported in part by Taiwan National Science Council (NSC 97-2621B-003-005-MY3). We also would like to thank Mark Garcia, Shu-Ping Huang, Gordon Schuett and Alan Watson for insightful discussion.

References Abbott, J.C., Dunbrack, R.L. & Orr, C.D. (1985) The interaction of size & experience in dominance relationships of juvenile steelhead trout (Salmo gairdneri). Behaviour, 92, 241–253. Abrahams, M.V., Robb, T.L. & Hare, J.F. (2005). Effect of hypoxia on opercular displays: evidence for an honest signal? Animal Behaviour, 70, 427–432. Alsop, D., Ings, J.S. & Vijayan, M.M. (2009) Adrenocorticotropic hormone suppresses gonadotropin-stimulated estradiol release from zebrafish ovarian follicles. PLoS ONE, 4, 6463. doi:10.1371/journal.pone.0006463. Altmann, S.A. (1981) Dominance relationships: the Chesire cat’s grin? Behavioral and Brain Sciences, 4, 430–431. Antunes, R.A. & Oliveira, R.F. (2009) Hormonal anticipation of territorial challenges in cichlid fish. Proceedings of the National Academy of Sciences USA, 106, 15985–15989. Arnott, G. & Elwood, R.W. (2008) Information gathering and decision making about resource value in animal contests. Animal Behaviour, 76, 529–542. Arnott, G. & Elwood, R.W. (2009) Assessment of fighting ability in animal contests. Animal Behaviour, 77, 991–1004. Austad, S.N. (1983) A game theoretical interpretation of male combat in the bowl and doily spider, Frontinella pyramitela. Animal Behaviour, 31, 59–73. Bakker, Th.C.M., Bruijn, E. & Sevenster, P. (1989) Asymmetrical effects of prior winning and losing on dominance in sticklebacks (Gasterosteus aculeatus). Ethology, 82, 224–229. Bakker, Th.C.M. & Sevenster, P. (1983) Determinants of dominance in male sticklebacks (Gasterosteus aculeatus L.). Behaviour, 86, 55–71. Barata, E.N., Fine, J.M., Hubbard, P.C., Almeida, O.G., Frade, P., Sorensen, P.W. & Canario, A.V.M. (2008) A sterol-like odorant in the urine of Mozambique tilapia males likely signals social dominance to females. Journal of Chemical Ecology, 34, 438–449. Barata, E.N., Hubbard, P.C., Almeida, O.G., Miranda, A. & Canario, A.V.M. (2007) Male urine signals social rank in the Mozambique tilapia (Oreochromis mossambicus). BMC Biology, 5, 54. Beacham, J.L. (1988) The relative importance of body size and aggressive experience as determinants of dominance in pumpkinseed sunfish, Lepomis gibbosus. Animal Behaviour, 36, 621–623. Beaugrand, J.P. & Goulet, C. (2000) Distinguishing kinds of prior dominance and subordination experiences in males of green swordtail fish (Xiphophorus helleri). Behavioural Processes, 50, 131–142. Beaugrand, J., Goulet, C. & Payette, D. (1991) Outcome of dyadic conflict in male green swordtail fish, Xiphophorus helleri: effects of body size and prior dominance. Animal Behaviour, 41, 417–424.

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

129

Beaugrand, J.P., Payette, D. & Goulet, C. (1996) Conflict outcome in male green swordtail fish dyads (Xiphophorus helleri): interaction of body size, prior dominance/subordination experience, and prior residency. Behaviour, 133, 303–319. B´egin, J., Beaugrand, J.P. & Zayan, R. (1996) Selecting dominants and subordinates at conflict outcome can confound the effects of prior dominance or subordination experience. Behavioural Processes, 36, 219–226. Bekoff, M. & Dugatkin, L.A. (2000) Winner and loser effects and the development of dominance relationships in young coyotes: an integration of data and theory. Evolutionary Ecology Research, 2, 871–883. Bonnie, K.E. & Earley, R.L. (2007) Expanding the scope for social information use. Animal Behaviour, 74, 171–181. Bouton, M.E. & Moody, E.W. (2004) Memory processes in classical conditioning. Neuroscience and Biobehavioral Reviews, 28, 663–674. Brick, O. (1998) Fighting behaviour, vigilance and predation risk in the cichlid fish Nannacara anomala. Animal Behaviour, 56, 309–317. Brick, O. (1999) A test of the sequential assessment game: the effect of increased cost of sampling. Behavioral Ecology, 10, 726–732. Briffa, M. & Sneddon, L.U. (2007) Physiological constraints on contest behaviour. Functional Ecology, 21, 627–637. Brown, C. & Laland, K.N. (2003) Social learning in fishes: a review. Fish and Fisheries, 4, 280–288. Brown, C., Laland, K.N. & Krause, J. (2006) Fish cognition and behaviour. In: C. Brown, K. Laland & J. Krause (eds) Fish Cognition and Behavior, pp. 1–8. Blackwell Publishing Ltd., Oxford. Buchanan, K.L. (2000) Stress and the evolution of condition-dependent signals. Trends in Ecology & Evolution, 15, 156–160. Burmeister, S.S., Kailasanath, V. & Fernald, R.D. (2007) Social dominance regulates androgen and estrogen receptor gene expression. Hormones and Behavior, 51, 164–170. Campbell, H.A., Handy, R.D. & Sims, D.W. (2005) Shifts in fish’s resource holding power during a contact paired interaction: the influence of a copper-contaminated diet in rainbow trout. Physiological and Biochemical Zoology, 78, 706–714. Carpenter, R.E., Korzan, W.J., Bockholt, C., Watt, M.J., Forster, G.L., Renner, K.J. & Summers, C.H. (2009) Corticotropin releasing factor influences aggression and monoamines: modulation of attacks and retreats. Neuroscience, 158, 412–425. Casto, J.M., Nolan, V. & Ketterson, E.D. (2001) Steroid hormones and immune function: experimental studies in wild and captive dark-eyed juncos (Junco hyemalis). American Naturalist, 157, 408–420. Charnov, E.L. (1976) Optimal foraging: the marginal value theorem. Theoretical Population Biology, 9, 129–136. Chase, I.D. (1974) Models of hierarchy formation in animal societies. Behavioral Science, 19, 374–382. Chase, I.D., Bartolomeo, C. & Dugatkin, L.A. (1994) Aggressive interactions and inter-contest interval: how long do winners keep winning? Animal Behaviour, 48, 393–400. Chase, I.D., Tovey, C. & Murch, P. (2003) Two’s company, three’s a crowd: differences in dominance relationships in isolated versus socially embedded pairs of fish. Behaviour, 140, 1193–1217. Clotfelter, E.D. & Paolino, A.D. (2003) Bystanders to contests between conspecifics are primed for increased aggression in male fighting fish. Animal Behaviour, 66, 343–347. Dall, S.R.X., Giraldeau, L.A., Olsson, O., McNamara, J.M. & Stephens, D.W. (2005) Information and its use by animals in evolutionary ecology. Trends in Ecology and Evolution, 20, 187–193. Domjan, M., Cusato, B. & Villarreal, R. (2000) Pavlovian feed-forward mechanisms in the control of social behavior. Behavioral and Brain Sciences, 23, 235–282. Dugatkin, L.A. (1997) Winner effects, loser effects, assessment strategies and the structure of dominance hierarchies. Behavioral Ecology, 8, 583–587. Dugatkin, L.A. (2001) Bystander effects and the structure of dominance hierarchies. Behavioral Ecology, 12, 348–352.

BLBK374-06

BLBK374-Brown

130

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Dugatkin, L.A. & Druen, M. (2004) The social implications of winner and loser effects. Proceedings of the Royal Society of London Series B, 271 (Suppl.), S488–S489. Dugatkin, L.A. & Earley, R.L. (2004) Individual recognition, dominance hierarchies and winner and loser effects. Proceedings of the Royal Society of London, Series B – Biological Sciences, 271, 1537–1540. Dugatkin, L.A. & Ohlsen, S.R. (1991) Contrasting asymmetries in value expectation and resource holding power: effects on attack behaviour and dominance in the pumpkinseed sunfish, Lepomis gibbosus. Animal Behaviour, 39, 801–804. Dunlap, A.S., McLinn, C.M., MacCormick, H.A., Scott, M.E. & Kerr, B. (2009) Why some memories do not last a lifetime: dynamic long-term retrieval in changing environments. Behavioral Ecology, 20, 1096–1105. Dunlap, K.D., Castellano, J.F. & Prendaj, E. (2006) Social interaction and cortisol treatment increase cell addition and radial glia fiber density in the diencephalic periventricular zone of adult electric fish, Apternonotus leptorhynchus. Hormones and Behavior, 50, 10–17. Earley, R.L., Druen, M. & Dugatkin, L.A. (2005) Watching fights does not alter a bystander’s response towards na¨ıve conspecifics in male green swordtail fish, Xiphohporus helleri. Animal Behaviour, 69, 1139–1145. Earley, R.L. & Dugatkin, L.A. (2002) Eavesdropping on visual cues in green swordtail (Xiphophorus helleri) fights: a case for networking. Proceedings of the Royal Society of London Series B, 269, 943–952. Earley, R.L. & Dugatkin, L.A. (2005) Fighting, mating and networking: pillars of poeciliid sociality. In: P.K. McGregor (ed) Communication Networks, pp. 84–113. Cambridge University Press, Cambridge. Earley, R.L., Edwards, J.T., Aseem, O., Felton, K., Blumer, L.S., Karom, M. & Grober, M.S. (2006) Social interactions tune aggression and stress responsiveness in a territorial cichlid fish (Archocentrus nigrofasciatus). Physiology & Behavior, 88, 353–363. Earley, R.L. & Hsu, Y. (2008) Reciprocity between endocrine state and contest behavior in the killifish, Kryptolebias marmoratus. Hormones and Behavior, 53, 442–451. Earley, R.L., Tinsley, M. & Dugatkin, L.A. (2003) To see or not to see: does previewing a future opponent affect the contest behavior of green swordtail males (Xiphophorus helleri)? Naturwissenschaften, 90, 226–230. Elofsson, U.O.E., Mayer, I., Damsg˚ard, B. & Winberg, S. (2000) Intermale competition in sexually mature arctic charr: effects on brain monoamines, endocrine stress responses, sex hormone levels, and behavior. General and Comparative Endocrinology, 118, 450–460. Engh, A.L., Siebert, E.R., Greenberg, D.A. & Holekamp, K.E. (2005) Patterns of alliance formation and postconflict aggression indicate spotted hyaenas recognize third-party relationships. Animal Behaviour, 69, 209–217. Enquist, M., Leimar, O., Ljungberg, T., Mallner, Y. & Segerdahl, N. (1990) A test of the sequential assessment game: fighting in the cichlid fish Nannacara anomala. Animal Behaviour, 40, 1–14. Fernald, R.D. (1976) The effect of testosterone on the behavior and coloration of adult male cichlid fish (Haplochromis burtoni, Gunther). Hormone Research, 7, 172–178. Fortune, E.S. & Chacron, M.J. (2009) From molecules to behavior: organismal-level regulation of ion channel trafficking. PLoS Biology, 7, e1000211, doi:10.1371/journal.pbio.1000211. Francis, R.C. (1983) Experiential effects on agnostic behavior in the paradise fish, Macropodus opercularis. Behaviour, 85, 292–313. Franck, D. & Ribowski, A. (1987) Influences of prior agonistic experiences on aggression measures in the male swordtail (Xiphophorus helleri). Behaviour, 103, 217–240. Frey, D.F. & Miller, R.J. (1972) Establishment of dominance relationships in blue gourami, Trichogaster trichopterus (Pallas). Behaviour, 42, 8–62. Gomez-Laplaza, L.M. & Morgan, E. (2000) Laboratory studies of the effects of short-term isolation on aggressive behaviour in fish. Marine Freshwater Behavior and Physiology, 33, 63–102. Goodson, J.L. (2005) The vertebrate social behavior network: evolutionary themes and variations. Hormones and Behavior, 48, 11–22.

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

131

Grant, J.W.A., Girard, I.L., Breau, C. & Weir, L.K. (2002) Influence of food abundance on competitive aggression in juvenile convict cichlids. Animal Behaviour, 63, 323–330. Griffiths, S.W. (2003) Learned recognition of conspecifics by fishes. Fish and Fisheries, 4, 256–268. Grosenick, L., Clement, T.S. & Fernald, R.D. (2007) Fish can infer social rank by observation alone. Nature, 445, 429–432. Haller, J., Mikl´osi, A., Csanyi, V. & Makara, G.B. (1996) Behavioral tactics control the energy costs of aggression: the example of Macropodus opercularis. Aggressive Behavior, 22, 437–446. Harvey-Girard, E., Dunn, R.J. & Maler, L. (2007) Regulated expression of N-methyl-D-aspartate receptors and associated proteins in teleost electrosensory system and telencephalon. Journal of Comparative Neurology, 505, 644–668. Hirschenhauser, K., Taborsky, M., Oliveira, T., Canario, A.V.M. & Oliveira, R.F. (2004) A test of the ‘challenge hypothesis’ in cichlid fish: simulated partner and territory intruder experiments. Animal Behaviour, 68, 541–550. Hock, K. & Huber, R. (2009) Models of winner and loser effects: a cost-benefit analysis. Behaviour, 146, 69–87. H¨oglund, E., Balm, P.H.M. & Winberg, S. (2000) Skin darkening, a potential social signal in subordinate arctic charr (Salvelinus alpinus): the regulatory role of brain monoamines and proopiomelanocortin-derived peptides. Journal of Experimental Biology, 203, 1711–1721. Hogue, M.E., Beaugrand, J.P. & Lague, P.C. (1996) Coherent use of information by hens observing their former dominant defeating or being defeated by a stranger. Behavioural Processes, 38, 241–252. Hojesjo, J., Johnsson, J.I., Petersson, E. & Jarvi, T. (1998) The importance of being familiar: individual recognition and social behavior in sea trout (Salmo trutta). Behavioral Ecology, 9, 445–451. Hollis, K.L., Dumas, M.J., Singh, P. & Fackelman, P. (1995) Pavlovian conditioning of aggressive behavior in blue gourami fish (Trichogaster trichopterus): winners become winners and losers stay losers. Journal of Comparative Psychology, 109, 125–133. Hsu, Y., Earley, R.L. & Wolf, L.L. (2006a) Modulation of aggressive behaviour by fighting experience: mechanisms and contest outcomes. Biological Reviews, 81, 33–74. Hsu, Y., Earley, R.L. & Wolf, L.L. (2006b) Modulation aggression through experience. In: C. Brown, K. Laland & J. Krause (eds) Fish Cognition and Behavior, pp. 96–118. Blackwell Publishing Ltd., Oxford. Hsu, Y., Lee, S.P., Chen, M.H., Yang, S.Y. & Cheng, K.C. (2008) Switching assessment strategy during a contest: fighting in killifish Kryptolebias marmoratus. Animal Behaviour, 75, 1641–1649. Hsu, Y., Lee, I.H. & Lu, C.K. (2009) Prior contest information: mechanisms underlying winner and loser effects. Behavioral Ecology and Sociobiology, 63, 1247–1257. Hsu, Y. & Wolf, L.L. (1999) The winner and loser effect: integrating multiple experiences. Animal Behaviour, 57, 903–910. Johnsson, J.I. (1997) Individual recognition affects aggression and dominance relations in rainbow trout, Oncorhynchus mykiss. Ethology, 103, 267–282. Johnsson, J. & Forser, A. (2002) Residence duration influences the outcome of territorial conflicts in brown trout (Salmo trutta). Behavioral Ecology and Sociobiology, 51, 282–286. Johnstone, R.A. (2001) Eavesdropping and animal conflict. Proceedings of the National Academy of Science USA, 98, 9177–9180. Jonides, J., Lewis, R.L., Nee, D.E., Lustig, C.A., Berman, M.G. & Moore, K.S. (2008) The mind and brain of short-term memory. Annual Review of Psychology, 59, 193–224. Kasai, H., Matsuzaki, M., Noguchi, J., Yasumatsu, N. & Nakahara, H. (2003) Structure-stabilityfunction relationships of dendritic spines. Trends in Neurosciences, 26, 360–368. Keeley, E. & Grant, J. (1993) Visual information, resource value, and sequential assessment in convict cichlid (Cichlasoma nigrofasciatum) contests. Behavioral Ecology, 4, 345–349. Ketterson, E.D., Atwell, J.W. & McGlothlin, J.W. (2009) Phenotypic integration and independence: hormones, performance, and response to environmental change. Integrative and Comparative Biology, 49, 365–379.

BLBK374-06

BLBK374-Brown

132

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Kim, T. & Zuk, M. (2000) The effects of age and previous experience on social rank in female red junglefowl, Gallus gallus spadiceus. Animal Behaviour, 60, 239–244. Kurtz, J., Kalbe, M., Langefors, A., Mayer, I., Milinski, M. & Hasselquist, D. (2007) An experimental test of the immunocompetence handicap hypothesis in a teleost fish: 11-ketotestosterone suppresses innate immunity in three-spined sticklebacks. American Naturalist, 170, 509–519. Leiser, J.K., Gagliardi, J.L. & Itzkowitz, M. (2004) Does size matter? Assessment and fighting in small and large size-matched pairs of adult male convict cichlids. Journal of Fish Biology, 64, 1339–1350. McDonald, A.L., Heimstra, N.W. & Damkot, D.K. (1968) Social modification of agonistic behaviour in fish. Animal Behaviour, 16, 437–441. McGregor, P.K. & Dabelsteen, T. (1996) Communication networks. In: D.E. Kroodsma & E.H. Miller (eds) Ecology and Evolution of Acoustic Communication in Birds, pp. 409–425. Ithaca, NY: Cornell University Press. McGregor, P.K., Peake, T.M. & Lampe, H.M. (2001) Fighting fish Betta splendens extract relative information from apparent interactions: what happens when what you see is not what you get. Animal Behaviour, 62, 1059–1065. Mesterton-Gibbons, M., Marden, J.H. & Dugatkin, L.A. (1996) On the war of attrition without assessment. Journal of theoretical Biology, 181, 65–83. Meyer, W.N., Keifer, J., Korzan, W.J. & Summers, C.H. (2004) Social stress and corticosterone regionally upregulate limbic N-methyl-D-aspartate receptor (NR) subunit type NR2A and NR2B in the lizard Anolis carolinensis. Neuroscience, 128, 675–684. Mikl´osi, A., Haller, J. & Csanyi, V. (1995) The influence of opponent-related and outcome-related memory on repeated aggressive encounters in the paradise fish (Macropodus opercularis). Biological Bulletin, 188, 83–88. Mikl´osi, A., Haller, J. & Csanyi, V. (1997) Learning about the opponent during aggressive encounters in paradise fish (Macropodus opercularis L.): when it takes place? Behavioural Processes, 40, 97–105. Milligan, C.L. (2003) A regulatory role for cortisol in muscle glycogen metabolism in rainbow trout Oncorhynchus mykiss Walbaum. Journal of Experimental Biology, 206, 3167–3173. Morris, M.R., Gass, L. & Ryan, M.J. (1995) Assessment and individual recognition of opponents in the pygmy swordtails Xiphophorus nigrensis and X. multilineatus. Behavioral Ecology and Sociobiology, 37, 303–310. Neat, F.C., Huntingford, F.A. & Beveridge, M.M.C. (1998a) Fighting and assessment in male cichlid fish: the effects of asymmetries in gonadal state and body size. Animal Behaviour, 55, 883–891. Neat, F.C. & Mayer, I. (1999) Plasma concentrations of sex steroids and fighting in male Tilapia zillii. Journal of Fish Biology, 54, 695–697. Neat, F.C., Taylor, A.C. & Huntingford, F.A. (1998b) Proximate costs of fighting in male cichlid fish: the role of injuries and energy metabolism. Animal Behaviour, 55, 875–882. O’Brien, J. & Unwin, N. (2006) Organization of spines on the dendrites of Purkinje cells. Proceedings of the National Academy of Sciences USA, 103, 1575–1580. O’Connor, K.I., Metcalfe, N.B. & Taylor, A.C. (2000) Familiarity influences body darkening in territorial disputes between juvenile salmon. Animal Behaviour, 59, 1095–1101. Oliveira, R.F. (2004) Social modulation of androgens in vertebrates: mechanisms and function. Advances in the Study of Behavior, 34, 165–239. Oliveira, R.F. (2009) Social behavior in context: hormonal modulation of behavioral plasticity and social competence. Integrative and Comparative Biology, 49, 423–440. Oliveira, R.F., Almada, V.C. & Canario, A.V.M. (1996) Social modulation of sex steroid concentrations in the urine of male cichlid fish Oreochromis mossambicus. Hormones and Behavior, 30, 2–12. Oliveira, R.F., Lopes, M., Carneiro, L.A. & Can´ario, A.V.M. (2001) Watching fights raises fish hormone levels. Nature, 409, 475. Oliveira, R.F., Mcgregor, P.K. & Latruffe, C. (1998) Know thine enemy: fighting fish gather information from observing conspecific interactions. Proceedings of the Royal Society of London Series B, 265, 1045–1049.

BLBK374-06

BLBK374-Brown

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Aggressive Behaviour in Fish: Integrating Information about Contest Costs

Char Count=

133

Oliveira, R.F., Silva, A. & Canario, A.V.M. (2009) Why do winners keep winning? Androgen mediation of winner but not loser effects in cichlid fish. Proceedings of the Royal Society, Series B – Biological Sciences, 276, 2249–2256. Øverli, Ø., Harris, C.A. & Winberg, S. (1999) Short-term effects of fights for social dominance and the establishment of dominant-subordinate relationships on monoamines and cortisol in rainbow trout. Brain, Behavior and Evolution, 54, 263–275. Oyegbile, T.O. & Marler, C.A. (2005) Winning fights elevates testosterone levels in California mice and enhances future ability to win fights. Hormones and Behavior, 48, 259–267. Oyegbile, T.O. & Marler, C.A. (2006) Weak winner effect in a less aggressive mammal: correlations with corticosterone but not testosterone. Physiology & Behavior, 89, 171–179. Pagel, M. & Dawkins, M.S. (1997) Peck orders and group size in laying hens: ‘future contracts’ for non-aggression. Behavioural Processes, 40, 13–25. Patnaik, B.K., Mahapatro, N. & Jena, B.S. (1994) Aging in fishes. Gerontology, 40, 113–132. Payne, R.J.H. & Pagel, M. (1996) Escalation and time costs in displays of endurance. Journal of Theoretical Biology, 183, 185–193. Paz-Y-Mi˜no, C.G., Bond, A.B., Kamil, A.C. & Balda, R.P. (2004) Pinyon jays use transitive inference to predict social dominance. Nature, 430, 778–781. Peake, T.M. (2005) Eavesdropping in communication networks. In: P.K. McGregor (ed), Animal Communication Networks, pp. 13–37. Cambridge University Press, Cambridge. Peake, T.M. & McGregor, P.K. (2004) Information and aggression in fishes. Learning and Behavior, 32, 114–121. Peake, T.M., Terry, A.M.R., Mcgregor, P.K. & Dabelsteen, T. (2002) Do great tits assess rivals by combining direct experience with information gathered by eavesdropping? Proceedings of the Royal Society of London, Series B – Biological Sciences, 269, 1925–1929. Prenter, J., Taylor, P.W. & Elwood, R.W. (2008) Large body size for winning and large swords for winning quickly in swordtail males, Xiphophorus helleri. Animal Behaviour, 75, 1981–1987. Remage-Healey, L. & Bass, A.H. (2006) From social behavior to neural circuitry: steroid hormones rapidly modulate advertisement calling via a vocal pattern generator. Hormones and Behavior, 50, 432–441. Ros, A.F.H., Becker, K. & Oliveira, R.F. (2006) Aggressive behaviour and energy metabolism in a cichlid fish, Oreochromis mossambicus. Physiology & Behavior, 89, 164–170. Rutte, C., Taborsky, M. & Brinkhof, M.W.G. (2006) What sets the odds of winning and losing? Trends in Ecology & Evolution, 21, 16–21. Santangelo, N. & Bass, A.H. (2006) New insights into neuropeptide modulation of aggression: field studies of arginine vasotocin in a territorial tropical damselfish. Proceedings of the Royal Society of London Series B, 273, 3085–3092. Schuett, G.W. (1997) Body size and agonistic experience affect dominance and mating success in male copperheads. Animal Behaviour, 54, 213–224. Shettleworth, S.J. (1998) Cognition, Evolution, and Behavior. Oxford University Press, New York. Sloman, K.A., Metcalfe, N.B., Taylor, A.C. & Gilmour, K.M. (2001) Plasma cortisol concentrations before and after social stress in rainbow trout and brown trout. Physiological and Biochemical Zoology, 74, 383–389. Snekser, J.L., Leese, J., Ganim, A. & Itzkowitz, M. (2009) Caribbean damselfish with varying territory quality: correlated behaviors but not a syndrome. Behavioral Ecology, 20, 124–130. Sørensen, C., Øverli, Ø., Summers, C.H. & Nilsson, G.E. (2007) Social regulation of neurogenesis in teleosts. Brain, Behavior and Evolution, 70, 239–246. Srivastava, D.P., Woolfrey, K.M., Jones, K.A., Shum, C.Y., Lash, L.L., Swanson, G.T. & Penzes, P. (2008) Rapid enhancement of two-step wiring plasticity by estrogen and NMDA receptor activity. Proceedings of the National Academy of Sciences USA, 105, 14650–14655. Summers, C.H. & Winberg, S. (2006) Interactions between the neural regulation of stress and aggression. Journal of Experimental Biology, 209, 4581–4589. Taylor, P.W. & Elwood, R.W. (2003) The mismeasure of animal contests. Animal Behaviour, 65, 1195–1202.

BLBK374-06

BLBK374-Brown

134

May 12, 2011

13:44

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Tibbetts, E.A. & Dale, J. (2007) Individual recognition: it is good to be different. Trends in Ecology & Evolution, 22, 529–537. Tinbergen, N. (1963) On the aims and methods of ethology. Zeitschrift fur Tierpsychologie, 20, 410–433. Trainor, B.C., Bird, I.M. & Marler, C.A. (2004) Opposing hormonal mechanisms of aggression revealed through short-lived testosterone manipulations and multiple winning experiences. Hormones and Behavior, 45, 115–121. Utne-Palm, A.C. & Hart, P.J.B. (2000) The effects of familiarity on competitive interactions between threespined sticklebacks. Oikos, 91, 225–232. Valone, T.J. (2007) From eavesdropping on performance to copying the behavior of others: a review of public information use. Behavioral Ecology and Sociobiology, 62, 1–14. Wallen, K. & Wojciechowski-Metzlar, C.I. (1985) Social conditioning and dominance in male Betta splendens. Behavioural Processes, 11, 181–188. White, S.A., Livingston, F.S. & Mooney, R. (1999) Androgens modulate NMDA receptor-mediated EPSCs in the zebra finch song system. Journal of Neurophysiology, 82, 2221–2234. Whitehouse, M.E.A. (1997) Experience influences male–male contests in the spider Argyrodes antipodiana (Theridiidae: Araneae). Animal Behaviour, 53, 913–923. Winberg, S. & Lepage, O. (1998) Elevation of brain 5-HT activity, POMC expression, and plasma cortisol in socially subordinate rainbow trout. American Journal of Physiology 274 (Regulatory, Integrative and Comparative Physiology), 43, R645–R654.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 7

Personality Traits and Behaviour Sergey Budaev and Culum Brown

7.1

Introduction

Individual differences in animal behaviour have been attracting the interest of researchers at least from the time of Darwin (Slater 1981; Caro & Bateson 1986; Clark & Ehlinger 1987; Sih et al. 2004; R´eale et al. 2007). Such an interest is justified because individual differences represent the raw material of natural selection and evolution, the main cornerstone of modern biology. Furthermore, the individual is, after all, the main unit of selection (Maynard Smith 1982). Within-population variation in alternative mating strategies, foraging tactics and other observed behaviours are now widely accepted in behavioural and ecological literature. Recent investigations, however, have revealed individual differences in behavioural traits that are consistent over time and across situations. Often, such variability cannot be easily described using observable behaviour and involves inference and interpretation in terms of internal physiological or psychological mechanisms such as fearfulness or aggressiveness. Essentially, such variation represents an analogue of human personality. Some people may accept personality in ‘higher’ animals such as primates or even in dogs, but seem to deny it in ‘lower’ species (such as fish) due to the underlying fear of anthropomorphism. Ironically, this is an example of anthropocentric thinking in terms of a ‘Scala Naturae’, which has long since been discredited (Hodos & Campbell 1991). Personality traits have now been identified in a variety of animals and in fact are actively manipulated by people working closely with them (e.g. police horses, guide dogs and domestic animals generally). A metaanalysis of the available animal literature suggests that about 35% of behavioural variability of single behavioural patterns can be ascribed to individuals (Bell et al. 2009). While there is still debate about the degree to which individual differences in behaviour are consistent across different situations (see Wilson et al. 1994; Coleman & Wilson 1998; Bell 2005; Wilson & Stevens 2005; Dingemanse et al. 2007), there is no doubt that consistency of behaviour exists within many situations. Fishes have rapidly become one of the most widely studied animals with respect to personality largely because of the utility of housing and breeding them in the laboratory, but

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-07

BLBK374-Brown

136

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

also because they can be collected from a wide variety of habitats making them particularly amenable to evolutionary research (Magurran 1993; Wilson et al. 1994; Coleman & Wilson 1998; Budaev & Zworykin 2002). Substantial differences between conspecifics have been found in feeding, defensive, sexual, and other behaviours (see reviews by Ringler 1983; Magurran 1993; Budaev & Zworykin 2002). Individual fish substantially differ even within a shoal (Helfman 1984; Magurran 1993; Pitcher & Parrish 1993; Ward et al. 2004; Leblond & Reebs 2006), which has for a long time been considered the most homogeneous social structure in fishes (Radakov 1972). Even the classical example of many ethological textbooks, the stereotypic response of male three-spined sticklebacks (Gasterosteus aculeatus) to the red belly of an opponent is very pronounced in some individuals but absent in others: it is so variable that the classical concept of innate releasing mechanisms (sign stimuli) can be questioned (Rowland 1982; Baerends 1985; Bolyard & Rowland 1996). Niko Tinbergen, in his classical work ‘On the aims and methods of ethology’ (Tinbergen 1963), outlined four questions that are fundamental for our understanding of any behaviour: (1) (2) (3) (4)

Causation: What is the cause of the behaviour in question? Function: What is its survival value? Ontogeny: How does it develop? Evolution: How did it evolve?

These questions can also be asked about individual differences in behaviour as well as behaviour itself. There is one important aspect of Tinbergen’s classical paper that has largely been overlooked in modern interpretations. Tinbergen starts his seminal paper with a section entitled ‘Observation and description’, pointing to the importance of observation in tackling the unexplored world of natural behavioural patterns and the analysis of the whole landscape of behaviour. He warns against a tendency to skip this preliminary ‘inductive’ stage, which would easily result in losing touch with natural phenomena. Thus, analysis of individual behavioural patterns in isolation from one another may cause us to lose sight of a more holistic interpretation in which multiple behavioural traits become intercorrelated in various situations. Indeed, not only can an individual’s behavioural patterns and strategies have proximate and ultimate causes, but so can the correlations and relationships between them. In this chapter we review recent studies of individual differences in fish behaviour using this approach. We also provide a general methodological framework for the observation, description and analysis of fish individuality, which is based on the concept of personality. Such an approach allows the application of concepts and methods developed in human psychology, where individual differences have been the primary focus over the last 50 years. There is no need to reinvent the wheel in the animal field because human personality psychologists have solved many similar issues. The personality approach is useful because it allows to analyse generalised behavioural individuality in terms of unobservable psychological constructs, abstracting across the species and disciplines, thereby providing a single comparative and evolutionary framework that could potentially benefit behavioural ecology, evolution and personality psychology. In particular, such a general integrative approach is required if we aim to examine why personality patterns are similar (or dissimilar) across species and higher taxonomic groups.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

7.2

Char Count=

137

Observation and description of personality

Biologists are accustomed to analysing differences between populations, species and other biological entities. Here, the basic unit of analysis is character. The concept of character includes any trait that can vary between species, populations or individuals (Michener & Sokal 1957; Langlet 1971; R´eale et al. 2007). In the context of morphological and physiological variation, characters are rather easy to define and measure. In behavioural studies, however, this is often not an easy task. The behaviour of each individual depends on both its motivational state and the immediate environmental stimuli (i.e. context). Even under controlled experimental conditions it is almost impossible to create identical environment for all individuals. They often respond differently to identical stimuli due to different experience. For example, exposure to a predator behind a clear partition may be exceptionally stressful to individuals with personal experience of predation but may simply be a curiosity to predator-na¨ıve individuals (Brown & Warburton 1999). Stochastic behavioural components represent a further caveat (Cooper & Kaplan 1982; Kaplan & Cooper 1984). One of the greatest misconceptions regarding animal personalities is the fact that they are absolutely stable over time or across contexts. At the same time, however, all behavioural ecologists recognise that behaviour is highly plastic and animals frequently adjust their behaviour to suit the prevailing conditions. How can these two concepts be reconciled? The possibility of stable characteristics of personality in a constantly changing behaviour first appeared in psychology at the beginning of the twentieth century. While many researchers were happy with the concept of stable personality traits, it also attracted substantial criticism. Among the most influential critics, Mischel (1973) argued that personality does not really exist, suggesting that human behaviour is flexible. This personality-flexibility debate has largely been resolved over the last 40 years (Kenrick & Funder 1988; Fleeson 2004; Funder 2009). It is now accepted that behavioural plasticity and personality traits are not mutually exclusive, rather both are important in shaping human behaviour. Human behaviour displays enormous flexibility and personality cannot predict every isolated behavioural act or decision; nonetheless, stable personality traits really do describe and predict trends, typical ways of acting, and behaviour over longer periods of time (Fleeson 2004). This general approach of inferring stable individual characteristics from a highly flexible behaviour can be applied to the study of non-human animal behaviour. Moreover, the concepts and techniques developed by human personality psychologists over a long period provide an ideal methodology for the description of the overall general landscape of animal individuality (see Gosling 2001; Budaev & Zworykin 2002; R´eale et al. 2007; Vazire et al. 2007).

7.2.1

Current terminology

If the basic model describing human personality variation can be applied to animal individuality, what hinders us from using the term personality? Personality, conceived as a broad domain of behavioural individuality involving the widest range of consistent and enduring behavioural traits can be legitimately applied to a wide range of species. It does not necessarily involve emotions or advanced cognitive ability. Theoretically, personality can even be applied to bacteria.

BLBK374-07

BLBK374-Brown

138

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Apart from applied research (Seaman et al. 2002; Svartberg 2002), application of the term personality to animals has been hampered by a widespread fear of anthropomorphism. Such a fear is largely unfounded, however, if animal personality is defined in descriptive, functional and motivational terms. Furthermore, when studying complex behaviour, some degree of frank anthropomorphism is inevitable (Dennet 1983). The best classical example is provided by Hebb (1946). When various behaviours were objectively recorded in chimpanzees, the resulting long list was virtually futile in predicting their behaviour: ‘All that resulted was an almost endless series of specific acts in which no order or meaning could be found.’ In contrast, more subjective anthropomorphic descriptions like ‘aggressive’ provided ‘an intelligible and practical guide to behavior’, which could be efficiently used even by persons inexperienced with the animals. In a similar vein, Konrad Lorenz, in his Nobel lecture, writes: ‘When we speak of falling in love, of friendship, personal enmity, or jealousy in these or other animals, we are not guilty of anthropomorphism. These terms refer to functionally determined concepts . . .’ (Lorenz 1974). To escape accusations in anthropomorphism, researchers tried to avoid personality by using a variety of presumably more ‘objective’ constructs like shyness–boldness (Wilson et al. 1994), behavioural syndrome (Sih et al. 2004), behavioural profile (Budaev et al. 1999a) or temperament (Francis 1990; R´eale et al. 2007) and coping style (Huntingford et al. 2010). This had another unfortunate consequence, namely that the literature on individual differences in animal behaviour has quickly become fragmented. It is necessary, therefore, to create a framework which reunites the various concepts adopted. Here we briefly summarise the terminology commonly used in the animal personality literature. 7.2.1.1

Shyness–boldness

A variety of related concepts have been used to describe individual differences in behaviour that are consistent over time and across situations. Wilson et al. (1994) proposed that the shy–bold continuum – the propensity to take risks – is a fundamental axis of behavioural variation in various species. The concept of boldness has been frequently applied to fishes. For example, Wilson et al. (1993) used it to describe individual differences in risk taking in the pumpkinseed sunfish, Lepomis gibbosus. In this study, the shyness–boldness trait was measured as a propensity to approach a novel object such as a minnow trap and a measuring stick. The position of individuals on the shy–bold continuum was consistent, predicting diet, acclimation to the laboratory, habitat utilisation and parasite fauna. The shyness–boldness continuum has been used in many subsequent studies. The tests and experiments used to measure boldness also varied substantially (Table 7.1). For example, researchers used empty novel environments (open field; higher locomotion indicative of boldness), novel objects, predator inspection (approach to predator or a novel object involves boldness), foraging in presence of a predator, latency to emerge into a novel environment from cover, time spent in open habitats and so on. In many studies, fishes behaved consistently when tested repeatedly over time and across situations (e.g. Huntingford 1976; Brick & Jakobsson 2002; Ward et al. 2004; Brown et al. 2007a; Wilson & Godin 2009), although this was not always the case (Coleman & Wilson 1998; Wilson & Stevens 2005; Dingemanse et al. 2007).

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Personality Traits and Behaviour

139

Table 7.1 A list of ‘boldness’ measures in fishes utilised by a range of authors. Reference

Measure

Species

Brown et al. (2007a)

Novel object inspection

Brachyraphis episcopi

Brown et al. (2007a, 2007b)

Open field

B. episcopi

Brown & Braithwaite (2004) and Brown et al. (2005a)

Latency to emerge from cover

B. episcopi

Bell & Stamps (2004)

Open field

Gasterosteus aculeatus

Bell & Stamps (2004) and Bell (2005)

Foraging under predation risk

G. aculeatus

Azuma et al. (2005)

Recovery from fright

Oncorhynchus mykiss

Brick & Jakobsson (2002)

Tendency to inspect mirror image

Nannacara anomala

Budaev (1997a)

Open field

Symphodus ocellatus

Budaev (1997b)

Open field

Poecilia reticulata

Budaev et al. (1999a)

Tendency to inspect novel fish

Steatocranus casaurius

Budaev et al. (1999a)

Open field

S. casaurius

Budaev et al. (1999b)

Open field

Cichlasoma nigrofasciatum

Budaev et al. (1999b)

Tendency to inspect novel fish

C. nigrofasciatum

Coleman & Wilson (1998)

Response to threatening stimuli

Lepomis gibbosus

Coleman & Wilson (1998)

Response to novel food source

L. gibbosus

Dugatkin & Alfieri (2003)

Predator inspection

P. reticulata

Dugatkin et al. (2005)

Predator inspection

Danio rerio

Fraser et al. (2001)

Tendency to cross open habitat

Rivulus hartii

Godin & Davis (1995)

Predator inspection

P. reticulata

Godin & Dugatkin (1996)

Predator inspection

P. reticulata

Huntingford (1976)

Response to predatory attack

G. aculeatus

Johnsson et al. (2001)

Response to predatory attack

Salmo trutta

Magnhagen & Staffan (2005) and Magnhagen (2006)

Foraging under predation threat

Perca fluviatilis

Schjolden et al. (2005)

Response to novel object

Oncorhynchus mykiss

Shaklee (1963)

Response to predators

Multiple species

Sneddon (2003)

Time spent in the open habitat

O. mykiss

Staffan et al. (2005)

Time spent in the open habitat

P. fluvitilis

Sundstrom et al. (2004)

Response to novel object

S. trutta

Ward et al. (2004)

Foraging under predation risk

G. aculeatus

Westerberg et al. (2004)

Time spent in the open habitat

P. fluvitilis

Wilson & Stevens (2005)

Latency to forage, pass through a net, feed under predation threat and open field

O. mykiss

Wilson et al. (1993)

Inspection of novel object; Open field

L. gibbosus

Wright et al. (2003, 2006)

Inspection of novel object

D. rerio

Yoshida et al. (2005)

Open field

L. macrochirus, Carassius langsdorfii, C. auratus

BLBK374-07

BLBK374-Brown

140

7.2.1.2

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Coping styles

Another concept frequently used to investigate individual differences in fish behaviour is coping styles or strategies, which often represent bimodal clusters of individuals with a number of similar behavioural traits rather than continuously distributed traits or dimensions (Budaev 1997a; Brelin et al. 2005; Øverli et al. 2007). Two alternative coping styles are frequently distinguished: proactive and reactive (Benus et al. 1991; Koolhaas et al. 1999; Øverli et al. 2007). Proactive individuals are more active, aggressive, bold, tend to form inflexible routines and hence learn more slowly about small changes in the environment. When presented with novel stimuli, they explore them quickly and superficially. Reactive individuals, in contrast, are shyer, non-aggressive, more sensitive to environmental changes, explore novel stimuli slowly and thoroughly and tend to adapt to the situational demands.

7.2.1.3

Behavioural syndromes

The third concept frequently implicated in the study of animal personality is behavioural syndrome: a suite of correlated behaviours that are expressed either within a given context or across contexts (e.g. correlations between activity levels, boldness and aggression in foraging and antipredator contexts) (Sih et al. 2004). Sih et al. pointed to a few behavioural syndromes that may be of particular importance: the aggression syndrome, activity syndrome, boldness, fearfulness and reactivity. In this approach, correlations between different contexts and across different types of behaviour are most interesting because they could generate trade-offs between contexts or behavioural traits and thereby may play an important role in the evolution of behaviour. The primary value of the syndrome approach, therefore, is that it recognises that various behavioural traits may be correlated, potentially providing constraints on behavioural flexibility. The approach also helps explain why some behavioural traits appear maladaptive in some contexts. For example, a highly aggressive individual may be a very successful forager, but may incidentally attack potential mates. When considering mating behaviour in isolation, a high level of aggressiveness may seem to be maladaptive.

7.2.2

Objectivity

A further problem with previous research on animal personality is that instead of carefully exploring the whole landscape of behavioural individuality, many researchers start by concentrating on a limited set of specific behavioural patterns, domains of situations or behaviours. Often, to gain more objectivity, the researcher provides a very specific (and narrow) definition for the individual trait under the study and then proceeds in developing methods to measure it. While there is nothing wrong with deductive hypothesis-led research, hasting from the first descriptive step is a potentially dangerous deviation from the ethological paradigm, which historically led certain areas of psychology to lose touch with the real phenomena due to loss of context (Tinbergen 1963). Such a danger can be illustrated by analysing boldness. Boldness was originally defined as a propensity to take risks (Wilson et al. 1994; Wilson 1998) and experimentally operationalised as an approach

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

141

to, or avoidance of, novel objects. However, the above definition of boldness could include virtually any behaviour. For example, locomotion is almost always risky because it would allow a potential predator to detect and discover the individual. Aggression is risky because it could result in physical injury and reduced attention to an approaching predator. Does it mean that all and any behaviour can be subsumed under the concept of boldness? The second potential problem is more subtle: when the overall personality landscape is obscure, it is easy to confuse different underlying traits. Imagine there are two independent personality traits based on different neurophysiological, hormonal or cognitive mechanisms: (1) fearfulness-reactivity and (2) curiosity. Some individuals could display behaviours indicative of heightened fear in a range of situations; also individuals could be either curious or uninquisitive in different contexts. Now imagine a researcher who decided to study ‘boldness’ operationalised as the propensity to take risks. The researcher developed two tests for boldness measuring an approach response to the stimulus, one involving a dangerous stimulus (e.g. sight of a predator) and another, involving novel object. It is likely that the first test would involve fearfulness-reactivity whereas the second, curiosity. For our blindly operationalist researcher, however, boldness just turns out to be non-existent because different tests presumed to measure boldness fail to detect any correlation! If each of these two kinds of boldness turns out to be consistent over time, however, the researcher may decide that boldness is domain- or situation-specific. The concept of behavioural syndrome may potentially have similar problems. Studies of behavioural syndromes often start from a hypothesis specifying the traits being correlated (e.g. boldness and aggression), whereas other possible relationships may be overlooked. Again, behavioural patterns that the researcher presumes to measure ‘aggression’ in two situations may in fact reflect different behavioural dimensions, motivational, cognitive and emotional mechanisms (e.g. aggression in one context but fear in another). On the other hand, it is possible that suites of traits correlate and form behavioural syndromes at two stages of the ontogeny (or just at two different moments of time) with little correlation across time. Some studies have found correlations between activity and boldness (Fraser et al. 2001; Dingemanse et al. 2007; Moretz et al. 2007). However, closer examination of many of these studies reveals that the correlation between personality traits may simply be a reflection of the techniques and methods employed. Fishes that are highly active, for example, are more likely to spend more time exploring a novel object, a novel environment or in risky locations simply because they are more likely, by chance alone, to score highly in these traits. In other words, the tests of each personality trait (boldness and activity) may not be measured independently. Indeed, activity levels are better quantified in a non-experimental context, such as the home aquaria, than in a novel experimental arena because the latter is a standard test for boldness (open field test; Crabbe et al. 1999; Brown et al. 2007a). Furthermore, analysis of partial correlations may be very helpful in controlling the moderating effect of locomotion on subtle behavioural differences (see Budaev & Andrew 2009a). Thus, studying animal personality inevitably involves certain psychological concepts that may be considered anthropomorphic. Avoiding anthropomorphism by using deliberately blind operational constructs may lead to even more serious problems. The putatively ‘objective’ labels applied to behavioural traits are often uninformative and at worst misleading with respect to their underlying mechanisms. It is hardly possible to completely

BLBK374-07

BLBK374-Brown

142

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

avoid descriptive labels with a degree of subjective connotation. However, the concept of construct validity could be instrumental in minimising interpretational anthropomorphism.

7.2.3

Labelling personality traits; construct validity

Once a measure of personality is obtained, its interpretation is often non-trivial. The descriptive label attached to such a measure must correspond to a particular theoretical concept. For example, if a trait is interpreted as fearfulness, the researcher must provide evidence that it is closely linked with fear (an emotional and/or motivational construct), if it is interpreted as curiosity, there must be evidence that it is linked with a predisposition to obtain novel information. In more formal terms, validity is ‘the degree to which the test actually measures what it purports to measure’ (Anastasi & Urbina 1997). The theoretical construct must specify concepts with which it is related (convergent validity) as well as those with which it is not related (discriminant validity) (Cronbach & Meehl 1955; Anastasi & Urbina 1997). The most popular approach to assess convergent and discriminant validity is the multitrait–multimethod matrix (Campbell & Fiske 1959). As its name suggests, this method involves correlation or factor analysis of a data matrix including several alternative measures of the construct under the study together with unrelated constructs. Then, convergent validity involves correlations between different measures of the same construct (ideally high) while discriminant validity involves correlations between measures of dissimilar constructs (ideally low). For example, in case of curiosity, convergent validity may require high correlations between tests involving responses to novel environment, novel object and novel food. Discriminant validity may involve the absence of high correlation between the tests for novelty and tangential measures such as locomotion or social tendency. In the field of animal behaviour, various experimental procedures and manipulations can be used to assess the validity of personality tests. Construct validity is rarely addressed in the animal personality field. Typically, the investigator chooses the tests and measures of personality traits and ascribes descriptive and interpretative labels to them arbitrarily (like boldness, fearfulness, exploration, sociability, etc.), based on whether they just appear persuasive. An exception in fish research where both convergent and discriminant validities were appropriately shown is the recent study by Burns (2008). In this study, scores the guppies Poecilia reticulata obtained in different open field tests correlated with emergence tests (convergent validity). Also, activity scores did not correlate with open field or emergence test behaviours (discriminant validity). While ecological validity of tests and stimuli (dictating that they should be compatible with the natural environment and behavioural repertoire of the species, see Tinbergen 1963; Lorenz 1974) is often an important concern in animal behaviour and personality research (R´eale et al. 2007), construct validity of tests that measure unobservable personality constructs is also crucial.

7.2.4

Objective and subjective measurements of personality

Even though behavioural consistency may seem a simple concept, measurement of consistent personality traits is usually a difficult task. First, such traits cannot be observed and

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

143

measured directly and must be inferred from multiple measures. Second, the same trait may be measured in a number of ways, using different tests. Furthermore, a single behavioural test may measure several different personality traits simultaneously. Two general approaches have been used to assess personality in animals that avoid these issues: (1) objective behavioural measurements and (2) subjective trait assessment by human observers using rating scales. The objective behavioural measurement method involves testing an animal in one or several tests (i.e. a test battery). A variety of behavioural measures could be recorded in each of these tests: latencies, durations and frequencies of various behaviours, locomotion scores and so on. A single measure or a combination of measures is then used to describe personality. For example, latency to approach a novel object is frequently used to measure boldness (Table 7.1). In the best scenario, several measures should be combined using the principal component or factor analysis or even just summed with unit weights. This approach is the one most commonly adopted by behavioural ecologists. In the subjective assessment method, one or more human observers rate their impressions of the animals using a set of adjectives or other similar descriptions. The observer ascribes quantities to the expression of the trait, which may be either binary (present/absent) or numeric (e.g. 1 for the lowest expression to 5 for the highest expression). For example, a personality assay of spotted hyenas (Crocuta crocuta) used by Gosling (1998) included numerous descriptive expressions like: ‘Bold, brave, not shy: Behaves in a positive, assured manner. Exhibits courage in the face of danger. Is daring, not restrained or tentative. Not timid, shy, or coy.’ (Gosling 1998, p. 117). To date, only the objective behavioural measurement method has been used to describe personality in fishes. Most studies of personality in primates (e.g. Stevenson-Hinde et al. 1980; King & Figueredo 1997; Capitanio 1999; King et al. 2005), and some studies of dogs (Gosling et al. 2003) and birds (Figueredo et al. 1995) used rating scales. Many scientists would argue that the ‘objective’ nature of behavioural measurement would make them superior over the subjective rating scales. However, this is not necessarily true. First, a significant amount of subjectivity is involved in the ‘objective’ methods for personality assessment: the choice of tests, procedures, selection of measures to record and analyse, etc. Second, whereas subjective ratings are based on a generalised perception of personality over many occasions, situations and observations, each of the objective measures is scored in a single context and, therefore, reflects a very significant contextdependent component. Most researchers do not think it essential to describe the protocol used (e.g. how many observers coded behaviour, were they experienced or undergraduate assistants, whether and how they were trained, etc., see Vazire et al. 2007) because objective behavioural measures are usually considered infallible. Reliability and stability of objective measures are usually rather low. However, the statistical power is significantly reduced with diminishing reliability of measurement. When single behavioural measures are used as a proxy for personality traits, large sample size is often necessary to detect moderate consistencies across situations, even when the measures are relatively reliable. Unfortunately, few researchers studying animal personality ever care about reliability. The average sample size used in studies of behavioural replicability was 39 (Bell et al. 2009), which is considered a relatively large number of subjects in

BLBK374-07

BLBK374-Brown

144

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

behavioural research. With such a sample size, however, the minimum correlation coefficient detectable with the power 0.8 at α = 0.05 is 0.43. But this minimum detectable correlation increases to 0.55 when the reliability of the measures equals to 0.8 and to 0.73 when the reliability is 0.6. Behavioural correlations across situations are rarely that strong and will remain undetected if the sample size is not very large and the measures are not perfectly reliable. Personality describes global individual differences, overall trends and predispositions that generalise across observations, measures and contexts. Therefore, single behavioural measurements are often too poor an approach for measuring personality. We argue that many reports that failed to find significant cross-situational consistency in fish boldness and behavioural syndromes (e.g. Wilson et al. 1994; Coleman & Wilson 1998; Bell 2005; Wilson & Stevens 2005; Dingemanse et al. 2007) may have been unable to do so because they used isolated behavioural measures with low reliability. Indeed, many studies of boldness and behavioural syndromes used only one or two measures of these behavioural constructs. To improve the reliability and relevance of personality measurement, psychologists often aggregate behavioural measures over time, situation, observers, etc. Such data aggregation reduces unstable situationally specific behavioural components, improves reliability and increases correlation coefficients measuring consistency (for more discussion, see Epstein 1983; Rushton et al. 1983; Funder 1995; Pruessner et al. 1997). A similar aggregation approach has been used in studies of personality in rats (Ossenkopp & Mazmanian 1985) and fishes (Budaev 1997a). However, sometimes multiple testing of the same individuals may also be difficult or even impossible; in addition to being costly and time-consuming, it may involve habituation, learning, high stress, and other undesirable effects. It has been shown that, when carefully designed, subjective rating scales have high inter-rater agreement, do not reflect anthropomorphic projections and usually agree with objective behavioural measurements (Vazire et al. 2007; Uher & Asendorpf 2008). However, subtle human cognitive biases cannot be ruled out. For example, humans can have specific adaptive cognitive mechanisms for rapid assessment of the human personality. The assessors could then match strangers, animals and even inanimate objects with a set of hardwired cognitive personality templates. Because the templates are species specific, they will not result in disagreement across observers so that subjectivity would not be easily noticed. This becomes increasingly problematic as we move further away from species closely related to ourselves (e.g. primates) towards the taxa with intuitively less familiar behaviour (e.g. fishes and invertebrates). Thus, while isolated behavioural measures are usually too poor an instrument for assessing animal personality, aggregation of many measures would improve personality assessment. Furthermore, subjective rating-bases assessment sometimes provides the most efficient (in terms of time and cost) approach to measure personality. Although human observers are likely to find it more difficult to rate fishes on subjective scales, such scales could still be used in studies of fish personality. When applied to measure personality in fishes, subjective rating scales should be validated using objective behavioural measures in the first instance. Ideally, if a smaller sample experiment using both objective measures and rating scales can be designed, then rating scale assessment could be used for rapid assessment of personality of a larger sample of fish.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

7.2.5

Char Count=

145

Modern terminology and statistical approaches

In order to appreciate the fact that animal personality traits are not incompatible with behavioural plasticity, it is important to adopt terminology developed in the human personality literature and apply modern statistical analyses. Two different aspects of stability are usually distinguished in psychology (Eysenck 1970; Eysenck and Eysenck 1985): (1) ‘stability’ and (2) ‘consistency’. Stability usually means that the absolute level of the behaviour in question does not change. Consistency, on the other hand, refers to predictability (or correlations) during repeated measurement of the same individuals in the same situation or across various situations (Nunnally 1967; Ozer 1986; R´eale et al. 2007). For example, an individual exhibiting higher level of aggressiveness than other individuals in one situation could also be more aggressive than others in another situation even if the absolute level of aggression measured for that individual changes from one situation to the next. Thus, the concept of stability involves the absolute value of a particular behavioural measure whereas consistency involves correlations and relative values within a population of individuals. The level of variability is another concept independent of the first two. Variability involves behavioural scatter in one situation relative to another situation. In effect, the behaviour can be situation-specific while individual differences are consistent. Further, the behaviour can be extremely variable in some situations while individual differences remain consistent (Fig. 7.1). This model can be reformulated in an ANOVA-like way, which is perhaps more familiar to biologists. We can consider two sources of variability: (1) ‘individual’ (random factor because we potentially have an infinite number of ‘random’ individuals) and (2) ‘situation’ (either fixed or random repeated measurement factor). In this way of thinking, consistency means that the individual factor accounts for a significant proportion of the total variance. Recently, Dingemanse et al. (2009) have proposed the concept of a behavioural reaction norm linking individual differences and behavioural plasticity. This approach accounts for individual behavioural response over an environmental gradient (stimulus value, predation

Fig. 7.1 Stability, consistency and variability in behavioural traits. Here the connected points on the left panel depict behavioural profiles of four individuals over three situations A, B and C; the right panel presents scatterplots of correlations between the behaviour scored in these situations. Low average level and high variability of the behaviour is observed in the situation A, high average level and low variability in the situation B and low average level and low variability in C. However, individual differences may be consistent (upper panel, strong correlations between situations) or inconsistent (lower panel, no correlations).

BLBK374-07

BLBK374-Brown

146

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Fig. 7.2 The concept of behavioural reaction norm linking individual differences and phenotypic plasticity (after Dingemanse et al. 2009). The behaviour can be measured over multiple environmental gradients (here A and B). The upper left panel shows consistent individual differences with zero plasticity represented by a collection of horizontal regression lines (slope = 0). Lower left panel displays consistent individual differences with identical plasticity described by parallel regression lines with the same slopes. The right-hand panel shows more complicated patterns involving correlation between elevation and slope, when shy individuals show higher plasticity.

risk, food availability, temperature, etc.). Individual behaviour is described by a linear regression line linking the response with the environmental condition rather than just the response value. Two aspects of the regression are then considered: (1) Elevation describing the average individual level of the behaviour and (2) slope describing individual degree of plasticity (Fig. 7.2). Linear mixed models (random regression model) can be used to estimate parameters of the individual responses, link them together (correlate the elevation and the slope) and with external variables, for example, indicators of fitness. The main advantage of the reaction norm approach is that it allows us to analyse individual differences and plasticity within the same adaptive framework. However, there are limitations. A single trait (measure) is usually analysed, making it less appropriate for the analysis of multivariate personality traits. Individuals are described by linear models requiring multiple measures for reliable parameter estimation (otherwise the standard errors of individual elevation and slope would be very large). Linearity is also not always a realistic assumption. While non-linear and multivariate models can be used, they increase complexity enormously. Furthermore, to achieve reasonable statistical power, random regression models require huge sample sizes (usually N>200, Martin et al. in press).

7.3

Proximate causation

If suites of correlated behavioural traits are observed, a reasonable hypothesis is that these correlations reflect specific genetic and physiological mechanisms that constrain behavioural variability. For example, genetic correlations could be brought about by pleiotropy (multiple action of a particular gene to more than one phenotypic trait) or linkage disequilibrium (non-random association of alleles at different loci, e.g., by physical linkage) (Falconer 1981). The simple existence of a phenotypic correlation could often suggest, not

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

147

necessarily however, that a genetic correlation could also be present between these traits (Falconer 1981; Cheverud 1988; Bakker 1994; Roff 1996). Several studies explored heritability of personality as well as genetic and phenotypic correlations between personality traits in humans (Livesley et al. 1998), dogs (Saetre et al. 2006) and birds (van Oers et al. 2004). Fish personality traits may also have a heritable component. Brown et al. (2007b) compared boldness scores in laboratory-raised offspring taken from two populations of a tropical poeciliid fish, Brachyraphis episcopi, with contrasting predation pressures. Fishes descended from high-predation populations were bolder than those descended from lowpredation population. Furthermore, the differences between the two groups of laboratoryreared fishes were of a similar magnitude as observed in the parental populations. Similarly, Wright et al. (2003) found differences in boldness in zebrafish, Brachydanio rerio, descendants from four wild populations. Some studies have attempted to obtain a measure of personality heritability. Bell (2005), for example, found that heritability estimates of boldness and aggression in two populations of sticklebacks, G. aculeatus, were rather low (<0.2), indicating that there may be a very strong selection depleting additive genetic variance or these traits are mostly under an environmental control. Similarly, Dingemanse et al. (2009) tested sticklebacks originating from high- and low-predation environments. Some of the fishes were also subjected to repeated predator experience allowing assessment of the experience effect. In this case, heritability of personality traits such as novel environment exploration, activity, sociability and boldness ranged from 0.06 to 0.32 and in most cases was higher in population sympatric with predators. The study of personality in fishes is frequently linked with the response to stress. Individuals that are relatively bold also show attenuated stress responses (Brown et al. 2005b); thus, there may be a link between personality traits and the expression of underlying hormones such as cortisol. Correlation between boldness and stress responses has been identified in several species of fishes (Koolhaas et al. 1999; Øverli et al. 2005; Schjolden et al. 2005). Recent studies have also shown relationships between stress responses and coping styles in carp (Huntingford et al. 2010). Metabolic rate was significantly higher in bold than in shy fishes, while expression of the cortisol receptor gene, plasma lactate and glucose concentrations was lower. Similar relationships between boldness and background blood cortisol concentrations have been observed in mulloway, Argyrosomus japonicus, where bold fishes have significantly lower cortisol concentrations than shy fishes (Raoult et al., in press). Moreover, fish lines selected for high- and low-stress responses differ in a range of behavioural tests, including dominance and boldness (Øverli et al. 2005). Such effects often distinguish hatchery-reared and wild fishes (Lepage et al. 2000; Sundstr¨om et al. 2004), which is likely a reflection of the vastly different rearing conditions (Brown & Day 2002). A selection programme was conducted on the rainbow trout, O. mykiss, where two lines of fishes were selectively bred for either high- or low-stress response (high- and low-plasma cortisol responsiveness, HR, i.e. high-response and LR, i.e. low-response lines). Subsequent studies indicated that these lines differ in numerous behavioural and physiological characteristics (Øverli et al. 2007). For example, the HR fishes demonstrated stress-induced anorexia: they did not eat during a stressful experimental period, whereas about 44% of

BLBK374-07

BLBK374-Brown

148

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

LR fishes ate (Øverli et al. 2002). LR trout were more aggressive and dominant than HR (Pottinger & Carrick 2001). Learning experiments indicated that LR trout were characterised by significantly longer extinction of a conditioned stress response than HR (Moreira et al. 2004). Thus, LR fishes show bold, active and inflexible behaviour characteristic of active coping style, almost exactly as previously found in mammals (mice, Benus et al. 1991; pigs, Bolhuis et al. 2005) and birds (tits, Groothuis & Carere 2005). The neuroendocrine differences between the HR and LR lines of trout include monoaminergic activity and metabolism (Øverli et al. 2001). Both the synthesis and metabolism of brain serotonin, norepinephrine and dopamine following stress were significantly more elevated in HR than in LR lines. LR fishes were characterised by increased level of 5-HIAA (serotonin metabolite) and MHPG (norepinephrine metabolite) in the hypothalamus and also had a significantly higher baseline 5-HIAA/5-HT ratios in the telencephalon. Overall, these patterns are similar to those found in mammals (see Øverli et al. 2007 for more discussion). The limbic system, including hippocampus, amygdala, hypothalamus and a few other nearby structures, plays an important role in controlling personality and emotions in humans and other mammals (Gray 1987; Lautin 2002). Even though the organisation of the brain in fishes is significantly different (Chapter 15), certain forebrain areas – lateral and medial telencephalic pallia – are considered homological to the mammalian limbic system (Flood et al. 1976). They also control similar behavioural mechanisms and processes, such as emotional learning (Broglio et al. 2005; Portavella & Vargas 2005). However, how these structures are involved in fish personality remains unknown. Recent studies on zebrafish highlighted a possible involvement of certain epithalamic structures, especially the habenula, in personality and laterality (for a full discussion, see Chapter 16). The habenula is a major component of the dorsal diencephalic conduction pathway connecting the limbic forebrain with midbrain and hindbrain (Sutherland 1982; Bianco & Wilson 2009). In fishes, it is asymmetric, the left lateral habenula significantly exceeds the right (see Bianco & Wilson 2009 for a review). Interestingly, spontaneous reversal of the habenular asymmetry in a selected zebrafish line resulted (along with reversed laterality) in heightened boldness (Dadda et al. 2010). Similarly, development of zebrafish embryos in darkness during early ontogeny, presumably affecting the habenular development (Budaev & Andrew 2009b), also results in differences in boldness (Budaev & Andrew 2009a). This is not surprising because the habenula is heavily involved in behavioural inhibition, pain, fear, anxiety and depression through modulation of the brain dopaminergic system (Shepard et al. 2006). Despite the body of research conducted to date, the genetic, neurophysiological and neuroanatomical causes of individual differences in fish personality remain poorly understood. Whereas there is a substantial interest in neurophysiology of personality in mammals, especially humans (e.g. Eysenck & Eysenck 1985; Gray 1987; Zuckerman 1994), fish provides an ideal model system for studying the role of single genes in the development of personality. For example, polymorphism of the D4 dopamine receptor (D4DR) gene expressed in the limbic areas of the brain in mammals predicts extraversion and novelty seeking in humans (Benjamin et al. 1996; Ebstein et al. 1996, 1998). A similar reduction of behavioural response to novelty was found in knockout mice, lacking the D4DR (Dulawa et al. 1999) and great tits, in which D4DR polymorphism predicts early exploratory behaviour (Fidler et al.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

149

2007). A similar analysis can be conducted on zebrafish or sticklebacks, whose genomes have been sequenced. The gene sequence of D4DR is highly conserved across a wide range of vertebrates and a BLAST (Basic Local Alignment Search Tool) search of the stickleback genome reveals a strong match on chromosome 19 (the stickleback sex chromosome). It is a commonly observed phenomenon across vertebrates that males are more prone to risk taking than females (Wilson & Daly 1985; van Oers et al. 2005; Brown et al. 2007b in fishes, birds and humans, respectively) so the link to a sex chromosome is not surprising. This observation also warrants closer examination in fishes, especially considering the wide range of sex determination mechanisms. In the genomic era it should be possible to identify the genes contributing to heritable personality traits. For example, a quantitative trait loci analysis of boldness scores in zebrafish revealed regions on chromosomes 9 and 16 significantly linked with boldness and a suggestive association with chromosome 21 showing signs of dominance and additive effects, respectively (Wright et al. 2006). To date, however, no further studies have attempted to identify the genes involved in fish personality traits.

7.4

Ontogeny and experience

As we have highlighted, personality traits are not entirely fixed for life, rather they are susceptible to adjustment through significant life experiences and developmental shifts. Only a handful of studies, however, have investigated the ontogeny of personality in fishes. Budaev et al. (1999b) analysed how behavioural consistency changes across the ontogeny in an African cichlid, Steatocranus casuarius. The fishes were tested for their responses to a novel environment, a novel fish, and a mirror (aggression test) at 4, 4.5, 12 and 13.5 months of age. Whereas the behavioural measures were not consistent in juveniles (4 and 4.5 months of age), consistency appeared in adult fishes (12–13.5 months). Behavioural consistency was found very early in newly emerged and larval fishes (Budaev & Andrew 2009a; Conrad & Sih 2009). Such early consistency could also be induced by exposure to predator and probably other environmental stimuli (Budaev & Andrew 2009a). Personality can be significantly modified by factors such as exposure to light (Budaev & Andrew 2009a), acting very early in the ontogeny and becomes more consistent during the individual’s development, which agrees with the data obtained in mammals (e.g. Loughry & Lazari 1994), including humans (see Roberts & DelVecchio 2000 for a review). Individual experience can significantly affect boldness. For example, differences between shy and bold pumpkinseed sunfish were significant in the field, but disappeared after a period of isolation in the laboratory (Wilson et al. 1993). Simulated predator attacks (repeated chasing with a net) increased boldness in captive bred B. episcopi originating from both high- and low-predation populations (Brown et al. 2007b). Magnhagen & Staffan (2005) found that changing group composition could significantly affect the behaviour of shy juvenile perch (Perca fluviatilis) and to a lesser degree, intermediate and bold perch. Shy fish, for example, become bolder when placed in a group of shy fishes. Bold individuals in a bold group tend to reduce their levels of boldness, whereas intermediate individuals did not change behaviour. Similar effects were confirmed by Magnhagen (2007): the correlation between risk taking (time spent in open water) and

BLBK374-07

BLBK374-Brown

150

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

exploration (entrance to novel environment) was significant only when the behaviour of other group members was taken into account. Experience of winning and losing a contest and simply observing a shy or bold conspecific could alter boldness in the rainbow trout, O. mykiss (Frost et al. 2007). The sex of the schoolmates may also significantly affect individual boldness. For example, male guppies are bolder after a simulated aerial predator attack when shoaling with males than with females (Piyapong et al. 2010). Such strong social influences on the expression of personality traits are expected in fishes that spend a considerable amount of their lives in schools (Brown and Laland 2002; Chapters 10 and 11). The pattern of social interactions can be significantly affected by personality traits, creating a further level of complexity. For example, social networks of Trinidadian guppies are characterised by significant assortment and shyer guppies have higher numbers of network connections (Croft et al. 2009). In addition to prior experience, fish personality could be affected by various physical factors. Ambient temperature, significantly affecting metabolism, would be one of the most important factors for fishes and other ectothermic animals. Even moderate changes in temperature could significantly affect boldness, aggressiveness and activity levels of damselfish, Pomacentrus moluccensis and P. bankanensis (Biro et al. 2009). Whereas correlations between personality traits were stable across temperature levels, individuals differed in their degree of plasticity. While some individuals significantly increased activity at higher temperature, others were more stable. Thus, as with the vast majority of traits, fish personality is not entirely genetically fixed, rather it can be affected during early ontogeny and modified by later experience. Furthermore, it is possible that the degree of flexibility and susceptibility to experience is itself a consistent individual trait affected by various developmental events and selection pressures within a particular population.

7.5

Is personality adaptive?

The starting point for the theory of evolution by natural selection is that the traits being selected have some genetic basis. We have seen above that this is likely to be true for personality traits in fishes as well as in many other species. The next question is: Are such consistent differences across many contexts, over time and developmental stages simply non-adaptive noise around a single adaptive mean? Can natural selection produce and maintain variation? Will natural selection also support correlations across behavioural domains and contexts? That is, is personality adaptive?

7.5.1

Frequency- and density-dependent selection

Early game theory models tended to neglect any variability except alternative strategies. It was accepted that two or more strategies could be evolutionarily stable (mixed strategies, see Maynard Smith 1982) when fitness of a strategy depends on the frequency of the alternative strategy. However, no assumption was made about correlation and consistency. For example, it was accepted that natural selection will produce aggressive (‘hawk’) and non-aggressive (‘dove’) strategies within a single population, but it was not clear whether

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

151

the particular individuals would use the same strategy consistently or switch randomly between them. Later models slowly began to incorporate individual differences. In a seminal paper on shyness and boldness, Wilson et al. (1994) provided a simple account for the existence of the shy–bold continuum in various species based on frequency- or density-dependent selection. Here, the optimal behaviour depends on the frequency or density of conspecifics displaying each trait. For example, it may be more adaptive to be shy and occupy safe habitats; however, when the density of such shy individuals is sufficiently high, the safe habitat becomes overcrowded. At some point, the fitness cost of competition within the safe habitat exceeds the predation risk in the risky habitat and some individuals start using the risky habitat (see Wilson et al. 1994). In a later paper, Wilson (1998) extended this framework: when the population exploits several habitats, a range of resources, etc., natural selection could maintain multiple-niche polymorphisms and stable personalities. However, such models still do not account for consistency: why should one individual be bold consistently in various contexts rather than have a broad adaptive norm of reaction? To explain consistency, it is assumed that phenotypic plasticity is limited by various genetic and epigenetic mechanisms and incurs significant fitness costs (see DeWitt et al. 1998; Sih et al. 2004). Many fish studies are consistent with such models. For example, a shyness–boldness continuum was found in wrasses, Symphodus ocellatus. In this species, fishes significantly differ in activity within a novel environment. Shy individuals (which do not explore novel environments) tended to stay in shoals and occupy relatively safe weeded habitats, whereas bold individuals (which are active in novel environments) were typically found in more dangerous open habitats and did not join shoals (Budaev 1997b). Similar patterns were found in bluegill and pumpkinseed sunfish (e.g. Ehlinger & Wilson 1988; Wilson et al. 1993). Feeding specialisations and alternative foraging tactics that may involve exploitation of distinct food resources and sub-niches by different individuals have long been reported in fishes (Bryan & Larkin 1972; Ringler 1983; Smith & Skulason 1996). While the early models explaining personality variation based on frequency- and densitydependent selection look convincing in many cases, they have difficulty in accounting for consistency over time and across situations: such consistency is still considered to result from constraints on adaptation. At the basic level, they are not different from the ‘hawk–dove’ game.

7.5.2

State-dependent models

More recent models have used a dynamic programming approach, where the pay-off of every behaviour is calculated iteratively and depends on the state of the animal, including energy reserves, territory size, etc. (Mangel & Clark 1988). It would be adaptive to avoid exposure to excessive predation risks if an animal has good energy reserves, but if such reserves are close to depletion (e.g. the animal is close to starvation) the potential benefits of obtaining food may outweigh the risk of being eaten (see Clark 1994; Dall et al. 2004). For example, sticklebacks characterised by higher weight loss following a 2-day food deprivation tended to emerge earlier from a refuge (Krause et al. 1999). Quite simple models predict consistency if the state and the history of behaviour over time are taken

BLBK374-07

BLBK374-Brown

152

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

into account (see Dall et al. 2004 for more discussion). Consistent differences in risk taking could be maintained within a foraging group if individuals differ in energy reserves and such differences are sustained by environmental factors (see Rands et al. 2003; Dall et al. 2004). Furthermore, when the individuals’ states are changeable but information about the world is uncertain, it may become adaptive to just ignore environmental cues and behave consistently. Consequently, environmental noise is predicted to facilitate consistent personalities (McElreath & Strimling 2006). Recent theoretical advances have emphasised the role of life-history trade-offs in generating consistent personalities (Stamps 2007; Biro & Stamps 2008). For example, Wolf et al. (2007) developed a series of state-dependent models based on a trade-off between current and future reproduction. Individuals with high expectations of future reproduction, who have much to lose, would be shy whereas those with low expectations would be bold. This is the case of a more general ‘asset protection principle’ (Clark 1994), stating that the larger the current reproductive asset, the more vital it becomes to avoid risks threatening reproduction. Therefore, accumulation of reproductive assets (body size, territory size, etc.) would lead to risk aversion (Brown & Braithwaite 2004). Similar considerations involving trade-offs have also been used to account for consistent individual differences in decision making. Specifically, a trade-off between speed and accuracy could lead to individual differences in impulsiveness: some individuals make fast and inaccurate decisions whereas others are careful but slow (Chittka et al. 2009). Guppies (P. reticulata) consistently differ in their ‘hastiness’ in a spatial memory maze task (with female as reinforcement): some individuals tend to make quick decisions with many errors while others are slow but accurate (Burns & Rodd 2008). Trade-offs between growth and mortality (Stamps 2007) and productivity and mortality (Biro & Stamps 2008) may be among the most important mechanisms maintaining consistent personality variation. In essence, the growth-mortality hypothesis argues that correlations between behaviours in various contexts may arise when these behaviours affect growth and mortality. Many fishes and other animals show consistent individual differences in growth rate (Biro et al. 2006; Stamps 2007). Personality traits such as risk taking and aggressiveness often affect both growth and mortality rates. Aggressiveness could increase growth rate by improving access to food resources, but would also increase the risk of injury and non-detection of a predator. Boldness in foraging context would also increase food intake, but simultaneously increase the risk of being eaten by a predator (Stamps 2007; Biro & Stamps 2008). From a more general perspective, any life-history trade-offs may be important: consistency appears when the behavioural tendencies contribute to individual differences in life-history productivity (Biro & Stamps 2008). The hypothesis developed by Biro & Stamps (2008) has serious limitations by assuming a simple positive relationship between personality (boldness or aggressiveness) and access to resources or food intake. Some studies reported a positive correlation between boldness and body mass in fishes (e.g. Magnhagen & Borcherding 2008), some reported no relationship (Kobler et al. 2009) while others documented a reverse relationship (e.g. Brown & Braithwaite 2004; Millot et al. 2009), which is indeed expected in many state-dependent models involving the asset protection principle (Clark 1994). Furthermore, there is no clear relationship between aggressiveness and social dominance (e.g. Bakker 1986; Francis

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

153

1988; Colleter & Brown 2011), and dominance and growth rate can also be unrelated (e.g. Harwood et al. 2003). Individual experience or similar mechanisms provide another important, although often neglected by evolutionary biologists, ingredient making individual differences consistent. A model developed by Wolf et al. (2008) accounts for consistent differences in responsiveness to environmental stimuli, with ‘responsive’ strategy assessing such stimuli and ‘unresponsive’, behaving rigidly. In addition to frequency-dependent selection (responsiveness pay-off reduces with increasing frequency of responsive individuals), it adds a positive feedback mechanism: responsiveness is less costly for individuals that were responsive before. It turns out that even a small positive feedback induces stable correlations between behavioural choices made at consecutive iterations. Such a feedback is realistic if responsive individuals are more efficient at discovering food with experience. Thus, personality is shaped by natural selection. The most basic ingredient usually invoked to account for alternative strategies, behavioural polymorphisms and personalities is frequency- and density-dependent selections. Other mechanisms predicting adaptive individual differences, for example involving optimal decisions in unpredictable conditions (‘adaptive coin-flipping’, Cooper & Kaplan 1982; Kaplan & Cooper 1984), have been given surprisingly little attention, even though they may be more general. Furthermore, consistent individual differences are likely to arise when individuals can exploit several resources, habitats and sub-niches. A variety of other mechanisms, including environmental noise, protection of reproductive assets, accumulation of individual experience and lifehistory trade-offs, would facilitate consistency over time and across contexts. All these ingredients are typically found in many fish species and populations, making consistent personality the ‘null hypothesis’. Nonetheless, in spite of the recent theoretical advances, our understanding of the adaptive factors contributing to the maintenance of consistent personalities and polymorphisms remains scant. The various existing models are often too simplistic and sometimes contradict each other and the empirical data. Thus, while the main adaptive factors producing personality in fishes are known, the exact mechanisms involved still remain a puzzle.

7.6

Evolution

Even though it is now clear that adaptive individual differences can be maintained by natural selection, very little is known about the phylogeny of personality. Because very similar personality factors have been found in a wide variety of vertebrates (Budaev 1998; Gosling & John 1999), they could represent a shared heritage involving homologous brain systems (for similar views, see Eysenck and Eysenck 1985; Gray 1987). This view depicts personality as a consequence of constraint on evolution. Alternatively, personality could be shaped by common adaptive mechanisms independently in each species or even in different populations; this view depicts personality as an adaptation. Further, if common adaptive factors are operating in a variety of species, we may expect similar patterns of personality to evolve (i.e. convergent evolution). Understanding the evolution of behaviour usually involves analysis of the patterns of similarities and differences across related species to

BLBK374-07

BLBK374-Brown

154

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

elucidate the history of their appearance and divergence throughout the phylogeny. In a sense, evolution of behavioural traits should be studied comparatively as is the case with the evolution of morphology and physiology; behavioural patterns here represent taxonomic traits similar to morphological characters (Tinbergen 1963) to which various methods of cladistic analysis are applicable (Brooks & McLennan 1991). It is tempting to use the same approach to animal personality. However, personality is extremely difficult to organise in such a phylogenetic framework. Personality traits and dimensions that appear as a result of correlation analysis are different from morphological characters and fixed behavioural patterns. Personality traits are artificial descriptive constructs, which have no ‘real’ physical existence and explanatory power (Revelle 1983; Eysenck & Eysenck 1985). Furthermore, there is no single optimal hierarchical level for personality traits. It is possible to define more narrow or more context-specific traits or fewer broader traits. In human personality research, second-order factor analysis has become popular (e.g. Cattell 1956; Digman 1997). Personality traits, dimensions and factors can be blended or split in various species or populations depending on characteristics of the sample of individuals, domains of situations, types of measures, characteristics of raters and various other causes. Such blending or splitting cannot be easily translated to any specific evolutionary events. Personality traits resulting from factor analysis can be rotated differently: the same correlation matrix may be equally well represented by an infinite number of factor loading patterns. In human personality psychology there is no single universal species-specific personality structure. The dominant Big Five model (postulating that human personality variation is encompassed by five basic dimensions: (1) extraversion, (2) neuroticism, (3) agreeableness, (4) conscientiousness, and (5) openness to experience) is merely a point of consensus as an optimal research framework (Digman 1990), and is not the only possible speciesspecific pattern for humans (McAdams 1992). It is possible to extract more narrow personality factors instead of fewer broader ones. Indeed, splittings (16 factors, Cattell 1973) and lumpings (two or three factors: Eysenck 1970; Eysenck & Eysenck 1985; Cloninger et al. 1991; single general personality factor: Rushton et al. 2008) as well as alternative rotations (e.g. Gray 1982; Zuckerman et al. 1993; Caprara and Perugini 1994; Budaev 1999) of the human Big Five personality dimensions have been proposed as alternative models. A further problem is that the dimensions presumed to be common in different species are not necessarily comparable. Comparison of personality in different species is totally based on the researcher’s intuitive interpretation in each case. For example, the researcher may interpret some behaviours as indicative of ‘boldness’ in one species. In a different species, ecological validity may dictate a different set of tests and measures but again a ‘boldness’ trait could be defined. The researcher then argues that ‘similar’ boldness traits are found in both species. This is, however, incorrect because what is compared here is intuitive interpretation of behaviours rather than behaviours themselves. Again, we emphasise that the labels ascribed to behavioural traits are often arbitrary and noninformative. While informal comparisons of personality dimensions across different species may be very helpful (see Gosling & John 1999), they cannot be used for formal phylogenetic analysis. The study of the evolutionary history of personality variations requires a new approach.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

7.7 7.7.1

Char Count=

155

Wider implications Fish production and reproduction

Personality and especially coping styles may have several important implications for aquaculture and conservation of fish stocks based on artificial rearing. Comparisons of wild and hatchery-reared fishes of several species reveal many important behavioural and physiological differences (e.g. Brown & Day 2002; Huntingford 2004; Huntingford & Adams 2005). For example, hatchery-reared brown trout, Salmo trutta, are significantly bolder and more aggressive (e.g. Sundstr¨om et al. 2004) and tend to dominate wild fishes (e.g. Sundstr¨om et al. 2003). These differences between the wild and domesticated fishes are heritable and result from deliberate selection for fast growth rate as well as from indirect selection of fishes best adapted to intensive high-density aquaculture (Huntingford 2004; Huntingford & Adams 2005). Possible implications of these personality differences for the welfare of farmed fishes have been considered by Huntingford & Adams (2005). Shy and non-aggressive fishes turn out to be disadvantaged in many high-density production systems, negatively affecting their welfare. Pre-screening, increasing the cost of fighting and competition by increasing the water current, more even distribution of food avoiding its monopolisation, the use of smart automatic feeders and other measures have been proposed (see Huntingford & Adams 2005 for further discussion). Restocking natural habitats with hatchery-reared juveniles has been used to replenish declining wild populations of various fishes, especially salmonids (Brown & Day 2002; Myers 2004; Bell et al. 2006). Restocking is very expensive (Beck et al. 1994) and its efficiency remains controversial. Most farm fishes die soon after release (Brown & Laland 2001) and some research suggests that the presence of hatchery-reared fishes in natural habitats may in fact facilitate extinction of wild stocks (e.g. McGinnity et al. 2003, 2009). Personality can be an important mediating factor here. Hatchery-reared fishes are more aggressive, bold, dominant and outcompete the wild fishes (Sundstr¨om et al. 2003, 2004; Huntingford 2004; Huntingford & Adams 2005). On the other hand, higher growth rate and the propensity to take risks would make them significantly more vulnerable to predators (Biro et al. 2004). Restocking natural environments with hatchery-adapted fishes, therefore, would depress rather than replenish the natural populations: hatchery fishes would competitively depress the wild fishes but would not contribute to reproduction due to increased mortality through predation.

7.7.2

Personality and population dynamics

Even though populations with contrasting predation regimes are frequently compared in fishes (e.g. Bell & Stamps 2004; Brown et al. 2007a, 2007b; Dingemanse et al. 2007), most studies of personality in animals strongly focus on individuals. Personality is, obviously, an attribute of the individual rather than the group. However, patterns of personality within a population may significantly affect higher order processes such as population dynamics. It has long been known that within-population variability in various traits, such as growth rate and body size, can increase the population stability, persistence and resistance to extinction

BLBK374-07

BLBK374-Brown

156

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(see Łomnicki 1988; Uchma´nski 2000; Grimm & Uchma´nski 2002). Such population models historically concentrated on individual variation in morphological and life-history traits. However, they can involve behavioural variability linked with competitive ability, resource utilisation and life-history traits (e.g. Biro & Stamps 2008; Colleter & Brown 2011). The presence of individual variation within competing species can facilitate their coexistence. For example, Begon & Wall (1987) developed a simple model based on the classical Lotka–Volterra equation to describe competition between two species. In the classical model without individual variability, the competitively superior species rapidly excluded the inferior competitor to extinction. However, when individuals of both species differ in competitive ability, species with different competitive ability can easily coexist. A few recent models investigated population effects of consistent personality traits. Petrovskii & Blackshaw (2003; see also Petrovskii et al. 2008) considered a population in a homogeneous environment under favourable environmental conditions. At some moment, the environment suddenly becomes harsh, causing environmental fragmentation (conditions are favourable within fragments but not in the rest of the environment) and significant mortality. Interestingly, if consistency is added to the model (the same individuals play fixed strategies, either aggressive or non-aggressive), the population density decreases significantly more slowly than in populations where all individuals play randomly. These studies have important implications for conservation by suggesting that personality variation represents a very important yet neglected dimension of biodiversity. In some cases, personality could be the key factor in maintaining the population stability, especially when the population size is small, such that stochastic oscillations or catastrophic events could bring the population to extinction.

7.8

Conclusions

For some time individual differences in behaviour have been ignored by ethologists and behavioural ecologists and ascribing personality traits them has been highly controversial. It is becoming increasingly apparent, however, that rather than representing annoying noise in population data sets, they represent the leading edge of evolution. Much could be gained by adopting game theory models. Both Darwin and Tinbergen recognised the importance of understanding individuality and the latter in particular cautioned against taking too narrow a view when studying animal behaviour. By taking a holistic view of behaviour, we begin to see important relationships between behavioural traits and, in some instances, recognise potential constraints and trade-offs that may limit plasticity. The study of fish personality is still very much in its infancy and suffers from a divided literature. There is a clear need to strengthen experimental methodology by taking advantage of the well-established human personality literature. We also recognise the importance of remaining faithful to the classical ethological framework. While some work has been directed at addressing ultimate questions, much remains to be done in terms of examining proximate causes of fish personality. It is clear that personality is subject to natural selection and research using fishes as model organisms has revealed that personality traits show a great deal of variability within populations, have fitness consequences and are heritable. In

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

157

the genomic era it may be possible to pinpoint the exact genes responsible for personality traits and identify underlying hormone cascades that may link behavioural traits. Most of the research to date has been conducted using the usual model organisms such as guppies and sticklebacks, but it is apparent that a better understanding of personality traits will have direct consequences for fisheries and aquaculture management. Thus, future studies are likely to be aimed at commercial and recreational species, such as salmonids, where selective manipulation of particular personality traits may significantly enhance productivity, as has been the case in terrestrial animal husbandry.

Acknowledgements Sergey Budaev was supported by the EU Sixth Framework Programme grant, ‘Evolution and Development of Cognitive, Behavioural and Neural Lateralization’. Culum Brown was supported by an Australian Research Fellowship from the Australian Research Council.

References Anastasi, A. & Urbina, S. (1997) Psychological Testing, 7th edn. Prentice Hall, Upper Saddle River, NJ. Azuma T., Dijkstra, J.M., Kiryu, I., Sekiguchi, T., Terada, Y., Asahina, K., Fischer, U. & Ototake, M. (2005) Growth and behavioral traits in Donaldson rainbow trout (Oncorhynchus mykiss) cosegregate with classical major histocompatibility complex (MHC) class I genotype. Behavior Genetics, 35, 463–478. Baerends, G.P. (1985) Do the dummy experiments with sticklebacks support the IRM-concept? Behaviour, 93, 258–277. Bakker, T.C.M. (1986) Aggressiveness in sticklebacks (Gasterosteus aculeatus L.): A behaviourgenetic study. Behaviour, 98, 1–144. Bakker, T.C.M. (1994) Genetic correlations and the control of behavior, exemplified by aggressiveness in sticklebacks. Advances in the Study of Behavior, 23, 135–171. Beck, B.B., Rapaport, L.G. & Wilson, A.C. (1994) Reintroduction of captive-born animals. In: P.J.S. Olney, G.M. Mace & A.T.C. Feister (eds) Creative Conservation: Interactive Management of Wild and Captive Animals, pp. 265–286. Chapman & Hall, London. Begon, M. & Wall, R. (1987) Individual variation and competitor coexistence: a model. Functional Ecology, 1, 237–241. Bell, A.M. (2005) Behavioural differences between individuals and two populations of stickleback (Gasterosteus aculeatus). Journal of Evolutionary Biology, 18, 464–473. Bell, A.M., Hankison, S.J. & Laskowski, K.L. (2009) The repeatability of behaviour: a meta-analysis. Animal Behaviour, 77, 771–783. Bell, A.M. & Stamps, J.A. (2004) Development of behavioural differences between individuals and populations of sticklebacks, Gasterosteus aculeatus. Animal Behaviour, 68, 1339–1348. Bell, J.D., Bartley, D.M., Lorenzen, K. & Loneragand, N.R. (2006) Restocking and stock enhancement of coastal fisheries: potential, problems and progress. Fisheries Research, 80, 1–8. Benjamin, J., Li, L., Patterson, C., Greenberg, B.D., Murphy, D.L. & Hamer, D.H. (1996) Population and familial association between the D4 dopamine receptor gene and measures of Novelty Seeking. Nature Genetics, 12, 78–80. Benus, R.F., Bohus, B., Koolhaas, J.M. & van Oortmerssen, G.A. (1991) Heritable variation for aggression as a reflection of individual coping strategies. Experientia, 47, 1008–1019.

BLBK374-07

BLBK374-Brown

158

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Bianco, I.H. & Wilson, S.W. (2009) The habenular nuclei: a conserved asymmetric relay station in the vertebrate brain. Philosophical Transactions of the Royal Society B, 364, 1005–1020. Biro, P.A., Abrahams, M.V., Post, J.R. & Parkinson, E.A. (2004) Predators select against high growth rates and risk-taking behaviour in domestic trout populations, Proceedings of the Royal Society B, 271, 2233–2237. Biro, P.A., Abrahams, M.V., Post, J.R. & Parkinson, E.A. (2006) Behavioural trade-offs between growth and mortality. Journal of Animal Ecology, 75, 1165–1171. Biro, P.A., Beckmann, C. & Stamps, J.A. (2009) Small within-day increases in temperature affects boldness and alters personality in coral reef fish. Proceedings of the Royal Society B, 277, 71–77. Biro, P.A. & Stamps, J.A. (2008) Are animal personality traits linked to life-history productivity? Trends in Ecology and Evolution, 23, 361–368. Bolhuis, J.E., Schouten, W.G.P., Schrama, J.W. & Wiegant, V.M. (2005) Individual coping characteristics, aggressiveness and fighting strategies in pigs. Animal Behaviour, 69, 1085–1091. Bolyard, K.J. & Rowland, W.J. (1996) Context-dependent response to red coloration in stickleback, Animal Behaviour, 52, 923–927. Brelin, D., Petersson, E. & Winberg, S. (2005) Divergent stress coping styles in juvenile brown trout (Salmo trutta). Annals of the New York Academy of Sciences, 1040, 239–245. Brick, O. & Jakobsson, S. (2002) Individual variation in risk taking: the effect of a predatory threat on fighting behavior in Nannacara anomala. Behavioral Ecology, 4, 439–442. Broglio, C., G´omez, A., Dur´an, E., Oca˜na, Jim´enez-Moya, F., Rodr´ıguez, F. & Salas, C. (2005) Hallmarks of a common forebrain vertebrate plan: specialized pallial areas for spatial, temporal and emotional memory in actinopterygian fish. Brain Research Bulletin, 66, 277–281. Brooks, D.R. & McLennan, D.A. (1991) Phylogeny, ecology, and behavior: a research program in comparative biology. Chicago, University of Chicago Press. Brown, C. & Braithwaite, V.A. (2004) Size matters: a test of boldness in eight populations of bishop, Brachyraphis episcopi. Animal Behaviour, 68, 1325–1329. Brown, C., Burgess, F. & Braithwaite, V.A. (2007b) Heritable and experiential effects on boldness in a tropical peociliid. Behavioral Ecology and Sociobiology, 62, 237–243. Brown, C. & Day, R.L. (2002) The future of stock enhancements: lessons for hatchery practice from conservation biology. Fish and Fisheries, 3, 79–94. Brown, C., Gardener, C. & Braithwaite, V.A. (2005b) Differential stress responses in fish from areas of high- and low-predation pressure. Journal of Comparative Physiology B, 175, 305–312. Brown, C., Jones, F. & Braithwaite, V.A. (2005a) In situ examination of boldness-shyness traits in the tropical poeciliid, Brachyraphis episcopi. Animal Behaviour, 70, 1003–1009. Brown, C., Jones, F. & Braithwaite, V.A. (2007a) Correlation between boldness and body mass in natural populations of the poeciliid Brachyrhaphis episcopi. Journal of Fish Biology, 71, 1590–1601. Brown, C. & Laland, K. (2001) Social learning and life skills training for hatchery reared fish. Journal of Fish Biology, 59, 471–493. Brown, C. & Laland, K. (2002) Social enhancement and social inhibition of foraging behaviour in hatchery-reared Atlantic salmon (Salmo salar). Journal of Fish Biology, 61, 987–998. Brown, C. & Warburton, K. (1999) Differences in timidity and escape responses between predatornaive and predator-sympatric rainbowfish populations. Ethology, 105, 491–502. Bryan, J.E. & Larkin, P.A. (1972) Food specialisation by individual trout. Journal of the Fisheries Research Board of Canada, 29, 1615–1624. Budaev, S.V. (1997a) “Personality” in the guppy (Poecilia reticulata): a correlational study of exploratory behavior and social tendency. Journal of Comparative Psychology, 111, 399–411. Budaev, S.V. (1997b) Alternative styles in the European wrasse, Symphodus ocellatus: boldnessrelated schooling tendency. Environmental Biology of Fishes, 49, 71–78. Budaev, S.V. (1998) How many dimensions are needed to describe temperament in animals: A factor reanalysis of two data sets. International Journal of Comparative Psychology, 11, 17–29. Budaev, S.V. (1999) Sex differences in the Big Five personality factors: Testing an evolutionary hypothesis. Personality and Individual Differences, 26, 801–813. Budaev, S.V. & Andrew, R.J. (2009a) Shyness and behavioural asymmetries in larval zebrafish (Brachydanio rerio) developed in light and dark. Behaviour, 146, 1037–1052.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

159

Budaev, S.V. & Andrew, R.J. (2009b) Patterns of early embryonic light exposure determine behavioural asymmetries in zebrafish: a habenular hypothesis. Behavioural Brain Research, 200, 91–94. Budaev, S.V. & Zworykin, D.D. (2002) Individuality in fish behavior: ecology and comparative psychology, Journal of Ichthyology, 42(Supplement 2), S189–S195. Budaev, S.V., Zworykin, D.D. & Mochek, A.D. (1999a) Individual differences in parental care and behavioural profile in the convict cichlid: A correlation study. Animal Behaviour, 58, 195–202. Budaev, S.V., Zworykin, D.D. & Mochek, A.D. (1999b) Consistency of individual differences in behaviour of the lion-headed cichlid, Steatocranus casuarius. Behavioural Processes, 48, 49–55. Burns, J. (2008) The validity of three tests of temperament in guppies (Poecilia reticulata). Journal of Comparative Psychology, 122, 344–356. Burns, J.G. & Rodd, F.H. (2008) Hastiness, brain size and predation regime affect the performance of wild guppies in a spatial memory task. Animal Behaviour, 76, 911–922. Campbell, D.T. & Fiske, D. (1959) Convergent and discriminant validation by the multitraitmultimethod matrix. Psychological Bulletin, 56, 81–104. Capitanio, J.P. (1999) Personality dimensions in adult male rhesus macaques: Prediction of behaviours across time and situation. American Journal of Primatology, 47, 299–320. Caprara, G.V. & Perugini, M. (1994) Personality described by adjectives: The generalizability of the Big Five to the Italian lexical context. European Journal of Personality, 8, 357–369. Caro, T.M. & Bateson, P. (1986) Organization and ontogeny of alternative tactics. Animal Behaviour, 34, 1483–1499. Cattell, R. (1956) Second-order personality factors. Journal of Consulting Psychology, 20, 411–418. Cattell, R. (1973) Personality and Mood by Questionnaire. Jossey-Bass, San Francisco, CA. Cheverud, J. (1988) A comparison of genetic and phenotypic correlations. Evolution, 42, 958–968. Chittka, L., Skorupski, P. & Raineet, N.E. (2009) Speed–accuracy tradeoffs in animal decision making. Trends in Ecology and Evolution, 24, 400–407. Clark, A.B. & Ehlinger, T.J. (1987) Pattern and adaptation in individual behavioral differences. In: P.P.G. Bateson & P.H. Klopfer (eds) Perspectives in Ethology, Vol. 7, pp. 1–47. Plenum Press, New York. Clark, C.W. (1994) Antipredator behavior and the asset-protection principle. Behavioral Ecology, 5, 159–170. Cloninger, C.R., Przybeck, T.R. & Svrakic, D.M. (1991) The Tridimensional Personality Questionnaire: US normative data. Psychological Reports, 69, 1047–1057. Coleman, K. & Wilson, D.S. (1998) Shyness and boldness in pumpkinseed sunfish: individual differences are context-specific. Animal Behaviour, 56, 927–936. Colleter, M. & Brown, C. (2011) Personality traits predict hierarchy rank in male rainbowfish (Melanotaenia duboulayi) social groups. Animal Behaviour, in press. Conrad, J.L. & Sih, A. (2009) Behavioural type in newly emerged steelhead Oncorhynchus mykiss does not predict growth rate in a conventional hatchery rearing environment. Journal of Fish Biology, 75, 1410–1426. Cooper, W.S. & Kaplan, R.H. (1982) “Adaptive coin-flipping”: a decision-theoretic examination of natural selection for random individual variation. Journal of Theoretical Biology, 94, 135–151. Crabbe, J.C., Wahlsten, D. & Dudek, B.C. (1999) Genetics of mouse behavior: interactions with laboratory environment. Science, 284, 1670–1672. Croft, D.P., Krause, J., Darden, S.F., Ramnarine, I.W., Faria, J.J. & James, R. (2009) Behavioural trait assortment in a social network: patterns and implications. Behavioral Ecology and Sociobiology, 63, 1495–1503. Cronbach, L.J. & Meehl, P.E. (1955) Construct validity in psychological tests. Psychological Bulletin, 52, 281–302. Dadda, M., Domenichini, A., Piffer, L., Argenton, F. & Bisazza, A. (2010) Early differences in epithalamic left–right asymmetry influence lateralization and personality of adult zebrafish. Behavioural Brain Research, 206, 208–215. Dall, S.R.X., Houston, A.I. & McNamara, J.M. (2004) The behavioural ecology of personality: consistent individual differences from an adaptive perspective. Ecology Letters, 7, 734–739.

BLBK374-07

BLBK374-Brown

160

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Dennett, D. (1983) Intentional systems in cognitive ethology: the ‘Panglossian paradigm’ defended. Behavioral and Brain Sciences, 6, 343–398. DeWitt, T.J., Sih, A. & Wilson, D.S. (1998) Costs and limits of phenotypic plasticity. Trends in Ecology and Evolution, 13, 77–81. Digman, J. (1990) Personality structure: the emergence of the Five-Factor model. Annual Review of Psychology, 41, 417–440. Digman, J. (1997) Higher order factors of the Big Five. Journal of Personality and Social Psychology, 73, 1246–1256. Dingemanse, N.J., Van Der Plas, F., Wright, J., R´eale, D., Schrama, M., Roff, D.A., Van Der Zee, E. & Barber, I. (2009) Individual experience and evolutionary history of predation affect expression of heritable variation in fish personality and morphology. Proceedings of the Royal Society B, 276, 1285–1293. Dingemanse, N.J., Wright, J., Kazem, A.N., Thomas, D.K., Hickling, R. & Dawnay, N. (2007) Behavioural syndromes differ predictably between 12 populations of three-spined stickleback. Journal of Animal Ecology, 76, 1128–1138. Dugatkin, L.A. & Alfieri, M.S. (2003) Boldness, behavioral inhibition and learning. Ethology Ecology & Evolution, 15, 43–49. Dugatkin, L.A., McCall, M.A., Gregg, R.G., Cavanaugh, A., Christensen, C. & Unseld, M. (2005) Zebrafish (Danio rerio) exhibit individual differences in risk-taking behavior during predator inspection. Ethology Ecology & Evolution, 17, 77–81. Dulawa, S.C., Grandy, D.K., Low, M.J., Paulus, M.P. & Geyer, M.A. (1999) Dopamine D4 receptorknock-out mice exhibit reduced exploration of novel stimuli. Journal of Neuroscience, 19, 9550–9556. Ebstein, R.P., Levine, J., Geller, V., Auerbach, J., Gritsenko, I. & Belmaker, R.H. (1998) Dopamine D4 receptor and serotonin transporter promoter in the determination of neonatal temperament. Molecular Psychiatry, 3, 238–246. Ebstein, R.P., Novick, O., Umansky, R., Priel, B., Osher, Y., Blaine, D., Bennett, E.R., Nemanov, L., Katz, M. & Belmaker, R.H. (1996) Dopamine D4 receptor (D4DR) exon III polymorphism associated with the human personality trait of novelty seeking. Nature Genetics, 12, 78–80. Ehlinger, T.J. & Wilson, D.S. (1988) Complex foraging polymorphism in bluegill sunfish. Proceedings of the National Academy of Sciences of the USA, 85, 1878–1882. Epstein, S. (1983) Aggregation and beyond: some basic issues in the prediction of behavior. Journal of Personality, 51, 360–392. Eysenck, H.J. (1970) The Structure of Human Personality. Methuen, London. Eysenck, H.J. & Eysenck, M.W. (1985) Personality and Individual Differences. A Natural Science Approach. Plenum Press, New York. Falconer, D.S. (1981) Introduction to Quantitative Genetics, 2nd edn. Longman, London. Fidler, A.E., van Oers, K., Drent, P.J., Kuhn, S., Mueller, J.C. & Kempenaers, B. (2007) Drd4 gene polymorphisms are associated with personality variation in a passerine bird. Proceedings of the Royal Society B, 274, 1685–1691. Fleeson, W. (2004) Moving personality beyond the person-situation debate. Current Directions in Psychological Science, 13, 83–87. Flood, N.B., Overmier, J.B. & Savage, G.E. (1976) Teleost telencephalon and learning: an interpretive review of data and hypotheses. Physiology and Behavior, 16, 783–798. Francis, R.C. (1988) On the relationship between aggression and social dominance. Ethology, 78, 223–237. Francis, R. (1990) Temperament in a fish: a longitudinal study of the development of individual differences in aggression and social rank in the Midas cichlid. Ethology, 86, 311–325. Fraser, D.F., Gilliam J.F., Daley, M.J., Le A.N. & Skalski, G.T. (2001) Explaining leptokurtic movement distributions: intrapopulation variation in boldness and exploration. American Naturalist, 158, 124–135. Frost, A.J., Winrow-Giffen, A., Ashley, P.J. & Sneddon, L.U. (2007) Plasticity in animal personality traits: does prior experience alter the degree of boldness? Proceedings of the Royal Society B, 274, 333–339.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

161

Funder, D.C. (1995) On the accuracy of personality judgement: a realistic approach. Psychological Review, 102, 652–670. Funder, D.C. (2009) Persons, behaviors and situations: an agenda for personality psychology in the postwar era. Journal of Research in Personality, 43, 120–126. Godin, J.G.J. & Davis, S.A. (1995) Boldness and predator deterrence – reply. Proceedings of the Royal Society of London Series B – Biological Sciences, 262, 107–112. Godin, J.G.J. & Dugatkin, L.A. (1996) Female mating preference for bold males in the guppy, Poecilia reticulata. Proceedings of the National Academy of Sciences of the United States of America, 93, 10262–10267. Gosling, S.D. (1998) Personality dimensions in spotted hyenas (Crocuta crocuta). Journal of Comparative Psychology, 112, 107–118. Gosling, S.D. (2001) From mice to men: What can we learn about personality from animal research? Psychological Bulletin, 127, 45–86. Gosling, S.D. & John, O.P. (1999) Personality dimensions in nonhuman animals: a cross-species review. Current Directions in Psychological Science, 8, 69–75. Gosling, S.D., Kwan, V.S.Y. & John, O.P. (2003) A dog’s got personality: a cross-species comparative approach to personality judgments in dogs and humans. Journal of Personality and Social Psychology, 85, 1161–1169. Gray, J.A. (1982) A critique of Eysenck’s theory of personality. In: H.J. Eysenck (ed) A Model for Personality, pp. 246–276. Springer, New York. Gray, J.A. (1987) The Psychology of Fear and Stress. Cambridge University Press, Cambridge. Grimm, V. & Uchma´nski, J. (2002) Individual variability and population regulation: a model of the significance of within-generation density dependence. Oecologia, 131, 196– 202. Groothuis, T.G.G. & Carere, C. (2005) Avian personalities: characterization and epigenesis. Neuroscience and Biobehavioral Reviews, 29, 137–150. Harwood, A.J., Armstrong, J.D., Metcalfe, N.B. & Griffiths, S.W. (2003) Does dominance status correlate with growth in wild stream-dwelling Atlantic salmon (Salmo salar)? Behavioral Ecology, 14, 902–908. Hebb, D.O. (1946) Emotion in man and animal: an analysis of the intuitive processes of recognition. Psychological Review, 53, 88–106. Helfman, G.S. (1984) Schooling fidelity in fishes: the yellow perch pattern. Animal Behaviour, 22, 663–672. Hodos, C.B.G. & Campbell, W. (1991) The Scala Naturae revisited: evolutionary scales and anagenesis in comparative psychology. Journal of Comparative Psychology, 105, 211–221. Huntingford, F.A. (1976) The relationship between anti-predator behaviour and aggression among conspecifics in the three-spined stickleback, Gasterosteus aculeatus. Animal Behaviour, 24, 245–260. Huntingford, F.A. (2004) Implications of domestication and rearing conditions for the behaviour of cultivated fishes. Journal of Fish Biology, 65, 122–142. Huntingford, F. & Adams, C. (2005) Behavioural syndromes in farmed fish: implications for production and welfare. Behaviour, 142, 1207–1221. Huntingford, F.A., Andrew, G., Mackenzie, S., Morera, D., Coyle, S.M., Pilarczyk, M. & Kadri, S. (2010) Coping strategies in a strongly schooling fish, the common carp Cyprinus carpio. Journal of Fish Biology, 76, 1576–1591. Johnsson, J.I., Sernland, E. & Blixt, M. (2001) Sex-specific aggression and antipredator behaviour in young brown trout. Ethology, 107, 587–599. Kaplan, R.H. & Cooper, W.S. (1984) The evolution of developmental plasticity in reproductive characteristics: an application of the “adaptive coin-flipping” principle. American Naturalist, 123, 393–410. Kenrick, D.T. & Funder, D.C. (1988) Lessons from the person-situation debate. American Psychologist, 43, 23–34. King, J.E. & Figueredo, A.J. (1997) The five-factor model plus dominance in chimpanzee personality. Journal of Research in Personality, 31, 257–271.

BLBK374-07

BLBK374-Brown

162

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

King, J.E., Weiss, A. & Farmer, K.H. (2005) A chimpanzee (Pan troglodytes) analogue of crossnational generalization of personality structure: zoological parks and an African sanctuary. Journal of Personality, 73, 389–410. Kobler, A., Engelen, B., Knaepkens, G. & Eens, M. (2009) Temperament in bullheads: do laboratory and field explorative behaviour variables correlate? Naturwissenschaften, 96, 1229– 1233. Koolhaas, J.M., Korte, S.M., De Boer, S.F., Van Der Vegt, B.J., Van Reenen, C.G., Hopster, H., De Jong, I.C., Ruis, M.A.W. & Blokhuis, H.J. (1999) Coping styles in animals: current status in behavior and stress-physiology. Neuroscience and Biobehaioral Reviews, 23, 925–935. Krause, J., Loader, S.P., Kirkman, E. & Ruxton, G.D. (1999) Refuge use by fish as a function of body weight changes. Acta Ethologica, 2, 29–34. Langlet, O. (1971) Revising some terms of intra-specific differentiation. Hereditas, 68, 277– 280. Lautin, A. (2002) The Limbic Brain. Kluwer, New York. Leblond, C. & Reebs, S. (2006) Individual leadership and boldness in shoals of golden shiners (Notemigonus crysoleucas). Behaviour, 143, 1263–1280. Lepage, O., Overli, O., Petersson, E., Jarvi, T. & Winberg, S. (2000) Differential stress coping in wild and domesticated sea trout. Brain Behavior and Evolution, 56, 259–268. Livesley, W.J., Jang, K.L. & Vernon, P.A. (1998) Phenotypic and genetic structure of traits delineating personality disorder. Archives of General Psychiatry, 55, 941–948. Łomnicki, A. (1988) Population Ecology of Individuals. Princeton University Press, Princeton, NJ. Lorenz, K.Z. (1974) Analogy as a source of knowledge. Science, 185, 229–234. Loughry, W.L. & Lazari, A. (1994) The ontogeny of individuality in black-tailed prairie dogs, Cynomys ludovicianus. Canadian Journal of Zoology, 72, 1280–1286. Magnhagen, C. (2007) Social influence on the correlation between behaviours in young-of-the-year perch. Behavioral Ecology and Sociobiology, 61, 525–531. Magnhagen, C. & Borcherding, J. (2008) Risk-taking behaviour in foraging perch: does predation pressure influence age-specific boldness? Animal Behaviour, 75, 509–517. Magnhagen, C. & Staffan, F. (2005) Is boldness affected by group composition in young-of-the-year perch (Perca fluviatilis)? Behavioral Ecology and Sociobiology, 57, 295–303. Magurran, A.E. (1993) Individual differences and alternative behaviours. In: T.J. Pitcher (ed) The Behaviour of Teleost Fishes, pp. 441–477. Chapman & Hall, London. Mangel, M. & Clark, C.W. (1988) Dynamic Modeling in Behavioral Ecology. Princeton University Press, Princeton, NJ. Martin, J.G.A., Nussey, D., Wilson, A. & R´eale, D. (in press). Measuring individual differences in reaction norms in field and experimental studies: a power analysis of random regression models. Methods in Ecology and Evolution, doi: 10.1111/j.2041-210X.2010.00084.x. Maynard Smith, J. (1982) Evolution and the Theory of Games. Cambridge University Press, Cambridge. McAdams, D.P. (1992) The Five-Factor model in personality: a critical appraisal. Journal of Personality, 60, 329–361. McElreath, R. & Strimling, P. (2006) How noisy information and individual asymmetries can make ‘personality’ an adaptation: a simple model. Animal Behaviour, 72, 1135–1139. McGinnity, P., Jennings, E., deEyto, E., Allott, N., Samuelsson, P., Rogan, G., Whelan, K. & Cross, T. (2009) Impact of naturally spawning captive-bred Atlantic salmon on wild populations: depressed recruitment and increased risk of climate-mediated extinction. Proceedings of the Royal Society B, 276, 3601–3610. McGinnity, P., Prod¨ohl, P., Ferguson, A., Hynes, R., Maoil´eidigh, N.O., Baker, N., Cotter, D., O’Hea, B., Cooke, D., Rogan, G., Taggart, J. & Crosset, T. (2003) Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon. Proceedings of the Royal Society B, 270, 2443–2450. Michener, C.D. & Sokal, R.R. (1957) A quantitative approach to the problem of classification. Evolution, 11, 130–162.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

163

Millot, S., Begout, M.L. & Chatain, B. (2009) Risk-taking behaviour variation over time in sea bass Dicentrarchus labrax: effects of day-night alternation, fish phenotypic characteristics and selection for growth. Journal of Fish Biology, 75, 1733–1749. Mischel, W. (1973) Toward a cognitive social learning reconceptualization of personality. Psychological Review, 80, 252–283. Moreira, P.S.A., Pulman, K.G.T. & Pottinger, T.G. (2004) Extinction of a conditioned response in rainbow trout selected for high or low responsiveness to stress. Hormones and Behavior, 46, 450–457. Moretz, J.A., Martins, E.P. & Robison, B.D. (2007) Behavioral syndromes and the evolution of correlated behavior in zebrafish. Behavioral Ecology, 18, 556–562. Myers, R.A. (2004) Hatcheries and endangered salmon. Science, 303, 1980. Nunnally, J.C. (1967) Psychometric Theory. McGraw-Hill, New York. Ossenkopp, K.-P. & Mazmanian, D.S. (1985) The principle of aggregation in psychobiological correlational research: an example from the open-field test. Animal Learning and Behavior, 13, 339–344. Øverli, Ø., Pottinger, T.G., Carrick, T.R., Øverli, E. & Winberg, S. (2001) Brain monoaminergic activity in rainbow trout selected for high and low stress responsiveness. Brain Behavior and Evolution, 57, 214–224. Øverli, Ø., Pottinger, T.G., Carrick, T.R., Øverli, E. & Winberg, S. (2002) Differences in behaviour between rainbow trout selected for high- and low-stress responsiveness. Journal of Experimental Biology, 205, 391–395. Øverli, Ø., Sørensen, C., Pulman, K.G.T., Pottinger, T.G., Korzan, W., Summers, C.H. & Nilsson, G.E. (2007) Evolutionary background for stress-coping styles: relationships between physiological, behavioral, and cognitive traits in non-mammalian vertebrates. Neuroscience and Biobehavioral Reviews, 31, 396–412. Øverli, Ø., Winberg, S. & Pottinger, T.G. (2005) Behavioral and neuroendocrine correlates of selection for stress responsiveness in rainbow trout – a review. Integrative and Comparative Biology, 45, 463–474. Ozer, D.J. (1986) Consistency in Personality: A Methodological Framework. Springer-Verlag, New York. Petrovskii, S.V. & Blackshaw, R.P. (2003) Behaviourally structured populations persist longer under harsh environmental conditions. Ecology Letters, 6, 455–462. Petrovskii, S.V., Blackshaw, R.P. & Li, B.L. (2008) Consequences of the Allee effect and intraspecific competition on population persistence under adverse environmental conditions. Bulletin of Mathematical Biology, 70, 412–437. Pitcher, T.J. & Parrish, J.K. (1993) Functions of schooling behaviour in teleosts. In: T.J. Pitcher (ed) The Behaviour of Teleost Fishes, pp. 363–439. Chapman, Hall, London. Piyapong, C., Krause, J., Chapman, B.B., Ramnarine, I.W., Louca, V. & Croft, D.P. (2010) Sex matters: A social context to boldness in guppies (Poecilia reticulata). Behavioral Ecology, 21, 3–8. Portavella, M. & Vargas, J.P. (2005) Emotional and spatial learning in goldfish is dependent on different telencephalic pallial systems. European Journal of Neuroscience, 21, 2800– 2806. Pottinger, T.G. & Carrick, T.R. (2001) Stress responsiveness affects dominant-subordinate relationships in rainbow trout. Hormones and Behavior, 40, 419–427. Pruessner, J.C., Gaab, J., Hellhammer, D.H., Lintz, D., Schommer, N. & Kirschbaum, C. (1997) Increasing correlations between personality traits and cortisol stress responses obtained by data aggregation. Psychoneuroendocrinology, 22, 615–625. Radakov, D.V. (1972) Schooling in the Ecology of Fishes. Nauka, Moscow. Rands, S.A., Cowlishaw, G., Pettifor, R.A., Rowcliffe, J.M. & Johnstone, R.A. (2003) Spontaneous emergence of leaders and followers in foraging pairs. Nature, 423, 432–434. Raoult, V., Brown, C., Zuberi, A. & Williamson, J.E. (2010) Personality traits are linked to stress hormones in a commercial marine fish species (Argyosomus japonicus). Biology Letters. (In press.)

BLBK374-07

BLBK374-Brown

164

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

R´eale, D., Reader, S.M., Sol, D., McDougall, P.T. & Dingemanse, N.J. (2007) Integrating animal temperament within ecology and evolution. Biological Reviews, 82, 291–318. Revelle, W. (1983) Factors are fictions, and other comments on individuality theory. Journal of Personality, 51, 707–714. Ringler, N.H. (1983) Variation in foraging tactics of fishes. In: D.L.G. Noakes (ed) Predators and Prey in Fishes, pp. 159–172. Junk Publishers, The Hague. Roberts, B.W. & DelVecchio, W.F. (2000) The rank-order consistency of personality traits from childhood to old age: a quantitative review of longitudinal studies. Psychological Bulletin, 126, 3–25. Roff, D.A. (1996) The evolution of genetic correlations: an analysis of patterns. Evolution, 50, 1392–1403. Rowland, W.J. (1982) The effects of male coloration on stickleback aggression: a reexamination. Behaviour, 80, 118–126. Rushton, J.P., Bons, T.A. & Hur, Y.M. (2008) The genetics and evolution of the general factor of personality. Journal of Research in Personality, 42, 1173–1185. Rushton, J.P., Brainerd, C.J. & Pressley, M. (1983) Behavioral development and construct validity: the principle of aggregation. Psychological Bulletin, 94, 18–38. Saetre, P., Strandberg, E., Sundgren, P.-E., Pettersson, U., Jazin, E. & Bergstrom, T.F. (2006) The genetic contribution to canine personality. Genes, Brain and Behavior, 5, 240–248. Schjolden, J., Stoskhus, A. & Winberg, S. (2005) Does individual variation in stress responses and agonistic behavior reflect divergent stress coping strategies in juvenile rainbow trout? Physiological and Biochemical Zoology, 78, 715–723. Seaman, S.C., Davidson, H.P.B. & Waran, N.K. (2002) How reliable is temperament assessment in the domestic horse (Equus caballus)? Applied Animal Behaviour Science, 78, 175–191. Shaklee, A.B. (1963) Comparative studies of temperament – fear responses in different species of fish. Journal of Genetic Psychology, 102, 295–310. Shepard, P.D., Holcomb, H.H. & Gold, J.M. (2006) The presence of absence: habenular regulation of dopamine neurons and the encoding of negative outcomes. Schizophrenia Bulletin, 32, 417– 421. Sih, A., Bell, A., Johnson, J.C. & Ziemba, R.E. (2004) Behavioral syndromes: an integrative overview. Quarterly Review of Biology, 79, 241–277. Slater, P.J.B. (1981) Individual differences in animal behaviour. In: P.P.G. Bateson & P.H. Klopfer (eds) Perspectives in Ethology Journal of Genetic Psychology, Vol. 4, pp. 35–49. Plenum Press, New York. Smith, T.B. & Skulason, S. (1996) Evolutionary significance of resource polymorphisms in fishes, amphibians. Annual Review of Ecology and Systematics, 27, 111–133. Sneddon, L.U. (2003) The bold and the shy: individual differences in rainbow trout. Journal of Fish Biology, 62, 971–975. Staffan, F., Magnhagen, C. & Alanara, A. (2005) Individual feeding success of juvenile perch is consistent over time in aquaria and under farming conditions. Journal of Fish Biology, 66, 798– 809. Stamps, J.A. (2007) Growth-mortality tradeoffs and ‘personality traits’ in animals. Ecology Letters, 10, 355–363. Stevenson-Hinde, J., Stillwell-Barns, R. & Zunz, M. (1980) Subjective assessment of rhesus monkeys over four successive years. Primates, 21, 66–82. Sundstr¨om, L.F., L˜ohmus, M. & Johnsson, J.I. (2003) Investment in territorial defence depends on rearing environment in brown trout (Salmo trutta). Behavioral Ecology and Sociobiology, 54, 249–255. Sundstr¨om, F.L., Petersson, E., H¨ojesj¨o, J., Johnsson, J.I. & J¨arvi, T. (2004) Hatchery selection promotes boldness in newly hatched brown trout (Salmo trutta): implications for dominance. Behavioral Ecology, 15, 192–198. Sutherland, R.J. (1982) The dorsal diencephalic conduction system: a review of the anatomy and functions of the habenular complex. Neuroscience and Biobehavioral Reviews, 6, 1–13.

BLBK374-07

BLBK374-Brown

May 13, 2011

17:9

Copyeditor’s Name:

Trim: 244mm X 172mm

Personality Traits and Behaviour

Char Count=

165

Svartberg, K. (2002) Shyness-boldness predicts performance in working dogs. Applied Animal Behaviour Science, 79, 157–174. Tinbergen, N. (1963) On aims and methods of ethology. Zeitschrift f¨ur Tierpsychologie, 20, 410–433. Uchma´nski, J. (2000) Individual variability and population regulation: an individual-based model. Oikos, 90, 539–548. Uher, J. & Asendorpf, J.B. (2008) Personality assessment in the Great Apes: comparing ecologically valid behaviour measures, behaviour ratings, and adjective ratings. Journal of Research in Personality, 42, 821–838. van Oers, K., de Jong, G., Drent, P.J. & van Noordwijk, A.J. (2004) A genetic analysis of avian personality traits: correlated, response to artificial selection. Behavior Genetics, 34, 611–619. van Oers, K., Klunder, M. & Drent, P.J. (2005) Context dependence of personalities: risk-taking behavior in a social and a nonsocial situation. Behavioral Ecology, 16, 716–723. Vazire, S., Gosling, S.D., Dickey, A.S. & Schaprio, S.J. (2007) Measuring personality in nonhuman animals. In: R.W. Robins, R.C. Fraley & R.F. Krueger (eds) Handbook of Research Methods in Personality Psychology, pp. 190–206. Guilford, New York. Ward, A.J.W., Thomas, P., Hart, P.J.B. & Krause, J. (2004) Correlates of boldness in three-spined sticklebacks (Gasterosteus aculeatus). Behavioral Ecology and Sociobiology, 55, 561–568. Westerberg, M., Staffan, F. & Magnhagen, C. (2004) Influence of predation risk on individual competitive ability and growth in Eurasian perch, Perca fluviatilis. Animal Behaviour, 67, 273–279. Wilson, A.D.M. & Godin, J.-G. (2009) Boldness and behavioral syndromes in the bluegill sunfish, Lepomis macrochirus. Behavioral Ecology, 20, 231–237. Wilson, A.D.M. & Stevens, E.D. (2005) Consistency in context-specific measures of shyness and boldness in rainbow trout, Oncorhynchus mykiss. Ethology, 111, 849–862. Wilson, D.S. (1998) Adaptive individual differences within single populations. Philosophical Transactions of the Royal Society of London B, 353, 199–205. Wilson, D.S., Clark, A.B., Coleman, K. & Dearstyne, T. (1994) Shyness and boldness in humans and other animals. Trends in Ecology and Evolution, 9, 442–446. Wilson, D.S., Coleman, K., Clark, A.B. & Biederman, L. (1993) Shy-bold continuum in pumpkinseed sunfish (Lepomis gibbosus): An ecological study of a psychological trait. Journal of Comparative Psychology, 107, 250–260. Wilson, M. & Daly, M. (1985) Competitiveness, risk-taking, and violence: the young male syndrome, Ethology and Sociobiology, 6, 59–73. Wolf, M., van Doorn, G.S., Leimar, O. & Weissing, F.J. (2007) Life-history trade-offs favour the evolution of animal personalities. Nature, 447, 581–585. Wolf, M., van Doorn, G.S. & Weissing, F.J. (2008) Evolutionary emergence of responsive and unresponsive personalities. Proceedings of the National Academy of Sciences of the USA, 105, 15825–15830. ¨ (2006) Epistatic regulation of behavioural and morphological Wright, D., Butlin, R. & Carlborg, O. traits in the zebrafish (Danio rerio). Behavior Genetics, 36, 914–922. Wright, D., Rimmer, L.B., Pritchard, V.L., Krause, J. & Butlin, R.K. (2003) Inter and intra-population variation in shoaling and boldness in the zebrafish (Danio rerio). Naturwissenschaften, 90, 374–377. Yoshida, M., Nagamine, M. & Uematsu, K. (2005) Comparison of behavioral responses to a novel environment between three teleosts, bluegill Lepomis macrochirus, crucian carp Carassius langsdorfii, and goldfish Carassius auratus. Fisheries Science, 71, 314–319. Zuckerman, M. (1994) Psychobiology of Personality. Cambridge University Press, Cambridge. Zuckerman, M., Kuhlman, D.M., Joireman, J., Teta, P. & Kraft, M. (1993) A comparison of three structural models for personality: The Big Three, the Big Five, and the Alternative Five. Journal of Personality and Social Psychology, 65, 757–768.

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 8

The Role of Learning in Fish Orientation Lucy Odling-Smee, Stephen D. Simpson and Victoria A. Braithwaite

8.1

Introduction

Until recently, the possibility that fishes are capable of flexible, learned orientation strategies has generally been overlooked. However, in most aquatic environments, the physical landscape and the need to relocate various biologically important locations (e.g. shelter or the position of a profitable food patch) should favour an ability to learn. Here, we review the evidence from both field and laboratory studies that fishes can and do use learning and memory to orientate within their natural environments. Comparisons between different species and populations indicate that fishes have a diverse array of sensory systems that they use to encode spatial information; however, what they learn is sometimes constrained by their environment, and/or their genotype. Much of our knowledge on animal orientation has typically come from terrestrial species such as birds and mammals (Dodson 1988; Healy 1998). However, miniaturisation of tracking devices and other telemetry techniques are now allowing us to investigate the movements and spatial behaviours exhibited by fishes (Armstrong et al. 1997; Hunter et al. 2003). Within the aquatic environment, fishes are exposed to an enormous diversity of potential cues from which they can extract information about their spatial location. Equally diverse is the range of spatial and navigational problems that may be encountered by different species, populations, individuals and even different developmental stages. Our aim in this chapter is not only to show that fishes are capable of orienting (i.e. keeping track of their location with respect to an external point of reference) using learning and memory, but also to convey that such flexibility is often constrained or modified by mechanisms that guide learning and associated perceptual processes in response to particular ecological conditions and navigational demands.

8.2

Why keep track of location?

Many behaviours, including foraging, reproduction, competition and predator avoidance, are dependent on a fish being able to plan and execute a route to guide it efficiently between Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

167

two locations. To achieve this, the fish needs to monitor its location with reference to an external point of reference as it moves through its environment. Resources are often widely separated in space or time such that individuals cannot rely purely on chance encounters to fulfil their energetic or reproductive requirements. Food is often distributed among sites that vary spatially and temporally in profitability and food patches may differ in the likelihood of renewal after depletion (Warburton 2003; Chapter 2). As long as there is some degree of predictability, foraging efficiency will increase if fishes can map the status and renewal rates of individual food patches onto their location (Noda et al. 1994; Hughes & Blight 1999, 2000) and use this information to guide their foraging. Spatial information may similarly be used to predict the location of receptive mates. For example, in certain reef fish, spawning aggregations draw fish to specific locations from disparate areas of the reef (Mazeroll & Montgomery 1998). Several species of fishes guard nest sites or other resources from competitors, displaying territoriality or a tendency to remain in a restricted area or ‘home range’ (Hallacher 1984; Matthews 1990a, 1990b; Kroon et al. 2000). Such familiarity with a particular area may allow the fishes to return to their territory, or re-establish territory boundaries if they become temporarily displaced (Hallacher 1984; Matthews 1990a, 1990b). Furthermore, spatial information may be used to pinpoint the location of shelters, or hiding places within a home range, allowing rapid escape in the event of an attack by a predator (Aronson 1951, 1971; Markel 1994). In addition to keeping track of site-specific resources, fishes may need to monitor locations associated with risk. Some predators are associated with particular microhabitats or locations that are best avoided by potential prey (Goodyear 1973; Jordan et al. 1997; Brown 2003). Others have predictable movements. For example, many reef-based piscivores concentrate on coral reefs at night and move away from the reefs during the day (Mazeroll & Montgomery 1995). Predator avoidance strategies may therefore involve daily migrations requiring fishes to keep track of their location with respect to both feeding areas and protective refuges (Ogden & Buckman 1973; Ogden & Quinn 1984; Mazeroll & Montgomery 1995). Moreover, predators themselves may need to be equipped with information about the spatial structure of their home range in order to avoid being recognised by potential prey (see Brown 2003). Some species of fishes return to their natal area for reproduction, otherwise known as ‘homing’ (Tsukamoto et al. 2003). Homing is likely to enhance reproductive success by allowing mature animals to return to their spawning grounds when conditions are optimal for egg and larval development. In addition, reproductive isolation achieved through homing may facilitate the development of adaptations specific to the particular habitat occupied (Dittman & Quinn 1996). In order to achieve accurate return to their natal sites, fishes must relocate their natal area from impressive distances, in some cases thousands of kilometres (Quinn & Dittman 1990; Dittman & Quinn 1996).

8.3

The use of learning and memory in orientation

Compared to terrestrial groups, considerably less is known about how, or indeed whether, fishes use learning and memory in spatial orientation (Healy 1998). Observations of fish movements within their natural habitats suggest impressive spatial abilities without

BLBK374-08

BLBK374-Brown

168

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

providing hard evidence for the role of learning and memory, or revealing what exactly might be learned. Various studies on fish orientation have used displacement of marked individuals over various distances from the place they were caught to investigate their capacity to home (e.g. Green 1971; Carlson & Haight 1972; Hallacher 1984; Quinn & Ogden 1984; Kolm et al. 2005). In the Banggai cardinal fish, homing from 50 m was found to be in direct response to the landscape, rather than to conspecifics in the colony (Kolm et al. 2005). Some species return home from as far as 22.5 km (Carlson & Haight 1972) or after 6 months in captivity (Green 1971), consistent with the suggestion that they use long-term spatial memory. These observations could, however, also be explained by the fish using reactive mechanisms to home, such as tracking changes in temperature or light levels. To track the search behaviour of a planktivorous reef fish, Chromis chrysurus, Noda et al. (1994) marked individual fish with acrylic paint and followed their movements after release. Individual fish repeatedly visited three distinct foraging areas, swimming slowly and in a stereotypic pattern within each foraging site before swiftly moving off to the next patch. Noda and colleagues suggested that spatial memory may allow the fish to concentrate foraging at relatively high densities of zooplankton, and avoid revisiting depleted areas. Again, alternative explanations cannot be ruled out. For example, C. chrysurus may locate food patches simply by responding to olfactory cues. We owe most of our knowledge about the role of learning and memory in fish orientation and the types of spatial information used, to controlled experiments. These involve manipulations of fish sensory systems or spatial cues, or laboratory based spatial tasks where fishes are trained to learn particular associations. Such approaches have revealed that fish can learn and remember routes and locations by using information from a variety of different sources, ranging from landmarks and compasses to spatially ‘informed’ conspecifics.

8.4

Learning about landmarks

In most aquatic landscapes, local features (such as visual or olfactory cues) are likely to be changeable components of the environment requiring that they be stored in memory and updated on the basis of experience. To support this, several species of fishes have been successfully trained to use landmark information to solve a range of laboratory-based spatial tasks (Huntingford & Wright 1989; Warburton 1990; Braithwaite et al. 1996; Salas et al. 1996a; Girvan & Braithwaite 1998; L´opez et al. 1999, 2000a, 2000b; Hughes & Blight 2000; Odling-Smee & Braithwaite 2003). However, aquatic environments are very diverse; there are several examples of fish species that still manage to orient even when their visual environment is not heterogeneous (e.g. Cain et al. 1994; Burt de Perera 2004a). This indicates that certain species of fishes are capable of learning about their spatial environment using non-visual senses. Less is know about the extent to which fishes use landmarks in the wild. One fieldbased approach has been to manipulate the position of landmarks in the natural migrating paths of fishes. Reese (1989) observed that butterflyfish (Chaetodontidae) spend some time searching in an area where coral heads have been removed, before continuing along their original foraging path. Migrating brown surgeonfish, Acanthurus nigrofuscus, similarly change direction in accordance with the new positions of displaced landmarks (Mazeroll & Montgomery 1998). Moreover, their reliance on particular landmarks is reduced when

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

169

landmarks are moved more than 6 m from their original location, suggesting that brown surgeonfish may be capable of assessing the stability and so reliability of visual landmarks (see Biegler & Morris 1996). However, the disruption of migratory routes by shifted landmarks does not prove landmark cues were originally learned to guide orientation. Shifting landmarks could disturb the sediment and increase food availability, promoting exploratory behaviour. Alternatively, fishes may simply be reacting to a change in their local environment. One successful approach to determining how fishes learn and use landmarks has been to combine laboratory and field work. In laboratory experiments, potentially confounding effects can be more easily controlled. For example, fish can be trained to use specific landmarks and then additional cues, such as compass or global place cues, can be manipulated or completely removed (by screening or rotating test tanks) to determine whether the fish used this additional information. The results from these trials can then be used to design and test the spatial cues that the fishes are using in a more natural setting (Braithwaite et al. 1996; Armstrong et al. 1997; Huntingford et al. 1998). In addition to using landmarks as direct cues or beacons, fishes appear to be capable of learning more complex spatial relationships. Goldfish (Carassius auratus, Cyprinidae) can locate a food reward by using landmarks as indirect reference points (Warburton 1990) and remember the spatial positions of three differently rewarded, hidden food patches in a tank (Pitcher & Magurran 1983). They can also learn to use visual cues to locate a goal, even if they approach the goals from a novel direction (Ingle & Sahagian 1973; Warburton 1990; Rodr´ıguez et al. 1994). This suggests a capacity to discriminate spatial relationships independently of any particular view of the surroundings, enabling the fishes to take shortcuts or choose between alternative routes to a goal. The level of accuracy achieved by Siamese fighting fish, Betta splendens, in an eight-arm radial maze similarly suggests that some amount of spatial memory is involved in recognising which of the eight arms have already been depleted of the food reward (Roitblat et al. 1982). The fact that fishes can detect environmental modifications (Welker & Welker 1958) and show an organised pattern of exploration when they are introduced into a novel environment (Kleerokoper et al. 1974) also suggests some degree of spatial memory. A classic demonstration of the ability of fishes to use spatial maps is provided by experiments on the gobiid fish, Bathygobius soporator (Aronson 1951, 1971). When threatened, gobiids jump from their home tide-pool to an adjacent pool with impressive accuracy. If required, the gobies can make a series of jumps leading them from one pool to the next and eventually into the open sea. In order to investigate whether gobiids acquire memories of the local topography around their home pools, Aronson constructed three artificial pools and manipulated the water level to simulate low and high tides (Fig. 8.1). Fishes that were given experience of the spatial distribution of the pools at ‘high tide’ successfully escaped a simulated attack at ‘low tide’ by jumping into the appropriate pool. There is a clear need for accuracy in this spatial task; a jump in the wrong direction could be the last one that a fish makes. It is surprising that such a simple model of spatial learning has not been investigated further. What types of landmark do the fishes use to recognise which particular pool they are in – local or global? Do the fishes remember a sequence of jumps, or are they able to make an appropriate escape response based on where they are at that point in time? For example, if they were moved to another pool would they be able to make the correct jump from this displaced position?

BLBK374-08

BLBK374-Brown

170

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Main pool

M

Water intake

High wall 30°

d

30°

Hi

gh

Sand

wa

ll

Pool

Drain

C 30° 30° 30°

d

Sand

dge

Pool

30°

h le

d

Hig

h Hig

ge led

A Pool

B

Fig. 8.1 Topography of tide pools used to test spatial learning ability by the gobiid fish, Bathybogius soporator (adapted from Aronson 1971). Fishes placed in the smaller pools A and B were stimulated to jump by being prodded with a stick. Trials were terminated when the fish had reached the main pool.

Markel (1994) used another species of goby to investigate the use of landmarks to find shelter when threatened with a simulated predator attack. Here, the time that blackeye gobies, Coryphopterus nicholsi, took to locate an artificial burrow was recorded. Fishes given additional exploration time in the test tank were quicker at relocating their burrow compared to fishes deprived of this additional experience. When the burrow was shifted to a new location, the experienced group took longer than the less experienced group to find the shifted burrow, suggesting that the fishes had learned and remembered the spatial location of the burrow, as opposed to its appearance. Where vision is precluded by lifestyle or habitat conditions, information about landmarks may be acquired though alternative sensory channels. The nocturnal African mormyrid, Gnathonemus petersii, can use electrolocation to learn to locate an opening or aperture in a wall (Cain et al. 1994; Cain 1995). Once the fishes have learned the position of an opening, they are able to locate it even after their electric organ has been denervated, suggesting that they develop a stored, internal representation of their environment. The detail that the fishes can perceive with their electroreceptors is impressive. For example, G. petersii and another species of electric fish Sternopygus macrurus can recognise the three-dimensional orientations and configurations of landmarks (Graff et al. 2004).

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

171

Blind Mexican cave fish (Astyanax fasciatus) also show some impressive spatial learning abilities. Eye formation in these fishes arrests during development, forcing them to use alternative sensory information to learn about the spatial organisation of their environment. The fishes use their lateral line organ (LLO) to provide them with topographical information. As they swim the fishes detect water displacement patterns with their LLO (Campenhausen et al. 1981). When a fish moves forwards, it displaces water and the lateral line detects small differences in water flow patterns as the displaced water is reflected off objects within the environment. The cave fish move slowly around their environment as they explore a new area, but then increase their swimming velocity once they have learned the position of objects (Teyke 1985, 1989). Teyke (1989) suggests that these fishes develop an internal map that allows them to swim at velocities below this optimal range, once they are familiarised with the environment (but see Bennet 1996). In support of this, A. fasciatus have been shown to encode spatial information about configurations of landmarks, including information about their order or sequence, and their three-dimensional relationships (Burt de Perera 2004a, 2004b; Burt de Perera et al. 2005). The way in which A. fasciatus learns about new landmarks suggests a lateralization in the way its brain processes spatial cues. For example, Burt de Perera & Braithwaite (2005) found that when the fishes were confronted with a new landmark they would preferentially swim past the landmark using the right side of the LLO. This right-side preference persisted whether they swam in a clockwise or anticlockwise direction. However, the preference waned once the fish became familiar with the landmark. The lateralized use of the right LLO may be related to use of the left hemisphere of the brain, because the neurons associated with the lateral line pass bilaterally from the hindbrain nucleus to the midbrain with contralateral predominance (McCormick 1989). An alternative explanation may be that the fishes react to novelty using their right LLO (Brown et al. 2004). Olfactory cues present an additional type of landmark information, which can be used to relocate home areas or natal streams. Perhaps most acclaimed is the ability of mature salmon to relocate their natal streams based on its unique olfactory composition, a feat described in more detail in Section 8.5. Another sensory modality that has largely been overlooked but nevertheless provides orientation information is sound. There is growing evidence that acoustic cues are highly important in the orientation of settlement-stage coral reef fishes when they return to reefs following a pelagic larval phase (Montgomery et al. 2006). Studies using light traps coupled with sound systems have shown that temperate (Tolimieri et al. 2000) and tropical (Leis et al. 2003; Simpson et al. 2004) reef fishes are attracted to reef sounds at settlement. Further, Simpson et al. (2005a) have demonstrated that this attraction is used by larvae at settlement to locate suitable habitat, demonstrating the ability in larvae, both to orient towards and to localise source sounds.

8.5

Compass orientation

In common with many terrestrial animals (e.g. Able 1993), when learning about locations and routes, fishes are likely to rely on more than one source of spatial information. The use of multiple cues provides back-up points of reference if changes in the environment make one cue unavailable or unreliable. Compasses provide another

BLBK374-08

BLBK374-Brown

172

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

source of spatial information, which may often be used in combination with landmarks (Goodyear 1973). Few studies have investigated the role of learning and memory in compass orientation. Early attempts demonstrated that novel sun-compass responses could be learned by training fishes under a clear sky, to seek cover in one of a number of circularly arranged refuges (Hasler et al. 1958; Schwassmann & Braemer 1961; Hasler 1971). Goodyear’s (1973) study on mosquitofish, Gambusia affinis, goes a step further by demonstrating the possible adaptive significance of having a flexible compass response (see also Goodyear & Bennet 1979). The directional preferences displayed by mosquitofish, when placed in test arenas under an artificial ‘sun’, suggested that they use a sun compass to move on a course perpendicular to the shore from which they were collected. This movement towards shallow water seems principally a mechanism for predator avoidance since it is absent in mosquitofish captured from predator-free environments (Goodyear & Ferguson 1969). Mosquitofish captured from a pond containing predators and from a predator-free pond were able to learn a new shoreward direction, after being conditioned to avoid a predatory largemouth bass, Micropterus salmoides (Centrachidae), under a clear sky in tanks with a sloped gravel bottom simulating an emergent shoreline (Goodyear 1973). In this case, a modifiable compass response may allow adaptation to changes in predation risk or to shifts in growth of protective vegetation. In addition to the sun, other cues may be available to fishes for use in compass navigation, such as polarised light and electromagnetic cues. Evidence that fishes learn to orient to polarised light is scant (e.g. Hawryshyn et al. 1990), although it is a likely candidate for fishes that perform daily migrations at sunrise and sunset when polarised light patterns are strongest (Quinn & Ogden 1984; Mazeroll & Montgomery 1998). Laboratorybased training procedures have shown that some species of fishes are capable of sensing and responding to magnetic compasses (Kalmijn 1982; Walker 1984; Walker et al. 1997) and possible anatomical sites for magnetic detection have been discovered (Kirschvink et al. 1985; Walker et al. 1988, 1997; Moore et al. 1990). As yet little is known about the extent to which learning and memory are involved in the use of polarised light and magnetic compasses. The use of polarised light and the sun for accurate orientation requires reference to an internal clock as the correct heading depends on the time of day (see Section 8.7).

8.6

Water movements

In habitats where visual and compass cues are unavailable or unreliable, fishes may resort to acquiring spatial information from the body of water that surrounds them. It has already been seen that some species can learn to locate landmarks based on information from water movements occurring between stationary objects in the environment and their own bodies (Teyke 1985, 1989; Burt de Perera 2004a, 2004b). There is also some evidence that three-spine sticklebacks, Gasterosteus aculeatus, can learn to use flow direction as an orientation cue (Girvan & Braithwaite 2000; Braithwaite & Girvan 2003). Tidal streams have been shown to be important in the movements made by migrating plaice (Hunter et al. 2004). Electronic data storage tags have been used to follow the movements of individual fish, and these have shown that some fishes use tidal currents in part of their migration,

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

173

but that the plaice also show other forms of navigation. This suggests that these fishes, which will perform several migrations in a lifetime, may be learning to use the tidal flows to guide their movements to spawning grounds. Other species appear to learn to sense and orient in response to shifts in hydrostatic pressure. For example, once an internal map-like representation is established in a familiar environment using active electrolocation, the electric elephantnose fish appears to orient using hydrostatic pressure cues (Cain 1995). Similarly, blind Mexican cave fish use hydrostatic pressure cues to learn the location of objects in their environment (Burt de Perera et al. 2005). A more widespread phenomenon is the directional preference or rheotaxis shown by newly hatched juveniles of a number of species when exposed to flowing water (Jonsson et al. 1994; Kaya & Jeanes 1995). This response is assumed to develop independently of experience, but a recent study suggests that some species of fishes may learn directional preferences. Smith & Smith (1998) tested the orientation directions of juveniles of two species of amphidromous gobies, Awaous guamensis and Sicyopterus stimpsoni, after exposure to a stream of falling water. Juveniles oriented their migratory activity in the direction corresponding to the upstream direction during the preceding period of water flow, indicating a rapid learning of directional cues. This may allow the fish to maintain oriented movement uprivers, even in areas with no flow and in backwaters where flow is opposite to the general gradient of the river.

8.7

Inertial guidance and internal ‘clocks’

Visual or olfactory landmarks are often transient features of the physical landscape and celestial compass cues will be periodically lost under cloud cover or at night. A ‘back-up’ strategy that may be more resistant to environmental fluctuations is the use of inertial or body-centred information, i.e. inertia-forced and other sensory signals generated by the animal’s own movements. Although it is not clear to what extent fishes use this information when orienting in the wild, a diverse array of species solve maze tasks by learning a bodycentred pattern of movement. For example, when spatial cues are absent, fifteen-spined sticklebacks, Spinachia spinachia, and corkwing wrasse, Crenilabrus melops, improve their foraging efficiency in an eight-arm radial maze by developing the algorithm of visiting every third arm (Hughes & Blight 1999; see also Roitblat et al. 1982). The use of a bodycentred turn response or a sequence of turns has similarly been observed in three-spine sticklebacks and goldfish (Rodr´ıguez et al. 1994; Salas et al. 1996b; Girvan & Braithwaite 1998; Odling-Smee & Braithwaite 2003). For some fishes, food availability or the risk of predation may vary spatio-temporally throughout the day but predictably from day to day. In the absence of reliable cues such as sun height, temperature or light intensity, fishes may learn to place themselves in appropriate locations at specific times of the day by consulting an internal circadian clock (Reebs 1999). In a study on golden shiners, Notemigonus crysoleucas, Reebs (1996) observed time–place learning when fishes were fed in one of two places at specific times of the day. More recently, the inanga, Galaxias maculatus, has been shown to be capable of time–place learning based on food but not on predation risk (Reebs 1999). However, the absence of time–place learning in laboratory tests (Reebs 1996, 1999) is difficult to interpret as fishes

BLBK374-08

BLBK374-Brown

174

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

may fail to learn because of the relatively low cost of being in the wrong place in spacelimited aquaria. More natural set-ups such as outdoor ponds may provide a better test of time–place learning in fishes.

8.8

Social cues

By observing and following the behaviour of ‘informed’ conspecifics, individuals may acquire spatial knowledge while avoiding many of the costs thought to be associated with individual learning, such as making mistakes or wasting time. Evidence that fishes can learn routes and locations by following conspecifics comes from both field- and laboratory-based studies in which na¨ıve fishes are allowed to follow informed ‘demonstrator’ fishes before being tested for route acquisition (Brown & Laland 2003; Chapter 10). Spatial information about the nature of certain sites – the profitability of one patch over another, for example (Coolen et al. 2003) – can also be learned.

8.9

How flexible is orientation behaviour?

Fishes are clearly exposed to an enormous diversity of potential cues from which they can extract information about their spatial location. Across the diverse array of aquatic habitats occupied by fishes, different sources of spatial information will differ in their availability and/or reliability. Equally diverse may be the range of spatial problems encountered. Individuals might therefore be expected to pay preferential attention to those cues that are most reliable within their particular habitats and invest time and energy in learning during developmental stages that require flexibility. In the Subsection 8.9.1, we discuss how learned orientation behaviour differs between species, populations and individuals exposed to different ecological conditions.

8.9.1

When to learn?

Fishes that remain in a restricted area or home range may need to update stored representations of local topography throughout life. However, in species that return to their natal site to breed, particularly those with an anadromous lifestyle (fishes that migrate from salt to freshwater to spawn), sensory contact with the home site may be lost for prolonged periods of life. In these fishes, ‘imprinting’ may be the mechanism by which young learn characteristics of the home site allowing recognition later in life (Morin et al. 1989a). Although the concept of imprinting and the validity of some of the criteria are controversial issues, imprinting is thought to be a specialised type of learning which takes place during a restricted period known as a sensitive period, and results in relatively long-lasting memory (Immelmann & Suomi 1981). In migratory fishes, long-term memory that is resistant to change is likely to be essential if the natal site is to be successfully recognised at the end of the return migration (Dodson 1988). Recent studies have shown that the larval stages of coral reef fishes do not undergo random dispersal. Despite a potentially dispersive larval phase, genetic heterogeneity over

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

175

small spatial scales is often high (Planes 2002), retention of larvae in near-shore waters is prevalent (Swearer et al. 1999), and self-recruitment back to the natal reef can occur (Jones et al. 1999). Imprinting of chemical cues from host anemones on embryonic clownfishes has been observed, where later larval orientation behaviour is related to earlier experiences (Arvedlund 1999). Since many reef fishes spawn their eggs demersally on the reef, there also exists the potential for acoustic imprinting. A study on the embryonic stages of clownfishes found that for several days during development the sounds of coral reefs would be audible prior to hatching (Simpson et al. 2005b). This possibility is the focus of ongoing research.

8.9.2

What to learn?

Control over when to learn may be accompanied by mechanisms that predispose fishes to use specific types of information or to learn certain associations in preference to others. Recent experiments suggest that three-spine sticklebacks originating from ponds and rivers differ in their propensity to use landmarks during spatial learning (Girvan & Braithwaite 1998; Odling-Smee & Braithwaite 2003). Odling-Smee & Braithwaite (2003) trained sticklebacks from five rivers and five ponds to locate a goal in one arm of a T-maze, either by learning a turn direction out of the start box, or by using plant landmarks as signposts indicating the rewarded end (Fig. 8.2). Pond fishes appear to use both turn direction and landmarks, while river fishes show a strong preference for using turn direction. In fast-flowing rivers, the visual landscape is likely to be continually disrupted by flow and turbulence making landmarks unreliable indicators of location. Therefore, even within a species, learned orientation behaviour may be adapted in response to specific habitat conditions. Huntingford & Wright (1989) observed similar population differences in the use of local visual cues by three-spine

Plastic plant landmark

Removable screen

Opaque PVC partition 2 Stimulus shoal Feeder Trap doors 1 Start box Fig. 8.2 Diagrammatic representation of the spatial task used to compare cue use by pond and river threespine sticklebacks. The arrow indicates the correct route a right-trained fish had to take to reach the goal (food and shoal mates). Landmarks were always placed in the rewarded arm. The numbers indicate the sequence of start box positions for a run of three consecutive trials starting at position 1.

BLBK374-08

BLBK374-Brown

176

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

sticklebacks collected from two sites of high- and low-predation risk. Fishes from the highrisk site used local landmark cues to learn an avoidance task more often than fishes from the low-risk site. However, other factors are likely to vary between two sites in addition to predation pressure, which could affect the use of visual cues. Many fishes experience conditions where visual cues do not vary greatly. This is true throughout the lives of cave-dwelling or deep-water species. It is also true for fishes that inhabit ‘blue water’ environments for all or part of their lives. For the pelagic larvae of coral reef fishes, cues that are relevant at spatial scales of hundreds of metres, such as auditory and olfactory signals, are important (Montgomery et al. 2001; Kingsford et al. 2002). Chemical cues can provide information about the environment they emanate from, while biological sounds such as snapping shrimp and fish vocalisations can indicate the proximity and direction of the local reef inhabitants. As such, they make ideal candidates for orientation in blue water. How and whether fishes learn to interpret these cues as they develop in the plankton and so improve their orientation is unknown, but an ability to discriminate between different habitats would provide valuable spatial information both at settlement and in later life.

8.9.3

Spatial learning capacity

Numerous studies have suggested that the telencephalon is involved in spatial learning in fishes (see Salas et al. 2003; Chapter 13). Preliminary evidence suggests that differences in telencephalon morphology may correlate with differences in the ecological demand for spatial learning ability. Male Azorean rock-pool blennies (Parablennius sanguinolentus, Blenniidae) establish nests. In this species, males establish nests in crevices and almost never leave their nest area during the entire breeding season, while females must travel relatively long distances in order to visit different nests and spawn with males. Females may need to retain a spatial map of the area and remember the location of previously visited nest sites, a requirement that may explain why the dorsolateral region of the telencephalon is larger in females (Carneiro et al. 2001). Similarly in cichlids, variation in telencephalon size appears to relate closely to the challenges of spatial, environmental complexity (Kotrschal et al. 1998). Van Staaden et al. (1994) and Huber et al. (1997) examined the brains of 189 species of cichlids from the East African lakes and Madagascar, and found that species living in complex habitats created by shallow rock and vegetation had comparatively large telencephalons than those living in pelagic zones. However, the telecephalon is likely to govern many cognitive abilities from spatial learning, and in none of these studies have spatial learning abilities of fishes been tested. Thus, until the relationship between spatial learning ability and specific features of the telencephalon is better understood, the results of these studies present suggestive, but inconclusive, evidence for a relationship between spatial learning ability and ecological demand. Ecologically driven differences in the demand for spatial learning may explain why two lake-dwelling sympatric species of three-spine stickleback differ in the rate at which they learn a spatial task (Odling-Smee & Braithwaite, unpublished data). ‘Benthic’ and ‘limnetic’ sticklebacks live in reproductive isolation in several lakes in South-western British Columbia (Schluter & McPhail 1992). The benthic species lives in the vegetated littoral zone, feeding predominantly on mud-dwelling invertebrates, while the limnetic

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

177

species lives in a comparatively homogenous environment in terms of spatial structure, out in the open water column, where they feed mainly on plankton. Two independently derived populations of benthics and limnetics from two lakes were trained to locate a goal arm in a T-maze using either plant landmarks or a turn direction out of the start box (Fig. 8.2). Although both species use both types of spatial cues, benthics learn the task almost twice as quickly as limnetics, consistent with the suggestion that pelagic and benthic lifestyles make differential demands on spatial learning ability. However, a classic problem in comparative studies of learning is the possibility that contextual variables rather than differences in learning ability are responsible for species differences in performance (Shettleworth 1993). For example, species may differ in motivation or in adaptability to laboratory conditions. Although attempts were made to minimise this possibility, further experiments are needed for it to be ruled out. Other, ecological variables have also been found to contribute to differences in fish spatial behaviour. For instance, observations from the Panamanian bishop (Brachyrhaphis episcpoi) indicate that the environment the fishes inhabit has a large effect on their spatial learning skills. Fishes from low-predation areas learned the location of a hidden food patch more quickly than those from high-predation areas (Brown & Braithwaite 2005). This result can partly be explained by differences in the strength of cerebral lateralization found in fishes from high- or low-predation sites, which generated a strong turning preference that influenced maze performance (Brown et al. 2004). Additionally, spatial ability may be more important to fishes from the low-predation areas (where the species dominates the fauna) as they will need to use spatial cues to help them maintain territories, whereas in high-predation areas the fishes are displaced into marginal habitats and do not appear to utilise territories. There is enormous potential for future research investigating the relationship between learned orientation strategies in fishes and their habitat ecology. So far, preliminary but suggestive evidence indicates that in fishes, as appears in terrestrial vertebrates (Sherry 1998), spatial learning may be modified or fine-tuned in response to particular ecological conditions. Fishes appear predisposed to ‘know’ when to learn or what stimuli to attend to. Furthermore, preliminary evidence suggests that fishes may invest only as much into spatial learning capacity as their ecologies and lifestyles demand.

8.10

Salmon homing – a case study

In this final section, we investigate homing in salmon, paying particular attention to the final return by reproductive adults to freshwater streams, a behaviour governed by olfactory recognition of home-stream water. There are several excellent reviews that deal in depth with salmonid migrations (Stabell 1984; Quinn & Dittman 1990; Dittman & Quinn 1996) and it is not our intention to make an exhaustive coverage of the literature. Rather, salmon homing presents a useful case study that illustrates the complex role of genetic, developmental and environmental influences in shaping orientation behaviour in response to particular ecological conditions. Although many species of fishes migrate at some point in their life, the study of fish migration has largely focused on the migratory behaviour of salmonids (salmon, trout and

BLBK374-08

BLBK374-Brown

178

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

charr species). The interest in understanding the migratory behaviour of this group is in part driven by their economic importance. However, the relative ease with which their migratory behaviour can be researched as they move up or downstream on their homeward or outward journey has also made them an amenable model species. Salmon therefore provide an ideal case study in which to determine what they learn as they undergo their migration. We now know, for example, that both genetic and environmental influences shape how and when the fishes learn information relevant to their migration. Typically, salmon spawn in freshwater and after a variable period of residence the offspring migrate to the ocean, to take advantage of the higher productivity of the marine environment (Gross et al. 1988). Once the fishes reach coastal waters some remain close to the mouth of their natal river over the summer, but many others migrate thousands of kilometres to feeding grounds. Virtually all the surviving fishes will at some point return to their natal stream to spawn. Homing is in many cases so precise that it has led to reproductive isolation of spawning populations and specialised adaptations for the natal habitat (Quinn & Dittman 1990; Dittman & Quinn 1996). Although a diverse array of sensory mechanisms and cues have been proposed, there is still remarkably little empirical information on how the salmon navigate once they are at sea (Quinn & Dittman 1990; Ueda et al. 1995; Dittman & Quinn 1996). It is likely that the fishes rely on a compass to migrate the long distances once they leave their natal coastal area (Dittman & Quinn 1996); one of the possible compasses is magnetic (Walker et al. 1997). Tracking work has shown that when the fishes are at sea they will swim both night and day, and magnetite, biogenically produced iron oxide, has been found in the lateral line of Atlantic salmon (Salmo salar) (Moore et al. 1990, 1991). The freshwater phase of the spawning migration is primarily governed by olfactory discrimination of home-stream waters (Hasler & Scholz 1983). The fishes learn the olfactory characteristics of their home site through imprinting during their outward migration. Recognition of the natal stream by returning adults has been investigated using two approaches. First, salmonids are exposed as juveniles to artificial odorants and these fishes are subsequently decoyed to unfamiliar streams scented with the odorants during their homing migrations (Scholz et al. 1976; Hasler & Scholz 1983). Second, hatchery fishes are transported from a rearing site and adult return patterns are monitored (Hansen et al. 1993). Monitoring the stray patterns of returning marked hatchery-reared salmon has also given clues about the imprinting process (Unwin & Quinn 1993; Pascual & Quinn 1994, 1995; Heard 1996; Hard & Heard 1999). One unresolved issue relates to when salmon imprint (Dittman et al. 1996). Experiments with artificial odours indicate that imprinting occurs at the time of peak thyroid hormone levels during smolt transformation, the physiological processes that prepare salmon for oceanic residence (Hasler & Scholz 1983; Morin et al. 1989a, 1989b; Morin & Døving 1992). However, several species of salmon commonly move from their natal site and develop elsewhere in freshwater prior to smolt transformation suggesting that olfactory imprinting and smolt transformation are not inextricably linked (Quinn 1985). Furthermore, kokanee (Oncorynchus spp.) are able to imprint on artificial odorants as alevins and emergent fry as well as at the smolt stage (Dittman & Quinn 1996). A likely resolution to this controversy is that species and populations will differ in the timing of imprinting, according to the nature of their ecologies and migratory patterns. There is extraordinary diversity of freshwater habitats and migratory patterns both within and among salmonid species and

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

179

individuals may need to be sensitive to odours at different stages of their development (Dittman et al. 1996). Salmon may also learn odours at more than one time and place and then use the sequence of learned odours to guide their homeward migration (Quinn 1985; Hansen et al. 1987; Heard 1996). One possibility is that thyroxine surges trigger transient neural changes in the peripheral olfactory system, permitting windows of sensitivity, which enable olfactory imprinting. Short-term increases in plasma thyroid hormones may enable wild salmon to learn key freshwater landmarks not only at developmentally regulated periods, but also during migrations in response to novel environmental stimuli (Dittman & Quinn 1996). Elevated levels of thyroxine that last for only a few hours have also been shown for migrating chum salmon (Oncorhynchus keta) when the salmon encounter different habitats (Iwata et al. 2003). Animals typically go through a phase of learning when they encounter new environments; it is possible then that thyroxine may in some way prepare the fish for a period of learning. Further evidence for a link between thyroxine and learning is that structural reorganisation in certain brain circuits has been shown to occur during the downstream migration of Atlantic salmon when levels of thyroxine are high (Ebbesson et al. 2003). Overall, these studies provide a convincing illustration of the extent to which developmental and genetic influences are likely to regulate and constrain flexible learned behaviour in response to species-typical and even population-typical habitat conditions and migratory patterns.

8.11

Conclusion

Most species of fishes experience environments in which biologically important locations as well as the physical landscape from which spatial information is extracted will change at rates that cannot be tracked by adaptive changes in the gene pool. A capacity to learn provides flexibility, allowing fishes to match their orientation strategy to a variable environment on the basis of experience. A combined approach of field- and laboratory-based experiments has shown that fishes are indeed capable of spatial learning and can use information from a vast array of different sources. Flexible learning appears to act in concert with genetic and developmental influences such that orientation behaviour is adapted for particular ecological conditions and navigational demands. Future studies that attempt to assess how and to what extent fishes learn to use different types of spatial information in orientation should involve experiments whose design is based on a detailed understanding of the habitat ecology and spatial problems likely to be encountered by fishes in their natural habitats.

Acknowledgements We would like to thank the editors for inviting us to make this contribution and for their helpful comments on the earlier versions of this chapter.

BLBK374-08

BLBK374-Brown

180

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

References Able, K.P. (1993) Orientation cues used by migratory birds – a review of cue-conflict experiments. Trends in Ecology & Evolution, 8, 367–371. Armstrong, J.D., Braithwaite, V.A. & Huntingford, F.A. (1997) Spatial strategies of wild Atlantic salmon parr: exploration and settlement in unfamiliar areas. Journal of Ecology, 66, 203–211. Aronson, L.R. (1951) Orientation and jumping behaviour in the Gobiid fish Bathygobius soporator. American Museum of Noviscotia, 1486, 1–22. Aronson, L.R. (1971) Further studies on orientation and jumping behaviour in the Gobiid fish Bathygobius sporator. Annals of the New York Academy of Science, 188, 378–392. Arvedlund, M., McCormick, M.I., Fautin, D.G. & Bildsøe, M. (1999) Host recognition and possible imprinting in the anemonefish Amphiprion melanopus (Pisces: Pomacentridae). Marine Ecology Progress Series, 188, 207–218. Bennet, A.T.D. (1996) Do animals have cognitive maps? Journal of Experimental Biology, 199, 219–224. Biegler, R. & Morris, R.G.M. (1996) Landmark stability: studies exploring whether perceived stability of the environment influences spatial representation. Journal of Experimental Biology, 199, 187–193. Braithwaite, V.A., Armstrong, J.D., McAdam, H.M. & Huntingford, F.A. (1996) Can juvenile Atlantic salmon use multiple cue systems in spatial learning? Animal Behaviour, 51, 1409–1415. Braithwaite, V.A. & Girvan, J.R. (2003) Use of waterflow to provide spatial information in a smallscale orientation task. Journal of Fish Biology, 63, 74–83. Brown, C. (2003) Habitat-predator association and avoidance in rainbowfish (Melanotaenia spp.). Ecology of Freshwater Fishes, 12, 118–126. Brown, C. & Braithwaite, V.A. (2005) Effects of predation pressure on the cognitive ability of the poeciliid Brachyraphis episcopi. Behavioral Ecology, 16, 482–497. Brown, C., Gardner, C. & Braithwaite, V.A. (2004) Population variation in lateralised eye use in the poeciliid Brachyraphis episcopi. Proceedings of the Royal Society of London Series B (Supplement), 271, S455–S457. Brown, C. & Laland, K.N. (2003) Social learning in fishes: a review. Fish and Fisheries, 4, 280–288. Burt de Perera, T. (2004a) Spatial parameters encoded in the spatial map of blind Mexican cave fish, Astynax fasciatus. Animal Behaviour, 68, 291–295. Burt de Perera, T. (2004b) Fish can encode order in their spatial map. Proceeedings of the Royal Society of London Series B – Biological Sciences, 271, 2131–2134. Burt de Perera, T. & Braithwaite, V.A. (2005) Laterality in a non-visual sensory modality – the lateral line of fish. Current Biology, 15, R241–R242. Burt de Perera, T., de Vos, A. & Guilford, T. (2005) The vertical component of a fish’s spatial map. Animal Behaviour, 70, 405–409. Cain, P. (1995) Navigation in familiar environments by the weakly electric elephantnose fish, Gnathonemus petersii L. (Mormyriformes, Teleostei). Ethology, 99, 332–349. Cain, P., Gerin, W. & Moller, P. (1994) Short-range navigation of the weakly electric fish, Gnathonemus petersii L. (Mormyridae, Teleostei), in novel and familiar environments. Ethology, 96, 33–45. Campenhausen, C.V., Reiss, I. & Weissert, R. (1981) Detection of stationary objects by the blind cave fish Anoptichthys jordani (Characidae). Journal of Comparative Physiology A, 143, 369–374. Carlson, H.R. & Haight, R.E. (1972) Evidence for a home site and homing of adult yellowtail rockfish, Sebastes flavidus. Journal of the Fisheries Research Board Canada, 29, 1011–1014. Carneiro, L.A., Andrade, R.P., Oliveira, R.F. & Kotrschal, K. (2001) Sex differences in home range and dorso-lateral telencephalon in the Azorean rock-pool blenny. Society for Neuroscience Abstract, 27, Program No. 535.4. Coolen, I., van Bergen, Y., Day, R.L. & Laland, K.N. (2003) Species differences in adaptive use of public information in sticklebacks. Proceedings of the Royal Society of London Series B – Biological Sciences. B, 270, 2413–2419.

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

181

Dittman, A.H. & Quinn, T.P. (1996) Homing in pacific salmon: mechanisms and ecological basis. Journal of Experimental Biology, 199, 83–91. Dittman, A.H., Quinn, T.P. & Nevitt, G.A. (1996) Timing of imprinting to natural and artificial odors by coho salmon (Oncorhynchus kisutch). Canadian Journal of Fisheries and Aquatic Sciences, 53, 434–442. Dodson, J.J. (1988) The nature and role of learning in the orientation and migratory behaviour of fishes. Environmental Biology of Fishes, 23, 161–182. Ebbesson, L.O.E., Ekstrom, P., Ebbesson, S.O.E., Stefansson, S.O. & Holmqvist, B. (2003) Neural circuits and their structural and chemical reorganisation in the light–brain–pituitary axis during parr–smolt tranformation. Aquaculture, 222, 59–70. Girvan, J.R. & Braithwaite, V.A. (1998) Population differences in spatial learning in three-spined sticklebacks. Proceedings of the Royal Society of London Series B – Biological Sciences, 265, 913–919. Girvan, J.R. & Braithwaite, V.A. (2000) Orientation behaviour in sticklebacks: modified by experience or population specific? Behaviour, 137, 833–843. Goodyear, C.P. & Ferguson, D.E. (1969) Sun-compass orientation in mosquitofish, Gambusia affinis. Animal Behaviour, 17, 636–640. Goodyear, P.C. (1973) Learned orientation in the predator avoidance of mosquitofish, Gambusia Affinis. Behaviour, 191–223. Goodyear, P.C. & Bennet, D.H. (1979) Sun compass orientation of immature bluegill. Transactions of the American Fisheries Society, 108, 555–559. Graff, C., Kaminski, G., Gresty, M. & Ohlmann, T. (2004) Fish perform spatial pattern recognition and abstraction by exclusive use of active electrolocation. Current Biology, 14, 181–123. Green, J.M. (1971) High tide movements and homing behaviour of the tidepool sculpin Oligocottus maculosus. Journal of the Fisheries Research Board Canada, 28, 383–389. Gross, M.R., Coleman, R.M. & McDowall, R.M. (1988) Aquatic productivity and the evolution of diadromous fish migration. Science, 239, 1291–1293. Hallacher, L.E. (1984) Relocation of original territories by displaced black-and-yellow rockfish, Sebastes chrysomelas, from Carmel Bay, California. California Fish and Game, 70, 158–162. Hansen, L.P., Døving, K.B. & Jonsson, B. (1987) Migration of farmed adult Atlantic salmon with and without olfactory sense, released on the Norwegian coast. Journal of Fish Biology, 30, 713–721. Hansen, L.P., Jonsson, N. & Jonsson, B. (1993) Oceanic migration in homing Atlantic salmon. Animal Behaviour, 45, 927–941. Hard, J.J. & Heard, W.R. (1999) Analysis of straying variation in Alaskan hatchery chinook salmon (Oncorhynchus tshawytscha) following transplantation. Canadian Journal of Fisheries and Aquatic Sciences, 56, 578–589. Hasler, A.D. (1971) Orientation and fish migration. In: W.S. Hoar & D.J. Randall (eds) Fish Physiology. Academic Press, London. Hasler, A.D., Horrall, R.M., Wisby, W.J. & Braemer, W. (1958) Sun-orientation and homing in fishes. Limnology and Oceanography, 111, 353–361. Hasler, A.D. & Scholz, A.T. (1983) Olfactory Imprinting and Homing in Salmon. Springer-Verlag, Berlin. Hawryshyn, C.W., Arnold, M.G., Bowering, E. & Cole, R.L. (1990) Spatial orientation of rainbow trout to plane-polarised light: the ontogeny of E-vector discrimination and spectral sensitivity characteristics. Journal of Comparative Physiology A, 166, 565–574. Healy, S.D. (1998) Spatial Representation in Animals. Oxford University Press, Oxford. Heard, W.R. (1996) Sequential imprinting in chinook salmon: is it essential for homing fidelity? Bulletin of National Research Institute of Aquaculture, Supplement, 2, 59–64. Huber, R., Van Staaden, M.J., Kaufman, L.S. & Liem, K.F. (1997) Microhabitat use, trophic patterns, and the evolution of brain structure in African cichlids. Brain, Behavior and Evolution, 50, 167–182. Hughes, R.N. & Blight, C.M. (1999) Algorithmic behaviour and spatial memory are used by two intertidal fish species to solve the radial maze. Animal Behaviour, 58, 601–613.

BLBK374-08

BLBK374-Brown

182

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Hughes, R.N. & Blight, C.M. (2000) Two intertidal fish species use visual association learning to track the status of food patches in a radial maze. Animal Behaviour, 59, 613–621. Hunter, E., Metcalfe, J.D., Arnold, G.P. & Reynolds, J.D. (2004) Impacts of migratory behaviour on population structure in North Sea plaice. Journal of Animal Ecology, 73, 377–385. Hunter, E., Metcalfe, J.D. & Reynolds, J.D. (2003) Migration route and spawning area fidelity by North Sea plaice. Proceedings of the Royal Society of London Series B – Biological Sciences, 270, 2097–2103. Huntingford, F.A., Braithwaite, V.A., Armstrong, J.D., Aird, D. & Joiner, P. (1998) Homing in juvenile salmon in response to imposed and spontaneous displacement: experiments in an artificial stream. Journal of Fish Biology, 53, 847–852. Huntingford, F.A. & Wright, P.J. (1989) How sticklebacks learn to avoid feeding patches. Behavioural Processes, 19, 181–189. Immelmann, K. & Suomi, S.J. (1981) Sensitive phases in development. In: K. Immelmann, G.W. Barlow, L. Petrinovich & M. Main (eds) Behavioural Development: The Bielefeld Interdisciplinary Project, pp. 395–431. Cambridge University Press, Cambridge. Ingle, D. & Sahagian, D. (1973) Solution of a spatial constancy problem by goldfish. Physiological Psychology, 1, 83–84. Iwata, M., Tsuboi, H., Yamashita, T., Amemiya, A., Yamada, H. & Chiba, H. (2003) Function and trigger of thyroxine surge in migrating chum salmon Oncorhynchus keta fry. Aquaculture, 222, 315–329. Jones, G.P., Milicich, M.J., Emslie, M.J. & Lunnow, C. (1999) Self-recruitment in a coral reef population. Nature, 402, 802–804. Jonsson, N., Jonsson, B., Skurdal, J. & Hansen, L.P. (1994) Differential response to water current in offspring of inlet- and outlet-spawning brown trout Salmo trutta. Journal of Fish Biology, 45, 356–359. Jordan, F., Bartolini, M., Nelson, C., Patterson, P.E. & Soulen, H.L. (1997) Risk of predation affects habitat selection by the pinfish Lagodon rhomboides (Linnaeus). Journal of Experimental Marine Biology and Ecology, 208, 45–56. Kalmijn, A.J. (1982) Electric and magnetic field detection in elasmobranch fishes. Science, 218, 916–918. Kaya, C.M. & Jeanes, E.D. (1995) Retention of adaptive rheotactic behaviour by F1 fluvial Arctic grayling. Transactions of the American Fisheries Society, 124, 453–457. Kingsford, M.J., Leis, J.M., Shanks, A., Lindeman, K.C., Morgan, S.G. & Pineda, J. (2002) Sensory environments, larval abilities and local self-recruitment. Bulletin of Marine Science, 70, 309–340. Kirschvink, J.L., Walker, M.M., Chang, S.-B.R. & Dizon, A.E. (1985) Chains of single domain magnetite particles in chinook salmon, Oncorhynchus tschawytscha. Journal of Comparative Physiology A, 157, 375–381. Kleerokoper, H., Matis, J., Gensler, P. & Maynard, P. (1974) Exploratory Behaviour of Goldfish, Carassius auratus. Animal Behaviour, 22, 124–132. Kolm, N., Hoffman, E.A., Olsson, J., Berglund, A. & Jones, A.G. (2005) Group stability and homing behaviour but no kin group structures in a coral reef fish. Behavioral Ecology, 16, 521–527. Kotrschal, K., Van Staaden, M.J. & Huber, R. (1998) Fish brains: evolution and environmental relationships. Reviews in Fish Biology and Fisheries, 8, 373–408. Kroon, F.J., de Graaf, M. & Liley, N.R. (2000) Social organisation and competition for refuges and nest sites in Coryphopterus nicholsii (Gobiidae), a temperate protogynous reef fish. Environmental Biology of Fishes, 57, 401–411. Leis, J.M., Carson-Ewart, B.M., Hay, A.C. & Cato, D.H. (2003) Coral-reef sounds enable nocturnal navigation by some reef-fish larvae in some places and at some times. Journal of Fish Biology, 63, 724–737. L´opez, J.C., Bingman, V.P., Rodr´ıguez, F., G´omez, Y. & Salas, C. (2000a) Dissociation of place and cue learning by telencephalic ablation in goldfish. Behavioural Neuroscience, 114, 687–699. L´opez, J.C., Broglio, C., Rodr´ıguez, F., Thinus-Blanc, C. & Salas, C. (1999) Multiple spatial learning strategies in goldfish (Carassius auratus). Animal Cognition, 2, 109–120.

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

183

L´opez, J.C., Broglio, C., Rodr´ıguez, F., Thinus-Blanc, C. & Salas, C. (2000b) Reversal learning deficit in a spatial task but not in a cued one after telencephalic ablation in goldfish. Behavioural Brain Research, 109, 91–98. Markel, R.W. (1994) An adaptive value of spatial learning and memory in the blackeye goby, Coryphoterus nicholsi. Animal Behaviour, 47, 1462–1464. Matthews, K.R. (1990a) An experimental study of the habitat preferences and movement patterns of copper, quillback, and brown rockfishes (Sebastes spp.). Environmental Biology of Fishes, 29, 161–178. Matthews, K.R. (1990b) A telemetric study of the home ranges and homing of copper and quillback rockfishes on shallow rocky reefs. Canadian Journal of Fisheries and Aquatic Sciences, 68, 2243–2250. Mazeroll, A.I. & Montgomery, W.L. (1995) Structure and organization of local migrations in brown surgeonfish (Acanthurus nigrofuscus). Ethology, 99, 89–106. Mazeroll, A.I. & Montgomery, W.L. (1998) Daily migrations of a coral reef fish in the Red Sea (Gulf of Aqaba, Israel): initiation and orientation. Copeia, 4, 893–905. McCormick, C.A. (1989) Central lateral line pathways in bony fish. In: S. Coombs, P. G¨orner & H. M¨unz (eds) The Lateral Line: Neurobiology and Evolution, pp. 341–363. Springer Verlag, New York. Montgomery, J.C., Jeffs, A.G., Simpson, S.D., Meekan, M.G. & Tindle, C. (2006) Sound as an orientation cue for the pelagic larvae of reef fishes and decapod crustaceans. Advances in Marine Biology, 51, 143–196. Montgomery, J.C., Tolimieri, N. & Haine, O.S. (2001) Active habitat selection by pre-settlement reef fishes. Fish and Fisheries, 2, 261–277. Moore, A., Bone, Q. & Ryan, K.P. (1991) The anatomy of the lateral line of the Atlantic salmon (Salmo salar L.). Journal of Marine Biology U.K., 71, 949. Moore, A., Freake, S.M. & Thomas, I.M. (1990) Magnetic particles in the lateral line of the Atlantic salmon (Salmo salar L.). Philosophical Transactions of the Royal Society of London Series B – Biological Sciences, 329, 11–15. Morin, P.-P., Dodson, J.J. & Dor´e, F.Y. (1989a) Cardiac responses to a natural odorant as evidence of a sensitive period for olfactory imprinting in young Atlantic salmon, Salmo salar. Canadian Journal of Fisheries and Aquatic Sciences, 46, 122–130. Morin, P.-P., Dodson, J.J. & Dor´e, F.Y. (1989b) Thyroid activity concomitant with olfactory learning and heart rate changes in Atlantic salmon, Salmo salar, during smoltification. Canadian Journal of Fisheries and Aquatic Sciences, 46, 131–136. Morin, P.-P. & Døving, K.B. (1992) Changes in the olfactory function of Atlantic salmon, Salmo salar, in the course of smoltification. Canadian Journal of Fisheries and Aquatic Sciences, 49, 1704–1713. Noda, M., Gushima, K. & Kakuda, S. (1994) Local prey search based on spatial memory and expectation in the planktivorous fish, Chromis chrysurus (Pomacentridae). Animal Behaviour, 47, 1413–1422. Odling-Smee, L. & Braithwaite, V.A. (2003) The influence of habitat stability on landmark use during spatial learning in the threespine stickleback. Animal Behaviour, 65, 701–707. Ogden, J.C. & Buckman, N.S. (1973) Movements, foraging groups, and diurnal migrations of the striped parrotfish Scarus croicensis Bloch (Scaridae). Ecology, 54, 598–596. Ogden, J.C. & Quinn, T.P. (1984) Migration in coral reef fishes: ecological significance and orientation mechanisms. In: J.D. McCleave, G.P. Arnold, J.J. Dodson & W.H. Neill (eds) Mechanisms of Migration in Fishes, pp. 293–308. Plenum Press, New York. Pascual, M.A. & Quinn, T.P. (1994) Geographical patterns of straying of fall chinook salmon, Oncorhynchus tshawytscha (Walbaum), from Columbia River (USA) hatcheries. Aquaculture and Fisheries Management, 25(Supplement), 17–30. Pascual, M.A. & Quinn, T.P. (1995) Factors affecting the homing of fall chinook salmon from Columbia River hatcheries. Transactions of the American Fisheries Society, 124, 308–320.

BLBK374-08

BLBK374-Brown

184

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Pitcher, T.L. & Magurran, A.E. (1983) Shoal size, patch profitability and information exchange in foraging goldfish. Animal Behaviour, 31, 546–555. Planes, S. (2002) Biogeography and larval dispersal inferred from population genetic analysis. In: P.F. Sale (ed) Coral Reef Fishes: Dynamics and Diversity in a Complex Ecosystem, pp. 201–220. Academic Press, San Diego, CA. Quinn, T.P. (1985) Salmon homing: is the puzzle complete? Environmental Biology of Fishes, 12, 315–317. Quinn, T.P. & Dittman, A.H. (1990) Pacific salmon migrations and homing: mechanisms and adaptive significance. Trends in Ecology and Evolution, 5, 174–177. Quinn, T.P. & Ogden, J.C. (1984) Field evidence of compass orientation in migrating juvenile grunts (Haemulidae). Journal of Experimental Marine Biology and Ecology, 81, 181–192. Reebs, S.G. (1996) Time-place learning in golden shiners (Pisces: Cyprinidae). Behavioural Processes, 36, 253–262. Reebs, S.G. (1999) Time-place learning based on food but not on predation risk in a fish, the inanga (Galaxias maculatus). Ethology, 105, 361–371. Reese, E.S. (1989) Orientation behaviour of butterflyfishes (family Chaetodontidae) on coral reefs: spatial learning of route specific landmarks and cognitive maps. Environmental Biology of Fishes, 25, 79–86. Rodr´ıguez, F., Dur´an, E., Vargas, J.P., Torres, B. & Salas, C. (1994) Performance of goldfish trained in allocentric and egocentric maze procedures suggest the presence of a cognitive mapping system in fishes. Animal Learning and Behaviour, 22, 409–420. Roitblat, H.L., Tham, W. & Golub, L. (1982) Performance of Betta splendens in a radial maze. Animal Learning and Behaviour, 10, 108–114. Salas, C., Broglio, C., Rodr´ıguez, F., L´opez, J.C., Portavella, M. & Torres, B. (1996a) Telencephalic ablation in goldfish impairs performance in a spatial constancy problem but not in a cued one. Behavioural Brain Research, 79, 193–200. Salas, C., Broglio, C. & Rodriguez, F. (2003) Evolution of forebrain and spatial cognition in vertebrates: conservation across diversity. Brain, Behavior and Evolution, 62, 72–82. Salas, C., Rodr´ıguez, F., Vargas, J.P., Dur´an, E. & Torres, B. (1996b) Spatial learning and memory deficits after telencephalic ablation in goldfish trained in place and turn maze procedures. Behavioural Neuroscience, 110, 965–980. Schluter, D. & McPhail, J.D. (1992) Ecological character displacement and speciation in sticklebacks. The American Naturalist, 140, 85–108. Scholz, A.T., Horrall, R.M., Cooper, J.C. & Hasler, A.D. (1976) Imprinting to chemical cues: the basis for home stream selection in salmon. Science, 192, 1247–1249. Schwassmann, H.O. & Braemer, W. (1961) The effect of experimentally changed photoperiod on the sun orientation rhythm of fish. Physiological Zoology, 34, 273–326. Sherry, D.F. (1998) The ecology and neurobiology of spatial memory. In: R. Dukas (ed) Cognitive Ecology, pp. 261–296. The University of Chicago Press, Chicago. Shettleworth, S.J. (1993) Where is the comparison in comparative cognition? American Psychological Society, 4, 179–184. Simpson, S.D., Meekan, M.G., McCauley, R.D. & Jeffs, A.G. (2004) Attraction of settlement-stage coral reef fishes to reef noise. Marine Ecology Progress Series, 276, 263–268. Simpson, S.D., Meekan, M.G., Montgomery, J.C., McCauley, R.D. & Jeffs, A.G. (2005a) Homeward Sound. Science, 308, 221. Simpson, S.D., Yan, H.Y., Wittenrich, M.L. & Meekan, M.G. (2005b) Response of embryonic coral reef fishes (Pomacentridae: Amphiprion spp.) to noise. Marine Ecology Progress Series, 287, 201–208. Smith, R.J.F. & Smith, M.J. (1998) Rapid acquisition of directional preferences by migratory juveniles of two amphidromous Hawaiin gobies, Awaous guamensis and Sicyopterus stimpsoni. Environmental Biology of Fishes, 53, 275–282. Stabell, O.B. (1984) Homing and olfaction in salmonids: a critical review with special reference to the Atlantic salmon. Biological Reviews, 59, 333–388.

BLBK374-08

BLBK374-Brown

May 12, 2011

10:41

Copyeditor’s Name:

Trim: 244mm X 172mm

The Role of Learning in Fish Orientation

Char Count=

185

Swearer, S.E., Caselle, J.E., Lea, D.W. & Warner, R.R. (1999) Larval retention and recruitment in an island population of a coral-reef fish. Nature, 402, 799–802. Teyke, T. (1985) Collision with and avoidance of obstacles by blind cave fish Anoptichthys jordani (Characidae). Journal of Comparative Physiology A, 157, 837–843. Teyke, T. (1989) Learning and remembering the environment in the blind cave fish Anoptichthys jordani. Journal of Comparative Physiology A, 164, 655–662. Tolimieri, N., Jeffs, A. & Montgomery, J.C. (2000) Ambient sound as a cue for navigation by the pelagic larvae of reef fishes. Marine Ecology Progress Series, 207, 219–224. Tsukamoto, K., Aoyama, J. & Miller, M.J. (2003) Migration, speciation, and the evolution of diadromy in anguillid eels. Canadian Journal of Fisheries and Aquatic Sciences, 59, 1989–1998. Ueda, H., Kaeriyama, M., Urano, A., Kurihara, K. & Yamauchi, K. (1995) Homing mechanism in salmon: roles of vision and olfaction. In: F.W. Goetz (ed) International Symposium on Reproductive Physiology of Fish, pp. 35–37. The University of Texas, Texas. Unwin, M.J. & Quinn, T.P. (1993) Homing and straying patterns of chinook salmon (Oncorhynchus tshawytscha) from a New Zealand hatchery: spatial distribution of strays and effects of release date. Canadian Journal of Fisheries and Aquatic Sciences, 50, 1168–1175. Van Staaden, M.J., Huber, R., Kaufman, L.S. & Karel, F.L. (1994) Brain evolution in cichlids of the African Great Lakes: brain and body size, general patterns, and evolutionary trends. Zoology, 98, 165–178. Walker, M.M. (1984) Learned magnetic field discrimination in yellowfin tuna, Thunnus albacares. Journal of Comparative Physiology A, 155, 673–679. Walker, M.M., Diebel, C.E., Haugh, C.V., Pankhurst, P.M., Montgomery, J.C. & Green, C.R. (1997) Structure and function of the vertebrate magnetic sense. Nature, 390, 371–376. Walker, M.M., Quinn, T.P., Kirschvink, J.L. & Groot, C. (1988) Production of single domain magnetite throughout life by sockeye salmon, Onchorynchus nerka. Journal of Experimental Biology, 140, 51–63. Warburton, K. (1990) The use of local landmarks by foraging goldfish. Animal Behaviour, 40, 500–505. Warburton, K. (2003) Learning of foraging skills by fish. Fish and Fisheries, 4, 203–215. Welker, W.I. & Welker, J. (1958) Reaction of fish (Eucinostomus gula) to environmental changes. Ecology, 39, 283–288.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 9

Social Recognition of Conspecifics Siˆan W. Griffiths and Ashley Ward

9.1

Introduction

Population-level patterns of social organisation are determined by localised interactions between conspecifics. Therefore, understanding the decision rules employed by individuals in these interactions is central to our understanding of species ecology. Research has shown that fishes are not equally socially attracted or aggressive towards all other conspecifics. For example, the preferential association of salmonids with relatives (and reduced aggression towards them) is now well established and, more recently, evidence has been obtained for the remarkable ability of fish to choose among unrelated conspecifics on the basis of shared habitat-based cues or prior experience (preferring familiar individuals). Here, we review the evidence for these discriminatory behavioural interactions and discuss how the nature of behavioural interactions and associations between fishes will have especially important implications for their dispersal and also their ability to recolonise after natural or human disturbances. Information regarding group membership (school structure or territorial assemblages) may also have implications for conservation of declining fish stocks and fisheries management policies because individuals that choose familiar schoolmates accrue antipredator and foraging benefits. Conferring these benefits on declining fish stocks has important economic and conservation implications. Therefore, future work may benefit from an exploration of the ecological contexts in which recognition of conspecifics is important in natural streams and rivers.

9.2

Recognition of familiars

It is well established that fishes are capable of discriminating between individuals with whom they interact. For example, shoaling fishes exert a high degree of choice with regard to their preferred shoalmates, selecting on the basis of species (Keenleyside 1955), colour (McRobert & Bradner 1998) and size (Ward & Krause 2001). In recent years, it has also become clear that fishes are capable of more fine-scale assessment and discrimination. For example, European minnows are able to discriminate between conspecifics Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

187

according to competitive ability and preferentially shoal with poor competitors to reduce their own competition costs (Metcalfe & Thomson 1995). Similarly, three-spined sticklebacks are able to determine which individuals may be most cooperative as predator inspection partners and preferentially associate with these individuals in this context (Milinski et al. 1990a, 1990b). A number of studies have now demonstrated that fishes recognise and behave differentially towards unrelated conspecifics, irrespective of context and in the absence of any obvious behavioural or phenotypic cues. Familiarity, as this ability is known, is a widespread phenomenon among fishes and has important implications for their social organisation.

9.2.1

Laboratory studies of familiarity

There is a large and increasing body of work documenting the phenomenon of conspecific familiarity in fishes (Table 9.1). Familiarity mediates the interactions of a wide range of species and plays a key role in their social dynamics. In shoaling species, the effects of familiarity may be seen chiefly in the fishes’ association decisions, affecting their choice of shoaling partners. Broadly, the function of this is to enhance the benefits of grouping and mediates association decisions in these species. By contrast, less gregarious fishes interact less frequently and familiarity is often confined to the contexts of territorial interactions, dominance hierarchies and mate recognition. Whilst the ability to recognise and subsequently associate with familiar conspecifics has been documented in a wide range of social fish species, this ability may also be expressed in a heterospecific context. Ward et al. (2003) reported that chub preferred to associate with familiar over unfamiliar individuals of the closely related and sympatric species, the European minnow. Furthermore, the association preference of chub for conspecifics disappeared when focal fishes were given a choice between familiar heterospecifics and unfamiliar conspecifics.

9.2.2

Mechanisms of familiarity recognition

Two differing mechanisms have been proposed to explain the phenomenon of familiarity. A very general form of recognition based on odour cues has been documented in fathead minnows (Brown & Smith 1994) and in sticklebacks (Ward et al. 2004, 2005). A more specific type of visual recognition, which may allow the recognition of particular individuals, has been reported in guppies (Griffiths & Magurran 1997a, 1998). Although further work is required to elucidate the role of each, the likelihood is that the two mechanisms operate in conjunction as each allows a different level of recognition. Chemical signalling is of great importance in the aquatic environment; water acts both as a solvent and as a medium to disperse cues. Chemical cues may be particularly useful when visual communication is limited, for example during darkness or in deep or turbid water, or where the environment is highly structured. Chemical communication mediates the interactions between fishes in a range of diverse contexts, including mate choice (Milinski et al. 2005), homing and migration (Hasler & Scholz 1983; Dittman & Quinn 1996), predator recognition (Wisenden 2000) and dominance relationships (Bryant & Atema 1987).

Year 2000 2001 2004 1986

1993 2002 1994 1995 2001 2004 1992 2004 1997 2003 2007 1997 1997a 1997b 1998 1999 2007 2002 1984 1991 2000 1998

Author (s)

Barber & Ruxton Barber & Wright Binoy & Thomas Brown & Colgan

Brown et al. Brown Brown & Smith Chivers et al. Courtenay et al. Croft et al. Dugatkin & Wilson Farmer et al.

Giaquinto & Volpato Godin et al. Gomez-Laplaza & Fuente Griffiths Griffiths & Magurran Griffiths & Magurran

Griffiths & Magurran

Griffiths & Magurran Griffiths et al. Hay & McKinell Helfman Hilborn Hoare et al. H¨ojesj¨o et al.

Two-choice test Shoal fidelity Shoal fidelity Shoal fidelity Shoal fidelity Shoal fidelity Resource competition

Copyeditor’s Name:

P. reticulata P. phoxinus Clupea pallasi Perca flavescens Katsuwonus pelamis Fundulus diaphanus Salmo trutta

16:37

Trinidadian guppy European minows Pacific herring Yellow perch Skipjack tuna Banded killifish Sea trout

Y Y Y Y N N N Y, Y Y Y Y Males, Y Y Y Y Y N Y & N mediated by social status Y Y Y N Females, Y Males, N Y Y Y N N N Y

Preference for familiar

May 14, 2011

P. reticulata

Two-choice test Two-choice test; near neighbour Two-choice test Shoal cohesion Two-choice test Two-choice test Two-choice test Two-choice test Two-choice test Two-choice test Two-choice test Two-choice test Resource competition Near neighbour Two-choice test Near neighbour Two-choice test

Shoal cohesion Two-choice test Two-choice test Two-choice test

Method

BLBK374-Brown

Trinidadian guppy

Gasterosteus aculeatus Phoxinus phoxinus Anabas testudineus Lepomis macrochirus Lepomis gibbosus Ambloplites rupestris Oncorhynchus mykiss Melanotaenia spp. Pimephales promales P. promales Oncorhynchus kisutch Poecilia reticulata L. macrochirus Notropis amabilis Cyprinella venusta Oreochromis niloticus P. reticulata Pterophyllum scalare P. phoxinus P. reticulata P. reticulata

Scientific name

188

Three-spined stickleback European minnows Climbing perch Bluegill sunfish Pumpkinseed sunfish Rock bass Hatchery rainbow trout Rainbowfish Fathead minnows Fathead minnows Coho salmon Trinidadian guppy Bluegill sunfish Texas shiner Blacktail shiner Nile tilapia Trinidadian guppy Angelfish European minnow Trinidadian guppy Trinidadian guppy

Common name

Table 9.1 Summary of investigations on the role of familiarity in fish schooling decisions.

BLBK374-09 Trim: 244mm X 172mm

Fish Cognition and Behavior

Char Count=

Hatchery rainbow trout Hatchery rainbow trout Trinidadian guppy Yellowfin tuna Domestic guppy Trinidadian guppy European minnow Paradise fish Atlantic salmon Hatchery Arctic charr Domestic guppy Three-spined stickleback Three-spined stickleback Domestic guppy Three-spined stickleback Fathead minnow

O. mykiss O. mykiss P. reticulata Thunnus albacares P. reticulata P. reticulata P. phoxinus Macropodus opercularis Salmo salar Salvelinus alpinus P. reticulata G. aculeatus G. aculeatus P. reticulata G. aculeatus P. promelas

Resource competition Resource competition Mating attempts Shoal fidelity Two-choice test Two-choice test Resource competition Resource competition Resource competition Growth and mortality Shoal cohesion Resource competition Two-choice test Two-choice test Shoal fidelity Alarm cell proliferation

Y Y Males, U Y Y Y Y Y Y Y Y Y Y Y Y Y

May 14, 2011 16:37 Copyeditor’s Name:

Preference for familiars (Y), no preference (N) or preference for unfamiliars (U) is indicated. The discriminatory ability of individual fish was tested in two-choice tests or by looking at association patterns of near neighbours. Familiarity has also been tested in the laboratory in the contexts of shoal cohesion, competition and aggression. Preference for familiars in the wild is inferred from shoal fidelity.

1997 1998 1999 1999 1998 1994 1995 1992 2000 2001 2001 2000 1988 1996 2002 1998

BLBK374-Brown

Johnsson ´ kerman Johnsson & U Kelley et al. Klimley & Holloway Lachlan et al. Magurran et al. Metcalfe & Thomson Miklosi et al. O’Connor et al. Sepp¨a et al. Swaney et al. Utne-Palm & Hart Van Havre & FitzGerald Warburton & Lees Ward et al. Wisenden & Smith

BLBK374-09 Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

189

BLBK374-Brown

190

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

The recognition of familiar individuals on the basis of chemical cues potentially allows a simple means of recognition. Brown & Smith (1994) investigated the sensory modalities involved in the recognition of familiar individuals in fathead minnows, finding that there was no preference for familiars when visual cues alone were available but that a preference for familiars was expressed when olfactory cues were available. The same pattern was also observed in three-spined sticklebacks (Ward et al. 2004) whilst in Nile tilapia olfactory cues were also required for complete recognition of conspecifics (Giaquinto & Volpato 1997). The olfactory cues expressed by individuals are known to be influenced by both recent habitat use and recent diet (Bryant & Atema 1987; Ols´en et al. 2003; Ward et al. 2004) and these cues in turn mediate the association preferences; fishes prefer individuals which smell similar to themselves (Ward et al. 2004, 2005), suggesting chemical selfreferencing. Ward et al. (2005) reported that three-spined sticklebacks recognised and subsequently expressed an association preference for conspecifics and heterospecifics that had experienced similar habitat and diet conditions as themselves over the previous 24 hours. Perhaps most interestingly, when direct experience was titrated against these general odour cues, fishes preferred individuals who expressed the same environmental cues as themselves to individuals with whom they had previously interacted but who expressed different environmental cues to themselves (Ward et al. 2004, 2005, 2007, 2009). A more specific form of recognition appears to underpin familiarity preferences in guppies. A study by Griffiths & Magurran (1997a) showed that familiarity develops gradually as individuals interact repeatedly over a period of 12 days (Fig. 9.1). Subsequent work by Griffiths & Magurran (1997b) showed that there is an upper limit of around 40 different individuals that a single fish can identify as familiar. Furthermore, in contrast to other species that require the presence of olfactory cues in order to demonstrate an association preference for familiars, guppies have been shown to recognise familiars on the basis of visual cues alone (Griffiths & Magurran 1998; Ward et al. 2009). 100

75 Time schooling (%)

BLBK374-09

Familiar Unfamiliar

50

25

0

0

10

20

30

Day Fig. 9.1 The schooling preference for familiar individuals develops gradually in guppies, becoming significant after 12 days.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

191

The gradual development of familiarity is a consistent feature amongst studies of this nature across different species and is suggestive of a learning process. A corollary of this is the gradual decrease in familiarity preferences that occurs if familiar individuals are separated. For example, Utne-Palm & Hart’s (2000) study of familiarity in three-spined stickleback revealed a gradual increase in familiarity over a 4-week period of association and a subsequent decrease in familiarity over a 4-week period of isolation. In a study of rainbow trout, Johnsson (1997) found that the separation of pairs of territorial combatants for a period of 3 days was sufficient for the familiarity-biased aggression levels to decay. Indeed, it may be that under certain circumstances, memory decay may be adaptive (Mikl´osi et al. 1992; Warburton 2003; Griffiths et al. 2007). In contrast to the gradual development of individual recognition over several days in a context-independent situation, fishes are capable of extremely rapid learning of identities in context-dependent circumstances. For example, Milinski et al. (1990a) observed that threespine sticklebacks accurately associated with individuals who were less likely to defect during predator inspection interactions after observing them perform predator inspection just four times previously. Similarly, Dugatkin & Alfieri (1991) observed a preference amongst guppies for the better of two inspectors after a period of less than 4 minutes. However, it is difficult to argue that this represents familiarity in its strictest sense given that the extent to which the documented recognition persists outside of the immediate context is not known. The variety of conspecific cues available to fish at any given time raises the possibility that more than one recognition mechanism may be used. It is plausible that fishes may use environmental habitat and diet cues, which are rapidly acquired, to provide general recognition of the individuals encountered whilst also using learned individual recognition, which takes longer to develop and may be associated with higher memory costs, for more fine-scale recognition in specific contexts. For example, it is known that three-spined sticklebacks are capable of general recognition by environmentally mediated chemical cues (Ward et al. 2005), and are capable of highly specific individual recognition in the context of predator inspection (Milinski et al. 1990a) and in a territorial context (Waas & Colgan 1994), but show no evidence of specific individual recognition in a shoal-choice context (Ward et al. 2009).

9.2.3

Functions of associating with familiar fish

By whichever means familiarity is achieved, substantial benefits may be obtained by associating with familiar individuals. Indeed, familiar shoals may be preferred over larger shoals under certain conditions, despite larger groups potentially offering safety in numbers. In a study on European minnows that titrated familiarity against the general preference for larger shoal sizes (Hager & Helfman 1991), focal fishes were given a choice of associating with a shoal of familiar conspecifics or a larger shoal of unfamiliar fishes. The results clearly indicate the importance of shoalmate familiarity as the number of fishes in the alternative, since the unfamiliar shoal needed to be almost double the size of the familiar shoal before the preference of the focal for the familiar shoal finally disappeared (Barber & Wright 2001).

BLBK374-09

BLBK374-Brown

192

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

In general, familiarity appears to enhance the benefits of group living. One of the first tests of the functions of familiarity revealed that shoal of familiars were more cohesive than shoals of unfamiliar fishes (Chivers et al. 1995; Webster et al. 2007), providing a potentially significant advantage to all group members when threatened with predation. Fish associating with familiars also display a reduced investment in epithelial alarm cells, potentially reflecting lower levels of perceived risk (Wisenden & Smith 1998). Greater shoal cohesion is thought to enhance the predator confusion effect, effectively reducing the probability that a predator will be able to make a successful attack (Landeau & Terborgh 1986). Given the clear antipredator benefits attached to familiarity, it has been predicted that the preference for familiars should increase under a predation threat. Despite this, Griffiths (1997) found no significant increase in the preference of minnows for familiar individuals on the appearance of a model predator. Brown (2002) reported similar results using rainbowfish and suggested that in habitats where fishes often experience a predation risk, it is adaptive to maintain a consistent preference for familiars. A uniform association preference is also suggestive of familiarity benefits beyond the scope of an antipredator context. Nevertheless, the tendency to associate with familiars may vary between populations according to predator densities (Magurran et al. 1994). The adaptive benefits of familiarity should theoretically increase in line with predation threat – where predation pressure is low, shoaling tendencies and the preference for familiar individuals may be correspondingly low (Godin et al. 2003). Recent research by Kydd & Brown (2009) on the rainbowfish, Melanotaenia duboulayi, lends support to this idea. They found that captive-reared fishes did not display any association preferences for familiars, while wild-caught specimens did. The implications of this are that domestication and artificial selection may profoundly affect the social behaviour of fishes, a fact which needs to be taken into consideration when designing captive breeding programmes for conservation or aquaculture purposes. Familiarity also has the very significant effect of reducing competition and aggression (Jakobsson 1987; Brick 1998). The ability to recognise and recall the outcome of previous exchanges with competitors should theoretically affect the strategy that an individual employs in subsequent interactions with the same competitor. Specifically, the expectation of future interactions should predispose animals to cooperative behaviour (Dugatkin 1997). Utne-Palm & Hart’s study of competition between sticklebacks revealed that familiar pairs were less likely to contest each prey item and, overall, were more likely to gain an equal share of the available prey items than pairs of unfamiliar fishes. Similarly, Webster & Hart (2007) found a lower instance of food kleptoparasitism between familiar fish than between unfamiliar fish. Despite this, Frommen et al. (2007a) demonstrated that the expression of a preference for familiar individuals in sticklebacks was condition-dependent: Satiated fish showed a consistent preference for familiars, whereas hungry individuals did not. Social status may also mediate the expression of an association preference for familiar fish. Subordinate angelfish, Pterophyllum scalare, prefer to associate with familiar individuals, but only when those individuals are also subordinate. When given a choice between familiar and unfamiliar dominant conspecifics, the subordinate angelfish preferred to associate with the unfamiliar fish. Dominant fish exposed to the same test did not demonstrate any clear preference (Gomez-Laplaza & Fuente 2007). Similar results were reported in the humbug damselfish (Dascyllus aruanus) where subordinate individuals are choosier than dominant

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

193

fish (Jordan et al. 2010). Therefore, the expression of preferences for familiar individuals is mediated by social status in some species. Some of the earliest studies of familiarity were concerned with territoriality and what came to be known as the ‘dear enemy’ effect (Jaeger 1981; Ydenberg et al. 1988) where familiar adjacent territory holders apparently declare an uneasy truce allowing them to deal with unfamiliar interlopers. Familiarity is known to stabilise dominance hierarchies, presumably because it allows individuals to remember the outcome of previous contests and to avoid further costly fighting. In a study on sea trout, H¨ojesj¨o et al. (1998) reported that familiarity stabilised dominance hierarchies and generally reduced the number of aggressive interactions between fishes. Juvenile Atlantic salmon signal their submissiveness to familiar dominant individuals by darkening their body colouration, thereby incurring less direct aggression (O’Connor et al. 2000). Similarly, the ability of rainbow trout to recognise individuals allows third-party observers to gauge the competitive ability of a pair of combatants by watching contests. The observer may then use this information to assess its own chances of prevailing against either of the two fighters in any subsequent contest ˚ kerman 1998). Perhaps as a consequence, Arctic charr that were maintained (Johnsson & A in familiar conspecific groups showed increased survivorship and better overall body condition than those maintained in non-familiar groups over a 21-day period (Sepp¨a et al. 2001). Recently, Griffiths et al. (2004) proposed a more general benefit of familiarity which is likely to apply across different contexts. They proposed that familiarity acts to release some of the constraints on an individual’s time budget. The stable social conditions and the attendant reduction in aggression found in familiar groups allow individuals to spend more time foraging or being vigilant. Empirical findings supported this hypothesis; juvenile brown trout reacted more quickly to a simulated predator attack and gathered more food items when in a group of familiars than when in a group of unfamiliar fishes (Griffiths et al. 2004). Familiarity may also lead to greater foraging efficiency in groups by enabling more effective information transfer, for example Swaney et al. (2001) reported that familiarity facilitated social learning of foraging information in guppies. This finding helps to explain the results of Ward et al. (2005) on sticklebacks where groups of familiar individuals located foraging patches more quickly than groups of unfamiliar fishes and consequently had a higher per capita feeding rate. Morrell et al. (2008) reported similar benefits of foraging efficiency in natural shoals when compared to artificially constructed shoals, potentially as a result of shoalmate familiarity in the former. Familiarity is documented as an important component in the mate-choice decisions of some species. Male guppies direct more of their courtship and mating effort towards unfamiliar females than towards familiar females, thereby reducing duplication of mating attempts towards the same individual females and maximising reproductive success (Kelley et al. 1999). This same preference for unfamiliar females has also been observed in another livebearer, the Panamanian bishop, in both laboratory and field trials (Simcox et al. 2005). By contrast, Thunken et al. (2007) showed that familiarity has no effect on the mate-choice decisions of the cichlid, Pelvicachromis taeniatus. Therefore, the importance of familiarity in mate-choice decisions may vary across different breeding systems, from driving the avoidance of familiar mates in promiscuous mating systems to a much less important role in other systems.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

194

Fish Cognition and Behavior

9.2.4

Familiarity in free-ranging fishes

Trim: 244mm X 172mm

Char Count=

Numerous demonstrations of familiarity-biased association have been reported from laboratory experiments; however, in common with other observed behavioural phenomena it is important to establish that these behaviour patterns also occur in the field, where they may be subject to natural selection. This is especially important in familiarity studies as it may be argued that the space constraints of laboratory aquariums produce environments where the number and frequency of repeated interactions between any two fishes are artificially high. Barber & Ruxton (2000) attempted to address this by designing a laboratory experiment where fishes could move freely between groups. Their results showed non-random patterns of association between fishes, indicating a significant preference for shoaling with familiar individuals. The more naturalistic conditions of Barber & Ruxton’s experiment clearly demonstrated the potential of familiarity to play a significant role in the social organisation of free-ranging fishes. There have been comparatively few investigations into the social dynamics of freeranging fishes, which probably reflect the logistical difficulties of such studies. The results of those studies that have been conducted in the field are somewhat equivocal. Some, such as Helfman’s (1984) study of yellow perch, Hoare et al.’s (2000) study of banded killifish and Hilborn’s (1991) study on skipjack tuna report no evidence for repeated association among particular fishes in a population. Others have provided at least tentative evidence of familiarity-biased association in the field. For example, Klimley & Holloway (1999) report repeated co-occurrences of electronically tagged yellowfin tuna at particular times and places. Ward et al. (2002) observed non-random patterns of association between pairs of free-ranging three-spined sticklebacks over consecutive days that could not be explained by site fidelity or phenotype matching. Advances in tagging technology offer the exciting prospect of gaining much greater understanding of the association patterns of fishes in the wild. A study by Griffiths et al. (2007) of the behaviour of tagged minnows in the seminatural conditions of large flow channel over a period of 3 weeks provides an intriguing insight into the way in which association patterns between familiar individuals change and develop over time. Large-scale studies of migrating populations also indicate a tendency for large groups of individuals to associate over considerable periods of time and distances (e.g. McKinnell et al. 1997; Hay & McKinnell 2002). McKinnell et al.’s (1997) study of migrating steelhead trout demonstrated that individuals form long-term associations in the North Pacific Ocean. Similarly, Hay & McKinnell (2002) documented the movement patterns of over half a million Pacific herring over a period of 14 years and concluded that individuals formed stable temporal and spatial associations. However, it seems unlikely that familiarity is entirely responsible for these patterns. Shoal fidelity in mass migrations of fishes may be promoted by one or more of several different mechanisms, including kin- or population-specific recognition (Quinn & Tolson 1986), activity synchronisation (cf. Conradt & Roper 2000), population specific migration traditions (Warner 1988) or pheromonal attraction (Baker & Montgomery 2001). Alternatively, because migrating fishes tend to remain in large, high-density shoals rather than forming a number of small, loose aggregations, patterns of association may be explained simply by long-term shoal cohesion.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

195

The application of network theory and analysis to free-ranging fish populations provide us with an excellent opportunity to gain a greater understanding of their social dynamics (Watts & Strogatz 1998; Croft et al. 2005). Using network analysis, we can determine the extent of the social connections between individuals in a population. This allows us to make predictions about a range of things including the transmission of information through a network and the alliances between individuals (e.g. Connor et al. 1999). In the context of familiarity, network analysis enables us to establish the conditions under which familiarity develops and potentially to observe the relevance of familiarity to the social organisation of free-ranging fish populations. Studies of populations using these methods, such as those carried out on the Trinidadian guppy (Croft et al. 2003) and the three-spine stickleback (Ward et al. 2002) show that the fishes exhibit highly structured and connected network patterns, consistent with ‘small world’ social networks. Such networks tend to be characterised by high cliquishness, meaning that pairs and small groups of fishes repeatedly co-occur. Such non-random patterns of association in the absence of phenotypic variability suggest that familiarity plays a significant role in structuring the interactions of fishes in the studied populations (Croft et al. 2005). The functional consequences of this were subsequently explored by Croft et al. (2006), who found that the strength of the social connections between individuals in a network was a predictor of their likelihood of them cooperating with one another in a predator-inspection context.

9.2.5

Determinants of familiarity

A number of factors affect the adaptive value of familiarity to wild fishes, including the ecological conditions that are experienced in the wild and the social dynamics of the species under consideration (Ward et al. 2009). For example, the social organisation of female guppies differs from that of males (e.g. Croft et al. 2003). Females spend a greater proportion of their time shoaling and repeatedly interacting with the same individuals whereas males devote more of their time to seeking mating opportunities. Therefore, females have more chance of developing shoalmate familiarity in the first place and may also enjoy greater benefits from such associations than males. This may explain the finding of Griffiths & Magurran (1998) that females demonstrated a preference for same-sex familiars, whereas males did not (although see Croft et al. 2004). It seems unlikely that this is due to limitations on the cognitive ability of males. Rather, males are likely to recognise familiars, but only behave differentially towards them in the context of mate choice. Male guppies are capable of recognising and avoiding familiar females (Kelley et al. 1999). Similarly, male Panamanian bishop fish use familiarity to avoid mating repeatedly with the same subset of females, although the tendency to be choosy decreases with increased predation risk (Simcox et al. 2005). Familiarity is most likely to develop under stable social conditions, which are dependent, in many instances, on ecological conditions. For example, during the dry season guppies are restricted to small pools with relatively small populations, which create ideal conditions for the development of familiarity, whereas during the rainy season the pools are joined into a more continuous river environment. This expands the potential home range of the fish and increases the numbers of different individuals that each might encounter.

BLBK374-09

BLBK374-Brown

196

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

As habitat complexity and the number of conspecifics encountered increases, the development of familiarity is constrained and the difficulties faced by individual fish in discriminating familiar from unfamiliar may also increase. Long-term shoal fidelity is typically rare amongst fishes of shallow freshwater habitats (e.g. Hoare et al. 2000). Whilst such conditions tend to militate against the development of learned individual recognition, it remains possible that small subsets of shoals may remain together for extended periods (Ward et al. 2002). It has been suggested that group membership may be more stable among many coral reef fishes which typically show a high degree of site fidelity, promoting high levels of local familiarity (Mapstone & Fowler 1988), yet despite this Kolm et al. (2005) found no evidence of recognition of familiars despite a high degree of site fidelity in Banggai cardinalfish.

9.3

Familiarity or kin recognition?

A large number of studies documenting strong partner-choice preferences by a wide range of species lead to the prediction that aggregations of fishes in the wild may be composed of non-random assortments of individuals. However, the possibility that fishes captured together are not only familiar with one another, but also related to one another, and that recognition is achieved on the basis of kinship, cannot be discounted. Indeed, many fishes are now understood to have the cognitive ability to distinguish kin from non-kin (Table 9.2). Here, we review kin recognition abilities among fishes and describe the putative benefits of kin-biased behaviour.

9.3.1

Kin recognition theory

In the 1960s Hamilton published his revolutionary model of the evolution of social behaviour (Hamilton 1964), solving the paradox of why animals behave in ways that benefit others but are costly to themselves. Hamilton explained that individuals only assist (or avoid impeding) close relatives, because by doing so they gain inclusive fitness (kin selection) advantages: Increasing the fitness of both the individual and the kin. By behaving altruistically, genes shared by close relatives, that are identical by descent, are preferentially propagated to the next generation. Traditionally, it has been presumed that a prerequisite to individuals accruing kin selection advantages is interaction among relatives, which usually involves kin aggregation (Fletcher & Mitchener 1987; Hepper 1991; Pfennig & Sherman 1995), although it is now accepted that kin-selection advantages may also be conferred through kin avoidance, and reduced competition among relatives (West et al. 2001, 2002). Therefore, initial studies of kin recognition and kin-biased association have focused on species with complex social structures, where individuals live in close proximity to family members. In birds, most work has been done with species that breed cooperatively: Some individuals relinquishing their own breeding attempts to help rear the offspring of kin (Brown 1970; Skutch 1987). More recently, fishes have also proved to be worthwhile study organisms. They clearly have the cognitive skills necessary for distinguishing kin from non-kin (Griffiths 2003;

Year 1999 2000 1986 2001 1982 1989 2006 2008 1992 1993a 1993b 1996b 1993 1996 2002 2004 2001 1986 2008 1980 1992 1999 2005 2007 2004 2000

Author (s)

Anderson & Sabado Arnold Avise & Shapiro Baras & d’Almeida Barnett Beacham Behrmann-Godel et al. Brodeur et al. Brown & Brown

Brown & Brown

Brown & Brown

Brown & Brown

Brown et al. Brown et al. Carlsson & Carlsson Carlsson et al. Courtenay et al. Dowling & Moore Evans & Kelley Ferguson & Noakes FitzGerald & Morrissette Fontaine & Dodson Fraser et al. Frommen et al. Frommen & Bakker Garant et al.

Kelp perch Rainbowfish Serranid reef fish Sharptooth catfish Midas cichlid Coho salmon Perch Atlantic salmon Hatchery rainbow trout Atlantic salmon Rainbow trout Atlantic salmon Hatchery rainbow trout Atlantic salmon Rainbow trout Atlantic salmon Hatchery rainbow trout Arctic charr Brown trout Brown trout Coho salmon Common shiner Trinidadian guppies Common shiner Three-spine stickleback Atlantic salmon Brook charr Three-spine stickleback Three-spine stickleback Atlantic salmon

Common name Brachyistus frenatus Melanotaenia eachamensis Anthias squamipinnis Clarias gariepinus Cichlasoma citrinellum Oncorhynchus kisutch Perca fluviatilis Salmo salar Oncorhynchus mykiss S. salar O. mykiss S. salar O. mykiss S. salar O. mykiss S. salar O. mykiss Salvelinus alpinus Salmo trutta S. trutta Oncorhyncus kisutch Notropis cornutus Poecilia reticulata N. cornutus Gasterosteus aculeatus S. salar Salvenius fontalis G. aculeatus G. aculeatus S. salar

Scientific name

Table 9.2 Summary of investigations of the role of kinship in fish schooling decisions.

Two-choice flow tank Aquarium tanks Molecular analysis Molecular analysis Two-choice flow tank Molecular analysis Aquarium tanks Molecular analysis Two-choice test Molecular analysis Molecular analysis Two-choice test Two-choice test Molecular analysis

Copyeditor’s Name: Trim: 244mm X 172mm

Social Recognition of Conspecifics

(Continued)

16:37

Stream tank

F C W F P/O C W&C W C C? F F C C C C C C W W F W C W C W W Y F W

Level of familiarity

May 14, 2011

Stream tank

A Y N N Y N Y N Y Y Y Y Y Y Y Y Y Y A Y Y N Y Y ? N Y Y Y N

Preference for kin

BLBK374-Brown

Stream tank

Aquarium tanks Two-choice test Molecular analysis Aquarium tanks Two-choice flow tank Aquarium tanks Molecular & two-choice Molecular analysis Two-choice flow tank

Method

BLBK374-09 Char Count=

197

Year 2006 2001 2007 2008 2002 2001 2000 1999 2003 2007 2009 2005 1998 1997 1999 2000 2005 2010 1996 1982 1994 2003 1976 2009 1999 1994 1993 2003 2005 1999

Author (s)

Gerlach & Lysiak Gerlach et al. Gerlach et al Gerlach et al. Griffiths & Armstrong Griffiths & Armstrong Griffiths & Armstrong Griffiths & Magurran Griffiths et al. Hain & Neff Hain & Neff Hansen & Jensen Hauser et al. Herbinger et al. Herbinger et al. Hiscock & Brown Kolm et al Le Vin et al. Lima & Vrijenhoek Loekle et al.

Magurran et al. Mann et al. McKaye & Barlow Mehlis et al. Mjølnerød et al. Moore et al. Naish et al. Neff Neff & Sherman Ojanguren & Bra˜na

Table 9.2 (Continued)

Y Y Y Y Y Y Y N A Y Y N N Y Y Y N Y Y Y Y Y Y Y A Y Y N A Y A&N

Preference for kin Y Y C W C C C C F W C C W W C C F W F P/O P/O W F F P/O W F W W Y C

Level of familiarity

May 14, 2011 16:37 Copyeditor’s Name:

Molecular analysis Two-choice test Two-choice test Two-choice test Molecular analysis Flow tank Molecular analysis Field test Two-choice test Two-choice flow tank

Two-choice test Molecular analysis Two-choice test Two-choice test Stream tank Field test Aquarium Two-choice test Stream tank Molecular & two-choice Aquarium tank Molecular analysis Molecular analysis Molecular analysis Aquarium tanks Two-choice flow tank Molecular analysis Two-choice test Aquarium tank Two-choice test

Method

BLBK374-Brown

Guppy Black molly Trinidadian guppy Zebrafish Midas cichlid Three-spine stickleback Atlantic salmon Atlantic salmon European minnow Bluegill sunfish Bluegill sunfish Brown trout

Danio rerio P. fluviatilis D. rerio D. rerio S. salar S. salar S. salar P. reticulata S. salar P. reticulata Lepomis macrochirus S. salar Limnonthrissa miodon Gadus morhua S. salar S. fontalis Pterapogon kauderni Neolamprologus pulcher Poecilia monarch P. reticulata Poecilia shenops P. reticulata D. rerio C. citrinellum G. aculeatus S. salar S. salar Phoxinus phoxinus Lepomis macrochirus L. macrochirus S. trutta

Scientific name

198

Zebrafish Eurasian perch Zebrafish Zebrafish Atlantic salmon Atlantic salmon Atlantic salmon Trinidadian guppies Atlantic salmon Trinidadian guppies Bluegill Brown trout Tanganyika sardine Cod Hatchery Atlantic salmon Brook trout Banggai cardinalfish Cichlid

Common name

BLBK374-09 Trim: 244mm X 172mm Char Count=

Fish Cognition and Behavior

1989 1997 1996 1996 1998 2002 2004 2008 1998 1999 1985 1986 2004 1999 1998 1988 1996 1992

Arctic charr Arctic charr Arctic charr Brown trout Arctic charr Arctic charr Atlantic salmon Atlantic salmon Three-spine stickleback Mouthbrooding tilapia Coho salmon Coho salmon Trinidadian guppy Three-spine stickleback Common goby Three-spine stickleback Domestic guppy Arctic charr

S. alpinus S. alpinus S. alpinus S. salar S. alpinus S. alpinus S. salar S. salar G. aculeatus Sarotherodon melanotheron O. kisutch O. kisutch P. reticulata G. aculeatus Pomatoschistus microps G. aculeatus P. reticulata S. alpinus

Two-choice flow tank Aquarium tanks Two-choice flow tank Aquarium tanks Two-choice flow tank Two-choice flow tank Molecular analysis Molecular analysis Molecular analysis Molecular analysis Two-choice flow tank Two-choice flow tank Molecular analysis Two-choice flow tank Aquarium tank Two-choice test Two-choice test Two-choice flow tank

Y Y Y Y Y N Y Y N Y Y N N N N Y Y Y

F? C C C W C C C W F C F W C N C F F

BLBK374-Brown May 14, 2011 16:37

Kin discrimination was absent (N), or resulted in kin-association (Y) or kin-avoidance (A). The discriminatory ability of individual fish was tested in two-choice tests or two-choice flow tanks (where water was throughflow). The social behaviour of groups of fishes was observed in aquaria and stream tanks. The possible confounding effect of familiarity on these results are indicated as follows: Test and stimulus fishes housed together before trial and therefore possibly familiar with one another (F), naturally occurring groups of fishes taken from the wild (W), test and stimulus fishes were parent and offspring (P/O) and levels of familiarity between stimulus fish controlled (C) and unknown (?).

Ols´en Ols´en & J¨arvi Ols´en & Winberg Ols´en et al. Ols´en et al. Ols´en et al. Ols´en et al. Palm et al. Peuhkuri & Seppa Pouyaud et al. Quinn & Busack Quinn & Hara Russel et al. Steck et al. Svensson et al. VanHavre & FitzGerald Warburton & Lees Winberg & Ols´en

BLBK374-09 Copyeditor’s Name: Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

199

BLBK374-09

BLBK374-Brown

200

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Ward & Hart 2003) and for the evolution of sociality and cooperative networks (Croft et al. 2006). Furthermore, the improved survival conferred on nepotistic fish has important implications not only for the fishes, but also for the fisheries’ managers and conservation biologists responsible for managing these fish stocks (Parrish et al. 1998; Fr´eon & Misund 1999).

9.3.2

Evidence for kin recognition from laboratory studies

The main tranche of evidence for kin discrimination in fishes comes from juvenile salmonids that prefer the water-borne odour of close relatives to unrelated conspecifics (Table 2; reviewed by Brown & Brown 1996a; Ols´en 1999; Ward & Hart 2003; Griffiths 2003). Juvenile salmonids are ideal candidates for these investigations, not only because their stocks are often economically valuable with financial implications for fisheries’ managers manipulating population structure according to relatedness of stocked individuals, but also because juvenile salmonids are territorial. The effect of group composition on patterns of association is likely to be particularly strong in fishes where dispersal from nests is limited (Elliott 1987) and where dominance hierarchies are stable (Huntingford & Turner 1987). Kinship is also likely to influence patterns of association in other nest-building or livebearing species such as stickleback (VanHavre & FitzGerald 1988; FitzGerald & Morrissette 1992) and Trinidadian guppies. In both examples, stickleback and guppy offspring are reared in sibling groups and thus have the opportunity to learn the identity of kin (Wootton 1985, Magurran 2005). Common rearing of siblings raises the interesting question of whether recognition of related conspecifics can more parsimoniously be attributed to prior experience than to kinship (Grafen 1990). In some studies, test fishes are not only related to, but are also familiar with the stimulus fishes (having previously been housed together or reared together in utero). Many studies have been unable to discount this straightforward explanation (see discussions in Quinn & Hara 1986; Winberg & Ols´en 1992; Brown & Brown 1996a; Griffiths & Magurran 1999; Steck et al. 1999; Krause et al. 2000; Hain & Neff 2007; Sikkel & Fuller 2010). Indeed, Courtenay et al. (1997) and more recently Frommen et al. (2007a) found that common rearing increased the preference for siblings among juvenile coho salmon and three-spined sticklebacks, respectively. Furthermore, Arctic charr reared in isolation from the egg stage do not discriminate kin from non-kin (Winberg & Ols´en 1992; Ols´en & Winberg 1996), suggesting that the recognition template by which fishes discriminate siblings from unrelated individuals is probably learned. Hain & Neff (2007) found that guppies discriminate kin from non-kin using both phenotype matching and familiarity. Nevertheless, studies that control for familiarity demonstrate conclusively that relatedness has an important influence on patterns of association in many species of fishes. Most recently Le Vin et al. (2010) found that kinship alone (and thus phenotype matching) was sufficient to elicit preferential association among cooperatively breeding juvenile African cichlid, Neolamprologus pulcher. In other two-choice tests, preference for unfamiliar kin has been demonstrated for species ranging from Atlantic salmon, coho salmon, Arctic charr, rainbow trout and brown trout to Trinidadian guppies, three-spine stickleback, zebrafish, Eurasian perch and rainbow fish (Table 9.2).

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

9.3.3

Char Count=

201

Advantages of kin discrimination

The advantages of such discriminatory behaviour are twofold. Direct fitness advantages are conferred by reduced levels of aggression among groups of kin compared to non-kin (Brown & Brown 1993a, 1993b; Brown et al. 1996; Ols´en et al. 1996), which lessens the risk of injury (Huntingford & Turner 1987) and reduces the risk of being preyed on (Jakobsson 1987; Brick 1998). The mechanisms by which these advantages are conferred include decreased conspicuousness and, in shoaling species, improved shoal cohesion among kin groups (Evans & Kelley 2008) or increased dispersion among mixed-relatedness groups (Hain & Neff 2007). Furthermore, when aggression is low, animals have more time to forage and attend to other activities (e.g. Roitblat 1987; Krause & Godin 1996). Therefore, the formation and maintenance of stable kin groups provide many potential advantages. Indeed, siblings housed together in aquaria or enclosures have reduced variation in size among individuals (Brown et al. 1996), at least during the early stages of their life (Anderson & Sabado 1999). They also have smaller territories and faster growth (Brown & Brown 1993a, 1996b; Greenberg et al. 2002; Gerlach et al. 2007). Clearly, there are significant incentives for individuals joining a group of related rather than unrelated conspecifics. However, in nature resources may be limited, and fishes compete with one another for food and suitable habitat. Under these conditions, individuals face a dilemma: To behave selfishly by monopolising resources, or to share a proportion of the available resources with close relatives. Future inclusive fitness benefits are weighed against the cost of immediate reduction in resources. Moreover, because the cost of reduced resource domination is likely to be disproportionately visited upon high-ranking individuals that are resource-rich, the decision to behave nepotistically is predicted to be mediated by dominance rank. Griffiths & Armstrong (2002) investigated this possibility in a study of juvenile Atlantic salmon parr in a large indoor stream. It was found that dominant individuals monopolised feeding territories and food until they became satiated. However, after achieving satiation dominant fish shared excess food with subordinate neighbouring territory holders if the neighbours were close relatives (Griffiths & Armstrong 2002). It seems that while the food intake of dominant fish was not reduced by either kin or nonkin neighbouring fish, subordinate individuals were able to increase their food intake by feeding within the territories of dominant kin (Griffiths & Armstrong 2002). Faster growth among subordinate kin, but not subordinate non-kin has also been documented for hatcheryreared Atlantic salmon and rainbow trout (Brown & Brown 1996b). Therefore, it seems that kin may accrue advantages from occasional association within a more fluid territorial structure rather than forming fixed associations. Indeed, recent work has shown that parr have large overlapping home ranges rather than fixed territories (Armstrong et al. 1999; Steingrimsson & Grant 2008).

9.3.4

Kin association in the wild

The strong laboratory evidence that fishes prefer to associate with siblings, gaining direct fitness advantages through increased feeding and growth (reviewed by Brown & Brown 1996a; Ols´en 1999; Ward & Hart 2003; Griffiths 2003) and inclusive fitness advantages by kin selection theory (Hamilton 1964), leads to the prediction that aggregations of fishes in

BLBK374-09

BLBK374-Brown

202

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

the wild should be composed of non-random assortments of related individuals (although see the following text for exception). In territorial species, individuals should preferentially occupy feeding territories adjacent to relatives in order to gain benefits from reduced aggression, while in shoaling species, genetic relatedness should be higher within than between shoals if fishes are to benefit from increased growth and survival. Paradoxically, there is little evidence for kin-biased association patterns in the wild among territorial fishes (Fontaine & Dodson 1999; Garant et al. 2000; Carlsson & Carlsson 2002; Brodeur et al. 2008), but see Mjølnerød et al. (1999) and Carlsson et al. (2004). Fraser et al. (2005) point out that part of the reason for this may lie in the unavoidable technical limitations of using molecular techniques such as allozymes or mitochondrial DNA and more recently in the use of too few microsatellite DNA loci to allow accurate determination of relatedness levels among sampled fishes. Further, Carlsson (2007) points out that kinbiased behaviours may not be expected in populations with large effective population size where the number of siblings and half-siblings is relatively low and opportunity for close kin to interact is reduced. In addition, most genetic analyses of fish populations have calculated the relatedness of fishes caught together in relatively long stream stretches varying from 50 to 300 m (Hansen et al. 1997; Carlsson et al. 1999, 2004; Mjølnerød et al. 1999; Garant et al. 2000). Only two previous studies have investigated how relatedness influences patterns of distribution of territorial fishes at high resolution. Fontaine & Dodson (1999) tested whether juvenile Atlantic salmon occupied territories adjacent to kin. They measured the relatedness of pairs of fishes within five 20 m-long stream stretches but contrary to their predictions found little evidence of siblings defending adjacent territories. Brodeur et al. (2008) also found no evidence for kin association despite mapping the location of parr using snorkelling in preference to electric fishing, in an attempt to minimise possible disturbance and displacement of fishes from their preferred location. It seems that salmon tend not to cluster in kin groups, except as they disperse from nests. Similarly, there is little evidence for kin-biased association patterns among shoals (Table 9.2), although there is some weak evidence of kin structuring within shoals of Eurasian perch (Pouyaud et al. 1999), tilapia (Gerlach et al. 2001), brook charr (Fraser et al. 2005) and Atlantic salmon smolts (Ols´en et al. 2004). Even in species where there is a predicted advantage to kin association, evidence is equivocal. For example, it has been suggested that shoaling with kin may be particularly advantageous for migratory species like salmonids, improving natal homing to breeding areas (Quinn & Busack 1985, Ols´en 1989). Nevertheless, a recent analysis of Atlantic salmon shoals caught in drift nets while foraging in the Baltic Sea did not demonstrate kin-biased association patterns (Palm et al. 2008). Interestingly, however, there was a weak tendency for individuals from the same river to co-occur, and the possibility remains that kin may have associated together in pairs, or small groups within shoals. Furthermore, the number of available kin may not match the optimal group size required for maximum antipredator vigilance or foraging success, so that schooling with non-kin is favoured, despite the associated costs of non-kin association (Aviles et al. 2004). For shoals of smolts migrating towards the sea, Ols´en’s work is interesting since it suggests that juvenile Atlantic salmon smolts migrate downstream in shoals that are largely composed of relatives (Ols´en et al. 2004). Individually tagged (and genetically typed) fishes were identified as they swam past a detector located at the downstream end of the

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

203

stream, and the time between fishes was used as an index of shoal cohesion (Ols´en et al. 2004). Related shoalmates were separated by approximately 35 seconds, while unrelated shoalmates were approximately 42 seconds apart, corresponding to spatial separation of 7.7–15 body lengths (based on fish migration speeds ranging from 0.044 ms−1 (drift speed) to 0.089 ms−1 (fastest fish)). These values exceed the customary definition of shoal membership (3–5 body lengths, Pitcher & Parrish 1993), and suggest, in agreement with Riley (2007), that smolts do not form cohesive schools. Nevertheless there is a clear difference in spacing between related and unrelated fishes. Perhaps this can be explained by work that suggests smolt migration occurs in two phases, beginning with solitary movement, followed by schooling as singletons join together (Bakshtanskiy et al. 1980, 1988; Wood et al. 1993; Riley et al. 2002). The smolts in Ols´en’s study may have been in the process of forming shoals, where the role of environment cues and habitat conditions may be of greater importance than kin-biased behaviour (Olsson et al. 2006), and the intriguing possibility remains that the effect of kinship on shoal composition strengthens as migration proceeds. In the future it may be informative to investigate temporal variation in shoal structure, as fishes make moment-by-moment decisions about whether to join or leave a group.

9.3.5

Explaining the discrepancies between laboratory and field

Discrepancies between behavioural observations of kin recognition among fishes under laboratory conditions and genetic studies of wild fish populations are perplexing. Even within species, laboratory observations of preferential association with related individuals (Van Havre & FitzGerald 1988; FitzGerald & Morrissette 1992) have not been confirmed by field experiments (Peuhkuri & Sepp¨a 1998). Part of the reason is that unnatural conditions are prevalent in many laboratory environments. The confinement of fishes to small and simple habitats for prolonged periods may allow tankmate identity to be learned and stronger associations to be formed than would occur naturally. Furthermore, recirculation of water, and the increased concentration of odour cues in laboratory aquaria, may allow kin recognition to be achieved relatively easily. Recirculation of water and increasing concentration of odour cues instigate heightened aggression in dominant fishes, but only towards non-kin subordinates (Griffiths & Armstrong 2000). When water is not recirculated there is no difference in aggression between kin and non-kin, which suggests that associating with kin may be advantageous, but only in areas of streams where water recirculates, for example in eddies behind rocks or in stream-bed refugia. In natural streams and rivers, kin selection benefits may be modulated by local patterns of water flow (Griffiths & Armstrong 2000). Interestingly, Brodeur et al. (2008) point out that high concentration of odours likely experienced by test fishes under laboratory conditions (of low water volume and flow) may be interpreted as an indicator of high levels of conspecific density and, therefore, competition. Indeed, they quote the densities of juvenile salmonids held in laboratory studies as ranging from 1.85 to 50 m−2 . By comparison, the densities of salmonids in field studies of kin discrimination are usually far lower, varying from 0.27 m−2 (Brodeur et al. 2008) to <1 m−2 (Fontaine & Dodson 1999; Carlsson & Carlsson 2002). It is certainly noteworthy that the only studies successful in documenting aspects of kin-biased behaviours in the wild have sampled populations of fishes where density was relatively high: 2.6 m−2

BLBK374-09

BLBK374-Brown

204

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(Carlsson et al. 2004). Perhaps the observed differences in kin-biased distribution and behaviour patterns between laboratory and field studies can be explained by perceived differences in levels of resource competition among fishes, and the subsequent benefits of sharing limited resources with kin (Fontaine & Dodson 1999; Brodeur et al. 2008). Future work will have to reconcile this possibility (of increased altruism at high levels of competition) with the predictions of, for example West et al. (2001, 2002) that increasing levels of competition between relatives under conditions of high density will reduce levels of kin selection for altruism. Further evidence that concentration of water-borne odour cues is important in discriminating conspecifics is given by Courtenay et al. (2001), Hiscock and Brown (2000) and Steck et al. (1999). Juvenile coho salmon, brook trout and three-spine sticklebacks when given the choice of water containing a high or low concentration of chemical cues from unrelated conspecifics prefer highly concentrated odour (Steck et al. 1999; Hiscock & Brown 2000; Courtenay et al. 2001). Interestingly, the preference by brook trout for strong odour of non-kin over weak odour of kin suggests that under certain conditions, signal strength (odour concentration) overrides information describing relatedness of the putative schoolmates (Hiscock & Brown 2000). Recent work on the relative importance of chemical and visual signals during antipredator behaviour shows that environmental conditions (e.g. water clarity) affect the response of fish to chemical alarm cues (Hartman & Abrahams 2000; Chapter 4). Therefore, it is possible that the clarity of kin odour cues relative to background levels of other chemical information may also change temporally and/or spatially, and may explain why some studies have failed to find any influence of kinship on patterns of growth and association (Beacham 1989; Ojanguren & Bra˜na 1999). The exact nature of the kinship signal is not fully understood. Moore et al. (1994) and Olsen (1987) showed that odours in the urine are important, at least for juvenile Atlantic salmon and Arctic charr. Recently, the role of major histocompatibility complex (MHC) has been highlighted. MHC genes are highly polymorphic (Rammensee et al. 1997), and allow detection and recognition of individulas (reviewed by Penn 2002). Information about kinship may be conveyed by MHC breakdown products as MHC molecules are replaced and degraded into small proteins and peptides that are released through the urine (Brown & Eklund 1994; Milinski et al. 2005, 2009). Olsen et al. (2002) reported that juvenile salmonids showed no preference in a choice between the water-borne odour of a sibling with a different MHC genotype and a non-sibling with an identical MHC genotype. When the choice was between two sibling odours, focal fish demonstrated a preference for siblings with matching MHC genotypes. In the context of mate choice, MHC is used by female three-spined stickleback to optimise the degree of MHC polymorphism and disease resistance in their offspring (Reusch et al. 2001; Aeschlimann et al. 2003; Wegner et al. 2003a, 2003b; Forsberg et al. 2007). It seems that while reduced aggression among kin may promote closely related juveniles to associate with one another, adults may strike an optimal balance between inbreeding and outbreeding by choosing a mate slightly different from immediate kin (Bateson 1983). These findings may have importance for fisheries’ managers aiming to maximise growth and survival of stocked fish since the preference for kin is likely to be lost by inbred or small populations of cultured fish (Frommen et al. 2007b; Mehlis et al. 2009), and growth rate of inbred fish is known to be reduced, at least in coho salmon (Gallardo & Neira 2005).

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

9.3.6

Char Count=

205

Kin avoidance

Most studies of kin discrimination have focused on patterns of kin association, because aggregation of relatives is usually a prerequisite to performing altruistic behaviour and accruing kin selection advantages (Fletcher & Mitchener 1987; Hepper 1991; Emlen 1995; Pfennig & Sherman 1995). However, patterns of distribution other than kin association may also provide opportunities for kin-directed behaviour. Under specific circumstances (such as mate choice), kin selection advantages may be afforded to individuals that avoid rather than associate with kin. Indeed, grouping with relatives can be disadvantageous due to the risks of inbreeding (Charsleworth & Charsleworth 1987). Evidence for kin avoidance during pre-copulatory mate-choice decisions in fishes is limited. The most convincing study shows that female stickleback avoid male siblings and preferring unrelated partners (Frommen & Bakker 2006). However, studies of adults of the cooperatively breeding African cichlid, N. pulcher, have failed to show either kin preference or kin avoidance (Striver et al. 2008) despite the documented ability of juveniles to use relatedness during shoaling decisions (Le Vin et al. 2010). Similarly, in the guppy there is at best only a marginal trend for mating partners to be unrelated (Pitcher et al. 2008; Johnsson et al. 2010) despite an increase in offspring number (Johnson et al. 2010) and quality (Pitcher et al. 2008) produced by unrelated partners. Perhaps kin-biased mate-choice decisions will only be found in populations where inbreeding depression is already present and there are considerable potential benefits to increasing heterozygosity and maintaining genetic variation. Apart from mate-choice decisions, patterns of kin avoidance are also predicted in species where cannibalism is prevalent, or where sibling competition is intense. Perhaps the best examples of kin avoidance in fishes come from cannibalistic species (reviewed by Manica 2002). For example, Loekle et al. (1982) observed that female Trinidadian guppies and black mollies cannibalised the offspring of other females in preference to their own. One explanation is that adult females gain kin selection advantages by avoiding close relatives, and eating only unrelated juveniles (Loekle et al. 1982). However, the possibility also exists that kin-biased dispersal is achieved as a result of the kin-recognition abilities of offspring, as is the case for Poecilia monarch juveniles, who actively disperse from their mother (Lima & Vrijenhoek 1996). In male fishes, filial cannibalism is expected to be most strongly influenced by kinship in species where cuckoldry is commonplace. Indeed in Bluegill sunfish, males distinguish between the water-borne odour of their own offspring and unrelated fry fathered by sneaker males (Neff & Sherman 2005). In contrast, however, Svensson et al. (1998) were unable to show any increases in filial cannibalism in male common gobies whose offspring were likely to have been sired by sneaker males. Other examples of kin-biased cannibalism come from work on three-spine stickleback, where females raid male nests to eat the eggs (Wootton 1985). Males only eat eggs from their own nests if other fishes have begun an attack and it becomes impossible to defend the nest (FitzGerald & Van Havre 1987). Females, on the other hand, avoid raiding nests in which they have spawned (FitzGerald & VanHavre 1987), and after spawning, switch their association preferences from males that provide offspring with optimal allelic diversity, to suboptimal males (Milinski et al. 2005). Presumably this behaviour restricts egg raiding to unrelated eggs (Milinski et al. 2005).

BLBK374-09

BLBK374-Brown

206

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Could sibling competition lead to kin avoidance? At low water temperatures, Atlantic salmon become nocturnal, hiding in streambed crevices during the day and emerging at night to forage in the water column. A study of winter sheltering behaviour in juvenile Atlantic salmon found that at this time salmon preferentially associate with non-kin and avoid sharing shelters with kin (Griffiths et al. 2003), perhaps to avoid imposing the costs of aggregation on close relatives (Waldman 1988). It is possible that kin avoidance may also reduce the degree of competition among siblings since different families of salmonid fishes specialise in different microhabitats (McLaughlin et al. 1999). In agreement with this prediction Carlsson & Carlsson (2002) found that for juvenile brown trout the distance between related territory holders was greater than predicted by chance. It seems that the decision of juvenile salmon to preferentially associate or avoid related conspecifics is not a simple one, but varies seasonally and with local environmental conditions. Indeed, in two field studies of Atlantic salmon (Griffiths & Armstrong 2001) and brown trout (Greenberg et al. 2002), the prevailing conditions favoured greatest advantages to be afforded to fishes that associate with their non-kin. When Atlantic salmon were released into sections of a stream in groups of full siblings or groups from a mixture of eight separate families, the density of fish in genetically diverse groups was almost twice as high as in groups where genetic diversity was low (Griffiths & Armstrong 2001). Similarly, Greenberg et al. (2002) found that juvenile brown trout had higher growth rates when reared in large enclosures with a mixture of siblings and non-siblings compared to single-family groups. It seems that under these conditions the conflicting outcomes of kin selection and heterogeneous advantage (where genetically dissimilar individuals compete less intensely) balance so as to encourage dispersal rather than preferential association of kin. These observations may explain why genetic studies have failed to find evidence of kin aggregating in the wild despite the apparent advantages of such behaviour implicit from the results of laboratory studies. Indeed, work on juvenile brown trout found that the distance between related territory holders was greater than predicted by chance (Carlsson & Carlsson 2002). Importantly, the results suggest not only an ecological advantage to individual juvenile salmon of avoiding relatives, but also an advantage to parents of producing genetically diverse progeny, as these offspring may realise higher densities than juveniles of low genetic diversity (Griffiths & Armstrong 2001).

9.4

Conclusion

Clearly, the ability to recognise and discriminate among conspecifics on the basis of relatedness and familiarity is widespread among fishes. The challenge for researchers in the future will be to examine these patterns in a broader context. Firstly, more data need to be obtained from field studies to provide us with a greater understanding of the importance of relatedness and familiarity for free-ranging fishes. Secondly, the studies of relatedness and familiarity have focused on a relatively small number of species. This has produced a skew in the literature in favour of freshwater over marine species and for those species that inhabit the shallow margins of aquatic habitats, as these are easier to obtain and usually to maintain. In studies of relatedness in fishes, most work has similarly focused on species where there are good a priori reasons for predicting that individuals will have the

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

207

opportunity to learn the identity of kin. For example, species whose life-history strategies result in offspring being reared together (e.g. mouth brooders, nest builders or species where juveniles shoal immediately after birth) are far more likely to have kin-biased behaviour patterns than species with broadcast spawning. Clearly, it is important to examine the importance of relatedness and familiarity in less-structured environments and among limnetic and pelagic species. Thirdly, as a corollary of this, the influence of the ecological conditions on the recognition and preference for related and familiar individuals requires elucidation. For example, habitat complexity and population density are both likely to affect encounter probabilities between related and/or familiar fishes (e.g. Pollock & Chivers 2003; Orpwood et al. 2008). This may be especially important in allowing us to predict the most effective conditions for the generation of familiarity preferences in particular. Fisheries’ managers can modify the physical structures of habitats and social structures of fish communities that will affect the cost-benefit ratios that may influence shoaling behaviour. This may then enable the fisheries industry to further harness the growth and survivorship benefits associated with familiarity. Recent technological advances may allow us to build up a clearer picture of patterns of relatedness and familiarity amongst free-ranging fishes. For example, field studies employing passive integrated transponder (PIT) tags may become more frequent as they are predicted to reduce in both size and cost and improve in terms of their spatial and temporal resolutions. Similarly, the use of microsatellite markers permits us to establish genetic relatedness with much greater accuracy. Furthermore, closer scrutiny of highly polymorphic areas of the genome, notably the MHC, has allowed us to gain a much greater understanding of the role of the genome in mediating association and mating decisions. Given that chemical cues mediated by the MHC are highly individualistic, it is a distinct possibility that this plays a major role in the recognition of both kin and familiar individuals (e.g. Olsen et al. 2002; Penn 2002).

References Aeschlimann, P.B., H¨aberli, M.A., Reusch, T.B.H., Boehm, T. & Milinski, M. (2003) Female sticklebacks Gasterosteus aculeatus use self-reference to optimize MHC allele number during mate selection. Behavioral Ecology and Sociobiology, 54, 119–126. Anderson, T.W. & Sabado, B.D. (1999) Effects of kinship on growth of a live-bearing reef fish. Marine Biology, 133, 115–121. Armstrong, J.D., Huntingford, F.A. & Herbert, N.A. (1999) Individual space use strategies of wild juvenile Atlantic salmon. Journal of Fish Biology, 55, 1201–1212. Arnold, K. (2000) Kin recognition in rainbowfish (Melanotaenia eachamensis): sex, sibs and shoaling. Behavioral Ecology and Sociobiology, 48, 385–391. Aviles, L., Fletcher, J.A. & Cutter, A.D. (2004) The kin composition of social groups: trading group size for degree of altruism. American Naturalist, 164, 132–144. Avise, J.C. & Shapiro, D.Y. (1986) Evaluating kinship of newly settled juveniles within social groups of the coral reef fish Anthia squamippinis. Evolution, 40, 1051–1059. Baker, C.F. & Montgomery, J.C. (2001) Species-specific attraction of migratory banded kokopu juveniles to adult pheromones. Journal of Fish Biology, 58, 1221–1229. Bakshtanskiy, E.L., Nesterov, V.S. & Neklyudov, M.N. (1980) The behaviour of young Atlantic salmon, Salmo salar, during downstream migration. Journal of Ichthyology, 20, 93–100.

BLBK374-09

BLBK374-Brown

208

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Bakshtanskiy, E.L., Nesterov, V.S. & Neklyudov, M.N. (1988) Development of schooling behavior in Atlantic salmon, Salmo salar, during seaward migration. Journal of Ichthyology, 28, 91–101. Baras, E. & d’Almeida, A. (2001) Size heterogeneity prevails over kinship in shaping cannibalism among larvae of sharptooth catfish Clarias gariepinus. Aquatic Living Resources, 14, 251–256. Barber, I. & Ruxton, G. (2000) The importance of stable schooling: do familiar sticklebacks stick together? Proceedings of the Royal Society of London Series B – Biological Sciences, 267, 151–155. Barber, I. & Wright, H. (2001) How strong are familiarity preferences in shoaling fish? Animal Behaviour, 61, 975–979. Barnett, C. (1982) The chemosensory responses of young cichlid fish to parents and predators. Animal Behaviour, 30, 35–42. Bateson, P. (1983) Optimal outbreeding. In: P. Bateson (ed) Mate Choice, pp. 257–277. Cambridge University Press, Cambridge. Beacham, T.D. (1989) Effect of siblings on growth of juvenile coho salmon (Oncorhynchus kisutch). Canadian Journal of Zoology, 67, 601–605. Behrmann-Godel, J., Gerlach, G. & Eckmann, R. (2006) Kin and population recognistion in sypatric Lake Constance perch (Perca fluvitalis L.): can assortative shoaling drive population divergence? Behavioral Ecology and Sociobiology, 59, 461–468. Binoy, V.V. & Thomas, K.J. (2004) The climbing perch (Anabas testudineus Bloch), a freshwater fish, prefers larger unfamiliar shoals to smaller familiar shoals. Current Science, 86, 207–211. Brick, O. (1998) Fighting behaviour, vigilance and predation risk in the cichlid fish, Nannacara anomala. Animal Behaviour, 56, 309–317. Brodeur, N.N., Noel, M.V., Venter, O., Bernatchez, L., Dayanandan, S. & Grant, J.W.A. (2008) No evidence of kin bias in dispersion of young-of-the-year Atlantic salmon Salmo salar L. in a natural stream. Journal of Fish Biology, 73, 2361–2370. Brown, C. (2002) Do female rainbow fish (Melanotaenia spp.) prefer to shoal with individuals under predation pressure? Journal of Ethology, 20, 89–94. Brown, G.E. & Brown, J.A. (1992) Do rainbow trout and Atlantic salmon discriminate kin? Canadian Journal of Zoology, 70, 1636–1640. Brown, G.E. & Brown, J.A. (1993a) Do kin always make better neighbours?: the effects of territory quality. Behavioral Ecology and Sociobiology, 33, 225–231. Brown, G.E. & Brown, J.A. (1993b) Social dynamics in salmonid fishes: do kin make better neighbours? Animal Behaviour, 45, 863–871. Brown, G.E. & Brown, J.A. (1996a) Kin discrimination in salmonids. Reviews in Fish Biology and Fisheries, 6, 201–219. Brown, G.E. & Brown, J.A. (1996b) Does kin-biased territorial behavior increase kin-biased foraging in juvenile salmonids? Behavioral Ecology, 7, 24–29. Brown, G.E., Brown, J.A. & Crosbie, A.M. (1993) Phenotype matching in juvenile rainbow trout. Animal Behaviour, 46, 1223–1225. Brown, G.E., Brown, J.A. & Wilson, W.R. (1996) The effects of kinship on the growth of juvenile Arctic charr. Journal of Fish Biology, 48, 313–320. Brown, G.E. & Smith, R.J.F. (1994) Fathead minnows use chemical cues to discriminate shoalmates from unfamiliar conspecifics. Journal of Chemical Ecology, 20, 3051–3061. Brown, J.A. & Colgan, P.W. (1986) Individual and species recognition in centrarchid fishes: evidence and hypotheses. Behavioral Ecology and Sociobiology, 19, 373–379. Brown, J.L. (1970) Cooperative breeding and altruistic behavior in the Mexican jay, Apelocoma ultramarina. Animal Behaviour, 18, 366–378. Brown, J.L. & Eklund, A. (1994) Kin recognition and the major histocompatibility complex: an integrative review. American Naturalist, 143, 435–461. Bryant, B.P. & Atema, J. (1987) Diet manipulation affects social-behavior of catfish – importance of body odor. Journal of Chemical Ecology, 13, 1645–1661. Carlsson, J. (2007) The effect of family structure on the likelihood for kin-biased distribution: an empirical study of brown trout populations. Journal of Fish Biology, 71(Supplement A), 98–110.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

209

Carlsson, J. & Carlsson, J.E.L. (2002) Micro-scale distribution of brown trout: an opportunity for kin selection? Ecology of Freshwater Fish, 11, 234–239. Carlsson, J., Carlsson, J.E.L., Ols´en KH, Hansen, M.M., Eriksson, T. & Nilsson, J. (2004) Kinbiased distribution in brown trout: an effect of redd location of kin recognition? Heredity, 92, 53–60. Carlsson, J., Ols´en, K.H., Nilsson, J., Øverli, Ø. & Stabell, O.B. (1999) Microsatellites reveal fine-scale genetic structure in stream-living brown trout. Journal of Fish Biology, 55, 1290–1303. Charlesworth, D. & Charlesworth, B. (1987) Inbreeding depression and its evolutionary consequences. Annual Review of Ecology and Systematics, 18, 237–268. Chivers, D., Brown, G. & Smith, R. (1995) Familiarity and shoal cohesion in fathead minnows (Pimephales promelas): implications for antipredator behaviour. Canadian Journal of Zoology, 73, 955–960. Connor, R.C., Heithaus, M.R. & Barre, L.M. (1999) Superalliance of bottlenose dolphins. Nature, 397, 571–572 Conradt, L. & Roper, T.J. (2000) Activity synchrony and social cohesion: a fission-fusion model. Proceedings of the Royal Society of London Series B – Biological Sciences, 267, 2213– 2218. Courtenay, S.C., Quinn, T.P., Dupuis, H.M.C., Groot, C. & Larkin, P.A. (1997) Factors affecting the recognition of population-specific odours by juvenile coho salmon. Journal of Fish Biology, 50, 1042–1060 Courtenay, S., Quinn, T., Dupuis, H., Groot, C. & Larkin, P. (2001) Discrimination of family-specific odours by juvenile coho salmon: roles of learning and odour concentration. Journal of Fish Biology, 58, 107–125. Croft, D.P., Arrowsmith, B.J., Bielby, J., Skinner, K., White, E., Couzin, I.D., Magurran, A.E., Ramnarine, I. & Krause, J. (2003) Mechanisms underlying shoal composition in the Trinidadian guppy, Poecilia reticulata. Oikos, 100, 429–438. Croft, D.P., Arrowsmith, B.J., Webster, M. & Krause, J. (2004) Intra-sexual preferences for familiar fish in male guppies. Journal of Fish Biology, 64, 279–283. Croft, D.P., James, R., Thomas, P.O.R., Hathaway, C., Mawdsley, D., Laland, K.N. & Krause, J. (2006) Social structure and co-operative interactions in a wild population of guppies (Poecilia reticulata). Behavioral Ecology and Sociobiology, 59, 644–650. Croft, D.P., James, R., Ward, A.J.W., Botham, M.S., Mawdsley, D. & Krause, J. (2005) Assortative interactions and social networks in fish. Oecologia, 143, 211–219. Dittman, A.H. & Quinn, T.P. (1996) Homing in Pacific salmon: mechanisms and ecological basis. Journal of Experimental Biology, 199, 83–91. Dowling, T.E. & Moore, W.S. (1986) Absence of population subdivision in the common shiner, Notropis cornutus (Cyprinidae). Environmental Biology of Fishes, 15, 151–155. Dugatkin, L. (1997) Cooperation Among Animals: An Evolutionary Perspective. Oxford University Press, Oxford. Dugatkin, L. & Alfieri, M. (1991) Tit-for-tat in guppies (Poecilia reticulata): the relative nature of cooperation and defection during predator inspection. Evolutionary Ecology, 5, 300–309. Dugatkin, L.A. & Wilson, D.S. (1992) The prerequisites for strategic behaviour in bluegill sunfish, Lepomis macrochirus. Animal Behaviour, 44, 223–230. Elliott, J.M. (1987) The distances travelled by downstream moving trout fry, Salmo trutta, in a lake district stream. Freshwater Biology, 17, 491–499. Emlen, S.T. (1995) An evolutionary theory of the family. Proceedings of the National Academy of Sciences, USA, 92, 8092–8099. Evans, J.P. & Kelley, J.L. (2008) Implications of multiple mating for offspring relatedness and shoaling behaviour in juvenile guppies. Biology Letters, 4, 623–626. Farmer, N.A., Ribble, D.O. & Miller, D.G. (2004) Influence of familiarity on shoaling behaviour in Texas and blacktail shiners. Journal of Fish Biology, 64, 776–782. Ferguson, M.M. & Noakes, D.L.G. (1980) Social grouping and genetic variation in common shiners, Notropis cornutus (Pisces, Cyprinidae). Environmental Biology of Fishes, 6, 357–360.

BLBK374-09

BLBK374-Brown

210

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

FitzGerald, G.J. & vanHavre, N. (1987) The adaptive significance of cannibalism in sticklebacks (Gasterosteidae: Pisces). Behavioural Ecology and Sociobiology, 20, 125–128. FitzGerald, G.J. & Morrissette, J. (1992) Kin recognition and choice of shoal mates by threespine sticklebacks. Ethology Ecology and Evolution, 4, 273–283. Fletcher, D. & Michener, C. (eds) (1987) Kin Recognition in Animals. John Wiley & Sons, New York. Fontaine, P.-M. & Dodson, J.J. (1999) An analysis of the distribution of juvenile Atlantic salmon (Salmo salar) in nature as a function of relatedness using microsatellites. Molecular Ecology, 8, 189–198. Forsberg, L.A., Dannewitz, J., Petersson, E. & Grahn, M. (2007) Influence of genetic dissimiliarity in the reproductive success and mate choice of brown trout – females fishing for optimal MHC dissimilarity. Journal of Evolutionary Biology, 20, 1859–1869. Fraser, D.J., Duchesne, P. & Bernatchez, L. (2005) Migratory charr schools exhibit population and kin associations beyond juvenile stages. Molecular Ecology, 14, 3133–3146. Fr´eon, P. & Misund, O. (1999) Dynamics of Pelagic Fish Distribution and Behaviour: Effects on Fisheries and Stock Assessment. Blackwell Publishing Ltd., Oxford. Frommen, J.G. & Bakker, T.C.M. (2004) Adult three-spined sticklebacks prefer to shoal with familiar kin. Behaviour, 141, 1401–1409. Frommen, J.G. & Bakker, T.C.M. (2006) Inbreeding avoidance through non-random mating in sticklebacks. Biology Letters, 2, 232–235. Frommen, J.G., Luz, C. & Bakker, T.C.M. (2007a) Nutritional state influences shoaling preference for familiars. Zoology, 110, 369–376. Frommen, J.G., Mehlis, M., Brendles, C. & Makker, T.C.M. (2007b) Shoaling decisions in threespined sticklebacks (Gasterosteus aculeatus) – familiarity, kinship and inbreeding. Behavioral Ecology and Sociobiology, 61, 533–539. Gallardo, J.A. & Neira, R. (2005) Environmental dependence of inbreeding depression in cultured Coho salmon (Oncorhynchus kisutch): aggressiveness, dominance and intraspecific competition. Heredity, 95, 449–456. Garant, D., Dodson, J.J. & Bernatchez, L. (2000) Ecological determinants and temporal stability of the within-river population structure in Atlantic salmon (Salmo salar L.). Molecular Ecology, 9, 615–628. Gerlach, G., Hodgins-Davis, A., Avolia, C. & Schunter, C. (2008) Kin recognition in zebrafish: a 24-hour window for olfactory imprinting. Proceedings of the Royal Society of London Series B – Biological Sciences, 275, 2165–2170. Gerlach, G., Hodgins-Davis, A., MacDonald, B. & Hannah, R.C. (2007) Benefits of kin association: related and familiar zebrafish larvae (Danio rerio) show improved growth. Behavioral Ecology and Sociobiology, 61, 1765–1770. Gerlach, G. & Lysiak, N. (2006) Kin recognition and inbreeding avoidance in zebrafish, Danio rerio, is based on phenotype matching. Animal Behaviour, 71, 1371–1377. Gerlach, G., Schardt, U., Eckmann, R. & Meyer, A. (2001) Kin-structured subpopulations in Eurasian perch (Perca fluviatilis L.). Heredity, 86, 213–221. Giaquinto, P. & Volpato, G. (1997) Chemical communication, aggression, and conspecific recognition in the fish Nile tilapia. Physiology and Behavior, 62, 1333–1338. Godin, J.-G.J., Alfieri, M.S., Hoare, D.J. & Sadowski, J.A. (2003) Conspecific familiarity and shoaling preferences in a wild guppy population. Canadian Journal of Zoology, 81, 1899–1904. Gomez-Laplaza, L.M. & Fuente, A. (2007) Shoaling decisions in Angelfish: the roles of social status and familiarity. Ethology, 113, 847–855. Grafen, A. (1990) Do animals really recognize kin? Animal Behaviour, 39, 42–54. Greenberg, L.A., Hernn¨as, B., Br¨onmark, C., Dahl, J., Ekl¨ov, A. & Ols´en, K.H. (2002) Effects of kinship on growth and movements of brown trout in field enclosures. Ecology of Freshwater Fish, 11, 251–259. Griffiths, S. (1997) Preferences for familiar fish do not vary with predation. Journal of Fish Biology, 51, 489–495.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

211

Griffiths, S. & Armstrong, J. (2000) Differential responses of kin and nonkin salmon to patterns of water flow: does recirculation influence aggression? Animal Behaviour, 59, 1019–1023. Griffiths, S. & Armstrong, J. (2002) Kin-biased territory overlap and food sharing among Atlantic salmon juveniles. Journal of Animal Ecology, 71, 480–486. Griffiths, S., Armstrong, J. & Metcalfe, N. (2003) The cost of aggregation: juvenile salmon avoid sharing winter refuges with siblings. Behavioral Ecology, 14, 602–606. Griffiths, S. & Magurran, A. (1997a) Familiarity in schooling fish: how long does it take to acquire? Animal Behaviour, 53, 945–949. Griffiths, S. & Magurran, A. (1997b) Schooling preferences for familiar fish vary with group size in a wild guppy population. Proceedings of the Royal Society of London Series B – Biological Sciences, 264, 547–551. Griffiths, S. & Magurran, A. (1998) Sex and schooling behaviour in the Trinidadian guppy. Animal Behaviour, 56, 689–693. Griffiths, S. & Magurran, A. (1999) Schooling decisions in guppies (Poecilia reticulata) are based on familiarity rather than kin recognition by phenotype matching. Behavioral Ecology and Sociobiology, 45, 437–443. Griffiths, S.W. (2003) Learned recognition of conspecifics by fishes. Fish and Fisheries, 4, 256–268. Griffiths, S.W. & Armstrong, J.D. (2001) The benefits of genetic diversity outweigh those of kin association in a territorial animal. Proceedings of the Royal Society of London Series B – Biological Sciences, 268, 1293–1296. Griffiths, S.W., Brockmark, S., Hojesjo, J. & Johnsson, J.I. (2004) Coping with divided attention: the advantage of familiarity. Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 695–699. Griffiths, S.W., Ojanguren, A.F., Orpwood, J.E., Magurran, A.E. & Armstrong, J.D. (2007) Familiaritybiased patterns of association shift with time among European minnows. Journal of Fish Biology, 71, 1602–1612. Hager, M.C. & Helfman, G.S. (1991) Safety in numbers – shoal size choice by minnows under predatory threat. Behavioral Ecology and Sociobiology, 29, 271–276. Hain, T.J.A. & Neff, B.D. (2007) Multiple paternity and kin recognition mechanisms in a guppy population. Molecular Ecology, 16, 3938–3946. Hain, T.J.A. & Neff, B.D. (2009) Kinship affects innate responses to a predator in bluegill Lepomis macrochirus larvae. Journal of Fish Biology, 75, 728–737. Hamilton, W.D. (1964) The genetical evolution of social behaviour. Journal of theoretical Biology, 7, 1–52. Hansen, M.M. & Jensen, L.F. (2005) Sibship within samples of brown trout (Salmo trutta) and implications for supportive breeding. Conservation Genetics, 6, 297–305. Hansen, M.M., Nielsen, E.E. & Mensberg, K.-L.D. (1997) The problem of sampling families rather than populations: relatedness among individuals in samples of juvenile brown trout Salmo trutta L. Molecular Ecology, 6, 469–474. Hartman, E.J. & Abrahams, M.V. (2000) Sensory compensation and the detection of predators: the interaction between chemical and visual information. Proceedings of the Royal Society of London Series B – Biological Sciences, 267, 571–575. Hasler, A.D. & Scholz, A.T. (1983) Olfactory Imprinting and Homing in Salmon. Springer-Verlag, Berlin. Hauser, L., Carvalho, G.R. & Pitcher, T.J. (1998) Genetic population structure in the Lake Tanganika sardine Limnothrissa miodon. Journal of Fish Biology, 53, 413–429. Hay, D.E. & McKinnell, S.M. (2002) Tagging along: association among individual Pacific herring (Clupea pallasi) revealed by tagging. Canadian Journal of Fisheries and Aquatic Sciences, 59, 1960–1968. Helfman, G.S. (1984) School fidelity in fishes: the yellow perch pattern. Animal Behaviour, 32, 663–672. Hepper, P.G. (ed) (1991) Kin Recognition. Cambridge University Press, Cambridge.

BLBK374-09

BLBK374-Brown

212

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Herbinger, C., Doyle, R., Taggart, C., Lochmann, S., Brooker, A., Wright, J. & Cook, D. (1997) Family relationships and effective population size in a natural cohort of Atlantic cod (Gadus morhua) larvae. Canadian Journal of Fisheries and Aquatic Sciences, 54, 11–18. Herbinger, C.M., O’Reilly, P.T., Doyle, R.W., Wright, J.M. & O’Flynn, F. (1999) Early growth performance of Atlantic salmon full-sib families reared in single family tanks versus in mixed family tanks. Aquaculture, 173, 105–116. Hilborn, R. (1991) Modelling the stability of fish schools: exchange of individual fish between schools of skipjack tuna (Katsuwonus pelamis). Canadian Journal of Fisheries and Aquatic Sciences, 48, 1081–1091. Hiscock, M.J. & Brown, J.A. (2000) Kin discrimination in juvenile brook trout (Salvelinus fontalis) and the effect of odour concentration on kin preferences. Canadian Journal of Zoology, 78, 278–282. Hoare, D.J., Ruxton, G.D., Godin, J.-G.J. & Krause, J. (2000) The social organization of free-ranging fish shoals. OIKOS, 89, 546–554. H¨ojesj¨o, J., Johnsson, J.I., Petersson, E. & J¨arvi, T. (1998) The importance of being familiar: individual recognition and social behavior in sea trout (Salmo trutta). Behavioral Ecology, 9, 445–451. Huntingford, F.A. & Turner, A. (1987) Animal Conflict. Chapman and Hall, London. Jaeger, R.G. (1981) Dear enemy recognition and the costs of aggression between salamanders. American Naturalist, 117, 962–974. Jakobsson, S. (1987) Male Behaviour in Conflicts Over Mates and Territories. PhD thesis, University of Stockholm, Stockholm. Johnson, A.M., Chappell, G., Price, A., Rodd, F.H., Olendorf, R. & Hughes, K.A. (2010) Inbreeding depression and inbreeding avoidance in a natural population of guppies (Poecilia reticulate). Ethology, 116, 448–457. Johnsson, J.J. (1997) Individual recognition affects aggression and dominance relations in rainbow trout, Oncorhynchus mykiss. Ethology, 103, 267–282. ˚ kerman, A. (1998) Watch and learn: preview of the fighting ability of opponents Johnsson, J.J. & A alters contest behaviour in rainbow trout. Animal Behaviour, 56, 771–776. Jordan, L.A., Avolio, C., Herbert-Read, J.E., Krause, J., Rubenstein, D.I. & Ward, A.J.W. (2010) Group structure in a restricted entry system is mediated by both resident and joiner preferences. Behavioral Ecology and Sociobiology, doi: 10.1007/s00265-010-0924-1. Keenleyside, M.H.A. (1955) Some aspects of the schooling behaviour of fish. Behavior, 8, 83–248. Kelley, J., Graves, J. & Magurran, A.E. (1999) Familiarity breeds contempt in guppies. Nature, 401, 661–662. Klimley, A.P. & Holloway, C.F. (1999) School fidelity and homing synchronicity of yellowfin tuna, Thunnus albacares. Marine Biology, 133, 307–317. Kolm, N., Hoffman, E.A., Olsson, J., Berglund, A. & Jones, A.G. (2005) Group stability and homing behavior but no kin group structures in a coral reef fish. Behavioral Ecology, 16, 521–527. Krause, J., Butlin, R., Peuhkuri, N. & Pritchard, V. (2000) The social organization of fish shoals: a test of the predictive power of laboratory experiments for the field. Biological Reviews, 75, 477–501. Krause, J. & Godin, J.-G.J. (1996) Influence of prey foraging posture on flight behavior and predation risk: predators take advantage on unwary prey. Behavioral Ecology, 7, 264–271. Kydd, E. & Brown, C. (2009) Loss of shoaling preference for familiar individuals in captive-reared crimson spotted rainbowfish Melanotaenia duboulayi. Journal of Fish Biology, 74, 2187–2195. Lachlan, R., Crooks, L. & Laland, K. (1998) Who follows whom? Shoaling preferences and social learning of foraging information in guppies. Animal Behaviour, 56, 181–190. Landeau, L. & Terborgh, J. (1986) Oddity and the ‘confusion effect’ in predation. Animal Behaviour, 34, 1372–1380. Le Vin, A.L., Mable, B.K. & Arnold, K.E. (2010) Kin recognition via phenotype matching in a cooperatively breeding cichlid, Neolamprologus pulcher. Animal Behaviour, 79, 1109–1114. Lima, N.R.W. & Vrijenhoek, R.C. (1996) Avoidance of filial cannibalism by sexual and clonal forms of Poeciliopsis (Pisces: Peociliidae). Animal Behaviour, 51, 293–301.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

213

Loeckle, D.M., Madison, D.M. & Christian, J.J. (1982) Time dependency and kin recognition of cannibalistic behaviour among Poeciliid fishes. Behavioural and Neural Biology, 35, 315–318. Magurran, A., Seghers, B., Shaw, P. & Carvalho, G. (1994) Schooling preferences for familiar fish in the guppy, Poecilia reticulata. Journal of Fish Biology, 45, 401–406. Magurran, A.E. (2005) Evolutionary Ecology: The Trinidadian Guppy. Oxford University Press, London. Manica, A. (2002) Filial cannibalism in teleost fish. Biological Reviews, 77, 261–277. Mann, K.D., Turnell, E.R., Atema, J. & Gerlach, G. (2003) Kin recognition in juvenile zebrafish (Danio rerio) based on olfactory cues. Biological Bulletin, 205, 224–225. Mapstone, B.D. & Fowler, A.J. (1988) Recruitment and structure of assemblages of fish on coral reefs. Trends in Ecology and Evolution, 3, 72–77. McKaye, K. & Barlow, G. (1976) Chemical recognition of young by the midas cichlid, Cichlasoma citrinellum. Copeia, 1965, 276–282. McKinnell, S., Pella, J.J. & Dahlberg, M.L. (1997) Population-specific aggregations of steelhead trout (Oncorhynchus mykiss) in the North Pacific Ocean. Canadian Journal of Fisheries and Aquatic Sciences, 54, 2368–2376. McLaughlin, R.L., Ferguson, M.M. & Noakes, D.L.G. (1999) Adaptive peaks and alternative foraging tactics in brook charr: evidence of short-term divergent selection for sitting-and-waiting and actively searching. Behavioral Ecology and Sociobiology, 45, 386–395. McRobert, S.P. & Bradner, J. (1998) The influence of body coloration on shoaling preferences in fish. Animal Behaviour, 56, 611–615. Mehlis, M., Bakker, T.C.M., Langen, K. & Frommen, J.G. (2009) Cain and Abel reloaded? Kin recognition and male–male aggression in three-spined sticklebacks. Gasterosteus aculeatus L. Journal of Fish Biology, 75, 2154–2162. Metcalfe, N.B. & Thomson, B.C. (1995) Fish recognize and prefer to shoal with poor competitors. Proceedings of the Royal Society of London Series B – Biological Sciences, 259, 207–210. ´ Haller, J. & Cs´anyi, V. (1992) Different duration of memory for consepcific and hetMikl´osi, A., erospecific fish in the paradise fish (Macropodus opercularis L.). Ethology, 90, 29–36. Milinski, M., Griffiths, S., Wegner, K.M., Reusch, T.B.H., Haas-Assenbaum, A. & Boehm, T. (2005) Mate choice decisions of stickleback females predictably modified by MHC peptide ligands. Proceedings of the National Academy of Sciences of the United States of America, 102, 4414–4418. Milinski, M., Griffiths, S.W., Reusch, T.B. & Boehm, T. (2009) Costly major histocompatibility complex signals produced only by reproductively active males, but not females, must be validated by a ‘maleness signal’ in three-spined sticklebacks. Proceedings of the Royal Society of London Series B – Biological Sciences, published online, doi:10.1098/rspb.2009.1501 Milinski, M., Kulling, D. & Kettler, R. (1990b) Tit for tat: sticklebacks ‘trusting’ a cooperating partner. Behavioral Ecology, 1, 7–12. Milinski, M., Pfluger, D., Kulling, D. & Kettler, R. (1990a) Do sticklebacks cooperate repeatedly in reciprocal pairs? Behavioral Ecology and Sociobiology, 27, 17–21. Mjølnerød, I.B., Refseth, U.H. & Hindar, K. (1999) Spatial association of genetically similar Atlantic salmon juveniles and sex bias in spatial patterns in a river. Journal of Fish Biology, 55, 1–8. Moore, A., Ives, M.J. & Kell, L.T. (1994) The role of urine in sibling recognition in Atlantic salmon Salmo salar (L.) parr. Proceedings of the Royal Society of London Series B – Biological Sciences, 255, 173–180. Morrell, L.J., Croft, D.P., Dyer, J.R.G., Chapman, B.B., Kelley, J.L., Laland, K.N. & Krause, J. (2008) Association patterns and foraging behaviour in natural and artificial guppy shoals. Animal Behaviour, 76, 855–864. Naish, K.-A., Carvalho, G.R. & Pitcher, T.J. (1993) The genetic structure and microdistribution of shoals of Phoxinus phoxinus, the European minnow. Journal of Fish Biology, 43(Supplement A), 75–89. Neff, B.D. (2003) Paternity and condition affect cannibalistic behavior in nest-tending bluegill sunfish. Behavioral Ecology and Sociobiology, 54, 377–384.

BLBK374-09

BLBK374-Brown

214

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Neff, B.D. & Sherman, P.W. (2005) In vitro fertilization reveals offspring recognition via selfreferencing in a fish with paternal care and cuckoldry. Ethology, 111, 425–438. O’Connor, K.I., Metcalfe, N.B. & Taylor, A.C. (2000) Familiarity influences body darkening in territorial disputes between juvenile salmon. Animal Behaviour, 59, 1095–1101. Ojanguren, A. & Bra˜na, F. (1999) Discrimination against water containing unrelated consepcifics and a marginal effect of relatedness on spacing behaviour and growth in juvenile brown trout, Salmo trutta L. Ethology, 105, 937–948. Ols´en, H. (1999) Present knowledge of kin discrimination in salmonids. Genetica, 104, 295–299. Ols´en, K., Grahn, M. & Lohm, J. (2002) Influence of mhc on sibling discrimination in Arctic char, Salvelinus alpinus (L.). Journal of Chemical Ecology, 28, 783–795. Ols´en, K. & Winberg, S. (1996) Learning and sibling odor preference in juvenile Arctic charr, Salvelinus alpinus (L.). Journal of Chemical Ecology, 22, 773–786. Olsen, K.H. (1987) Chemoattraction of juvenile Arctic charr (Salvelinus alpinus, L) to water scented by conspecific intestinal content and urine. Comparative Biochemistry and Physiology A, 87, 641–643. Ols´en, K.H. (1989) Sibling recognition in juvenile Arctic charr, Salvelinus alpinus (L.). Journal of Fish Biology, 34, 571–581. Olsen, K.H., Grahn, M. & Lohm, J. (2003) The influence of dominance and diet on individual odours in MHC identical juvenile Arctic charr siblings. Journal of Fish Biology, 63, 855–862. Ols´en, K.H., Grahn, M., Lohm, J. & Langefors, A. (1998) MHC and kin discrimination in juvenile Arctic charr, Salvelinus alpinus (L.). Animal Behaviour, 56, 319–327. Ols´en, K.H. & J¨arvi, T. (1997) Effects of kinship on aggression and RNA content in juvenile Arctic charr. Journal of Fish Biology, 51, 422–435. Ols´en, K.H., J¨arvi, T. & L¨of, A.-C. (1996) Aggressiveness and kinship in brown trout. Behavioral Ecology, 7, 445–450. Ols´en, K.H., Petersson, E., Ragnarsson, B., Lundqvist, H. & Jarvi, T. (2004) Downstream migration in Atlantic salmon (Salmo salar) smolt sibling groups. Canadian Journal of Fisheries and Aquatic Sciences, 61, 328–331. Olsson, I.C., Greenberg, L.A., Bergman, E. & Wysujack, K. (2006) Environmentally induced migration: the importance of food. Ecology Letters, 9, 645–651. Orpwood, J.E., Magurran, A.E., Armstrong, J.D. & Griffiths, S.W. (2008) Minnows and the selfish herd: effects on predation risk on shoaling behaviour are dependent on habitat complexity. Animal Behaviour, 76, 143–152. Palm, S., Dannewitz, J., Jarvi, T., Koljonen, M.-L., Prestegaard, T. & Ols´en, K.H. (2008) No indications of Atlantic salmon (Salmo salar) shoaling with kin in the Baltic Sea. Canadian Journal of Fisheries and Aquatic Sciences, 65, 1738–1748. Parrish, D., Behnke, R., Gephard, S., McCormick, S. & Reeves, G. (1998) Why aren’t there more Atlantic salmon (Salmo salar)? Canadian Journal of Fisheries and Aquatic Sciences, 55, 281–287. Penn, D.J. (2002) The scent of genetic compatibility: sexual selection and the major histocompatibility complex. Ethology, 108, 1–21. Peuhkuri, N. & Sepp¨a, P. (1998) Do three-spined sticklebacks group with kin? Annales Zoologici Fennici, 35, 21–27. Pfennig, D.W. & Sherman, P.W. (1995) Kin recognition. Scientific American, June, 98–103. Pitcher, T.E., Rodd, F.H. & Rowe, L. (2008) Female choice and the relatedness of mates in the guppy (Poecilia reticulata). Genetica, 134, 137–146. Pitcher, T.J. & Parrish, J.K. (1993) Functions of shoaling behaviour in teleosts. In: T.J. Pitcher (ed) Behaviour of Teleost Fishes, pp. 363–439. Chapman & Hall, London. Pollock, M.S. & Chivers, D.P. (2003) Does habitat complexity influence the ability of fathead minnows to learn heterospecific chemical alarm cues? Canadian Journal of Zoology, 81, 923–927. Pouyaud, L., Desmarais, E., Chenuil, A., Agnese, J. & Bonhomme, F. (1999) Kin cohesiveness and possible inbreeding in the mouthbrooding tilapia Sarotherodon melanotheron (Pisces Cichlidae). Molecular Ecology, 8, 803–812.

BLBK374-09

BLBK374-Brown

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Recognition of Conspecifics

Char Count=

215

Quinn, T. & Hara, T. (1986) Sibling recognition and olfactory sensitivity in juvenile coho salmon (Oncorhynchus kisutch). Canadian Journal of Zoology, 64, 921–925. Quinn, T.P. & Busack, C.A. (1985) Chemosensory recognition of siblings in juvenile coho salmon (Oncorhynchus kisutch). Animal Behaviour, 33, 51–56. Quinn, T.P. & Tolson, G.M. (1986) Evidence of chemically mediated population recognition in coho salmon (Oncorhynchus kisutch). Canadian Journal of Zoology, 64, 84–87. Rammensee, H.G., Bachmann, J. & Stefanovic, S. (1997) MHC Ligands and Peptide Motifs. Georgetown, Landes Bioscience. Reusch, T.B.H., H¨aberli, M.A., Aeschlimann, P.B. & Milinski, M. (2001) Female sticklebacks count alleles in a strategy of sexual selection explaining MHC polymorphism. Nature, 414, 300–302. Riley, W.D. (2007) Seasonal downstream movements of juvenile Atlantic salmon, Salmo salar L., with evidence of solitary migration of smolts. Aquaculture, 273, 194–199. Riley, W.D., Eagle, M.O. & Ives, S.J. (2002) The onset of downstream movement of juvenile Atlantic salmon, Salmo salar L., in a chalk stream. Fisheries Management & Ecology, 9, 87–94. Roitblat, H.L. (1987) Introduction to Comparative Cognition. W. H. Freeman & Co, New York. Russel, S.T., Kelley, J.L., Graves, J.A. & Magurran, A.E. (2004) Kin structure and shoal composition dynamics in the guppy, Poecilia reticulata. Oikos, 106, 520–526. Sepp¨a, T., Laurila, A., Peuhkuri, N., Piironen, J. & Lower, N. (2001) Early familiarity has fitness consequences for Arctic charr (Salvelinus alpinus) juveniles. Canadian Journal of Fisheries and Aquatic Science, 58, 1380–1385. Sikkel, P.C. & Fuller, C.A. (2010) Shoaling preference and evidence for maintenance of sibling groups by juvenile black perch Embiotoca jacksoni. Journal of Fish Biology, 76, 1671–1616. Simcox, H., Colegrave, N., Heenan, A., Howard, C. & Braithwaite, V.A. (2005) Context-dependent male mating preferences for unfamiliar females. Animal Behaviour, 70, 1429–1437. Skutch, A.F. (1987) Helpers at Bird’s Nests. University of Iowa Press, Iowa City. Steck, N., Wedekind, C. & Milinski, M. (1999) No sibling odor preference in juvenile threespined sticklebacks. Behavioral Ecology, 10, 493–497. ´ & Grant, J.W.A. (2008) Multiple central-place territories in wild young-of-theSteingr´ımsson, S.O. year Atlantic salmon Salmo salar. Journal of Animal Ecology, 77, 448–457. Striver, K.A., Fitzpatrick, J.L., Desjardins, J.K., Neff, B.D., Quinn, J.S. & Blashine, S. (2008) The role of genetic relatedness among social mates in a cooperative breeder. Behavioral Ecology, 19, 816–823. Svensson, O., Magnhagen, C., Forsgren, E. & Kvarnemo, C. (1998) Parental behaviour in relation to the occurrence of sneaking in the common goby. Animal Behaviour, 56, 175–179. Swaney, W., Kendal, J., Capon, H., Brown, C. & Laland, K. (2001) Familarity facilitates learning of foraging behaviour in the guppy. Animal Behaviour, 62, 591–598. Thunken, T., Bakker, T.C.M., Baldauf, S.A. & Kullmann, H. (2007) Direct familiarity does not alter mating preference for sisters in male Pelvicachromis taeniatus (Cichlidae). Ethology, 113, 1107–1112. Utne-Palm, A. & Hart, P. (2000) The effects of familiarity on competitive interactions between threespined sticklebacks. OIKOS, 91, 225–232. VanHavre, N. & FitzGerald, G.J. (1988) Shoaling and kin recognition in the threespine stickleback (Gasterosteus aculeatus L.) Biological Behaviour, 13, 190–201. Waas, J.R. & Colgan, P.W. (1994) Male sticklebacks can distinguish between familiar rivals on the basis of visual cues alone. Animal Behaviour, 47, 7–13. Waldman, B. (1988) The ecology of kin recognition. Annual Review of Ecology and Systematics, 19, 543–571. Warburton, K. (2003) Learning of foraging skills by fish. Fish and Fisheries, 4, 203–215. Warburton, K. & Lees, N. (1996) Species discrimination in guppies: learned responses to visual cues. Animal Behaviour, 52, 371–378. Ward, A., Axford, S. & Krause, J. (2003) Cross-species familiarity in shoaling fishes. Proceedings of the Royal Society of London Series B – Biological Sciences, 270, 1157–1161.

BLBK374-09

BLBK374-Brown

216

May 14, 2011

16:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Ward, A., Botham, M., Hoare, D., James, R., Broom, M., Godin, J. & Krause, J. (2002) Association patterns and shoal fidelity in the three-spined stickleback. Proceedings of the Royal Society of London Series B – Biological Sciences, 269, 2451–2455. Ward, A.J.W. & Hart, P. (2003) The effects of kin and familiarity on interactions between fish. Fish and Fisheries, 4, 348–358. Ward, A.J.W., Hart, P. & Krause, J. (2004) The effects of habitat- and diet-based cues on association preferences in three-spined sticklebacks. Behavioral Ecology, 15, 925–929. Ward, A.J.W., Holbrook, R., Krause, J. & Hart, P. (2005) Social recognition in sticklebacks: the role of direct experience and habitat cues. Behavioral Ecology and Sociobiology, 57, 575–583. Ward, A.J.W. & Krause, J. (2001) Body length assortative shoaling in the European minnow, Phoxinus phoxinus. Animal Behaviour, 62, 617–621. Ward, A.J.W., Webster, M.M. & Hart, P.J.B. (2007) Social recognition in wild fish populations. Proceedings of the Royal Society Series B – Biological Sciences, 274, 1071–1077. Ward, A.J.W., Webster, M.M., Magurran, A.E., Currie, S. & Krause, J. (2009) Species and population differences in social recognition between fishes: a role for ecology? Behavioral Ecology, 20, 511–516. Warner, R. (1988) Traditionality of mating-site preferences in a coral reef fish. Nature, 335, 719–721. Watts, D. & Strogatz, S. (1998) Collective dynamics of ‘small-world’ networks. Nature, 393, 440–442. Webster, M.M., Goldsmith, J., Ward, A.J.W. & Hart, P.J.B. (2007) Habitat-specific chemical cues influence association preferences and shoal cohesion in fish. Behavioral Ecology and Sociobiology, 62, 273–280. Webster, M.M. & Hart, P.J.B. (2007) Prior association reduces kleptoparasitic prey competition in shoals of three-spined sticklebacks. Animal Behaviour, 74, 253–258. Wegner, K.M., Kalbe, M., Kurtz, J., Reusch, T.B.H. & Milinski, M. (2003b) Parasite selection for immunogenetic optimality. Science, 301, 1343. Wegner, K.M., Reusch, T.B.H. & Kalbe, M. (2003a) Multiple parasite species are driving major histocompatibility complex polymorphism in the wild. Journal of Evolutionary Biology, 16, 233–241. West, S.A., Murray, M.G., Machado, C.A., Griffin, A.S. & Herre, E.A. (2001) Testing Hamilton’s rule with competition between relatives. Nature, 409, 510–513. West, S.A., Pen, I. & Griffin, A.S. (2002) Cooperation and competition between relatives. Science, 296, 72–75. Winberg, S. & Ols´en, K.H. (1992) The influence of rearing conditions on the sibling odour preference of juvenile Arctic charr Salvelinus alpinus (L.). Animal Behaviour, 44, 157–164. Wisenden, B. & Smith, R. (1998) A re-evaluation of the effect of shoalmate familiarity on the proliferation of alarm substance cells in ostariophysan fishes. Journal of Fish Biology, 53, 841–846. Wisenden, B.D. (2000) Olfactory assessment of predation risk in the aquatic environment. Philosophical Transactions of the Royal Society of London Series B – Biological Sciences, 355, 1205–1208. Wood, C.C., Hargreaves, N.B., Rutherford, D.T. & Emmett, B.T. (1993) Downstream and early marine migratory behaviour of sockeye salmon (Oncorhynchus nerka) smolts entering Barkley Sound, Vancouver Island. Canadian Journal of Fisheries and Aquatic Sciences, 50, 1329–1337. West, S.A., Murray, M.G., Machado, C.A., Griffin, A.S. & Herre, E.A. (2001) Testing Hamilton’s rule with competition between relatives. Nature, 409, 510–513. West, S.A., Pen, I. & Griffin, A.S. (2002) Cooperation and competition between relatives. Science, 296, 72–75. Wootton, R.J.A. (1985) Functional Biology of Sticklebacks, Croom Helm, London. Ydenberg, R.C., Giraldeau, L.A. & Falls, J.B. (1988) Neighbors, strangers, and the asymmetric war of attrition. Animal Behaviour, 36, 343–347.

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 10

Social Organisation and Information Transfer in Schooling Fish Christos C. Ioannou, Iain D. Couzin, Richard James, Darren P. Croft and Jens Krause

10.1

Introduction

Information acquisition and processing is integral to the lives of most animals (Dall et al. 2005) and is especially important for animals such as fishes that live in spatially and temporally dynamic environments. Changes in predator and food distributions occur at varying scales; for example, fish may respond in fractions of a second to an imminently attacking predator (Magurran & Pitcher 1987) or may migrate over miles to spawn in a lowpredator habitat (Makris et al. 2009). Most of the cognitive tasks necessary for survival can be achieved by individual fish; however, social interactions can greatly improve the absolute ability to complete a task and/or its efficiency by reducing associated costs. Social learning research has made significant advances in our understanding of acquiring and processing information by individuals (see Chapter 11). Whereas this research concerns learning, i.e. information that is committed to memory, a more transient form of information transfer can occur within coordinated groups of individuals. In shoals of fish, where membership is often fluid and groups large, individuals are unlikely to be aware of the information held by others within their sensory range. Instead, the biasing of each individual’s directed motion with that of near neighbours can allow for a flexible structure with an improved ability to acquire and disseminate information. Although much has been written on the use of ‘inadvertent social information’ or ‘public information’ (e.g. Danchin et al. 2004), less is known about issues relating to the collective capacity for information acquisition, the filtering of environmental and/or sensory noise, information storage or collective decisionmaking in animal groups. It is particularly unclear what kind of information processing can be achieved by the collective that is not possible, or is extremely difficult, for individuals in isolation. Such collective processing, previously little recognised, could provide additional insights, especially if integrated into current social learning theory. We also consider the potential for information transfer in wild populations of a fish species (the guppy, Poecilia reticulata) for which studies of the networks of social Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-10

BLBK374-Brown

218

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

interactions can be informative. The number of possible pair-wise interactions becomes enormous in such networks, yet, as we shall demonstrate, guppy behaviour results in nonrandom associations correlated both with individual sex, phenotype and environmental structure that may facilitate effective transfer of pertinent information among individuals (Krause et al. 2007; Croft et al. 2008).

10.2

Collective motion

Understanding how individual fish interact with one another in the absence of stimuli is a necessary precursor to understanding information transfer within shoals. Many fish schools perform highly integrated collective motion (Radakov 1973; Partridge 1982; Magurran & Pitcher 1987), so much so that they can appear like some animate fluid as they change volume, direction and shape with density waves propagating across the group (Radakov 1973), yet still remaining a cohesive whole. The movement rules, which these usually unrelated individuals employ, are based on two general observations of interacting individuals (Parr 1927; Breder Jr 1954). First, individuals avoid collisions with others if they get too close (as shown by Partridge et al. 1980; Tien et al. 2004), and secondly, if individuals are not performing avoidance behaviour, they will tend to be both attracted to (Tegeder & Krause 1995), and to align with (Viscido et al. 2004), others. Although intuitive, mathematics and computer simulation become necessary to predict what group-level behaviours emerge from these simple interactions when group size is in excess of a few individuals. It is the recursive nature of these interactions that makes them particularly hard to investigate without the use of these tools; the behaviour of a particular individual influences that of others (e.g. near neighbours), which in turn influence the focal individual, and so on. Possibly due to the intriguing formations produced by some species, fish shoals have often provided the inspiration for these ‘equations of motion’ (Parr 1927; Breder Jr 1954; Romey 1996; Couzin et al. 2002), which makes these models particularly applicable to, and testable using, fish. To gain empirical support for any model, predictions can be compared to experimental or observational data (Hoare et al. 2004; Viscido et al. 2004; Hensor et al. 2005). Alternatively, a complementary approach is to test the assumptions underlying the model, again through either experimentation or observation (Gr¨unbaum et al. 2004; Tien et al. 2004). Testing predictions of collective behaviour models against data has the advantage that the positions and directions of often large numbers of individuals can be determined and compared to the model predictions. To date, this has been done most successfully with locusts, Schistocerca gregaria (Buhl et al. 2006), although see Huth & Wissel (1994) for an example using fish. However, fish shoals have an additional problem that they are almost always threedimensional, which makes automated tracking particularly difficult (Gr¨unbaum et al. 2004, for example, managed to track shoals of eight giant danios, Danio aequipinnatus, in three dimensions). A weakness of this prediction-testing approach is that there is always a possibility of convergence between model and empirical data, where different mechanisms generate the same pattern (Parrish & Edelstein-Keshet 1999; Gr¨unbaum et al. 2004). Hence, explicit testing of the assumptions made must be done where possible to truly validate a model.

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Social Organisation and Information Transfer in Schooling Fish

219

A number of experiments have been carried out which attempt to identify the rules fish use when interacting, often without any specific reference to models of collective behaviour. Three-spined sticklebacks (Gasterosteus aculeatus) have been shown to consider both distance from a conspecific and their turning speed to minimise approach time (Krause & Tegeder 1994). Tegeder & Krause (1995) then showed that these fishes determined their direction based on the number of potential neighbours, similar to the averaging of neighbour headings to determine direction seen in models of collective behaviour (e.g. Couzin et al. 2002). The fishes also traded off the number of potential neighbours with the distance to them, presumably to minimise risk. Averaging of velocities between individuals within shoals also seems to occur, as the responding velocity of focal fish to the acceleration of a stimulus fish (jack mackerel, Trachurus symmetricus) was lower in shoals of 6 compared with a shoal of 2 (Hunter 1969). Interestingly, the opposing forces of social interaction may be driven by different sensory modalities, at least in some fishes, with the zone of repulsion being maintained by the lateral line system while vision mediates attraction (see also Partridge & Pitcher 1980; Burgess & Shaw 1981). Unfortunately, due to limitations on technology in collecting the data, and statistical and analytic tools in analysing the data, the results of these often ingenious studies leave their findings less than conclusive. Sample sizes also tend to be low, in terms of both the number of interacting agents (e.g. Hunter 1969) and the number of replicates (e.g. Tien et al. 2004). Partridge (1980) presents data suggesting that shoals of two fishes show a fundamentally different structure than larger shoals and that shoal formations are highly variable in structure and dynamics, both of which are problematic for determining experimentally the movement rules fishes use to form and maintain shoals. A further issue is that the shoals of different species show different structures (Partridge et al. 1980), which can make generalising results difficult, although it does make comparative analyses of collective behaviour possible.

10.3

Emergent collective motion in the absence of external stimuli

The similarity between patterns generated from models of collective motion and those found in naturally occurring fish shoals suggests that simple movement rules, like those in Section 10.2, capture the essential properties exhibited by shoaling fishes. The models reveal that variation between individuals, such as that generated from differences in body size (Reebs 2001) or internal state (Krause 1993a), is not necessary for collective motion (see Couzin et al. 2002 for the predicted effects of within-shoal variation; Romey 1996). Even more importantly, only local interactions are required without the need for global information or control (Couzin et al. 2002). This can then explain how fishes and other animals can form and maintain groups with dimensions far in excess of the sensory range of any individual (e.g. Makris et al. 2009). The model of Couzin et al. (2002) predicts four types of collective behaviour found in natural fish shoals as the orientation (i.e. alignment) and attraction zones are varied in size. In the swarm state, the group is cohesive but unpolarised when individuals only show attraction (Fig. 10.1a). When alignment is low relative to attraction, a torus formation is generated, with individuals rotating around an empty core (Fig. 10.1b). Increasing alignment

BLBK374-10

BLBK374-Brown

220

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

further polarises individuals and the group shows directional movement, but individuals are still able to change their relative position (the dynamic parallel state; Fig. 10.1c). Finally, in an arrangement visually similar to Fig. 10.1c, relative positions become fixed in the highly parallel state as polarisation increases further. The important thing to note is that the patterns can result from individuals following relatively simple generative behavioural rules that do not explicitly describe this pattern nor rely on an external attractor; for example, there is nothing explicit in the rules that generate the torus that suggests ‘move in a circle’ (although of course in other cases attractors may facilitate a similar behaviour). The possible states are surprisingly well defined, i.e. intermediate states are not observed, as the transitions between states occur rapidly across small changes in parameter space (see Figs 3E and F in Couzin et al. 2002). All the predicted states have been observed: barracuda, jack and tuna demonstrate the torus formation (as in Fig. 1 in Parrish & Edelstein-Keshet 1999), while Partridge (1980) demonstrated that minnows (Phoxinus phoxinus) switch quickly between a disordered, swarm-like state and a state showing high polarisation.

10.4 Response to internal state and external stimuli: Information processing within schools 10.4.1

Collective response to predators

A major factor believed to drive the evolution of group living is predation (reviewed by Krause & Ruxton 2002). Changes in the size of the interaction zones used in the model of Couzin et al. (2002) and elsewhere provide a framework to explain responses to predation, both over evolutionary time where the size of the zones becomes heritable (Seghers 1974), and as a phenotypically plastic short-term response to immediate threat (Krause 1993c; Hoare et al. 2004). Larger zones of attraction lead to larger groups, benefiting through attack abatement as the visibility of groups scales less than linearly with group size, unlike risk dilution (Turner & Pitcher 1986). Predatory confusion (the reduced ability to successfully catch prey due to an increased number of similar targets) has also been shown to be beneficial to fish shoals (Neill et al. 1974), and although coordination between individuals has shown not to be necessary for the effect to operate (at least in humans; Ruxton et al. 2007), it is still likely that highly coordinated motion increases the benefit further. In addition to shoal size and cohesion, the shape transitions discussed in Section 10.2 could be important ways by which shoals assume a collective strategy appropriate to their internal state, such as when herring adopt a loose swarm-like formation when feeding, but a highly polarised group if predators are detected (Radakov 1973). The highly integrated social interactions among fishes in such schools means that, in some sense, they can be considered a ‘social medium’ which can rapidly propagate local information (such as the location and direction of a predator attack) simply through the tendency of individuals to group and to align with near neighbours. Escape behaviour can thereby be transmitted through the entire shoal. A number of empirical studies have demonstrated a wide range of shoal-level formations as a reaction to predator attack (juvenille herring (Clupea harengus): Axelsen et al. 2001; minnow: Magurran & Pitcher 1987; adult herring: Nøttestad & Axelsen

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

221

(a)

(b)

(c)

Fig. 10.1 Collective motion in fish: (a) ‘Swarm-like’, loose and unpolarised aggregation; (b) ‘torus’ formation; (c) ‘dynamic polarised group’, whereby individuals become polarised, but can dynamically change position. (For details of the model, see Couzin et al. (2002).)

1999), and it has been shown that simple individual-based rules can explain the variety of group behaviours observed (Vabø & Nøttestad 1997). The benefit of group vigilance as a strategy to decrease predation risk in groups (the ‘many-eyes’ effect) is reliant on this social transfer of information. Godin et al. (1988) demonstrated that increasing shoal size of glowlight tetras (Hemigrammus erythrozonus) increased the probability of responding to a fright stimulus, as expected. Moreover, as

BLBK374-10

BLBK374-Brown

222

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

the proportion of fishes that responded to the stimulus increased with shoal size, the study explicitly showed that the information gained by the few was transmitted to other individuals. The responding individuals do not necessarily have to be able to detect the primary cue themselves (Magurran & Higham 1988). For example, the fright response of chub (Leuciscus cephalus) to Schreckstoff is transmitted to three-spined sticklebacks, which are unable to detect the cue, and helps explain the co-occurrence of the latter species within heterospecific shoals (Krause 1993b). That sticklebacks responded with behaviour associated with predator detection, such as fast-starts, suggests that a predator was perceived, rather than the fish being influenced solely by the motion of their neighbours (as would occur in the current models of collective motion). In some cases the transmission speed of information about a predator can be faster than the approach speed of that predator. This ‘Trafalgar effect’ (Treherne & Foster 1981) means that not only is detection of a predator more likely within groups and can occur at a greater distance, but individuals further from the predator detect it even sooner than they would have done if solitary and at the same distance from the predator. Convincing evidence for this effect comes from shoals of banded killifish (Fundulus diaphanus) being approached by a model predator moving at a realistic speed under field conditions (Godin & Morgan 1985). The fact that transmission speed of social information can greatly exceed swimming speed suggests that the Trafalgar effect may be a common feature of fish shoals. Radakov (1973), for example, found that the speed of propagation of turning waves reached 11.8–15.1 m/s in Atherinomorus compared to the maximum burst swimming speed of this species of silverside of around 1 m/s. In an impressive study of natural fish shoals containing tens of millions of fishes, Makris et al. (2006) observed density waves again travelling at speeds of an order of magnitude greater than the typical swimming speed of fishes. This finding was supported in a second study where the growth of oceanic fish shoals again occurred ten times faster than swimming speed, and in this study Atlantic herring could be identified as making up the majority of the shoal (Makris et al. 2009). That the movement speed of the shoals was consistent with the typical swimming speed of schooling herring (approximately 0.2 m/s) helps to validate the remote sensing technique used in these studies to elucidate these emergent behaviours at massive scales.

10.4.2

Mechanisms and feedback in information transfer

Current models have investigated a variety of ways in which information is transferred within animal groups, mostly focusing on focal individuals biasing their own motion with the average of many, or a subset of, detectable neighbours (Couzin et al. 2005). However, the way in which information is propagated in groups such as fish schools is poorly understood, with little experimental data to support theoretical work. Response latencies between neighbouring saithe (Pollachius virens) have been observed to be in the order of 0.1 second (Partridge & Pitcher 1980; Partridge 1981). Latencies have also been shown to be longer in shoals cruising at greater speeds (Partridge 1981), which suggests reduced activity may be beneficial to facilitate information transfer as well as making the shoal less conspicuous to predators (Krause & Godin 1995). In one of the few specific tests of the mechanisms underlying information transfer between fishes, Hunter (1969) used jack

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

223

mackerel (T. symmetricus) to demonstrate that the speed of response between a stimulus fish and a focal individual was greatest when the fishes were alongside one another, or when the stimulus fish was in the binocular region directly in front of the focal. Additional factors influencing the response latency of the focal fish included the velocity, relative angle and distance from the stimulus fish (Hunter 1969), which suggests that the movement perceived by the focal fish (i.e. the change of image on the retina) determines the latency in information transfer. In addition to vision, both olfaction and the lateral line have been shown to be important to form and maintain shoals, and to transfer information between individuals. Blinded fishes are still able to shoal, although responses to neighbours were delayed and occurred only if the neighbour was within one body length (Pitcher et al. 1976). Individual saithe which had their lateral lines disabled compensated by increasing the frequency in which neighbours were directly alongside them (Partridge & Pitcher 1980), probably to maximise information flow on the retina (as suggested by Hunter 1969). However, this change in structure may have reduced the ability of information to transfer across the school, as the direction of a fright stimulus relative to body length only had an effect on response latency when the lateral line was disabled (Partridge & Pitcher 1980). Olfactory cues are used to locate other individuals and maintain contact over larger spatial scales (Hemmings 1966), and are unlikely to be of much importance to information transfer within shoals over short time scales. However, olfaction does appear to play a strong role in shoal choice (Ward et al. 2002a), and if information transfer is affected by shoal composition, olfaction will have an indirect effect on short-term information transfer. As a side note, the findings of Parr (1927), Pitcher et al. (1976), Partridge & Pitcher (1980) and Burgess & Shaw (1981), amongst others that are not referenced here, should be treated with caution. The disabling of any sensory modality of importance in a healthy fish is likely to change behaviour in ways unintended by the experimenter. In addition to the ethical concerns these methods raise, we suggest that the use of genetic manipulation may be a more fruitful approach, and has recently been used successfully in behavioural studies in fishes (e.g. Snekser et al. 2006). Fewer confounding effects of the treatment are likely to occur, in part because the fishes are born and raised under the experimental condition. Furthermore, the genetic manipulations can be seen as experimental mutations that are under selective pressure, such as the ability to respond to information gained by other members of a shoal. Screening of the manipulation to examine potential confounding effects remains necessary, of course. Responding to the motion of near neighbours is clearly important to the functionality of schools and the perceptual range over which they interact has important effects on collective behaviour. If this range is relatively small, an individual will behave more or less independent of those around it and changes in direction spread poorly or cannot spread (nor do groups remain cohesive). However, as this range increases, individuals respond to a greater number of neighbours, cohesive groups can form, and such groups become capable of transferring directional information, facilitating collective response to external environmental stimuli. However, if this range becomes too large, individual response includes distant individuals whose direction and position may not encode relevant information. Thus, the focal individual may not respond in an optimal manner (see also

BLBK374-10

BLBK374-Brown

224

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Inada & Kawachi 2002). Consequently, in the case where the range of perception is not restricted by shadowing effects (as it is in some pelagic schools; Partridge 1982), individuals may deliberately restrict the range over which they respond to others. Even where global information is available it may benefit individuals to respond only to local information. Observational data does suggest that individuals pay greater attention to those closer by and it has been suggested that this can dampen small perturbations of movement in the shoal but amplify any perturbations above a threshold strength (Partridge 1981). In shoals of banded killifish, although flight reaction distance from an approaching predator was constant across shoal sizes, the variance in reaction distance decreased with increasing shoal size (Godin & Morgan 1985). This suggests that rather than being separate agents each with a given probability of detecting a threat, which would increase the reaction distance but not its variability, interactions between individuals increase the reliability of responding correctly to a potential threat. There may be important collective feedback systems whereby individuals change their probability of responding to neighbours depending on context. Even without leaders, i.e. those with additional knowledge, some individuals may have a greater probability to be located in parts of the school with more influence on group movement, or to swim faster or with more turns (Partridge 1981). These may be driven by differences in motivational state (e.g. hunger; Krause 1993a), body size (Atz 1953) or consistent personality differences (Harcourt et al. 2009). Individual variation itself is not necessary to cause feedback. Such ‘informational cascades’, to use the term from economics (Bikhchandani et al. 1992; Giraldeau et al. 2002), can result in an autocatalytic response with a defined response threshold. This type of response has been observed in a range of collective systems (Ame et al. 2004; Jeanson et al. 2005; Sumpter 2006) and can allow for a rapid adaptive response to stimuli. Ward et al. (2008) provide strong evidence that informational cascades are important for collective decision-making in fishes. Using three-spined sticklebacks, they demonstrated that the decision taken by an individual is dependent on the number of other fishes which have already made a decision, but in a non-linear manner (a quorum response, see Section 10.5), which is typical of non-linear positive feedback (Sumpter 2006). Moreover, the time frame over which a previous decision could influence a focal individual decreased with the presence of a predator, suggesting that the fishes adjusted their behaviour in a context-specific manner. In the case where fishes are in a state of heightened response (such as if alarm substance is detected, or following an attack), chance turning events can frequently lead to spontaneous turning by the group, even in the absence of further attacks (Radakov 1973). This heightened response presumably allows the group to be extremely sensitive to weak or ambiguous external stimuli (essentially increasing the ‘system gain’) but also becomes more susceptible to noise (stochastic effects) and false alarms. At the opposite extreme, such as when fishes adopt a poorly polarised group structure (Fig. 10.1a), signals (changes in the direction of a subset of individuals) are more readily damped; the signal to noise ratio becomes relatively high, but relatively low gain means the group takes longer, or may even not, respond to true signals. The way in which individuals adopt behaviours adaptively to tune the sensitivity of collective response, adjusting gain and thus the signal to noise ratio, would make a very interesting study and there may well be an important learned component to such responses.

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Social Organisation and Information Transfer in Schooling Fish

10.4.3

225

Information transfer during group foraging and migration

Alignment and cohesion is of importance not only in the evasion of predators, but can also facilitate information transmission in other contexts such as during foraging or migration. Migration has attracted a lot of attention due to the commercial importance of some migratory fish species (Quinn & Dittman 1990), and although fish migration can be density dependent (Makris et al. 2009), little is known about how interactions at the individual level facilitate long-range movement. Gr¨unbaum (1998) demonstrated theoretically that schooling can facilitate more accurate taxis in noisy environments (such as when responding to thermal or nutritional gradients). Schools act as an integrated array of sensors allowing individuals to increase the ‘effective’ range of their interaction with the environment, and also to average errors in individual motion (error being a fundamental property of sensory systems). Thus, schools can dampen small-scale fluctuations to allow perception of gradients not possible for individuals in isolation. For a conceptually similar model for large fish schools, see Niwa (1998b). Grouping in this way could allow individuals to find more effectively, and subsequently remember, areas of appropriate habitat, such as food or spawning grounds. This property of schools has also recently inspired new search algorithms in computer science, termed ‘particle swarm optimization’, allowing for efficient and effective search of complex multidimensional function space (Kennedy et al. 2001) and further investigations of the collective computational ability of groups in this context may provide new insights into schooling ecology. Radakov (1973) performed experiments on juvenile pollock (Pollachius pollachius) to investigate social transfer of information. A total of 25–35 specimens were kept in an aquarium and he investigated how information about the location of a discrete food source (a patch of gammarids), only visible to a small proportion of the fishes, spread throughout the group. In some trials he used an opaque partition (which extended from the roof of the aquarium to near the base, leaving a 20-cm gap in 100 cm of water) under which fish could swim to move between sections. By adding food to only one section he could control which fish could directly see the introduced food. On average six fishes directly detected food, whereas the rest did not. However, 15 seconds after its introduction, on average, 14 fishes had moved into the section with food. Tracking the motion of fish from film he could quantify the response of ‘observers’ to those ‘informed individuals’ they could see under the partition. Forage-area copying is a common behaviour in many fish species and the head-down posture (or any other movement characteristic of feeding) is frequently used as an indicator that a conspecific has found food (Pitcher & Parrish 1993).

10.5 Informational status, leadership and collective decision-making in fish schools The environment through which fish schools move is often complex and changing; resources and predators are distributed probabilistically in space and time. Since individuals also move between groups (a topic we will return to later), aggregates of social animals often consist of individuals that can differ with respect to their informational status (such as when only some individuals have pertinent information about a currently viable resource) or internal

BLBK374-10

BLBK374-Brown

226

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

state (such as degree of hunger). Fishes are capable of associating specific locations with rewards (Laland & Williams 1997) and can be trained to move to a known food source at specific times of day (Reebs 2000). By schooling with such ‘informed’ individuals, those that are na¨ıve can more readily find food and learn about their environment. For example, transplanted French grunts can learn daily migration patterns from resident fishes within a couple of days (Helfman & Schultz 1984). Thus, by learning within a social context, individuals can become informed themselves, and in turn bias the motion of other na¨ıve individuals, and allowing knowledge to spread within a population (Reader & Laland 2000). What makes this form of information transfer so effective is that only a minority of individuals need to be informed: Reebs (2000), Swaney et al. (2001) and Ward et al. (2008) all found that relatively few individuals are capable of guiding a group of na¨ıve individuals. Couzin et al. (2005) developed a generic model of this type of guidance in animal groups. Following classic (Miller 1944; Brown 1948; Tinbergen 1952) and more recent (McClure et al. 2004; Usher & McClelland 2004; Livnat & Pippenger 2006) studies, they proposed that individuals exhibit internal conflict whereby they may have to reconcile different behavioural tendencies, such as schooling, with a desire to move in a specific (or general) direction. Na¨ıve individuals, since they have no particular directional preference, predominantly school whereas informed individuals balance schooling with a desire to move in a preferred direction (such as towards the location of a food source, or along a section of a migration route). The degree to which this balance is maintained is controlled through a simple weighting, termed ω. If ω = 0, then individuals are completely na¨ıve (and therefore have no desire to move in any specific direction). If individuals develop a preference to move in a particular direction (through learning), ω becomes greater than 0 and individuals increasingly bias their tendency to school with others with their competing tendency to direct their motion in a relatively specific desired direction. Therefore, a single term can represent different strengths of directional preference within a group, and thereby represent na¨ıve individuals (ω = 0) and those that are informed to various degrees (ω > 0). Using this framework, Couzin et al. (2005) demonstrated that informed individuals can guide those that are na¨ıve in an effective manner without requiring either signalling or the ability for individuals to establish who has, and who has not got, information (Thomas D. Seeley of Cornell University has termed this the ‘subtle guide hypothesis’). In agreement with empirical work, a small, group-size dependent proportion of informed individuals can guide groups accurately. Furthermore, as group size increased, the required proportion of informed individuals needed to achieve a given accuracy decreases. The result is that a more or less constant number of individuals is all that is required to lead the group (a quorum), irrespective of group size (see Fig. 1A in Couzin et al. 2005). Shoals of threespined sticklebacks can be led by replica conspecific ‘leaders’, which allowed Ward et al. (2008) to demonstrate experimentally that a minimum of two leaders is necessary to guide shoals of 4 or 8 fishes, with little difference between these two shoal sizes. However, a single leader could lead a solitary fish or a shoal of 2, which suggests the quorum rule does not apply well to very small numbers of individuals. In this case, the game theoretic approach used by Harcourt et al. (2009), again with experimental support from three-spined sticklebacks, may be more suitable. In the subtle guide model (Couzin et al. 2005), when only a few individuals within a group were informed, the accuracy of guidance increased with increasing ω, but this came

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

227

at a cost due to a trade-off between increasing accuracy of group motion and the increasing probability that the group splits (informed individuals leaving those that are na¨ıve behind). Experimental support of this effect is provided by Swaney et al. (2001) with shoals of guppies, demonstrating that well-trained individuals tend to lose the group, whereas poorly trained individuals remain with the group but might be less capable of locating the target. Where informed individuals differ in preference, Couzin et al. (2005) demonstrated that even though informed individuals do not know if they are in the majority or minority (they cannot assess the informational status of others) groups can still make consensus decisions, and select collectively the direction associated with the majority. The generic, and simple, nature of this model means that this collective computational ability could be a fundamental property of cohesive fish schools. Furthermore, groups were also shown to be able to discriminate on the basis of the quality of information (without requiring any modification of the model) selecting collectively the direction associated with least error, even though the individuals themselves were not aware of how their information compared with that of others, or even if there were any other informed individuals. Consequently, the model predicts that individuals can respond spontaneously to those with information and that informational status can change freely as individuals acquire information from their environment, either socially or asocially, or as their needs change. Although this is currently a deliberately simplified version of reality, including a limited number of preferred directions, the framework is highly flexible and can easily be extended to consider more complex cases, or the ability for individuals to dynamically modify both multiple directional preferences and the ability to learn about the environment. In the initial study, Couzin et al. (2005) deliberately investigated the case where groups would typically remain cohesive (to simplify the analysis). However, this is only one area of parameter space in this model. When ω is high, for example, informed individuals are less willing to compromise and will leave the group to move in their desired direction, often taking a subset of na¨ıve individuals with them. Under such conditions groups frequently split if there is informational heterogeneity. Similarly, when the strength of attraction is reduced, groups will readily split and fuse. If combined with memory of new resource locations, this can allow information to spread through populations.

10.6

The structure of fish schools and populations

Informational status can lead to assortment within animal groups as shoals of fish tend to be led from the front (Bumann & Krause 1993; Krause et al. 2000b; Reebs 2000). Emergence of inter-individual differences in the tendency to lead (Leblond & Reebs 2006; Harcourt et al. 2009) produces non-random assorting even in groups with stable membership where variation in information would be thought to be minimal. In addition to informational status, a number of other passive mechanisms can lead to assortment: individual differences in speed, interaction range and turning rate can spontaneously assort individuals both within (Couzin et al. 2002) and between (Krause et al. 2005) groups. For example, smaller golden shiners (Notemigonus crysoleucas) took front positions in shoals during training, even though the large individuals were then capable of leading na¨ıve individuals during testing, which they did from the front (Reebs 2001). Assortment due to these effects may only

BLBK374-10

BLBK374-Brown

228

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

be apparent when informational status is important, i.e. when groups are on the move. It would be interesting to see whether the size assortment apparent during leadership is present when the shoals were at rest, which could be used to test between passive and active mechanisms of phenotypic assortment. Active assortment itself may be driven by individuals assessing others and choosing to bias the strength of their associations (e.g. by modifying their weighting of attraction towards some phenotypes, or actively repelling themselves from others). By modifying the size of the zones of attraction and repulsion relative to others, Couzin et al. (2002) showed that an individual can influence their position within a group without knowledge of their relative position. This is likely to be of importance in larger and/or densely packed shoals, especially those found in pelagic systems which may consist of tens of thousands or millions of individuals and extend over kilometres (DeBlois & Rose 1996). As the costs and benefits of shoaling vary with spatial position within a group, differences between individuals in age (DeBlois & Rose 1996), nutritional status (DeBlois & Rose 1996) or parasite load (Ward et al. 2002b) may all influence the optimal position adopted by an individual within a group (Krause 1994), leading to non-random assortment. Krause (1993a), for example, demonstrated that starved roach (Rutilus rutilus) tend to occupy frontal positions where they are presumably more likely to encounter food. However, such positions may be relatively dangerous with such individuals more likely to encounter sitand-wait predators (Bumann et al. 1997). A further property of groups that can be controlled by modification of simple local rules is group size. As shown with comparison to experimental data, individuals can accurately change their probability of being in a group of a certain size at a certain time. Hoare et al. (2004), for example, showed that group size distributions in banded killifish changed in response to environmental stimuli, increasing group size when an alarm odour was detected, and decreasing group size versus controls in response to food odour. Adapting the range over which individuals responded with one another in an individual-based model produced good agreement with these experimental results, suggesting information of the positions and orientations of all shoal members is not necessary to produce an adaptive response to ecological changes. At a population-level, group-size distributions in fish populations tend to be stable (Godin & Morgan 1985; Makris et al. 2006). Individual interactions can result in a fission–fusion system (Bonabeau & Dagorn 1995; Gueron & Levin 1995; Niwa 1998a; Bonabeau et al. 1999), where rates of amalgamation (fusion) and splitting (fission) define the group-size distribution in the population. Where fusion rates are high relative to fission, the number of groups with relatively few individuals tends to decrease (larger groups can persist for longer) and the group-size distribution assumes a long tail. This is characteristic of some pelagic fish schools (skipjack tuna, Katsuwonus pelamis: Bayliff 1988; yellowfin tuna, Thunnus albacares: Klimley & Holloway 1999), where the half-life of schools can be in the order of weeks. In such populations the timescale over which processes such as information transfer occurs may be completely different to freshwater fishes where fusion rates tend to be low relative to fission and groups tend to be much more unstable, resulting in group-size distributions which tend to be more rapidly decreasing (e.g. Godin & Morgan 1985). Krause et al. (2000a), for example, found that in wild populations of killifish the rate of exchange of individuals between schools was high and that schools encountered one

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

229

another very frequently, on average every 1.1 minute. In wild guppy populations, Croft et al. (2003) found that encounters between shoals occurred approximately every 14 seconds. See also Seghers (1981) and Helfman (1984) for further evidence of such behaviour. When considering information transfer at a population level, one has to consider the timescales of biological relevance. For pelagic groups, which range widely, food patches may be separated by distances in the order of kilometres. In freshwater fishes the analogous spatial properties may be an order of magnitude less. Consequently, one needs to scale assumptions about information transfer and individual memory accordingly. In many cases the distribution of group-size can be recorded in a population (see Couzin & Krause 2003 for a review), but this provides little information about how information transmission could occur within a wild population as it is likely some phenotypes are more likely to change groups than others (Ward et al. 2002b; Croft et al. 2003). However, there are methods for assessing the social structure of animal populations that can be used to make predictions for information transmission within and between populations. It has been well documented that social interactions in animals rarely occur randomly (Whitehead & Dufault 1999; Croft et al. 2008). As discussed in the preceding text, particular fish shoals are frequently assorted by a number of phenotypic characteristics including body size, sex, species and parasite load (Krause et al. 2000a). Such non-random interactions may have implications for the transmission of social information, with information being more likely to be transmitted between individuals of a similar phenotype (Hoare & Krause 2003). To truly understand the transmission of information in a population, we must consider ‘who learns from whom’. One starting point is to consider ‘who associates with whom’, as social learning may be more likely between individuals that have strong social ties (Swaney et al. 2001; Franz & Nunn 2009; Hoppitt et al. 2010). However, other factors may also have a role to play such as dominance, size, sex, etc. (Laland 2004). Recent progress has been made in investigating the fine-scale structure of animal populations by representing social interactions between individuals as a social network, a method that allows us to study the social structure of a population as a whole.

10.7

Social networks and individual identities

A social network for fishes may be constructed from information about who shoals with whom, with ties in the network representing individuals that have co-occurred in the same shoal. To collect the information required to construct social networks, it is necessary to know the identities of each fish (in a group or population). The last few years have seen significant progress with various marking techniques (e.g. fluorescent elastomer) that allow such individual marking even of relatively small fishes of just 15–20 mm body length (e.g. Ward et al. 2002c; Croft et al. 2004). Such a technique can be used to mark fishes in both the laboratory and field (Croft et al. 2004). Once all the individuals in a whole population are marked in this way, we can observe the social contacts (individuals within the same shoal are usually defined as those within four body lengths of one another) between individuals on a nearly continuous basis by using a digital tracking system (Balch et al. 2001) or at given time intervals (point-sampling, see Martin & Bateson 1993). The latter is logistically simpler and usually the only option when collecting data in the field.

BLBK374-10

BLBK374-Brown

230

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

For example, in an investigation of guppies (Croft et al. 2004), all adult individuals were marked within a wild population (n = 199) (see also Ward et al. 2002c for an earlier investigation using three-spine sticklebacks). The population was then sampled by capturing entire shoals once per day over a period of 15 days, recording the composition of the shoals. Based on this information, a social network of fish association patterns was constructed

(a)

(b) Fig. 10.2 (a) A social network of a guppy population in Trinidad. All guppies from two interconnected pools were marked and released. Over the next 2 weeks, approximately 20 shoals were captured daily and fishes that belonged to the same shoal were connected in the network. Over time a completely connected network developed that comprises 197 fishes. Each circle represents an individual male fish and each square an individual female. The size of the symbol is indicative of the body length of the fish. Individuals interconnected by lines were found at least twice together. (b) Five distinct communities (indicated by grayscale) were identified in the guppy network.

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

231

(Fig. 10.2a). Social networks provide us with information on who had social contact with whom and how often, and can form the basis for predictions regarding information flow (Franz & Nunn 2009; Hoppitt et al. 2010) or disease transmission within a population or even between populations. A useful first step is usually to generate a graph of the social network (using one of the widely available software programs such as UCINET). To this graph we can add attribute data of individuals such as body size, sex and information status (in case of learning experiments). Often the inspection of the data will generate many useful ideas (Krause et al. 2007; Croft et al. 2008). In a second step we can characterise the structure of the social network, using simple descriptive statistics such as mean path length (L), mean clustering coefficient (C) and mean degree of connectedness (k) (Albert & Barab´asi 2002; Croft et al. 2003; Newman 2003). L is calculated as the mean number of connections in the shortest path between two individuals in the network (see Newman 2003) and describes a global property of the network which gives a simple indicator of how quickly social information might spread in an animal population (information can be expected to spread more quickly for lower values of L). C is a measure of the mean cliquishness of the network, calculated for each individual as the fraction of connections that exist between animals in its immediate network neighbourhood (Newman 2003). C describes an average local property of the network, and is of particular interest because it measures the extent to which two of one’s network neighbours are themselves neighbours. In social animals this local structure may be due to active associations (e.g. phenotypic assortment or associations between familiar individuals). The degree of connectedness of an individual (from which k is derived) is simply the number of direct social connections (edges) that an individual has (Fig. 10.2). The mean degree for the network (k) can then be calculated as the average of the individual degrees. L, C and k are just some examples of commonly used descriptive statistics that can be calculated to characterise the social network structure. Over the last 10 years or so a large number of network descriptors and statistical techniques have been developed that cannot be discussed in detail here. Interested readers are referred to Croft et al. (2008) and James et al. (2009). Standard network theory (as used in the preceding text for calculating L, C and k) is primarily based on unweighted associations between individuals and thus does not allow the analysis of one of the biological features of most interest, which is the occurrence of stable pairs of fishes in wild populations. To measure the persistence of pair-wise associations, Croft et al. (2004) used an ‘association strength’ (AS), which indicates the number of days that a pair of fish was caught together in a shoal and compared this value to a randomisation test that preserved the shoal-size distribution and the number of recaptures for individual fish. In guppies the social networks were all found to have a non-random structure and exhibited ‘social cliquishness’ (Croft et al. 2005). A number of factors were observed to contribute to this structuring. Firstly, social network structure was influenced by body length and shoaling tendency, with individuals interacting more frequently with conspecifics of similar body length and shoaling tendency. Secondly, individuals with many social contacts were found to interact with each other more often than with other conspecifics, a phenomenon known as a ‘positive degree correlation’. Finally, repeated interactions between pairs of individuals occurred within the networks more often than expected by random interactions and were a good predictor of cooperative relationships between fishes

BLBK374-10

BLBK374-Brown

232

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(Croft et al. 2006). It remains to be tested whether such strong associations between pairs of fishes are also good predictors of information flow between individual fish within the network (see Franz & Nunn 2009 for model predictions on association-based information flow). This illustrates the point that we can generate more than one type of network for a population depending on whether we measure association, cooperative behaviours or other types of interactions with statistical tools now available that allow to test for correlations between networks (Croft et al. 2008). A number of studies have looked at individual identities of fishes and their social interactions in the field (guppy: Croft et al. 2004; yellow perch: Helfman 1984; yellowfin tuna: Klimley & Holloway 1999; three-spined stickleback: Ward et al. 2002b). Apart from Helfman’s study on the yellow perch, they all found good support for a significant co-occurrence of particular individuals with one another suggesting active shoalmate choice. However, the only paper that directly tested this idea is Croft et al.’s (2006) study on the guppy providing support for the hypothesis that this cooccurrence of pairs of fishes is indeed based on active social preference rather than similar swimming speeds (Krause et al. 2005) or microhabitat preferences. Additional research revealed that cooperative individuals (cooperation was tested in a predator-inspection context) have a tendency to associate with other cooperators whereas defectors had only weak connections with cooperators (Croft et al. 2009). The fact that behavioural phenotype can affect social network structures was also demonstrated by Pike et al. (2008), who showed that in three-spine sticklebacks bold individuals had fewer social connections than shy ones but distributed their connections more evenly among conspecifics. These studies indicate that fishes do not only assort by morphological traits but also by behavioural strategies, which has important implications for the evolution of behavioural strategies. Game-theoretic models conventionally make the assumption that there is random mixing among individuals (and therefore behavioural strategies) in a population (Maynard-Smith 1982). However, the above findings regarding fish populations (see Section 10.6) highlight the importance of the social fine structure of animal populations for the evolution of behavioural strategies – a point that is increasingly taken up in a new generation of game-theoretic models that are played on networks (see Krause et al. 2010 for an overview). Recent work by Ward et al. (2009) suggests that both habitat-based familiarity (Ward et al. 2004) and individual recognition take place in the guppy but found no evidence for the latter in sticklebacks. It has been suggested that fishes learn preferentially from familiar conspecifics (Brown et al. 2003; Chapters 8 and 10), which means that strong associations might be indicative of likely pathways of information transmission. Regardless of the underlying recognition mechanism, strong preferential association patterns provide an interesting test bed for predictions on information flow within populations (see Franz & Nunn 2009; Hoppitt et al. 2010). In a similar way we might expect that other factors that lead to phenotypic assortment might play a role for information transmission within populations and network analysis can be useful in identifying pathways for information transmission.

10.8

Community structure in social networks

When considering the social structure of group-living animals one generally thinks of individuals forming groups and various numbers of groups making up a population. However,

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

233

recent developments in network theory (see e.g. Lusseau & Newman 2004; Newman 2004; Newman & Girvan 2004; Croft et al. 2008) have enabled us to explore the possibility that there are meaningful communities, at a level between the group and the population, which may play an important role in determining the social organisation, and thus the routes for information transmission, in animal populations. James & Mawdsley (unpublished) investigated the community structure of a population of wild guppies (Fig. 10.2b). There is evidence that the communities are structured by body length differences and to a lesser extent by sex. Assortment of fish communities by phenotypic characters (such as body size, colour, pattern, parasite load, etc.) is well established (Radakov 1973; Krause et al. 2000a; Couzin & Krause 2003). However, so far studies on assortment have principally focused on group composition (although see Radakov 1973; Couzin & Krause 2003). In contrast, what the community analysis suggests is that individuals can be freely exchanged between groups provided that they belong to the same community. Furthermore, we would predict that information transfer is more rapid within communities than between them. This is a yet untested prediction and should provide an interesting field for future experimental work. Another important consequence of this type of analysis is that we can identify individuals and their position in the network. For instance, individuals that interconnect communities should play a key role in information transmission in the population as a whole. In the guppy network there is an indication that fishes of intermediate body lengths that cannot be clearly assigned to any particular community take up this position. Whether these individuals are indeed pivotal to information transmission remains to be ascertained. All of the social network analysis we have presented has assumed the social network to be a static entity derived from the accumulation of data over a period of time. The reality, especially in fluid social systems, is much more dynamic and interesting. Though the structure of the static network gives an indication of the information-carrying capacity of a population, there are certain features that depend strongly on the order in which pair-wise connections are made and broken (Moody 2002).

10.9

Conclusions and future directions

Much of the functional complexity of fish schools arises from repeated interactions among individuals. Since they can be readily observed and manipulated, fishes present an excellent opportunity to quantify the behaviour of individuals and to develop new mathematical approaches that link the behaviour of individuals to the higher order properties at the group and population levels that result from these interactions. This approach is still in a relatively early stage of development, but as we have discussed in the chapter, is likely to provide new insights into how information is acquired, processed and stored. Therefore, combining this approach with current ideas in the field of social learning is likely to be an important direction for future research. What distinguishes information transfer within shoals from social learning, however, is that the receiver does not need to remember any information. In fact, there is no need for the receiver to even be aware of what information it receives for collective motion to be adaptive. It is likely that movement of a fish in response to a predator or food have characteristic patterns that neighbours can detect, although it is also possible that this extra cognitive

BLBK374-10

BLBK374-Brown

234

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

task might slow the response enough to be selected against. Therefore, the behaviour of shoaling may represent an alternative evolutionary strategy for solving complex problems compared to developing a more advanced, individual, cognitive ability. For example, in the study of Krause (1993b), the sticklebacks circumvent having to evolve the ability to detect Schreckstoff because they shoal with a species that does. In a sense, the relevant cognitive unit may be the group rather than the individual, whilst selection remains at the individual level as members of fish shoals tend not to be stable kin groups. Linking experimental studies of fishes to those in natural populations remains an important goal. Here a key challenge will be to scale from individual motion and interactions to the temporal scale of practical observation in the field. One way this may be possible is to create intermediate meso-scale experiments in the laboratory where automated digital tracking could allow scaling from sub-second intervals to that of several hours. This could facilitate studies designed to reveal the important timescales, and temporal ordering, over which social networks form and function.

Acknowledgements Financial support was provided by an Office of Naval Research Award #N00014-09-1-1074 (Christos C. Ioannou and Iain D. Couzin), a Searle Scholar Award (Iain D. Couzin) and aNational Science Foundation Award PHY-0848755 (Iain D. Couzin). Jens Krause was supported by NERC (NE/D011035/1).

References Albert, R. & Barab´asi, A.L. (2002) Statistical mechanics of complex networks. Reviews of Modern Physics, 74, 47–97. Ame, J.M., Rivault, C. & Deneubourg, J.L. (2004) Cockroach aggregation based on strain odour recognition. Animal Behaviour, 68, 793–801. Atz, J.W. (1953) Orientation in Schooling Fishes, pp. 103. Office of Naval Research, Department of the Navy. Axelsen, B.E., Anker-Nilssen, T., Fossum, P., Kvamme, C. & Nøttestad, L. (2001) Pretty patterns but a simple strategy: predator–prey interactions between juvenile herring and Atlantic puffins observed with multibeam sonar. Canadian Journal of Zoology, 79, 1586–1596. Balch, T., Khan, Z. & Veloso, M. (2001) Automatically Tracking and Analyzing the Behavior of Live Insect Colonies, pp. 521–528. ACM, New York, NY. Bayliff, W.H. (1988) Integrity of schools of skipjack tuna, Katsuwonus pelamis, in the eastern Pacific Ocean, as determined from tagging data. Fishery Bulletin, 86, 631–643. Bikhchandani, S., Hirshleifer, D. & Welch, I. (1992) A theory of fads, fashion, custom, and cultural change as informational cascades. Journal of Political Economy, 100, 992. Bonabeau, E. & Dagorn, L. (1995) Possible universality in the size distribution of fish schools. Physical Review E, 51, 5220–5223. Bonabeau, E., Dagorn, L. & Fr´eon, P. (1999) Scaling in animal group-size distributions. Proceedings of the National Academy of Sciences of the United States of America, 96, 4472. Breder Jr, C.M. (1954) Equations descriptive of fish schools and other animal aggregations. Ecology, 35, 361–370. Brown, C., Laland, K.N. & Krause, J. (2003) Learning in Fishes: Why They Are Smarter than You Think, Special edition of Fish and Fisheries. Blackwell Publishing Ltd., Oxford.

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

235

Brown, J.S. (1948) Gradients of approach and avoidance responses and their relation to level of motivation. Journal of Comparative and Physiological Psychology, 41, 450–465. Buhl, J., Sumpter, D.J.T., Couzin, I.D., Hale, J.J., Despland, E., Miller, E.R. & Simpson, S.J. (2006) From disorder to order in marching locusts. Science, 312, 1402–1406. Bumann, D. & Krause, J. (1993) Front individuals lead in shoals of three-spined sticklebacks (Gasterosteus aculeatus) and juvenile roach (Rutilus rutilus). Behaviour, 125, 189–198. Bumann, D., Krause, J. & Rubenstein, D. (1997) Mortality risk of spatial positions in animal groups: the danger of being in the front. Behaviour, 134, 1063–1076. Burgess, J.W. & Shaw, E. (1981) Effects of acoustico-lateralis denervation in a facultative schooling fish: a nearest-neighbor matrix analysis. Behavioral and Neural Biology, 33, 488. Couzin, I.D. & Krause, J. (2003) Self-organization and collective behavior in vertebrates. Advances in the Study of Behavior, 32, 1–75. Couzin, I.D., Krause, J., Franks, N.R. & Levin, S.A. (2005) Effective leadership and decision-making in animal groups on the move. Nature, 433, 513–516. Couzin, I.D., Krause, J., James, R., Ruxton, G.D. & Franks, N.R. (2002) Collective memory and spatial sorting in animal groups. Journal of Theoretical Biology, 218, 1–11. Croft, D.P., Arrowsmith, B.J., Bielby, J., Skinner, K., White, E., Couzin, I.D., Magurran, A.E., Ramnarine, I. & Krause, J. (2003) Mechanisms underlying shoal composition in the Trinidadian guppy (Poecilia reticulata). Oikos, 100, 429–438. Croft, D.P., James, R. & Krause, J. (2008) Exploring Animal Social Networks. Princeton University Press, Princeton, NJ. Croft, D.P., James, R., Ward, A.J.W., Botham, M.S., Mawdsley, D. & Krause, J. (2005) Assortative interactions and social networks in fish. Oecologia, 143, 211–219. Croft, D.P., Krause, J., Darden, S.K., Ramnarine, I.W., Faria, J.J. & James, R. (2009) Behavioural trait assortment in a social network: patterns and implications. Behavioral Ecology and Sociobiology, 63, 1495–1503. Croft, D.P., Krause, J. & James, R. (2004) Social networks in the guppy (Poecilia reticulata). Proceedings of the Royal Society of London Biology Letters, 271, 516–519. Dall, S.R.X., Giraldeau, L.A., Olsson, O., McNamara, J.M. & Stephens, D.W. (2005) Information and its use by animals in evolutionary ecology. Trends in Ecology & Evolution, 20, 187–193. Danchin, E., Giraldeau, L.A., Valone, T.J. & Wagner, R.H. (2004) Public information: from nosy neighbors to cultural evolution. Science, 305, 487. DeBlois, E.M. & Rose, G.A. (1996) Cross-shoal variability in the feeding habits of migrating Atlantic cod (Gadus morhua). Oecologia, 108, 192–196. Franz, M. & Nunn, C. (2009) Network-based diffusion analysis: a new method for detecting social learning. Proceedings of the Royal Society of London Series B – Biological Sciences, 276(1663), 1829–1836. Giraldeau, L.A., Valone, T.J. & Templeton, J.J. (2002) Potential disadvantages of using socially acquired information. Philosophical Transactions: Biological Sciences, 357, 1559–1566. Godin, J.G.J., Classon, L.J. & Abrahams, M.V. (1988) Group vigilance and shoal size in a small characin fish. Behaviour, 104, 29–40. Godin, J.G.J. & Morgan, M.J. (1985) Predator avoidance and school size in a cyprinodontid fish, the banded killifish (Fundulus diaphanus Lesueur). Behavioral Ecology and Sociobiology, 16, 105–110. Gr¨unbaum, D. (1998) Schooling as a strategy for taxis in a noisy environment. Evolutionary Ecology, 12, 503–522. Gr¨unbaum, D., Viscido, S. & Parrish, J.K. (2004) Extracting interactive control algorithms from group dynamics of schooling fish. Lecture Notes in Control and Information Sciences, 309, 103–117. Gueron, S. & Levin, S.A. (1995) The dynamics of group formation. Mathematical Biosciences, 128, 243–264. Harcourt, J.L., Ang, T.Z., Sweetman, G., Johnstone, R.A. & Manica, A. (2009) Social feedback and the emergence of leaders and followers. Current Biology, 19, 248–252.

BLBK374-10

BLBK374-Brown

236

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Helfman, G.S. (1984) School fidelity in fishes: the yellow perch pattern. Animal Behaviour, 32, 663–672. Helfman, G.S. & Schultz, E.T. (1984) Social transmission of behavioural traditions in a coral reef fish. Animal Behaviour, 32, 379–384. Hemmings, C.C. (1966) Olfaction and vision in fish schooling. Journal of Experimental Biology, 45, 449–464. Hensor, E., Couzin, I.D., James, R. & Krause, J. (2005) Modelling density-dependent fish shoal distributions in the laboratory and field. Oikos, 110, 344–352. Hoare, D.J., Couzin, I.D., Godin, J.G.J. & Krause, J. (2004) Context-dependent group size choice in fish. Animal Behaviour, 67, 155–164. Hoare, D.J. & Krause, J. (2003) Social organisation, shoal structure and information transfer. Fish and Fisheries, 4, 269–279. Hoppitt, W., Boogert, N. & Laland, K. (2010) Detecting social transmission in networks. Journal of Theoretical Biology, 263(4), 544–555. Hunter, J.R. (1969) Communication of velocity changes in jack mackerel (Trachurus symmetricus) schools. Animal Behaviour, 17, 507–514. Huth, A. & Wissel, C. (1994) The simulation of fish schools in comparison with experimental data. Ecological Modelling, 75, 135–146. Inada, Y. & Kawachi, K. (2002) Order and flexibility in the motion of fish schools. Journal of Theoretical Biology, 214, 371–387. James, R., Croft, D.P. & Krause, J. (2009) Potential banana skins in animal social network analysis. Behavioral Ecology and Sociobiology, 63, 989–997. Jeanson, R., Rivault, C., Deneubourg, J.L., Blanco, S., Fournier, R., Jost, C. & Theraulaz, G. (2005) Self-organized aggregation in cockroaches. Animal Behaviour, 69, 169–180. Kennedy, J., Eberhart, R.C. & Shi, Y. (2001) Swarm Intelligence. Springer, Berlin. Klimley, A.P. & Holloway, C.F. (1999) School fidelity and homing synchronicity of yellowfin tuna, Thunnus albacares. Marine Biology, 133, 307–317. Krause, J. (1993a) The relationship between foraging and shoal position in a mixed shoal of roach (Rutilus rutilus) and chub (Leuciscus cephalus): a field study. Oecologia, 93, 356–359. Krause, J. (1993b) Transmission of fright reaction between different species of fish. Behaviour, 127, 37–48. Krause, J. (1993c) The effect of ‘Schreckstoff ’ on the shoaling behaviour of the minnow: a test of Hamilton’s selfish herd theory. Animal Behaviour, 45, 1019–1024. Krause, J. (1994) Differential fitness returns in relation to spatial position in groups. Biological Reviews of the Cambridge Philosophical Society, 69, 187. Krause, J., Butlin, R.K., Peuhkuri, N. & Pritchard, V.L. (2000a) The social organization of fish shoals: a test of the predictive power of laboratory experiments for the field. Biological Reviews, 75, 477–501. Krause, J., Croft, D.P. & James, R. (2007) Social network theory in the behavioural sciences: potential applications. Behavioral Ecology and Sociobiology, 62, 15–27. Krause, J., Croft, D.P. & James, R. (2010) Personality in the Context of Social Networks. Philosophical Transactions of the Royal Society of London Series B – Biological Sciences, 365(1560), 4099–4106, doi: 10.1098/rstb.2010.0216. Krause, J. & Godin, J.G.J. (1995) Predator preferences for attacking particular prey group sizes: consequences for predator hunting success and prey predation risk. Animal Behaviour, 50, 465–473. Krause, J., Hoare, D., Krause, S., Hemelrijk, C.K. & Rubenstein, D.I. (2000b) Leadership in fish shoals. Fish and Fisheries, 1, 82–89. Krause, J. & Ruxton, G.D. (2002) Living in Groups. Oxford University Press, Oxford, USA. Krause, J. & Tegeder, R.W. (1994) The mechanism of aggregation behaviour in fish shoals: individuals minimize approach time to neighbours. Animal Behaviour, 48, 353–359. Krause, J., Ward, A.J.W., Jackson, A.L., Ruxton, G.D., James, R. & Currie, S. (2005) The influence of differential swimming speeds on composition of multi-species fish shoals. Journal of Fish Biology, 67, 866–872.

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

237

Laland, K.N. (2004) Social learning strategies. Learning & Behavior, 32, 4–14. Laland, K.N. & Williams, K. (1997) Shoaling generates social learning of foraging information in guppies. Animal Behaviour, 53, 1161–1169. Leblond, C. & Reebs, S.G. (2006) Individual leadership and boldness in shoals of golden shiners (Notemigonus crysoleucas). Behaviour, 143, 1263–1280. Livnat, A. & Pippenger, N. (2006) An optimal brain can be composed of conflicting agents. Proceedings of the National Academy of Sciences, 103, 3198. Lusseau, D. & Newman, M.E.J. (2004) Identifying the role that animals play in their social networks. Proceedings: Biological Sciences, 271, 477–481. Magurran, A.E. & Higham, A. (1988) Information transfer across fish shoals under predator threat. Ethology, 78, 153–158. Magurran, A.E. & Pitcher, T.J. (1987) Provenance, shoal size and the sociobiology of predator-evasion behaviour in minnow shoals. Proceedings of the Royal Society of London Series B – Biological Sciences, 229, 439–465. Makris, N.C., Ratilal, P., Jagannathan, S., Gong, Z., Andrews, M., Bertsatos, I., Godo, O.R., Nero, R.W. & Jech, J.M. (2009) Critical population density triggers rapid formation of vast oceanic fish shoals. Science, 323, 1734–1737. Makris, N.C., Ratilal, P., Symonds, D.T., Jagannathan, S., Lee, S. & Nero, R.W. (2006) Fish population and behavior revealed by instantaneous continental shelf-scale imaging. Science, 311, 660–663. Martin, P. & Bateson, P. (1993) Measuring Behaviour: An Introductory Guide. Cambridge University Press, Cambridge. Maynard-Smith, J. (1982) Evolution and the Theory of Games. Cambridge University Press, Cambridge. McClure, S.M., Laibson, D.I., Loewenstein, G. & Cohen, J.D. (2004) Separate neural systems value immediate and delayed monetary rewards. Science, 306, 503. Miller, N.E. (1944) Experimental studies of conflict. Personality and the Behavior Disorders, 1, 431–465. Moody, J. (2002) The importance of relationship timing for diffusion. Social Forces, 81, 25–56. Neill, S.R., St, J. & Cullen, J.M. (1974) Experiments on whether schooling by their prey affects the hunting behaviour of cephalopods and fish predators. Journal of Zoology (London), 172, 569. Newman, M.E.J. (2003) The structure and function of complex networks. Siam Review, 45, 167–256. Newman, M.E.J. (2004) Detecting community structure in networks. The European Physical Journal B – Condensed Matter and Complex Systems, 38, 321–330. Newman, M.E.J. & Girvan, M. (2004) Finding and evaluating community structure in networks. Physical Review E, 69, 26113. Niwa, H.S. (1998a) School size statistics of fish. Journal of Theoretical Biology, 195, 351–361. Niwa, H.S. (1998b) Migration dynamics of fish schools in heterothermal environments. Journal of Theoretical Biology, 193, 215–231. Nøttestad, L. & Axelsen, B.E. (1999) Herring schooling manoeuvres in response to killer whale attacks. Canadian Journal of Zoology, 77, 1540–1546. Parr, A.E. (1927) A contribution to the theoretical analysis of the schooling behavior of fishes. Occasional Papers of the Bingham Oceanographic Collection, 1, 1–32. Parrish, J.K. & Edelstein-Keshet, L. (1999) Complexity, pattern, and evolutionary trade-offs in animal aggregation. Science, 284, 99. Partridge, B.L. (1980) The effect of school size on the structure and dynamics of minnow schools. Animal Behaviour, 28, 68–77. Partridge, B.L. (1981) Internal dynamics and the interrelations of fish in schools. Journal of Comparative Physiology A: Neuroethology, Sensory, Neural, and Behavioral Physiology, 144, 313–325. Partridge, B.L. (1982) The structure and function of fish schools. Scientific American, 246, 114–123. Partridge, B.L. & Pitcher, T.J. (1980) The sensory basis of fish schools: relative roles of lateral line and vision. Journal of Comparative Physiology A: Neuroethology, Sensory, Neural, and Behavioral Physiology, 135, 315–325.

BLBK374-10

BLBK374-Brown

238

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Partridge, B.L., Pitcher, T., Cullen, J.M. & Wilson, J. (1980) The three-dimensional structure of fish schools. Behavioral Ecology and Sociobiology, 6, 277–288. Pike, T.W., Samanta, M., Lindstr¨om, J. & Royle, N.J. (2008) Behavioural phenotype affects social interactions in an animal network. Proceedings of the Royal Society of London Series B – Biological Sciences, 275, 2515. Pitcher, T.J. & Parrish, J.K. (1993) Functions of shoaling behaviour in teleosts. Behaviour of Teleost Fishes, 2, 369–439. Pitcher, T.J., Partridge, B.L. & Wardle, C.S. (1976) A blind fish can school. Science, 194, 963. Quinn, T.P. & Dittman, A.H. (1990) Pacific salmon migrations and homing: mechanisms and adaptive significance. Trends in Ecology & Evolution, 5, 174–177. Radakov, D.V. (1973) Schooling in the Ecology of Fish. John Wiley & Sons Inc., New York. Reader, S.M. & Laland, K.N. (2000) Diffusion of foraging innovations in the guppy. Animal Behaviour, 60, 175–180. Reebs, S.G. (2000) Can a minority of informed leaders determine the foraging movements of a fish shoal? Animal Behaviour, 59, 403–409. Reebs, S.G. (2001) Influence of body size on leadership in shoals of golden shiners, Notemigonus crysoleucas. Behaviour, 138, 797–809. Romey, W.L. (1996) Individual differences make a difference in the trajectories of simulated schools of fish. Ecological Modelling, 92, 65–77. Ruxton, G.D., Jackson, A.L. & Tosh, C.R. (2007) Confusion of predators does not rely on specialist coordinated behavior. Behavioral Ecology, 18(3), 590–596. Seghers, B.H. (1974) Schooling behavior in the guppy (Poecilia reticulata): an evolutionary response to predation. Evolution, 28, 486–489. Seghers, B.H. (1981) Facultative schooling behavior in the spottail shiner (Notropis hudsonius): possible costs and benefits. Environmental Biology of Fishes, 6, 21–24. Snekser, J.L., McRobert, S.P., Murphy, C.E. & Clotfelter, E.D. (2006) Aggregation behavior in wildtype and transgenic zebrafish. Ethology, 112, 181–187. Sumpter, D.J.T. (2006) The principles of collective animal behaviour. Philosophical Transactions of the Royal Society of London Series B – Biological Sciences, 361, 5–22. Swaney, W., Kendal, J., Capon, H., Brown, C. & Laland, K.N. (2001) Familiarity facilitates social learning of foraging behaviour in the guppy. Animal Behaviour, 62, 591–598. Tegeder, R.W. & Krause, J. (1995) Density dependence and numerosity in fright stimulated aggregation behaviour of shoaling fish. Philosophical Transactions: Biological Sciences, 350, 381–390. Tien, J.H., Levin, S.A. & Rubenstein, D.I. (2004) Dynamics of fish shoals: identifying key decision rules. Evolutionary Ecology Research, 6, 555–565. Tinbergen, N. (1952) “Derived” activities; their causation, biological significance, origin, and emancipation during evolution. The Quarterly Review of Biology, 27, 1–32. Treherne, J.E. & Foster, W.A. (1981) Group transmission of predator avoidance behaviour in a marine insect: The Trafalgar effect. Animal Behaviour, 29, 911–917. Turner, G.F. & Pitcher, T.J. (1986) Attack abatement: a model for group protection by combined avoidance and dilution. American Naturalist, 128, 228. Usher, M. & McClelland, J.L. (2004) Loss aversion and inhibition in dynamical models of multialternative choice. Psychological Review, 111, 757–769. Vabø, R. & Nøttestad, L. (1997) An individual based model of fish school reactions: predicting antipredator behaviour as observed in nature. Fisheries Oceanography, 6, 155–171. Viscido, S.V., Parrish, J.K. & Gr¨unbaum, D. (2004) Individual behavior and emergent properties of fish schools: a comparison of observation and theory. Marine Ecology Progress Series, 273, 239–249. Ward, A., Axford, S. & Krause, J. (2002a) Mixed-species shoaling in fish: the sensory mechanisms and costs of shoal choice. Behavioral Ecology and Sociobiology, 52, 182–187. Ward, A.J.W., Hoare, D.J., Couzin, I.D., Broom, M. & Krause, J. (2002b) The effects of parasitism and body length on positioning within wild fish shoals. Journal of Animal Ecology, 71, 10–14.

BLBK374-10

BLBK374-Brown

May 12, 2011

4:31

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Organisation and Information Transfer in Schooling Fish

Char Count=

239

Ward, A.J.W., Sumpter, D.J.T., Couzin, I.D., Hart, P.J.B. & Krause, J. (2008) Quorum decision-making facilitates information transfer in fish shoals. Proceedings of the National Academy of Sciences, 105, 6948. Ward, A.J.W., Thomas, P., Hart, P.J.B. & Krause, J. (2004) Correlates of boldness in three-spined sticklebacks (Gasterosteus aculeatus). Behavioral Ecology and Sociobiology, 55, 561–568. Ward, A.J.W., Botham, M.S., Hoare, D.J., James, R., Broom, M., Godin, J.G.J. & Krause, J. (2002c) Association patterns and shoal fidelity in the three-spined stickleback. Proceedings of the Royal Society of London Series B – Biological Sciences, 269, 2451. Ward, A.J.W., Webster, M.M., Magurran, A.E., Currie, S. & Krause, J. (2009) Species and population differences in social recognition between fishes: a role for ecology? Behavioral Ecology, 20(3), 511–516. Whitehead, H. & Dufault, S. (1999) Techniques for analyzing vertebrate social structure using identified individuals: review and recommendations. Advances in the Study of Behavior, Vol. 28, pp. 33–74. Academic Press, San Diego, CA.

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 11

Social Learning in Fishes Culum Brown and Kevin Laland

11.1

Introduction

In making decisions, such as how to find food and mates, or avoid predators, many animals utilise information that is produced by others. Such individuals are referred to as ‘observers’ in the social learning literature (Heyes & Galef 1996) or ‘eavesdroppers’ in the signalreceiver literature (McGregor 1993). Socially transmitted information may simply be an inadvertent by-product of the ‘demonstrating’ individual’s behaviour, or a signal targeted towards a particular individual or audience. Any learning that involves the use of socially provided information is termed ‘social learning’. Social learning refers to any incidence in which individuals acquire new behaviour or information about their environment via observation of, or interaction with, other animals or their products. It is generally assumed that social learning is beneficial because na¨ıve individuals can acquire locally adaptive behaviour quickly and efficiently from more knowledgeable individuals (Boyd & Richerson 1985), for instance, without having to incur the costs of exploration or the risks of learning about predators. Social learning is sometimes assumed to be more common in, of a more sophisticated form in, or even restricted to, ‘intelligent’ or ‘large-brained’ taxa. However, research over the last 50 years has demonstrated that social learning is common amongst fishes, birds and mammals, and even invertebrates (Leadbeater & Chittka 2007), and should now be regarded as a regular feature of animal life (Lefebvre & Palameta 1988; Heyes & Galef 1996). Moreover, recent studies have revealed that individuals, including fishes, will not always utilise socially available information, and will switch between reliance on social and asocial sources of information according to the costs and benefits (Laland 2004; Kendal et al. 2005). Many of the processes that underpin social learning lend themselves particularly well to shoals of fishes, although are by no means restricted to shoaling species. For example, social learning through ‘local enhancement’ or ‘stimulus enhancement’ occurs when the behaviour (or simply the presence) of one individual attracts the attention of another individual to a particular location or stimulus, about which the na¨ıve individual subsequently learns something. For instance, observers may learn that food is found at that location. ‘Social

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

241

facilitation’ occurs when the behaviour of one individual induces an identical behaviour in another individual and the latter then learns something via the expression of this behaviour, for instance it learns about the consequences of producing the behaviour in that context. ‘Guided learning’ or ‘exposure’ refers to instances in which, by following or being with a knowledgeable animal, an individual is exposed to similar features of the environment and comes to learn the same behaviour. ‘Observational conditioning’ occurs when the response of a demonstrator to a stimulus elicits a matching response on the part of an observer, who simultaneously perceives the original stimulus, and effectively learns that the response is an appropriate response to it. For instance, rhesus monkeys (Macaca mulatta) can acquire a fear of snakes when they witness experienced monkeys behaving fearfully in the presence of a snake (Mineka & Cook 1988). A variety of other processes, some assuming more sophisticated psychological underpinnings, can also facilitate social learning in animals, described variously as imitation, goal emulation, copying, etc. (see Whiten & Ham 1992; Heyes 1994; Hoppitt & Laland 2008, for reviews). We know of no laboratory evidence that fishes are capable of imitation, i.e. learning to produce particular bodily movements through observation of others, although there is plenty of experimental evidence for observational learning in fishes (Coolen et al. 2003; Schuster et al. 2006; Grosnick et al. 2007). However, suggestive circumstantial evidence of imitation is provided by Mazeroll & Montgomery’s (1995) report that in the local migrations of Brown surgeonfish (Acanthurus nigrofuscus, Acanthuridae), followers not only take the route of leaders but reproduce their postural changes (e.g. dips and rolls). However, it is a feature of animal social learning that simple processes are sufficient both to allow individuals to acquire adaptive information and to mediate behavioural traditions in populations. Over the last 20 years the study of social learning has expanded greatly and is now a major topic of research in ethology, behavioural ecology and comparative psychology (Galef & Giraldeau 2001; Shettleworth 2001). Documented cases of social learning in fishes are now commonplace and our growing understanding of the underlying mechanisms is enabling further identification and clarification of such phenomena. Here we review evidence that social learning plays a role in fish (1) antipredator behaviour, (2) migration and orientation, (3) foraging, (4) mate choice, and (5) aggression, all of which may be facilitated by eavesdropping.

11.2

Antipredator behaviour

It is well known that the antipredator behaviour of many fishes has a learned component, and countless species of fishes improve their antipredator response with experience (Kieffer & Colgan 1992; Kelley & Magurran Chapter 3 of this volume). However, learning about predators is a risky business, and there is little room for mistakes or extensive training (Kelley & Magurran 2003). Given the large risks and limited opportunities for asocial learning, it is easy to see the potential benefits of learning about predators from others. Not surprisingly, social learning of predator presence or predator identity appears widespread in fishes (Suboski & Templeton 1989; Brown & Laland 2001) as well as in other animals (reviewed by Griffin 2004).

BLBK374-11

BLBK374-Brown

242

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

The use of socially transmitted information enables individuals to respond to threats without having to verify the presence of danger independently. Fishes are particularly well equipped for rapid information transfer via sight and the lateral-line system. Such information may pass through a shoal much more quickly than the speed of an approaching predator or the diffusion of an alarm substance (Webb 1980; Potts 1984) resulting in what appears to be a synchronous response of shoal members. While such responses are not in themselves social learning (being merely manifestations of communication between shoal members), in the process young and inexperienced individuals frequently learn to identify predators, acquire appropriate antipredator responses, or refine these responses. For example, Kelley et al. (2003) found that wild-caught guppies captured in low-predation areas altered their response to predators when paired with conspecifics from high-, but not low-, predation locations in the presence of a predator model. Low-predation fishes paired with their more experienced conspecifics increased their shoaling cohesion and inspected the predator model from a greater distance. Members of fish shoals are able to make decisions about predators by observing changes in the behaviour of shoalmates. When fishes return from a predator-inspection visit it appears that the behaviour of the rest of the shoal changes in response to the level of threat perceived by the inspectors as manifest in their behaviour (Pitcher et al. 1986). When fishes are attacked or startled by the presence of a predator they frequently display a fright response. It has long been known that the fright responses of many fishes can be precipitated by a chemical alarm substance emitted by fishes (von Frisch, cited in Pfeiffer 1974; see Brown 2003 and Chapter 4). However, von Frisch discovered that anosmic fish joined the fright reactions of conspecifics and subsequent research has established that such fright responses can be visually transmitted (Verheijen 1956). The fright response provides an obvious visual cue that a predator is in the vicinity. For example, European minnows (Phoxinus phoxinus, Cyprinidae) increased the frequency of flight responses after observing the flight response of conspecifics in a neighbouring tank that had been threatened by a predator (Magurran & Higham 1988). Not only is the speed of communication considerably increased over the diffusion of a chemical, but predator presence can be communicated to and by fishes that have no direct experience with the predator. Social learning of fright responses has also been reported in mixed-species shoals, where the fright response of one species elicits a similar response in a second (Krause 1993; Mathis et al. 1996). Conceivably the elevated antipredator responses of fishes that have recently witnessed conspecifics and heterospecifics behaving fearfully in the presence of a predator could be the result of a number of processes. Central to the story would seem to be the formation of an association between the fright reaction of conspecifics and the predator. Observers may learn the identity of a predator by avoiding anything that elicits fright responses in conspecifics. Evidence for this comes from Suboski et al. (1990), who demonstrated that minnows could learn a fright response to olfactory cues from a novel predator if these cues were presented at the same time as either the visual fright reactions or the alarm substance of conspecifics. A similar mechanism can result in fish learning about risky habitats (Chivers & Smith 1995). These are examples of observational conditioning. It would seem that many fishes exhibit an unlearned fright response to the alarm substance (von Frisch, cited in Pfeiffer 1974), which would be sufficient for it to act as an unconditioned stimulus and would, through Pavlovian conditioning, explain the conditioned response to the predator

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

243

when the alarm substance and predator are paired. It is less clear how the fright reactions of conspecifics should play this role. Perhaps they too elicit an unlearned fear response, but more plausibly they could have acquired higher order conditioning properties through prior association with the alarm substance, or with other predators. In addition, observers may learn to flee, seek refuge, freeze or shoal tightly when others do so and that these behaviour patterns are appropriate responses to the predator. Furthermore, observers may have their motivational or physiological state changed as a result of perception of fearful conspecifics in a manner that leaves them more likely to learn about all aspects of their environment. Clearly a complete understanding of this apparently simple example of social learning awaits further analysis. With the exception of the Kelley et al. (2003) study described in the preceding text, as yet there is little direct evidence that fishes learn antipredator responses to natural predators through social learning (i.e. that they learn to respond differently to different threats, such as shoaling vs. hiding). Magurran & Higham (1988) reported that minnows (P. phoxinus) decreased their activity (an antipredatory response to reduce conspicuousness) after observing the startle responses of predator-inspecting group mates, indicating that antipredator information is transmitted between individuals. Moreover, investigations into the avoidance responses of guppies and zebrafish to artificial predators (Sugita 1980; Brown & Laland 2002a; Lindeyer & Reader 2010) suggest that the learning of specific antipredator behaviour is possible. Sugita (1980) found that guppies (Poecilia reticulata, Poeciliidae) learn to avoid an electric shock by following demonstrator fish into one of two safe compartments in a shuttle box. Brown & Laland (2002a) exposed na¨ıve fishes to a model trawl apparatus, which simulates a predator. Half of the na¨ıve fishes were placed with demonstrators trained to take one escape route and the others to demonstrators trained to take an alternative route. Both groups remained faithful to the demonstrated route and escaped more frequently than control groups while the demonstrators were present. However, once the demonstrators were removed, although the experimental groups were still far more efficient at escaping than control groups that had had no demonstrators, their fidelity to the demonstrated route degraded. These results suggest that the fish learnt an appropriate escape response by following the example set by demonstrators. While, in the absence of clear demonstration, functional aspects of traditional behaviour were maintained (i.e. to escape by swimming through a hole in the trawl), more arbitrary components were lost (i.e. utilisation of the previously demonstrated route when equally viable alternatives were available). Lindeyer & Reader (2010) achieved equivalent results with zebrafish (Danio rerio). Similar experiments carried out on natural populations of Trinidadian guppies in situ found that even in the absence of demonstrators the route preference was maintained (Reader et al. 2003; Fig. 11.1), suggesting that social learning about antipredator responses is heightened in wild populations compared to pet-shop strains. Population variation in social learning performance has yet to receive much attention from researchers, but may shed further light on the evolution of social learning by revealing the ecological contexts in which it is favoured (see the following text for further discussion). Given the mounting evidence that social learning about predators appears to occur readily in fish populations, there is increased interest in the possibility that social learning may be used as a tool to train na¨ıve, hatchery-reared fishes to recognise predators (Brown & Laland 2001; Brown & Day 2002). Vilhunen and colleagues have examined social learning of

BLBK374-Brown

244

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

0 –10 Escape latency (s)

BLBK374-11

–20 –30 –40 –50 –60 –70 –80 Red observers

Yellow observers

Fig. 11.1 The mean (SE) difference in escape latency (red route escape latency –yellow escape route latency) in wild guppies paired with demonstrators trained to escape via the red and yellow escape routes. Negative scores indicate a preference to escape via the red route. (Based on Reader et al. 2003.)

antipredator responses in endangered, land-locked salmonids. These populations are fully maintained by reintroductions from hatchery stocks after hydroelectric schemes destroyed their primary spawning grounds. Initial studies revealed that socially acquired avoidance of predator odours is particularly effective means of training na¨ıve fishes to recognise predators (Vilhunen et al. 2005) and these early exposures to predators do lead to improved survival upon subsequent contact with live predators (Vilhunen 2006).

11.3

Migration and orientation

A number of studies have explored how through social learning fishes may learn to orientate around their environment and, in the process, learn the route to food sites, resting sites, schooling sites or mating sites. One of the most elegant demonstrations of social learning in a natural population of fish was carried out by Helfman & Schultz (1984). Helfman et al. (1982) had discovered that specific resting sites in coral reefs contained groups of French grunts (Haemulon flavolineatum, Haemulidae) that made daily migrations to feeding grounds. These groups appeared to be joined occasionally by newly recruited juveniles that had followed older individuals and seemingly had subsequently learned the migration path. Helfman & Schultz (1984) tested this by transplanting individuals between resting locations and then recorded their path towards foraging grounds. In the experimental condition the transplanted juveniles were allowed to follow the residents for 2 days before the residents were removed. In a control condition the resident population were removed prior to transplanting. While the experimental fishes learnt the same migration path as the resident adults, the control fishes continued to use paths appropriate to their original resting site. Bluehead wrasse (Thalassoma bifasciatum, Labridae) show similar migratory traditions. These fishes have mating-site locations that remain constant over many generations. Warner (1988, 1990) removed entire populations and replaced them with transplanted populations. Not only did the new fishes establish entirely new mating grounds, but these new locations remained constant over subsequent generations. In the 12 years of studying 22 patches on

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

245

the reef not once was a new mating site established or lost despite fluctuations in the wrasse population size. Combined with the observation that reef populations are not subject to significant genetic differentiation (Warner 1988), this finding provides strong evidence of cultural variation. In laboratory experiments, Laland & Williams (1997, 1998) found that na¨ıve ‘observer’ guppies could learn a route to a foraging patch by following more knowledgeable ‘demonstrators’. Observers were placed in the presence of demonstrators that had previously been trained to take one of two routes to feed. Observers typically shoaled with the demonstrators and fed at the feeding site. When the demonstrators were removed, the observers continued to utilise the same route to feed, despite the availability of an alternative route. Traditions were established in small populations in which experienced fishes were repeatedly replaced with na¨ıve conspecifics, and yet the route preferences remained. Laland & Williams (1997) found that as the number of demonstrators increased the more likely it was that observers would remain faithful to the route. This is consistent with the hypothesis that animal social learning is commonly ‘conformist’, with the rate of information transmission increasing with the number of individuals displaying the behaviour. The strength of social learning in Laland & Williams’ (1997) experiment was such that guppies would even maintain traditions for maladaptive1 circuitous routes for brief periods. This conformity hypothesis is particularly important in group-living species that are under heavy selective pressure from predators to look and behave similarly to other group members (Brown & Laland 2002a). What are the mechanisms underlying such natural- and laboratory-based migratory traditions? Some recent experimental and theoretical studies shed light on this issue. Reebs (2000, 2001) has carried out experimental studies of leading and following in fishes which demonstrate how small numbers of individuals can direct the movements of entire groups. Among shoals of 12 golden shiners, Notemigonus crysoleucas (Cyprinidae), some individuals were trained to swim to a specific location in a large tank where they were fed at the same time each day. The tank was divided into two areas, one of which was shaded and preferred by the fishes, the other brightly lit but where food was delivered. In experiments with shoals composed of a mix of trained and na¨ıve fishes, in which the trained fishes were in the minority (down to a single trained individual in the shoal), Reebs demonstrated that even a few trained fishes were capable of leading the entire shoal to the feeding area at the time that food was delivered. Reebs’ empirical findings receive theoretical support from an agent-based model developed by Couzin et al. (2005). These researchers programmed virtual individuals with an aggregation rule, which endowed them with a tendency to be close to and aligned with neighbouring individuals, and provided a subset of individuals with information about a preferred direction, representing, for example, the direction to a known resource or segment of a migration route. Couzin et al. showed that this information could be transferred within the group without signalling and when group members do not know which individuals have information. Moreover, the larger the group, the smaller the proportion of informed

1

Note this experiment provides evidence for the social transmission of maladaptive information (i.e. take the long route), and for a suboptimal behavioural tradition (for taking the long route), but neither the behaviour of the fish (where it pays to shoal for protection from predators) nor the general capacity for social learning (which is typically advantageous) should be described as maladaptive.

BLBK374-Brown

246

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

individuals needed to guide the group (see also Chapter 10). The model suggests that, merely by shoaling, a small group of demonstrator fishes that are highly motivated to swim to a particular location can lead a much larger group of na¨ıve individuals to that location, even if the na¨ıve individuals do not pay particular attention to the demonstrators or even know who they are. Previous hypotheses put forward to account for learned migratory traditions in animals often assumed that young or na¨ıve individuals paid particular attention to, or learned to follow, older or more experienced conspecifics. The Couzin et al. explanation is attractive since it relies on what would appear to be a much simpler mechanism than such accounts, and helps to explain why blocking does not prevent na¨ıve individuals from learning an association between spatial cues and reinforcement. Further insight comes from a state-dependent model of foraging by Rands et al. (2003, 2004) which found that, assuming some advantage to foraging together, the individual with lower reserves determines when a pair of individuals should forage. They suggest that group coordination may be resolved through the spontaneous emergence of temporary ‘leaders’ and ‘followers’, owing to the build-up of differences in energetic state. In support of this, studies of foraging in groups of fishes have shown that leadership decisions may often be made by individuals with lower reserves (Krause et al. 1998). The analysis, once again, implies that simple rules of thumb, which require no detailed knowledge of the state of other individuals, can lead to coordinated behaviour on the part of a group. There is also laboratory experimental evidence that the natural shoaling tendency exhibited by many fishes can generate a simple form of conformity, which enhances the stability of group behaviour. This is illustrated by an experiment by Day et al. (2001) investigating how shoal size affects foraging efficiency in fishes. It is generally accepted that fishes in large shoals find food sooner than those in small shoals, largely as a consequence of having more socially transmitted information available to them. However, Day et al. (2001) discovered that where the food patch is visually isolated from the shoal, smaller groups of fishes discover it more quickly than larger groups (Fig. 11.2). Day et al. interpret this Opaque partition

Transparent partition 600

600 Shoal foraging latency (s)

BLBK374-11

4 16

500

500

400

400

300

300

200

200

100

100

0 1

5

10

0 1 15 Trial number

5

10

15

Fig. 11.2 Large shoals find the location of a foraging patch faster when the patch is located behind a transparent partition. However, the reverse is true when the patch is located behind an opaque partition. (After Day et al. 2001.)

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

247

apparent contradiction as stemming from the fact that fishes are generally reluctant to leave the safety of the shoal and forage on their own at a patch that is hidden. The larger the shoal the stronger the apparent compulsion to remain with the shoal. In small shoals, the protection conferred by the shoal is low and individuals are more likely to break away and search for food. This reasoning is the logical inverse of the observation that fish will be more likely to join a large than a small shoal (Lachlan et al. 1998). Supporting this interpretation is Day et al.’s finding that individuals in large groups located food behind an otherwise identical transparent barrier faster than individuals in small groups. Consensus decision-making has also been reported in three-spine sticklebacks (Sumpter et al. 2008). We anticipate that the conformity illustrated in the preceding text might operate to maintain the stability of migratory traditions in fishes, since each individual would be penalised by increased vulnerability to predation if they were to take an alternative route to the rest of the shoal, even if it were, in some respects, more efficient. The simple shoaling and learning processes suggested by these findings probably underlie the traditions of coral reef fish, and perhaps the migrations of walleye (Stizostedion vitreum, Percidae) (Olson et al. 1978) and surgeon fish (Mazeroll & Montgomery 1995). It is still largely unknown if fishes use social learning to aid in the navigation during largescale migrations, such as those seen in salmonids and eels (see Odling-Smee & Braithwaite 2003 and Chapter 8; Tsukamoto et al. 2003).

11.4

Foraging

Given the interest among behavioural ecologists in prey choice (Stephens & Krebs 1986; see Warburton 2003 and Chapter 2), there has been surprisingly little attention given to how social learning might facilitate the acquisition of dietary preferences. However, social learning is implicated in patch choice, where the use of socially transmitted information is well established in fishes (Pitcher & House 1987; Ryer & Olla 1991). ‘Forage area copying’ (Barnard & Sibly 1981) appears to operate through a local enhancement process. Once discovered, the foraging behaviour of the finder of a patch acts as a cue to the rest of the shoal, which quickly join it. Groups of fishes find food faster than individuals because the probability of detecting a patch varies directly with group size while the time spent on vigilance is inversely related to group size (Morgan & Colgan 1987). In open water, goldfish (Carassius auratusauratus, Cyprinidae), bluntnose minnows (Pimephales notatus, Cyprinidae), Alaska pollock (Theragra chalcogramma, Gadidae), sticklebacks (Gasterosteus aculeatus, Gasterosteidae) and guppies all forage more efficiently in social groups than alone (Pitcher et al. 1982; Pitcher & House 1987; Morgan 1988; Ryer & Olla 1992; Peuhkuri et al. 1995; Day et al. 2001). There is also evidence that some fishes will adjust their copying of patch choice adaptively, according to the predation threat. Webster & Laland (2008) found that European minnows spent more time in a socially demonstrated patch than in a non-demonstrated alternative when simulated risk was high compared to when it was low, implying that foraging minnows disproportionately copy when learning for themselves would be costly. Increased foraging efficiency in a social context is not only restricted to shoaling species. Juvenile Atlantic salmon (Salmo salar, Salmonidae) take up benthic foraging stations from

BLBK374-11

BLBK374-Brown

248

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

which they dart to the surface to intercept prey items, then return to the river bed. Brown & Laland (2002b) set out to determine if the darting motion sends a message to other fishes that food is available, and whether this cue could be utilised by na¨ıve fishes to learn to forage on novel prey items. They found that, over the course of the experiment, 100% of individuals paired with pre-trained demonstrators learned to accept the novel prey. Na¨ıve fishes paired with equally na¨ıve individuals actually perform worse (50% of individuals learned to feed on novel prey items) than individuals learning in isolation (73%). The presence of an inactive conspecific provided negative feedback to both individuals and thus both were reluctant to begin feeding, a finding they labelled ‘social inhibition’. Social enhancement of foraging has also been reported in hatchery-reared juvenile chum salmon (Oncorhynchus keta, Salmonidae), Alaska pollock and rock bass (Ambloplites rupestris, Centrarchidae) (Templeton 1987; Ryer & Olla 1991, 1992). Sundstrom & Johnsson (2001) found an increase in foraging performance when in visual contact with another feeding conspecific in wild but not hatchery-reared brown trout (Salmo trutta, Salmonidae) suggesting that the conditions under which hatchery fishes are raised may diminish their ability to exploit social cues. Finally, there is experimental evidence that fishes can learn novel foraging behaviour through observation of conspecifics. Anthouard (1987) reports that juvenile European seabass (Dicentrarchus labrax, Moronidae) learned to press a lever to receive a food reward through observation of proficient trained demonstrators. These self-feeders are now commonly employed in aquaculture facilities (Shima et al. 2003). In a more naturalistic context, Schuster et al. (2006) provide striking evidence that archer fish (Toxotes jaculatrix) can learn to shoot down moving aerial prey through observation of the successful performance of conspecifics.

11.5

Mate choice

The role of learning in mate-choice decisions is discussed in detail elsewhere (Chapter 5). Here we provide only an overview of the evidence of social learning in mate choice. There is increasing evidence that social factors play a role in mate choice in many fishes (Westneat et al. 2000) including mollies (Poecilia latipinna, Poeciliidae) (Witte & Ryan 2002), guppies (Dugatkin 1992), gobies (Pomatoschistus microps, Gobiidae) (Reynolds & Jones 1999) and the Japanese rice fish (Oryzias latipes, Adrianichthyidae) (Grant & Green 1996). ‘Mate-choice copying’ is said to have occurred when the probability of an individual selecting another as a sexual partner increases because other individuals (of the same sex) have selected the same partner (Gibson & Hoglund 1992). In the paradigm experiment (e.g. Dugatkin 1992), two males are secured at the ends of an aquarium, one with a demonstrator female nearby. The observer, another female, placed centrally, watches the other female interact with one of the males. When, after the demonstrator has been removed, the observer is allowed to choose between the two males, she consistently chooses the male that had the female nearby. In mollies, similar observations consistent with mate-choice copying have been reported for males (Schlupp & Ryan 1997). In addition, Plath et al. (2008) report that male Atlantic mollies will change their courtship behaviour when observed by another male, to no longer court their preferred female and to approach a non-preferred female. he

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

249

authors interpret this as an act of deception, designed to mislead observers into pursuing the wrong female, as an adaptation to guard against the threat of sperm competition. There is also evidence suggesting that mate-choice copying occurs in the wild (Witte & Ryan 2002). One interpretation of these finding is that the observing female utilises the presence of the female near a male as an indication of his quality, and biases her male choice accordingly (Dugatkin & Godin 1992). It is assumed, that like other forms of social learning, matechoice copying reduces the time individuals spend sampling potential mates before coming to a decision on which mate is of the highest quality. However, there are a number of other possible interpretations of the data (e.g. shoalmate choice), and researchers have struggled to replicate some prominent findings (Brooks 1996, 1999; Lafleur et al. 1997). If fish choose mates on the basis of major histocompatability complex (MHC) compatibility (Milinski 2003), choosing mates based on the choice of others seems to make little sense, since optimal matches of MHC genotypes between partners will vary from individual to individual (Milinski et al. 2005). Mate-choice copying is by no means the only way that socially influenced mate choice may occur. For example, in a number of species females prefer males with a greater number of eggs present in their nest (sticklebacks, Goldsmidt et al. 1993; bullhead (Cotus gobio, Cottidae) Bisazza & Marconato 1988; darter (Etheostoma flabellare, Percidae), Knapp & Sergeant 1989; fathead minnows (Pimephales promelas, Cyprinidae), Unger & Sergent 1988, blennys (Aidablennius sphinx, Blenniidae), Kraak & Weissing 1996; sand gobies (Pomatoschistus minutus, Gobiidae), Forsgren et al. 1996; and others). Although the observing female may not have witnessed the laying of these eggs, their presence may act as a cue suggestive of mating success, or prior female choice. Once again, there are other explanations. In the case of the sand goby, females seem to use egg number as an estimate of a male’s ability to defend their nests rather than prior female choice (Lindstrom & Kangas 1996). In sticklebacks, males steal eggs from neighbouring nests in an attempt to bolster their attractiveness (Largiader et al. 2001). However, females might select nests with large numbers of eggs already present for reasons that have nothing to do with mate choice per se; for instance, predator risk dilution (Jamieson 1995). While these other processes can generate mating patterns similar to mate-choice copying, they do not necessarily constitute social learning. Isolating the exact mechanism responsible for such behavioural patterns remains a challenge for future investigation. For further discussion, see Lafleur et al. (1997); Westneat et al. (2000) and Witte (Chapter 5).

11.6

Aggression

Male Siamese fighting fish, Betta splendens (Belontiidae), monitor aggressive interactions between neighbouring conspecifics and use the information on relative fighting ability in subsequent aggressive interactions with the males they have observed (Oliveira et al. 1998). Similar observations have been made in rainbow trout, Oncorhynchus mykiss (Salmonidae) (Johnsson & Akerman 1999). This exploitation of communicated signals in a network has come to be known as ‘eavesdropping’ (McGregor 1993). Essentially, information from a transmitter that is directed at a particular conspecific, the receiver, can be ‘overheard’

BLBK374-11

BLBK374-Brown

250

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

by peripheral individuals or ‘bystanders’. The bystander then uses that information in later interactions with either the signaller or receiver. Oliveira et al.’s findings suggest that the level of aggression that eavesdroppers observe in interactions between a pair of demonstrators strongly affects their subsequent agonistic interactions. Similarly, green swordtails (Xiphophorus helleri, Poeciliidae) were less likely to initiate fights, escalate fights or win against the winners of the contests they had observed (Earley & Dugatkin 2002). Grosnick et al. (2007) report that, following observation of conspecific fights, the cichlid (Astatotilapia burtoni) deploys an impressive capability for transitive inference to compute a linear hierarchy amongst its rivals, which then guides with which fish it interacts. Eavesdropping provides a method whereby individuals can gain information about the social status of others without having to expend energy or risk injury in social contests. Theoretical modelling suggests that the advantages of such a system are most apparent when the potential costs of combat (death or severe injury) are high (Johnstone 2001). For further discussion of the role of prior experience in agonistic interactions, see Chapter 6. Early studies assumed that the signaller and receiver were unaware of the eavesdropper (hence the terminology); indeed, most investigations in the signal-receiver literature tend to focus on the dyad alone, ignoring the social context. However, further investigation in more realistic social surroundings revealed that the signal can be deliberately altered to take the presence of bystanders into account. Male Siamese fighting fishes alter their threat displays depending on the audience because females also eavesdrop on male–male displays. In the presence of female observers, males reduce the number of aggressive components (e.g. bites) in their display and tailor it more towards a sexual display (Doutrelant et al. 2001). Male rainbowfish, Melanotaenia duboulayi (Castelnau), in contrast seem to pay little attention to whether a female is observing male–male contests or not. In the presence of a female observer, jousting males made fewer charges but the number of bites, chases and lateral displays remained the same as when a female was absent (Colleter & Brown 2011). Thus, the focus of research has swiftly switched from dyads to the examination of communication networks. Much of the work on eavesdropping has tended to concentrate on agnostic or competitive displays; however, Johnstone & Bshary (2004) highlight the fact that the theory is equally applicable to a wide range of social contexts where individuals are likely to keep track of the ‘social image’ of conspecifics or heterospecifics. It could, for instance, be equally applicable to altruistic or cooperative encounters like those seen in cleaner–client relationships (Bshary 2002). Despite the obvious correspondence in subject matter, unfortunately eavesdropping has been comparatively neglected within the social learning literature.

11.7 Trade-offs in reliance on social and asocial sources of information The belief that social learning is restricted to, or of particularly importance to, large-brained species of vertebrates is widespread, but is inconsistent with the data presented herein. Many researchers have suggested that social learning abilities may be more strongly associated with ecology than taxonomy (Klopfer 1959; Lefebvre & Palameta 1988), and we endorse

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

251

this argument. Another common assumption is that social learning will be particularly prevalent in highly social animals. This assumption would anticipate that social learning is likely to be restricted to shoaling fishes; however, this again is not the case. For instance, Atlantic salmon parr are known for their highly aggressive nature, yet are capable of social learning (Brown & Laland 2002b). We suggest that consideration of the costs and benefits associated with reliance on social and asocial sources of information is more likely to explain variation in reliance on social learning within and between species than crude inferences based on assumptions of cleverness or sociality. If individuals use social information when personal information is costly, unreliable or likely to be out of date, then there may be differing propensities for social learning in populations for which survival demands vary along these dimensions. Populations at greater risk of predation when they collect personal information will be more likely to use social information than others less at risk. A good example is provided by Coolen et al.’s (2003) study of public information use in sticklebacks. Both three-spined sticklebacks and nine-spined sticklebacks (Pungitius pungitius, Gasterosteidae) can use cues provided by others (public information) to locate foraging patches, but only nine-spined sticklebacks could assess patch quality (Coolen et al. 2003; Fig. 11.3). This difference in the information these two closely related species obtained by observing conspecifics may depend on the costs of relying on asocial learning to obtain information about the environment. Nine-spines lack body armour and have far smaller spines than do three-spines and are more vulnerable to predation while sampling food patches. It is likely that, because of their increased vulnerability, nine-spines rely more heavily on observing the behaviour of others from the safety of cover before deciding on which patch to forage. Whilst this assumption is yet to be empirically tested, it does suggest that the propensity to rely on public information varies considerably between species and perhaps even within species depending on the relative costs and benefits associated with an individual’s motivational state. Van Bergen et al. (2004) revealed that nine-spined sticklebacks switch between reliance on public and private information (information gained by personally sampling the

Proportion of time spent in feeding zone (%)

BLBK374-11

30 Low patch quality High patch quality 20

10

Nine-spined sticklebacks

Three-spined sticklebacks

Fig. 11.3 Nine-spined sticklebacks were able to assess the quality of a foraging patch by observing conspecifics whereas three-spined sticklebacks could not. (After Coolen et al. 2003.)

BLBK374-11

BLBK374-Brown

252

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

environment) depending on its costs. When private information was reliable and recently acquired, sticklebacks ignored public information and based their foraging decision on information gained from their own personal experiences. However, when private information was less reliable or outdated, sticklebacks tended to switch to social learning. Seemingly these fishes prefer to base their decisions on information recently gained from personal experience, but reliance on this information decays over time as it becomes outdated. Sticklebacks switched to reliance on social learning if their own information was older than 7 days. These studies are consistent with the hypothesis that the relative costs and benefits of reliance on social and asocial sources of information will explain a substantial proportion of the variance in social learning across fishes.

11.8

Concluding remarks

In summary, there is now unequivocal evidence, from laboratory and field studies, that a variety of different species of fishes are capable of social learning, including learning how to find food, which foods to eat, to recognise predators and assessing mate and rival quality. When viewed in the context of the burgeoning literature on fish cognition (Bshary et al. 2002; Chapter 13) it is quite apparent that the abilities and complexity of social behaviour of this group have previously been seriously underestimated. We end by noting that the widespread use of social learning by fishes may have important implications for conservation and fisheries reintroductions (Suboski & Templeton 1989; Brown & Laland 2001; Brown & Day 2002). Typically, over 95% of all fishes released from hatcheries die from predation or starvation in the first few weeks following release (Brown & Laland 2001) – an enormous waste of resources. It is conceivable that hatcheryreared fishes could be trained en masse to recognise predators and prey using social learning protocols. The evidence presented in the chapter suggests that it may be possible to cut post-release mortality figures dramatically by allowing hatchery fishes to learn from more experienced or wild conspecifics. The manners in which the cognitive abilities of fishes can be exploited are discussed at length by Fern¨o et al. in Chapter 16.

Acknowledgements We are grateful to the Biotechnology and Biological Sciences Research Council and Royal Society for financial support and to Stephen Reebs and Mike Webster for helpful comments on the manuscript. Culum Brown was supported by an Australian Research Fellowship from the Australian Research Council.

References Anthouard, M. (1987) A study of social transmission in juvenile Dicentrarchus labrax (pisces, serranidae), in an operant-conditioning situation. Behaviour, 103, 266–275. Barnard, C. & Sibly, R. (1981) Producers and scroungers: a general model and its application to feeding flocks of house sparrows. Animal Behaviour, 29, 543–550.

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

253

Bisazza, A. & Marconato, A. (1988) Female mate choice, male–male competition and parental care in the river bullhead, Cotus gobio. Animal Behaviour, 36, 1352–1360. Boyd, R. & Richerson, P.J. (1985) Culture and the Evolutionary Process. Chicago University Press, Chicago. Brooks, R. (1996) Copying and the repeatability of mate choice. Behavioral Ecology and Sociobiology, 39, 323–329. Brooks, R. (1999) Mate choice copying in guppies: females avoid the place where they saw courtship. Behaviour, 136, 411–421. Brown, C. & Day, R. (2002) The future of stock enhancements: bridging the gap between hatchery practice and conservation biology. Fish and Fisheries, 3, 79–94. Brown, C. & Laland, K.N. (2001) Social learning and life skills training for hatchery reared fish. Journal of Fish Biology, 59, 471–493. Brown, C. & Laland, K.N. (2002a) Social learning of a novel avoidance task in the guppy, P. reticulata: conformity and social release. Animal Behaviour, 64, 41–47. Brown, C. & Laland, K.N. (2002b) Social enhancement and social inhibition of foraging behaviour in hatchery-reared Atlantic salmon. Journal of Fish Biology, 61, 987–998. Brown, G.E. (2003) Learning about danger: chemical alarm cues and local risk assessment in prey fishes. Fish and Fisheries, 4, 227–234. Bshary, R. (2002) Biting cleaner fish use altruism to deceive image-scoring client reef fish. Proceedings of the Royal Society of London Series B – Biological Sciences, 269, 2087–2093. Bshary, R., Wickler, W. & Fricke, H. (2002) Fish cognition: a primate’s eye view. Animal Cognition, 5, 1–13. Chivers, D.P. & Smith, J.F. (1995) Chemical recognition of risky habitats is culturally transmitted among fathead minnows, Pimephales promelas (Osteichthyes, Cyprinidae). Ethology, 99, 286–296. Colleter, M. & Brown, C. (2011) Personality traits predict hierarchy rank in male social groups. Animal Behavior (In press.) Coolen, I., van Bergen, Y., Day, R.L. & Laland, K.N. (2003) Species difference in adaptive use of public information in sticklebacks. Proceedings of the Royal Society of London Series B – Biological Sciences, 270, 2413–2419. Couzin, I.D., Krause, J., Franks, N.R. & Levin, S.A. (2005) Effective leadership and decision-making in animal groups on the move. Nature, 433, 513–516. Day, R., MacDonald, T., Brown, C., Laland, K. & Reader, S.M. (2001) Interactions between shoal size and conformity in guppy social foraging. Animal Behaviour, 62, 917–925. Doutrelant, C., McGregor, P.K. & Oliveira, R.F. (2001) The effect of an audience on intrasexual communication in male Siamese fighting fish, Betta splendens. Behavioral Ecology, 12, 283–286. Dugatkin, L.A. (1992) Sexual selection and imitation: females copy the mate choice of others. American Naturalist, 139, 1384–1489. Dugatkin, L.A. & Godin, J-G.J. (1992) Reversal of mate choice by copying in the guppy (Poecilia reticulata). Proceedings of the Royal Society of London Series B – Biological Sciences, 249, 179–184. Earley, R.L. & Dugatkin, L.A. (2002) Eavesdropping on visual cues in green swordtail (Xiphophorus helleri) fights: a case for networking. Proceedings of the Royal Society of London Series B – Biological Sciences, 269, 943–952. Forsgren, E., Karlsson, A. & Kvarnemo, C. (1996) Female sand gobies gain direct benefits by choosing males with eggs in their nests. Behavioral Ecology and Sociobiology, 39, 91–96. Galef, B.G. Jr. & Giraldeau, L.-A. (2001) Social influences on foraging in vertebrates: causal mechanisms and adaptive functions. Animal Behaviour, 61, 3–15. Gibson, R.M. & Hoglund, J. (1992) Copying and sexual selection. Trends in Ecology and Evolution, 7, 229–232. Goldsmidt, T., Bakker, T.C.M. & Feuth-de Bruijn, E. (1993) Selective choice in copying of female sticklebacks. Animal Behaviour, 45, 541–547. Grant, J.W. & Green, L.D. (1996) Mate copying versus preferences for actively courting males by female Japanese medaka (Oryzias latipes). Behavioral Ecology, 7, 165–167.

BLBK374-11

BLBK374-Brown

254

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Griffin, A.S. (2004) Social learning about predators: a review and prospectus. Learning & Behavior, 32, 131–140. Grosnick, L., Clement, T.S. & Fernald, R.D. (2007) Fish can infer social rank by observation alone. Nature, 445, 429–432. Helfman, G.S., Meyer, J.L. & McFarland, W.N. (1982) The ontogeny of twilight migration patterns in grunts (Pisces: Haemulidae). Animal Behaviour, 30, 379–384. Helfman, G.S. & Schultz, E.T. (1984) Social tradition of behavioural traditions in a coral reef fish. Animal Behaviour, 32, 379–384. Heyes, C.M. (1994) Social learning in animals: categories and mechanisms. Biological Reviews, 69, 207–231. Heyes, C.M. & Galef, B.G. (1996) Social Learning in Animals: The Roots of Culture. Academic Press, London. Hoppitt, W. & Laland, K.N. (2008) Social processes influencing learning in animals: a review of the evidence. Advances in the Study of Behavior, 38, 105–165. Jamieson, I. (1995) Do female fish prefer to spawn in nests with eggs for reasons of mate choice copying or egg survival. American Naturalist, 145, 824–832. Johnsson, J.I. & Akerman, A. (1999) Watch and learn: preview of the fighting ability of opponents alters contest behaviour in rainbow trout. Animal Behaviour, 56, 771–776. Johnstone, R.A. (2001) Eavesdropping and animal conflict. Proceedings of the National Academy of Science, 98, 9177–9180. Johnstone, R.A. & Bshary, R. (2004) Evolution of spite through indirect reciprocity. Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 1917–1922. Kelley, J.L., Evans, J.P., Ramnarine, I.W. & Magurran, A.E. (2003) Back to school: can antipredator behaviour in guppies be enhanced through social learning? Animal Behaviour, 65, 655–662. Kelley, J.L. & Magurran, A.E. (2003) Learning of predator recognition and anti-predator responses in fishes. Fish and Fisheries, 4, 216–226. Kendal, R.L., Coolen, I., van Bergen, Y. & Laland, K.N. (2005) Tradeoffs in the adaptive use of social and asocial learning. Advances in the Study of Behaviour, 35, 333–379. Kieffer, J.D. & Colgan, P.W. (1992) The role of learning in fish behaviour. Reviews in Fish Biology and Fisheries, 2, 125–143. Klopfer, P.H. (1959) Social interactions in discrimination learning with special reference to feeding behaviour in birds. Behaviour, 14, 282–299. Knapp, R.A. & Sergeant, R.C. (1989) Egg mimicry as a mating strategy in the faintail darter, Etheostoma flabellare: females prefer males with eggs. Behavioral Ecology and Sociobiology, 25, 321–326. Kraak, S.B.M. & Weissing, F.J. (1996) Female preference for nests with many eggs: a cost-benefit analysis of female choice in fish with paternal care. Behavioral Ecology, 7, 353–361. Krause, J. (1993) Transmission of fright reaction between different species of fish. Behaviour, 127, 37–48. Krause, J., Reeves, P. & Hoare, D. (1998) Positioning behaviour in roach shoals: the role of body length and nutritional state. Behaviour, 135, 1031–1039. Lachlan, R.F., Crooks, L. & Laland, K.N. (1998) Who follows whom? Shoaling preferences and social learning of foraging information in guppies. Animal Behaviour, 56, 181–190. Lafleur, D.L., Lozano, G.A. & Sclafani, M. (1997) Female mate-choice copying in guppies, Poecilia reticulata: a re-evaluation. Animal Behaviour, 54, 579–586. Laland, K.N. (2004) Social learning strategies. Learning and Behavior, 32, 4–14. Laland, K.N. & Williams, K. (1997) Shoaling generates social learning of foraging information in guppies. Animal Behaviour, 53, 1161–1169. Laland, K.N. & Williams, K. (1998) Social transmission of maladaptive information in the guppy. Behavioral Ecology, 9, 493–499. Largiader, C.R., Fries, V. & Bakker, T.C.M. (2001) Genetic analysis of sneaking and egg-thievery in a natural population of the three-spined stickleback (Gasterosteus aculeatus L.). Heredity, 48, 459–468.

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

255

Leadbeater, E. & Chittka, L. (2007) The dynamics of social learning in an insect model, the bumblebee (Bombus terrestris). Behavioral Ecology and Sociobiology, 61, 1789–1796. Lefebvre, L. & Palameta, B. (1988) Mechanisms, ecology and population diffusion of sociallylearned, food-finding behavior in feral pigeons. In: T.R. Zentall & B.G. Galef (eds) Social Learning: Psychological and Biological Perspectives, pp. 141–164. Lawrence Erlbaum Associates, New Jersey. Lindeyer, C.M. & Reader, S.M. (2010) Social learning of escape routes in zebrafish and the stability of behavioural traditions. Animal Behaviour, 79, 827–834. Lindstrom, K. & Kangas, N. (1996) Egg presence, egg loss, and female mate preferences in the sand goby (Pomatoschistus minutus). Behavioral Ecology, 7, 213–217. Magurran, A.E. & Higham, A. (1988) Information transfer across fish shoals under predator threat. Ethology, 78, 153–158. Mathis, A., Chivers, D.P. & Smith, R.J.F. (1996) Cultural transmission of predator recognition in fishes, intraspecific and interspecific learning. Animal Behaviour, 51, 185–201. Mazeroll, A.I. & Montgomery, W.L. (1995) Structure and organization of local migrations in brown surgeonfish (Acanthurus nigrofuscus). Ethology, 99, 89–106. McGregor, P.K. (1993) Signaling in territorial systems – a context for individual identification, ranging and eavesdropping. Philosophical Transactions of the Royal Society of London Series B – Biological Sciences, 340, 237–244. Milinski, M. (2003) The function of mate choice in sticklebacks: optimizing MHC genetics. Journal of Fish Biology, 63, 1–16. Milinski, M., Griffiths, S., Wegner, K.M., Reusch, T.B.H., Haas-Assenbaum, A. & Boehm, T. (2005) Mate choice decisions of stickleback females predictably modified by MHC peptide ligands. Proceedings of the National Academy of Sciences of the United States of America, 102, 4414–4418. Mineka, S. & Cook, M. (1988) Social learning and the acquisition of snake fear in monkeys. In: T.R. Zentall & B.G. Galef (eds) Social Learning:Psychological and Biological Perspectives, pp. 51–74. Lawrence Erlbaum Associates, New Jersey. Morgan, M.J. (1988) The influence of hunger, shoal size and predator presence on foraging in bluntnose minnows. Animal Behaviour, 36, 1317–1322. Morgan, M.J. & Colgan, P.W. (1987) The effects of predator presence and shoal size on bluntnose minnows, Pimephales notatus. Environmental Biology of Fishes, 20, 105–111. Odling-Smee, L. & Braithwaite, V.A. (2003) The role of learning in fish orientation. Fish and Fisheries, 4, 235–246. Oliveira, R.F., McGregor, P.K. & Latruffe, C. (1998) Know thine enemy: fighting fish gather information from observing conspecific interactions. Proceedings of the Royal Society of London Series B – Biological Sciences, 265, 1045–1049. Olson, D.E., Schupp, D.H. & Macins, V. (1978) A hypothesis of homing behaviour of walleyes as related to observed patterns of passive and active movement. American Fisheries Society Special Publication, 11, 52–57. Peuhkuri, N., Ranta, E., Juvonen, S.K. & Lindstrom, K. (1995) Schooling affects growth in the 3-spined stickleback, Gasterosteus aculeatus. Journal of Fish Biology, 46, 221–226. Pfeiffer, W. (1974) Pheromones in fish and amphibia. In: M.C. Birch (ed) Pheromones, pp. 269–296. North-Holland, Amsterdam. Pitcher, T.J., Green, D.A. & Magurran, A.E. (1986) Dicing with death: predator inspection behaviour in minnow shoals. Journal of Fish Biology, 28, 439–448. Pitcher, T.J. & House, A. (1987) Foraging rules for group feeders: area copying depends upon density in shoaling goldfish. Ethology, 76, 161–167. Pitcher, T.J., Magurran, A.E. & Winfield, I.J. (1982) Fish in larger shoals find food faster. Behavioral Ecology Sociobiology, 10, 149–151. Plath, M., Richter, S., Tiedemann, R. & Schlupp, I. (2008) Male fish deceive competitors about mating preferences. Current Biology, 18, 1–4. Potts, W.K. (1984) The chorus-line hypothesis of manoeuvre coordination in avian flocks. Nature, 309, 344–345.

BLBK374-11

BLBK374-Brown

256

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Rands, S.A., Cowlishaw, G., Pettifor, R.A., Rowcliffe, J.M. & Johnstone, R.A. (2003) Spontaneous emergence of leaders and followers in foraging pairs. Nature, 423, 432–434. Rands, S.A., Pettifor, R.A., Rowcliffe, J.M. & Cowlishaw, G. (2004) State-dependent foraging rules for social animals in selfish herds. Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 2613–2620 Reader, S.M., Kendal, J.R. & Laland, K.N. (2003) Social learning of foraging sites and escape routes in wild Trinidadian guppies. Animal Behaviour, 66, 729–739. Reebs, S.G. (2000) Can a minority of informed leaders determine the foraging movements of a fish shoal? Animal Behaviour, 59, 403–409. Reebs, S.G. (2001) Influence of body size on leadership in shoals of golden shiners, Notemigonus crysoleucas. Behaviour, 138, 797–809. Reynolds, J.D. & Jones, J.C. (1999) Female preference for preferred males is reversed under low oxygen conditions in the common goby (Pomatoschistus microps). Behavioral Ecology, 10, 149–154. Ryer, C.H. & Olla, B.L. (1991) Information transfer and the facilitation and inhibition of feeding in a schooling fish. Environmental Biology of Fishes, 30, 317–323. Ryer, C.H. & Olla, B.L. (1992) Social mechanisms facilitating exploitation of spatially variable ephemeral food patches in a pelagic marine fish. Animal Behaviour, 44, 69–74. Schlupp, I. & Ryan, M.J. (1997) Male sailfin mollies (Poecilia latipinna) copy the mate choice of other males. Behavioral Ecology, 8, 104–107. Schuster, S., Wohl, S., Griebsch, M. & Klostermeier, I. (2006) Animal cognition: how archer fish learn to down rapidly moving targets. Current Biology, 16, 378–383. Shettleworth, S.J. (2001) Animal cognition and animal behaviour. Animal Behaviour, 61, 277–286. Shima, T., Yamamoto, T., Furuita, H. & Sumiki, N. (2003) Effect of the response interval of self-feeders on the self-regulation of feed demand by rainbow trout (Oncorhynchus mykiss) fry. Aquaculture, 224, 181–191. Stephens, D.W. & Krebs, J.R. (1986) Foraging Theory. Princeton University Press, Princeton, NJ. Suboski, M.D., Bain, S., Carty, A.E., McQuoid, L.M., Seelen, M.I. & Seifert, M. (1990) Alarm reaction in acquisition and social transmission of simulated predator recognition by zebra danio fish (Brachydanio rerio). Journal Comparative Psychology, 104, 101–112. Suboski, M.D. & Templeton, J.J. (1989) Life skills training for hatchery fish: social learning and survival. Fisheries Research, 7, 343–352. Sugita, Y. (1980) Imitative choice behaviour in guppies. Japan Psychological Research, 22, 7–12. Sumpter, D.J.T., Krause, J., James, R., Couzin, I.D. & Ward, A.J.W. (2008) Consensus decision making by fish. Current Biology, 18, 1773–1777. Sundstrom, L.F. & Johnsson, J.I. (2001) Experience and social environment influence the ability of young brown trout to forage on live novel prey. Animal Behaviour, 61, 249–255. Templeton, J. (1987) Individual Differences in the Behaviour of Juvenile Rock Bass (Ambloplites rupestris): Causes and Consequences. Master’s thesis, Queen’s University, Ontario. Tsukamoto, K., Aoyama, J. & Miller, M.J. (2003) Migration, speciation, and the evolution of diadromy in anguillid eels. Canadian Journal of Fisheries and Aquatic Science, 59, 1989–1998. Unger, L.M. & Sergent, R.C. (1988) Alloparental care in the fathead minnow, Pimehales promales: females prefer males with eggs. Behavioral Ecology and Sociobiology, 23, 27–32. van Bergen, Y., Coolen, I. & Laland, K.N. (2004) Nine-spined sticklebacks exploit the most reliable source when public and private information conflict. Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 957–962. Verheijen, F.J. (1956) Transmission of a flight reaction amongst a school of fish and the underlying sensory mechanisms. Experientia, 12, 202–204. Vilhunen, S. (2006) Repeated anti-predator conditioning: a pathway to habituation or to better avoidance? Journal of Fish Biology, 68, 25–43. Vilhunen, S., Hirvonen, H. & Laakkonen, M.V.M. (2005) Less is more: social learning of predator recognition requires a low demonstrator to observer ratio in Arctic charr (Salvelinus alpinus). Behavioral Ecology and Sociobiology, 57, 275–282. Warburton, K. (2003) Learning of foraging skills by fish. Fish and Fisheries, 4, 203–215.

BLBK374-11

BLBK374-Brown

May 13, 2011

17:14

Copyeditor’s Name:

Trim: 244mm X 172mm

Social Learning in Fishes

Char Count=

257

Warner, R.R. (1988) Traditionality of mating-site preferences in a coral reef fish. Nature, 335, 719–721. Warner, R.R. (1990) Male versus female influences on mating-site determination in a coral-reef fish. Animal Behaviour, 39, 540–548. Webb, P.W. (1980) Does schooling reduce fast-start response latencies in teleosts? Comparative Biochemistry and Physiology Part A: Comparative Physiology, 65, 231–234. Webster, M.M. & Laland, K.N. (2008) Minnows copy only when asocial learning would be costly. Proceedings of the Royal Society of London Series B – Biological Sciences, 275, 2869–2876. Westneat, D.F., Walters, A., McCarthy, T.M., Hatch, M.I. & Hein, W.K. (2000) Alternative mechanisms of nonindependent mate choice. Animal Behaviour, 59, 467–476. Whiten, A. & Ham, R. (1992) On the nature and evolution of imitation in the animal kingdom: reappraisal of a century of research. Advances in the Study of Behaviour, 21, 239–283. Witte, D.J. & Ryan, M.J. (2002) Mate-choice copying in the sailfin molly, Poecilia latipinna, in the wild. Animal Behaviour, 63, 943–949.

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 12

Cooperation and Cognition in Fishes Michael S. Alfieri and Lee A. Dugatkin

12.1

Introduction

‘The theory of evolution is based on the struggle for life and the survival of the fittest. Yet cooperation is common between members of the same species and even between members of different species’ (Axelrod & Hamilton 1981, p. 1390). In this simple, but powerful quote, Robert Axelrod (a professor of political science) and William Hamilton (a professor of evolutionary biology) illustrate a principal difficulty in understanding cooperation in light of evolutionary theory. Why should any organism help another at risk to oneself if there is no apparent benefit in doing so? This question has perplexed evolutionary biologists since the inception of the field. Indeed, Charles Darwin initially struggled to explain how sterile insects, individuals that sacrifice reproduction to contribute to the production of the hive or colony without gaining obvious benefits from their cooperative behaviours, could fit into his theory of natural selection (Darwin 1859). A key question regarding cooperative behaviours and natural selection faced Darwin, namely how could natural selection favour cooperation if the cooperation does not increase the fitness of those that express this trait? Darwin posited one possible solution a few paragraphs later when he noted that ‘this difficulty, though appearing insuperable, is lessened, or, as I believe, disappears, when it is remembered that selection may be applied to the family, as well as to the individual, and may thus gain the desired end’ (Darwin 1859, p. 237). Here Darwin described one of the main paths by which cooperation can arise and be maintained in a population, later described and formalised as kin selection or inclusive fitness by Hamilton (1964a, 1964b). In this chapter, we briefly describe kin selection, plus three other categories of cooperation, suggest the necessary cognitive prerequisites for cooperation to occur, and provide empirical examples that illustrate each category of cooperation in fishes. There are four commonly recognised categories invoked to explain the origin and maintenance of cooperation (Table 12.1), namely (1) kin selection (Hamilton 1964a, 1964b), (2) reciprocity (Trivers 1971), (3) by-product mutualism (West Eberhard 1975; Brown 1983), and (4) trait group selection (Wilson 1980) (categories reviewed in Dugatkin et al. 1992; Dugatkin 1997). These categories are not mutually exclusive and more than one category

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

Char Count=

259

may accurately describe cooperative behaviours within or among species. This is evident by our example of predator inspection that appears twice, in two different categories of cooperation in this chapter. Because cooperation is often envisioned as a cognitively complex trait, the cognitive abilities that are needed for cooperation to persist in populations have recently been examined (Dugatkin 1997; Gadagkar 1997; Bshary 2002a; Dugatkin & Alfieri 2002; Sachs et al. 2004) and at least some of the controversy surrounding the study of cooperative behaviour has focused on these abilities (Dugatkin 1997; Clutton-Brock 2002; Stevens and Hauser 2004). Disagreement about the cognitive abilities of fishes in particular may have begun in some of the earlier works on animal social behaviour. In his pioneering book, Sociobiology: The New Synthesis, E.O. Wilson addresses the evolutionary mechanisms behind several social behaviours in the animal kingdom including cooperation (Wilson 1975b). Wilson suggests that the ‘lack of intelligence’ (among other possibilities) in lower vertebrates including fishes is a reason why he did not see specific cooperative behaviours in this group. In the 35 years since the publication of this book, the evidence is now quite clear that fishes possess impressive cognitive abilities worthy of study in the field of cooperative behaviours (Dugatkin 1997; Bshary 2002a; Dugatkin & Alfieri 2002; Brown & Laland 2003; Griffiths 2003). However, as evident by the ongoing debates (Riolo et al. 2001; Hammerstein 2002; Stevens & Hauser 2004; Trivers 2004; Pfeiffer et al. 2005), more directed empirical studies focusing on the specific cognitive prerequisites necessary for cooperation are needed.

12.2

Why study cooperation in fishes?

The study of cooperative behaviours can benefit greatly from research focusing on fishes. Fishes display enormous diversity in morphology, behaviour, physiology, life histories, and can be found in almost all aquatic environments (Godin 1997). Among vertebrates, teleost fishes are the most abundant and diverse group representing over 50% of all known vertebrate species (Diana 2004). Half of all fish species spend at least part of their lives in groups (Shaw 1978) allowing for at least the potential for cooperative interactions to evolve under many diverse conditions (although social groups are not a necessary prerequisite for cooperative behaviour as seen in the example between cleaner and client fishes in Subsection 12.3.2.4). Examples of cooperation in the animal kingdom are widespread and many of these examples include fishes (Dugatkin 1997). Recently, fishes have been examined as a model system for the study of cognition (Bshary 2002a; Laland & Hoppitt 2003), as well as models for the study of cooperation and the role of cognition in cooperative behaviours (Dugatkin & Mesterton-Gibbons 1996; Dugatkin 1997; Tebbich et al. 2002). However, even with our ever-improving understanding of fish systems, the need for rigorous, directed, empirical studies examining the cognitive abilities necessary for cooperation in fishes is still great. In this chapter, we present selected examples of cooperation from the fish literature that highlight the categories of cooperation and the potential cognitive abilities needed for each (Table 12.1). Each example may be the result of a single or multiple categories of cooperation; however, we use this framework to illustrate the diverse cognitive abilities and varied methods by which cooperation can be found in fishes.

Individuals act together to achieve a beneficial outcome that could not have been achieved as efficiently by any single individual

Within-group costs to cooperator are Categorical recognition of others less than between-group benefits (e.g., cooperator vs. noncooperator)

Trait Group Selection Wilson (1975a, 1980)

Dugatkin and Mesterton-Gibbons (1996)

Copyeditor’s Name:

Predator inspection G. aculeatus and P. reticulata

Foster (1985, 1987)

Trivers (1971), Tebbich et al. (2002)

Cleaning between Labroides dimidiatus and their clients Epinephelus striatus Cooperative foraging Acanthurus coeruleus and Thalassoma lucasanum

Milinski (1987) and Dugatkin (1988)

Predator inspection Gasterosteus aculeatus and Poecilia reticulata

11:2

By-product Mutualism West Eberhard (1975), Brown (1983)

Fischer (1980, 1988)

Egg trading in Hypoplectrus nigricans

May 27, 2011

Categorical recognition of environment (e.g., harsh vs. mild)

A costly act of cooperation is repaid Memory of past events and to the cooperator by the recipient at a individual recognition future time

Reciprocity Trivers (1971)

Territorial defence in Salmo salar and Oncorhynchus mykiss

Cooperative breeding in Lamprologus brichardi

Categorical recognition of kin

An act of cooperation is directed towards kin with costs to the cooperator

Kin selection Hamilton (1964a, 1964b)

Taborsky (1984) (1985) Brown & Brown (1993)

Representative examples and species References of examples

Possible cognitive prerequisites

Definition of cooperative category

BLBK374-Brown

Category of cooperation

260

Table 12.1 Categories and cognition prerequisites of selected cooperative behaviours in fishes.

BLBK374-12 Trim: 244mm X 172mm

Fish Cognition and Behavior

Char Count=

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

12.3

Char Count=

261

Cooperation and its categories

The phrase ‘to cooperate’ has many meanings. It may, for example, imply to achieve cooperation; a result that is realised at the group level. Alternatively, to cooperate may refer to the actions of the individual. That is, to behave in a manner that makes cooperation possible, although cooperation may not be realised unless other members of a group also act cooperatively. For example, imagine two individuals trapped in a cave whose exit is blocked by a large boulder. We may speak of cooperation as the outcome of the two individuals working together to move the boulder, or may define cooperation by the act of helping to move the boulder (something an individual does). In our review of cooperation, we refer to the latter, the behavioural act that allows for cooperation. That is, cooperation results when two or more individuals behave in a coordinated manner and the outcome is that participants gain some type of benefit (Dugatkin 1997). We define ‘benefit’ as a positive contribution towards one’s fitness (see Wilson & Dugatkin 1992 for a discussion of relevant definitions). The benefit may be direct (e.g. gaining a meal, a valuable piece of information, or protection from predators), or indirect, (e.g. improving the survival of kin or increasing the possibility of future benefits). Recently, authors have defined cooperation in a similar manner (Sachs et al. 2004; Stevens & Hauser 2004).

12.3.1

Category 1 – kin selection

Theoretically, kin selection is perhaps the most intuitive category of cooperation. To see why, we need to take a ‘gene’s eye’ perspective on cooperation. By definition, blood kin share many genes that are identical by descent, i.e. they are derived from some common ancestor. A gene that codes for helping one’s blood kin is increasing the probability that copies of itself that reside in blood kin are transmitted to the next generation (Hamilton 1964a, 1964b). ‘Hamilton’s rule’ posits that cooperation will be selected for in a population when rb – c>0, where ‘b’ is the benefits to the recipient of a cooperator’s action, ‘r’ is the coefficient of relatedness (i.e. the proportion of genes shared between two individuals from a common ancestor), and ‘c’ is the cost to cooperator for cooperating. Phrased in the cold language of natural selection, relatives are worth helping in direct proportion to their genetic (blood) relatedness.

12.3.1.1

Cognition and kin selection

When an individual’s blood kin are scattered throughout a given environment, kin recognition allows the benefits of cooperation to be differentially allocated to such blood kin. Such recognition may be based on some behavioural attribute or on so-called ‘recognition genes’ that may allow identification of kin based on odour, morphological marks, etc. A large body of research has demonstrated that fishes are able to recognise others (including kin) as well as alter behaviours based on this recognition (reviewed by Krause et al. 2000; Griffiths 2003; Ward & Hart 2003). In their recent approach to studying cooperation, Sachs et al. (2004) refer to kin-selected cooperation based on recognition of kin, learned or heritable, as ‘kin choice’. However, individuals need not be able to recognise others as

BLBK374-12

BLBK374-Brown

262

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

kin for cooperation via kin selection to occur. When an individual is always (or almost always) surrounded by close relatives, then strategies to treat any encounter as an encounter with kin may evolve. These cases of kin-selected cooperation have been referred to as ‘kin fidelity’ (Sachs et al. 2004). In such cases, cooperation via kin selection is possible without the cognitive abilities to recognise an individual as kin. However, this ‘rule of thumb’ – i.e. treat others as kin – is vulnerable to invasion as non-kin may receive the cooperative benefits reserved for kin. 12.3.1.2

Example of kin selected cooperation: Cooperative breeding

In cooperative breeding societies, groups are composed of breeding individuals and helpers that either delay their own breeding or completely forego reproduction to assist breeding pairs (Brown 1987). Although the literature on cooperative breeding is dominated by examples involving birds, insects, and mammals, a few species of fishes have been studied in light of cooperative breeding. Michael Taborsky’s research (1984, 1985) on the cooperative breeding cichlid Lamprologus brichardi provides a case example of kin-selected cooperation in fishes. In L. brichardi, sexually mature offspring stay at the nest and help maintain and defend eggs despite the fact that they pay a cost of staying and helping. This cost may be realised through reduced growth rates or lost breeding opportunities. After several other possible hypotheses were tested as to why helpers stay and help at the nest (including the potential benefits of gaining experiences for future success in raising their own young, safety in the natal territory, possibility of taking over the parental territory, improved diet due to cannibalism), kin-selected benefits were determined to be the best explanation for L. brichardi behaviour (reviewed in Dugatkin 1997). Recently, Griffin & West (2003) used meta-analysis to determine the relative importance of kin selection among 18 cooperatively breeding vertebrates (birds and mammals). They found a significant pattern of kin recognition and preferential treatment of closely related kin among helpers. Although their study did not include fish, it highlights the importance of kin recognition and kin selection in other cooperatively breeding vertebrates (for alternative explanations, see Clutton-Brock 2002). Recent research on cooperative breeding in fishes has focused on the cichlid, Neolamprologus pulcher (Werner et al. 2003; Bergmuller et al. 2005; Bergmuller & Taborsky 2005; Brouwer et al. 2005). In this species smaller (younger) helper fishes are more closely related to both the breeding pair and their brood than are larger (older) helpers (Stiver et al. 2004; Bergmuller et al. 2005; Bergmuller & Taborsky 2005; Brouwer et al. 2005). As such, kin selection predicts that cooperation should be more common in smaller helpers than larger helpers. Brouwer et al. (2005) present compelling experimental evidence supporting kin-selected cooperation in smaller fishes (see Bergmuller & Taborsky 2005; Brouwer et al. 2005 for explanations of cooperation among larger fishes). 12.3.1.3

Example of kin selected cooperation: Conditional territory defence

Another cooperative interaction in fishes that is likely based on kin selection is conditional territory defence behaviours. A territory should be defended if the benefits of owning an area (e.g. food, shelter) are greater than the costs of defending it (e.g. energy expenditure,

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

Char Count=

263

potential injury) (Brown 1964; see Grant 1997 for costs and benefits of territoriality in fishes). Brown & Brown (1993) found that both Atlantic salmon, Salmo salar, and rainbow trout, Oncorhynchus mykiss, two species in which kin recognition has been documented (Brown & Brown 1992), displayed reduced aggressive behaviours, increased tolerance and reduction in territory size when their territorial neighbour holder was kin compared to nonkin. In this example of kin-selected cooperation, both indirect fitness benefits (the improved survival of a close relative) and direct fitness benefits (reduction in energy and potential harm occurred in aggressive defence) are achieved. Similar kin-selection benefits including territory sharing and latency to acquire food patches were found in Atlantic salmon in a study by Griffiths & Armstrong (2002).

12.3.2

Category 2 – reciprocity

A second category of cooperation is reciprocity, also called direct reciprocity, directed reciprocation or reciprocal altruism (Dugatkin 1997; Sachs et al. 2004). In his 1971 work, Robert Trivers describes a model by which cooperative behaviours can persist in a population both in the absence of, or in conjunction with, kin selection. During reciprocity, an act of cooperation is repaid to the cooperator by the recipient at a future time. Cooperation via reciprocity is vulnerable to ‘cheating’, i.e. after receiving the beneficial act from a cooperator, a recipient receives a higher pay-off from simply not returning the favour. Trivers addresses the ‘cheater problem’ by discussing cooperation in light of a game theory model, the Prisoner’s Dilemma (Fig. 12.1; Luce & Raiffa 1957; Rapoport & Chammah 1965, cited in Trivers 1971). During the game, two individuals are faced with the choice to either cooperate or defect (i.e. not cooperate). During a single encounter, each player receives a greater pay-off if they defect. To see why, consider the pay-off matrix (Fig. 12.1). If player 2 cooperates, player 1 receives the greatest pay-off if it defects (‘T’ the temptation Player 2 Cooperate

Defect

Cooperate

R=3

S=0

Defect

T=5

P=1

Player 1

Fig. 12.1 The Prisoner’s Dilemma game. Each cell represents the pay-off to player 1 given its interaction with player 2. The game is constructed so that the Temptation to cheat (T) > Reward for mutual cooperation (R) > the Punishment for mutual defection (P) > Sucker’s pay-off for cooperating when your opponent defects (S).

BLBK374-12

BLBK374-Brown

264

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

to cheat) than if it also cooperates (‘R’ the reward for mutual cooperation). By defecting, player 1 receives the benefit of a cooperative act towards itself but does not pay any of the costs associated with cooperating. Alternatively, if player 2 defects and player 1 cooperates, player 1 pays all the costs of cooperation and receives none of the benefits. In this scenario, player 1 receives the lowest possible pay-off (‘S’ the sucker’s pay-off). Thus, if player 2 defects player 1 should also defect and both players receive (‘P’) the punishment for mutual defection. While playing the Prisoner’s Dilemma game, it appears that the strategy to defect is the best strategy for each player – so where is the ‘dilemma’? The dilemma exists in the fact that if both players cooperate they each receive a reward for mutual cooperation (‘R’) that is greater than the punishment for mutual defection (‘P’). To achieve cooperation, Trivers (1971) suggests that the game is not played only once but iterated games must be considered between the same two individuals so that each has an opportunity to respond to the others to ‘mimic real life’. In 1981, Axelrod and Hamilton used the iterated Prisoner’s Dilemma (iPD) to examine numerous strategies in a computer tournament in which they invited experts to submit a set of behavioural rules that would interact with other sets of rules to determine how cooperation can arise and be maintained in a population. In two separate tournaments, the strategy that outcompeted all others and allowed for cooperation to thrive in a population (if the probability of meeting a given partner was above a critical threshold) was Tit-for-Tat (TFT), submitted by Anatol Rapport. TFT is a simple set of rules that directs a player to cooperate on one’s first move, and subsequently copy an opponent’s last move (Axelrod & Hamilton 1981; Axelrod 1984). The iPD game continues to be used as a focal model to study reciprocity (Axelrod 1984; Dugatkin 1997; Dugatkin & Reeve 1998). 12.3.2.1

Cognition and reciprocity

In Section 12.3.2, we described reciprocity and how the Prisoner’s Dilemma has been used to examine the conflict between cooperation and the temptation to defect (i.e. not cooperate). To play TFT (which includes copying your partner’s last move) an individual must be able to recognise who they are paired with, as well as how that partner acted during their last encounter. As such, the cognitive requirements associated with playing TFT are the ability to recognise partners and remember the outcome of previous encounters. However, these are not strict prerequisites as individual recognition would not be required when individuals interact with only one partner for a longer time (Axelrod & Hamilton 1981). The consistency between partners may be due to spatial constraints during interactions (e.g. individuals are not very mobile), or a lack of alternative partners in a population. Recent authors have further stressed the importance of memory and learning when individuals are playing the iPD game. For example, Milinski & Wedekind (1998) showed that in humans, constraints on memory can affect strategies used when playing the iPD game. Additionally, Gutnisky & Zanutto (2004) present a model highlighting scenarios during which operant learning can benefit individuals during the iPD game. The study of cooperation via reciprocity has recently been explored in the wild. These works suggest that in some natural populations of freshwater fish species, the conditions necessary for cooperation via reciprocity are found (Ward et al. 2002; Croft et al. 2004; Croft et al. 2005, 2006). For example, Ward et al. (2002) and Croft et al. (2004, 2005, 2006)

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

Char Count=

265

have demonstrated that sticklebacks and guppies respectively form stable social affiliations, an important condition for cooperation via reciprocity (Dugatkin 1997). 12.3.2.2

Example of reciprocity: Egg trading

Teleost fishes represent the only known vertebrates that are capable of simultaneous hermaphroditism, i.e. the possession by a single individual of both eggs and sperm at the same time. While this is most prevalent in deep-sea fishes (Smith 1975), it also occurs in the Serranidae, a shallow-water family of fishes that includes the subfamily Serraninae, the seabasses (Fischer & Petersen 1987; Fischer 1988). Reciprocity via egg trading has been well studied in black hamlet fish, Hypoplectrus nigricans (Fischer 1980, 1981, 1987), zebra goby, Lythrypnus zebra (St. Mary 1996), belted sandfish, Serranus subligarius (Oliver 1997; Cheek 1998), tobacco fish, Serranus tabacarius (Petersen 1995) and chalk bass, Serranus tortugarum (Fischer 1984; Petersen & Fischer 1996). Fischer (1980) described simultaneous hermaphroditism in the black hamlet fish, H. nigricans (Serranidae). During the last 2 hours of the day before sunset, fishes come together in pairs at the reef edge or slope to spawn, usually away from their foraging territories. After several courtship displays, one fish initiates the spawn by releasing eggs which are externally fertilised by its partner. Eggs are much more expensive to produce than sperm, and so ‘cheaters’ could benefit by limiting their role in spawning to that of sperm donor. All available eggs are not released, but instead are parcelled out over four to five releases on average during a spawning period. The parcelling of eggs allows fishes to alternate their role as either male or female (i.e. egg trading) with one individual providing eggs to be fertilised in exchange for eggs from a partner to fertilise. Fischer (1988) refers to this exchange as ‘delayed reciprocity’. But why should a partner not defect (not cooperate) – i.e. not switch roles – in this system? Kin selection is not a likely explanation as these fishes are obligate outbreeders and eggs are planktonic (Fischer 1988). In the black hamlet, cooperation appears to be maintained because these fishes are playing the iPD game in which a cooperative act is parcelling eggs, defection is not providing eggs to a partner, and iterations are the exchange of parcels of eggs during the spawning period. Fischer (1988) argues that the costs and benefits associated with cooperation and defection during egg trading are consistent with the pay-off matrix in the iPD game. Additionally, he presents evidence that there exists the temptation to cheat and this temptation is met with retaliation. Fischer (1980) finds that fishes wait significantly longer to provide eggs to a partner that has failed to reciprocate compared to a partner that provided eggs on the previous move. This evidence suggests that the black hamlet uses a strategy similar to TFT in the iPD game – a strategy that is more forgiving than TFT called ‘Generous Tit-for-Tat’ (Nowak & Sigmund 1992). Although this evidence is compelling, alternative explanations may exist. One alternative to the TFT strategy has been suggested by Connor (1992), namely pseudo-reciprocity (Connor 1986). In pseudo-reciprocity (aka by-product mutualism), unlike reciprocity, there is no incentive to cheat as the benefits of mutual cooperation are greater than the temptation to cheat and the cognitive requirements of pseudo-reciprocity do not include individual recognition or memory. Connor (1992) argues that fishes engaging in egg trading are benefiting themselves and only as an incidental outcome also benefiting others. This remains to be tested.

BLBK374-12

BLBK374-Brown

266

12.3.2.3

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Example of reciprocity: Predator inspection

Small fishes have been observed swimming away from the relative safety of their shoal and moving towards a potentially dangerous predator. This behaviour has been called predator inspection (Pitcher et al. 1986) and approaching (Dugatkin & Godin 1992a) and has been observed in many fish species including guppies (Poecilia reticulata), stickleback (Gasterosteus aculeatus), minnows (Phoxinus phoxinus), paradise fish (Macropodus opersularis), damselfish (Stegastes planifrons), bluegill sunfish (Lepomis macrochirus) and mosquitofish (Gambusia affinis) (see Dugatkin & Godin 1992a; Pitcher 1992; Smith 1997 for reviews on predator inspection in fishes). The phenomenon of prey inspecting a potential predator is not limited to fishes and has been observed in other taxa including mammals (Walther 1969, Cheney & Seyfarth 1990; Fitzgibbon 1994) and birds (Altmann 1956; Curio 1978; Olendorf et al. 2004). The advantages of predator inspection include: Signalling to the predator that it has been seen and an attack would be unsuccessful, gathering information about the potential threat that may be transmitted back to the shoal and advertising individual quality to conspecifics (see Dugatkin & Godin 1992b for a review of costs and benefits of predator inspection). If multiple individuals inspect together, it can be considered a form of cooperation. In fact, predator inspection in fishes has become one of the most popular experimental systems to study reciprocity (see Dugatkin 1997 and Stevens & Hauser 2004 for controversies surrounding predator inspection and cooperation). Debates notwithstanding, it appears that predator-inspection behaviour fits the assumed pay-offs from the Prisoner’s Dilemma, namely T > R > P > S. Specifically, a fish benefits the most if it allows its partner to move closer to a potential predator to either signal to the predator it has lost the element of surprise and/or to gain information about the threat. As such, the temptation to defect (‘T’), i.e. staying out of harm’s way while another individual inspects, is greater than the pay-off for both fishes if they inspect together and gain the reward for mutual cooperation (‘R’). However, the fishes receive a greater reward if they both inspect than if they both remain in the shoal and not approach thereby getting the punishment for mutual defection (‘P’). Finally, it would be most dangerous (i.e. least rewarding) for a fish to approach a predator alone, receiving the sucker’s pay-off (‘S’). Early works by Milinski (1987) and Dugatkin (1988) on sticklebacks and guppies, respectively, have provided a strong foundation to suggest that not only are these fishes cooperating via reciprocity, but also likely using the TFT strategy, or a similar strategy, during predator inspections. Milinski (1987) and Dugatkin (1988) used a series of experiments during which cooperation or defection was simulated by placing a mirror parallel to (cooperation) or at an angle away from (defection) an individual inspecting a predator. In the case of simulated cooperation, the mirror image of the inspecting fish stayed next to the subject during its movement toward the predator. However, the mirror placed at an angle made the image appear to swim away from the subject as it moved toward the predator, i.e. it mimicked an act of defection. Both Milinski (1987) and Dugatkin (1988) found that in the simulated cooperation trials, fish approached a predator more closely when an inspector perceived its inspecting partner (its mirror image) cooperating as defecting from the inspection. This suggests that fishes are copying a partner’s last move as predicted by the TFT strategy. In addition to this, Milinski (1987) and Dugatkin (1988) tested specific predictions of the TFT strategy. For example, they found that fishes retaliated against defectors (fishes moved away

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

Char Count=

267

from the predator when their partner appeared to defect), and were forgiving (continued to inspect once their partner reappeared close to the subject) (see Dugatkin 1997 for a review). As stated in Subsection 12.3.2.1, the cognitive abilities required for cooperation via reciprocity and the iPD include recognition of partners and remembering their previous move. These prerequisites were tested in both sticklebacks (Milinski et al. 1990a, 1990b) and guppies (Dugatkin & Alfieri 1991a, 1991b). These works present evidence that sticklebacks and guppies both show preferences for partners that were more likely to cooperate with them during previous inspection bouts suggesting that individuals are able to discriminate between past partners as well as remember their previous encounters with them. Recently, the work of Dugatkin & Alfieri (1991a) has been repeated but with a current focus on the preference of an observer guppy that was restricted to only watch, but not interact with potential cooperators and defectors (Brosnan et al. 2003). This work did not find any significant preferences between observer guppies and either perceived cooperators or defectors and suggests that the repeated act of inspection between co-inspectors, as opposed to merely watching a conspecific inspect, is necessary for cooperation between inspectors to form. There has been much debate in the literature over the role of reciprocity during predation inspection in fishes. Although the scope of this discussion is too extensive to fully address in this chapter, aspects of the debate are evident by the exchange between Connor (1996), who argues that predator inspection behaviour can be explained via by-product mutualism, and the responses to this claim by Dugatkin (1996) and Milinski (1996) (for more on this topic, see Dugatkin 1997). 12.3.2.4

Example of reciprocity: Interspecific cleaning behaviour

Trivers (1971) provided an early example of cooperation via reciprocity by describing the interactions between a cleaner fish (e.g. the wrasse, Labroides dimidiatus) and their clients (e.g. the grouper, Epinephelus striatus). During this cooperative interaction, ‘cleaner fish repeatedly interact with ‘client’ fish, during which time cleaners feed on the parasites and unhealthy tissue of clients. Clients often swim to a specific location – a cleaning station – where a specific cleaner fish will likely swim into the gill chambers and mouth of a client or host fish to feed on ectoparasites (Trivers 1971). Cleaning stations are often found in the same place with the same cleaners and specific clients repeatedly return to these locations, so that there are repeated interactions between specific clients and cleaners. There is also a substantial cost (e.g. lost time, increased predation) to both cleaners and clients for repeatedly establishing new pairs. For this to be an example of cooperative behaviours via reciprocity, cleaner–client pairs must repeatedly interact and therefore must either have some form of individual recognition or have only very limited opportunity to interact with different partners at a cleaning station (Axelrod & Hamilton 1981). Trivers presents compelling evidence that these assumptions are met in at least some cleaner–client systems and that this is a primary example of reciprocal altruism (1971), although his example has been met with much debate. For example, Gorlick et al. (1978) directly question Trivers’ example. They describe behaviours of several Labroides spp. that do not meet the assumptions set forth by Trivers including evidence that (1) cleaners will feed on the healthy tissue, mucus, and fins of clients in addition to ectoparasites, (2) clients may not feed on cleaners not because of a cooperative interaction but rather because cleaners often

BLBK374-12

BLBK374-Brown

268

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

avoid the mouth of some piscivorous clients as well as avoid clients in search of food and (3) some cleaners are distasteful to clients and are avoided as a food item (Gorlick et al. 1978 and references within). Further examination of some of these issues including biting by and diet preferences of cleaner fishes have been examined in studies of cooperative behaviours (Bshary 2002b; Bshary & Grutter 2002; Grutter & Bshary 2004). Recently, the study of cleaning fishes has focused on specific aspects of cooperative behaviours. For example, Tebbich et al. (2002) studied individual recognition and preference for familiar versus unfamiliar fish in the cleaner fish, L. dimidiatus, and its client, the surgeon fish, Ctenochaetus striatus. They provide evidence that the cleaner fish spend significantly more time with familiar versus unfamiliar clients suggesting an important role of individual recognition in cleaning behaviour. However, Tebbich et al. (2002) did not find a preference for either familiar or unfamiliar cleaners in the client fish. They suggest that this may be due to experimental artefacts including low statistical power and/or low motivation on the part of the client due to insufficient time to allow client and cleaners to establish a ‘significant relationship’. Additionally, they suggest that there is not a need for recognition of individuals in this system but rather a need for recognition of a site where clients return and thereby facilitate repeated interactions with the same cleaners (Tebbich et al. 2002). Studies including the development of cleaner-client relationships (Bshary 2002c) and the role that tactile stimulations have on developing cooperative behaviours between cleaners and clients (Bshary & Wurth 2001) continue to advance our knowledge of cooperative behaviour.

12.3.3

Category 3 – by-product mutualism

By-product mutualism is a form of cooperation during which two or more individuals act together to achieve an outcome that could not have been achieved as efficiently (or at all) by any single individual (West Eberhard 1975; Brown 1983; Connor 1995) and there is no temptation for either individual not to cooperate. Using the definition we have described here, by-product mutualism has also been called no-cost cooperation (Dugatkin 1997), pseudo-reciprocity (Connor 1986), selfish cooperation (Stevens & Hauser 2004) and two-way by-products (Sachs et al. 2004). In this category of cooperation, cheaters (i.e. non-cooperators) receive a lower pay-off than cooperators. Although this makes by-product mutualism conceptually very easy to understand, it has been argued that it should not be considered cooperation at all because the temptation to cheat and the cost to act cooperatively are absent. However, given coordinated actions are needed between individuals to achieve an outcome of a greater reward than any one individual could obtain, we feel by-product mutualism does indeed fulfil our definition of cooperation. Cooperation via by-product mutualism occurs when environmental situations dictate that acting together yields greater rewards than acting alone. In this model, the environment is categorised as either ‘harsh’, in which case the best strategy is to cooperate, or ‘mild’ in which the best strategy is not to cooperate (Mesterton-Gibbons & Dugatkin 1992). 12.3.3.1

Cognition and by-product mutualism

Cognitive requirements are often less demanding in by-product mutualism compared to other categories of cooperation because memory and recognition of individuals are not

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

Char Count=

269

needed. The cognitive requirement of by-product mutualism is a categorical recognition of one’s environment – if you are in a ‘harsh’ environment, then cooperate, if in a ‘mild’ environment, do not cooperate (Dugatkin 1997; Dugatkin & Alfieri 2002). However, in situations where populations have been subjected to a particular type of environmental condition (harsh or mild) for long periods of time, the ability to recognise the environmental type may not be necessary. As a result of the lesser cognitive requirements associated with by-product mutualism (as compared to reciprocity), it has been suggested that by-product mutualism is a more common form of cooperation in the animal kingdom. Stevens & Hauser (2004) state that ‘selfish cooperation’ (i.e. by-product mutualism) is in fact common in animal societies mostly because the cognitive requirements are severe enough to act as a barrier to the evolution of most types of reciprocity. 12.3.3.2

Example of by-product mutualism: Cooperative foraging

The feeding habits of adult blue tang surgeonfish (Acanthurus coeruleus) present an example of cooperation via by-product mutualism. Blue tang surgeonfish feed on algae and form feeding schools (Foster 1985). Often, highly desirable algal mats are defended by territorial dusky damselfish (Stegastes dorsopunicans). A solitary blue tang cannot overcome the defence of a damselfish; however, feeding schools can overcome a territorial damselfish and feed on the algal resource (Foster 1985). In this example, the ‘harsh’ environment that stimulates cooperation via by-product mutualism in the surgeonfish is the territorial defence of the damselfish. Dugatkin (1997) and Dugatkin & Mesterton-Gibbons (1996) describe similar examples of cooperative foraging through by-product mutualism including work by Foster (1987) on the wrasse (Thalassoma lucasanum). In this example, the sergeant major damselfish (Abudefduf troschelii) defends its embryos successfully from small groups of wrasses (fewer than 30 individuals); however, it cannot prevent the predation of the embryos from larger groups (hundreds of individuals). Here again, the harsh environment is defined by territorial defence and only through cooperative actions can a reward not obtainable by one (or a few) be obtained by the group. Interestingly, large groups of wrasses were seen only when damselfish were nesting with embryos, suggesting that group-size formation in the wrasse is a direct response to achieving a cooperative reward. Similar examples of by-product mutualism cooperation resulting from the harsh environment of a territory holder are cited in Dugatkin (1997). Bshary (2002a) presents interesting anecdotal evidence of complex cooperative hunting behaviours between giant moray eels (Gymnothorax javanicus) and two different groupers, the red sea coral groupers (Plectropomus pessuliferus) and lunartail groupers (Variola louti). Groupers have been observed following eels and octopuses while they hunt in an attempt to capture prey the eels and octopuses flush out of hiding places in corals (Diamant & Shpigel 1985). Bshary (2002a) describes his observations of the two grouper species mentioned in the preceding text, i.e. P. pessuliferus and V. louti, approaching a resting moray eel, and shaking their bodies within close proximity to the eel. In half of these close encounters, the eel and grouper would then swim off together in close proximity to hunt for prey. The eel would swim into a coral while the grouper waited above the coral. In one instance, Bshary (2002a) witnessed a grouper wait outside a coral head for an escaped prey item to emerge, swim away from the coral, then return with a grey moray eel (Siderea grisea). Although

BLBK374-12

BLBK374-Brown

270

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

he did not witness capture of prey by any eels or groupers, he suggests that groupers are recruiting eels to hunt cooperatively and that groupers and eels assume different roles during a hunt (Bshary 2002a). Interactions between eels and groupers are interesting examples of complex, by-product mutualism cooperative foraging behaviours assuming the harsh environment is the difficulty in capturing a prey item.

12.3.4

Category 4 – trait group selection

Until the 1960s, the term ‘group selection’ was associated with Wynne-Edwards’ ‘for the good of the species’ argument (1962). Wynne-Edwards argued that only groups with cooperators that benefited their species (i.e. self-sacrificing individuals that would control population size to safeguard against overexploitation) would be selected as the benefits to the group outweighed the costs to cooperators. In general, this view of selection has not been supported by theoretical or empirical works and has never gained favour within the scientific community as evident by criticisms from Williams (1966). Williams argued that individuals that possessed the trait for self-sacrifice (e.g. restricting reproduction for the good of the group) would be selected against due to natural selection favouring individuals that were not sacrificing, producing relatively more offspring than sacrificing individuals and passing on this trait to a larger percentage of subsequent generations. Over evolutionary time, self-sacrificing individuals would be outcompeted within populations (Alcock 2001). However, more complex models of group selection have been proposed. Modern or ‘trait group selection’ (Wilson 1975a) models describe cooperative behaviours in populations by examining fitness based on the productivity of local groups or ‘trait groups’. Here, the effects of cooperative acts are examined both at the level of the individuals within the trait group (where a cooperator pays a cost that non-cooperators do not) and at the level of the trait groups within the global population or deme (Dugatkin & Mesterton-Gibbons 1996). Cooperation is possible (even if there is a cost to the cooperator) if the within-group costs are less than between-group benefits so that groups with cooperators are more productive than groups without cooperators (Sober & Wilson 1998) (for more on this, see Wilson 1990; Mesterton-Gibbons & Dugatkin 1992; Wilson & Sober 1994; Dugatkin 1997; Sober & Wilson 1998). 12.3.4.1

Cognition and trait group selection

Although it is possible to have cooperation via trait group selection without individual recognition or memory of events (Wilson 1980), trait group cooperation would be favored if individuals recognised and associated with other cooperators (Peck 1993; Wilson & Dugatkin 1997; Roberts & Sherratt 1998). Such recognition would allow for the formation of trait groups with many cooperators, and such groups should have greater productivity than groups with proportionally more cheaters. This would require individuals to recognise others as belonging to a general category of cooperators or cheaters but could also include individual recognition (Dugatkin & Alfieri 2002). 12.3.4.2

Example of trait group selected cooperation: Predator inspection

Predator inspection (see description in Subsection 12.3.2.3) involves the movement of a few individuals (inspectors) away from a larger group towards a potential predator.

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

Char Count=

271

For trait group cooperation to be the mechanism of cooperation during predator inspection, one needs to document costs to individuals who cooperate when they leave the group to inspect, document benefits accrued by groups from having inspectors and show that groups with inspectors should have an advantage over groups without inspectors. Dugatkin & Mesterton-Gibbons (1996) review several empirical studies in the guppy (P. reticulata) that provide evidence for each of these requirements. First, there are costs associated with inspection. Inspectors have been shown to be at greater risk of predation (Dugatkin 1992) and inspectors obtain less food than non-inspectors (Dugatkin & Godin 1992a). Secondly, groups benefit from the action of inspectors. The information obtained by inspectors is transmitted back to the group and the entire group benefits from this information (Magurran & Higgam 1988). Finally, groups with inspectors may have some fitness advantage over groups without inspectors. Although they did not test this directly, the results from Dugatkin & Godin (1992a) suggest that this is indeed the case, that groups with inspectors are attacked less often by predators than groups without inspectors. While more work needs to be undertaken in this system, the within-group costs of inspection may be less than the between-group benefits suggesting predator inspection as an example of trait group selection in fishes. At equilibrium, selection would balance the within-group cost to cheating (i.e. cheaters get information about predators, but pay no cost) against the between-group benefit of having many inspectors (e.g. groups are very vigilant against predators). A higher proportion of inspectors to cheaters in a group may result in greater productivity, i.e. reduced attacks, for the group (Dugatkin & Godin 1992a). Two fish species that have been extensively studied regarding predator inspection are the guppy (P. reticulata) and the stickleback (G. aculeatus). Empirical studies on both species have shown that guppies and sticklebacks are able to recognise familiar individuals in general (Magurran et al. 1994; Barber & Ruxton 2000, respectively) as well as in the specific context of preference for a familiar cooperative conspecific (Dugatkin & Alfieri 1991a, 1991b in guppies and Milinski et al. 1990a, 1990b in sticklebacks). However, in guppies preferential assortment seems to be limited to individuals in small groups (Dugatkin & Alfieri 1991a), and is absent when group size becomes too large (Dugatkin & Wilson 2000). Additionally, several studies of guppies and sticklebacks have shown that kin selection is likely not a selective force on cooperative behaviours in these species due to low relatedness in sampled populations (Griffiths & Magurran 1999; Russell et al. 2004 in guppies and FitzGerald & Morisette 1992 in sticklebacks; see also Ward & Hart 2003 and chapter 8 for a general review of familiarity and kin recognition). The work on predation inspection shows nicely that different categories of cooperation are not mutually exclusive. As we have now seen, cooperation during predator inspection may include elements of both reciprocity and group-selected behaviour.

12.4

Conclusion

We have selected specific examples of empirical studies in fishes that have addressed each of the four categories of cooperation and their cognitive requirements (Table 12.1). Through further investigation, each example may help refine models of cooperation or perhaps inspire the creation of new models. We feel that this is a productive way to best continue

BLBK374-12

BLBK374-Brown

272

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

our understanding of the relationships among cooperation, cognition and fish behaviour. We hope that this work will stimulate research that directly addresses the cognitive requirements in all forms of cooperative behaviours.

Acknowledgements We would like to thank Jennifer Sadowski, Glena Temple and Darren Croft for their helpful and thoughtful comments on this chapter.

References Alcock, J. (2001) Animal Behavior: An Evolutionary Approach, 7th ed. Sinauer Associates Inc., Massachusetts. Altmann, S.A. (1956) Avian mobbing behavior and predator recognition. Condor, 58, 241–253. Axelrod, R. (1984) The Evolution of Cooperation. Basic Books Inc., New York. Axelrod, R. & Hamilton, W.D. (1981) The evolution of cooperation. Science, 211, 1390–1396. Barber, I. & Ruxton, G.D. (2000) The importance of stable schooling: do familiar sticklebacks stick together? Proceedings of the Royal Society of London Series B – Biological Sciences, 267, 151–155. Bergmuller, R., Heg, D. & Taborsky, M. (2005) Helpers in a cooperatively breeding cichlid stay and pay or disperse and breed, depending on ecological constraints. Proceedings of the Royal Society of London Series B – Biological Sciences, 272, 325–331. Bergmuller, R. & Taborsky, M. (2005) Experimental manipulation of helping in a cooperative breeder: helpers ‘pay to stay’ by pre-emptive appeasement. Animal Behaviour, 69, 19–28. Brosnan, S.F., Earley, R.L. & Dugatkin, L.A. (2003) Observational learning and predator inspection in guppies (Poecilia reticulata). Ethology, 109, 823–833. Brouwer, L., Heg, D. & Taborsky, M. (2005) Experimental evidence for helper effects in a cooperatively breeding cichlid. Behavioral Ecology, 16, 667–673. Brown, C. & Laland, K.N. (2003) Social learning in fishes: a review. Fish and Fisheries, 4, 280–288. Brown, G.E. & Brown, J.A. (1992) Do rainbow trout and Atlantic salmon discriminate kin? Canadian Journal of Zoology, 70, 1636–1640. Brown, G.E. & Brown, J.A. (1993) Social dynamics in salmonid fishes: do kin make better neighbours? Animal Behaviour, 45, 863–871. Brown, J.L. (1964) The evolution of diversity in avian territorial systems. Wilson Bulletin, 76, 160–169. Brown, J.L. (1983) Cooperation: a biologist’s dilemma. In: J.S. Rosenblatt (ed) Advances in the Study of Behaviour, pp. 1–37. Academic Press, New York. Brown, J.L. (1987) Helping and Communal Breeding in Birds. Princeton University Press, Princeton, NJ. Bshary, R. (2002a) Fish cognition: a primate’s eye view. Animal Cognition, 5, 1–13. Bshary, R. (2002b) Biting cleaner fish use altruism to deceive image-scoring client reef fish. Proceedings of the Royal Society of London Series B – Biological Sciences, 269, 2087–2093. Bshary, R. (2002c) Building up relationships in asymmetric co-operation games between the cleaner wrasse Labroides dimidiatus and client reef fish. Behavioral Ecology and Sociobiology, 52, 365–371. Bshary, R. & Grutter, A.S. (2002) Asymmetric cheating opportunities and partner control in a cleaner fish mutualism. Animal Behaviour, 63, 547–555. Bshary, R. & Wurth, M. (2001) Cleaner fish Labroides dimidiatus manipulate client reef fish by providing tactile stimulation. Proceedings of the Royal Society of London Series B – Biological Sciences, 268, 1495–1501.

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

Char Count=

273

Cheek, A.O. (1998) Ovulation does not constrain egg parcel size in the simultaneous hermaphrodite Serranus subligarius. Environmental Biology of Fishes, 52, 435–442. Cheney, D.L. & Seyfarth, R.M. (1990) How Monkeys See the World. University of Chicago Press, Chicago. Clutton-Brock, T. (2002) Breeding together: kin selection and mutualism in cooperative vertebrates. Science, 296, 69–72. Connor, R.C. (1986) Pseudo-reciprocity: investing in mutualism. Animal Behaviour, 34, 1562–1584. Connor, R.C. (1992) Egg-trading in simultaneous hermaphrodites: an alternative to tit-for-tat. Journal of Evolutionary Biology, 5, 523–528. Connor, R.C. (1995) The benefits of mutualism: a conceptual framework. Biological Reviews, 70, 427–257. Connor, R.C. (1996) Partner preferences in by-product mutualisms and the case for predator inspection in fish. Animal Behaviour, 51, 451–454. Croft, D.P., Krause, J. & James, R. (2004) Social networks in the guppy (Poecilia reticulata). Proceedings of the Royal Society of London Biology Letters, 271, 516–519. Croft, D.P., James, R., Ward, A.J.W., Botham, M.S., Mawdsley, D. & Krause, J. (2005) Assortative interactions and social networks in fish. Oecologia, 143, 211–219. Croft, D.P., James, R., Thomas, P.O.R., Hathaway, C., Mawdsley, D., Laland, K.N. & Krause, J. (2006) Social structure and co-operative interactions in a wild population of guppies (Poecilia reticulata) Behavioral Ecology and Sociobiology, 59, 644–650. Curio, E. (1978) The adaptive significance of avian mobbing. I. Teleonomic hypotheses and predictions. Zeitschrift fur Tierpsychologie, 48, 175–183. Darwin, C. (1859) On the Origin of Species. J. Murray, London. Diamant, A. & Shpigel, M. (1985) Interspecific feeding associations of groupers (Teleostei: Serranidae) with octopuses and moral eels in the Gulf of Eilat (Aqaba). Environmental Biology of Fishes, 13, 153–159. Diana, J.S. (2004) Biology and Ecology of Fishes, 2nd ed. Cooper Publishing Group, LLC, USA. Dugatkin, L.A. (1988) Do guppies play tit for tat during predator inspection visits? Behavioral Ecology and Sociobiology, 25, 395–399. Dugatkin, L.A. (1992) Tendency to inspect predators predicts mortality risk in the guppy. Poecilia reticulata. Behavioral Ecology, 3, 124–128. Dugatkin, L.A. (1996) Tit for Tat, by-product mutualism and predator inspection: a reply to Connor. Animal Behaviour, 51, 455–457. Dugatkin, L.A. (1997) Cooperation Among Animals: An Evolutionary Perspective. Oxford University Press, New York. Dugatkin, L.A. & Alfieri, M. (1991a) Guppies and the tit for tat strategy: preference based on past interaction. Behavioral Ecology and Sociobiology, 28, 243–246. Dugatkin, L.A. & Alfieri, M. (1991b) Tit for tat in guppies: the relative nature of cooperation and defection during predator inspection. Evolutionary Ecology, 5, 300–309. Dugatkin, L.A. & Alfieri, M.S. (2002) A cognitive approach to the study of animal cooperation. In: M. Bekoff & C. Allen (eds) The Cognitive Animal, pp. 413–419. M.I.T. Press, Cambridge, MA. Dugatkin, L.A. & Godin, J-G.J. (1992a) Predator inspection, shoaling and foraging under predation hazard in the Trinidadian guppy, Peocilia reticulata. Environmental Biology of Fishes, 34, 265–276. Dugatkin, L.A. & Godin, J-G.J. (1992b) Prey approaching predators: a cost-benefit perspective. Annales Zoologici Fennici, 29, 233–252. Dugatkin, L.A. & Mesterton-Gibbons, M. (1996) Cooperation among unrelated individuals: reciprocal altruism, byproduct mutualism, and group selection in fishes. Biosystems, 37, 19–30. Dugatkin, L.A., Mesterton-Gibbons, M. & Houston, A.I. (1992) Beyond the prisoner’s dilemma: towards models to discriminate among mechanisms of cooperation in nature. Trends in Ecology and Evolution, 7, 202–205. Dugatkin, L.A. & Reeve, H.K. (eds) (1998) Game Theory and Animal Behavior. Oxford University Press, Oxford.

BLBK374-12

BLBK374-Brown

274

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Dugatkin, L.A. & Wilson, D.S. (2000) Assortative interactions and the evolution of cooperation during predator inspection in guppies (Poecilia reticulata). Evolutionary Ecology Research, 2, 761–767. Fischer, E.A. (1980) The relationship between mating system and simultaneous hermaphroditism in the coral reef fish, Hypoplectrus nigricans (Serranidae). Animal Behaviour, 28, 620–633. Fischer, E.A. (1981) Sexual allocation in a simultaneously hermaphroditic coral reef fish. American Naturalist, 117, 64–82. Fischer, E.A. (1984) Egg trading in the chalk bass, Serranus tortugarum, a simultaneous hermaphrodite. Zeitschrift fuer Tierpsychologie, 66, 143–151. Fischer, E.A. (1987) Mating behavior in the black hamlet-gamete trading or egg trading? Environmental Biology of Fishes, 18, 143–148. Fischer, E.A. (1988) Simultaneous hermaphroditism, tit-for-tat and the evolutionary stability of social systems. Ethology and Sociobiology, 9, 119–136. Fischer, E.A. & Petersen, C.W. (1987) The evolution of sexual patterns in the seabasses. BioScience, 37, 482–489. FitzGerald, G.J. & Morisette, J. (1992) Kin recognition and choice of shoal mates by threespine sticklebacks. Ethology Ecology and Evolution, 4, 273–283. Fitzgibbon, C.D. (1994) The costs and benefits of predator inspection behavior in Thomson gazelles. Behavioral Ecology and Sociobiology, 34, 139–148. Foster, S.A. (1985) Group foraging by a coral reef fish: mechanism for gaining access to defended resources. Animal Behaviour, 33, 782–792. Foster, S.A. (1987) Acquisition of a defended resource: a benefit of group foraging for the neotropical wrasse, Thalassoma lucasanum. Environmental Biology of Fishes, 19, 215–222. Gadagkar, R. (1997) Survival Strategies: Cooperation and Conflict in Animal Societies. Harvard University Press, Cambridge, MA. Godin, J.-G.J. (1997) Beahvioural ecology of fishes: adaptations for survival and reproduction In: J-G.J. Godin (ed) Behavioural Ecology of Teleost Fishes, pp. 1–9. Oxford University Press, New York. Gorlick, D.L., Atkins, P.D. & Losey, G.S., Jr. (1978) Cleaning stations as water holes, garbage dumps, and sites for the evolution of reciprocal altruism? The American Naturalist, 112, 341–353. Grant, J.W.A. (1997) Territoriality. In: J-G.J. Godin (ed) Behavioural Ecology of Teleost Fishes, pp. 81–103. Oxford University Press, New York. Griffin, A.S. & West, S.A. (2003) Kin discrimination and the benefit of helping in cooperatively breeding vertebrates. Science, 302, 634–636. Griffiths, S.W. (2003) Learned recognition of conspecifics by fishes. Fish and Fisheries, 4, 256–268. Griffiths, S.W. & Armstrong, J.D. (2002) Kin-biased territory overlap and food sharing among Atlantic salmon juveniles. Journal of Animal Ecology, 71, 480–486. Griffiths, S.W. & Magurran, A.E. (1999) Schooling decisions in guppies (Poecilia reticulata) are based on familiarity rather than kin recognition by phenotype matching. Behavioral Ecology and Sociobiology, 45, 437–443. Grutter, A.S. & Bshary, R. (2004) Cleaner fish, Labroides dimidiatus, diet preference for different types of mucus and parasitic gnathiid isopods. Animal Behaviour, 68, 583–588. Gutnisky, D.A. & Zanutto, B.S. (2004) Cooperation in the iterated prisoner’s dilemma is learned by operant conditioning mechanisms. Artificial Life, 10, 433–461. Hamilton, W.D. (1964a) The genetical evolution of social behaviour I. Journal of Theoretical Biology, 7, 1–16. Hamilton, W.D. (1964b) The genetical evolution of social behaviour II. Journal of Theoretical Biology, 7, 17–52. Hammerstein, P. (2002) Why is reciprocity so rare in social animals? A Protestant appeal. In: P. Hammerstein (ed) Genetic and Cultural Evolution of Cooperation, pp. 83–93. Dahlem University Press. Krause, J., Butlin, R., Peuhkuri, N. & Pritchard, V. (2000) The social organization of fish shoals: a test of the predictive power of laboratory experiments for the field. Biological Reviews, 75, 477–501.

BLBK374-12

BLBK374-Brown

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Cooperation and Cognition in Fishes

Char Count=

275

Laland, K.N. & Hoppitt, W. (2003) Do animal’s have culture? Evolutionary Anthropology, 3, 150–159. Luce, R.D. & Raiffa, H. (1957) Games and Decisions. John Wiley & Sons, New York. Magurran, A.E. & Higgam, A. (1988) Information transfer across fish shoals under predator threat. Ethology, 78, 153–158. Magurran, A.E., Seghers, B., Shaw, P. & Carvalho, G. (1994) Schooling preferences for familiar fish in the guppy, Poecilia reticulata. Journal of Fish Biology, 45, 401–406. Mesterton-Gibbons, M. & Dugatkin, L.A. (1992) Cooperation among unrelated individuals: evolutionary factors. Quarterly Review of Biology, 67, 267–281. Milinski, M. (1987) Tit for tat and the evolution of cooperation in sticklebacks. Nature, 325, 433–435. Milinski, M. (1996) By-product mutualism, tit for tat reciprocity and cooperative predator inspection: a reply to Connor. Animal Behaviour, 51, 458–461. Milinski, M., Kulling, D. & Kettler, R. (1990a) Tit for tat: sticklebacks “trusting” a cooperating partner. Behavioral Ecology, 1, 7–12. Milinski, M., Pfugler, D. Kulling, D. & Kettler, R. (1990b) Do sticklebacks cooperate repeatedly in reciprocal pairs? Behavioral Ecology and Sociobiology, 27, 17–23. Milinski, M. & Wedekind, C. (1998) Working memory constrains human cooperation in the prisoner’s dilemma. Proceedings of the National Academy of Science United States of America, 95, 13755–13758. Nowak, M.A. & Sigmund, K. (1992) Tit for tat in heterogeneous populations. Nature, 355, 250–252. Olendorf, R., Getty, T. & Scribner, K. (2004) Cooperative nest defense in red-winged blackbirds: reciprocal altruism, kinship or by-product mutualism? Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 177–182. Oliver, A.S. (1997) Size and density dependent mating strategies in the simultaneously hermaphroditic seabass Serranus subligarius (Cope, 1870). Behaviour, 134, 563–594. Peck, J.R. (1993) Friendship and the evolution of cooperation. Journal of Theoretical Biology, 162, 195–228. Petersen, C.W. (1995) Reproductive behavior, egg trading and correlates of male mating success in the simultaneous hermaphrodite. Serranus tabacarius. Environmental Biology of Fishes, 43, 351–361. Petersen, C.W. & Fischer, E.A. (1996) Intraspecific variation in sex allocation in a simultaneous hermaphrodite: the effect of individual size. Evolution, 50, 636–645. Pfeiffer, T., Rutte, C., Killingback, T., Taborsky, M. & Bonhoeffer, S. (2005) Evolution of cooperation by generalized reciprocity. Proceedings of the Royal Society of London Series B – Biological Sciences, 272, 1115–1120. Pitcher, T.J. (1992) Who dares wins: the function and evolution of predator inspection behaviour in shoaling fish. Netherlands Journal of Zoology, 42, 371–391. Pitcher, T.J., Green, D.A. & Magurran, A.E. (1986) Dicing with death: predator inspection behaviour in minnow shoals. Journal of Fish Biology, 28, 438–448. Rapoport, A. & Chammah, A. (1965) Prisoner’s Dilemma. Univeristy of Michigan Press, Ann Arbor, MI. Riolo, R.L., Cohen, M.D. & Axelrod, R. (2001) Evolution of cooperation without reciprocity. Nature, 414, 441–443. Roberts, G. & Sherratt, T. (1998) Development of cooperative relationships through increasing investment. Nature, 394, 175–179. Russell, S.T., Kelley, J.L., Graves, A. & Magurran, A.E. (2004) Kin selection and shoal composition dynamics in the guppy, Poecilia reticulata. Oikos, 106, 520–526. Sachs, J.L., Mueller, U.G., Wilcox, T.P. & Bull, J.J. (2004) The evolution of cooperation. The Quarterly Review of Biology, 79, 135–160. Shaw, E. (1978) Schooling fishes. American Scientist, 66, 166–175. Smith, C.L. (1975) The evolution of hermaphroditism in fishes. In: R. Reinboth (ed) Intersexuality in the Animal Kingdom, pp. 295–310. Springer-Verlag, New York. Smith, R.J.F. (1997) Avoiding and deterring predators In: J-G.J. Godin (ed) Behavioural Ecology of Teleost Fishes, pp. 163–190. Oxford University Press, Oxford.

BLBK374-12

BLBK374-Brown

276

May 27, 2011

11:2

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Sober, E. & Wilson, D.S. (1998) Unto Others: The Evolution and Psychology of Unselfish Behavior. Harvard University Press, Cambridge, MA. St. Mary, C.M. (1996) Sex allocation in a simultaneous hermaphrodite, the zebra goby Lythrypnus zebra: insights gained through a comparison with its sympatric congener, Lythrypnus dalli. Environmental Biology of Fishes, 45, 177–190. Stevens, J.R. & Hauser, M.D. (2004) Why be nice? Psychological constraints on the evolution of cooperation. Trends in Cognitive Sciences, 8, 60–65. Stiver, K.A., Dierkes, P., Taborsky, M. & Balshine, S. (2004) Dispersal patterns and status change in a co-operatively breeding fish Neolamprologus pulcher: evidence from micro-satellite analyses and behavioural observations. Journal of Fish Biology, 65, 91–105. Taborsky, M. (1984) Broodcare helpers in the cichlid fish Lamprologus bricharid: their costs and benefits. Animal Behaviour, 32, 1236–1252. Taborsky, M. (1985) Breeder-helper conflict in a child fish with broodcare helpers: an experimental analysis. Behaviour, 95, 45–75. Tebbich, S., Bshary, R. & Grutter, A.S. (2002) Cleaner fish Labroides dimidiatus recognize familiar clients. Animal Cognition, 5, 139–145. Trivers, R. (1971) The evolution of reciprocal altruism. Quarterly Review of Biology, 46, 35–57. Trivers, R. (2004) Genetic and cultural evolution of cooperation. Science, 304, 964–965. Walther, F.R. (1969) Flight behaviour and avoidance of predators in Thomson’s gazelles (Gazella thomsoni Guenther 1884). Behaviour, 34, 184–221. Ward, A.J.W., Botham, M.S., Hoare, D.J., James, R., Broom, M., Godin, J.-G.J. & Krause, J. (2002) Association patterns and shoal fidelity in the three-spined stickleback. Proceedings of the Royal Society of London Series B – Biological Sciences, 269, 2451–2455. Ward, A.J.W. & Hart, P.J.B. (2003) The effects of kin and familiarity on interactions between fish. Fish and Fisheries, 4, 348–358. Werner, N.Y., Balshine, S., Leach, B. & Lotem, A. (2003) Helping opportunities and space segregation in cooperatively breeding cichlids. Behavioral Ecology, 14, 749–756. West Eberhard, M.J. (1975) The evolution of social behavior by kin selection. Quarterly Review of Biology, 50, 1–33. Williams, G.C. (1966) Adaptation and Natural Selection. Princeton University Press, Princeton, NJ. Wilson, D.S. (1975a) A theory of group selection. Proceeding of the National Academy of Sciences of the United States of America, 72, 143–146. Wilson, E.O. (1975b) Sociobiology: The New Synthesis. Harvard University Press, Cambridge, MA. Wilson, D.S. (1980) The Natural Selection of Populations and Communities. Benjamin Cummings, Menlo Park. Wilson, D.S. (1990) Weak altruism, strong group selection. Oikos, 59, 135–140. Wilson, D.S. & Dugatkin, L.A. (1992) Altruism. In: E.F. Keller & E.A. Lloyd (eds) Keywords in Evolutionary Biology, pp. 28–33. Harvard University Press, Cambridge, MA. Wilson, D.S. & Dugatkin, L.A. (1997) Group selection and assortative interactions. The American Naturalist, 149, 336–351. Wilson, D.S. & Sober, E. (1994) Re-introducing group selection to the human behavioral sciences. Behavioral and Brain Science, 17, 585–654. Wynne-Edwards, V.C. (1962) Animal Dispersion in Relation to Social Behavior. Oliver and Boyd, Edinburgh.

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 13

Machiavellian Intelligence in Fishes Redouan Bshary

13.1

Introduction

The aim of this chapter is to present an updated overview of the social strategic behaviour of fishes. The content was inspired by the ‘Machiavellian intelligence hypothesis’ (Byrne & Whiten 1988). In its initial form the hypothesis was formulated very broadly and stated that social life offers ample opportunities for the evolution of cognitive skills (Byrne & Whiten 1988). However, a more refined hypothesis emerged that proposes that the main cognitive challenge for individual primates is to cope with and exploit the complexity of the social structure in a manner that enhances their fitness. More specifically, an individual has to know all group members and their (genetic and social) relationships in order to find the right coalition partners and to prevent opponents from building successful coalitions. Key cognitive abilities of individuals are thus the ability to understand and remember complex relationships, to cooperate and skills in manipulation and deception of group members. Other social cognitive abilities like social learning and the formation of traditions became less prominent in the Machiavellian intelligence hypothesis, but play a role in the closely related social brain hypothesis (Dunbar 1992; Barton & Dunbar 1997), which stresses a link between social complexity and neocortex size evolution in mammals. There are positive correlations between group size (a correlate of social complexity) and neocortex ratio (neocortex size regressed against the size of the rest of the brain) in primates, carnivores and bats (Barton & Dunbar 1997). More recently, similar results have been found in birds but with an emphasis on the importance of pair bonds (Burish et al. 2004; Emery & Clayton 2004). Thus, both the Machiavellian intelligence hypothesis and the social brain hypothesis seem to be applicable to a variety of taxa. This chapter applies this reasoning to social strategies of fishes. To be able to present the evidence for Machiavellian intelligence in fishes, an important issue has to be clarified. Many primatologists interpret the social behaviour of primates as the result of complex cognitive mechanisms (Byrne & Whiten 1988). The hypothesis that Machiavellian intelligence is linked to neocortex size (or the size of functionally homologous structures) is based on the assumption that social animals not only successfully solve

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-13

BLBK374-Brown

278

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

the complexity of their social environment but that they actually evolve some understanding about why a certain behaviour is successful or not. This has spurred a major debate and much research effort into trying to find out whether primates are able, for example, to understand what other individuals want, feel, believe, i.e. whether they have a theory of mind (Premack & Woodruff 1978). The research effort has provided some good evidence for at least some basic aspects of theory of mind in primates and corvids. For example, the ability to take another individual’s perspective (being able to predict what another individual may or may not be able to see) has been documented in several primates (Tomasello et al. 1998), other mammals (Br¨auer et al. 2004; Kaminski et al. 2005) and in ravens (Bugnyar et al. 2004). In scrub jays, there is experimental evidence for so-called experience projection: Young jays develop an aversion for being watched during food caching if they are given the opportunity to raid the caches of others (Emery & Clayton 2001). With respect to strategic competence, it has been demonstrated in chimpanzees that they immediately understand whether or not they need a partner to solve a task (Melis et al. 2006). No similar tests have ever been conducted on fishes. Therefore, it is important to make a clear distinction between two components that are, unfortunately, mixed up in most definitions of ‘cunning’ social behaviour, namely the phenomenon and the underlying mechanism. This issue can be illustrated by a definition of tactical deception. On the phenomenological level, tactical deception is defined as the production of a signal out of its normal context, causing contextspecific behaviour of the signal’s recipient to its own disadvantage and to the signaller’s advantage (Hauser 1998). A good example is a false alarm call that yields access to food monopolised by others (Munn 1986). However, everyday language and social scientists would use the term ‘tactical deception’ only if the actor knows why the signal works and that it has detrimental effects for the recipient. In other words, the signaller has to be conscious about his action and conscious about what the recipient perceives, so the signaller must have a theory of mind. However, theory of mind is not the only mechanism that may produce this behavioural pattern. Animals may learn to produce signals out of context in a much simpler way, via operant conditioning (Thorndike 1917; Heyes 1998). An initial error (an alarm call that was immediately perceived to be unjustified because, for example, a log was misidentified as a crouching leopard) is positively reinforced (others flee, yielding access to food), which increases the probability that a false alarm call will be produced in the future. In the absence of knowledge about the underlying mechanism, I propose that it is most fruitful to restrict the definition of behavioural sequences to the phenomenological level. On this level, many different species from different taxa can be compared. As knowledge of underlying mechanisms increases, researchers can exploit this to distinguish between cases: Tactical deception based on theory of mind; tactical deception based on operant conditioning; or tactical deception based on a single error. An alternative way of defining animal behaviour in terms of cognition has been promoted by Clayton et al. (2003), who faced the problem that although animals may remember the ‘what’, ‘when’ and ‘where’ of particular events, current theories of human episodic memory refer, in addition, to the conscious experience of self that accompanies episodic recall, a state that has no obvious manifestation in non-linguistic behaviour (Tulving 1983; Tulving & Markowitsch 1998). As a solution, Clayton et al. (2003) promote the term ‘episodic-like memory’. In analogy, one could use ‘tactical deception-like’ behaviour for cases where the underlying mechanism is not theory of mind. My personal feeling is that adding ‘like’ may solve some definition

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Machiavellian Intelligence in Fishes

279

problems (in particular those where the definition is linked to language) but not all. Most problematic is the use of ‘like’ definitions when a phenomenon is described in the absence of definite knowledge about the underlying mechanism. Should we classify the acquisition of tool use in wild chimpanzees ‘social learning-like’ as long as we do not have experimental evidence for a social learning mechanism? This chapter sticks to the distinction between phenomenon and underlying mechanism. The basis for more sophisticated social behaviour is certainly the ability to recognise individuals and to interact repeatedly. The latter condition is certainly fulfilled if animals live in stable groups. However, interspecific relationships can also be quite stable and hence select for sophisticated behaviour. Individual recognition has been shown for a variety of fish species (see Chapter 9 for a full review). Therefore, it seems reasonable to argue that individual recognition will have evolved whenever it was useful, and that behavioural actions of fishes are unlikely to be constrained by an absence of individual recognition. This statement includes fishes that live in aggregations, as indicated by data on partner choice in predator-inspection behaviour. With respect to group living, fishes often live in diverse stable groups of varying sizes and sex composition and defend their territories and/or their eggs and larvae. Damselfish (genera Dascyllus and Amphiprion) of the Red Sea live in stable social assemblages as pairs, harems or solitary neighbouring individuals, where unknown individuals are treated differently to established neighbours (Fricke 1975). Cichlids in particular are well known for their uniparental or biparental brood care (reviewed by Keenleyside 1991). In the first and larger part of this chapter, I describe cases that highlight functional aspects of Machiavellian intelligence: Cooperation, manipulation, reconciliation and deception in fishes. In the second part, I have incorporated recent new insights on cognitive mechanisms described in fishes. This part presents the major difference to the chapter on Machiavellian intelligence in fish in the first edition of this book.

13.2

Evidence for functional aspects of Machiavellian intelligence

13.2.1 Information gathering about relationships between other group members A key ability for Machiavellian behaviour would be to know the relationships between other individuals and use this information for one’s own decisions. Fishes are known to ‘eavesdrop’, i.e. to use information from observations of interactions between conspecifics (McGregor 1993; and see Chapters 5 and 11). Dugatkin & Godin (1992a) provided experimental evidence for female guppies changing their preferences between two males if they observed another female being courted by the less-preferred male. These results have been replicated with sailfin mollies (Poecilia latipinna, Poeciliidae) under field conditions (Witte & Noltemeier 2002; Witte & Ryan 2002). Earlier, Schlupp et al. (1994) had found that male sailfin mollies may improve their reproductive success by mating with females of the parthenogenetic Amazon molly (Poecilia formosa, Poeciliidae), because it increases the probability that females of their own species select them as partners. Thus, these females extract information from observed interactions between males and other females. Oliveira

BLBK374-13

BLBK374-Brown

280

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

et al. (1998) showed experimentally that given the opportunity to observe fights between conspecifics, Siamese fighting fish (Betta splendens, Osphronemidae) attack ‘losers’ in a previous fight more vigorously than ‘winners’. In this experiment, both observed fishes were actually winners in fights with two other conspecifics that were hidden from the observer’s perspective; thus, to the observer it looked like the two winners were interacting, and the one that stopped threat behaviour first was the ‘loser’. Similar results were obtained by Earley & Dugatkin on fighting assessment in swordtails (Xiphophorus helleri, Poeciliidae) (Earley & Dugatkin 2002; and see Earley & Dugatkin 2005 on other poeciliid fishes). Eavesdropping should generate selection for behavioural changes in the individuals that are being observed (so-called ‘audience effects’), as the outcome of the current interaction will affect future outcomes. Audience effects could be achieved genetically through a general increase/decrease in the frequency of behaviour like aggression (Johnstone 2001). However, Siamese fighting fish solve the problem in a smart way, i.e. they increase aggression only if observers are present but not if they are absent (Dutreland et al. 2001). Observations of interactions between third parties and audience effects also play a major role in cleaning mutualism (Bshary 2002a, and see Subsection 13.2.4). To summarise so far, fishes have several cognitive abilities that are necessary for cunning Machiavellian-intelligence-like behaviour: They can recognise each other on an individual basis and adapt their behaviour within a communication network by both monitoring relationships between third parties and by adjusting own behaviour to the presence of bystanders. Subsection 13.2.2 concentrates on three systems that have yielded results that may justify the application of the Machiavellian intelligence hypothesis to fishes. The first example concerns predator inspection, the second one the group-living cichlids in Lake Tanganyika and the final example deals with marine cleaning mutualism.

13.2.2

Predator inspection

This topic is also discussed in Chapter 12 on intraspecific cooperation in fishes. Therefore, I will just summarise briefly some key findings to demonstrate that they are interesting for the Machiavellian intelligence hypothesis. Predator inspection is probably the most famous cooperative behaviour described in fishes. It involves individuals, pairs or several individuals leaving the safety of a shoal to inspect a nearby predator (Pitcher et al. 1986; Magurran & Higham 1988). During inspection, pairs of three-spined sticklebacks (Gasterosteus aculeatus, Gasterosteidae) and guppies (Poecilia reticulata, Poeciliidae), among others, approach the predator in alternating moves. Although the exact game structure is still debated, inspection seems to fit a Prisoner’s Dilemma game (Luce & Raiffa 1957), in the sense that cheating a partner by lagging behind seems to be a profitable option as predators are more likely to attack the leading individual (Milinski et al. 1997). Croft et al. (2005) found that pairs of guppies that frequently engaged in predator inspection did so in a more cooperative way, exchanging lead position more often than other pairs. Given the increased risk of predation for the lead fish when inspecting in a pair, changes in the lead position strongly suggest an act of cooperation based on reciprocity, in a manner predicted by the ‘Tit-for-Tat’ strategy. Two results yield important insights about the cognitive abilities underlying predator inspection in fishes. First, Milinski et al. (1990a) showed that individual sticklebacks prefer specific partners to others. This result implies that (a) school

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

281

members recognise each other, and (b) there are better (more cooperative) partners than others. Second, partners build up ‘trust’ in each other if they have cooperated repeatedly (Milinski et al. 1990b). The two fishes were actually in different aquaria, and an opaque partition would make the partner disappear all of a sudden (= cheating), while the removal of the partition would make the partner cooperative from the focal individual’s point of view. In the decisive experiment, all partners seemed to cheat because the partition was always present. Nevertheless, fishes approached a predator more closely when accompanied by a partner that had ‘cooperated’ in the past than when accompanied by a partner that had ‘cheated’ in the past. In addition, fishes also take into consideration what a partner currently does, as demonstrated by Milinski (1987). In this experiment, individual sticklebacks had their mirror image as a partner. In one condition, the mirror was placed parallel to the direction of inspection, mimicking a partner that invariably matches the behaviour of the focal individual. In the other conditions, the mirror was placed at an angle so that the mirror image suggested that the partner was lagging behind during inspections. This latter situation led to larger inspection distances. Taken together, the experiments demonstrate that sticklebacks fine-tune their behaviour by using information both on current behaviour of co-inspecting individuals and on how these have behaved in the past. The latter result implies that they are capable of bookkeeping (remembering their partners’ behaviour during previous interactions), with several partners simultaneously.

13.2.3

Group-living cichlids

Stable social groups are best known for cichlids of the great African lakes (Keenleyside 1991). Cooperative breeding, i.e. the presence of individuals that help the breeding pair raise its offspring, has been described in eight species of fishes to date (Taborsky 1994), of which six species are Lamprologine cichlids and endemic to Lake Tanganyika. In the best-studied species, the Princess of Burundi (Neolamprologus pulcher/brichardi, Cichlidae), there are, on average, five helpers of both sexes and of various sizes (Taborsky & Limberger 1981; Balshine et al. 2001). Helpers may be related or unrelated to the breeding pair, as breeding individuals are often replaced from outside (Taborsky & Limberger 1981) and because helpers may switch between groups (Stiver et al. 2004; Bergm¨uller et al. 2005a; Dierkes et al. 2005). Another cichlid, Neolamprologus multifasciatus (Cichlidae), endemic to Lake Tanganyika, lives in extended family groups (sensu Emlen 1997); i.e. in stable groups with two or more sexually active members of both sexes (Kohler 1997). In these social species, individual recognition of group members is very likely, and experimentally shown for N. brichardi (Hert 1985; Balshine-Earn & Lotem 1998). From a Machiavellian intelligence perspective, these systems are very interesting because there are several important conflicts between individual group members, which should promote social intelligence. Conflicts occur over: (1) Various group tasks such as sand digging, nest and egg/larvae maintenance, and defence against competitors or predators; (2) snail shells or crevices that are used for shelter; and (3) most importantly, over reproduction. In cooperatively breeding species, helpers have some reproductive success at the expense of the dominant fish (Dierkes et al. 1999, 2008), and in N. multifasciatus, males and females may disagree over the optimal group composition. This is because males could benefit from the presence of additional females under certain circumstances, because it would increase their reproductive success

BLBK374-13

BLBK374-Brown

282

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(Kohler 1997). Moreover, the costs of helping are measurable. Sexually mature helpers face various costs when delaying dispersal: Reduced growth rates (Taborsky 1984; Heg et al. 2004); delayed reproduction; and increased energy expenditure as a result of helping and from costly social interactions (Grantner & Taborsky 1998; Taborsky & Grantner 1998). In conclusion, any (genetic or cognitive) factor that helps to reduce these costs or increases reproductive success might be expected to be favoured by natural selection. A variety of interesting behaviours and strategies have been found in the Neolamprologus cichlids. Schradin & Lamprecht (2000) showed experimentally that N. multifasciatus males actively intervene in female–female aggression in favour of the unfamiliar female, and that this intervention increased the probability that the new females would settle in the group. Males of the dwarf cichlid (Apistogramma trifasciata, Cichlidae) seem to behave similarly (Burchard 1965). A general feature of Neolamprologus species is that subordinates frequently show submissive behaviours towards high-ranking group members (Kohler 1997; Bergm¨uller et al. 2005a). This submissive behaviour apparently functions as pre-emptive conflict avoidance (Bergm¨uller & Taborsky 2005). In N. pulcher/brichardi, strategic options are highly fine-tuned and variable: 1. Dominants are more likely to tolerate large helpers if competition for space with other species is very high (Taborsky 1985). 2. Temporarily removed helpers assisted more in territory maintenance and defence and visited the brood chamber more often after they were returned (Balshine-Earn et al. 1998) and helpers that could potentially breed independently reduced helping and submissive behaviour in the home territory (Bergm¨uller et al. 2005b). 3. In the field, Balshine-Earn et al. (1998) found that residents attacked temporarily removed helpers when the latter were returned to their groups’ territories. Thus, there is some social pressure on individuals to contribute to group tasks. 4. Helpers often visit neighbouring territories and may either switch or reduce working load at home if conditions outside the home territory are favourable (Bergm¨uller et al. 2005a). 5. There is a positive correlation between the amount of submissive behaviour performed by helpers and the amount of aggression they receive from dominants. In an experiment, helpers showed less submissive behaviour towards breeders per received aggression if they had defended the group than if they were prevented from doing so (Bergm¨uller & Taborsky 2005). The authors concluded that helping and submissive behaviour are at least partly interchangeable options for the same goal: Appeasement to avoid being punished or evicted from the group. 6. Experimental manipulations revealed that unrelated helpers responded to territory owner’s aggression with increased territorial defence if perceived density of alternative helpers was high and with mere submissive gestures if perceived density of alternative helpers was low (Bruintjes & Taborsky 2008). 7. Small helpers showed less submissive behaviour towards large helpers when they helped more, indicating that there are also conflicts between the helpers over individual contributions to group tasks (Bergm¨uller & Taborsky 2005). Further conflicts between helpers are a result of apparent competition for space (Balshine-Earn et al. 1998; Werner et al. 2004).

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

283

In summary, individual fishes in cooperatively breeding species seem to be capable of monitoring the behaviour of several group members as well as the social dynamics of neighbouring groups. They use this information to fine-tune their behaviour. Appeasement behaviour is an interesting component of interactions, as the behaviour itself provides no direct benefits to the recipient, so subordinates seem to use it to manipulate the decisions of dominant group members. But then why do dominants accept the appeasement behaviour of the subordinates? Bender et al. (2006) analysed both cortisol levels as an indicator of stress and androgen levels as indicators of reproductive activity in territorial males and helpers. They found a correlation between frequent submissive behaviour and low androgen levels, suggesting that submissive behaviour is an honest indicator of reduced sexual competition.

13.2.4

Machiavellian intelligence in cleaning mutualisms

In a cleaning mutualism, so-called client fish trade the removal of parasites and dead or infected tissue for an easy meal for so-called ‘cleaner fish’ (Losey et al. 1999; Cˆot´e 2000; Bshary & No¨e 2003; Bshary & Cˆot´e 2008). The best-studied cleaner fish species, Labroides spp. and Elacatinus spp., have small territories (‘cleaning stations’) that clients actively visit. Over the last few years, plenty of evidence has accumulated suggesting that these interactions are indeed mutualistic (Grutter 1999; Cheney & Cˆot´e 2001; Bshary 2003; Grutter et al. 2003; Bshary et al. 2007). While the cleaning goby mutualism appears to be rather free of conflicts (Soares et al. 2008), research on the cleaner wrasse revealed that there are a variety of potential conflicts between cleaners and clients, namely the risk for cleaners of being eaten by predatory clients (Trivers 1971), timing and durations of interactions (Bshary 2001), partner choice (Bshary & Grutter 2002a) and the cleaners’ preference for at least some clients’ mucus over ectoparasites (Grutter & Bshary 2003, 2004). The latter point is crucial because it shows that interactions are not a simple by-product mutualism (Brown 1983) but clients need to control the behaviour of cleaner fish in order to prevent being cheated (‘cleaner eats mucus’) and to make cleaners cooperate (‘cleaner eats ectoparasites’). The control mechanism depends critically on the strategic options of the clients: Predators may retaliate by eating a cheating cleaner, while non-predatory clients lack this option, and visiting clients with access to several cleaning stations may switch partner if cheated, while resident clients with access to the local cleaner only lack that option. As predicted by a generalised and phenomenon-based Machiavellian intelligence hypothesis, the social conflicts between cleaners and clients and the variety of strategic options have promoted an array of interesting behaviour that raises questions about the underlying cognitive abilities. 13.2.4.1

Categorisation and individual recognition of clients

Observations on interactions suggest that cleaner wrasse (Labroides dimidiatus, Labridae) discriminate between three client categories (Bshary 2001). First, they discriminate between predatory and non-predatory client species, the former being virtually never cheated, in contrast to the latter. Among the non-predatory clients, cleaners further distinguish between resident clients and visiting clients, giving the latter priority of service. This is advantageous because visiting clients would not wait for inspection and would visit a different cleaning station instead, while residents simply have to wait if they want to be serviced at all

BLBK374-13

BLBK374-Brown

284

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(Bshary & Sch¨affer 2002). An important difference between visitors and residents is their response to cleaner fish cheating: Visitors simply swim off, while residents chase the cleaners around (Bshary & Grutter 2002a). The aggression functions as punishment (sensu Clutton-Brock & Parker 1995). Aggressive chasing bears immediate energetic costs to both client and cleaner but it makes cleaners more cooperative in future interactions. According to theory, punishment can function only if there is individual recognition (Ostrom 1990). Choice experiments with one-way mirrors confirmed that cleaners can recognise clients individually – in the absence of environmental clues, they spent more time near a familiar client than near an unfamiliar client of the same species and of similar size (Tebbich et al. 2002). Individual cleaners may interact with more than 100 individual resident clients belonging to various species (R. Bshary, unpublished observations). Thus, they may well be capable of recognising more than 100 clients on an individual basis and to remember their last interaction with each of them (see Subsection 13.2.4.2), although this clearly has to be tested in detail in a future study. The closely related cleaner wrasse L. bicolor differs from L. dimidiatus in that it roves over much larger areas and consequently seeks clients more actively (Oates 2010a). Roving works against a repeated game structure, and indeed L. bicolor individuals cheat more frequently (Oates et al. 2010a). However, L. bicolor individuals have core areas in their home ranges, where delays between successive interactions are relatively short. The cleaners apparently incorporate location into service quality as they are more cooperative in core areas than elsewhere, thus taking into account ‘the shadow of the future’ (Oates et al. 2010b). A potential role of individual recognition has not yet been investigated. 13.2.4.2

Building up relationships between cleaners and resident clients

If cleaner wrasse are experimentally translocated to a new site, resident clients first chase them around (Bshary 2002b). The cleaners spend most of the first day exhibiting very peculiar behaviour, which has been observed so far only among cleaners of the genus Labroides: They ride on the clients’ back, snout pointing into the blue while applying tactile stimulation to the clients’ back with the pelvic and pectoral fins (Potts 1973). Only with time does client aggression decrease and the cleaner start to inspect their resident clients for food (Bshary 2002b). Thus, cleaners have to build up relationships with their resident clients, through initial investment rather than through increasing investment (Roberts & Sherratt 1998). Interactions with visiting clients, in contrast, are ‘normal’ right from the beginning, while predators seem to receive some extra tactile stimulation (see Subsection 13.2.4.3) compared to what they normally receive (Bshary 2002b). 13.2.4.3 Use of tactile stimulation by cleaners to manipulate client decisions and reconcile after conflicts As mentioned in Subsection 13.2.4.2, cleaners of the genus Labroides apply tactile stimulation to the clients’ back (Potts 1973). Cleaner wrasse use this massage in particular: During interactions with predatory clients; to make moving clients stop so that the cleaner can inspect them; and functionally to reconcile (de Waal & van Roosmalen 1979) with resident clients in follow-up interactions after client punishment (Bshary & W¨urth 2001). The specific behaviour of cleaners in follow-up interactions makes it plausible that they

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

285

really can recognise their clients and remember their last interaction with each of them. Tactile stimulation during interactions with predators can be seen as pre-conflict management (Aureli & de Waal 2000) that reduces the probability of predators cheating (see also Grutter 2004). The benefits of manipulating predators extend to clients that could be prey; predators are less likely to attack prey with cleaners around and the less so if the cleaners provide predators with tactile stimulation (Cheney et al. 2008). Finally, the application of tactile stimulation to swimming clients is a means to manipulate the clients’ behaviour in favour of the cleaner as it gains access to a food source. 13.2.4.4

Audience effects in response to image scoring and tactical deception

Field observations revealed that clients arriving at a cleaning station invite inspections if the cleaner’s current interaction ends without apparent conflict, but avoid an interaction if the current client flees or chases the cleaner (Bshary 2002a). Thus, the clients seem to observe ongoing interactions and extract information on the behaviour of the cleaner fish, which they use to make their decision on whether to invite inspection. In other words, cleaners are given an image score or a social prestige (Alexander 1987; Zahavi 1995; Nowak & Sigmund 1998; Lotem et al. 2003) from bystanders or eavesdroppers (McGregor 1993) that is positive if they cooperate and negative if they cheat. As a consequence of image-scoring clients, cleaners seem to be more cooperative to their current client if eavesdropping clients are around (Bshary & D’Souza 2005). The benefits of this more cooperative behaviour are not accrued directly by the interacting client but indirectly, via access to the observing clients. The main reason why clients image score might be that some cleaners bite rather than clean (Bshary 2002a), and these biting cleaners can be avoided by image scoring without having to make a personal negative experience first. Image scoring is thus immediately self-serving. This contrasts with image scoring in humans where bystanders are willing to pay in order to help cooperative individuals (Wedekind & Milinski 2000). Therefore, the cleaner fish example is a case of indirect pseudo-reciprocity (Bshary & Bergm¨uller 2008): Cleaners invest in current clients to benefit from self-serving decisions of bystanders, a scenario that was also confirmed experimentally (Bshary & Grutter 2006). As a consequence of image scoring, clients biting cleaners lose access to some potential victims but have found a way partly to cope – they may raise their image by seeking interactions with small residents and giving them tactile stimulation (Bshary 2002). These interactions are purely costly for the cleaners but benefits might be accrued through deceiving larger clients into visiting so that they can be exploited. Giving tactile stimulation to small residents thus functions as tactical deception (Hauser 1998) – it is a signal out of context, directed at observing clients (rather than at the small resident that receives the stimulation) to make them choose an option that is costly to them but beneficial to the cleaner. 13.2.4.5

Punishment by males during pair inspections

Cleaner wrasses L. dimidiatus are protogynous hermaphrodytes: They start their reproductive career as females and eventually switch sex and become males. Males typically defend a harem (Robertson 1972) and spend most of the time at the cleaning station of their largest female. There, large clients are often inspected together, which leads to an interesting conflict of interest: As clients often leave in response to cheating by a cleaner

BLBK374-13

BLBK374-Brown

286

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

the costs of leaving are shared between partners while only one had the benefit of biting. A game theoretic analysis shows that the problem is akin to an iterated Prisoner’s Dilemma (Bshary et al. 2008). Cleaner pairs solve the dilemma but not in a ‘fair’ way. Males use their size advantage and punish females through aggressive chasing if the latter cheated (Bshary et al. 2008). Experiments demonstrate that the chasing causes females to behave more cooperatively during future joint inspections, which benefits both the male and future clients (Raihani et al. 2010). Such behaviour by males is probably linked to the fact that the female may become his future rival, which causes conflicts of interest. In cleaning gobies, a species without sex change and stable male–female pairs, conflicts of interests are largely absent and coercion does not occur (Soares et al. 2009).

13.3

Evidence for cognitive mechanisms in fishes

In this section, I go beyond the description of behavioural interaction patterns and ask what cognitive abilities may produce the observed behaviours. This part is still rather preliminary, and plenty of progress could still be made in putting together the cognitive toolbox of particular species and of fishes in general.

13.3.1 What cognitive abilities might cleaners need to deal with their clients? As developed in Section 13.2.4, cleaners can recognise clients individually, categorise clients according to predator, resident and visitor and they can remember the outcome of past interactions with specific clients. Also, their behaviour appears to be fine-tuned to specific circumstances. There are good arguments to assume that these fine-tuned behavioural tactics are the result of associative learning, without invoking any deeper understanding associated with theory of mind. This is because a cleaner wrasse has about 2000 interactions per day (Grutter 1995) and, therefore, constantly receives feedback about its actions. Laboratory experiments confirmed that cleaners can easily learn to behave in a way that increases their foraging success (Bshary & Grutter 2002b, 2005). In these experiments, cleaners interacted with Plexiglas plates attached to levers that allowed them to make the plates ‘behave’ in certain ways. In one set of experiments, cleaners learned to choose the less preferred food items in order to stop the plates moving away or chasing them (Bshary & Grutter 2005). This strongly suggests that cleaners have some self-control (as opposed to impulsiveness), defined as the ability to inhibit a natural tendency to reach for the greater or more attractive of two food items (Anderson 2001). This ability has been linked to self-awareness in primates (Genty et al. 2004). The typical experimental set-up to test for self-control is the reverse-reward-contingency-task, where subjects can choose between two food quantities. They will receive the opposite of what they chose and should therefore choose the lesspreferred option. Danisman et al. (2010) tested eight cleaner wrasses in this paradigm. Like most primates (Genty et al. 2004) they all failed in the initial task, while one individual learned the solution with some extra trials in which choosing the large reward yielded no food at all. While one individual out of eight is not a great number and without discussing in detail potential methodological issues, the results show that cleaner wrasses are at least in principle capable of some self-control (Danisman et al. 2010). In another experiment, cleaners learned to inspect a ‘visiting’ plate before a ‘resident’ plate because the former

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

287

would not wait for inspection while the latter would (Bshary & Grutter 2002b). There were equal amounts of food on both plates and cleaners could invariably feed on the plate they visited first. So they learned to prefer the visitor plate even though they were rewarded for any action they took. Furthermore, field observations of one particular cleaner female revealed that a cleaner’s behaviour could be completely independent of its internal state. This cleaner fish female was biting clients and, therefore, was not tolerated by the larger male at the cleaning station. Therefore, the female installed her station 3 m away but regularly visited the male’s area until chased away. It turned out that she was a ‘normal’ cleaner while at her station and a biting cleaner whenever she visited the male’s area (Bshary & D’Souza 2005). Thus, cleaners might be able to adjust their behaviour to a specific external context (‘interacting with the male’s client or with own client’) rather than merely behaving according to an internal state (‘if I am hungry I bite’). L. bicolor individuals certainly adjust service quality to the location in their home range (Oates et al. 2010b). Finally, cleaners seem to have some knowledge of the interspecific relationships between their clients. They exploit the presence of a predatory client to prevent punishing clients from chasing them further. They swim to the predator and provide tactile stimulation while the aggressive client apparently does not dare to approach the predator that close and terminates its pursuit (Bshary et al. 2002).

13.3.2

Other cognitive mechanisms

An ability that is generally considered to be cognitively demanding is transitive inference. Transitive inference is used if an individual deduces from the observations that A > B and B > C so that it must follow that A > C. Grosenick et al. (2007) tested male cichlids (Astatotilapia burtoni) for their ability to infer a dominance hierarchy between neighbouring territory owners by using transitive inference. The hierarchy was constructed artificially by placing individuals into another male’s territory, which invariably caused the territory owner to win. With this technique, bystanders observed over successive days that A > B, B > C, C > D and D > E. In the experiments, the bystanders were confronted with individuals B and D at opposite ends of their tank, and the time spent near each individual was evaluated. Previous tests had shown that weaker males were preferred (Clement et al. 2005). As expected if fishes use transitive inference, individuals spent more time near the individual D (Grosenick et al. 2007). The ability to plan ahead is another ability of broad interest. In chimpanzees, for example, it has been noted that only individuals of one population signal to each other to start a cooperative hunt (Boesch & Boesch 1989). The critical aspect is that the decision to hunt is taken in the absence of prey, which has to be searched after the decision was made. Boesch & Boesch (1989) defined this as ‘intentional hunting’. According to this operational definition, also Red Sea coral groupers Plectropomus pessuliferus are capable of intentional hunting. These groupers often approach giant moray eels (Gymnothorax javanicus, Muraenidae) (which usually rest in crevices during daytime) and shake their head repeatedly (Bshary et al. 2006). In response, the morays might swim off with the grouper. Morays hunt in the crevices that are inaccessible to the groupers, while the groupers hunt in the water above the reef. Coordinating the hunt thus yields a double predation effect. The key issue is that groupers elicit joint hunting to individuals of a different species without having previously spotted potential prey. Only once the partners swim off together, they may attack a fish

BLBK374-13

BLBK374-Brown

288

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

that does not behave ‘properly’, while the other great numbers of fishes in the reef are ignored. More generally, cooperative hunting, in the sense that several predators hunt the same prey simultaneously, is widespread in fishes and may merit attention also because of the cognitive demands imminent in the coordination. Mackerels (Carangidae) have been described herding their prey (Hiatt & Brock 1948; Sette 1950). Schmitt & Strand (1982) even argued that in yellowtails (Seriola lalandei, Carangidae), individuals play different roles during such hunts (splitting the school of prey, herding the prey), and refrain from single hunting attempts until the prey is in a favourable position. In addition, Schmitt & Strand (1982) mention that the hunting strategies are variable and that they depend on the prey species. It is of interest that in cooperative hunting each individual has to monitor the movements of partners relative to the prey to bring itself into the best position for an attack. If individuals assume different roles during hunts, it would be helpful to know whether they specialise in different roles and whether there is some reciprocity between group members if different roles yield different success rates (Gazda et al. 2005). As a final point, recent studies on nine-spined sticklebacks have uncovered amazingly sophisticated decision rules based on social learning (van Bergen et al. 2004; Coolen et al. 2005; Kendal et al. 2009a). Social learning is presented elsewhere in much detail (Chapter 11 among others), and I will therefore focus only on an aspect that I consider to be essential for Machiavellian intelligence, namely the ability to identify suitable demonstrators for social learning. Social learning is not advantageous per se (Boyd & Richerson 1985; Giraldeau et al. 2002) but depend on the reliability of social information relative to private information (Kendal et al. 2009b). Laland (2004) distinguished between two classes of social learning strategy: ‘When’ strategies, which dictate the circumstances under which individuals copy others, and ‘who’ strategies, which specify from whom individuals learn. Coolen et al. (2005) tested food-patch choice in nine-spined sticklebacks when given social information about patch quality. They were allowed to observe food patches simultaneously, where one patch was visited by six conspecifics and the other one by only two. In half of the subjects, this public information conflicted with personal information as the subjects had received six food items at the poorly visited patch relative to two food items at the highly used patch. Sticklebacks used personal information if available and public information otherwise. An earlier experiment had demonstrated that these fishes rely on personal information only if this information is highly accurate and recent (van Bergen et al. 2004). Finally, Kendal et al. (2009a) demonstrated that nine-spined sticklebacks are able to directly compare their own foraging success with that of conspecifics and thereby learn to exploit the most profitable food patches. This ability is the basis for social learning leading to fitness maximising (‘hill climbing’) behaviour, a prerequisite for cumulative culture (Kendal et al. 2009a). At the same time, monitoring the behaviour and success of others and adjusting own behaviour accordingly is also the basis for successful Machiavellian strategies.

13.4

Discussion

Since the initial article on Machiavellian intelligence in fishes (Bshary 2006) the list of phenomena that are of interest with respect to the Machiavellian intelligence hypothesis has grown even longer while it is still true that most phenomena are poorly studied from

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

289

a cognitive perspective. Therefore, we still know little about the kind and quantity of information that fishes use to make these decisions. Nevertheless, the little we do know, and recent results indicate, that a wealth of interesting results awaits description by scientists. But it is evident that the preconditions for Machiavellian-type intelligence, i.e. individual recognition, living in groups/repeated interactions and knowledge about relations between other group members/third parties, is widespread in fishes. Therefore, it is probable that an array of interesting cognitive social abilities waits to be found and studied in more detail in fishes. The research on the cleaner fish mutualism shows that if we use phenomenonbased definitions, we find reconciliation, punishment, coercion, audience effects based on image scoring and tactical deception. The laboratory experiments show that cleaners can easily learn to respond adaptively to various behavioural responses of Plexiglas plates (thus excluding purely genetic strategies). They also show some self-control (feed against their preference) and are flexible in their decisions rather than just dependent on their internal state. The various studies on cooperative breeding in fishes suggest that cichlids are also capable of fine-tuned adjustment of behaviour to ecological and social factors, and that social skills may enhance survival and reproductive success. The studies on nine-spined sticklebacks demonstrate a sophistication in decision rules that is hitherto unmatched outside humans. While this may be due to a lack of decisive experiments in other taxa, it still demonstrates the high level of cognitive abilities that can be found in fishes.

13.4.1

Future avenues I: How Machiavellian is fish behaviour?

Cognitive research on Machiavellian aspects of social behaviour could address detailed questions about the precise mechanism that is used for learning and decision-making, as well as producing quantitative information on the following: 1. 2. 3. 4.

The number of individuals that can be known and their behaviours tracked. How fast individual fish can learn social tasks. How accurately they can learn to behave. How fast they can integrate new additional or contradicting information.

These questions could be embedded in a comparative approach that compares many different species that differ with respect to the complexity of their social life (Kamil 1998). A comparison between two species might yield significant differences that are in line with predictions based on social complexity. However, alternative explanations for the differences cannot be excluded. Only if enough species are tested on the same question to allow statistical analysis on the number of species investigated, can we be able to find potential links between social complexity and cognitive capacities. A highly relevant study on corvids with respect to learning mechanisms and speed of knowledge acquisition was produced by Bond et al. (2003). They compared semi-territorial western scrub jays (Aphelocoma californica, Corvidae), with the highly social pinyon jays (Gymnorhinus cyanocephalus, Corvidae). The birds had to learn an arbitrary dominance hierarchy between symbols to obtain food. The more social pinyon jays learned some relations between symbols through transitive inference (if A > B and B > C, then A > C), while western scrub jays seem to have learned each pair through associative learning. The use of these different

BLBK374-13

BLBK374-Brown

290

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

mechanisms may explain why the pinyon jays learned the tasks faster than the western scrub jays (Bond et al. 2003). Furthermore, it is argued that the capacity of transitive inference is more important in large social groups in order to be able to track relationships between other group members. Data on more species are needed to test whether the difference in leaning mechanism is indeed linked to the differences in sociality. Now that we know that a fish species can use transitive inference for decision-making (Grosenick et al. 2007) we can expand the number of species tested and evaluate potential links between ecology and cognitive mechanisms in more detail. Fishes seem to be particularly suitable for study of interspecific social cognition in detail, as they often occur in mixed-species shoals (Krause 1993; and see Chapter 10), forage in mixed-species parties (Dugatkin & Godin 1992b), hunt with interspecific partners (Bshary et al. 2006) and often have cleaning interactions (Feder 1966). Often, closely related species differ markedly with respect to the importance of social (interspecific) interactions, offering ideal opportunities for large-scale interspecific comparisons that link a species’ social environment and factors like habitat structure, diet or antipredator behaviour to specific cognitive capacities. Taking cleaning mutualism as an example, cleaner fishes are found in many different fish families and can differ markedly in the degree to which they depend on interactions with clients for their diet (Feder 1966). Therefore, it would be interesting to investigate whether the degree of cleaning ‘professionalism’ correlates with the cognitive abilities of species in social contexts. Experiments using Plexiglas plates (Bshary & Grutter 2002b, 2005) could be used to test a variety of cleaning and non-cleaning species. First preliminary results (Bergm¨uller et al. unpublished) indicate that cleaning indeed affects the ability to solve a variety of tasks. Three key questions have to be addressed. First, seemingly cunning social behaviour might have nothing to do with learning/cognition but is produced with complex chains of key stimuli and automatic responses (Bronstein et al. 2010), or with behaviour modified through endocrine responses (Oliveira 2005; Soares et al. 2010). So we have to show that learning plays a role. Second, any differences between species with respect to performance in the same task might be of a quantitative nature (speed of learning, precision of behaviour). How widespread are more complex cognitive mechanisms like transitive inference? Third, the social tasks could be transferred into environmental tasks to investigate whether social cognitive abilities correlate with environmental cognitive abilities or whether these abilities exist independently of each other. In a classic study on food caching in three corvid bird species (Balda & Kamil 1989; review by Balda & Kamil 1998), the amount of food caching correlated well with a species’ spatial memory capacity but not with colour memory.

13.4.2 Future avenues II: Relating Machiavellian-type behaviour to brain size evolution Currently, we do not know whether seemingly complex social behaviour in fishes is correlated with the size of relevant brain areas, as has been documented in several mammalian taxa and in birds (Barton & Dunbar 1997; Burish et al. 2004). Apparently, the telencephalon is an important area for higher decision-making processes and thus a prime candidate for future investigations (Roth & Wullimann 2001; and see Chapter 15). With experiments on three-spined sticklebacks, Sch¨onherr (1954) and Seegar (1956) could show that after the

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

291

lesion of the telencephalon, all behaviour that is necessary to build a nest is still performed but the proper coordination necessary to build the nest is lacking. Also, the mesencephalon seems to be, at least in part, involved in learning and decision making (Healey 1957). The link between the ecology of fish species and brain size has already been investigated (reviewed by Kotrschal et al. 1998). Cichlids of the East African Lakes vary greatly in the size of their telencephalon (van Staaden et al. 1995), but this variation has not yet been linked to the complexity of a species’ social life (but see Gonzalez-Voyer et al. 2009 for whole brain size). Cichlids are in general a very promising fish family for study of the relation between social organisation and brain anatomy, because of the large interspecific variance in the amount of brood care and corresponding social organisation. In addition, they also live in a great variety of habitats, and there is evidence that spatial complexity of habitat correlates with telencephalon size (van Staaden et al. 1995; Huber et al. 1997). While Gonzalez-Voyer et al. (2009) have shown that brain size is related to the type of parental care and to diet in females, nothing is known about the relative size of specific brain parts and the cognitive ability of the fish tested. What is still missing, then, is the link between such information on fish ecology, brain structure and cognitive skills, which includes social complexity as a parameter. With respect to intraspecific variation in social cognitive skills and the link to brain anatomy, fishes seem to be very suitable subjects. In particular, effects of ontogeny should become evident in brain structure as fishes can generate new brain cells throughout their entire life (reviewed in Kotrschal et al. 1998).

13.4.3 Extending the Machiavellian intelligence hypothesis to general social intelligence The Machiavellian intelligence hypothesis currently puts the emphasis on the Prisoner’s Dilemma-type cooperation (where deception yields higher short-term benefits), deception and manipulation. However, group living may also promote the cognitive abilities for social learning that are generally of a cooperative nature. Aspects of social learning are discussed in several chapters of this volume (Chapters 3, 4, 10 and 11). These chapters provide plenty of evidence that fishes learn socially, although true imitation of tutor behaviour has not yet been documented. Nevertheless, social life in fishes may select as much for the evolution of cognitive capacities, which allow the acquisition of freely available social information, as it does for Machiavellian intelligence.

Acknowledgements I thank Culum Brown, Kevin Laland and Jens Krause for inviting me to write this chapter and their comments on the manuscript. My research is currently supported by the Swiss National Science Foundation.

References Alexander, R.D. (1987) The Biology of Moral Systems. Aldine de Gruiter, New York. Anderson, J.R. (2001) Self- and other-control in squirrel monkeys. In: T. Matsuzawa (ed) Primate Origins of Human Cognition and Behavior, pp. 330–347. Springer-Verlag, Tokyo.

BLBK374-13

BLBK374-Brown

292

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Aureli, F. & de Waal, F.B.M. (2000) Natural Conflict Resolution. Cambridge University Press, Cambridge. Balda, R.P. & Kamil, A.C. (1989) A comparative study on cache recovery by three corvid species. Animal Behaviour, 38, 486–495. Balda, R.P. & Kamil, A.C. (1998) The ecology and evolution of spatial memory in corvids of the southwestern USA: the perplexing pinyon jay. In: R.P. Balda, P.A. Bednekoff & A.C. Kamil (eds) Animal Cognition in Nature, pp. 29–64. Academic Press, New York. Balshine, S., Leach, B., Neat, F.C., Reid, H., Taborsky, M. & Werner, N. (2001) Correlates of group size in a cooperatively breeding cichlid fish. Behavioral Ecology and Sociobiology, 50, 134–140. Balshine-Earn, S. & Lotem, A. (1998) Individual recognition in a cooperatively breeding cichlid: evidence from video playback experiments. Behaviour, 135, 369–386. Balshine-Earn, S., Neat, F.C., Reid, H. & Taborsky, M. (1998) Paying to stay or paying to breed? Field evidence for direct benefits of helping behaviour in a cooperatively breeding fish. Behavioral Ecology, 9, 432–438. Barton, R.A. & Dunbar, R.I.M. (1997) Evolution of the social brain. In: A. Whiten & D.W. Byrne (eds) Machiavellian Intelligence II, pp. 240–263. Cambridge University Press, Cambridge. Bender, N., Heg, D., Hamilton, I.M., Bachar, Z., Taborsky, M. & Oliveira, R.F. (2006) The relationship between social status, behaviour, growth and steroids in male helpers and breeders of a cooperatively breeding cichlid. Hormones and Behavior, 50, 173–182. Bergm¨uller, R., Heg, D., Peer, K. & Taborsky, M. (2005a) Extended safe havens and between group dispersal of helpers in a cooperatively breeding cichlid. Behaviour, 142, 1643–1667. Bergm¨uller, R., Heg, D. & Taborsky, M. (2005b) Helpers in a cooperatively breeding cichlid fish stay and pay, but choose to leave when independent breeding options are available. Proceedings of the Royal Society of London Series B – Biological Sciences, 272, 325–331. Bergm¨uller, R. & Taborsky, M. (2005) Experimental manipulation of helping in a cooperative breeder: helpers ‘pay-to-stay’ by pre-emptive appeasement. Animal Behaviour, 69, 19–28. Boesch, C. & Boesch, H. (1989) Hunting behaviour of wild chimpanzees in the Ta¨ı National Park. American Journal of Physical Anthropology, 78, 547–573. Bond, A.B., Kamil, A.C. & Balda, R.P. (2003) Social complexity and transitive inference in corvids. Animal Behaviour, 65, 479–487. Boyd, R. & Richerson, P.J. (1985) Culture and the Evolutionary Process. University of Chicago Press, Chicago. Br¨auer, J., Call, J. & Tomasello, M. (2004) Visual perspective taking in dogs (Canis familiaris) in the presence of barriers. Applied Animal Behaviour Science, 88, 299–317. Brown, J.L. (1983). Cooperation: a biologist’s dilemma. In: J.S. Rosenblatt (ed) Advances in the Study of Behaviour, pp. 1–37. Academic Press, New York. Brosnan, S.F., Salwiczek, L. & Bshary, R. (2010) The interplay of cognition and cooperation. Philosophical Transactions of the Royal Society, Series B – Biological Sciences, 365, 2699–2710. Bruintjes, R. & Taborsky, M. (2008) Helpers in a cooperative breeder pay a high price to stay: effects of demand, helper size and sex. Animal Behaviour, 75, 1843–1850. Bshary, R. (2001) The cleaner fish market. In: R. No¨e, J.A.R.A.M. van Hooff & P. Hammerstein (eds) Economics in Nature, pp. 146–172. Cambridge University Press, Cambridge. Bshary, R. (2002a) Biting cleaner fish use altruism to deceive image scoring clients. Proceedings of the Royal Society of London Series B – Biological Sciences, 269, 2087–2093. Bshary, R. (2002b) Building up relationships in asymmetric cooperation games between the cleaner wrasse Labroides dimidiatus and client reef fish. Behavioral Ecology and Sociobiology, 52, 365–371. Bshary, R. (2003) The cleaner wrasse, Labroides dimidiatus, is a key organism for reef fish diversity at Ras Mohammed National Park, Egypt. Journal of Animal Ecology, 72, 169–176. Bshary, R. (2006) Machiavellian intelligence in fishes. In: C. Brown, K.N. Laland & J. Krause (eds) Learning and Cognition in Fishes, pp. 223–242. Blackwell Publishing Ltd., Oxford. Bshary, R. & Bergm¨uller, R. (2008) Distinguishing four fundamental approaches to the evolution of helping. Journal of Evolutionary Biology, 21, 405–420.

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

293

Bshary, R. & Cˆot´e, I.M. (2008) New perspectives on marine cleaning mutualism. In: C. Magnhagen, V.A. Braithwaite, E. Forsgren & B.G. Kapoor (eds) Fish Behaviour, pp. 563–592. Science Publishers, Enfield. Bshary, R. & D’Souza, A. (2005) Cooperation in communication networks: indirect reciprocity in interactions between cleaner fish and client reef fish. In: P.K. McGregor (ed) Communication Networks, pp. 521–539. Cambridge University Press, Cambridge. Bshary, R. & Grutter, A.S. (2002a) Asymmetric cheating opportunities and partner control in a cleaner fish mutualism. Animal Behaviour, 63, 547–555. Bshary, R. & Grutter, A.S. (2002b) Experimental evidence that partner choice is the driving force in the payoff distribution among cooperators or mutualists: the cleaner fish case. Ecology Letters, 5, 130–136. Bshary, R. & Grutter, A.S. (2005) Punishment and partner choice cause cooperation in a cleaning mutualism. Biology Letters, 1, 396–399. Bshary, R. & Grutter, A.S. (2006) Image scoring causes cooperation in a cleaning mutualism. Nature, 441, 975–978 Bshary, R., Grutter, A.S., Willener, A.S.T. & Leimar, O. (2008) Pairs of cooperating cleaner fish provide better service quality than singletons. Nature, 455, 964–967. Bshary, R., Hohner, A., Ait-El-Djoudi, K. & Fricke, H. (2006) Interspecific communicative and coordinated hunting between groupers and giant moray eels in the Red Sea. PLoS Biology, 4, e431. Bshary, R. & No¨e, R. (2003) Biological markets: the ubiquitous influence of partner choice on cooperation and mutualism. In: P. Hammerstein (ed) Genetic and Cultural Evolution of Cooperation, pp. 167–184. MIT Press, Cambridge. Bshary R, Oliveira RF, Oliveira T & Canario AVM (2007) Do cleaning organisms reduce the stress response of client reef fish? Frontiers in Zoology, 4, 21. Bshary, R. & Sch¨affer, D. (2002) Choosy reef fish select cleaner fish that provide high service quality. Animal Behaviour, 63, 557–564. Bshary, R., Wickler, W. & Fricke, H. (2002) Fish cognition: a primate’s eye view. Animal Cognition, 5, 1–13. Bshary, R. & W¨urth, M. (2001) Cleaner fish Labroides dimidiatus manipulate client reef fish by providing tactile stimulation. Proceedings of the Royal Society of London Series B – Biological Sciences, 268, 1495–1501. Bugnyar, T., St¨owe, M. & Heinrich, B. (2004) Ravens, Corvus corax, follow gaze direction of humans around obstacles. Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 1331–1336. Burchard, J.E. (1965) Family structure in the dwarf cichlid Apistogramma trifasciatum. Zeitschrift f¨ur Tierpsychologie, 22, 150–162. Burish, M.J., Kueh, H.Y. & Wang, S.S.-H. (2004) Brain architecture and social complexity in modern and ancient birds. Brain, Behavior and Evolution, 63, 107–124. Byrne, R.W. & Whiten, A. (1988) Machiavellian Intelligence. Clarendon Press, Oxford. Cheney, K.L., Bshary, R. & Grutter, A.S. (2008) Cleaner fish cause predators to reduce aggression towards bystanders at cleaning stations. Behavioral Ecology, 19, 1063–1067. Cheney, K.L. & Cˆot´e, I.M. (2001) Are Caribbean cleaning symbioses mutualistic? Costs and benefits of visiting cleaning stations to longfin damselfish. Animal Behaviour, 62, 927–933. Clayton, N.S., Bussey, T.J., Emery, N.J. & Dickinson, A. (2003) Prometheus to Proust: the case for behavioural criteria for ‘mental time travel’. Trends in Cognitive Sciences, 7, 436–437. Clement, T.S., Grens, K.E. & Fernald, R.D. (2005) Female affiliative preference depends on reproductive state in the African cichlid fish, A. burtoni. Behavioral Ecology, 16, 83–88. Clutton-Brock, T.H. & Parker, G.A. (1995) Punishment in animal societies. Nature, 373, 209–215. Coolen, I., Ward, A.J.W., Hart, P. & Laland, K.N. (2005) Foraging nine-spined sticklebacks prefer to rely on public information over simpler social cues. Behavioral Ecology, 16, 865–870. Cˆot´e, I.M. (2000) Evolution and ecology of cleaning symbioses in the sea. Oceanography and Marine Biology: An Annual Review, 38, 311–355.

BLBK374-13

BLBK374-Brown

294

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Croft, D.P., James, R., Ward, A.J.W., Botham, M.S., Mawdsley, D. & Krause, J. (2005) Assortative interactions and social networks in fish. Oecologia, 143, 211–219. Danisman, E., Bshary, R. & Bergm¨uller, R. (2010) Do cleaner fish learn to feed against their preference in a reverse reward contingency task? Animal Cognition, 13, 41–49. Dierkes, P., Heg, D., Taborsky, M., Skubic, E. & Achmann, R. (2005) Genetic relatedness in groups is sex-specific and declines with age of helpers in a cooperatively breeding cichlid. Ecology Letters, 8, 968–975. Dierkes, P., Taborsky, M. & Achmann, R. (2008) Multiple paternity in the cooperatively breeding fish Neolamprologus pulcher. Behavioral Ecology and Sociobiology, 62, 1581–1589. Dierkes, P., Taborsky. M. & Kohler, U. (1999) Reproductive parasitism of broodcare helpers in a cooperatively breeding fish. Behavioral Ecology, 10, 510–515. Dugatkin, L.A. & Godin, J.G.J. (1992a) Reversal of female mate choice by copying in the guppy Poecilia reticulata. Proceedings of the Royal Society of London Series B – Biological Sciences, 249, 179–184. Dugatkin, L.A. & Godin, J.G.J. (1992b) Prey approaching predators: a cost-benefit perspective. Annales Zoologici Fennici, 29, 233–252. Dunbar, R.I.M. (1992) Neocortex size as a constraint on group size in primates. Journal of Human Evolution, 20, 287–296. Dutreland, C., McGregor, P.K. & Oliveira, R.F. (2001) The effect of an audience on intrasexual communication in male Siamese fighting fish, Betta splendens. Behavioral Ecology, 12, 283–286. Earley, R.L. & Dugatkin, L.A. (2002) Eavesdropping on visual cues in swordtail (Xiphophorus helleri) fights – a case for networking. Proceedings of the Royal Society of London Series B – Biological Sciences, 269, 943–952. Earley, R.L. & Dugatkin, L.A. (2005) Fighting, mating and networking: pillars of poeciliid sociality. In: P.K. McGregor (ed) Animal Communication Networks, pp. 84–113. Cambridge University Press, Cambridge. Emery, N.J. & Clayton, N.S. (2001) Effects of experience and social context on prospective caching strategies in scrub jays. Nature, 414, 443–446. Emery, N.J. & Clayton, N.S. (2004) The mentality of crows: convergent evolution of intelligence in corvids and apes. Science, 306, 1903–1907. Emlen, S.T. (1997) Predicting family dynamics in social vertebrates. In: J.R. Krebs & N.B. Davies (eds) Behavioural Ecology, 4th ed., pp. 228–253. Blackwell Publishing Ltd., Oxford. Feder, H.M. (1966) Cleaning symbiosis in the marine environment. Symbiosis, 1, 327–380. Fricke, H. (1975) Sozialstruktur und oekologische spezialisierung von verwandten fischen (Pomacentridae). Zeitschrift f¨ur Tierpsychologie, 39, 492–520. Gazda, S.K., Connor, R.C., Edgar, R.K. & Cox, F. (2005) A division of labour with role specialization in group-hunting bottlenose dolphins (Tursiops truncatus) off Cedar Key, Florida. Proceedings of the Royal Society of London Series B – Biological Sciences, 272, 135–140. Genty, E., Palmier, C. & Roeder, J.J. (2004) Learning to suppress responses to the larger of two rewards in two species of lemurs, Eulemur fulvus and E. macaco. Animal Behaviour, 67, 925–932. Giraldeau., L.-A., Valone, T.J. & Templeton, J.J. (2002) Potential disadvantages of using socially acquired information. Philosophical Transactions of the Royal Society of London Series B – Biological Sciences, 357, 1559–1566. Gonzalez-Voyer, A., Winberg, S. & Kolm, N. (2009) Social fishes and single mothers: brain evolution in African cichlids. Proceedings of the Royal Society of London Series B – Biological Sciences, 276, 161–167. Grantner, A. & Taborsky, M. (1998) The metabolic rates associated with resting, and with the performance of agonistic, submissive and digging behaviours in the cichlid fish Neolamprologus pulcher. Journal of Comparative Psychology, 168, 427–433. Grosenick, L., Clement, T.S. & Fernald, R.D. (2007) Fish can infer social rank by observation alone. Nature, 445, 429–432. Grutter, A.S. (1995) Relationship between cleaning rates and ectoparasite loads in coral reef fishes. Marine Ecology Progress Series, 118, 51–58.

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

295

Grutter, A.S. (1999) Cleaners really do clean. Nature, 398, 672–673. Grutter, A.S. (2004) Cleaner fish use tactile dancing behavior as a preconflict management strategy. Current Biology, 14, 1080–1083. Grutter, A.S. & Bshary, R. (2003) Cleaner wrasse prefer client mucus: support for partner control mechanisms in cleaning interactions. Proceedings of the Royal Society of London Series B – Biological Sciences, 270, 242–244. Grutter, A.S. & Bshary, R. (2004) Cleaner fish Labroides dimidiatus diet preferences for different types of mucus and parasitic gnathiid isopods. Animal Behaviour, 68, 583–588. Grutter, A.S., Murphy, J.M. & Choat, J.H. (2003) Cleaner fish drives local fish diversity on coral reefs. Current Biology, 13, 64–67. Hauser, M.D. (1998) Minding the behaviour of deception. In: A. Whiten & D.W. Byrne (eds) Machiavellian Intelligence II, pp. 112–143. Cambridge University Press, Cambridge. Healey, E.G. (1957) The nervous system. In: M.E. Brown (ed) The Physiology of Fishes, Vol. 2, pp. 1–119. Academic Press, New York. Heg, D., Bachar, Z., Brouwer, L. & Taborsky, M. (2004) Predation risk is an ecological constraint for helper dispersal in a cooperatively breeding cichlid. Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 2367–2374. Hert, E. (1985) Individual recognition of helpers by the breeders in the cichlid fish Lamprologus brichardi (Poll, 1974). Zeitschrift f¨ur Tierpsychologie, 68, 313–325. Heyes, C.M. (1998) Theory of mind in nonhuman primates. Behavioral and Brain Sciences, 21, 101–134. Hiatt, R.W. & Brock, V.E. (1948) On the herding of prey and the schooling of the black skipjack, Euthynnus yaito Kishinouye. Pacific Science, 2, 297–298. Huber, R., van Staaden, M.J., Kaufman, L.S. & Liem, K.F. (1997) Microhabitat use, trophic patterns, and the evolution of brain structure in African cichlids. Brain, Behavior and Evolution, 50, 167–182. Johnstone, R.A. (2001) Eavesdropping and animal conflict. Proceedings of the National Academy of Sciences of the USA, 98, 9177–9180. Kamil, A.C. (1998) On the proper definition of cognitive ethology. In: R.P. Balda, P.A. Bednekoff & A.C. Kamil (eds) Animal Cognition in Nature, pp. 1–28. Academic Press, New York. Kaminski, J., Riedel, J., Call, J. & Tomasello, M. (2005) Domestic goats (Capra hircus) follow gaze direction and use social cues in an object choice task. Animal Behaviour, 69, 11–18. Keenleyside, M.H.A. (1991) Cichlid Fishes. Chapman & Hall, London. Kendal, J.R., Rendell, L., Pike, T.W. & Laland, K.N. (2009a) Nine-spined sticklebacks deploy a hill-climbing social learning strategy. Behavioral Ecology, 20, 238–244. Kendal, R.L., Coolen, I. & Laland, K.N. (2009b) Adaptive trade-offs in the use of social and personal information. In: R. Dukas & J.M. Ratcliffe (eds) Cognitive Ecology II, pp. 249–271. Kohler, U. (1997) Zur struktur und evolution des sozialsystems von Neolamprologus multifasciatus (Cichlidae, Pisces), dem kleinsten schneckenbuntbarsch des tanganjikasees. PhD thesis, LudwigMaximilian-Universit¨at, Munich. Kotrschal, K., van Staaden, M.J. & Huber, R. (1998) Fish brains: evolution and environmental relationships. Reviews in Fish Biology and Fisheries, 8, 373–408. Krause, J. (1993) Transmission of fright reaction between different species of fish. Behaviour, 127, 37–48. Laland, K.N. (2004) Social learning strategies. Learning and Behavior, 32, 4–14. Losey, G.C., Grutter, A.S., Rosenquist, G., Mahon, J.L. & Zamzow, J.P. (1999) Cleaning symbiosis: a review. In: V.C. Almada, R.F. Oliveira & E.J. Goncalves (eds) Behaviour and Conservation of Littoral Fish, pp. 379–395. Instituto Superior de Psicologia Aplicada, Lisbon. Lotem, A., Fishman, M.A. & Stone, L. (2003) From reciprocity to unconditional altruism through signalling benefits. Proceedings of the Royal Society London, Series B – Biological Sciences, 270, 199–205. Luce, R.D. & Raiffa, H. (1957) Games and Decisions. John Wiley & Sons, New York. Magurran, A.E. & Higham, A. (1988) Information transfer across fish shoals under predator threat. Ethology, 78, 153–158.

BLBK374-13

BLBK374-Brown

296

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

McGregor, P.K. (1993) Signalling in territorial systems: a context for individual identification, ranging and eavesdropping. Philosophical Transactions of the Royal Society of London Series B – Biological Sciences, 340, 237–244. Melis, A.P., Hare, B. & Tomasello, M. (2006) Chimpanzees recruit the best collaborators. Science, 311, 1297–1300. Milinski, M. (1987) Tit for tat in sticklebacks and the evolution of cooperation. Nature, 325, 433–435. Milinski, M., K¨ulling, D. & Kettler, R. (1990a) Tit for tat: sticklebacks Gasterosteus aculeatus ‘trusting’ a cooperative partner. Behavioral Ecology, 1, 7–11. Milinski, M., Luthi, J.H., Eggler, R. & Parker, G.A. (1997) Cooperation under predation risk: experiments on costs and benefits. Proceedings of the Royal Society of London Series B – Biological Sciences, 264, 831–837. Milinski, M., Pfluger, D., K¨ulling, D. & Kettler, R. (1990b) Do sticklebacks cooperate repeatedly in reciprocal pairs? Behavioral Ecology and Sociobiology, 27, 17–21. Munn, C.A. (1986) Birds that ‘cry wolf ’. Nature, 319, 143–145. Nowak, M.A. & Sigmund, K. (1998) Evolution of indirect reciprocity by image scoring. Nature, 393, 573–577. Oates, J., Manica, A. & Bshary, R. (2010) Roving decreases service quality in the cleaner wrasse Labroides bicolor. Ethology, 116, 309–315. Oates, J., Manica, A. & Bshary, R. (2010) The Shadow of the Future affects cooperation in cleaner fish. Current Biology, 20, R472–R473. Oliveira, R.F. (2005) Hormones, social context and animal communication. In: P.K. McGregor (ed) Communication Networks, pp. 481–520. Cambridge University Press, Cambridge. Oliveira, R.F., McGregor, P.K. & Latruffe, C. (1998) Know thine enemy: fighting fish gather information from observing conspecific interactions. Proceedings of the Royal Society of London Series B – Biological Sciences, 265, 1045–1049. Ostrom, E. (1990) Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press, New York. Pitcher, T.J., Green, D.A. & Magurran, A.E. (1986) Dicing with death: predator inspection behavior in minnow Phoxinus phoxinus shoals. Journal of Fish Biology, 28, 439–448. Potts, G.W. (1973) The ethology of Labroides dimidiatus (Cuv. and Val.) (Labridae, Pisces) on Aldabra. Animal Behaviour, 21, 250–291. Premack, D. & Woodruff, G. (1978) Does the chimpanzee have a theory of mind? Behavior and Brain Science, 4, 515–526. Raihani, N.J., Grutter, A.S. & Bshary, R. (2010) Punishers benefit from third-party punishment in fish. Science, 327, 171. Robertson, D.R. (1972) Social control of sex reversal in a coral-reef fish. Science, 177, 1007–1009. Roberts, G. & Sherratt, T.N. (1998) Development of cooperative relationships through increasing investment. Nature, 394, 175–179. Roth, G. & Wullimann, M.F. (2001) Brain Evolution and Cognition. John Wiley & Sons, Inc., and Spektrum Akademischer Verlag, New York/Heidelberg. Schlupp, I., Marler, C. & Ryan, M.J. (1994) Benefit to male sailfin mollies of mating with heterospecific females. Science, 263, 373–374. Schmitt, R.J. & Strand, S.W. (1982) Cooperative foraging by yellowtail Seriola lalandei (Carangidae) on two species of fish prey. Copeia, 1982, 714–717. ¨ Sch¨onherr, J. (1954) Uber die abh¨angigkeit der instinkthandlungen vom vorderhirn und zwischenhirn (Epiphyse) bei Gasterosteus aculeatus L. Zoologische Jahrb¨ucher. Abteilung f¨ur Allgemeine Zoologie und Physiologie der Tiere, 65, 357–386. Schradin, C. & Lamprecht, J. (2000) Female-biased immigration and male peace-keeping in groups of the shell-dwelling cichlid fish Neolamprologus multifasciatus. Behavioral Ecology and Sociobiology, 48, 236–242. Seegar, J. (1956) Brain and instinct with Gaterosteus aculeatus. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen, Series C– Biological and Medical Sciences, 59, 738–749.

BLBK374-13

BLBK374-Brown

May 27, 2011

11:5

Copyeditor’s Name:

Trim: 244mm X 172mm

Machiavellian Intelligence in Fishes

Char Count=

297

Sette, O.E. (1950) Biology of the Atlantic mackerel (Scomber scombrus) of North America II: migration and habits. Fish Bulletin 49, U.S. Fish Wildlife Service, 51, 251–358. Soares, M.C., Bshary, R. & Cˆot´e, I.M. (2009) Cleaning in pairs enhances honesty in male cleaning gobies. Behavioral Ecology, 20, 1343–1347. Soares, M.C., Cˆot´e, I.M., Cardoso, S.C. & Bshary, R. (2008) On the absence of punishment, partner switching and tactile stimulation in the cleaning goby – client mutualism. Journal of Zoology, 276, 306–312. Soares, M.C., Bshary, R., Fusani, L., Goymann, W., Hau, M., Hirschenhauser, K. & Oliveira, R.F. (2010) Hormonal mechanisms of cooperative behaviour. Philosophical Transactions of the Royal Society, London B – Biological Sciences, 365, 2737–2750. Stiver, K.A., Dierkes, P., Taborsky, M. & Balshine, S. (2004) Dispersal patterns and status change in a co-operatively breeding fish Neolamprologus pulcher: evidence from microsatellite analyses and behavioural observations. Journal of Fish Biology, 65, 91–105. Taborsky, M. (1984) Broodcare helpers in the cichlid fish, Lamprologus brichardi: their costs and benefits. Animal Behaviour, 32, 1236–1252. Taborsky, M. (1985) Breeder–helper conflict in a cichlid fish with broodcare helpers: an experimental analysis. Behaviour, 95, 45–75. Taborsky, M. (1994) Sneakers, satellites, and helpers: parasitic and cooperative behavior in fish reproduction. Advances in the Study of Behavior, 23, 1–100. Taborsky, M. & Grantner, A. (1998) Behavioural time-energy budgets of cooperatively breeding Neolamprologus pulcher (Pisces: Cichlidae). Animal Behaviour, 56, 1375–1382. Taborsky, M. & Limberger, D. (1981) Helpers in fish. Behavioral Ecology and Sociobiology, 8, 143–145. Tebbich, S., Bshary, R. & Grutter, A.S. (2002) Cleaner fish Labroides dimidiatus recognize familiar clients. Animal Cognition, 5, 139–145. Thorndike, E.L. (1917) Animal Intelligence: Experimental Studies. Macmillan, New York. Tomasello, M., Call, J. & Hare, B.H. (1998) Five primate species follow the visual gaze of conspecifics. Animal Behaviour, 55, 1063–1069. Trivers, R.L. (1971) The evolution of reciprocal altruism. Quarterly Review of Biology, 46, 35–57. Tulving, E. (1983) Elements of Episodic Memory. Oxford University Press, New York. Tulving, E. & Markowitsch, H.J. (1998) Episodic and declarative memory: role of the hippocampus. Hippocampus, 8, 198–204. Van Bergen, Y. , Coolen, I. & Laland, K.N. (2004) Nine-spined sticklebacks exploit the most reliable source when public and private information conflict. Proceedings of the Royal Society of London Series B – Biological Sciences, 271, 957–962. van Staaden, M.J., Huber, R., Kaufman, L. & Liem, K. (1995) Brain evolution in cichlids of the African Great Lakes: brain and body size, general patterns and evolutionary trends. Zoology, 98, 165–178. de Waal, F.B.M. & van Roosmalen, A. (1979) Reconciliation and consolation among chimpanzees. Behavioral Ecology and Sociobiology, 5, 55–66. Wedekind, C. & Milinski, M. (2000) Cooperation through image scoring in humans. Nature, 288, 850–852. Werner, N.Y., Balshine-Earn, S., Leach, B. & Lotem, A. (2004) Helping opportunities and space segregation among helpers in cooperatively breeding cichlids. Behavioral Ecology, 14, 749–756. Witte, K. & Noltemeier, B. (2002) The role of information in mate-choice copying in female sailfin mollies (Poecilia latipinna). Behavioral Ecology and Sociobiology, 52, 194–202. Witte, K. & Ryan, M.J. (2002) Mate choice copying in the sailfin molly, Poecilia latipinna, in the wild. Animal Behaviour, 63, 943–949. Zahavi, A. (1995) Altruism as a handicap – the limitations of kin selection and reciprocity. Journal of Avian Biology, 26, 1–3.

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 14

Lateralization of Cognitive Functions in Fish Angelo Bisazza and Culum Brown

14.1

Introduction

From an anatomical point of view the left and right sides of the brain of a vertebrate appear fundamentally identical. Yet, more than a century ago it was discovered that the human brain is functionally asymmetric, with some functions being represented in one hemisphere but not in the other (Broca 1861). Asymmetric control of cognitive functions is called cerebral lateralization. Many cognitive functions are known to be lateralized in humans including language, face recognition, spatial abilities, mathematical abilities, and emotional response (Chochon et al. 1999; Floel et al. 2001; Phelps et al. 2001). In a few cases the consequences of cerebral lateralization are observable in everyday behavior (usually called “behavioral lateralization” or “laterality”) and hence can be easily measured. The most notable example of laterality in human is handedness, the preferential use of one hand for fine manipulation of objects. Much of the early progress made in understanding cerebral lateralization in humans was made using patients with brain damage in one hemisphere or the other, caused, for example, by stroke. However, for most cognitive functions the investigation of lateralization in our species requires complex procedures such as presenting stimuli to one portion of the retina for few milliseconds or invasive methods such as unihemispheric anaesthesia or brain imaging. For more than a century cerebral asymmetries were thought to be a uniquely human trait linked to the evolution of language, handedness, and tool use. The first clear evidence for functional lateralization in nonhuman species was provided in the early 1970s by Fernando Nottebohm, who showed that severing the left hypoglossal nerve impaired singing in the canary, whereas severing the right nerve had no effect on song production (Nottebohm 1971). During the following decade numerous other examples of functional hemispheric specialization were provided for both mammals and birds. The first report of lateralization in a nonhuman mammal is by Denenberg et al. (1978), who showed that lesions of the left and right hemispheres differently affected exploratory behavior in the rat; this was followed shortly by demonstration of a left-hemisphere advantage in processing species-specific Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Lateralization of Cognitive Functions in Fish

299

calls in Japanese macaques (Petersen et al. 1978). Rogers & Anson (1979) showed that administration of an inhibitor of protein synthesis in the left hemisphere of chicks impaired learning in a food versus nonfood discrimination task, whilst treatment of right hemisphere was ineffective. A few years later, Mench & Andrew (1986) testing chicks in condition of monocular vision showed that they could learn the same food discrimination while only using their right eye, but not while using their left eye. The latter study showed that the procedure for revealing lateralization could be greatly simplified by the use of behavioral techniques, paving the way for a new generation of studies. Investigation on fishes, amphibians, and reptiles have appeared more recently (Bauer 1993; Cantalupo et al. 1995; Deckel 1995), but in the last 15 years there has been a rapid accumulation of evidence in this field and teleost fishes in particular have become favorite models for studying many aspects of cerebral lateralization. There are several advantages to investigating lateralization in fishes, especially in comparison with classical model organisms such as primates or rodents. Consider, for example, visual lateralization (by far the most frequent type of lateralization studied) in humans and in a fish. In most fishes, the two eyes are laterally placed and, with the exception of a small frontal overlapping portion, each eye largely sees a different portion of the external world. All the axons from one eye cross to the contralateral side so that a stimulus seen with the left eye is primarily analyzed by the right side of the brain and vice versa (Fig. 14.1). By covering one eye or the other or presenting one stimulus unilaterally, we can draw inferences about the way hemispheres function by measuring the differences in the behavioral response of the subject. In contrast, human eyes are placed frontally so that both eyes see nearly the same scene. The optic nerve fibers coming from one eye partially decussate at the optic chiasm and approximately 50% of the fibers from each eye reach each hemisphere. As a consequence, our perception of the external world changes little when we cover one eye or the other. Even when the information reaches predominantly one hemisphere (e.g., for other sensory modalities), a fast and efficient interhemispheric communication is enabled

Left visual hemifield

Right visual hemifield

Right visual hemifield Left visual hemifield

Fig. 14.1 A comparison of the overlap in left and right visual hemifields in primates and fish. Not only do fishes have relatively little overlap in the visual fields but they also show almost 100% crossover to the contralateral hemisphere at the optic chiasm. However, in primates objects falling in the visual field are seen by both eyes, each eye projects to both hemispheres and the connection between them is further facilitated by the corpus callosum.

BLBK374-14

BLBK374-Brown

300

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

by the corpus callosum. As consequence of these differences, human research is confined almost exclusively to the domain of neuroscience while lateralization can easily be studied in fishes using behavioral methods. The differences outlined in the preceding text also imply that functional left–right asymmetries are much more relevant in everyday behavior for a fish or a bird than for a primate. For example, a domestic chick can discriminate a companion from an unfamiliar individual visually if it sees the bird with the left eye or with both eyes but not if it sees it with the right eye only (Vallortigara & Andrew 1994). Although face recognition is mainly performed by the right hemisphere in humans too, we have no difficulty in discriminating familiar people placed on the left or the right of our visual field. Birds, and to some extent reptiles, can partially counter this drawback by using their mobile necks to scan in sequence the relevant scene with both eyes (e.g., marsh tits, Clayton & Krebs 1994), but such a solution is far more difficult for fishes which have to reorientate their entire body.

14.2

Lateralized functions in fish

In contrast with mammalian research that has often directly looked at left–right differences in brain functioning, most studies on fishes have investigated left–right differences in behavior, assuming that they reflect an underlying asymmetry in the functioning of the nervous system. As a consequence of this approach, laterality of ray-finned fishes has been studied in a large variety of contexts and the behavioral effects of lateralization are known for a large number of functions. As in the other vertebrates, very few studies have investigated lateralization in sensory modalities other than vision (but see olfactory navigation in eel and lateral line use during exploration in blind cave fish; Westin 1998; Burt de Perera & Braithwaite 2005). Some studies have attempted to examine motor asymmetries, but since it is difficult to exclude the influence of visual lateralization in tests for motor biases, few can be considered successful in isolating the motor component of behavior. Among these studies are sound production in catfish, fin use by gourami fish, and turning biases in the dark by topminnows (Fine et al. 1996; Bisazza et al. 2001a, 2001b). However, of all the contexts examined thus far antipredator behavior has received the most attention.

14.2.1

Antipredator behavior

A first indirect evidence for lateralization in antipredator behavior was provided by the observation that external scars resulting from predation attempts on whitefish (Coregonus spp.) were located far more commonly on the left side than on the right side (Reist et al. 1987). The asymmetry in the location of the scars was the same for all types of scars (parasitic copepods, lampreys, terrestrial vertebrates), suggesting that asymmetry was probably the consequence of lateralization in evasive actions by the whitefish rather than to lateralization of its predators. Subsequent studies conducted in the laboratory have confirmed the lateralization of antipredator responses in many teleosts. Three main lines of research have been investigated: (1) Lateralization of predator evasion, (2) predator inspection, and (3) asymmetry in fast escape response.

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

14.2.1.1

Char Count=

301

Predator inspection

In many species of fishes, two or more individuals leave their shoal in order to approach and inspect a potential predator (Magurran & Pitcher 1987; Magurran & Seghers 1990; Chapter 4). Bisazza et al. (1999) have studied the laterality of cooperative predator inspection in female eastern mosquitofish, Gambusia holbrooki, using a procedure introduced by Milinski (1987) in which a mirror is placed parallel to the tank during inspection so that the image appears to swim along with the fish, simulating a cooperative partner. Mosquitofish performed significantly closer inspections when the mirror image was visible on the left rather than on the right side of the subject (Bisazza et al. 1999). A subsequent study (De Santi et al. 2001) found that the mirror on the left provides to subject the best arrangement of complementary monocular lateral stimulation with the virtual partner seen with the left eye and the predator seen with the right eye. The studies discussed in the preceding text imply that in mosquitofish, the left hemisphere is specialized for processing information about predators, while the other hemisphere is specialized for other functions, for example processing social stimuli. However, analyses at the population level may mask underlying individual variability. For example, in mosquitofish, about two-thirds of the subjects preferentially used the right eye to inspect the predator, while the remaining third showed an opposite preference (De Santi et al. 2001; Fig. 14.2). Whilst these studies used captive-reared fishes, the pattern of laterality whilst inspecting predators in the wild may vary depending on the relative experience of the individual. Fishes from high-predation areas that have a history of predator exposure may view the predator as a threat, whilst those from low-predation areas may view it as a novel object (Brown & Warburton 1999), which may cause a switch in the eye used to scrutinize

6 5 Number of individuals

BLBK374-14

4 3

Left eye

Right eye

2 1 0

0,18

0,26

0,34 0,42 0,5 0,58 0,66 Right-eye use in predator inspection

0,74

0,82

Fig. 14.2 Frequency distribution of mosquitofish viewing a predator through a clear partition with the left or right eye. About two-thirds of the population used their right eye to view the predator while the remaining third used their left eye. (Data from De Santi et al. 2001.)

BLBK374-14

BLBK374-Brown

302

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

the predator. Even within high-predation populations, exposure to a novel predator may invoke an alternative response to that produced by a frequently encountered predator, but these experiments have yet to be conducted. 14.2.1.2

Predator evasion

The lateralization of predator evasion has been the object of several studies that employed the “detour test.” The detour test records the percentage of right and left turns taken by a fish over ten consecutive trials when leaving a runway and facing a vertical-bar barrier behind which a visual stimulus is located. The direction of detour is determined by the nature of the object that is seen beyond the barrier (Bisazza et al. 1997a, 1997b, 1998a). While leaving the runaway on the left, a fish is monitoring the stimulus with the right eye (and vice versa). Several lines of evidence now confirm that the different direction taken is determined by specific eye preferences to observe different classes of stimuli (Facchin et al. 1999; Bisazza et al. 2001b; Brown et al. 2004; Bisazza et al. 2007). Using a realistic predator model as stimulus, Bisazza et al. (1997a) found a significant leftward bias at the population level in the western mosquifish, G. holbrooki. This result was later confirmed in another poeciliid, the goldbelly topminnow, Girardinus falcatus (Bisazza et al. 1998a). Although this population bias was consistent across species and studies, within each sample large interindividual variation was observed and while many fishes turned preferentially to the left, several others showed the opposite tendency, or no turning preference. Facchin et al. (1999) showed that individual differences in lateralization of detour behavior arise from individual differences in the asymmetry of eye use to observe a predator. Fishes with different scores in the detour test were singly exposed in their home tank to a dummy predator or a neutral stimulus, with eye preference recorded while fixating the two novel objects. Fishes that tended to detour the barrier on the left side used the right eye to scrutinize the predator and the left eye to scrutinize the neutral stimulus, while the fishes that tended to detour the barrier on the right side showed the reverse eye preference. Fishes that turned 50% in each direction had an equal tendency to look at a potential predator or at a neutral stimulus with both eyes. Recent studies have confirmed that the eye used to look at the dummy predator in the detour test is the same as that used when viewing a real predator (Brown et al. 2004, 2007a; Bisazza et al. 2007). Bisazza et al. (2000a) compared 16 species from different fish families in the detour test. Fishes of all 16 species were trained to escape from the same stimulus (a dip net provided with two eyes) that mimicked a predator and was used as the target stimulus in the detour test. Ten out of the 16 species showed lateralization at the population level. Not all lateralized species were biased in the same direction, although closely related species tended to have similar patterns of lateralization, providing some evidence of phylogentic constraint. Interestingly, species with and without population biases did not differ in the absolute values of the laterality index, indicating that they were composed of individuals lateralized, on average, to the same degree and that they differ with respect to the proportions of right- and left-biased individuals. Thus, variation exists at both the species and individual levels with respect to which eye is used to view predators. Research by Brown et al. (2004) examined the response of the poeciliid, Brachyrhaphis episcopi, collected from the wild to live predators placed behind clear perspex. Several

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

303

populations from high- and low-predation streams were examined. Fishes from highpredation streams were more strongly lateralized and used their right eye to view predators whilst the reverse was true for fishes derived from low-predation areas. When taken together, the research to date strongly suggests that the pattern and strength of laterality in fishes responding to predators varies from species to species, between populations within species and at the individual level. The data suggest that laterality is influenced by both environmental variables and phylogeny (Brown 2005).

14.2.1.3

Fast escape response

Most teleost are capable of performing a fast-start escape response following a visual or acoustic stimulation. This response consists of a rapid unilateral muscle contraction, which bends the body into a “C” followed by a strong propulsive tail stroke that moves the fish away from danger. The neural circuits controlling C-start response are distinct from those controlling other antipredator responses, being mediated by a pair of giant reticulospinal neurons, the Mauthner cells, which allow a response latency of less than 20 ms (Domenici & Blake 1997). Cantalupo et al. (1995) measured the direction of turning during escape responses evoked by a rapidly approaching simulated predator in a poeciliid fish, the goldbelly topminnow. Newborn minnows showed a significant bias to escape rightwards in the first week when repeatedly exposed to a simulated predator. The bias was progressively reduced in subsequent weeks until, in the fifth week, it reversed, yielding a preference to escape leftwards. A similar pattern was observed in adults, suggesting that the familiarity with the situation rather than developmental mechanisms was responsible for the variation in escape direction over time. A significant leftward population bias in escape response elicited by generating pressure waves was also found in a nonteleost fish, the Australian lungfish (Neoceratodus forsteri) (Lippolis et al. 2009). By contrast, in the one-sided livebearer, Jenynsia multidentata, the distribution of laterality score was bimodal with approximately the same number of individual escaping on left or right (Bisazza et al. 1997c). Heuts (1999) found a significant rightward population bias in C-start direction in zebrafish and goldfish, but not in guppies or in four species of cichlid. Interestingly, in both goldfish and zebrafish, slow-turns were significantly left-biased, thus opposite in direction in respect to fast responses. In zebrafish, this specialization was paralleled by an asymmetry at the muscular level, the left side of the trunk having a significantly larger proportion of larger red-muscle mass (mainly recruited in slow swimming) compared with the left side. To summarize, once again we see considerable variation in the fast start response at all levels and at least some of the variability can be explained by variation in experience or familiarity with the testing context including the stimuli presented to the fish. Intriguingly, no correlation was found between the turning direction in the detour test and the direction of fast escape response in the shiner perch, Cymatogaster aggregata (Dadda et al. 2010a). Laterality of fast escape response in G. falcatus was unaffected by artificial selection for antipredator response in the detour test, suggesting that they are regulated by different genes or develop independently (Bisazza et al. 2005).

BLBK374-14

BLBK374-Brown

304

14.2.2

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Mating behavior

The manner in which fishes view conspecifics may vary depending on their motivation. The detour test was also used in poeciliids to measure eye preference for looking at fish of the opposite sex (Bisazza et al. 1997b, 1998a). Right-eye population biases for looking at opposite-sex fish were found in both males and females, but only if they were sexually motivated. Females G. holbrooki and G. falcatus showed right-eye preferences after being deprived of males for two months, a procedure that exhausts stored sperm reserves and enhances their willingness to mate. Females kept in mixed sex groups showed no such bias (Bisazza et al. 1998a). Similarly, male-deprived female guppies showed a stronger preference for the right eye when viewing bright-colored than dull males (Kaarthigeyan & Dharmaretnam 2005). Males of bold species like G. holbrooki and Poecilia reticulata that resume mating behavior soon after being moved in a novel place, showed right-eye bias to look at females. Males of shy species such as Brachyrhaphis roseni and G. falcatus showed the same bias but only if they were familiar with the apparatus. However, if tested in an unknown environment they showed a significant left-eye bias as if they were considering females as potential shoalmates (Bisazza et al. 1997b, 1998a). In contrast to the above studies, no lateral bias was found in courting display of Betta splendens and P. reticulata (Cantalupo et al. 1996; Gross et al. 2007). However, in the latter study males were found to individually bias their displays toward their more colorful side.

14.2.3

Aggression

Aggression is controlled by regions in the right hemisphere in humans and other vertebrates (Rogers 2002); thus, laterality is also likely to have large effects on the expression of aggression in a wide range of species. Gelada baboons, domestic chicks, Anolis lizard, and common toads are more likely to attack a rival male seen on their left than on their right side (Rogers et al. 1985; Deckel 1995; Casperd & Dunbar 1996). However, the opposite pattern appears to occur in fishes. Bisazza & De Santi (2003) investigated laterality of aggression in three teleosts, G. holbrooki, Xenotoca eiseni, and B. splendens, and found a significant population bias for looking at rivals with the right eye prior to an attack in all species. In convict cichlids, an intriguing interaction occurs between aggression and personality. Highly aggressive males and nonaggressive females tended to show a right-eye bias in the detour test, whereas nonaggressive males and aggressive females had a slight tendency towards a left-eye bias (Reddon & Hurd 2008). The varying results in fishes compared to the rest of the vertebrates suggest a change in lateralization of the mechanisms controlling aggression after the separation of land vertebrates from fishes.

14.2.4

Shoaling and social recognition

Using social stimuli (females of the same size) as target in the detour test, female mosquitofish and Panamanian bishop were found to have a strong bias to use their left eye to detour the barrier (Bisazza et al. 1998a; Brown et al. 2007a). This result was later confirmed using a more direct measure of shoaling preference, the mirror test (De Santi

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

305

4

School position

BLBK374-14

3 Epa nigrans wild 2

1 Left

Right Flume preference

Fig. 14.3 The relationship between eye preference in a mirror test and the position adopted in a school of four individuals for two species of rainbowfish. Position 1 is the far right and position 4 is the far left school position. The dotted line represents the central position in the school. Captive-reared M. duboulayi (Epa) with left- or right-eye bias in the mirror test show no preference for either side of the school. In comparison, both the first generation laboratory-reared M. nigrans (nigrans) and the wild M. duboulayi (wild) populations show the expected pattern; fishes that prefer to shoal with their mirror image on the right side prefer to take up positions of the left side of a shoal and vice versa.

et al. 2001) in which the fish is placed in a tank with mirror walls and the time spent shoaling with the virtual companion on left or right was recorded. Left-eye preference in the mirror test was shown to occur in species from different orders (Osteoglossiformes, Cypriniformes, Cyprinodontiformes, Beloniformes) suggesting that right hemisphere specialization for analyzing social stimuli may be the general pattern in teleosts (Sovrano et al. 1999, 2001). However, research using rainbowfish, Melanotaenia duboulayi and Melanotaenia nigrans (Atheriniformes), found significant preferences for the right eye while schooling with their mirror image in a flume. Interestingly, in rainbowfish correlations were found between eye preferences in the mirror tests and the location that the fish subsequently took up while schooling (Fig. 14.3). Rainbowfish that showed right-eye preferences in the mirror test preferentially adopt positions on the left side of a school but the strength of this association varied between species and the environment in which they were reared (Bibost & Brown, unpublished data). Other vertebrates show right-hemisphere dominance for analyzing certain classes of social stimuli. For example, chicks can distinguish familiar from unfamiliar conspecifics using solely the left eye but not when solely using the right eye (Vallortigara & Andrew 1994) and face recognition is mainly performed by the right hemisphere in primates (Hamilton & Vermeire 1988). It is apparent that different categories of social stimuli may be analyzed by different hemispheres. In a study of guppies, females were found to look preferentially with the right eye at a familiar fish while they preferentially used the left eye when the stimulus was an unfamiliar female (Kaarthigeyan & Dharmaretnam 2005). As a mirror image equates to an unfamiliar fish, this may explain the fact that laterality in mirror response often vanishes or even reverses after the first minutes of testing as the stimulus becomes

BLBK374-14

BLBK374-Brown

306

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

increasingly familiar (Sovrano et al. 2001). Moreover, the duration of this familiarization phase may differ across species and perhaps between populations. For example, rainbowfish bred in captivity show no preferences for shoaling with familiar individuals (Kydd & Brown 2009), which contrasts with wild populations (Brown 2002; Chapter 9). Thus, preferences for familiarity and the time it takes to develop are factors that should be taken into consideration when comparing laterality preferences in different species and likely add a further source of variability to the equation.

14.2.5

Foraging behavior

Fossil records suggest that lateralization of predation is an ancient phenomenon. Babcock (1993) found evidence of asymmetry in attack patterns on Cambrian trilobite prey which was reminiscent of that observed in whitefish (Reist et al. 1987). While this evidence is suggestive of lateralized predatory attacks by fish in ancient times, in the modern era few studies have examined lateralization of foraging behavior experimentally. Australian Lungfish were found to bend to the right side more often to eat inanimate food items (Lippolis et al. 2009) and the authors suggest the greater involvement of the left hemisphere in feeding behavior is analogous to that observed in birds and anurans (Mench & Andrew 1986; Vallortigara et al. 1998; see also Andrew 2002 for a discussion of the evolution of lateralized feeding responses). A similar study that compared goldbelly topminnow selected for opposite direction of lateralization at the detour test also found opposite lateralization in prey capture but only when individuals were distracted by a concurrent task – monitoring a predator (Dadda & Bisazza 2006a). This suggests that in some instances strong lateralization at the behavioral level may only be overtly expressed in cognitively demanding contexts. Among the scale-eating cichlids of genus Perissodus, the mouth opens on either the left or right side. Leftward individuals feed on the right body side of their prey and vice versa. In a population studied by Hori (1993), the predominant direction of mouth opening showed cyclical fluctuations in frequency over the years, suggesting the possibility that lateralization may generate an evolutionary dynamic between predator and prey driven by frequency dependent selection (see Brown 2005 for a discussion). Comparisons of parents and offspring indicate that mouth laterality is determined by a one-locus, two-allele system. While mouth asymmetry per se might be independent from cognitive lateralization, it is likely that it is associated with cognitive asymmetries in terms of eye preferences while viewing potential prey targets and the direction of attack approach. In a related species, Neolamprologus fasciatus, mouth asymmetry was only weakly correlated with the preferred side of prey capture. However, fishes with right-opening mouths were significantly more efficient at capturing prey than “lefties,” which suggest a possible interaction of cerebral and morphological asymmetries in determining catch success (Takeuchi & Hori 2008).

14.2.6

Exploration and response to novelty

The results of a study by Miklosi & Andrew (1999) suggest that in zebrafish the response to novelty is lateralized. After zebrafish were trained to bite at a small bead, the appearance of the target was changed. Strong right-eye use was found to be associated with the decision to bite the newly appearing object. In this species, laterality was also found to influence

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

307

exploratory behavior in a novel location (Dadda et al. 2010b). When Panamanian bishops were confronted with novel objects, fishes from high-predation areas tended to view them with their left eye, while fishes from low-predation areas showed no eye preference (Brown et al. 2004, 2007a). The laboratory-reared offspring derived from these populations showed similar strengths of laterality, but in the case of the high-predation fishes, laboratory-reared offspring had strong preferences for the opposite eye (Brown et al. 2007a). These results suggest that rearing environment can influence the pattern of laterality in fishes likely via experiential effects. The response to novelty in a number of species has also been tested where the fishes are relying on alternative sensory inputs. Gouramis have a pair of modified ventral fins that are used to contact objects to obtain tactile and chemical information. In the blue gourami, Trichogaster trichopterus, fin use during initial exploration of novel objects has been found to be lateralized and the type of “handedness” shown by the fishes was found to be influenced by the nature of the object explored (Bisazza et al. 2001a). Burt de Perera and Braithwaite (2005) studied the asymmetry of lateral-line use during exploration in the Mexican blind cave fish. When confronted with a new stimulus for the first time, fishes used their right flank significantly more often than the left but this preference disappeared if the same object was presented again the next day. Taken together, these results suggest that the strength of laterality exhibited by individuals is likely to be an inherited trait (see Section 14.3), but the direction of laterality is heavily dependent on individual experience, which likely determines how objects are classified and, thus, which hemisphere is used to analyze the information.

14.2.7

Homing and spatial abilities

Fishes use a wide range of sensory systems to navigate in their environment (Chapter 8), and many of these systems are lateralized. In a study on the role of olfaction on spawning migration of European silver eel, intact control subject eels were compared with eels with one or both nostrils experimentally blocked (Westin 1998). The group with left nostril blocked behaved like eels with both nostrils blocked while that with right nostril blocked behaved like the controls, suggesting that homing is a completely lateralized function in eels. In a spatial task requiring the use of visual cues and geometric features of the environment, strongly lateralized goldbelly topminnow learn faster than weakly lateralized counterparts (Sovrano et al. 2005). This may be in relation with a crucial role of brain asymmetry in discriminating left from right (Chiandetti & Vallortigara 2008). Conversely, B. episcopi from high-predation populations, which are, on average, more lateralized than their low-predation counterparts (Brown et al. 2004), take longer to complete a maze task because their laterality interfered with an efficient exploratory behavior (Brown & Braithwaite 2005). Thus, laterality can both enhance and inhibit spatial navigation.

14.2.8

Communication

The left hemisphere advantage for producing and processing species-specific vocalizations have been reported for a range of vertebrates, including primates (Petersen et al. 1978), rodents (Ehret 1987), passerine birds (Nottebohm 1971), and anurans (Bauer 1993), and

BLBK374-14

BLBK374-Brown

308

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

some have argued that this specialization may represent a precursor for the evolution of lateralized control of speech in humans (Corballis 2003). Fine et al. (1996) studied the laterality of sound production in the channel catfish, Ictalurus punctatus. Catfish produce sounds by rubbing the pectoral spine against a groove in the pectoral girdle. In about half of the individuals there was a significant fin preference with 90% of fishes preferring the right fin. It is difficult to assess the generality of this finding because no other study of laterality of communication has ever been conducted in fishes, despite the fact that perhaps 50% of all fish species produce some kind of meaningful noise. Thus, the possibilities for future study in this area are immense.

14.3

Individual differences in lateralization

Around 90% of people are right-handed with little variation across cultures and historical periods since the Neolithic (McManus 2009). Such large population biases appear to be rare in the animal kingdom with the exception of parrots (Magat & Brown 2009). Classically, animal studies focused attention on population-level lateralization. However, many examples discussed in Section 14.1 suggest that significant individual variation for lateralization exist within and between animal populations as well (Hori 1993; De Santi et al. 2001; Brown et al. 2004, 2007a). Why should such variability exist? Considerable effort has been made in recent years, especially using fishes as model organisms, to understand the proximate causes and the evolutionary significance of laterality polymorphisms.

14.3.1

Hereditary basis of lateralization

Several studies have investigated the genetics of lateralization in mammals but the causes of individual differences are far from being understood (reviewed in Corballis 2009; McManus 2009). In contrast, research in fishes is starting to indicate some of the genes responsible for these traits. Recent work in the poeciliid B. episcopi examined the influences of both genes and rearing environment on the expression of laterality in the detour task (Brown et al. 2007a). Fishes from multiple populations were captured from the wild, bred and their offspring tested in the laboratory. While the strength of laterality was clearly heritable, the direction of laterality was not. Using a parent–offspring regression approach, Bisazza et al. (2000b) found heritability exceeding 0.5 for the detour test using a predator model as stimulus in G. falcatus. In a subsequent experiment in which the progeny were separated at birth to minimize nongenetic influences (e.g., social learning), similar heritability values where observed. Consequently, a selection experiment was undertaken with two lines selected for right-turning (right detour = RD), two for left-turning (LD), one for no turning preferences (nonlateralized = NL), and one unselected line as a control. A prompt response to directional selection was observed in all lines and the response was similar in left- and right-turning lines with no gender difference in response (Bisazza et al. 2007). In the NL line the proportion of nonlateralized individuals increased significantly from approximately 25% in the unselected population to approximately 50%. Intriguingly, RD and LD lines ceased to diverge after the first two

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

309

generations and significant phenotypic variation was maintained after several generations of directional selection, implying that other, probably nongenetic, factors contribute to the determination of strength and direction of laterality in this species. A series of subsequent experiments set out to assess whether fish belonging to the selected lines only differed for eye preference while viewing a predator in the detour test or if wider differences in the organization and localization of cognitive functions had also been generated. Fishes turning 80% or more to the left (from LD lines) or to the right (from RD lines) and fishes turning 50% to either direction (from NL line) were compared in other laterality tests. LD and RD fishes were strongly biased in opposite directions in almost all the laterality measures (Bisazza et al. 2001b, 2005; Dadda et al. 2007, 2009). For example, LD males conducted more mating attempts on their left side and more intrasexual attacks on their right side while RD males did exactly the reverse. The only exception in a dozen tests performed is that LD and RD fishes showed no differences in fast escape response to a visual stimulus, both being biased toward leftward turn. Belonging to LD or RD line was found in many cases to be nearly 100% predictive of the laterality score indicating that LD and RD fishes may be similar but with complete mirror-reversed organization of cerebral functions. By converse NL fishes appeared poorly lateralized in other laterality tests too, suggesting that they have a bilateral representation of most cognitive functions. Molecular studies conducted on zebrafish and other model organisms have revealed that genes of the Nodal signaling pathway are implicated in the development of the left–right axis of the body and cause asymmetric positioning of visceral organs. Early in development, Nodal-related signals also regulate asymmetric gene expression in the forebrain of zebrafish and are involved in particular in asymmetric positioning of parapineal in the epithalamic region of the dorsal diencephalon (on the left side in most individuals). Barth et al. (2005) found a correlation between visceral asymmetries and certain lateralized behaviors in the frequent-situs-inversus (fsi) line of zebrafish. The situs inversus mutation causes a reversal of position of visceral organs and also determines high rate of reversal of parapineal position (on the right) and these authors have suggested that there might be a causal relationship between asymmetries of the epithalamus and behavioral laterality. In two other strains of zebrafish, Facchin et al. (2009) found that artificial selection for right-eye use when looking at own mirror image significantly increased the frequency of reversed asymmetry in epithalamus while selection for left-eye use tended to decrease it, thus providing another indirect evidence of an association of epithalamic asymmetry and lateralization of cognitive functions. Recently Dadda et al. (2010b) have tested this hypothesis directly by comparing fishes with left- or right-positioned parapineal in a series of laterality test. Although significant differences between fishes with opposite parapineal position were found in all laterality tests, it is clear from this study that early asymmetric parapineal positioning is not an all-or-nothing determinant of cerebral lateralization in zebrafish and that other genetic or environmental factors must be involved in the determination of this character.

14.3.2

Sex differences in lateralization

Males and females often respond to the same stimuli in completely different ways. Males, for example, are motivated by sex and are less risk averse, whereas females tend to be highly motivated by food and highly risk sensitive which is related to their varying life-history

BLBK374-14

BLBK374-Brown

310

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

priorities. Reddon & Hurd (2009) investigated the lateralized response of males and females to emotionally conditioned stimuli in the convict cichlids (Archocentrus nigrofasciatus). One stimulus was associated with food reward while the other was repeatedly paired with a chemical alarm substance. When tested for eye preference while observing these two stimuli, no significant population bias in either sex, nor sex difference in laterality, were observed. However, when considering the absolute laterality index, a measure of the strength of lateralization, males tended to be more strongly lateralized to aversive stimuli while females were more strongly lateralized when responding to positively reinforced stimuli. In a related study, the authors found that sex and a personality trait (aggression) interact in determining turning direction in the detour test (Reddon & Hurd 2008; see Chapter 7 and Subsection 14.3.3 for further discussion). In two poeciliid fishes, little or no sex difference was found in the laterality tests where male and female fishes were tested in comparable situations. These include fast escape response in G. falcatus (Cantalupo et al. 1995), detour test in G. falcatus and G. holbrooki (Bisazza et al. 1998a), viewing test in G. falcatus (Facchin et al. 1999), mirror test in G. holbrooki (Sovrano et al. 1999), turning direction in T-maze, and rotational preference in G. falcatus (Bisazza et al. 2001b). Many of these tests were not specifically designed to test for the presence of sex differences and might have lacked sufficient statistical power to evidence it. In addition, the absolute laterality index was not deployed in these studies, which may explain why they have failed to find evidence for sex differences in the strength of lateralization. In future studies, tests that specifically target variation in motivational factors between sexes are likely to be fruitful in this regard.

14.3.3 Environmental factors influencing development of lateralization Although a hereditary basis of lateralization has been evidenced in fishes and to some extent in other vertebrates, genetic factors seem to account for only a fraction of the interindividual variation in laterality, suggesting that additional environmental factors may be important in the development of lateralization. Growing evidence now indicates that the development and expression of lateralization in vertebrates can be modulated by several environmental factors including maternal effects such as prenatal stress (Fride & Weinstock 1988), androgen exposure (Zappia & Rogers 1987), and features of early rearing environment (Collins 1975; Bibost & Brown 2010). These effects may represent adaptive plastic responses allowing parents to adjust the developmental trajectories of their offspring to the environmental conditions they will experience (Deng & Rogers 2002; Andrew 2009). For instance, the amount of light that enters through the eggshell in the days prior to hatching greatly affects development of lateralized visual behavior in domestic chicks (Deng & Rogers 2002). This has a profound impact on the ability of chicks to perform two concurrent tasks, such as feeding and predator vigilance (Rogers et al. 2004). Some authors have suggested that ecological conditions at the time of incubation (e.g., social density or predator abundance), through influencing nesting site choice or time spent at the nest, may determine the lateralization of the offspring and ultimately generate phenotypes that can better cope with current conditions (Deng & Rogers 2002; Vallortigara & Rogers 2005; Andrew et al. 2009).

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

311

Recent evidence indicates that light effects on laterality are not restricted to birds. In zebrafish, differential exposure to light early in development causes large differences in lateralization that have effects on many aspects of behavior including social responses, level of activity in a novel arena, and tendency to approach new, potentially dangerous, objects (i.e., boldness) (Andrew et al. 2009; Budaev & Andrew 2009a, 2009b). As suggested for birds, this might constitute a proximate mechanism enabling parents to affect the behavioral phenotype of their progeny by spawning in sites with different light exposure. Maternal steroid hormones (both glucocorticoids and androgens) deposited in the egg or crossing the placenta is another important factor affecting development of lateralization in birds and mammals (Diaz et al. 1995; Rogers & Deng 2005). Here again, there is the potential for a mechanism enabling a mother experiencing stressful situations (e.g., predatory attacks) to adaptively modulate the development of laterality in their offspring (Deng & Rogers 2002; Halpern et al. 2005). Currently no study investigated whether early hormonal exposure affects laterality in fishes, but it is known that elevated maternal stress hormones transferred to the egg yolk influence other aspects of fry development (McCormick 1998). Brown and colleagues (Brown et al. 2004, 2007a) have clearly shown that the level of predation pressure is a key environmental variable in determining lateralization in wild populations. By conducting common-garden experiments, it was revealed that early rearing environment (in particular the lack of predators) influenced turn biases in the detour test (Brown et al. 2007a). In related experiments, rainbowfish were reared in structurally enriched and impoverished conditions to examine the influence of environmental complexity of the development of laterality (Bibost & Brown 2010). Interestingly, males showed enhanced laterality under enriched conditions but females did not. It seems reasonable to conclude that laterality is partly heritable, but the remaining variability is determined by exposure to contemporary environmental conditions and thereby altered through individual experience. It makes sense to allow for some degree of plasticity in any generation to fortify offspring against potential environmental heterogeneity.

14.3.4

Lateralization and personality

It is now widely recognized that many animals have consistent individual differences, arguably with parallels to human personality (Wilson et al. 1994; R´eale et al. 2007; Bell et al. 2009; see Chapter 7 for a review). While there is some disagreement regarding the dimensions over which personality should be measured, there are, nonetheless, several dimensions that are regularly measured in fishes. Perhaps the most common measure of fish personality is boldness. Boldness refers to the tendency to take risks particularly in novel situations and has been measured using a number of methodologies. However, most approaches include a context that invokes fear in subjects. For example, Brown et al. (2007b) measured the latency to emerge from cover and explore a novel, potentially dangerous, environment. This measure also correlates with other measures such as the tendency to leave the safety of a shoal and approach a novel object (Brown et al. 2005). Others have measured the predator inspection behavior (e.g., Godin & Dugatkin 1996; Johnsson et al. 2001), or the tendency to forage

BLBK374-14

BLBK374-Brown

312

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

under predation threat (Ward et al. 2004; Magnhagen & Staffan 2005), both of which clearly involve an element of risk and are designed to induce a fear response. To date there have been no direct studies investigating the link between personality and laterality in fishes, but many studies report an association between personality and lateralization. Moreover, there is a clear theoretical link between lateralization and personality in relation to the emotive content of visual stimuli and the contexts in which they are encountered. In particular, personality may vary with laterality because of individual variation in the predominant hemisphere (left or right) that controls the processing of certain types of information. In many vertebrates, fear is encoded in the left or right hemisphere depending on the contingency (e.g., in the left amygdala with visual fear conditioning in humans; Phelps et al. 2001). Fear appears to be lateralized at both the individual and population levels as indicated by eye preferences for viewing predators and novel objects (Brown et al. 2004, 2007a). Studies of the poeciliid, B. episcopi, have revealed covariation of laterality and boldness across populations (Brown et al. 2004, 2005). Fishes collected from regions of highpredation pressure were consistently bolder and more strongly lateralized than those from low-predation areas. This link between boldness and laterality is further illustrated by research conducted on fsi lines of zebrafish. Using fishes with left (L) and right (R) asymmetries, Barth et al. (2005) showed that the lines varied in a number of behavioral traits. For example, when emerging into a novel environment both L and R lines show similar turning biases but their latency to emerge (boldness) differed significantly. Along similar lines, Dadda et al. (2010b) recorded the position of the parapineal organ with reference to the left or right habenula in young zebrafish using the foxD3:GFP marker. The habenulae and associated brain circuitry are known to play important roles in a number of behavioral functions including avoidance learning. Fishes that differed in the location of the parapineal organ consistently differed in a range of laterality tests. More importantly, however, fishes with right-parapineal position tended to be bolder when inspecting a predator and spent less time in the peripheral portion of an open field (generally assumed to be the safer location) than fishes with parapineal organ located on the left. Andrew et al. (2009) reared zebrafish in light and dark conditions and successfully manipulated laterality with similar effect to that seen in the chick. Interestingly, rearing eggs and larvae in darkness for the first six days after fertilization increased shyness as measured by a tendency to avoid a predator model and reduced locomotion in its presence (Budaev & Andrew 2009a). Partial reversal of lateralized functions (either genetic or environmental) may serve to generate new behavioral phenotypes. In the current context, it may be that reversal of lateralized function generates different personality types each of which may be optimal under a range of ecological conditions (Reddon & Hurd 2009).

14.4 14.4.1

Ecological consequences of lateralization of cognitive functions Selective advantages of cerebral lateralization

Given its taxonomic ubiquity, lateralization is expected to provide some evolutionary advantage. Several possible advantages of an asymmetric brain have been suggested, including

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

313

enabling left–right discrimination (Benton & Menefee 1957), avoiding costly duplication of functions in the two hemispheres (Levy 1977), and preventing the simultaneous activation of incompatible responses in organisms with laterally placed eyes (Andrew 1991). Recently, Rogers (2000, 2002) has suggested that hemispheric specialization evolved mainly because it favors multitasking. Research on humans, fishes, birds, and spiders (Metcalfe et al. 1987; Dukas & Kamil 2000; Kastner & Ungerleider 2003; Hebets 2005) indicates that animals are normally constrained in how much attention they can focus on different activities simultaneously, so that individuals engaged in a complex task are unable to allocate enough attention to a second concurrent task (limited attention hypothesis; Dukas 2004). The ecological implications of limited attention have mainly focused on the trade-off between foraging and vigilance. Guppies catching live Daphnia, for example, were found to be captured more often by a predator when the density of prey increased and more attention was devoted to foraging (Godin & Smith 1988). According to Rogers’ hypothesis, cerebral lateralization allows an increase in the brain’s capacity to carry out simultaneous processing, by channeling different types of information into the two separate halves of the brain and by enabling separate and parallel processing to take place in the two hemispheres. To test her hypothesis, Rogers et al. (2004) compared normally and weakly lateralized chicks that had to learn discrimination between food and nonfood while a model of an avian predator was moved overhead. Lateralized chicks learned faster and were more responsive to the model predator compared with weakly lateralized chicks, while, in the control experiment without the predator, no difference in learning ability was found. Similar cognitive advantages have also been identified in strongly lateralized parrots (Magat & Brown 2009). The dual processing hypothesis was recently tested in fishes comparing topminnow from lines selected for high and low degrees of laterality in conditions requiring attention to be shared between two concurrent tasks. In one experiment (Dadda & Bisazza 2006a), fishes were trained to enter in a compartment adjacent to the home tank to capture live brine shrimp and then were tested in either the presence or absence of a live predator situated at some distance. When the predator was present, and subjects were required to share attention between vigilance and prey capture, topminnow of both lateralized lines (LD and RD, collectively named LAT) were twice as fast at catching shrimp than nonlateralized fishes of the nonlateralized (NL) line, while no difference in capture rate was recorded in the condition without predator. A detailed analysis of fish movements revealed that LAT fishes tended to monitor the predator with one eye (the right eye in LD and the left in RD fishes) and to use the opposite eye for catching prey, whereas NL fishes switched between tasks, using either eye for both functions. In a second study (Dadda & Bisazza 2006b), LAT female topminnows were found to be more efficient foragers than NL females when they had to share attention between finding food and avoiding unsolicited male mating attempts, whereas no difference was found when females could forage undisturbed. A better performance in multitasking may not be the sole advantage of cerebral lateralization in topminnows. In one study, schools of LAT fishes showed significantly more cohesion and coordination than schools of NL fishes (Bisazza & Dadda 2005). Moreover, in schools composed of both LAT and NL fishes, the NL fishes were more often at the periphery of the school, while LAT fishes occupied the center, a position normally safer and energetically less expensive (Bumann et al. 1997; Svendsen et al. 2003). These experiments

BLBK374-14

BLBK374-Brown

314

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

were conducted in a novel tank, and therefore the possibility that fishes were performing a concurrent task (predator vigilance) while schooling cannot entirely be ruled out. In another experiment, LAT fishes proved to be better than NL fishes when using geometric cues to reorient themselves in a small environment (Sovrano et al. 2005). When taken together, these results suggest that lateralized fishes are better at coordinating information that arrives from each eye and integrating it into an appropriate behavioral response. Escape performance has been found to be positively correlated with the degree of lateralization in another teleost, the shiner perch, C. aggregata. Fishes showing higher scores in the detour test had shorter latencies in C-start response and hence traveled longer escape distances compared with less lateralized fishes (Dadda et al. 2010a). Differentiation between individuals and populations that are repeatedly observed in fish begs the question as to what kinds of environments are likely to favor the evolution of strongly lateralized phenotypes. Studies conducted on poeciliids certainly suggest that predation pressure is a key ecological factor (Brown et al. 2004), but even the simple addition of physical complexity can influence laterality (Bibost & Brown 2010). Indeed, any factor that enhances the cognitive load on a given species should theoretically favor strongly lateralized individuals. For example, examination of a wide range of parrot species shows that strongly lateralized parrots nearly always rely on beak–foot coordination during foraging, whereas those that use the beak alone tend to be nonlateralized (Magat & Brown 2009). Whether the cognitive load of eye–foot–beak coordination or the complexity of the foraging task is driving this differentiation is unclear. Research conducted on cichlids suggests that social and environmental complexities are correlated with larger brains (Gonzalez-Voyer et al. 2009), and we hypothesize that these species are also likely to be more strongly lateralized. Theory suggests that maximizing the efficiency of the brain should take place before the energetically extreme measure of enlarging it takes place. It may well be that habitat complexity in general and the manner in which species engage with complexity drives the evolution of laterality.

14.4.2

Costs of cerebral lateralization

In organisms with laterally placed eyes such as fish, the complementary specialization of hemispheres translates into differential responsiveness to sensory input on the left and right sides of the body. Toads, for example, are more likely to strike at a prey moving in their right lateral field of vision while agonistic responses are delivered preferentially to a conspecific seen on their left side (Vallortigara et al. 1998). Remarkably, toads are more likely to react when a predator appears from their left side than their right side (Lippolis et al. 2002). Similar asymmetries in behavioral responses have been found in other vertebrates (e.g., Vallortigara & Andrew 1994; Deckel 1995). The position in which biologically relevant stimuli will appear within the visual field is frequently unpredictable and it is easy to imagine the potential disadvantages derived from having side biases in the latency and efficiency to react to a particular class of stimuli, as well as the possibility for competitors, predators, or prey to exploit such asymmetries (see, for example, Takahashi & Hori 1994). However, until recently, no studies have directly tested whether the left–right differences in the way an animal analyzes and responds to environmental stimuli translates into a disadvantage for more lateralized individuals. Agrillo et al. (2009) compared topminnow from selected lines for their latency to react to a predatory stimulus appearing on the right or the left visual

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

315

hemifield. No difference was found between LAT and NL fishes or between the two eyes in escape performance. However, the setting used in this study was very simplified and did not permit a fine measurement, so further investigation is needed before a firm conclusion can be reached. However, two recent experiments on the same species have provided some evidence that marked lateralization can hinder performance when a task requires communication and cooperation among hemispheres (Dadda et al. 2009). In both situations the visual inputs were divided between the visual fields so that each eye (and contralateral hemisphere) had access to only half of the information needed to perform the task. The first experiment was an adaptation of the line-bisecting task, a widely used neuropsychological test. When right-handed human subjects are required to indicate the middle of a straight line, they tend to bisect slightly to the left, an effect that is commonly ascribed to the right-hemisphere dominance for spatial tasks (reviewed in Jewell & McCourt 2000). Subjects had to select the middle door in a row of nine in order to rejoin their social group. In this test, while NL fishes quickly learnt to use the central door, fishes from LD and RD lines made systematic errors of around 10% to the left or the right of the central door, respectively. During the task each half of the row was normally seen with a different eye and, as suggested for humans, it is likely that this drawback is the consequence of a greater degree of hemispheric dominance for spatial tasks in LAT fishes. In the second experiment, an isolated individual could choose between two shoals differing in quality (number and size of fishes). The subject emerged into the choice area from a narrow corridor so that at the moment of choice it saw each shoal with a different eye. NL fishes chose the high-quality shoal significantly more often than the LAT fishes, which in most cases chose the option seen with the eye that in their selection line was dominant for analyzing social stimuli, irrespective of its relative quality. The most likely explanation for these results is that information relative to the properties of the stimulus is confined, at least initially, to the hemisphere that directly receives the visual input. Therefore, in LAT fishes, during the decision phase, the hemisphere dominant for analyzing social stimuli can only, or predominantly, access to the information it receives from the contralateral eye. Similar negative effects of strong laterality occur in another fish, B. episcopi. Fishes from high-predation populations are, on average, more lateralized than their low-predation counterparts (Brown et al. 2004, 2007a), but they take longer to complete radial maze task as their laterality hinders an efficient movement toward the rewarded arm (Brown & Braithwaite 2005). A reader not familiar with lateralization literature might be puzzled by the poor integration of information reaching the two eyes that was observed among lateralized fishes in these experiments. However, the results are consistent with current knowledge of the way the visual system of fish integrates the two lateral inputs. The left- and right-eye systems can operate quite independently, as shown by the fact that fish trained monocularly to discriminate between two stimuli can simultaneously learn one stimulus as positive with one eye and negative with opposite eye (Ingle 1968). In general, experiments involving subjects trained monocularly to discriminate patterns have shown that interocular information transfer is slow and incomplete (McCleary 1960; Mark 1966; Ingle 1968). It is not easy to guess how relevant these drawbacks are to an individual’s fitness in its natural environment. Laboratory experiments suggest that lateralized fishes may make frequent suboptimal decisions about mates, prey, shoals, or refuges when they have to

BLBK374-14

BLBK374-Brown

316

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

take a quick decision and the alternative options are placed at the opposite sides of the body. Perhaps this situation is not so uncommon in their natural habitat. Most fishes have a visual field covering almost 360◦ and the frontal overlap of the opposite visual fields is usually around 10◦ (Collin & Shand 2003; Fig. 14.1). Thus, the probability that two stimuli fall into the two opposite visual hemifields should be relatively high. Despite the potential drawbacks, the cost–benefit analysis does seem to favor laterality most of the time as evidenced by its widespread occurrence across and increasingly large array of taxa.

14.4.3 Maintenance of intraspecific variability in the degree of lateralization Altogether, the picture emerging from the studies of laterality in fishes indicates that the advantages associated with having an asymmetric brain, such as the possibility of processing multiple information flows in parallel, may be balanced by some ecological disadvantages associated with left–right differences in the response to stimuli. The relative costs and benefits of lateralization are expected to vary with ecological factors (structure of habitat, predation risk, social density, food abundance, etc.) and different degrees of lateralization should be favored in different populations in relation to the relative importance of the various factors. For example, better schooling performance and efficiency in multitasking, and therefore a strong lateralization, should be favored under high-predation regimes but not necessarily in low-predation populations, a hypothesis that received support from a field study (Brown et al. 2004). The advantages and disadvantages of laterality may also vary temporally or spatially within a single population or with sex and developmental stages, thus contributing to maintenance of substantial phenotypic diversity for this trait. In some instances where social environmental factors are playing a key role in maintaining laterality within a population (perhaps in strongly schooling species for example), laterality may be under frequency-dependent selection (Brown 2005). Clearly, further research is required to determine the relative costs and benefits of lateralization in wild fishes and to identify ecological correlates. Further variability in laterality is likely generated through the manner in which individuals perceive and analyze stimuli both in the laboratory and in the natural world. Fishes with experience with predators will perceive them as threats, but predator-na¨ıve fishes may not (Brown & Warburton 1999). The latter may gradually shift their perception of threat as they learn to associate predators with danger and thus the hemisphere they use to analyze that information may also shift (De Santi et al. 2000). Similar changes are likely to be associated with habituation (Cantalupo et al. 1995). We have already discussed a related phenomenon as social partners become increasingly familiar and subjects shift their eye preferences accordingly. We have barely begun to explore the source of variance that relates to changes in perception as a result of learning but it may well explain variation in laterality that has previously been attributed to size, age, or developmental stage. In any case, it clearly adds yet another level of complexity to the puzzle.

14.4.4

Evolutionary significance of population biases in laterality

Experiments with fishes and birds suggest that a lateralized brain confers enhanced cognitive efficiency although this advantage is sometimes countered by disadvantages associated with

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

317

poorer interhemispheric integration. While these factor could explain the large prevalence of lateralization as well as the maintenance of variation in the degree of asymmetry, they cannot account for the fact that many populations show directional asymmetries in laterality, i.e., that most individuals show the same direction of bias (reviews in Bisazza et al. 1998b; Vallortigara & Rogers 2005). Indeed, the greater cognitive efficiency of lateralized brains appears to be independent of the direction of asymmetry, LD and RD individuals being normally equally efficient. The existence of consistent population biases might even convey a specific disadvantage, not present at the individual level, as it could make individual behavior more predictable to other organisms, for example allowing predators to learn the most frequent direction of escape in their prey. Equally, in strongly shoaling species where social synchrony is paramount, predictability may convey a very significant fitness advantage. Rogers (1989) has proposed an adaptive explanation, suggesting that alignment of laterality direction may occur as the consequence of the need to coordinate social activities with other individuals. Analysis has indicated that population-level lateralization can arise as an evolutionarily stable strategy when the benefit to an asymmetrical individual of coordinating with others of its same laterality equals the costs arising from predators having more success with the more common prey type (Ghirlanda & Vallortigara 2004). However, empirical evidence in support of this hypothesis is equivocal. Bisazza et al. (2000a) examined 16 species of fishes and found evidence that population biases were more frequent in species with a strong shoaling tendency than in solitary ones, but this finding requires confirmation using a larger sample size and correcting for phylogeny. Conversely, Bisazza & Dadda (2005) found no difference in schooling efficiency when comparing groups composed of female topminnow of mixed laterality (LD and RD) and groups of females with the same laterality. Several authors have argued that there might be common patterns of lateralization across different vertebrate classes; for example, social stimuli are processed by the right hemisphere in fishes, amphibians, birds, and mammals (Andrew 2000; Rogers 2002). Emerging data on fishes certainly challenges this theory showing that strength and direction of lateralization can vary not only among species but also among natural populations or laboratory strains of the same species. It is possible that some similarity among distantly related species is a consequence of the genetic mechanisms implicated in the development of the left–right axis, which is fundamentally the same in all vertebrates. Genes of the Nodal family determine the same directional asymmetries (e.g., liver on the right) in more than 99% of individuals. These genes are also implicated, at least in fishes, in early establishment of left–right anatomical asymmetries of the brain (Barth et al. 2005; see also Subsection 14.3.1). We argue that in the absence of strong selective forces acting on the direction of lateralization (as probably occurs for captive-reared populations), these shared genetic mechanisms may produce some phenotypes more commonly than others in all species.

14.5

Summary and future research

It is clear that laterality is a highly pervasive trait in the behavior of fishes shaping everything from personality traits to aggression and navigation. We have shown that laterality is at

BLBK374-14

BLBK374-Brown

318

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

least partially heritable and also influenced by exposure to various environmental variables during development (e.g., predators or complex environments). As such it has all the trademarks of a trait that is likely subjected to natural selection and indeed the vast amount of variability observed at both the individual and population levels suggest that this is the case. While some of the potential benefits of laterality are starting to emerge, we know very little about the potential costs. Clearly, any cost–benefit analysis has to be conducted in the appropriate context and therefore we expect laterality to reflect contemporary environmental conditions as well as the phylogenetic history of the organism. Indeed, trying to establish what environmental factors are associated with laterality in wild populations will certainly aid us in identifying likely constraints. Given the emphasis on laboratory-reared fishes to date, future work needs to examine laterality in a wide range of wild populations in a diverse array of species. Strong population skews in laterality suggest that it may well be under directional selection in some environments, while in others bimodal distributions are indicative of disruptive selection. In the case of the latter, the theorized benefits of having a strongly lateralized brain could drive laterality to the left or right with equal probability. It may well be that the direction of skew observed in a given population is simply a by-product of founder effects, genetic drift, or related factors. Alternatively, population skew may be an emergent property of group living, particularly in those populations where social synchrony is vital. Further analyses of ecological factors shaping laterality at the population level need to be conducted. Laterality on an individual level may also be viewed within a variable life-history framework, a possibility that has largely been ignored to date. Fishes have long been studied for the existence of alternative life-history strategies, (e.g., sneaker male in guppies and salmon) and it may well be that variation in laterality could provide another example of evolution at work on this fine scale. Examination of a suit of behavioral traits in strongly and weakly lateralized fishes occupying the same habitat is likely to be fruitful in this respect. Thus, merging the personality and laterality frameworks presents an exciting avenue for future research. Researchers still know very little about the proximate mechanisms of laterality. In the genomic era, it is possible to scan entire genomes looking quantitative trait loci and determine if the same genes are responsible for lateralization in a range of fish species (zebrafish, sticklebacks, and guppies are obvious target species). Moreover, once this information is assimilated, we can begin to guess what the downstream consequences for gene expression might be. Additionally, as researchers become increasingly familiar with the structure and function of the fish brain, they may be able to make further inroads into the neural and developmental bases of laterality.

Acknowledgments Culum Brown was supported by an Australian Research Fellowship from the Australian Research Council. Angelo Bisazza was supported by an Italian Ministry of Education grant. We thank Marco Dadda for help with the figures and Lesley Rogers and Kevin Laland for their helpful comments on the chapter.

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

319

References Agrillo, C., Dadda, M. & Bisazza, A. (2009) Escape behaviour elicited by a visual stimulus. A comparison between lateralised and non-lateralised female topminnows. Laterality, 14, 300–314. Andrew, R.J. (1991) The nature of behavioural lateralization in the chick. In: R.J. Andrew (ed) Neural and Behavioural Plasticity: The Use of the Chick as a Model. Oxford University Press, Oxford, pp. 536–554. Andrew, R.J. (2000) Origins of cerebral lateralisation: evidence from fish and birds. European Journal of Neuroscience, 12, 192–192. Andrew, R.J. (2002) The earliest origins and subsequent evolution of lateralization. In: L.J. Rogers & R.J. Andrew (eds) Comparative Vertebrate Lateralization, pp. 70–93. Cambridge University Press, Cambridge. Andrew, R.J. (2009) Origins of asymmetry in the CNS. Seminars in Cell Developmental Biology, 20, 485–490. Andrew, R.J., Osorio, D. & Budaev, S. (2009) Light during embryonic development modulates patterns of lateralization strongly and similarly in both zebrafish and chick. Philosophical Transactions of the Royal Society of London, Series B – Biological Sciences, 364, 983–989. Babcock, L.E. (1993) Trilobite malformations and the fossil record of behavioral asymmetry. Journal of Palaeontology, 67, 217–229. Barth, K.A., Miklosi, A., Watkins, J., Bianco, I.H., Wilson, S.W. & Andrew, R.J. (2005) fsi Zebrafish show concordant reversal of laterality of viscera, neuroanatomy, and a subset of behavioral responses. Current Biology, 15, 844–850. Bauer, R.H. (1993) Lateralization of neural control for vocalisation by the frog (Rana pipiens). Psychobiology, 21, 243–248. Bell, A.M., Hankison, S.J. & Laskowski, K.L. (2009) The repeatability of behaviour: a meta-analysis. Animal Behaviour, 77, 771–783. Benton, A.I. & Menefee, F.L. (1957) Handedness and right-left discrimination. Child Development, 28, 237–242. Bibost, A-L. & Brown, C. (2010) Environmental enrichment affects laterality in rainbowfish, Melanotaenia spp. Unpublished data. Bisazza, A., Cantalupo, C., Capocchiano, M. & Vallortigara, G. (2000a) Population lateralisation and social behaviour: a study with 16 species of fish. Laterality, 5, 269–284. Bisazza, A., Cantalupo, C. & Vallortigara, G. (1997c) Lateral asymmetries during escape behaviour in a species of teleost fish (Jenynsia lineata). Physiology and Behavior, 61, 31–35. Bisazza, A. & Dadda, M. (2005) Enhanced schooling performance in lateralized fishes. Proceedings of the Royal Society of London, Series B – Biological Sciences, 272, 1677–1681. Bisazza, A., Dadda, M. & Cantalupo, C. (2005) Further evidence for mirror-reversed laterality in lines of fish selected for leftward or rightward turning when facing a predator model. Behavioural Brain Research, 156, 165–171. Bisazza, A., Dadda, M., Facchin, L. & Vigo, F. (2007) Artificial selection on laterality in the teleost fish Girardinus falcatus. Behavioural Brain Research, 178, 29–38. Bisazza, A. & De Santi, A. (2003) Lateralization of aggression in fish. Behavioural Brain Research, 141, 131–136. Bisazza, A., De Santi, A. & Vallortigara, G. (1999) Laterality and cooperation: Mosquitofish move closer to a predator when the companion is on their left side. Animal Behavior, 57, 1145–1149. Bisazza, A., Facchin, L., Pignatti, R. & Vallortigara, G. (1998a) Lateralization of detour behaviour in poeciliid fish: The effect of species, gender and sexual motivation. Behavioural Brain Research, 91, 157–164. Bisazza, A., Facchin, L. & Vallortigara, G. (2000b) Heritability of lateralization in fish: concordance of right-left asymmetry between parents and offspring. Neuropsychologia, 38, 907–912. Bisazza, A., Lippolis, G. & Vallortigara, G. (2001a) Lateralization of ventral fins use during object exploration in the blue gourami (Trichogaster trichopterus). Physiology & Behavior, 72, 575–578.

BLBK374-14

BLBK374-Brown

320

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Bisazza, A., Pignatti, R. & Vallortigara, G. (1997a) Detour tests reveal task- and stimulus-specific neural lateralization in mosquitofish (Gambusia holbrooki). Behavioural Brain Research, 89, 237–242. Bisazza, A., Pignatti, R. & Vallortigara, G. (1997b) Laterality in detour behaviour: interspecific variation in poeciliid fish. Animal Behaviour, 54, 1273–1281. Bisazza, A., Rogers, L.J. & Vallortigara, G. (1998b) The origins of cerebral asymmetry: A review of evidence of behavioural and brain lateralization in fishes, reptiles and amphibians. Neuroscience and Biobehavioral Reviews, 22, 411–426. Bisazza, A., Sovrano, V.A. & Vallortigara, G. (2001b) Consistency among different tasks of left-right asymmetries in lines of fish originally selected for opposite direction of lateralization in a detour task. Neuropsychologia, 39, 1077–1085. Broca, P. (1861) Remarques sur le si´ege de la facult´e du langage articul´e, suivies d’une observation d’aph´emie (perte de la parole). Bulletin de la Soci´et´e Anatomique de Paris, 6, 330–357. Brown, C. (2002) Do female rainbowfish (Melanotaenia spp.) prefer to shoal with familiar individuals under predation pressure? Journal of Ethology, 20, 89–94. Brown, C. (2005) Cerebral lateralisation; social constraints and coordinated antipredator responses. Behavioral and Brain Sciences, 28, 591–592. Brown, C. & Braithwaite, V.A. (2005) Effects of predation pressure on the cognitive ability of the poeciliid Brachyraphis episcopi. Behavioral Ecology, 16, 482–497. Brown, C., Gardner, C. & Braithwaite, V.A. (2004) Population variation in lateralised eye use in the poeciliid Brachyraphis episcopi. Proceedings of the Royal Society of London, Series B – Biological Sciences, 271, S455–S457. Brown, C., Jones, F. & Braithwaite, V.A. (2005) In situ examination of boldness-shyness traits in the tropical poeciliid, Brachyraphis episcopi. Animal Behaviour, 70, 1003–1009. Brown, C., Jones, F.C. & Braithwaite, V.A. (2007b) Correlation between boldness and body mass in natural populations of the poeciliid Brachyrhaphis episcopi. Journal of Fish Biology, 71, 1590–1601. Brown, C. & Warburton, K. (1999) Differences in timidity and escape responses between predatorna¨ıve and predator-sympatric rainbowfish populations. Ethology, 105, 491–502. Brown, C., Western, J. & Braithwaite, V.A. (2007a) The influence of early experience on, and inheritance of, cerebral lateralization. Animal Behaviour, 74, 231–238. Budaev, S.V. & Andrew, R.J. (2009a) Shyness and behavioural asymmetries in larval zebrafish (Brachydanio rerio) developed in light and dark. Behaviour, 146, 1037–1052. Budaev, S.V. & Andrew, R.J. (2009b) Patterns of early embryonic light exposure determine behavioural asymmetries in zebrafish: a habenular hypothesis. Behavioural Brain Research, 200, 91–94. Bumann, D., Krause, J. & Rubenstein, D. (1997) Mortality risk of spatial positions in animal groups: the danger of being in front. Behaviour, 134, 1063–1076. Burt de Perera, T. & Braithwaite, V.A. (2005) Laterality in a non-visual sensory modality – the lateral line of fish. Current Biology, 15, R241–R242. Cantalupo, C., Bisazza, A. & Vallortigara, G. (1995) Lateralization of predator-evasion response in a teleost fish (Girardinus falcatus). Neuropsychologia, 33, 1637–1646. Cantalupo, C., Bisazza, A. & Vallortigara, G. (1996) Lateralization of dispalys during aggressive and courtship behaviour in the Siamese fighting-fish Betta splendens. Physiology and Behavior, 61, 249–252. Casperd, L.M. & Dunbar, R.I.M. (1996) Asymmetries in the visual processing of emotional cues during agonistic interactions by Gelada baboons. Behavioural Processes, 37, 57–65. Chiandetti, C. & Vallortigara, G. (2008) Effects of embryonic light stimulation on the ability to discriminate left from right in the domestic chick. Behavioural Brain Research, 198, 240–246. Chochon, F., Cohen, L., van de Moortele, P.F. & Dehaene, S. (1999) Differential contributions of the left and right inferior parietal lobules to number processing. Journal of Cognitive Neuroscience, 11, 617–630. Clayton, N. & Krebs, J. (1994) Memory for spatial and object-specific cues in food-storing and non-storing birds. Journal of Comparative Physiology A, 174, 371–379.

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

321

Collin, S.P. & Shand, J. (2003) Retinal sampling and the visual field in fishes In: S.P. Collin & J. Marchal (eds) Sensory Processing in Aquatic Environments, pp. 139–169. Springer, New York. Collins, R.L. (1975) When left-handed mice live in right-handed worlds. Science, 187, 181–184. Corballis, M.C. (2003) From mouth to hand: gesture, speech and the evolution of right-handedness. Behavioral and Brain Sciences, 26, 199–208. Corballis, M.C. (2009) The evolution and genetics of cerebral asymmetry. Philosophical Transactions of the Royal Society of London, Series B – Biological Sciences, 364, 867–879. Dadda, M. & Bisazza, A. (2006a) Does brain asymmetry allow efficient performance of simultaneous tasks? Animal Behaviour, 72, 523–529. Dadda, M. & Bisazza, A. (2006b) Lateralized female topminnows can forage and attend to a harassing male simultaneously. Behavioral Ecology, 17, 358–363. Dadda, M., Domenichini, A., Piffer, L., Argenton, F. & Bisazza, A. (2010b) Early differences in epithalamic left-right asymmetry influence lateralization and personality of adult zebrafish. Behavioural Brain Research, 206, 208–215. Dadda, M., Koolhaas, W.H. & Domenici, P. (2010a) Behavioural asymmetry affects escape performance in a teleost fish. Biology Letters, 6, 414–417. Dadda, M., Zandon`a, E., Agrillo, C. & Bisazza, A. (2009) The costs of hemispheric specialization in a fish. Proceeding of the Royal Society of London, Series B – Biological Sciences, 276, 4399–4407. Dadda, M., Zandona, E. & Bisazza, A. (2007) Emotional responsiveness in fish from lines artificially selected for a high or low degree of laterality. Physiology & Behavior, 92, 764–772. Deckel, A.W. (1995) Laterality of aggressive responses in Anolis. Journal of Experimental Zoology, 272, 194–200. Denenberg, V.H., Garbanati, J., Sherman, G., Yutzey, D.A. & Kaplan, R. (1978) Infantile stimulation induces brain lateralization in rats. Science, 201, 1150–1152. Deng, C. & Rogers, L.J. (2002) Factors affecting the development of lateralization in chicks. In: L.J. Rogers & R.J. Andrew (eds) Comparative Vertebrate Lateralization, pp. 206–246. Cambridge University Press, Cambridge. De Santi, A. , Sovrano, V.A., Bisazza, A. & Vallortigara, G. (2001) Mosquitofish display differential left- and right-eye use during mirror-image scrutiny and predator-inspection responses. Animal Behaviour, 61, 305–310. De Santi, A., Bisazza, A., Cappelletti, M. & Vallortigara, G. (2000) Prior exposure to a predator influences lateralization of cooperative predator inspection in the guppy, Poecilia reticulata. Italian Journal of Zoology, 67, 175–178. Diaz, R., Agren, S.O., Blum, M. & Fuxe, K. (1995) Prenatal corticosterone increases spontaneous and D-amphetamine induced locomotor activity and brain dopamine metabolism in prepubertal male and female rats. Neuroscience, 66, 467–473. Domenici, P. & Blake, R.W. (1997) The kinematics and performance of fish fast-start swimming. Journal of Experimental Biology, 200, 1165–1178. Dukas, R. (2004) Causes and consequences of limited attention. Brain, Behavior and Evolution, 63, 197–210. Dukas, R. & Kamil, A.C. (2000) The cost of limited attention in blue jays. Behavioral Ecology, 11, 502–506. Ehret, G. (1987) Left hemisphere advantage in the mouse brain for recognizing ultrasonic communication calls. Nature, 325, 249–251. Facchin, L., Argenton, F. & Bisazza, A. (2009) Lines of Danio rerio selected for opposite behavioural lateralization show differences in anatomical left-right asymmetries. Behavioural Brain Research, 197, 157–165. Facchin, L., Bisazza, A. & Vallortigara, G. (1999) What causes lateralization of detour behavior in fish? Evidence for asymmetries in eye use. Behavioural Brain Research, 103, 229–234. Fine, M.L., McElroy, D., Rafi, J., King, C.B., Loesser, K.E. & Newton, S. (1996) Lateralization of pectoral stridulation sound production in the channel catfish. Physiology & Behavior, 60, 753–757. Floel, A., Knecht, S., Lohmann, H., Deppe, M., Sommer, J., Drager, B., Ringelstein, E.B. & Henningsen, H. (2001) Language and spatial attention can lateralize to the same hemisphere in healthy humans. Neurology, 57, 1018–1024.

BLBK374-14

BLBK374-Brown

322

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Fride, E. & Weinstock, M. (1988) Prenatal stress increases anxiety related behavior and alters cerebral lateralization of dopamine activity. Life Sciences, 42, 1059–1065. Ghirlanda, S. & Vallortigara, G. (2004) The evolution of brain lateralization: a game-theoretical analysis of population structure. Proceedings of the Royal Society of London, Series B – Biological Sciences, 271, 853–857. Godin, J.G.J. & Dugatkin, L.A. (1996) Female mating preference for bold males in the guppy, Poecilia reticulata. Proceedings of the National Academy of Science USA, 93, 10262–10267. Godin, J.G.J. & Smith, S.A. (1988) A fitness cost of foraging in the guppy. Nature, 333, 69–71. Gonzalez-Voyer, A., Winberg, S. & Kolm, N. (2009) Social fishes and single mothers: brain evolution in African cichlids. Proceedings of the Royal Society of London, Series B – Biological Sciences, 276, 161–167. Gross, M.R., Suk, H.Y. & Robertson, C.T. (2007) Courtship and genetic quality: asymmetric males how their best side. Proceedings of the Royal Society of London, Series B – Biological Sciences, 274, 2115–2122. Halpern, M.E., Gunturkun, O., Hopkins, W.D. & Rogers, L.J. (2005) Lateralization of the vertebrate brain: taking the side of model systems. Journal of Neuroscience, 25, 10351–10357. Hamilton, C.R. & Vermeire, B.A. (1988) Complementary hemispheric specialization in monkeys. Science, 242, 1691–1694. Hebets, E.A. (2005) Attention-altering interactions among signals in multimodal wolf spider courtship displays. Behavioral Ecology, 16, 75–82. Heuts, B.A. (1999) Lateralization of trunk muscle volume, and lateralization of swimming turns of fish responding to external stimuli. Behavioural Processes, 47, 113–124. Hori, M. (1993) Frequency-dependent natural selection in the handedness of scale-eating cichlid fish. Science, 260, 216–219. Ingle, D. (1968) Interocular integration of visual learning by goldfish. Brain, Behavior and Evolution, 1, 58–85. Jewell, G. & McCourt, M.E. (2000) Pseudoneglect: a review and meta-analysis of performance factors in line bisection tasks. Neuropsychologia, 38, 93–110. Johnsson, J.I., Sernland, E. & Blixt, M. (2001) Sex-specific aggression and antipredator behaviour in young brown trout. Ethology, 107, 587–599. Kaarthigeyan, J. & Dharmaretnam, M. (2005) Relative levels of motivation and asymmetries of viewing and detour task in guppies (Poecillia reticulata). Behavioral Brain Research, 159, 37–41. Kastner, S. & Ungerleider, L.G. (2003) Mechanisms of visual attention in the human cortex. Annual Review of Neuroscience, 23, 315–341. Kydd, E. & Brown, C. (2009) Loss of shoaling preference for familiar individuals in captive-reared crimson spotted rainbowfish Melanotaenia duboulayi. Journal of Fish Biology, 74, 2187–2195. Levy, J. (1977). The mammalian brain and the adaptive advantage of cerebral asymmetry. The New York Academy of Science, 299, 264–272. Lippolis, G., Bisazza, A., Rogers, L.J. & Vallortigara, G. (2002) Lateralization of predator avoidance in three species of toads. Laterality, 7, 163–183. Lippolis, G., Joss, J.M.P. & Rogers, L.J. (2009) Australian lungfish (Neoceratodus forsteri): a missing link in the evolution of complementary side biases for predator avoidance and prey capture. Brain, Behavior and Evolution, 73, 295–303. Magat, D. & Brown, C. (2009) Laterality enhances cognition in Australian parrots. Proceedings of the Royal Society of London, Series B – Biological Sciences, 276, 4155–4162. Magnhagen, C. & Staffan, F. (2005) Is boldness affected by group composition in young-of-the-year perch (Perca fluviatilis)? Behavioral Ecology and Sociobiology, 57, 295–303. Magurran, A.E. & Pitcher, T.J. (1987) Provenance, shoal size and the sociobiology of predator-evasion behaviour in minnow shoals. Proceedings of the Royal Society of London, Series B – Biological Sciences, 229, 439–465. Magurran, A.E. & Seghers, B.H. (1990) Population differences in predator recognition and attack cone avoidance in the guppy Poecilia reticulata. Animal Behaviour, 40, 443–452.

BLBK374-14

BLBK374-Brown

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Lateralization of Cognitive Functions in Fish

Char Count=

323

Mark, R. F. (1966) The tectal commissure and interocular transfer of pattern discrimination in cichlid fish. Experimental Neurology, 16, 215–225. McCleary, R.A. (1960) Type of response as a factor in interocular transfer in the fish. Journal of Comparative Physiological Psychology, 53, 311–321. McCormick, M.I. (1998) Behaviorally induced maternal stress in a fish influences progeny quality by a hormonal mechanism. Ecology, 79, 1873–1883. McManus, I.C. (2009) The history and geography of human handedness. In: I.E.C. Sommer & R.S. Kahn (eds) Language Lateralization and Psychosis, pp. 37–57. Cambridge University Press, Cambridge. Mench, J.A. & Andrew, R.J. (1986) Lateralization of a food search task in the domestic chick. Behavioral and Neural Biology, 46, 107–114. Metcalfe, N.B., Huntingford, F.A. & Thorpe, J.E. (1987) Predation risk impairs diet selection in juvenile salmon. Animal Behaviour, 35, 931–933. Miklosi, A. & Andrew, R.J. (1999) Right eye use associated with decision to bite in zebrafish. Behavioural Brain Research, 105, 199–205. Milinski, M. (1987) Tit for tat in sticklebacks and the evolution of cooperation. Nature, 325, 433–435. Nottebohm, F. (1971) Neural lateralization of vocal control in a passerine bird. I. Song. Journal of Experimental Zoology, 177, 229–261. Petersen, M.R., Beecher, M.D., Zoloth, S.R., Moody, D.B. & Stebbins, W.C. (1978) Neural lateralization of species-specific vocalizations by Japanese macaques (Macaca fuscata). Science, 202, 324–327. Phelps, E.A. O’Connor, K.J., Gatenby, J.C., Gore, J.C., Grillon, C. & Davis, M. (2001) Activation of the left amygdala to a cognitive representation of fear. Nature Neuroscience, 4, 437–441. Reddon, A.R. & Hurd, P.L. (2008) Aggression, sex and individual differences in cerebral lateralization in a cichlid fish. Biology Letters, 4, 338–340. Reddon, A.R. & Hurd, P.L. (2009) Acting unilaterally: why do animals with strongly lateralized brains behave differently than those with weakly lateralized brains? Bioscience Hypotheses, 2, 383– 387. R´eale, D., Reader, S.M., Sol, D., McDougall, P.T. & Dingemanse, N.J. (2007) Integrating animal temperament within ecology and evolution. Biological Reviews, 82, 291–318. Reist, J.D., Bodaly, R.A., Fudge, R.J.P., Cash, K.J. & Stevens, T.V. (1987) External scarring of whitefish, Coregonus nasus and C. clupeaformis complex, from the western Northwest Territories, Canada. Canadian Journal of Zoology, 65, 1230–1239. Rogers, L.J. (1989) Laterality in animals. International Journal of Comparative Psychology, 3, 5–25. Rogers, L.J. (2000) Evolution of hemispheric specialization: advantages and disadvantages. Brain and Language, 73, 236–253. Rogers, L.J. (2002) Advantages and disadvantages of lateralization. In: Comparative Vertebrate Lateralization, pp. 126–153. Cambridge University Press, Cambridge. Rogers, L.J. & Anson, J.M. (1979) Lateralization of function in chicken forebrain. Pharmacology Biochemistry and Behavior, 10, 679–686. Rogers, L. J. & Deng, C. (2005) Corticosterone treatment of the chick embryo affects light-stimulated development of the thalamofugal visual pathways, Behavioural Brain Research, 159, 63–71. Rogers, L.J., Zappia, J.V. & Bullock, S.P. (1985) Testosterone and eye-brain asymmetry for copulation in chickens. Experientia, 1, 1447–1449. Rogers, L.J., Zucca, P. & Vallortigara, G. (2004) Advantages of having a lateralized brain. Proceedings of the Royal Society of London, Series B – Biological Sciences, 271, S420–S422. Sovrano, V.A., Bisazza, A. & Vallortigara, G. (2001) Lateralization of response to social stimuli in fishes: a comparison between different methods and species. Physiology & Behavior, 74, 237–244. Sovrano, V.A., Dadda, M. & Bisazza, A. (2005) Lateralized fish perform better than nonlateralized fish in spatial reorientation tasks. Behavioural Brain Research, 163, 122–127. Sovrano, V.A., Rainoldi, C., Bisazza, A. & Vallortigara, G. (1999) Roots of brain specializations: preferential left-eye use during mirror-image inspection in six species of teleost fish. Behavioural Brain Research, 106, 175–180.

BLBK374-14

BLBK374-Brown

324

May 13, 2011

17:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Svendsen, J.C.S., Kov, J.S., Bildsoe, M. & Teffensen, J.F.S. (2003) Intra-school positional preference and reduced tail beat frequency in trailing positions in schooling roach under experimental conditions. Journal of Fish Biology, 62, 834–846. Takahashi, S. & Hori, M. (1994) Unstable evolutionary strategy and oscillation: a model of lateral asymmetry in scale eating cichlids. The American Naturalist, 144, 1001–1020. Takeuchi, Y. & Hori, M. (2008). Behavioural laterality in the shrimp-eating cichlid fish Neolamprologus fasciatus in Lake Tanganyika. Animal Behaviour, 75, 1359–1366. Vallortigara, G. & Andrew, R. J. (1994) Differential involvement of right and left hemisphere in individual recognition in the domestic chick. Behavioural Processes, 33, 41–58. Vallortigara, G. & Rogers, L. J. (2005) Survival with an asymmetrical brain: advantages and disadvantages of cerebral lateralization. Behavioral and Brain Sciences, 28, 575–633. Vallortigara, G., Rogers, L.J., Bisazza, A., Lippolis, G. & Robins, A. (1998) Complementary right and left hemifield use for predatory and agonistic behaviour in toads. Neuroreport, 9, 3341–3344. Ward, A. J.W., Thomas, P., Hart, P.J.B. & Krause, J. (2004) Correlates of boldness in three-spined sticklebacks (Gasterosteus aculeatus). Behavioral Ecology and Sociobiology, 55, 561–568. Wilson, D.S., Clark, A.B., Coleman, K. & Dearstyne, T. (1994) Shyness and boldness in humans and other animals. Trends in Ecology and Evolution, 9, 442–446. Westin, L. (1998) The spawning migration of European silver eel (Anguilla anguilla L.) with particular reference to stocked eel in the Baltic. Fisheries Research, 38, 257–270. Zappia, J.V. & Rogers, L.J. (1987) Sex differences and reversal of brain asymmetry by testosterone in chicks. Behavioral Brain Research, 23, 261–267.

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 15

Brain and Cognition in Teleost Fish Cristina Broglio, Antonia G´omez, Emilio Dur´an, Cosme Salas and Fernando Rodr´ıguez

15.1

Introduction

The evolution of brain and cognition has been tacitly treated as a special case by the traditional theories of vertebrate evolution. It has been described as a linear series of increasing complexity and advancement, from ‘lower’ to ‘higher’ groups – fishes, amphibians, reptiles, birds and finally mammals, to reach the superior cerebral and cognitive levels of humans with the incorporation of new neural structures, circuits and mechanisms in successive steps or evolutionary stages (Papez 1929; Ari¨ens-Kappers et al. 1936; Crosby & Schnitzlein 1983; MacLean 1990). According to these evolutionary notions that dominated neuroscience until recent times, fishes would be the ‘most primitive’ or ‘least evolved’ vertebrate group, with only relatively simple neural circuits sustaining elemental ‘instinctive’ or ‘reflex’ forms of behaviour, in clear contrast with ‘more recent’ vertebrate groups. Mammals, and specially humans, would distinguish by sophisticated cognitive capacities and intelligent behaviour, particularly associated with the expansion of the six-layered neocortex (Papez 1929; Romer 1962; Jerison 1973; MacLean 1990). In fact, the forebrain of fishes was viewed as consisting of a sub-pallium (‘paleostriatum’) and a very small pallium (‘paleocortex’), both dominated by olfactory inputs (Papez 1929; Ari¨ens-Kappers et al. 1936; Crosby & Schnitzlein 1983; MacLean 1990). Therefore, the ‘archistriatum’, ‘neostriatum’ and ‘archicortex’ (the proposed antecedents of the mammalian pallial amygdala, caudate and putamen, and hippocampus, respectively), structures that would have evolved later, would be completely absent in the telencephalon of fishes. Fortunately, these anagenetic, anthropocentristic ideas, which were consequence of the hybridising of some poorly understood Darwinian concepts of evolution with the Aristotelian idea of Scala naturae (Hodos & Campbell 1969; Deacon 1990), are being replaced by a new understanding on the evolution of brain, behaviour and cognition in vertebrates. Recent developmental, neuroanatomical and functional data show that vertebrate brain and behaviour evolution has been far more conservative than previously thought (for revisions, see Nieuwenhuys et al. 1998; Salas et al. 2003; Broglio et al. 2005; Butler &

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-15

BLBK374-Brown

326

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

OB Tel OT CB

Shark

Chondrichthyans

Sturgeon

Teleost

Osteichthyes

Frog

Amphibians

Turtle

Bird

Amniotes

Fig. 15.1 Schematic representation of the main brain divisions in some groups of extant vertebrates. Cb, Cerebellum; OB, Olfactory bulbs; OT, Optic tectum; Tel, Telencephalon.

Hodos 2005; Striedter 2005). Despite some notable morphological and cytoarchitectural differences, the central nervous system of every vertebrate group is organised in equivalent broad areas (Fig. 15.1), and even in the telencephalon, there is a high degree of conservation in the genes that specify the different brain regions, the cell groups and their connections, the main ascending and descending pathways and the pattern of distribution of histochemical and molecular markers (Northcutt & Braford 1980; Braford 1995; Northcutt 1995, 2008; Nieuwenhuys et al. 1998; Wulliman & Mueller 2004; Nieuwenhuys 2009). The most notable morphological variation is observed in the telencephalon of ray-finned fishes (e.g. teleosts), which consists of two massive hemispheres separated by a single ventricular cavity and most of its surface ependymal instead of pial (Nieuwenhuys et al. 1998; Butler & Hodos 2005; Nieuwenhuys 2011). These differences are due to a major variation during the embryonic development, i.e. the eversion (or bending outward) of the prosencephalic alar plate in ray-finned fishes, instead of the evagination (bending inward) that characterises the development of this brain structure in every other vertebrate group, i.e. agnathans, chondryctians and sarcopterygians, including the lobe-finned fishes and tetrapods (Fig. 15.2). However, as indicated in the preceding text, the telencephalon, whether everted or evaginated, presents a comparable pattern of organisation, with pallial and sub-pallial zones (Northcutt & Braford 1980; Wulliman & Mueller 2004; Northcutt 2008; Braford 2009; Nieuwenhuys 2011), and, also in fishes, the olfactory areas represent only a limited portion of this structure. Moreover, phylogenetic analysis indicates that the main subdivisions of the dorsal pallium in the actinopterygian telencephalon, although topologically inverted in the medial-to-lateral axis, are likely homologous to the hippocampus, the amygdala and the neocortex of tetrapods (Northcutt & Braford 1980; Wulliman & Mueller 2004; Butler & Hodos 2005; Yamamoto et al. 2007; Northcutt 2008; Braford 2009; Mueller & Wulliman 2009; Fig. 15.2). In the present work, we review recent behavioural and functional data that also challenge the traditional notions on brain and cognition evolution, as it shows that fishes share

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Brain and Cognition in Teleost Fish

Neural tube

327

Dp

P3 P2 P1 V

Mp Lp S V

V Str

Evagination

Eversion P3 P2

P3

P2

Frog

P1

V

P1

V

Acc

Dm Vd P2 P1

V

P2

P3

P1 V

Vv

Actinopterygians

Non-actinopterygians (a)

Dc

Dld

V

P3

V

Dd

Dlv

Goldfish (b)

Fig. 15.2 (a) Schematic representation of the processes involved in the embryonary development of the telencephalon in actinopterygian and non-actinopterygian vertebrates. P1, P2, and P3 correspond to the three main subdivisions of the pallium. (b) Coronal sections of brains representative of the evaginated (frog) and everted (goldfish) patterns of telencephalon development. Note that in actinopterygian fish the inner telencephalic ventricles are not present, and that the position of the main dorsal pallium subdivisions appears inverted relative to the evaginated telencephalon. Acc, nucleus accumbens; Dc, central division of area dorsalis; Dd, dorsal division of area dorsalis; Dld, dorsal subdivision of lateral division of area dorsalis; Dlv, ventral subdivision of lateral division of area dorsalis; Dm, medial division of area dorsalis; Dp, posterior zone of area dorsalis; Lp, lateral pallium: Mp, medial pallium; S, septal nuclei; Str, striatum; v, ventricle; Vd, dorsal nucleus of area ventralis; Vv, ventral nucleus of area ventralis.

complex learning and memory capabilities with land vertebrates, based on equivalent neural mechanisms and brain systems. In the following sections, we discuss recent evidence on the neural basis of associative, emotional and spatial learning in teleost fishes, showing that particular brain areas are components of separate memory systems.

15.2

Classical conditioning

Fishes show reliable classical conditioning in a variety of reflexes and response systems, and in a wide range of conditions. Similar to mammals, they show sensitivity to the predictive relationship between the conditioned and the unconditioned stimuli and exhibit overshadowing, blocking, autoshaping, and higher order conditioning (Davey 1989; Overmier & Hollis 1990). In addition, recent evidence suggests that at least some of the neural mechanisms underlying these learning phenomena in teleost fishes are shared with other vertebrates (G´omez et al. 2010).

BLBK374-15

BLBK374-Brown

328

15.2.1

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Delay motor classical conditioning and teleost fish cerebellum

In a typical classical conditioning paradigm (e.g. eyeblink classical conditioning), animals learn to express a conditioned eyeblink response (CR) to a predictive or conditioned stimulus (CS), for example a tone or a light, that is paired with a significant unconditioned stimulus (US), such as a periorbital shock or an air puff, which reliably elicits a reflexive unconditioned eyeblink response (UR). Following repeated CS–US pairings, the presentation of the previously neutral CS now elicits a learned eyeblink (conditioned response, CR) prior to US onset. In mammals, the essential circuit for acquisition and performance of this simple, learned reflex resides in the cerebellum and related brainstem structures (for a review, see Thompson & Steinmetz 2009). It is well known that the neural basis of classical conditioning involves the convergence, on individual cells, of synaptic inputs from pathways that process the CS and the US (Christian & Thompson 2003). The essential CS pathway for eyeblink conditioning consists of mossy fibres that arise from many sources in the brainstem and spinal cord and project onto the cerebellar cortex and interpositus nucleus. Instead, the US pathway involves the climbing fibres that originate in the inferior olivary nucleus and reach the Purkinje cells as well as the deep cerebellar nuclei. So, the CS and the US converge in specific regions of the cerebellum that transmit the entire output to the pathways responsible for motor control. In this sense, several lines of evidence indicate that the plastic changes underlying eyeblink conditioning occur in both the anterior interpositus nucleus and the cerebellar cortex (Christian & Thompson 2003; Thompson & Steinmetz 2009). The basic connectional organisation pattern in the cerebellum of teleosts is similar to that reported for mammals (Wullimann & Northcutt 1988; Meek & Nieuwenhuys 1998; Ikenaga et al. 2006; Meek et al. 2008; Xue et al. 2008), suggesting that it could also support the convergence of CS and US required for classical conditioning to occur. In fact, the paired stimulation of climbing and parallel fibres in the mormyrid fish cerebellum produces plastic changes in the synapses of the parallel fibres onto Purkinje cells (Han et al. 2007). A set of recent studies shows that, as in mammals, the teleost fish cerebellum is involved in the classical conditioning of motor responses. One of these experiments (G´omez et al. 2010) was aimed at determining if the goldfish cerebellum plays a critical role in classical conditioning, and whether the forebrain structures are, as has been demonstrated in mammals, not necessary for delay conditioning. Goldfish were trained in a delay classical conditioning procedure analogous to the eyeblink model commonly used with mammals. A red light was employed as CS and a mild electric shock as US (Fig. 15.3a). A delay paradigm was used, i.e. the CS onset preceded the US, but both stimuli overlapped in time and co-terminated. The US evoked a robust defensive withdrawal reflex, characterised by a consistent saccadic eye movement (unconditioned eyeblink-like response, UR). Over the course of training the control goldfish showed a progressive and significant increase in the percentage of CRs to the CS presentation (Fig. 15.3b). In the control goldfish, as in mammals, the percentage of CRs increased with paired CS–US presentations and decreased with CS alone (extinction sessions) or unpaired CS–US presentations (pseudoconditioning procedure; Fig. 15.3d). The sensitivity of the fish’s performance to these variations in training conditions (i.e. in the CS–US relationships) indicates that in teleost fishes, as in mammals, the acquisition of the CR is governed by associative rules, enabling to discard the possibility of pseudoconditioning biases or other non-associative mechanisms (Kehoe & Macrae 2002).

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Brain and Cognition in Teleost Fish

(a)

(c)

US

Magnet

329

(d) CCb VL

OT

Tel

60

HL

Conditioned responses (%)

BLBK374-15

CS Hall-effect sensor UR (b) CR

Sh Cb Tel Pseudo

50

40

30

20 5° 100 ms

CS

US

1

1 2 3 4

1 2 3 4

Habituation

Acquisition

Extinction

Sessions

Fig. 15.3 Effects of cerebellum and telencephalon lesions on delay classical conditioning of an eyeblink-like response in goldfish. (a) Diagram of the experimental setting showing the CS, US and the eye movement recording system. The CS was a light (350 ms in duration) and the US a mild electric shock (0.15 ms in duration). (b) Illustrative examples of URs and CRs in paired CS–US trials. The grey box and the dotted line indicate CS duration and US onset, respectively. (c) Telencephalon and cerebellum lesions. Lateral view of the brain of a control, a telencephalon-ablated and a cerebellum-ablated animals. The arrows mark the boundaries of tissue removal. CCb, corpus cerebelli; HL, hypothalamic lobes; Tel, telencephalon; OT, optic tectum, VL, vagal lobes. (d) Percentage of CRs from the control (Sh), cerebellum-ablated (Cb), telencephalon-ablated (Tel) and pseudoconditioned (Pseudo) groups during the different phases of training. Note that cerebellum lesions produced a severe impairment on delay conditioning, and that telencephalon ablation did not. (Modified from G´omez et al. 2010.)

The performance of the lesioned groups shows that, like in mammals, cerebellar lesions in goldfish severely impair the acquisition of a conditioned response in a delay eyeblink-like procedure, whereas telencephalon ablation does not prevent learning this motor response (Fig. 15.3c–d). Cerebellum-ablated goldfish were not able to learn the association between the CS and the US, as no increase in the level of CRs was observed following conditioning. Furthermore, the performance of this lesioned group did not differ from that of the unoperated animals trained in unpaired CS–US presentations. In addition, the deficit observed in the cerebellum-lesioned group was selective to the CRs; in fact, no differences were observed in the performance of the UR or in the percentages of spontaneous and alpha responses relative to the control and telencephalic animals. In contrast to cerebellum ablation, telencephalon lesion did not impair the acquisition of the conditioned eyeblink-like response, although it did increase resistance to extinction (Fig. 15.3d). In mammals, it has been demonstrated that the participation of forebrain structures is not required for delay eyeblink conditioning to occur, as no deficit in acquisition has been found in decerebrated animals (Norman et al. 1977; Mauk & Thompson 1987; Kotani et al. 2002). Interestingly, several studies have shown slower extinction of eyeblink conditioning in the delay paradigm also in well-trained rabbits after hippocampus removal (Powell & Buchanan 1980; Akase et al. 1989), thus revealing a functional similarity in the extinction deficits observed

BLBK374-15

BLBK374-Brown

330

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

following hippocampal lesion in mammals and telencephalic ablation in teleost fishes. Thus, the performance of telencephalon-ablated goldfish suggests that in teleost fishes, like in mammals, the ability to make the association between CS and US in a delay paradigm does not depend upon forebrain structures; however, some telencephalic structures, in particular the hippocampal pallium or homologue areas, could be relevant for other learning processes such as the extinction of the conditioned response or the learning in a trace paradigm (see the following text). The involvement of the teleost cerebellum in the eyeblink-like classical conditioning has been confirmed in an experiment in which possible learning-related changes in the metabolic activity of the cerebellum of goldfish were studied by means of cytochrome oxidase (CO) histochemistry (Rodr´ıguez et al. 2005). The CO activity in neurons reflects sustained energy demands; therefore, differences in CO activity of particular brain regions reveal changes in their functioning levels (Wong Riley 1989; Gonz´alez-Lima & Cada 1998). In this work, one group of goldfish was trained in the delay eyeblink-like procedure described in the preceding text with paired presentations of the CS and US (paired group), and another received explicitly unpaired CS and US presentations (unpaired group). Following a 4hour training session, the animals were perfused immediately and their brains processed for CO histochemistry. Optical densitometry analysis showed an increase in the level of CO activity in the molecular and granular layers of the cerebellum, which was selective to the goldfish trained in the paired condition. That is, the CO activity did not increase in the cerebellum of the fish in the unpaired condition. As the only difference between both groups was restricted to the mode in which the CS and the US were presented, the remarkable metabolic increment observed in the cerebellum of the animals in the paired group could not be caused by unspecific sensory or emotional factors. These data reveal specific learning-related changes in the cerebellum of goldfish. In fact, in the paired animals, a positive correlation was found between the relative CO staining intensity of both cerebellar layers and their percentage of conditioned responses. The similarity in the basic cerebellar circuitry organisation (Meek & Nieuwenhuys 1998; Butler & Hodos 2005; Meek et al. 2008) and physiology (Kotchabhakdi 1976; Han & Bell 2003; Bell et al. 2008) in mammals and teleosts suggests that some learning functions of the cerebellum appeared early in vertebrate evolution, being conserved through the phylogenetic history of the extant vertebrates (Rodr´ıguez et al. 2005). In addition, recent evidence from experimental and neuropsychological studies indicates that the cerebellum, traditionally associated with motor control, is implicated in a variety of cognitive and emotional functions in humans and other mammals (Petrosini et al. 1998; Sacchetti et al. 2004, 2005; Thompson & Steinmetz 2009). Similarly, the cerebellum of teleost fishes is not only involved in the classical conditioning of motor responses, but also, as we will see later in this chapter, involved in emotional conditioning and in more complex, higher order processes such as spatial cognition.

15.2.2 Role of the teleost cerebellum and telencephalic pallium in trace motor classical conditioning In mammals, the cerebellum and related brainstem circuits mediate eyeblink classical conditioning, as brain structures above the level of the midbrain are not required for conditioning

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

331

this simple motor response (Norman et al. 1977; Mauk & Thompson 1987; Kotani et al. 2002). However, this is valid for delay conditioning, but not for trace conditioning. Trace conditioning imposes additional task requirements, as in this procedure the CS and the US do not overlap; instead, the end of the CS is separated from the onset of the US by a stimulus-free time gap (trace interval). In these conditions, some telencephalic structures, especially the hippocampus, become engaged in mammals (Moyer et al. 1990; WoodruffPak & Disterhoft 2008). Like in mammals, the acquisition and maintenance of eyeblink-like classical conditioning in goldfish are critically dependent on the cerebellum, irrespective of whether a delay or a trace procedure is used (G´omez et al. 2004). The cerebellum and associated brainstem circuitry seem to be sufficient for eyeblink-like conditioning in the delay paradigm, as the complete ablation of the telencephalon spares this form of learning in goldfish (G´omez et al. 2004). However, lateral pallium (LP) lesions in goldfish, like hippocampus lesions in mammals, selectively impair eyeblink-like conditioning when a trace interval is introduced between the end of the CS and the onset of the US without producing any significant deficit in delay conditioning (G´omez et al. 2004). Interestingly, the LP of teleost fishes has been proposed as the homologue of the medial cortex or hippocampus of amniotes on the basis of developmental, neuroanatomical and behavioural data (Rodr´ıguez et al. 2002; Wulliman & Mueller 2004; Northcutt 2006, 2008; Yamamoto et al. 2007; see Fig. 15.2). Thus, the findings reviewed in this section reveal that the brain networks involved in the classical conditioning of discrete motor responses are very similar in teleosts and mammals, i.e. the cerebellum plays an essential role in classical conditioning independently of the temporal requirements of the procedure, whereas the hippocampal pallium is specifically involved in trace conditioning but not in delay conditioning.

15.3

Emotional learning

It has long been known that the telencephalon of teleost fishes is involved in emotional and reproductive behaviour (Segaar & Nieuwenhuys 1963; Overmier & Gross 1974; Shapiro et al. 1974; de Bruin 1980). In particular, the medial pallium is a telencephalic structure that seems to play an important role in those aspects of behaviour in which the motivational and emotional factors must be taken into account. The medial telencephalic pallium of the actinopterygian fishes is considered homologous to the pallial amygdala of amphibians and land vertebrates on the basis of developmental evidence and similarities in the pattern of gene expression, neurochemical distribution and neuroanatomical comparative evidence as well as behavioural data (Portavella et al. 2004; Wulliman & Mueller 2004; Northcutt 2006, 2008; Braford 2009; Mueller & Wullimann 2009; Desjardins & Fernald 2010). The pallial amygdala in mammals is an essential component of the neural circuits responsible for emotional learning and memory (LeDoux 2000; Maren 2001; McGaugh 2004). Also, lesions to the teleost medial pallium disrupt or disorganise aggressive, sexual and parental behaviour (Segaar & Nieuwenhuys 1963; de Bruin 1983). In addition, electrical stimulation in the medial pallium in free-swimming fishes evokes arousal, defensive behaviour and escape responses (Savage 1971; Quick & Laming 1988). The cerebellum has also been related to emotional learning. In fact, recent studies show that the cerebellum

BLBK374-15

BLBK374-Brown

332

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

of mammals, besides its well-known role in motor control, also participates in emotional conditioning (for a review, see Sacchetti et al. 2009). Interestingly, the corpus cerebelli of teleost fish, considered homologous to the cerebellar vermis of land vertebrates (Wullimann & Northcutt 1988; Meek 1992; Meek & Nieuwenhuys 1998), is likewise involved in fear conditioning. In the following sections, we summarise recent data showing that the medial pallium and the corpus cerebellum of teleost fishes are involved in emotional learning and memory.

15.3.1 Role of the medial pallium in avoidance conditioning and taste aversion learning The initial evidences concerning the neural basis of emotional conditioning in teleost fishes showed that complete telencephalic ablation produces devastating effects on the acquisition and maintenance of conditioned avoidance (Overmier & Papini 1986; Overmier & Hollis 1990), suggesting that the teleost telencephalon is involved in the use of emotional states as conditioned reinforcers to produce instrumental responses (Mowrer 1960; Flood et al. 1976). In the avoidance conditioning paradigm, the animals learn to prevent the presentation of an unpleasant unconditioned stimulus (US; usually a mild electric shock) by producing a particular conditioned response (CR; such as jumping to a safe area) at the presentation of a conditioned stimulus (CS; usually a light), which signals the presentation of the US. There is good evidence that avoidance learning is based on the acquisition of a mediational state of fear in goldfish. For example, preceding Pavlovian pairings of the warning stimulus with a shock facilitate subsequent avoidance conditioning (Gallon 1972; Overmier & Starkman 1974), suggesting that the Pavlovian contingency leads to the acquisition of an internal state of fear (conditioned fear). This fear would contribute to the development of the instrumental stimulus-response association in as much as the avoidance response reduces the state of fear (Flood et al. 1976; Overmier & Hollis 1990; Zhuikov et al. 1994). A recent series of lesion studies addressed the possibility that the medial pallium of teleost fishes, proposed as homologous to the pallial amygdala of mammals, may play an important role in emotional learning and memory. And, in fact, these experiments showed that medial but not lateral pallium lesions impaired acquisition and retention of conditioned avoidance in goldfish (Portavella et al. 2003, 2004). Moreover, the deficit in the retention of conditioned avoidance consequent to medial pallium lesions is as severe as that produced by the ablation of the whole telencephalon (Fig. 15.4a). The impairment observed in medial pallium lesioned goldfish is similar to that found in mammals with amygdalar lesions (Aggleton 1992; Davis et al. 1992). It is important to note that the effects of medial pallium damage are selective to emotional memory, as lesioned animals are impaired in their ability to produce avoidance responses, but are still able to produce escape responses. These experiments also showed that the lateral pallium of goldfish, proposed as homologous to the hippocampal formation of tetrapods on the basis of anatomical and developmental evidence, plays an important role in trace avoidance conditioning (Portavella et al. 2004). When a temporal gap separating the end of the CS from the onset of the US (trace interval) is introduced in the two-way active avoidance procedure, lateral pallium lesions severely impair goldfish performance (Fig. 15.4b). In mammals, lesions of the hippocampal formation produce deficits in avoidance conditioning when

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

100

CS (light)

15 s 5s

100 80

60

60

40

40 Sh MP LP

Surgery

80

20 0

15 s

US (shock)

20

Char Count=

333

Trace (5 s) 5s

Surgery

CS (light) US (shock)

% Avoidance

BLBK374-15

0 Pre Post 10-trials blocks (a)

Pre Post 10-trials blocks (b)

Fig. 15.4 Neural bases of avoidance learning in goldfish. This experiment analysed the effects of medial (MP) and lateral (LP) telencephalic pallium lesions on the retention of an avoidance response previously acquired in two different conditioning situations, one with stimuli overlapping (a) and the other with an interstimuli gap (b). A two-way active avoidance paradigm was used in a shuttle box adapted for goldfish conditioning (see cartoon representation). In the delay experiment (a) the discriminative stimulus (light) was turned on for a maximum of 15 seconds in the compartment where the fish was located in that moment. If the fish did not swim across the barrier within 10 seconds of light onset, a mild electric shock was turned on for a maximum of 5 seconds. A response within the first 10 seconds (avoidance response) prevented the presentation of the shock and terminated the warning stimulus (light). A response during the 10–15-second period cancelled both the warning stimulus and the shock. In the trace experiment (b), the light was turned on for a maximum of 10 seconds in the compartment where the fish was located, followed by a gap period of 5 seconds after termination of the light. Thus, the temporal separation between warning onset and shock onset was 15 seconds. If the fish did not swim across the barrier within 15 seconds, the mild electric shock was turned on for a maximum of 5 seconds. An avoidance response during the first 15 seconds terminated the warning stimulus (light), and the shock was not delivered. A response during the 15–20-second period cancelled both the warning stimulus and the shock. Both figures show the mean percentages of avoidance responses in the last three acquisition sessions and the six post-surgery retention sessions. Note that MP lesions produced a severe deficit in the retention of conditioned avoidance in both, the non-trace (a) and the trace (b) procedures, whereas LP lesions impaired retention only in the trace procedure. (Modified from Portavella et al. 2004; Broglio et al. 2005.)

contextual or temporal cues are significant for the conditioning situation (Woodruff & Kantor 1983; Moyer et al. 1990; Phillips & LeDoux 1992; LeDoux 2000; Maren 2001; McGaugh 2004). The deficits found in lateral pallium-lesioned goldfish in trace avoidance conditioning are similar to those observed in the trace motor classical conditioning experiments (see Subsection 15.2.2) and add further support to the evidence showing that the lateral pallium of actinopterygian fishes, like the hippocampus of mammals, is involved in trace memories. Another learning process that has been recently related to the dorsomedial pallium in teleosts is taste aversion learning (Mart´ın et al. 2011). Goldfish learn to avoid the ingestion of a flavour paired with visceral discomfort, when trained in a delayed procedure which consists of the presentation of two flavours on different days, one followed by lithium chloride and the other by saline, both after a 10-minute delay. Dorsomedial pallium lesions impair the acquisition of taste aversion, whereas damage to the dorsolateral pallium, the most likely homologue of the hippocampus, does not produce significant changes in this learning. The deficit caused by the dorsomedial pallium lesions in goldfish, as severe as

BLBK374-15

BLBK374-Brown

334

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

that produced by the complete ablation of the telencephalon, is similar to that observed by damage to the amygdala in mammals (Yamamoto et al. 1994; Bernstein 1999; Lamprecht & Dudai 2000; Berm´udez-Rattoni 2004). Recent neuroanatomical data showing that gustatory and general visceral inputs converge in the dorsomedial pallium add support to the evidence on its critical role in taste aversion learning (Folgueira et al. 2003, 2004; Northcutt 2006; Yoshimoto & Yamamoto 2010), suggesting that this region, like the amygdala of mammals, could be a site for the taste-malaise integration necessary for the formation of taste aversion memory in teleosts. The data summarised here show that the medial pallium plays a critical role in avoidance conditioning and taste aversion learning and provide additional evidence of its homology with the amygdala of mammals. The results showing that the medial pallium of goldfish is an essential component of an emotional memory system and that the lateral pallium is necessary for trace conditioning indicate that in teleost fishes, as in land vertebrates, particular telencephalic areas are involved in different learning functions as components of separate specialised memory systems.

15.3.2

Teleost cerebellum and fear conditioning

An increasing number of studies suggest that the cerebellum of mammals participates not only in the conditioning of simple motor responses but also in emotional learning, for instance in the formation of fear memory (for a review, see Sacchetti et al. 2009). Fear conditioning involves learning that a previously neutral stimulus (CS) predicts an aversive unconditioned stimulus (US). In mammals, lesions in the cerebellar vermis affect fear memory without altering baseline autonomic responses to the frightening stimuli (Supple & Leaton 1990a, 1990b; Supple & Kapp 1993; Gherladucci & Sebastiani 1996; Bob´ee et al. 2000), and reversible inactivation of the vermis during the consolidation period impairs retention of fear memory (Nader et al. 2000; Sacchetti et al. 2002, 2007). Interestingly, recent data show that the corpus cerebelli of goldfish, proposed as homologous to the vermis of mammals (Wullimann & Northcutt 1988; Meek 1992; Meek & Nieuwenhuys 1998), is also involved in emotional learning. The role of the teleost fish cerebellum in emotional learning has been postulated assessing the effects of cerebellar lesions and temporary inactivation on delay (Yoshida et al. 2004; Rodr´ıguez et al. 2005; Martin et al. 2009) and trace fear heart rate conditioning in goldfish (Martin et al. 2009). In the control goldfish, paired presentations of a CS (light) and a US (shock) consistently produced a rapid increase in the percentage of conditioned bradycardia responses (a deceleration of the heart rate during the interval between the onset of the CS and the US relative to pre-CS baseline), which decreased quickly during extinction training. In contrast, goldfish with corpus cerebelli lesions failed to acquire the conditioned bradycardia response, independently of the procedure employed (Fig. 15.5a). It is important to note that no deficit was observed either in the reflex response to the US or in the autonomic orientation response to the CS in ablated animals, indicating that the sensorial and motor neural circuits underlying the expression of the unconditioned cardiac responses were spared in cerebellum-ablated goldfish (Yoshida et al. 2004; Martin et al. 2009). Thus, the effects of corpus cerebelli lesions on the cardiac activity of goldfish seem to be selective to the conditioned bradycardia response. Similarly, vermis lesions impair the acquisition

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Brain and Cognition in Teleost Fish

(a) 30

Bradycardia (%)

335

Sh

Cb Pseudo Sh

Cb

20

1s

CS US

CCb LC

10

VCb Tel

VL

Habituation Acquisition Extinction CS alone CS alone CS – US

(b) Habituation trial

CS 4 2

US Acquisition trial

CS

2 mv

500 ms

Standard scores

BLBK374-15

0 –2 –4 –6 –8

–10 0

1000

Habituation CS alone

0

1000 0

Acquisition Paired CS – US

1000 (ms) Extinction CS alone

Fig. 15.5 Involvement of the goldfish cerebellum in emotional learning. (a) Effects of goldfish cerebellum ablation on delay heart rate conditioning. The paired presentations of the conditioned (CS, light, 1000 ms) and unconditioned stimuli (US, shock, 0.15 ms) consistently produce a conditioned bradycardia in sham-operated fish (Sh). The histogram shows the percentage of deceleration of the heart rate during the CS–US interval relative to pre-CS baseline during habituation, acquisition and extinction. Goldfish with cerebellum lesions (Cb) failed to acquire the conditioned bradycardia response showing a performance level similar to the animals in the unpaired CS–US condition (pseudoconditioned). Right: Electrocardiograms recorded during a paired CS–US trial of representative cerebellum lesioned (Cb) and sham (Sh) animals at the end of acquisition training. Note that the Cb fish do not exhibit the normal bradycardia response to the CS. The photograph shows a saggital section of a lesioned animal. Note that the ablation was restricted to corpus cerebelli (arrows). CCb, corpus cerebelli; LC, lobus caudalis; Tel, telencephalon; VL, vagal lobes. (b) Learning-related changes in Purkinje cells activity during heart rate classical conditioning. Left: Illustrative recording of the activity of a Purkinje cell during habituation and acquisition trials. Note the change in the cell response to the CS following paired presentations. Right: Standard scores of a representative Purkinje cell during the three phases of the experiment. Each bar represents a 200-ms sub-period of the CS duration. (Modified from Rodr´ıguez et al. 2005 and Martin 2009.)

of the conditioned bradycardia response in rats and rabbits, without altering the heart rate baseline or the orientation response to the CS (Supple & Kapp 1993; Gherladucci & Sebastiani 1996). Also, in humans, neuropsychological data show that conditioned bradycardia is impaired in patients with medial cerebellar lesions (Maschke et al. 2002) and that the vermis is necessary to learn a new association between sensory and aversive stimuli, while it is not required for the regulation of baseline fear responses (Turner et al. 2007). Additional evidence on the involvement of goldfish cerebellum in fear heart rate conditioning has been obtained recently in an extracellular recording experiment (Martin 2009). The activity of single Purkinje cells was tracked in goldfish during heart rate conditioning

BLBK374-15

BLBK374-Brown

336

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

and extinction. The results showed changes in the simple spike firing of the Purkinje cells after CS presentations that match the features of the bradycardia conditioned response (Fig. 15.5b). The changes were directly related to the paired presentation of CS and US, as they disappeared with unpaired presentations of the stimuli (extinction). Learning-related changes in Purkinje cells’ activity during classical conditioning have been observed in mammals trained in fear and eyeblink learning procedures (Supple et al 1993; Sacchetti et al. 2004; Jirenhed et al. 2007; Zhu et al. 2007; Scelfo et al. 2008). Thus, the results from the cerebellum-lesioned animals and the Purkinje cells’ activity during heart rate conditioning suggest that the cerebellum of teleost fishes, like the cerebellum of tetrapods, plays an essential role in emotional learning.

15.4

Spatial cognition

In land vertebrates, spatial cognition comprises a variety of perceptive and cognitive mechanisms, subserved by distinct brain networks. In teleost fishes, the initial research on the neural basis of spatial learning and memory, focussed almost exclusively on the role of the telencephalon, produced contradictory results. Thus, telencephalon ablation was reported to produce no deficits, impairments or even improvement in acquisition and reversal of spatial learning (for review, see Flood et al. 1976). The inconsistency of these results could be attributed to the fact that the tasks used in these pioneering ablation studies were not specifically aimed at analysing spatial cognition, and consequently lacked a precise definition of the spatial requirements of the tasks and specific control tests, essential for characterising behavioural impairments. Moreover, widely used terms such as ‘spatial learning and memory’ and ‘spatial cognition’ involve indeed a variety of perceptive and cognitive processes, actually dependent on separate neural substrata, which play different roles in spatial orientation and navigation. Thus, a variety of brain mechanisms are required for processing, encoding and integrating the sensory-motor information, and for translating it into a series of body-centred coordinate systems, and finally to an allocentric, world-centred coordinate system. For example, in mammals the perception and action based on egocentric frames of spatial reference depend on brain circuits that extend from the superior colliculus and the cerebellum to the parietal, somatosensory cortex and the frontal motor cortical areas (Stein & Meredith 1993; Burgess et al. 1999). Instead, the use of allocentric frames of spatial reference for navigation depends on other neural systems, mainly the hippocampal formation (O’Keefe & Nadel 1978; Nadel 1991; Bingman 1992; Burgess et al. 1999; Salas et al. 2003). The allocentric frames of reference (cognitive maps), based on the encoding of the reciprocal spatial relationships between the goal and multiple sensory features, endow spatial behaviour with remarkable flexibility (O’Keefe & Nadel 1978; Burgess et al. 1999). A cognitive map is defined as a map-like, ‘world-centred’ representation of the objective space that provides a stable framework, allowing the subject to reach the goal independently of its own actual position and local view, and irrespective of the loss of some conspicuous environmental information. Further evidence of behavioural flexibility is that the animals can learn rapidly new goal locations (reversal learning), indicating that prior experiences do not conflict with the demands of a new task. These representations, which are true relational

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

337

memories and can be considered the clearest animal equivalent of human declarative or episodic memory (Clayton & Dickinson 1998; Eichenbaum 2000), have been thought as an exclusive attribute of vertebrate groups that supposedly have evolved more complex associational brain structures (i.e. mammals and birds). An increasing amount of evidence from naturalistic and experimental studies indicates that fishes, like amniotes, are able to orient and navigate using a variety of spatial strategies based on distinct, parallel spatial learning and memory systems, including cognitive mapping (Odling-Smee et al. 2006; Salas et al. 2008; Guttridge et al. 2009; for further discussion, see Chapter 8 of the present book).

15.4.1

Allocentric spatial memory representations in teleost fishes

Spatial learning and memory capabilities of teleost fishes have been thoroughly examined in controlled laboratory experiments, designed to provide optimal conditions to reveal the strategies, the cues and the mechanisms used by these vertebrates for orientation and navigation. Moreover, standard tasks closely matching those used to analyse spatial cognition in mammals and birds, analogous laboratory settings and behavioural procedures, and thorough probe and transfer trials, have been used in order to increment the comparative value of the data. The first clear-cut data concerning the capacity of teleost fishes to use cognitive mapping, in addition to egocentric orientation strategies, were provided by Rodr´ıguez et al. (1994). In this study, goldfish were trained to locate a baited feeder in a four-arm maze surrounded by an array of distal visual cues using three different procedures: (1) A particular turn response (egocentric strategy), (2) the information provided by the distal visual cues (allocentric strategy), and (3) of both sources of information, i.e., both types of strategies, simultaneously. Interestingly, although the fishes in every group achieved similar levels of performance, the transfer and probe tests revealed that they used different strategies (Fig. 15.6). Goldfish trained in the allocentric procedure navigated directly to the rewarded place from previously unvisited start locations, adopting spontaneously the most direct routes to the goal although the new paths involved navigating in different or even opposite directions (Rodr´ıguez et al. 1994). The use of appropriate trajectories without a history of previous training, even when these imply new (never experienced before) egocentric relations to landmarks and local views, provides distinct evidence for the capacity of these animals to represent the environment independently of a body-centred reference system (Fig. 15.6c). In addition, the goldfish could use orientation (egocentric) strategies, as indicated by their ability to reach the goal by using a fixed body turn (i.e. turn right or left) disregarding environmental information, or could implement simultaneously both body-centred and allocentric strategies and use one or the other according to experimental conditions (Fig. 15.6c). Recently, Schluessel & Bleckmann (2005) obtained experimental results that suggest that also elasmobranchs, which are characterised, as land vertebrates, by the evaginated pattern of telencephalon morphology, can use allocentric strategies for navigation. Like goldfish, rays (Potamotrygon motoro) are able to reach the goal using novel routes starting from unfamiliar locations. Another experiment, in which goldfish were trained in a mixed place–cue procedure in a plus-maze similar to that used by Rodr´ıguez et al. (1994), also provided evidence on the

BLBK374-Brown

338

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(a)

(c) Place

Bookcase

42.5

45.3

7.5

4.7

85.7 9.7

Dustbin Chair

*

44.3

3.7

13.9

72.5

* 9.8

76.5

14.2

14.8

11.7

*

*

74.4

* Place–Turn

Sink

46.6

*

* 96.1 3.9

45.7

*

4.6 2.8

63.1

32.3

*

51.4

80

66.8 31.6

1.3

*

*

*

60

30.9 Place-turn Turn Place Control

15 30 45 3-session blocks

Control

20

73.3

*

8.5

11.9

14.7

12.3

5.3

0.5 m

(b) 100

40

70.7

13.9

* Turn

Table

13.4

*

Maze

Aquarium Door

**

N

Poster

PC

Door

Percent correct

BLBK374-15

48.6

36.3

53.4

30.4

33.3

35.7

34.3 28.7 36.9

Type 1 test

Training

33.4

Type 2 test

Transfer tests: New start postitions

Fig. 15.6 Allocentric and egocentric strategies used by goldfish to find a goal in a plus maze. (a) Experimental room and maze. (b) Mean percentage of correct choices of goldfish trained in four different experimental conditions. In the place group, two different start positions were used at random but the goal was situated always in the same place in the room. Thus, animals in this group could only use the array of extra-maze cues to solve the task. In the turn group, two different start positions were used at random, but in every trial the goal arm was determined by a fixed turn response (e.g. always left). In this task extra-maze cues were irrelevant for task solution. The fishes in the place–turn group started always from the same place of the room and the goal arm was always in another constant place. Thus, this task allowed the selection of the correct arm on the basis of a specific turn direction and/or extra-maze cues. Finally, in the control group the location of the goal varied randomly across trials. (c) Transfer tests run to elucidate whether the animals solved their respective tasks on the basis of turn (egocentric) or place (allocentric) strategies. In type 1 transfer tests, the maze remained in its usual position but the animals were released from a novel start position. In type 2 tests, the maze was displaced in the room (see dashed lines in (a)) such that the end of one arm reached the place of the room where the goal was located during training trials, but the start positions were different to those used during training. The figure shows the trajectories chosen by the animals during training and transfer trials. The numbers and the relative thickness of the arrows denote the percentage of times that a particular choice was made. The dashed lines denote the original position of the maze before it was displaced for type 2 tests and the asterisks mark the goal location. Note that the animals in the place group consistently chose the arm with the distal extreme at the same place of the room where they were rewarded during training trials. In contrast, the most chosen arm by the fish in the turn group was the one coinciding with the learned turn, independently of the location of the start arm. (Modified from Rodr´ıguez et al. 1994.)

simultaneous and cooperative use of multiple spatial learning strategies in fishes (L´opez et al. 2000a). In this study, the goal was located in a constant room location and was, in addition, signalled by a distinct intra-maze visual cue, such that the fishes could simultaneously implement place learning and egocentric strategies to solve the task. The probe trials showed that when either the distal cues, or the intra-maze, proximal cues, were removed, the fishes relied on the available environmental information. When the intra-maze visual cue was eliminated, the goldfish were still able to locate the goal on the basis of the information provided by the array of extra-maze visual cues (place strategy). Conversely, when

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

339

the use of the distal visual cues was precluded by a curtain that surrounded the maze, the animals readily reached the goal approaching the intra-maze cue (orientation strategy). The ability of goldfish to implement relational or map-like spatial strategies is not linked to the use of distal visual cues. For example, goldfish trained in a spatial constancy task to locate a goal in a small enclosure where only proximal visual cues were available (Salas et al. 1996b) navigated accurately to the goal from different start locations, regardless of route direction and response requirements. A distinctive piece of evidence concerning cognitive mapping in goldfish is provided by their efficiency to locate a goal in the absence of some familiar conspicuous environmental cues (Rodr´ıguez et al. 1994; L´opez et al. 1999; Dur´an et al. 2008). As cognitive maps store redundant environmental information, when a subset of spatial cues becomes unavailable, accurate navigation is still possible on the basis of those that remain (O’Keefe & Nadel 1978; Thinus-Blanc 1996). The study by Rodr´ıguez et al. (1994) provided interesting evidence in this regard: Although the performance of the fishes trained in the allocentric task became as poor as that of the control fishes when all the cues were simultaneously excluded, indicating that they used the information provided by those cues to solve the task, it did not deteriorate when the most salient ones were individually removed or hidden. Similar results were provided by a study in which goldfish were trained in a spatial procedure analogous to the hole-board task commonly used with rodents (Dur´an et al. 2008). In this procedure, the fishes had to locate the only baited feeder (goal) in a 5 × 5 feeder-matrix surrounded by an array of intra-maze visual cues, which maintained stable spatial relationships with the goal. The probe tests showed that the fishes relied on the array of visual cues to solve the task, as indicated by the deficit observed when these were eliminated in a whole, or disarranged (Fig. 15.7). Interestingly, their performance was not impaired when any one of the individual cues or subsets of cues were eliminated. Convergent results were obtained when goldfish were trained in the spatial constancy task or in a cued version of the same procedure (L´opez et al. 1999). Whereas the performance of the fishes in the cue task was dramatically impaired when the cue associated directly with the goal was removed, goldfish in the spatial constancy task navigated accurately to the goal despite the deletion of any one of them (Fig. 15.8). All of these results indicate that the performance of goldfish trained in allocentric, relational spatial tasks is based on the knowledge of the relationships among the goal location and many environmental cues, such that when some are missing, the remaining ones are sufficient to locate the goal. Moreover, fishes can simultaneously encode the spatial relationships among landmarks and the shape (geometry) of the environmental boundaries (Broglio et al. 2000; Sovrano et al. 2002, 2003, 2007; Vargas et al. 2004). The first evidence was provided by L´opez et al. (1999), who observed that the performance of goldfish trained in the spatial constancy task was impaired when a modification was introduced in the experimental apparatus that altered its shape and global topography, but left unchanged the local views of the areas corresponding to each of the doors (Fig. 15.8). Sovrano et al. (2003) show that redtail splitfin (Xenotoca eiseni) are able to reorient relying on the shape of the environment, and to combine geometric with non-geometric information of the environment such as the colour of the walls or the features provided by the visual cues. In addition, Vargas et al. (2004) show that goldfish locate a place in an environment by encoding the goal location with respect to the geometrical features of the experimental space, even in absence of

BLBK374-15

BLBK374-Brown

340

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(a)

(c)

Probe tests c

b

e a

c

a

d

5% 10% Cues 15% disorganisation 20% b

e

d

Food removal

Cues removal

b

b c

d

e

e

Cue a removal

Training

S2

S1 a

S1 S3

e

b c

a

d

S4

e

Cue d removal

Cue c removal

S2 b

b c

25% c

a

a

d

(b)

b

c

a

d e

b c

a

d e

b c

a

d

d e

S4

e

S3

Fig. 15.7 Goldfish use map-like strategies to locate a goal in a hole-board analogue task. (a) View of the experimental apparatus with the 25-feeder matrix and the arrangement of the landmarks. (b) Schematic representation of the apparatus and procedure used during training. The grey filled circle indicates the goal (baited feeder). The positions of the five cues are marked by a, b, c, d and e. S1–S4 indicate the four different starting positions used during training. The four diagrams on the right show two representative searching trajectories of a well-trained animal departing from each start position. (c) Schematic representations of the pattern of choice distribution in the probe tests used in this experiment. Each diagram shows the maze and the location of the cues relative to the goal. The diameter of each point denotes the percentage of choices. In the food removal probe test no experimental change was introduced, except replacing the reward by a fake food stick. In the cues removal and cues disorganisation probe tests, all the cues were removed or disarranged, respectively. In the single cue removal probe tests, only one cue was removed at a time. (Modified from Dur´an et al. 2008.)

objects (Fig. 15.9). These results reveal that fishes, like mammals and birds, are able to integrate spatial information of different nature and from various sources, for allocentric navigation (Cheng & Gallistel 1984; Cheng 1986; Gallistel 1990; Cheng & Newcombe 2005; Chiesa et al. 2006).

15.4.2 Role of the teleost telencephalon in egocentric and allocentric spatial navigation Fishes trained in a variety of spatial procedures display behavioural abilities probably based on multiple learning and memory systems, and their similarity to those described in mammals and birds outlines the central issue of whether these cognitive capabilities are supported by neural centres and circuits equivalent to those that underlie spatial cognition in land vertebrates. Lesion studies aimed at examining this question have provided strong evidence of the importance of the telencephalon for allocentric spatial cognition in teleost fishes (Broglio et al. 2003). Thus, telencephalon ablation dramatically and irreversibly impaired the ability of the fishes trained in a place procedure to reach goal location (Salas et al. 1996a; Fig. 15.10). Conversely, telencephalon ablation did not alter the performance of the animals using egocentric (body-centred) orientation strategies. Moreover, although

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Brain and Cognition in Teleost Fish

Spatial constancy Reversal

Glass barrier

50%

Exit

Exit

50%

100 50%

Glass barrier

50%

Percent Corrrect

Training

50%

50%

Exit

60

Sessions

Training

Glass barrier

Reversal Exit

50%

Glass barrier

50%

(b)

100

*

Transfer test 1

* *

*

Spatial constancy Cue

80

40

Cue

341

Reversal

(a)

Percent Corrrect

BLBK374-15

Spatial constancy Cue

80 60 40

Transfer test 1

Transfer test 2

Transfer test 2 Fig. 15.8 Goldfish can use multiple spatial learning strategies to locate a goal in a small stimulus-controlled enclosure. (a) Animals were trained in a spatial constancy or a cue version of the same task. The schemas show the distribution of the experimental visual cues (black circles and hollow squares), the position of the glass barrier and the location of the exit (goal, asterisk) for both tasks. Following acquisition the fishes were trained in the reversal of the task (schemas on the right). The numbers indicate the percentage of trials initiated from each starting compartment. The arrows show the most efficient routes to the goal. The figure shows the percentage of correct responses during acquisition and reversal. The photograph shows a fish solving the spatial constancy task. (b) The schemas show the different transfer tests conducted to elucidate the strategies employed by the animals in each group. The figure on the right shows the percentage of correct responses during each transfer test. (Modified from L´opez et al. 1999, 2000a.)

the ablation did not disrupt the overt post-surgery performance of the fishes trained in a mixed place–turn procedure in the same experiment, test trials revealed a notable memory deficit in these animals: Whereas before ablation they used either place (allocentric) or turn (egocentric) strategies in a flexible and cooperative manner according to experimental conditions, after surgery they only used turn responses (Salas et al. 1996a).

BLBK374-15

BLBK374-Brown

342

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(a)

Rotational error

(b)

Fig. 15.9 Goldfish can encode geometric and featural information to navigate. (a) In the geometry task, fishes were trained to find the exit door (goal) placed in a corner (A) of a rectangular arena that had three identical but blocked openings in the other three corners (B, C and D). Because of the geometric properties of the apparatus, the correct corner was indistinguishable from the diagonally opposite corner (rotational error). The curves show the percentage of choices for the four corners during training and the diagrams in the right show the percentage of choices (numbers) for each corner during the geometry and the invalidated geometry tests. In the geometry test, when the glass barrier was removed, the animals chose more frequently the geometrically correct doors. In the invalidated geometry test, when the geometric cues were removed by using a new (square) apparatus, no significant differences were observed in the percentage of choices for the four doors. (b) In the geometry + feature task, fishes were trained in the same rectangular box but in which additional feature information was provided by alternate dark grey and white vertical stripes on two walls. The curves show the percentage of choices for the four corners during training. The diagrams on the right show the percentage of choices for each corner during the three different probe tests conducted for this group. In the geometry test, the striped panels were removed and the animals chose more frequently the two geometrically correct doors (A and C). In the feature test, the fish chose the correct door significantly more (A). Finally, in the dissociation test, when the information provided by geometry and features was set in conflict by rotating the striped panels 90◦ , the fish did not show a preference for any particular door. (Modified from Vargas et al. 2004.)

Consistent results were obtained in a subsequent experiment in which intact and telencephalon-ablated goldfish were trained in a mixed place–cue procedure (L´opez et al. 2000a). Interestingly, telencephalon-ablated goldfish showed better performance during training relative to sham animals; nonetheless, they were unable to reach the goal when arriving to it demanded flexible responses (Fig. 15.11). In the place–cue dissociation tests, when the two sources of information were set in conflict (the place and the cue responses were incompatible), the control fishes did not show a significant preference for either the cue or the place responses. In contrast, the performance of the ablated fishes was notably

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

343

Fig. 15.10 Telencephalic ablation in goldfish produces allocentric spatial learning and memory deficits. Goldfish were trained in the same procedures described in Fig. 15.6. Following acquisition, complete ablation of both telencephalic hemispheres was carried out. The curves show the mean percentage of correct choices of telencephalon-ablated and sham-operated goldfish. Note that ablation impaired performance exclusively in the animals using place strategies. Bottom: Trajectories chosen by the animals in the place group during training, and on transfer tests when new start positions were used, before and after surgery. Note that after ablation the fishes in the place task failed to navigate accurately to the goal from familiar locations (training trials) and from novel ones (transfer trials). (Modified from Salas et al. 1996b.)

biased; they showed a significant preference for the arm containing the cue that signalled the goal during training. Moreover, in the intra-maze cue removal test, whereas the control goldfish consistently chose the arm placed at the location of the room where they were rewarded during training, the telencephalon-ablated fishes chose at random between the maze-arms, indicating that they lacked the ability to use the array of extra-maze cues as a source of spatial information. These results reveal that although both groups learned the task, the telencephalon-ablated animals suffered a profound allocentric learning deficit. In fact, the sham and the ablated goldfish differed in their capacity to use different types of navigational strategies; the control animals used both place and cue strategies, but the telencephalon-ablated animals solved the task exclusively on basis of a cue strategy (Fig. 5.11). In addition, telencephalon ablation disrupted the post-surgery performance (Salas et al. 1996b) and reversal learning (L´opez et al. 2000b) of goldfish trained in a spatial constancy task, but did not produce any observable deficit in a cue procedure. As mentioned in Section 15.4, the use of allocentric or relational frames of reference for navigation endows spatial behaviour with remarkable flexibility, evident also when the goal

BLBK374-15

BLBK374-Brown

344

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Fig. 15.11 Telencephalic ablation produces spatial learning deficits in goldfish trained in a mixed place–cue procedure. The fish could solve the task by relying on the information provided simultaneously by the distal landmarks that surrounded the maze, and by one particular intra-maze cue that signalled directly the location of food. The insert shows a schematic representation of the maze, the training procedure, and the position of the two intra-maze cues (dotted and striped panels). The photographs show a normal brain and another with a complete telencephalon ablation. The curves show the percentage of correct responses of each group during training. Note that the telencephalon-ablated goldfish learned the task faster than controls. The figure on the right shows the trajectories chosen by sham and telencephalon-ablated animals in probe tests. In the intra-maze cues reversal test, the two sources of information were set in conflict; thus, place and cue responses were incompatible. In the extra-maze cues occlusion test, the maze was completely surrounded by a curtain. In the intra-maze cues removal test, the fish was released from a new start position. Note that the control fish used the information provided by the intra- and the extra-maze cues, but telencephalon-ablated fishes consistently chose the arm containing the cue that signalled the goal during training, and, when this was excluded, they navigated at random although the distal cues were visible. (Modified from L´opez et al. 2000a.)

location is changed after the mastering of the task (reversal learning). Thus, goldfish trained in allocentric or relational spatial tasks show faster reversal learning than fishes trained in egocentric procedures, and relative to their own acquisition learning (Rodr´ıguez et al. 1994; L´opez et al. 1999, 2000b). Interestingly, telencephalon ablation selectively impaired reversal learning in the animals trained in the former tasks; in both cases, the ablated goldfish needed more trials to learn the new goal location relative to their own initial learning and to the sham group (Salas et al. 1996a, 1996b). Indeed, the reversal performance of the ablated fishes trained in the place and the spatial constancy task did not differ from that of the animals trained in the turn procedure (Salas et al. 1996a) or in a cue version of the spatial constancy task (Salas et al. 1996b), indicating that they had lost the behavioural flexibility that characterises the use of allocentric frames of reference. These reversal learning deficits are similar to those observed in mammals and birds with hippocampal lesions (Hirsch & Segal 1972; Nonneman et al. 1974; Good 1987; Good & MacPhail 1994; Hampton et al. 2004). In summary, the place memory impairments observed in these experiments, showing a selective but severe disruption in spatial cognition after telencephalon ablation, provide significant evidence concerning the presence of a telencephalon-dependent spatial memory system in teleost fishes.

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Brain and Cognition in Teleost Fish

15.4.3

345

Map-like memories and hippocampal pallium in teleost fishes

In mammals, birds and reptiles, the hippocampal formation is critical for encoding the features of the environment in map-like or relational memory representations (O’Keefe & Nadel 1978; Sherry & Duff 1996; Bingman et al. 1998; Burgess et al. 1999; Eichenbaum et al. 1999; Rodr´ıguez et al. 2002; L´opez et al. 2003a, 2003b; Squire et al. 2004). As mentioned in Section 15.1, the telencephalon of ray-finned fishes (e.g. teleosts) presents a medial-to-lateral inversion in the topological position of the main pallial subdivisions. Therefore, the lateral pallium of ray-finned fishes, and in particular the ventral subdivision (Dlv), is the most likely homologue of the hippocampus or medial pallium of land vertebrates, as it occupies the most distal topological position in the pallium homology that is confirmed by its extensive interconnections with the likely homologues of the septal nuclei and preoptic area (Butler & Hodos 2005; Northcutt 2006) by the distribution of histochemical and molecular markers (Kapsimali et al. 2000) and by the pattern of neurogenesis and migration of interneurons (Zupanc et al. 2005; Grandel et al. 2006). Also, data derived from lesion and functional studies agree with these embryological and anatomical peculiarities, showing that the lateral telencephalic pallium of teleost fishes is, like the hippocampus of land vertebrates, involved in spatial cognition. A series of experiments analysing the effects of selective pallial lesions on the performance of goldfish trained in a variety of spatial learning and memory tasks provided clear evidence concerning the role of the teleost lateral pallium (LP) in spatial cognition. LP lesions produced a dramatic memory impairment in goldfish trained in a place task (Rodr´ıguez et al. 2002; Dur´an et al. 2010; Fig. 15.12a). In fact, the place memory deficit observed after lesions selective to the ventral lateral pallium in goldfish is as severe as that produced by the complete ablation of both telencephalic hemispheres (Salas et al. 1996a; L´opez et al. 2000a; Rodr´ıguez et al. 2002). In contrast, medial (MP) or dorsal (DP) pallium lesions did not produce any observable impairment in place memory (Rodr´ıguez et al. 2002; Fig. 15.12a). Furthermore, the involvement of the LP of goldfish in spatial cognition seems to be selective to the allocentric memory system, as damage to this area does not impair the use of cue or other egocentric strategies (Salas et al. 1996a, 1996b; L´opez et al. 2000a; Rodr´ıguez et al. 2002; Dur´an et al. 2010; Fig. 15.12b). Similar results were obtained when goldfish with medial, dorsal or lateral pallium lesions were trained in the spatial constancy task or in the cued version of the same procedure (Broglio et al. 2010). The results showed that lesions in the lateral pallium, but not those in the medial or the dorsal pallium, produced spatial memory and reversal learning impairments (Fig. 15.12c–d). Moreover, these deficits only occurred in the animals trained in the allocentric or relational procedure. Thus, damage in the ventral zone of the lateral pallium, but not in the medial or the dorsal pallium, produced a significant impairment on the capability of goldfish to solve the spatial constancy problem previously learned (Fig. 15.12c). In addition, the ventral lateral pallium-lesioned animals needed a significantly higher number of trials to learn the new goal location following reversal, relative to the fishes with medial or dorsal pallium lesions and relative to their own initial learning rate (Fig. 15.12d). Similar results have been observed following lesions in the hippocampal formation of land vertebrates: Although the reversed discrimination is eventually learned, the number of trials to reach criterion is substantially increased (Hirsch & Segal 1972; Nonneman et al. 1974; Good 1987; Good & MacPhail 1994; L´opez et al. 2003a).

BLBK374-Brown

346

(a) 100

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Goal 50%

50%

Place

80

Vv VI

Sh MP DP LP

40

Pre-surgery

Exit

Dc

(c) Spatial constancy

50%

100

Exit

Dlv Dp

oc

LP

Post-surgery

50%

100

Dlv

Dld Dmv Dc Vd Dlv Vv VI Dc Dld Vs Dlv Nt Dp ac

60

(b)

Dld

Dmv Dc Vd

Surgery

Percent correct

Dmd Dd

50%

DP

Spatial reversal

50%

50%

Exit

Sh MP DP LP

40

Exit

50%

Cue 100 50%

80

60 60

MP

(d)

50%

80

80

Surgery

Percent correct

BLBK374-15

60 Sh MP DP LP

Sh MP DP LP

40

20

40 Acquisition

Retention

Acquisition

Reversal

Post-surgery

Fig. 15.12 Spatial memory deficits after lateral telencephalic pallium lesion in goldfish. In these experiments, the effects of lesions to the lateral (LP), dorsal (DP) or medial (MP) pallium of goldfish were analysed in spatial (a and b) and cue (d) tasks and in the reversal learning of a spatial task (c). The insert on each curve shows a schematic representation of the training procedures. LP lesion produced a dramatic impairment in goldfish trained in the place and the spatial constancy tasks (a and b), whereas MP and DP lesions did not decrease accuracy. LP lesions also produced a profound deficit in the reversal learning of the spatial constancy task (c). In contrast, none of the pallial lesions produced deficits in the cued task (d). The drawings show a schematic representation of the largest (grey shading) and smallest (black shading) extent of the LP, DP and MP lesions in goldfish, reconstructed in coronal sections. (Modified from Rodr´ıguez et al. 2002; Broglio et al. 2010.)

Additional evidence has been obtained from a recent experiment aimed to analyse possible learning-related changes in the transcriptive activity of the pallial neurons in goldfish trained in a spatial task or in a cue version of the same procedure (Broglio et al. 2010). The results showed that training in the spatial constancy task produced an increment in the transcriptive activity in the neurons of the ventral lateral pallium, as indicated by increases in the size of the nucleolar organising region (NOR), i.e. the nucleolar organelles associated with the synthesis of ribosomal proteins (Derenzini 2000). Moreover, these changes were selective to training in the spatial constancy task. In fact, training in the cue version of the same procedure did not produce observable changes, although the tasks used in this experiment were characterised by identical visual complexity and response requirements, and only differed on the type of spatial representations necessary to succeed (Salas et al. 1996b; L´opez et al. 1999). In addition, the enlargement of the NORs was limited to the neurons of the ventral lateral pallium (DLv-d and DLv-v), as the NORs size of the neurons in the medial pallium (DM) did not increase with training either in the spatial or in

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

347

the cue procedures. These data indicate, on one side, that the lateral pallium is involved in allocentric but not in guidance strategies, and, on the other, that the medial pallium does not play an important role, if any, in spatial cognition. Recent studies have also found increases in the relative size of the lateral pallium associated with elaborated spatial behaviour in several species of teleost fishes (Shumway 2008; Costa et al. 2011). These results showing that training in spatial tasks and differential experience in natural environments produce morphological and neurochemical learning-dependent plastic changes in the lateral pallium are similar to those found in the hippocampus of amniotes (Jacobs & Spencer 1994; Hampton et al. 1995; Basil et al. 1996; Healy et al. 1996; Ramirez-Amaya et al. 1999, 2001; Holahan et al. 2006; Roth et al. 2006; Cnotka et al. 2008). In summary, the data presented here clearly show that the ventral subdivision of the lateral telencephalic pallium of teleost fishes, like the medial cortex or hippocampus of amniotes, underlies the ability of fishes to navigate on the basis of allocentric representations of the environment. These results, consistent with developmental and neuroanatomical data, provide additional evidence regarding the homology of the teleost lateral pallium with the hippocampal pallium in vertebrates with evaginated telencephala.

15.4.4

Neural mechanisms for egocentric spatial orientation

As described in the previous sections, hippocampal pallium lesions in teleost fishes do not impair the use of egocentric strategies for spatial orientation, clearly indicating that other, non-telencephalic brain structures and circuits such as the optic tectum and the cerebellum, could be implicated in these processes, as it occurs in mammals. The neuroanatomical and functional organisation of the optic tectum (superior colliculus in mammals) are notably conserved in vertebrates. For example, marked similarities can be observed in the specialised cytoarchitecture and microcircuitry and the profuse connectivity with other motor and sensory centres (Vanegas 1984) as well as in the mechanisms for generating coordinated eye, head and body movements, and for coding the metric and kinetic features of these movements (Du Lac & Knudsen 1990; Salas et al. 1997; Herrero et al. 1998; Sparks 2002; Perrault et al. 2003; Luque et al. 2005). In fact, this structure provides a common body-centred framework for multisensory integration and sensorymotor transformations (Stein & Meredith 1993; Sparks 2002) and is crucial for generating actions within an egocentric frame of spatial reference (Burnett et al. 2004). As in other vertebrates, focal electrical stimulation in the optic tectum elicits coordinated eye and body movements, postural adjustments and other motor patterns in teleost fishes (Demski 1983; Vanegas 1984; Al-Akel et al. 1986; Salas et al. 1997; Herrero et al. 1998; Fig. 15.13a–b). Similarly, there is a topographically ordered motor map in the deep tectal layers of teleosts in correspondence with the retinotopic visual map in the superficial layers, as revealed by the fact that the characteristics of the orienting eye movements depend on the active tectal site (Salas et al. 1997; Herrero et al. 1998; Sparks 2002; Torres et al. 2005). Additional evidence on the role of the teleost fish optic tectum in egocentric orientation has been obtained from lesion experiments. Tectal ablation in goldfish abolishes the orienting response towards the contralateral visual hemifield (Davis & Klinger 1987; Torres et al. 2005), even though they can detect visual stimuli in that hemifield and can turn towards the contralateral

BLBK374-15

BLBK374-Brown

348

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(a)

(b)

u 10° 5°

200 100 150

c

250 300

CCb

350 400 500 Hz 1000 800 600

VCb

OT

0.1 m 1 mm

(c)

(d) b

S1

Optic tectum

Sham

Sham c

a

S2 b c

a

d

e

b

d

e

S1

Cerebellum c

a

S2 b c

a

d

e

d

e

Fig. 15.13 The teleost optic tectum and cerebellum are also implicated in spatial behaviour. The optic tectum is a crucial centre for the generation of egocentrically referenced actions in space. Focal electrical stimulation in the optic tectum of goldfish elicits coordinated eye and body movements, postural adjustments and other motor patterns. Variation of the stimulation site in the rostro-caudal axis produces a systematic change in the amplitude of the horizontal component of the saccade (a), whereas variation of the stimulation site in the medial–lateral axis produces an increase in the vertical component of the eye movements (not shown). Cb, cerebellum; CCb, corpus cerebellum; OT, optic tectum; VCb, valvula cerebellum; u, c, upward and contraversive direction of evoked eye saccades, respectively. (Modified from Salas et al. 1997; Herrero et al. 1998.) (b) The electrical microstimulation of the optic tectum in free-swimming fishes also produces body movements. Evoked movements consisted of complete orientation responses including coordinated movements of the axial musculature, fins and eyes, which closely resemble the natural responses. The direction and amplitude of the orienting responses depend on the tectal stimulation site and on the stimulus parameters (the figure shows the effects of stimulus frequency variation). (c) Tectal ablation disrupts the normal exploratory behaviour in goldfish. Examples of the exploratory patterns of representative sham and tectum-ablated animals during a 4-h session of free exploration in a large hexagonal open field. The figures show the trajectories during the first 5 minutes of each hour throughout the 4-h session. Note that the organised and systematic pattern of exploration area by area is disrupted in tectum-ablated fishes. (d) Goldfish with cerebellum ablation show a severe spatial deficit. Diagrams show some examples of trajectories of representative sham and cerebellum-ablated animals from the start positions (S1–S2) to the baited feeder (black circle); a–e show the location of the visual cues, and the small points the location of the feeders.

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

349

side. Tectal ablation also disrupts the normal exploratory activity that goldfish display when they enter an unfamiliar environment, in order to acquire spatial knowledge. Control goldfish engage in an organised and systematic pattern of exploration when introduced into a novel environment, performing initially ample turns near the edges and across the tank and later exploring thoroughly a particular area before moving to the next one, until exploring the entire tank (Kleerekoper et al. 1974; Rodr´ıguez 1996). Remarkably, tectum ablation produces a profound disorganisation of this exploratory pattern (Rodr´ıguez 1996; Fig. 15.13c). In summary, these data indicate that the tectal mechanisms of teleost fishes are strikingly similar to those of other vertebrates and that the tectum participates in the generation of orienting responses, providing egocentric frames of reference for perception and action. Recent neuropsychological and experimental evidence indicate that the mammalian cerebellum, in addition to being a crucial centre for motor coordination, is also involved in spatial cognition and other cognitive processes (Petrosini et al. 1998; Thompson & Steinmetz 2009). Interestingly, recent experiments show that the teleost cerebellum, similar to that of mammals, participates in spatial cognition, as indicated by lesion studies using a variety of spatial tasks. For example, cerebellum-lesioned goldfish were trained to locate the only baited feeder within a 25-feeder matrix surrounded by an array of visual cues, which maintained stable spatial relationships relative to the goal (Dur´an et al. 2004; see Fig. 15.7). Although the cerebellum-lesioned animals improved slightly during training, their performance revealed a profound spatial cognition deficit: They showed a stereotyped and inefficient search pattern, and never reached the level of accuracy of the control and sham operated animals (Fig. 15.13d). Similar results have been observed in mammals with cerebellar lesions when trained in spatial tasks, such as the Morris water maze or the Tmaze (Petrosini et al. 1998; Rondhi-Reig et al. 2002; Molinari et al. 2008). Moreover, when goldfish with cerebellum lesions, telencephalic lesions or sham operations, were trained in a spatial or a cue-learning task (Dur´an et al. 2004), cerebellar, but not telencephalic lesions, were equally disruptive regardless of training conditions. As described in Subsection 15.4.2, in teleost fishes, telencephalic lesions, and in particular those damaging the hippocampal pallium, impair the performance in the allocentric spatial tasks, but spare cue learning (Salas et al. 1996a; Rodr´ıguez et al. 2002; Broglio et al. 2005). In contrast, the performance of cerebellum-ablated fishes decayed to random levels in both the spatial and the cue tasks, indicating that the teleost cerebellum is involved also in the association of oriented motor responses with single landmarks and in other egocentric mechanisms. It is important to note that although the cerebellum lesions impair the use of both allocentric and egocentric strategies, the deficits are restricted to spatial abilities. In fact, posture and swimming ability, as well as distance travelled, are not affected under these training conditions (Dur´an et al. 2004; Rodr´ıguez et al. 2005; Matsumoto et al. 2007).

15.5

Concluding remarks

We have reviewed recent psychobiological and neurobiological evidence that clearly contradict the traditional theories about brain and cognition evolution in vertebrates, which regarded fishes as situated at the bottom of the so-called ‘phylogenetic scale’ and lacking most of the brain centres and neural circuits that support cognitive capabilities in the

BLBK374-15

BLBK374-Brown

350

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

‘superior’ vertebrate groups, i.e. birds and especially mammals. The results we discussed in this chapter indicate that at least some learning and memory systems (including motor, emotional and spatial learning) of teleost fishes are strikingly similar to those of reptiles, birds and mammals. These data suggest that these complex memory systems, far from being a distinguishing characteristic of the ‘more recent’ and ‘advanced’ vertebrates, occurred early in evolution and were already present in the ancestral fish group that gave way to the extant vertebrates, having being conserved through their phylogenetic history notwithstanding major variations in brain morphology.

Acknowledgements We thank Mr. Gerardo Labrador for technical help. This work was supported by grants BFU2007-62228 F.E.D.E.R. from the Spanish CICYT, and P08-CVI-03934 from Junta de Andaluc´ıa.

References Aggleton, J.P. (1992) The Amygdala: Neurobiological Aspects of Emotion, Memory, and Mental Dysfunction. Wiley-Liss, New York. Akase, E., Alkon, D.L. & Disterhoft, J.F. (1989) Hippocampal lesions impair memory of short-delay conditioned eye blink in rabbits. Behavioral Neuroscience, 103, 935–943. Al-Akel, A.S., Guthrie, D.M. & Banks, J.R. (1986) Motor responses to localized electrical stimulation of the tectum in the freshwater perch (Perca fluviatilis). Neuroscience, 19, 1381–1391. Ari¨ens-Kappers, C.U., Huber, G.C. & Crosby, E.C. (1936) The Comparative Anatomy of the Nervous System of Vertebrates, Including Man. Macmillan Publishing Co., New York. Basil, J.A., Kamil, A.C., Balda, R.P. & Fite, K.V. (1996) Differences in hippocampal volume among food storing corvids. Brain, Behavior and Evolution, 47, 156–164. Bell, C.C., Han, V. & Sawtell, N.B. (2008) Cerebellum-like structures and their implications for cerebellar function. Annual Review of Neuroscience, 31, 1–24. Berm´udez-Rattoni, F. (2004) Molecular mechanisms of taste-recognition memory. Nature Reviews Neuroscience, 5, 209–217. Bernstein, I.L. (1999) Taste aversion learning: a contemporary perspective. Nutrition, Immunology, Neuroscience and Behavior, 15, 229–234. Bingman, V.P. (1992) The importance of comparative studies and ecological validity for understanding hippocampal structure and cognitive function. Hippocampus, 2, 213–220. Bingman, V.P., Riters, L.V., Strasser, R. & Gagliardo, A. (1998) Neuroethology of avian navigation. In: R. Balda, I. Pepperberg & A. Kamil (eds) Animal Cognition in Nature, pp. 201–226. Academic Press, New York. Bitterman, M.E. (1975) The comparative analysis of learning. Science, 188, 699–709. Bob´ee, S., Mariette, E., Tremblay-Leveau, H. & Caston, J. (2000) Effects of early midline cerebellar lesion on cognitive and emotional functions in the rat. Behavioural Brain Research, 112, 107–117. Braford, M.R. (1995) Comparative aspects of forebrain organization in the ray-finned fishes: touchstones or not? Brain, Behavior and Evolution, 46, 259–274. Braford, M.R. (2009) Stalking the everted telencephalon: comparisons of forebrain organization in basal ray-finned fishes and teleosts. Brain, Behavior and Evolution, 74, 56–76. Broglio, C., G´omez, A., Dur´an, E., Oca˜na, F.M., Jim´enez-Moya, F., Rodr´ıguez, F. & Salas, C. (2005) Hallmarks of a common forebrain vertebrate plan: specialized pallial areas for spatial, temporal and emotional memory in actinopterygian fish. Brain Research Bulletin, 66, 277–281.

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

351

Broglio, C., G´omez, Y., L´opez, J.C., Rodr´ıguez, F., Salas, C. & Vargas, J.P. (2000) Encoding of geometric and featural properties of a spatial environment in teleostean fish (Carassius auratus). XXVII International Congress of Psychology, Stockholm. Broglio, C., Rodr´ıguez, F., G´omez, A., Arias, J.L. & Salas, C. (2010) Selective involvement of the goldfish lateral pallium in spatial memory. Behavioural Brain Research, 210, 191–201. Broglio, C., Rodr´ıguez, F. & Salas, C. (2003) Spatial cognition and its neural basis in teleost fishes. Fish & Fisheries, 4, 247–255. Burgess, N., Jeffery, K.J. & O’Keefe, J. (1999) The Hippocampal and Parietal Foundations of Spatial Cognition. Oxford University Press, London. Burnett, L.R., Stein, B.E., Chaponis, D. & Wallace, M.T. (2004) Superior colliculus lesions preferentially disrupt multisensory orientation. Neuroscience, 124, 535–547. Butler, A.B. (2000) Topography and topology of the teleost telencephalon: a paradox resolved. Neuroscience Letters, 293, 95–98. Butler, A.B. & Hodos, H. (2005) Comparative Vertebrate Neuroanatomy: Evolution and Adaptation, 2nd edn. Wiley-Liss, New York. Cheng, K. (1986) A purely geometric module in the rat’s spatial representation. Cognition, 23, 149–178. Cheng, K. & Gallistel, C.R. (1984) Testing the geometric power of an animal’s spatial representation. In: H.L. Roitblat, T.G. Bever & H.S. Terrace (eds) Animal Cognition, pp. 409–423. Erlbaum, Hillsdale, NJ. Cheng, K. & Newcombe, N.S. (2005) Is there a geometric module for spatial orientation? Squaring theory and evidence. Psychonomic Bulletin & Review, 12, 1–23. Chiesa, A.D., Speranza, M., Tommasi, L. & Vallortigara, G. (2006) Spatial cognition based on geometry and landmarks in the domestic chick (Gallus gallus). Behavioural Brain Research, 175, 119–127. Christian, K.M. & Thompson, R.F. (2003) Neural substrates of eyeblink conditioning: acquisition and retention. Learning and Memory, 10, 427–455. Clayton, N.S. & Dickinson, A. (1998) Episodic-like memory during cache recovery by scrub jays. Nature, 395, 272–274. Cnotka, J., Mohle, M. & Rehkamper, G. (2008) Navigational experience affects hippocampus size in homing pigeons. Brain, Behavior and Evolution, 72, 233–238. Costa, S.S., Andrade, R., Carneiro, L.A., Gonc¸alves, E.J., Kotrschal, K. & Oliveira, R.F. (2011) Sex differences in the dorsolateral telencephalon correlate with home range size in blenniid fish. Brain, Behavior and Evolution, 77, 55–64. Crosby, E.C. & Schnitzlein, H.N. (1983) Comparative Correlative Neuroanatomy of the Vertebrate Telencephalon. Macmillan Publishing Co., New York. Davey, G. (1989) Comparative aspects of conditioning: Pavlovian learning. In: G. Davey (ed) Ecological Learning Theory, pp. 23–57. Routledge, London. Davis, M., Hitchcock, J.M. & Rosen, J.B. (1992) A neural analysis of fear conditioning. In: E. Gormezano & E. Wasserman (eds) Learning and Memory: the Behavioral and Biological Substrates, pp. 153–181. Lawrence Erlbaum, Hillsdale, NJ. Davis, R.E. & Klinger, P.D. (1987) Spatial discrimination in goldfish following bilateral tectal ablation. Behavioural Brain Research, 25, 255–260. de Bruin, J.P.C. (1980) Telencephalon and behavior in teleost fish. A neuroethological approach. In: S.O.E. Ebbesson (ed) Comparative Neurology of the Telencephalon. Plenum Press, New York. de Bruin, J.P.C. (1983) Neural correlates of motivated behavior in fish. In: J.P. Ewert, R.R. Capranica & D.J. Ingle (eds) Advances in Vertebrate Neuroethology, pp. 969–995. Plenum Press, New York. Deacon, T.W. (1990) Rethinking mammalian brain evolution. American Zoologist, 30, 629–705. Demski, L.S. (1983) Behavioral effects of electrical stimulation of the brain. In: R.E. Davis & R.G. Northcutt (eds) Fish Neurobiology. Volume 2. Higher Brain Areas and Functions, pp. 317–359. The Michigan University Press, Ann Arbor.

BLBK374-15

BLBK374-Brown

352

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Derenzini, M. (2000) The AgNORs. Micron, 31, 117–120. Desjardins, J.K. & Fernald, R.D. (2010) What do fish make of mirror images? Biology Letters, 6, 744–747. Du Lac, S. & Knudsen, E.I. (1990) Neural maps of head movement vector and speed in the optic tectum of the barn owl. Journal of Neurophysiology, 63, 131–146. ´ Dur´an, E., G´omez, A., Oca˜na, F.M., Alvarez, E., Broglio, C., Jim´enez-Moya, F., Rodr´ıguez, F. & Salas, C. (2004) Cerebellum and spatial learning in teleost fish. FENS Forum Abstracts, p. A112.15. Dur´an, E., Oca˜na, F.M., Broglio, C., Rodr´ıguez, F. & Salas, C. (2010) Lateral but not medial telencephalic pallium ablation impairs the use of goldfish spatial allocentric strategies in a “hole-board” task. Behavioural Brain Research, 214, 480–487. Dur´an, E., Oca˜na, F.M., G´omez, A., Jim´enez-Moya, F., Broglio, C., Rodr´ıguez, F. & Salas, C. (2008) Telencephalon ablation impairs goldfish allocentric spatial learning in a “hole-board” task. Acta Neurobiologiae Experimentalis, 68, 1–13. Eichenbaum, H. (2000) A cortical-hippocampal system for declarative memory. Nature Reviews Neuroscience, 1, 41–50. Eichenbaum, H., Dudchenko, P., Wood, E., Shapiro, M. & Tanila, H. (1999) The hippocampus, memory, and place cells: is it spatial memory or memory space? Neuron, 23, 1–20. Flood, N.B., Overmier, J.B. & Savage, G.E. (1976) The teleost telencephalon and learning: an interpretative review of data and hypotheses. Physiology & Behavior, 16, 783–798. Folgueira, M., Anad´on, R. & Y´an˜ ez, J. (2003) Experimental study of the connections of the gustatory system in the rainbow trout, Oncorhynchus mykiss. Journal of Comparative Neurology, 465, 604–619. Folgueira, M., Anad´on, R. & Y´an˜ ez, J. (2004) Experimental study of the connections of the telencephalon in the rainbow trout (Oncorhynchus mykiss). II: dorsal area and preoptic region. Journal of Comparative Neurology, 480, 204–233. Gallistel, C.R. (1990) The Organization of Learning. MIT Press, Cambridge. Gallon, R.L. (1972) Effects of pre-training with fear and escape conditioning on shuttle-box avoidance acquisition by goldfish. Psychological Reports, 31, 919–924. Gherladucci, B. & Sebastiani, L. (1996) Contribution of the cerebellar vermis to cardiovascular control. Journal of Autonomic Nervous System, 56, 149–156. ´ G´omez, A., Alvarez, E., Dur´an, E., Oca˜na, F.M., Broglio, C., Jim´enez-Moya, F., Salas, C. & Rodr´ıguez, F. (2004) Delay vs trace conditioning following pallium ablation in goldfish. FENS Forum Abstracts, p. A042.10. G´omez, A., Dur´an, E., Salas, C. & Rodr´ıguez, F. (2010) Cerebellum lesion impairs eyeblink-like classical conditioning in goldfish. Neuroscience, 166, 49–60. Gonz´alez-Lima, F. & Cada, A. (1998) Quantitative histochemistry of cytochrome oxidase activity: theory, methods, and regional brain vulnerability. In: Gonzalez-Lima, F. (ed) Cytochrome Oxidase in Neuronal Metabolism and Alzheimer’s Disease, pp. 263–280. Plenum, New York. Good, M. (1987) The effects of hippocampal-area parahippocampalis lesions on discrimination learning in the pigeon. Behavioural Brain Research, 31, 207–220. Good, M. & Macphail, E.M. (1994) The avian hippocampus and short-term memory for spatial and non-spatial information. The Quarterly Journal of Experimental Psychology. B, 47, 293–317. Grandel, H., Kaslin, J., Ganz, J., Wenzel, I. & Brand, M. (2006) Neural stem cells and neurogenesis in the adult zebrafish brain: origin, proliferation dynamics, migration and cell fate. Developmental Biology, 295, 263–277. Guttridge, T.L., Myrberg, A.A., Porcher, I.F., Sims, D.W. & Krause, J. (2009) The role of learning in shark behaviour. Fish & Fisheries, 10, 450–469. Hampton, R.R., Hampstead, B.M. & Murray, E.A. (2004) Selective hippocampal damage in rhesus monkeys impairs spatial memory in an open-field test. Hippocampus, 14, 808–818. Hampton, R.R., Sherry, D.F., Shettleworth, S.J., Khurgel, M. & Ivy, G. (1995) Hippocampal volume and food-storing behavior are related in parids. Brain, Behavior and Evolution, 45, 54–61.

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

353

Han, V.Z. & Bell, C.C. (2003) Physiology of cells in the central lobes of the mormyrid cerebellum. The Journal of Neuroscience, 23, 11147–11157. Han, V.Z., Zhang, Y., Bell, C.C. & Hansel, C. (2007) Synaptic plasticity and calcium signaling in Purkinje cells of the central cerebellar lobes of mormyrid fish. The Journal of Neuroscience, 27, 13499–13512. Healy, S.D., Gwinner, E. & Krebs, J.R. (1996) Hippocampal volume in migratory and non-migratory warblers: effects of age and experience. Behavioral Brain Research, 81, 61–68. Herrero, L., Rodr´ıguez, F., Salas, C. & Torres, B. (1998) Tail and eye movements evoked by electrical microstimulation of the optic tectum in goldfish. Experimental Brain Research, 120, 291–305. Hirsch, R. & Segal, M. (1972) Complete transection of the fornix and reversal of position in the rat. Physiology & Behavior, 8, 1051–1054. Hodos, W. & Campbell, C.B.G. (1969) The scala naturae: why there is no theory in comparative psychology. Psychological Review, 76, 337–350. Holahan, M.R., Rekart, J.L., Sandoval, J. & Routtenberg, A. (2006) Spatial learning induces presynaptic structural remodeling in the hippocampal mossy fiber system of two rat strains. Hippocampus, 16, 560–570. Ikenaga, T., Yoshida, M. & Uematsu, K. (2006) Cerebellar efferent neurons in teleost fish. Cerebellum, 5, 268–274. Jacobs, L.F. & Spencer, W.D. (1994) Natural space-use patterns and hippocampal size in kangaroo rats. Brain, Behavior and Evolution, 44, 125–132. Jerison, H. (1973) Evolution of the Brain and Intelligence. Academic Press, New York. Jirenhed, D.A., Bengtsson, F. & Hesslow, G. (2007) Acquisition, extinction, and reacquisition of a cerebellar cortical memory trace. The Journal of Neuroscience, 27, 2493–2502. Kapsimali, M., Vidal, B., Gonzalez, A., Dufour, S. & Vernier, P. (2000) Distribution of the mRNA encoding the four dopamine D(1) receptor subtypes in the brain of the European eel (Anguilla anguilla): comparative approach to the function of D(1) receptors in vertebrates. Journal of Comparative Neurology, 419, 20–43. Kehoe, E.J. & Macrae, M. (2002) Fundamental behavioral methods and findings in classical conditioning. In: J.W. Moore (ed) A Neuroscientist’s Guide to Classical Conditioning, pp. 171–231. Springer, New York. Kotani, S., Kawahara, S. & Kirino, Y. (2002) Classical eyeblink conditioning in decerebrate guinea pigs. The European Journal of Neuroscience, 15, 1267–1270. Kotchabhakdi, N. (1976) Functional circuitry of the goldfish cerebellum. Journal of Comparative Physiology, 112, 47–73. Lamprecht, R. & Dudai, Y. (2000) The amygdala in conditioned taste aversion: it’s there but where? In: J. Aggleton (ed) The Amygdala, pp. 331–351. Oxford University Press, New York. LeDoux, J.E. (2000) Emotion circuits in the brain. Annual Review of Neuroscience, 23, 155–184. L´opez, J.C., Bingman, V.P., Rodr´ıguez, F., G´omez, Y. & Salas, C. (2000a) Dissociation of place and cue learning by telencephalic ablation in goldfish. Behavioral Neuroscience, 114, 687–699. L´opez, J.C., Broglio, C., Rodr´ıguez, F., Thinus-Blanc, C. & Salas, C. (1999) Multiple spatial learning strategies in goldfish (Carassius auratus). Animal Cognition, 2, 109–120. L´opez, J.C., Broglio, C., Rodr´ıguez, F., Thinus-Blanc, C. & Salas, C. (2000b) Reversal learning deficit in a spatial task but not in a cued one after telencephalic ablation in goldfish. Behavioural Brain Research, 109, 91–98. L´opez, J.C., Gomez, Y., Vargas, J.P. & Salas, C. (2003a) Spatial reversal learning deficit after medial cortex lesion in turtles. Neuroscience Letters, 341, 197–200. L´opez, J.C., Vargas, J.P., G´omez, Y. & Salas, C. (2003b) Spatial and non-spatial learning in turtles: the role of medial cortex. Behavioural Brain Research, 143, 109–120. Luque, M.A. P´erez-P´erez, M.P. Herrero, L. & Torres, B. (2005) Involvement of the optic tectum and mesencephalic reticular formation in the generation of saccadic eye movements in goldfish. Brain Research Reviews, 49, 388–397. MacLean, P. (1990) The Triune Brain in Evolution. Plenum Press, New York.

BLBK374-15

BLBK374-Brown

354

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Maren, S. (2001) Neurobiology of Pavlovian fear conditioning. Annual Review of Neuroscience, 24, 897–931. Martin, I. (2009) Cerebellum and classical conditioning in the teleost Carassius auratus. Study with lesion and electrophysiological recording techniques. PhD thesis, University of Sevilla, Seville. Martin, I., G´omez, A., Salas, C. & Rodr´ıguez, F. (2009) Teleost cerebellum and heart rate conditioning. 41st European Brain and Behaviour Society Meeting and 23rd Hellenic Society for Neuroscience Meeting, Rodas, September 13–19, 2009. Frontiers in Behavioral Neuroscience. Conference Abstract: 41st European Brain and Behaviour Society Meeting. doi: 10.3389/conf. neuro.08.2009.09.224. Mart´ın, I., G´omez, A., Salas, C., Puerto, A. & Rodr´ıguez, F. (2011) Dorsomedial pallium lesions impair taste aversion learning in goldfish. Neurobiology of Learning & Memory. (In press.) Maschke, M., Schugens, M., Kindsvater, K., Drepper, J., Kolb, F.P., Diener, H.C., Daum, I. & Timmann, D. (2002) Fear conditioned changes of heart rate in patients with medial cerebellar lesions. Journal of Neurology, Neurosurgery, and Psychiatry, 72, 116–118. Matsumoto, N., Yoshida, M. & Uematsu, K. (2007) Effects of partial ablation of the cerebellum on sustained swimming in goldfish. Brain, Behavior and Evolution, 70, 105–114. Mauk, M.D. & Thompson, R.F. (1987) Retention of classically conditioned eyelid responses following acute decerebration. Brain Research, 403, 89–95. McGaugh, J.L. (2004) The amygdala modulates the consolidation of memories of emotionally arousing experiences. Annual Review of Neuroscience, 27, 1–28. Meek, J. (1992) Comparative aspects of cerebellar organization. From mormyrids to mammals. European Journal of Morphology, 30, 37–51. Meek, J. & Nieuwenhuys, R. (1998) Holosteans and teleosts. In: R. Nieuwenhuys, H.J. ten Donkelaar & C. Nicholson (eds) The Central Nervous System of Vertebrates, pp. 759–937. SpringerVerlag, Berlin. Meek, J., Yang, J.Y., Han, V.Z. & Bell, C.C. (2008) Morphological analysis of the mormyrid cerebellum using immunohistochemistry, with emphasis on the unusual neuronal organization of the valvula. The Journal of Comparative Neurology, 510, 396–421. Molinari, M., Chiricozzi, F.R., Clausi, S., Tedesco, A.M., De Lisa, M. & Leggio, M.G. (2008) Cerebellum and detection of sequences, from perception to cognition. Cerebellum, 7, 611–615. Mowrer, O.H. (1960) Learning Theory and Behavior. John Wiley & Sons, New York. Moyer, J.R., Deyo, R.A. & Disterhoft, J.F. (1990) Hippocampectomy disrupts trace eyeblink conditioning in rabbits. Behavioral Neuroscience, 104, 243–252. Mueller, T. & Wullimann, M.F. (2009) An evolutionary interpretation of teleostean forebrain anatomy. Brain, Behavior and Evolution, 74, 30–42. Nadel, L. (1991) The hippocampus and space revisited. Hippocampus, 1, 221–229. Nader, K., Schafe, G.E. & Le Doux, E. (2000) Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. Nature, 406, 722–726. Nieuwenhuys, R. (2009) The forebrain of actinopterygians revisited. Brain, Behavior and Evolution, 73, 229–252. Nieuwenhuys, R. (2011) The development and general morphology of the telencephalon of actinopterygian fishes: synopsis, documentation and commentary. Brain Structure and Function, 215, 141–157. Nieuwenhuys, R., ten Donkelaar, H.J. & Nicholson, C. (1998) The Central Nervous System of Vertebrates. Springer-Verlag, Berlin. Nonneman, A.J., Voigt, J. & Kolb, B.E. (1974) Comparisons of behavioural effects of hippocampal and prefrontal cortex lesions in the rat. Journal of Comparative and Physiological Psychology, 87, 249–260. Norman, R.J., Buchwald, J.S. & Villablanca, J.R. (1977) Classical conditioning with auditory discrimination of the eyeblink in decerebrate cats. Science, 196, 551–553. Northcutt, R.G. (1995) The forebrain of gnathostomes: in search of a morphotype. Brain, Behavior and Evolution, 46, 275–318.

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

355

Northcutt, R.G. (2006) Connections of the lateral and medial divisions of the goldfish telencephalic pallium. Journal of Comparative Neurology, 494, 903–943. Northcutt, R.G. (2008) Forebrain evolution in ray-finned fishes. Brain Research Bulletin, 75, 191–205. Northcutt, R.G. & Braford, M.R. (1980) New observations on the organization and evolution of the telencephalon in actinopterygian fishes. In: S.O.E. Ebbesson (ed) Comparative Neurology of the Telencephalon, pp. 41–98. Plenum Press, New York. Odling-Smee, L., Simpson, S.D. & Braithwaite, V.A. (2006) The role of learning in fish orientation. In: C. Brown, K. Laland & J. Krause (eds) Fish Cognition and Behavior, pp. 119–138. Blackwell Publishing Ltd., Oxford. O’Keefe, J. & Nadel, L. (1978) The Hippocampus as a Cognitive Map. Clarendon Press, Oxford. Overmier, J.B. & Gross, D. (1974) Effects of telencephalic ablation upon nest-building and avoidance behaviors in East African mouth breeding fish, Tilapia mossambica. Behavioral Biology, 12, 211–222. Overmier, J.B. & Hollis, K.L. (1990) Fish in the think tank: learning, memory and integrated behavior. In: R.P. Kesner & D.S. Olton (eds) Neurobiology of Comparative Cognition, pp. 204–236. Lawrence Erlbaum Associates, Hillsdale, NJ. Overmier, J.B. & Papini, M.R. (1986) Factors modulating the effects of teleost telencephalon ablation on retention, relearning, and extinction of instrumental avoidance behavior. Behavioral Neuroscience, 100, 190–199. Overmier, J.B. & Starkman, N. (1974) Transfer of control of avoidance behavior in normal and telencephalon ablated goldfish (Carassius auratus). Physiology & Behavior, 12, 605–608. Papez, J. (1929) Comparative Neurology. Crowell, New York. Perrault, T.J., Vaughan, J.W., Stein, B.E. & Wallace, M.T. (2003) Neuron-specific response characteristics predict the magnitude of multisensory integration. Journal of Neurophysiology, 90, 4022–4026. Petrosini, L., Leggio, M.G. & Molinari, M. (1998) The cerebellum in spatial problem solving: a co-start or a guest start? Progress in Neurobiology, 56, 191–210. Phillips, R.G. & LeDoux, J.E. (1992) Differential contribution of amygdala and hippocampus to cued and contextual fear conditioning. Behavioral Neuroscience, 106, 274–285. Portavella, M., Torres, B. & Salas, C. (2004) Avoidance response in goldfish: emotional and temporal involvement of medial and lateral telencephalic pallium. Journal of Neuroscience, 24, 2335–2342. Portavella, M., Vargas, J.P., Salas, C. & Papini, M. (2003) Involvement of the telencephalon in spaced-trial avoidance learning in the goldfish (Carassius auratus). Physiology & Behavior, 80, 49–56. Powell, D.A. & Buchanan, S. (1980) Autonomic–somatic relationships in the rabbit (Oryctolagas cuniculus): effects of hippocampal lesions. Physiological Psychology, 8, 455–462. Quick, I.A. & Laming, P.R. (1988) Cardiac, ventilatory and behavioural arousal responses evoked by electrical stimulation in the goldfish (Carassius auratus). Physiology & Behavior, 43, 715–727. Ramirez-Amaya, V., Balderas, I., Sandoval, J., Escobar, M.L. & Berm´udez-Rattoni, F. (2001) Spatial long-term memory is related to mossy fiber synaptogenesis. Journal of Neuroscience, 21, 7340–7348. Ramirez-Amaya, V., Escobar, M.L., Chao, V. & Berm´udez-Rattoni, F. (1999) Synaptogenesis of mossy fibers induced by spatial water maze overtraining. Hippocampus, 9, 631–636. Rodr´ıguez, F. (1996) Mecanismos tectales implicados en la orientaci´on espacial en el carp´ın dorado (Carassius auratus): un estudio mediante t´ecnicas de lesi´on y de microestimulaci´on el´ectrica localizada. PhD thesis, Universidad de Sevilla, Seville. ´ Rodr´ıguez, F., Dur´an, E., G´omez, A., Oca˜na, F.M., Avarez, E., Jim´enez-Moya, F. Broglio, C. & Salas, C. (2005) Cognitive and emotional functions of the teleost fish cerebellum. Brain Research Bulletin, 66, 365–370. Rodr´ıguez, F., Dur´an, E., Vargas, J.P., Torres, B. & Salas, C. (1994) Performance of goldfish trained in allocentric and egocentric maze procedures suggests the presence of a cognitive mapping system in fishes. Animal Learning and Behavior, 22, 409–420.

BLBK374-15

BLBK374-Brown

356

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Rodr´ıguez, F., L´opez, J.C., Vargas, J.P., G´omez, Y., Broglio, C. & Salas, C. (2002) Conservation of spatial memory function in the pallial forebrain of amniotes and ray-finned fishes. Journal of Neuroscience, 22, 2894–2903. Romer, A.S. (1962) The Vertebrate Body. Saunders, Philadelphia. Rondhi-Reig, L., Le Marec, N., Caston, J. & Mariani, J. (2002) The role of climbing and parallel fibers inputs to cerebellar cortex in navigation. Behavioral Brain Research, 132, 11–18. Roth, E.D., Lutterschmidt, W.I. & Wilson, D.A. (2006) Relative medial and dorsal cortex volume in relation to sex differences in spatial ecology of a snake population. Brain, Behavior and Evolution, 67, 103–110. Sacchetti, B., Baldi, E., Lorenzini, C.A. & Bucherelli, C. (2002) Cerebellar role in fear-conditioning consolidation. Proceedings of the National Academic of Sciences USA, 99, 8406–8411. Sacchetti, B., Sacco, T. & Strata, P. (2007) Reversible inactivation of amygdala, cerebellum, but not perirhinal cortex, impairs reactivated fear memories. The European Journal of Neuroscience, 25, 2875–2884. Sacchetti, B., Scelfo, B. & Strata, P. (2009) Cerebellum and emotional behavior. Neuroscience, 162, 756–762. Sacchetti, B., Scelfo, B. & Strata, P. (2005) The cerebellum: synaptic changes and fear conditioning. The Neuroscientist, 11, 217–227. Sacchetti, B., Scelfo, B., Tempia, F. & Strata, P. (2004) Long-term synaptic changes induced in the cerebellar cortex by fear conditioning. Neuron, 42, 973–982. Salas, C., Broglio, C., Dur´an, E., G´omez, A. & Rodr´ıguez, F. (2008) Spatial learning in fish. In: R. Menzel (ed) Learning Theory and Behavior. Vol. 1. In: J. Byrne (ed) Learning and Memory: A Comprehensive Reference, pp. 499–528. Elsevier, Oxford. Salas, C., Broglio, C. & Rodr´ıguez, F. (2003) Evolution of forebrain and spatial cognition in vertebrates: conservation across diversity. Brain, Behavior and Evolution, 62, 72–82. Salas, C., Broglio, C., Rodr´ıguez, F., L´opez, J.C., Portavella, M. & Torres, B. (1996a) Telencephalic ablation in goldfish impairs performance in a spatial constancy problem but not in a cued one. Behavioural Brain Research, 79, 193–200. Salas, C., Herrero, L., Rodr´ıguez, F. & Torres, B. (1997) Tectal codification of eye movements in goldfish studied by electrical microstimulation. Neuroscience, 78, 271–288. Salas, C., Rodr´ıguez, F., Vargas, J.P., Dur´an, E. & Torres, B. (1996b) Spatial learning and memory deficits after telencephalic ablation in goldfish trained in place and turn maze procedures. Behavioral Neuroscience, 110, 965–980. Savage, G.E. (1971) Behavioural effects of electrical stimulation of the telencephalon of the goldfish, Carassius auratus. Animal Behaviour, 19, 661–668. Scelfo, B., Sacchetti, B. & Strata, P. (2008) Learning-related long-term potentiation of inhibitory synapses in the cerebellar cortex. Proceedings of the National Academy of Sciences of the USA, 105, 769–774. Schluessel, V. & Bleckmann, H. (2005) Spatial memory and orientation strategies in the elasmobranch Potamotrygon motoro. Journal of Comparative Physiology A, 191, 695–706. Segaar, J. & Nieywenhuys, R. (1963) New etho-physiological experiments with male Gasterosteus aculeatus, with anatomical comment. Animal Behaviour, 11, 331–344. Shapiro, S.M., Schuckman, H., Sussman, D. & Tucker, A.M. (1974) Effect of telencephalic lesions on the gill cover response of Siamese fighting fish. Physiology & Behavior, 13, 749–755. Sherry, D.F. & Duff, S.J. (1996) Behavioral and neural bases of orientation in food storing birds. Journal of Experimental Biology, 199, 165–172. Shumway, C.A. (2008) Habitat complexity, brain, and behavior. Brain, Behavior and Evolution, 72, 123–134. Sovrano, V.A., Bisazza, A. & Vallortigara, G. (2002) Modularity and spatial reorientation in a simple mind: encoding of geometric and nongeometric properties of a spatial environment by fish. Cognition, 85, 51–59. Sovrano, V.A., Bisazza, A. & Vallortigara, G. (2003) Modularity as a fish (Xenotoca eiseni) views it: conjoining geometric and nongeometric information for spatial reorientation. Journal of Experimental Psychology. Animal Behavior Processes, 29, 199–210.

BLBK374-15

BLBK374-Brown

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Brain and Cognition in Teleost Fish

Char Count=

357

Sovrano, V.A., Bisazza, A. & Vallortigara, G. (2007) How fish do geometry in large and in small spaces. Animal Cognition, 10, 47–54. Sparks, D.L. (2002) The brainstem control of saccadic eye movements. Nature Reviews Neuroscience, 3, 952–964. Squire, L.R., Stark, C.E. & Clark RE. (2004) The medial temporal lobe. Annual Review of Neuroscience, 27, 279–306. Stein, B.E. & Meredith, M.A. (1993) The Merging of the Senses. MIT Press, Cambridge. Striedter, G.F. (2005) Principles of Brain Evolution. Sinauer Associates, Sunderland. Supple, W.P. & Kapp, B.S. (1993) The anterior cerebellar vermis: essential involvement in classically conditioned bradycardia in the rabbit. Journal of Neuroscience, 13, 3705–3711. Supple, W.F. & Leaton, R.N. (1990a) Cerebellar vermis: essential for classically conditioned bradycardia in the rat. Brain Research, 509, 17–23. Supple, Jr W.F. & Leaton, R.N. (1990b) Lesions of the cerebellar vermis and cerebellar hemispheres: effects on heart rate conditioning in rats. Behavioral Neuroscience, 104, 934–947. Supple, W.F., Sebastiani, L. & Kapp, B.S. (1993) Purkinje cell responses in the anterior cerebellar vermis during Pavlovian fear conditioning in the rabbit. Neuroreport, 4, 975–978. Thinus-Blanc, C. (1996) Animal Spatial Cognition. Behavioral and Neural Approaches. World Scientific Publishing Co., Singapore. Thompson, R.F. & Steinmetz, J.E. (2009) The role of the cerebellum in classical conditioning of discrete behavioral responses. Neuroscience, 162, 732–755. Torres, B., Luque, M.A. P´erez-P´erez, M.P. & Herrero L. (2005) Visual orienting response in goldfish: a multidisciplinary study. Brain Research Bulletin, 66, 376–380. Turner, B.M., Paradiso, S., Marvel, C.L., Pierson, R., Boles Ponto, L.L., Hichwa, R.D. & Robinson, R.G. (2007) The cerebellum and emotional experience. Neuropsychologia, 45, 1331– 1341. Vanegas, H. (1984) Comparative Neurology of the Optic Tectum. Plenum Press, New York. Vargas, J.P., Lopez, J.C., Salas, C. & Thinus-Blanc, C. (2004) Encoding of geometric and featural spatial information by goldfish (Carassius auratus). Journal of Comparative Psychology, 118, 206–216. Wong Riley, M.T.T. (1989) Cytochrome oxidase: an endogenous metabolic marker for neuronal activity. Trends in Neurosciences, 12, 94–101. Woodruff, M.L. & Kantor, H. (1983) Fornix lesions, plasma ACTH levels, and shuttle box avoidance in rats. Behavioral Neuroscience, 97, 897–907. Woodruff-Pak, D.S. & Disterhoft, J.F. (2008) Where is the trace in trace conditioning? Trends in Neurosciences, 3, 105–112. Wulliman, M.F. & Mueller, T. (2004) Teleostean and mammalian forebrains contrasted: evidence from genes to behavior. Journal of Comparative Neurology, 75, 143–162. Wullimann, M.F. & Northcutt, R.G. (1988) Connections of the corpus cerebelli in the green sunfish and common goldfish: a comparison of perciform and cypriniform teleost. Brain, Behavior and Evolution, 32, 293–316. Xue, H.G., Yang, C.Y. & Yamamoto, N. (2008) Afferent sources to the inferior olive and distribution of the olivocerebellar climbing fibers in cyprinids. Journal of Comparative Neurology, 507, 1409–1427. Yamamoto, T., Shimura, T., Sako, N., Yasoshima, Y. & Sakai, N. (1994) Neural substrates for conditioned taste aversion in the rat. Behavioural Brain Research, 65, 123–137. Yamamoto, N., Ishikawa, Y., Yoshimoto, M., Xue, H.G., Bahaxar, N., Sawai, N., Yang, C.Y., Ozawa, H. & Ito, H. (2007) A new interpretation on the homology of the teleostean telencephalon based on hodology and a new eversion model. Brain, Behavior and Evolution, 69, 96–104. Yoshida, M., Okamura, I. & Uematsu, K. (2004) Involvement of the cerebellum in classical fear conditioning in goldfish. Behavioural Brain Research, 153, 143–148. Yoshimoto, M. & Yamamoto, N. (2010) Ascending general visceral sensory pathways from the brainstem to the forebrain in a cichlid fish, Oreochromis niloticus (Tilapia). Journal of Comparative Neurology, 518, 3570–3603.

BLBK374-15

BLBK374-Brown

358

May 27, 2011

11:13

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Zhu, L., Scelfo, B., Hartell, N.A., Strata, P. & Sacchetti, B. (2007) The effects of fear conditioning on cerebellar LTP and LTD. European Journal of Neuroscience, 26, 219–227. Zhuikov, A.Y., Couvillon, P.A. & Bitterman, M.E. (1994) Quantitative two-process analysis of avoidance conditioning in goldfish. Journal of Experimental Psychology: Animal Behavior Processes, 20, 32–43. Zupanc, G.K.H., Hinsch, K. & Gage, F.H. (2005) Proliferation, migration, neuronal differentiation, and long-term survival of new cells in the adult zebrafish brain. Journal of Comparative Neurology, 488, 290–319.

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 16

Fish Behaviour, Learning, Aquaculture and Fisheries Anders Fern¨o, Geir Huse, Per Johan Jakobsen, Tore S. Kristiansen and Jonatan Nilsson

16.1

Fish learning skills in the human world

Human beings have harvested fish resources for thousands of years, but only during the past century has the development of new fishing technologies produced detrimental impacts on fishery resources. The relative importance of commercial fisheries in regulating populations and marine ecosystems has increased dramatically. In the course of the last century, a worldwide aquaculture industry has developed, and within a few decades the production of farmed fish will presumably exceed landings of wild fish. The rapidly changing technology in both fisheries and aquaculture challenges fishes exposed to human impact in novel ways that differ from those that threaten their survival in the ancestral environments. Animals categorise the world that surrounds them into objects of relevance to their survival and reproduction and objects of less or no relevance. The evolved ability of animals to categorise depends upon their sensory apparatus and their capacity to learn about and manipulate objects. In other words, animals are commonly seen as neural units shaped by the ghost of selection past, and the way in which animals categorise events and objects may be preadapted and tailored to their ecological niche (Rozin & Kalat 1972). For instance, by transplanting pieces of quail neural tissue into chicken embryos, Balaban (1997) was able to demonstrate two different, but specific, quail behaviours in chickens, suggesting that the functions of at least some neural circuits are innate, species-specific, and may thus limit the capacity to categorise evolutionarily novel situations adequately. As for automated responses, learning skills may be pre-programmed and adapted to environmental factors. The development of cognitive skills (spatial learning, problem solving, etc.) in fishes seems to be associated with visual orientation and well-structured habitats (Kotrschal et al. 1998; Brown et al. 2003; Odling-Smee et al. 2008). Although classical conditioning is widely applicable, various constraints may make it more adaptive. Fishes can be more readily conditioned to moving than to stationary stimuli (Wisenden & Harter 2001), and certain key stimuli are crucial when fishes learn to recognise Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-16

BLBK374-Brown

360

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

and avoid predators (Cs´anyi 1986; Altb¨acker & Cs`anyi 1990). Rats are able to associate taste with sickness, but not with electrical shock (Garcia & Koelling 1966), as inner pain presumably results from something ingested and surface pain is inflicted by an external agent (Johnston 1999). Similar surprises can be found in fisheries. A fish may perceive a fishing net, but associate it with vegetation. Since fishes can pass through vegetation, they may also attempt to pass through the net rather than avoid it. The lack of stimulus relevance of man-made constructions such as fishing gear and aquaculture installations may thus have profound consequences for the ability of fish to learn. Perceptual concepts encountered in modern fisheries and fish farming may not resemble those that occur in the natural environment, in that selected responses to man-made facilities will coincide with innate responses in some cases and contrast with them in others (Brown & Warburton 1999a and references therein for responses to trawls). The result may be a mismatch between the fishes and recent anthropogenic changes in environments, whereby prior experience, innate behaviour, locomotor ability and morphology together render them less adapted to the new situation. Knowledge of how to perform a given behaviour includes procedural representations, which are a set of inner instructions related to a specific procedure (McFarland 1985). This will depend on the particular neural network of the species and combine the genetically preprogrammed circuits with experience. Hence, experience may be fixed as procedural representations that are dependent on the ‘value system’ of each association. When environmental cues are novel, it is not obvious that the animal has predefined value systems as in the Hebb type of learning (Hebb 1949), because the content does not provide for appropriate links between behaviour and environment. When this value system is changed due to an introduced and novel mismatch with the former value system, it is difficult to reliably predict which patterns will emerge. Demands for cognitive control fluctuate over time and must be calibrated to appropriate levels on every occasion. Thus, high-conflict events may arise when stimuli mismatch in ‘colour and world’ on consecutive trials (Kerns et al. 2004). Viewed as a multilevel value system, emergent properties often arise due to mismatched inter-item competition (Morris et al. 2009). On the other hand, for behaviour and learning in natural systems where such predefined value system has been under selection, predictions on this level can be derived from evolutionary models. Examples of emergent properties in aquaculture and fisheries might be self-organisation on new levels based on artificial and hence novel associations. This may produce large, long-lasting effects as a result of small differences in environments. Additionally, cultural breakdowns of fish populations due to overfishing may create a completely new culture after population restoration due to emergent properties of the individual procedural representations. The learning abilities of fishes could create problems for humans if fishes learn to avoid vessels and gear. Learning could also lead to less predictable spatial distributions and migration routes, making it more difficult to localise fish aggregations with consequent implications for fisheries and stock assessment and management. Decadal changes in climatic conditions may favour the learning of migration patterns rather than more fixed migration strategies. Learning is particularly valuable for long-lived species that may rely on previous experience when making decisions. Thus, a horse mackerel (Trachurus trachurus, Carangidae) that lives up to 20 years could benefit from previous knowledge when making decisions about its summer migration path, while a capelin (Mallotus villosus, Osmeridae)

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

361

FISH LEARNING Social

Solitary

Individual learning Habituation Classical conditioning Operant conditioning Imprinting

Social learning Horizontal transmission Vertical transmission SPATIAL–TEMPORAL SCALE

Constraints Sensory capacity Mental capacity Species and context dependent Interaction process limitations Fig. 16.1 Different fish learning mechanisms and their constraints playing a role in fisheries and aquaculture. The neural network of individual fish is specified by a set of innate parameters that must learn from a continuous stream of input patterns about how to classify future input patterns. Learning thus has a temporal scale. Individual fitness will be determined by how well a fish classifies the new input before discovering the correct classes and learning from them. This is the spatial scale. Since most fishes are precocial – they do not learn from experience of their parents, so that learning and lifetime experience are products of their own experience, which is individual learning. Social learning is predetermined on the basis of innate traits and modified by learning transmission from the surrounding assemblage of conspecifics. Social learning is dependent on effective interactions between experienced and na¨ıve fish, and requires spatial–temporal overlap between them and similar local levels of abundance.

that spawns only once in its lifetime has to rely on genetically inherited strategies to make its migratory decisions. At the most extreme, the implications of overfishing might be that we could wipe out entire cultural units of long-lived fish populations because there are no longer enough experienced fishes to follow (see Chapter 11). Therefore, frequencydependent social learning may be taken into account all the way up to a holistic ecosystem management perspective. On the other hand, in aquaculture and sea-ranching programmes the ability of fishes to learn may be beneficial, because we wish the fishes to adapt to both a farming environment and free-range conditions. The human ambition is to teach the fishes to deal with environmental cues and limit their distribution to a restricted area. Figure 16.1 outlines the different types of individual and social learning and their constraints that can influence interactions between fishes and man. The fish species studied in applied research are a diverse group that has not been systematically studied with respect to learning skills. Some species have also been studied in particular detail, biasing comparisons between species. Most of the research of fish learning has tended to focus on small freshwater species (guppy (Poecilia reticulata, Poeciliidae), goldfish (Carassius auratus, Cyprinidae), pumpkinseed sunfish (Lepomis gibbosus, Centrarchidae), etc.) that are easily maintained in small laboratories. In this chapter, we offer examples in which different types of learning play a role in fisheries, aquaculture and stock enhancement. The main topics that we discuss here deal with the implications for the efficiency of the equipment used and the influence of fisheries on learned migration patterns. We also discuss the role of learning in reared fishes in aquaculture facilities and following release into the wild.

BLBK374-16

BLBK374-Brown

May 27, 2011

362

Fish Cognition and Behavior

16.2

Fisheries

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

The adapted life history traits and behaviour of fishes in exploited fish populations are contested by fishing activity, and this may lead to the selection of new traits and modification of behaviour by learning. From the human point of view, the aim is to exploit fish populations in a responsible way in order to avoid population collapses and ecosystem breakdowns. To that end it is vital to maintain control of population size as well as to understand how populations interact with other components of the ecosystem at higher and lower trophic levels. The next step is to exploit fish stocks using selective and environmentally friendly fishing gear. If we are to achieve these goals, both natural behaviour and reactions to stimuli from fishing operations and surveys must be taken into account. Fish antipredator and feeding behaviour are shaped by interactions between innate and learned components (see Chapters 2, 3 and 4). Thus, learning influences the species interactions and dynamics of ecosystems with potentially powerful effects on the variations in abundance of commercially exploited species. The effect of relevant ecological processes, including learning, should ideally be incorporated into ecosystem-based management, but this is a complex task that demands quantitative knowledge of modifications of species interactions at several different levels. However, the effect of fish learning skills on the actual interactions between target and non-target species in an ecosystem should be borne in mind when evaluating the realism of population dynamic models and management reference points (Hall & Mainprize 2004). We now consider in particular the role played by learning processes in spatial dynamics, fish capture and abundance estimation.

16.2.1

Spatial dynamics

16.2.1.1

Learning skills and movement

The capacity to learn provides flexibility that is crucial for the ability to utilise a variable environment (Dodson 1988). Habitat utilisation and migration routes in fishes are presumably often influenced by experience, but our knowledge of this field in commercially exploited species is limited. Learning can control movements within a stationary home range as well as long-distance migrations (see Chapter 8). Our understanding of fish migrations is further constrained by the close interplay between navigation and orientation mechanisms, both of which rely to some extent on unknown sensory organs. Much of our insight into migrations in commercial fish thus remains at a fairly descriptive level, and has until recently largely been gathered by means of acoustic methods and tags, with limited knowledge of the motivation of the fish and the environment. However, archival tags that record pressure and temperature can in combination with GPS data now tell us a lot of about where the fish ˚ dlandsvik et al. 2007; Evans et al. 2008; was with respect to the physical environment (A Pedersen et al. 2008; Neuenfeldt et al. 2009). Demersal fishes often stay within a certain home range (Matthews 1990) and may, therefore, be assumed to rely on learned spatial cues (Reese 1989; Braithwaite et al. 1996). For instance, the gadoid ling (Molva molva, Lotidae) occupies a home range in fjords and remains in a core area within the home range about 65% of the time (Løkkeborg et al. 2000). The pelagic environment is more homogeneous than the demersal zone, and pelagic

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

363

fishes are, therefore, often assumed to display less fidelity to a home range. Tuna occur over large areas in tropical and subtropical waters, and may also make excursions into boreal and subArctic waters during the summer (Block et al. 2005). Studies using electronic tags have demonstrated the ability of yellowfin tuna (Thunnus albacares, Scombridae) to navigate accurately between fish aggregating devices (Brill et al. 1999) and to remain in an area over long periods. This suggests that yellowfin tuna are able to maintain core areas around fish aggregating devices through learning much in the same manner as demersal fishes, although they are probably relying on different cues to aid navigation (e.g. polarised light or magnetic fields). On a larger scale, pelagic fishes may locate favourable habitats by using a combination of predictive orientation mechanisms based upon genetic factors and learning, and reactive mechanisms, such as memory-based state-location comparisons and orientation to gradients in the sea (Neill 1979; Fern¨o et al. 1998; Kvamme et al. 2003). Field experiments on Atlantic salmon (Salmo salar, Salmonidae) have further demonstrated that smolts are imprinted on olfactory cues during seawards migration and that they use these cues on returning to their home river for spawning (see Chapter 8). Therefore, learning is important to the movements of a wide range of commercial species. 16.2.1.2

Social learning of migration pattern

Social learning involves the transfer of knowledge between individuals (Chapter 11). The migration patterns of many fishes are learned at an early age, often from older individuals in the stock, so-called ‘guided learning’ (Brown & Laland 2003), and thereafter maintained over time. Helfman & Schultz (1984) performed a classic demonstration of social learning in fishes by showing that daily migrations by French grunts (Haemulon flavolineatum, Haemulidae) between resting and feeding grounds were maintained by guided learning by transplanting individuals from their home ground to a new resting area. The transplanted fishes soon adopted the migration route to the feeding area utilised by the resident fishes. In a control experiment in which fishes were transplanted to a resting area from which the resident fishes had been removed, the transplanted fishes continued to use paths appropriate to their original resting site. There is growing evidence to suggest that guided learning plays a role in migration of important resources such as herring and cod. A few centuries ago, Norwegian fishermen thought that the massive herring schools were guided to the coast by the oarfish (Regalecus glesne, Regalecidae), known as the ‘herring king’ in Norwegian. Indeed, the herring (Clupea harengus, Clupeidae) do seem to be guided, but by their own elders rather than by serpent-like creatures. In some herring populations, there appears to be strong inter-cohort learning (McQuinn 1997; Fern¨o et al. 1998; Corten 1999, 2001). Normally, a recruiting year class learns the migration pattern of the stock by schooling with the older component of the stock (McQuinn 1997). Under non-harvesting conditions, this is likely to be a fairly robust population mechanism. But if the stock collapses and most of the old individuals are lost, the interaction between recruiting and old cohorts may be disrupted, thus breaking the vertical social transmission chain (Chapter 11). This is parallel to the experiment by Helfman & Schultz (1984) in which transplanted fishes did not encounter resident fishes. More generally, the adoption of the adult migration pattern by the recruiting cohorts can be interrupted if the proportion of recruits relative to the adult population is high (Huse et al. 2002a) or when there is lack

BLBK374-16

BLBK374-Brown

364

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

of spatial overlap between cohorts (Corten 2001). The process whereby abundant cohorts ‘repel’ socially transmitted information from older cohorts, and thus inhibit leadership by experienced individuals, has been coined ‘numerical domination’ (Huse et al. 2002a). Numerically dominant herring cohorts might not follow the older herring, simply because there are insufficient old individuals compared to abundant first-time spawners to elicit responses in the na¨ıve fishes. Reebs (2000) showed experimentally that a minority of informed golden shiners (Notemigonus crysoleucas, Cyprinidae) were able to determine foraging movement in a shoal, and in guppies, ‘observers’ were more faithful to the route to a foraging path with increasing number of ‘demonstrators’ (Laland & Williams 1997). A new model of group behaviour shows that the need to reach a target can be an important factor motivating individuals in a group differently so that a minority of the most ‘needy’ individuals become the leaders of large groups (Conradt et al. 2009; Sumpter 2009). Sticklebacks copy the patch choice of large demonstrators more than they copy the patch choice of small demonstrators (Duffy et al. 2009). In addition, a slower rate of turnover producing a more stable composition of the school results in more stable foraging traditions in guppies and southern platyfish (Xiphophorus maculatus, Poecilliidae) (Stanley et al. 2008). The probability of transmission of a novel behaviour to na¨ıve observers is generally believed to be a function of the number of demonstrators that perform the behaviour. Numerical domination, on the other hand, refers to the relative proportion of na¨ıve observers to demonstrators. The disagreement about absolute and relative numbers is partly a matter of scale; more demonstrators will be needed to transmit novel behaviour to na¨ıve fishes in a school of thousands of fishes than in a school of tens of fishes. This matter can be most readily investigated by means of model simulations. In a simulation study, Couzin et al. (2005) found that the proportion of demonstrators needed to exert effective leadership in a school decreased as group size rose, while in a similar study, Huse et al. (2002a) found the proportion to be independent of group size. More studies are needed to sort out the dependence of transmission on absolute and relative numbers of demonstrators, focussing on a range of representative group sizes, variation in motivation and sensitivity to simulation specifications. There are now strong indications that lack of leadership due to numerical domination is a key factor in the change of migration patterns of herring stocks. The wide spread of overwintering grounds of NSS herring in the past 50 years (Fig. 16.2) illustrates the great differences in the spatial location of overwintering grounds in the course of time, with the changes co-occurring with the recruitment of large year classes to the spawning stock (Huse et al. 2010, Fig. 16.3). When evaluating the realism of models on the influence of determined individuals, it is essential to take into account how experienced fishes may influence na¨ıve fishes and the degree of contact between the different categories of fishes. Within a school, smaller fishes may be more influenced by the movements of larger fishes than vice versa as observed in golden shiners (Reebs 2001). Migrating herring schools are usually elongated (Misund 1993) and experienced herring concentrated towards their preferred migration direction within the school could generate a wave of moving masses (Misund 1993) that may exert a powerful influence on na¨ıve fishes. Simulations with a line of fishes have in fact shown that determined fishes have a great influence on the rest of the school (Huse et al. 2002a).

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Behaviour, Learning, Aquaculture and Fisheries

15W

10W

5W

0

5E

10E

5f

74N

365

15E 74N

2f 2

72N

72N

70N

4f

Jan Mayen

6

Norwegian 3 Lofoten Sea Islands 3f

1f

68N

2s

6f

5

66N

70N

68N

4 66N

Iceland

1 3 3–6s 1s Møre

64N

62N

64N

Norway

62N

60N

60N

North Sea

58N

56N 15W

10W

5W

0

5E

58N

56N 10E

15E

Fig. 16.2 Distribution of wintering (red), spawning (blue, s, mainly in the Møre area) and feeding (green, f) areas of NSS herring during the last 50 years: Iceland (1, 1952–1956), Bear Island (2, 1965), Lofoten Islands (3, 1973–1986), Ofotfjorden-Tysfjord (4, 1988–1994), Vestfjorden area (5, 1995–2001), and Vester˚alen (6, 2002–2005). The numbers indicate the chronological development of wintering distributions. (Locations 1–3 were taken from Dragesund et al. (1997), 4 from Anon. (1997), 5 from Jacobsen et al. (2002), and 6 from unpublished recent IMR surveys. Modified from Huse et al. (2010)

However, in order to transmit migration routes from experienced to na¨ıve fishes, the two categories of fishes must come into functional contact. Herring usually swim in distinct schools. One difficulty for a transfer of information is that fishes of different size have different swimming speeds (Videler 1993) and preferentially swim in separate schools. Large year-classes experience strong intraspecific competition for food, resulting in slow

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

4 100

5+

Abundance ratio

3

2

4

5

6

1000 100

10

10

1

1 0.1

0.1

0.01 0.001 1950

Ratio

366

Abundance (billion individuals)

BLBK374-16

0.01 1960

1970

1980 Year

1990

2000

0.001

Fig. 16.3 Abundance of NSS herring aged 4 years (solid line) and 5 years and older (broken line), and their ratio (solid bold line), respectively (ICES 2007). The numbers refer to the locations in Fig. 16.2.

growth and small size (Toresen & Østvedt 2000). This could lead to increased size and age segregation with limited contact between young and old fishes. On the other hand, herring schools are often aggregated in clusters with short interschool distance (Mackinson et al. 1999; Haugland & Misund 2004), and attractive forces between schools may transmit migration tendencies from one school to another. In fact, schools may interact in a related manner to individuals within schools, but such interactions have been little studied to date, and they need a firmer observational basis before they can be addressed in model simulations. All the same, the available evidence suggests that herring have a potential for learning, and that na¨ıve and experienced individuals have opportunities to interact, thus permitting acquired migration patterns to be disseminated. The maintenance of spawning areas may also be based on learning and tradition in cod (Gadus morhua, Gadidae). Harden Jones (1968) refers to evidence for ‘dummy’ spawning runs by juvenile cod, and spawning grounds seem to be learned by young Newfoundland cod by following the older spawning stock on a migration ‘highway’ (Rose 1993) in a similar manner to herring. Tagging studies have revealed that cod repeatedly home to the same spawning grounds (Robichaud & Rose 2001). Like those of herring, cod traditions are therefore vulnerable to stock collapses, and there is reason to believe that local cod and tuna spawning grounds may be lost when populations are fished down (Cury & Anneville 1998).

16.2.1.3

Implications of learning for fisheries management

Spatial management and the establishment of protected marine areas demand knowledge of spatial dynamics (Babcock et al. 2005), including the role of learning. Changes in migration patterns, as described in Subsection 16.2.1.2 for herring and cod, appear to take place when the population is unstable or collapsing. While collapsed fish stocks may show signs of depensation or Allee effects (Shelton & Healey 1999), the loss of culture associated with population crashes may well hamper the rebuilding of fish stocks.

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

367

Thus, careful attention should be paid to the question of whether spatial patterns in fish stocks are governed by learning or through genetically mediated strategies, since the consequences of mismanagement might be greater if culture is lost. The northern cod stock now shows sign of rebuilding (www.dfo-mpo.gc.ca), but it is more than 15 years since the fishing moratorium was introduced (Rose & O’Driscoll 2002). The collapse of this stock was accompanied by substantial changes in its distribution (Rose et al. 2000). Given the importance of collective movements in this stock (Rose 1993), the loss of culture associated with the demise of the spawning stock may have played a role in the prolonged collapsed state of the whole stock. In fact, there are many reasons for leaving large fish in harvested populations, for example for the sake of increased fecundity, larval survival and guided learning (Birkeland & Dayton 2005). Cultural diversity in fish stocks can be regarded as one aspect of biodiversity. While horizontal cultures can adapt rapidly within generations, oblique cultures that are transmitted across generations are more sensitive to anthropogenic effects (Whitehead et al. 2004). Therefore, the question of culture needs to be integrated into the management and conservation of fish populations.

16.2.2

Fish capture

Fishing can be regarded as an arms race between the fishes and the fishermen, but it is hardly a close race. Fishes have evolved a repertoire of behavioural patterns to meet the challenges of the natural environment, while fishermen utilise these adaptations by designing fishing gear that releases these normally adaptive reactions and ‘cheat’ the fishes into behaving maladaptively during the capture process (Fern¨o 1993). Given the rapid development of technology, man is generally bound to be the winner of this race. Many fish stocks are currently under serious pressure (Christensen et al. 2003). In view of the many experiences of failure to exploit fish stocks in a way that will sustain them (Mullon et al. 2005), we may be tempted to classify human as slow learners, at least on the collective level. Fishing is often the most important cause of mortality in exploited fish populations after recruitment to the fishery (Moav et al. 1978), and over time, natural selection may result in evolutionary changes in genetically determined characteristics (Law & Grey 1989; Policansky 1993). However, many fish stocks are exploited by several types of fishing gear that change over time, resulting in conflicting and variable selection pressures. All the same, fishing has been shown to influence fish life-history parameters (Kristiansen & Sv˚asand 1998; Olsen et al. 2004; Grift et al. 2007). Few studies have examined the physiological and behavioural consequences of fisheries-induced selection, but Cooke et al. (2007) have shown that four generations of selection for vulnerability to recreational angling in largemouth bass (Micropterus salmoides, Centrarchidae) resulted in clear differences in physiological and energetic attributes and may thus select for certain personality types. Learning can modify the behaviour on a shorter time scale. A substantial number of fishes survive contact with fishing gear, creating the potential for learning (Fr´eon & Misund ¨ 1998). Ozbilgin & Glass (2004) have calculated that demersal fishes in the heavily fished North Sea may encounter towed fishing gears many times in the course of a year. If learning results in gear avoidance, the consequence could be reduced efficiency followed by increased fishing effort, with negative economic and environmental effects.

BLBK374-16

BLBK374-Brown

368

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Fig. 16.4 Schematic representation of how the gadoids cod and ling are believed to experience stimuli emitted by baits. Both species react to both visual and chemical cues, but visual stimuli are regarded as being relatively more important in ling. This is based on a more pronounced diel activity rhythm, shorter attraction distance to baits and random swimming in relation to the current. In contrast, cod have a tendency to swim perpendicular to the current, thus increasing the probability of encountering odour plumes. Ling are more stationary, hunting for mobile prey that emit visual stimuli (see Skajaa 1997; Løkkeborg 1998; Løkkeborg & Fern¨o 1999; Løkkeborg et al. 2000; Vabø et al. 2004) and have a more elongated body permitting fast acceleration. Cod have a greater inclination of the mouth facilitating chemical food search (Mattson 1990). Finally, in a classification of gadoid species in a gradient from fish feeders relying on visual stimuli to invertebrate feeders relying on chemical stimuli, based on the anatomy of the brain, Kotrschal et al. (1998) placed ling closer to piscivory.

Fishing vessels and fishing gear emit a variety of stimuli. The response of the fishes is determined by innate reactions to stimuli, learning based on innate predispositions (Kieffer & Colgan 1992) and more specific experience of the gear. To elicit a response, a stimulus must first fall within the sensory capacity of a species. Secondly, both for the initial response and for its modification after experience, what the perceived stimuli represents in the perceptible world of an animal (Fig. 16.4) is of central importance. How fishes cognitively categorise objects they encounter is critical for the response. Some stimuli may resemble stimuli from naturally occurring objects and release reactions adapted to these objects. Others may be novel in an evolutionary perspective and not release any initial response. Even such minor changes such as a change in net colour in trawls and seine nets may be experienced as more novel and release a different response. The reactions to gear reflect the natural repertoire of behaviour patterns. The different phases in the reaction to baited fishing gear are similar to the reactions to food items (see Chapter 2) that emit chemical and visual stimuli (Løkkeborg 1994). Reactions to active fishing gear may be more complex and consist of a mixture of different responses. Avoidance and freezing released by trawls (Eng˚as et al. 1998) may reflect antipredator behaviour. Fishes that keep pace with the net wall of a trawl are displaying a visually released optomotor response (Kim & Wardle 2003) that is involved in the synchronisation of individuals within schools. But the behaviour is not necessary relevant. Hedgehogs defend themselves by freezing and raising their spines. This is effective vis-`a-vis cats and dogs but maladaptive as far as road vehicles are concerned. Thus, reaction towards gear can seldom be explained as a clear response to a single stimulus. First, novel objects may activate several motivational systems resulting in complex responses. Secondly, fishes usually encounter several stimuli, either simultaneously or in sequence. Thirdly, there are presumably often emergent properties in the response as a

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

369

result of stimuli-reactions interactions. For instance, responses to gear-induced stimuli acting in concert with stimuli from neighbouring fishes may lead to unforeseen events via self-organisation (Camazine et al. 2001). In this way individual fishes might lose control to the group and be unable to make decisions based on their previous experience. We have, for example, observed the first fish in a school of saithe (Pollachius virens, Gadidae) swim into a net, get entangled and show panic swimming. This was seemingly experienced by the following fish as an escape reaction from something behind, which led the rest of the fish to rapidly swim into the net. We can identify five stages at which learning can have an effect in fish capture: 16.2.2.1

Natural variations in spatial distribution and behaviour

A variable migration pattern influenced by learning (see Subsection 16.2.1) could make it difficult to locate fishable concentrations. The level of activity and spatial use can change when there are shifts in learned migration routes and influence the probability of encountering static gear (Eng˚as & Løkkeborg 1994). On a smaller scale, the spatial dynamics within a home range influenced by learned spatial cues (Reese 1989) could have an effect on the probability of capture. A simulation study based on in situ movements of cod and ling in fjords showed that there was a low risk of encountering a fleet of gillnets set in a fixed area, and explained this finding by the limited search range of the fish (Eng˚as & Jørgensen 1997). 16.2.2.2

Avoidance and attraction before fishing

Fishes seem to be able to learn about risky habitats from their own experience or from the reactions of other fishes (Chivers & Smith 1995; Brown 2003). At present we have little evidence that fishes avoid heavily fished areas, because they lack relevant stimuli for avoidance. Na¨ıve brown trout (S. trutta, Salmonidae) avoided a section of a river where other individuals had been hooked the previous day (Young & Hayes 2004), and attraction of dolphins to a fishing area induced different avoidance reactions in different species of fishes (Rocklin et al. 2009). Chemosensory cues are generally more reliable than visual cues in this context (see Chapter 4), and chemical stimuli from fishing gear could provide information about the local predation (fishing) risk. The release of alarm substance by injured fish may reduce the number of fish in an area and enhance the response to visual indicators of predation risk (Wisenden et al. 2004). Auditory stimuli and turbidity (Humborstad et al. 2005) generated by trawling activity may also provide fish with reliable information. Interestingly, high concentrations of fishes have been observed in recently trawled areas, where fishing vessels have been observed to re-fish the same tow route, taking advantage of immigrating scavenging fishes (Kaiser & Spencer 1994). This illustrates that even a trawl may reward the fish and possibly attract fish during subsequent encounters. 16.2.2.3

Before physical contact with the gear

The search image of the fishes and degree of novelty of the perceived stimuli can be crucial. Bait attraction may be influenced by the olfactory and visual stimuli emitted by the prey organisms present. A possible example is the low efficiency experienced in the longline

BLBK374-16

BLBK374-Brown

370

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Reward (Food)

Cognition of chemical stimuli

Cognition of visual stimuli

Cognition of taste and texture Punishment (Pain inflicted by the hook)

New approach

Dynamic evaluation of costs and benefits

Swim away

Fig. 16.5 A fish approaching a baited hooks successively experiences and cognitively categorises different stimuli and objects. After swimming away from the gear, a fish seems to process and integrate information from its experiences, which can result in a new approach.

fishery for cod feeding on capelin in the Barents Sea, where cod appear to react to visual stimuli from moving prey rather than to chemical stimuli from baits. Pots that have been set close to structures and which, therefore, differ less from their surroundings than pots further away are more efficient (High & Beardsley 1970) and are presumably perceived by fishes as less novel being a part of the complex environment. Artificial baits are reported to be more effective for large cod than for small ones, which suggests a learned preference (Løkkeborg 1990). Small cod may experience the artificial bait as being more novel, as large cod have a more extensive diet (P´alsson 1994) and thus more experience of various prey organisms. Cod can react to the olfactory plume from a baited longline from a distance of 700 m (Løkkeborg 1998) and must then cognitively categorise the chemical stimuli emitted from the gear (Fig. 16.5). At shorter distances fishes also categorise the gear on the basis of its visual image (Fig. 16.5), which challenges them to adjust the categorisation of the object based on the additional stimuli it perceives. ‘Does it look like it smells?’ Socially transmitted learning may also lead to avoidance. Copying fishes may associate the response of others with the sound of trawler engines and thus acquire a modified response by observational conditioning (see Chapter 11). In the laboratory, fishes learn from experienced demonstrators the escape route from a model trawl apparatus (Brown & Laland 2002a). Fishes in schools display learned avoidance of a model trawl, whereas pairs of fishes show no evidence of learning (Brown & Warburton 1999b). Tropical sardines show conditioned avoidance reactions in the laboratory, with transmission of reactions from conditioned to na¨ıve fishes (Soria et al. 1993). On the other hand, attacks on bait by other fishes and movements of hooked fishes and nearby baits seem to stimulate attacks in cod, haddock (Melanogrammus aeglefinus, Gadidae) and whiting (G. merlangus, Gadidae) (Fern¨o et al. 1986; Løkkeborg et al. 1989). This partly explains the patchy distribution of hooked fishes along a longline (Sigler 2000). Struggling fishes do not appear to trigger fear responses, and gadoid fishes thus do not seem to have any preprogrammed or learnt response to constrained fishes – a situation that they presumably never encounter under natural conditions. Species reacting to chemical

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

371

alarm cues released by injured fishes (Mathis et al. 1996) may behave differently. Entrapped fishes could socially reinforce fish approaching a pot. Munro et al. (1971) observed that when some grunts (H. plumieri, Haemulidae) had entered a pot, the daily ingress of fishes rose sharply. 16.2.2.4

After physical contact with the gear

Fishes coming into physical contact with the gear encounter new stimuli that must be cognitively integrated with the previously perceived stimuli. Possible emergent properties of this integration may create irrational reactions if the stimuli combination is not a product of selection past. The different stimuli may also generate conflicting tendencies, such as when fishes that come into physical contact with baited hooks encounter taste and texture stimuli from the bait and pain stimuli from the point of the hook (Fig. 16.5). Response modification and learning after contact with gear may occur on various timescales. Within a single encounter with a gear, fishes may modify their response based on events that took place seconds or minutes previously. The modification can be caused either by a change in motivational state, such as fright, or by learning. During classical (Pavlovian) conditioning (Lieberman 2000) fishes usually receive negative reinforcement from the gear. An example of classical fear conditioning may be fishes that learn to associate the sound of a trawl with later physical contact with the gear. Reduced catchability by trawls during repeated hauls has been demonstrated in bream (Abramis brama, Cyprinidae) and explained by trawl avoidance observed in acoustically tagged fishes caught by trawl (Pyanov 1993). Avoidance learning has an element of operant learning (an association between a behavioural action and its outcome): to make a certain response in order to avoid a fishing gear, such as diving below the bobbins of a trawl or swimming over the ropes of a Danish seine. Fishes can learn to penetrate the meshes in a net and in this way perhaps ¨ learn to swim through trawl nets (Ozbilgin & Glass 2004). Operant conditioning may also take place when fishes attack baited hooks or find their way through the funnel of a pot and escape. Fishes seldom receive positive reinforcement apart from escaping the gear, but haddock are known to ‘steal’ baits from long lines and thus be rewarded by bait attacks. The behaviour resulting from learning can vary on the individual level. After biting a baited hook in the laboratory cod response intensity falls dramatically, irrespective of whether the fish was hooked or not (Fern¨o & Huse 1983, Fig. 16.6). However, whereas some cod displayed only a few intensive attacks several days apart, other individuals after some initial strong responses displayed several hundred low-intensity responses (approaches and tastings without biting). Previous experience, dominance rank or coping style (Chapter 7) may explain the variation. There are individual differences in the way in which fishes respond to novelty (Brown et al. 2005), and wide individual variations in behaviour have also been observed in fishes learning to avoid toxic prey (Crossland 2001). The sharply decreased response intensity and ‘neurotic’ behaviour of some cod that made a long series of approaches without bait contact indicate an active trade-off between the benefit of food and the cost of hook contact and strongly suggest that fishes experience a baited hook as aversive and painful – a possibility that has been much debated (Rose 2002, 2007; Sneddon 2003a; Huntingford et al. 2006, 2007; Arlinghaus et al. 2007). Individual variations in catchability have also been found in largemouth bass (Katano 2009).

BLBK374-16

BLBK374-Brown

372

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Fig. 16.6 The effect of experience on the behaviour of cod vis-`a-vis a baited hook. The response strength is calculated as the ratio between the number of strong (complete bite, jerk, shake, pull, chew, rush) and weak (approach, taste, incomplete bite) responses on each day of the experiments (see Fern¨o & Huse 1983). Group 1 (solid line, 15 fishes) were caught by trawl and kept in the laboratory for 3 years and subsequently subjected to several bait bag preference tests, while group 2 (broken line, 20 fishes) were caught in traps and kept in the laboratory for 2 months prior to testing. The total number of responses per day ranged from 22 to 178.

16.2.2.5

Behaviour after escaping the gear and long-term consequences

Fishes that escape from fishing gear may suffer swimming impairment and behavioural deficits that subject them to elevated predation risk and reduced feeding behaviour (Ryer 2004). The degree to which learning influences later encounters depends on how fishes experience the reward or punishment from the gear and the number of encounters needed to establish an association. A combination of irrelevant signals and high stress level during the process may also reduce learning ability. How long the modification persists may depend on the total reinforcement schedule in the variable situation in the sea, where fishes are exposed to a multitude of naturally occurring stimuli in addition to vessels and gear. Learning and memory can be affected in different ways by contrasting ecological factors (Brydges et al. 2007) with the memory window, for instance, adapted to the stability of prey in different habitats (Mackney & Hughes 1995). In some cases the memory window can be long, but as the ability to forget in a dynamic environment may be as important as the ability to learn (Kraemer & Golding 1997), infrequent encounters or reinforcement can lead to forgetting or extinction of a learned response. A question also arises to what extent the fish has the ability of generalisation (Lieberman 2000) and to associate similar cues or sensory inputs as the conditioned signal associated with a certain type of danger. In fishes, learning about predators usually occurs after just one simultaneous presentation of the cue and the stimulus (Magurran 1989) and the response can be retained for several months (Chivers & Smith 1994; Brown & Warburton 1999a). Fishes seem to be able to learn to avoid trawls in the course of one-trial learning (Pyanov 1993). During fishing experiments with hook and lines and in catch-and-release fishing, lower catchability during the coarse of fishing has been observed in many species (Beukema & de Vos 1974; Hackney & Linkous 1978; O’Grady & Huges 1980; Yoneyama et al. 1996; Young & Hayes 2004; Askey et al. 2006), and trout are easier to catch in rivers with low-fishing pressure than in more heavily fished rivers (Young & Hayes 2004). However, Tsuboi & Morita (2004) found that whitespot char (Salvelinus leucomaenis, Salmonidae) that had been hooked and released were

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

373

more likely to be caught than previously uncaught fishes, perhaps since they were initially the least risk-averse individuals. Cod remember the association between a stimulus and food for at least 3 months (Nilsson et al. 2008a). One single hooking experience made carp (Cyprinus carpio, Cyprinidae) more difficult to catch for at least a year (Beukema 1970), and learned avoidance of a trawl apparatus in the laboratory persisted for at least 11 months (Brown 2001). We assume that learning and subsequent modification of the behaviour take place at the very moment a fish experiences something. However, it appears that the full evaluation of the positive and negative experiences from a contact with a gear takes some time while the fish cognitively processes the information with the final decision possessing emergent properties. In aggressive contests, individuals seem to process and integrate information from different experiences (Hsu & Wolf 2001; Chapter 6). During observations of the behaviour of fishes vis-`a-vis baited test lines, we often observe cod entering and leaving the observation field at intervals of some minutes. These individuals are presumably in a conflict situation between attraction and avoidance to the gear making a dynamic evaluation of the costs and benefits of responding that may result in repeated visits (Fig. 16.5). A clearcut example of this is reported for cod in fjords (Kallayil et al. 2003, Fig. 16.7), where several cod were observed swimming in loops and, after a sudden turn, return to nets baited with large bait bags from distances up to 400 m. The cod seemed to initially evaluate the bait bags as too large to swallow and thus swam away from the net. It could be argued that the fishes forgot the situation and were simply attracted again, an emergent property arisen from reactions to combined stimuli that are evolutionarily irrelevant. However, as the cod returned from different directions in relation to the current, they were not merely repeatedly

Cod 3

Cod 2

Cod 1

600

Distance (m)

BLBK374-16

2

400

18 12 14317 16 6 7 15 5 8

200

16 4 17121 15318

13 114 1 12 11 3

5

6

14

10

13 9

4

14

712

58

13 7

11 10

8 9

6

12 9 11

0

10

0

200

400

Distance (m)

600

0

200

400

Distance (m)

600

0

200

400

600

Distance (m)

Fig. 16.7 Cod moving in loops relative to a net baited with large bait bags too large to swallow. Dark bands indicate baited nets. Direction of fish movements is indicated by progressing serial numbers. (Modified from Kallayil et al. 2003.)

BLBK374-16

BLBK374-Brown

374

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

attracted by the odour plume, but presumably returned to the food source utilising a spatial map of their home range after ‘thinking it over’. An interesting question is to what extent fishes can learn from experience based on the outcome of interactions with gear. In cod, 2 minutes between the stimulus and the reward seem to be the maximum time that permits an association to be formed (Nilsson et al. 2008a). A fish that reacted to the odour plume from a baited pot is not initially punished. If the fish enter the funnel it will encounter a reward (the bait) and may not experience the confinement as stressing until some time later. It is not certain whether a fish that manage to escape can associate the eventual negative reinforcement of confinement with their initial approach to the pot and entering the funnel. Fishermen have reported an easily recognisable blind cod found in pots over and over again even though it was repeatedly caught and handled, and we have also caught the same tagged cod six times in a fish trap (Kristiansen, personal observations). Similarly, the complex reaction of schooling fish to a moving trawl with the fishes exposed to multiple stimuli from the gear and school mates (Wardle 1993) may create emergent properties that make the fish go through a chain of events that make it difficult for them to associate their initial reaction to the herding ropes with the subsequent negative experience of the trawl.

16.2.3

Abundance estimation

Learning influences spatial dynamics and can thereby affect abundance estimates of fish stocks. Surveys should cover areas with high fish concentrations, and shifts in migration patterns present a challenge to survey design. The pronounced abrupt changes in the spatial pattern in herring could have severe consequences for surveys, which need to cover the entire distribution of the stock. Variations in fish avoidance of survey vessels and gear create uncontrolled biases in abundance estimates (Fr´eon & Misund 1998). Horizontal avoidance of the vessel moves the fishes out of the acoustic beam. Vertical escape influences the tilt angle and reduces the volume of the swimbladder, and thereby decreases the horizontal projection area and echo level. Modification of the response to fishing gear will result in biases when catch per unit effort (CPUE) is employed as an index of abundance (Fr´eon & Misund 1998). Small fishes can learn to penetrate meshes and in this way learn to escape through nets in trawls ¨ (Ozbilgin & Glass 2004) and this can influence the selective properties of trawls in heavily fished areas with multiple gear contacts. In order to actually understand and predict the behaviour of fishes with respect to manmade constructions, it is vital to know the biological basis of the response (Fern¨o 1993). Vessel avoidance reactions presumably reflect predator avoidance and are determined by a balance between the costs and benefits of responding. Previous experience should be expected to influence the estimate of risk. Herring in net pens are reported to habituate to playback of vessel sounds (Schwarz & Greer 1984), and repeated exposures to a passing vessel seem to result in response waning (Vabø et al. 2002). Fishes may habituate to vessel sounds in areas with heavy boat traffic as in the North Sea, although fishes in this area ¨ experience vessel sound in combination with fishing gear relatively often (Ozbilgin & Glass 2004), a circumstance that should counteract habituation. It would be interesting to find out whether there is a correlation between boat traffic and vessel avoidance in different

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

375

areas, for instance, based on transects from areas with heavy boat traffic close to the coast to more offshore areas.

16.3

Aquaculture

The farming environment ranges from semi-natural ponds to high-technology, intensive, recirculation farms. While pond-farmed fishes experience near-natural environments, fishes reared in tanks and cages are kept in an environment very different from the natural habitats to which they are evolutionarily adapted. Intensively farmed fishes must be able to adapt to high densities, restricted space, and artificial and uniform food and not least to frequent disturbances and handling by man. The rearing environment needs to be within the species’ physiological range of tolerance, but the fish must also cognitively process the sensory information presented by the farming systems. Fish species selected for intensive aquaculture need flexible behaviour to be able to adapt. Considering only biological performance, the best aquaculture candidate species, therefore, should be social, generalist species with broad niches and large environmental tolerance intervals, such as Tilapia spp. However, high market prices for more demanding species can also make it profitable to tailor the rearing environment to the species’ preferences (e.g. Atlantic halibut, Hippoglossus hippoglossus, Pleuronectidae). Most aquaculture environments are structurally very simple and the challenges for the fish seem to be equally simple, with food easy to catch and an absence of predators. However, high stocking densities with frequent social interactions, suboptimal environmental conditions, very limited choice of habitat and food and abundant, noisy and unpredictable sensory stimuli, may make the environment cognitively demanding with few learning opportunities and a high stress level that may even influence the quality of fillet (Olsen et al. 2008). One way to reduce cognitive stress in these environments would be to hand over individual control and decision-making to the group. Information about rewards (food) and punishments (handling, social aggression) may be received indirectly via the behaviour of other fishes through social learning, and the behaviour of individuals will trigger selforganising group behaviour that may be either adaptive or maladaptive.

16.3.1

Ontogeny

Most fishes are only a few millimetres long at hatching and their weight grows by several orders of magnitude before they reach maturity. They pass through several developmental stages in different environmental niches and habitats with different predators, prey and adaptive demands (Balon 1984; Fuiman 1994). The tiny larvae have few resources to allocate to brain development and learning, and must mostly rely on preprogrammed behaviour. For instance, halibut larvae that approach the feeding stage are positively phototactic, which in nature leads them towards the prey-rich surface layers of the sea (Naas & Mangor Jensen 1990). However, in tanks such phototactic behaviour results in halibut larvae that are butting against light tank walls and the surface, as they are trapped in their preprogrammed behaviour without the capacity to learn to adapt this behaviour to the novel environment (Naas & Mangor Jensen 1990). Learning ability may be better suited to associations that

BLBK374-16

BLBK374-Brown

376

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

have a strong survival advantage. For example, most fish larvae rapidly develop clear prey preferences and improved prey-catching skills (Checkley 1982; Cox & Pankhurst 2000). Taste preferences are usually genetically controlled (Kasumyan & Døving 2003), but the association between the taste and visual characteristics of the prey and improved preycatching skills ought to be a learned process with high survival value (Chapter 2). The ability to discriminate between dangerous and harmless objects may also be learned. While adults of the damselfish, Dascyllus marginatus (Pomacentridae), are able to discriminate between predatory and non-predatory fish, juveniles lack this ability (Karplus et al. 2006). It is crucial that behaviour promoting growth and welfare is established as early as possible, otherwise a suboptimal culture may result. Group behaviour often differs between identical rearing units, with fishes in one tank, for example, being consistently more fearful than those in other tanks. Mixing fish between tanks or the removal of aggressive individuals may be able to turn a ‘bad tank culture’ into a ‘good’ one. However, a ‘bad culture’ may also occur when fishes from two ‘good cultures’ with different experience and habits are mixed (Fern¨o et al. 1988; Juell 1995, see Fig. 16.8). Further studies on how early experience influences the development of the behaviour in farmed fishes are needed.

16.3.2

Habituation, conditioning and anticipation

Farmed fishes are exposed to many sudden stimuli that elicit the same behavioural reactions (reflexes and fixed action patterns). These reactions can also trigger a physiological stress response (Conte 2004). An example is the startle reflex caused by sudden noise or approaching objects. If the stimulus is not associated with harmful events (e.g. people passing the tanks), repeated exposure results in response waning – a mechanism known as ‘habituation’ (Lieberman 2000). Rapid habituation should be a preferable trait in farmed fishes. However, in most species certain stimuli, such as silhouettes of bird or fish predators and approaching objects, release strong predisposed responses with slow habituation. The habituation rate may be reduced further if a stimulus that is usually harmless is occasionally followed by an aversive event, or resembles other stimuli that are associated with aversive events. Classical (Pavlovian) conditioning occurs when two events are associated in time so that an originally neutral stimulus can be associated with an aversive or rewarding stimulus (Lieberman 2000). For a stimulus that induces predisposed fear responses, such as a large moving object, learned associations between that stimulus and an aversive event (e.g. dip net and confinement) would strengthen the fear response. A complication for habituation to harmless stimuli is that they are often similar to stimuli associated with fearful events. For instance, removal of dead fish and cleaning tanks and netting all involve a moving object that suddenly appears. It may be difficult for the fishes to discriminate between such stimuli, and the effect of generalisation (Lieberman 2000) may override the effect of habituation. As long as similar stimuli are associated both with neutral and negative consequences, the average outcome is likely to be negative. Rewarding harmless stimuli, like feeding with a preferred food during cleaning, could balance this equation and reduce the stress level. Chinook salmon (Oncorhynchus tshawytscha, Salmonidae) may be positively conditioned to the emptying of the tank until the water just covers the fish, with a lowered physiological stress response to subsequent handling and transport stressors (Schreck et al. 1995). Not only can the stress level be reduced when the stressor is paired with a reward,

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

377

Goal not achieved Stereotypy

Goal not achieved

Goal achieved

Stressor Reactive fish state Past experience

Food

Classical conditioning CS-US Trace conditioning CS US

Proactive fish exploring

INDIVIDUAL LEVEL

Self-organisation

on

Stimulus

Reacti

Choice 1 Remain in group

re L ind oss ivid of ua lity

rig

g

in a no l gro t e up xi s st tr in uc g

tu

in py

Co

Choice 2 Leave group

O

BLBK374-16

GROUP LEVEL Fig. 16.8 Individual decisions and learning in a rearing environment of fishes with different coping styles (above) and consequences of copying and self-organisation on collective behaviour (below). Maladaptive behaviour with loss of control and chronic stereotyped behaviours may develop on the individual level if the goal of the behaviour is not achieved. An individual that joins a group of fishes that copies and mixes with another group with a subsequent change in group structure by self-organisation may be trapped in a maladaptive group structure.

BLBK374-16

BLBK374-Brown

378

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

but the previous stressor may even induce anticipatory behaviour. Groups of cod that were rewarded with food after each exposure to a dip net moving in the tank developed a strong anticipatory behaviour after net exposure and gathered below the dip net when it stopped moving and then swam to the feeding area (Nilsson et al. unpublished data). Similarly, an initial escape response of salmon post-smolts to a flashing light in the feeding area was replaced with an approach response when the flashes were followed by food (Bratland et al. 2010). The transition from fear to anticipation occurred gradually from trial to trial, with a reduced startle response after the light blinks and increasing anticipatory behaviour and approach to the feeding area before food delivery. High levels of stress can lead to a shift of attention and impaired learning (Olla et al. 1998), and reducing stress by reward conditioning may be assumed to be less effective for severe stressors such as netting and air exposure. Other methods of stress control of farmed fishes that involve learning have been reviewed by Lines & Frost (1999). Reward conditioning often occurs automatically in connection with feeding. Fishes learn, for example, to associate the footsteps of the farmer or the sound of pellets in the feeding pipes with food and may display strong anticipatory behaviour. However, low or even contradictory predictability of such ‘natural’ stimuli may involve uncertainty, e.g. the farmer’s footsteps may sometimes mean handling instead of food. The uncertainty related to reward or punishment associated with husbandry routines may increase stress levels. Therefore, fish farmers should make their routines predictable and associated with positive events. One way of doing so would be to clearly signal what is coming, for example, by sounding a tone or turning on a light. This strategy was used by Bakken et al. (1993) in fox farm by using differently coloured clothes during handling and during feeding.

16.3.3

Pavlovian learning – delay and trace conditioning

The fundamental capacities and constraints of learning (the cognitive tool box) set the limits on what tasks fishes are capable of solving. A learning task is usually complex and involves a number of variables and events, but to what extent the fishes are able to solve the task may be limited by a single factor, for instance how long a time-gap between two events the fishes are able to bridge or how spatially close events must be. Fishes have excellent delay conditioning ability (Bull 1928; Overmier & Hollis 1990). In this standard procedure of classical conditioning, there is an overlap in time between the conditioned stimulus (CS) and the unconditioned stimulus (US) (Fig. 16.9). This is a primitive type of learning that

DELAY PROCEDURE

TRACE PROCEDURE

CS US

CS

Trace interval

US

Time Fig. 16.9 The two main procedures employed in Pavlovian conditioning. In delay conditioning, the CS and the US are temporally adjacent or overlapping. In trace conditioning, the offset of the CS and the onset of the US are separated by a stimulus-free trace interval.

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

379

is also found in invertebrates. The offset of the CS and the onset of the US can also be separated by a space in time – a paradigm called trace conditioning (Fig. 16.9). The CS must then leave some ‘trace’ in the nervous system. Trace conditioning can be crucial in a rearing unit to enable fish to associate events separated in time (like self-feeding). The ability of trace conditioning was previously little investigated in fishes, but both cod and halibut have recently been shown to associate a light flash with food separated by period of up to 2 minutes (Nilsson et al. 2008a, 2010). This trace length is impressive even for mammals (Lieberman 2000). How fishes respond to a CS announcing a food reward is dependent on a range of factors, including the temporal and spatial relationship between the CS and the reward, social interactions and the species. The response usually reflects the natural behaviour when expecting a prey. For instance, archer fish (Toxotes chatareus, Toxotidae) that naturally forage by squirting water at insects above the surface, respond to a CS light above the surface by squirting at it when it is paired with fruit flies delivered on the surface (Waxman & McCleave 1978). The reef damselfish, Pomacentrus amboinensis (Pomacentridae), that lives in a complex environment easily learn to discriminate between a pattern associated with food and similar patterns unrelated to food and is able to dissociate the location of the signal pattern from the location of food delivery (goal-tracking, Boakes 1977) (Siebeck et al. 2009). Cruising predators that apply a hunting strategy that involves approach and pursuit of the prey upon detection would usually benefit from rapid reflex-like approach responses to stimuli associated with prey. For instance, cod respond by immediately approaching the CS (sign-tracking, Hearst & Jenkins 1974) regardless of whether they are delay or trace conditioned or whether the CS is located in the feeding area or not (Figs. 16.10 and 16.11, Nilsson et al. 2008a, 2008b). Species with other hunting strategies, such as sit-and-wait, may take advantage of more flexible anticipatory responses and postpone the attack until the prey is at a suitable distance for an ambush attack. Halibut that apply a sit-and-wait strategy do not always approach the signal (Nilsson et al. 2010, Fig. 16.11). Delay-conditioned halibut (expecting food very soon) respond immediately at the onset of the CS and approach the site of food delivery, an attack-like behaviour. In contrast, trace-conditioned halibut (expecting food after a while) respond with cautious movements near the bottom (at the longest trace intervals so subtle that they were not detectable by the naked eye) as if preparing for a future attack (Figs. 16.10 and 16.11, Nilsson et al. 2010). A lesson to learn here is that the degree to which fishes clearly express what they have learned does not necessarily reflect how much they have learned (‘behavioural silence’). The differences in anticipatory response in cod and halibut resemble the differences between the rat that applies a ‘search behaviour’ and the domestic cat that is a ‘sit-and-wait’ predator. In the same appetitive conditioning procedure, rats become hyperactive while cats become hypoactive, with the different anticipatory responses reflecting their foraging strategies (van den Bos et al. 2003).

16.3.4

Potential use of reward conditioning in aquaculture

In planned reward conditioning procedures, light or sound signals are used to trigger anticipatory behaviour. Signalling may advertise not only what is going to happen, but also where, and reward conditioning can be used to lead fishes to a feeding area (Midling

BLBK374-16

BLBK374-Brown

380

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

(A)

(a)

(B) B

(b) b

(C)

(c)

Fig. 16.10 Anticipatory behaviour of two fish species with different hunting strategies. Cod are distributed throughout the tank before CS presentation (A) and gather in the CS/US area when the CS is switched on (a). The responses in trace and delay procedures are similar, with cod approaching the CS immediately at onset. The picture is from a 60-s trace procedure. Halibut lying motionless on the tank floor before CS presentation in a delay conditioning procedure (B) swim towards the feeding area at the surface when the CS is switched on (b). Trace-conditioned halibut respond differently. The fish lying on the tank floor before CS presentation (C) reposition themselves at the floor after the CS is switched on (c), but do not swim towards the surface. The trace interval here was 60 seconds. Photos a, b and c were taken 10 seconds after the onset of the CS. (Adapted from Nilsson 2008.)

et al. 1987). When transferred from tanks to sea cages, fishes are introduced to a very different environment with much larger volume and depth, different light conditions and different feeding procedures. Many individuals fail to cope with the new environment and end up as ‘losers’ with poor growth and welfare. Training fishes to associate a localised CS, for example light flashes, with food while still in their tanks and using a similar CS as a temporal and spatial guide in the cage environment would increase predictability and possibly facilitate feeding after transfer to cages. Species differences should be taken into account when designing the signalling system, i.e. which types of CSs are most suitable and

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

381

75 50 25 CS

0 0

10

20 Time (s)

100

30

40

6

CS area Repositions

75

4 50 2

25 CS

0 0

10

20 Time (s)

30

40

Repositions per 5 s

Percent fish in CS area

100

Percent fish in CS area

BLBK374-16

0

Fig. 16.11 Response of groups of cod (upper panel) and halibut (lower panel) to a food-announcing CS in trace conditioning with a 20-s trace interval. For the cod, the feeding area and CS area (25% of the tank) were on opposite sides of a 3-m tank, with CS and food delivery at the surface. For the halibut, the CS was located on the tank floor with the CS area representing 25% of the floor, and food spread out on the surface at delivery. Cod were actively swimming throughout the interval from the onset of the CS to food delivery, first gathering in the CS area (sign-tracking) and then moving to the feeding area (goal-tracking). In contrast, trace-conditioned halibut neither sign-tracked nor goal-tracked, but responded with cautious repositions at the tank floor. The differences in anticipatory response of cod and halibut reflect their different hunting strategies, with cod searching actively for food and pursue prey before capture and halibut applying a ‘sit-and-wait’ strategy. The foraging strategies of cod and halibut resemble those of the rat and the cat respectively, in which similar differences have been found (van den Bos et al. 2003). (Based on data from Nilsson et al. 2008b; Nilsson et al. 2010.)

how the temporal relationship between the CS and food delivery may influence anticipatory behaviour (Nilsson et al. 2010). Most learning studies with fishes have used individual subjects or relatively small groups, and little is known about learning and anticipatory responses in larger groups and larger arenas, such as in sea cages. On the one hand, individual learning can be improved by the presence of other fishes, for example by social learning (Chapter 11), but socially learned information may be maladaptive (Laland & Williams 1998), or a majority of lesscompetent individuals may mislead better learners (Nilsson et al. 2008b). Fishes may not have an overview of the entire arena and the large and monotonous environment may restrict spatial learning. Moreover, social interactions among the large numbers of fishes may lead to emergent properties where the anticipatory response is modified or absent (see Subsection 16.3.6). For instance, Atlantic salmon were successfully trained to associate and

BLBK374-16

BLBK374-Brown

382

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

respond to blinking lights in tanks with a group of 550 individuals (Bratland et al. 2010), but no evident anticipatory responses were observed in groups of 3000 salmon in a large deep sea cage, possibly due to the unwillingness of individual fishes to break out of the school (O. Folkedal, Institute of Marine Research, Bergen, unpublished data; conformity effect, Day et al. 2001). More knowledge of learning and group dynamics under full-scale farm conditions is needed to permit efficient use of learning programmes in the aquaculture industry.

16.3.5

Operant learning

Exploratory behaviour is crucial to obtain information about food resources and hazards in a constantly changing environment (Inglis et al. 2001), and fishes in a tank or cage are often active in the absence of external stimulation. On reward, the probability that the behaviour will be repeated gradually increases resulting in operant conditioning. In fish farming this way of learning is used in various types of self-feeding (demand-feeding) technology where the exploratory behaviour starts a feeding automate (Jobling et al. 2001). A trigger consisting of a rod to be pushed or a string to be pulled is usually involved (Alan¨ar¨a 1996; Rubio et al. 2004). Many farmed fish species can learn to operate self-feeders (Divanach et al. 1986, 1993; Alan¨ar¨a 1996; Sunuma et al. 2007), but most experiments have focused on feeding activity rather than learning ability and the various aspects of learning rate and effects of reward value have not been described in any detail. The time different species need to learn to operate the self-feeder is claimed to vary from 10 to 45 days (Jobling et al. 2001), but the learning situation in a group is complex, with several learning aspects operating together and social interactions influencing the degree to which fishes express what they have learned. In contrast to the more active operant training of dolphins (Pryor 1975), self-feeding fishes are not rewarded for sub-goals. Thus, individuals must explore the trigger in a manner that leads to food delivery, for example push a rod or pull a string. The fishes soon learn that there is an association between the trigger and food, but since the fishes are performing many different exploratory behaviours like unrewarded ‘nosing’ on the trigger and the gap in time between the trigger is released and the food is seen in the water, it may be difficult to learn exactly which behaviours are actually rewarded. Since many fishes may be rewarded by the action of a single fish, while the fish that actually operates the trigger may not be rewarded, the behaviour–outcome relationship is not straightforward. A further complicating factor is that relatively few fishes perform most operations (Alan¨ar¨a 1996; G´elineau et al. 1998; Cov`es et al. 2006; Millot et al. 2008). Still, low-trigging and even non-trigging individuals may have learned an association between triggering and food delivery. A high-triggering individual may abruptly become a non-trigger or low-trigger and a new fish take the position as high-trigger (Millot & B´egout 2009). In a self-feeding study with cod, we frequently observed that when one individual attempted to pull the trigger other fishes immediately approached the feeding area even if the trigging attempt failed (own observations). One could assume that high-triggers are large and dominant individuals with the ability to monopolise the self-feeder and that they would have higher growth rate. This may be the case in some species, for example Arctic charr (S. alpinus, Salmonidae) and rainbow trout (O. mykiss, Salmonidae) (Alan¨ar¨a & Br¨ann¨as 1997). However, in other species such as sea bass (Dicentrarchus labrax, Moronidae) there

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

383

is no initial tendency for the high-triggers to be larger individuals, and they do not exhibit higher growth rate than low- or non-triggers (Cov`es et al. 2006; Millot et al. 2008). In some situations the operation of the trigger may also be rewarding in its own right and release ‘false triggering’, leading to accidental feeding (Rubio et al. 2004). Many fishes develop clear circadian rhythms in self-feeding activity, with one or more daily peaks (Chen et al. 2002, 2007; Amano et al. 2006; Millot & B´egout 2009). In sea bass, the rhythm of the most-triggering fish has been found to determine the rhythm of the other individuals of the group (Millot & B´egout 2009). Most fish farmers have observed that fishes have an impressive ability to anticipate scheduled feeding times and that they must have some kind of an internal circadian clock (time–place learning, Reebs 1996, 2000). This is clearly demonstrated by sea bass fed by time-restricted self-feeders (Azzaydi et al. 1998; Sanchez-V´asques & Madrid 2001).

16.3.6

Individual decisions and collective behaviour

The behaviour of other fishes is a powerful stimulus for group-living fishes, and copying behaviour has both inherent (preprogrammed) and learned components (Kieffer & Colgan 1992; Chapter 11). In fish tanks and cages, access to food is often unpredictable in time and space and visibility may be low due to turbidity and high fish density. Therefore, the behaviour of other fishes is an important source of information, and learning by a combination of classical and operant conditioning occurs when fishes learn to associate the behaviour of their shoalmates with rewards or aversive events. Understanding the consequences of the behavioural decisions of an individual fish in a rearing unit is a challenge, particularly since social interactions and emergent school structures will influence the end result with regard to distribution and group behaviour. There is presumably a complicated intercalation of rigid and flexible components when fish encounter a situation, respond to stimuli and subsequently enter a new situation that releases a new stimulus-behaviour complex. If the sensory input is too cognitively complex, a shift from a high-level ‘off-line’ cognitive control of individual choices to low-level ‘online’ direct stimulus-response controlled schooling behaviour may occur (Toates 2004). However, if the actions of an individual fish are not rewarded, a ‘loss of control’ situation may occur and the fish becomes a ‘loser fish’ with a ‘learned helplessness’ showing little response to food and other fish. Like mammals, fishes have personalities or behavioural syndromes (a suite of correlated behaviours expressed across different contexts) such as bold and timid individuals (Magurran 1993; Sneddon 2003b; Brown et al. 2005; Chapter 7) and proactive and reactive stress coping styles (Koolhaas et al. 1999, 2007; Kristiansen & Fern¨o 2007; Øverli et al. 2007, Fig. 16.8). We should assume that fishes with different coping styles have different learning abilities and motivations that should be characterised by different appraisal of rewards, anticipatory behaviour and speed of learning. However, very few studies exist in this field. Moreira et al. (2004) showed that fishes from a low cortisol responsive line had slower extinction rates of the conditioned response to a stressful stimulus, while Dugatkin & Alfieri (2003) observed a positive relationship between male boldness and a simple associative learning task in guppy.

BLBK374-16

BLBK374-Brown

384

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Group decisions can reduce aggression and facilitate food localisation, but can also have negative consequences. High densities can result in loss of individual cognitive control with the fishes switching from individual to school behaviour rules, leading to emergent school structures and self-organisation (Camazine et al. 2001; Couzin et al. 2005; Chapter 10). In individual-based models, small changes in individual decision rules have a powerful impact on emergent school behaviour (Nøttestad et al. 2004). In an environment in which fishes are primarily influenced by social stimulation and food, copying other fish should strongly influence culturally mediated group behaviour, with various and unpredictable outcomes in species that are not adapted to a schooling lifestyle. In a way, farmed fishes are at the mercy of copying and self-organisation. When a group structure has been established as a result of individual decisions, the original decisions may no longer be beneficial and fishes may find themselves trapped in a collective maladaptive pattern of behaviour (Fig. 16.8). However, in some cases fishes can have a choice between different group structures. Two divisions of fishes have been observed to coexist in salmon net pens, with fishes at the centre swimming in different directions and fishes along the net walls in a polarised schoollike way (Fern¨o et al. 1988; Juell 1995, Fig. 16.12). Over time the polarised group became larger and eventually took over the whole cage, indicating that individual fishes chose to join the structured division. School-like swimming should lead to fewer physical encounters, resulting in a lower perceived fish density and stress level and thus higher pay-off. In fact, schooling seems to be correlated with high food intake (Fern¨o et al. 1988), indicating that an adaptive collective behaviour developed in this case.

16.4

Stock enhancement and sea-ranching

For the past century releases of hatchery-reared fishes have been an important fisheries management tool in the restoration or creation of new fisheries and in increasing recruitment in areas in which the natural recruitment of juveniles is, or is believed to be, less than the body of water can sustain (Shelbourne 1964; Bowen 1970; Cowx 1994; White et al. 1995; Munro & Bell 1997). In the course of the past few decades rearing techniques for the mass production of many freshwater and marine fish species have been developed and a great deal of effort has been put into mass releases of more than 250 species (Howell et al. 1999; Brown & Day 2002). Today, reared juveniles are released all over the world, from small-scale releases to enhance recreational fisheries in rivers and lakes to industrial-scale releases of billions of salmonids in the North Pacific (Howell et al. 1999; Brown & Laland 2001; Leber et al. 2004). In spite of some documented success stories (Aprahamian et al. 2003), such as the releases of chum salmon (O. keta, Salmonidae) in Japan (Kaeriyama 1999), the benefits of the releases have not usually been properly evaluated, and the use of reared fish for stock enhancement is still a controversial issue (Cowx 1994; Blankenship & Leber 1995; White et al. 1995; Hilborn 1998; Welcomme 1998). When hatchery-reared fishes are released into the wild, they must immediately cope with a novel and complex environment, identify and catch live prey, and avoid the risk of predator attacks. Therefore, it is not surprising that reared fishes run into problems after release and a major problem with restocking is the high mortality of hatchery-reared fishes during the first period after release. Their poor performance can be explained by a

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

385

FRESH WATER

Pay-off A down to B

SEA WATER February–May

Unstructured

Schooling Pay-off C

June

Pay-off B

Divisions

Schooling

Pay-off C

July–September

Schooling

Schooling

Fig. 16.12 A possible example of long-lasting effects of the early rearing environment on group structure in salmon in marine net pens (Fern¨o et al. 1988). Parr on the left were reared in small tanks with consistently different current directions and parr to the right in larger tanks with a changing current direction. The fishes were transferred to marine net pens in May/June and the group structure recorded each weekday from February until September the following year. Unlike the parr reared in large tanks, those from small tanks spent a long period in unstructured swimming before eventually changing to school-like swimming. This was presumably caused by a conflict between different swimming directions. The shift in group structure can be explained by a change in the pay-off matrix, which was supported by data on food intake. The pay-off for unstructured fishes might have gradually decreased from A to B in connection with high and increasing perceived density as the fish grew at the same time as the growth potential should increase when the temperature rises in the spring. Some fishes then began to school along the net wall, thereby attaining a higher pay-off (C) than the unstructured fishes (B) at the centre. Unstructured fishes may thus have chosen to join the structured division, and a frequency-dependent pay-off may have led to a rapid change until almost all fishes swam in a polarised way, which co-occurred with increased food intake.

BLBK374-16

BLBK374-Brown

386

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

combination of incomplete neural and sensory development due to prenatal stress and lack of appropriate environmental stimuli during critical periods in ontogeny, impaired preycatching and predator-avoidance skills and physiological and psychological stress responses after release, in addition to genetic selection (Blaxter 1976; Browman 1989; Sv˚asand et al. 1998; Huntingford 2004; Sv˚asand 2004). Mortality rates of juvenile fishes are difficult to measure in nature and direct comparisons between reared and wild fishes are few (Kristiansen 2001; Sv˚asand 2004). However, on the basis of comparisons of recapture rates of tagged wild fishes and released reared fishes and analyses of stomach contents of predators caught in the release area, it seems clear that released fishes are more vulnerable to predation than their wild counterparts (Miller 1954; Elson 1975; Blaxter 1976; Olla et al. 1998; Brown & Day 2002). Predator-avoidance skills improve rapidly with experience, and several laboratory studies have found better predator avoidance after relatively short experience of predators (Ginetz & Larkin 1976; Jakobsson & J¨arvi 1976; Patten 1977; Olla & Davis 1989; Suboski & Templeton 1989; J¨arvi & Uglem 1993; Hossain et al. 2002). Vilhunen et al. (2005) demonstrated social learning of predator recognition in hatchery-reared Arctic charr and Arai et al. (2007) showed that flounder juveniles are capable of predator conditioning through both direct and observational learning processes. Surprisingly few examples exist of predator-trained fishes being released into the wild, so we need to be cautious in extrapolating findings from laboratory studies to natural conditions. In laboratory experiments, Nødtvedt et al. (1999) demonstrated differences in antipredatory behaviour between predator-na¨ıve reared cod and wild cod. Reared cod went from a stage of excessively weak reactions to predators to a subsequent stage of excessively strong reactions, but presumably over time developed the ‘balanced’ response displayed by wild cod. However, in a release experiment with reared cod exposed just before release to large predators in their cages, no effects of predator training on later predation or recapture rates were found (Otter˚a et al. 1999). One explanation for this could be that the appropriate responses to predators are learned within a short period of time after release, and that the difference in mortality between trained and untrained fishes in that short period is too small to be detected in field studies. The fact that these were pond-reared fishes, raised in a semi-natural environment, may also help to explain the similar mortality of reared and wild cod and the negligible effect of training. Hawkins et al. (2007) also found that hatchery-reared salmon smolts that had been conditioned to respond to pike odour were no more likely than control fishes to survive following release into a Scottish river system in which pike is the main predator. Foraging skills are acquired through learning based on innate predispositions (Kieffer & Colgan 1992; Chapter 2). After release, reared cod chose immobile prey the first days after release (Nordeide & Salvanes 1991), but within 2 weeks they developed the same prey preferences as wild conspecifics (Kristiansen & Sv˚asand 1992, Fig. 16.13). However, released reared fishes usually need several weeks or months to reach the same daily food energy intake as wild fish (Sosiak et al. 1979; Johnsen & Ugedal 1986; Kristiansen & Sv˚asand 1992; Nordeide & Foss˚a 1992). The overlap in prey choice can be explained by inherited habitat choices, for example hiding in sea weed areas and visual predation on moving prey resulting in similar prey availability as wild fish, but learning to find the same prey by attraction to areas with feeding wild fish may also be involved. The lower feed intake can be explained by poorer prey-catching skills, but also by factors such as lack of

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

387

Niche overlap index reared vs. wild

160

Stomach contents (mean dry weight, mg)

BLBK374-16

140 120 100 80 60 40 20 0 R1

R2

R3

Ra

W

Fig. 16.13 Mean dry weight of stomach contents from released reared cod recaptured 1 (R1), 2 (R2) and 3 weeks (R3), and 6 months (Ra) after release, compared to the stomach contents of wild cod (W) of the same size. The sector diagrams indicate the niche overlap index based on prey numbers (Pianka 1973) between the reared groups and the wild cod. (Redrawn from Kristiansen & Sv˚asand 1992.)

appropriate ‘searching images’ (Ware 1971), higher fearfulness (e.g. towards larger fish) leading to reduced feeding (Nordeide & Sv˚asand 1990), higher migratory activity and less food search and fewer associations with good feeding areas (Kristiansen & Sv˚asand 1992). Poor prey-catching skills will also have indirect effects on predator avoidance, since more time and attention need to be devoted to prey search and hunger may increase risk-taking behaviour (Hossain et al. 2002). Steingrund & Fern¨o (1997) showed that reared cod displayed much more active prey-catching behaviour than wild cod, which could lead to less predator attention and avoidance. Fishes trained to catch live prey in the hatchery before release have been shown to have better prey-catching abilities after release (Suboski & Templeton 1989), and observation of a trained Atlantic salmon increased the rate at which na¨ıve hatchery-reared fish accepted novel, live prey (Brown & Laland 2002b). Several studies of reared fishes have showed that the rearing process may lead to other morphological and behavioural deficiencies that increase mortality after release (see reviews by Blaxter 1976; Howell 1994; Ellis et al. 1997; Olla et al. 1998; Sv˚asand et al. 1998; Tsukamoto et al. 1999; Huntingford 2004; Masuda 2004; Sv˚asand 2004). For example, reared Japanese flounder (Paralichthys olivaceus, Paralichthyidae) swim longer distances and spend more time in the water column after take-off from the bottom and display a low incidence of burrowing behaviour (Yamashita & Yamada 1999). The larger the differences from natural food and environment in the rearing environment, the greater will be the morphological and behavioural differences found and such differences also increase with age at release (Tsukamoto et al. 1999; Masuda 2004). In many cases reared fishes are in a poorer physiological state due to inappropriate nutrition, less exercise and high stress level in the tanks. Poor larval and juvenile nutrition might affect the development of the brain and nervous system with lifelong effects

BLBK374-16

BLBK374-Brown

388

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

on the sensory system and learning ability. For example, yellowtail (Seriola quinqueradiata, Carangidae) fed Artemia enriched on a diet without the fatty acid DHA did not school (Ishizaki et al. 2001). While knowledge of the effects of stress on neural plasticity in fishes is sparse, social stress in trout has been shown to reduce cell proliferation in the telencephalon (Sørensen et al. 2007). In mammals, many different forms of acute and chronic stress reduce hippocampal neurogenesis that is often related to elevated glucocorticoid levels (reviewed in Pittenger & Duman 2008). Rearing fishes in enriched or more natural environments has positive effects on prey-catching and antipredator behaviour (Berejikian et al. 2001; Brown & Day 2002). Cod from enriched environments learn to switch earlier to new prey types, are more aggressive when feeding, use cover more often and recover sooner from a novel fright stimulus (Braithwaite & Salvanes 2005; Salvanes & Braithwaite 2005) and displayed different shoaling responses (Salvanes et al. 2007). It thus seems that simulated natural environments may help newly released fishes to categorise their environmental cues more correctly and hence learn to adjust their behaviour vis-`a-vis their new environmental concepts more readily. Therefore, it is reasonable to assume that some behavioural differences are caused by lack of appropriate stimuli during ontogeny (Sv˚asand et al. 1998). Hawkins et al. (2008) showed that reared salmon fry and parr had periods during which they recognised predator odours better and had better learning ability. The onset of learning occurred at the age when wild fishes undergo a habitat shift that greatly increases their encounters with predators, and they suggested that attempts to improve predator recognition skills of fishes should take greater account of life history and focus on the ontogenetic stage where learning is favoured. Several studies have shown that the brain size is reduced in hatchery-reared fishes (Marchetti & Nevitt 2003; Kihslinger & Nevitt 2006), which may explain some of the poorer coping ability in the wild. In comparison with the enormous costs and efforts that have been allocated to restocking activities, surprisingly little has been done to develop full-scale methods for production of ecologically viable fry. As we have shown, juvenile quality with survival skills can be improved by enriching the rearing environment and by various kinds of predator and livefeed training, and methods for large-scale survival skill training using social (transmission chain) learning protocols should be developed (Brown & Laland 2001; Brown & Day 2002).

16.5

Escapees from aquaculture

Many fishes unintentionally escape from aquaculture facilities to the wild. In such cases we actually prefer the fishes not to adapt to the natural conditions, even if suboptimal adaptation may be a fish welfare problem. Escapees can compete with wild fish and disrupt spawning, and in particular produce a risk of genetic introgression from hybridisation between wild and farmed populations resulting in fitness depression (Naylor et al. 2005). The behaviour of escapees should be expected to be even more maladaptive than the behaviour of searanched fish, as escapees often encounter the wild environment at a later stage and have thus been in a situation with restricted exposure to natural stimuli for longer time during ontogeny. Nevertheless, large numbers of escaped salmon have been reported to successfully locate natural spawning sites and spawn. Although farmed salmonids generally do poorly in spawning competition with wild fish, they interbreed with wild fish, with farmed females

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

389

acting as the main vector for farmed genes into wild populations (Fleming et al. 1996; Weir et al. 2004). Similarly, several studies have shown that farmed cod can navigate to local spawning grounds (Sv˚asand et al. 1990; Wroblewski et al. 1996; Nøstvik & Pedersen 1999; Uglem et al. 2008) and may remain in the vicinity of spawning sites for a considerable time (Meager et al. 2009). Innate mechanisms such as movement towards odours from spawning products or social learning may lead escaped cod to the sites. Although marked differences in the vertical distribution of farmed and wild cod on a spawning ground have been observed, interbreeding does seem to take place (Meager et al. 2009; Meager et al. 2010). The genetic effect of domestication on the fitness of farmed fish in the wild may be extremely rapid (Araki et al. 2007).

16.6

Capture-based aquaculture

Fishes that are caught and subsequently kept in net cages for on-growth in capture-based aquaculture ought to face the most serious difficulties when transferred between different environments. Such individuals have spent much of their life in the wild and must adapt to a markedly different environment after experiencing severe capture stress. Habituated fishes are often used to ease the transfer of the newcomers from live to dead food, but in many species the transfer to dry food may still be difficult. In spite of these difficulties, large-scale capture-based aquaculture for species such as eels (Anguilla spp.), groupers (Epinephelus spp.), tunas and yellowtails (Seriola spp.) exist (Ottolenghi et al. 2004). Although the wild fishes often show rapid growth in cages (Ticina et al. 2007), the behaviour and welfare after the transition from nature to culture have barely been studied. The transmission between the two environments can be assisted by the use of social learning. In sea trout, farmed mentors facilitate the switch between natural food and dry food pellets (Birkeland & Jakobsen 1997), and groupers (Serranidae spp.) are generally reared with sea bream that stimulate the groupers to feed (Boonyaratpalin 1997).

16.7

Conclusions and perspectives

Fishes may be more traumatised by stimuli relevant to their ancient survival traits than by novel stimuli. Hence, the ability to learn and speed of learning may be greatly influenced by the novelty of fishing gear, farming facilities and general breeding conditions. Everchanging fishery technologies and fish culture environments have presumably brought about greater variance in individual fitness in commercially exploited populations. This may favour flexible traits (Caraco 1980; Real 1980), and among those, improved learning skill is one clear candidate. However, there is a discrepancy in the time scale involved, and fishes are not necessarily able to change from preprogrammed behaviours that have evolved to deal with a previously more stable world to behaviour appropriate to a new and rapidly changing environment. Another complication is that fisheries and fish stocking may modify competition between trophic levels and create indirect changes that are difficult to track and relate to consequences of man-made changes. Finally, since behavioural traits may be transferred socially, selective fisheries may change whole cultures, and consequently

BLBK374-16

BLBK374-Brown

390

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

change migratory routes, national ownership of fishery resources and thereby have a great impact on fishery practice. To identify and interpret reliable signs and scaffolding mechanisms (Hoffmeyer 2007) that can lead the animal to correct action is the main challenge. Behaviour is an emergent property of the interaction of the brain, body and environment. When evolved under natural conditions, behaviour is expected to be adaptive, while under novel conditions the behaviour expressed may also be novel or maladaptive. Hence, it is difficult to predict the outcomes of different treatments based on conventional theory. One benefit under artificial conditions is that learning can compensate for inadequacies in an animal’s genotype with respect to its new environment and result in adaptive behaviour (Sutter & Kawecki 2009). Therefore, it is essential to ensure that the learning toolbox matches its environment (Marples & Roper 1996) and further, to reduce the number of conflicting stimuli. As pointed out by Rodr´ıguez et al. (2006), studies of the mechanisms of learning in fishes have been neglected. Ecological studies seldom take into account the constraints imposed by the toolbox of learning mechanisms that are available to each species. Nobody would conclude that a blind cavefish actually make a behavioural decision when it cannot learn to react to the visual image of an attacking predator, but similar misinterpretations can be made if a species that does not modify its behaviour actually lacks the mental capacity to make an association between the conditioned and unconditioned stimulus under a given learning schedule. The difference between delay conditioning and trace conditioning must be taken into account. Inappropriate conclusions can also be drawn if we forget that an individual of one species initially approach a stimulus (sign-tracking, Nilsson et al. 2008a, 2008b), whereas a member of another species initially approaches the reward (goaltracking). Moreover, the overt response of a conditioned fish can be very subtle and easily overlooked (Nilsson et al. 2010), which can lead to misunderstandings about the ability to learn. Such species differences could reflect different foraging strategies and are crucial to our understanding of species-specific adaptations. It is also essential to be aware of the capacity of different species to learn different tasks if we wish to understand the transmission of migration patterns in fishes as well as in sea-ranching when they are shaping a functional behaviour and in fish farming when they attempt to utilise learning in various contexts. It is not a straightforward matter to transfer observations on the role of learning in fish capture in a well-defined laboratory situation to the highly variable marine environment (Løkkeborg et al. 1993), and one bottleneck in understanding the role of learning in fish capture is the lack of field observations of individual fish over time. In sea-ranching too, more verification tests of laboratory experiments need to be performed in full-scale experiments in natural environments. Our knowledge of the role of learning in the spatial dynamics of fishes and of where in the brain spatial learning is controlled has greatly expanded during the past decade (see Chapter 8; Broglio et al. 2003; Chapter 15). However, research of this sort is predominantly based on laboratory experiments, and our steadily growing knowledge of the meso- and macro-scale dynamics of commercially exploited species in the field is seldom considered. Likewise, field studies relate only to a limited extent to detailed laboratory findings. The interpretations of field data have been hampered by a primarily descriptive approach and by relatively crude data resulting from limited tracking accuracy, short observation periods and few replications. Recent technological advances, such as geographical information systems (GIS), multi-beam sonars, PIT tags, miniature

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Behaviour, Learning, Aquaculture and Fisheries

391

acoustic tags and archival tags should enable us to acquire a better understanding of spatial dynamics and the role of learning. Individual-based modelling provides an ideal platform for linking individual fish behaviour and learning to population dynamics (Huston et al. 1988; Huse et al. 2002b). We are looking forward to a productive synthesis between the different scientific traditions in the laboratory, field and modelling. In areas of strong fishing pressure where man challenges the fishes by repeatedly ‘attacking’ with an arsenal of fishing gear, it is even more vital for the fishes to react in an adaptive and balanced way than when it is under threat from natural predators alone (Lima & Bednekop 1999). In order to avoid gear while continuing fitness-related activities essential for growth and reproduction, it is crucial for the fish to distinguish Danger from No danger. The benefits of learning skills should then be greater. It has been demonstrated that fishing has led to a rapid evolution of life-history traits (Olsen et al. 2004). Similarly, the idea cannot be excluded that an improved genetically controlled learning capacity can evolve in heavily exploited species. The general reaction to novel stimuli can also change if selection favours stronger neophobia. In fish farming there can also be genetically based changes in learning ability. The selection of fishes on the basis of rapid growth can lead to changes in behavioural traits (Huntingford 2004), possibly also influencing innate learning skills. One might expect that a high capacity to learn should facilitate adaptation to the farming environment and thereby enhance growth and be selected for. However, the problems that farmed fishes face may still be simple compared to those of wild fishes. In fact, the brains of hatchery-reared fishes are relatively smaller in several critical measures than their wild counterparts (Marchetti & Nevitt 2003), partly as a result of the influence of the early rearing environment (Kihslinger & Nevitt 2003, 2006). More effort ought to be put into studying both basic and applied aspects of learning.

Acknowledgements We thank Svein Løkkeborg and Culum Brown for valuable comments on the chapter.

References ˚ dlandsvik, B., Huse, G. & Michalsen, K. (2007) Introducing a method for extracting horizontal A migration patterns from data storage tags. Hydrobiologia, 582, 187–197. Alan¨ar¨a, A. (1996) The use of self-feeders in rainbow trout (Oncorhynchus mykiss) production. Aquaculture, 145, 1–20. Alan¨ar¨a, A. & Br¨ann¨as, E. (1997) Diurnal and nocturnal feeding activity in Arctic char (Salvelinus alpinus) and rainbow trout (Oncorhynchus mykiss). Canadian Journal of Fisheries and Aquatic Sciences, 54, 2894–2900. Altb¨acker, V. & Cs´anyi, V. (1990) The role of eye-spots in predator recognition and antipredatory behaviour in the paradise fish (Macropodus opercularis). Ethology, 85, 51–57. Amano, M., Iigo, M., Furukawa, K., Tabata, M. & Yamamori, K. (2006) Photic and circadian regulation of self-feeding activity in ayu. Fisheries Science, 72, 250–255. Aprahamian, M.W., Martin-Smith, K., McGinnity, P., McKelvey, S. & Taylor, J. (2003) Restocking of salmonids – opportunities and limitation. Fisheries Research, 62, 211–227.

BLBK374-16

BLBK374-Brown

392

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Arai, T., Tominaga, O., Seikaia, T. & Masuda, R. (2007) Observational learning improves predator avoidance in hatchery-reared Japanese flounder Paralichthys olivaceus juveniles. Journal of Sea Research, 58, 59–64. Araki, H., Cooper, B. & Blouin, M.S. (2007) Genetic effects of captive breeding cause a rapid, cumulative fitness decline in the wild. Science, 318, 100–103. Arlinghaus, R., Cooke, S.J., Schwab, A. & Cowx, I.G. (2007) Fish welfare: a challenge to the feelings-based approach, with implications for recreational fishing. Fish and Fisheries, 8, 57–71. Askey, P.J., Richards, S.A. & Post, J.R. (2006) Linking angling catch rates and fish learning under catch-and-release regulations. North American Journal of Fisheries Management, 26, 1020–1029. Azzaydi, M., Madrid, J.A., Zamora, S., S´anchez-V´azquez, F.J. & Mart´ınez, F.J. (1998) Effect of three feeding strategies (automatic, ad libitum demand-feeding and time-restricted demand-feeding) on feeding rhythms and growth in European sea bass (Dicentrarchus labrax L.). Aquaculture, 163, 285–296. Babcock, E.A., Pikitch, E.K., McAllister, M.K., Apostolaki, P. & Santora, C. (2005) A perspective on the use of spatialized indicators for ecosystem-based fishery management through spatial zoning. ICES Journal of Marine Science, 62, 469–476. Bakken, M., Moe, R. & Smith, A. (1993) Radio telemetry: A method of evaluation stress and learning ability in the silver fox (Vulpes vulpes). Proceedings International Society of Applied Ethology, 3rd Joint Meeting, 1993, 591–594. Balaban, E. (1997) Changes in multiple brain regions underlie species differences in a complex, congenital behaviour. Proceedings of the National Academy of Science, USA, 94, 2001–2006. Balon, E.K. (1984) Reflections on some decisive events in the early life of fishes. Transactions of the American Fisheries Society, 113, 178–185. Berejikian, B.A., Tezak, E.P., Risley, S.C. & LaRae, A. (2001) Competitive ability and social behaviour of juvenile steelhead reared in enriched and conventional hatchery tanks and a stream environment. Journal of Fish Biology, 59, 1600–1613. Beukema, J.J. (1970) Angling experiments with carp (Cyprinus carpio L.). II. Decreasing catchability through one-trial learning. Netherlands Journal of Zoology, 20, 81–92. Beukema, J.J. & de Vos, G.J. (1974) Experimental tests on a basic assumption of the capturerecapture method in pond populations of carp Cyprinus carpio L. Journal of Fish Biology, 6, 317–329. Birkeland, C. & Dayton, P.K. (2005) The importance in fishery management of leaving the big ones. Trends in Ecology & Evolution, 20, 356. Birkeland, K. & Jakobsen, P.J. (1997) Salmon lice, Lepeophtheirus salmonis, infestation as a causal agent of premature return to rivers and estuaries by sea trout, Salmo trutta, juveniles. Environmental Biology of Fishes, 49, 129–137. Blankenship, H.L. & Leber, K.M. (1995) A responsible approach to marine stock enhancement. American Fisheries Society Symposium, 15, 167–175. Blaxter, J.H.S. (1976) Reared and wild fish – how do they compare? In: G. Persone & E. Jaspers (eds) Proceedings of the 10th European Symposium on Marine Biology, Vol. 1, pp. 11–26. Universal Press, Wettern, Belgium. Block, B.A., Teo, S.L.H., Walli, A., Boustany, A., Stokesbury, M.J.W., Farwell, C.J., Weng, K.C., Dewar, H. & Williams, T.D. (2005) Electronic tagging and population structure of Atlantic bluefin tuna. Nature, 434, 1121–1127. Boakes, R.A. (1977) Performance on learning to associate a stimulus with positive reinforcement. In: H. Davis & H.M. Hurwitz (eds) Operant–Pavlovian Interactions, pp. 67–97. Erlbaum, Hillsdale, NJ. Boonyaratpalin, M. (1997) Nutrient requirements of marine food fish cultured in Southeast Asia. Aquaculture, 51, 283–313. Bowen, J.T. (1970) A history of fish culture as related to the development of fishery programs. American Fisheries Society Special Publication, 7, 71–93.

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

393

Braithwaite, V.A., Armstrong, J.D., Mcadam, H.M. & Huntingford, F.A. (1996) Can juvenile Atlantic salmon use multiple cue systems in spatial learning? Animal Behaviour, 51, 1409–1415. Braithwaite, V.A. & Salvanes, A.G.V. (2005) Environmental variability in early rearing environment generates behaviourally flexible cod: Implications for rehabilitating wild populations. Proceedings of the Royal Society of London, Series B-Biological Sciences, 272, 1107–1113. Bratland, S., Stien, L.H., Braithwaite, V.A., Juell, J.-E., Folkedal, O., Nilsson, J., Oppedal, F. & Kristiansen, T.S. (2010) From fright to anticipation: Using aversive light stimuli to investigate reward conditioning in large groups of Atlantic salmon (Salmo salar L.). Aquaculture International, 18, 991–1001. Brill, R.W., Block, B.A., Boggs, C.H., Bigelow, K.A., Freund, E.V. & Marcinek, D.J. (1999) Horizontal movements and depth distribution of large adult yellowfin tuna (Thunnus albacares) near the Hawaiian Islands, recorded using ultrasonic telemetry: implications for the physiological ecology of pelagic fishes. Marine Biology, 133, 395–408. Broglio, C., Rodr´ıguez, F. & Salas, C. (2003) Spatial cognition and its neural basis in teleost fishes. Fish and Fisheries, 4, 247–255. Browman, H.I. (1989) Embryology, ethiology and ecology of ontogenetic critical periods in fish. Brain, Behaviour and Evolution, 34, 5–12. Brown, C. (2001) Familiarity with the test environment improves escape responses in the crimson spotted rainbowfish, Melanotaenia duboulayi. Animal Cognition, 4, 109–113. Brown, C., Davidson, T. & Laland, K. (2003) Environmental enrichment and prior experience improve foraging behaviour in hatchery-reared Atlantic salmon. Journal of Fish Biology, 63 (s1), 187–196. Brown, C. & Day, R. (2002) The future of stock enhancements: bridging the gap between hatchery practice and conservation biology. Fish and Fisheries, 3, 79–94. Brown, C., Jones, F. & Braithwaite, V. (2005) In situ examination of boldness-shyness traits in the tropical poeciliid, Brachyraphis episcopi. Animal Behaviour, 70, 1003–1009. Brown, C. & Laland, K.N. (2001) Social learning and life skills training for hatchery reared fish. Journal of Fish Biology, 59, 471–493. Brown, C. & Laland, K.N. (2002a) Social learning of a novel avoidance task in the guppy, P. reticulata: Conformity and social release. Animal Behaviour, 64, 41–47. Brown, C. & Laland, K. (2002b) Social enhancement and social inhibition of foraging behaviour in hatchery-reared Atlantic salmon. Journal of Fish Biology, 61, 987–998. Brown, C. & Laland, K.N. (2003) Social learning in fishes: a review. Fish and Fisheries, 4, 280–288. Brown, C. & Warburton, K. (1999a) Differences in timidity and escape responses between predatorna¨ıve and predator-sympatric rainbowfish populations. Ethology, 105, 491–502. Brown, C. & Warburton, K. (1999b) Social mechanisms enhance escape responses in shoals of rainbowfish, Melanotaenia duboulayi. Environmental Biology of Fishes, 56, 455–459. Brown, G.E. (2003) Learning about danger: chemical alarm cues and local risk assessment in prey fishes. Fish and Fisheries, 4, 227–234. Brydges, N.M., Heathcote, R.J.P. & Braithwaite, V.A. (2007) Habitat stability and predation pressure influence learning and memory in populations of three-spined sticklebacks. Animal Behaviour, 75, 935–942. Bull, H.O. (1928) Studies on conditioned responses in fishes. Journal of the Marine Biological Association of the UK, 15, 485–533. Camazine, S., Deneubourg, J.L., Franks, N.G., Sneyd, J., Theraulaz, G. & Bonebeau, E. (2001) Self-organization in Biological Systems. Princeton University Press, Princeton, NJ. Caraco, T. (1980) On foraging time allocation in a stochastic environment. Ecology, 61, 119–128. Checkley Jr., D.M. (1982) Selective feeding by Atlantic herring (Clupea harengus) larvae on zooplankton in natural assemblages. Marine Ecology Progress Series, 9, 245–253. Chen, W.M., Naruse, M. & Tabata, M. (2002) Circadian rhythms and individual variability of selffeeding activity in groups of rainbow trout Oncorhynchus mykiss (Walbaum). Aquaculture Research, 33, 491–500. Chen, W.M., Umeda, N., Mitsuboshi, T. & Hirazawa, N. (2007) Circadian self-feeding rhythms in greater amberjack Seriola dumerili (Risso). Journal of Fish Biology, 70, 451–461.

BLBK374-16

BLBK374-Brown

394

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Chivers, D.P. & Smith, R.J.F. (1994) Fathead minnows, Pimephales promelas, acquire predator recognition when alarm substance is associated with the sight of unfamiliar fish. Animal Behaviour, 48, 597–605. Chivers, D.P. & Smith, J.F. (1995) Chemical recognition of risky habitats is culturally transmitted among fathead minnows, Pimephales promelas (Osteichthyes, Cyprinidae). Ethology, 99, 286–296. Christensen, V., G´uanette, S., Heymans, J.J., Walters, C.J., Watson, R., Zeller, D. & Pauly, D. (2003) Hundred-year decline of North Atlantic predatory fishes. Fish and Fisheries, 4, 1–24. Conradt, L, Krause, J., Couzin, I.D. & Roper, T.J. (2009) “Leading According to Need” in SelfOrganizing Groups. American Naturalist, 173, 304–312. Conte, F.S. (2004) Stress and the welfare of cultured fish. Applied Animal Behaviour Science, 86, 205–223. Cooke, S.J., Suski, C.D., Ostrand, K.G., Wahl, D.H. & Philipp, D.P. (2007) Physiological and behavioral consequences of long-term artificial selection for vulnerability to recreational angling in a teleost fish. Physiological and Biochemical Zoology, 80, 480–490. Corten, A. (1999) The reappearance of spawning Atlantic herring (Clupea harengus) on Aberdeen Bank (North Sea) in 1983 and its relationship to environmental conditions. Canadian Journal of Fisheries and Aquatic Sciences, 56, 2051–2061. Corten, A. (2001) The role of “conservatism” in herring migrations. Reviews in Fish Biology and Fisheries, 11, 339–361. Couzin, I.D., Krause J., Franks, N.R. & Levin, S.A. (2005) Effective leadership and decision-making in animal groups on the move. Nature, 433, 513–516. Cov`es, D., Beauchaud, M., Attia, J., Dutto, G., Bouchut, C. & B´egout Anras, M.L. (2006) Longterm monitoring of individual fish triggering activity on a self-feeding system: An example using European sea bass (Dicentrarchus labrax). Aquaculture, 385, 385–392. Cowx, I.G. (1994) Stocking strategies. Fisheries Management and Ecology, 1, 15–30. Cox, E.S. & Pankhurst, P.M. (2000) Feeding behaviour of greenback flounder larvae, Rhombosolea tapirina (Gunther) with differing exposure histories to live prey. Aquaculture, 183, 285–297. Crossland, M.R. (2001) Ability of predatory native Australian fishes to learn to avoid toxic larvae of the introduced toad Bufo marinus. Journal of Fish Biology, 59, 319–329. Cs´anyi, V. (1986) Ethological analysis of predator avoidance by the Paradise fish (Macropodus opercularis). II. Key stimuli in avoidance learning. Animal Learning and Behaviour, 14, 101–109. Cury, P. & Anneville, O. (1998) Fisheries resources as diminishing assets: marine diversity threatened by anectotes. In: M.H. Durand, P. Cury, R. Mendelssohn, C. Roy, A. Bakun & D. Pauly (eds) Global versus Local Changes in Upwelling Systems, pp. 537–548. Orstom Editions, Paris. Day, R. MacDonald, T., Brown, C., Laland, K. & Reader, S.M. (2001) Interactions between shoal size and conformity in guppy social foraging. Animal Behaviour, 62, 917–925. Divanach, P., Kentouri, M., Charalambakis, G., Pouget, F. & Sterioti, A. (1993) Comparison of growth performance of six Mediterranean fish species reared under intensive farming conditions in Crete (Greece), in raceways with the use of self feeders. Special Publication, European Aquaculture Society, 18, 285–297. Divanach, P., Kentouri, M. & Dewavrin, G. (1986) The weaning and the development of biological performance of extensively reared sea bream, Sparus auratus, fry after replacing continuous feeders by self-feeding distributors. Aquaculture, 52, 21–29. Dodson, J.J. (1988) The nature and role of learning in the orientation and migratory behaviour of fishes. Environmental Biology of Fishes, 23, 161–182. Dragesund, O., Johannessen, A. & Ulltang, Ø. (1997) Variation in migration and abundance of Norwegian spring spawning herring (Clupea harengus L.). Sarsia, 82, 97–105. Duffy, G.A., Pike, T.W. & Laland, K.N. (2009) Size-dependent directed social learning in nine-spined sticklebacks. Animal Behaviour, 78, 371–375. Dugatkin, L.A. & Alfieri, M.S. (2003) Boldness, behavioural inhibition and learning. Ethology Ecology and Evolution, 15, 43–49. Ellis, T., Howell, B.R. & Hayes, J. (1997) Morphological differences between wild and hatchery-reared turbot. Journal of Fish Biology, 50, 1124–1128.

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

395

Elson, P.F. (1975) Atlantic salmon rivers, smolt production and optimal spawning: an overview of natural production. Special Publication Serie International Atlantic Salmon Foundation, 6, 96–119. Eng˚as, A. & Jørgensen, T. (1997) The probability of cod (Gadus morhua) and ling (Molva molva) to encounter gillnets based on in situ fish movements. ICES C.M. 1997/W:17. Eng˚as, A., Kyrkjebø Haugland, E. & Øvredal, J.T. (1998) Reactions of cod (Gadus morhua L.) in the pre-vessel zone to an approaching trawler under different light conditions. Preliminary results. Hydrobiologia, 371/372, 199–206. Eng˚as, A. & Løkkeborg, S. (1994) Abundance estimation using bottom gillnet and longline – the role of fish behaviour. In: A. Fern¨o & S. Olsen (eds) Marine Fish Behaviour Related to Capture and Abundance Estimation, pp. 134–165. Fishing News Books, London. Evans, K., Langley, A., Clear, N.P., Williams, P., Patterson, T., Sibert, J., Hampton, J. & Gunn, J.S. (2008) Behaviour and habitat preferences of bigeye tuna (Thunnus obesus) and their influence on longline fishery catches in the western Coras Sea. Canadian Journal of Fisheries and Aquatic Sciences, 65, 2427–2443. Fern¨o, A. (1993) Advances in understanding of basic behaviour – Consequences for fish capture. ICES Marine Science Symposia, 196, 5–11. ˚ . (1988) A multiple approach to behaviour studies of Fern¨o, A., Furevik, D.M., Huse, I. & Bjordal, A salmon reared in marine net pens. ICES C.M. 1988/F:15. Fern¨o, A. & Huse, I. (1983) The effect of experience on the behaviour of cod (Gadus morhua L. ) towards a baited hook. Fisheries Research, 2, 19–28. Fern¨o, A., Pitcher, T.J., Melle, V., Nøttestad, L., Mackinson, S., Hollingworth, C. & Misund, O.A. (1998) The challenge of the herring in the Norwegian Sea: making optimal collective spatial decisions. Sarsia, 83, 149–167. Fern¨o, A., Solemdal, P. & Tilseth, S. (1986) Field studies on the behaviour of whiting (Gadus merlangus L.) towards baited hooks. Fiskeridirektoratets Skrifter Serie Havundersøkelser, 18, 113–122. Fleming, I.A., Jonsson, B., Gross, M.R. & Lamberg, A. (1996) An experimental study of the reproductive behaviour and success of farmed and wild Atlantic salmon (Salmo salar). Journal of Applied Ecology, 33, 893–905. Fr´eon, P. & Misund, O.A. (1998) Dynamics of Pelagic Fish Distribution and Behaviour – Effects on Fisheries and Stock Assessment. Blackwell Publishing Ltd., Oxford. Fuiman, L.A. (1994) The interplay of ontogeny and scaling in the interactions of fish larvae and their predators. Journal of Fish Biology, 45, 55–79. Garcia, J. & Koelling, R.A. (1966) Relation of cue to consequence in avoidance learning. Psychonomic Sciences, 4, 123–124. G´elineau, A., Corraze, G. & Boujard, T. (1998) Effects of restricted ration, time-restricted access and reward level on voluntary food intake, growth and growth heterogeneity of rainbow trout (Oncorhynchus mykiss) fed on demand with self-feeders. Aquaculture, 167, 247–258. Ginetz, R.M. & Larkin, P.A. (1976) Factors affecting rainbow trout (Salmo gairdneri) predation on migrant fry of sockeye salmon (Oncorhynchus nerka). Canadian Journal of Fisheries and Aquatic Sciences, 33, 19–24. Grift, R.E., Heino, M., Rijnsdorf, A.D., Kraaak, S.B.N. & Dieckmann, U. (2007) Three-dimensional maturation reaction norms for North Sea plaice. Marine Ecology Progress Series, 334, 213–224. Hackney, P.A. & Linkous, P.E. (1978) Striking behaviour of the largemouth bass and use of the binomial distribution for its analysis. Transactions of the American Fisheries Society, 107, 682–688. Hall, S.J. & Mainprize, B. (2004) Towards ecosystem-based fisheries management. Fish and Fisheries, 5, 1–20. Harden Jones, F.R. (1968) Fish Migration. Edward Arnold Ltd., London. Haugland, E.K. & Misund, O.A. (2004) Evidence for a clustered spatial distribution of clupeid fish schools in the Norwegian Sea and off the coast of southwest Africa. ICES Journal of Marine Science, 61, 1088–1092.

BLBK374-16

BLBK374-Brown

396

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Hawkins, L.A., Armstrong, J.A. & Magurran, A.E. (2007) A test of how predator conditioning influences survival of hatchery-reared Atlantic salmon, Salmo salar, in restocking programmes. Fisheries Management and Ecology, 14, 291–293. Hawkins, L.A., Magurran, A.E. & Armstrong, J.D. (2008) Ontogenetic learning of predator recognition in hatchery-reared Atlantic salmon, Salmo salar. Animal Behaviour, 75, 1663–1671. Hearst, E. & Jenkins, H.M. (1974) Sign-tracking: The Stimulus-Reinforcer Relation and Directed Actions. Psychonomic Society, Austin, TX. Hebb, D.O. (1949) The Organization of Behavior. John Wiley & Sons, New York. Helfman, G.S. & Schultz, E.T. (1984) Social tradition of behavioural traditions in a coral reef fish. Animal Behaviour, 32, 379–384. High, W.L. & Beardsley, A.J. (1970) Fish behaviour studies from an undersea habitat. Commercial Fisheries Review, 1970, 31–37. Hilborn, R. (1998) The economic performance of marine stock enhancement projects. Bulletin of Marine Science, 62, 661–674. Hoffmeyer, J. (2007) Semiotic scaffolding in living systems. In: M. Barbieri (ed) Introduction to Biosemiotics. The New Biological Synthesis, pp. 149–166. Springer, Dordrecht. Hossain, M.A.R., Tanaka, M. & Masuda, R. (2002) Predator–prey interactions between hatchery reared Japanese flounder juvenile, Paralichthys olivaceus, and sandy shore crab, Matuta lunaris: daily rhythms, anti-predator conditioning and starvation. Journal of Experimental Marine Biology and Ecology, 267, 1–14. Howell, B.R. (1994) Fitness of hatchery reared fish for survival in the sea. Aquaculture and Fisheries Management, 25, 3–17. Howell, B.R., Moksness, E. & Sv˚asand, T. (1999) Stock Enhancement and Sea Ranching. Fishing News Books, Blackwell Publishing Ltd., Oxford. Hsu, Y. & Wolf, L.L. (2001) The winner and loser effect: what fighting behaviour is influenced? Animal Behaviour, 61, 777–786. Humborstad, O.B., Jørgensen, T. & Grotmol, S. (2005) Exposure of cod Gadus morhua to resuspended sediment: an experimental study of the impact of bottom trawling. Marine Ecology Progress Series, 309, 247–254. Huntingford, F.A. (2004) Implications of domestication and rearing conditions for the behaviour of cultivated fishes. Journal of Fish Biology, 65 (Supplement A), 122–142. Huntingford, F.A., Adams, C., Braithwaite, V.A., Kadri, S., Pottinger, T.G., Sandoe, P. & Turnbull, J.F. (2006) Current issues in fish welfare. Journal of Fish Biology, 68, 332–372. Huntingford, F., Adams, C., Braithwaite, V.A., Kadri, S., Pottinger, T.G., Sandoe, P. & Turnbull, J.F. (2007) The implications of a feelings-based approach to fish welfare: a reply to Arlinghaus et al. Fish and Fisheries, 8, 277–280. Huse, G., Fern¨o, A. & Holst, J.G. (2010) Establishment of new wintering areas in herring co-occur with peaks in recruit to repeat spawner ratio. Marine Ecology Progress Series, 409, 189–198. Huse, G., Giske, J. & Salvanes A.G.V. (2002b) Individual-based models. In: P.J.B. Hart & J. Reynolds (eds) Handbook of Fish and Fisheries, Vol. 2, pp. 228–248. Blackwell Publishing Ltd., Oxford. Huse, G., Railsback, S.F. & Fern¨o, A. (2002a) Modelling changes in migration pattern of herring: collective behaviour and numerical domination. Journal of Fish Biology, 60, 571–582. Huston, M., DeAngelis, D. & Post, W. (1988) New computer models unify ecological theory. BioScience, 38, 682–691. ICES (l997) Report of the ICES planning group on surveys on pelagic fish in the Norwegian Sea (PGSPEN). ICES CM 1997/H:3. ICES, Copenhagen. ICES (2007) Report of the working group on northern pelagic and blue whiting fisheries (WGNPBW). ICES CM 2007/ACFM:29. ICES, Copenhagen. Inglis, I.R., Langton, S., Forkman, B. & Lazarus, J. (2001) An information primacy model of exploratory and foraging behaviour. Animal Behaviour, 62, 543–557. Ishizaki, Y., Masuda, R., Uematsu, K., Shimizu, K., Arimoto, M. & Takeuchi, T. (2001) The effect of dietary docosahexaenoic acid on schooling behaviour and brain development in larval yellowtail. Journal of Fish Biology, 58, 1691–1703.

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

397

Jakobsson, S. & J¨arvi, T. (1976) Anti-predator behaviour of two-year-old hatchery-reared Atlantic salmon (Salmo salar) and a description of the predatory behaviour of burbot (Lota lota). Zoologisk Revy, 38, 57–70. Jacobsen, J.A., Gudmundsdottir, A., Heino, M. et al. (2002) Report of the planning group on surveys on pelagic fish in the Norwegian Sea. ICES CM 2002/D:07. ICES, Copenhagen. J¨arvi, T. & Uglem, I. (1993) Predator training improves the anti-predator behaviour of hatchery reared Atlantic salmon (Salmo salar) smolts. Nordic Journal of Freshwater Research, 68, 63–71. Jobling, M., Cov´es, D., Damsg˚ard, B., Kristiansen, H., Koskela, J., Petursdottir, T.E., Kadri, S. & Gudmunson, O. (2001) Techniques for measuring feed intake. In: D. Houlihan, T. Boujard & M. Jobling (eds) Food Intake in Fish, pp. 49–87. Blackwell Publishing Ltd., Oxford. Johnsen, B.O. & Ugedal, O. (1986) Feeding by hatchery-reared and wild brown trout, Salmo trutta L., in a Norwegian stream. Aquaculture and Fisheries Management, 17, 281–287. Johnston, V.S. (1999) Why We Feel: The Science of Human Emotions. Perseus Books, US. Juell, J.-E. (1995) The behaviour of Atlantic salmon (Salmo salar) in relation to efficient cage rearing. Reviews in Fish Biology and Fisheries, 5, 320–335. Kaeriyama, M. (1999) Hatchery programmes and stock management of salmonid populations in Japan. In: B.R. Howell, E. Moksnes & T. Sv˚asand (eds) Stock Enhancement and Sea Ranching, pp. 153–167. Fishing News Books, Blackwell Publishing Ltd., Oxford. Kaiser, M.J. & Spencer, B.E. (1994) Fish scavenging behaviour in recently trawled areas. Marine Ecology Progress Series, 112, 41–49. ˚ . & Fern¨o, A. (2003) Behaviour of cod toward baited gill nets. Kallayil, J.K., Jørgensen, T., Eng˚as, A Fisheries Research, 61, 125–133. Karplus, I, Katzenstein, R. & Goren, M. (2006) Predator recognition and social facilitation predator avoidance in coral reef fish Dascyllus marginatus juveniles. Marine Ecology Progress Series, 319, 215–223. Kasumyan, A.O. & Døving, K.B. (2003) Taste preferences in fish. Fish and Fisheries, 4, 289–347. Katano, O. (2009) Individual differences in catchability of largemouth bass Micropterus salmoides by fishing in an experimental pond. Nippon Suisan Gakkaishi, 75, 425–431. Kerns, J.G., Cohen, J.D., MacDonald, A.W., Cho, R.Y. & Carter, C.S. (2004) Anterior Cingulate conflict monitoring and adjustment in control. Science, 303, 1023–1026. Kieffer, J.D. & Colgan, P.W. (1992) The role of learning in fish behaviour. Reviews in Fish Biology and Fisheries, 2, 125–143. Kihslinger, R.L. & Nevitt, G.A. (2003) The early rearing environment produces variation in the size of the brain subdivisions in steelhead trout (Oncorhynchus mykiss). Integrative and Comparative Biology, 43, 944. Kihslinger, R.L. & Nevitt, G.A. (2006) Early rearing environment impacts cerebellar growth in juvenile salmon. Journal of Experimental Biology, 209, 504–509. Kim, Y.-H. & Wardle, C.S. (2003) Optomotor response and erratic response: quantitative analysis of fish reaction to towed fishing gears. Fisheries Research, 60, 455–470. Koolhaas, J.M., de Boer, S.F., Buwalda, B. & van Reenen, K. (2007) Individual variation in coping with stress: A multidimensional approach of ultimate and proximate mechanisms. Brain, Behaviour and Evolution, 70, 218–226. Koolhaas, J.M., Korte, S.M., De Boer, S.F., Van Der Vegt, B.J., Van Reenen, C.G., Hopster, H., De Jonga, I.C., Ruis, M.A.W. & Blokhuis, H.J. (1999) Coping styles in animals: current status in behaviour and stress-physiology. Neuroscience & Biobehavioural Reviews, 23, 925–935. Kotrschal, K., van Staaden, M.J. & Huber, R. (1998) Fish brains: evolution and environmental relationships. Reviews in Fish Biology and Fisheries, 8, 373–408. Kraemer, P.J. & Golding, J.M. (1997) Adaptive forgetting in animals. Psychonomic Bullentin and Review, 4, 480–491. Kristiansen, T.S. (2001) Enhancement studies of coastal cod in Norway. Stage and size dependent mortality of reared and wild cod (Gadus morhua L.). PhD thesis, University of Bergen.

BLBK374-16

BLBK374-Brown

398

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Kristiansen, T.S. & Fern¨o, A. (2007) Individual behaviour and growth of halibut (Hippoglossus hippoglossus L.) fed sinking and floating feed: evidence of different coping styles. Applied Animal Behaviour Science, 104, 236–250. Kristiansen, T.S. & Sv˚asand, T. (1992) Comparative analysis of stomach contents of cultured and wild cod, Gadus morhua L. Aquaculture and Fisheries Management, 23, 661–668. Kristiansen, T.S. & Sv˚asand, T. (1998) Effect of size-selective mortality on growth of coastal cod illustrated by tagging data and an individual-based growth and mortality model. Journal of Fish Biology, 52, 688–705. Kvamme, C., Nøttestad, L., Fern¨o, A., Misund, O.A., Dommasnes, A., Axelsen, B.E., Dalpadado, P. & Melle, W. (2003) Age-specific migration patterns in Norwegian spring-spawning herring: why young fish swim away from the wintering area in late summer. Marine Ecology Progress Series, 247, 197–210. Laland, K.N. & Williams, K. (1997) Shoaling generates social learning of foraging information in guppies. Animal Behaviour, 53, 1161–1169. Laland, K.N. & Williams, K. (1998) Social transmission of maladaptive information in the guppy. Behavioral Ecology, 9, 493–499. Law, R. & Grey, D.R. (1989) Evolution and yields from populations with age-specific cropping. Evolutionary Ecology, 3, 343–359. Leber, K.M., Kitada, S., Blankenship, H.L. & Sv˚asand, T. (2004) Stock Enhancement and Sea Ranching. Developments, Pitfalls and Opportunities. Blackwell Publishing Ltd., Oxford. Lieberman, D.A. (2000) Learning: Behavior and Cognition, 3rd edn. Wadsworth, Belmont, CA. Lima, S.L. & Bednekoff, P.A. (1999) Temporal variation in danger drives antipredator behavior: the predation risk allocation hypothesis. American Naturalist, 153, 649–659. Lines, J.A. & Frost, A.R. (1999) Review of opportunities for low stress and selective control of fish. Aquacultural Engineering, 20, 211–230. Løkkeborg, S. (1990) Reduced catch of under-sized cod (Gadus morhua) in longlining by using artificial bait. Canadian Journal of Fisheries and Aquatic Sciences, 47, 1112–1115. Løkkeborg, S. (1994) Fish behaviour and longlining. In: A. Fern¨o & S. Olsen (eds) Marine Fish Behaviour Related to Capture and Abundance Estimation, pp. 9–27. Fishing News Books, London. Løkkeborg, S. (1998) Feeding behaviour of cod, Gadus morhua: activity rhythm and chemically mediated food search. Animal Behaviour, 56, 371–378. ˚ . & Fern¨o, A. (1989) Responses of cod (Gadus morhua) and haddock Løkkeborg, S., Bjordal, A (Melanogrammus aeglefinus) to baited hooks in the natural environment. Canadian Journal of Fisheries and Aquatic Sciences, 46, 1478–1483. ˚ . & Fern¨o, A. (1993) The reliability and value of behaviour studies in Løkkeborg, S., Bjordal, A longline gear research. ICES marine Science Symposia, 196, 41–46. Løkkeborg, S. & Fern¨o, A. (1999) Diel activity pattern and food search behaviour in cod, Gadus morhua. Environmental Biology of Fishes, 54, 345–353. Løkkeborg, S., Skajaa, K. & Fern¨o, A. (2000) Food-search strategy in ling (Molva molva L.): crepuscular activity and use of space. Journal of Experimental Marine Biology and Ecology, 247, 195–208. Mackinson, S., Nøttestad, L., Gu´enette, S., Pitcher, T., Misund, O.A. & Fern¨o, A. (1999) Cross-scale observations on distribution and behavioural dynamics of ocean feeding Norwegian spring spawning herring (Clupea harengus). ICES Journal of Marine Science, 56, 613–626. Mackney, P.A. & Hughes, R.N. (1995) Foraging behaviour and memory window in sticklebacks. Behaviour, 132, 1241–1253. Magurran, A.E. (1989) Acquired recognition of predator odour in the European minnow (Phoxinus phoxinus). Ethology, 82, 216–233. Magurran, A.E. (1993) Individual differences and alternative behaviours. In: T.J. Pitcher (ed) The Behaviour of Teleost Fishes, pp. 441–477. Chapman and Hall, London. Marchetti, M.P. & Nevitt, G.A. (2003) Effects of hatchery rearing on brain structures of rainbow trout, Oncorhynchus mykiss. Environmental Biology of Fishes, 66, 9–14.

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

399

Marples, N.M. & Roper, T.J. (1996) Effects of novel colour and smell on the response of na¨ıve chicks towards food and water. Animal Behaviour, 51, 1417–1424. Masuda, R. (2004) Behavioural approaches to fish stock enhancement: a practical review. In: K.M. Leber, S. Kitada, H.L. Blankenship & T. Sv˚asand (eds) Stock Enhancement and Sea Ranching. Developments, Pitfalls and Opportunities, 2nd edn, pp. 83–90. Blackwell Publishing Ltd., Oxford. Mathis, A.D.P., Chivers, D.P. & Smith, J.F. (1996) Cultural transmission of predator recognition in fishes: intraspecific and interspecific learning. Animal Behaviour, 51, 185–201. Matthews, K.R. (1990) An experimental study of the habitat preferences and movement patterns of copper, quillback, and brown rockfishes (Sebastes spp.). Environmental Biology of Fishes, 29, 161–178. Mattson, S. (1990) Food and feeding habitats of fish species over a soft sublittoral bottom in the northeast Atlantic. Sarsia, 75, 247–260. McFarland, D. (1985) Animal Behaviour. Longman Scientific and Technical, Essex. McQuinn, I.H. (1997) Metapopulations and the Atlantic herring. Reviews in Fish Biology and Fisheries, 7, 297–329. Meager, J.J., Skjæraasen, J.E., Fern¨o, A., Karlsen, Ø., Løkkeborg, S., Michalsen, K. & Utskot, S.O. (2009) Vertical dynamics and reproductive behaviour of farmed and wild Atlantic cod Gadus morhua. Marine Ecology Progress Series, 389, 233–243. Meager, J., Skjæraasen, J.E., Fern¨o, A. & Løkkeborg, S. (2010) Reproductive interactions between fugitive farmed and wild cod (Gadus morhua L) in the field. Canadian Journal of Fisheries and Aquatic Sciences, 67, 1221–1231. Midling, K.Ø., Kristiansen T.S, Ona, E. & Øiestad, V. (1987) Fjordranching with conditioned cod (Gadus morhua L). ICES CM 1987/F:10. Miller, R.B. (1954) Comparative survival of wild and hatchery-reared cutthroat trout in a stream. Transactions of the American Fisheries Society, 83, 120–130. Millot, S., B´egout, M.L., Person-Le Ruyet, J., Breuil, G., Di-Po¨ı, C., Pineau, P., Rou´e, M. & S´ev`ere, A. 2008 Feed demand behaviour in sea bass juveniles: effects on individual specific growth rate variation and health (inter-individual and inter-group variation). Aquaculture, 274, 87–95. Millot, S. & B´egout, M.L. (2009) Individual fish rhythm directs group feeding: a case study with sea bass juveniles (Dicentrarchus labrax) under self-demand feeding conditions. Aquatic Living Resources, 22, 363–370. Misund, O.A. (1993) Dynamics of moving masses: variability in packing density, shape, and size among herring, sprat, and saithe schools. ICES Journal of Marine Science, 50, 145–160. Moav, R., Brody, T. & Hulatin, G. (1978) Genetic improvement of wild fish populations. Science, 201, 1090–1094. Moreira, P.S.A., Pulman, K.G.T. & Pottinger, T.G. (2004) Extinction of a conditioned response in rainbow trout selected for high or low responsiveness to stress. Hormones and Behaviour, 46, 450–457. Morris, A.L., Still, M.L. & Harris, C.L.C. (2009) Repitition blindness; an emergent property of inter item competition. Cognitive Psychology, 58, 338–375. Mullon, C., Fr´eon, P. & Cury, P. (2005) The dynamics of collapse in world fisheries. Fish and Fisheries, 6, 111–120. Munro, J.L. & Bell, D. (1997) Enhancement of marine fisheries resources. Reviews in Fisheries Sciences, 5, 185–222. Munro, J.L., Reeson, P.H. & Gant, V.C. (1971) Dynamic factors affecting the performance of the Antillean fish trap. Proceedings of Gulf Carribian Fisheries Institute, 23, 184–194. Naas, K. & Mangor Jensen, A. (1990) Positive phototaxis during late yolk sack stage of Atlantic halibut larvae Hippoglossus hippoglossus. Sarsia, 75, 243–246. Naylor, R., Hindar, K., Fleming, I.A., Goldburg, R., Williams, R., Volpe, J., Whoriskey, F., Eagle, J., Kelso, D. & Mangel, M. (2005) Fugitive salmon: assessing the risks of escaped fish from net-pen aquaculture. Bioscience, 55, 427–437. Neill, W.H. (1979) Mechanisms of fish distribution in heterothermal environments. American Zoologist, 19, 305–317.

BLBK374-16

BLBK374-Brown

400

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Neuenfeldt, S., Andersen, K.H. & Hinrichsen, H.-H. (2009) Some Atlantic cod Gadus morhua in the Baltic Sea visit hypoxic water briefly but often. Journal of Fish Biology, 75, 290–294. Nilsson, J. (2008) Learning and anticipatory behaviour in cod and halibut. PhD thesis, University of Bergen. Nilsson, J., Kristiansen, T.S., Fosseidengen, J.E., Fern¨o, A. & van den Bos, R. (2008a) Learning in cod (Gadus morhua): long trace interval retention. Animal Cognition, 11, 215–222. Nilsson, J., Kristiansen, T.S., Fosseidengen, J.E., Fern¨o, A. & van den Bos, R. (2008b) Sign- and goal-tracking in Atlantic cod (Gadus morhua). Animal Cognition, 11, 651–659. Nilsson, J., Kristiansen, T.S., Fosseidengen, J.E., Stien, L.H., Fern¨o, A. & van den Bos, R. (2010) Learning and anticipatory behaviour in a “sit-and-wait” predator: the Atlantic halibut. Behavioural Processes, 83, 257–266. Nødtvedt, M., Fern¨o, A., Gjøsæter, J. & Steingrund, P. (1999) Anti-predator behaviour of hatcheryreared and wild juvenile Atlantic cod (Gadus morhua L.) and the effect of predator training. In: B.R. Howell, E. Moksness & T. Sv˚asand (eds) Stock Enhancement and Sea Ranching, pp. 350–362. Fishing News Books, Blackwell Publishing Ltd., Oxford. Nordeide, J.T. & Foss˚a, J.H. (1992) Diet overlap between two subsequent year-classes of juvenile coastal cod (Gadus morhua L.) and wild and released cod. Sarsia, 77, 111–117. Nordeide, J.T. & Salvanes, A.G.V. (1991) Observations of reared newly released and wild cod (Gadus morhua L.) and their potential predators. ICES marine Science Symposia, 192, 139–146. Nordeide, J.T. & Sv˚asand, T. (1990) The behaviour of wild and reared juvenile cod, Gadus morhua L., towards a potential predator. Aquatic Fisheries and Management, 21, 317–325. Nøstvik, F. & Pedersen, T. (1999) Movement patterns and growth of wild cod (Gadus morhua L.) and hatchery-reared cod released as 1-group. In: B.R. Howell, E. Moksness & T. Sv˚asand (eds) Stock Enhancement and Sea Ranching, pp. 315–333. Fishing News Books, Blackwell Publishing Ltd., Oxford. Nøttestad, L., Fern¨o, A., Misund, O.A. & Vabø, R. (2004) Understanding herring behaviour: linking individual decisions, school patterns and population distribution. In: H.R. Skjoldal, R. Sætre, A. Fern¨o, O.A. Misund & I. Røttingen (eds) The Norwegian Sea Ecosystem, pp. 227–262. Tapir, Trondheim. Odling-Smee, L.C., Boughman, J.W. & Braithwaite, V.A. (2008) Sympatric species of threespine stickleback differ in their performance in a spatial learning task. Behavioral Ecology and Sociobiology, 62, 1935–1945. O’Grady, K.T. & Huges, P.C.R. (1980) Factorial analysis of an experimental comparison of three methods of fishing for rainbow trout, Salmo gairdneri Richardson, in still water. Journal of Fish Biology, 16, 257–264. Olla, B.L. & Davis, M.W. (1989) The role of learning and stress in predator avoidance of hatchery reared coho salmon (Oncorhynchus kisutch) juveniles. Aquaculture, 76, 209–214. Olla, B., Davis, M. & Ryer, C. (1998) Understanding how the hatchery environment represses or promotes the development of behavioral survival skills. Bulletin of Marine Science, 62, 531–550. Olsen, E.M., Heino, M., Lilly, G.R., Morgan, M.J., Brattey, J., Ernande, B. & Dieckmann, U. (2004) Maturation trends indicative of rapid evolution preceded the collapse of northern cod. Nature, 428, 932–935. Olsen, S.H., Sorensen, N.K., Larsen, R., Elvevoll, E.O. & Nilsen, H. (2008) Impact of pre-slaughter stress on residual blood in fillet portions of farmed Atlantic cod (Gadus morhua) – measured chemically and by visible and near-infrared spectroscopy. Aquaculture, 284, 90–97. Otter˚a, H., Kristiansen, T.S., Sv˚asand, T., Nødtvedt, M. & Borge, A. (1999) Sea-ranching of Atlantic cod (Gadus morhua L.), effects of release strategy on survival. In: B.R. Howell, E. Moksness & T. Sv˚asand (eds) Stock Enhancement and Sea Ranching, pp. 293–305. Fishing News Books, Blackwell Publishing Ltd., Oxford. Ottolenghi, F., Silvestri, C., Giordano, P., Lovatelli, A. & New, M.B. (2004) Capture-based aquaculture. The fattening of eels, groupers, tunas and yellowtails. FAO report, Rome. Øverli, Ø., Sørensen, C., Pulman, K.G.T., Pottinger, T.G., Korzan, W., Summers, C.H. & Nilsson, G.E. (2007) Evolutionary background for stress-coping styles: relationships between physiological,

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

401

behavioral, and cognitive traits in nonmammalian vertebrates. Neuroscience and Biobehavioral Reviews, 31, 396–412. Overmier, J. & Hollis, K. (1990) Fish in the think tank: learning, memory, and integrated behavior. In: R. Kesner & D. Olton (eds) Neurobiology of Comparative Cognition, pp. 205–236. Lawrence Erlbaum, Hillsdale, NJ. ¨ Ozbilgin, H. & Glass, C.W. (2004) Role of learning in mesh penetration behaviour of haddock (Melanogrammus aeglefinus). ICES Journal of Marine Science, 61, 1190–1194. ´ P´alsson, O.K. (1994) A review of trophic interactions of cod stocks in the North Atlantic. ICES Marine Science Symposia, 198, 553–575. Patten, B.G. (1977) Body size and learned avoidance as factors affecting predation on coho salmon, Oncorhynchus kisutch, fry by torrent sculpin, Cottus rhotheus. Fishery Bullentin, 75, 457–459. Pedersen, M.W., Righton, D., Thygesen, H.J., Andersen, K.H. & Madsen, H. (2008) Geolocation of North Sea cod using hidden Markov models and behavioural switching. Canadian Journal of Fisheries and Aquatic Sciences, 65, 2367–2377. Pianka, E.R. (1973) The structure of lizard communities. Annual Review of Ecology and Systematics, 4, 53–73. Pittenger, C. & Duman, R.S. (2008) Stress, depression and neoroplasticity: a convergance of mechanisms. Neuropsychopharmacology, 33, 88–109. Policansky, D. (1993) Fishing as a cause of evolution in fishes. In: T.K.A. Stokes, R. Law & J. McGlade (eds) The Exploitation of Evolving Resources, pp. 2–13. Springer Verlag, Berlin. Pryor, K. (1975) Lads before the Wind. Harper & Row, Evanston, New York. Pyanov, A.I. (1993) Fish learning in response to trawl fishing. ICES Marine Science Symposia, 196, 12–16. Real, L.A. (1980) Fitness, uncertainty, and the role of diversification in evolution and behavior. American Naturalist, 115, 623–638. Reebs, S.G. (1996) Time-place learning in golden shiners (Pisces: Cyprinidae). Behavioural Processes, 36, 253–262. Reebs, S.G. (2000) Can a minority of informed leaders determine the foraging movements of a fish shoal? Animal Behaviour, 59, 403–409. Reebs, S.G. (2001) Influence of body size on leadership in shoals of golden shiners, Notemigonus crysoleucas. Behaviour, 138, 797–809. Reese, E.S. (1989) Orientation behaviour of butterflyfishes (family Chaetodontidae) on coral reefs: spatial learning of route specific landmarks and cognitive maps. Environmental Biology of Fishes, 25, 79–86. Robichaud, D. & Rose, G.A. (2001) Multiyear homing of Atlantic cod to a spawning ground. Canadian Journal of Fisheries and Aquatic Sciences, 58, 2325–2329. Rocklin, D., Santoni, M.-C., Culiolo, J.-M., Tomasini, J.-A., Pelletier, D. & Mouillot, D. (2009) Changes in the catch composition of artisanal fisheries attributable to dolphin depredation in a Mediterranean marine reserve. ICES Journal of Marine Science, 66, 699–707. Rodr´ıguez, F., Broglio, C., Dur´an, E., G´omez, A. & Salas, C. (2006) Neural mechanisms of learning in teleost fish. In: C. Brown, K. Laland & J. Krause (eds) Fish Cognition and Behavior, pp. 243–277. Blackwell Publishing Ltd, Oxford. Rose, G., deYoung, B., Kulka, D., Goddard, S. & Fletcher, G. (2000) Distribution shifts and overfishing the northern cod (Gadus morhua): a view from the ocean. Canadian Journal of Fisheries and Aquatic Sciences, 57, 644–663. Rose, G.A. (1993) Cod spawning on a migration highway in the North-West Atlantic. Nature, 366, 458–461. Rose, G.A. & O’Driscoll, R.L. (2002) Capelin are good for cod: can the northern stock rebuild without them? ICES Journal of Marine Science, 59, 1018–1026. Rose, J.D. (2002) The neurobehavioral nature of fishes and the question of awareness and pain. Reviews in Fisheries Science, 10, 1–38. Rose, J.D. (2007) Anthropomorphism and ‘mental welfare’ of fishes. Diseases of Aquatic Organisms, 75, 139–154.

BLBK374-16

BLBK374-Brown

402

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Rozin, P. & Kalat, J. (1972) Learning as a situation-specific adaption. In: M.E.P. Seligman & J. Hager (eds), Biological Boundaries of Learning, pp. 66–97. Appelton, New York. Rubio, C., Vivasa, M., Sanchez-Mut, A., Sanchez-Vazquez, F.J., Coves, D., Dutto, G. & Madrid J.A. (2004) Self-feeding of European sea bass (Dicentrarchus labrax, L.) under laboratory and farming conditions using a string sensor. Aquaculture, 233, 393–403. Ryer, C.H. (2004) Laboratory evidence of behavioural impairments of fish escaping trawls. ICES Journal of Marine Science, 61, 1157–1164. Salvanes, A.G.V. & Braithwaite, V.A. (2005) Exposure to variable spatial information in the early rearing environment generates asymmetries in social interactions in cod (Gadus morhua). Behavioral Ecology & Sociobiology, 59, 250–257. Salvanes, A.G.V., Moberg, O. & Braithwaite, V.A. (2007) Effects of early experience on group behaviour in fish. Animal Behaviour, 74, 805–811. Sanchez-V´asques, F.J. & Madrid, J.A. (2001) Feeding anticipatory activity. In: D. Houlihan, T. Boujard & M. Jobling (eds) Food Intake in Fish, pp. 217–232. Blackwell Publishing Ltd., Oxford. Schreck, C.B., Jonsson, L., Feist, G. & Reno, P. (1995) Conditioning improves performance of juvenile Chinook salmon, Oncorhynchus tshawytscha, to transportation stress. Aquaculture, 135, 99–110. Schwarz, A.L. & Greer, G.I. (1984) Responses of pacific herring, Clupea harangus pallasi to some underwater sounds. Canadian Journal of Fisheries and Aquatic Sciences, 41, 1183–1192. Shelbourne, J.E. (1964) The artificial propagation of marine fish. Advances in Marine Biology, 2, 1–83. Shelton, P.A. & Healey, B.P. (1999) Should depensation be dismissed as a possible explanation for the lack of recovery of the northern cod (Gadus morhua) stock? Canadian Journal of Fisheries and Aquatic Sciences, 56, 1521–1524. Siebeck, U.E., Litherland, L. & Wallis, G.M. (2009) Shape learning and discrimination in reef fish. Journal of Experimental Biology, 212, 2112–2118. Sigler, M.F. (2000) Abundance estimation and capture of sablefish (Anoplopoma fimbria) by longline gear. Canadian Journal of Fisheries and Aquatic Sciences, 57, 1270–1283. Sih, A., Bell, A. & Johnson, J.C. (2004) Behavioral syndromes: an ecological and evolutionary overview. Trends in Ecology & Evolution, 19, 372–378. Skajaa, K. (1997) Basic movement pattern and chemo-oriented search towards baited gears in demersal species: a field study on ling and edible crab. Master thesis, University of Bergen. Sneddon, L.U. (2003a) The evidence for pain in fish: the use of morphine as an analgesic. Applied Animal Behaviour Science, 83, 153–162. Sneddon, L.U. (2003b) The bold and the shy: individual differences in rainbow trout. Journal of Fish Biology, 62, 971–975. Sørensen, C., Øverli, Ø., Summers, C.H. & Nilsson, G.E. (2007) Social regulation of neurogenesis in teleosts. Brain, Behavior and Evolution, 70, 239–246. Soria, M., Gerlotto, F. & Fr´eon, P. (1993) Study of learning capabilities of tropical clupeoids using an artificial stimulus. ICES marine Science Symposia, 196, 17–20. Sosiak, A.J., Randall, R.G. & McKenzie, J.A. (1979) Feeding by hatchery-reared and wild Atlantic salmon (Salmo salar) parr in streams. Journal of the Fisheries Research Board of Canada, 36, 1408–1412. Stanley, E.L., Kendal, R.L., Kendal, J.R., Grounds, S. & Laland, K.N. (2008) The effect of group size, rate of turnover and disruption to demonstration on the stability of foraging traditions in fish. Animal Behaviour, 75, 565–572. Steingrund, P. & Fern¨o, A. (1997) Feeding behaviour of reared and wild cod and the effect of learning: two strategies of feeding on two spotted goby. Journal of Fish Biology, 51, 334–348. Suboski, M.D. & Templeton, J.J. (1989) Life skills training for hatchery fish: social learning and survival. Fisheries Research, 7, 343–352. Sumpter, D.J.T. (2009) Group behaviour: leadership by those in need. Current Biology, 19, 325–327.

BLBK374-16

BLBK374-Brown

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Fish Behaviour, Learning, Aquaculture and Fisheries

Char Count=

403

Sunuma, T., Amano, M., Yamanome, T., Kiyoshi Furukawa, K. & Yamamori, K. (2007) Self-feeding activity of a pleuronectiform fish, the barfin flounder. Aquaculture, 270, 566– 569. Sutter, M. & Kawecki, T.J. (2009) Influence of learning on range expansion and adaption to novel habitats. Journal of Evolutionary Biology, 22, 2201–2214. Sv˚asand, T. (2004) Why juvenile quality and release strategies are important factors for stock enhancement and sea ranching. In: K.M. Leber, S. Kitada, H.L. Blankenship & T. Sv˚asand (eds) Stock Enhancement and Sea Ranching. Developments, Pitfalls and Opportunities, 2nd edn, pp. 61–70. Blackwell Publishing Ltd., Oxford. Sv˚asand, T., Jørstad, K.E. & Kristiansen, T.S. (1990) Enhancement studies of coastal cod in western Norway. Part I. Recruitment of wild and reared cod to a local spawning stock. Conseil international pour l’exploration de la mer, 47, 5–12. Sv˚asand, T., Skilbrei, O., van der Meeren, G.I. & Holm, M. (1998) Review of morphological and behavioral differences between reared and wild individuals: implications for sea-ranching of Atlantic salmon, Salmo salar L., Atlantic cod, Gadus morhua L., and European lobster, Homarus gammarus L. Fisheries Management and Ecology, 5, 1–18. Ticina, V., Katavic, I. & Grubisic, L. (2007) Growth indices of small northern bluefin tuna (Thynnus thynnus, L.) in growth-out rearing cages. Aquaculture, 269, 538–543. Toates, F. (2004) Cognition, motivation, emotion and action: a dynamic and vulnerable interdependence. Applied Animal Behaviour Science, 86, 173–204. Toresen, R. & Østvedt, O.J. (2000) Variation in abundance of Norwegian spring-spawning herring (Clupea harengus, Clupeidae) throughout the 20th century and the influence of climatic fluctuations. Fish and Fisheries, 1, 231–256. Tsuboi, J. & Morita, K. (2004) Selectivity effects on wild white-spotted char (Salvelinus leucomaensis) during a catch and release fishery. Fisheries Research, 69, 229–238. Tsukamoto, K., Kuwada, H., Uchida, K., Masuda, R. & Sakakura, Y. (1999) Fish quality and stocking effectiveness: behavioural approach. In: B.R. Howell, E. Moksness & T. Sv˚asand (eds) Stock Enhancement and Sea Ranching, pp. 306–314. Fishing News Books, Blackwell Publishing Ltd., Oxford. Uglem, I., Bjorn, P.A., Dale, T., Kerwath, S., Okland, F., Nilsen, R., Aas, K., Fleming, I. & McKinley, R.S. (2008) Movements and spatiotemporal distribution of escaped farmed and local wild Atlantic cod (Gadus morhua L.). Aquaculture Research, 39, 158–170. Vabø, R., Huse, G., Fern¨o, A., Jørgensen, T., Løkkeborg, S. & Skaret, G. (2004) Simulating search behaviour of fish towards bait. ICES Journal of Marine Science, 61, 1224–1232. Vabø, R., Olsen, K. & Huse, I. (2002) The effect of vessel avoidance of wintering Norwegian spring spawning herring. Fisheries Research, 58, 59–77. van den Bos, R., Meijer, M., van Renselaar, J., van der Harst, J. & Spruijt, B. (2003) Anticipation is differently expressed in rats (Rattus norvegicus) and domestic cats (Felis silvestris catus) in the same Pavlovian conditioning paradigm. Behavioural Brain Research, 141, 83–89. Videler, J.J. (1993) Fish Swimming. Chapman & Hall, London. Vilhunen, S., Hirvonen, H. & Laakkonen, M.V.M. (2005) Less is more: social learning of predator recognition requires a low demonstrator to observer ratio in Arctic charr (Salvelinus alpinus). Behavioral Ecology and Sociobiology, 57, 275–282. Wardle, C.S. (1993) Fish behaviour and fishing gear. In: T. Pitcher (ed) Behaviour of Teleost Fishes, 2nd edn, pp. 609–644. Chapman & Hall, London. Ware, D.M. (1971) Predation by rainbow trout (Salmo gairdnieri): the effect of experience. Canadian Journal of Fisheries and Aquatic Sciences, 28, 1847–1852. Waxman, H.M. & McCleave, J.D. (1978) Auto-shaping in the archer fish (Toxotes chatareus). Behavioral Biology, 22, 541–544. Weir, L.K., Hutchings, J.A., Fleming, I.A. & Einum, S. (2004) Dominance relationships and behavioural correlates of individual spawning success in farmed and wild male Atlantic salmon, Salmo salar. Journal of Animal Ecology, 73, 1069–1079.

BLBK374-16

BLBK374-Brown

404

May 27, 2011

11:42

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Welcomme, R.L. (1998) Evaluation of stocking and introductions as management tools. In: I.G. Cowxed (ed) Stocking and Introduction of Fish, pp. 397–413. Fishing News Books, Blackwell Publishing Ltd., Oxford. White, R.J., Karr, J.R. & Nehlsen, W. (1995) Better roles for fish stocking in aquatic resource management. In: H.L. Schramm & R.G. Piper (eds) Uses and Effect of Cultured Fishes in Aquatic Ecosystems. American Fisheries Society Symposium, 15, 527–547. Whitehead, H., Rendell, L., Osborne, R.W. & Wursig, B. (2004) Culture and conservation of nonhumans with reference to whales and dolphins: review and new directions. Biological Conservation, 120, 427–437. Wisenden, B.D. & Harter, K.R. (2001) Motion, not shape, facilitates association of predation risk with novel objects by fathead minnows. Ethology, 107, 357–364. Wisenden, B.D., Vollbrecht, K.A. & Brown, J.L. (2004) Is there a fish alarm cue? Affirming evidence from a wild study. Animal Behaviour, 67, 59–67. Wroblewski, J.S., Smedbol, R.K., Taggart, C.T. & Goddard, S.V. (1996) Movements of farmed and wild Atlantic cod (Gadus morhua) released in Trinity Bay, Newfoundland. Marine Biology, 124, 619–627. Yamashita, Y. & Yamada, H. (1999) Release strategy for Japanese flounder fry in stock enhancement programmes. In: B.R. Howell, E. Moksness & T. Sv˚asand (eds) Stock Enhancement and Sea Ranching, pp. 191–204. Fishing News Books, Blackwell Publishing Ltd., Oxford. Yoneyama, K., Matsuoka, T. & Kawamura, G. (1996) The effect of starvation on individual catchability and hook-avoidance learning of rainbow trout. Nippon Suisan Gakkaishi, 62, 236–242. Young, R.G. & Hayes, J.W. (2004) Angling pressure and trout catchability: behavioral observations of brown trout in two New Zealand backcountry rivers. North American Journal of Fisheries Management, 24, 1203–1213.

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Chapter 17

Cognition and Welfare Lynne U. Sneddon

17.1

Introduction

There has been a tremendous growth in scientific studies addressing fish welfare questions over recent years. These studies provide clear evidence for fish experiencing ‘negative’ welfare states such as pain, fear and stress (Conte 2004; Sneddon 2006; Ashley 2007; Ashley & Sneddon 2007; Sneddon 2009). However, there has also been much controversy and debate as to whether fishes are consciously aware of these complex emotions due to their smaller brain size and lack of a neocortex (Section 17.2; Rose 2002; Iwama 2007). These reviewers also suggest that fishes are stimulus-response beings lacking any thought processing or decision-making; are incapable of any complex learning and memory; and are restricted to simple forms of learning such as non-associative learning and classical conditioning (Rose 2002; Iwama 2007). The preceding chapters in this book have detailed a myriad of advanced behaviours reliant upon recognition of external biotic factors including profitable prey and danger in the form of predatory threat; differentiating conspecifics to identify suitable mates (Chapter 5) and discrimination of kin from non-kin (Chapter 9); adopting adaptive behaviours in others through social learning (Chapter 11); learning to avoid aversive stimuli (Chapters 15 and 16); and recall of navigation routes (Chapter 8). Much of this information must be learned and remembered to make behavioural decisions that improve survivorship and ultimately fitness. Fishes are also capable of modulating their social behaviour through prior experience (Chapter 11) and engage in complicated interspecific and intraspecific relationships that involve cooperation and reciprocation (Chapter 12). The subjects in these social relationships, in most instances, will only engage with others when it is in their own best interest as clearly illustrated by specific examples of manipulation of others (Chapter 13). Such advanced behaviours were thought to occur only in mammals and possibly birds, but due to clever experimental approaches we are now beginning to understand that cognition and so-called higher mental functions do occur in fishes even with their relatively smaller brains and less differentiated cortex (Section 17.2). Current research has shown that fishes have clear preferences for particular items and resources, exhibit mate choice, favour social interactions with related individuals and select the most preferable environmental conditions (Section 17.3). Conversely, fishes also Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-17

BLBK374-Brown

406

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

display strong aversive responses and avoid unfavourable items, conspecifics and habitat conditions (Section 17.4). These studies provide an insight into what choices fishes may make through cognitive testing. Moreover, this review outlines the accumulating evidence that fishes experience pain and fear in a similar manner to other vertebrates (Section 17.5). These findings have significant implications for fish welfare (Section 17.2) and raise a considerable number of controversial questions (Section 17.7). For example, how should fishes be killed in commercial fishing operations? Is recreational fishing morally defendable? Should aquarium owners be licensed? The strict welfare and ethics regulations that are applied to mammals, birds and reptiles have not been routinely applied to fishes. The time is rapidly approaching when legislation regarding fish welfare will need to be revisited.

17.1.1

Fish welfare

Fishes are the third most popular experimental model in the United Kingdom after rats and mice (APC, UK, http://apc.homeoffice.gov.uk/reference/apc-05-26.pdf) and in Australia fishes comprise approximately 80% of all study animals subject to licensed scientific procedures. Indeed, over the last 12 months 10% of all animal behaviour publications in internationally peer-reviewed journals use fishes as model organisms (Brown personal communication). Fishes are an important source of protein with over half a million tonnes of fishes produced by aquaculture in Europe (FEAP, http://www.feap. info/feap/presentations/EUparliament en.asp) annually and approximately 73 million tonne of fishes caught in marine waters globally (FAO 2006). Finally, in developed nations, fishes are the third most popular pet after dogs and cats although they easily outnumber number them in absolute terms (Iwama 2007). Thus, fishes play significant roles in our lives and the way in which we interact with them warrants careful thought if they are capable of experiencing negative affective states and, as a consequence, suffer. Not only should we consider minimising adverse emotional states but we should also seek to cater for increasing positive experiences. For example, what does a fish of a given species want or need within its aquarium to allow it to freely express its normal behavioural repertoire? Cognitive approaches can inform us as to what animals prefer, avoid and what particular internal or mental states are important to the individual. Understanding the subjective and emotional lives of animals should enable us to improve fish welfare. Emotions in fishes are not as apparent as in mammals making such study problematic. Fishes do not audibly vocalise (although they can be heard only with special sound equipment) nor do they have recognisable facial expressions linked to positive or negative affective states. However, this lack of an overt and recognisable response does not preclude the possibility that fishes have emotions. A variety of studies have produced significant data demonstrating fear in fishes (Yue et al. 2004; Ashley & Sneddon 2007; Sneddon 2009). While fishes may lack obvious facial expressions, they do exhibit subtle changes in fin posture and colouration that can be readily linked to a variety of states such as disease, fear and stress (Gibson et al. 2009; Korsoen et al. 2009). Species-specific responses to a potentially painful event have been identified (Reilly et al. 2008), so the welfare context and interspecific variation should be considered. More research is needed to clearly identify and characterise reliable indictors that can be easily assessed visually when determining fish welfare.

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

407

Opinions regarding the status of fish cognition are divided with some authors suggesting that fishes lack cognitive and decision-making processes (Rose 2002; Iwama 2007) and many are unwilling to apply the capacity for experiencing even basic emotions to fishes due to morphological and neuroanatomical differences. In comparison, other aquatic animals, such as cetaceans, are readily accepted as being second to humans in their cognitive ability. The rules of evolution dictate that no function suddenly arises in an animal group without that function evolving in ancestral animals. Therefore, cognitive functions and emotional experiences should be considered on a phylogenetic sliding scale (Bekoff & Sherman 2004); humans in this case may have the most advanced cognition and fishes, in comparison, may be considered more rudimentary, but as we have read in the preceding chapters, certainly not lacking in cognitive power entirely. In Section 17.5, studies demonstrating pain and fear in fishes are discussed in terms of what these data tell us about cognition and welfare.

17.1.2

Preference and avoidance testing

Using preference and avoidance testing, we can gain a better understanding of what animals prefer as well as identifying stimuli that are avoided (Section 17.3). Informed decisions on positive and negative ‘feelings’ or improving welfare can be made based upon the preferences and avoidance decisions made directly by an animal (Bekoff 2006, 2007). There are many examples of fishes in their natural habitat displaying preferences for specific temperatures, oxygen levels, habitat types and other abiotic factors (Fangue et al. 2009; Ludsin et al. 2009; Plumb & Blanchfield 2009). Extrapolating from natural habitat preferences can assist in advising the optimum environmental conditions in which to hold these species in captivity. These conditions can be modified further by using preference tests. Reliable preference data is vital to understand what choices are made by the animal and is indicative of the fact that access to a specific resource is rewarding or beneficial (McMillan & Lance 2004; Balcombe 2006; Bekoff 2006). For example, iguanas prefer a warm environment with no food rather than experiencing cold conditions where food is available (Ramirez & Cabanac 2003). Thus, preference testing provides an insight into what animals want or need and the rank order of importance. In other experimental paradigms, one can examine how much effort they are willing to expend to obtain desirable resources (Carbone 2004). This latter approach involves the animal paying a cost to accessing resources that they value. The effort extended to gain access to the resource can be used as a form of standardised currency. Studies in pigs and mink have used these approaches where, for example, increasing weights were applied to doors that provided access to water. The heavier the weight the more effort must be spent on gaining access to the resource and thus the researcher is able to determine how much the animal will work to obtain access to the water. Similarly, one can train animals to press a lever to gain access to a resource and the number of lever presses required to unlock the door can be progressively increased. This type of testing has not been applied to fishes but obviously could tell us how much a fish would be willing to pay in order to obtain a preferred resource. Avoidance testing, in contrast, demonstrates how important aversive stimuli are to an animal (Section 17.4). If the subject quickly learns to avoid a situation or event, then this is one that may have a negative impact on welfare. The animal may be so highly motivated by this adverse experience that it acquires an avoidance response after just one trial or

BLBK374-17

BLBK374-Brown

408

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

exposure. Examples of such stimuli are nearly always obviously aversive including electric shock or fear-eliciting paradigms. Learning after a single exposure to a potentially damaging event such as hooking during capture (Chapter 16) or predators (Chapter 3) provides clear evidence of the importance of these negative events to the animal.

17.1.3

Behavioural flexibility and intraspecific variation

With fishes being the most diverse vertebrate group, there will obviously be species differences in what fishes need in terms of habitat content as well as differences in how they display negative emotions. Currently, studies on welfare in fishes are limited to a handful of species; therefore, it is not possible to discuss large-scale group-specific requirements such as comparing cyprinids and gadoids with salmonids. Instead, a discussion on intraspecific variation in behavioural reactions highlights the argument against fishes being simply stimulus-response automatons. If fishes do lack the ability to be flexible in their response and use only reflexive behaviour, then it is difficult to explain why we see individual variation or ‘personality’ traits within a species (Wilson et al. 1993; Chapter 7). Personality, in philosophical terms, is not restricted to humans but is defined as the characteristics of an individual that makes the individual distinct from another and thus recognisable. When applied to fishes, personality is often expressed in multiple dimensions such as boldness, the propensity to take risks in the face of novel challenges (e.g. Brown et al. 2005; Frost et al. 2007). Fishes in wild and captive situations often have a dichotomous distribution consisting of individuals showing bold, risky behaviours or conversely shy, cautious behaviours. These intraspecific differences are also discussed with relevance to welfare and cognition in Section 17.6.

17.2

What is welfare?

Animal welfare has a number of definitions depending upon how one perceives what good welfare is. The meaning and definition of animal welfare and how best to objectively measure it are subject to much discussion among scientists (Broom 1991a, 1991b; Dawkins 1998a; Mendl & Paul 2004; Broom 2007). There are three main definitions: (1) Good biological functioning is an accepted means of measuring welfare and unquestionably provides scientific data regarding the physical condition of the animal. Poor health can be caused by, and may be a consequence of, suboptimal welfare, which can be reliably quantified. (2) The ‘feelings’-based definition of animal welfare goes beyond physical parameters and includes a psychological component. This assumes that animals are sentient and have subjective experiences and so can experience suffering on an emotional level (Broom 1991b). (3) Finally, ‘nature’-based concepts of animal welfare suggest that animals should be able to freely perform their natural behavioural repertoire. It is impossible to know if animals are aware of their emotions and consciously suffer since one would have to be an animal and know exactly how it feels. Indeed, we only know

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

409

how other humans feel because they are able to communicate their experience. Therefore, many welfare scientists recommend that we should apply the precautionary principle and that fishes should be given the benefit of the doubt and treated as if they do suffer (Bateson 1991; Broom 1991a, 2007; Bekoff & Sherman 2004; Mendl & Paul 2004; Balcombe 2006; Arlinghaus et al. 2007, 2009; Bekoff 2007; Sneddon 2009). However, we can scientifically and objectively measure responses to a painful event and use analgesics to return behaviour back to normal (Sneddon 2003a; Sneddon et al. 2003a) to prove that there are deleterious changes in fish behaviour that are alleviated by the administration of analgesia. There is growing evidence from scientific studies that make a robust case for fishes perceiving pain and that this is an important and detrimental state for them to experience (Sneddon 2009). Well-being is a term often used interchangeably with welfare and can be defined as the state or condition of being well and content. From a well-being perspective, animals should be disease free, kept in optimum conditions (if we have identified what these are) and be free from pain or suffering. This not only requires the use of analgesia and anaesthesia to minimise pain during invasive procedures but also being able to assess pain in animals.

17.2.1

Sentience and consciousness

For animals to have poor welfare or well-being, it is assumed they are at least sentient beings. Sentience and consciousness have many definitions due to their complex nature and are readily debated amongst scientists and philosophers. Here, I define sentience as the ability to detect and respond to external stimuli and having an awareness of pain, fear and stress (Dawkins 1998b; Broom 2007). Consciousness can be defined as an internal mental image and a sense of ‘I’ and how ‘I’ relate to the world (Beckoff & Sherman 2004). A variety of studies have demonstrated that animals can detect and respond to painful stimuli (e.g. receptors, nociceptors that can detect noxious stimuli with subsequent changes in behaviour and physiology; Sneddon 2002; Sneddon 2003b; Sneddon et al. 2003b; Ashley et al. 2006, 2007, 2009; Reilly et al. 2008), but the crucial point is whether these animals are conscious of the painful stimulus (i.e. do they know that they are in pain and hence suffer?). Do animals have conscious thoughts where they relate to their own experience and think about this in their minds? Some suggest that non-primate animals lack the cognitive abilities of conscious beings because human conscious thought arises in the neocortex and, therefore, animals that lack a neocortex do not consciously experience the affective states of pain (Rose 2002). If one accepts this opinion, then this means mammals such as dogs, cats, rodents as well as birds and amphibians are also not aware of the negative affective component of pain yet there is a plethora of published studies demonstrating that pain is an adverse experience for these animals (Gentle 1992; Flecknell et al. 2007). However, this opinion defies the laws of evolution that suggests that each function has an evolutionary history that can be traced back in related taxa (Bekoff & Sherman 2004; Bekoff 2007). Moreover, animals with a completely different life history, ecology and evolutionary trajectories may evolve the same functions in completely different areas of the brain. This concept has already been demonstrated in the avian and fish brains (Jarvis et al. 2005; Chapter 16). Therefore, it may not be valid to attribute a function such as being consciousness to a particular brain area when comparing humans with other animals (Molyneux 2010).

BLBK374-17

BLBK374-Brown

410

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

It is impossible to measure emotion directly in any animal (including humans); therefore, the indirect evidence used to establish conscious motivational affective states must come from the study of neuroanatomy, neurophysiology and particularly behaviour (Duncan 2002; Sneddon 2004, 2009).

17.2.2

Cognition and welfare

Specific types of behaviour are thought to be indicative of an animal’s ability to form internal representations and act upon these depictions of its internal and external environments. This state of basic or rudimentary consciousness is only thought to have been achieved in species that have nervous systems that have attained a sufficient level of complexity during evolution (Shettleworth 2001). Therefore, cognitive studies of an animal may be used to assess consciousness and sentience. Self-recognition is one of the key criteria for consciousness. There is evidence that fishes can indeed recognise themselves and discriminate themselves from others, which was previously thought to be restricted to mammalian species. The cichlid, Pelvicachromis taeniatus, recognises its own odour and prefers this over the odour of other fishes regardless if they are familiar or related (Thunken et al. 2009). To be able to engage in self-assessment of one’s own status and make behavioural decisions based upon this, fishes must be able to compare themselves to others. Evidence from studies where fishes watched conspecifics fight suggests that fishes are capable of making third-party assessments of themselves relative to others since their performance is affected when faced with the victor or loser of the observed interaction (Oliveira et al. 1998). These data suggest that fishes are capable of self-recognition and self-assessment, which are central concepts in higher cognitive processing and consciousness. If fishes have cognitive functioning to such a degree that they exhibit clear preferences, avoid aversive stimuli, use tools (Pasko 2010), are able to learn complex tasks and have long-term memory (Chapters 8, 11, 15 and 16), then this infers that they have some form of internal decision-making process as well as the capacity to remember negative events that impact upon their welfare. When considered in conjunction with the emerging evidence from behavioural studies that fishes are likely to have some form of consciousness and, therefore, may suffer on an emotional level, it is apparent that fish welfare becomes an important issue. A greater understanding of the cognitive functioning of fishes provides us with insight into what they may experience when their welfare is compromised. While further research is clearly required, it is apparent that fishes have the capacity for pain and suffering and this ought to be addressed whenever it is encountered. Therefore, it is vital to minimise or avoid procedures that cause pain, fear or stress to fishes to ensure good welfare. However, positive welfare should also be enhanced rather than just trying to minimise aversive events. Providing the ideal habitat in captivity might be a first step in addressing these issues and determining what fishes need can be revealed by using preference tests (Section 17.3).

17.3

What fishes want

Concepts such as the five freedoms proposed for farmed animals are usually applied to all captive situations where animals are held by humans (FAWC 1996). These are (1) freedom

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

411

from hunger and thirst, (2) freedom from environmental challenge, (3) freedom from disease, fear, pain, injury and discomfort, (4) freedom from behavioural restriction to allow expression of the normal repertoire and (5) freedom from mental suffering. Therefore, when applying these freedoms to fishes, we require an understanding of how they are affected when they are subject to these factors. Fishes are undoubtedly capable of making decisions in a variety of contexts and this may involve cognitive processing to varying extents. The preference approach provides a reliable indicator of what a fish wants by means of exposing them to a choice test. This is based upon the premise that a sentient animal would choose a beneficial option over one that is less beneficial or has a negative impact upon welfare. Therefore, the choice they make presumably has a perceived benefit over the alternative. However, this assumption may not always hold true. When given a choice between fruit and chocolate, for example, a child may choose chocolate even though it is not the most appropriate dietary item. Caution should be taken since the preference depends upon what choices are presented to the fish and could be significantly affected by internal or physiological state at that time. Carefully designed experimentation should account for individual variation in sex, age, breeding condition and so on and also to pair specific resources with one another so that a hierarchy of needs can be ascertained. Preference tests must be designed to have biological relevance and, owing to the idiosyncrasies of various species, extrapolation between species is not always advisable. For example, a highly territorial species such as rainbow trout, Oncorhynchus mykiss, thrives in isolation from others whereas a sociably species such as common carp, Cyprinus carpio, do not fare well when kept alone. Large groups of schooling or shoaling species, such as the zebrafish, Danio rerio, show little aggression but when held in smaller groups of four or less, significant amounts of aggression results in stress in subordinate individuals. Therefore, social groupings must be carefully differentiated between gregarious and territorial species (Wirtz & Davenport 1976; Volpato et al. 2007). Therefore, with careful design preference data can be extremely valuable in understanding the subjective needs of fishes and can be applied to captive husbandry in terms of space requirements and usage, diet and time of feeding, light regime and intensity, oxygenation, temperature, social context, water quality and so on.

17.3.1

Preference tests

17.3.1.1

Physical habitat

There are many studies demonstrating preference for favourable environmental conditions in a variety of fish species. Using observations from the wild environment, Sessa et al. (2008) designed an experiment to alter depth gradient in tanks of reproducing zebrafish to reflect natural conditions. Natural populations of zebrafish spawn in the shallows of water bodies, which contrasts with captive conditions where they are expected to spawn in the deepest area at the bottom of a tank. By providing a depth gradient, mating behaviour was unaffected but ovoposition did occur in shallower depths with a significant increase in the number of embryos deposited and surviving compared with standard conditions (Sessa et al. 2008). This demonstrates that provision of a depth gradient can improve reproductive success. Other examples include laboratory-held weakfish, Cynoscion regalis, who only

BLBK374-Brown

412

17:37

Copyeditor’s Name:

* Percentage of fish

Percentage of fish

May 13, 2011

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

100 75 50 25 0

1 vs. 2

2 vs. 4

4 vs. 6

100 90 80 70 60 50 40 30 20 10 0

Choice of dissolved oxygen (mg/L)

a

Under shelter On shelter Out of shelter

a

b b b

b

b b

122 dpf

137 dpf

157 dpf

(b)

(a) 1400 Bivalve-fed fishes

1200

Algae-fed fishes

1000

*

500

800

400

600

Time (s)

Time (s)

BLBK374-17

400 200

*

300 200 100

0 Low 10–13

Intermediate 16–19 Temperature (°C)

(c)

High 25–30

0 5W

5B

5B

1B

(d)

Fig. 17.1 (a) The percentage of weakfish choosing the higher dissolved oxygen when given a choice between 1 versus 2 mg/L (1 vs. 2); 2 versus 4 mg/L (2 vs. 4) and 4 versus 6 mg/L (4 vs. 6). (Adapted from Stierhoff et al. 2009; * P < 0.05.) (b) Mean ± SD percentage of Arctic charr under, on or out of shelter during development expressed in days post-fertilisation (dpf), significantly different fish locations (P < 0.05) are represented by a different letter above the bar (Benhaim et al. 2009). (c) Mean amount of time(s) ±1 SE that intertidal fish fed on low-nutrition algae (black bar) and on higher nutrition bivalves (white bar) stayed at low-, intermediate- and high-temperature cells of the thermal gradient. Total duration of each trial (fish) was 60 minutes, and 11 replicates were conducted per experimental group (Pulgar et al. 2003). (d) Mean amount of time(s) ±1 SE that individually housed brown striped female fighting fish spent when presented with a choice of five white fishes (5W) versus five brown fishes (5B) and a choice of five brown fishes (5B) versus one brown fish (1B). (* P < 0.05.) (Adapted from Blakeslee et al. 2009. All figures are reproduced by kind permission from Elsevier.)

avoid low oxygen concentrations that negatively affect their growth rate and demonstrate no preference between oxygen levels that appear to have no impact on their physiology (Stierhoff et al. 2009; Fig. 17.1a). In broad-nosed pipefish, Syngnathus typhle, brooding males chose to spend more time in higher temperature areas of an aquarium than females, the latter of which exhibited no preference (Ahnesjo 2008). This behavioural temperature preference could be linked to increased male brooding rate. Reproductive status also affected temperature preference in Japanese eels, Anguilla japonica, where they choose to spawn in temperatures of 18◦ C–22◦ C in captivity, similar to ambient temperatures, when given a choice ranging between 14◦ C and 27◦ C (Dou et al. 2008). When given a range of salinities, temperatures and substrates to choose from, the mudskipper, Boleophthalmus pectinirostris, exhibited profound preferences for approximately 31◦ C, salinity of 5 ppm and sandy mud substrate which governs this species’ choice of microhabitat (Chen et al. 2008).

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

413

Habitat composition can have dramatic effects on the biological functioning of fishes with positive implications for welfare. Arctic charr, Salvelinus alpinus, when reared from birth with a shelter in their tanks exhibited better growth, lower mortality and commenced first feeding about 6 days later compared with fishes without shelter (Benhaim et al. 2009; Fig. 17.1b). This study also adopted preference tests during development and found that a higher proportion of fishes used the shelter rather than being visible in the open. The animals in these studies based their choice upon the relative value of the resources to which they were exposed. The animal must judge its preferred choice to be better than the alternative and this provides some insight into what fishes want and need in their environment. These studies provide clear preferences that are species specific and that can be used to inform guidelines as to the captive husbandry of fish species in terms of physical environmental and ecological conditions. 17.3.1.2

Breeding

Many animals including fishes construct nests for breeding purposes. For example, male sticklebacks build elaborate nests in the breeding season and females use this to judge male quality and to decide which male to mate with (Rushbrook et al. 2008). The female lays her eggs within the nest and the male fertilises them and performs parental care until the eggs hatch. Many captive fish species in aquaculture and experimentation are not provided with such material since in vitro fertilisation is used. However, this possibly does not fulfil the behavioural needs of the fishes if animals are highly motivated to build nests and leads to frustration, stress and associated maladaptive behaviours. In other intensively farmed animals such as pigs and chickens, providing females with nesting material had a dramatic improvement on their behaviour and well-being (Cronin et al. 1998; Kruschwitz et al. 2008). Yet these materials are generally not provided to captive fishes even though they may have a positive impact upon behaviour and welfare. Indeed, the first recorded spawning behaviour of the round goby, Neogobius melanostomus, was observed in captivity by providing males with nesting material (Meunier et al. 2009). When nesting material was available, the male fishes spent a substantial amount of time engaged in nest building, courtship behaviour to attract females and parental care. By applying knowledge of the natural behaviour of fishes, such as the requirement for nest building for reproduction to take place (Galhardo et al. 2009), significant improvements to the captive environment can provide what fishes need to express their natural behavioural repertoire. This infers that welfare could be improved via a reduction in abnormal behaviours linked to frustration from being unable to perform natural behaviours. However, to date there have been no studies specifically examining this in fishes. Future studies should investigate whether the welfare of captive species is enhanced by providing suitable substrates and materials necessary for mating to occur. 17.3.1.3

Diet

Dietary preferences are another way of ensuring that the appropriate nutrition and foodstuff is given to fishes. Fishes can be herbivores, carnivores or omnivorous (see Chapters 2 and 3); therefore, provision of the correct diet is vital if we accept that normal biological functioning

BLBK374-17

BLBK374-Brown

414

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

equates to good welfare. Diet can affect behaviour and preferences in fishes; therefore, the animal’s hunger level must be accounted for when designing such experiments. In the intertidal fish, Girella laevifrons, food quality had a major impact on temperature preference. A highly nutritious diet resulted in these fishes selecting temperatures of 16◦ C–18◦ C whereas a lower quality diet resulted in fishes choosing lower temperatures of 10◦ C–12◦ C (Pulgar et al. 2003; Fig. 17.1c). These results were explained by optimisation of digestion and mechanisms of energy conservation in the low-quality diet treatment. In rainbow trout, the lack of vitamin C in the diet appears to lead to poor competitive ability and as such these fishes are mainly seen occupying the top layer of the aquarium tanks (Blom et al. 1999). Supplementation of the diet with vitamin C resulted in these fishes occupying the preferred lower areas of the tank possibly through an increase in competitive status. When operating self-feeders, goldfish, Carassius auratus, were given the choice between three diets differing in macronutrient composition (Sanchez-Vasquez et al. 1998). Goldfish actively chose the diet highest in carbohydrate and fat composition and selected against the high protein diet despite the fact that this may not be the ‘healthiest’ diet. These examples illustrate how easy it is to determine dietary preferences in fishes and that improper diet can cause significant shifts in behaviour. They also provide more evidence that fishes can make choices by actively evaluating resources. One area that has been neglected in studies of fish feeding is time spent foraging under natural conditions versus how they are fed in captivity. If a particular species spends most of its daily time budget foraging, does providing it with one or two feeds in a day in captivity result in frustration? This has been observed in stabled horses which are fed monotonous food stuff once to three times per day. In contrast, feral horses will spend up to 16 hours in a day foraging and individuals are highly motivated to perform this behaviour. Therefore, stabled horses exhibit signs of frustration and stereotypical behaviours, which can be significantly reduced by altering the means and type of food provision. For example, hiding food in different areas of the stable results in more time spent foraging accompanied by a reduction in abnormal behaviour (Ninomiya et al. 2004). The motivation to forage for long periods may have complicated effects on captive fishes but these have yet to be elucidated. Another factor to consider is whether live prey should be provided to predatory fish species. If individuals are highly motivated to hunt, then perhaps providing dead food is inadequate. Feeding of live invertebrates to fishes is generally considered acceptable and such food is available at most local pet shops; however, feeding of live vertebrates such as smaller fish is frowned upon. Research in large cats, such as tigers and servals, has shown improved behavioural indicators in captivity by providing these predators with hunting opportunities using artificial prey (Markowitz & LaForse 1987). Future research should target these questions since they may provide evidence of how fishes are affected cognitively by the lack of feeding opportunities and as such how their welfare is compromised. 17.3.1.4

Social interactions

Fishes range from being territorial to highly gregarious; therefore, their preferences for being in close proximity with others or with related or unrelated individuals may vary according to the natural behaviour of a given species. The social context in which animals are housed, as mentioned in Section 17.3, is vitally important since it can have a negative

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

415

effect on fishes of low-social status in aggressive or territorial species when housed in groups (Gilmour et al. 2005). Salmonid behaviour has received much attention since they are a commercially important aquaculture and fisheries species. Holding salmonids in a group results in the formation of a dominance hierarchy, where low-status fishes suffer acute to chronic stress that can have deleterious effects upon growth and reproduction (Gilmour et al. 2005). In terms of biological function, this has a negative impact upon welfare, so holding this species in captivity can be problematic due to aggression (Conte 2004). In the natural situation, fishes are not confined to a relatively small area and can withdraw from aggressive interactions. Equally, holding gregarious species in isolation may result in stress. Therefore, it is important to fully understand the social context of the natural behaviour of each fish species so that the correct decisions are made regarding stocking densities, as well as the composition of group members. Again, this can be explored using preference tests. In shoaling species, specific social preferences are known to exist where shoal composition can be affected by relatedness, sex, age, dominance status, personality and phenotype of polymorphic species (see Chapter 10). Therefore, it may be wise to better understand the natural composition of shoals before placing fishes into groups in captivity to promote positive welfare. For example, gender may have implications for the welfare of group members where females are harassed by males for breeding as observed in guppies. This harassment is believed to diminish female fitness via reduced foraging, augmenting predation risk, energy costs and disease transmission (Smith & Sargeant 2006). Therefore, female guppies may have better well-being when housed only with females. However, in the western mosquitofish, Gambusia affinis, male harassment had no negative effects on female growth or fecundity but increasing the density of females whilst reducing male density had significant detrimental effects on female fitness (Smith & Sargent 2006). These contrasting results from two closely related species highlight the important of species-specific requirements. Therefore, future studies need to provide information on a variety of species to fully understand how group composition affects welfare. Theory predicts that shoals should be composed of similar individuals and that any individual that differs from the norm makes the shoal more conspicuous thereby increasing predation risk (Gomez-Laplaza 2009). Female Siamese fighting fish, Betta splendens, show significant preference for spending time near females rather than being in a chamber on their own (Blakeslee et al. 2009; Fig. 17.1d). These females also prefer to shoal with similarly coloured females. However, placing females with males can result in high levels of aggression and females are frequently killed by potential suitors. Such assortative shoaling has also investigated in juvenile angelfish, Pterophyllum scalare, where individuals of the uniformly black and golden colour morphs were held in groups with conspecifics of similar and dissimilar body colours to themselves, as well as in mixed-colour groups. These fishes were given a binary choice to shoal with a group of conspecifics composed of unfamiliar fishes of either a similar or dissimilar colour phenotype to themselves. Fishes from the similar- and mixed-colour groups showed a significant preference for the similar shoal; however, those fishes in dissimilar groups showed no preference (Gomez-Laplaza 2009). Thus, previous experience affected assortative shoal choice and it is important to consider past housing history when interpreting the results of conspecific preference tests.

BLBK374-17

BLBK374-Brown

416

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Captivity does present a rather different environment with contrasting selective pressures. For example, since there are no natural predators in laboratory situations, this strong selective force in nature has been removed and may affect the antipredator advantages to shoaling. In rainbowfish, Melanotonia duboulayi, captive individuals show no preference for familiar shoal neighbours yet wild population show a strong preference for familiar individuals (Kydd & Brown 2009). This may reflect a relaxed approach to shoal composition in captive fishes when predator threat is negligible. In contrast, wild populations may need to create tight cohesive shoals, which would be enhanced by forming shoals with fishes that are familiar with one another since each fish would know its place within the shoal. Therefore, when considering the results from captive-reared fishes, it is important to consider whether motivational drivers have been removed under artificial conditions either via habituation or via artificial selection. For example, providing food daily may reduce the need to express foraging behaviour since the fish never experiences hunger or has to search over large distances for food. This question needs to be fully explored by careful experimentation to determine how captivity affects the well-being of fish and how it may impact upon their cognitive function. Very few studies have explored whether conspecific choice in territorial fishes under captive conditions could be relevant to improving welfare. Some species have a preference for forming groups, breeding cooperatively or sharing refuges with related individuals whilst others selectively avoid kin (see Chapter 9). Brown trout, Salmo trutta, are naturally territorial fishes and show a strong preference for stream water containing no conspecific cues over that containing conspecific cues. When juveniles were given a choice between water scented with siblings or non-siblings, the results were highly variable but most individuals showed kin avoidance (Ojanguran & Brana 1999). This may be adaptive in territorial species so that they do not compete with closely related individuals. In contrast, juvenile brook trout, Salvelinus fontinalis, show a strong preference for the odour of kin compared with non-kin in preference tests (Hiscock & Brown 2000). In Atlantic salmon, Salmo salar, kin preference was tested by holding pairs of related and unrelated individuals with either recirculating water to increase the concentration of chemicals cues involved in kin recognition or a flow-through system where chemical cues were removed. The pairs quickly formed a dominance relationship; however, water recirculation heightened aggression, especially against unrelated fishes (Griffiths & Armstrong 2000). Therefore, water recirculation appears to have a negative impact on the welfare of the subordinate fishes if they are unrelated and, in this case, a flow-through system was beneficial in reducing aggression. Research into methods of reducing aggression in territorial species is necessary to improve conditions for these species. Placing fish into novel, unfamiliar groups may promote aggression or stress; therefore, this should be considered when moving fish between aquaria.

17.4

What fishes do not want

Considering what an animal actively avoids is a useful way to understand what stimuli or experiences a fish would seek to steer clear of. These must be important and, therefore, have negative consequences for the individual. Fishes in classical conditioning experiments with

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

417

negative reinforcement usually learn to avoid an aversive stimulus in a few trials or less (e.g. electric shock that may be painful, Ehrensing et al. 1982). Zebrafish can be trained to remain in a dark compartment of a shuttle box to avoid a weight dropping into the vicinity causing the fish to display a startle response (Kim et al. 2009) and show avoidance behaviour at the sight of an animated predator (Gerlai et al. 2009). Fishes also show aversion to unpalatable food demonstrated by studies in the medaka, Oryzias latipes, where transgenic fishes with a lack of taste receptor genes fail to show aversion (Aihara et al. 2009). Many studies have used electric shock as an aversive stimulus that would be painful to humans. Goldfish demonstrate a prolonged unwillingness to enter an area where they received a shock; however, when this area also contained food, eventually hungry fishes were willing to risk being shocked in order to eat (Millsopp & Laming 2008). Even though the fishes were still given a shock in the area, they traded off the acute pain caused by the shock to get a resource that is crucial for survival. Therefore, these hungry goldfish were willing to pay a cost in this case to gain access to a highly valued resource. The fishes may also have learned that the shock is a very short stimulus and not life-threatening, and therefore, made the decision based upon another important motivational state, hunger. Studies on wild populations have provided many cases of fish avoiding certain stimuli, which may be used to discourage fishes from occupying areas where they are not wanted. In dams and reservoirs, where fishes are damaged in equipment that impairs their welfare but also causes problems for humans, these avoidance approaches can be employed to deter fishes from being injured. Such an approach was used in a study on vendace, Coregonus albula, where continuous artificial light was employed to prevent fishes from aggregating (Schmidt et al. 2009). Vendace exhibited strong avoidance behaviour by swimming downwards when the light was turned on. With the careful deployment of artificial lights they stopped congregating in problem areas of the reservoir. Similar approaches are being adopted where acoustic repellents are employed to prevent sharks getting entangled in nets and to prevent fishes entering power station coolant intakes (Maes et al. 2004). Thus, understanding what fishes do not want or avoid can be a useful tool in the management and improved welfare of natural populations. Avoiding such aversive stimuli as a management approach during captivity should also improve well-being. Welfare studies need to determine what routine stimuli that fishes are exposed to in captivity (e.g. disturbance during cleaning, vibration, excessive noise, etc.) would normally elicit an avoidance response and use this information to develop better husbandry procedures. Behavioural observations made during these events could determine if there are any detrimental effects upon the fishes in terms of suspension of normal behaviour coupled with avoidance testing to obtain an insight into whether these experiences are aversive to the fishes.

17.5

Pain and fear in fish

The key question in negative emotional states is whether fishes consciously experience them. Do fishes suffer when damaged? Does a fearful situation result in mental suffering in fishes? These are difficult questions since they require knowledge of how fishes feel. Since we cannot tell what another human is feeling unless they communicate it to us, how are we supposed to know how an animal feels? Rather than addressing these questions

BLBK374-Brown

418

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

directly, behavioural observations and cognitive approaches have been adopted to understand how significant these negative affective states are to the animal. In other words, fishes can give us some indication of how they feel by altering their behaviour in some quantifiable way. So far, it has been demonstrated that teleost fishes possess nociceptors, receptors that preferentially detect painful stimuli; have pathways from the periphery to the brain; the brain is active during painful stimuli; the fishes display adverse changes in behaviour and physiology indicative of suffering that are ameliorated by morphine; and can learn to avoid painful events like electric shock and hooking usually in one trial (reviews in Sneddon 2004, 2006, 2009; Chapters 15 and 16). Possible anticipation and learned avoidance of painful events may demonstrate that fishes will react strongly to evade these behaviours or stimuli that will result in an aversive or painful state. This suggests some level of consciousness in fish since the fish must have such a strong negative experience associated with these events that they are motivated to avoid them often after one exposure. One can prove that fish can detect, react and show complicated, prolonged behavioural changes that are not simple reflexes, but are these indicative of how important pain is to them. Using selective attention strategies based upon the idea that individuals have a limited pool or capacity to their attention may provide a means of gauging the significance of pain. If the fish’s attention cannot be diverted away from pain-related responses to apparently important competing stimuli, then pain is more important. When given a pain stimulus, rainbow trout did not show an appropriate fear response to novel object testing whereas controls exhibited a significant neophobia. Administering morphine resulted in pain-treated fishes returning to a normal neophobic response (Sneddon et al. 2003b). This suggests that pain was dominant over fear with respect to the attention of the fish. Similarly, when fishes experiencing pain were given a predator cue, they did not show the typical antipredator responses that the controls performed, i.e. increased escape attempts and increased refuge use (Ashley et al. 2009; Fig. 17.2a). Again, pain takes priority over diverting attention to predation, which would be detrimental in a natural context. Interestingly, dominant trout in

80 Refuge Escape 40

**

20 0 –20 Control

Acid

Frequency of chases

120

**

60 Change (%)

BLBK374-17

Normal Treatment

100 80

**

60 40 20 0 CF

–40 (a)

CU

AF

AU

(b)

Fig. 17.2 (a) The mean percentage change in refuge use and escape behaviour in rainbow trout that were injected subcutaneously with saline as a control or acetic acid. (b) The mean frequency of aggressive chases performed by dominant rainbow trout before (normal) and after (Treatment) injection with saline and placed into a familiar (CF) or unfamiliar (CU) social group or injected with acid. (AF, familiar; AU, unfamiliar; ** P < 0.001.) (Modified from Ashley et al. 2009. Copyright 2009, with permission from Elsevier.)

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

419

social groups exhibited none of the physiological signs of pain (enhanced ventilation rate, increased cortisol), but when returned to their familiar group they decreased the amount of aggression over three hours. By contrast, when dominants were returned to a novel group, no suspension in aggressive behaviour was seen whilst they were in pain (Ashley et al. 2009; Fig. 17.2b). Given the importance of dominance status in trout which governs access to resources and results in enhanced fitness, it would seem exerting dominance is more important than exhibiting signs of pain. Studies on birds and mammals have demonstrated that they do not show signs of pain in novel situations or when conspecifics are present and this has been interpreted as the animals being motivated to avoid showing signs of weakness in a risky situation (Gentle 2001; Arras et al. 2007). Therefore, in the trout study, it may be that the presence of unfamiliar conspecifics results in noxiously treated trout refraining from exhibiting any signs of pain in order to maintain their social status. Together, the trout studies confirm that the trout’s behavioural responses to pain are indicative of some form of discomfort and suffering and are not the simple reflex responses that critics claim. Species-specific differences have been identified where trout and zebrafish show an elevated ventilation rate and reduced activity in response to a standard pain test, but common carp do not. Carp and trout show anomalous behaviours such as rubbing the affected site and rocking from either pectoral fin (similar to fishes attempting to maintain an upright position); however, zebrafish do not (Reilly et al. 2008). Collectively, these studies indicate that rather than being only able to show stimulus-response behaviour, different species of fishes are capable of prolonged, complex responses suggesting that they experience the negative affective component of pain which is distracting enough to prevent them from performing other behaviours. Emotional conditioning using fear paradigms are well documented in fishes (see Chapter 13). Fear behaviours such as the startle response, freezing, escape and so on can easily be measured. Very few of these studies have been applied to an animal welfare context or have explored the negative feelings associated with fear. Fear responses in rainbow trout have been investigated using classical conditioning with negative reinforcement (Yue et al. 2004). Here, fish associated a light cue with an aversive chase by a plunging net. After training, fish responded to the light cue before net presentation by swimming to another compartment to avoid being chased. This experience was significant to the fish since memory recall was demonstrated after seven days. Other studies have demonstrated escape responses lasting for approximately 11 months in rainbowfish (Brown 2001). Using this approach it may be possible to quantify the motivation of the fish to avoid a fearcausing stimuli by using the length of memory recall to gauge how important the event was. Caution has to be used here since it has been suggested that true decision-making processes or very high-level cognitive processes are not involved in classical conditioning (Rose 2002). However, subsequent behaviour is affected by these training paradigms. If fishes are able to anticipate negative events and produce a response prior to it occurring, then this is evidence of cognitive processing rather than a reflex response. Experiments using classical conditioning between an aversive stimulus and a neutral cue should be designed to carefully differentiate between those behaviours motivated by the animal’s affective state or higher cognitive functions rather than the stimulus response processes of associative learning.

BLBK374-17

BLBK374-Brown

May 13, 2011

420

Fish Cognition and Behavior

17.6

Personality in fish

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Personality is often applied to humans; however, the philosophical and psychological definition is the possession of traits that characterise one individual from another (Eysenck 1946) and can readily by applied to fishes (Chapter 7). One of the most commonly studied axes of fish personality is the shyness–boldness continuum. Boldness can be measured as the willingness to take risks when encountering a novel challenge (e.g. Frost et al. 2007). These divergent phenotypes also exist in natural populations. Brook charr collected from the wild continued to exhibit bold and shy personalities in the laboratory with bold fishes performing more exploration and activity in a novel environment. However, in response to a fear test, a uniform startle response was observed that was independent of personality type (Wilson & McLaughlin 2007). Therefore, the responses to threatening stimuli was not linked to personality in this species. However, in studies on pain bold fishes appear to recover more quickly compared with shy fishes and in the antipredator experiments bold fishes experiencing pain actually decreased their use of refuges after a predator cue was presented which is contrary to the responses of control fishes (Ashley et al. 2009). Whether this demonstrates a real difference in cognitive processing between the two phenotypes or is reflective of how they cope with stress is yet to be established. Behavioural needs can be considered as those resources in an animal’s environment that allow the expression of a normal behavioural repertoire (Jensen & Toates 1993). In terms of behavioural needs and welfare, bold fishes may have different requirements to shy fishes, especially in species where boldness correlates with aggression. In captive experiments bold trout generally dominate shy trout (Frost et al. 2007); therefore, shy fishes are likely to become stressed if overt aggression is continually used by bold, dominant fishes (Gilmour et al. 2005). Since faster growing fishes have been selected for in aquaculture, this strategy has co-selected bold, aggressive fishes which can present problems for any subordinates in terms of chronic stress (Huntingford 2004). This difference in dominance status also affects captive fish behaviour and use of substrate in the Mozambique tilapia, Oreochromis mossambicus, where dominant males preferred a soft substrate as opposed to no substrate but subordinate males did not show such a preference (Galhardo et al. 2009). Dominant males are likely to secure matings and use the substrate as nesting sites; therefore, the provision of a substrate may have positive implications for breeding male welfare. Fishes do have the capacity to modulate their behaviour and make an adaptive response in different contexts, which is suggestive of higher cognition – the ability to make decision dependent upon an evaluation of external factors (Chapter 7).

17.7

Wider implications for the use of fish

Considering the evidence for cognitive ability and awareness in fishes, for their capacity to respond to and learn to avoid negative events such as pain and fear, and for their ability to choose between resources so that clear preferences can be identified, this suggests that fishes should be considered capable of experiencing poor welfare states and that these should be minimised. Even if some doubts remain which may be waylaid by future research, the precautionary principle should be adopted as best practice.

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

421

Fishes are subject to practices that would be unacceptable in other vertebrates, yet many more individuals are harvested in aquaculture, recreational and commercial fisheries than in production for terrestrial meat, poultry and dairy products. Fishes are also subject to invasive procedures in scientific experimentation although there are regulations protecting their use including humane guidelines in most countries. Finally, fish can be purchased as a companion animal by members of the public who may have no experience in fish husbandry. These issues are discussed in the following sections with reference to fish welfare and cognition.

17.7.1

Aquaculture

Many practices associated with aquaculture such as high stocking density, transportation in confined vessels and tanks, food withdrawal prior to slaughter and slaughter itself also result in stress and are considered problematic from a welfare perspective (Ashley 2007; Table 17.1). Farmed fishes are generally held in a very simple, monotonous environment in high densities (Chapter 16). They are not provided with the opportunity to select their habitat, food, mates or perform their normal behavioural repertoire such as nest building. High stocking density can lead to high transmission of diseases that are necrotic and cause tissue damage that may be painful. All of these factors are known to lead to abnormal behaviours, such as functionless, repetitive stereotypies, in farmed and captive mammals (Jensen 2009). Yet relatively little information is available as to whether fishes are detrimentally affected by many of these procedures. Questions regarding the cognitive functioning of farmed fishes are rare so future studies should tackle the impact of aquaculture procedures upon the behaviour and welfare of these fishes. Farmed fishes are also subject to unpredictable, stressful disturbances such as vaccination, size grading, cleaning and movement between tanks (Conte 2004; Ashley 2007). These are likely to impair welfare and increased stress or mortality has been recorded after these events. This also results in frequent changes in social composition within the tank environment. The impact of being unable to associate with preferred or familiar individuals in captivity needs to be investigated since high stocking density reduces normal dominance hierarchy formation. This could result in increased aggression when fishes are competing for dominance status as seen in pig farming. Mixing of unfamiliar pigs results in substantial overt aggression and injuries. Simple practices such as providing ad libitum food to reduce competition (Barnett et al. 1994) and mixing pigs of different sizes and thus aggressiveness substantially reduce these problems (Erhard et al. 1997). To resolve these issues of subjecting fish to uncontrollable and unpredictable stressful events, stimuli could be introduced to indicate when aversive stimuli are about to occur. For example, a tone or light could turn on to pre-empt common procedures such as cleaning. If these events are more predictable, it enables fishes to prepare themselves for the impending event and may reduce stress levels. Other strategies, such as using an avoidance response to facilitate fishes to move of their own accord, have been successful. Trout avoid carbon dioxide as demonstrated by a fish farm based study where increased dissolved carbon dioxide levels resulted in fish swimming through a pipe to the neighbouring tanks (Clingerman et al. 2007). This approach was extremely successful and avoided any mechanical disturbance or air emersion thereby improving welfare.

BLBK374-17

BLBK374-Brown

422

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Table 17.1 The major welfare issues fish experience in aquaculture and suggested improvements. Area of welfare concern

Welfare issues involved

Improvements

Although many are clearly associated with specific bacterial pathogens, immunosuppression during winter may play a large role.

Immunisation

Fin rot Abrasion with the environment and/or aggressive interactions cause fin damage and secondary infection may follow

Injectable vaccines have superseded antibiotics although vaccines and adjuvants are associated with inflammation and granuloma, as well as the stress of handling anaesthesia and injection.

Vaccines with improved efficacy and reduced side effects as well as oral application.

Sea lice Parasitic copepods may cause severe tissue damage

Lice have developed resistance to traditional chemical treatments.

Potential alternative controls include vaccination and selective breeding towards louse resistance

Winter diseases Several diseases associated with low temperatures

Adapted diet providing a supplementary dosage of vitamins and trace minerals to assist the immune system and altered feeding regime controlling level of nutrients available to the pathogen.

Biological control with cleaner wrasse but should consider wrasse welfare. Viral diseases Examples: Infectious pancreatic necrosis, infectious haematopoietic necrosis, viral haemorrhagic septicaemia, infectious salmon anaemia, sleeping disease

Traditional vaccines developed over the past 20 years have shown only moderate success and there are relatively few commercial vaccines and specific therapeutics with adequate efficacy.

Non-infectious production-related deformities Deformities of the heart, Fish with heart deformities show a swim bladder and spine high mortality rate during stress due to impaired cardiovascular function, cardiac failure or heart rupture. Both genetic and environmental factors may contribute to spinal deformities.

Grading, handling and crowding Inherently stressful Many procedures, such as grading, are aimed at improving welfare. There is a large variation between species in stress response to procedures and handling stressors can affect subsequent stress response.

The development of alternative anti-viral treatments such as DNA vaccines and selection for disease resistance.

High temperatures during incubation of salmon should be avoided. Spinal deformities may be reduced by increasing smolt weight at seawater introduction, vaccinating and reducing salinity and temperature variations. Fish from families showing a high incidence of deformities should not be used for breeding. Appropriate supplementation of dietary vitamins C + E and glucan may protect against the adverse effects of chronic stress. The appropriate use of good crowd management, suitable nets, careful handling, recovery periods and movement using fish pumps and transfer pipes is preferable.

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Cognition and Welfare

423

Table 17.1 (Continued) Area of welfare concern

Welfare issues involved

Improvements Appropriate feeding technique and stocking densities may avoid frequent grading.

Transportation Inherently stressful as it may involve capture, loading, transport, unloading and stocking

Food withdrawal Starvation prior to slaughter, transportation and other management practices

Slaughter Slaughter should be as humane as possible – fish should be stunned prior to slaughter, causing an immediate loss of consciousness that lasts until death

Stocking density Pivotal factor affecting welfare in a number of different ways (e.g. through aggression, water quality, and activity/feeding patterns)

Transport stressors can affect fish over a prolonged period.

Adverse effects may be reduced by suitable acclimation and recovery periods as well as species appropriate use of anaesthesia and dilute salt solutions.

May benefit welfare by reducing metabolism, oxygen demand and waste production. Although Atlantic salmon and rainbow trout show long anorexic periods in the wild so the welfare effect of food deprivation in aquaculture is not known. Deprivation for short periods under appropriate conditions may not diminish welfare.

Starvation for up to 72 h for Atlantic salmon and 48 h for rainbow trout should only occur where beneficial to welfare and empirical studies on the effects of starvation on stress physiology or behaviour are required.

Dewatering followed by asphyxiation in ice slurry of rainbow trout and gilt head sea bream; immersion in CO2 saturated water followed by gill cut or gill cutting alone for Atlantic salmon and rainbow trout; and de-sliming followed by evisceration of eels do not meet the criteria for humane slaughter.

Percussive or electrical stunning methods appear to achieve humane slaughter in Atlantic salmon, gilt-head sea bream, turbot, and rainbow trout.

The effect of stocking density comprises of numerous interacting and case-specific factors. Sea bass show high stress levels at high densities. Arctic charr show low growth and food intake at low and very high densities. Halibut tolerance for high-stocking density appears to be stage dependent. Rainbow trout show a decrease in welfare at high densities, water quality being a key factor. High-stocking densities, above a given threshold, are associated with reduced welfare in Atlantic salmon in sea cages. Site-specific factors also have an effect on welfare.

Feeding pattern and floor space may be altered to improve the effect of density on welfare in halibut, also see ‘Aggression’.

The use of electric stunning tongues or electrically stunning batches of eels in freshwater combined with nitrogen flushing can cause immediate unconsciousness.

Salmon-swimming depth and shoal density can be manipulated by artificial light levels, and feeding patterns can alter aggressive interactions in several species including Atlantic salmon.

(continued)

BLBK374-17

BLBK374-Brown

424

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Table 17.1 (Continued) Area of welfare concern

Welfare issues involved

Improvements

Aggression Formation of social hierarchies may lead to injuries, chronic social stress and size heterogeneity

Sociobiology, stocking density and feeding technique have strong influences on the levels of social interactions.

Feeding technique should be species appropriate to avoid excess competition and aggression. The presence of a small number of larger fish may reduce aggression within groups of smaller fish. Increased dietary levels of l-tryptophan has been shown to suppress aggressive activity. Substrate or background colour may be used to influence aggressive behaviour in some species.

Abnormal behaviour and the freedom to express normal behaviour Abnormal behaviour includes repetitive behaviour and abnormal swimming activity/patterns

Understanding the functional origin of apparently abnormal behaviour is important. Empirical studies are required to establish whether abnormal behaviours represent diminished welfare or adaptive responses with no effect on welfare.

Enriched rearing environments may improve welfare following release to augment wild populations. Without empirical studies the importance of a given behavioural pattern to a given species is unclear. Studies of the mechanism of control and/or the behavioural and physiological consequences of denial of expression of key behaviours are required. Choice studies may allow assessment of the value associated with a given behaviour or resource.

Source: Adapted from Ashley (2007), with permission from Elsevier.

When harvesting, fishes are crowded into a small area and become observably stressed, showing flanks and have increased cortisol concentrations. This can be ameliorated by moving farm sea cages to a much lower depth reducing light levels and catching and killing the fishes more quickly (Brown et al. 2010). Reduced stress in the fishes during harvest not only has ethical benefits but may also improve the quality of the fillet thereby increasing economic return (e.g. Bahuaud et al. 2010). Aquaculture has also promoted the selection for fast growth by breeding the quickest growing individuals and producing bolder, more aggressive fishes (Sundstrom et al. 2004). As personality traits affect survivorship, this has resulted in farmed fishes released for restocking purposes outcompeting wild populations due to their aggressiveness (Sundstrom et al. 2003), but has made these bold fishes more vulnerable to predators (Chapter 16). This

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

425

raises questions as to whether it is ethically sound to use farmed fishes that have no experience of spatially complex environments or how to react to predators for restocking depleted rivers and lakes (Brown & Day 2002). If they are inept at surviving in the natural environment and have a detrimental effect on wild populations, it would seem ineffective and morally corrupt to use them in this way. Studies have explored the possibility of ameliorating these effects by teaching farmed fish antipredator and foraging skills prior to release to enhance their survivorship (Chapter 16; Brown & Laland 2001). This provides an excellent example as to how cognitive approaches can play important roles in applied management contexts and should be further explored in future studies. Altogether, one may consider the welfare of farmed fishes not to be ideal but fishes do provide an important source of protein. The aquaculture industry has attempted to understand and improve the well-being of fish (Table 17.1). However, implementing changes such as reduced stocking density is likely to enhance the costs of production similar to that seen in free range, welfare-friendly meat where the increase in price can be considerable. Consumers should demand to know where the fish they buy came from, under what conditions they were kept and how they were caught to make ethically based decisions.

17.7.2

Fisheries

Large-scale fisheries also employ procedures that are likely to impair the welfare of caught fishes (Chapter 16). Trawling, dredging, hooking on long line, and capture in nets (e.g. cast, drift, ghost, gill, seine) are known to cause stress and under some conditions mortality (Turunen et al. 1994; Chopin & Arimoto 1995; Metcalfe 2009). Fishes are usually hauled out of the water causing suffocation, deposited on deck or within the vessel and killed by a variety of means including suffocation on ice (Ashley 2007). If fishes have the capacity to experience some form of pain and fear, we must consider this unacceptable. Live chilling results in increased physiological stress (Lambooij et al. 2002) and suffocation on ice can take up to 200 minutes for brain death to occur (Robb & Kestin 2002). Humane killing should be quick and effective and although percussive stunning which causes brain destruction could be used as seen in fish farming (Roth et al. 2007), this may be impractical for the enormous numbers of fish caught in large-scale fisheries. Much research remains to be conducted to improve fisheries practices and to understand whether these do indeed impair welfare, cause suffering and to what extent. Many fishes that are caught are non-target species and are considered a by-catch, which are discarded. Studies aimed at understanding whether these fishes have impaired welfare as a consequence of being caught are necessary since some species suffer 50% mortality as a result of the capture process (Mandelman & Farrington 2007).

17.7.3

Recreational fishing

In many respects recreational fishing is similar to any other form of hunting but fishing does not receive the same sort of social stigmatism or raise as much objection because fish are not perceived in the same way as other animals. Recreational fisheries or angling for sport is often conducted purely for enjoyment and may involve catch and release of fishes. Fishes are hooked, landed and then released and are known to suffer stress (Arlinghaus et al.

BLBK374-17

BLBK374-Brown

426

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

2007, 2009) and impaired behaviour post-release (e.g. Cooke & Philipp 2004; Danylchuk et al. 2007). This is in contrast to catch and kill where fishes are killed presumably to be eaten by the angler. However, lengthy capture causes stress and fish should not be allowed to suffocate in air since this takes approximately 15 minutes for death to occur in trout (Robb & Kestin 2002). Given that the catch and release of fish may result in injury from hooking, exhaustion, air emersion, stress, removal from their natural environment and mortality, then the fish’s welfare is clearly compromised. Damage to the fish during hooking and net abrasion when landed (Butcher et al. 2008) are likely to be painful events and result in a stress ‘fight or flight’ response, producing fear and/or negative states associated with stress. One can accept there is a benefit to humans when fishes are humanely killed for food so long as the fish is killed rapidly after capture; however, the question of practising catch and release must be considered from a moral and ethical perspective. Fishes are able to learn to avoid hooks after being caught and released (Ferno & Huse 1983; Pyanov 1993; Young & Haye 2004; Chapter 16); therefore, this aversive stimulus has a significant effect upon subsequent fish behaviour. Many angling organisations and scientists have made recommendations for improved catch-and-release practices to improve the welfare of caught fish (Schupplid 1999; Freshwater anglers, Australia in; Cooke & Sneddon 2007; Table 17.2). Such practices include the use of barbless hooks and knotless nets. Germany has responded to this moral question by restricting the capture of fish for food purposes only (Arlinghaus & Mehner 2003). However, more research is needed to fully understand how catch-and-release practices affect the fish in both the long and short terms.

17.7.4

Research

Unlike many other realms where fishes are widely used, the use of fish in research is heavily controlled by legislation in most developed nations. Nevertheless some issues remain. Experimentation of fish does involve maintaining large numbers in captivity. This is often conducted using barren stock tanks. Since fishes have specific preferences for substrate, refuge, nesting material, related individuals and so on, environmental enrichment should Table 17.2 The main recommendations made to anglers to improve welfare to the fish in recreational fishing. Welfare concern

Welfare issues

Improvement

Exhaustion and stress during long angling event

Stress, fear and prolonged recovery impairing subsequent behaviour

Minimise the duration of the angling event; reduce play time

Air emersion and handling

Stress, suffocation and damage during handling

Minimise or eliminate handling and exposure to air by keeping fish in water

High water temperatures associated with increased stress and mortality

Stress and prolonged recovery impairing subsequent behaviour or death

Restrictions in angling at higher water temperatures

Damage caused by hooking can result in injury or mortality

Pain, fear, stress or death

Using hooks that reduce injury, stress or mortality

Impaired reproductive success

Stress affects reproductive capacity

Avoid angling during reproductive period

Source: Adapted from Arlinghaus et al. (2007) and Cooke & Sneddon (2007), with permission from Elsevier.

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

427

be explored as a means of improving welfare. This would obviously need to be specific to the species requirements. Moving fishes between tanks should also be carefully considered since if fishes prefer to shoal or hide with familiar individuals, placing them in new groups or in novel environments is likely to be stressful. Aggressive species which form dominance hierarchies may need to be held in high numbers to reduce such behaviour so group composition and density are also important factors. If fishes rely upon information transfer and learn from conspecifics, then if one fish within a tank is stressed and acting abnormally then surely this will affect all fishes within the tank. Future studies should address these welfare concerns within the realm of the laboratory aquaria by using cognitive approaches such as preference testing. Comprehensive knowledge of the life history and normal behaviour of each species is vital to understand how captive conditions may impair fish behaviour.

17.7.5

Companion fish

Finally, our use of fish as pets should be highly scrutinised given their increasing popularity. Part of the problem stems from the very low cost of purchasing fish as a pet and the ease of availability. Fortunately, in most countries, goldfish in a plastic bag can no longer be offered as a fairground prize. The classic image of a solitary goldfish in a barren, small goldfish bowl seems unreasonable given the complicated nature of fish behaviour and preferences for social and environmental stimulations. It is also believed that due to a low surface area they provide insufficient dissolved oxygen for fish to breath and indeed these spherical bowls have been banned in Rome, Italy. No licensing or training is required to set up an aquarium tank and keep fish, yet it is clear that an understanding of these aquatic animals is necessary to maintain them in good health. This is in stark contrast to the situation with most other pets where, in many countries, owners of cats, dogs and reptiles need to be registered. Another aspect of the ornamental fish trade is identifying where the fishes have come from. Some fishes are bred in captivity specifically for the pet trade but many are taken from the wild depleting natural populations. They are frequently harvested in unsustainable ways, for example the use of clove oil to stun entire populations on coral reef bombies. These fishes are placed into plastic bags containing aerated water but can be transported for extended periods of time without further aeration and with deteriorating water quality (Walster 2008; IATA 2009). In guppies, post-transport mortality is linked to high stress levels (Lim et al. 2003). Research aimed at understanding and improving the procedures in the ornamental fish trade is necessary since currently relatively little is known about what specific welfare problems exist.

17.8

Conclusion

Cognitive experimental approaches are extremely useful in understanding the subjective state of fish and determining how negative welfare states impact upon fish behaviour. We use fishes in a variety of ways subjecting them to practices that would not be considered acceptable in other vertebrates. Therefore, careful consideration of fish behaviour, physiological functioning and welfare are needed. Preference testing can inform husbandry of

BLBK374-17

BLBK374-Brown

428

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

captive fish since if fish actively choose specific items such as shelter or substrates then they must be beneficial in terms of what options were given. Avoidance tests employing aversive learning paradigms can also deliver important information on what fishes find harmful and can also be informative for fish husbandry. Housing conditions in captive fish have received attention from regulatory bodies with environmental enrichment now considered necessary in the laboratory aquarium (e.g. Europe, Sauer 2004). However, it is vital that species-specific requirements are developed since what one fish species prefers another may not and this could be detrimental to welfare. Environmental enrichment is known to promote brain development in fishes (Lema et al. 2005; Kihslinger et al. 2006) and may stimulate learning ability (Brown et al. 2003; Salvanes et al. 2007). Providing habitat complexity may be a valuable tool for improving welfare of captive fishes. Rearing fishes in isolation can also affect brain size (Gonda et al. 2009); therefore, social housing may have stimulatory effects directly upon brain development, which has consequences for behaviour and possibly welfare. More research is needed on social and environmental enrichment to make reliable and valid conclusions. Negative emotional states such as pain and fear do result in profound changes in behaviour and physiology. Pain may prevent animals showing normal fear and antipredator responses; however, the context is very important and one must consider species-specific differences. Studies on fear and pain should be designed to allow animals to make a choice to avoid these states but more advanced paradigms other than classical conditioning could be used to truly understand the cognitive implications of these negative states. Studies directly tackling higher cognitive functions such as self-recognition should be designed with the biology of fishes in mind rather than applying paradigms from mammals. Fishes rely on different sensory systems and may recognise themselves through smell rather than through vision in mirror tests. If higher cognitive abilities are identified in fishes, this would make a real advance in our understanding of whether fishes are conscious of pain and fear and as a consequence suffer. Assessment of negative welfare may be confounded by individual differences where bold fishes recover more quickly from stressful events. Housing conditions may promote welfare in bold fishes but as a consequence may be detrimental to shy fishes where they are dominated by bold individuals. Cognitive approaches should be developed to understand whether differences in learning ability are due to underlying mental processes or are simply a result of the risk-taking bold phenotype being more willing or motivated to engage in learning trials (Roult et al. unpublished data). Finally, the question of whether captive conditions remove certain motivationally driven behaviour is an important one. Does regular feeding of caged Atlantic salmon in aquaculture remove the need for these fishes to travel long distances to find food as they would do in natural populations? Or do the fishes swim continuously in a circular fashion to express this behaviour in the confines of the sea cage? Does the absence of predators in laboratory aquaria result in reduced or absent antipredator behaviours in captive fish populations and subsequently their fear concepts? These and many other questions should be addressed in future cognitive experiments where comparisons with the natural behavioural repertoire of the wild counterpart will provide insight into the behaviour and welfare of captive housed fishes. Only then can we really understand how our use of fishes impacts upon their well-being.

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

429

Acknowledgements I am grateful to Culum Brown and two anonymous referees for their useful comments on this chapter. I wish to thank NERC, The Leverhulme Trust and UFAW for funding.

References Ahnesjo, I. (2008) Behavioural temperature preference in a brooding male pipefish Syngnathus typhle. Journal of Fish Biology, 73, 1039–1045. Aihara, Y., Yasuoka, A., Iwamoto, S., Yoshida, Y., Misaka, T. & Abe, K. (2009) Construction of a taste-blind medaka fish and quantitative assay of its preference-aversion behaviour. Genes Brain and Behavior, 7, 924–932. Arlinghaus, R., Cooke, S.J., Schwab, A. & Cowx, I.G. (2007) Fish welfare: a challenge to the feelings-based approach, with implications for recreational fishing. Fish and Fisheries, 8, 57–71. Arlinghaus, R. & Mehner, T. (2003) Socio-economic characterisation of specialised common carp (Cyprinus carpio L.) anglers in Germany, and implications for inland fisheries management and eutrophication control. Fisheries Research, 61, 19–33. Arlinghaus, R., Schwab, A., Cooke, S.J. & Cowx, I.G. (2009) Contrasting pragmatic and sufferingcentred approaches to fish welfare in recreational angling. Journal of Fish Biology, 75, 2448– 2463. Arras, M., Rettich, A., Cinelli, P., Kasermann, H.P. & Burki, K. (2007) Assessment of post-laparotomy pain in laboratory mice by telemetric recording of heart rate and heart rate variability. BMC Veterinary Research, 3, 16. Ashley, P.J. (2007) Fish welfare: current issues in aquaculture. Applied Animal Behaviour Science, 104, 199–235. Ashley, P.J., Ringrose, S., Edwards, K.L., Wallington, E., McCrohan, C.R. & Sneddon, L.U. (2009) Which is more important in fish: pain, anti-predator responses or dominance status? Animal Behaviour, 77, 403–410. Ashley, P.J. & Sneddon, L.U. (2007) Pain and fear in fish. In: E. Branson (ed) Fish Welfare. Blackwell Publishing Ltd., Oxford. Ashley, P.J., Sneddon, L.U. & McCrohan, C.R. (2006) Properties of corneal receptors in a teleost fish. Neuroscience Letters, 410, 165–168. Ashley, P.J., Sneddon, L,U. & McCrohan, C.R. (2007) Nociception in fish: stimulus-response properties of receptors on the head of trout Oncorhynchus mykiss. Brain Research, 1166, 47–54. Bahuaud, D., Mørkøre, T., Østbye, T.K. Veiseth-Kent, E., Thomassen, M.S. & Ofstad, R. (2010) Muscle structure responses and lysosomal cathepsins B and L in farmed Atlantic salmon (Salmo salar L.) pre- and post-rigor fillets exposed to short and long-term crowding stress. Food Chemistry, 118, 602–615. Balcombe, J. (2006) Pleasurable Kingdom: Animals and the Nature of Feeling Good. Macmillan, London. Barnett, J.L. Cronin, G.M., McCallum, T.H. & Newman, E.A. (1994) Effects of food and time of day on aggression when grouping unfamiliar adult pigs. Applied Animal Behaviour Science, 39, 339–347. Bateson, P. (1991) Assessment of pain in animals. Animal Behaviour, 42, 827–839. Bekoff, M. (2006) Animal Passions and Beastly Virtues: Reflections on Redecorating Nature. Temple University Press, Philadelphia. Bekoff, M. (2007) Aquatic animals, cognitive ethology, and ethics: questions about sentience and other troubling issues that lurk in turbid water. Diseases of Aquatic Organisms, 75, 87–98. Bekoff, M. & Sherman, P.W. (2004) Reflections on animal selves. Trends in Ecology and Evolution, 19, 176–180.

BLBK374-17

BLBK374-Brown

430

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Benhaim, D., Leblanc, C.A. & Lucas, G. (2009) Impact of a new artificial shelter on Arctic charr (Salvelinus alpinus, L.) behaviour and culture performance during the endogenous feeding period. Aquaculture, 295, 38–43. Blakeslee, C., McRobert, S.P., Brown, A.C. & Clotfelter, E.D. (2009) The effect of body coloration and group size on social partner preferences in female fighting fish (Betta splendens). Behavioural Processes, 80, 157–161. Blom, J.H., Dabrowski, K., Rapp, J.D., Sakakura, Y. & Tsukamoto, K. (1999) Competition for space and food in rainbow trout, Oncorhynchus mykiss, as related to ascorbic acid status. Aquaculture, 180, 79–87. Broom, D.M. (1991a) Assessing welfare and suffering. Behavioral Processes, 25, 117–123. Broom, D.M. (1991b) Animal welfare: concepts and measurement. Journal of Animal Science, 69, 4167–4175. Broom, D.M. (2007) Cognitive ability and sentience: which aquatic animals should be protected? Diseases of Aquatic Organisms, 75, 99–108. Brown, C. (2001) Familiarity with the test environment improves escape responses in the crimson spotted rainbowfish, Melanotaenia duboulayi. Animal Cognition, 4, 109–113. Brown, C., Davidson, T. & Laland, K. (2003) Environmental enrichment and prior experience of live prey improve foraging behaviour in hatchery-reared Atlantic salmon. Journal of Fish Biology, 63, 187–196. Brown, C. & Day, R. (2002) The future of stock enhancements: bridging the gap between hatchery practice and conservation biology. Fish and Fisheries, 3, 79–94. Brown, C., Jones, F. & Braithwaite, V. (2005) In situ examination of boldness–shyness traits in the tropical poeciliid, Brachyraphis episcopi. Animal Behaviour, 70, 1003–1009. Brown, C. & Laland, K.N. (2001) Social learning and life skills training for hatchery reared fish. Journal of Fish Biology, 59, 471–493. Brown, J.A., Watson, J., Bourhill, A. & Wall, T. (2010) Physiological welfare of commercially reared cod and effects of crowding for harvesting. Aquaculture, 298, 315–324. Butcher, P.A., Broadhurst, M.K. & Cairns, S.C. (2008) Mortality and physical damage of angled-and released dusky flathead, Platycephalus fuscus. Diseases of Aquatic Organisms, 81, 127–134. Carbone, L. (2004) What Animals Want: Expertise and Advocacy in Laboratory Animal Welfare Policy. Oxford University Press, New York. Chen, S.X., Hong, W.S., Su, Y.Q. & Zhang, Q.Y. (2008) Microhabitat selection in the early juvenile mudskipper Boleophthalmus pectinirostris (L.). Journal of Fish Biology, 72, 585–593. Clingerman, J., Bebak, J., Mazik, P.M. & Summerfelt, S.T. (2007) Use of avoidance response by rainbow trout to carbon dioxide for fish self-transfer between tanks. Aquacultural Engineering, 37, 234–251. Conte, F.S. (2004) Stress and welfare of cultured fish. Applied Animal Behaviour Science, 86, 205–223. Cooke, S.J. & Philipp, D.P. (2004) Behavior and mortality of caught-and-released bonefish (Albula spp.) in Bahamian waters with implications for a sustainable recreational fishery. Biological Conservation, 118, 599–607. Cooke, S.J. & Sneddon, L.U. (2007) Animal welfare perspectives on recreational angling. Applied Animal Behaviour Science, 104, 176–198. Cronin, G.M., Dunsmore, B. & Leeson, E. (1998) The effects of farrowing nest size and width on sow and piglet behaviour and piglet survival. Applied Animal Behaviour Science, 60, 331– 345. Danylchuk, S.E., Danylchuk, A.J., Cooke, S.J., Goldberg, T.L., Koppelman, J. & Phillip, D.P. 2007. Effects of recreational angling on the post-release behaviour and predation of bonefish (Albula vulpes): the role of equilibrium status at the time of release. Journal of Experimental Marine Biology and Ecology, 346, 127–133. Dawkins, M.S. (1998a) Evolution and animal welfare. Quarterly Review of Biology, 73, 1–21. Dawkins, M.S. (1998b) Through Our Eyes Only? The Search for Animal Consciousness. Oxford University Press, Oxford.

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

431

Dou, S.Z., Yamada, Y., Okamura, A., Shinoda, A., Tanaka, S. & Tsukamoto, K. (2008) Temperature influence on the spawning performance of artificially-matured Japanese eel, Anguilla japonica, in captivity. Environmental Biology of Fishes, 82, 151–164. Duncan, I.J.H. (2002) Gordon memorial lecture. Poultry welfare: science or subjectivity? British Poultry Science, 43, 643–652. Ehrensing, R.H., Michell, G.F. & Kastin, A.J. (1982) Similar antagonism of morphine analgesia by MIF-1 and naloxone in Carassius auratus. Pharmacology, Biochemistry and Behaviour, 17, 757–761. Erhard, H.W., Mendl, M. & Ashley, D.D. (1997) Individual aggressiveness of pigs can be measured and used to reduce aggression after mixing. Applied Animal Behaviour Science, 54, 137–151. Eysenck, H.J. (1946) The measurement of personality. Proceedings of the Royal Society of Medicine, 40, 75–80. Fangue, N.A., Podrabsky, J.E., Crawshaw, L.I. & Schulte, P.M. (2009) Countergradient variation in temperature preference in populations of killifish, Fundulus heteroclitus. Physiological and Biochemical Zoology, 82, 776–786. FAO (2006) ftp://ftp.fao.org/fi/stat/summary/summ 06/a1a.pdf FAWC (1996) Report on the Welfare of Farmed Fish. Farmed Animal Welfare Council, Surbiton, Surrey. Ferno, A. & Huse, I. (1983) The effect of experience on the behaviour of cod (Gadus morhua L.) towards baited hook. Fisheries Research, 2, 19–28. Flecknell, P., Gledhill, J. & Richardson, C. (2007) Assessing animal health and welfare and recognising pain and distress. Altex, 24, 82–83. Frost, A.J., Ashley, P.J, Winrow-Giffen, A. & Sneddon, L.U. (2007) Plasticity in animal personality traits: does prior experience alter the degree of boldness? Proceedings of the Royal Society of London Series B – Biological Sciences. 274, 333–339. Galhardo, L., Almeida, O. & Oliveira, R.F. (2009) Preference for the presence of substrate in male cichlid fish: effects of social dominance and context. Applied Animal Behaviour Science, 120, 224–230. Gentle, M.J. (1992) Pain in Birds. Animal Welfare, 1, 235–247. Gentle, M.J. (2001) Attentional shifts alter pain perception in the chicken. Animal Welfare, 10, S187–194. Gerlai, R., Fernandes, Y. & Pereira, T. (2009) Zebrafish (Danio rerio) responds to the animated image of a predator: towards the development of an automated aversive task. Behavioural Brain Research, 201, 318–324. Gibson, R., Burns, J.G. & Rodd, F.H. (2009) Flexibility in the colouration of the meninx (brain covering) in the guppy (Poecilia reticulata): investigations of potential function. Canadian Journal of Zoology, 87, 529–536. Gilmour, K.M., DiBattista, J.D. & Thomas, J.B. (2005) Physiological causes and consequences of social status in salmonid fish. Integrative and Comparative Biology, 45, 263–273. Gomez-Laplaza, L.M. (2009) Recent social environment affects colour-assortative shoaling in juvenile angelfish (Pterophyllum scalare). Behavioural Processes, 82, 39–44. Gonda, A., Herczeg, G. & Merila, J. (2009) Habitat-dependent and -independent plastic responses to social environment in the nine-spined stickleback (Pungitius pungitius) brain. Proceedings of the Royal Society of London Series B – Biological Sciences, 276, 2085–2092. Griffiths, S.W. & Armstrong, J.D. (2000) Differential responses of kin and nonkin salmon to patterns of water flow: does recirculation influence aggression? Animal Behaviour, 59, 1019–1023. Hiscock, M.J. & Brown, J.A. (2000) Kin discrimination in juvenile brook trout (Salvelinus fontinalis) and the effect of odour concentration on kin preferences. Canadian Journal of Zoology, 78, 278–282. Huntingford, F.A. (2004) Implications of domestication and rearing conditions for the behaviour of cultivated fishes. Journal of Fish Biology, 65, 122–142. IATA (2009) http://www.iata.org/ps/publications/live-animals.htm. Iwama, G.K. (2007) The welfare of fish. Diseases of Aquatic Organisms, 75, 155–158.

BLBK374-17

BLBK374-Brown

432

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Jarvis, E.D., G¨unt¨urk¨un, O., Bruce, L., Csillag, A., Karten, H., Kuenzel, W., Medina, L., Paxinos, G., Perkel, D.J., Shimizu, T., Striedter, G., Wild, J.M., Ball, G.F., Dugas-Ford, J., Durand, S.E., Hough, G.E., Husband, S., Kubikova, L., Lee, D.W., Mello, C.V., Powers, A., Siang, C., Smulders, T.V., Wada, K., White, S.A., Yamamoto, K., Yu, J., Reiner, A. & Butler, A.B. (2005) Avian brains and a new understanding of vertebrate brain evolution. Nature Reviews Neuroscience, 6, 151–159. Jensen, P. (2009) The Ethology of Domestic Animals: An Introductory Text, 2nd ed. CABI, Oxfordshire. Jensen, P. & Toates, F.M. (1993) Who needs ‘behavioural needs’? Motivational aspects of the needs of animals. Applied Animal Behaviour Science, 37, 161–181. Kihslinger, R.L., Lema, S.C. & Nevitt, G.A. (2006) Environmental rearing conditions produce forebrain differences in wild Chinook salmon Oncorhynchus tshawytscha. Comparative Biochemistry and Physiology A, 145, 145–151. Kim, Y.H., Lee, Y., Lee, H., Jung, M.W. & Lee, C.J. (2009) Impaired avoidance learning and increased hsp70 mRNA expression in Pentylenetetrazol-treated zebrafish. Animal Cells and Systems, 13, 275–281. Korsøen, O.J., Dempster, T., Fjelldal, P.G., Oppedal, F. & Kristiansen, T.S. (2009) Long-term culture of Atlantic salmon (Salmo salar L.) in submerged cages during winter affects behaviour, growth and condition. Aquaculture, 296, 373–381. Kruschwitz, A., Zupan, M., Buchwalder, T. & Huber-Eicher, B. (2008) Nest preference of laying hens (Gallus gallus domesticus) and their motivation to exert themselves to gain nest access. Applied Animal Behaviour Science, 112, 321–330. Kydd, E. & Brown, C. (2009) Loss of shoaling preference for familiar individuals in captive-reared crimson spotted rainbowfish Melanotaenia duboulayi. Journal of Fish Biology, 74, 2187–2195. Lambooij, E., van de Vis, J.W., Kloosterboer, R.J. & Pieterse, C. (2002) Welfare aspects of live chilling and freezing of farmed eel (Anguilla anguilla L.): neurological and behavioural assessment. Aquaculture, 210, 159–169. Lema, S.C., Hodges, M.J., Marchetti, M.P. & Nevitt, G.A. (2005) Proliferation zones in the salmon telencephalon and evidence for environmental influence on proliferation rate. Comparative Biochemistry and Physiology A, 141, 327–335. Lim, L.C., Dhert, P. & Sorgeloos, P. (2003) Recent developments and improvements for ornamental fish packaging systems for air transport. Aquaculture Research, 34, 923–935. Ludsin, S.A., Zhang, X., Brandt, S.B., Roman, M.R., Boicourt, W.C., Mason, D.M. & Costantini, M. (2009) Hypoxia-avoidance by planktivorous fish in Chesapeake Bay: implications for food web interactions and fish recruitment. Journal of Experimental Marine Biology and Ecology, 381, S121–S131. Maes, J., Turnpenny, A.W.H., Lambert, D.R., Nedwell, J.R., Parmentier, A. & Ollevier, F. (2004) Field evaluation of a sound system to reduce estuarine fish intake rates at a power cooling water inlet. Journal of Fish Biology, 64, 938–946. Mandelman, J.W. & Farrington, M.A. (2007) The estimated short-term discard mortality of a trawled elasmobranch, the spiny dogfish (Squalus acanthias). Fisheries Research, 83, 238–245. Markowitz, H. & LaForce, S. (1987) Artificial prey as behavioral enrichment devices for felines. Applied Animal Behaviour Science, 18, 31–43. McMillan, F.D. & Lance, K. (2004) Unlocking the Animal Mind: How Your Pet’s Feelings Hold the Key to His Health and Happiness. Rodale, Emmaus, PA. Mendl, M. & Paul, E. (2004) Consciousness, emotion and animal welfare: insights from cognitive science. Animal Welfare, 13, S17–S25. Metcalfe, J.D. (2009) Welfare in wild-capture marine fisheries. Journal of Fish Biology, 75, 2855–2861. Meunier, B., Yavno, S., Ahmed, S. & Corkum, L.D. (2009) First documentation of spawning and nest guarding in the laboratory by the invasive fish, the round goby (Neogobius melanostomus). Journal of Great Lakes Research, 35, 608–612. Millsopp, S. & Laming, P. (2008) Trade-offs between feeding and shock avoidance in goldfish (Carassius auratus). Applied Animal Behaviour Science, 113, 247–254.

BLBK374-17

BLBK374-Brown

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Cognition and Welfare

Char Count=

433

Molyneux, B. (2010) Why the neural correlates of consciousness cannot be found? Journal of Consciousness Studies, 17, 168–188. Ninomiya, S., Kusunose, R., Sato, S., Terada, M. & Sugawara, K. (2004) Effects of feeding methods on eating frustration in stabled horses. Animal Science Journal, 75, 465–469. Ojanguren, A.F. & Bra˜na, F. (1999) Discrimination against water containing unrelated conspecifics and a marginal effect of relatedness on spacing behaviour and growth in juvenile brown trout, Salmo trutta L. Ethology, 105, 937–948. Oliveira, R.F., McGregor, P.K. & Latruffe, C. (1998) Know thine enemy: fighting fish gather information from observing conspecific interactions. Proceedings of the Royal Society of London Series B – Biological Sciences, 265, 1045–1049. Pasko, L. (2010) Tool-Like behavior in the sixbar wrasse, Thalassoma hardwicke (Bennett, 1830). Zoo Biology, 28, 1–7. Plumb, J.M. & Blanchfield, P.J. (2009) Performance of temperature and dissolved oxygen criteria to predict habitat use by lake trout (Salvelinus namaycush). Canadian Journal of Fisheries and Aquatic Sciences, 66, 2011–2023. Pulgar, J.M., Aldana, M., Bozinovic, F. & Ojeda, F.P. (2003) Does food quality influence thermoregulatory behavior in the intertidal fish Girella laevifrons? Journal of Thermal Biology, 28, 539–544. Pyanov, A.I. (1993) Fish learning in response to trawl fishing. ICES Marine Science Symposia, 196, 12–14. Ramirez, J. & Cabanac, M. (2003) Pleasure, the common currency of emotions. Annals of the New York Academy of Science, 1000, 203–295. Reilly, S.C., Quinn, J.P., Cossins, A.R. & Sneddon, L.U. (2008) Behavioural analysis of a nociceptive event in fish: comparisons between three species demonstrate specific responses. Applied Animal Behaviour Science, 114, 248–249. Robb, D.H.F. & Kestin, S.C. (2002) Methods used to kill fish: field observations and literature reviewed. Animal Welfare, 11, 269–282. Rose, J.D. (2002) The neurobehavioral nature of fishes and the question of awareness and pain. Reviews in Fisheries Science, 10, 1–38. Roth, B., Slinde, E. & Robb, D.H.F. (2007) Percussive stunning of Atlantic salmon (Salmo salar) and the relation between force and stunning. Aquacultural Engineering, 36, 192–197. Rushbrook, B.J., Dingemanse, N.J. & Barber, I. (2008) Repeatability in nest construction by male three-spined sticklebacks. Animal Behaviour, 75, 547–553. Salvanes, A.G.V., Moberg, O. & Braithwaite, V.A. (2007) Effects of eary experience on group behaviour in fish. Animal Behaviour, 74, 805–811. S´anchez-V´azquez, F.J., Yamamoto, T., Akiyama, T., Madrid, J.A. & Tabata, M. (1998) Selection of macronutrients by goldfish operating self-feeders. Physiology and Behavior, 65, 211–218. Sauer, U.G. (2004) The revision of European housing guidelines for laboratory animals: expectations from the point of view of animal welfare. Alternatives to Laboratory Animals, 32, 187–190. Schmidt, M.B., Balk, H. & Gassner, H. (2009) Testing in situ avoidance reaction of vendace, Coregonus albula, in relation to continuous artificial light from stationary vertical split-beam echosounding. Fisheries Management and Ecology, 16, 376–385. Schupplid, C.A. (1999) Report and Recommended Actions for Humane Angling in Canada. Prepared for the Animal Welfare Foundation of Canada, Vancouver, BC, p. 23. Sessa, A.K., White, R., Houvras, Y., Burke, C., Pugach, E., Baker, B., Gilbert, R., Look, A.T. & Zon, L.I. (2008) The effect of a depth gradient on the mating behavior, oviposition site preference, and embryo production in the zebrafish, Danio rerio. Zebrafish, 5, 335–339. Shettleworth, S.J. (2001) Animal cognition and animal behaviour. Animal Behaviour, 61, 277–286. Smith, C.C. & Sargent, R.C. (2006) Female fitness declines with increasing female density but not male harassment in the western mosquitofish, Gambusia affinis. Animal Behaviour, 71, 401– 407. Sneddon, L.U. (2002) Anatomical and electrophysiological analysis of the trigeminal nerve in a teleost fish, Oncorhynchus mykiss. Neuroscience Letters, 319, 167–171.

BLBK374-17

BLBK374-Brown

434

May 13, 2011

17:37

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Fish Cognition and Behavior

Sneddon, L.U. (2003a) The evidence for pain in fish: the use of morphine as an analgesic. Applied Animal Behaviour Science, 83, 153–162. Sneddon, L.U. (2003b) Trigeminal somatosensory innervation of the head of a teleost fish with particular reference to nociception. Brain Research, 972, 44–52. Sneddon, L.U. (2004) Evolution of nociception in vertebrates: comparative analysis of lower vertebrates. Brain Research Reviews, 46, 123–130. Sneddon, L.U. (2006) Ethics and welfare: pain perception in fish. Bulletin of the European Association of Fish Pathologists, 26, 6–10. Sneddon, L.U. (2009) Pain perception in fish: indicators and endpoints. ILAR Journal, 50, 338–342. Sneddon, L.U., Braithwaite, V.A. & Gentle, M.J. (2003a) Do fishes have nociceptors? Evidence for the evolution of a vertebrate sensory system. Proceedings of the Royal Society of London Series B – Biological Sciences, 270, 1115–1121. Sneddon, L.U., Braithwaite, V.A. & Gentle, M.J. (2003b) Novel object test: examining nociception and fear in the rainbow trout. Journal of Pain, 4, 431–440. Stierhoff, K.L., Tyler, R.M. & Targett, T.E. (2009) Hypoxia tolerance of juvenile weakfish (Cynoscion regalis): laboratory assessment of growth and behavioral avoidance responses. Journal of Experimental Marine Biology and Ecology, 381, S173–S179. Sundstr¨om, L.F., L˜ohmus, M. & Johnsson, J.I. (2003) Investment in territorial defence depends on rearing environment in brown trout (Salmo trutta). Behavioral Ecology and Sociobiology, 54, 249–255. Sundstr¨om, F.L., Petersson, E., H¨ojesj¨o, J., Johnsson, J.I. & J¨arvi, T. (2004) Hatchery selection promotes boldness in newly hatched brown trout (Salmo trutta): implications for dominance. Behavioral Ecology, 15, 192–198. Thunken, T., Waltschyk, N., Bakker, T.C.M. & Kullmann, H. (2009) Olfactory self-recognition in a cichlid fish. Animal Cognition, 12, 717–724. Volpato, G.L., Gonc¸alves-de-Freitas, E. & Fernandes-de-Castilho, M. (2007) Insights into the concept of fish welfare. Diseases of Aquatic Organisms, 75, 165–171. Walster, C. (2008) The welfare of ornamental fish. In: E. Branson (ed) Fish Welfare. Blackwell Publishing Ltd., Oxford. Wilson, A.D.L. & McLaughlin, R.L. (2007) Behavioural syndromes in brook charr, Salvelinus fontinalis: prey-search in the field corresponds with space use in novel laboratory situations. Animal Behaviour, 74, 689–698. Wilson, D.S., Coleman, K., Clark, A.B. & Biederman, L. (1993) Shy-bold continuum in pumpkinseed sunfish (Lepomis gibbosus): an ecological study of a psychological trait. Journal of Comparative Psychology, 107, 250–260. Wirtz, P. & Davenport, J. (1976) Increased oxygen consumption in blennies (Blennius pholis L.) exposed to their mirror images. Journal of Fish Biology, 9, 67–74. Young, R.G. & Haye, J.W. (2004) Angling pressure and trout catchability: behavioral observations of brown trout in two New Zealand backcountry rivers. North American Journal of Fisheries Management, 24, 1203–1213. Yue, S., Moccia, R.D. & Duncan, I.J.H. (2004) Investigating fear in domestic rainbow trout, Oncorhynchus mykiss, using an avoidance learning task. Applied Animal Behaviour Science, 87, 343–354.

BLBK374-species

BLBK374-Brown

June 1, 2011

17:12

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Species List

Scientific name Abramis brama Abudefduf troschelii Acanthurus coeruleus Acanthurus nigrofuscus Aequidens pulcher Aidablennius sphynx (sphinx) Ambloplites rupestris Amblyglyphidodon leucogaster Amphiprion percula Anabas testudineus Anguilla anguilla Anguilla japonica Anguilla spp. Antennarius marmaoratus Anthias squamipinnis Apistogramma trifasciatum Apteronotus leptorhynchus Archocentrus nigrofasciatus Argyrosomus japonicus Astatotilapia burtoni Asterropteryx semipunctatus Astyanax fasciatus Awaous guamensis Brachydanio rerio Bathygobius soporator Betta splendens

Common name freshwater bream Panamic sergeant major blue tang surgeonfish brown surgeonfish blue acara

Bidyanus bidyanus

silver perch

rock bass white-belly damselfish orange clownfish climbing perch European silver eel Japanese eel eels frogfish sea goldie dwarf cichlid brown ghost knife fish convict cichlid mulloway starry goby banded astyanax zebra danio frillfin goby Siamese fighting fish

Pages 16, 371 269 260, 269 168, 241 49 249 68, 188, 248 89–90 175 188 307 412 247, 389, 423 41 197 282 125 64, 110, 112, 304, 310 147 118, 125, 250, 287 47 171 173 44, 50, 61, 62, 147 39, 169–170 84–6, 113, 116, 117, 119, 169, 249–50, 280, 304, 415 15, 16, 22, 25

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

BLBK374-species

436

BLBK374-Brown

June 1, 2011

17:12

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Species List

Scientific name Boleophthalmus pectinirostris Brachyrhaphis episcopi

Brachyrhaphis roseni Canthigaster valentini Carassius auratus

Carassius carassius Carassius langsdorfii Catostomus commersoni Cheatadon spp. Chromis caeruleus Chromis chrysurus Cichlasoma citrinellum Cichlasoma nigrofasciatum Clarias gariepinus Clupea harengus Clupea pallasi Coregonus albula Coregonus spp. Coryphopterus nicholsi Cottus gobio Crenilabrus melops Ctenochaetus striatus Culaea inconstans Cymatogaster aggregate Cynoscion regalis Cyprinella venusta Cyprinus carpio Danio aequipinnatu Danio rerio Dascyllus aruanus Dascyllus marginatus Dicentrarchus labrax Elacatinus genie Epinephelus striatus Esox lucius Esox niger Etheostoma flabellare Fundulus diaphanus

Common name mudskipper Panamanian bishop

Valentini’s sharpnose pufferfish goldfish

crucian carp white sucker butterflyfish green chromis stout chromis Midas cichlid convict cichlid sharptooth catfish herring Pacific herring vendace whitefish blackeye goby bullhead goby corkwing wrasse surgeonfish brook stickleback shiner perch weakfish blacktail shiner common carp giant danio zebra danio humbug damselfish marginate dascyllus seabass cleaning goby Nassau grouper Northern pike chain pickerel fantail darter banded killifish

Pages 412 139, 147, 149, 177, 193, 195, 302, 304, 307–308, 312, 315 304 46 12–14, 16, 20, 27, 44, 45, 49, 63, 70, 139, 169, 173, 247, 303, 327–335, 337– 349, 361, 414, 417, 427 41 139 66–67 168 45 16, 168 197–198 139 197 220, 222, 363–366, 374 188, 194 417 300 170 88, 249 14, 24, 173 268 48, 61, 62 303, 314 411, 412 188 373, 411, 419 218 61, 139, 198, 243, 411 192 376 248, 382 283, 286 260, 267 60–64, 66, 67, 70, 72 51 88, 249 188, 194, 222, 224, 228

BLBK374-species

BLBK374-Brown

June 1, 2011

17:12

Copyeditor’s Name:

Trim: 244mm X 172mm

Species List

Scientific name Gadus merlangus Gadus morhua Galaxias maculatus Gambusia affinis Gambusia holbrooki Gasterosteus aculeatus

Common name whiting Atlantic cod inanga Western mosquitofish Eastern mosquitofish three-spined stickleback

Geophagus brasiliensis Girardinus falcatus Girella laevifrons Glossamia aprion Gnatholepis anjerensis Gnathonemus petersii Gobiomorphus cotidianus Gobiusculus flavescens Gymnocephalus cernuus Gymnothorax javanicus Haemulon flavolineatum Haemulon plumieri Hemigrammus erythrozonus Hephaestus fuliginosus Hippoglossus hippoglossus Hoplias malabaricus Hypoplectrus nigricans Ictalurus punctatus Jenynsia multidentata Katsuwonus pelamis Kryptolebias marmoratus Labroides dimidiatus

pearl cichlid goldbelly topminnow mouth almighty ear-bar goby elephantnose fish common bully two-spotted goby ruffe giant moray eel French grunt white grunt glowlight tetra sooty grunter Atlantic halibut trahira black hamlet fish channel catfish one-sided livebearer skipjack tuna mangrove killifish bluestreak cleaner wrasse

Lamprologus brichardi Lates calcarifer Lepomis cyanellus Lepomis gibbosus

barramundi green sunfish pumpkinseed sunfish

Lepomis macrochirus

bluegill sunfish

Lepomis megalotis Leuciscus cephalus Limia nigrofasciata

longear sunfish chub humpback limia

Char Count=

437

Pages 370 17, 25, 39, 198, 366 41, 64, 173 65, 172, 266, 415 51, 301, 302, 304, 310 12, 24, 26, 39, 50, 82, 89, 113, 136, 139, 172, 175, 176, 188–189, 191, 195, 197–199, 200, 204, 205, 219, 232, 247, 260, 266, 271, 280 16 50, 302–304, 308, 310 414 39 47 170, 173 61, 62 39 64 269, 287 40, 244, 363 371 47, 221 23 375, 379, 380, 381, 423 41 260, 265 308 193, 303 188, 194, 228 111–114, 116 46, 260, 267, 268, 283, 285, 422 260, 262 23 113 23, 110, 113, 114, 138– 139, 151, 188, 361 16, 20, 22, 45, 139, 188, 198, 205, 266 68 48, 222 88

BLBK374-species

438

BLBK374-Brown

June 1, 2011

17:12

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Species List

Scientific name Limia perugiae Limnothrissa miodon Lota lota Lythrypnus zebra Macropodus opercularis Mallotus villosus Melanochromis auratus Melanogrammus aeglefinus Melanotaenia duboulayi Melanotaenia eachamensis Melanotaenia nigrans Melanotaenia spp. Micropterus salmoides Molva molva Monocirrhus polyacanthus Nannacara anomala Neoceratodus forsteri Neogobius melanostomus Neolamprologus fasciatus Neolamprologus multifasciatus Neolamprologus pulcher Notemigonus crysoleucas Notropis amabilis Notropis cornutus Oncorhynchus keta Oncorhynchus kisutch Oncorhynchus mykiss

Oncorhynchus mykiss Oncorhynchus tshawytscha Oncorhynchus nerka Oreochromis mossambicus Oreochromis niloticus Oryzias latipes Parablennius sanguinolentus Paralichthys olivaceus Paraluteres prionurus

Common name Perugia’s limia Tanganyika sardine burbot zebra goby paradise fish

Pages 89 198 64 265 45, 49, 113, 114, 189, 266 capelin 360, 370 golden mbuna 113 haddock 370, 371 crimson spotted rainbowfish 22, 192, 250, 305, 416 Lake Eacham rainbowfish 43, 197 black-striped rainbowfish 305 rainbowfish 39, 188, 192, 197, 306, 311, 419 largemouth bass 172, 367, 371 gadoid ling 362 barbeled leaf fish 41, 42 goldeneye cichlid 139 Australian lungfish 303, 306 round goby 413 306 281, 282 princess of Burundi 198, 200, 205, 262, 281 golden shiner 15, 173, 227, 245, 364 Texas shiner 188 common shiner 197 chum salmon 179, 248, 384 coho salmon 22, 50, 188, 197, 199 rainbow trout 25, 26, 64, 65, 66, 67, 68, 70, 71, 125, 139, 147, 150, 188, 189, 191, 193, 197, 200, 201, 249, 260, 263, 382, 411, 414, 418, 419, 423 steelhead trout 51, 61, 113, 194 chinook salmon 61, 72, 376 kokanee 178 Mozambique tilapia 113, 116, 420 Nile tilapia 61, 67, 188, 190, 202 Japanese medaka 88, 89, 248, 417 rusty blenny 176 Japanese flounder 387 leatherjacket 46

BLBK374-species

BLBK374-Brown

June 1, 2011

17:12

Copyeditor’s Name:

Trim: 244mm X 172mm

Species List

Scientific name Pelvicachromis taeniatus Perca fluviatilis

Common name

Phoxinus neogaeus Phoxinus phoxinus

finescale dace European minnow

Phyllopteryx eques Pimephales notatus Pimephales promelas

seadragon bluntnose minnow fathead minnow

Plagiotremus rhinorhynchos Plectropomus pessuliferus Poecilia formosa Poecilia latipinna Poecilia mexicana Poecilia picta Poecilia reticulate

blue striped fangblenny Red Sea coral grouper Amazon molly sailfin molly Atlantic molly swamp guppy guppy (Trinidadian guppy)

Poecilia sphenops Poecilia vivipara Pollachius pollachius Pollachius virens Pomacentrus amboinensis

black molly

Eurasian perch

pollock saithe ambon damselfish (reef damselfish) Pomacentrus bankanensis speckled damselfish Pomacentrus moluccensis lemon damsel Pomatoschistus microps common goby Pomatoschistus minutus sand goby Potamotrygon motoro South American freshwater stingray Pseudopleuronectes americanus winter flounder Pterapogon kauderni banggai cardinalfish Pterophyllum scalare freshwater angelfish Pundamilia nyererei Pundamilia pundamilia Pungitius pungitius nine-spined stickleback Regalecus glesne oarfish Rivulus hartii Hart’s rivulus Rutilus rutilus roach

Char Count=

439

Pages 193, 410 64, 139, 149, 197, 198, 200, 202 47 44, 62, 67, 187, 188, 189, 198, 220, 242, 243, 266 41 247 39, 40, 44, 48, 50, 60, 62, 63, 66, 67, 70, 88, 187, 188, 189, 190, 249 46 269, 287 86, 91, 92, 93, 279 88–100, 248, 279 86–7, 91, 92 83–4 23, 41, 83, 92–7, 100, 188, 189, 195, 198, 199, 200, 205, 217–18, 229, 230, 232–3, 267, 271, 361, 383 198 83 225 222, 223, 369 379 150 150 199, 248 249 337 42 196, 198 16, 188, 192, 415 82 82 251, 288, 289 363 139 16, 65, 66, 228

BLBK374-species

440

BLBK374-Brown

June 1, 2011

17:12

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Species List

Scientific name Salmo salar

Common name Atlantic salmon

Salmo trutta

brown trout

Salvelinus alpinus

Arctic charr

Salvelinus fontinalis

brook charr

Salvelinus leucomaenis Salvelinus malma Salvelinus namaycush Sander lucioperca Sarotherodon melanotheron Sciaenops ocellatus Scophthalmus maximus Scortum barcoo Seriola lalandei

white-spot char dolly varden lake trout pikeperch blackchin tilapia red drum turbot jade perch yellowtail amberjack (seriola lalandei.)

Seriola quinqueradiata Serranus subligarius Serranus tabacarius Serranus tortugarum Sicyopterus stimpsoni Sillago maculata Solea solea Sphyraena spp. Spinachia spinachia Steatocranus casuarius Stegastes leucostictus Stegastes planifrons Sternopygus macrurus Stizostedion vitreum Symphodus ocellatus Syngnathus spp. Syngnathus typhle Synodus intermedius Thalassoma bifasciatum

belted sandfish tobacco fish chalk bass Simpson’s goby trumpeter sillago sole barracuda fifteen-spined stickleback lionhead cichlid beaugregory damselfish threespot damselfish longtail knifefish walleye ocellated wrasse pipefish broadnosed pipefish lizardfish bluehead wrasse

Pages 19, 21, 22, 24, 61, 62, 65, 72, 73, 178, 179, 189, 193, 197, 198, 199, 201, 202, 204, 206, 247, 251, 260, 263, 363, 381, 387, 416, 423, 428 12, 22, 61, 110, 139, 155, 188, 197, 198, 248, 369, 416 15, 61, 72, 189, 193, 197, 199, 200, 204, 382, 386, 412, 413, 423 51, 62, 63, 66, 67, 197, 202, 416, 420 372 61 15, 61, 66 72 199 19, 22 22, 423 51 288 388 265 265 265 173 25 42 220 14, 16, 22–26, 173 149 110, 125 47, 266 170 247 89, 91, 139, 151 41 84, 89, 412 40 40, 244

BLBK374-species

BLBK374-Brown

June 1, 2011

17:12

Copyeditor’s Name:

Trim: 244mm X 172mm

Species List

Scientific name Thalassoma lucasanum Thalassoma pavo Theragra chalcogramma Thunnus albacares Tilapia zillii Toxotes chatareus Toxotes jaculatrix Trachurus symmetricus Trachurus trachurus Trichogaster trichopterus Variola louti Xenotoca eiseni Xiphophorus birchmanni Xiphophorus helleri

Common name Cortez rainbow wrasse ornate wrasse Alaska pollock yellowfin tuna redbelly tilapia spotted archer fish banded archer fish jack mackerel horse mackerel blue gourami lunartail grouper redtail splitfin Northern swordtail green swordtail

Xiphophorus maculatus

Southern platyfish

Char Count=

441

Pages 260, 269 46 247 189, 194, 228, 232, 363 110 379 248 219, 223 360 113, 300, 307 269 304, 339 86 63, 112, 114, 117, 119, 250, 280 364

BLBK374-IND

BLBK374-Brown

May 27, 2011

11:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Index

Note: Page numbers with italicized f’s and t’s refer to figures and tables, respectively. abundance estimation, 374–5 adaptation, 153 aggression, 108–28 behavioral changes, 120–21 experiences, 108–9 familiarity, 192 hormones, 120 information contest costs, 110–11 cost-related, 118–19 fighting ability, 111–13 multiple contest experiences, 116 past contests, 113–18 resource value, 110 lateralization, 304 physiological mechanisms, 119–26 prior contests, 109 social learning, 249–50 alarm cues damage-released, 40 learning, 44 sensory perception, 43. See also chemical cues; olfactory cues Allee effects, 366 allocentric orientation, 337–44 allopatry, 91–2 amniotes, 326f AMPA receptors, 124–5 amphibians, 326f androgens, 124, 311 anthropogenic constraints, 73 anticipatory behavior, 376–8 antipredator behavior, 241–3 fast escape response, 303 lateralization, 300–303 predator evasion, 302–3 predator inspection, 301–2. See also predator–prey interactions antipredator response, 36–7 aposematism, 46 aquaculture, 375–84 anticipation, 376–8 capture-based, 389 collective behavior, 383–4 conditioning, 376–8

delay conditioning, 378–9 escapees, 388–9 group level, 377f habituation, 376–8 individual decisions, 383–4 individual level, 377f mortality rates, 386–7 ontogeny, 375–6 operant learning, 382–3 Pavlovian learning, 378–9 personality traits, 155 reward conditioning, 379–82 self-feeding, 383–4 stock enhancement, 384–8 trace conditioning, 378–9 welfare issues, 421–5. See also fisheries archicortex, 325 archistriatum, 325 arginine vasotocin, 125–6 association kin-based, 201–3 learned, 12 simple, 14–15 strength, 231–2 associative learning, 12, 43–4 attack inhibition, 47 attention, 14–15 audience effect, 85–7, 280, 285 autoshaping, 17 avoidance behaviors, 39–40, 369, 371, 416–17 avoidance conditioning, 332–4 baits, 369–70 Batesian mimicry, 46 behavioral syndromes, 140–41 behaviors causation, 136 consistency, 145 coping styles, 140 evolution, 136, 153–4 experiences, 149–50 flexibility, 137 function, 136 growth-mortality hypothesis, 152 mating, 304

Fish Cognition and Behavior, Second Edition. Edited by Culum Brown, Kevin Laland and Jens Krause. C 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.

Char Count=

BLBK374-IND

BLBK374-Brown

444

May 27, 2011

11:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Index

behaviors (Continued ) objective measurements, 142–4 objectivity, 140 ontogeny, 136, 149–50 physical factors, 150 proximate causation, 146 spatial, 177 stability, 145 state-dependent models, 151–3 statistical models, 145–6 subjective measurements, 142–4 variability, 135, 145 blocking, 14 boldness, 138, 140–41 activity, 141 density-dependent selection, 151 experiences, 149–50 measures, 139t brain, 325–49 development, 325–6 divisions, 326f evolution, 325–6 size, 290–91 spatial cognition, 336–7 spatial memory, 337–40 breeding cooperative, 262, 281 preference tests, 413 burrow, 170 by-product hypothesis, 122f, 124 by-product mutualism, 260t, 268–70 cognition, 268–9 foraging, 269–70. See also cooperation bystander effect, 85–7 bystanders, 250 cannibalism, kin-biased, 205 capture success, 50–51 capture-based aquaculture, 389 cerebellum classical conditioning, 327–31 trace, 330–31 delay motor classical conditioning, 328–30 egocentric orientation, 347–9 emotional learning, 331–2 fear conditioning, 334–6 lesions, 329f spatial cognition, 336–7 trace motor classical conditioning, 330–31. See also brain; telencephalon chemical cues, 70–72 anthropogenic constraints, 73 damage-released alarm cues, 60 disturbance cues, 60 familiarity recognition, 187, 190 field-based studies, 73 flexible learning, 62–4 imprinting, 175 information, 176 innate responses, 60–61 learned predator recognition, 70–72 learning, 60–61 neophobia, 60–61 non-predator cues, 66–7 predation risk, 62–4 predator cues, 66–7

predator recognition, 62, 72–3 predator recognition continuum hypothesis, 68–70 risk assessment, 64 risk generalization, 66–8 sensory complementation, 65–6 threat-sensitive learning, 65–6 threat-sensitive responses, 59. See also alarm cues; olfactory cues chondrichthyans, 326f cichlids, 281–3 classical conditioning, 15, 327–31 delay motor, 328–30 trace motor, 330–31 cleaner-client relationships, 283–6 audience effect, 285 categorization of clients, 283–4 cognitive abilities, 286–7 conflicts, 284–5 decision-making, 284–5 eavesdropping, 285 image scoring, 285 individual recognition of clients, 283–4 interspecific, 267–8 interspecific cleaning behavior, 267–8 Machiavellian intelligence, 283–6 pair inspections, 285–6 punishments, 285–6 relationship building, 284 tactile stimulation, 284–5 territories, 283 cognition, 15–17 by-product mutualism, 268–9 kin selection, 261–2 reciprocity, 264–5 spatial, 336–7 trait group selection, 270 welfare, 410 cognitive mapping, 339 collective motion, 218–20 absence of external stimuli, 219–20 dynamic polarized group, 221f models, 218–9 statistical analysis, 219 swarm state, 219–20, 221f torus formation, 221f communication, 307–8 companion fish, 427 comparative psychology, 10 compass orientation, 171–2 competition, 26, 192 familiarity, 26 siblings, 206 conditioned response, 328–30 conditioned stimulus, 328–30, 378–9 conditioning avoidance, 332–4 classical, 327–31 cue competition, 13 delay, 378–9 farmed fishes, 376–8 reward, 378, 379–82 trace, 378–9 conformity, 246 consciousness, 409–10 consistency, 145

Char Count=

BLBK374-IND

BLBK374-Brown

May 27, 2011

11:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Index

conspecifics association, 191–3 familiarity recognition, 186–91 free-ranging fishes, 194–5 kin association in the wild, 201–3 kin avoidance, 205–6 kin discrimination, 201 kin recognition, 196–201 convergent validity, 142 cooperation, 258–71 breeding, 262, 281 by-product mutualism, 268–70 categories, 258–9, 260t, 261–71 egg trading, 265 foraging, 269–70 kin selection, 261–3 predator inspection, 266–7 reciprocity, 263–8 study, 259 territory defense, 263–4 trait group selection, 270–71 cooperative foraging, 269–70 coping styles, 140 corpus cerebelli, 334 corticotropin-releasing factor (CRF), 125 crypsis, 42 cues alarm, 40, 43–4 competition, 13–14 electromagnetic, 172 olfactory, 171, 178–9, 187, 190, 203–4, 223 social, 174 visual, 175–6 cultural inheritance, 96 curiosity, 141–2 D4 dopamine receptor (D4DR) gene, 148–9 damage-released alarm cues, 40, 60 Darwinian evolution, 2 dear enemy effect, 193 delay conditioning, 378–9 delay motor classical conditioning, 328–30 demersal fishes, 362–3 demonstrators, 245 density-dependent selection, 150–51 deprivation level, 12 detection, 41–3 diet, 413–14 dilemma, 264 discriminant validity, 142 discrimination, 201 disturbance cues, 60 dominance hierarchies, 117, 193, 287 dopamine, 148 drive, 12 eavesdropping, 84–7 aggression, 249–50 audience effect, 85–7 benefits, 84–5 bystanders, 250 cleaner-client relationships, 285 Machiavellian intelligence, 279 mate choice, 84 social, 117–18 social learning, 240

ecological selection, 69–70 egg trading, 265 egocentric orientation, 337, 340–44, 347–9 electrolocation, 170 electromagnetic cues, 172 emotional learning, 331–2 endocrine, 120 environmental variation, tracking, 23–6 episodic-like memory, 278 escape behavior, 220 escape speed, 50 escape trajectory, 50 evasion, 49–51 evolution, 153–4 experiences, 108–9 behavioral mechanisms, 115–16 behaviors, 149–50 individual recognition, 117 multiple contests, 116 personality traits, 149–50 projection, 278 winner and loser effects, 113–15 exploration, 14, 306–7 exposure, 241 eyeblink classical conditioning, 328–30 familiarity, 186–96 association function, 191–3 benefits, 192–3 chemical cues, 187, 190 determinants, 195–6 development, 191 free-ranging fishes, 194–5 group living, 192 habitat-based, 231 laboratory studies, 187, 188–9t mate choice, 193 mechanisms, 187–91 network analysis, 195 olfactory cues, 187, 190 schooling preference, 190f territoriality, 193. See also kin recognition fast escape response, 303 fear, 142 fear conditioning, 334–6 fearfulness-reactivity, 141 fish capture, 367–74 attraction, 369 avoidance, 369 baits, 369–70 behaviors, 369 escaping, 372–4 before physical contact with gear, 369–71 after physical contact with gear, 371 spatial distribution, 369 fish schools. See schooling fish fisheries, 362–75 abundance estimation, 374–5 fish capture, 367–74 learning, 366–7 learning skills, 362–3 migration pattern, 363–6 movement, 362–3 sea-ranching, 384–8 social learning, 363–6 spatial dynamics, 362–7

Char Count=

445

BLBK374-IND

BLBK374-Brown

446

May 27, 2011

11:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Index

fisheries (Continued ) stock enhancement, 384–8 welfare, 425. See also aquaculture fishing gear avoidance, 368 escaping, 372–4 before physical contact, 369–71 after physical contact, 371 stimuli, 368–9 fishing vessels, 368 followers, 245 food patch, 21–3 discrimination, 23–4 sampling, 14 foraging, 10–28 competition, 26 conceptual framework, 11f cooperative, 269–70 environmental variation, 23–6 exploration, 14 information transfer, 225 lateralization, 306 leadership, 245 learning, 12–9 patch use, 19–21 performance, 21–3 probability matching, 19–21 reared fishes, 386 sampling, 14 social learning, 247–8 forgetting, 25 free-ranging fishes, 194–5 frequency-dependent selection, 150–51 fright reaction, 40 generalization, 13 generalized learning, 69–70 Generous Tit-for-Tat strategy, 265 giving-up time (GUT), 20 glucocorticoids, 311 group selection, 270 growth-mortality hypothesis, 152 guided learning, 241 guppies, 94–5 habenula, 148 habitats dangerous, avoiding, 40 predation risk, 40 preference tests, 411–13 restocking, 155 habituation, 49, 376–8 Hamilton’s rule, 261–2 Hebb learning, 360 hermaphroditism, 265 homing, 167, 177–9, 307 hormones, 124–5 hunting behavior, 269, 287–8 image scoring, 285 imprinting, 174 incentive value, 12 individual recognition, 117 inertial guidance, 173–4 information asocial sources, 250–52 cost-related, 118–19

eavesdropping, 117–18 individual recognition, 117 past contests, 113–18 social sources, 250–52 winner and loser effects, 113–16 information primacy thesis, 14 information transfer collective response to predators, 220–22 feedback, 222–4 group foraging, 225 mechanisms, 222–4 migration, 225 informational cascades, 224 innate recognition, 69 inspection behavior, 47, 49 intentional hunting, 287–8 interference, 16 internal clocks, 173–4 interspecific cleaning behavior, 267–8 isolation stress, 12 iterated Prisoner’s Dilemma (iPD), 264–5, 267, 286 kin association in wild, 201–3 avoidance, 205–6 cannibalism, 205 discrimination, 201 shoal behavior, 201, 202–3 sibling competition, 206 kin recognition field studies, 203–4 laboratory studies, 199, 203–4 olfactory cues, 203–4 schooling decisions, 197–9t. See also familiarity kin recognition theory, 196 kin selection, 260t, 261–3 cognition, 261–2 cooperative breeding, 262 Hamilton’s rule, 261–2 territory defense, 262–3 kleptoparasitism, 192 landmarks, 168–71, 175f, 339, 340f latent learning, 14 lateral line organ, 171 lateralization, 298–318 aggression, 304 antipredator behavior, 300–303 communication, 307–8 costs, 314–16 environmental factors, 310–11 evidence, 298–9 exploration, 306–7 fast escape response, 303 foraging behavior, 306 hereditary basis, 308–9 homing, 307 individual differences, 308–12 intraspecific variability, 316 mating behavior, 304 personality, 311–12 population biases, 316–17 predator evasion, 302–3 response to novelty, 306–7 selective advantages, 312–14 sex differences, 309–10 shoaling, 304–6

Char Count=

BLBK374-IND

BLBK374-Brown

May 27, 2011

11:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Char Count=

Index

social recognition, 304–6 spatial abilities, 307 visual, 299. See also brain leaders, 245 learned association, 12 learning association, 14–15 associative, 12, 43–4 attention, 14–15 chemical cues, 60–61 cognition, 15–17 drive, 12 emotional learning, 331–2 exploration, 14 and fish feeding, 27 generalized, 69–70 guided, 241 Hebb type, 360 landmarks, 168–71 latent, 14 Machiavellian intelligence, 288 mate choice, 83–4 after maturity, 83–4 memory, 18–19 operant, 382–3 orientation, 167–8, 174–6 Pavlovian, 378–9 predator, 37 predator–prey interactions, 38–9 reinforcement, 12 retention, 70–72 sampling, 14 skill transfer, 18–19 spatial, 169–71, 176–7 spatial-temporal scale, 361f specificity, 44–5 stimulus attractiveness, 12–14 taste aversion, 332–4 threat-sensitive, 65–6 time-place, 173–4 limbic system, 148 limited entry, 37 linear regression line, 146 linkage disequilibrium, 146 location, tracking, 166–7 loser effects behavioral changes, 120–21 by-product hypothesis, 122f, 124 metabolic costs, 121, 123 organizational hypothesis, 123f, 124–5 physiological deviations, 121 losing experience, 113–15 Machiavellian intelligence, 277–91 brain size, 290–91 cleaning behavior, 283–6 cognitive abilities, 286–7, 291 cognitive mechanisms, 287–8 decision-making, 284–5, 289–90 evidence, 279–86 group-living cichlids, 281–3 hypothesis, 277–8, 291 individual recognition, 283–5 information gathering, 279–80, 289–90 intentional hunting, 287–8 learning, 289–90 predator inspection, 280–81

447

social learning, 288 transitive inference, 287. See also learning major histocompatibility complex (MHC), 204–5 male traits, 96–9 male–male contests, 250 mate choice, 81–102 adaptive strategy, 99–101 eavesdropping, 84–7 familiarity, 193 genetic preferences, 94–6 learning after maturity, 83–4 sexual imprinting, 82–3 social learning, 248–9 social preferences, 94–6 mate-choice copying, 88–94 adaptive strategy, 99–101 allopatry, 91–2 benefits, 99–100 costs, 100–101 cultural evolution, 96 early environment, 92–3 experimental evidence, 88–9, 98–9 male traits, 96–9 model fish quality, 93–4 occurrence, 89 social learning, 248–9 sympatry, 91–2 theoretical approaches, 97–8 wild fish studies, 89–91 mating behavior, 304 memory, 18–19 episodic-like, 278 map-like, 345–7 orientation, 167–8 retrieval, 16 spatial, 24–5, 337–40. See also cognition migration information transfer, 225 landmarks, 168–9 olfactory cues, 178–9 predation risk, 40–41 social learning, 244–7, 363–6 tidal streams, 172–3 mimicry, 46 multitrait–multimethod matrix, 142 multivariate personality traits, 146 natural selection, 150–51, 153 neophobia, 60–61, 69 neostriatum, 325 network analysis, 195 neuroendocrine, 148 NMDA receptors, 124–5 non-predator cues, 67–8 norepinephrine, 148 novelty, response to, 306–7 objective measurements, 142–4 observational conditioning, 241–2 observers, 240, 243, 245 olfactory cues, 171 familiarity recognition, 187, 190 information transfer, 223 kin recognition, 203–4 migration, 178–9. See also alarm cues; chemical cues ontogeny, 375–6 operant learning, 382–3

BLBK374-IND

BLBK374-Brown

448

May 27, 2011

11:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Index

optic tectum, 347–9 optimal foraging theory (OFT), 10–11, 20 organizational hypothesis, 123f orientation, 166–79 compass, 171–2 egocentric, 337, 347–9 flexibility, 174 inertial guidance, 173–4 internal clocks, 173–4 landmarks, 168–71 learning, 167–8, 174–6 location tracking, 166–7 memory use, 167–8 salmon homing, 177–9 social cues, 174 social learning, 244–7 spatial learning, 175f spatial learning capacity, 176–7 visual, 175–6 water movements, 172–3 osteichthyes, 326f overshadowing, 13 pain in fish, 417–19 pair inspections, 285–6 paleocortex, 325 paleostriatum, 325 pallium, 330–31 avoidance conditioning, 332–4 hippocampal, 345–7 taste aversion learning, 332–4 past contests, 113–18 winner and loser effects, 113–15 patch use, 19–21 Pavlovian learning, 378–9 pelagic fishes, 363 peptide neuromodulators, 125–6 performance, 21–3 personality traits, 135–57 adaptability, 150–53 anthropocentric thinking, 135 anthropomorphic interpretation, 141–2 consistency, 145 construct validity, 142 coping styles, 140 curiosity, 141 density-dependent selection, 150–51 description, 137 evolution, 153–4 experiences, 149–50 fearfulness-reactivity, 141 fish production and reproduction, 155 frequency-dependent selection, 150–51 growth-mortality hypothesis, 152 heritability, 147 labeling, 142 lateralization, 311–12 multivariate, 146 natural selection, 153 objective measurements, 142–4 objectivity, 140–42 observation, 137 ontogeny, 149–50 physical factors, 150 population dynamics, 155–6 proximate causation, 146–9

shyness-boldness, 138, 139f stability, 145 stable, 137 state-dependent models, 151–3 statistical models, 145–6 stress responses, 147–8 subjective measurements, 142–4 terminology, 137–40 variability, 145 welfare, 420 pet fish, 427 pleiotropy, 146 polarized light, 172 population dynamics, 155–6 populations, fish, structure, 227–9 positive degree correlation, 231 predation risk assessment, 59, 64 flexible learning, 62–4 location tracking, 167 predator–prey interactions, 36–7 sensory complementation, 65–6 threat-sensitive learning, 65–6 predator evasion, 302–3 predator inspection lateralization, 301–2 Machiavellian intelligence, 280–81 reciprocity, 266–7 trait group selection, 270–71 predator recognition anthropogenic constraints, 73 conditioning, 72–3 field-based studies, 73 generalized learning, 69–70 innate vs. learned learning, 69 predator recognition continuum hypothesis, 68–70 predator–prey interactions, 36–52 approach, 47–9 avoidance, 38 crypsis, 42 detection, 41–3 encounter, 39–41 evasion, 49–51 learning, 38–9 migration, 40–41 predation risk in, 36 recognition, 43–6 stages, 37f, 38–51 predators activity pattern changes, 36, 40–41 adaptations, 37f chemical cues, 66–7 counterdefenses, 38–51 information gaining, 47 innate recognition, 69 learned recognition, 62, 69 learning, 37, 69–70 odor, 62–4, 68f sensory perception, 43–4 preparedness, 12 prey antipredator response, 36–7 aposematism, 46 avoidance behaviors, 38, 39–40 defenses, 37f detection avoidance, 41–3

Char Count=

BLBK374-IND

BLBK374-Brown

May 27, 2011

11:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Index

inspection behavior, 47 learned predator recognition, 36, 43–4, 70–72 learning specificity, 44–5 pursuit deterrence, 47 sensory complementation, 65–6 sensory perception, 43–4 social learning, 47–8 threat-sensitive learning, 65–6 prey-subjugation skills, 17–18 prior contests, 109 Prisoner’s Dilemma, 263–4, 280 proactive coping, 140 proactive interference, 16 probability matching, 19–21 proximate causation, 146–9 pseudo-reciprocity, 265 psychology, 10 punishments, 285–6 reactive coping, 140 reactive distance, 50 reactive speed, 50 reciprocity, 260t, 263–8 cognition, 264–5 egg trading, 265 interspecific cleaning behavior, 267–8 predator inspection, 266–7 Prisoner’s Dilemma, 263–4 recognition, 43–6, 186–207 alarm cues, 62 aposematism, 46 associative learning, 43–4 development, 191 familiarity, 186–96 innate, 69 kin, 200 kin association in the wild, 201–3 kin avoidance, 205–6 kin discrimination, 201 kin recognition, 196–201 lateralization, 304–6 learned, 69 learning specificity, 44–5 mimicry, 46 predator, 69–70, 72–3 search images, 45. See also predator–prey interactions recognition genes, 261 recreational fishing, 425–6 reinforcement, 12 relative pay-off sum (RPS) learning, 26 Rescorla–Wagner theory, 18 research, 426–7 restocking, 155 retention, 70–72 retroactive interference, 16 reward conditioning, 378, 379–82 rheotaxis, 173 sailfin mollies, 94–5 salmon homing, 177–9 sampling, 14 schooling fish collective decision-making, 225–7 collective motion, 218–20 collective response to predators, 220–22

familiarity, 190f familiarity recognition, 188–9t foraging, 225 individual identities, 229–32 information transfer, 222–4 feedback, 222–4 mechanisms, 222–4 informational status, 225–7 kinship, 197–9t leadership, 225–7 migration, 225 models, 218–19 population, 227–9 social networks, 229–32 structure, 226 sea-ranching, 384–8 search images, 45 self-feeding, 27, 382–3 sensory perception, 43–4 sensory plasticity, 43 sentience, 409–10 serotonin, 125, 148 sexual imprinting, 82–3 Shepard’s law of generalization, 13 shoals antipredator behavior, 241–3 collective behavior, 219–20 collective decision-making, 225–7 collective motion, 218–20 collective response to predators, 220–2 conformity, 246 escape behavior, 220 fidelity, 196 foraging, 246–7 informational status, 225–7 kin discrimination, 201 kin-based association, 202–3 lateralization, 304–6 leadership, 225–7 population, 227–9 social networks, 229–32 structure, 227–9 Trafalgar effect, 222 shyness-boldness, 149–50 density-dependent selection, 151 frequency-dependent selection, 151 sign-tracking, 17 skill transfer, 17f, 18–19 social cues, 174 social eavesdropping, 117–18 social learning, 47–8, 240–52 aggression, 249–50 antipredator behavior, 241–3 benefits, 240 foraging, 247–8 local enhancement, 240 Machiavellian intelligence, 288 mate choice, 248–9 migration, 244–7 migration pattern, 363–6 orientation, 244–7 stimulus enhancement, 240 trade-offs, 250–52. See also learning social networks, 229–32 community structure, 232–3 social recognition, 304–6

Char Count=

449

BLBK374-IND

BLBK374-Brown

450

May 27, 2011

11:20

Copyeditor’s Name:

Trim: 244mm X 172mm

Index

sound, 171, 176 spatial cognition, 336–7 spatial learning, 169–71, 175f, 176–7, 345–7 spatial memory, 24–5, 337–40, 345–7 spatial navigation, 307, 340–44 spawning, 265 spawning migration, 178 stability, 145 standard network theory, 231 statistical models, 145–6 steroid hormones, 124–5 stimulus attractiveness, 12–4 foraging, 12–4 substitution, 15 stock enhancement, 384–8 stress responses, 147–8 subjective measurements, 142–4 subtle guide hypothesis, 226 sun-compass response, 172 survival benefits, 50–51 S-wiggles, 23 sympatric speciation, 82–3 sympatry, 91–2 tactical deception, 278 taste aversion learning, 332–4 telencephalon ablation, 340–44 embryonary development, 327f emotional learning, 331–2 evolution, 325–6 lesions, 329f map-like memories, 345–7 pallium, 330–31 avoidance conditioning, 332–4 hippocampal, 345–7 taste aversion learning, 332–4 spatial cognition, 336–7 spatial learning, 337–40, 341f, 343f, 344f, 345–7 spatial navigation, 340–44, 342f variation, 326. See also brain; cerebellum temperatures, 150 territoriality, 193, 262–3 territory defense, 262–3 threat-sensitive assessment, 59 tidal streams, 172–3 tide pools, 169, 170f time-place learning, 173–4 Tit-for-Tat (TFT) strategy, 264–6, 280 trace conditioning, 378–9 trace motor classical conditioning, 330–31 Trafalgar effect, 222 trait group selection, 260t, 270–71 cognition, 270 predator inspection, 270–71 transitive inference, 287

unconditioned response, 328–30 unconditioned stimulus, 328–30, 378–9 validity, 142 visual cues, 175–6 visual lateralization, 299 water movements, 172–3 welfare, 405–28 abnormal behavior, 424t aggression, 424t avoidance behaviors, 416–17 behavioral flexibility, 408 cognition, 410 consciousness, 409–10 crowding, 422t definitions, 408–9 deformities, 422t fear in fish, 417–19 fin rot, 422t fish use implications, 420–27 aquaculture, 421–5 companion fish, 427 fisheries, 425 recreational fishing, 425–6 research, 426–7 fish welfare, 406–7 food withdrawal, 423t grading, 422t handling, 422t intraspecific variation, 408 pain in fish, 417–19 personality in fish, 420 preference tests, 407–8 breeding, 413 diet, 413–14 physical habitat, 411–13 social interactions, 414–16 sentience, 409–10 slaughter, 423t stocking density, 423t transportation, 423t viral diseases, 422t winter diseases, 422t winner and loser effects, 113–15 behavioral changes, 120–21 behavioral mechanisms, 115–16 by-product hypothesis, 122f, 124 information integration, 116 Machiavellian intelligence, 280 metabolic costs, 121 organizational hypothesis, 123f, 124–5 physiological deviations, 121 physiological mechanisms, 120. See also information winter sheltering, 206

Char Count=