The Ethology of Domestic Animals An Introductory T...

Copyright © 2017. CAB International. All rights reserved.
The Ethology of Domestic Animals : An Introductory Text, edited by Per Jensen, CAB International, 2017. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/slub-ebooks/detail.action?docID=6273887.
Created from slub-ebooks on 2021-09-29 09:35:58.
The Ethology of Domestic Animals, 3rd Edition

An Introductory Text
The Ethology of Domestic Animals, 3rd
Edition
An Introductory Text
Per Jensen
IFM Biology
Linköping University
SE-58183 Linköping
Sweden
CABI is a trading name of CAB International
CABI CABI
Nosworthy Way 745 Atlantic Avenue
Wallingford 8th Floor
Oxfordshire OX10 8DE Boston, MA 02111
UK USA
Tel: +44 (0)1491 832111 Tel: +1 (617)682-9015

Fax: +44 (0)1491 833508
E-mail: cabi-nao@cabi.org E-mail: info@cabi.org
Website: www.cabi.org
© CAB International, 2017. All rights reserved. No part of this publication may be reproduced in any form or by any means,
electronically, mechanically, by photocopying, recording or otherwise, without the prior permission of the copyright owners.
A catalogue record for this book is available from the British Library, London, UK.
Library of Congress Cataloging-in-Publication Data
Names: Jensen, Per, editor.

Title: The ethology of domestic animals : an introductory text / Per Jensen, IFM Biology, Linköping University.
Description: 3rd edition. | Wallingford, Oxfordshire, UK ; Boston, MA : CABI, [2017] | Includes bibliographical
references and index.
Identifiers: LCCN 2017015882 (print) | LCCN 2017029212 (ebook) | ISBN 9781786391667 (ebook) | ISBN
9781786391674 (epub) | ISBN 9781786391650 (pbk. : alk. paper)
Subjects: LCSH: Livestock--Behavior. | Domestic animals--Behavior.
Classification: LCC SF756.7 (ebook) | LCC SF756.7 .E838 2017 (print) | DDC 636--dc23
LC record available at https://lccn.loc.gov/2017015882
ISBN-13: 9781786391650
Commissioning editor: Caroline Makepeace

Editorial assistant: Alexandra Lainsbury
Production editor: Marta Patiño
Typeset by SPi, Pondicherry, India

Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY
Contents
Contributors
Preface
PART A: BASIC ELEMENTS OF ANIMAL BEHAVIOUR
1 The Study of Animal Behaviour and Its Applications
Per Jensen
1.1 Introduction
1.2 The History of Animal Behaviour Studies
1.3 The Schools of the 20th Century
1.4 Modern Approaches to Ethology
1.5 Applied Ethology
1.6 The Field of Applied Ethology
2 Behaviour Genetics, Evolution and Domestication
Per Jensen
2.1 Introduction
2.2 To What Degree is Behaviour Genetically Inherited?
2.3 Genetic Versus Environmental Influence on Behaviour
2.4 Single-gene Influences
2.5 How Do Genes Affect Behaviour?
2.6 Localizing the Genes
2.7 Epigenetics
2.8 Evolution of Behaviour
2.9 Tracing the Evolutionary History of Behaviour
2.10 Modification and Ritualization
2.11 The Function of Behaviour

2.12 Domestication – An Evolutionary Case History
2.13 Which Animals Were Domesticated?
2.14 Behavioural Effects of Domestication
3 Behaviour and Physiology
Anna Valros and Laura Hänninen
3.1 Introduction
3.2 Central Regulation of Behaviour
3.3 Sleep
3.4 Regulation of Food Intake and Eating
3.5 Reproductive and Maternal Behaviour
3.6 Stress-related Behaviour
4 Motivation and the Organization of Behaviour
Georgia Mason and Melissa Bateson
4.1 What Are Motivational States and Why Are They Important?
4.2 Motivational States and the Organization of Behaviour
4.3 Conceptualizing Motivational Systems and Investigating Their Neurological
Mechanisms
4.4 Motivation and Emotion
4.5 Motivation and Applied Ethology
4.6 Conclusions and Links to Other Chapters
5 Learning and Cognition
Michael Mendl and Christine J. Nicol
5.1 Introduction
5.2 Associative Learning: Pavlovian and Instrumental Conditioning
5.3 Associative Learning: Reinforcement, Punishment, Rewards and Punishers
5.4 Associative Learning: What is Learned?
5.5 Memory
5.6 Discrimination, Generalization, Categorization and Concept Formation
5.7 Social Cognition
5.8 Cognition and Emotion
5.9 Complex Cognition in Early Life
5.10 Metacognition
5.11 Animal Cognition and Animal Welfare
6 Social and Reproductive Behaviour
Daniel M. Weary and David Fraser
6.1 What is Social Behaviour?
6.2 Group Living
6.3 Communication
6.4 Natural Selection

6.5 Common Social Behaviours
6.6 Conclusions
7 Play Behaviour
Susan Held
7.1 What is Play and Who Plays?
7.2 Types of Play and Changes in Play over Time
7.3 ‘Playfulness’
7.4 Mechanisms
7.5 The Functional Benefits of Play
7.6 Play and Welfare in Domestic Animals
8 Introduction to Animal Personality
Hanne Løvlie
8.1 What is Animal Personality?
8.2 Terms Used to Describe Animal Personality
8.3 Assays Used and Methodological Considerations
8.4 Why do Animals have Personality?
8.5 Consequences of Animal Personality
9 Abnormal Behaviour, Stress and Welfare
Linda Keeling and Per Jensen
9.1 Introduction
9.2 Behavioural Disorders
9.3 Stress
9.4 Animal Welfare
9.5 Concluding Remarks
10 Human–Animal Relations
Susanne Waiblinger
10.1 The Human in the Animals’ World
10.2 What is the Human–Animal Relationship?
10.3 Factors Affecting the Quality of an Animal’s Relationship with Humans
10.4 The Human as a Social Partner
10.5 Human–Animal Communication
10.6 Practical Relevance – The Significance of Human–Animal Relationships for
Welfare
PART B: SPECIES-SPECIFIC BEHAVIOUR OF SOME IMPORTANT
DOMESTIC ANIMALS
11 Behaviour of Domesticated Birds: Chickens, Turkeys and Ducks
Joy A. Mench
11.1 Origins
11.2 Social Behaviour

11.3 Foraging, Feeding and Drinking
11.4 Body Maintenance
11.5 Diurnal Rhythm
11.6 Sexual Behaviour
11.7 Egg-laying, Incubation and Behaviour at Hatching
11.8 Care of Offspring
11.9 Offspring Development
12 Behaviour of Horses
Daniel Mills and Sarah Redgate
12.1 Origin
12.2 Time Budgets
12.3 Foraging and Feeding Behaviour
12.5 Mating/Sexual Behaviour
12.6 Behaviour at Birth
12.7 Care of Offspring, Including Nursing
12.8 Offspring Development and Management
12.9 Perception
13 Behaviour of Cattle
Cassandra B. Tucker
13.1 Origins of Cattle
13.4 Diurnal Rhythm
13.5 Mating and Sexual Behaviour
13.6 Parturition
14 The Behaviour of Sheep and Goats
Cathy Dwyer
14.1 Origins of Sheep and Goats
14.4 Habitat Use
14.6 Maternal Behaviour at Birth
14.7 Maternal Behaviour during Lactation
14.9 Management and Welfare
15 Behaviour of Pigs
Marek Špinka
15.1 Origins of Pigs
15.3 Perception and Communication
15.4 Cognition, Emotions and Personalities
15.6 Coping with the Environment
15.7 Mating Behaviour
15.8 Periparturient Behaviour
15.9 Behaviour during Lactation
15.10 Piglet Behavioural Development
16 Behaviour of Dogs
Deborah L. Wells
16.1 Origins and Domestication
16.4 Communication
16.5 Biological Rhythms
16.7 Behaviour at Birth and Parental Care
16.8 Development of Behaviour
16.9 Canine Cognition
17 Behaviour of Cats
John Bradshaw
17.2 Effects of Domestication
17.3 Hunting Behaviour and Object Play
17.4 ‘Umwelt’
17.5 Social Behaviour and Communication
17.6 Courtship and Reproductive Behaviour
17.7 Dominance
17.8 Ontogeny of Behaviour
17.9 Socialization
17.10 The Ethology of Cat–Human Relationships
18 Behaviour of Foxes and Mink Kept for Fur Production
Anne-Lene Hovland, Leena Ahola and Jens Malmkvist
18.1 General Biology
18.2 Origin and Domestication

18.4 Sexual Behaviour, Reproduction and Nursing
18.5 Diurnal Rhythms, Activity Patterns and Foraging Behaviour
18.6 Management and Welfare of Foxes and Mink
18.7 Adaption and Domestication – Concluding Remarks
19 The Behaviour of Laboratory Mice and Rats
Hanno Würbel, Charlotte Burn and Naomi Latham
19.2 Origin and Domestication History
19.3 Diurnal Rhythm, Sleep and Grooming Behaviour
19.4 General Activity, Foraging and Feeding Behaviour
19.5 Social Organization and Social Behaviour
19.6 Mate Choice, Mating Behaviour and Parental Care
19.7 Housing and Welfare in Laboratory Rodents
19.8 Conclusions
Index
Contributors
Leena Ahola, Kuopio, Finland. Email: l.k.ahola@dnainternet.net

Melissa Bateson, Division of Psychology, School of Biology and Psychology, Newcastle
University, Henry Wellcome Building for Neuroecology, Framlington Place, Newcastle
upon Tyne NE2 4HH, UK. Email: Melissa.Bateson@newcastle.ac.uk
John Bradshaw, Honorary Senior Research Fellow, University of Bristol, School of
Clinical Veterinary Science, Langford BS40 5DU, UK. Email:
J.W.S.Bradshaw@bristol.ac.uk
Charlotte Burn, Department of Veterinary Clinical Science, Royal Veterinary College,
University of London, London NW1 0TU, UK.
Cathy Dwyer, Animal Behaviour and Welfare, Animal and Veterinary Sciences, SRUC,
Roslin Institute Building, Easter Bush Campus, Edinburgh EH26 9RG, UK. Email:
cathy.dwyer@sruc.ac.uk
David Fraser, 2357 Main Mall, Faculty of Land and Food Systems,
University of British Columbia, Vancouver BC, Canada V6T 1Z4. Email:
david.fraser@ubc.ca
Laura Hänninen, Research Centre for Animal Welfare, Department of Production
Animal Medicine, PO Box 57, 00014 University of Helsinki, Finland.
Suzanne Held, School of Veterinary Sciences, Animal Welfare and Behaviour Group,
University of Bristol. Email: suzanne.held@bristol.ac.uk
Anne-Lene Hovland, Norwegian University of Life Sciences, Department of Animal and
Aquacultural Sciences, PO Box 5003, N-1432 Ås, Norway. Email:
anne.hovland@nmbu.no
Per Jensen, IFM Biology, Linköping University, SE581 83 Linköping, Sweden. Email:
perje@ifm.liu.se
Linda Keeling, Department of Animal Environment and Health, Swedish University
of Agricultural Sciences, 755 07 Uppsala, Sweden. Email: linda.keeling@slu.se
Naomi Latham, Formerly Department of Zoology, University of Oxford OX1 3PS, UK.
Hanne Løvlie, IFM Biology, Linköping University, SE58183 Linköping, Sweden. Email:
hanne.lovlie@liu.se
Jens Malmkvist, Department of Animal Science, Aarhus University, PO Box 50, DK-8830
Tjele, Denmark. Email: jens.malmkvist@anis.au.dk
Georgia Mason, Department of Animal Biosciences, University of Guelph,
Guelph, Ontario, Canada N1G 2W1. Email: gmason@uoguelph.ca
Joy Mench, Department of Animal Science, University of California, Davis, One
Shields Avenue, Davis, CA 95616-8521, USA. E-mail: jamench@ucdavis.edu
Mike Mendl, School of Veterinary Sciences, University of Bristol,
Langford House, Langford, UK. Email: mike.mendl@bristol.ac.uk
Daniel Mills, Dept of Biological Sciences, University of Lincoln, Riseholme Park,
Lincoln LN2 2LG, UK. Email: dmills@lincoln.ac.uk
Christine Nicol, School of Veterinary Sciences, University of Bristol,
Langford House, Langford, UK. Email: c.j.nicol@bristol.ac.uk
Sarah Redgate, School of Animal Rural and Environmental Sciences, Nottingham Trent
University, Brackenhurst Campus, Southwell NG25 0QF, UK. Email:
sarah.redgate@ntu.ac.uk
Marek Špinka, Department of Ethology, Institute of Animal Science, 104 01 Prague -
Uhříněves, Czech Republic. Email: spinka.marek@vuzv.cz
Cassandra Tucker, Center for Animal Welfare, Department of Animal Science, University
of California, Davis,
One Shields Avenue, Davis, CA 95616-8521, USA. Email: cbtucker@ucdavis.edu
Anna Valros, Research Centre for Animal Welfare, Department of Production Animal
Medicine, PO Box 57, 00014 University of Helsinki, Finland. Email:
anna.valros@helsinki.fi
Susanne Waiblinger, University of Veterinary Medicine Vienna, Institute of Animal
Husbandry
and Animal Welfare, Veterinärplatz 1, Vienna A-1210, Austria. Email:
Susanne.Waiblinger@vetmeduni.ac.at
Daniel M. Weary, 2357 Main Mall, Faculty of Land and Food
Systems, University of British Columbia, Vancouver BC, Canada V6T 1Z4. Email:
dan.weary@ubc.ca
Deborah Wells, School of Psychology, Queen’s University Belfast, Belfast BT7 1NN,
UK. Email: d.wells@qub.ac.uk
Hanno Würbel, Division of Animal Welfare, Vetsuisse Faculty, University of Bern,
Länggassstrasse 120, CH-3012 Bern, Switzerland. Email:
hanno.wuerbel@vetsuisse.unibe.ch
Preface
Modern farm environments are profoundly different from the natural habitats of the ancestors
of today’s farm animals, and through genetic selection both the appearance and behaviour of
the animals themselves have changed. However, the legacy of the ancestors is still obvious,
and some apparently bizarre actions are only possible to understand in the light of the
evolutionary history of the species. On the other hand, some of the behaviours we can
observe in animals in a modern farm or in a laboratory are not part of the normal, species-
specific behaviour at all. They may even indicate that the animal is under stress and that its
welfare is poor. Distinguishing between these possibilities is one important goal of applied
ethology.
In this third edition of The Ethology of Domestic Animals: An Introductory Text, efforts
have been made to expand the subject and to include novel and intriguing topics and
research. For example, a new chapter deals with the behaviour of animals kept for fur
production, and the topics of animal play and animal personality receive detailed attention.
All chapters have been extensively reviewed and updated to include new parts on, for
example, behaviour epigenetics and the role of applied ethology in conservation biology. The
readers we have had in mind when writing the book are, for example, students of animal
science or veterinary medicine, or biology students taking introductory courses in animal
behaviour or applied zoology. We have assumed that the readers will have some fundamental
knowledge of basic biology, for example, physiology and genetics, and probably of modern
animal husbandry, but that this book will be one of their first texts in animal behaviour.
Some authors of the book are new compared with previous editions. Still, all authors are
active in research, in addition to being experienced teachers. Hopefully, this will mean that
the text has a suitable level and content. In order to make the text easier to read, the authors
have reduced the references to a minimum and refer mainly to central publications which will
provide a basis for further studies. Hence, not all cited experimental data are explicitly
referenced, but can mostly be located via some of the other publications in the reference lists.
Per Jensen
Editor
Linköping, April 2017
Part A
Basic Elements of Animal Behaviour
Editor’s Introduction
The first nine chapters of this book introduce the basic concepts and the central subject
matters required for a firm understanding of the biological bases of animal behaviour. The
first chapter provides a historical background, which may help in understanding the questions
that occupy contemporary ethology and its applied branches. In the second chapter, we
approach the important question of how behaviour is controlled by genes (the nature–nurture
problem) and also what this means for behavioural evolution. This chapter also describes the
process of domestication, which is essential for understanding present-day domestic animals.
The third chapter goes to some depth in describing how observable behaviour is a result of
physiological processes throughout the body. The concept of motivation has a long history in
ethology, and has proven to be essential for understanding the cognitive capacities and needs
of animals in captivity, and the concept receives a detailed treatment in the fourth chapter.
This leads naturally on to the fifth chapter, where learning and cognition are in focus. Here,
considerable scientific advance has been made over the last decade, and this is introduced to
the reader together with its relevance for animal welfare.
In Chapter 6, we move towards a more evolutionary and ecological approach to
behaviour. Social and reproductive behaviour are important elements of applied ethology,
since animals are normally kept in groups and they are expected to reproduce. The topic of
the seventh chapter is play behaviour, a phenomenon well-known to all, but until recently
poorly understood by science. In Chapter 8, the reader receives a thorough introduction to the
concepts of individual variation and personality, of high relevance for ethology and animal
welfare. Chapter 9 moves into the important aspects of abnormal behaviour and stress, and
raises some central topics in contemporary applied ethology, such as how the behaviour of an
animal can be used to assess its welfare and whether it is under stress. Furthermore, the
relatively novel insights into positive emotions and welfare are covered. In the tenth and last
chapter of this section, focus shifts to social interactions with the perhaps most important
counterpart of domesticated animals – the human. Extensive research has produced some
exciting new aspects on this relationship, and the chapter provides a broad and up-to-date
introduction to this.
1 The Study of Animal Behaviour and Its
Applications
P. Jensen
1.1 Introduction
Most people have a clear conception of the meaning of the word ‘behaviour’, yet it is
strikingly difficult to define it in a precise way. Since ethology is the science of animal
behaviour, its causation and function, it is worthwhile to start with some consideration of
what type of biological phenomena may be included in the concept. In its simplest form,
behaviour may be series of muscle contractions, perhaps performed in clear response to a
specific stimulus, such as in the case of a reflex. However, at the other extreme end we find
immensely complex activities, such as birds migrating across the world, continuously
assessing their direction and position with the help of various cues from stars, landmarks and
geomagneticism. It may not be obvious which stimuli actually trigger the onset of this
behaviour. Indeed, a bird kept in a cage in a windowless room with constant light will show
strong attempts to escape and move towards south at the right timing, without any relevant
external cues at all.
We would use the word behaviour for both these extremes and for many other activities
in between in complexity. It will include all types of activities that animals engage in, such as
locomotion, grooming, reproduction, caring for young, communication, etc. Behaviour may
involve one individual reacting to a stimulus or a physiological change, but may also involve
two individuals, each responding to the activities of the other. And why stop there? We would
also call it behaviour when animals in a herd or an aggregation coordinate their activities or
compete for resources with one another. No wonder ethology is such a complex science,
when the phenomena we study are so disparate.
But how did it all begin and how has ethology developed into the science it is today?
This chapter will provide a brief overview of some landmarks in history and of the various
fields into which the science has branched over the last decades. The field that will interest us
most in this book is of course the application of ethology to the study of animals utilized by
man.
1.2 The History of Animal Behaviour Studies

No doubt, knowledge of animal behaviour must have been critical for the survival of early
Homo sapiens. How could you construct a trap, or kill dangerous prey weighing several
times your own weight, unless you have a genuine feeling for animal behaviour? So it should
not be any surprise that the earliest ‘documents’ available from humans – cave paintings up
to 30,000 years old – are dominated by pictures of animals in various situations. Aristotle
published written, systematic observations and ideas about animal behaviour more than 300
years bc (Thorpe, 1979).
One of the first to write about animal behaviour in a modern fashion was the British
zoologist John Ray. He published in 1676 a scientific text on the study of ‘instinctive
behaviour’ in birds. He was astonished by the fact that birds, removed from their nests as
young, would still build species-typical nests when adult. Ray was unable to explain the
phenomenon, but noted the fact that very complex behaviour could develop without learning
or practice. Almost 100 years later, French naturalists had an important influence on the
development of the science. For example, Charles Georges Leroy, who was not a formally
trained zoologist, published a book on intelligence and adaptation in animals. Leroy heavily
criticized those philosophers who spent their time in their chambers, thinking about the
world, rather than observing animals in their natural environments. Only then, he argued,
would it be possible to fully appreciate the adaptive capacity and flexibility in the behaviour
of animals (Thorpe, 1979). Figure 1.1 illustrates the typical nest-building behaviour of a sow.
Fig. 1.1. A sow needs no prior experience to be able to construct an elaborate nest before
farrowing. ‘Instinctive behaviour’ such as this fascinated early behavioural researchers.
Another 100 years on, two important scientists deserve to be mentioned. The first is the
British biologist Douglas Spalding, who published a series of papers on the relationship
between instinct and experience. Spalding was way ahead of his time in experimental
approaches. For example, he hatched eggs from hens by using the heat from a steaming
kettle, in order to examine the development of the visual and acoustic senses without the
influences of a mother hen (Thorpe, 1979). The second important scientist is no one less than
Charles Darwin.
Darwin is probably the person who has had the most significant influence on the
development of modern ethology – in fact on all modern biology. Most people know him as
the father of the theory of evolution, which in itself, of course, is the foundation for any study
of animal behaviour. However, he also approached the subject more directly, and his last
published work in 1872, The Expression of the Emotions in Man and Animals, was probably
the first modern work on comparative ethology.
1.3 The Schools of the 20th Century

In the beginning of the 20th century, behavioural research grew fast. However, the
development in the USA and Europe took different directions. American researchers were
influenced by the behaviouristic approach, developed by people such as John B. Watson and
later Burrhus Frederic Skinner. Their work was focused primarily on controlled experiments
in laboratory environments and their subject species par préférence were rats and mice. At
the centre of their interest were the mechanisms of learning and acquisition of behaviour
through reinforcement or punishment (Goodenough et al., 1993). The behaviouristic research
was concerned with finding general rules and principles of learning, and there was a strong
belief that such rules were independent of context. Therefore, the evolutionary histories of
the study subjects, or their ecological ways of life, were regarded as irrelevant for the
research.
In contrast, in Europe the development of the science was dominated by naturalistic
biologists, who spent most of their time observing wild animals in nature. Birds and insects
were favourite subjects, and these researchers were mostly interested in instinctive, innate
and adaptive behaviour. One of the pioneers was Oskar Heinroth, who first started to use the
term ‘ethology’ with the meaning we give it today (Thorpe, 1979). The naturalistic
behavioural biologists shared an important approach with the behaviourists. They were not
particularly interested in mental processes or emotions that may be associated with
behaviour. Such processes were often regarded as unavailable for scientific research, since
they were not considered to be observable. Only much later has a scientific interest in mental
processes emerged, something that will be dealt with more in later chapters in this book.
In the footsteps of Heinroth, we meet two scientists whose influence over modern
ethology cannot be overemphasized: Niko Tinbergen in Holland and Britain, and Konrad
Lorenz in Austria. Tinbergen developed a field methodology of high exactness. He designed
experiments in which details of the environments of free-living animals were altered and
their behaviour could be recorded. He was a true pioneer in experimental ethology (Dawkins
et al., 1991). Lorenz, on the other hand, did not go much into nature, but rather bred his
experimental animals himself and kept many of them almost as pets. He rarely conducted
elaborate experiments and was not prone to quantitative recordings. The strength of Lorenz
was on the theoretical level. He formulated many of the fundamental ideas in ethology, and
developed the first coherent theory of instinct and innate behaviour (Goodenough et al.,
1993).
Lorenz and Tinbergen definitely placed ethology on the solid ground of well-accepted
sciences when they, together with the German researcher Karl von Frisch, were awarded the
1973 Nobel Prize in medicine and physiology.
1.4 Modern Approaches to Ethology

From the 1960s onwards, ethology developed into the science it is today. This was guided to
a large extent by the research programme formulated by Tinbergen, which is still generally
accepted as the fundamentals of ethology (Tinbergen, 1963; Dawkins et al., 1991). This
programme is frequently referred to as ‘Tinbergen’s four questions’ and the four aspects of
behaviour that he used to define the field of ethology were:
1. What is the causation of the behaviour? The answer to this question refers to the
immediate causes, such as which stimuli elicit or stimulate certain behaviour, or which
physiological variables, such as hormones, are important in the causation.
2. What is the function of the behaviour? In this case, the answer describes how the
behaviour adds to the reproductive success, the fitness, of the animal. It therefore has to do
with evolutionary aspects and consequences.
3. How does the behaviour develop during ontogeny? Studies on this question aim at
describing the way behaviour is modified by individual experiences.
4. How does the behaviour develop during phylogeny? This is a clearly evolutionary
question, and usually calls for comparative studies of related species.
Whereas early ethology was occupied mainly with causation, ontogeny and phylogeny,
the research during the 1960s and onwards became increasingly focused on the functional
question. Researchers have outlined new theories on how behaviour evolves through
individual selection at the gene level, and have provided formal mathematical models for
how the functional aspects of behaviour could be determined. The impact of this approach on
contemporary animal behaviour science has been tremendous.
One aspect that was not covered by Tinbergen’s questions is what animals perceive, feel
and know in relation to their own behaviour. As mentioned earlier, this aspect of animal
behaviour was largely considered to be inaccessible to science. However, other scientists
have developed methods and concepts to allow investigation into this area. This has led to a
new branch of the science, emerging in the 1970s, known as cognitive ethology (Bekoff,
2000) (the word cognition means subjective, mental processes – or thinking).
1.5 Applied Ethology

Even early in the development of ethology, it was apparent that the new insights into the
biology of behaviour could be of great value in understanding more of the behaviour of
domestic animals. This branch of science saw a dramatic expansion as the debate on animal
welfare in so-called factory farming (a concept coined by the influential writer Ruth
Harrison) took off in the 1960s. However, applied ethology is not only concerned with
animal welfare. Let us look at a few areas of interest.
Welfare assessment
There is no doubt that the welfare of animals on farms, in zoos and in laboratories dominates
the interest of most researchers in the area. The problems may be formulated for example like
this: many laying hens in the world are kept in cages made of wire mesh, with very little
space available for the animals and almost no substrates for carrying out many of the species-
typical behaviour patterns of poultry (see Fig. 1.2). So, which are the most essential
behaviour patterns for laying hens? Perhaps it is being able to dust-bathe, or to perch during
night, or to perform nest building and lay eggs in a secluded area. All these are typical
poultry behaviours, and there may be others as well. How are the animals affected if they
cannot behave like this? Can the activities be rated in any order of importance to the animals
(Appleby et al., 1993)?
Fig. 1.2. Battery cages (a) and floor housing systems (b) both cause behavioural problems for
laying hens. To estimate the relative importance of different behaviours to animals, thereby
allowing better decisions regarding choice of housing systems, is one important goal of many
researchers in applied ethology. (Photos courtesy of Per Jensen.)
On the other hand, a common alternative to battery cages is a floor system with
thousands of hens in one big group, sometimes with quite high stocking rates. In this
situation, some unwanted behaviour (which can be present both in cages and in floor
systems), such as feather-pecking or cannibalism, might cause great harm to the animals. So,
is it better for the hens to be in a situation where they can perform all the activities mentioned
above, but where the social system may collapse and abnormal behaviour may spread widely
(Hansen, 1994)?
Difficult questions such as these are important aspects of welfare assessment – only
rarely do all indices point in the same direction. In this book, Chapter 9 will examine these
aspects further and describe some of the methods researchers have developed to try to answer
such questions.
Optimizing production
Farm animals are kept to produce food and other essentials for humans, and farmers need
their enterprises to be profitable. It is therefore important that the difference between the
value of what the animals produce (e.g. amount of milk or meat) and the cost the farmer
incurs for this production (e.g. feed, investments and labour) is sufficiently high.
By taking knowledge of animal behaviour into account, such optimization may be easier
to achieve. For example, animals may utilize their feed better if they are fed according to
their species-specific feeding rhythm and in a social context that is adapted to the species
(Nielsen et al., 1996). Social animals may eat more and digest the food better when all in a
group are allowed to feed at the same time.
Social animals that are kept in individual housing systems may be poorer at transforming
feed into valuable products. Likewise, husbandry routines applied at a biologically
inadequate time may decrease the production rate of the animals. Young piglets that are
weaned from their mothers too early and in an abrupt manner show a decreased growth
curve, and mixing of piglets after weaning may also have negative results on production
(Algers et al., 1990; Pajor et al., 1991).
Behavioural control
The essence of keeping animals in captivity is to control their behaviour – by preventing
them from escaping, to control their breeding and making them adapt to the housing
environment. The control is achieved largely by direct human actions, but also by using
technical equipment.
A growing interest has been paid to the nature of human–animal interactions. For
example, researchers have investigated how animals perceive humans and how they
remember experiences with human behaviour. This may help farmers and others to interact
more smoothly with their animals (Hemsworth and Barnett, 1987).
An increasing trend in animal farming is to use technical inventions in animal husbandry.
For example, group-housed pregnant sows are often fed from an electronic feeding station,
which the animals to some extent are required to control themselves. The sows are equipped
with transponders that allow them to open the feeding stations and obtain their individual
feed rations. However, such systems must be carefully designed to avoid problems. For
example, the social hierarchy of a group of sows may lead to some individuals occupying the
feed entrance, biting and wounding other animals and thereby destroying the functionality of
the system. By means of ethological knowledge, technical equipment can be designed to
work better for the animals (Broom et al., 1995).

Behavioural disorders
Housing systems such as those described earlier, malfunctioning technical equipment or poor
human management may all lead to various behavioural disorders. Aggression levels may
become excessively high, dramatic behaviour such as cannibalism may develop, and several
other types of abnormal behaviour may be seen as well (Lawrence and Rushen, 1993). This
is not only the case for farm animals. Many pets develop unwanted and abnormal behaviour,
such as owner-directed aggression, uncontrolled urination and defecation in the home, or
anxiety-like states.
The characterization and understanding of abnormal behaviour is a central aspect of
applied ethology. Sometimes the behaviour can be cured by behavioural therapies, such as
enrichment of the home environment or stimulation of other behaviour. At other times,
research can provide insights that may help prevent the behaviours from ever developing.
Behaviour and conservation biology
Animal species are going extinct or becoming threatened at an accelerating rate, largely as a
consequence of human activities. This includes hunting, pollution, climate change, habitat
loss and introduction of foreign species and diseases. Any countermeasure against this will
be greatly helped by a thorough understanding of the behaviour of the species under concern.
There are essentially two main fields of animal conservation biology. In in-situ
conservation, the focus is on improving the environmental conditions in the natural habitat of
the species to the extent that the species will be able to survive where it lives. In ex-situ
conservation, on the other hand, animals are taken from the wild to be reared and bred in
captivity with the goal of eventually reintroducing their progeny when suitable habitats are
available.
For in-situ programmes, knowledge is required about, for example, normal social
structures, foraging strategies and mating behaviour of the species. Unless such things are
known, efforts to rescue the environmental conditions may prove futile, since it is not
possible to prioritize measures in relation to the needs of the species. Furthermore, behaviour
can provide essential indicators for monitoring populations (Berger-Tal et al., 2011).
Knowing the normal movement patterns, vigilance and territoriality of a species can provide
alarm signals when something in the habitat causes drastic changes in the population.
With respect to ex-situ conservation, the above-mentioned topics of ethological
knowledge are of course equally important but, on top of that, applied ethology can provide
other important insights. For example, the captive environment will always be a compromise
between the needs of human caretakers and those of the animals. In addition, not only is
behaviour important to consider, but also things such as health and hygiene. Sometimes, the
ambition to keep animals fit and free from disease can be in conflict with their behavioural
needs. It may be difficult to offer the right food in the species-specific manner, and the
possibility to use sufficiently complex cage environments may be in conflict with cleaning
needs. Here, applied ethology can offer helpful insights to aid in designing the best captive
environments possible.
A fine example of successful contributions from ethology to animal conservation is the

case of the North American whooping crane (see Fig. 1.3), which at one point was almost
extinct. An extensive ex-situ conservation programme utilizing ethologically founded rearing
methods was coupled with ambitious in-situ field studies after reintroduction to gradually
improve the breeding (Kreger et al., 2005). This has led to an impressive recovery of the
population.
Fig. 1.3. The whooping crane, a charismatic North American bird species (a), was saved
from extinction after intense work with ex-situ breeding and ethological fieldwork (b) to get a
detailed picture of the natural behaviour of the species. (Photos courtesy of Inma Estevez (a)
and Michael Kreger (b).)
1.6 The Field of Applied Ethology

As should be clear from these accounts, ethology is a science that may offer many different
sorts of applications in situations where humans utilize or are responsible for animals for
various purposes. Whereas animal welfare assessment clearly dominates in the applications
of this science, it is by no means the only way in which knowledge of behaviour can be used.
Applied ethologists are normally concerned with all four of Tinbergen’s questions. The
causation and ontogeny of behaviour are essential aspects of understanding, for example,
how abnormal behaviours develop and how they can be prevented. Phylogeny and function
of behaviour are often less emphasized, but many studies have advanced our understanding
of domestic animal behaviour greatly by considering how it can have evolved in the
ancestors and how it may have been affected by domestication (Fraser et al., 1995).
Experimental studies tend to dominate, but important scientific data have been made
available through studies of domestic animals in wild-like conditions (which will become
obvious in Chapters 11–19, where accounts are given of the normal behaviour of some
important domestic species).
In the optimal situation, applied ethology research concerns all the fields outlined above.
They are all interlinked: poor human–animal, or equipment–animal, interaction may cause
poor welfare, which in turn leads to behavioural disorders and reduced production, or
reduced ability to reproduce in captivity in the case of ex-situ conservation. Applied ethology
is therefore an essential part of the proper keeping and rearing of animals. And last but not
least: as will become obvious in this book, understanding the behaviour of domestic animals
is a fascinating aspect of biology in its own right.
References
Algers, B., Jensen, P. and Steinwall, L. (1990) Behaviour and weight changes at weaning and regrouping of pigs in relation
to teat quality. Applied Animal Behaviour Science 26, 143–155.
Appleby, M.C., Smith, S.F. and Hughes, B.O. (1993) Nesting, dust bathing and perching by laying hens in cages: effects of
design on behaviour and welfare. British Poultry Science 34, 835–847.
Bekoff, M. (2000) Animal emotions: exploring passionate natures. BioScience 50, 861–870.
Berger-Tal, O., Polak, T., Oron, A., Lubin, Y., Kotler, B.P. and Saltz, D. (2011) Integrating animal behavior and conservation
biology: a conceptual framework. Behavioral Ecology 22, 236–239.
Broom, D.M., Mendl, M.T. and Zanella, A.J. (1995) A comparison of the welfare of sows in different housing conditions.
Animal Science 61, 369–385.
Dawkins, M.S., Halliday, T.R. and Dawkins, R. (eds) (1991) The Tinbergen Legacy. Chapman & Hall, London.
Fraser, D., Kramer, D.L., Pajor, E.A. and Weary, D.M. (1995) Conflict and cooperation: sociobiological principles and the
behaviour of pigs. Applied Animal Behaviour Science 44, 139–157.
Goodenough, J., McGuire, B. and Wallace, R.A. (1993) Perspectives on Animal Behavior. Wiley, New York.
Hansen, I. (1994) Behavioural expression of laying hens in aviaries and cages: frequency, time budgets and facility
utilisation. British Poultry Science 35, 491–508.
Hemsworth, P.H. and Barnett, J.L. (1987) Human–animal interactions. In: Price, E.O. (ed.) Veterinary Clinics of North
America: Food Animal Practice, Vol. 3. W.B. Saunders, Philadelphia, Pennsylvania, pp. 339–356.
Kreger, M.D., Hatfield, J.S., Estevez, I., Gee, G.F. and Clugston, D.A. (2005) The effects of captive rearing on the behavior
of newly-released whooping cranes (Grus americana). Applied Animal Behaviour Science 93, 165–178.
Lawrence, A.B. and Rushen, J. (eds) (1993) Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare.
CAB International, Wallingford, UK.
Nielsen, B.L., Lawrence, A.B. and Whittemore, C.T. (1996) Feeding behaviour of growing pigs using single or multi-space
feeders. Applied Animal Behaviour Science 47, 235–246.
Pajor, E.A., Fraser, D. and Kramer, D.L. (1991) Consumption of solid food by suckling pigs: individual variation and
relation to weight gain. Applied Animal Behaviour Science 32, 139–156.
Thorpe, W.H. (1979) The Origins and Rise of Ethology. Heinemann Educational Books, London.
Tinbergen, N. (1963) On aims and methods of ethology. Zeitschrift für Tierpsychologie 20, 410–433.
2 Behaviour Genetics, Evolution and
Domestication
P. Jensen
2.1 Introduction
As we noted in the first chapter, people have been amazed for hundreds of years by the fact
that animals are often able to perform extensive and complex, seemingly goal-directed,
behaviour with no prior learning possibilities. Birds may build elaborate nests and migrate to
the typical wintering sites even if they are raised out of contact with other members of the
same species. Darwin suggested that such ‘instinctive’ behaviour was somehow transferred
from the parents without the need for any learning processes and, hence, ‘instincts’ would be
suitable raw material for evolution. Behaviour would then evolve as adaptive traits just like
any morphological or physiological characteristic.
When Darwin formulated his ideas, he knew nothing about genes. The work of the
‘father of genetics’, Gregor Mendel, did not become known to the scientific world until
several decades after Darwin’s death, and the discovery of the chemical structure of DNA
was still almost 100 years away. However, Darwin’s suggestion was prophetic: we now know
that genes, built up by DNA, contain the chemical instructions necessary for the development
of behaviour in an animal, and that evolution modifies the frequency of genes over
generations. Evolution therefore moulds the behaviour of species and individuals. In this
chapter, we will examine some of the evidence for this and its implications.
The focus of this book is on domesticated animals, so of course we will need to ask how
domestication has come about and how animals have been changed as a consequence of
being domesticated.
2.2 To What Degree is Behaviour Genetically Inherited?

Anyone interested in dogs and dog breeds would have no problem in accepting the idea that
even complex behaviour is genetically inherited from the parents. Retrievers have puppies
that are, on the whole, more inclined to retrieve and carry things around than other breeds,
while border collies have puppies with a strong tendency to herd. Some breeds are more
famous for being aggressive, while others are generally viewed as friendly or docile.
Observations such as these are suggestive, but they could perhaps also be explained, for
example, by the offspring learning a specific behaviour from their parents, or by the fact that
some types of owners tend to keep some types of dogs and therefore may affect their dogs’
behaviour differently. So how can we tell whether and to what extent behaviour is inherited
genetically?
One of the more famous examples of a clear genetic influence on behaviour comes from
studies of lovebirds, Agapornis spp. (Dilger, 1962). One species, Fischer’s lovebird, carries
nest material (e.g. strips of paper) one piece at a time in the beak. The closely related peach-
faced lovebird tucks the strips in between the rump feathers and is therefore able to carry
more nest material at each flight. Hybrids between the two show a poorly functioning
mixture of the behaviour. They attempt to tuck material between the feathers, fail to let go of
it, pull it out again, and then start the sequence over again. After several months of practising,
the behaviour can become at least partly successful, in that the birds manage to transport
some material back to the nest site, but not in a manner typical of any of the parents. The
strange behaviour of the hybrids is consistent with an intermediate (non-dominant)
inheritance pattern of one or more alleles (an allele is one variant of a gene that is present in
the population; hence every gene may have several alleles).
Another source of evidence is from the bulk of experiments where the tendency to
behave in a certain manner has been artificially selected for over generations. Mostly, this
leads to individuals of the selected lines being more and more inclined to behave according
to the selection criteria, indicating a strong genetic basis for the behaviour. For example, in
fruit flies (Drosophila melanogaster) mating time may vary considerably between
individuals. In a classical experiment (see Fig. 2.1), 100 fruit flies of both sexes were placed
in a chamber and the ten fastest and the ten slowest maters were determined. They were
allowed to breed separately, and the procedure was repeated in each new generation. After 25
generations, mating time differed between populations by about 30 min (Manning, 1961).
Similarly, the tendency for positive versus negative geotaxis (i.e. to move towards or against
gravitation on a vertical surface) varies between fruit flies, and selection for the strength of
this tendency produced rapid changes in the population over a few generations (Hirsch,
1967).
Fig. 2.1. Results of a selection experiment where fruit flies were selected for long or short
mating time. (From Manning, 1961.)
Experiments such as these indicate that genes can control even complex behavioural
responses. When there is a quantitative response – behaviour becomes gradually faster or
slower, or a tendency is gradually strengthened or weakened – it is most likely that several
genes interact in a quantitative fashion in producing the observed behaviour phenotype. Later
we will examine some cases where a few or even single genes affect the behavioural output.
Do studies of insects have any relevance for applied ethology, which is mostly concerned
with mammals and birds? Yes; genetic control is universal among animals. Not only are the
tools – DNA, RNA, etc. – identical, the way in which the control is exerted is largely the
same (we will deal with the mechanisms later in the chapter). Moreover, the sequencing of
the complete, or almost complete, genomes of organisms ranging from yeast via the
nematode Caenorhabditis elegans and fruit flies to chickens and various mammals, including
humans, has revealed large similarities in structures and functions of many genes. Hence,
genetic instructions for behaviour may sometimes be very similar across species and phyla.
There is of course also a lot of direct evidence of genetically determined behavioural
differences among domestic animals. For example, laboratory mice were selected for their
tendency to build nests. After 15 generations, one line collected about 50 g of cotton for use
as nest material, compared with 5–10 g in the other line (Lynch, 1980). As another example,
laying hens sometimes develop a problematic behaviour called feather-pecking, where they
damage the feathers of flockmates. The tendency to develop the behaviour differs between
strains, and genetic selection against the behaviour can cause it to decrease over a few
generations (Kjaer and Sorensen, 1997). This indicates a genetic basis for this abnormal
behaviour.
Genetic selection of desired behaviour, or against undesired behaviour, is therefore also
an important means for improving animal welfare. For example, the level of fearfulness has
been reduced by selection in both poultry and quail (Jones and Hocking, 1999). Fearfulness
has also been reduced in some fur animals, along with undesirable behaviours like chewing
and destroying the fur in mink (Malmkvist and Hansens, 2001).
How can there be genes controlling onset of abnormal behaviour such as feather-
pecking? Probably this is not how it works. We must always remember that even though
selection is performed on a well-defined trait, the observed responses can often be explained
by indirect effects. The reduction of feather-pecking by selection has been suggested to be
indirectly caused by a general decrease in foraging activity, which reduces the overall
tendency of the birds to peck (Klein et al., 2000). Therefore, in this case we cannot be certain
that there is a direct genetic control of feather-pecking as such – in fact it may seem more
logical to assume that the effect is often indirect.
2.3 Genetic Versus Environmental Influence on Behaviour

Sometimes, people misinterpret the fact that genes contain instructions for the behaviour of
an animal. The misinterpretation is often referred to as genetic determinism: the belief that if
there is a genetic control, the behaviour of an individual will be inflexible and determined
from the point of fertilization. The following example illustrates nicely why this is wrong.
Rats will normally learn easily to run through a maze, in order to reach a goal box
containing some food. With increasing number of trials, they will make fewer and fewer
errors, in the sense that they become less and less likely to run into blind alleys. However,
there is considerable individual variation in how easily this task is accomplished. In a famous
experiment, rats were selected depending on how fast they learned a specific maze. In only a
few generations, the selected lines had separated, with almost no overlap, into a line of so-
called ‘bright’ rats and another of so-called ‘dull’ rats (Tryon, 1940). These lines could be
preserved over generations and generations, and the offspring would show the same pattern
in learning ability as their parents.
A possible interpretation of this experiment is that learning capacity is controlled by
genetic factors. Once it is known which population an animal descends from, whether its
father and mother are ‘dull’ or ‘bright’, it would be possible to predict accurately how well
this animal would manage a maze-learning test. But this conclusion is oversimplified, as
shown by the following later study.
Rats from the same populations were raised in three different environments: (i) the
standard laboratory environment in which the populations had been maintained for
generations; (ii) an impoverished environment without bedding material or any other
interesting stimuli; and (iii) an enriched environment, where different substrates for
manipulation and stimulation were added to the cage (Cooper and Zubek, 1958). When these
animals were tested in the maze, those reared in the standard environment showed the same
differences between lines as expected: the ‘dull’ line had considerably larger difficulties in
mastering the problem.
However, when the behaviour of the rats from the two lines reared in a restricted
environment was compared, an interesting result emerged. The differences between the
strains had disappeared and they both performed as poorly as the ‘dull’ rats reared in the
standard environment. Furthermore, there were also no differences between the strains when
they had been reared in enriched environments, but this time they both performed as good as
the ‘bright’ strain from standard cages (see Fig. 2.2).
Fig. 2.2. (a) Results of selection for maze-learning capacity in rats (from Tryon, 1940). (b)
Results of a maze-learning experiment in which individuals from each of the selected lines
were raised in the standard laboratory environment, an impoverished laboratory environment
or an enriched one (from Cooper and Zubek, 1958).
These results show how careful we must be in inferring deterministic genetic control over
behaviour when a genetic effect has been demonstrated. Clearly, genes supply the organism
with the necessary basis for a particular behaviour – the limbs, muscles, nerves and sensory
organs, and a central nervous system – but any behaviour is likely to develop in interaction
with the environment in which the animal lives. Rather than considering genes as
determinants of behaviour, we should consider genetic traits to be predispositions, which bias

animals towards certain reactions and developmental pathways.
This approach also helps to understand the possible indirect genetic influence on feather-
pecking, described earlier. If there is a genetic predisposition to peck at different substrates,
feather-pecking may develop when the environment lacks necessary pecking stimuli. Hens in
a rich environment may therefore never develop the behavioural disturbance, although the
same individuals would do so under poorer conditions.
2.4 Single-gene Influences
Since behaviour is a complex trait, involving identification of stimuli acting on different
sensory channels, central nervous processing of information and the concerted actions of
groups of muscles, it may seem implausible that single genes would be able to control the
expression of behaviour. Nevertheless, there are several examples of such effects of a single
gene, or at least of a very few genes.
In a comprehensive study of behaviour genetics, two breeds of dogs with very divergent
behaviour, the basenji and the cocker spaniel, were crossed (Scott and Fuller, 1965). By
producing two successive generations coupled with backcrosses and standard genetic
analysis methods, it was possible to identify simple genetic control mechanisms for rather
complex behaviour patterns. For example, struggling when being restrained – more
pronounced in basenjis – was consistent with inheritance by one single allele, with the
basenji trait being dominant. The tendency to bark was found to be inherited through two
genes, with dominant alleles for a low stimulus threshold (basenjis of course are famous for
having a very high barking threshold, i.e. they rarely bark).
In fruit flies there are many examples of effects of separate loci on complex behaviour.
One mutant strain, called ‘dunce’, has a mutation in a single gene on the X chromosome.
These flies cannot learn to avoid an odour when trained in an experiment that allows normal
flies to associate the odour with an electric shock (Quinn et al., 1974).
2.5 How Do Genes Affect Behaviour?

Studies like those summarized above demonstrate clearly that genes have a strong influence
over behaviour. But what does a gene actually do to affect behaviour, trapped as it is inside
the nucleus of the cell? How is it possible that mutations in one or a few genes can change
complex patterns of behaviour? DNA is a fantastic molecule for storing information, but the
information in a gene is used only for providing the instructions for how and when a protein
should be produced. What is the link between proteins and behaviour?
Since genes carry the codes for synthesis of proteins, they thereby regulate important
aspects of metabolism, hormone concentrations, etc. Of course, the direct interaction between
the environment and the animal, including its behaviour, usually happens via the nervous
system. However, the winding up of neurons into an entire nervous system, as well as the
design and sensitivity of receptors and sense organs, are all processes that are regulated by
genes during the embryology and further ontogeny. In the adult animal, levels of hormones,
neurotransmittors and other substances that are extremely important for behavioural control
are regulated by genes.

The detailed pathway from the synthesis of a specific protein to an observed behaviour is
long and complex, and we have actually only started to understand some of it. For example,
the ‘dunce’ fruit flies, discussed above, have a defect gene that is normally responsible for the
production or activity of an enzyme called cyclic AMP phosphodiesterase, which degrades
cyclic AMP (a common intermediate molecule in cell metabolism). So either the enzyme or
the cyclic AMP itself is involved in olfactory learning. Once pathways such as these are more
fully understood, they will no doubt provide deep insights into the biology of behaviour.
Even if we have seen examples of single-gene effects, most behaviour is usually not
controlled by a single gene. However, a single gene can have a crucial role for the normal
appearance of a certain behaviour pattern. Single mutations may alter the behaviour
drastically, for example by changing the development of important structures (Huntingford,
1984). One mutation in the nematode C. elegans changes the myosin filaments in parts of the
body and produces uncoordinated movements. Another mutation affects the structure of
chemoreceptors in the head and changes the receptivity of the worm to chemical stimuli. The
slow-mating fruit flies, which we discussed earlier, were suggested to have an altered
production of juvenile hormone as a result of selecting genes for this.
2.6 Localizing the Genes

Even if it is clear that genes predispose animals to specific types of behaviour, the
mechanisms whereby this is happening are still largely unknown, particularly in vertebrates.
Usually we do not even know exactly which genes are involved in controlling a behaviour
pattern and we therefore cannot understand what these genes might do to exert their effects.
With the rapid development of genetics, identifying the genes and their functions is likely to
be an important area of research in the immediate future.
A first step towards identifying genes tied to specific behavioural traits is to localize a
position on the chromosomes that is related to the expression of the behaviour. This can
sometimes be determined because a behaviour may be inherited in close correlation with
some other trait, for which the gene and its location are known – what geneticists refer to as
linkage. It could, for example, be a colour pattern or some other observable trait. One can
then conclude that the behaviourally linked gene must be located close to the gene for the
other trait.
With modern gene technology, new possibilities for gene identification have emerged.
One common method is to use quantitative trait loci (QTL) analysis or genome-wide
association studies (GWAS). With these analyses, the genomes of animals are mapped with
respect to the occurrence of specific DNA markers, usually non-coding sequences of base
pairs, which vary between individuals in specific chromosomal locations. The correlation
between a particular phenotypic trait – for example, a particular behaviour pattern – and the
occurrence of different markers can then be assessed (this sounds simple, but requires many
animals, a lot of laboratory magic and advanced statistical computing). When a certain trait
has been found to be correlated with the occurrence of a certain marker, one can imply that
the gene that affects the trait is located close to that marker. By using the increasingly
detailed sequenced genomes that are available for more and more species (e.g. humans),
relevant genes in that chromosome region can be identified and their products studied (see
Fig. 2.3).
Fig. 2.3. When junglefowl are crossed with modern laying hens, and the F1 generation is
subsequently intercrossed, the resulting F2 generation exhibits a large genetic and phenotypic
variation. Such breeding experiments are important tools in QTL studies (i.e. experiments
designed to localize the chromosomal loci of genes controlling specific traits).
Genetic association analyses have identified chromosomal regions influencing such

variable aspects of behaviour as the tendency of honeybees to sting, the cyclicity of activity
in mice, the preference for alcohol in mice and the hyperactivity syndrome of rats.
Methods of identifying the actual genes and their products are developing rapidly, and in
the next few years we will most likely see a tremendous increase in our understanding of
behaviour genetics. Once the genes are established, it is necessary to move from correlation
to causation. This requires experimental methods where the suspected gene is manipulated in
order to verify its effects on the observed behaviour. One important source is so-called
‘knockout’ animals. These are strains of animals where specific genes have been made silent.
The development of such an animal, compared with normal ones, therefore tells us important
things about the function of the knocked-out genes. As an example, mice lacking the gene for
oxytocin (a hormone involved in many different processes such as parturition and milk
ejection) not only – as expected – lack the capacity to eject milk, but also show a reduction in
aggressive behaviour (Crawley, 1999). Recently, a novel cheap and comparatively simple
method has been developed for genetic engineering, the so-called CRIPSR-Cas9 technique.
This uses molecules from a bacterial virus defence system that can be engineered to modify
(e.g. to silence) any DNA sequence in a cell (Konermann et al., 2015). The low price and
simple technology mean that this method is likely to be implemented even in small
laboratories, and may help unravel the effects of single genes as well as of gene complexes
on behaviour.
2.7 Epigenetics
Lately, it has become clear that the genome is not as insensitive to environmental impact as
we used to think. In fact, the regulation of gene activity is highly dynamic and responds in
intricate ways to events encountered during the lifetime of an individual. An increasing
amount of research even shows that such changes may sometimes be transferred across
generations through the germ cells. In essence, this means that some acquired genetic
variants can be inherited. This insight is totally contradictory to established textbook facts
from the 20th century. What then are the mechanisms that might lead to this, and that are
grouped under the common heading ‘epigenetics’ – something that is ‘over’ or ‘in addition
to’ genetics?
A common definition of epigenetics refers to any chemical modification of DNA that
affects gene expression without affecting the DNA sequence. These changes should also be
heritable at least over a normal cell division, and sometimes also over meiosis (gamete
formation). Two such chemical changes have attracted particular attention: (i) the addition of
a methyl group (CH3) to the cytosine base at specific positions (referred to as DNA
methylation); and (ii) modifications of the histone proteins that pack DNA in the nucleus.
The consequence of both is to affect the expression of one or several genes in the vicinity of
the epigenetic change.
One of the first demonstrations of the importance of epigenetics for animal behaviour
was a pioneering study of maternal behaviour in rats (summarized by Kappeler and Meaney,
2010). Offspring raised by mothers with a more intense maternal behaviour had an increased
resilience to stress later in life, and also became more careful mothers themselves. A central
factor underlying this was shown to be epigenetic modifications (DNA methylation and
histone alterations) in genomic regions affecting the expression of the glucocorticoid receptor
gene, which in turn affected the entire stress response (see Chapter 9). Hence, the experiences
of the pups during their early postnatal period caused a modification in the genome, affecting
important behaviour throughout life.
Several studies verified this later and also showed that some epigenetic changes may
persist into coming generations (Jensen, 2015). For example, chickens stressed during their
first weeks of life later showed changes in learning ability and modified physiological stress
responses, both of which were transferred to their offspring. This was associated with a
modified gene expression profile in the brain, which was also mirrored in their chicks.
2.8 Evolution of Behaviour

Now that we have seen how behaviour depends on genetic predispositions, we have two of
the necessary keys to understand how behaviour can have developed during evolution.
Nowadays it is basic biological knowledge that animals are products of a long evolutionary
history, whereby their anatomy and physiology have been adaptively shaped to what we see
today. But is that also the case for behaviour?
In order for any trait to be modified by evolution, three principles are required, which can
be deducted from Darwin’s original writings:
1. The principle of variation. This states that a trait must vary between the individuals of a
population. If all individuals are identical, no evolution of the trait is possible.
2. The principle of genetic inheritance. This principle requires that some of the variation in
the population must be of genetic origin. It is not necessary that the trait is genetically
determined; only that genes have some influence over the phenotypical expression of the
trait. It follows from this principle that, on average, when a trait is genetically inherited,
individuals resemble their parents more than they resemble other randomly chosen
individuals of the population with respect to this trait. Furthermore, the closer the
relationship between any two individuals, the larger the resemblance. Variation has
traditionally been attributed to genetic mutations that cannot be transmitted to the offspring
but, as we have seen above, could also be caused by epigenetic changes induced by
environmental conditions.
3. The principle of natural selection. According to this, some variants of the trait must cause
different abilities of the individuals to reproduce. If the reproduction capacity is enhanced,
the trait will increase in frequency over generations; if it is reduced, the frequency will
decrease.
In order for evolution to modify any trait, all of these principles must be fulfilled
simultaneously. In fact, when they are all fulfilled, evolution is bound to happen. In the case
of behaviour, we have already seen that there is often a large variation within populations
(e.g. in the case of maze-running ability in rats), and that this variation is often partly caused
by genetic differences between individuals. What remains is to show that this genetically
based influence causes different reproductive success.
Indeed, most contemporary evolutionary research in behaviour is devoted to examining
the reproductive advantages of having a certain behaviour rather than any other possible
alternative. Obtaining and defending a territory may increase the chance of getting a mate
and reproducing, and territory quality and size are often closely related to reproductive
success. Searching for food in a manner that causes minimal exposure to predators may seem
an obvious example, and different food-searching patterns do have effects on the efficiency
of reproduction. Some decades of intense research have produced innumerable similar
examples of behaviour patterns that follow the principle of natural selection, and this is often
covered in the subject called behavioural ecology (Krebs and Davies, 1991).
2.9 Tracing the Evolutionary History of Behaviour

Behavioural archaeology would be a rather fruitless field of science, but occasionally fossils
actually carry some information about the behaviour of animals long extinct. For example,
close examination of the skeleton and feather structure of the oldest bird fossil,
Archaeopteryx, has led to the suggestion that bird flight developed through gliding from tree
branches, and running and leaping after prey on the ground (Alcock, 2001) (although other
interpretations have also been suggested). Mostly, however, we have to rely on comparisons
between animals that are living and behaving around us today.
The method applied in these kinds of studies is to take advantage of phylogenetic trees,
which may be deduced from traits other than behavioural, and then map the behaviour seen
in closely related species on to this tree. For example, in one study of courtship displays in
manakins, a group of small, fruit-eating tropical birds, a phylogenetic tree was constructed
based on similarities between species in the structure of the vocal apparatus (syrinx). The
male courtship behaviour of the different species was then examined in detail. There were 44
behavioural elements that occurred in at least one of the 28 examined species. By mapping
the species to the phylogenetic tree based on how many of the behavioural traits they shared,
it was possible to deduce a possible history of how present-day courtship has evolved from
simpler behaviour in different ancestors. Mostly, evolution has led to more elements being
included in the repertoire but, in some cases, behavioural patterns have been lost during
evolutionary development (Prum, 1990).
Sometimes, the mechanisms forming behavioural differences between species have been
possible to elucidate by comparative studies. The greenish warbler, Phylloscopus
trochiloides, comes in several subspecies throughout Russia, Siberia and China. In its
western range, the subspecies interbreed readily. However, during evolutionary time,
subpopulations have been separated by the Tibetan Plateau, and where the populations meet
again, they are in effect two different species. They produce completely different songs, and
the females do not recognize the songs of the other population as species-specific (Irwin et
al., 2001). Geographical separation and female partner choice have driven the formation of a
new warbler species.
2.10 Modification and Ritualization

Reconstruction of the evolutionary history of behavioural traits leaves us with some
important lessons. First, evolution works only by selection of variants of traits that are
already present in a population. This means that evolution of course cannot invent any new
behaviour in the face of a ‘need’. Hence, the reason why an animal uses a particular
behaviour for a certain purpose is often to be found in the history of the species.
Inevitably, all behaviours we see today are therefore modified versions of behaviour
patterns that may have served very different purposes in ancestral forms. A behaviour that
serves a certain function gets slightly modified and may then serve a slightly new function,
and as time and generations pass, a new behaviour has developed. We can verify this
causative chain since the original behaviour may be conserved in closely related species, with
its original function still intact.
A special case of this is the evolution of animal signals by the process of ritualization.
One of the most quoted examples of a ritualized display is that of the peacock’s courtship.
The enormous and colourful tail is spread like a fan, vibrating, at the same time as the male is
bowing towards the female. A likely evolutionary history looks like this: ancestors attracted
females by pecking at feed items on the ground and emitting a special call – this is still the
common courtship of one close relative, domestic poultry. These movements have become
exaggerated, as in pheasants, and finally become unrelated to feed presentation, as in the
peacock. The tail has developed and increased in size at the same time. A behaviour that
originally served as food presentation has now become the courtship of peacocks (Schenkel,
1956). Ritualization is therefore the process by which a certain behaviour evolves into a
signal by becoming exaggerated and losing its original function.
Ritualization, combined with exaggeration of anatomical traits (such as the male peacock
tail; see Fig. 2.4), is also driven by preferences of the females, which Darwin referred to as
sexual selection. A trait (e.g. a colour or an ornament) may evolve because it indicates
important qualities in the males. Perhaps only males that are genetically healthy and resistant
to parasites can grow large tail feathers – the female that prefers to mate with males carrying
large tails may then get offspring that carry the favourable genes (Alcock, 2001).
Fig. 2.4. Domestic cockerels court females in the same way as junglefowl. One important
element is the ‘tidbitting’, where the male emits a special call at the same time as he pecks
towards the ground, sometimes towards food particles. This has by ritualization developed
into the spectacular peacock courtship.
2.11 The Function of Behaviour

Another important lesson we have learnt from evolutionary biology is that behaviour does
not evolve for ‘the good of the species’. Since all evolution can do is to work on variation
between individuals, there is simply no mechanism that can take the view of a group of
animals, or a species, into account (but there are noteworthy exceptions, where the group
may be an efficient unit of selection; see Chapter 6 for more details). When we look for the
function of any particular behaviour, we therefore always have to consider the benefits to the
individual. That benefit ultimately has to be measured in reproduction. The reproductive
success of an individual is also referred to as its fitness.
Behaviour not only has potential fitness benefits to the performer, but also costs. It may
consume valuable energy, which could otherwise be used for reproduction, or expose the
animal to predators. Once we realize that behaviour has both costs and benefits, it becomes
obvious that evolution will select the behaviour that maximizes the difference between
fitness benefits and costs. This is called the optimal behaviour (see Fig. 2.5).
Fig. 2.5. Costs and benefits of a territory both increase as a function of the size of the
territory, as shown in the principle diagram here. The functions are not identical – cost often
increases linearly or with an increasing function, whereas benefits follow a decreasing
function. This is because an animal cannot gain infinitely from larger territories. There are,
for example, time constraints. Evolution has favoured animals which choose to defend
territories with an optimal size (i.e. where the difference between benefits and costs is
largest).
For example, big territories are usually better, since they provide more food and other
resources. Consequently, females often prefer males that hold large territories. Should a male
therefore always attempt to defend as big a territory as possible? No; big territories require
much more energy and time to defend, which leaves less for reproductive efforts. Therefore,
males should aim for optimal territories that maximize the benefits received from attracting
more females in relation to the costs. Here is another example: how long should an animal
continue to search for food in the same place rather than moving to another, perhaps more
profitable, site? The answer is that the animal should weigh together the information
available regarding how depleted the present food patch is (number of food items ingested
per time unit), how profitable another average patch in the habitat is likely to be and how
long it will take to find another patch. Then it should choose to leave the present patch when
net energy intake over time is maximized. This general idea is central in what is called
optimal foraging theory (Krebs and Davies, 1991).
It may seem implausible that animals would be able to perform complicated calculations
like this. Nevertheless, many studies have shown that animals generally behave in an optimal
manner (Krebs and Davies, 1991). But, of course, they do not perform the mathematics
implied in the theory. Evolution has selected animals with the highest fitness, and they are the
optimal ones. If we could ask the animal why it leaves a patch after a certain time, it would
not be able to tell us the evolutionary reason. Could it speak, it would probably say it ‘felt
like leaving now’. In the technical jargon of behavioural ecology, the animal has a
behavioural strategy that causes it to react to stimuli as if it was consciously maximizing
fitness by optimizing its behaviour.
The same reasoning can be used to understand social behaviour. What is best for an
animal to do in a social contest (e.g. to attack or retreat) depends strongly on what the
opponent does or is likely to do. Even though it may superficially appear beneficial to always
attack and drive away intruders from a food source, some further thinking shows us that this
is not necessarily so. Since the tendency to attack depends on genes, a population where all
individuals always attack each other may be very unstable. An individual carrying a mutation
that causes it to avoid fights may gain a lot in fitness because it is not wounded, and the
mutated gene may therefore spread quickly in the population. The optimal social strategy is
therefore not always easy to prescribe without the use of formal mathematical modelling. The
strategy that, on average, confers the largest benefit to the individuals of the population is
usually referred to as the evolutionary stable strategy, or ESS (Krebs and Davies, 1991).
2.12 Domestication – An Evolutionary Case History

Domestication, the process whereby an animal is transformed from a life in the wild to a life
under some control of humans, is one of the most dramatic evolutionary processes that are
accessible to scientific investigation. It is often treated in the literature as a cultural
phenomenon, and the story often goes something like this (Clutton-Brock, 1999): when man
became sedentary and agriculture started, some 10,000 years ago, those animals that were
available for exploitation were tamed by man, so they could provide food, clothing, etc. This
led to the first domesticated species, dogs, goats and sheep, later followed by the main
agricultural species, such as pigs, cattle and chicken. The reason why the story is often told in
this way is that bones with typical domesticated traits (shortened legs, compressed skulls,
etc.) are found in excavations of early agricultural sites. According to this ‘standard model’
of domestication, each domestic species emerged from one single ancestral species and from
one single domestication event, or from a few events. However, a biological approach to the
subject reveals a somewhat different story.
Modern DNA technology has provided new tools for examining a variety of questions.
For example, differences between animals in sequences of DNA in the mitochondria
(mtDNA, which is inherited only from the mother) reveal their degree of relatedness and also
the time that has passed since the populations became reproductively isolated from each
other. Examinations of such data show that, for example, dogs and pigs dissociated from
wolves and wild boars much earlier than archaeological evidence has suggested. Perhaps as
early as more than 50,000 years ago, wolves that were later going to develop into dogs
became reproductively isolated from the line that led to today’s wild wolves. Domestication
of pigs seems to have happened independently in Europe and Asia, and again reproductive
isolation of the populations occurred much earlier than indicated by archaeology (Giuffra et
al., 2000). Independent events of horse domestication have been suggested from similar
types of data (see Chapter 12).
This is consistent with a theory of domestication claiming that at least some animals in
effect ‘chose to be domesticated’ (Budyansky, 1992), in the following sense. Some
populations seem to have adapted to take advantage of newly arising, highly productive yet
ephemeral habitats. The new species, Homo sapiens, emerging on the Eurasian continent
some 50,000–60,000 years ago, may have provided a potentially rich and fruitful niche for
such animals. They may have enjoyed large fitness benefits from associating with man,
simultaneously increasing human fitness by being relatively easy to hunt or by guarding
against danger (in the case of wolves). It appears that these populations became
reproductively isolated from the rest of the species long before domestication has been
thought to begin, and this period is sometimes referred to as ‘proto-domestication’. Perhaps
domestication took a leap when humans became sedentary and agriculture started, and
perhaps what archaeologists discover may be the results of the onset of active breeding.
However, domestication as an evolutionary process seems to have been going on for perhaps
five or ten times longer than this.
2.13 Which Animals Were Domesticated?

One feature of domestication that requires a biological explanation is the similarity in some
specific behavioural and ecological traits of the animals. It appears that not every animal
species is suitable for domestication. From a systematic perspective, there is a large bias
among domesticated animals towards ungulate mammals and gallinaceous birds. From a
behavioural aspect, there is a huge dominance of gregarious omnivores or herbivores without
strong mating bonds.
These behavioural traits (which are abundant among ungulates and gallinaceous birds)
may be crucial for a successful domestication (Price, 1997). Social life allows many animals
to live together in the human setting, and predisposes for hierarchical systems, where humans
can more easily adopt the role of a dominant group leader. Feeding habits that do not
compete with humans would be essential for successful cohabitation. Weak bonds between
mates would make selective breeding much easier.

The first wave of domestication comprised the major present-day farm animals and
animals such as the dog and the horse. When agriculture was established in large parts of the
world, many of these animals were already present and from a biological perspective could
be considered domesticated. For thousands of years, few new species were added (but of
course a variety of breeds within each species developed). During the last centuries, a second
wave occurred and a number of new species were domesticated. This time, there is no doubt
that this was a process controlled completely by man, dictated by specific needs or wishes.
The process gave us fur animals (mink, foxes, racoon dogs, chinchillas, etc.), laboratory
animals (mainly mice and rats) and several new meat producers (buffaloes, ostriches,
salmon). The animals domesticated in this second wave do not necessarily show the typical
traits outlined earlier. For example, mink are solitary, territorial animals, and foxes have
strong mate bonds.
2.14 Behavioural Effects of Domestication

Since domestication involves a genetic modification of a population of animals, we would
expect domestic animals to differ in a number of traits from their wild ancestors. Typical
morphological changes in colour (more white and spotted individuals), shape (size and
relative leg length) and function (less pronounced seasonality in reproduction) may lead us to
expect that there will be pronounced behavioural differences as well. However, most research
has found only subtle differences between domestic and wild animals (Price, 1997; see Fig.
2.6). Most typically, these differences can be attributed to modified stimulus thresholds,
causing some behaviour patterns to become more common and others to become rarer during
domestication. No new behaviours seem to have been added to the behavioural repertoire of
any domestic species, and few of the ancestral behaviour patterns have disappeared
completely. Hence, pigs kept for generations in restricted indoor housing systems still build
elaborate farrowing nests if released into a forest, and laying hens kept in battery cages will
attempt to perch high up during the night if given the slightest possibility.
Fig. 2.6. Domestic pigs (a) differ in appearance from wild boars (b), but their behaviour is
very similar.
The threshold differences that have occurred are sometimes caused by active selection by
man. For example, most dogs have been bred to bark in response to very low stimulation,
whereas basenjis, used for sneak-hunting in Africa, have been bred for the opposite. Other
behavioural changes may be a result of an evolutionary adaptation to a life with humans. For
example, energetically costly behavioural strategies appear in some cases to have been
reduced in pigs and poultry, allowing energy resources to be used for growth and
reproduction, which have been favoured during domestication (Gustafsson et al., 1999;
Schütz and Jensen, 2001). Furthermore, domestic animals are generally less fearful towards
humans and more socially tolerant towards conspecifics. Dogs present some striking
adaptations to living with us. For example, unlike wolves, they are highly sensitive to human
ostensive cues, they excel in word understanding and they solicit human attention when
facing difficult problems (Jensen et al., 2016). Recently, it has been found that the human-
directed social behaviour appears to be linked to genes that are also associated with human
social disorders such as autism and attention-deficit hyperactivity disorder.
The changes in appearance and physiology, and to some extent behaviour, that have
occurred during domestication are the results of selection of genetic variants already present
in the wild ancestral populations or mutations occurring spontaneously. Such new mutations
seem to underlie many of the colour changes that are common in domesticated breeds of
different species. However, epigenetic modifications may also contribute. In chickens, DNA
methylation as well as gene expression in the brain differ substantially between a modern egg
layer and its ancestor, the red junglefowl (Nätt et al., 2012). Epigenetic variation may in fact
be highly important in driving fast evolutionary change of a population, such as that
occurring during domestication.
Although there are behavioural differences between wild and domestic animals, it is clear
that they are not as large as we sometimes tend to believe. It is also frequently suggested that
domestic animals are less responsive to their environment than wild animals, even that they
are ‘more stupid’. However, detailed studies of the behaviour of domestic animals in natural
conditions reveal that their behaviour is very similar to that of the ancestors. The fact that
they often develop abnormal behaviour, and even pathologies, when they are prevented from
performing a normal behaviour, indicates a strong responsiveness towards the environment in
which they are kept. Regardless of whether we think that welfare concerns are central in
animal husbandry, or whether we are guided by concerns over the productivity of the
animals, we need to remember that the behaviour of domestic animals is controlled by
genetic mechanisms shaped over thousands and thousands of generations of evolution in the
wild and altered only slightly during domestication. The evolutionary history and adaptations
of the ancestors, and the natural behaviour of the present-day animals, are therefore important
pieces of information if we want to understand the animals we keep for our utility.
References
Alcock, J. (2001) Animal Behaviour: An Evolutionary Approach. Sinauer Associates, Inc., Sunderland, Massachusetts.
Budyansky, S. (1992) The Covenant of the Wild: Why Animals Chose Domestication. William Morrow and Co., Inc., New
York.
Clutton-Brock, J. (1999) A Natural History of Domesticated Mammals. Cambridge University Press, Cambridge.
Cooper, R.M. and Zubek, J.P. (1958) Effects of enriched and restricted early environments on the learning ability of bright
and dull rats. Canadian Journal of Psychology 12, 159–164.
Crawley, J.N. (1999) Behavioral phenotyping of transgenic and knockout mice. In: Jones, B.C. and Mormède, P. (eds)
Neurobehavioral Genetics: Methods and Applications. CRC Press, Washington, DC, pp. 105–119.
Dilger, W.C. (1962) The behavior of lovebirds. Scientific American 206, 88–98.
Giuffra, E., Kijas, J.M.H., Amarger, V., Carlborg, Ö., Jeon, J.-T. and Andersson, L. (2000) The origin of the domestic pig:
independent domestication and subsequent introgression. Genetics 154, 1785–1791.
Gustafsson, M., Jensen, P., de Jonge, F.H. and Schuurman, T. (1999) Domestication effects on foraging strategies in pigs
(Sus scrofa). Applied Animal Behaviour Science 62, 305–317.
Hirsch, J. (1967) Behavior-Genetic Analysis. McGraw-Hill, New York.
Huntingford, F. (1984) The Study of Animal Behaviour. Chapman & Hall, London.
Irwin, D.E., Bensch, S. and Price, T.D. (2001) Speciation in a ring. Nature 409, 333–337.
Jensen, P. (2015) Adding ‘epi-’ to behaviour genetics: implications for animal domestication. Journal of Experimental
Biology 218, 32–40.
Jensen, P., Persson, M.E., Wright, D., Johnsson, M., Sundman, A.-S. and Roth, L.S.V. (2016) The genetics of how dogs
became our social allies. Current Directions in Psychological Science 15, 750–756.
Jones, R.B. and Hocking, P.M. (1999) Genetic selection for poultry behaviour: big bad wolf or friend in need? Animal
Welfare 8, 343–359.
Kappeler, L. and Meaney, M.J. (2010) Epigenetics and parental effects. Bioessays 32, 818–827.
Kjaer, J.P. and Sorensen, P. (1997) Feather pecking behaviour in white leghorns, a genetic study. British Poultry Science 38,
333–341.
Klein, T., Zeltner, E. and Huber-Eicher, B. (2000) Are genetic differences in foraging behaviour of laying hen chicks
paralleled by hybrid-specific differences in feather pecking? Applied Animal Behaviour Science 70, 143–155.
Konermann, S., Brigham, M.D., Trevino, A.E., Joung, J., Abudayyeh, O.O., Barcena, C., Hsu, P.D., Habib, N., Gootenberg,
J.S., Nishimasu, H., Nureki, O. and Zhang, F. (2015) Genome-scale transcriptional activation by an engineered CRISPR-
Cas9 complex. Nature 517, 583–588.
Krebs, J.R. and Davies, D.B. (1991) Behavioural Ecology: An Evolutionary Approach. Blackwell Scientific Publications,
Oxford.
Lynch, C.B. (1980) Response to divergent selection for nesting behaviour in Mus musculus. Genetics 96, 757–765.
Malmkvist, J. and Hansens, S.W. (2001) The welfare of farmed mink (Mustela vison) in relation to behavioural selection: a
review. Animal Welfare 10, 41–52.
Manning, A. (1961) The effects of artificial selection for mating speed in Drosophila melanogaster. Animal Behaviour 16,
108–113.
Nätt, D., Rubin, C.-J., Wright, D., Johnsson, M., Beltéky, J., Andersson, L. and Jensen, P. (2012) Heritable genome-wide
variation of gene expression and promoter methylation between wild and domesticated chickens. BMC Genomics 13, 59.
Price, E.O. (1997) Behavioral genetics and the process of animal domestication. In: Grandin, T. (ed.) Genetics and the
Behavior of Domestic Animals. Academic Press, San Diego, California, pp. 31–65.
Prum, R.O. (1990) Phylogenetic analysis of the evolution of display behavior in the neotropical manakins (Aves: Pipridae).
Ethology 84, 202–231.
Quinn, W.G., Harris, W.A. and Benzer, S. (1974) Conditioned behavior in Drosophila melanogaster. Proceedings of the
National Academy of Sciences USA 71, 708–712.
Schenkel, R. (1956) Zur Deutung der Phasianidenbalz. Ornithologische Beobachtungen 53, 182.
Schütz, K. and Jensen, P. (2001) Effects of resource allocation on behavioural strategies: a comparison of red junglefowl
(Gallus gallus) and two domesticated breeds of poultry. Ethology 107, 753–765.
Scott, J.P. and Fuller, J.L. (1965) Genetics and the Social Behavior of the Dog. University of Chicago Press, Chicago,
Illinois.
Tryon, R.C. (1940) Studies in individual differences in maze ability VII. The specific components of maze ability and a
general theory of psychological components. Journal of Comparative Physiology and Psychology 30, 283–335.
3 Behaviour and Physiology
A. Valros and L. Hänninen
3.1 Introduction
The behaviour of an animal reflects a complex series of underlying physiological events,
both neurological and endocrinological. The motivation to perform certain behaviours is
affected by both acute changes in the homeostatic balance of the individual, such as the
reaction to a sudden change in environmental temperature, and long-term internal or external
processes, such as the reproductive state of the individual or the nutritional level of the
environment. The physiology of an animal has an amazing capacity to adapt to changing
situations. In addition, as the mechanisms behind the shown behaviour are extremely
complex and often involve several interrelated mechanisms, it is often easier to measure and
interpret the behaviour of an animal than to understand the underlying physiological
mechanisms.
Behaviour can be considered a black box – showing us what is going on without always
revealing why. For example, studies by Janczak et al. (2006, 2007) have shown that similar
changes in the behaviour of the chicken, reflecting adaptations to prenatal stress, can be
achieved both by injecting corticosterone into fertilized eggs and by stressing the mother
hens during the laying period. However, the two ways of ‘stressing’ the eggs did not seem to
be governed by the same physiological mechanisms, as the eggs from the stressed hens failed
to show an increased level of corticosterone. Nevertheless, despite the complex nature of the
physiological mechanisms governing an individual’s behaviour, we do need a basic
understanding of these if we are to understand the reasons for the animal’s behaviour.
3.2 Central Regulation of Behaviour

Hypothalamus and hypophysis – the ‘master gland’
One of the most central brain regions governing the motivational state, and thus the
behaviour, of animals is the hypothalamus. The hypothalamus is situated at the base of the
brain and is extremely important in regulating several homeostatic mechanisms and related
emotional reactions, such as body temperature, ingestion of food and water, sexual behaviour
and rage (Eckert et al., 1988). The hypothalamus regulates the autonomic nervous system
and links the nervous system to the endocrine system, thus being at the top of the endocrine
hierarchy. At the basis of the hypothalamus can be found a small, but very important
appendage, the hypophysis or the pituitary gland, which has been named the ‘master gland’
for its important role in governing peripheral endocrine functions (see Fig. 3.1).
Fig. 3.1. Cross-section of a pig brain. A, hippocampus; B, hypothalamus; C, hypophysis; D,
cerebellum; E, nucleus accumbens; F, suprachiasmatic nucleus; G, optic chiasm; H, pineal
gland.
The hypothalamus largely regulates the hypophysis through neurosecretory cells (i.e.
nerve cells that produce hormones and secrete them into the bloodstream). These cells react
to input from different parts of the body related to such mechanisms as temperature
regulation, osmoregulation and sexual cycles. Some of these cells reach throughout the
posterior lobe of the hypophysis or the neurohypophysis, which is really more like an
extension of the hypothalamus, and release so-called neurohypophyseal hormones directly
into the bloodstream. These hormones thus affect peripheral target tissues directly and
include antidiuretic hormone (ADH), also called vasopressin, that increases water
reabsorption in the kidney, and oxytocin, which effects smooth muscular contractions and is
positively related to both social and maternal behaviour of mammals, as well as to milk
production and ejection.
The other type of neurosecretory cells in the hypothalamus release hypothalamic
releasing and inhibiting hormones into the bloodstream. From the hypothalamus the blood is
directed into a capillary network in the anterior lobe of the hypophysis, or the
adenohypophysis. Here the hypothalamic releasing and inhibiting hormones attach to cell
receptors and thus regulate the release of hormones from the non-neural endocrine cells in
the adenohypophysis. Three of these adenohypophyseal hormones are regulated by both
stimulatory and inhibitory hypothalamic hormones, and act directly on non-endocrine target
tissues (see Table 3.1). These include growth hormone (GH), prolactin (PL) and melanocyte-
stimulating hormone (MSH). GH is important for growth, milk production and immunology,
and PL for milk production and maternal behaviour. MSH is actually secreted from the pars
intermedia, a section between the anterior and posterior lobes of the hypothalamus, and
regulates pigment formation.
Table 3.1. Hormones secreted from the hypothalamus and the adenohypophysis and their
target tissues or target glands.
Hypothalamic hormones (and Adenohypophyseal hormones (and Target tissues or target glands (and their
abbreviations) abbreviations) hormones)
Growth hormone RH (GHRH) Growth hormone (GH) All tissues
Growth hormone RIH (GHRIH)
Prolactin RH (PRH) Prolactin (PL) Mammary tissue
Prolactin RIH (PRIH)
Melanocyte-stimulating hormone RH Melanocyte-stimulating hormone Pigment cells
(MSHRH) (MSH)
Melanocyte-stimulating hormone RIH
(MSHRIH)
Corticotropin RH (CRH) Adrenocorticotropin (ACTH) Adrenal cortex (corticosteroids)
Thyrotropin RH (TRH) Thyrotropin (TSH) Thyroid gland (thyroid hormones)
Luteinizing hormone RH (LHRH) Luteinizing hormone (LH) Gonads (sex steroids)
Follicle-stimulating hormone RH Follicle-stimulating hormone (FSH) Gonads (sex steroids)
(FSHRH)
RH, releasing hormone; RIH, release-inhibiting hormone.
The release of the remaining four adenohypohyseal hormones, adrenocorticotropin

(ACTH), thyrotropin (TSH), luteinizing hormone (LH) and follicle-stimulating hormone
(FSH), is stimulated by a hypothalamic releasing hormone, while their inhibition is
dependent on a negative feedback system. This means that their further release is inhibited
via the hypothalamus, by the hormones released from their target glands into the blood
circulation. All of these hormones are important for the behaviour of animals: FSH and LH
regulate reproductive behaviour by affecting the secretion of sex steroids; ACTH is a central
part of the stress reaction, as it governs the secretion of corticosteroids; and TSH stimulates
the release of thyroid hormones which in turn increase oxygen consumption and the
metabolic rate, thereby also affecting thermoregulation. In addition, thyroid hormones are
very important for the normal development and growth of individuals.
Emotion and consciousness

The limbic system involves several, closely interacting parts of the brain. The main parts
include the hypothalamus, the hippocampus and the amygdala, all situated just beneath the
cerebrum, or the forebrain (see Fig. 3.1). In addition, several other structures are often
considered part of the limbic system, such as the orbitofrontal cortex, which is important for
decision making, and the nucleus accumbens, which is involved in producing feelings of
pleasure and in addiction. The limbic system reacts to external and internal pleasant or
unpleasant stimuli and coordinates changes needed to maintain homeostasis.
It has been shown that the amygdala is one of the key brain structures involved in
emotional processes both in humans and other animals (Cardinal et al., 2002). However, in
order for a stimulus to cause a conscious feeling or emotion it is believed that also other parts
of the brain, such as the forebrain, need to be involved: for a stimulus to result in an emotion,
it needs to be processed and related to expectations and associations, thus involving higher-
order brain functions (Ressler, 2004). In many cases, non-human animals appear to show
similar brain reactions to stimuli as do humans and exhibit similar brain structures to those
proven to be related to experiences in humans. But as Dawkins (2006) states, even though
there are plenty of data on how animals of different species react to stimuli and which brain
areas and physiological processes are involved in these reactions, there is still no commonly
accepted answer to the question of which physiological and neurological processes result in
consciousness and in sentience, and which species have these abilities.
Decision making, the reward system and neurotransmission
One of the most important mechanisms steering the behaviour of an animal is the reward
system. The simplified mechanism of this system is that an animal reacts differently to
external or internal stimuli causing pleasant (rewards) or unpleasant (punishment) sensations
(Ressler, 2004). Thus, a reward is something that an animal will work towards achieving,
such as food or comfort, and a punishment something an animal will work to avoid, such as a
predator or discomfort. The limbic system is involved in regulating the process of acting on
potential rewards and punishments.
The amygdala is an important component of the reward system, which is largely based on
dopaminergic neurons (i.e. neurons with dopamine as their transmitter substance). Dopamine
is a monoamine, a similar compound to other established neurotransmitters, such as serotonin
and noradrenaline. Dopaminergic neurons are involved in several behaviour-related
regulative systems of the brain, regulating, for example, functions related to cognition,
motivation, sleep and mood. The dopaminergic system has also been connected to the
performance of stereotypies (e.g. in hens).
The way an individual processes stimuli to make decisions can be simplified as seen in
Fig. 3.2. Sensory input is processed by different areas of the brain, such as the amygdala and
the prefrontal cortex, thus allowing the individual to evaluate the input in relation to the
expected reward value and prior experiences. At least in higher vertebrates, such as primates,
different reaction possibilities are then evaluated and compared, and a voluntary choice is
made on how to act in the current situation (Opris and Bruce, 2005).
Fig. 3.2. Schematic description of the decision-making system. (From Opris and Bruce,
2005.)
Serotonin, or 5-hydroxytryptamine (5-HT), is another important monoamine involved in
the regulation of behaviour. Serotonin has been studied extensively in connection with
human mental disorders such as depression and compulsive-obsessive disorders. In other
animals, serotonin has been connected to the level of fear and anxiety, stress reactivity and
stereotypic behaviour. In pigs, serotonin has recently been connected to tail-biting, which is a
common and very problematic behaviour disturbance in pigs (Valros et al., 2015). By feeding
tryptophan, an amino acid used by the body to synthesize serotonin, the behavioural reactions
of animals can be altered. For example, tryptophan feeding or administration has been shown
to reduce fear and anxiety in fur animals and reduce stress reactivity in pigs probably due to
an increase in brain serotonin levels.
Also, neuropeptides are involved in the reward system, producing pleasure- and
euphoria-like states in animals. Endogenous opioids, such as endorphins, found mainly in the
hypophysis, and encephalins, found throughout the nervous system, have been studied
extensively in this context. These neuropeptides are produced by the central nervous system
and bind to opioid receptors. They cause a reduction in the sense of pain (i.e. have an
analgesic effect) and they rise in response to pleasurable events, such as eating. This system
is also involved in drug addictions in humans: narcotic opiates such as opium and morphine
bind to the same opioid receptors, thus causing a pleasurable feeling. Repeated use of
narcotic opiates causes a change in the neuronal metabolism, resulting in the individual
feeling extreme discomfort until more drugs are administered. This is called addiction. In
animals, it has been shown that endogenous opioids are related to the level of stress response
and to the performance of stereotypies.
Control of circadian rhythms
The body’s biological clock is a hierarchal system, governed by the suprachiasmatic nucleus
(SCN), which is the main regulator for biological rhythms. The SCN is situated in the
hypothalamus, directly above the optic chiasm (see Fig. 3.1). Even though most of the
peripheral tissues and organs also generate their own circadian rhythms, their molecular
clock is entrained by the SCN. In mammals, the SCN is regulated mainly by light via the
retinohypothalamic tract (RHT), which is a pathway of photoreceptive cells from the retina to
the SCN. In mammals, the hormone melatonin is secreted mainly during the dark period from
the pineal gland, a small pine-like structure in the central part of the brain (see Fig. 3.1). The
secretion is driven by the SCN, and can be thought of as a neuroendocrine transducer of the
light–dark cycle. Melatonin is an important modulator of circadian rhythms. It promotes
sleep and can induce phase shifts of biological rhythms. Other factors that synchronize the
circadian system are, for example, nutrition, hormone feedback mechanisms, activity and
social cues (Buijs et al., 2003).
Hormonal fluctuation and rest–activity or sleep rhythms are examples of biological
rhythms. The rhythms are classified as circadian, with a cycle of approximately 24 h, and
ultradian, with a cycle of less than 24 h. Many of the hormones mentioned above have
circadian rhythms, such as ACTH, cortisol, GH and prolactin. Also, body temperature
follows a circadian cycle. This of course has a marked effect on the individual’s behaviour,
alertness and cognitive output.
Examples of circadian rhythms involving both physiology and behaviour include the
connection between sleep and hormones such as GH and glucocorticoids. Sleep is a
physiological regulator for GH in laboratory animals and humans, and sleep is associated
with GH secretion also in sheep. Typically, sleep onset stimulates GH secretion. The
secretion of GH increases during sleep independent of the circadian sleeping cycle, and sleep
deprivation diminishes the GH release. In humans and many of the animal species studied so
far, blood glucocorticoids are highest when waking up in the subjective morning and lowest
in the subjective evening and early part of the sleep period. Thus, sleep normally sets in when
the corticotropic activity is quiescent, at least in humans and laboratory rodents. Cortisol
begins to rise a few hours before the usual waking time. The cortisol rhythm is very stable
but sleep deprivation changes the pulse amplitude.
3.3 Sleep
All animals benefit from restricting their different activities to times of the day when optimal
environmental conditions exist – for example, when it is not too dangerous to eat or sleep.
Thus, during evolution, each species has developed its unique sleeping method, amount and
rhythm. Primates often exhibit a mono- or biphasic sleeping rhythms while several other
species have a polyphasic rhythm, sleeping all through the day. Typical examples of
polyphasic, but also short sleepers are our domestic ruminants and horses. Birds, on the other
hand, have developed a unique sleeping style – many of them, such as domestic hens, are
capable of unihemispheric sleep (i.e. one cerebral hemisphere sleeps while the other remains
awake). When one hemisphere is asleep, the opposite, contralateral, eye is closed while the
other eye is open. This type of a sleep makes it possible for birds to keep an eye on predators,
even when asleep. Birds are capable of controlling which brain side is asleep according to
predator pressure.
Sleep can be defined electrophysiologically as consisting of two main phases: (i) rapid
eye movement sleep (REM), in earlier literature also called paradoxical sleep or active sleep;
and (ii) non-rapid eye movement sleep (NREM), also called quiet sleep, orthodoxical sleep
or slow-wave sleep. The stimulus threshold to wake up from REM sleep is higher in many
species than that of NREM sleep, and thus a long REM sleep period can be a threat to
survival for prey species other than those rodent species that sleep in nests.
The sleep cycle consists of one or several REM and NREM phases. The cycle length is
species-specific. The cycles are short in farm animals, as is usual in prey animal species. For
example, calves sleep in several 5 min periods throughout the 24 h day (Hänninen, 2007; see
Fig. 3.3). Also, total sleep duration varies a lot between species (see Fig. 3.4). These
differences between species depend on several factors, such as time spent eating, available
food sources, digestion and rumination, and ecological niche. Grazing animals, for example,
are assumed to sleep less, as they need more time to consume large amounts of low-energy
foods (Siegel, 2005).
Fig. 3.3. Young calves sleep for approximately 20% of the day in short bouts throughout the
24 h period. The image shows a calf with its head to the side and neck relaxed, indicating
REM sleep. (Photo courtesy of Pate Pesonius.)
Fig. 3.4. Total daily sleeping time (REM and NREM sleep) in hours in different animal
species. (Data from Tobler, 2005, except cattle from Ruckebusch, 1972.)
In mature mammals and birds, the sleep phase usually starts with NREM and deepens
during the REM phase. The young of many terrestrial mammalian species sleep more and
have more REM sleep than do older animals. Sleep is essential for the development of the
brain, and REM sleep is connected to the early developmental phase. For that reason, altricial
young mammals, such as rat pups, spend 70% of their day in REM sleep, which is
proportionally more than, for example, the 20% of bovine calves (Hänninen et al., 2008).
Young animals also have a greater need than older animals for energy retention, which can be
achieved through more sleep.
3.4 Regulation of Food Intake and Eating

In order for an individual to get the appropriate amount and quality of nutrients, the
behaviour needs to be optimized and regulated accordingly. Important nutrients include, in
addition to energy, such compounds as essential amino acids, vitamins and minerals.
Depending on the digestive system and needs of an animal, several behavioural patterns have
evolved to make sure these nutritional needs are met.
Finding and ingesting food is a prerequisite for staying alive and thus evolution has
resulted in very complex systems to make sure the individual is very motivated to eat. Energy
intake is dependent on several internal and external factors, such as the reproductive state of
the animal, the amount of exercise and the environmental temperature. Food intake is
governed mainly by the hypothalamus, but also other mechanisms are of importance in
regulating the food intake. These mechanisms include the level of blood sugar (glucose), the
level of filling of the stomach, endocrinological control and cognitive aspects (Schmidt-
Nielsen, 1997). As separate mechanisms, these cannot regulate the food intake efficiently, but
the coordination of several systems is needed for this complex process. For example,
ingesting a large bulk of low-energy feed will quickly fill up the stomach, but will not satisfy
the energy need of the animal, which will then compensate by eating an even larger amount
of the food. In contrast, a diet with a very high energy content can result in the animal not
feeling satisfied due to the lack of filling of the stomach, even though the energy intake might
have been appropriate. Under risk of starvation, hypothalamic functions as well as activity in
several other parts of the brain will cause food to be an extremely pleasurable reward and
thus make the motivation to feed top priority.
Hormonal control of feeding behaviour
A number of hormones are involved in the regulation of hunger, satiation and feed intake,
including cholecystokinin (CCK), leptin, somatostatin, insulin and ghrelin (Geary, 2004).
CCK inhibits eating and is secreted during meals from the small intestine. The rise in CCK
acts to limit eating at a certain meal, while it does not appear to influence total food
consumption. If CCK is administered, meal size is reduced while inter-meal intervals are
unaffected or even reduced to compensate for the smaller intake per meal.
Also, injections of leptin can reduce meal size, but do not appear to affect the inter-meal
interval. Leptin is therefore probably more important for determining body weight than CCK.
Leptin is secreted mainly from fat tissue and the level of leptin is higher in obese individuals.
Interestingly, leptin follows a diurnal rhythm and its secretion is not affected by food intake
or meal timing. Receptors are found mainly in the hypothalamus, suggesting a role in the
homeostatic system, but also in many other parts of the brain. A total lack of leptin will cause
immediate overeating and severe obesity. In situations where preferred foods are freely
available, individuals can also develop leptin resistance, resulting in obesity. The
evolutionary reason for this could be that in situations of high food availability (e.g. summer)
it is of benefit for the animal to consume as much as possible, in order to survive the
following period of lower food availability (e.g. winter). Leptin resistance is also seen in
animals preparing for hibernation, thus allowing them to eat more that their current
nutritional needs call for. Leptin resistance has been found both in obese humans and pet
animals, such as dogs and cats, suggesting either that the homeostatic function of leptin is
disrupted for some reason in these individuals or that its function simply is not to protect
from obesity but to make sure animals of low energy status are highly motivated to eat.
Interestingly, low plasma leptin concentration has been associated with the occurrence of
crib-biting in horses, being especially low before expected concentrate delivery (Hemmann et
al., 2013), thus indicating a central role of feeding motivation in at least this type of oral
stereotypic behaviour.
Ghrelin is produced mainly by endocrine cells in the stomach and it is the hormone with
the strongest known positive effect on feeding behaviour. Ghrelin stimulates eating by
increasing the feeling of hunger and the ingested meal size. It appears to signal hunger, as it
increases before meals and decreases after eating. In addition, the level of fat tissue also
affects ghrelin – obese individuals have lower levels of ghrelin. Ghrelin receptors in the
hypothalamus mediate its effect on eating behaviour, but receptors also exist peripherally,
such as in the abdomen.
Insulin is an important anabolic hormone, signalling that the body should store nutrients.
Insulin is secreted from the pancreas in response to an increase in blood glucose level. Thus,
insulin, following the level of glucose, increases rapidly after the ingestion of food.
In addition to the mainly short-term regulation of feeding behaviour discussed above,
feeding behaviour is also hormonally regulated on an annual level. For example, melatonin
has a marked effect on metabolism and hunger in certain species where there is a large annual
variation in feed availability, such as reindeer.
Feeding and decision making
Not only changes in the homeostatic balance, such as those induced by food or water
restriction, affect feeding behaviour. Many of the decisions related to feeding are related to
more complex brain functions, involving such things as memory and the reward system.
Throughout evolution, finding and consuming food has been of uttermost importance and
thus a large portion of the nervous system is involved in this process (Berthoud, 2007). For
example, parts of the amygdala appear to be important for learning cues related to food,
while the orbitofrontal cortex of the brain is important for memorizing these. This enables an
animal to learn which foods are palatable, beneficial or harmful, as well as to memorize
where preferred food sources can be found. In humans, simply thinking about food and in
animals giving a cue related to food causes physiological reactions, such as saliva and insulin
secretion (Berthoud, 2007). The dopaminergic system is involved in the process of wanting
foods an animal has learnt to like. When eating a preferred food, dopaminergic neuron
activity causes a pleasurable feeling, a reward, which later will cause the animal to remember
this food as preferred. The cognitive processes described here can cause an animal to want a
food even though its metabolic needs do not indicate the need for energy. Therefore, in an
environment where food is not a limited resource, animals, as well as humans, easily become
obese due to strong evolutionary benefits from a strong motivation to feed.
In addition to meeting the energetic needs of an animal, the feeding behaviour needs to
make sure the animal’s needs for specific nutrients are met, a process called selective
satiation (Berthoud, 2007). It has been shown that certain neurons that become active when
ingesting food can distinguish different energy sources from each other. For example, the
predictive value of a preferred food is influenced by what foods have been ingested recently:
after a high-glucose meal, there is less neural activity when presented with further sugar-rich
foods, while other foods might still cause neural activity of similar magnitude.
3.5 Reproductive and Maternal Behaviour

Hormonal regulation of reproductive behaviour
Reproductive behaviour is largely controlled by endocrinological sequences, although these
processes are also influenced by external factors such as season, food availability and social
factors. In most animals, timing of reproduction is set to a certain time of the year, when the
survival possibilities for the offspring are maximized. Typically, offspring are born in the
spring or early summer, when food resources are usually available. The seasonality of the
reproductive cycle is regulated mainly by light–dark cycles, something that is utilized for
example in the egg industry: under natural lighting conditions laying hens decrease egg-
laying during the winter but, by increasing the daylight with artificial lighting, egg
production can be kept more or less stable throughout the year. Also, by using different light
rhythms during rearing, the lifetime production capacity of modern laying hens can be further
refined and optimized. In other production animals, the seasonality of reproduction has
negative impacts on optimizing the production capacity: in sheep, all-year breeding is very
challenging; and in pigs, remnants of seasonality can be seen as a lower piglet production
during the autumn period, even though the domestic pig is no longer a purely seasonal
breeder. Also in the boar, a seasonal variation can be seen as higher levels of testosterone,
higher amount and quality of sperm, and a related higher level of sexual activity in the
autumn, which is the main breeding season of the pig. It appears that it is the length of the
light period, not the intensity, which is important for governing this system.
The effect of light is most probably transmitted via melatonin to the hypothalamus,
which, as stated earlier in this chapter, is the main regulator of the reproductive functions of
animals. Two of the main hormones involved in the regulation of reproductive cycles and
behaviour are the gonadotropins, LH and FSH. The main target organs of the gonadotropins
include the ovaries in females and the testes in males. In females, the release of
gonadotropins mainly affects the release of oestrogen, influencing the development of
secondary female sex characteristics and sexual behaviour. Infusions of oestrogen into the
brain of castrated cats can, for example, induce behaviour indicative of willingness to mate.
In the male, the most potent androgen is testosterone, which affects secondary sex
characteristics, such as plumage and facial hair growth, as well as sexual behaviour.
Testosterone affects the level of aggressiveness in the male, as well as the level of sexual
activity.
Hormonal regulation of maternal behaviour
The maternal behaviour of animals is largely under endocrinological control. Behaviour
during the prepartum period (such as nest building in the sow), during parturition itself and
during the postpartum period (including both nursing behaviour and behaviour related to the
mother–young bond) has been shown to have a clear hormonal background. This does not, of
course, mean that environmental factors would be of no importance, but does indicate that
behaviours related to maternal abilities are highly motivated in the dam. In addition,
continuous feedback from the offspring is needed for the continuance of maternal behaviour.
For example, as a result of reduced udder manipulation, milk production and related hormone
secretion will decrease prematurely.
During pregnancy, the level of the steroid hormone progesterone is high, acting to
maintain the pregnancy. Progesterone originates from the corpora lutea in the ovary of a
pregnant female. Prior to parturition, progesterone levels drop, thus allowing milk production
to begin and metabolic functions to favour milk production instead of storage of nutrients in
adipose tissue. If the level of progesterone does not decrease appropriately, this can have
adverse effects on milk production and thus growth of the nursed offspring.
Before parturition, levels of prolactin start to increase. Prolactin is one of the main
hormones involved in initiating and maintaining milk production, as well as in influencing
the maternal and nursing behaviour of the dam. Prolactin also affects lactational metabolism
by increasing the amount of insulin receptors in the mammary gland. High levels of prolactin
are maintained by stimulus from the offspring and, if this is reduced, prolactin levels will also
decrease.
In the pig, one of the most marked effects of prepartum prolactin is on the sow’s nest-
building behaviour (see Fig. 3.5). A rise in prolactin a few days before farrowing causes the
initiation of nest-building behaviour. This pre-farrowing increase in prolactin is related to an
increase in prostaglandin (PGF2α), which also increases significantly prior to farrowing.
Different aspects of the nest-building behaviour of the sow have also been shown to be
correlated with concentrations of other hormones, such as progesterone, somatostatin and
oxytocin. The sow alters her nest-building behaviour depending on environmental cues, such
as temperature, while the possibility to perform nest building appears to affect the level of
important maternal hormones, oxytocin and prolactin, which further relate to postpartum
maternal behaviour (Yun, 2015). The mechanism behind the termination of nest building is
less well known, and probably involves complex hormonal and environmental interactions. A
high level of oxytocin early before parturition has been shown to correlate with an early
cessation of nest building, but the mechanism for this is still unknown.
Fig. 3.5. The domestic sow, like its ancestor the wild boar, builds a nest prior to farrowing in
order to provide shelter to the piglets. The nest-building behaviour is largely regulated by
prepartum endocrinological changes. (Photo courtesy of Heli Castrèn.)
Oxytocin is secreted from the hypophysis and is important for several different aspects of
maternal behaviour in mammals. For example, grooming of pups in the rat is induced by
oxytocin, and oxytocin has been correlated with carefulness of the sow towards her piglets.
Oxytocin is needed for the initiation of lactation and for milk ejection during separate
nursings. Stimulation by the offspring, such as udder manipulation, stimulates oxytocin
release, which then triggers milk ejection. In addition to this all-or-nothing effect on milk
ejection, oxytocin influences metabolism during lactation, making sure resources are
available for milk production. In the pig, a high basal oxytocin level has been connected with
a high catabolic level (negative energy balance) of the sow and a correlated high piglet
weight gain (Valros, 2003). In addition, oxytocin has an important role in the parturition
process, by stimulating uterine contractions. In general, the expulsion of each fetus is related
to a peak in oxytocin, but it also appears that in litter-bearing species, the birth of the first
fetus stimulates the parturition process further. In pigs and rats, a high oxytocin level has
been linked to a shorter duration of parturition (Algers and Uvnäs-Moberg, 2007).
Hormones related to feed intake and metabolism have also been shown to be important in
the regulation of lactation and nursing behaviour. As insulin promotes the utilization of
nutrients for peripheral body tissues, its secretion is lower during stages where catabolism
can be seen as an advantage, such as during lactation, where nutrients should be directed
mainly towards milk production. For example in the sow, insulin pre-feeding levels decrease
with proceeding lactation and a correlation between sow nursing behaviour and insulin level
has been reported. Sows with high levels of pre-feeding insulin show more behaviour related
to avoiding udder massage by their piglets. It is possible that a high level of nursing stimulus
increases the number of insulin receptors in the mammary gland, thus decreasing the level of
circulating insulin, but it also cannot be excluded that the high insulin level, as such,
influences the nursing motivation of the sow, possibly via interaction with prolactin.
Whatever the cause and the effects, it appears important for the sow to be in a highly
catabolic state during lactation, as this appears to be a prerequisite for efficient piglet growth
and high piglet survival rate (Valros, 2003). Hormones such as vasoactive intestinal
polypeptide (VIP), gastrin, somatostatin and glucagon are also influenced by suckling. All of
this causes the optimization of nutrient utilization for milk production by affecting
gastrointestinal functions and allocation of nutrients.
3.6 Stress-related Behaviour

When an individual is subjected to a stressor, the blood concentrations of glucocorticoids
increase as a result of activation of a specific endocrinological pathway, the hypothalamus–
pituitary–adrenocortical (HPA) axis. When exposed to a stressor, CRH from the
hypothalamus stimulates pituitary ACTH secretion, which leads to an increased
glucocorticoid release from the adrenal cortex. In addition, the sympathetic nervous system
increases an individual’s performance in demanding situations by increasing the secretion of
catecholamines, such as adrenaline and noradrenaline, from the adrenal medulla.
Noradrenaline is also secreted directly from the sympathetic nervous system, being the
terminal transmitter substance of this system. These reactions will cause the so-called fight-
or-flight response, preparing the individual to face physical or physiological challenges.
During the acute flight-or-fight response, blood pressure increases and blood sugar
concentration rises. Heart frequency increases and blood oxygen is effectively transported,
especially to the skeletal muscles. Individuals are bright and alert, preparing themselves to
fight, take flight or freeze. However, the secretion of glucocorticoids in itself is not sign of a
negative experience as it is also stimulated by physical exercise, feeding or mating. It is also
worth noting that if the stressful situation is prolonged, it disturbs the circadian variation of
glucocorticoids in the blood.
For more information on stress-related mechanisms, see Chapter 9.
References
Algers, B. and Uvnäs-Moberg, K. (2007) Maternal behaviour in the pig. Hormones and Behavior 52, 78–85.
Berthoud, H.-R. (2007) Interactions between the ‘cognitive’ and ‘metabolic’ brain in the control of food intake. Physiology
& Behavior 91, 486–498.
Buijs, R.M., van Eden, C.G., Goncharuk, V.D. and Kalsbeck, A. (2003) The biological tunes the organs of the body: timing
by hormones and autonomic nervous system. Journal of Endocrinology 177, 17–26.
Cardinal, R.N., Parkinson, J.A., Hall, J. and Everitt, B.J. (2002) Emotion and motivation: the role of the amygdala, ventral
striatum and prefrontal cortex. Neuroscience and Biobehavioral Reviews 26, 321–352.
Dawkins, M.S. (2006) Through animal eyes: what behaviour tells us. Applied Animal Behaviour Science 100, 4–10.
Eckert, R., Randall, D. and Augustine, G. (1988) Animal Physiology: Mechanisms and Adaptation, 3rd edn. W.H. Freeman,
New York.
Geary, N. (2004) Endocrine controls of eating: CCK, leptin and ghrelin. Physiology & Behavior 81, 719–733.
Hänninen, L. (2007) Sleep and rest in calves – relationship to welfare, housing and hormonal activity. PhD thesis, Helsinki
University, Helsinki.
Hänninen, L., Hepola, H., Raussi, S. and Saloniemi, H. (2008) Effect of colostrum feeding method and presence of dam on
the sleep, rest and sucking behaviour of newborn calves. Applied Animal Behaviour Science 112, 213–222.
Hemmann, L., Koho, N., Vainio, O. and Raekallio, M. (2013) Effects of feed on plasma leptin and ghrelin concentrations in
crib-biting horses. The Veterinary Journal 98, 122–126.
Janczak, A.M., Braastad, B.O. and Bakken, M. (2006) Behavioural effects of embryonic exposure to corticosterone in
chickens. Applied Animal Behaviour Science 96, 69–82.
Janczak, A.M., Torjesen, P., Palme, R. and Bakken, M. (2007) Effects of stress in hens on the behaviour of their offspring.
Applied Animal Behaviour Science 107, 66–77.
Opris, I. and Bruce, C. (2005) Neural circuitry of judgement and decision mechanisms. Brain Research Reviews 48, 509–
526.
Ressler, N. (2004) Rewards and punishments, goal-directed behaviour and consciousness. Neuroscience and Biobehavioral
Reviews 28, 27–39.
Ruckebush, Y. (1972) The relevance of drowsiness in the circadian cycle of farm animals. Animal Behaviour 20, 637–643.
Schmidt-Nielsen, K. (1997) Animal Physiology: Adaptation and Environment, 5th edn. Cambridge University Press,
Cambridge.
Siegel, J.M. (2005) Clues to the functions of mammalian sleep. Nature 437, 1264–1271.
Tobler, I. (2005) Phylogeny of the sleep regulation. In: Kryger, M.H., Roth, T. and Dement, W.C. (eds) Principles and
Practice in Sleep Medicine, 4th edn. Elsevier Saunders, Philadelphia, Pennsylvania, pp. 77–90.
Valros, A. (2003) Behaviour and physiology of lactating sows – associations with piglet performance and sow postweaning
reproductive success. PhD thesis, Helsinki University, Helsinki.
Valros, A., Palander, P., Heinonen, M., Munsterhjelm, C., Brunberg, E., Keeling, L. and Piepponen, P. (2015) Evidence for a
link between tail biting and central monoamine metabolism in pigs (Sus scrofa domestica). Physiology & Behavior 143,
151–157.
Yun, J. (2015) Importance of maternal behaviour and circulating oxytocin for successful lactation in sows – effects of
prepartum housing environment. PhD thesis, Helsinki University, Helsinki.
4 Motivation and the Organization of Behaviour
G. Mason and M. Bateson
4.1 What Are Motivational States and Why Are They

Important?
Motivational states as ‘causal’ explanations for animal behaviour
Motivational states are concepts that help explain why animals’ behaviour patterns change
from one moment to the next, especially in terms of which external stimuli are responded to
(or instead ignored), and how rapid or effortful their behaviour is. Historically termed
‘drives’ in older psychological and ethological research, these states are envisaged as
occurring within the brain and include those we call ‘thirst’, ‘hunger’ and ‘fear’, as well as
urges to migrate, play, explore, sleep, mate, nest-build, dust-bathe and so on. Motivational
states are reversible, and vary in magnitude from one moment to the next, often in response
to the passage of time since a behaviour was last performed and/or the quality of potentially
reinforcing stimuli appearing in the animals’ environment (e.g. the nutritional value of
presented food, the fertility of a potential sexual partner or the threat posed by a potential
predator).
Motivational states influence several aspects of behaviour which typically change in
tandem. They thus provide causal explanations (see Chapter 1): proximate, mechanistic
accounts for why an animal is currently performing a particular behaviour pattern. At the
physiological level, changes in motivation are often associated with changes in hormonal
state. But this is not always true, and motivational states are thus diverse in both their
underlying mechanisms as well as the functions of the resulting behaviour patterns. For those
driving the restoration or protection of homeostasis (e.g. via eating or drinking), or related to
reproduction (e.g. motivations to mate or nest-build), the crucial roles of specific hormonal
systems are often understood very well. In contrast, hormones play no known role in many
other motivational states, including young animals’ motivations to play and adult animals’
motivations to interact with peers or environmental enrichments. Here, the underlying

mechanisms instead are likely to be solely ‘central’ (i.e. occurring within the brain) and
mediated by oxytocin and other neurotransmitters via mechanisms that are generally little
understood yet.
Why it is important to understand domestic animals’ motivations
Understanding of the properties of motivational states has many practical advantages for
animal management. For example, it can help us design training regimes where the
reinforcers offered are most likely to cause animals to rapidly learn what we want them to
learn. It can help us encourage animals to move from one location to another (e.g. to the
milking parlour, in the case of cattle). Understanding motivation can also help us promote the
feeding, maternal and sexual behaviours that we wish to see in the animals we are looking
after.
Understanding motivational states is also important for helping us improve animal well-
being. It can help us better assess welfare-relevant states like sickness, pain, depression and
boredom. It can help us understand where some abnormal repetitive behaviours (e.g. feather-
plucking in birds) come from and why they are so common in captive animals. Last but not
least, motivational research helps us identify – and potentially alleviate – the sources of some
important frustration-related welfare problems. This is because allowing motivations to be
satisfied is an important source of positive emotions, while preventing them from being met
is an important source of negative ones.
The scope of this chapter
Section 4.2 reviews the properties of highly motivated behaviour, looks at how internal and
external stimuli interact to affect these states and presents various ways in which
motivational changes affect the timing and sequencing of behaviour. Subsequently, Section
4.3 first shows how ethologists have tried to capture these various properties with an array of
different conceptual models, before summarizing neuroscientific research into the
mechanisms underlying how animals choose between different goals and allocate effort into
reaching them. Then, Section 4.4 briefly addresses the links between these two types of state.
Finally, Section 4.5 builds on this, looking in more detail at how understanding motivation
can help us understand and improve animal welfare, as well as at how it can help us manage
animals more efficiently.
4.2 Motivational States and the Organization of Behaviour

The properties of highly motivated behaviour
‘Drives’ and motivational states were first studied by psychologists and ethologists
principally in the 1950s–1990s, with influential research and insights coming from Neal
Miller (e.g. Bower and Miller, 1960; Miller, 1961), Robert Hinde (e.g. Hinde, 1956), Gerard
Baerends (e.g. Baerends, 1976), Fred Toates (e.g. Toates, 1986), David McFarland (e.g.
McFarland, 1985, 1987, 1993) and many others. The subsequent decades saw applied
ethologists taking over, with Marian Dawkins (e.g. Dawkins, 1983, 1990), Christine Nicol
(e.g. Nicol and Guilford, 1991), Jan Ladewig (e.g. Ladewig et al., 2002) and many of the
authors of this book being among the leading figures here. Together this work has resulted in
a corpus of knowledge and ideas about how causal explanations of behaviour are helped by
positing underlying motivational states and what the properties of highly motivated
behaviour are. This is summarized below, with a few modern illustrations added too.
First, changes in motivational state help explain the different decisions made by
individual animals when faced with choices about what activity to do next: how their
priorities are variable and how the stimuli they react to, versus those they do not, change over
time and with circumstances. If you watch an animal for any period of time you will notice
that it sequentially engages in different behaviour patterns, apparently spontaneously
switching between these at intervals. For example, a domestic cat during the course of its day
will switch from feeding to drinking, from hunting to resting and so on. Correspondingly,
changes in motivational state also help explain why sometimes external stimuli act as
powerful triggers for behaviour, but at other times they do not. For instance, in the middle of
the day, domestic chickens switch from foraging to dust-bathing; but before and after this
time a dust bath in the enclosure will be little used. Likewise, if a hen is deprived of a
suitable substrate to dust-bathe in, she will dust-bathe vigorously when she is finally able to
(this vigour being another key feature of highly motivated behaviour, as we outline below),
but once she has finished doing this, for a while afterwards that very same substrate will lose
its ability to stimulate her to perform further dust-bathing (e.g. Hogan, 1997; Olsson and
Keeling, 2005; Nicol, 2015; see also Chapter 11). Such changes in responsiveness can occur
over the course of several days too; for example, some animals go through phases where they
ignore food, even when it is readily available, with red junglefowl brooding their eggs being
one classic example (e.g. Mrosovsky and Sherry, 1980). These reversible changes in
responsiveness typify changes in motivational state.
A second property of motivational states is that they help determine the reinforcing value
of external stimuli. Thus changes in motivational state do not just affect the likelihood that
certain external stimuli will trigger a change in ongoing behaviour, but also the likelihood
that the animal will form associations between these stimuli and either predictive cues (via
classical conditioning; see Chapter 5) or activities (via instrumental or operant learning; see
Chapter 5). Thus it is much easier to train a hungry animal to learn a novel operant response
for food rewards (such as lever-pressing or key-pecking) than it is to train a sated animal
(which may well not learn the task at all). One innovative recent study from Norway
illustrates this well, albeit by manipulating a different aspect of homeostatic state.
Researchers trained horses to touch symbols to have a blanket either placed on them, or, if
they were already wearing one, removed. The horses were trained successfully in just 14
days, because during the trials the team made sure to fully expose the animals to both warm
conditions (to motivate them to seek becoming cooler) and cold conditions (to motivate them
to seek becoming warmer) (Mejdell et al., 2016). Manipulating motivational states to
enhance learning is a topic we come back to in Section 4.5.
Third, motivational states explain which goals animals currently prioritize, with a seeking
or preparatory phase (sometimes termed the ‘appetitive’ phase) often being a key component
of motivated behaviour. Thus as an animal switches to a new activity, it may travel long
distances or overcome other obstacles to reach its new goals. Caged mink and foxes will push
heavily weighted doors to reach certain environmental enrichments, for example (e.g. Mason
et al., 2001; Koistinen et al., 2016); sheep will walk long runways to reach clover and other
food rewards, especially if they have been food-deprived (Champion et al., 2007; Doughty et
al., 2016); and hungry chickens will wade though water just for the chance to forage (even if
this does not result in food) (Dixon et al., 2014). Such appetitive behaviour may well involve
learnt components. Thus operant responses are forms of appetitive behaviour, and animals
are not only more likely to learn these when highly motivated to obtain the reinforcers they
yield, as we saw above, but also, once learned, they are also more likely to perform them, and
to do so with speed and persistence. One illustration of how operant responses can vary in
vigour due to motivational effects comes from foxes. If taught to pull a loop as an operant to
open a door to reach a goal, female foxes will do this more persistently to reach food than to
reach a companion. Furthermore, when working for a companion, some animals in this study
proved more compatible than others and this too affected how persistently they pulled: the
foxes that worked the hardest for social contact were those which would then rest peaceably
with the female they gained access to (Hovland et al., 2011).
The effort or intensity with which animals seek certain reinforcers is not the only thing to
change with motivational state. A fourth characteristic of motivated behaviour is that the
intensity of interaction with the sought-out reinforcer changes: thus the eating of food, the
drinking of water and other so-called ‘consummatory’ phases of motivated behaviour
(meaning those which succeed in reducing their own underlying motivational states) all
become faster, more prolonged and/or harder to interrupt. For example, when watching our
deprived hen finally being allowed to indulge in dust-bathing, we will notice that this
activity’s duration and intensity both increase (e.g. Hogan, 1997; Olsson and Keeling, 2005;
Nicol, 2015). A related change is that poorer-quality stimuli become acceptable (just as we
may be happy to eat stale bread, say, when very hungry). Miller was one of the first
researchers to show this, demonstrating that rats which were previously water-deprived
would drink water tainted with bitter quinine; water that less thirsty animals would reject
(e.g. Miller, 1961).
The fifth and final key property of highly motivated behaviours is that they seem to be
accompanied by emotions: negative emotions if proper consummation is not possible (e.g. if
hungry people or animals cannot access sufficient food to reduce their hunger) and positive
emotions once consummation is in progress. This aspect of motivation is potentially very
important for animal welfare and is discussed in more detail in Section 4.4.
Thus overall, along with the increased probability of a specific behaviour pattern
occurring, we see other correlated changes as well: if obstacles are placed in an animal’s way,
increased efforts will be made to overcome these and perform the behaviour, and once
consummatory behaviour can occur, we see further changes like increased rates of
performance. Motivational states explain this coordination of a suite of changes in the quality
and quantity of specific behaviour patterns, a principle we revisit in Section 4.3 further on.
Note that these properties mean that not all behaviour patterns have a motivational
component. For example, we do not invoke motivation to explain behaviours that are always
elicited in the same way by the same stimuli (e.g. reflexes like blinking, or rapid limb
withdrawal from painful stimuli). Motivational states also play little or no role in behaviours
involving interaction with stimuli that do not act as reinforcers, or in behaviours that are
never preceded by a searching or preparatory phase (like defecation or urination in animals
that do not seek out particular sites for this behaviour, such as cattle and sheep; or suckling in
newborn rats, where it is the mother, rather than the suckling animals themselves, that
controls the onset and cessation of this behaviour) (Hall and Williams, 1983). Motivational
states also do not explain irreversible changes in behaviour (such as those resulting from
maturation).
The effects of internal and external stimuli on motivational states
As should be clear by now, motivational states can be induced by internal stimuli (signals
from the body, such as circulating hormones), external stimuli (cues in the outside world) and
the interactions between them (i.e. how they combine to influence the animal).
Perhaps the most obvious types of internal stimuli to affect motivation are ones reflecting
disrupted homeostasis. These lead to motivations to perform behaviours that will restore
physiological equilibrium. Thus cues from the body signalling dehydration induce strong
motivations to drink, while cues from the body signalling low energy induce motivations to
eat (with the underlying mechanisms for such effects being described well in any good
physiology or biological psychology textbook, e.g. Ferguson, 2000; Carlson, 2005).
Other common examples of internal cues important for motivation include reproductive
hormones and similar chemical signals circulating in the bloodstream (e.g. prostaglandins).
To illustrate, a crucial role of hormones and prostaglandins has been shown for nest-building
motivation in female pigs (as described in Chapter 3): in seminal work on peri-parturient
sows, Colin Gilbert showed that their motivations are elicited within the brain by
prostaglandins released as part of physiological preparation for birth (see e.g. Boulton et al.,
1997). As another example, female marmosets in the right hormonal state (heavily pregnant
or lactating) will also lever-press for the sight and sounds of an infant: effects mediated by
oestrogen (Pryce et al., 1993). Finally, in hens, as oviposition (egg-laying, a behaviour that
typically happens once a day) approaches, birds likewise become increasingly motivated to
sit in a nest, as can be seen from their increasing willingness to push weighted doors or
squeeze through narrows gaps to reach one (reviewed in Nicol, 2015).
External stimuli that influence motivational states include signals of time of day or
season of the year (the effects of the latter often being mediated by hormones); and the sights
or sounds of other animals performing the relevant behaviour, something called ‘social
facilitation’. Other external stimuli that affect motivation are cues signalling a motivationally
relevant resource or threat. We are all familiar with the way the smell of a delicious meal can
make us ravenously hungry, for example. These particular types of motivating external cue
are sometimes known as ‘eliciting stimuli’. Classically conditioned learnt cues (e.g. the sight
of a favourite restaurant for humans or the sound of the fridge opening for a cat) can enhance
motivational states too. Such eliciting stimuli can have powerful effects on animals’
motivations to perform behaviours as diverse as feeding and mating. The strength of these
effects generally relates to the quality or magnitude of the external stimulus: larger amounts
of food induce stronger motivations than smaller amounts, more fertile females are more
motivating to male guppies than are less fertile ones (as we will see below) and so on.
Novelty can be important here too: sometimes animals will habituate to one type of eliciting
cue, only to find a new one powerfully motivating. Thus the presentation of new sexual
partners or of new types or flavours of food, may boost or renew mating or feeding
motivations (a phenomenon known as the ‘Coolidge effect’, after the American president of
the same name, which helps explain why there’s always room for dessert).
How influential such eliciting stimuli are varies between different motivational states and
types of behaviour. External stimuli are particularly important in triggering escape
motivations (which are elicited by predator cues), aggressive motivations (elicited by cues
from rival conspecifics) and predatory responses, like pouncing and biting, in cats and other
carnivores (which are elicited by the movements, squeaks and struggles of prey). Exploratory
behaviour may be another example in some animals; in mice, for instance, this behaviour,
often accompanied by characteristic cautious-looking ‘stretched attend’ postures, seems not
to occur within the home cage, but is elicited strongly when animals are placed in novel
environments (but in mink, as we will see later, exploration may be more intrinsically
motivated). In contrast, internal stimuli (e.g. changes in state that build up over time or reflect
increasing levels of relevant hormones) are relatively more important for many reproductive
behaviours and for behaviours that regulate the body’s energy or water balance.
An experiment with caged mink clearly demonstrates how the role of external factors
varies across different motivational systems. Warburton and Mason (2003) compared
individual minks’ motivation to reach four different resources in parallel (food, swimming-
water, toys and social contact). Animals had to push one of four weighted doors (depending
on which resource they were motivated to access) and then travel along a tunnel 7.5 m long
to reach the resource. This tunnel either doubled around to reach resources that were visible
from the animal’s home cage, and thus visible when it pushed the weighted door (‘cues
treatment’); or it went behind an opaque screen to where the various resources were placed
out of sight from the home cage (‘no cues treatment’; see Fig. 4.1a). Whether or not cues
were present when the mink chose to push the door affected their motivation to visit some of
the resources. When their cues were not detectable from the home cage, mink visited toys
and a conspecific less often than they did food, but when cues were detectable they visited
them as often. In contrast, there was no effect of cues on visits to food or swimming-water
(see Fig. 4.1b). Furthermore, the minks’ ‘consumer surplus’ (a measure of value or
motivational importance) for toys, but no other resource, was significantly lower when the
toys’ cues were screened compared with when they were visible at the choice point.
Fig. 4.1. An experiment to explore the effect of the availability of external cues on minks’
use of four resources (food, swimming-bath, access to social contact and toys). (a) The
apparatus used in the two treatments of the experiment; in the ‘no cues’ treatment the screen
blocks visual cues from the resources at the time the mink decides whether to work to access
a resource. (b) The mean visits per day (one simple measure of motivation) made by the
mink to each of the resources, with one standard error represented by vertical bars.
*Significant difference, P < 0.05. (From Warburton and Mason, 2003, who also show further
data on other indices of motivation.)
Thus the effect of external stimuli varies depending on which motivation you are dealing
with. Indeed, most motivated behaviours are affected by a combination of external cues and
internal states acting together. Classic work by Gerard Baerends and colleagues (described in
Baerends, 1976 and McFarland, 1985) on courtship behaviour in male guppies illustrates this
well by demonstrating how the size of a female fish (an external stimulus relating to her
fecundity and thence attractiveness) and dynamic colour patterns on a male (an indication of
his rapidly changing internal sexual state) combine to affect his courtship behaviour. Male
guppy courtship proceeds through a series of stages indicating increased sexual motivation
and readiness to mate, from following a female, through to ‘posturing’ (facing her and
moving forward and backwards), via through to full sigmoid displays: the most intense forms
of courtship involving bending his body into an S shape. Figure 4.2 shows that the
attractiveness (i.e. size) of the female required to elicit a given display declines as the
relevant internal state of the male changes: the readier to mate he is, the more likely even
small females are to trigger courtship displays, while for males still in more quiescent states,
only the very largest females will succeed in eliciting this behaviour.
Fig. 4.2. The size of a female guppy needed to elicit each of three different displays in male
guppies (posturing, intentional sigmoid and full sigmoid, which occur in a sequence that
leads up to mating) as a function of the markings on the male. The size of the female is an
external stimulus, whereas the markings on the male are an indication of his current internal
state. The smooth lines plot ‘motivational isoclines’ – lines that connect points of equal
likelihood that a male will perform a given display. (From Baerends et al., 1955, described in
McFarland, 1985; see also Baerends, 1976.)
The principles revealed in this experiment apply to many other types of motivational
state too. Miller’s quinine-tolerant thirsty rats are one example; as another, food-deprived
animals need only poor-quality cues to elicit feeding, ingesting even unpalatable food, while
sated animals will ignore most food, responding only to eliciting stimuli that signal the most
highly preferred, palatable or novel food items. That very highly motivated animals will
respond to even the poorest-quality external stimuli may help us understand some abnormal
behaviours in captive animals too, as we will discuss in Section 4.5.
The organization of behaviour

Behavioural sequencing I: appetitive versus consummatory behaviour
Bouts of motivated behaviour usually have some kind of internal structure. For example, as
outlined above, a distinction is often made between appetitive and consummatory phases
within a behavioural sequence. The appetitive phase comes first and comprises active,
flexible, searching behaviours; as we have seen, appetitive behaviour can also include
operant responses that the animal has learnt. Appetitive behaviour may appear quite similar
even when motivated by quite different states. For example, the appetitive phases of food
search, mate search and nesting-site search are all superficially quite similar both within and
between species. Appetitive behaviour is then followed by consummatory behaviour, and this
is typically more stereotyped, unlearnt, species-typical and motivation-typical. Examples
include the specific movements used in foraging (e.g. cows and giraffes tongue-twirl, sheep
and goats bite, pigeons peck); drinking (e.g. cats and dogs lap, pigs and humans suck,
starlings dip and tilt); and mating (e.g. lordosis by female quadrupeds, thrusting by males).
The term ‘fixed action pattern’ is often used for the stereotyped, species-typical forms
involved. Analysing fixed action patterns can be a useful way of identifying the motivational
bases of abnormal behaviour, as we discuss later in Section 4.5.
The appetitive/consummatory distinction is not a rigid one, however. For example, as we
saw in male guppies, courtship is appetitive for mating and yet clearly as stereotyped and
species-typical as many consummatory responses. It is also the case that learnt appetitive
behaviour, such as key-pecking, often takes on the form of the associated consummatory act:
pigeons trained to peck a key for a reinforcement of water ‘suck’ the key with a closed bill as
if drinking, whereas pigeons trained for grain peck with an open bill characteristic of eating.
The movements comprising consummatory behaviour may themselves have a distinct
sequence (with all of the phases within the sequence appearing to be motivated by a single
state). For example, a bout of dust-bathing in a chicken starts with the bird scratching and
bill-raking at the substrate. Next it erects its feathers, squats in the substrate and performs a
sequence of vigorous dust-bathing behaviours. This phase is followed by the bird flattening
its feathers and lying on its side. Finally, it stands up and shakes the substrate from its
plumage before switching to another activity.
Bout lengths and the roles of negative feedback, hysteresis, positive feedback and
switching costs
Consummatory behaviour usually occurs in discrete bouts that last a certain length of time.
For example: animals typically eat in ‘meals’, as opposed to nibbling constantly (lions can
eat 30% of their own body weight in one sitting!); in hens, dust-bathing typically occurs in
bouts of about 20 min that are performed every two days. Therefore, some motivational
mechanisms are needed to explain why a bout of behaviour continues as it does and then
ends. The important ones are negative feedback, ‘hysteresis’, positive feedback and the
effects of ‘switching costs’.
Negative feedback is the process whereby execution of a behaviour pattern reduces the
motivation to perform it and is important in limiting the length of bouts of many different
behaviours. Negative feedback may come from the direct consequences of the behaviour. For
example, the absorption of water into the body post-drinking causes a cascade of
physiological consequences that reduce motivation to drink. Negative feedback may also
come from the performance of a behaviour pattern per se. For example, the satiating effect of
water injected into the body intravenously is not as great as water drunk by mouth. Similarly,
mutant featherless hens will still go through the actions of dust-bathing despite the fact that
the behaviour can be having no effect on the condition of their (non-existent) plumage. This
evidence supports the hypothesis that dust-bathing has rewarding consequences for chickens
above and beyond its direct benefits for feather maintenance.
Whereas negative feedback is important in stopping animals from performing a
behaviour pattern indefinitely, mechanisms are also necessary to make sure that animals
persist at a behaviour long enough to achieve its functional goal. A simple ‘competition’
model of motivation whereby an animal always immediately switches to the behaviour with
the highest motivation suffers from the problem that it will leave animals constantly dithering
between, for example, taking tiny bites of food and tiny sips of water (Fig. 4.3). A range of
solutions has been suggested to deal with this problem, since real animals do not typically
behave like this.
Fig. 4.3. Dithering between eating and drinking behaviour predicted by a simple competition
theory of motivation. The animal is assumed to reduce its initially higher motivation, here
hunger, until thirst is greater than hunger (i.e. the dotted line has been crossed) at which point
it starts drinking. This model does not describe well what animals typically do in realistic
situations (for explanations, see text). (From McFarland, 1985.)
One is hysteresis, a term meaning delayed negative feedback, wherein it takes a period of
time for negative feedback to start reducing motivation. This is one feature of motivational
control that stops dithering and gives bouts some stability. Furthermore, this is often aided by
positive feedback which, although similar in effect, differs in that motivation actually
increases after the start of a bout. One good example comes from the careful studies of eating
in mice by Piet Wiepkema (1971). He analysed the 20–30 mini-bouts of eating that comprise
each single meal a mouse eats. When mice were food-deprived for 24 h, mini-bouts increased
in length (from about 8 to 13 s) and the gaps between them decreased (from about 14 to 9 s),
illustrating an effect of deprivation. However, within each meal, and particularly at the start,
mini-bouts initially also increased in length and occurred closer together (thus e.g. the fifth
mini-bout of a meal was longer than the first), nicely illustrating the effects of positive
feedback. Later on in the meal, the mini-bouts then decreased in duration again, illustrating
the negative feedback that eventually terminates the meal.
Another factor affecting the length of behavioural bouts is the cost of switching to a
different behaviour. Animals stick longer at a single type of behaviour (A) if it is costly (in
terms of time, energy or potential danger) to switch to another (B), perhaps because the
motivation for B has to build up to a higher level before the animal will be prepared to pay
high costs to switch. A classic example of this phenomenon is provided by McFarland’s
(1971) study (described in McFarland, 1985) of switching between eating and drinking in
doves. He showed that if the birds had to negotiate a barrier to switch between these two
behaviours, then they switched less often when the barrier was long (requiring a protracted
detour) than when it was short (making switching easy). It appears, therefore, that animals
take into account the cost of switching activities when deciding when to stop one type of
activity and change to another.
If the explanation for the delay in switching is indeed that motivation has to build to a
higher level before the animal is prepared to pay the switching cost, then this higher
motivation should be reflected in the behaviour of the animal when it finally makes the
switch. This prediction is supported by data from an experiment in which mink were offered
an array of different enrichments to interact with, but the costs of reaching each were
increased by weighting the doors that the mink had to push to reach them. Increased costs
resulted in two main changes in behaviour: (i) as with McFarland’s doves, the animals
switched between enrichments less often, using each enrichment for longer in each bout; and
(ii) when they eventually paid the cost to use a new enrichment, they had a shorter latency to
interact with it after pushing open the door. This latter result illustrates that when the costs of
switching were higher, a higher motivation had built up before the animal would choose to
pay the cost of switching (Cooper and Mason, 2000).
Behavioural sequencing II: ‘feed-forward’ processes

Transitions between one type of motivated behaviour and another (e.g. eating versus
drinking) are often rather predictable and seem designed to usefully anticipate (rather than
merely react to) deviations from homeostasis. Thus often one behaviour pattern follows
another: drinking typically follows eating (in e.g. rats and humans), and may be followed by
thermoregulatory behaviour (if drinking reduces core body temperature); while grooming
follows eating in cats; and preening follows water-bathing in birds.
Such sequences of behaviour are typically thought to be adaptive and may be partially
the product of so-called ‘feed-forward’ mechanisms. Thus while, traditionally, feeding and
drinking behaviour were thought to be controlled by feedback mechanisms that tracked the
levels of physiological indices of hunger and thirst, in reality eating typically motivates
subsequent drinking even before the osmotic consequences of food intake have commenced,
indicating that feed-forward mechanisms are important in the control of feeding and
drinking. In at least some instances, such feed-forward effects are the products of past
experience and learning (e.g. Ramsay et al., 1996). Similar effects, in which deficits are
cleverly anticipated rather than merely reacted to post hoc, occur when animals are exposed
to hot temperatures: these typically motivate drinking, and do so before much water has been
lost to evaporative cooling responses like panting or sweating, to pre-empt any resulting
dehydration (e.g. McFarland, 1987).
4.3 Conceptualizing Motivational Systems and Investigating

Their Neurological Mechanisms
‘Black box’ approaches
In order to summarize and picture how internal and external factors are integrated, and how
negative feedback, positive feedback and other sequencing processes occur, ethologists have
conceptualized or ‘modelled’ motivational states in a number of different ways. As we will
see, different models concentrate on explaining different aspects of the observed phenomena
described above. Ethologists’ approaches have been termed ‘black box’ approaches, as they
do not seek to empirically investigate the underlying neurobiological mechanisms (which are
regarded as opaque and unknown, hence the ‘black box’), but instead to generate broad
patterns or rules that explain the relationship between the events to which an animal is
exposed (inputs to the black box) and its behavioural output. These types of approach are
helpful when the underlying mechanisms are not yet understood, and/or when we want to
extract and illustrate clearly what the key principles at work appear to be. In addition, such
models can also help us generate novel hypotheses to test (as we will illustrate at the end
with the case of dust-bathing).
Motivational states as ‘intervening variables’

The simplest, least controversial and most basic conception is that motivational states are
‘intervening variables’. Summarizing Neal Miller’s influential work and arguments, Michael
Morgan wrote:
Miller argued that a drive could summarize economically the relationship between a number of different inputs and
outputs. To quote one of his examples, an animal can be given a thirst drive by depriving it of water, giving it dry
food, or by an intraperitoneal injection of hypertonic saline. The effect on its behaviour will be to make it drink
more, to press a lever more vigorously for water, to reduce its food intake, or to drink water adulterated with a bitter-
tasting quinine. To explain these findings without postulating a single intervening variable we should need a large
number of separate causal connections … and the drive construct is conceptually more economical. (Morgan, 1979)
Thus if we were to draw all the various possible causal links between all the influences or
inputs, and all the consequences or outputs, outlined above, we would come up with a
complicated diagram with nine possible connections (see Fig. 4.4a). And this would be more
complicated still if we added yet further possible relevant inputs (e.g. the feed-forward
effects of eating any food, not just dry food, or of being exposed to hot temperatures; the
eliciting effects of the sight of another animal drinking; the sight or sound of other external
cues that, through learning, have come to be associated with drinking; sickness; and acute
stress) and yet further possible consequences in terms of how drinking behaviour is altered
(e.g. running speed for water, latency to drink, bout length of drinking, enhanced tendency to
learn new cues or activities that predict access to water, perceived pleasantness of drinking
and so on). Miller’s simpler view instead summarizes the many observed relationships
between these inputs and outputs by postulating just one single intervening variable – in this
case a motivational state called ‘thirst’ – that is influenced by all the many input variables
(dry mouth, external cues and so on) and that in turn influences all the output variables
(amount consumed, quinine tolerated and so on; see Fig. 4.4b). This general concept of
motivational states as intervening variables underlies the more complex models of motivation
described next.
Fig. 4.4. Two alternative models for the explanation of behaviour. Three causal factors on the
left (deprivation of water, ingestion of dry food and injection of saline) lead to three types of
response on the right (increased drinking, increased bar-pressing for water and increased
toleration of quinine in available drinking-water). Miller noted that explanations could
involve postulating nine stimulus–response links (a) or a smaller number of links if an
intervening variable (‘thirst’) is included (b). (From Miller, 1956, described in Barnard, 2004
and Mazur, 2017.)
The psychohydraulic model

One early model of motivation attempted to capture the negative feedback properties of
motivational outputs, as well as to depict how internal and external motivational influences
combine together to affect behaviour. This was Lorenz’s (1950) ‘psychohydraulic model’
(described in McFarland, 1985), which likens behavioural control to a cistern of water that
can overflow via a valve, this overflow representing the performance of behaviour (see Fig.
4.5). The cistern steadily fills from an in-pipe, representing steadily increasing levels of
internal stimuli when a behaviour is not performed. Overflow via the valve occurs if a
combination of the pressure due to the build-up of water in the cistern (representing the
build-up of internal stimuli) and the weights pulling on the valve (representing the influence
of external stimuli) reaches a critical threshold. Flow ceases again once the fluid level in the
cistern, and hence the internal pressure, has decreased.
Fig. 4.5. Lorenz’s psychohydraulic model of motivation. Behaviour occurs when water exits
the reservoir into the trough. The spring holds the valve shut until the valve is opened by a
combination of the pressure of water that has built up in the reservoir (i.e. internal state) and
the weight in the pan (i.e. external stimuli). The behaviour pattern displayed (1–6) depends
on the level of water in the trough, with higher patterns requiring higher levels of water.
(From Lorenz, 1950, described in McFarland, 1985.)
This model admirably represents the ways that internal and external factors combine to
determine motivation; that behaviour patterns occur in predictable sequences; and that
several aspects of behaviour change with increased motivational state, including duration and
rate. However, a valid criticism of this model is that it implies that the motivation always
remains high unless the behaviour is performed, because overflow via the valve is the only
way for the cistern to empty. While this may be appropriate for certain behaviours (e.g.
eating and drinking), it clearly is not the case for others: as we have already discussed,
internal factors seem unimportant for some behaviours. In the absence of rivals, for instance,
an animal will not become more and more likely to fight; and likewise, if predators are
absent, an animal will not become more scared and more likely to hide or make alarm calls.
Thus for many behaviours, if the goal of a behaviour exists already, the motivation to
perform that behaviour will remain low. This type of model also cannot capture hysteresis or
positive feedback.
Thermostat-like models
Other models used the thermostat and similar control systems as their inspiration. These did
not assume an inevitable build-up of internal causal factors in the absence of behavioural
expression: they allow motivations to be weak or even absent altogether as long as the end
point of the behaviour (be that safety, a well-filled stomach or a well-built nest) is already in
existence.
For example, von Holst and Mittelstaedt (1950, described in Mason et al., 1997)
developed a so-called ‘Sollwert–Istwert’ model which conceptualized motivational systems
as acting like a thermostat, causing behaviour that brings an animal’s current state (its
‘Istwert’ – literally, ‘the way the world is’) closer to a desired end point (the ‘Sollwert’ – ‘the
way the world should be’). The greater the discrepancy between Sollwert and Istwert, the
stronger the motivation to perform the behaviour.
Like the psychohydraulic model, in this conception negative feedback is the key means
of behavioural control. But because this is too simple to account for how real animals
partition their time between different activities, more sophisticated representations of the
control and sequencing of behaviour were then developed: so-called ‘control systems’ models
of motivation (McFarland, 1971, described in McFarland, 1985; also Toates, 1986). These
models incorporate means of control other than negative feedback, including hysteresis,
positive feedback and feed-forward (see e.g. Fig. 4.6). They can also incorporate behavioural
sequences (e.g. appetitive and consummatory phases).
Fig. 4.6. A control systems model of drinking in doves. In this model, oral stimulation during
drinking provides positive feedback and gut factors provide negative feedback. The circles
represent points in the control system at which different variables are assumed to be summed
up; a black quadrant in one of these circles changes the sign of an input. (From McFarland
and McFarland, 1968, described in McFarland, 1985.)
A real example of the value of black box models: modelling dust-bathing

In this chapter we have described several features of dust-bathing in hens: it occurs in
discrete bouts about 20 min in duration, most often in the middle of the day. Hogan and Van
Boxel (1993) suggested a simple motivational model of dust-bathing that captures these basic
characteristics. They began by assuming a simple Lorenzian psychohydraulic process
whereby an internal factor for dust-bathing builds up over time, but dissipates rapidly during
dust-bathing bouts (Fig 4.7). But they added a modification: the level of the factor necessary
for dust-bathing to occur was assumed to depend on a starting threshold that is high in the
morning but much lower in the middle of the day. If dust-bathing continued until the internal
factor returned to zero then the model would predict dust-bathing bouts to be longer in the
morning than at midday (which is not observed); therefore it was also necessary to assume a
lower stopping threshold that varies in parallel to the starting threshold.
Fig. 4.7. A motivational model of dust-bathing. The lower panel shows the internal factor for
dust-bathing that is assumed to increase over time in a decelerating manner, and decrease
rapidly when a bout of dust-bathing (B) occurs. The upper panel shows the parallel upper and
lower thresholds assumed to start and stop dust-bathing, respectively. Dust-bathing starts
when the level of the internal factor exceeds the start threshold, and stops when it falls below
the lower threshold. External stimuli such as light, heat and also qualities of the substrate are
assumed to affect the thresholds. (From Hogan and Van Boxel, 1993.)
In ethology, a good model of any kind not only captures facts that we already know to be
true, but will also make novel predictions that we can test. If the circadian variation in the
height of the starting threshold is due to changes in light and temperature (a reasonable
assumption to make), then we would predict that applying additional light and heat should
shift the timing of dust-bathing, but not the total amount performed. Hogan and Van Boxel
(1993) were able to confirm these predictions in a study of Burmese red junglefowl.
However, subsequently other ethologists have argued that the above model does not capture
all the features of dust-bathing and have suggested alternative approaches (for more on this
interesting topic, see Olsson and Keeling, 2005; Nicol, 2015).
Neurobiological approaches
Research by neuroscientists ‘opens the black box’ to study the mechanisms involved as the
brain integrates internal and external information and prioritizes competing behaviour
patterns. In terms of the different physiological mechanisms underlying specific motivational
systems, how the brain detects and responds to various threats to homeostasis (e.g.
dehydration, low sodium or an empty stomach), as well as to seasonal changes that then
influence aggression or reproductive readiness, has been understood rather well for many
years, and for the sake of space we will not cover these topics here (see e.g. Ferguson, 2000;
Carlson, 2005). But, more recently, neurobiological insights have been gained into some of
the processes common to all motivational systems: first, that internal state often influences
the motivational value of potential eliciting stimuli (the phenomenon illustrated earlier with
Baerend’s guppies); second, that to influence behaviour, one motivation must successfully
outcompete others (such that the choice is made to eat, say, rather than drink, play, burrow,
nest-build or perform any of the other behaviours an animal could allocate time to); and third,
that decisions must be made as to the effort worth allocating to meeting any given motivation
(e.g. the size of barrier an animal is willing to overcome to eat). Much of these data come
from brain-scanning humans using functional MRI, although supplemented by animal
research (e.g. single-cell recording in monkeys and lesion studies in rats). Interestingly,
because motivational states are concepts rather than clearly defined biological entities,
neuroscientists working on such topics often do not use the term ‘motivation’ at all. Instead,
they generally refer to these three aspects of highly motivated behaviour as respectively
‘stimulus valuation’ (or sometimes ‘incentive salience’ or ‘reward valuation’); choice; and
the expenditure of effort or work.
Numerous studies have now shown that the value of any potential reward is computed in
a part of the prefrontal cortex that is variously termed the ventromedial prefrontal cortex or
the orbitofrontal cortex (with other interconnecting parts of the forebrain sometimes involved
too, notably the anterior cingulate and nucleus accumbens). The quality and quantity of the
reward factor into this assessment, as does any cost involved in obtaining it; but, importantly,
so too does the subject’s internal state, such that water is given more value during a state of
water deprivation, food rewards are given more value during a state of food deprivation and
so on (e.g. Hare et al., 2011; Rolls, 2014; Saker et al., 2014; Sescousse et al., 2015). This
same brain region also plays a central role in subjects faced with a choice, when different
potential rewards must be compared (as Levy and Glimcher, 2012 put it, ‘when we need to
choose between a large amount of water and a single apple’). The relative net values of food,
water, sexual stimulation and even monetary rewards can be weighed up here, allowing a
subject to choose between them (e.g. Hare et al., 2011; Rolls, 2014).
Once this decision has been made, appropriate signals are then sent to the motor cortex
so that the right behaviour patterns can now be performed. These signals are relayed via two
brain regions that play a crucial role in the quality of these behaviour patterns, especially
their vigour, speed and ‘effortfulness’. These two regions are the anterior cingulate and the
nucleus accumbens (a part of the basal ganglia sometimes also called the ventral striatum)
(e.g. Der-Avakian and Markou, 2012; Salamone et al., 2016). Lesions to either of these
regions leave animals’ consummatory behaviours unchanged, but selectively compromise
their abilities to perform effortful appetitive behaviours. Thus rats with such lesions will not
climb barriers or persistently lever-press to obtain even very high-value rewards (despite still
interacting intensely with these rewards if they can access them without effort): a form of
apathy superficially resembling the ‘anhedonia’ discussed in Section 4.5 further on.
This exciting, growing area of research is primarily fuelled not by a desire to understand
animal behaviour, but instead by the need to understand problematic ‘over-motivated’
behaviours in humans (compulsions and addictions such as overeating, gambling and drug-
taking), as well as the generalized loss of motivation typical of other disorders like
depression. However, applied ethologists are now starting to find this mechanistic
information useful, for instance to understand abnormal behaviour in captive animals (see
Section 4.5 and Chapter 9).
4.4 Motivation and Emotion

The concepts of motivation and emotion are closely tied together. In this section we therefore
describe why understanding motivation is likely to be important for animal welfare, by
helping us promote positive emotions and avoid negative ones. But we will also show that
gaining a deep conceptual understanding of precisely how these two constructs interrelate
may be somewhat tricky.
In humans, emotions involve subjective feelings that are either positive or negative,
pleasant or unpleasant. And both everyday experience and controlled laboratory studies show
us that such feelings often accompany states of high motivation. Extreme hunger and thirst
are distressing, for example (some researchers even call these states ‘homeostatic emotions’),
as is being frustrated of food or water access when you are in such states (e.g. by being
shown food or water that you cannot reach). Consummating a motivational state, especially if
that state is strong, in contrast, is subjectively positive. For example, it feels very pleasant to
drink water when you are thirsty (but rather neutral if you are not, and distinctly unpleasant if
you have already drunk water to excess; Saker et al., 2014). In humans, the positive
emotional impact of rewards (how pleasant we find getting them) therefore typically mirrors
their motivational importance: how much we want them and how hard we will work for
them. Conversely, the negative emotional impact of punishers (how unpleasant we find
stimuli like spiders or electric shocks) motivates us to avoid or escape from them.
The neurobiological bases of these responses are well understood (with e.g. the nucleus
accumbens and orbitofrontal cortex being involved in positive emotional responses to highly
motivating rewards) and these effects appear to be homologous between humans and other
mammals. This makes it possible (although not all agree; see e.g. Rolls, 2014) that non-
human animals have similar pleasant feelings to us when performing consummatory
behaviour and similar negative feelings to us when prevented from performing highly
motivated activities. In support of this hypothesis, behavioural signs of aversion (such as
escape attempts and aggression) and physiological signs of high arousal (e.g. elevated
corticosteroids) are common in frustrated animals, and, furthermore, these responses can be
alleviated with tranquillizing drugs (e.g. McFarland, 1987; Mason and Burn, 2011).
This is not to say that all motivational states have an emotional component. For many,
this will only be true if the motivation is very strong. ‘Hunger’, for example, is a
motivational state that is only sometimes emotionally negative, often instead being simply an
emotionally neutral reason to change your behaviour (e.g. triggering you to stop working and
go to the fridge). But if not all motivational states have an emotional component, do all
emotional states have a motivational one? This is somewhat harder to clarify, partly because
the way in which people use these concepts varies in the literature, and also because
emotional states (like motivational states) are diverse. Rolls (2014), for example, states
boldly that ‘emotions are motivating’, but like many neuroscientists he uses ‘emotion’ in a
rather non-intuitive way that has no connotations of associated subjective feelings. Cognitive
psychologists working on human emotion (subjective feelings and all) have instead proposed
more generally that the function of emotions is to alter behavioural priorities (e.g. Oatley and
Johnson-Laird, 1987). But such alterations may well involve changes in motivated behaviour
(e.g. Russell, 2003). Indeed, some emotional states have such clear motivational effects that
changes in motivated behaviour can be used to infer the presence of the underlying emotion
(e.g. elevated motivations to explore being used to infer the emotion of boredom of mink, as
we show in Section 4.5). Blurring the boundaries between emotions and motivations yet
further, certain states can be labelled as either emotions or motivational states in the
literature; this is true of ‘fear’ for example, which is variously called an emotion and a
motivation in papers written by different authors (because it always involves negative
subjective feelings – in us, at least – and has the behavioural properties of a motivational
state discussed in this chapter).
This terminological quagmire reflects the fact that both emotions and motivations are
merely concepts or constructs: both are hypothetical intervening variables with somewhat
fuzzy borders. Furthermore, different researchers in different fields do not use the terms in
fully consistent ways. But, the take-home messages are simply these. First, be aware of this
complexity when reading widely across different fields. Second, recognize that preventing or
allowing animals to satisfy strong motivations is likely to have emotional consequences – the
topic we cover next.
4.5 Motivation and Applied Ethology

There are many reasons why those interested in domestic animal behaviour should
understand motivation, and we end the chapter with these.
Unfulfilled motivations and poor animal welfare
The strong links between motivation and emotion led Marian Dawkins (1990) and others to
suggest that assessing motivational priorities is crucial for maximizing the welfare of animals
kept in captivity: if they prove highly motivated for resources they are denied in their typical
housing conditions, then we should assume they have poor welfare due to negative emotions
and we should then meet those motivations if we wish to solve these welfare problems.
All too often, captive animals have insufficient resources or opportunities for
homeostasis, being exposed to, for example, food restriction (e.g. pregnant sows; broiler
breeders; some dairy calves), excessive cold/heat (mink underfed during the winter; cattle in
shadeless fields or feedlots; pigs transported on trucks) and other threats to homeostasis.
Furthermore, when presented with opportunities to restore homeostasis, these animals show
themselves highly motivated. For example, breeding sows and broiler chickens, typically fed
less than half the ration they would choose to consume, will work very hard to obtain extra
food on operant schedules, so prompting extensive research into how to mitigate their hunger
without allowing them to eat so much that they become fat (the problem their normally
restricted diets aims to avoid) (e.g. D’Eath et al., 2009). Thermal homeostasis is important
too: hot dairy cows, for instance, will actively seek areas of shade even if this comes at the
expense of being able to lie down (Schütz et al., 2008) – itself another highly motivated
activity, as we will see next.
Animals are not only motivated to restore homeostasis. When ‘asked’– experimentally –
if they will work for the chance to perform naturalistic activities, often they answer ‘yes’,
even when these behaviours offer no direct nutritional or physiological benefits. For
example, starting with the case of lying for dairy cattle, these animals will perform operant
responses to be able to do this, especially if they have already been deprived of lying for a
period (Jensen et al., 2005). They also prefer soft lying surfaces to hard ones (e.g. Norring et
al., 2010). Ferrets, likewise, will push very heavy weights to access comfortable places to
sleep and rest, preferring hammocks particularly strongly (Reijgwart et al., 2016). Turning to
other motivated behaviours that may not always be possible in captivity, pregnant sows will
work to access and/or build nests, as will hens about to lay (reviewed by e.g. Mason and
Burn, 2011). In the former, the critical role of prostaglandins released during physiological
preparation for birth suggests that probably the only way to reduce this motivation is by
allowing sows to interact with nesting materials (see e.g. Boulton et al., 1997). Cats placed in
a novel environment appear highly motivated to hide (Vinke et al., 2014). As we saw earlier
with vixens as the example, interactions with same-age conspecifics are also valued by some
animals, with isolated primates, rats and ferrets also working hard for social companionship
(Mason and Burn, 2011; Reijgwart et al., 2016). Human contact may be valued too; for
example, calves will seek opportunities to be gently brushed (Westerath et al., 2014). And
again, as we saw earlier, hungry broiler breeders will surmount obstacles just for the chance
to rake and scratch in litter, even when this yields no food, while naturally semi-aquatic mink
will also work for water to swim, wade and ‘head-dip’ in (all aspects of their natural foraging
behaviour). Thus some aspects of natural foraging behaviour appear intrinsically motivated,
independent of feeding itself.
Specific natural behaviours whose prevention causes suffering are termed ‘ethological’ or
‘behavioural needs’. Of course, to be sure that the findings reported above really do indicate
poor welfare in animals unable to perform these motivated activities, we need to check that
these measured motivations are not simply elicited by cues from the resources animals are
exposed to while being tested (since these, as we saw, could act to enhance motivation).
However, in at least some of these examples (e.g. nest-building for sows, hiding for cats,
social isolation for rats and primates and perhaps also swimming for some mink), good
additional stress-related data (see Chapter 7 and Mason and Burn, 2011) show that preventing
these natural behaviours is a real cause for concern. Thus motivations to meet certain
ethological needs can cause poor welfare when opportunities to consummate these are
absent. Furthermore, frustrated motivations to meet ethological, as well as homeostatic,
needs can help explain the origins of some forms of abnormal repetitive behaviour (e.g.
stereotypic behaviours), as we discuss in the following subsection.
Two other sorts of elevated motivation are of welfare concern too. First, as well as these
specific ethological needs, some animals have more non-specific motivations to be exposed
to and interact with diverse stimuli, even ones that induce fear responses. This might seem
strange, but strong urges to seek out stimulation of all kinds (even forms linked with danger)
typify bored humans, and similar ‘boredom-like’ states appear to occur in mink (Meagher
and Mason, 2012; Meagher et al., 2017). These animals are highly motivated to access
enriched environments. If kept in small barren cages instead, they become hyper-exploratory
to any stimulus presented outside the cage (even cues of predators or the aversive leather
gloves that farmers use to catch them), showing shorter latencies to investigate these as well
as longer interaction times. This suggests that mink have intrinsic motivations to explore
which are frustrated when their environments are unstimulating (Meagher and Mason, 2012;
Meagher et al., 2017). (Note too how, in this example, the emotional state of boredom was
‘operationalized’ – made measurable – via its effects on motivation.) The second form of
generally elevated motivation relevant here is that of ‘stress sensitization’, wherein acute
stress can make both humans and animals highly motivated to seek out rewards in general, be
these food rewards like sucrose or recreational drugs (e.g. van der Harst et al., 2003; Sinha
and Jastreboff, 2013). Here, unlike in all the previous examples in this subsection, these
elevated motivations do not show that these specific motivation systems had been frustrated,
but instead that rewards and positive experiences in general have been recently missing from
the animal’s life. (Note that the opposite effect, a generalized decreased responsiveness to
rewards or ‘anhedonia’, can also appear instead when animals are chronically stressed, as we
will review below.)
Understanding chronic behavioural changes indicative of poor welfare
The motivational underpinnings of stereotypic behaviour
A second major reason to investigate domestic animals’ motivations out of concern for their
welfare is that this can help us understand why abnormal behaviour patterns like stereotypic
behaviours are so common (see Chapter 9). ‘Stereotypies’, traditionally defined as abnormal
repetitive activities with no apparent goal or function, for example, are very prevalent in
captive animals: over 85 million perform them worldwide, and in some populations (e.g. zoo-
housed giraffes and single-housed laboratory primates) they are nearly ubiquitous. Often
these seem to be expressions of sustained, high motivations to perform species-typical
behaviours, directed towards suboptimal substrates that do not permit the proper
consummation of normal behaviour; in other cases, they seem to derive from attempts to
escape frustrating barren housing conditions. Powerless to effect real change, the animal is
able only to repeatedly attempt to perform a substitute for the relevant behaviour, or try and
try again to escape the frustrating situation (see e.g. Hughes and Duncan, 1988).
Thus caged wild-caught birds, for instance, typically show more route-tracing than
captive-bred birds, possibly because they are trying to escape; premature maternal separation
typically induces similar stereotypic attempts to escape, or instead to suckle, in young
mammals; and removing or delaying expected food rewards elevates stereotypic pacing,
weaving and oral behaviours in captive pigs and carnivores (Mason, 1991; Latham and
Mason, 2008; Feenders and Bateson, 2012). Observing the stereotypic digging of caged
gerbils, Wiedenmayer (1997) reasoned that, in the wild, the consummatory stimulus that
would bring digging to a close would be a tunnel leading to a nesting chamber. He tested this
idea in the laboratory and found that gerbils that were raised with extra space, sand to dig in
or a simple plastic nest chamber still showed the abnormal behaviour – but animals raised
with a nest chamber reached via a plastic tunnel did not. Nevison et al. (1999) similarly
manipulated external stimuli to test the idea that bar-mouthing by laboratory mice derives
from repeated escape attempts. They found that if a door in the cage lid was regularly opened
and mice sometimes allowed to exit from it, their bar-chewing became directed to that site;
on the other hand, if clear Plexiglas was placed over a region of the cage lid, to prevent odour
cues from entering at that point, mice would move their bar-mouthing elsewhere. More
recent work has focused on feather-pecking by hens, demonstrating that it derives from a
very specific form of redirected foraging (rather than from redirected dust-bathing pecks, as
had been suggested in the 1980s). Harlander-Matauschek and colleagues (e.g. Harlander-
Matauschek et al., 2007) discovered that individual hens which display a lot of feather-
pecking are actually highly motivated to ingest feathers: offered bowls of food, shaving and
feathers, they approach the feathers more rapidly, stay at the feather-containing bowls for
longer and ingest more feathers than non-feather-pecking individuals. They will even
perform an operant to gain feathers to eat. Meanwhile other researchers (Dixon et al., 2008)
examined the ‘fixed action patterns’ involved in feather-pecking: their careful quantitative
comparison of the morphology of feather pecks with foraging pecks, dust-bathing pecks,
drinking pecks and exploratory pecks showed that foraging and feather-pecking involved
near-identical fixed action patterns.
These explanations, in terms of specific motivational frustrations, may thus explain the
form and timing of many abnormal behaviours. However, they do not seem to be the whole
story. There is growing evidence that these strange behaviours are exacerbated by alterations
within the brain – perhaps induced by stress and/or by abnormal developmental environments
– that predispose animals to excessive levels of inappropriate repetition (see Chapter 9).
Furthermore, some of these changes may well involve broad motivational abnormalities:
generalized tendencies to inappropriately, excessively respond to potential rewards that seem
to be like prolonged forms of stress sensitization, perhaps related to the tendencies involved
in ‘overly motivated’ human compulsions and addictions. The evidence for this hypothesis is
that some highly stereotypic animals show elevated activity in the nucleus accumbens, a
region that, as we saw earlier, plays a critical role in energetic appetitive behaviour. In
particular, highly stereotypic horses have greater densities of dopamine receptors here
(McBride and Hemmings, 2005), while highly stereotypic mice show other evidence of
sustained accumbens activation: increased levels there of ‘ΔFosB’, a transcription factor that
accumulates over time with repeated neuronal stimulation (Phillips et al., 2016). This will
continue to be an exciting area for future research.
Anhedonia and poor animal welfare

In contrast to what we discussed in ‘Unfulfilled motivations and poor animal welfare’ earlier,
where poor welfare was associated with increases in motivation, poor welfare is sometimes
also associated with a decrease in motivation termed anhedonia. In humans, anhedonia
(derived from the Greek for ‘without pleasure’) is characterized by an inability to pursue and
partake in usually pleasurable activities. It is believed to be caused by dysfunction of the
brain’s reward systems critical for sensing pleasure and is a key symptom of depression (Der-
Avakian and Markou, 2012). In omnivorous or herbivorous animals, anhedonia is detected by
measuring the quantity of sweetened water or sweetened food consumed: a drop in sucrose
consumption compared with a baseline condition is classed as anhedonia in such species.
Animals housed in chronically stressful conditions assumed to promote depressive-like states
often show the predicted drop in sucrose consumption indicative of anhedonia. For example,
rats and mice exposed unpredictably for 3 weeks to frequencies of ultrasound corresponding
to the screams of conspecifics exposed to life-threatening conditions develop a suite of
indicators of depression including reduced sucrose consumption (Morozova et al., 2016). In
another relevant study, a subset of riding-school horses with symptoms of a depression-like
state including withdrawn behaviour also showed reduced sucrose consumption (Fureix et
al., 2015).
These results raise the question: why is poor welfare sometimes associated with increases
in motivation but sometimes instead with anhedonia? The answer seems to lie at least in part
in the duration of exposure to poor conditions, with long-lasting, chronically unpleasant
environments or unmet needs being most prone to inducing depressive-like states and
anhedonia. It seems possible too that these depression-like effects are an alternative response
to becoming highly stereotypic (see e.g. Fureix et al., 2016).
Practical benefits of understanding motivation: encouraging desired
behaviour patterns
Welfare aside, understanding motivation can assist in many other ways with animal
management. It can help us design better training regimes when we wish animals to learn
particular associations, help us move animals more efficiently from one location to another
and help us encourage animals to eat and to care for their young, to list but a few examples.
Motivation and learning are, as we have seen, intimately linked: the opportunity to
perform highly motivated activities is powerfully reinforcing – far more so than the
opportunity to perform activities that are not highly motivated. Thus in choosing, say, how to
use food treats to reward desired behaviours during training, we need to understand that these
will only be effective if the animal is motivated to obtain them. The use of unpalatable food
rewards, or working with animals that are sated owing to just having eaten a large meal, will
therefore impede training. In contrast, the use of highly palatable treats, or prior moderate
food deprivation, combined with the use of small treats that do not sate the animal, will all
facilitate training by helping to ensure the animal is sufficiently motivated to learn the
behaviours we desire so that it can obtain the food rewards (e.g. McGreevy and Boakes,
2008). This can have animal welfare implications too; rather than depriving rats or mice of
food to get them to run tasks in the laboratory, researchers who understand motivation can
choose instead to use highly palatable sweet breakfast cereals as rewards. Similarly,
undeprived hens and starlings will work for highly desirable mealworms, while the
motivation of dairy cattle to eat highly palatable grain pellets has been exploited to train these
animals to voluntarily enter the milking parlour. This is essential for the smooth running of
automated ‘robotic’ systems, which operate with minimal human labour and rely on cows
choosing to visit the parlour two or three times daily.
Turning to species-typical behaviours, understanding motivation can help us enhance and
manipulate these too. For example, by appreciating how a ewe’s hormonal state and lamb
odour combine to motivate maternal care in sheep, farmers can devise effective ways to
manipulate ewes so that they will accept orphan lambs (by timing their introduction attempts
appropriately relative to parturition; and by modifying olfactory cues from the orphaned lamb
so as to provide the strongest possible eliciting stimulus). Allowing sows to express their
strong motivations to interact with nesting materials can help promote maternal care in this
species as well: being able to nest-build elevates plasma oxytocin and prolactin (two
hormones important in lactation) and triggers more careful maternal care, potentially
resulting in faster-growing piglets (Yun et al., 2014).
Knowing that highly motivated animals will be less choosy about how to satisfy their
motivations can also be practically useful. For example, we can use this knowledge to
persuade sexually aroused stallions to ejaculate into an artificial mare for semen collection,
by first using a ‘teaser’ mare, in oestrous, to increase the males’ sexual motivations. We can
also use this knowledge to understand why hens that are very highly motivated to lay will
sometimes choose to do so in very un-nest-like locations such as the floor of an aviary. Since
such ‘floor eggs’ can end up dirty or damaged, this in turn has prompted lots of research into
improving nest-box design to create ones that are more readily accepted and used by all hens
(e.g. Ringgenberg et al., 2015). In addition, understanding the eliciting role of external
stimuli can be important in other ways, for instance in persuading animals to eat. Warming
food so that its odours are stronger is often recommended as a way to help motivate sick cats
and dogs to eat. Exploiting the positive feedback phase typical of the start of bouts of
motivated behaviour can be useful here too. Tempting a sick animal to eat particularly
palatable ‘treats’ can be used to elevate feeding motivation and so trigger a bout of feeding.
4.6 Conclusions and Links to Other Chapters

This chapter introduced data and theories relating to the concept of motivation and discussed
these in the context of applied ethology. We would like to end by returning to Tinbergen’s
four questions and emphasizing the strengths of this ethological approach. Although the
study of motivation focuses on questions of proximate causation, in trying to understand the
motivation for a particular behaviour pattern it will often be important to consider
Tinbergen’s other three questions (see Chapter 1). The causal mechanisms responsible for a
behaviour pattern are often affected by the developmental experience of an animal, meaning
that a good understanding of the ontogeny of a behaviour pattern could be crucial in
understanding why a captive animal is behaving abnormally. The causal mechanisms
responsible for a behaviour pattern have also evolved in order to motivate animals to perform
behaviour patterns that would, at least in their ancestral environments, optimize survival and
reproduction. Therefore, understanding the ultimate goal and survival value of a behaviour
pattern, and its phylogenetic roots, is also important in understanding how it is controlled.
This is particularly important when solving problems in applied ethology because, as we
have seen, the unfulfilled ethological needs and abnormal behaviour patterns of captive
animals may often have their origins in adaptive natural behaviours.
Acknowledgements
Thanks to Laura Dixon and Naomi Latham for some useful examples, to Aitor Arrazola for
Fig. 4.2 and to David Fraser for Fig. 4.4. Also to Tina Widowski, Michelle Hunniford, Jeff
Rau and Lee Niel for help with practical management examples.
References
Baerends, G.P. (1976) The significance of colour patterns in fish for the study of some fundamental issues in behaviour.
Revue des Travaux de l’Institut des Pêches Maritimes 40, 413–423.
Barnard, C. (2004) Animal Behaviour: Mechanism, Development, Function and Evolution. Pearson, Harlow, UK.
Boulton, M.I., Wickens, A., Brown, D., Goode, J.A. and Gilbert, C.L. (1997) Prostaglandin F2α-induced nest-building in
pseudopregnant pigs. 1. Effects of environment on behaviour and cortisol secretion. Physiology & Behavior 62, 1071–
1078.
Bower, G.H. and Miller, N.E. (1960) Effects of amount of reward on strength of approach in an approach-avoidance conflict.
Journal of Comparative and Physiological Psychology 53, 59–62.
Carlson, N.R. (2005) Foundations of Physiological Psychology, 6th edn. Pearson Allyn & Bacon, Boston, Massachusetts.
Champion, R.A., Lagstrom, N.A. and Rook, A.J. (2007) Motivation of sheep to eat clover offered in a short-term closed
economy test. Applied Animal Behaviour Science 108, 263–275.
Cooper, J.J. and Mason, G.J. (2000) Increasing costs of access to resources cause re-scheduling of behaviour in American
mink Mustela vison: implications for the assessment of behavioural priorities. Applied Animal Behavior Science 66,
135–151.
Dawkins, M.S. (1983) Battery hens name their price: consumer demand theory and the measurement of ethological ‘needs’.
Animal Behaviour 31, 1195–1205.
Dawkins, M.S. (1990) From an animal’s point of view: motivation, fitness and animal welfare. Behavioural and Brain
Sciences 13, 1–61.
D’Eath, R.B., Tolkamp, B.J., Kyriazakis, I. and Lawrence, A.B. (2009) ‘Freedom from hunger’ and preventing obesity: the
animal welfare implications of reducing food quantity or quality. Animal Behaviour 77, 275–288.
Der-Avakian, A. and Markou, A. (2012) The neurobiology of anhedonia and other reward-related deficits. Trends in
Neurosciences 35, 68–77.
Dixon, L., Duncan, I.J.H. and Mason, G.J. (2008) What’s in a peck? Using fixed action patterns to identify the motivation
behind feather-pecking. Animal Behaviour 76, 1035–1042.
Dixon, L.M., Brocklehurst, S., Sandilands, V., Bateson, M., Tolkamp, B.J. and D’Eath, R.B. (2014) Measuring motivation
for appetitive behaviour: food-restricted broiler breeder chickens cross a water barrier to forage in an area of wood
shavings without food. PLoS One 9(7), e102322.
Doughty, A.K., Ferguson, D., Matthews, L.R. and Hinch, G.N. (2016) Assessing feeding motivation in sheep using different
behavioural demand models and measures. Applied Animal Behaviour Science 180, 43–50.
Feenders, G. and Bateson, M. (2012) The development of stereotypic behavior in caged European starlings, Sturnus
vulgaris. Developmental Psychobiology 54, 773–784.
Ferguson, E.D. (2000) Motivation: A Biosocial and Cognitive Integration of Motivation and Emotion. Oxford University
Press, Oxford.
Fureix, C., Beaulieu, C., Argaud, S., Rochais, C., Quinton, M., Henry, S., Hausberger, M. and Mason, G. (2015)
Investigating anhedonia in a non-conventional species: do some riding horses Equus caballus display symptoms of
depression? Applied Animal Behaviour Science 162, 26–36.
Fureix, C., Walker, M., Harper, L., Reynolds, K., Saldivia-Woo, A. and Mason, G. (2016) Stereotypic behaviour in standard
non-enriched cages is an alternative to depression-like responses in C57BL/6 mice. Behavioural Brain Research 305,
186–190.
Hall, W.G. and Williams, C.L. (1983) Suckling isn’t feeding, or is it? A search for developmental continuities. Advances in
the Study of Behavior 13, 219–254.
Hare, T.A., Schultz, W., Camerer, C.F., O’Doherty, J.P. and Rangel, A. (2011) Transformation of stimulus value signals into
motor commands during simple choice. Proceedings of the National Academy of Sciences USA 108, 18120–18125.
Harlander-Matauschek, A., Benda, I., Lavetti, C., Djukic, M. and Bessei, W. (2007) The relative preferences for wood
shavings or feathers in high and low feather pecking birds. Applied Animal Behaviour Science 107, 78–87.
Hinde, R.A. (1956) Ethological models and the concept of ‘drive’. British Journal for the Philosophy of Science 6, 321–331.
Hogan, J.A. (1997) Energy models of motivation: a reconsideration. Applied Animal Behaviour Science 53, 89–105.
Hogan, J.A. and Van Boxel, F. (1993) Causal factors controlling dustbathing in Burmese red junglefowl: some results and a
model. Animal Behaviour 46, 627–635.
Hovland, A.L., Akre, A.K., Flø, A., Bakken, M., Koistinen, T. and Mason, G.J. (2011) Two’s company? Solitary vixens’
motivations for seeking social contact. Applied Animal Behaviour Science 135, 110–120.
Hughes, B.O. and Duncan, I.J.H. (1988) The notion of ethological ‘need’, models of motivation and animal welfare. Animal
Behaviour 36, 1696–1707.
Jensen, M.B., Pedersen, L.J. and Munksgaard, L. (2005) The effect of reward duration on demand functions for rest in dairy
heifers and lying requirements as measured by demand functions. Applied Animal Behaviour Science 90, 207–217.
Koistinen, T., Korhonen, H.T., Hämäläinen, E. and Mononen, J. (2016) Blue foxes’ (Vulpes lagopus) motivation to gain
access and interact with various resources. Applied Animal Behaviour Science 176, 105–111.
Ladewig, J., Sørensen, D.B., Nielsen, P.P. and Matthews, L.R. (2002) The quantitative measurement of motivation:
generation of demand functions under open versus closed economies. Applied Animal Behaviour Science 79, 325–331.
Latham, N.R. and Mason, G.J. (2008) Maternal deprivation and the development of stereotypic behaviour. Applied Animal
Behaviour Science 110, 84–108.
Levy, D.J. and Glimcher, P.W. (2012) The root of all value: a neural common currency for choice. Current Opinion in
Neurobiology 22, 1027–1038.
Mason, G.J. (1991) Stereotypies: a critical review. Animal Behaviour 41, 1015–1037.
Mason, G.J. and Burn, C. (2011) Behavioural deprivation. In: Appleby, M., Mench, J.A., Olsson, A. and Hughes, B.O. (eds)
Animal Welfare. CAB International, Wallingford, UK, pp. 98–119.
Mason, G., Cooper, J. and Garner, J. (1997) Models of motivational decision-making and how they affect the experimental
assessment of motivational priorities. In: Forbes, J.M., Lawrence, T.L.J., Rodway, R.G. and Varley, M.A. (eds) Animal
Choices. Occasional Publication No. 20. British Society of Animal Science, Penicuik, UK, pp. 9–17.
Mason, G., Cooper, J. and Clarebrough, C. (2001) Frustrations of fur-farmed mink. Nature 410, 35–36.
Mazur, J.E. (2017) Learning and Behavior, 8th edn. Routledge, New York.
McBride, S.D. and Hemmings, A. (2005) Altered mesoaccumbens and nigro-striatal dopamine physiology is associated with
stereotypy development in a non-rodent species. Behavioural Brain Research 159, 113–118.
McFarland, D. (1985) Animal Behaviour. Longman Scientific & Technical, Harlow, UK.
McFarland, D. (1987) Oxford Companion to Animal Behaviour. Oxford University Press, Oxford.
McFarland, D. (1993) Animal Behaviour: Psychobiology, Ethology and Evolution, 2nd edn. Oxford University Press,
Oxford.
McGreevy, P.D. and Boakes, R.A. (2008) Carrots and Sticks: Principles of Animal Training. Cambridge University Press,
Cambridge.
Meagher, R.K. and Mason, G.J. (2012) Environmental enrichment reduces signs of boredom in caged mink. PLoS One
7(11), e49180.
Meagher, R., Campbell, D. and Mason, G.J. (2017) Effects of enrichment on boredom-like states in mink and their
behavioural correlates: a replicate. Applied Animal Behaviour Science (under review).
Mejdell, C.M., Buvik, T., Jørgensen, G.H. and Bøe, K.E. (2016) Horses can learn to use symbols to communicate their
preferences. Applied Animal Behaviour Science 184, 66–73.
Miller, N.E. (1961) Analytical studies of drive and reward. American Psychologist 16, 739–754.
Morgan, M.J. (1979) The concept of drive. Trends in Neurosciences 2, 240–242.
Morozova, A., Zubkov, E., Strekalova, T., Kekelidze, Z., Storozeva, Z., Schroeter, C.A., Bazhenova, N., Lesch, K.P., Cline,
B.H. and Chekhonin, V. (2016) Ultrasound of alternating frequencies and variable emotional impact evokes depressive
syndrome in mice and rats. Progress in Neuro-psychopharmacology and Biological Psychiatry 68, 52–63.
Mrosovsky, N. and Sherry, D.F. (1980) Animal anorexias. Science 207, 837–841.
Nevison, C.M., Hurst, J.L. and Barnard, C.J. (1999) Why do male ICR(CD-1) mice perform bar-related (stereotypic)
behaviour? Behavioural Processes 47, 95–111.
Nicol, C.J. (2015) The Behavioural Biology of Chickens. CAB International, Wallingford, UK.
Nicol, C.J. and Guilford, T. (1991) Exploratory activity as a measure of motivation in deprived hens. Animal Behaviour 41,
333–341.
Norring, M., Manninen, E., de Passille, A.M., Rushen, J. and Saloniemi, H. (2010) Preferences of dairy cows for three stall
surface materials with small amounts of bedding. Journal of Dairy Science 93, 70–74.
Oatley, K. and Johnson-Laird, P.N. (1987) Towards a cognitive theory of emotions. Cognition and Emotion 1, 29–50.
Olsson, I.A.S. and Keeling, L.J. (2005) Why in earth? Dustbathing behaviour in jungle and domestic fowl reviewed from a
Tinbergian and animal welfare perspective. Applied Animal Behaviour Science 93, 259–282.
Phillips, D., Choleris, E., Ervin, K.S., Fureix, C., Harper, L., Reynolds, K., Niel, L. and Mason, G.J. (2016) Cage-induced
stereotypic behaviour in laboratory mice covaries with nucleus accumbens FosB/ΔFosB expression. Behavioural Brain
Research 301, 238–242.
Pryce, C.R., Döbeli, M. and Martin, R.D. (1993) Effects of sex steroids on maternal motivation in the common marmoset
(Callithrix jacchus): development and application of an operant system with maternal reinforcement. Journal of
Comparative Psychology 107, 99–115.
Ramsay, D.S., Seeley, R.J., Bolles, R.C. and Woods, S.C. (1996) Ingestive homeostasis: the primacy of learning. In: Capaldi,
E.D. (ed.) Why We Eat What We Eat: The Psychology of Eating. American Psychological Association, Washington, DC,
pp. 11–30.
Reijgwart, M.L., Vinke, C.M., Hendriksen, C.F.M., van der Meer, M., Schoemaker, N.J. and van Zeeland, Y.R.A. (2016)
Ferrets’ (Mustela putorius furo) enrichment priorities and preferences as determined in a seven-chamber consumer
demand study. Applied Animal Behaviour Science 180, 114–121.
Ringgenberg, N., Fröhlich, E.K.F., Harlander-Matauschek, A., Toscano, M.J., Würbel, H. and Roth, B.A. (2015) Effects of
variation in nest curtain design on pre-laying behaviour of domestic hens. Applied Animal Behaviour Science 170, 34–
43.
Rolls, E.T. (2014) Emotion and Decision-making Explained. Oxford University Press, Oxford.
Russell, J.A. (2003) Core affect and the psychological construction of emotion. Psychological Review 110, 145–172.
Saker, P., Farrell, M.J., Adib, F.R., Egan, G.F., McKinley, M.J. and Denton, D.A. (2014) Regional brain responses associated
with drinking water during thirst and after its satiation. Proceedings of the National Academy of Sciences USA 111,
5379–5384.
Salamone, J.D., Yohn, S.E., López-Cruz, L., San Miguel, N. and Correa, M. (2016) Activational and effort-related aspects of
motivation: neural mechanisms and implications for psychopathology. Brain 139, 1325–1347.
Schütz, K.E., Cox, N.R. and Matthews, L.R. (2008) How important is shade to dairy cattle? Choice between shade or lying
following different levels of lying deprivation. Applied Animal Behaviour Science 114, 307–318.
Sescousse, G., Li, Y. and Dreher, J.C. (2015) A common currency for the computation of motivational values in the human
striatum. Social Cognitive and Affective Neuroscience 10, 467–473.
Sinha, R. and Jastreboff, A.M. (2013) Stress as a common risk factor for obesity and addiction. Biological Psychiatry 73,
827–835.
Toates, F.M. (1986) Motivational Systems. Cambridge University Press, Cambridge.

van der Harst, J.E., Baars, A.M. and Spruijt, B.M. (2003) Standard housed rats are more sensitive to rewards than enriched
housed rats as reflected by their anticipatory behaviour. Behavioural Brain Research 142, 151–156.
Vinke, C.M., Godijn, L.M. and Van der Leij, W.J.R. (2014) Will a hiding box provide stress reduction for shelter cats?
Warburton, H. and Mason, G. (2003) Is out of sight, out of mind? The effects of resource cues on motivation in mink,
Mustela vison. Animal Behaviour 65, 755–762.
Westerath, H.S., Gygax, L. and Hillmann, E. (2014) Are special feed and being brushed judged as positive by calves?
Wiedenmayer, C. (1997) Causation of the ontogenetic development of stereotypic digging in gerbils. Animal Behaviour 53,
461–470.
Wiepkema, P.R. (1971) Positive feedback at work during feeding. Behaviour 39, 2–4.
Yun, J., Swan, K.M., Farmer, C., Oliviero, C., Peltoniemi, O. and Valros, A. (2014) Prepartum nest-building has an impact
on postpartum nursing performance and maternal behaviour in early lactating sows. Applied Animal Behaviour Science
160, 31–37.
5 Learning and Cognition
M. Mendl and C.J. Nicol
5.1 Introduction
A young sow approaches a site on the boundary between forest and open field where she
found some tasty roots the previous day. She is poised to start rooting the ground when she
notices one of her group-mates, a large dominant female, approaching. Instead of digging for
the roots, she turns away and continues along the forest boundary. Close by, a field
ethologist, having studied the behaviour of this group of sows for several weeks, scribbles in
his notepad. The young sow appears to have remembered where the good food source was,
something that he has observed previously, but why didn’t she feed this time? Is it possible
that she didn’t want to reveal the location of the source to the approaching dominant sow? If
so, wouldn’t this suggest that she had some understanding of what would happen if she gave
away her secrets?
This is not just a fanciful example; pigs show intriguing behaviour like this in controlled
experimental studies (e.g. Held et al., 2002). But it raises a number of important questions
about the processes underlying the behaviour that we observe. How does the sow remember
where food was? Does she form an association between a particular visual stimulus, a clump
of trees say, and food? Or does she have a complete mental representation of the area? Why
does she not feed? Assuming that she recognizes her dominant group-mate, does she
associate the dominant with previous painful skirmishes over resources and simply avoid
her? Or does she actually understand what will happen if she started to feed, and perhaps
even attribute food-pilfering intentions to the dominant sow?
These questions reveal that apparently ‘clever’ behaviour can be explained by a variety
of putative underlying mental or cognitive processes. These range from associative learning
through to the capacity to form ‘cognitive maps’ or to have a ‘theory of mind’, abilities which
involve complex mental representation of the outside world. Distinguishing between these
explanations for observed behaviour is often not easy and is a theme that we will return to.
However, our principal aims in this chapter are to introduce aspects of cognition and learning
theory, and to consider the evidence for a number of different types of cognitive ability in
domestic animals. Work in this area has progressed considerably since the previous edition.
Domestic animals (especially dogs and horses) now feature often in studies of animal
cognition, partly because their long-term shared history with humans is of interest in itself.
This is a welcome development in a field previously dominated by studies of laboratory rats,
pigeons and non-human primates.
Before we go any further, what exactly do we mean by ‘cognition’? Historically, the
behaviourist school of psychology that dominated the study of animal learning for much of
the 20th century, through the work of J.B. Watson, B.F. Skinner and others, sought to explain
all behaviour without referring to unobserved mental processes. Indeed, it denied that it was
possible to study these at all. However, the cognitive revolution of the latter part of the 20th
century has re-legitimized the study of hidden mental processes through careful inference
from observed behaviour. Consequently, for some authors cognition refers to these mental
processes, and the study of cognition seeks to understand how information is represented and
manipulated in the mind, in stark contrast to the behaviourist approach. Others, however, use
the term in a broader sense to refer to all processes by which ‘animals acquire, process, store
and act on information from the environment’ (Shettleworth, 1998). In this chapter, we
discuss a number of examples of animal cognition (sensu Shettleworth) but, where relevant,
also consider the extent to which they can be explained by behaviourist and cognitivist
theories. We should also emphasize that when we refer to ‘mental representations’, we do not
imply that these are consciously experienced by non-human animals in the way that we
humans experience a thought or a feeling. This may be the case, but at present we do not
know. It is important to remember that the question of whether animals are consciously aware
of their mental processes is distinct from the question of how these processes work. The
study of cognition is concerned with the latter question only.
Learning and cognitive abilities should enhance the inclusive fitness of an individual,
increasing its chances of contributing genetically to the next generation. Different aspects of
learning will be of adaptive benefit to different species, making it very difficult to say
whether one species is ‘more intelligent’ than another, although the capacity to learn will
certainly be influenced by environmental and social complexity. Animals born into constant
environments can survive well using innate responses, but when environments are varied and
unpredictable, a wide range of learning abilities will be essential. The ancestors of our most
common domestic mammals and birds will have faced many uncertainties, including food of
variable quality or unpredictable distribution, predators with different appearances, location
and habits, and complex social environments where the identities and roles of group
members vary over time. This varied heritage means that domestic animals should be easily
able to establish relationships between events and thus form associations that guide
behavioural change. We start by considering the associative learning processes underlying
these abilities. We then consider memory processes, before discussing discrimination,
generalization and category formation, social cognitive abilities, the relationship between
cognition and emotion, complex cognitive abilities present in early life, and metacognition.
We conclude the chapter by briefly discussing how an understanding of animal cognition
influences our attitudes towards and understanding of animal welfare.
5.2 Associative Learning: Pavlovian and Instrumental

Conditioning
Two main types of associative learning have been described and extensively studied. In
classical, or Pavlovian, conditioning, an environmental event or stimulus is followed
predictably by some other occurrence. Pavlov studied the effects of sounding a bell just prior
to food arrival on salivation responses in dogs. The dogs naturally salivated to the smell or
appearance of food, so salivation was described as an unconditioned response (UR) and food
as an unconditioned stimulus (US). However, after repeated pairings of the bell with the
food, the dogs salivated in response to the sound, i.e. the conditioned stimulus (CS). When an
association is classically conditioned animals do not acquire new responses or behaviours, so
the conditioned response (CR) takes the same form as the UR. Such associations are acquired
with ease. Companion dogs, for example, can become expert at detecting the subtlest signs
that might predict their daily walk, without any formal training. The sight of a human picking
up a lead (the CS) can set off a volley of anticipatory responses such as barking and leaping.
Classical conditioning therefore allows animals to predict events, but it gives them little
control or influence over the actual timing, quality or duration of their meals or walks.
Instrumental, or operant, conditioning is a second type of associative learning, whereby
an animal directs a behaviour to a new part of its environment or learns to perform a new
behaviour, to obtain a reward. Imagine, for example, a captive zoo-housed coati (Nasua
nasua), presented for the first time with a cylindrical environmental enrichment device
containing small food pellets that fall through holes only when the device is manipulated and
moved in a certain way. The coati might initially approach the cylinder and perform a range
of natural exploratory movements (URs). But if one particular movement results in the
acquisition of a food pellet (US), the coati will repeat the last few movements it made, until it
learns precisely which action (now a CR) reliably produces food. In this case, the animal
gains not only the ability to predict what might happen next, but also the ability to control the
timing and delivery of its own reward.
An influential model providing insight into the process of association formation was
outlined by Rescorla and Wagner in 1972 (reviewed by Hall, 1994). This model describes
many features of associative learning, although it provides no insight as to the neurological
mechanisms involved. Repeated pairings of events result in associations with strengths that
reflect the exact statistical probability that a CS (e.g. a bell sounding) or a CR (e.g. an action
directed at a cylinder) will result in US arrival. The contiguity between two events is
therefore more important in establishing an association than the absolute temporal
relationship between them. Usually, if a stimulus or response is followed rapidly by US
arrival, a good association will be formed, because there is no time for intervening events to
interfere with the predictive relationship. However, under some conditions, an association
can be formed even if the CS precedes the US by many hours, provided the relationship
between the two is statistically strong. The best examples of delayed association formation
come from ecologically appropriate situations. In food aversion learning, for example, an
animal can ingest a novel (slightly toxic) substance and not feel sick for many hours. Once
nausea sets in, however, it will be able to associate the sickness with the novel substance it
ingested, rather than with any of the sights or sounds that it encountered in the intervening
period. The novel food is biologically the most plausible predictor of sickness.
The Rescorla–Wagner model has a number of other attributes that have been valuable in
explaining empirical data. It provides a framework that integrates the animal’s expectations
about the world with its potential to learn. Prior to association formation the presence of a CS
may be registered but it will not convey any meaning and so the animal will therefore be
surprised by the subsequent arrival of the US. After the association is fully formed, the
arrival of the US is not in the least bit surprising to the animal, indeed it is utterly predictable
and expected. Learning is thus intricately bound up with expectations and further learning
occurs most readily when contingencies change and the animal’s expectations about the
world no longer hold true. A consequence of this is that the presence of existing associations
(e.g. the bell that predicts food) can ‘interfere’ with the formation of new associations (e.g.
that a light also predicts food). The bell is a perfect predictor so the animal is never surprised
by the arrival of food and can never learn its association with light. Various processes of this
type, such as overshadowing or blocking (Dickinson, 1980), can inhibit further learning.
5.3 Associative Learning: Reinforcement, Punishment,

Rewards and Punishers
The chance of strong associative learning is greatest if the size or importance of the US to the
animal is high. Indeed, associative learning takes place only when a US is biologically
meaningful; something the animal either wants, or wants to avoid. Such stimuli are often said
to reinforce (increase performance of) the behaviour under consideration, or punish (decrease
performance of) the behaviour. Positive reinforcement occurs when provision of a desired US
(e.g. food) causes the animal to increase performance of the behaviour of interest, while
negative reinforcement occurs when removal of an undesired US increases performance of
that behaviour. In horse training, negative reinforcement is used to encourage a horse to
move forwards or sideways to avoid the rider’s leg pressure. In standardized tests, Icelandic
horses learnt rapidly within ten trials on one day to step away from pressure applied by an
algometer (an instrument that applies pressure at an increasing force until a response is
obtained). The pressure required to elicit a lateral movement decreased from approximately
30 N during the first couple of trials to approximately 15 N by the end of the first day
(Ahrendt et al., 2015).
Punishment, on the other hand, occurs when presentation or removal of a US decreases
behaviour. Positive punishment occurs when presentation of an undesired US decreases the
behaviour under consideration. For example, if a dog is shouted at (undesired US) whenever
it barks, this may lead to a reduction in barking. Conversely, negative punishment occurs
when removal of a desired US decreases the targeted behaviour. For example, if friendly
contact with a dog is withheld whenever it is lying on the sofa, this may decrease sofa-
hogging behaviour. Positive and negative thus refer, respectively, to the presentation or
removal of USs. This terminology can be confusing, so we provide a simple summary table
in Fig. 5.1.
Fig. 5.1. Terminology used to describe the different effects of stimuli on behaviour that
underlies associative learning. Note that these terms refer specifically to changes in the
behaviour of interest.
To add to the potential for confusion, USs are sometimes referred to as rewards or
punishers. It is helpful to think of rewards as things that animals want and will work for, and
punishers as aversive stimuli that they don’t want and will work to avoid (Rolls, 2005). It is
impossible to predict fully which stimuli or situations will be rewarding or aversive for any
given individual; while one dog may decrease barking when shouted at, another may like its
owner shouting and bark more in response. For this second dog, a yelling owner is a reward
and not a punisher, whatever the intentions of the owner, and the result is positive
reinforcement of barking!
Establishing which stimuli or resources are rewarding or aversive to domestic animals is
an important goal in animal welfare science, driven by the assumption that animals will
suffer if important rewards are missing from their environment or aversive stimuli are present
(Dawkins, 1990). Appropriate social contact (horses; Sondergaard et al., 2011), comfortable
bedding and the opportunity to rest (cows; Jensen et al., 2005), and the opportunity for a
good scratch (cows; Mandel et al., 2013) have all been shown experimentally to be
rewarding. For some individuals rewards may be cognitively complex, almost abstract. Dogs
that have been trained to give their paw to a human handler to obtain a food reward will
continue to do this behaviour even when food is omitted. However, they will stop if their
food reward is withheld and a partner dog continues to receive food for the same behaviour.
The overall rewarding value of giving a paw appears to be greatly reduced in this social
context, suggesting a degree of sensitivity to the fairness of outcomes within a social group
(Range et al., 2009). This demonstrates that what is perceived as rewarding can alter in very
sophisticated and subtle ways in animals with highly developed social sensibilities.
Complex rewards such as social contact or, indeed, social justice may be important in
guiding learned behaviour, but they are difficult or impossible to use in many practical
training situations. In some situations, it may even be impossible to deliver a food reward at
the appropriate moment. Because of this, secondary reinforcers, initially neutral but paired
with a US, are often used when training animals to perform a particular behaviour. A good
example is the training of giant African pouched rats to detect landmines. The rats are trained
to scratch the ground when they have detected traces of the explosive 2,4,6-trinitrotoluene
(TNT) below the ground. Because they do not naturally scratch at the ground in response to
the smell of TNT, this response has to be carefully elicited. Initially, they are trained that a
click (secondary reinforcer) reliably precedes a food reward (banana). Rats are then given
extensive odour discrimination training using a range of different substances placed beneath
sniffer holes, and are rewarded with a click followed by banana only when they sniff at TNT.
Once a rat is accurately discriminating TNT scent, it is trained to search an area of sand or
loose earth and rewarded by a click (and a later food reward) whenever it scratches to
indicate hidden TNT. Training targets are hidden at progressively sparser intervals and
greater depths until the rat is ready to start work in the field. Rats trained like this show very
high detection rates coupled with low false positives. In the dangerous real-world field
situation, it is impossible to know whether the rat has accurately indicated the presence of a
landmine or whether it has made a mistake. Rewarding the rat for a mistake will reduce
future accuracy, but omitting rewards may lead to extinction. For this reason, the rats may be
intermittently rewarded for detecting bags of TNT hidden below the soil, even during field
detection (Mahoney et al., 2014).
The use of secondary reinforcers is essential during a procedure such as landmine
detection where it would be impossible to throw pieces of banana to a rat working at some
distance from the handler. However, experimental studies have not always shown that
secondary reinforcers promote faster learning or greater retention. Williams et al. (2004), for
example, found that secondary reinforcement did not improve learning acquisition, or
resistance to extinction, in horses.
5.4 Associative Learning: What is Learned?

Studies of associative learning were the bedrock of behaviourist research and, in many cases,
it was possible to explain changes in behaviour in terms of stimulus–response connections
that required no reference to unobserved mental processes. However, this was not always so.
For example, rats exposed to the coincidence of two biologically neutral stimuli – ‘tone
predicts light’ – showed no change in behaviour, indicating to a behaviourist that nothing had
been learnt. However, if the light then became predictive of electric shock, the rats
subsequently showed avoidance behaviour to the tone as well. This could be explained by
postulating that the rats formed mental representations of the relationships between tone,
light and shock, and were thus able to combine the information ‘tone predicts light’ and ‘light
predicts shock’ (see Dickinson, 1980). A cognitivist perspective therefore accounted for the
observed behaviour.
The cognitivist view further postulates that information acquired during associative
learning can be stored in two different ways. Declarative representations, as in the above
example, involve knowledge about things or relationships (e.g. the light predicts shock) that
allow different representations to be combined in different ways and confer flexibility in how
the animal uses the information to guide a behavioural response (Dickinson, 1980).
Procedural representations, on the other hand, involve knowledge about what to do (e.g.
when the light is on, crouch), as in a simple stimulus–response connection, and can only be
used in an inflexible way. Related to this distinction is whether learned responses are goal-
directed or habitual. Goal-directed behaviour can be likened to the animal having a
representation or model of the various behaviours that it can perform and their resulting
outcomes. If the value of an outcome is altered in some way (e.g. a previously palatable food
reward unexpectedly makes the animal sick), the animal updates its model and, when offered
the opportunity to perform a behaviour to acquire the food once more, it suppresses this
behaviour in a flexible and adaptable way. However, if the behaviour has become habitual
and ‘model-free’, for example due to over-training, devaluation of an outcome cannot be
incorporated into a decision model and the animal continues to perform the food-acquisition
behaviour in an inflexible way. Habitual behaviour does alter slowly with repeated
experience of the changed outcome, and is computationally cheaper than goal-directed
model-based learning, and hence advantageous in this respect (Dolan and Dayan, 2013).
Studies of chickens suggest that they are able to incorporate new information on the value of
food into foraging decisions, indicating flexible goal-directed behaviour and declarative
representations of information (Forkman, 2001).
5.5 Memory
From a cognitivist perspective, associative learning involves changes to representations of
information, and the storage and retention of these – memory. Studies of learning and
memory thus overlap, and to a certain extent the two phenomena are inseparable. However,
memory research can be characterized by the types of question that it addresses, which
include how information is acquired, stored and retrieved and what factors affect this, how
long information can be retained and what type of information is stored. A number of
different types of ‘memory system’ have been identified. In animal studies, the terms
working memory and short-term memory are often used interchangeably to describe the
storage of information over a few minutes to hours, as in when an animal is foraging for food
and uses its memory to avoid revisiting locations that it has just searched. Long-term or
reference memory holds information for much longer periods (days, months, years) and
seems to have a virtually unlimited capacity. It is associated with molecular and cellular
events, including protein synthesis, in areas of the brain such as the hippocampus, amygdala
and medial temporal lobes. Although far from being fully understood, these can be
conceptualized as the laying down of a ‘memory trace’.
Working, short- and long-term memory abilities have been demonstrated in a variety of
domestic species, often using ecologically relevant spatial tasks such as remembering the
location of food. For example, cattle develop good long-term memory of where high- or low-
quality food is in a radial-arm maze, showing a preference for high-quality sites and avoiding
low-quality ones. In the case of high-quality food, this memory may persist for at least 30
days (Bailey and Sims, 1998). Cows can also remember how to traverse a maze for at least 6
weeks (Hirata et al., 2016), and so can sheep when tested over a 40-day period (Lee et al.,
2006). In terms of short-term memory, dogs are able to remember the location of food buried
in a sand arena over a 90 min interval, with younger dogs (1–4 years) performing
significantly better than those over 8 years old (Salvin et al., 2011). Pigs can be readily
trained to search an arena each day for food in a novel location (Fig. 5.2), retain that memory
over the next 1–2 h and revisit the same (re-baited) location when allowed back into the
arena (Mendl et al., 1997). They can also remember the location of two different-sized food
sources and preferentially relocate the larger one first when foraging on their own or with a
non-competing pen-mate, but not when foraging with a competitor who follows them around
and displaces them from it. It is conceivable that this is an intentional attempt to misdirect the
competitor but, if so, they might be expected to lead the competitor to the smaller source
first, and this was not observed. More likely, the pigs learn that they will be displaced from
their preferred food source and hence decrease their propensity to visit it first (Held et al.,
2010). Although we do not know the underlying cognitive mechanisms, it appears that pigs
can store relatively complex spatial information in memory and use it in a flexible way.
These examples of spatial memory probably involve associative learning processes that link
visual, olfactory or other spatial cues with food reward. Whether these species can develop
‘cognitive maps’, mental representations of the spatial environment which allow them to
make novel shortcuts, remains to be elucidated.
Fig. 5.2. Pigs in a competitive foraging task. Pigs are trained and tested daily in a foraging
arena. (a) On the first trial of the day a pig enters the arena to search for food placed
randomly in one of eight buckets. (b) It finds the food and eats it. It has previously been
trained that, on the next trial a few minutes later, food will be in the same location as on the
first trial. (c) The ‘informed’ pig is now introduced with a heavier companion who has been
trained that there is food in the arena, but does not know where it is. (d) If the companion
follows the informed pig to the correct bucket, it is able to displace it and steal the food. Over
many days it is possible to observe how the behaviour of the two pigs changes as first the
companion learns to follow the informed pig to the food, and then the informed pig develops
behaviours which minimize the chances of the companion pig stealing its food (see Held et
al., 2002). Furthermore, there is evidence that pigs can learn and remember the location of
differently sized food sources and preferentially relocate the larger one first when foraging on
their own or with a non-competitor, but not when a competitor who can displace them from
the food is foraging with them (Held et al., 2010).
Another ecologically relevant ability, namely memory for other individuals, or at least for
cues from those individuals such as facial appearance, has also been studied in domestic
species. One study demonstrated that sheep could discriminate between pairs of conspecific
faces, generalize from frontal views to profiles and perform accurately on these
discriminations after a gap of 2 years, although a decline in performance may have indicated
that they were mainly remembering the task per se and could have performed equally well on
new face pairs (this was not reported). Neurophysiological evidence indicated that cells in the
temporal and medial prefrontal cortex of the brain that fire when sheep view a familiar
conspecific continued to fire in response to specific facial images even after 8–12 months in
which that individual had not been encountered (Kendrick et al., 2007). Another study
showed that female goats remembered the vocalizations of their kids for a year or more after
weaning and permanent separation, even when they had given birth to a new kid in the
intervening period (Briefer et al., 2012).
The factors that affect the selection, storage and retrieval of memory are critically
important in determining what is remembered and how well (Fig. 5.3), but have received less
attention in domestic species. There is increasing evidence that, just as only biologically
meaningful unconditioned stimuli trigger associative learning, the events that are prioritized
for storage in memory are also ‘important’ ones, those which are most likely to impinge on
survival and reproductive success. A study of pigs indicated that information that was more
costly to forget, in this case because forgetting it increased the time taken to obtain food, was
indeed encoded more effectively in memory (Laughlin and Mendl, 2004). Information stored
in memory may also ‘decay’ with time in that it is less readily retrieved as time passes.
Retroactive interference, which occurs when new information interferes with the stored
representation of similar information, may contribute to this memory decay. For example,
pigs trained to remember the location of food in an arena appeared to have their memory
disrupted if exposed to the arena again (without being allowed to search it) prior to being
tested for memory retrieval (Mendl et al., 2001). One explanation is that, when the animal
encounters a relevant cue, the memory trace of the original information is ‘reactivated’ from
a ‘dormant state’ into a ‘labile’ one that is vulnerable to alteration by new information before
being stored again in a dormant form. It is possible to investigate this idea because the
(re)storage of longer-term memories requires protein synthesis in specific brain areas which
can be temporarily stopped using protein-synthesis blockers. Perrin et al. (2007) used these
techniques to show that a ewe’s memory trace of her lamb does indeed appear to be
reactivated by cues from the lamb, and can then be disrupted such that her subsequent ability
to recognize the lamb deteriorates. These and other findings emphasize the complexity of
memory processes (Fig. 5.3) and help to explain how memories can be interfered with, decay,
and become embellished with time.
Fig. 5.3. Schematic representation of some of the events that are thought to occur when
information is acquired, stored and retrieved from memory. A stimulus or event is perceived
and a representation of this is consolidated into long-term memory as a neural ‘memory
trace’. This involves molecular processes including protein synthesis in cells, and can be
facilitated by moderate elevations in ‘stress hormones’ ensuring that important or emotionally
salient events are given priority in memory. When a cue related to the original stimulus
and/or a mood state associated with the original stimulus is encountered, the ‘dormant’
memory trace appears to be reactivated, as evidenced by behaviour indicating that the
memory has been retrieved. New information present at this time may then be incorporated
into a newly consolidated memory trace which may thus be slightly, or even greatly, altered
from the original memory.
Exactly what is held in memory remains mysterious. Whether animals have vivid
conscious recall of events that happened to them in the past, termed ‘episodic memory’ in
humans, remains unknown and perhaps unknowable. However, there is increasing evidence
that some species may be able to remember not just what happened where, but also when it
happened. ‘What–where–when’ (www) memory has been likened to human episodic memory
and investigated in a variety of species following seminal research by Clayton and Dickinson
(1998) on scrub jays. A study of mini-pigs indicated that they could remember what objects
they had previously encountered where and in which context, by showing more interest in the
one out of two objects that was presented in a new context–location combination in a test
trial. The authors argued that an ability to combine what, where and which information,
similar to www memory, was needed to solve the task (Kouwenberg et al., 2009). Episodic
memory processes may also play a role in future planning – the ability to prospect forwards
in time and even to imagine future scenarios. A study of two border collies suggested that
one of the dogs was able to combine information about what object needed to be retrieved
(object name spoken by researcher) and from which location (one of two rooms in which the
dog had previously seen the objects) to accurately plan its search for that object (Kaminski et
al., 2008). These are interesting early findings but more studies are needed. And even if
strong evidence for www memory is found, we still do not know whether such memories are
consciously experienced.
5.6 Discrimination, Generalization, Categorization and

Concept Formation
Discrimination allows distinctions to be drawn between objects or stimuli that differ in
particular features. There appear to be natural biases in the ease with which different species
attend to and use different stimulus features for discrimination, which almost certainly
depend on adaptive ecological history. For example, in tests where both the appearance and
relative position of buckets are potential cues indicating food availability, horses
preferentially utilize the positional cues (Hothersall et al., 2010; see Fig. 5.4).
Fig. 5.4. Foals can learn to select a bucket containing food based either on its relative
location within an array or on visual cues such as the black markings shown here, but they
learn to use relative location as a discriminative cue far more rapidly than they learn to use
visual cues. In addition, foals that have learnt a visual-cue discrimination can readily relearn
to use relative location instead, but foals that have first acquired a relative-location
discrimination appear unable to relearn a visual-cue discrimination (Hothersall et al., 2010).
In these experiments olfactory cues are always evenly distributed across all buckets, and the
array can be placed anywhere within a stable or outdoor enclosure to exclude the use of
global position cues.
Discrimination also enables animals to fine-tune their behaviour in response to

environmental cues. A dog may learn that barking results in attention from the owner when
the TV is off, but not when the TV is on. The status of the TV therefore controls the
behaviour of dog, much as a laboratory rat might discriminate by performing an operant
response in the presence of one stimulus but not another. Generalization is the converse
tendency to attend to shared features within a range of stimuli. Generalization enables
animals to respond to new stimuli adaptively, provided sufficient shared information exists
with previously encountered stimuli. The adaptive balance between discrimination and
generalization depends on precise circumstances. A riding-school pony, for example, has to
respond appropriately to a variety of riders using leg, seat and hand cues in different ways.
The adaptive response of the pony is to generalize and respond in the same way to a broad
range of different cues. For an advanced dressage horse, such generalization would be
disastrous: dressage riders aim for precision and consistency in their aids in order to foster
discrimination and avoid generalization.
If the shared features of a class of stimuli are based on physical similarities such as
colour, size or shape, then the processes of discrimination and generalization can result in
category formation. Horses can categorize trained and novel images according to whether
they have open centres or are solidly black, when procedures are used that control for
confounding variables such as the overall areas of black and white in the stimuli (Hanggi,
1999). Categorizing stimuli according to abstract principles such as ‘bigger than’ or ‘paler
than’ is cognitively more demanding. Well-designed experiments have demonstrated that
primates, parrots and pigeons can form such relational categories, but there is limited
information about whether domestic animals can do so. It is salutary to note that honeybees
can use relational categories of this type as a cue to locate food (Avargues-Weber et al.,
2014), suggesting that we should be cautious in making any assumption that mammals are
especially cognitively complex or in assuming that categorization is especially cognitively
demanding. The only relational category that has been studied to any real extent in domestic
animals is the ‘same/different’ category. This can be examined using matching-to-sample or
non-matching-to-sample tests, where an animal capable of this degree of abstraction should
be able to select a stimulus that is the ‘same as’ or ‘different from’ one previously viewed.
Same/different categorizations have been difficult to demonstrate using some paradigms but,
with improved methodologies, the ability to categorize geometric shapes as the ‘same’ or
‘different’ has been shown convincingly in Shetland ponies. Four of seven subjects were
trained to select the ‘same’ stimulus as a preceding sample, using a training set of circles and
crosses. All four were then able to transfer this categorization to a novel set of triangular and
rectangular stimuli (Gabor and Gerken, 2012).
An entirely abstract category could be called a ‘concept’ or an ‘idea of a class of objects’
where stimuli are included on the basis that they stand for the same idea, and not that they
resemble each other physically in any way. Concept formation, especially in animals that
have no language, is of great theoretical interest; however, few studies have been conducted
with domestic animals and developing viable methodologies remains a challenge (Lea et al.,
2006).
5.7 Social Cognition

For many species, the social environment is highly complex and changeable, and some of the
most interesting examples of cognitive abilities arise in this context. Notably, social learning
facilitates the acquisition of new behaviours in many domestic animals. Social learning
occurs when a naïve animal (the observer) acquires information from a knowledgeable
conspecific (the demonstrator), resulting in faster or more effective learning by the observer,
often with greatly reduced costs. Sometimes, the demonstrator simply draws the observer’s
attention to a previously unnoticed stimulus, and subsequent acquisition of a new response by
the observer behaviour then takes place by normal instrumental conditioning. But social
learning can also occur via processes that are cognitively more demanding, including
imitation, where an animal copies the physical movements of a demonstrator, and emulation,
where it reproduces the results of a demonstrator’s actions (Whiten et al., 2004).
There is good evidence that most domestic animals employ social learning in some
situations, but the processes involved have perhaps been most studied in chickens (reviewed
by Nicol, 2004) where, in early life, chicks are sensitive to social guidance about which foods
they should and should not ingest. Hens attract their young to food with a complex display of
staccato food calls and pecks to the ground. The hens’ display is more intense in the presence
of high-quality food items, if the chicks move away or if the chicks make apparent ‘errors’ in
the objects they peck at. Older chickens will learn to eat, but not to avoid, novel coloured
food after watching a knowledgeable demonstrator. The relationship between the observer
and demonstrator birds is very important, as hens are more likely to acquire new behaviours
after watching socially dominant demonstrators than social subordinates.
As well as transmitting information about food availability or quality, animals also
communicate with each other about predator presence and identity. Communication can be
both subtle and flexible as demonstrated by the alarm-calling behaviour of cockerels, who
give different types of call in response to perceived aerial or ground predators, and modulate
their calling depending on the type of ‘audience’ that may be listening. Cockerels are most
keen to impress novel female chickens and call more in their presence than in the presence of
familiar females or novel birds of a different species. In addition, both male and female
(broody) bantam chickens adjust their alarm calling according to the size of the aerial
predator relative to their own growing chicks. Small hawks invoke parental alarm calls only
when chicks are very young and vulnerable (Palleroni et al., 2005).
Complex and flexible communication does not, by itself, mean that domestic animals
possess full language abilities. There is no evidence, for example, that individuals of any
domestic species can rearrange discrete symbols (words, or a visual equivalent) to generate
new meanings. That said, the closer we look, the more difficult it is to draw really clear
distinctions between human language abilities and the capabilities of dogs. Not only has one
study highlighted a dog that could accurately use human verbal labels to distinguish between
hundreds of different toys (Pilley and Reid, 2011), but another recent study has found that
dogs, like humans, use the left brain hemisphere to process word meaning but the right
hemisphere to process the intonation of human speech (Andics et al., 2016).
Perhaps the most advanced form of social cognition is the ability to ‘put oneself in
another’s shoes’; for example, to take the visual perspective of another individual, to
understand that this individual has a mental state like oneself and even to use this knowledge
to deceive them. Not surprisingly, it is very difficult to be sure that any behavioural evidence
for these abilities cannot be explained by ‘simpler’ associative learning processes.
Primatologists research and argue about this area, but there have been only a few studies in
domestic animals. As we have seen, pigs can show behaviour which allows them to minimize
exploitation of their knowledge of where food is in a foraging arena by another individual
(Fig. 5.2), but it is probable that this ‘clever’ behaviour is the result of the knowledgeable
animal learning that it is likely to lose out to its larger companion and thus avoiding
competitive encounters (Held et al., 2002).
A technique designed by primatologists to reveal more precisely what one animal
understands about another’s knowledge has been adapted for pigs and dogs. Here the subject
animal chooses to use one of two individuals as a guide to where food is. The subject is
unable to see where food is being placed in an arena, but can see that one of the individuals
(the ‘knower’) has visual access to this baiting event while the other (the ‘guesser’) does not.
The question is whether the subject then uses information provided by the ‘knower’ to locate
food, which would indicate some understanding of the relative knowledge of the two
individuals. In a carefully controlled study, one out of ten pigs made the ‘correct’ choice to
follow ‘knower’ rather than ‘guesser’ pigs to a food source (Held et al., 2001). In a similar
study using human ‘guessers’ and ‘knowers’, dogs preferred to use information provided by
the ‘knower’ as would be predicted if they were able to take their visual perspective
(Maginnity and Grace, 2014). Despite well-designed experiments, it is still difficult to rule
out that animals showing the correct response were doing so on the basis of previously
learned associations rather than by taking the visual perspective of the other individual. For
example, they may have learnt that another’s gaze direction or visual access to an event is a
good predictor of subsequent behaviour. Dogs at least are highly sensitive to gestures and
even the glancing behaviour of humans.
5.8 Cognition and Emotion

A growing area of interest is the interaction between cognition and emotion in areas such as
the development of new indicators of welfare and the study of empathy. Drawing on the
terminology introduced earlier, we can operationally define emotions as states induced by
rewards or punishers, where rewards are things for which animals will work and punishers
are things they will work to avoid (Rolls, 2005). Defining emotions in this way allows us to
objectively identify positively and negatively valenced states in other species, while
remaining agnostic as to whether these are consciously experienced or not – an issue which,
as we have seen, is currently impossible to resolve. But if we assume that emotions are
consciously experienced in at least mammalian and bird species (e.g. The Cambridge
Declaration on Consciousness; available at:
http://fcmconference.org/img/CambridgeDeclarationOnConsciousness.pdf (accessed 10 April
2017)), then being able to accurately measure the valence of these states is clearly of
paramount importance from an animal welfare perspective.
Assessment of emotional valence is challenging – for example, traditional indicators of
‘stress’ such as heart rate and glucocorticoid hormones may be elevated in individuals
experiencing both negative (e.g. fear) and also positive (e.g. excitement) states. Recently,
there has been interest in the possibility that changes in cognitive function reliably reflect
emotional valence and hence can be used as proxy indicators of these states. For example,
people in a negative affective state are more likely to interpret an ambiguous event negatively
and to anticipate negative rather than positive things happening (Paul et al., 2005). These
emotion-related ‘cognitive biases’ have been explored in animals using discrimination
learning protocols in which subjects are trained to make response A to stimulus X in order to
get a good thing (e.g. food), and response B to stimulus Y (in the same sensory modality as
X) in order to get a less good thing (e.g. less food) or avoid a bad thing (e.g. noise). Once
trained, the subjects are presented with ambiguous stimuli (intermediate between X and Y) to
see whether they perform response A, indicating that they categorize the stimulus as
predicting a good thing (‘optimistic’), or response B indicating a ‘pessimistic’ judgement of
the stimulus (Fig. 5.5). The prediction is that subjects in a putative negative affective state are
more likely to show ‘pessimistic’ responses (Harding et al., 2004). This prediction is upheld
in the majority of over 90 published studies on a variety of domestic species, indicating that
cognitive bias is a promising indicator of affective valence in animals. However,
methodological issues (e.g. Roelofs et al., 2016) and null and opposing findings require
further investigation.
Fig. 5.5. One example of a procedure for a cognitive bias task. In this ‘judgement bias’ task,
the animal is trained that when training cue A (e.g. a tone of a specific frequency) is
presented, it must perform response P to get a good reward (e.g. several pellets of food),
while when training cue B (e.g. a tone of a different frequency) is presented it must perform
response N to get a less good reward (e.g. one pellet of food). Once trained on this task, it
can be assumed that performance of response P indicates anticipation of a good event, while
performance of response N indicates anticipation of a less good event. The animal is then
presented with ambiguous cues (tones of intermediate frequency) and its responses to these
are recorded. The prediction is that animals in a negative mood state are more likely to
categorize the ambiguous cues as predicting a relatively negative event and hence are more
likely to show response N. This ‘pessimistic’ judgement bias compares with negative
cognitive biases observed in people in negative mood states. Variants on this procedure
include making training cue B predictive of a negative event (e.g. noise) which can be
avoided by performing response N (e.g. Harding et al., 2004).
Another potential ‘cognitive indicator’ of emotion is anticipatory behaviour occurring in

Pavlovian conditioning during the period between the onset of a cue and the delivery of the
predicted reward. Spruijt et al. (2001) hypothesized that reward sensitivity alters according to
affective state, increasing when animals are in a short-term negative state and decreasing as
this state becomes chronic, and that anticipatory behaviour during Pavlovian conditioning is
positively correlated with reward sensitivity. Thus animals in a short-term negative state
would be predicted to show increased anticipatory behaviour. Studies of domestic species
have supported this prediction (van der Harst and Spruijt, 2007), although the biphasic nature
of the relationship between emotion and reward sensitivity makes interpretation potentially
difficult.
Interest has also been growing in the idea that lateralization of brain processes may
reflect emotional states. For example, data from humans suggest that the right hemisphere
dominates processing during negative situations or states and therefore that such states can be
detected by a bias to show contralateral, left-dominated, visual or auditory responses.
Although limited work has been done in this area, there are data, particularly from studies of
dogs (e.g. Siniscalchi et al., 2010), that provide some support for this hypothesis (see
Leliveld et al., 2013).
Having seen how cognitive changes can be used to interpret emotional states in animals
we turn briefly to a consideration of the role of cognition in generating complex social
behaviours such as empathy. In humans, a full empathic response may involve theory of
mind and an understanding of the situation or plight of another, but it also requires the
involvement of other underlying processes shared by a wider range of species (Edgar et al.,
2012). These different processes can be disentangled through careful study. For example, an
experiment was designed to determine what cues generated the apparently empathic
physiological and behavioural changes seen in mother hens when they considered their
chicks to be in danger of receiving a mildly aversive puff of air. Some chicks were pre-
trained that they were in a safe environment (thus providing no behavioural cues to the hen)
while others were pre-trained that this same environment was dangerous. The hens’
behavioural responses (increased vocalization, walking and decreased preening) were found
to depend solely on their own perception and knowledge of the situation, whereas certain
physiological responses (stress-induced hyperthermia) were activated only when this
knowledge concurred with that of the chicks (Edgar et al., 2013). Thus both behavioural cues
from chicks and an understanding of their situation contributed to the hen’s empathic
behaviour.
5.9 Complex Cognition in Early Life

Some surprisingly complex cognitive abilities appear to be present very early in life. For
example, newly hatched young chicks show an awareness that objects are separate entities
that continue to exist when out of sight of the observer. They will successfully make detours
around obstacles to rejoin a hidden object on which they have imprinted (Freire et al., 2004).
In human babies, this ‘object permanence’ awareness develops at about 8 months, and can be
thought of as a specific type of memory. Not only do very young chicks outperform human
babies in this regard, they also hatch with surprising numerical abilities and expectations
about the physical nature of their world. For example, they can distinguish sets of one versus
two, and two versus three, on the basis of number, in carefully controlled experiments in
which they cannot solve the problem by a simple rule concerning surface area (Rugani et al.,
2008). They can also keep track of simple additions and subtractions, with numbers up to
five or six, demonstrating a capacity for basic arithmetic (Rugani et al., 2009). The young
chick even has an innate ability in physics, showing significant preferences for realistic
diagrams of three-dimensional objects over non-realistic diagrams (Regolin et al., 2011).
5.10 Metacognition
An area of growing interest in comparative psychology is ‘metacognition’, partly because
this capacity to be aware of one’s own knowledge has been equated with conscious
experience. One way of testing for metacognition is to see whether animals actively search
for information in situations where they are uncertain, hence suggesting metacognitive
awareness of their lack of knowledge. There is evidence that children, apes and monkeys that
are unable to see a food-baiting event will actively seek further information about whether
and where it happened, for example by looking into tubes where the food may have been
hidden, before deciding to choose a tube or to opt for a less desired foodstuff in a known
location. A recent study has shown that dogs too will seek further information if they are
uncertain where a reward is hidden, but only by selecting a human who assists them, and not
by looking themselves (McMahon et al., 2010). However, the prediction that a preference for
an informative as opposed to an uninformative person should be stronger only when the dog
is uncertain has not been tested. These types of experimental protocols can provide
suggestive evidence about metacognitive abilities, but have rarely been applied to domestic
species.
5.11 Animal Cognition and Animal Welfare

The welfare of domestic animals may influence their cognitive abilities. For example, in
concordance with results from laboratory animal studies, raising pigs in an enriched
environment with straw, a rooting area and a changing variety of objects to explore results in
enhanced performance in a spatial learning task (Grimberg-Henrici et al., 2016). But we
should also consider the relationship between animal cognition and animal welfare from a
broader perspective.
There is growing interest in understanding the cognitive abilities of domestic animals.
Once we appreciate that they do not simply respond to their environments, to each other or to
us with a set of simple, fixed or ‘unthinking’ actions, we may want to rethink their place
within our ethical frameworks. We may admire and appreciate their complexity just as we
might admire great paintings or diverse and complex landscapes, and perhaps accord them
greater respect or protection on these grounds alone.
The cognitive abilities of animals can also have a direct impact on their welfare. For
example, behaviours such as cannibalism in chickens are facilitated by prior observation of
companions (Cloutier et al., 2002). Only by understanding the mechanisms of social learning
in large flocks might it be possible to reduce the transmission rates of these harmful
behaviours. The ability to measure or at least discriminate the passage of time could also be
used positively to improve welfare. Pigs can learn to choose a location where they will be
confined for a shorter period of time over one where confinement lasts longer (Spinka et al.,
1998). Such pigs could, theoretically, use cues designed to ‘reassure’ them that an upcoming
procedure will be over soon. In a different context, it is sometimes argued that animals
cannot be trusted to make rational or considered choices in preference tests, because they are
incapable of ‘weighing up’ the consequences. This concern may be misplaced, as empirical
studies of chickens’ preferences across sets of varied environments not only remain
consistent and stable across time, but are also rational in the mathematical sense of the word.
A chicken that prefers environment A over B, and B over C, is also statistically more likely
than chance to show a logical preference for A over C (Browne et al., 2010). This
demonstrates a stable coding of the rewarding properties of each environment over the long
term. This type of evidence gives us more confidence in the use of preference tests to assess
animal needs and desires.
Finally, although there is no compelling reason why any of the complex learning or
cognitive abilities already described should be accompanied by subjective experiences or
consciousness, it is usual to give domestic animals the benefit of the doubt and concede that
consciousness is logically possible, or even likely. If so, then an animal that has an episodic
memory of past, stressful events may have reduced welfare many days or weeks after the
event itself ended. It is even possible that animals may generate ‘innate’ mental
representations of things they have never encountered in their own lifetimes, so that an
animal kept in social isolation could still, conceivably, ‘miss’ the company of others. And an
animal that can prospect about future events may be able to anticipate its own future (Mendl
and Paul, 2008). Much work remains to be done.
References
Ahrendt, L.P., Labouriau, R., Malmkvist, J., Nicol, C.J. and Christensen, J.W. (2015) Development of a standard test to
assess negative reinforcement learning in horses. Applied Animal Behaviour Science 169, 38–42.
Andics, A., Gabor, A., Gacsi, M., Farago, T., Szabo, D. and Miklosi, A. (2016) Neural mechanisms for lexical processing in
dogs. Science 353, 1030–1032.
Avargues-Weber, A., d’Amaro, D., Metzler, M. and Dyer, A.G. (2014) Conceptualization of relative size by honeybees.
Frontiers in Behavioral Neuroscience 8, 80.
Bailey, D.W. and Sims, P.L. (1998) Association of food quality and locations by cattle. Journal of Range Management 51,
2–8.
Briefer, E.F., de la Torre, M.P. and McElligott, A.G. (2012) Mother goats do not forget their kids’ calls. Proceedings of the
Royal Society B: Biological Sciences 279, 3749–3755.
Browne, W.J., Caplen, G., Edgar, J., Wilson, L.R. and Nicol, C.J. (2010) Consistency, transitivity and inter-relationships
between measures of choice in environmental preference tests with chickens. Behavioural Processes 83, 72–78.
Clayton, N.S. and Dickinson, A. (1998) Episodic-like memory during cache recovery in scrub jays. Nature 395, 272–274.
Cloutier, S., Newberry, R.C., Honda, K. and Alldredge, J.R. (2002) Cannibalistic behaviour spread by social learning.
Dawkins, M.S. (1990) From an animal’s point of view: motivation, fitness and animal welfare. Behavioral and Brain
Sciences 13, 1–61.
Dickinson, A. (1980) Contemporary Animal Learning Theory. Cambridge University Press, Cambridge.
Dolan, R.J. and Dayan, P. (2013) Goals and habits in the brain. Neuron 80, 312–325.
Edgar, J.L., Nicol, C.J., Clark, C.C.A. and Paul, E.S. (2012) Measuring empathic responses in animals. Applied Animal
Edgar, J., Paul, E. and Nicol, C.J. (2013) Protective mother hens: cognitive influences on the avian maternal response.
Forkman, B. (2001) Domestic hens have declarative representations. Animal Cognition 3, 135–137.
Freire, R., Cheng, H.W. and Nicol, C.J. (2004) Development of spatial memory in occlusion experienced domestic chicks.
Gabor, V. and Gerken, M. (2012) Cognitive testing in horses using a computer based apparatus. Applied Animal Behaviour
Science 139, 242–250.
Grimberg-Henrici, C.G.E., Vermaak, P., Bolhuis, J.E., Nordquist, R.E. and van der Staay, F.J. (2016) Effects of
environmental enrichment on cognitive performance of pigs in a spatial holeboard discrimination task. Animal Cognition
19, 271–283.
Hall, G. (1994) Pavlovian conditioning. In: Mackintosh, N.J. (ed.) Animal Learning and Cognition. Academic Press, San
Diego, California, pp. 15–43.
Hanggi, E.B. (1999) Categorization learning in horses (Equus caballus). Journal of Comparative Psychology 113, 243–252.
Harding, E.J., Paul, E.S. and Mendl, M. (2004) Cognitive bias and affective state. Nature 427, 312.
Held, S., Mendl, M., Devereux, C. and Byrne, R.W. (2001) Behaviour of domestic pigs in a visual perspective taking task.
Behaviour 138, 1337–1354.
Held, S., Mendl, M., Devereux, C. and Byrne, R.W. (2002) Foraging pigs alter their behaviour in response to exploitation.
Held, S.D.E., Byrne, R.W., Jones, S., Murphy, E., Friel, M. and Mendl, M.T. (2010) Domestic pigs, Sus scrofa, adjust their
foraging behaviour to whom they are foraging with. Animal Behaviour 79, 857–862.
Hirata, M., Tomita, C. and Yamada, K. (2016) Use of a maze test to assess spatial learning and memory in cattle: can cattle
traverse a complex maze? Applied Animal Behaviour Science 180, 18–25.
Hothersall, B., Gale, E.V., Harris, P. and Nicol, C.J. (2010) Cue use by foals (Equus caballus) in a discrimination learning
task. Animal Cognition 13, 63–74.
Jensen, M.B., Pedersen, L.J. and Munksgaard, L. (2005) The effect of reward duration on demand functions for rest in dairy
heifers and lying requirements as measured by demand functions. Applied Animal Behaviour Science 90, 207–217.
Kaminski, J., Fischer, J. and Call, J. (2008) Prospective object search in dogs: mixed evidence for knowledge of what and
where. Animal Cognition 11, 367–371.
Kendrick, K.M., da Costa, A.P., Leigh, A.E., Hinton, M.R. and Peirce, J.W. (2007) Sheep don’t forget a face. Nature 447,
346.
Kouwenberg, A.-L., Walsh, C.J., Morgan, B.E. and Martin, G.M. (2009) Episodic-like memory in crossbred Yucatan
minipigs (Sus scrofa). Applied Animal Behaviour Science 117, 165–172.
Laughlin, K. and Mendl, M. (2004) Costs of acquiring and forgetting information affect spatial memory and its
susceptibility to interference. Animal Behaviour 68, 97–103.
Lea, S.E.G., Wills, A.J. and Ryan, C.M.E. (2006) Why are artificial polymorphous concepts so hard for birds to learn?
Quarterly Journal of Experimental Psychology 59, 251–267.
Lee, C., Colegate, S. and Fisher, A.D. (2006) Development of a maze test and its application to assess spatial learning and
memory in Merino sheep. Applied Animal Behaviour Science 96, 43–51.
Leliveld, L.M.C., Langbein, J. and Puppe, B. (2013) The emergence of emotional lateralization: evidence in non-human
vertebrates and implications for farm animals. Applied Animal Behaviour Science 145, 1–14.
Maginnity, M.E. and Grace, R.C. (2014) Visual perspective taking by dogs (Canis familiaris) in a guesser–knower task:
evidence for a canine theory of mind? Animal Cognition 17, 1375–1392.
Mahoney, A., Lalonde, K., Edwards, T., Cox, C., Weetjens, B. and Poling, A. (2014) Landmine-detection rats: an evaluation
of reinforcement procedures under simulated operational conditions. Journal of the Experimental Analysis of Behavior
101, 450–456.
Mandel, R., Whay, H.R., Nicol, C.J. and Klement, E. (2013) The effect of food location, heat load, and intrusive medical
procedures on brushing activity in dairy cows. Journal of Dairy Science 96, 6506–6513.
Manteuffel, G., Puppe, B. and Schon, P.C. (2004) Vocalisation of farm animals as a measure of welfare. Applied Animal
McMahon, S., Macpherson, K. and Roberts, W.A. (2010) Dogs choose a human informant: metacognition in canines.
Behavioural Processes 85, 293–298.
Mendl, M. and Paul, E.S. (2008) Do animals live in the present? Current evidence and implications for welfare. Applied
Animal Behaviour Science 113, 357–382.
Mendl, M., Laughlin, K. and Hitchcock, D. (1997) Pigs in space: spatial memory and its susceptibility to interference.
Mendl, M., Burman, O., Laughlin, K. and Paul, E. (2001) Animal memory and animal welfare. Animal Welfare 10(Suppl. 1),
S141–S159.
Nicol, C.J. (2004) Development, direction and damage limitation: social learning in domestic fowl. Learning and Behavior
32, 72–81.
Palleroni, A., Hauser, M. and Marler, P. (2005) Do responses of galliform birds vary adaptively with predator size? Animal
Cognition 8, 200–210.
Paul, E.S., Harding, E.J. and Mendl, M. (2005) Measuring emotional processes in animals: the utility of a cognitive
approach. Neuroscience and Biobehavioral Reviews 29, 469–491.
Perrin, G., Ferreira, G., Meurisse, M., Verdin, S., Mouly, A.M. and Levy, F. (2007) Social recognition memory requires
protein synthesis after reactivation. Behavioral Neuroscience 121, 148–155.
Pilley, J.W. and Reid, A.K. (2011) Border collie comprehends object names as verbal referents. Behavioural Processes 86,
184–196.
Range, F., Horn, L., Viranyi, Z. and Huber, L. (2009). The absence of reward induces inequity aversion in dogs. Proceedings
of the National Academy of Sciences USA 106, 340–345.
Regolin, L., Rugani, R., Stancher, G. and Vallortigara, G. (2011) Spontaneous discrimination of possible and impossible
objects by newly hatched chicks. Biology Letters 7, 654–657.
Roelofs, S., Boleij, H., Nordquist, R.E. and van der Staay, F.J. (2016) Making decisions under ambiguity: judgment bias
tasks for assessing emotional state in animals. Frontiers in Behavioral Neuroscience 10, 119.
Rolls, E.T. (2005) Emotion Explained. Oxford University Press, Oxford.
Rugani, R., Regolin, L. and Vallortigara, G. (2008) Discrimination of small numerosities in young chicks. Journal of
Experimental Psychology: Animal Behavior Processes 34, 388–399.
Rugani, R., Fontanari, L., Simoni, E., Regolin, L. and Vallortigara, G. (2009) Arithmetic in newborn chicks. Proceedings of
the Royal Society B: Biological Sciences 276, 2451–2460.
Salvin, H.E., McGreevy, P.D., Sachdev, P.S. and Valenzuela, M.J. (2011) The canine sand maze: an appetitive spatial
memory paradigm sensitive to age-related change in dogs. Journal of the Experimental Analysis of Behavior 95, 109–
118.
Shettleworth, S. (1998) Cognition, Evolution, and Behavior. Oxford University Press, Oxford.
Siniscalchi, M., Sasso, R., Pepe, A.-M., Vallortigara, G. and Quaranta, A. (2010) Dogs turn left to emotional stimuli.
Behavioural Brain Research 208, 516–521.
Sondergaard, E., Jensen, M.B. and Nicol, C.J. (2011) Motivation for social contact in horses measured by operant
conditioning. Applied Animal Behaviour Science 132, 131–137.
Spinka, M., Duncan, I.J.H. and Widowski, T.M. (1998) Do pigs prefer short-term to medium-term confinement? Applied
Spruijt, B.M., van den Bos, R. and Pijlman, F.T.A. (2001) A concept of welfare based on reward evaluating mechanisms in
the brain: anticipatory behaviour as an indicator for the state of reward systems. Applied Animal Behaviour Science 72,
145–171.
van der Harst, J.E. and Spruijt, B.M. (2007) Tools to measure and improve animal welfare: reward-related behaviour. Animal
Welfare 16(Suppl. 1), 67–73.
Whiten, A., Horner, I., Litchfield, C.A. and Marshall-Pescini, S. (2004) How do apes ape? Learning and Behavior 32, 36–
52.
Williams, J.L., Friend, T.H., Nevill, C.H. and Archer, G. (2004) The efficacy of a secondary reinforcer (clicker) during
acquisition and extinction of an operant task in horses. Applied Animal Behaviour Science 88, 331–341.
6 Social and Reproductive Behaviour
D.M. Weary and D. Fraser
6.1 What is Social Behaviour?

Social behaviour is the glue that allows groups of animals to function and that allows us to
interact with the domestic animals we care for. Social behaviour happens whenever animals
interact: crows fighting over picnic scraps, two people kissing at a bus stop, a dairy cow
licking her newborn calf or dogs wrestling in the park. In this chapter, we provide an
introduction to the basics of social interactions among animals and we review some of the
main types of social behaviour: competition, sexual behaviour, parent–offspring interactions
and play. Let’s begin by considering three basic questions: how do animals benefit from
living in groups, how do they communicate with each other and how did these social
behaviours evolve?
6.2 Group Living

Domestic animals are typically kept in groups. Indeed, given that essentially all domestic
species are descended from group-living ancestors, the ability to live in groups may have
been a prerequisite for domestication. Animals in the wild may benefit in several ways from
forming groups. If all else is equal, a member of a pair should be half as likely to be eaten by
a predator as would the same animal on its own. In this way animals may be able to dilute
their risk of predation by forming simple aggregations, although this benefit will be reduced
if larger groups are more likely to be detected. Risk dilution may also be shared unequally
among group members. For example, animals may benefit more from group membership if
they are at the centre of the herd rather than on the periphery, or by associating with more
vulnerable herd-mates. According to Canadian folk wisdom, it is best to travel in bear
country with companions you can outrun.
Groups may also be of value in the defence of vulnerable young. Predators will
sometimes attack the young of large animals such as muskoxen or buffalo, but a group of
adults acting in concert can create a formidable deterrent.
Group membership may also help in detecting danger. In particular, large groups have
many eyes, ears and noses for early detection of predators. Thus groups may be able to detect
danger more reliably or more quickly than solitary individuals. Also, frequent scanning for
predators reduces the efficiency of other behaviour such as foraging. By forming larger
groups, animals can ‘share’ the cost of scanning (Beauchamp, 2008) such that each group
member scans less and thus can forage more efficiently.
An obvious cost of foraging in a group is that animals may have to share the food that
they discover (Giraldeau and Caraco, 2000). This cost can be exploited by some individuals
through the adoption of distinct strategies or combinations of strategies: some may be food
finders while others are ‘scroungers’ that simply exploit the food. For example, a foraging
bald eagle can look for salmon on its own, or simply wait for another eagle to find the fish
and then steal or share in the meal. The choice of strategy will depend partly of the
proportion of individuals using the two strategies (scroungers will do well when they are
few) and any inherent advantage of being the finder (such as gaining longer access).
Group foraging can also lead to more efficient detection of food, especially when food
sources are distributed in patches with relatively barren areas in between. This benefit will
reduce the variation in intake over a fixed period of time, a factor that may be particularly
important to animals such as small birds that face the risk of starvation overnight if they do
not find food on a given winter’s day. Feeding in groups can also improve the probability of
capturing and consuming prey. For example, packs of wolves can kill moose that no single
wolf could manage, and groups of orcas can surround and capture schools of salmon.
Group living can provide other foraging benefits. For example, dairy calves reared in
social groups begin to eat solid food sooner than do calves reared individually (Costa et al.,
2016). In part this is because social rearing makes calves less neophobic (i.e. fearful of new
food items), but the higher intakes may also be due to social facilitation (i.e. seeing other
group members eating makes calves more interested in the feed) and social learning (i.e.
learning from group-mates what is good to eat).
6.3 Communication
Almost all social behaviour involves some form of communication (Bradbury and
Vehrencamp, 2011). When a young calf becomes separated from the cow it will call
repeatedly. When the cow hears these calls, it will turn towards the sound, attempt to
approach the caller and vocalize. Thus, signallers can affect the behaviour of receivers by the
signals they produce. Communication can occur through a range of modalities – by sound as
in the current example, but also by smell (or other chemical detection), sight and touch.
Humans, having a poorly developed sense of smell, find it hard to appreciate the
importance of smell for many of the domestic mammals. Dogs, for example, are 100 million
times more sensitive than humans to some chemicals. Many species use chemical signals to
communicate such things as territorial boundaries and reproductive status. Scents have the
advantage that they can be deposited and serve as a marker long after the signaller has left.
Some mammals also possess a vomeronasal organ, located at the base of their nasal cavity,
which is used to detect chemical signals; air is passed over this organ when animals show
‘flehmen’ or lip-curling, such as when a stallion detects a mare in heat.
When animals are in relatively close range or live in an open habitat they can use visual
signals such as antlers on deer, colourful plumage on some birds, or body movements such as
holding the tail erect by dominant dogs. Body movements have the advantage of being
flexible and hence well suited for signalling information that is likely to change depending on
the signaller’s condition. More static displays, such as plumage coloration, are useful for
signalling more stable information such an animal’s species, sex and individual identity.
Tactile signals include social grooming in primates, suckling and nuzzling in mammals,
and many of the behaviours associated with social play. One complex example is the
nuzzling of the sow’s udder by her piglets during a nursing bout. Litters signal their desire to
nurse by gathering at the udder and nuzzling it with their snouts, even though no milk is
available at that time. In response to the nuzzling, the sow will often lay on her side, rotate
her udder so that the teats are exposed and grunt rhythmically. If enough piglets are present
and continue massaging the teats, the sow will likely respond by releasing oxytocin from the
pituitary gland. The hormone then travels through the bloodstream to the udder where it
results in an ejection of milk. However, if the nuzzling stimulus is too weak, for example
because some of the piglets have not assembled, then the release of oxytocin will not occur
and the piglets will have to try again later. This prevents a few piglets from triggering a milk
ejection (and making pigs of themselves!) when the rest of the litter is asleep. Interestingly,
the sow increases her rate of grunting at the time of the oxytocin release (Fig. 6.1). When the
piglets hear this vocal signal, they stop their vigorous nuzzling of the udder and switch to
sucking on the teats in anticipation of the arrival of milk.
Fig. 6.1. A nursing bout in domestic pigs, showing the relationship between the sow’s grunt
rate and milk ejection. At the beginning of the episode, the sow grunts rhythmically as the
piglets massage the udder with their snouts. The rate of grunting then increases rapidly at or
near the time when oxytocin is released from the pituitary gland into the bloodstream. About
20 s later, the oxytocin reaches the udder and causes an ejection of milk, shown between the
two vertical lines. The piglets generally continue to massage the udder during the slow
grunts, and then switch to sucking on the teats, apparently in anticipation of the milk, when
the rate of grunting increases. (From Whittemore and Fraser, 1974; photo courtesy of Per
Jensen.)
The relative cost and benefit of communication may help to explain why animals
communicate in some situations but not others. For example, some animals vocalize when
they are in pain, whereas others are stoic. The difference in behaviour probably reflects
differences in the potential audience. Dependent young, such as newborn piglets, may benefit
by attracting a parent if they signal pain. In contrast, an adult dairy cow may gain little from
signalling pain associated with lameness. Indeed, in a natural setting, signalling such pain
might alert predators to the cow’s higher vulnerability. Thus, we think of communication,
like other types of behaviour, as having been shaped by natural selection to maximize the
animals’ chances of survival and reproduction.
6.4 Natural Selection

How did social behaviour evolve? To answer this question, it helps to divide social behaviour
into three broad categories:
1. Mutualistic behaviour (+/+): where both the ‘actor’ and the ‘recipients’ of the behaviour
benefit.
2. Selfish behaviour (+/–): where the actor benefits but recipients experience a cost.
3. Altruistic behaviour (–/+): where the actor experiences a cost while recipients benefit.
The evolution of mutualistic behaviour comes as no surprise. We would expect natural

selection to favour types of social behaviour which, on average, benefit all the animals that
show it. Thus, for example, we might expect that fish would evolve the behaviour of
schooling, assuming that all the animals are typically better off for behaving in this way.
Selfish behaviour is also easy to explain. Because natural selection operates powerfully at
the level of the individual, we expect behavioural traits to evolve if they increase the
reproductive success of the individuals possessing them, even if they have negative
consequences for other members of their groups. For example, when a vulture joins other
vultures at a carcass, it obviously benefits from access to food even if this means that other
group members get less.
Altruistic behaviour, however, seems puzzling. In altruistic interactions individuals
behave in ways that are costly to themselves but offer benefits to group-mates. How do such
behaviours persist in populations when they reduce the reproductive success of the individual
performing them? There are, in fact, two common ways in which these behaviours can be
selected for: through kin selection and through reciprocation.
In kin selection, heritable altruistic behaviours persist when they benefit an individual’s
relatives. The genes that code for these behaviours increase in the population because they
increase the reproductive success of these relatives rather than of the individual performing
the behaviour. The evolutionary biologist W.D. Hamilton summarized this selection process
in a simple formula (Box 6.1). He posited that kin-selected altruism should occur where the
reproductive cost of an act to the individual is less than the reproductive benefit to the
recipient divided by the probability that the recipient also carries the genes for the altruistic
trait. Full siblings share half of their genes; thus, the probability of one sibling carrying an
altruistic trait exhibited by the other is 0.5 (r = 0.5). As the famous biologist J.B.S. Haldane
explained, we should be willing to lay down our lives for two full siblings or eight first
cousins!
Box 6.1. Hamilton’s Rule

B(r) > C
Kin-selected altruism should occur where the reproductive cost (C) of an act to the
individual is less than the reproductive benefit (B) to the recipient divided by the
probability that the recipient also carries the genes for the altruistic trait (called the
coefficient of relatedness between the two individuals (r)).
In the example shown, a 1-year-old scrub jay (Aphelocoma coerulescens) feeds his
parents’ current offspring, a common behaviour in the species. The gene for this
altruistic ‘helping’ behaviour is represented as the shaded portion of the circles imposed
on each bird. The juvenile helper inherited the trait from one parent (upper left) and, as
the nestlings have the same parents, the probability of each nestling carrying this trait is
0.5 (r). For this behaviour to persist in the population, the benefit must be more than
double the cost. Expressed mathematically, B(0.5) > C or B > 2C.
Kin-selected altruism is common in species that form groups of closely related

individuals (e.g. bees, wasps, ants). Worker honeybees from the same patriline, for example,
share 75% of their genes with their sisters. These workers spend their lives helping to rear the
queen’s daughters and do not reproduce themselves. This behaviour can be understood in part
because of the benefits derived from kin selection: because the workers are closely related to
the queen’s offspring, their behaviour increases the frequency of these altruistic genes via the
fertile offspring of the queen.
Altruism may occur between non-relatives under conditions that allow for reciprocation.
In ‘reciprocal altruism’ an individual will incur a cost to help others but will later benefit
when these individuals come to their aid. Because cheaters can easily exploit this behaviour,
we would expect it to occur only where social networks are stable enough for individuals to
encounter each other frequently and where individuals can identify and punish cheaters.
Some of the most interesting instances of altruistic behaviour appear to be due to natural
selection acting at the level of the group rather than the individual. In these cases individuals
pay some cost, such as reduced reproduction, to the benefit of the group. The conditions
under which such ‘group selection’ can occur in nature are limited, but artificial selection in
domestic animals can be arranged so as to favour altruistic behaviour among group members.
If poultry geneticists select individual birds with the highest egg production in a group, they
may breed inadvertently for aggressive, competitive behaviour. If, however, they keep
closely related birds together in groups, and then breed selectively from the groups that
achieve high average production, then they will breed for an ability to do well in a social
setting. Given that laying hens are typically housed in groups on commercial egg farms, the
latter would seem to be a better strategy. Indeed, experimental work has shown that such
group-level selection can lead in just a few generations to birds that are both highly
productive and relatively non-aggressive (called ‘KGB’ or ‘kind, gentle birds’ by Cheng,
2010). This is a promising area for new research on other domestic animals we house in
groups.
6.5 Common Social Behaviours

Competition
Why do some animals defend resources from competitors whereas others simply use
resources without trying to defend them? Imagine a group of sows foraging for worms in a
freshly ploughed field. The worms will be distributed fairly evenly in space, so an animal
would gain little from defending any specific part of the field – in fact, the time spent in
defence would only be lost from foraging. The sows may still be engaged in a kind of
competition, in the sense that the animals that find and eat the worms most efficiently leave
fewer for the others. However, this so-called ‘scramble competition’ is indirect and not
overtly aggressive. Now imagine the same group of sows fed a concentrated diet from a
single feeder. In this situation, an aggressive sow may well benefit from defending the food
source from other sows (‘territorial defence’), because the time and effort needed to exclude
competitors may allow the territory holder to consume extra food.
Which of these two patterns we witness will depend partly on how resources are
distributed in both time and space. As we see in this example, it is generally less beneficial to
defend resources that are spread over a large area. How resources are distributed over time
will also influence the type of competition. For example, feed delivered gradually over the
day as a slow trickle is more easily defended than an equivalent quantity delivered all at
once.
When resources are defended, we expect the outcome of contests to be decided by
differences in factors such as body size, strength or competitive abilities. For example,
piglets must compete for access to one of the sow’s teats to obtain milk and to survive and
grow, and piglets that are heavier at birth have an advantage at acquiring a productive teat.
However, in this case body size is not the only factor affecting competitive success. Piglets
are also born with incisors that protrude laterally – making effective weapons for competition
with their siblings. Piglets that lack these weapons (e.g. as a result of the common husbandry
procedure of teeth clipping) are less able to compete for a teat if other littermates have their
teeth intact (reviewed in Drake et al., 2008).
Sometimes other sorts of asymmetries can affect which animal wins a contest. One of the
best known relates to the advantage of being the current resource holder. Territory holders
may benefit from better knowledge of the territory. Also, intruders must often compete not
only with the resident but also with the neighbouring territory holders – a cost that an
established resident does not have to pay. Animals can also vary in their need for a specific
resource, and this too can affect competitive behaviour. For example, a lioness may defend a
portion of a carcass for a period of time, but as she becomes satiated she is more likely to be
usurped by hungry pride-mates.
When individual animals are in frequent competition over resources, they can avoid the
cost of continual aggression by establishing which one is ‘boss’. In the 1920s an ethologist
studying small flocks of chickens noticed that one bird in each flock tended to peck all the
others, whereas a second bird pecked all but the first, a third pecked all but the first two, and
so on. This simple relationship came to be called the ‘peck order’ or ‘dominance order’ of the
flock (Fig. 6.2a). This idea took wing and soon animal behaviourists were describing social
behaviour in other situations, from rat colonies to office politics, in terms of dominance
orders, often by testing animals to determine which would have priority of access to
resources such as food.
Fig. 6.2. Classic examples of the ‘peck order’ in two small flocks of chickens. Birds are
identified by letters corresponding to ‘rank’ where bird A pecked the greatest number of other
birds, B pecked the second greatest number, and so on. Each bird pecked all birds below it in
the diagram, except as shown by arrows pointing upwards. (a) Thirteen young females
showed an almost perfectly linear order except that bird J pecked the higher-ranking H. (b)
Ten young males showed a complex set of relationships with more than half of the birds
pecking higher-ranking flockmates. (From Masure and Allee, 1934.)
In reality, social relationships between animals are generally far more varied and
complex than a simple dominance order would suggest. For many species, including
domestic dogs, a simple hierarchy may misrepresent the kind of social behaviour seen in the
group (Bradshaw et al., 2009). In long established groups of pigs, for example, the animals
often show mild mutual aggression with no clear hierarchy. Other species have multiple
forms of social relationships: in cattle, for instance, individuals with high priority of access to
food may not be the ones that normally lead the herd to new locations. Even dominance
relations among chickens can depart greatly from any simple pecking order (Fig. 6.2b). A
focus on dominance also tends to emphasize competition between animals, whereas some
relationships are characterized more by cooperation and affiliative behaviours, such as the
mutual licking (allogrooming) shown by dairy cows kept in groups.
Sexual behaviour
When people breed animals on farms or in zoos, they generally select which males will mate
with which females in order to achieve the desired combinations of genes in the offspring,
and then either use artificial insemination or hope that the chosen parents will cooperate by
mating. In the wild, however, the animals themselves make these choices. The scientific
study of mate selection has begun to uncover the underlying principles.
The importance of mate choice to males and females depends on the time, energy and
other resources that the animals invest in producing the offspring. Some animals (particularly
the males in many mammalian species) contribute only gametes (sperm) that are relatively
inexpensive to produce, whereas other animals (mammalian females) incur much larger
reproductive costs such as maintaining pregnancy and feeding the young by producing milk.
In general, we expect that the sex that makes the larger investment (and thus has a lower
potential reproductive rate) will be the most selective. In all of the common domestic
animals, the male is able to father young at a much faster rate than females can produce them,
so we expect females to be selective in their choice of mates. In the wild, the elaborate
breeding displays and feather coloration of males in many bird species – traits that are
attractive to females but may also increase the birds’ vulnerability to predation – illustrate the
strong selective pressures that have shaped males to become attractive to females.
Mating systems can be divided roughly into four categories: monogamy, polygyny,
polyandry and promiscuity. In monogamous systems, male and female bond for some period
of time and often both parents contribute to care of the offspring. Polgynous systems are
characterized by males mating with multiple females and females mating only with a single
male and normally caring for the young. Polyandrous systems, seen especially in some
species of fish, are the reverse. Promiscuous systems involve a mixture of polygyny and
polyandry. Although the ancestors of domestic species may have employed a different mating
system, domestic animals are typically promiscuous or polygynous. For example, whereas
wolves are monogamous, domestic dogs tend to be promiscuous. The seasonality of breeding
is also reduced in domestic animals with most domestic species able to breed year-round.
Polygynous males can gain mating opportunities by controlling access to resources, such
as food or nesting sites, that are important for female reproduction. This ‘resource defence
polygyny’ is facilitated by factors that help make resources defendable, such as the clumping
of suitable resources in space. Males that control key resources will be more likely to mate or
find multiple mates. For example, male red-winged blackbirds that have territories containing
the most suitable nesting sites are more likely to attract multiple females as mates.
Alternatively, polygynous males can attempt to defend a group of females as a harem. This is
facilitated when females group for other reasons such as for anti-predation benefits as
discussed above. Among red deer, for example, stags compete with other males for control of
harems, and somewhat similar competition among males is common in some domestic
species.
Parent–offspring interactions
Some of the most important social interactions occur between parents and their dependent
young. Parents typically prepare suitable locations for the young, provide them with nutrients
and protect them from harm, although how these goals are achieved varies enormously from
species to species.
Pigs and sheep provide an interesting contrast. The sow gives birth to a large number of
small, fragile piglets. Typically she separates herself from her usual social group a day or so
before farrowing, and seeks a secluded nest site. She prepares the site by rooting the soil to
form a soft depression, and then lines it with resilient material such as branches and with soft,
insulating material such as grass. While the piglets are being born she lies relatively
immobile with the udder exposed, leaving the piglets to find the teats unassisted. The
carefully prepared nest site protects the piglets from cold and keeps the young in a single
location that the mother can protect. With her relatively passive behaviour during parturition,
the sow avoids harming the piglets by excessive movement, but does not learn to
discriminate between her own and foreign young for perhaps a day or more. Hence, it is
relatively easy to foster day-old piglets from a sow that has too many young to one with a
smaller litter.
Ewes, which normally give birth to only one or two large, mobile offspring, follow a
different pattern. A ewe may move away from other sheep for parturition, or give birth in the
midst of the flock. Once a lamb is born, the ewe licks it vigorously; this stimulates the lamb
to rise and suckle, and it exposes the mother to the odour of the lamb, such that she can tell
her lamb from others soon after birth. Moreover, lambs are exposed to the mother’s voice
while still in the uterus and they seem to respond specifically to the calls of their own
mothers, more or less from birth. Thus, although the lamb and ewe are highly mobile, they
remain close together in the flock through mutual recognition and attraction. If farmers want
to foster an orphan lamb to a ewe whose own lamb has died, they often have to go to lengths
to disguise the appearance and odour of the orphan or the ewe will refuse to let it suckle.
Given its importance to fitness, parental behaviour is underlain by strong motivations.
Under the influence of hormonal changes before parturition, the sow becomes extremely
restless and even if confined in a narrow stall will still greatly increase her level of activity,
standing and lying repeatedly and showing a strong interest in nesting material. Once the
young are born, the sow may become protective if intruders approach the nest; and if a piglet
escapes from the nest and gives characteristic ‘separation calls’, the sow becomes very
attentive and seeks out the lost animal. Maternal motivation in the ewe takes a somewhat
different form. Ewes that are about to lamb often become intensely attracted to newborn
lambs, and may even ‘steal’ newborns from other ewes, but this behaviour ceases once her
own lamb is born.
Traditionally, many animal behaviourists studied parent–offspring behaviour as if the
relationship were all one-way, with the parent providing care for relatively passive young. In
reality, the young have signals – sometimes subtle ones – that solicit care from the parent.
Many young mammals and birds have some form of distress call when separated and hunger
signals that encourage the parents to provide more food. In a famous demonstration,
ethologist Lars von Haartman showed the power of these begging calls. He studied pied
flycatchers that were raising their young in specially designed nest boxes containing a hidden
compartment where von Haartman could temporarily conceal a second brood of hungry
nestlings. Exposed to a double dose of begging calls, from both their own brood and the
concealed brood, the parents increased their rate of bringing food to the nest, to the benefit of
their fortunate offspring.
Given the ability of young to stimulate parental care, parents need to strike a balance
between providing too much care and too little. If parents devote too many resources to their
present young, they may be slow to rebreed or not be in good enough condition to raise their
next young successfully. The young, however, having a greater stake in their own success
than in that of future siblings, may solicit a higher level of care than the parents ought to
provide for their own maximum reproductive success. One obvious outcome of these
conflicting interests is ‘weaning conflict’ whereby the parent may repel attempts by young to
suckle or solicit food. Some housing conditions make it difficult for the mother to escape the
offsprings’ demands. For example, if sows are able to get away from their piglets, they
gradually reduce the frequency of nursing and thus force the piglets to eat more solid food,
but if the sow and litter are kept permanently together, the sow fails to reduce the frequency
of nursing and may lose much more body condition through the heavy demands of lactation
(Fig. 6.3).
Fig. 6.3. The number of nursing episodes per day by sows in relation to the age of the piglets.
Sows confined with their piglets (control, circles) maintained a similar number of nursings
per day throughout the 27 days of lactation. Those that could leave the young by stepping
over a low barrier (experimental, squares) gradually reduced the number of nursings starting
in the third week of lactation. (From Weary et al., 2002.)
Play
One of the most interesting aspects of social behaviour is play, and differences in play are
sometimes used to make inferences about animal welfare (Held and Špinka, 2011). Social
play often consists of interactions that imitate the process, if not the end point, of more
clearly functional behaviours like fighting and hunting. Sometimes animals use specific
behaviours that signal the onset of play. Domestic dogs, for example, will often ‘bow’ during
play sequences, especially if play includes fighting behaviours that could trigger aggression
from the playmate.
How animals benefit from play is not well known. Play may serve as practice for adult
activities, allow animals to become familiar with their environment, and develop social skills
and relationships. Social skills are especially important for species that need practice to
develop effective courtship, appeasement or competitive behaviour. In a series of
experiments that were both famous and infamous, psychologist Harry Harlow raised infant
rhesus macaques in individual cages where they had no physical contact with either their
mothers or other young monkeys. The objective was to raise very healthy animals by keeping
them free from infectious diseases, but Harlow soon discovered that the animals showed
signs of emotional disturbance and dysfunctional social behaviour: they stared into space,
circled in their cages in a stereotyped manner and rocked repetitively. However, Harlow
found that as little as 20 min of play per day with other monkeys in an enriched environment
was enough for reasonably normal social behaviour to develop.
In the farmed animals, play is likely less critical than it is for primates, but it may still
serve a valuable role. For example, pre-weaned dairy calves that are reared in groups (as
compared with the more conventional individual housing) spend time playing and these
calves are more likely to become dominant when mixed with animals that have been
individually reared. Play may also prepare animals for coping with unusual situations, such
as maintaining balance on a slippery surface. Indeed, play sequences often involve some
aspect of self-handicap, such as adopting ungainly postures, and this may allow young
animals to develop physical skills that will prove useful later in life. Play specifically, and
social interactions more generally, may also benefit the mental skills of animals. For
example, dairy calves that have been reared in groups show better performance in certain
learning tasks (Meagher et al., 2015).
Chapter 7 contains a more comprehensive text on play behaviour and its significance.
6.6 Conclusions
For millennia the domestication of animals has depended on the ability of humans to
understand the social behaviour of animals, to accommodate this behaviour in controlled
circumstances and sometimes become part of the animal’s social world. There is great scope
for improving our care and handling of domestic animals by further improving our
understanding of their social behaviour.
Most of the theories, concepts and vocabulary described in this chapter were developed
by behavioural scientists to help us understand how behaviour equips animals to live in the
environment in which the species evolved. With most domestic animals, many of the
behavioural adaptations of the species are still present, but may be ill suited to the unnatural
physical and social environment in which the animals are kept. Like its wild counterpart, the
newly hatched domestic chick appears predisposed to become imprinted on a parental figure,
but what form does imprinting take when hundreds of chicks are hatched together in an
incubator? Like the wild sow, the domestic sow seems predisposed to wean her young
gradually by spending less and less time with them as they age, but what are the
consequences if the sow and litter are penned together continuously for several weeks, and
then abruptly and permanently separated? Females of many species seem predisposed to
select mates based on certain attributes, but if they are penned with only a single male, does
this affect their willingness to mate or how strongly they display oestrus? One area of special
interest is how social behaviours, which likely evolved to facilitate interactions among small
groups of related individuals, can break down when animals are housed in pens containing
hundreds of unrelated animals (Croney and Newberry, 2007). Thus a key challenge for future
research is to understand how rearing conditions can be altered to better suit the animals’
social environment, and thus help avoid problems for both the animals and the people who
work with them.
References
Beauchamp, G. (2008) What is the magnitude of the group-size effect on vigilance? Behavioral Ecology 19, 1361–1368.
Bradbury, J.W. and Vehrencamp, S.L. (2011) Principles of Animal Communication, 2nd edn. Sinauer Associates,
Sunderland, Massachusetts.
Bradshaw, J.W., Blackwell, E.J. and Casey, R.A. (2009) Dominance in domestic dogs – useful construct or bad habit?
Journal of Veterinary Behavior 4, 135–144.
Cheng, H.W. (2010) Breeding of tomorrow’s chickens to improve well-being. Poultry Science 89, 805–813.
Costa, J.H.C., von Keyserlingk, M.A.G. and Weary, D.M. (2016) Effects of group housing of dairy calves on behavior,
cognition, performance and health. Journal of Dairy Science 99, 2453–2467.
Croney, C.C. and Newberry, R.C. (2007) Group size and cognitive processes. Applied Animal Behaviour Science 103, 215–
228.
Drake, A., Fraser, D. and Weary, D.M. (2008) Parent–offspring resource allocation in domestic pigs. Behavioral Ecology
and Sociobiology 62, 309–319.
Giraldeau, L.A, and Caraco, T. (2000) Social Foraging Theory. Princeton University Press, Princeton, New Jersey.
Held, S.D. and Špinka, M. (2011) Animal play and animal welfare. Animal Behaviour 81, 891–899.
Masure, R.H. and Allee, W.C. (1934) The social order in flocks of the common chicken and pigeon. The Auk 51, 306–327.
Meagher, R.K., Daros, R.R., Costa, J.H.C., von Keyserlingk, M.A.G., Hötzel, M. and Weary, D.M. (2015) Effects of degree
and timing of social housing on reversal learning and response to novel objects in dairy calves. PLoS One 10, e0132828.
Weary, D.M., Pajor, E.A., Bonenfant, M., Fraser, D. and Kramer, D.L. (2002) Alternative housing for sows and litters. Part
4. Effects of sow-controlled housing combined with a communal piglet area on pre- and post-weaning behaviour and
performance. Applied Animal Behaviour Science 76, 279–290.
Whittemore, C.T. and Fraser, D. (1974) The nursing and suckling behaviour of pigs. II. Vocalization of the sow in relation to
suckling behaviour and milk ejection. British Veterinary Journal 130, 346–356.
7 Play Behaviour
S. Held
7.1 What is Play and Who Plays?

Piglets sparring, gangs of lambs gambolling up and down a field, calves leaping about in
straw pens, laboratory rats pouncing on each other or kittens batting balls of wool: these are
easily recognized instances of domestic animals at play. In many domestic animals, play is
usually easy to spot and common, especially in infants and juveniles. Considering the whole
animal kingdom, it is most widespread in mammals. Almost all mammals play, and among
them it is most common in ungulates, carnivores, primates and rodents (Burghardt, 2005).
This covers the classes that make up the bulk of mammalian domestic species (ungulates,
including cattle, horses, pigs, sheep, goats; carnivores, including dogs and cats).
Comparisons across mammal and bird species suggest that total brain size may explain
differences in the prevalence and complexity of play, with larger-brained animals playing
more (e.g. Iwaniuk et al., 2001). Play is also positively correlated to the relative size of brain
areas that are stimulated by gonadal hormones to trigger sexually dimorphic behaviours (e.g.
Lewis and Barton, 2006 for social play in primates). Within birds play is also common, and
most varied and complex in corvids and parrots (Diamond and Bond, 2003; Burghardt,
2005). To what extent galliform (chicken, quail) and anseriform birds (ducks, geese) play is
less clear. These make up the other main domesticated animal group, and it has been
suggested that they express some but not all types of play (Burghardt, 2005).
Whether play is part of a species’ behavioural repertoire is important for our
understanding of the evolution and functions of play – and clearly depends largely on how
we define play. Play may often be easy to spot but properly defining it has proved difficult
because of its enormous variety, variability and a lack of clarity about its purpose.
Clearly, play can take quite different forms in different species. Extreme examples are
fantasy role-play in children compared with ravens dropping and catching objects in flight
(Heinrich and Smolker, 1998). Not only does play vary considerably between species, it also
does so within species. In many of them, the animals exhibit different types of play; males
and females may express play differently, especially in sexually dimorphic species; and play
types and frequencies typically change with age (Fagen, 1981). This variety means that
defining play on the basis of its structure or morphology (i.e. on the basis of what it looks
like) is an impossible task since it takes so many different forms. It has proved similarly
difficult to define play on the basis of its purpose or function (i.e. on the basis of what it is
for), that is its biological survival value. This is largely because it remains unclear whether
play has evolved for a single or several functional. Given the difficulty in structurally or
functionally defining play, biologists have instead characterized it using lists of key features
shared across all instances of play. One such comprehensive, useful list is provided by
Burghardt (2005). It consists of five criteria essential to identifying a behaviour as play. The
criteria are that the behaviour: (i) is not fully functional in the sense that it contains elements
‘that do not contribute to current survival’; (ii) is self-rewarding (i.e. it is performed for its
own sake rather than to achieve some other goal; the immediate reward to the animal is in
performing the behaviour itself); (iii) differs in structure or timing from its adult, serious
form (e.g. gambolling versus running, play-fighting versus fighting; or occurring at an earlier
age than needed for survival); (iv) is performed repeatedly but not stereotypically (i.e. it will
typically vary across repeats); and (v) is expressed when the animal is in a ‘relaxed field’ (i.e.
when there are no immediate environmental risks to the animal’s fitness) (Burghardt, 2005).
7.2 Types of Play and Changes in Play over Time

Piglets play-fighting, calves leaping about in straw pens and kittens batting balls of wool are
examples of the three main types of play: social play, locomotor-rotational (LR) play and
object play.
Social play is play directed at another animal, or also a human in the case of companion
animals (Burghardt, 2005; Bradshaw et al., 2015). LR play involves vigorous running,
jumping, headshakes and/or body twists. These movements are often combined and repeated,
with the animals characteristically returning to where the play started as opposed to moving
from or towards something as in non-play locomotion (Wilson and Kleiman, 1974). LR play
is done alone, and also in groups (‘parallel play’, Jensen et al., 1998) as often seen in calves,
lambs, piglets and foals. Object play involves manipulating an inanimate object, for example
pawing or mouthing it, and again can be done alone or in company. A single play bout often
contains elements of the different play types with quick changes between them, as for
example in piglets or calves chasing around in groups, stopping to play-mount or play-fight,
then chasing on, stopping again and so on (e.g. Newberry et al., 1988; Jensen et al., 1998).
The three types of play are also often combined into new forms.
Dogs, for example, appear to prefer using play objects when socially playing with
humans (e.g. tug-of-war, throw–fetch). They also play with objects by themselves in ways
that resemble predatory behaviour (biting, shaking, ripping; Bradshaw et al., 2015).
However, their play is predominantly social as may be expected from both their evolutionary
and domestication history. Dogs are descendants of wolves, and genetic analyses confirm
they belong to the wolf-like grouping within the canids (Lindblad-Toh et al., 2005; see also
Chapter 16). Canids, generally, are social carnivores but among them, the wolf-like species
have the highest sociality levels, for example in terms of hunting strategy (cooperative
hunting in packs) and social organization (complex dominance hierarchies) (Marshall-Pescini
and Kaminski, 2014). Social play, particularly play-fighting, can play a role in asserting
dominance, and the specific play movements that help to signal the harmless intent of play
interactions to prevent misinterpretation and potential escalation into serious fighting are
more prevalent in younger and subordinate play partners (Bauer and Smuts, 2007).
Domestication has meant an increase in playfulness in dogs compared with their wild
ancestors, possibly as a consequence of selective breeding for neoteny and/or for sociability
with humans (Bradshaw et al., 2015).
Domestic cats, too, play more than their wild ancestor and closest living relative, the
wildcat. However, in contrast to dogs, domestic cats are naturally solitary and territorial, and
still highly motivated to hunt (see also Chapter 17). Their play is thus predominantly object
play, resembling hunting, prey-catching and killing (see e.g. Biben, 1979; Leyhausen, 1979).
Domestic cattle, horses, sheep, goats and pigs are social herbivores or omnivores (pigs)
with their play behaviour similarly still reflecting ancestral natural selection pressures. Living
in groups and being fast-moving reduce predation risk. Play in these domestic ungulates is
thus mainly LR and social play characterized by gambolling, pivoting, kicking out and
running (all done alone or in groups; see Fig. 7.1), as well as play-fighting and mock-
mounting which is often combined with LR play in the same bout (cattle: see Reinhardt,
1980; Jensen et al., 1998; horses: see Crowell-Davies et al., 1987; Cameron et al., 2008;
sheep: see Sachs and Harris, 1978; Berger, 1980; pigs: see Newberry et al., 1988; Blackshaw
et al., 1997).
Fig. 7.1. Example of locomotor-rotational play.
Young animals typically play more than adults, and in mammals play frequency tends to
peak twice: once in infancy just after play first emerges and then again later in the juvenile
period. Play in many domestic mammals follows a similar developmental course. One
possible explanation is that the timing of their play peaks reflects behavioural adaptations in
their respective wild ancestors and relatives. In domestic lambs, for example, play peaks
around the third week of life, and then again during weeks 7–9 (Sachs and Harris, 1978).
Ancestral and related wild species live in mountainous cliff terrain where ewes would
separate from the group to lamb in secluded spots on the cliffs, inaccessible to predators.
Much playful physical activity on a cliff would be dangerous to a newborn in the first few
days of its life and may therefore have been selected against. This could explain the low play
rates still found in domestic lambs in the first few days of life (see Dwyer, 2003). Towards
the end of the ancestral mountain lambs’ first week, they leave the cliffs with their mothers to
join other ewes and new lambs lower down the mountains (Sachs and Harris, 1978). They
then stay in these mother–lamb groups for 3–4 weeks until they start their migration to the
summer grounds. During this time, the risk of being predated is reduced for each lamb
because there are greater overall numbers of lambs, the terrain is flatter and thus safer, and
lambs meet many peers, all of which facilitates play. The first play peak in domestic lambs
thus largely coincides with the first safe and social period in a mountain lamb’s life. The
second play peak corresponds to the timing of the arrival at the summer grounds when
reduced travel and increased nutritional quality again create conditions that facilitate play
(Sachs and Harris, 1978).
LR play precedes social play in development, as demonstrated in Welsh pony foals
(Crowell-Davies et al., 1987), which start LR-playing by themselves from the first day of
life. Solitary LR play stays the main type of play during the first 4 weeks with a marked
decrease thereafter. Parallel LR play in groups is first seen in week 2 but stays rare until it
rises in weeks 5–8 when it can make up 22% of all play bouts. Social play emerges during
the first 4 weeks with colts playing more than fillies from the beginning. By weeks 5–8 social
play rises to 52% of the colts’ play bouts and 22% of the fillies’.
7.3 ‘Playfulness’
The terms ‘playful’ and ‘playfulness’ are used in at least two different ways in the ethological
and psychological literature, each one referring to an important aspect of domestic animals’
play.
In one sense, ‘playful’ refers to individuals and groups with high play frequencies or high
propensities to play. Where such playfulness appears to be a stable, consistent trait of an
individual, or characteristic of a group, it may help explain the proximate mechanisms
underlying play. We therefore return to this again later when we introduce some mechanisms.
In the other sense, ‘playful’ obviously refers to a particular behaviour that is being
performed in a playful manner. Burghardt’s criteria are designed to help with deciding
whether a certain behaviour is performed playfully but not all instances are clear-cut, and
telling play from non-play can be difficult. Social interactions, especially, can be ambiguous,
for example when distinguishing play-fighting (also called ‘rough-and-tumble’ (R-T) play)
from genuine fighting (see e.g. Palagi et al., 2016). The use of play markers or play signals
can help here. Play markers are movements and actions that only occur during the playful
version of a behaviour, never in the serious form. Piglets tossing their heads and pivoting
while fighting or running would be an example: it characterizes play-fighting or LR-play
bouts but not serious fighting or movement from A to B (Newberry et al., 1988). Head-
tossing and pivoting are two of a number of play markers identified for pigs by Newberry et
al. (1988). The low-pitched, bark-like vocalizations that pigs often produce during play are
not unambiguously play markers as they also occur in alarm situations (Newberry et al.,
1988; Chan et al., 2011; Reimert et al., 2013). Laboratory rats can also produce characteristic
vocalizations during play (at 50 kHz); however, they are, again, not unique to play situations
but instead are thought to be produced in situations that induce a positive affective state (such
as during anticipation of a reward, during mating, R-T play, being tickled; reviewed in
Burgdorf et al., 2011; see also Melotti et al., 2014). Play signals per se are specific
communication signals around social play, especially R-T play (Palagi et al., 2016). A classic
domestic animal example is the dog’s play-bow (Fig. 7.2), which signals an intention to play
to a potential partner (Bekoff, 1995). Play signals are key to soliciting and maintaining social
play without it appearing challenging to established dominance rank relationships or
escalating into injurious fights (Rooney et al., 2001; Palagi et al., 2016).
Fig. 7.2. Play-bow. (Photo courtesy of Julie Bedford.)
7.4 Mechanisms
Play levels change in response to many diverse factors, suggesting a range of proximate
mechanisms. A growing body of work is now available on the brain mechanisms that
modulate social play in laboratory rats. It provides the strongest evidence yet for a specific
proximate mechanism underlying play. The brain’s opioid, endocannabinoid, dopamine and
noradrenaline systems together regulate the rewarding properties of play through their
actions in a network of corticolimbic structures (see e.g. Vanderschuren, 2010;
Vanderschuren et al., 2016). These include structures which are involved in the regulation of
social behaviour and/or the generation and perception of emotions (nucleus accumbens,
ventral tegmental area, pallidum, frontal cortex and amygdala; Vanderschuren et al., 2016).
Dopamine is the main neurotransmitter responsible for modulating the rat’s motivation to
play, while opioids are key in determining its hedonic impact (i.e. the extent to which social
play is experienced as rewarding and pleasurable in the rat) (as in Kringelbach and Berridge,
2012; see also below).
Studying individual differences provides another approach to investigating the
mechanisms underlying different play behaviours. In animals, individual differences in
playfulness have been most intensively studied in domestic dogs, where playfulness appears
to be a stable individual trait consistent across different situations (e.g. Svartberg and
Forkman, 2002) and related to several other personality traits including ‘sociability’ and
‘responsiveness to training’ (e.g. Jones and Gosling, 2005), which could point to a genetic
predisposition. A genetic component is also suggested for playfulness in two strains of inbred
laboratory mice (Balb/c, DBA/2) where play frequencies were found to be moderately
heritable (Walker and Byers, 1991).
In domestic piglets, playfulness has recently been shown to be more variable between
litters than between individuals within litters (Brown et al., 2015). These litter differences
could not be explained by differences in general activity, which suggests different underlying
causes. Litters that played more also had higher postnatal growth rates, pointing to energy
availability as the underlying factor, here most likely via differences in milk supply between
the different sows (Brown et al., 2015). Litter differences in playfulness were also found in
cats and feral dogs (Martin and Bateson, 1985; Pal, 2010) and in older, weaned piglets
(Rauw, 2013), suggesting pre-weaning differences can persist into the post-weaning period.
Energy availability generally is a key factor determining play rates, which we return to later
when we consider how environmental conditions affect play in domestic animals related to
welfare. Note that maternal effects have also been described in feral horses and farmed
sheep, where differences in maternal investment are similarly associated with differences in
the play rates of lambs and foals (Dwyer, 2003; Cameron et al., 2008).
7.5 The Functional Benefits of Play

Arguably, play should result in some fitness costs (for some discussion see Martin and Caro,
1985): the physical and mental activity involved in playing requires metabolic energy
resulting in an energy cost; playing also takes up time that might otherwise be used for some
alternative fitness-enhancing activity such as feeding or rest resulting in an opportunity cost;
and finally, playing in the wild can increase predation and injury risk if it means vigilance is
lowered or exposure is increased, for example by playing at the edge of a herd and away
from the main group (any wild-living ungulate), on cliff edges (e.g. mountain lambs; Sachs
and Harris, 1978) or congregating in a confined space which attracts predators to the play
spot and can increase strike rates (e.g. fur seal pups in tidal pools; Harcourt, 1991). Exactly
how big such fitness costs are is difficult to measure precisely, but appears to range from
negligible to substantial in different situations and species, and depending on the type of cost.
In domestic kittens, for example, the energy costs of social or object play have been
calculated as less than 10% of total daily energy expenditure, and time allocated to play as
averaging 9% of total time per 24 h (Martin and Caro, 1985 and references therein).
Predation costs in wild animals, on the other hand, can be very high as shown for the fur seal
pups (Harcourt, 1991). In that study, 87% of pups killed outside the breeding season were
killed by sea lions while playing in tidal pools. In most captive conditions (farms, homes,
zoos, laboratories) animals, including domesticated species, are well food-provisioned, and
protected from predators and other poor environmental conditions. They thus have surplus
time and energy available such that missed opportunity costs are reduced, and play should
(and often does) flourish.
Given that play thus most likely carries some fitness cost, it must also have brought some
functional benefit to the wild ancestors of our domestic animals or it would have put the
playing individuals at a disadvantage during natural selection.
What, then, is animal play good for? Many functions have been proposed. Play occurs
mainly when animals are young, it has no immediately apparent purpose and typically
therefore looks like a not-completely-functional version of a biologically important adult
behaviour with play markers added (e.g. play-mounting, play-fighting, LR play). Most, and
especially the early, hypotheses regarding function therefore focus on play as preparation for
adult life, thus having delayed, long-term benefits (e.g. Groos, 1898 cited in Špinka et al.,
2001; Fagen, 1981): what is trained in a playful form in a safe environment as youngsters
under the protection of adults will increase efficiency and competence in the serious
equivalent adult behaviour. Byers and Walker (1995) suggested that play specifically
enhances neural development in the part of the brain responsible for coordinating and fine-
tuning muscle activity (cerebellum), and also the differentiation into slow and fast skeletal
muscle fibres during early muscle growth within particular time windows in development
(‘sensitive periods’). Play is also thought to more generally train species-typical behaviours
and physical and social competencies important later in life. R-T play in male juvenile rats,
for example, has been suggested to improve their success in adult sexual mounting through
being able to more efficiently adjust their own body positions in response to the females’
movements (Pellis and Pellis, 2009). A recent study on farmed mink confirms the suggested
relationship between R-T play and adult sexual behaviour: more frequent R-T play as
juveniles was correlated with longer-lasting copulation as adults in the males (as might be
expected to increase their copulation success) and longer latencies to copulate in females (as
might be expected to facilitate greater mate choice) (Dallaire and Mason, 2017). In many
domestic species, male juveniles play-fight more than females, which has been similarly
explained as preparation for the different roles they would assume later in life under natural
conditions. In Welsh ponies living on pasture, for example, colts and fillies played equally
much, but in the colts 38% of all play was social including play-fighting as opposed to only
12% in the fillies (Crowell-Davies et al., 1987). Adult stallions have to compete, often by
fighting, for mating access to stable groups of mares. In the mares, competition for resources
is low and regulated non-aggressively via established dominance relationships (see also other
ungulates – pigs: e.g. Dobao et al., 1985; sheep: e.g. Sachs and Harris, 1978; cattle: e.g.
Reinhardt et al., 1978). Correspondingly, social play differs little between male and female
wolves, which share the same roles in the pack (Cordoni, 2009). Finally, playing has also
been proposed to improve the animals’ general physical and emotional ability to cope with
the changing, stressful, sometimes unpredictable social and physical situations they will
encounter as adults (‘training for the unexpected’, Špinka et al., 2001; ‘coping with stress’,
Pellis and Pellis, 2009).
Play, in addition, is likely to have some immediate benefits linked to the animals’ current
situation, since adults also play. Furthermore, play often accompanies behaviour that matters
currently rather than later in development, for example in socially stressful situations such as
around feeding or social mixing (e.g. Palagi, 2006; Antonacci et al., 2010). One potential
function with immediate benefits, then, is to positively influence a current social situation,
for example to reduce social tension, or also to reinforce the dominance status of the player
(e.g. in oryx antelopes: Feuerriegel, 1997; in dogs: Bauer and Smuts, 2007). A further
suggested immediate benefit is that play may help animals to obtain useful information
about: (i) its current physical environment, to improve its ability to quickly and efficiently
move around (LR play; Stamps, 1995); (ii) its group members, such as information on the
fighting skills of potential competitors (e.g. social play in wolves, Cordoni, 2009); or
information about (iii) its own developing manipulative, physical and social capabilities
(through continuous self-assessment via object, LR and social play; Thompson, 1998).
Finally, play may be a way of self-administering endogenous opioids which works, for
example, to reduce cortisol levels during periods of stress in rats (Pellis and Pellis, 2009), and
maybe also at times of low stimulation to alleviate boredom (Held and Špinka, 2011).
A couple of key points are important for concluding this introduction on proposed play
functions. First, the evidence is correlational rather than causal, and for good reason:
experimental approaches that have the potential to provide direct evidence for causal effects
are hampered by the practical challenge of manipulating play without affecting other
behaviours (see Donaldson et al., 2002; Pellis and Pellis, 2009; Dallaire, 2015). Even
correlational studies that have convincingly demonstrated an association between play levels
and survival (or other fitness parameter) still acknowledge the possibility of a third,
confounding factor influencing both play and fitness (e.g. Fagen and Fagen, 2004). Second,
the proposed functions are most likely to be complementary rather than alternative given the
variety of play across and within animals. Play may be best understood as multifunctional,
with many functions overlapping, rather than serving one overarching one (Graham and
Burghardt, 2010; Held and Špinka, 2011).

And lastly, not all explanations for the existence of play presume it initially brought a
selective advantage. Burghardt’s Resource Surplus Theory (2005) proposes that play
originally emerged as a by-product of surplus energy in many predisposed species and at
different times in evolutionary history. Endothermy, for example, would be one such
prerequisite that facilitates the build-up of excess energy. Once play had emerged, it could
then be shaped by natural selection to provide a variety of immediate and longer-term
benefits in different species (Burghardt, 2005).
7.6 Play and Welfare in Domestic Animals
As environmental conditions fluctuate, so do play rates. They are typically highest when
animals are free from fitness threats (Burghardt’s ‘relaxed field’ criterion); and play can
disappear completely from the behavioural repertoire even in young animals when conditions
become increasingly challenging (reviewed in Held and Špinka, 2011). Furthermore, because
of its self-rewarding nature, play is frequently linked to the experience of positive emotions
(as e.g. in Burghardt, 2005). Play has therefore long been proposed as a welfare indicator for
animals (Fagen, 1981; Lawrence, 1987). This section therefore considers the links between
play and welfare and introduces some behavioural and neurobiological evidence.
Play responding to changes in environmental conditions
In wild animals, food and energy availability crucially affects play frequencies. Loy (1970),
for example, observed a 17-fold drop in play behaviour in a free-living colony of rhesus
monkeys resulting from acute food shortage lasting 22 days. In bottle-fed white-tailed deer
fawns, a 33% reduction in milk supply led to a 35% reduction in play behaviour (Müller-
Schwarze et al., 1982). Conversely, play rates have been shown to increase when extra food
is made available, for example in free-living meerkats where experimental food
supplementation can more than double play behaviour (Sharpe et al., 2002).
Weather conditions influence play in the wild through their impact on food availability
and ambient temperature. In Bighorn sheep living in the Colorado Desert, lambs stopped
playing at temperatures above 32°C and 88% of lamb play occurred below 26°C (Berger,
1980). In a study on Ethiopian Gelada baboons, play rates predictably tracked rainfall but,
interestingly, the quality of play also changed: declining rainfall resulted in a shift from high-
energy play elements such as play-boxing and wrestling towards actions involving less
physical energy (e.g. object play and play-biting; Barrett et al., 1992). Injury hazards (e.g.
cactus spines in Bighorn sheep; Berger, 1980), and actual injury and disease can temporarily
reduce play (Fagen, 1981). An example is from young lambs where play can be depressed for
at least 3 days after castration (e.g. Thornton and Waterman-Pearson, 2002). Play can also be
depressed when animals are strongly motivated to perform other behaviours because of their
current high fitness consequences such as during the mating season or around feeding.
Captive male wolves, for example, show less solitary and social play during mating than in
non-mating periods, and during feeding than in post-feeding periods (Cordoni, 2009). In
harbour seals, too, solitary (locomotor) play was found to be reduced during the mating
season (Renouf, 1993).

Many domestic animals are protected from natural variation in fitness challenges but
their play behaviour still responds to environmental conditions largely as predicted from wild
animal studies. Environmental conditions influencing play in domestic animals include space
provision, group size, food availability, substrate and environmental change or novelty.
In group-housed dairy calves, for example, LR play but not play-fighting decreased
between the ages of 5 and 9 weeks (Jensen and Kyhn, 2000). Increasing space allowance in
the home pen from 1.5 or 2.2 m2 per calf to 3.0 or 4.0 m2 significantly increased LR play at 5
weeks old but not thereafter. However, when released into a large test pen (4.8 m × 9.6 m) at
10 weeks old, calves kept at 1.5 m2 showed the most LR play, suggesting increased
motivational rebound (Jensen and Kyhn, 2000; see also Dellmeier et al., 1985, Rushen and
de Passillé, 2014). Single compared with pair or group housing, lack of straw as floor
substrate, reduced milk provision and dehorning without appropriate anaesthesia also reduce
play levels in calves (e.g. Brownlee, 1954; Jensen et al., 1998; Duve et al., 2012; Krachun et
al., 2010; Mintline et al., 2013).
In piglets, too, play is promoted by straw provision and increased space. In Chaloupková
et al.’s (2007) study, for example, young piglets in standard farrowing crates tended to show
less LR and social play than piglets in 20% more spacious, strawed crates or piglets in 60%
larger, strawed farrowing pens. Weaned piglets expressed more running, gambolling,
pivoting and playing with straw using head-shaking (LR and object play) in larger test pens
enriched with straw and food treats with a companion than in smaller, barren pens (Reimert
et al., 2013). The enriched test pens in Reimert et al.’s (2013) study offered favourable
environmental conditions as well as change from the home pen environment. Novelty or
‘sudden stimulus change’ (Newberry et al., 1988) such as novel objects, or new bedding as in
Reimert et al. (2013), also typically lead to temporary increases in play rates (see also Wood-
Gush and Vestergaard, 1991).
While evidence for the negative effects of poor environmental conditions on play is
widespread and strong, examples to the contrary also exist. A classic one is the finding by
Bateson and colleagues that reductions in maternal care and milk provisioning increase play
in kittens. The restriction of maternal investment in such cases has been proposed as an
adaptive response signalling poor environmental conditions to the young (Bateson et al.,
1981, 1990; see also Smith, 1991 for rats; Devinney et al., 2003 for rhesus monkeys). Under
poor conditions, the young would be weaned and become independent from their mothers at
an earlier age. They should therefore increase their playing while still under the protection of
their mothers and in the company of their siblings. Furthermore, social play can increase
rather than decrease in certain socially stressful situations such as before feeding in captive
bonobos (Palagi, 2006). And finally, play can increase as a result of challenging conditions
when there is a new opportunity to play, as was described above for calves released from
small into much larger pens. In piglets, similarly, play rates were higher in litters released
into a play arena from barren pens than from enriched pens (Wood-Gush et al., 1990).
We can conclude, therefore, that under fitness-challenging conditions play rates in
domestic animals may: (i) indeed decrease to be replaced by behaviours that are more urgent
because of their immediate fitness benefits, thus indicating current welfare challenge; (ii)
increase because some play protects the animal against challenging conditions now or in the
future; or (iii) be depressed initially but increase above baseline when conditions improve
later (Held and Špinka, 2011).
Play as self-rewarding and indicating positive emotions
Early evidence for the self-rewarding nature of play came from conditioning studies with
chimpanzees and rats where social play was used as a conditioning reward. Mason et al.
((1963), cited in Vanderschuren, 2010), for example, showed that social play was as strong a
reinforcer as the chimpanzees’ most favoured food (fresh fruit), and a stronger reinforcer than
being petted or getting a less favoured food such as chow. When tested in a T-maze set-up,
rats learned to discriminate between non-playing and playful companions as quickly as a
food/no food discrimination (Humphreys and Einon, 1981). Calcagnetti and Schechter (1992)
showed that the opportunity for social play can reverse a previously learned place preference.
Many studies since have confirmed that social play can be used as a reward for operant
conditioning, maze learning and place preference conditioning in rats (reviewed in
Vanderschuren, 2010). In domestic animals, the rewarding nature of social play has most
closely been studied in dogs where humans act as play partners, and social play is extensively
used as a reward in training (Bradshaw et al., 2015), for example in detection dogs (Rooney
et al., 2004).
Rewards, generally, are thought to have motivational (‘wanting’) and hedonic (‘liking’)
psychological properties (e.g. Berridge and Robinson, 2003; see Rolls, 2005 for a different
view) mediated by separate, though interacting, neural pathways and mechanisms as
summarized above (see also Chapter 4). As there is strong evidence that social play acts as a
reward, we should expect it, too, to have differentially neuromodulated ‘wanting’ and ‘liking’
properties. However, much of our understanding of the neurobiology of rewards comes from
other rewarding behaviours in rats such as the consumption of sucrose solution. These studies
suggest that the brain areas and opioid transmitter systems that mediate the ‘liking’ properties
of sucrose rewards to some extent overlap with those involved in modulating social play as
introduced earlier. Centrally administered opioid agonists such as morphine, for example,
increase tongue protrusion and paw-licking in rats after tasting a sucrose solution (Peciña and
Berridge, 2000). These are characteristic behavioural hedonic responses, that is they are
behavioural expressions of ‘liking’, in rats. Important for the case for play being
accompanied by positive emotions in animals generally is the fact that the same opioid
agonists also increase social play in rats and opioid antagonists decrease it (e.g. Normansell
and Panksepp, 1990). Social play also, conversely, increases opioidergic activity in the brain
suggesting that it may not only reflect changes in opioid levels but also cause them
(Vanderschuren et al., 1995). Since then, many studies have confirmed the central role of
brain opioids in the regulation of social play in rats (Vanderschuren, 2010), pointing to an
immediately rewarding, pleasurable positive emotion mediated by brain opioids.
We should expect opioid involvement leading to rewarding, pleasurable emotions during
social play also in many domestic animals on the basis of homologous or analogous
neurobiological features. That said, social play, and play-fighting in particular, plays a role in
maintaining dominance relationships in many domestic animals (e.g. pigs: Newberry et al.,
1988; dogs: Bauer and Smuts, 2007), as mentioned before. In such instances, social play is
likely to be accompanied by negative emotions due to (social) stress, not just positive,
rewarding emotions (Enkel et al., 2010; Mendl et al., 2010).
It is also worth noting that neurobiological evidence is currently limited to social play in
rats with one study on common marmosets (Guard et al., 2002). It suggests a proximate
mechanism for reinforcing a behaviour for which fitness-enhancing benefits (and associated
rewarding experiences) may be delayed (such as reinforcing play-fighting in juveniles to
increase adult sexual competence, as in Pellis and Pellis, 2009). Since delayed fitness
benefits are also proposed for object and LR play, they, too, might be accompanied by a
rewarding, pleasurable emotion (Held and Špinka, 2011).
Play as improving welfare in domestic animals
Play clearly directly contributes to good welfare in domestic animals where it positively
influences current social situations, for example by diffusing social tension and maintaining
rank relationships in dogs (see above). Play of course also improves welfare where it helps
with developing physical skills, and social and emotional competencies important for a
healthy, successful life. In domestic animals, though, healthy and successful lives typically
depend on fewer specific survival skills than in wild animals. While young domestic animals
may thus still be motivated to play in ways that would train their species-specific predation,
predator escape, foraging or mating skills, the welfare consequences for their later lives may
be limited. That said, play may still have long-term welfare benefits in domestic animals
where it contributes to developing more generic competencies such as being able to flexibly
and successfully adjust to changing, stressful conditions (e.g. pigs: Donaldson et al., 2002;
rats: von Frijtag et al., 2002; see also Špinka et al., 2001, Pellis and Pellis, 2009).
Finally, we have already considered behavioural and neurobiological evidence for play
being rewarding and likely to be accompanied by pleasurable emotions. If we consider
welfare as being crucially dependent on an animal’s subjective experiences, we might thus
conclude that a playing animal typically is in a good state of welfare. On that basis, we can
improve domestic animal welfare by identifying and providing the social and physical
conditions that promote their play behaviour. In addition, play is contagious in that it spreads
across individuals, through groups. Seeing one animal play often triggers play in others (see
e.g. Špinka, 2012 and references therein). Play may thus act, and be ultimately useable, as an
emotional contagion spreading pleasurable emotions through groups of domestic animals,
thereby improving their welfare (Held and Špinka, 2011; Špinka, 2012).
Acknowledgements
Sections of this chapter owe much to discussions with Marek Špinka while preparing Held
and Špinka (2011).
References
Antonacci, D., Norscia, I. and Palagi, E. (2010) Stranger to familiar: wild strepsirhines manage xenophobia by playing.
PLoS One 5, e13218.
Barrett, L., Dunbar, R.I.M. and Dunbar, P. (1992) Environmental influences on play behaviour in immature gelada baboons.
Bateson, P., Martin, P. and Young, M. (1981) Effects of interrupting cat mothers’ lactation with bromocriptine on the
subsequent play of their kittens. Physiology & Behavior 27, 841–845.
Bateson, P., Mendl, M. and Feaver, J. (1990) Play in the domestic cat is enhanced by rationing of the mother during
lactation. Animal Behaviour 40, 514–525.
Bauer, E.B. and Smuts, B.B. (2007) Cooperation and competition during dyadic play in domestic dogs, Canis familiaris.
Bekoff, M. (1995) Play signals as punctuation: the structure of social play in canids. Behaviour 132, 419–429.
Berger, J. (1980) The ecology, structure and functions of social play in bighorn sheep (Ovis canadensis). Journal of Zoology
192, 531–542.
Berridge, K.C. and Robinson, T.E. (2003) Parsing reward. Trends in Neurosciences 26, 507–513.
Biben, M. (1979) Predation and predatory play behaviour of domestic cats. Animal Behaviour 27, 81–94.
Blackshaw, J.K., Swain, A.J., Blackshaw, A.W., Thomas, F.J.M. and Gillies, K.J. (1997) The development of playful
behaviour in piglets from birth to weaning in three farrowing environments. Applied Animal Behaviour Science 55, 37–
49.
Bradshaw, J.W., Pullen, A.J. and Rooney, N.J. (2015) Why do adult dogs ‘play’? Behavioural Processes 110, 82–87.
Brown, S.M., Klaffenböck, M., Nevison, I.M. and Lawrence, A.B. (2015) Evidence for litter differences in play behaviour in
pre-weaned pigs. Applied Animal Behaviour Science 172, 17–25.
Brownlee, A. (1954) Play in domestic cattle in Britain: an analysis of its nature. British Veterinary Journal 110, 48–68.
Burgdorf, J., Panksepp, J. and Moskal, J.R. (2011) Frequency-modulated 50 kHz ultrasonic vocalizations: a tool for
uncovering the molecular substrates of positive affect. Neuroscience and Biobehavioral Reviews 35, 1831–1836.
Burghardt, G.M. (2005) The Genesis of Animal Play: Testing the Limits. MIT Press, Cambridge, Massachusetts.
Byers, J.A. and Walker, C. (1995) Refining the motor training hypothesis for the evolution of play. American Naturalist 146,
25–40.
Calcagnetti, D.C. and Schechter, M.D. (1992) Place conditioning reveals the rewarding aspect of social interaction in
juvenile rats. Physiology & Behavior 51, 667–672.
Cameron, E.Z., Linklater, W.L., Stafford, K.J. and Minot, E.O. (2008) Maternal investment results in better foal condition
through increased play behaviour in horses. Animal Behaviour 76, 1511–1518.
Chaloupková, H., Illmann, G., Bartoš, L. and Špinka, M. (2007) The effect of pre-weaning housing on the play and agonistic
behaviour of domestic pigs. Applied Animal Behaviour Science 103, 25–34.
Chan, W.Y., Cloutier, S. and Newberry, R.C. (2011) Barking pigs: differences in acoustic morphology predict juvenile
responses to alarm calls. Animal Behaviour 82, 767–774.
Cordoni, G. (2009) Social play in captive wolves (Canis lupus): not only an immature affair. Behaviour 146, 1363–1385.
Crowell-Davies, S.L., Houpt, K.A. and Kane, L. (1987) Play development in Welsh pony (Equus caballus) foals. Applied
Dallaire, J.A. (2015) Investigating the functions of rough-and-tumble play in American mink, Neovison vison. Doctoral
dissertation, University of Guelph, Canada.
Dallaire, J.A. and Mason, G.J. (2017) Juvenile rough-and-tumble play predicts adult sexual behaviour in American mink.
Dellmeier, G.R., Friend, T.H. and Gbur, E.E. (1985) Comparison of four methods of calf confinement. II. Behavior. Journal
of Animal Science 60, 1102–1109.
Devinney, B.J., Berman, C.M. and Rasmussen, K.L.R. (2003) Individual differences in response to sibling birth among free-
ranging yearling rhesus monkeys (Macaca mulatta) on Cayo Santiago. Behaviour 140, 899–924.
Diamond, J. and Bond, A.B. (2003) A comparative analysis of social play in birds. Behaviour 140, 1091–1115.
Dobao, M.T., Rodriganez, J. and Silio, L. (1985) Choice of companions in social play in piglets. Applied Animal Behaviour
Science 13, 259–266.
Donaldson, T.M., Newberry, R.C., Špinka, M. and Cloutier, S. (2002) Effects of early play experience on play behaviour of
piglets after weaning. Applied Animal Behaviour Science 79, 221–231.
Duve, L.R., Weary, D.M., Halekoh, U. and Jensen, M.B. (2012) The effects of social contact and milk allowance on
responses to handling, play, and social behavior in young dairy calves. Journal of Dairy Science 95, 6571–6581.
Dwyer, C.M. (2003) Behavioural development in the neonatal lamb: effect of maternal and birth-related factors.
Theriogenology 59, 1027–1050.
Enkel, T., Gholizadeh, D., von Bohlen und Halbach, O., Sanchis-Segura, C., Hurlemann, R., Spanagel, R., Gass, P. and
Vollmayr, B. (2010) Ambiguous-cue interpretation is biased under stress and depression-like states in rats.
Neuropsychopharmacology 35, 1008–1015.

Fagen, R. (1981) Animal Play Behavior. Oxford University Press, Oxford.
Fagen, R. and Fagen, J. (2004) Juvenile survival and benefits of play behaviour in brown bears, Ursus arctos. Evolutionary
Ecology Research 6, 89–102.
Feuerriegel, K. (1997) The role of play and agonistic behaviour for herding of oryx antelopes (Oryx gazella callotis).
Zeitschrift fur Säugetierkunde 62, 52–58.
Graham, K.L. and Burghardt, G.M. (2010) Current perspectives on the biological study of play: signs of progress. The
Quarterly Review of Biology 85, 393–418.
Groos, K. (1898) The Play of Animals. Appleton, New York.
Guard, H.J., Newman, J.D. and Roberts, R.L. (2002) Morphine administration selectively facilitates social play in common
marmosets. Developmental Psychobiology 41, 37–49.
Harcourt, R. (1991) Survivorship costs of play in the South American fur-seal. Animal Behaviour 42, 509–511.
Heinrich, B. and Smolker, R. (1998) Play in common ravens (Corvus corax). In: Bekoff, M. and Byers, J.A. (eds) Animal
Play: Evolutionary, Comparative and Ecological Perspectives. Cambridge University Press, Cambridge, pp. 183–204.
Held, S.D. and Špinka, M. (2011) Animal play and animal welfare. Animal Behaviour 81, 891–899.
Humphreys, A.P. and Einon, D.F. (1981) Play as a reinforcer for maze-learning in juvenile rats. Animal Behaviour 29, 259–
270.
Iwaniuk, A.N., Nelson, J.E. and Pellis, S.M. (2001) Do big-brained animals play more? Comparative analyses of play and
relative brain size in mammals. Journal of Comparative Psychology 115, 29–41.
Jensen, M.B. and Kyhn, R. (2000) Play behaviour in group-housed dairy calves, the effect of space allowance. Applied
Jensen, M.B., Vestergaard, K.S. and Krohn, C.C. (1998) Play behaviour in dairy calves kept in pens: the effect of social
contact and space allowance. Applied Animal Behaviour Science 56, 97–108.
Jones, A.C. and Gosling, S.D. (2005) Temperament and personality in dogs (Canis familiaris): a review and evaluation of
past research. Applied Animal Behaviour Science 95, 1–53.
Krachun, C., Rushen, J. and de Passillé, A.M. (2010) Play behaviour in dairy calves is reduced by weaning and by a low
energy intake. Applied Animal Behaviour Science 122, 71–76.
Kringelbach, M.L. and Berridge, K.C. (2012) The joyful mind. Scientific American 307(2), 40–45.
Lawrence, A. (1987) Consumer demand theory and the assessment of animal welfare. Animal Behaviour 35, 293–295.
Lewis, K.P. and Barton, R.A. (2006) Amygdala size and hypothalamus size predict social play frequency in nonhuman
primates: a comparative analysis using independent contrasts. Journal of Comparative Psychology 120, 31–37.
Leyhausen, P. (1979) Cat Behavior: The Predatory and Social Behaviour of Domestic and Wild Cats. Garland STPM Press,
New York.
Lindblad-Toh, K., Wade, C.M., Mikkelsen, T.S., Karlsson, E.K., Jaffe, D.B., Kamal, M., Clamp, M., Chang, J.L., Kulbokas,
E.J. 3rd, Zody, M.C. et al. (2005) Genome sequence, comparative analysis and haplotype structure of the domestic dog.
Nature 438, 803–819.
Loy, J. (1970) Behavioral responses of free-ranging rhesus monkeys to food shortage. American Journal of Physical
Anthropology 33, 263–271.
Marshall-Pescini, S. and Kaminski, J. (2014) The social dog: history and evolution. In: Kaminski, J. and Marshall-Pescini,
S. (eds) The Social Dog: Behavior and Cognition. Academic Press, San Diego, California, pp. 3–34.
Martin, P. and Bateson, P. (1985) The ontogeny of locomotor play behaviour in the domestic cat. Animal Behaviour 33, 502–
510.
Martin, P. and Caro, T.M. (1985) On the functions of play and its role in behavioral development. Advances in the Study of
Behavior 15, 59–103.
Mason, W.A., Sharpe, L.G. and Saxon, S.V. (1963) Preferential responses of young chimpanzees to food and social rewards.
Psychological Record 13, 341–345.
Melotti, L., Bailoo, J.D., Murphy, E., Burman, O. and Würbel, H. (2014) Play in rats: association across contexts and types,
and analysis of structure. Animal Behavior and Cognition 1, 489–501.
Mendl, M., Burman, O.H. and Paul, E.S. (2010) An integrative and functional framework for the study of animal emotion
and mood. Proceedings of the Royal Society B 277, 2895–2904.
Mintline, E.M., Stewart, M., Rogers, A.R., Cox, N.R., Verkerk, G.A., Stookey, J.M., Webster, J.R. and Tucker, C.B. (2013)
Play behavior as an indicator of animal welfare: disbudding in dairy calves. Applied Animal Behaviour Science 144, 22–
30.
Müller-Schwarze, D., Stagge, B. and Müller-Schwarze, C. (1982) Play behavior: persistence, decrease, and energetic
compensation during food shortage in deer fawns. Science 215, 85–87.
Newberry, R.C., Wood-Gush, D.G.M. and Hall, J.W. (1988) Playful behaviour in piglets. Behavioural Processes 17, 205–
216.
Normansell, J. and Panksepp, J. (1990) Effects of morphine and naloxone on play-rewarded spatial discrimination in
juvenile rats. Developmental Psychobiology 23, 75–83.
Pal, S.K. (2010) Play behaviour during early ontogeny in free-ranging dogs (Canis familiaris). Applied Animal Behaviour
Science 126, 140–153.
Palagi, E. (2006) Social play in bonobos (Pan paniscus) and chimpanzees (Pan troglodytes): implications for natural social
systems and interindividual relationships. American Journal of Physical Anthropology 129, 418–442.
Palagi, E., Burghardt, G.M., Smuts, B., Cordoni, G., Dall’Olio, S., Fouts, H.N., Réháková-Petru, M., Siviy, S.M. and Pellis,
S.M. (2016) Rough-and-tumble play as a window on animal communication. Biological Reviews 91, 311–327.
Peciña, S. and Berridge, K.C. (2000) Opioid eating site in accumbens shell mediates food intake and hedonic ‘liking’: map
based on microinjection Fos plumes. Brain Research 863, 71–86.
Pellis, S. and Pellis, V. (2009) The Playful Brain. Oneworld Publications, Oxford.
Rauw, M.W. (2013) A note on the consistency of a behavioural play marker in piglets. Journal of Animal Science and
Biotechnology 4, 33.
Reimert, I., Bolhuis, J.E., Kemp, B. and Rodenburg, T.B. (2013) Indicators of positive and negative emotions and emotional
contagion in pigs. Physiology & Behavior 109, 42–50.
Reinhardt, V. (1980) Untersuchung zum Sozialverhalten des Rindes: Eine zweijährige Beobachtung an einer halb-wilden
Rinderherde, Bos indicus. Birkhäuser Verlag, Basel.
Reinhardt, V., Mutiso, F.M. and Reinhardt, A. (1978) Social behaviour and social relationships between female and male
prepubertal bovine calves (Bos indicus). Applied Animal Ethology 4, 43–54.
Renouf, D. (1993) Play in a captive breeding colony of harbour seals (Phoca vitulina) constrained by time or by energy.
Journal of Zoology 231, 351–363.
Rolls, E.T. (2005) Emotion Explained. Oxford University Press, Oxford.
Rooney, N.J., Bradshaw, J.W.S. and Robinson, I.H. (2001) Do dogs respond to play signals given by humans? Animal
Behaviour 61, 715–722.
Rooney, N.J., Bradshaw, J.W. and Almey, H. (2004) Attributes of specialist search dogs – a questionnaire survey of UK dog
handlers and trainers. Journal of Forensic Science 49(2), 1–7.
Rushen, J. and de Passillé, A.M. (2014) Locomotor play of veal calves in an arena: are effects of feed level and spatial
restriction mediated by responses to novelty? Applied Animal Behaviour Science 155, 34–41.
Sachs, B.D. and Harris, V.S. (1978) Sex differences and developmental changes in selected juvenile activities (play) of
domestic lambs. Animal Behaviour 26, 678–684.
Sharpe, L.L., Clutton-Brock, T.H., Brotherton, P.N.M., Cameron, E.Z. and Cherry, M.I. (2002) Experimental provisioning
increases play in free-ranging meerkats. Animal Behaviour 64, 113–121.
Smith, E.F.S. (1991) The influence of nutrition and postpartum mating on weaning and subsequent play behaviour of hooded
rats. Animal Behaviour 41, 513–524.
Špinka, M. (2012) Social dimension of emotions and its implication for animal welfare. Applied Animal Behaviour Science
138, 170–181.
Špinka, M., Newberry, R.C. and Bekoff, M. (2001) Play: training for the unexpected. The Quarterly Review of Biology 76,
141–168.
Stamps, J. (1995) Motor learning and the value of familiar space. American Naturalist 146, 41–58.
Svartberg, K. and Forkman, B. (2002) Traits in the domestic dog (Canis familiaris). Applied Animal Behaviour Science 79,
133–157.
Thompson, K.V. (1998) Self assessment in juvenile play. In: Bekoff, M. and Byers, J.A. (eds) Animal Play: Evolutionary,
Comparative and Ecological Perspectives. Cambridge University Press, Cambridge, pp. 183–204.
Thornton, P.D. and Waterman-Pearson, A.E. (2002) Behavioural responses to castration in lambs. Animal Welfare 11, 203–
212.
Vanderschuren, L.J.M.J. (2010) How the brain makes play fun. American Journal of Play 2, 315–337.
Vanderschuren, L.J.M.J., Stein, E.A., Wiegant, V. and Van Ree, J.M. (1995) Social play alters regional brain opioid receptor
binding in juvenile rats. Brain Research 680, 148–156.
Vanderschuren, L.J., Achterberg, E.M. and Trezza, V. (2016) The neurobiology of social play and its rewarding value in rats.
Neuroscience and Biobehavioral Reviews 70, 86–105.
von Frijtag, J.C., Schot, M., van den Bos, R. and Spruijt, B.M. (2002) Individual housing during the play period results in
changed responses to and consequences of a psychosocial stress situation in rats. Developmental Psychobiology 41, 58–
69.
Walker, C. and Byers, J.A. (1991) Heritability of locomotor play in house mice, Mus domesticus. Animal Behaviour 42,
891–897.
Wilson, S.C. and Kleiman, D.G. (1974) Eliciting play: a comparative study. American Zoologist 14, 341–370.
Wood-Gush, D.G.M. and Vestergaard, K. (1991) The seeking of novelty and its relation to play. Animal Behaviour 42, 599–
606.
Wood-Gush, D.G.M., Vestergaard, K. and Petersen, H.V. (1990) The significance of motivation and environment in the
development of exploration in pigs. Biology of Behaviour 15, 39–52.
8 Introduction to Animal Personality
H. Løvlie
8.1 What is Animal Personality?

Behavioural variation is ubiquitous, both between and within species. Moreover, behavioural
differences among individuals of the same species can sometimes be consistent within
individuals across observations. Such among-individual consistency in behaviour is often
recognized by people who interact closely with animals. As a result, an owner of, say, two
dogs can probably quite accurately predict which one will bark at a stranger and which one
will be most scared by fireworks. These observations of individual differences have
similarities to human personality and lay the foundation for what we call animal personality.
The concept of personality comes from psychology, where human personality can be defined
broadly as ‘those characteristics of individuals that describe and account for consistent
patterns of feeling, cognition and behaving’ (Pervin and John, 1997). When defining animal
personality, the focus is on behaviour and we refer to ‘behavioural responses that vary among
individuals of the same species, and that are consistent within individuals over time and or
context’ (Gosling, 2001; Dall et al., 2004; Réale et al., 2007; Carere and Maestripieri, 2013).
Animal personality thus refers to behavioural characteristics describing some aspect of an
individual that is best explained by the identity of the animal.
In biology, research on animal personality has become very popular over the last two
decades. Animal personality is now described in a broad range of species, ranging from
insects to primates (Gosling, 2001; Réale et al., 2007; Carere and Maestripieri, 2013). That
animals have personality is fascinating in its own right. However, intriguingly, animal
personality raises fundamental questions about our view of behaviour. Behaviour has
traditionally been regarded as a very plastic trait. The observation of personality still accepts
behaviour to be plastic, but suggests that there may be limitations to this plasticity generating
consistent behavioural responses. As a consequence, our traditional view of adaptive
behaviour – where individuals should behave optimally and adjust their behavioural
responses to each single situation they face – is somewhat challenged. Observed behavioural
variation among individuals within a species or population was traditionally viewed as non-
adaptive variation around an adaptive population mean of a response. The observations of
consistency in behaviour therefore suggest that animals do not always produce optimal or
adaptive behaviour. This is likely due to the often multiple, underlying mechanisms and
correlated nature of personality. Because animal personality has been demonstrated to have
ecological and evolutionary significance, variation in personality may not be random and
understanding individual variation becomes of interest to researchers. Combined, this has
implications for behavioural research. Animal personality research is thus important because
it highlights an integrated and holistic approach to the study of animal behaviour, which is a
realistic view of behaviour. It also offers improved understanding of how individual variation
in behaviour (including seemingly non-adaptive behaviour) arises as well as how behavioural
consistency can be explained.
Despite the current interest and research effort, the evolution of personality is still not
fully understood and key questions remain: What is the origin of variation and consistency in
behavioural responses, and how are variation and consistency maintained, particularly over
evolutionary time? In other words, why do animals have personality, and why has the one
most ‘perfect’ type not been selected? Before discussing these questions further, the different
terms and model species typically used for research on consistent behavioural responses will
be presented to provide some background to the history and span of the research field.
Thereafter methods and caveats associated with how to measure personality will be
discussed, before some of the current explanations for why animals have personality are
presented. The chapter will be closed by comments on the broad consequences of animals
having personality.
8.2 Terms Used to Describe Animal Personality

For research on consistent individual behavioural differences, several different terms have
been used, such as ‘coping style’, ‘behavioural syndrome’, ‘behavioural type’, ‘temperament’
and ‘animal personality’. These terms are to a large extent interchangeable and are used to
describe individual consistency in behaviour. The different terms are often used separately
for specific disciplines within biology, which in turn commonly use different species as their
primary models.
Human personality
The interest in human personality has probably existed for as long as we have been humans,
with the strict science on human personality being at least over 100 years old (e.g. described
by Galton, 1883). More recently, human personality has been described by the ‘the big five’,
a five-factor model representing five main personality traits: extraversion, neuroticism,
conscientiousness, agreeableness and openness (to experience; McCrae and John, 1992;
Carere and Maestripieri, 2013). More gradients are sometimes used, but these five ones are
able to capture variation in our personality quite holistically. Obtained scores are based on
questionnaires answered by people themselves or on observations of our behaviour.
Individuals can score differentially along all five gradients independently (which we will see
later is different from, for example, ‘coping style’ and ‘behavioural syndromes’). Scoring
high or low on a gradient can be both beneficial and carry costs. For example, scoring high
on extraversion can result in facing more opportunities like access to sexual partners, but can
also increase exposure to harm from taking more risks. Scoring low on extraversion can, on
the other hand, result in increased safety, but missed opportunities. Further, the combination
of scores along all gradients together can affect the benefits and costs of a more extreme
score. For example, scoring low on extraversion and high on neuroticism is often associated
with depression. A high score on extroversion and a low score on neuroticism, in contrast, is
associated with happiness. In other words, the combination of scores on the different
personality gradients together matter and there is no clear ‘best’ personality type (which can
potentially help us understand maintenance of variation in personality over evolutionary
time).
The details of the underlying mechanisms of human personality gradients are generally
unknown. All five gradients have a genetic contribution and heritability explains about half
of the observed variation. Polymorphisms and expression differences of genes of the
monoaminergic systems (see Chapter 3) are often explored and have been found to be linked
to variation in personality. The monoaminergic systems can have fundamental influences on
behaviour in general; for personality, the dopaminergic reward system of the brain is linked
to variation in extraversion, and the serotonergic system to neuroticism. However,
interactions with the environment can modify an individual’s personality score, and
responses can depend on ontogeny, social context, prior experiences, etc. (Carere and
Maestripieri, 2013).
The big five are universal in humans (McCrae et al., 2005) and have similarities with
personality gradients used to describe other animals (Gosling and John, 1999; Gosling,
2001). As we will see, due to species-specific differences, the gradients used often differ and
may not directly be comparable across species.
Although important information of relevance to animal personality can be obtained from
the human literature, one should be aware of the potential constraints this research can have.
For example, self-reported answers may not be objective, and controlled experiments on
humans are not always possible.
Coping styles
Initiated as a model for human stress research, the personality of laboratory rodents (mice
and rats) has been studied extensively under the term ‘coping style’. Coping style refers to
the different ways, or styles, in which individuals cope with stressors and how they differ in
their stress responses. Coping style is formally defined as ‘a coherent set of behavioural and
physiological stress responses which is consistent over time and which is characteristic to a
certain group of individuals’ (Koolhaas et al., 1999, 2010). Coping styles differ somewhat
from descriptions of human and animal personality, and overlap in some aspects more with
how behavioural syndromes are described. This is because responses often cluster and form
one reactive–proactive gradient describing coping styles along one suite of correlated
behavioural and physiological responses. Discrete variation is often observed, which differs
from the often more continuous personality variation seen in other species. This is because
the observed laboratory rodents typically either adopt a more reactive or a more proactive
response, and less often show intermediate responses. Started by seminal early work by
Benus, reactive individuals are often observed to act more on external cues and are more
careful, passive and less aggressive, compared with proactive individuals that act more on
internal cues and are more bold, risk-taking, aggressive and active. Reactive individuals are
more observant of environmental changes and show more plastic responses, whereas
proactive individuals tend to be more rigid and more quickly form routines (Koolhaas et al.,
1999, 2010).
Research on variation in coping styles often links physiology to behavioural variation
and has demonstrated relationships between responses of the hypothalamic–pituitary–adrenal
(HPA) axis (see Chapters 3 and 9), the monoaminergic systems and behavioural responses
(Coppens et al., 2010). Reactivity–proactivity is associated with individual differences in
stress responses and stress responsivity, where proactive copers show a more active response
trying to counteract the stressor (flight–fight), an adrenaline-based response to challenges,
greater sympathetic activation and less HPA activation. Reactive individuals show a more
passive response to a stressful situation, rather trying to avoid the stressor (freeze–hide).
Passive copers have higher HPA response to stressors, a lower sympathetic adrenomedullary
reactivity, a higher humoral immunity and a higher vulnerability to stress-induced illness
(Koolhaas et al., 1999; Coppens et al., 2010; Carere and Maestripieri, 2013).
To behaviourally describe the two coping styles, several assays are used, several showing
similarities with those used in other species. For example, reactive or proactive responses to a
so-called ‘forced swim test’, variation in defensive burying of an electric probe introduced to
an animal’s home cage and aggression measured as attack latency towards a live intruder
describe aspects of coping styles. Variation in anxiety can be described based on behaviour in
an open field test, which is also sometimes used to observe variation in exploration and
activity. Anxiety and fear can be explored in an elevated plus maze. Sociability can be
measured as frequency of contact and amount of play. Responses to alteration of learned
tasks, often through spatial or operant conditioning tests, enable measures of variation in
behavioural plasticity.
One caveat to bear in mind when reading the coping styles literature is that research is
normally carried out in a laboratory on captive individuals, and thus may lack ecological and
evolutionary relevance. Furthermore, commonly only males are used. This is because
females show large variation both between and within individuals caused by hormonal
cycles, variation that can be difficult to measure and control. This can make predictions of
expected variation among females hard.
Temperament
The term ‘temperament’ has its origin in human developmental psychology, where it is
considered closely related to personality. It describes aspects of behaviour that are expected
to emerge early in life and thought to have closer ties to genetics, compared with the case for
personality in a broader sense (Carere and Maestripieri, 2013). Also, when describing
animals, temperament can be somewhat more restricted compared with other definitions of
consistent behavioural variation; ‘the characteristic style of emotional and behavioural

response of an individual in a variety of different situations that is often, but not invariably,
demonstrated very early in life’ (Box, 1999). However, the distinction between temperament
and personality is not consistent in the literature, not even for humans. Definitions of
temperament also tend to have links to emotionality and often include variables like
fearfulness, anxiety or nervousness (Boissy, 1995; Réale et al., 2007).
The term temperament is often used for domesticated animals, both production and pet
animals, although not solely. The questions investigated are often linked to production, for
example aspects of human–animal interaction and responses to handling, or how to better
match pet owners and their pets. In the early 1900s, Ivan Pavlov categorized the temperament
of dogs into four categories according to their observed reactions to stress along the two axes
of ‘passivity’ (active–passive) and ‘extremeness’ (in response to stress; extreme to moderate
response). More contemporary descriptions of dog temperament commonly use
questionnaires filled out by owners covering various aspects of the behaviour of their dogs
(e.g. activity levels, fear, play behaviour; see Fig. 8.1). For this, temperament can also be
subjectively scored by trained personnel in standardized behavioural tests, commonly
including a battery of tests such as responses to approaching strangers, gun shots and
proneness to chasing or hunting. The outcome of both questionnaires and behavioural tests is
used to describe aspects of dog behaviour similar to what is used under other personality
labels, like activity, aggression and reactivity, but also includes terms like self-confidence,
friendliness and responsiveness to training, which are rarely used for other animals.
Differences among breeds of dogs are often in focus and are commonly observed (Jones and
Gosling, 2005; see Fig. 8.1).
Fig. 8.1. Dogs can differ in various personality traits, such as playfulness, and breed
differences are often observed. (Photos courtesy of Tom Løvlie.)
A potential caveat here is that the description of individual behaviour may have reduced
objectivity when based on questionnaires filled out by, for example, owners. There may also
be limitations on results obtained from domesticated animals and in production settings in

terms of implications for the ecology and evolution of personality. On the other hand, we
often spend much time in close proximity with our pet animals, enabling detailed
observations of behaviour not available for wild animals to the same extent.
Behavioural syndromes
Often with an ecological and evolutionary approach, more recent research has focused on
consistency in behaviour under the term ‘behavioural syndromes’. Behavioural syndromes
are defined as ‘suites of correlated behaviours across situations’ (Sih et al., 2004) and
describe behavioural correlations across different functional contexts. If individuals behave
in a consistent way through time or across situations they are said to have a ‘behavioural
type’. Behavioural syndromes can be one-dimensional if individuals exhibit within- and
between-individual consistency along one axis describing behavioural types, or
multidimensional when two or more such axes are correlated.
Research describing behavioural syndromes often uses wild species and in an ecological
context with the aim to understand the selective factors that favour variation in individual
consistency in behaviour and how it depends on environmental conditions. A small fish, the
three-spined stickleback (Gasterosteus aculeatus; see Fig. 8.2), is an important model species
for this research. In the stickleback, it is observed that individuals that are more active are
often also more explorative, bold and aggressive (Sih et al., 2004; Carere and Maestripieri,
2013). This was first demonstrated in landmark work by Huntingford (1976) where fish that
were bolder towards predators were also more aggressive towards conspecifics and more
active in an unfamiliar environment. How this correlational nature of behavioural responses
is formed, and differs within populations, has been in focus. It has for example been shown
that the bold–exploratory–aggressive syndrome observed in sticklebacks develops under
increased risk of predation, while more plasticity in the relationship between these
behavioural responses is observed in populations with less predation (Bell and Sih, 2007;
Dingemanse et al., 2007).
Fig. 8.2. A three-spined stickleback (Gasterosteus aculeatus) in a mirror test. (Photo courtesy
of Hanne Løvlie.)
The assays used to describe variation in behavioural syndromes include measures of

aggression towards live intruders or mirror images (Fig. 8.2). Also, exploration in novel
arenas, exposure to novel objects or response to simulated predator attacks is used. This
allows scoring of variation in risk-taking and the trade-offs that may underlie observed
behavioural variation.
For students reading the behavioural syndrome literature it is worth remembering that
individuals are often wild-caught, thus their ontogenetic background and accumulated
experiences are rarely known. And for many species, there may be a bias towards those
animals that will go into a trap, or those animals that strive in the new laboratory conditions.
Similarly, the sex and age of test animals, together with the selective history of the population
from where they are obtained, may not be known.
Animal personality
Most recently, research on consistent individual differences in behaviour, particularly in an
ecological and evolutionary context, has been using the term ‘animal personality’. This
research field springs mainly from behavioural ecology and is centred on evolutionary theory.
Research is typically hypothesis-driven and based on observations and experiments that
attempt to address how variation in personality traits relates to ecological and social
variables. An important model species for this research is a small passerine bird, the great tit
(Parus major; see Fig. 8.3), starting with work by Verbeek et al. (1994) and carried out in the
wild or on wild-caught individuals brought into the laboratory. Behavioural responses are
often scored along some of the five gradients set out by Réale et al. (2007) (see below, which
differ from the five gradients used to describe human personality, see above), where
boldness–aggression–exploration often correlate positively and form a behavioural syndrome
describing more proactive birds. Animal personality thus has similarities with both coping
style and behavioural syndrome literature. Personality gradients do not, however, need to be
interrelated and, in that respect, is similar to human research that also describes individual
variation along independent variables. Variation in personality relates to ecological factors
that can have fitness consequences, such as dispersal patterns, territory size, feeding
strategies, mate choice, number of offspring produced, etc. The consequences of having a
certain personality may change with fluctuations in the environment (Dingemanse et al.,
2004), offering explanations for how personality variation may be maintained over
evolutionary time (since the ‘best’ personality type changes over time, see below).
Fig. 8.3. The great tit (Parus major) has been an important species in many personality
studies. (Photo courtesy of Tom Løvlie.)
Selection lines of great tits with diverging levels of exploration show different
physiological profiles, suggesting similarities with description of coping styles. Studies have
found associations between exploration propensity and polymorphisms in a dopaminergic
receptor, although this association differs among populations (Kosten et al., 2010).
Tests used are typically aggression towards a stuffed intruder or playback of the song of
unfamiliar individuals, responses to simulated predator attacks or to human interference at
nest boxes, exploration of novel arenas and fear of novel objects.
Caveats for this research can be similar to those described for wild-caught animals; age
and previous history may be unknown, together with potential biases in the animals which
are caught.
8.3 Assays Used and Methodological Considerations

Assays used and traits measured
There are many types of assays and tests used to describe the personality of animals. Overall,
there are historical differences between disciplines and between the species that have
commonly been used. Personality often crystallizes under mild stress and several personality
assays use the context of exposing the individual to some sort of stress. Such tests can be a
barren, novel environment (open field); a ‘furnished’ version of the open field test often
aimed to simulate a more natural, yet novel environment for the focal individual (novel
arena); or exposure to a novel object for measuring neophobia. Other examples are responses
to intrusion (of a conspecific or an electrical probe) and more acutely stressful situations (like
a forced swim test or an elevated maze), startle tests or exposure to predator models. How
easy an animal is to catch or trap (‘trapability’), and how it responds to being caught
(‘docility’), are also sometimes used for wild species. These assays and several more (see
above for examples of more species-specific assays; Fig. 8.4) have been used to describe
personality of a range of different species.
Fig. 8.4. Personality assays used to score individual variation in behavioural responses in the
red junglefowl (Gallus gallus). (a) A female fowl in a novel arena test. Bushes are added with
the aim to obscure the view of the bird and encourage exploration. Bushes, rather than trees,
are used because the fowl is primarily moving on the ground, instead of flying. (b) A male
fowl in a novel object test. Food is provided in a familiar food bowl in front of the novel
object. The addition of the novel object enables measures of the delay until feeding caused by
the addition of the object. The fowl has very good vision, and an object with eyes is chosen
with the aim of obtaining responses of aversion rather than attraction to the object because
the object more resembles a predator than food. (Photos courtesy of Hanne Løvlie.)
In a seminal review of Réale et al. (2007), five personality gradients were suggested to be
of main relevance across species. These are exploration, boldness, activity, aggression and
sociability (and, as can be seen, are not aimed to resemble the ‘big five’ personality gradients
used for humans). Exploration describes an individual’s reaction to a novel situation (e.g.
novel arena, novel object). Boldness describes an individual’s reaction to risky situations that
are not novel (e.g. potential danger). Activity refers to general level of activity and should be
scored in a familiar (i.e. not novel) situation, such as for example the home environment of
the individual or after habituation to a test set-up. Aggression refers to an individual’s
agonistic response to conspecifics (e.g. live or stuffed, playback sounds, own mirror image),
scoring variation in antagonistic interactions initiated by the focal individual. Sociability is
described as the reaction of individuals to presence or absence of conspecifics, excluding
aggression. These five gradients are probably an oversimplification of the diversity of
personality traits, but are a good starting point when designing studies. However, there are
species-specific differences in what behavioural responses an individual shows and what
behaviour that may be biologically meaningful to use in describing personality.
When behavioural responses are obtained, one has to check the independence of
variables. If the measured behaviours are all interrelated, observations describe variation in
only one behavioural syndrome, and not of different personality traits. This discrepancy has
consequences for the interpretation of the observed behaviour, such as the potential
underlying mechanisms and the evolution of the observed variation.
How to measure animal personality
There are several ways to measure personality, which differ between study species and
subdisciplines of biology. The commonly used methods can be summarized as subjective
ratings, where observers who know the subjects well rate the behaviour of the animals
according to questionnaires, or behavioural recordings made by an observer of a focal
individual that either is performing undisturbed behaviour or is responding to an
experimental set-up. Independent of method, the key aspects are to capture both behavioural
variation among individuals and consistency within individuals in their behavioural
responses. Observations of several individuals from a population or species are therefore
needed to test for between-individual differences, with two or more observations of the same
individual to test for within-individual consistency in the behaviour scored.
Repeated observations of the same individual can be obtained by either repeating the use
of the same or similar assays, or the use of different assays capturing functionally similar or
different behaviour (dependent on the focus of the study). Both approaches have potential
problems. For example, repeated use of the same assay can generate altered responses due to
familiarization to the set-up. Familiarization can result in a stronger, or weaker, response on a
later test occasion, and response to familiarization can itself be personality-dependent. An
example of this can be if an individual is too scared to explore a novel arena on its first test
occasion, but does so on a later test occasion. However, if the individual explores the set-up
thoroughly on its first visit it may show no interest in the set-up on a later occasion because
the set-up is already explored. The use of different set-ups, for example the use of different
novel arenas, can potentially reduce this problem. However, the use of different set-ups can
cause other problems. For example, the use of different novel objects can result in the
problem of scoring behaviour belonging to different personality gradients because different
objects can trigger different (and even opposing) responses (e.g. attraction versus aversion to
a novel object).
Another potential issue to consider is that behaviour recorded even in the same test, but
where an animal is forced into a test arena, compared with if it is freely entering it, may not
necessarily correspond to the same personality trait. One should also consider the time
interval between test occasions. There are currently no clear guidelines for how persisting
consistency needs to be to constitute personality. Observations closer in time and in the same
context are intuitively more likely to generate a similar response in an individual, thus a
species-specific and biologically relevant timespan should be considered. Another aspect is
whether individuals are assayed alone or with conspecifics. Many species are social;
however, if several individuals are assayed at the same time, an individual’s response is not
independent of the behaviour of the others and individual responses can be obscured. This
should be considered if a live intruder or opponent is used, for example when scoring
variation in sociability and aggression. In this case, the behaviour of the non-focal individual
can often affect the behaviour of the focal individual.
To reduce these problems, one should carefully consider the biology of the study species,
and perform pilot studies to explore and investigate the responses observed. Moreover,
meaningful – both biologically and statistically – test set-ups should be designed, and the
observed responses evaluated. For example, exposing a night-active animal that does not use
vision as its primary sense to colourful novel objects may not generate biologically relevant
responses. Furthermore, if the situational strength is too strong, all individuals may be scared
and show similar responses with very little variation. A bias caused by a ‘floor effect’ can
occur when the first unit of recording is so broad that existing differences between
individuals are obscured. Cutting off the data, for example at too early a time point, can
generate a ‘ceiling effect’ where many individuals all receive the highest score, which also
can cause a bias. There is an ongoing discussion in the research community regarding how to
design personality assays, how to describe personality traits and the importance of evaluating
if the method chosen really measures the targeted trait in the study species (e.g. Carter et al.,
2013; Dall and Griffith, 2014). One should preferentially use multiple tests for each trait to
evaluate if the predicted correlations occur across all, which in turn can confirm the function
of the response one aims to capture by a specific assay.
Recorded traits may correlate because they reflect the same underlying trait (e.g. if both
boldness and exploration are linked to general activity), or as an artefact caused by the design
or method used to record them. By adopting an experimental and hypothesis-driven approach
these problems can be evaluated and potentially reduced. Other relevant issues discussed in
the research community are how to reduce the confusion of the use of the same test to score
different behaviour or personality gradients (e.g. open field to measure boldness, anxiety or
exploration), the use of different tests to describe the same personality trait (e.g. open field,
novel object, novel arena and startle tests to describe variation in boldness), labelling of
different behaviours as the same personality trait (e.g. handling a novel object and movement
in a familiar arena both labelled as exploration) and labelling of the same behavioural
response as different personality gradients (e.g. a short latency to start moving in an open
field labelled exploration or boldness). The latter two are called the jingle-jangle fallacy,
where ‘jingle’ refers to when a single label is used to describe functionally different traits
measured with different tests and ‘jangle’ when more than one label describes the same trait.
There is yet no clear consensus regarding these issues, but readers and researchers interested
in the field are urged to follow the discussion carefully (e.g. Carter et al., 2013; Dall and
Griffith, 2014). In general, a clear description of what has been tested and observed
(preferentially both why and how) is needed to improve clarity of the research carried out.
Finally, one should bear in mind that personality is very much a statistical phenomenon.
This is because the way we measure whether a behavioural response describes an
individual’s personality is based on whether two or more behavioural measures are relatively
consistent within or among the observed individuals. This can be done by correlating
responses obtained in observation 1 with responses obtained in observation 2. Alternatively,
repeatability can be used to estimate how similar responses of one individual are compared
with responses of other individuals in the population. Consequently, one observation of a
response is not sufficient to describe variation in personality among individuals of a
population.
Some general knowledge on how ‘significant’ correlations are generated is also needed.
One problem to be aware of is that potential mean-level differences between subgroups
within the data (e.g. sex, age, infection status) can generate statistically significant
correlations with no relevance for personality (Fig. 8.5). There is an increasing interest and
development of statistical tools and approaches concerning how to best measure personality.
Students interested in animal personality research should familiarize themselves with the
current literature regarding this.
Fig. 8.5. A hypothetical example of how a significant correlation can be generated by

diverging responses of two subgroups in the data (here blue could symbolize males, red
females). Within each group (blue and red, solid lines), there are no relationships between the
responses obtained in trial 1 and trial 2. However, due to the mean difference between the
groups, a correlation between the two clusters of data generated by the subgroups gives the
overall apparent consistency in responses on the population level (black, dashed line), even if
individuals do not in fact show consistency in their responses across trials.
8.4 Why do Animals have Personality?

Currently, there are two main, non-exclusive sets of explanations for why animals differ in
their personality. The first set of explanations states that personality is caused by constraints
to behavioural plasticity. The second suggests that different personality types are alternative,
adaptive strategies. Many aspects of these suggested explanations still remain to be
investigated empirically.
Origin of personality
Consistency in behaviour over time, or correlated behaviour across contexts, can be caused
by constraints to behavioural plasticity. This can potentially be explained by pleiotropy (e.g.
caused by genes or hormones), covariance among traits (e.g. caused by life-history trade-
offs) or other factors (internal or external) that generate consistency in behavioural responses
(e.g. stable ‘states’). Behaviour, like most traits, is affected by genetic and environmental
factors, and the interaction between the two. The same is true for personality. In principle,
any genetic or environmental effect that causes consistency in behaviour has the potential to
explain (at least to some extent) the existence of personality.
The genetic contribution to behavioural variation is commonly polygenic, and many
genes contribute with small effects. This is likely to be the case for personality traits as well,
and the genetics of personality may be complex due to multiple contributing effects and
complex interrelated relationships among traits. When the heritability of animal personality is
estimated, this is often about 0.26, with heritability of personality observed in wild
populations being on average 0.36 and in domestic species being lower, 0.24 (Carere and
Maestripieri, 2013). The conclusion from this is that genetic factors contribute to personality.
The association between animal personality and specific genes has been explored mainly by
focusing on candidate genes from the monoaminergic systems. For example, polymorphism
in a dopamine receptor gene can explain up to about 5% of the variation in exploration
among great tits (Kosten et al., 2010), similarly to the approximately 10% of variation
accounted for in novelty-seeking in humans (Savitz and Ramesar, 2004). Knocking out the
serotonin transporter gene alters personality in mice and produces a stress-prone phenotype
with anxious behaviour and exaggerated responses to stressors. The potential role for
pleiotropic effects of genes affecting personality has been shown, for example, in barn owls
(Tyto alba), where individuals with one form of a gene from the melanocortin system have
darker plumage and are also more active, aggressive and bolder (Ducrest et al., 2008).
Personality of offspring can also be affected by indirect genetic effects, for example through
parental effects by altered levels of hormones in eggs, or differential investment in offspring.
Pleiotropic effects can also be explained by emotional and cognitive systems affecting
several traits simultaneously. For example, fearfulness can affect an individual’s responses to
a range of potentially threatening situations and exposure to habitats, objects and
conspecifics (Boissy, 1995).
Underlying life-history trade-offs are often used to explain observe personality variation.
This is in part based on the expected trade-off between growth and mortality (Stamps, 2007).
Individuals that grow fast are expected to die earlier and thus should have been selected to
take more risks to ensure that they reproduce before dying. Individuals that grow slowly, on
the other hand, are predicted to be more careful to ensure that they survive until they have
reproduced and secured their fitness. Boldness and aggression both represent ‘high risk, high
gain’ lifestyles. Variation in metabolism can sometimes explain why animals grow at
different speeds, setting individuals off on different life-history trajectories from which it
may be costly to deviate. This integrated view is called the pace-of-life syndrome (Réale et
al., 2010), which integrates life history, physiology and behaviour, and where variation in
metabolism is suggested to have pleiotropic effects and explain variation in personality. With
different paces of life, individuals face different resource-acquisition or assimilation trade-
offs relevant for evolution and maintenance of personality. Several of these explanations
suggest a rigid architecture of behaviour, which can be explained by more plastic architecture
not being evolutionarily possible or less advantageous (e.g. in an unpredictable or noisy
environment; Carere and Maestripieri, 2013).
Another set of explanations for individual variation and consistency in behaviour is based
on differences among individuals in ‘state’ and state-dependent behaviour. If states (such as
body size, energy reserves) change more slowly than behaviour, they can cause short-term
stability in state-dependent behaviour (Dall et al., 2004). There are multiple causes for why
individuals initially may differ in state. Examples can be alternative mating tactics (e.g.
territoriality versus sneaking), developmental stage (young versus old) or life-history trade-
offs. These individual differences can change with time. The initial difference can lead to
specialization if there are benefits of specializing, or costs of switching between behavioural
types, or positive feedback loops between behaviour and the underlying state variable.
Maintenance of personality
In an evolutionary framework, several processes can explain maintenance of personality
differences. Some of those are linked to explanations of how personality differences arise in
the first place. They often entail that different personality types have overall equal fitness. On
the other hand, if there is one ‘best type’ we would expect directional selection to favour this
one, and the frequencies of different personality types should change over time. In general,
several types of selection can potentially maintain polymorphism within a population,
collectively referred to as balancing selection (e.g. temporal or spatial environmental
variation, frequency-dependent selection, heterozygous advantage).
The influence of environmental heterogeneity on the success of different personality
types has been demonstrated in several ways. Environmental variation along a spatial or
temporal scale can explain maintenance of personality variation because the optimal
phenotype should vary in space (e.g. differs among territories) or in time (e.g. over a season
or years). For example, in great tits, survival of adult birds depends on their personality and
the availability of food in a year, and shows opposing patterns for males and females. More
explorative females survive better in poorer years (i.e. years with low food availability),
while explorative males survive better in good years. The reproductive success of these birds
also depends on interactions with their personality, since extreme personality types (the least
or most explorative) recruited most offspring in good years (Dingemanse et al., 2004).
Personality can also influence other fitness-related traits. For example, in guppies,
personality affects mate choice patterns and females prefer bolder males (Godin and
Dugatkin, 1996). Assortative (the preference for a similar type) or disassortative (where
different personality types mate) mating patterns can to some extent explain why different
types exist at the same time. However, a meta-analysis on links between variation in
personality and fitness (Smith and Blumstein, 2008) shows that relationships among
personality and fitness-related traits are often complex. For example, bolder individuals may
have an increased reproductive success, but survive less well. This suggests that there are
life-history trade-offs associated with personality variation, with fitness consequences.
8.5 Consequences of Animal Personality
That animals can have personality has consequences on how we as researchers view
behaviour and can explain apparently suboptimal behaviour. For an individual with a certain
behavioural type, the consequences can affect all aspects of life. In wild animals, different
personality types may have different foraging strategies, territory choice, group position,
dispersal patterns, mate choice, parental care strategies, survival and fitness. In applied
ethology, the personality of an individual can have implications, for example, on human–
animal interactions, animal welfare, success of conservation programmes, production
efficiency and population consequences of hunting (McDougall et al., 2006; Biro and
Stamps, 2008). This can be linked to personality types varying in their proneness to stress
and responses to novelty.
Interactions with humans can affect personality types in different ways. For example,
animal personality can affect human–animal interactions such as the handling of production
animals by farmers, the performance of working dogs, or pet owners and the matching of
suitable pets (e.g. riders and their horses). Variation in personality can in several manners
affect various aspects of conservation. For example, different personality types may vary in
their response to captivity and may differ in their ability to colonize, establish and spread
after release due to different vulnerabilities to human activity, novel environments and
exposure to predation. Reintroduction programmes unfortunately often suffer from very high
mortality of reintroduced individuals (see Chapter 1). In captive-bred swift foxes (Vulpes
velox), bolder foxes were more likely to be found dead 6 months following release from
captivity (Bremner-Harrison et al., 2004). Captivity may also select for more docile and
tolerant individuals, which may not be well adapted to a life in the wild. Moreover, due to
differential reproduction, the release of certain types may affect the long-term success of
reintroduction and the establishment of new populations. This in turn may lead to loss of
variation, with the result that populations may thus have reduced ability to adapt to changing
environmental conditions and have increased risk of extinction. Overall, improved
understanding of which personality types may thrive in a zoo, or be successful in
reintroduction programmes, can affect the success of such projects.
Under commercial conditions, personality may be important for the production capacity
of animals. With respect to fish in aquaculture, personality can affect feeding strategies,
growth, stress sensitivity and social interactions. Production conditions can also affect the
personality types observed in a population due to artificial selection pressures favouring
certain types. In farmed fish, domestication has selected for fast, bold individuals. This is
likely also happening in populations that are harvested, where there is expected to be a bias
in which personality types are targeted by humans and also which types are caught. This has
been explored in fish, suggesting fishing to result in selection for smaller and more careful
types (Biro and Stamps, 2008).
Taken together, it is clear that animals, like humans, can have personality and behave
consistently over time or context. This has been demonstrated in all taxa investigated,
sometimes under labels other than animal personality. Despite large research efforts, it is still
not fully understood why animals have personality. Nevertheless, personality differences can
have significant implications and affect our view of animal behaviour. For individual
animals, personality differences can have consequences directly or indirectly related to our
interactions with them and for their success in captivity and in the wild. This in turn can have
consequences for populations and ecosystems.
References
Bell, A.M. and Sih, A. (2007) Exposure to predation generates personality in threespined sticklebacks (Gasterosteus
aculeatus). Ecology Letters 10, 828–834. Available at: http://dx.doi.org/10.1111/j.1461-0248.2007.01081.x (accessed 11
April 2017).
Biro, P.A. and Stamps, J.A. (2008) Are animal personality traits linked to life-history productivity? Trends in Ecology &
Evolution 23, 361–368. Available at: http://dx.doi.org/10.1016/j.tree.2008.04.003 (accessed 11 April 2017).
Boissy, A. (1995) Fear and fearfulness in animals. The Quarterly Review of Biology 70, 165–191. Available at:
http://dx.doi.org/10.1086/418981 (accessed 11 April 2017).
Box, H.O. (1999) Temperament and socially mediated learning among primates. In: Box, H.O. and Gibson, K.R. (eds)
Mammalian Social Learning: Comparative and Ecological Perspectives. Cambridge University Press, Cambridge, pp.
33–56.
Bremner-Harrison, S., Prodohl, P.A. and Elwood, R.W. (2004) Behavioural trait assessment as a release criterion: boldness
predicts early death in a reintroduction programme of captive-breed swift fox (Vulpes velox). Animal Conservation 7,
313–320. Available at: http://dx.doi.org/10.1017/S1367943004001490 (accessed 11 April 2017).
Carere, C. and Maestripieri, D. (eds) (2013) Animal Personalities. University of Chicago Press, Chicago, Illinois.
Carter, A.J., Feeney, W.E., Marshall, H.H., Cowlishaw, G. and Heinsohn, R. (2013) Animal personality: what are
behavioural ecologists measuring? Biological Reviews of the Cambridge Philosophical Society 88, 465–475. Available
at: http://dx.doi.org/10.1111/brv.12007 (accessed 11 April 2017).
Coppens, C.M., de Boer, S.F. and Koolhaas, J.M. (2010) Coping styles and behavioural flexibility: towards underlying
mechanisms. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 365, 4021–4028.
Available at: http://dx.doi.org/10.1098/rstb.2010.0217 (accessed 11 April 2017).
Dall, S.R.X. and Griffith, S.C. (2014) An empiricist’s guide to animal personality variation in ecology and evolution.
Frontiers in Ecology and Evolution 2, 1–7. Available at: http://dx.doi.org/10.3389/fevo.2014.00003 (accessed 11 April
2017).
Dall, S.R.X., Houston, A.I. and McNamara, J.M. (2004) The behavioural ecology of personality: consistent individual
differences from an adaptive perspective. Ecology Letters 7, 734–739. Available at: http://dx.doi.org/10.1111/j.1461-
0248.2004.00618.x (accessed 11 April 2017).
Dingemanse, N.J., Both, C., Drent, P.J. and Tinbergen, J.M. (2004) Fitness consequences of avian personalities in a
fluctuating environment. Proceedings of the Royal Society of London B. Biological Sciences 271, 847–852. Available at:
http://dx.doi.org/10.1098/rspb.2004.2680 (accessed 11 April 2017).
Dingemanse, N.J., Wright, J., Kazem, A.J., Thomas, D.K., Hickling, R. and Dawnay, N. (2007) Behavioural syndromes
differ predictably between 12 populations of three-spined stickleback. Journal of Animal Ecology 76, 1128–1138.
Available at: http://dx.doi.org/10.1111/j.1365-2656.2007.01284.x (accessed 11 April 2017).
Ducrest, A.L., Keller, L. and Roulin, A. (2008) Pleiotropy in the melanocortin system, coloration and behavioral syndromes.
Trends in Ecology & Evolution 23, 502–510. Available at: http://dx.doi.org/10.1016/j.tree.2008.06.001 (accessed 11
April 2017).
Galton, F. (1883) Inquiries into Human Faculty and Its Development. J.K. Dent and Company, London.
Godin, J.-G.J. and Dugatkin, L.A. (1996) Female mating preference for bold males in the guppy, Poecilia reticulate.
Proceedings of the National Academy of Sciences USA 93, 10262–10267.
Gosling, S.D. (2001) From mice to men: what can we learn about personality from animal research? Psychological Bulletin
127, 45–86. Available at: http://dx.doi.org/10.1037//0033-2909.127.1.45 (accessed 11 April 2017).
Gosling, S.D. and John, O.P. (1999) Personality dimensions in nonhuman animals: a cross-species review. Current
Directions in Psychological Science 8, 69–75. Available at: http://dx.doi.org/10.1111/1467-8721.00017 (accessed 11
April 2017).
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. Available at:
http://dx.doi.org/10.1016/S0003-3472(76)80034-6 (accessed 11 April 2017).
Jones, A.C. and Gosling, S.D. (2005) Temperament and personality in dogs (Canis familiaris): a review and evaluation of
past research. Applied Animal Behaviour Science 95, 1–53. Available at:
http://dx.doi.org/10.1016/j.applanim.2005.04.008 (accessed 11 April 2017).
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. and
Blokhuis, H.J. (1999) Coping styles in animals: current status in behavior and stress-physiology. Neuroscience and
Biobehavioral Reviews 23, 925–935. Available at: http://dx.doi.org/10.1016/S0149-7634(99)00026-3 (accessed 11 April
2017).
Koolhaas, J.M., de Boer, S.F., Coppens, C.M. and Buwalda, B. (2010) Neuroendocrinology of coping styles: towards
understanding the biology of individual variation. Frontiers in Neuroendocrinology 31, 307–321. Available at:
http://dx.doi.org/10.1016/j.yfrne.2010.04.001 (accessed 11 April 2017).
Kosten, P., Mueller, J.C., Hermannstädter, C., Bouwman, K.M., Dingemanse, N.J., Drent, P.J., Liedvogel, M., Matthysen, E.,
van Oers, K., van Overveld, T., Patrick, S.C., Quinn, J.L., Sheldon, B.C., Tinbergen, J.M. and Kempenaers, B. (2010)
Association between DRD4 gene polymorphism and personality variation in great tits: a test across four wild
populations. Molecular Ecology 19, 832–843. Available at: http://dx.doi.org/10.1111/j.1365-294X.2009.04518.x
(accessed 11 April 2017).
McCrae, R.R. and John, O.P. (1992) An introduction to the five-factor model and its applications. Journal of Personality 60,
175–215. Available at: http://dx.doi.org/10.1111/j.1467-6494.1992.tb00970.x (accessed 11 April 2017).
McCrae, R.R., Terracciano, A. and 78 members of the Personality Profiles of Cultures Project (2005) Universal features of
personality traits from the observer’s perspective: data from 50 cultures. Journal of Personality and Social Psychology
88, 547–561. Available at: http://dx.doi.org/10.1037/0022-3514.88.3.547 (accessed 19 April 2017).
McDougall, P.T., Réale, D., Sol, D. and Reader, S.M. (2006) Wildlife conservation and animal temperament: causes and
consequences of evolutionary change for captive, reintroduced, and wild populations. Animal Conservation 9, 39–48.
Available at: http://dx.doi.org/10.1111/j.1469-1795.2005.00004.x (accessed 11 April 2017).
Pervin, L.A. and John, O.P. (1997) Personality: Theory and Research, 10th edn. Wiley, New York.
Réale, D., Reader, S.M., Sol, D., McDougall, P.T. and Dingemanse, N.J. (2007) Integrating animal temperament within
ecology and evolution. Biological Reviews of the Cambridge Philosophical Society 82, 291–318. Available at:
http://dx.doi.org/10.1111/j.1469-185X.2007.00010.x (accessed 11 April 2017).
Réale, D., Garant, D., Humphries, M.M., Bergeron, P., Careau, V. and Montiglio, P.O. (2010) Personality and emergence of
the pace-of-life syndrome concept at the population level. Philosophical Transactions of the Royal Society of London.
Series B, Biological Sciences 365, 4051–4063. Available at: http://dx.doi.org/10.1098/rstb.2010.0208 (accessed 11 April
2017).
Savitz, J.B. and Ramesar, R.S. (2004) Genetic variants implicated in personality: a review of the more promising candidates.
American Journal of Medical Genetics Part B – Neuropsychiatric Genetics 131B, 20–32. Available at:
http://dx.doi.org/10.1002/ajmg.b.20155 (accessed 19 April 2017).
Sih, A., Bell, A.M., Johnson, J.C. and Ziemba, R.E. (2004) Behavioral syndromes: an integrative overview. The Quarterly
Review of Biology 79, 241–558. Available at: http://dx.doi.org/10.1086/422893 (accessed 11 April 2017).
Smith, B.R. and Blumstein, D.T. (2008) Fitness consequences of personality: a meta-analysis. Behavioral Ecology 19, 448–
455. Available at: http://dx.doi.org/10.1093/beheco/arm144 (accessed 11 April 2017).
Stamps, J.A. (2007) Growth–mortality tradeoffs and ‘personality traits’ in animals. Ecology Letters 10, 355–363. Available
at: http://dx.doi.org/10.1111/j.1461-0248.2007.01034.x (accessed 11 April 2017).
Verbeek, M.E.M., Drent, P.J. and Wiepkema, P.R. (1994) Consistent individual differences in early exploratory behaviour of
male great tits. Animal Behaviour 48, 1113–1121. Available at: http://dx.doi.org/10.1006/anbe.1994.1344 (accessed 11
April 2017).
9 Abnormal Behaviour, Stress and Welfare
L. Keeling and P. Jensen
9.1 Introduction
When humans first started keeping animals in captivity, their concerns were probably
restricted to how the animals could be prevented from escaping and how they could be kept
alive and healthy. Later, we started to be concerned about production and how we could get
our farm animals to produce more milk and eggs, grow faster and have more offspring; or, in
the case of our sport and companion animals, run fast and look more beautiful. During these
times behavioural disturbances and stress were only a problem in so far as they affected
health and performance, and good health was usually considered synonymous with good
welfare. It wasn’t until the 1960s and 1970s that we started to question these assumptions and
to consider behaviour an important component of welfare. This chapter deals with the
important aspects of behavioural disorders and stress and ends with a short review of current
theories on what animal welfare is and how to measure it. Other books taking up this subject
in more detail include Appleby et al. (2011) and Broom and Fraser (2007).
9.2 Behavioural Disorders

Normal and abnormal behaviour
One way to define normal – or natural – behaviour of an animal could be the behaviour that
has developed during evolutionary adaptation. This would of course include any learnt
behaviour that serves the function of promoting the health, survival and reproduction of an
animal in a certain environment. For domesticated animals, there are three important sources
of information about normal behaviour: (i) the behaviour of the wild ancestors; (ii) the
behaviour of feral animals (i.e. domestic animals which have escaped or been released and
have adapted to a life without dependence on humans); and (iii) the behaviour of domestic
animals when placed (usually by researchers) in environments similar to those of the
ancestors (see Fig. 9.1).
Fig. 9.1. When pigs are released into nature, they will form groups of closely related sows,
just like wild boars, and spend most of their time in pursuit of food, mainly by rooting.
Since learning and adaptation will modify the behaviour of any individual, there will
always be a range of behavioural profiles that can be considered normal. Sometimes, people
therefore dismiss the very concept of normal behaviour and consider any deviations to be
adaptive. However, in spite of a large variation, there are aspects of the behaviour of any
animal that are typical for the species, and there may be aspects which are outside the range
usually observed in members of the species in non-captive situations. It is therefore necessary

to try to understand which behaviour patterns are species-typical and which ones can be
regarded as not normal for an animal of a given species, sex and age.
To avoid semantic confusion, it is important to remember that, in this context, an
abnormal behaviour may be very common among individuals of a species. For example,
since most poultry in the world are kept in captivity, and most of the egg-laying hens are kept
in cages, any abnormal behaviour caused by the cage environment is going to be shown by a
majority of the hens. Frequency of occurrence, therefore, is no synonym for normality with
respect to behaviour. When we consider abnormal behaviour and behavioural disturbances,
we must always remember that the norm is the behaviour as it has evolved in the natural
habitats of the species.
Stereotypies
Stereotypies are one particular form of abnormal behaviour. They can be described as
movements which are repeated in the same manner over and over again and may occupy a
substantial part of the time an animal is awake. They are sometimes conspicuous, but
sometimes less obvious to a casual observer. Stereotypies have been defined as movement
patterns which are unvarying, repetitive and lack an obvious goal or function (see Chapter 4
for some further details on this).
Stereotypies are thought to develop when an animal is prevented from performing certain
normal and strongly motivated behaviour patterns, for example those related to foraging and
exploration. A restrictive and stimulus-poor environment has been shown to affect the
probability of an animal developing stereotyped behaviour, although in some cases the
crucial factor appears to be whether or not the animal has the opportunity to forage. For
example, experiments with tethered and loose-housed sows found that animals on a restricted
diet developed more stereotyped behaviour patterns, such as sham-chewing, bar-biting and
chain-chewing, than sows on the higher ration, irrespective of housing condition (see Fig.
9.2). Differences between animals in the way they eat are reflected in the range of
stereotypies that are seen, such as chewing stereotypies in pigs, tongue-rolling stereotypies in
cattle, pecking stereotypies in chickens and crib-biting in horses. In general, it appears that
grazing and omnivorous animals, which are likely to be motivated to spend much time
feeding, tend to mainly develop oral stereotypies (eg, biting, chewing, pecking), whereas
predators, which are likely to be motivated for moving and chasing, more often develop
locomotory stereotypies (e.g. pacing, route-tracing).
Fig. 9.2. Sows which are confined or tethered during pregnancy spend a large part of their
time performing stereotyped bar-biting.
Efforts to investigate the link between the performance of stereotypic behaviour and
welfare, however, have proved to be problematic, with some researchers finding a good
correlation and others no relationship between stereotypies and other welfare measures. It is
now thought that there are different phases in the development of stereotypies. At first the
behaviour is context-specific, but it is later performed in an ever-increasing range of
situations until, when fully established, it continues even when the animal is moved to a more
enriched environment. Stereotyped behaviour is evidence therefore that the animal has had
reduced welfare at some time, but it does not necessarily imply that its welfare is poor now.
However, stereotypies remain important indicators that the environment is not providing
sufficient possibilities for the animals to perform their normal behaviour. The number of
animals showing stereotypies in a particular environment and the proportion of time they
spend performing stereotypies are therefore important welfare indicators.
Cannibalism
Although dramatic and something most people would want to prevent, cannibalism in
animals is not necessarily an abnormal behaviour. It occurs in the wild, but only under
conditions of adversity, when the benefits of removing competitors from an area and the
nutritional benefits of eating them outweigh the potential risks. Of the species that are dealt
with in this book, cannibalism is most common in pigs, poultry and rodents. Studies in
behavioural ecology have shown that, in nature, it is usually larger individuals cannibalizing
smaller ones, often young, or healthy individuals cannibalizing weaker ones.
However, in captivity, cannibalism may develop in poor environments, for example
where the animals live in crowded or impoverished housing systems. It is a common
misconception that the behaviour is a redirected form of aggression; mostly aggressive
motivation has nothing to do with cannibalism. In chickens, the most serious form of
cannibalism is when pecks are directed at the cloaca, which may perhaps sometimes be a
redirected investigative behaviour. There is evidence from outbreaks in commercial flocks of
laying hens that it is the large birds that become cannibals and the small birds that become
victims. But the effects are subtle, and a chance peck resulting in skin damage followed by
learning for the reward of blood probably also plays a role (see Fig. 9.3).
Fig. 9.3. Feather-pecking in laying hens is a serious welfare problem that may sometimes
develop into cannibalism.
Tail-biting in fattening pigs is more common when the pens are small, crowded and
lacking in appropriate stimuli. Giving the pigs sufficient access to straw for exploration and
chewing is a relatively good way of preventing the behaviour. It may also have some chance
component, although in pigs given the opportunity to chew on blood-soaked ‘imitation tails’
(tassels of rope hung in the pen), nutritional deficiencies, and in particular salt deficiencies,
were found to increase the incidence of chewing.
That there is some nutritional basis for cannibalism would be in accordance with
outbreaks in the wild, which often occur in times of food shortage. In rodents, pup-
cannibalism is common. Here the cannibalism is by the parent, which can sometimes be
adaptive in nature. Often the victim would have had little chance of surviving in any case.
However, it may also be caused by stress. When performed by an unrelated individual in
nature, the benefits of infanticide (i.e. eating the young) may be increased food acquisition,
but they can also be decreased resource competition or increased male reproductive success.
Despite the work on cannibalism from a behavioural ecology perspective, we have only a
poor understanding of all aspects of cannibalism in farm animals. In the main it seems to be
multifactorial, with factors such as barrenness of the environment (e.g. lack of straw for pigs
or litter for poultry), increasing stocking density and group size all increasing the risk for
cannibalism. There has been some success in reducing it by genetic selection, but whether
this is a direct effect on cannibalistic behaviour or an indirect effect on some other linked trait
is unclear.
Abnormal aggression
Aggression is used in the establishment of dominance relationships, and so the level is low
and rarely a problem in long-established groups. Aggression therefore is a normal behaviour,
but in domestic animals we regard it as undesirable if it reaches proportions where it causes
injury or practical problems. Unfortunately, it is often the way we house or manage animals
that results in high levels of aggression. For example, we often mix animals, or keep them in
groups that would be unlikely to occur under natural conditions.
Horses, as presented Chapter 12, live in bands in the wild. However, at a typical riding
stable, horses may be moved from one paddock to another on an almost daily basis and may
even be part-time members, being outside only during the daytime in fine weather.
Comparisons show that the level of aggression in these groups is much higher than in groups
in the wild. The most common reason for outbreaks of aggression in pigs is also the mixing
with unfamiliar animals. Attempts to prevent newly mixed animals from fighting have not
been very successful, not even by using drugs.
Stocking density also influences aggression, and an aggressive interaction may persist if
a submissive individual is not able to signal its submission effectively because there is
insufficient space for it to remove itself from the aggressor. This has been shown to be a
contributing factor in aggression among fattening pigs or sows.
Finally, aggressive problems in dogs are widely reported. Here the problem may be
because of the high level, but it can equally well be that the owner perceives even a low level
of aggression in their pet as undesirable, for example when there are small children in the
family.
9.3 Stress
The concept of stress was developed over a number of decades in the middle of the 20th
century. The leading persons were two physiologists, Walter B. Cannon and Hans Selye, who
both studied bodily reaction patterns to potentially harmful stimuli. The reactions appeared to
be more or less the same regardless of which type of threat they studied. Their combined
findings gave rise to what we might call the standard stress model (Toates, 1995).
According to this standard model, a threatening stimulus (often referred to as a ‘stressor’)
evokes mainly two sets of physiological responses. One is the activation of the sympathetic
branch of the autonomic nervous system causing, for example, increased heart rate, increased
blood pressure, decreased gastrointestinal activity and increased secretion of catecholamines
(adrenaline and noradrenaline) from the medulla of the adrenal cortex. Catecholamines in
turn emphasize and prolong similar effects on the circulation and intestines. The other set of
reactions consists of an increased secretion of the hormone ACTH (adrenocorticotropic
hormone) from the pituitary (regulated by CRH, corticotropin-releasing hormone), which
stimulates secretion of corticosteroids (e.g. cortisol and corticosterone) from the adrenal
cortex. Corticosteroids are mainly metabolic hormones, stimulating the recruitment of energy
in the form of glucose and fatty acids from depots in the body (see Fig. 9.4).
Fig. 9.4. The ‘standard stress model’, showing the two main physiological pathways: the
hypothalamic–pituitary–adrenal cortex (HPA) axis and the sympathetic–adrenal medulla
(SA) axis (ACTH, adrenocorticotropic hormone).
Prolonged activation of this system was found to be harmful to the organism in a number
of ways, causing syndromes such as stomach ulcerations, cardiovascular diseases and
alterations in the efficiency of the immune defence. The systems have been found to be
extremely complex and there are also several other systems involved, but basically stress in
this standard model is used as a concept describing the adverse bodily effects of certain
stimuli.
Why should ethologists bother about this? There are a number of reasons. One is that the
stimuli evoking the responses, the stressors, have often been described as ‘psychological’.
For example, a stressor could be to live in a crowded environment, or to be exposed to
sudden and unexpected novelties, both of which are of interest to applied ethologists.
Another reason is that the last few decades of stress research has shown that the standard
model is not complete – stress responses are intimately connected with how animals perceive
stressors, and to what extent they can act in a functional way in response to them. Hence,
stress responses depend on the mechanisms controlling normal behaviour. Stress is therefore
an important link between environment, behavioural disorders and disease.
The concept of allostasis, which means stability through change, is increasingly being
used in the context of stress and animal welfare. This approach suggests that changes in body
functioning following exposure to a given stressor actually help an animal anticipate further
challenges and so respond to them. In this way, the animal can better adapt to its
environment.
Predictability and controllability – key concepts in stress responses
The pathological consequences of a certain stressor have long been known to depend on how
the stressor is perceived. The crucial experiment demonstrating the salient factors in
perception of threats was performed by Weiss (1971). He exposed rats to well-controlled
stressors in the form of electric shocks, delivered via an electrode attached to the tail. In each
test, Weiss had three rats in single cages. Two of the rats were yoked in the same electric
circuitry, so when one rat was shocked, the other received exactly the same sensation. The
third rat was a control, which never received any shocks. One of the yoked rats was then
given a predictive cue, in the form of a sound signal going off a few seconds before each
shock.
The possibility to predict the shock significantly decreased the development of stomach
ulcers, even though both yoked rats received exactly the same physical pain experience.
When Weiss allowed one rat to exert some control over the shocks, the effect was even
stronger. The rat could turn the current off once it had started (by turning a wheel), and
thereby shorten the shock for itself and its unknown yoked partner, or postpone the shock
somewhat. Again, ulcerations decreased dramatically, to the extent that the rats with
possibilities to predict and control had only small differences in ulcerations compared with
the animals that never received any shock at all.
This experiment and others afterwards have shown that the effects of a stressor do not
depend so much on the physical characteristics of the stressor (intensity, duration, frequency,
etc.) as on whether or not the animal can predict and, above all, control the stressor. Control
is normally exerted by means of performing a behaviour which is relevant for the stimulus.
Individual differences in stress reactions
Even in historical stress research, it was clear that individuals often reacted quite differently
to the same stressor under identical conditions. This obviously reflected some constitutional
differences between individuals. In humans, psychologists would refer to ‘Type A’ persons
for those more prone to react with activity in the sympathetic nervous system and ‘Type B’
for those with less sympathetic reactivity. Behaviourally, Type A were described as more
extrovert and aggressive, and more inclined to react with active attempts to control a stressor,
while Type B would behave much in the opposite way.
This was paralleled by similar observations of differences in animals. Male tree shrews
(Tupaia belangeri) were found to be either actively subordinate to a dominant male or
passively submissive. If a submissive individual was placed in a cage close to a dominant
male, it would lose weight and die within a rather short time, whereas the actively
subordinate would survive and live fairly well in the same situation. In wild-caught mice,
resident males (males which are the ‘owners’ of a cage) were found to behave in one of two
distinct and different ways towards an intruder: either they attacked the new mouse within a
very short time, or they did not attack at all. Very few intermediates, attacking for example
after a few minutes, were observed, leading the researchers to conclude that there were two
non-overlapping subpopulations among the original population of wild mice (Sluyter et al.,
1996).
A strong genetic basis for the behavioural differences among the mice is shown by the
fact that selection for either short or long attack latency in the resident–intruder set-up is
possible. The young of short-latency parents on average attack intruders immediately, and
vice versa.
The mice differ not only in their response to strangers, but also in more or less any
situation where they are exposed to stressful events, even very minor ones. For example,
mice of both lines take the same amount of time to learn to run down a maze to reach a goal
box with food. But, if a very small modification of the environment is introduced, such as
placing a small piece of tape on the floor, the animals from the two populations differ in their
reactions. Those selected for short attack latency seem not to pay any attention to the new
item, whereas the mice with long attack latency get disrupted in their task. They seem to
forget what they are supposed to do and spend their time exploring the tape instead.
The reactions of short-attack-latency mice follow a general pattern, which has been
termed proactive coping, whereas the pattern of the others has been called reactive coping.
Proactive copers will attempt to deal with any challenge by finding routines and performing
behaviour patterns that have proved to be successful earlier. Reactive copers face the same
challenges by attempting to modify their behaviour to find the best way of handling every
new potential stressor in its own way. Physiologically, proactive copers react on challenges
mainly with activation of the sympathetic nervous system, while this system is involved
much less in the reactions of reactive copers (see Table 9.1).
Table 9.1. A scheme showing the major differences between proactive and reactive mice.
Character Proactive copers Reactive copers

Reaction to social intruder High aggression Low aggression
Time to learn a new task, e.g. running through a maze Same as reactive Same as proactive
Reactions to small changes in environment Small, pays little attention High, investigative
Time to adapt to major changes in environment, e.g. shift in Long Short
diurnal light cycle
Activation of HPA axis in stressful situation Medium–high Medium–high

Activation of SA axis in stressful situation High Low
Typical pathological consequences of stress Cardiac disease, stomach Infection, stomach
ulceration ulceration
HPA, hypothalamic–pituitary–adrenal cortex; SA, sympathetic–adrenal medulla.
It seems as if there is some generality to these observations across species and even phyla
– there is some evidence of similar within-species differences among birds. In farm animals,
the reactions of young piglets to a simple restraint test have been used to predict a variety of
reactions to challenges later in life. It even has some predictive value for disease
susceptibility and some aspects of meat quality after slaughter many months later (Ruis et al.,
2000).
A more thorough account of individual differences, personalities, in animals is found in
Chapter 8.
Is stress a physiological or a behavioural phenomenon?
Since stress was originally discovered by physiologists and described in physiological terms,
it has often been treated mainly as a physiological phenomenon. However, the earlier
described studies of Weiss, and many other similar observations, have shown that whether a
certain stressor is harmful or not to an animal depends on the animal’s perception of the
stressor and on its possibility to behave in a relevant way.
This has led some researchers to suggest that stress, in some aspects, can be regarded as
mainly a behavioural phenomenon (Jensen and Toates, 1997). The common denominator of
events that may become harmful to animals, according to this view, is that a behavioural
system is motivated (see Chapter 4), but the animal cannot perform the motivationally
adequate behaviour or, at least, it cannot achieve the relevant functional consequence of the
behaviour. The rats in the experiment of Weiss were exposed to electric shocks in the tails.
This would have motivated their flight behaviour. By performing flight motions with the
forelimbs, the rats acted with the relevant behaviour (so effectively ‘controlling’ the
situation) and when the shock was turned off, they received the desired functional
consequence. As a result, those that did not achieve the desired consequence (the ones
without control) were more harmed by the stressor.
In the example above, the animals were motivated to perform a behaviour by external
stimuli (in this case, the electric shock), but they can also be motivated to perform a
behaviour by mainly internal stimuli, such as hormones. For example, a sow due to farrow is
motivated to build a nest, and this motivation is stimulated almost completely by internal
hormonal events. If the sow is prevented from carrying out the nest-building activities, for
example by being tethered in a crate, the increase in cortisol is much higher than if she is free
to move around. By this measure, the sow would be stressed because she is not able to act
according to her motivational state.
9.4 Animal Welfare

The discussions about abnormal behaviour in domestic animals and improved knowledge of
stress both contributed to a heightened awareness of animal welfare. But in reality it is not so
simple, and not even correct, to think that good animal welfare (or animal well-being as it is
sometimes called) is merely the absence of behavioural disorders and stress.
What is animal welfare?
The Brambell Committee was the first to attempt a scientific definition of animal welfare in
1965. They were rather farsighted in three ways. First, they drew attention to the importance
of behaviour in animal welfare. Up until that time, good welfare had almost been
synonymous with good health. Second, they stressed the importance of the scientific study of
animal welfare, paving the way for future experimental studies. Third, they accepted that
animals had feelings, which went against the behaviourist trend of the time (see Chapter 1).
The committee also proposed five freedoms that every animal should have irrespective of
how or why it is kept. These were that the animal could lie down, stand up, turn around,
stretch and groom. While these may appear self-evident freedoms, there are many examples
even today where they are not achievable; for example, tethered cows and pigs which cannot
turn around, or caged laying hens which cannot flap their wings. The British Farm Animal
Welfare Council later revised these five freedoms to define ideal states rather than standards
for acceptable welfare. They are now:
• freedom from thirst and hunger;

• freedom from discomfort;
• freedom from pain, injury and disease;
• freedom to express normal behaviour;
• freedom from fear and distress.
Since the Brambell Committee’s first attempts, many scientists have attempted to define
welfare but, with hindsight, they can be classified into two main categories. One category
emphasizes the biological functioning of the animal (its health, reproductive success, etc.)
whereas the other emphasizes the subjective experiences of the animal (suffering, pleasure,
etc.). The species-specific natural behaviour of the animal is of course related to both these.
In combination, these three views seem to cover the most important concerns that people
have about animal welfare (Fraser, 2008).
A well-accepted welfare definition of the first category is proposed by Broom
(summarized in Broom, 1996), who defines it by saying ‘the welfare of an animal is its state
as regards its attempts to cope with its environment’. Here it is proposed that we can assess
welfare by recording disease, injury, abnormal behavioural patterns and physiological
changes related to stress (a high incidence of which would imply that the animal was not
coping) and growth or reproduction (high rates of which would imply that the animal was
coping). It is argued that by taking many measures from behaviour, physiology and health we
get a balanced view of where the animal is on the scale from coping easily to not coping at
all. A benefit of this approach is that we already have many techniques for measuring these
parameters. The disadvantage is that we are not sure how to combine them in a way that
takes into consideration that some measures should be given more weight than others.
A well-accepted definition of the second category is proposed by Duncan (summarized in
Duncan, 1996), who says that ‘welfare is all to do with what the animal feels’. Here it is
proposed that feelings have evolved in animals to improve survival and fitness. Pain is
obviously adaptive; by making it unpleasant for an animal to put weight on an injured leg,
the chance that the leg heals increases. Similarly, the frantic behaviour seen in frustrating
situations may increase the chance that the animal eventually changes its situation. It may
even be that positive states, such as pleasure, are used for rewarding the animal for
performing an appropriate behaviour, so increasing the chance that the behaviour is shown
again.
It is argued that, while under natural conditions feelings will closely reflect the health and
physiology of an animal, there is a risk for domesticated animals kept in captivity that the
feeling and its usual physiological correlate become separated. In view of this risk, it is
argued that we must pay attention to whether the animals feels hungry or feels stressed, not
necessarily to whether the animal has a nutritional deficit or elevated corticosteroid levels.
The advantage of this approach to defining animal welfare is that it is intuitively appealing,
reflecting concerns in studies related to quality of life in humans. The disadvantage is that as
yet we do not have good techniques to measure emotional states in animals. To have any
indication of what an animal feels we have to take indirect measures, mainly by examining
their behaviour.
Assessing welfare
Whereas the two lines of scientific approach to welfare assessment may at first appear rather
different, they differ mostly in where the emphasis is placed: the immediate subjective
experiences of the animal or its long-term biological functioning. When it comes to the actual
methods used in the scientific assessment of animal welfare, there is actually good consensus
between researchers. For example, all agree that the focus should be on the animal. In quality
assurance schemes the term animal-based measures is often used to differentiate them from
measures focusing on the environment of the animal. These resource-based and management-
based measures tell us about the risk to the animal’s welfare from its environment.
In the following we will only briefly go through the health, production and physiology
indicators used when assessing welfare; instead, in this book on ethology, we will spend most
time on the behavioural indicators.
Health and production as indicators of animal welfare
It is obvious that animal health is an important aspect of animal welfare. An animal with a
broken leg, bleeding wound or other physical injury clearly has poorer welfare than an
animal without that damage. But not all health issues are so clear and it is important to
remember that the border between health and disease is often indistinct. Complete health, that
is to say 100% disease-free, probably doesn’t even exist since the body is continually
reacting to bacteria and viruses. Furthermore, even an unhealthy animal does not necessarily
experience pain or distress. A dog with a small tumour probably does not experience any pain
or distress and a chicken with weak bones will not experience pain until there is a fracture.
But the dog will probably suffer reduced welfare later, if the tumour is not treated
successfully, and the hen is at risk of reduced welfare since bones are often broken during
handling and transport.
A simple conclusion therefore is that an unhealthy animal either already has its welfare
reduced or is at risk some time in the future of having its welfare reduced. But as will be
explained later, there is more to good welfare than being healthy.
Just as historically a healthy animal has been regarded as an animal with good welfare, so
it has been argued that a farm animal with a high production has good welfare. But this view
is criticized because production traits (e.g. number of offspring, growth rate, etc.) have been
the focus of intense selection for many generations. They differ between breeds and strains
even under identical housing and management conditions and they can be easily manipulated
by changes in diet, lighting schedules and so on. In fact, very high production itself can lead
to problems, such as udder inflammation in high-producing dairy cows or leg problems in
fast-growing broiler meat chickens and fattening pigs. These have been called production
diseases and are a relatively new category of welfare problem resulting from high production.
A final, albeit extreme, example to demonstrate the risk of using traditional production
parameters as welfare indicators is the case of number of offspring. Using this measure, the
animal with highest welfare would be a breeding bull at a semen station whose sperm are
used to inseminate cows around the world. In fact, this bull may even be dead!
In summary, while poor production performance may be associated with poor welfare,
good production per se is no guarantee of good welfare.
Physiological indicators of welfare
The most commonly used physiological indicators are those associated with the stress
response outlined earlier in this chapter. Commonly used variables therefore include heart
rate, corticosteroid levels and adrenal gland weights. Often these measures are taken
remotely using radiotelemetry or by measuring corticosteroids in non-invasive ways, from
the saliva, urine or faeces, instead of from blood samples. Not disturbing the animal while
taking the measurement is particularly important when the physiological response is rapid,
such as the case with heart rate.
The frequently used interpretation is the greater the stress, the poorer the welfare. Usually
this is true, but not always. The physiological response to short-term (acute) stress is different
from that to long-term (chronic) stress because the system adapts and down-regulates. Also,
some so-called stress responses can reflect positive experiences such as excitement and
arousal.
To help overcome the problem of there being no obvious cut-off point where the level of
stress can be said to be ‘bad’, some researchers have proposed that there is a welfare problem
when the stress response is such that it leads to a change in the biological functioning of the
animal and the animal enters a prepathological state. In short, it is argued that instead of
measuring the stress response per se, the consequences of the stress are measured. Examples
of such consequences can be immunosuppression or reduced reproductive success.
Behavioural indicators of welfare

The advantage of behavioural indicators of welfare is that they are the easiest to obtain and
they probably reflect an animal’s first attempts to cope with a less than optimal situation.
Thus responses such as huddling together when the temperature is too low or vocalizing
when hungry probably occur long before there is a threat to welfare from hypothermia or
starvation. Natural selection has led to mechanisms that operate earlier and earlier, thus
giving us more sensitive indicators that welfare is at risk than the obvious indicators such as
injury and disease (Dawkins, 1998).
Comparison with normal behaviour
Early work on behaviour related to welfare was often associated with developing new
housing systems and it was common practice to compare the time budgets of animals in the
new systems with that in the traditional system (a time budget refers to a quantification of
how much time an animal allocates to different activities over the day). Comparisons were
also made between animals kept under traditional systems and those kept under extensive
conditions in outdoor enclosures. These extensive conditions were assumed to be better for
welfare and the standard upon which other systems could be compared. But there are several
problems with such comparisons. That the time budgets of animals differ according to the
environment in which they are kept is not surprising. Much more difficult is to interpret this
difference in terms of the animal’s welfare. If the range of behaviour patterns seen in one
system is less than in the other, it may mean that the behaviour is not released by that
particular environment or it may mean that the animal is prevented from performing the
behaviour.
An example of the first type may be anti-predator behaviour. Under extensive conditions
poultry can give alarm vocalizations and escape to cover several times per day. If this
behaviour is not seen under more confined conditions, we would not recommend that the
caretaker takes steps to elicit it. The behaviour is triggered by external stimuli in the
environment and there are unlikely to be welfare problems associated with its absence.
An example of the second type may be roosting behaviour in chickens, which is also in
part an anti-predator behaviour, since predation losses are lower for birds roosting up off the
ground. Given perches birds will usually roost on them during the dark period, otherwise
they roost on the floor. Roosting behaviour is triggered by external stimuli, in this case the
decreasing light intensity. Birds have a dark period even in commercial poultry housing, but
there are no, or at least should not be, any real predators. The question is whether roosting up
off the ground is still important for the birds and so whether providing perches under
commercial conditions would improve welfare. We could not have answered this question
without further studies which showed that indeed birds are motivated to get access to a perch
for night-time roosting and so motivated for this behaviour (Olsson and Keeling, 2002). Thus
comparisons of behaviour in different environments can often highlight issues to be
investigated experimentally.
An additional benefit of studies comparing the behaviour of the same animals in both
extensive and intensive conditions, or even of modern breeds and hybrids with their
ancestors, is the important information we obtain on where, when and how different
behaviours are performed. Such studies help us understand some of the behaviours we see in
captivity but have not been able to interpret. For example, it is common knowledge that
many birds build nests. It is therefore likely that the restless behaviour a laying hen performs
in the 1–2 h before she lays her egg may be related to nest building. However, who would
propose a similar explanation for the movement patterns shown by a sow about to farrow, if
there had not been observations on nest-building behaviour in nature? After all, no other
ungulates are known to build nests.
Experimental studies of behaviour as welfare indicators
Experimental studies on behaviour have contributed greatly to our understanding of animal
welfare. Rather than observing the undisturbed animal, as described above, an artificial
situation is created where the behaviour of the animal helps us answer questions about its
state. Imagine that we did not know how a cow with a pain in its leg behaved. To determine
this, we could inject uric acid in to a joint (to create a form of gout which we know is
painful) and in follow-up behaviour observations we would observe that the animal limped,
putting less weight on the treated limb. In future when we saw an animal limping in the same
way, we might conclude that it too was experiencing pain.
Now we do not need to do this experiment because we already know the answer, but we
can artificially create other states where we have less knowledge. Frustration is likely to
result when an animal is prevented from performing a behaviour it is motivated to perform.
By experimentally creating a frustrating situation it is possible to study what behaviour
patterns are shown by animals. Examples of frustration-related behaviour in chickens, for
example, range from displacement preening when the level of frustration is mild, to increased
aggression and stereotyped pacing in more frustrating situations. It is then possible to observe
chickens in a variety of different situations to identify times of day or situations when there
seems to be evidence of frustration so that these situations can be studied in more detail. This
idea of artificially creating emotional states has mostly been used to study frustration so far,
but could probably be used to help investigate other states in animals. Recently, for example,
similar methods have been used to show that animals develop expectations, both for positive
and negative events. Signalling a later positive or negative event will evoke a range of
various anticipatory behaviours, which can then be used in other situations to determine
whether the animal finds some particular procedure rewarding or punishing.
Another methodology that has been used frequently is that of preference testing. Here the
animal is given a choice between two or more resources or situations and the assumption is
that it chooses in its best interest. The concept of preference testing is simple, but it is
important that the tests are controlled well if the results are to be reliable. For example, many
of our domesticated animals are wary of unfamiliar environments and food. Thus in a
preference test the animal should have had equal experience of both choices. Animals also
almost always choose to maximize short-term welfare over long-term, for example a sweet
tasty food over a nutritionally balanced one. Age, strain, previous experience and time of day
will also influence the outcome.
Even with a well-designed preference test, though, there is rarely a 100% choice for one
option over the other. Partial preferences may reflect that all animals rank the choices
relatively similarly, or that a proportion of the animals prefer one option whereas the other
proportion prefers the other option. Partial preferences may even reflect that one choice is
preferred on one occasion and the other the next time, depending on the state of the animal. A
modification is to leave the animal in the test apparatus for a longer time and then compare
the total time spent in each of the available compartments. While this may overcome some
difficulties with the simple choice test, one must remember that time spent using a resource
does not necessarily reflect the whole picture of its importance to the animal.
Despite the improvements in preference tests’ experimental design, one cannot
compensate for the fact that in preference tests the choice is always relative. That is to say,
while one option may be strongly preferred over the other, both may be poor for welfare or
both may be good. Training the animal to perform an operant response such as pecking a key
or pushing a lever to get access to a resource or an environment allows us to quantify the
importance of a choice. The assumptions are that if an animal will work for the resource then
it must be of at least some importance to it and, presumably, the harder the animal will work
(i.e. more pecks or pushes) the more important it is (see Fig. 9.5).
Fig. 9.5. A laying hen pushes against a weighted door to open it and so gain access to
whatever resource is provided on the other side. The weight of the door, and so the effort
needed to open it, can be manipulated systematically.
A further refinement of this idea of quantifying how hard the animal will work for the
resource is to incorporate it into consumer demand theory used by economists to assess the
importance of a commodity to human consumers. Commodities where demand decreases as
price increases are said to have an elastic demand and those where the demand changes little
as price increases are said to have inelastic demand. In humans, an example of the first may
be fish and the second petrol. If the price of fish goes up people eat something else, but the
majority of people still buy petrol for their car even if they complain about the price.
In animal studies, food is normally used as the standard example of an inelastic
commodity and a regression line of data on amount of work to gain access to food with
increasing price is usually rather horizontal. For other resources the regression line usually
slopes downwards, showing that animals become less willing to obtain the resources as the
amount of necessary work increases. By comparing the slopes of the lines one can rank the
importance of difference resources to the animals (see Fig. 9.6). Just as this technique can be
used to measure strength of motivation to obtain a resource, it can also be used to measure
motivation to avoid something unpleasant.
Fig. 9.6. Demand curves for different commodities in pigs. The pigs were required to
perform an increased amount of work (numbers of snout presses on a bar) for access to a
given commodity. The slope of the line indicates the relative importance of the commodity to
the pig, a more horizontal slope indicating a higher importance. (From Ladewig and
Matthews, 1996.)
Positive welfare indicators

While research has been quite successful in developing tools to assess the negative side of
animal welfare (e.g. if animals suffer from frustration or are in pain), less attention in the past
has been given the positive side of welfare. Nevertheless, most people would agree that good
animal welfare is not only signified by absence of suffering, but also should encompass
positive feelings and experiences. In order to say that an animal has good welfare, we need
some indicator of it actually feeling well or even happy. This has proved to be a difficult
scientific task.
One possible indicator is the ability to perform motivated, species-specific behaviour.
The assumption here is that when an animal is able to act in accordance with its urges or
drives, it probably feels well. Hence, a sow building a nest, or a laying hen engaging in dust-
bathing, would experience positive emotions. However, this would ideally need to be
assessed with some independent measure of the experiences of the animals. Whereas there
are physiological factors persistently correlating with negative welfare (e.g. levels of
corticosteroids), similar factors indicative of the opposite are more elusive. It has been
suggested that oxytocin, the neuropeptide associated with social bonding and affiliative
behaviour, could serve as such an indicator, since it is closely associated with positive social
experiences and social support, and reduces heart rate and blood pressure (Rault, 2012).
Another possible physiological indicator of positive welfare is the endogenous opioid
system (Rault, 2012). Opioids, such as endorphins, activate rewarding neural circuits in the
brain and reduce stress-related behaviours in many species. In humans, they induce a feeling
of relaxation, pain relief and even euphoria. However, so far, no systematic use has been
made of either oxytocin or opioid levels as a measure of positive welfare in animals, partly
because of inherent problems in sampling without inducing stress.
Behaviour-wise, play has become a major candidate indicator of positive welfare. This is
dealt with in depth in Chapter 7. The logic behind this is that animals are thought to engage
in play only when all other essential needs are satisfied. That is to say, they are clearly not in
a state of poor welfare when playing. One problem with play as an indicator is that this is not
commonly observed in adult animals, so it would work primarily in studies of juveniles.
More generally, it has been suggested that positive emotional states are most likely to have
become associated with performing a behaviour where the fitness benefits are in the future
(Boissy et al., 2007). This is the case for play, but may also support that activities related to
hygiene and body care, such as grooming and dust-bathing, are experienced positively.
Research is also starting to show that very subtle behaviour, such as certain body postures,
facial expressions and vocalizations, is shown in situations thought to be pleasurable. For
example, rats emit a specific high-frequency vocalization, a type of laugh, when tickled
(Panksepp and Burgdorf, 2000).
Two other behavioural approaches that have become increasingly popular attempts to
assess positive welfare are cognitive bias (Mendl et al., 2009) and anticipatory behaviour
(Spruijt et al., 2001). Cognitive bias is based on the knowledge from human psychology that
humans in a positive state of mind tend to judge ambiguous stimuli more optimistically than
those in a more negative state. A range of experiments has shown that this appears to be the
case in animals as well. For example, an animal taught that a black bowl contains food and a
white one is empty will tend to run faster to a grey bowl if it has a history of positive
experiences.
Anticipatory behaviour refers to the activities of an animal when it is expecting a reward.
It can be measured by pairing a stimulus with an event (a reward or a punishment), and then
gradually increasing the time lag from the signal until the event. The behaviour of the animal
during the time lag will usually differ depending on whether it anticipates a positive or a
negative event after the stimulus. Positive anticipatory behaviour can therefore be identified
and used as an indicator of good welfare.
Both cognitive bias and anticipatory behaviour are treated in greater depth in Chapter 5.
9.5 Concluding Remarks
This chapter has shown that scientific studies of animal behaviour can provide some of the
necessary keys for understanding the welfare of animals in captivity. Behaviour is often the
first reaction used by an animal to adapt in a specific environment. Whereas data on
physiological reactions and health are of course informative, they are often difficult to
interpret in welfare terms without the necessary knowledge of accompanying behavioural
reactions. For example, an elevated corticosteroid level may simply reflect an increased
energy demand on the body, because the animal is actively playing, but in another situation it
may signal that the animal is stressed, with an associated increased risk for disease and
suffering. Observing behaviour can also help inform us about whether the animal is in pain or
otherwise in a poor state of health.
Of course, the ultimate decision of whether a certain level of welfare is acceptable or not
is an ethical one. It requires perspectives and arguments that cannot be obtained from
scientific studies alone. However, any ethical discussion about what we ought or ought not do
to animals should preferably be based on objective scientific knowledge about their welfare.
References
Appleby, M.C., Mench, J.A., Olsson, A. and Hughes, B.O. (2011) Animal Welfare, 2nd ed. CAB International, Wallingford,
UK.
Boissy, A., Manteuffel, G., Jensen, M.B., Moe, R.O., Spruijt, B., Keeling, L.J., Winckler, C., Forkman, B., Dimitrov, I.,
Langbein, J., Bakken, M., Veissier, M. and Aubert, A. (2007) Assessment of positive emotions in animals to improve
their welfare. Physiology & Behavior 92, 375–397.
Broom, D.M. (1996) Animal welfare defined in terms of attempts to cope with the environment. Acta Agriculturae
Scandinavica, Section A – Animal Science, Supplementus 27, 22–28.
Broom, D.M. and Fraser, A.F. (2007) Domestic Animal Behaviour and Welfare, 4th edn. CAB International, Wallingford,
UK.
Dawkins, M.S. (1998) Evolution and animal welfare. The Quarterly Review of Biology 73, 305–327.
Duncan, I.J.H. (1996) Animal welfare defined in terms of feelings. Acta Agriculturae Scandinavica, Section A – Animal
Science, Supplementus 27, 28–36.
Fraser, D. (2008) Toward a global perspective on farm animal welfare. Applied Animal Behaviour Science 113, 330–339.
Jensen, P. and Toates, F.M. (1997) Stress as a state of motivational systems. Applied Animal Behaviour Science 54, 235–243.
Ladewig, J. and Matthews, L.R. (1996) The role of operant conditioning in animal welfare research. Acta Agriculturae
Scandinavica, Section A – Animal Science, Supplementus 27, 64–68.
Mendl, M., Burman, O.H.P., Parker, R.M.A. and Paul, E.S. (2009) Cognitive bias as an indicator of animal emotion and
welfare: emerging evidence and underlying mechanisms. Applied Animal Behaviour Science 118, 161–181.
Olsson, I.A.S. and Keeling, L.J. (2002) The push-door for measuring motivation in hens: II Laying hens are motivated to
perch at night. Animal Welfare 11, 11–19.
Panksepp, J. and Burgdorf, J. (2000) 50-kHz chirping (laughter?) in response to conditioned and unconditioned tickle-
induced reward in rats: effects of social housing and genetic variables. Behavioural Brain Research 115, 25–38.
Rault, J.-L. (2012) Friends with benefits: social support and its relevance for farm animal welfare. Applied Animal
Ruis, M.A.W., te Brake, J.H.A., van de Burgwal, J.A., de Jong, I., Blokhuis, H.J. and Koolhaas, J.M. (2000) Personalities in
female domesticated pigs: behavioural and physiological indications. Applied Animal Behaviour Science 66, 31–47.
Sluyter, F., van Oortsmerssen, G.A., de Ruiter, A.J.H. and Koolhaas, J.M. (1996) Aggression in wild house mice: current
state of affairs. Behavior Genetics 26, 489–496.
Spruijt, B.M., van den Bos, R. and Pijlman, F.T.A. (2001) A concept of welfare based on reward evaluating mechanisms in
the brain: anticipatory behaviour as an indicator for the state of reward systems. Applied Animal Behaviour Science 72,
145–171.
Toates, F. (1995) Stress: Conceptual and Biological Aspects. Wiley, Chichester, UK.
Weiss, J.M. (1971) Effects of coping behaviour in different warning signal conditions on stress pathology in rats. Journal of
Comparative and Physiological Psychology 77, 1–13.
10 Human–Animal Relations
S. Waiblinger
10.1 The Human in the Animals’ World

Throughout the history of the human race, animals have played an important part in human
life and vice versa. In the early hunter-gatherer/nomadic cultures, animals were viewed as
prey but also as dangerous predators. The same holds true for the animals’ perspective: some
species might have considered the human a predator, some a prey. For other species, humans
might just have been a neutral part of the environment. When domestication started, the
human–animal relationship developed towards a symbiosis in which the human provided, for
instance, protection from predators or food in exchange for animal products (food and fur)
and power. For some individuals, such as animals ‘adopted’ by humans at an early age, the
human was a social partner with whom they played and exchanged affiliative behaviour.
Hediger (1965) first described the five possible roles (predator, prey, neutral part of
environment, symbiont, social partner) of humans for animals in captivity. Estep and Hetts
(1992) suggested that some of these roles may not be mutually exclusive, and that an animal
probably perceives a human in terms of a combination of the above roles and according to
the current situational factors. An example would be a dog that may perceive a child as prey
(Miklósi, 2007) only in a particular situation.
Like their human counterparts, domestic animals now take on a wide variety of roles:
production animals (for food, wool, fur, leather); working animals (e.g. oxen or horses for
transportation or ploughing the field; dogs for hunting, herding, protection; cats for catching
rodents); animals used in sports (often just another form of working), in shows or as beloved
pets. Obviously, differing relationships between humans and animals can be expected
according to these roles. However, the relationship can also vary widely within those
categories of animal utilization. The aim of this chapter is to discuss the reasons for variation
in the relationship as well as its relevance, mainly for farm and pet animals. But first, we will
deal with the definition of the human–animal relationship.
10.2 What is the Human–Animal Relationship?

When visiting different farms, for instance as a veterinarian, the differences in the behaviour
of both animals and humans are striking: on some farms, animals are confident in the
presence of humans, they can be examined and treated easily, whereas on others, animals are
fearful of humans, treatment is difficult and one has to be careful to avoid being injured.
Accordingly, on the former, the farmers talk to and stroke their animals and handle them
calmly, whereas on the latter, farmers hardly show such gentle behaviour but are often
impatient, shout at or even hit their animals (Fig. 10.1). Such variation was shown in
different farmed species (Hemsworth and Coleman, 2011). Similar differences can be found
in dog–human or cat–human interactions.
Fig. 10.1. Variation in human and animal behaviour on dairy cow farms. For exact definition
of stockpeople behaviours and test procedure, see Waiblinger et al. (2006). (a) Number of
different behaviours used by stockpeople per cow during one evening milking on 30 farms:
Pos, friendly, positive interactions (stroking, touching, talking calmly); Neu, neutral or
moderately negative interactions (slight hit with hand or stick, talking dominantly); Neg,
clearly negative behaviour (strong hits with hand or stick, shouting, talking impatiently). (b)
Percentage of cows that can be approached until being touched by an unfamiliar test person
or that withdraw from the test person already at a distance of more than 1 m.
Obviously, the relationships between humans and animals differ distinctly in those
examples. They are reflected in the behaviour of both animals and humans (see Fig. 10.2).
These behavioural differences are expressions of underlying emotions and motivations
during human–animal encounters that are caused by the previous experiences with each
other. Thus, the history of previous interactions forms the foundation of the current
relationship. The quality of the relationship then affects the nature and perception of future
interactions. That means a relationship is always dynamic; both positive and negative
changes can happen throughout a lifetime, even though in some phases of life changes are
more likely to occur or to be longer-lasting (as will be discussed below in more detail). In
sum, the human–animal relationship can be defined as the degree of relatedness or distance
between the animal and the human (i.e. the mutual perception), which develops and
expresses itself in their mutual behaviour (Estep and Hetts, 1992; Waiblinger et al., 2006).
The relationship can range on a continuum from poor or negative, where the human is
perceived as frightening and unpleasant emotions are involved during interactions, to good or
positive, where the human is perceived as a social partner and interacting with him or her is
often pleasurable. The relative strength of pleasant or unpleasant emotions in the perception
of humans constitutes an animal’s relationship with humans and vice versa (Waiblinger et al.,
2006; Fig. 10.3). In this chapter we concentrate on the animal’s perspective. More
information on the human side can be found in Hemsworth and Coleman (2011).
Fig. 10.2. The mutual relationship is reflected in the behaviour of the cows (enjoying
stroking or searching for contact) and the person (stroking). (Photo courtesy of C. Menke.)
Fig. 10.3. The two dimensions (unpleasant, pleasant) contributing to the human–animal
relationship and the corresponding perception of the human or animal. Increased levels of
pleasant emotions improve the relationship and are characteristic for a very good
relationship, and vice versa. (From Waiblinger et al., 2006.)
Individual or general relationship

In principle, a relationship develops between two individuals knowing each other; for
instance, a farm animal and its caretaker or a pet animal and its owner. Such relationships
require mutual individual recognition. While in pet animals such personalized relationships
are most common, individual relationships between farm animals and humans are limited to
systems enabling sufficient contact. Keeping large groups of farm animals, for example a
group of several thousand laying hens, impedes the recognition of individual animals by the
stockperson and thus individual relationships.
However, animals often generalize their experience with familiar humans to unfamiliar
humans (Bernstein, 2005; Waiblinger et al., 2006; Fig. 10.4). Farm animals that withdraw
from their own stockperson at a great distance react in the same manner to an unknown
human, whereas animals accepting to be touched by their familiar stockperson also allow a
strange person to do so. Humans as well may show generalized attitudes and behaviour
towards animals of the same species or even in general (Hemsworth and Coleman, 2011). As
long as individual recognition is not precluded, both individual relationships and
generalization exist in parallel and influence animal behaviour. For example, lambs that were
bottle-fed and received gentle handling showed less isolation distress when a familiar or
unfamiliar shepherd was present, even though the effect was greater with a familiar shepherd.
Handled piglets also interacted more with familiar and unfamiliar humans than non-handled
ones, but made contact sooner and more often with the familiar handler and were less
agitated when caught by her or him than by an unknown person (reviewed in Waiblinger et
al., 2006).
Fig. 10.4. Gentle tactile contact in a generalized relationship: the lambs and the human met
for the first time.
Recognition of humans by animals

Individual relationships need individual recognition. The examples above have already
shown that animals are able to differentiate between humans. But what cues do they use?
When people are asked this question with respect to farm animals, they often mention smell
first. However, similar to the recognition of conspecifics, mainly visual cues seem to be used
when possible (Rushen et al., 1999, 2001). Pigs can differentiate humans by the mere use of
olfactory, acoustic or visual cues, but they are worst with olfactory cues alone and best when
they can use the three senses concurrently. Like pigs, cattle are able to differentiate humans
by their faces, but if clearer cues such as colour of clothes are available, they make use of
them. Previous experience with humans dressed in a specific colour during specific situations
(be it pleasant or unpleasant) may then be translated into respective expectations when the
same or other humans are dressed equally. Sometimes, the sight of clothes of a negative
handler alone elicits stress responses (Waiblinger et al., 2006). Sheep can not only
differentiate the faces of different humans, but also remember them 2 years later (Kendrick,
2006). This agrees with anecdotal reports of bulls remembering an aversive person more than
a year later. Horses and dogs can discriminate between individual humans by their faces
(photographs) or voices and also show cross-modal recognition: they can match familiar
voices with the respective faces (Adachi et al., 2007; Huber et al., 2013; Brubaker and Udell,
2016).
10.3 Factors Affecting the Quality of an Animal’s Relationship

with Humans
The previous history of human–animal interactions, the quality and quantity of the
interactions and the period of life when they happened, is the main determinant of an
animal’s relationship with humans. However, the genetic background of the animal as well as
the physical and social environment for rearing and interacting with humans are also
influential. These factors are discussed in this section.
In domestic animals, it is mainly the human who determines the possibilities of
interactions, their type and timing within a given environment. Often he or she also provides
the environmental conditions of interactions. Therefore, the human contributes to a great
extent to the animal’s developing relationship with him or her. The main factors underlying
human behaviour are personality and attitudes, but empathy, knowledge, experience and the
actual situation (e.g. time pressure, workload) also play a role. These human factors have
been reviewed, for instance, by Serpell and Paul (1994), Podberscek et al. (2000), Spoolder
and Waiblinger (2008) and Hemsworth and Coleman (2011), and will not be considered here.
Quality of human–animal interactions
Human–animal interactions can involve visual, tactile, olfactory and auditory perception. An
animal may perceive an interaction as negative, neutral or positive. As mentioned above, this
perception is influenced by the existing relationship with humans. A simple approach of a
human can be perceived by different animals in any of the three ways mentioned above
according to previous experience. However, some interactions are aversive by their very
nature because they are painful or otherwise distressing (e.g. dehorning, beak trimming,
forceful hitting by hand or stick). Shouting and hitting are perceived as aversive, as shown in
cattle by aversion-learning techniques and preference tests or handling studies, where acute
and chronic stress responses were found (Waiblinger et al., 2006). Frequent or predominant
use of negative interactions such as hitting (even with low force) or shouting increases
animals’ fear of humans, whereas neutral interactions (such as visual presence) allow
habituation to humans with decreased levels of fear. Positive interactions (such as stroking,
feeding, gentle talking) reduce fear and enhance confidence in humans reflected in behaviour
(more approach and less avoidance) and physiology (lower heart rate and cortisol levels)
(Boivin et al., 2003; Hemsworth, 2003; Waiblinger et al., 2006).
Different sensory channels provide different possibilities. Even though visual presence
alone can be powerful in habituating animals to humans and thus in improving the
relationship, gentle tactile interactions are more powerful, especially when provided in a
manner imitating intraspecific social grooming (McMillan, 1999; Schmied et al., 2008a).
Stroking elicits quite distinct behavioural responses (e.g. lowering eyelids, hanging ears)
indicating a state of relaxation. In several species, it was also shown to trigger oxytocin
release and thereby to lead not only to immediate effects of a lower heart rate, lower blood
pressure and lower cortisol levels, but also to long-term positive effects on social bonding
and health (Uvnäs-Moberg, 1997; Uvnäs-Moberg and Petersson, 2005; Nowak and Boivin,
2015). However, a certain level of confidence in humans seems necessary to elicit such
positive physiological responses.
Gentle contact may not be perceived as positive by every individual and in any situation.
Besides the existing relationship, the prevailing motivation, affective state and environment
are likely to influence the perception. It would be interesting to know if forced stroking,
where animals are restrained and cannot avoid being touched by the human, could also have
positive effects. Results in cattle and horses suggest this to be the case; at least in animals
used to some form of positive or neutral human contact (Ligout et al., 2008; Schmied et al.,
2008b). In newborn foals, forced handling had the opposite effect (Hausberger et al., 2008).
However, the reaction of the handled animal may be a reliable indicator of its perception: the
early handled foals showed strong escape attempts, whereas defensive behaviours were rare
in the example of Schmied’s cows. Furthermore, subtle differences in the quality of
interaction (pressure during stroking, accompanied by a gentle voice or not) could be
important for the effects of stroking (Hennessy et al., 1998).
Food is often recommended as a means to establish relationships with animals. Research
suggests that, even though offering food is actually a good start to enter into a relationship, it
has a less rewarding component than other positive interactions such as gentle tactile
interactions or playing, and is thus not sufficient for maintaining a positive relationship. Cats,
for instance, seem to know which family member is most likely to provide food and direct
their interactions towards that person when they are hungry. Nevertheless, they are likely to
be just as affectionate to other family members at other times (Bradshaw, 1992). Another
experiment shows that, although feeding first attracts cats to the person feeding them, the
preference for this person over a non-feeder (outside the feeding situation) ceased after a few
days – other interaction such as playing, petting, etc. are required to cement a newly founded
relationship (Karsh and Turner, 1988). In lambs and cattle, results are inconclusive. Hand-
feeding was as effective as just gentle interactions in some experiments but less effective in
others; the exact way of feeding and interactions may be relevant (Nowak and Boivin, 2015).
Gentle, soothing talking is often used by farmers as well as in experiments combined
with other forms of contact, but there are few investigations on the effects of talking alone. In
cats the socialization of kittens by handling them during early development was especially
effective when handling included talking (Bernstein, 2005; see also Section 10.5 below). For
reducing stress, interactions further should be predictable and controllable. For instance, the
random incorporation of negative interactions within positive contact had similar negative
effects on welfare as a consistent regime of negative handling (Boivin et al., 2003).
Quantity of human–animal interactions
Apparently, the amount of interaction is also crucial. Fear responses to humans can not only
be caused by a learnt negative association due to aversive interactions, but also by an absence
of habituation to human contact. The animal’s relationship with humans will improve with an
increasing amount of neutral or positive contact. For example, calves handled daily during
the first 3 months of life and running with their mother from the 4th month on were easier to
handle, showed less flight behaviour, accepted longer tactile contact and, importantly, were
never aggressive in a ‘docility test’ at the age of 8 months compared with calves running with
their mother from birth onwards (Boivin et al., 1994). Kittens handled for over 1 h/day went
directly to a familiar person, climbed into his or her lap, purred and either played or slept. In
contrast, a kitten that was handled less, such as a mere 15 min/day, tended only to approach
and to head-rub and then moved off (Karsh and Turner, 1988).
Sensitive periods
In dogs and cats, there is clear evidence that the sensitive period of socialization is extremely
important for their later relationship with humans; in other domestic animals too, sensitive
periods seem to exist. Socialization is the process by which an animal develops appropriate
social behaviour towards conspecifics (Karsh and Turner, 1988). During the socialization
period, the animal forms primary social relationships or social attachments. In the wild, they
attach themselves to conspecifics; in domesticated animals, contact with humans and other
non-conspecifics allows the formation of relationships with them (Serpell and Jagoe, 1995).
Being raised throughout the socialization period with non-conspecifics only (e.g. cross-
fostering of kittens and pups with the opposite species) leads to a preference for this species
for later positive social interactions (Serpell and Jagoe, 1995). Moreover, hand-reared
ruminants and pigs display courtship behaviour preferably or exclusively towards humans,
depending on the duration of isolation from conspecifics (Sambraus and Sambraus, 1975).
If raised with their conspecifics, the amount and type of contact with humans (or other
non-conspecifics) during the socialization period will strongly impact on the later social
behaviour towards humans (or other non-conspecifics) in dogs and cats. It is difficult to
compensate a lack of contact with humans during the sensitive period of socialization by later
interactions. In cats, for example, animals handled in the sensitive period of 2–7 weeks of age
are generally friendly to people, whereas non-handled kittens will be extremely difficult to
socialize at a later date (Karsh and Turner, 1988). In dogs, the primary socialization period is
thought to last from about the 3rd to the 12th week after birth, with a peak of sensitivity
between 6 and 8 weeks (Serpell and Jagoe, 1995). However regular social reinforcement
until the age of about 8 months is necessary to keep socialization intact, as dogs being well
socialized at 3 months of age will become fearful again without further contact. Socialization
of dogs later in life is very difficult and requires considerable patience (Serpell and Jagoe,
1995).
The amount of contact, as well as the number of different humans during the sensitive
period, is likely to affect the later relationship with humans. As mentioned above, kittens
handled for longer show more positive behaviour and stay longer in contact with a familiar
person compared with kittens handled only for 15 min/day. Cats handled by one person only
could be held longer by the familiar person than by a stranger; while cats with experience of
four handlers stayed with any person for the same amount of time. However, while kittens
handled by five people were more outgoing and showed least fear of strangers, kittens with
one handler only were more affectionate to their handler, purred more often and played for
longer (Karsh and Turner, 1988; Bradshaw et al., 2012).
In farm animals, the period of an animal’s life during which human contact occurs can be
important as well, even though the early socialization period seems less crucial than in cats or
dogs. In herbivores, soon after birth but also during weaning as well as the hours after
parturition seem to constitute sensitive periods, and are especially effective in improving the
relationship: the same amount and type of contact has stronger impact (Rushen et al., 2001;
Boivin et al., 2003; Waiblinger et al., 2006). For example, heifers handled gently in the first
2 weeks after birth or weaning accepted being stroked and hand-fed to a much higher degree
than heifers handled in the same manner 6 weeks after birth or weaning. The latter, however,
were still less fearful than animals not handled at all. What these periods have in common is
that they are related to the formation of or changes in social relationships (to the mother, the
peers or the offspring). Again, also in farm animals, contact during those sensitive periods
helps to establish a good animal–human relationship, but a prolonged period of regular
handling seems to be necessary to maintain it (Boivin et al., 2003; Hausberger et al., 2008).
Furthermore, in foals, it was even argued that there is no clear evidence for the existence of
sensitive periods throughout their development facilitating the establishment of a foal–human
bond. Two-year-old non-handled animals become as familiar as handled animals (Hausberger
et al., 2008). However, they were exposed daily to caretakers bringing food, which may be
sufficient for a primary socialization. Similarly, fear of humans was reduced in heifers by 2
weeks of gentle interactions regardless of whether they had experienced gentle human
contact in the first 2 weeks of life or not in addition to routine care (Lürzel et al., 2016).
Thus, the relative importance of handling during sensitive periods and prolonged or lifelong
regular handling in farm animals is still ambiguous.
In sum, for both pet and farm animals, the best relationship can develop and be
maintained when positive contact with humans takes place during sensitive periods early in
life and regular positive interactions also happen later on (see Fig. 10.5). This is consistent
with intraspecific social relationships, where the strongest bonds are formed early in life and
are regularly reinforced by positive social interactions. It has to be stressed, however, that
also an adequate socialization with conspecifics is important for later life, and premature
separation of the offspring from mother and siblings can cause problems later.
Fig. 10.5. Early positive contact such as play is crucial for a good relationship later in life.
Genetic background and personality

A major characteristic of domestication is a change in animals’ reactions to humans, which
can be summarized by decreased fear and aggression towards humans (i.e. greater
‘tameness’). The powerful role of genetics can be seen in selection experiments in silver
foxes, where behaviour towards humans changed within one generation due to selection for
tameness. In later generations, behaviour changed more and more to patterns resembling
those of dogs (e.g. wagging tail, licking the human’s hand; Price, 2002; Miklósi, 2007). In
domesticated animals, the genetic influence on animal–human interactions and thus the
relationship becomes evident from differences between breeds and individual differences
within breeds (Serpell and Jagoe, 1995; Turner, 2000; Rushen et al., 2001; Price, 2002;
Boissy et al., 2005; Hausberger et al., 2008).
Results from various species suggest that the influence of the genetic background on the
animals’ relationship with humans is due to underlying changes in different personality traits,
mainly fearfulness/boldness, aggressiveness, curiosity and sociability (Turner, 2000;
Bernstein, 2005; Boissy et al., 2005; Miklósi, 2007). Personality can be defined as an
individual’s unique set of traits which is relatively stable over time and affects how it
interacts with the environment. In cats, paternity was shown to affect animals’ behaviour
towards humans via influences on the trait of boldness. Reported heritability of fear of
humans or reactivity to handling varies widely in cattle, sheep and pigs from low heritability
(h2 ≅ 0.1) to quite high heritability (0.53). Presumably, such variations result from
differences in test procedures measuring reactions to humans. The fact that the estimated
heritability of traits related to reactions to handling or humans can change with the age of the
animals (lower heritability in adult versus young animals) points to the importance of
experience.
Genetic background and experience with humans interact in the development of the
animal–human relationship. Cats that were socialized to humans in the first 3 months of life
and had a father friendly to humans were more likely to show relaxed behaviour and less
likely to show defensive behaviour towards a stranger (Bernstein, 2005). In some cases, the
impact of personality seems to be quite strong: in cats, some individuals seem resistant to
changes of their original type of interaction with humans (i.e. some kittens are quite friendly
to humans from the beginning and remain so), whereas others remain quite fearful despite
handling (Bernstein, 2005). Individual personality types among cats with a comparable
history of socialization were one of the most significant factors influencing their behaviour
towards people in experimental settings (Turner, 2000).
Social learning and the social environment
In social species such as our domestic animals, social influences on interactions and the
relationship of a given individual with humans are not surprising. In groups of animals, fear
of humans of some animals may be transmitted via behavioural and/or olfactory cues.
Likewise, relaxed and bold behaviour of some individuals towards humans may also
facilitate approach by others. The mother’s reactions to humans may especially be
transmitted to her young through social facilitation, as shown in goats, quails and horses
(Hausberger et al., 2008). Young goats’ responses towards humans were shown to correlate
with the positive responses of an adult goat, especially the dam, towards humans. In horses,
gentle handling of the mother (during the first 5 days after birth) revealed a long-lasting (at
least 1 year) improvement of the foal’s relationship with the familiar human as well as a
generalization to unfamiliar humans (Hausberger et al., 2008). Thus, tame mothers, but
probably also other social models, may play a crucial role in the development of a positive
human–animal relationship in their young.
These effects, however, have received more attention only recently, whereas earlier
studies mostly focused on the effects of direct handling of the young (as discussed above).
Research in sheep and cattle showed that the presence of the dam can limit the efficiency of
early human contact given to lambs or calves on their later relationship with humans (Boivin
et al., 2003; Nowak and Boivin, 2015). In contrast, early gentle interactions provided to beef
suckler calves in the presence of their mothers caused a long-lasting improvement of the
calves’ responses to humans (Probst et al., 2012). The reaction of the dam, the exact time of
treatment relative to birth, other environmental characteristics and the experiences after the
treatment might have led to the different findings. In cats, the presence of the (calm) mother
facilitates interactions of kittens with humans after a short period when kittens prefer to stay
close the mother (Karsh and Turner, 1988). In horses, a direct comparison of early handling
of foals in the presence of the mother, exposure to a motionless human and gentle handling of
the mother (all during the first 5 days after birth) revealed a long-lasting improvement of the
foal’s relationship with humans only for those foals having observed their dam being
handled; exposure to a motionless human had some limited, short-lasting improving effect
(Hausberger et al., 2008). In cats, the mother may have another influential role when
preventing kittens from human contact during sensitive periods by hiding them.
10.4 The Human as a Social Partner

The social nature of a human–animal relationship, where the human is a social partner of the
animals, becomes apparent from animals behaving equally to humans and conspecifics. This
comprises affiliative behaviours but also dominance-related behaviours and species-specific
communication as well as the ability to provide social support (Rushen et al., 2001). In 1950,
Scott (cited in Karsh and Turner, 1988) stated that ‘the behaviour patterns exhibited by dogs
toward human beings are essentially the same as those exhibited toward dogs’. Turner (2000)
states that the human–cat relationship is indeed a ‘two-way partnership, with both parties
adjusting their behaviour to that of their partners’. Cats and dogs seek proximity of humans,
direct affiliative social behaviour towards them and initiate grooming (stroking) interactions
or play – obviously enjoying these interactions. This is also true for farm animals with good
relationships (Figs 10.2 and 10.4), but there is far less scientific evidence. In the 1970s, the
Fulani herdsmen and their cattle were described to form a dual-species social system
fulfilling the following characteristics: after an early mutual socialization of cattle and
humans, the herdsmen are integrated into the herd by taking the role of a leader, a dominant
(humans clearly exhibit dominance over the animals and learn how to do so early in
childhood) and a companion (humans invest much time in stroking animals at body regions
where they often lick each other) (Lott and Hart, 1977, 1979). As a consequence, animals
will follow the call by their herdsmen to be restrained (Hinrichsen, 1979). Furthermore, the
human can give the animal social support and reduce stress in aversive situations, such as
isolation from conspecifics or a veterinary procedure (McMillan, 1999; Waiblinger et al.,
2006). While attachment is not necessarily involved in those behaviours and effects, dogs and
cats are often attached to their owner (Karsh and Turner, 1988; Miklósi, 2015; Vitale Shreve
and Udell, 2015). This may occur also in farm animals that had intensive contact with
humans, especially hand-reared animals (Nowak and Boivin, 2015).
10.5 Human–Animal Communication

Herding a group of cows, moving pigs, riding a horse, calling a cat to come for feeding, a
dog guiding a blind human or a dog and his/her owner playing together – all these situations
involve communication between humans and animals. Different sensory channels are
involved to a different degree according to the situation. Vocalizations, body postures, facial
expressions, movements and the release of pheromones all have communicative value. When
living in a social group, effective communication is extremely important, and thus animals
have developed special communicative behaviours (e.g. threatening and submissive postures
or facial expressions), which can be inborn or learned. Communication between species bears
an enhanced risk of misinterpretation of signals. However, different species can learn to
understand each other, at best via mutual socialization. In human–animal interactions,
imitating species-specific animal signals has been recommended for effective control of farm
animals as well as in the training of dogs. Miklósi (2007) suggests that, besides the use of
species-specific signals, the consideration of the rules of intraspecific social interactions also
seems to improve human–animal communication.
However, there seem to be species-independent features of signalling behaviour. For
instance, threatening or submissive behaviours are often associated with making the body
appear taller or smaller, respectively, and with gazing directly at the opponent or avoiding
gaze (Eibl-Eibesfeldt, 1999). Indeed, avoiding direct gaze and only glancing at the animals,
as well as minimizing body shape such as adopting a quadrupedal posture, squatting, sitting
or lying, allow closer and quicker approach than standing upright and gazing directly towards
the animals (Kendrick, 1998; Waiblinger et al., 2006; Spoolder and Waiblinger, 2008). Many
humans use such signals unconsciously (e.g. squatting to make animals approach). Moreover,
the direction of approach, movement style (quick or slow, deliberately) and even more subtle
cues such as muscle tension and breathing patterns may all play a role (Boivin et al., 2003;
Waiblinger et al., 2006). Dogs react appropriately to a stranger approaching in a friendly or
threatening way, seemingly reacting to the whole pattern of human behaviour (Miklósi,
2007). Facial expression alone seems to be sufficient in discrimination of human emotions in
dogs (Albuquerque et al., 2016). Cats were sensitive to their owner’s emotions, but not to a
stranger’s emotions, in an experiment where the person was sitting stationary, with a happy
or angry facial expression, and body tension: the cats showed more often positive behaviours
and stayed longer in contact with the happy-looking owner than with the angry-looking
owner (Galvan and Vonk, 2016). If other species do have this ability as well is not yet
investigated. It is conceivable that animals reacting to small changes in body posture or facial
expressions within their own species can also do so in the human–animal context provided
sufficient experience with humans, even if there are species-independent features of
emotional signalling.
Conversely, humans can assess aspects of animal emotional state quite reliably by
integrating the whole range of animal behaviour including body postures, movement, tension
and facial expressions, without quantification of single behavioural patterns, as research on
‘qualitative behavioural assessment’ showed (Wemelsfelder and Mullan, 2014; Fleming et
al., 2016).
In addition, with respect to the characteristics of vocalizations, species-independent

communalities seem to exist. There is evidence that some physical properties of sound have
consistent, species-independent effects on the response of an animal receiver and that humans
use these features for effective communication with dogs or horses (McConnell, 1991).
Short, rapidly repeated broad-band notes are used to stimulate motor activity, whereas longer,
continuous narrow-band notes are used to inhibit motor activity by animal trainers with a
wide geographical and linguistic background. Dog pups learn better to come when trained
with short, rapidly repeated sounds (McConnell, 1991). Further, universal acoustic features
seem to be associated with different motivational and affective states in both humans and
animals, facilitating interspecific understanding. Dogs can discriminate between positive
(happy) and negative (angry) vocalizations and cats reacted differently to a novel object
depending on the emotional information in their owners’ voice (Vitale Shreve and Udell,
2015; Albuquerque et al., 2016).
Vice versa, also the vocalizations of animals carry emotional information for the human
receiver. Humans can categorize barks from different situations over chance level
independently of having owned a dog or not (Miklósi, 2015). Similarly, humans could
classify piglet calls from positive or negative situations, especially if they were recorded
during castration. Experience influenced this ability: ethologists working with pigs were
better at recognizing the individual situations (Tallet et al., 2010).
Besides such universal communicative signals, human–animal communication can reach
a very sophisticated level of ‘understanding’ depending on the level of socialization and
intensity of human–animal contact, as well as on individual and breed differences (for review
see Miklósi and Soproni, 2006; Miklósi, 2015; Vitale Shreve and Udell, 2015; Brubaker and
Udell, 2016). A famous example was the horse ‘Kluger Hans’, who was able to ‘calculate’
mathematical tasks. By sensing and reacting to changes in humans’ body tension, he stamped
the correct number of times with his foreleg. This is also a good example that humans may
unconsciously emit signals.
Dogs were shown to be able to comprehend human communicative signals and are aware
of human attentional states. For example, dogs follow the pointing or gaze of humans to find
hidden food or toys. They use the direction of attention of humans to obtain food from the
appropriate person and to perform forbidden actions. They can also communicate the
location of a hidden target to the human. A particular border collie was found to be able to
associate 340 spoken words with objects or specific persons. Again, genetic potential and
experience interact: working dogs have better communicative skills than breeds not selected
for work, and trained working dogs are more skilled than pet working dogs. Interestingly,
cooperative hunting breeds performed better in comprehending pointing gestures than
independently hunting breeds.
The cooperative aspect of human–animal relationship and communication becomes
apparent in guide dogs and their owners. During work, guide dogs have the ability to
exchange roles interactively with the human as the initiator of actions. Dogs also adapt their
communicative behaviour to the visual state of their owner: during a test situation where food
was unreachable for the dogs, guide dogs used sonorous mouth-licking behaviour when
interacting with their blind owners, in contrast to pet dogs with sighted owners (Serpell and
Paul, 1994; Gaunet, 2008).

Not only dogs understand communicative signals of humans; cats, horses and goats were
shown to do so, too, and it is likely that this is the case also for other species given a certain
level of experience. Interestingly, wolves with the same history of socialization with humans
during the first 3–4 months of life are mostly unable to understand and use pointing gestures
at the same age as dogs, but they are able to do so later in life. This supports the hypothesis
that dogs (and other domesticated species) have been selected for enhanced sociocognitive
abilities for living in human social settings, but also points to the effects of socialization and
experience.
10.6 Practical Relevance – The Significance of Human–Animal
Relationships for Welfare
In both farm and pet animals, research has shown that the human–animal relationship has a
huge impact on animal welfare and performance, as well as on human safety and welfare. In
a poor relationship, interactions with humans elicit negative emotions such as fear, causing
behavioural and physiological stress reactions. Further consequences are a higher risk of
injuries and more difficult handling due to escape, defensive or aggressive behaviour. In
addition, health can be impaired due to depression of the immune system or lesions (e.g. claw
lesions). Finally, increased stress also negatively affects reproduction, performance and meat
quality. A good relationship not only minimizes such stress reactions and their negative
consequences, but even has an anti-stress potential, alleviating stress from other sources such
as a necessary veterinary treatment or restraint (Seabrook and Bartle, 1992; McMillan, 1999;
Rushen et al., 1999; Boivin et al., 2003; Hemsworth, 2003, 2004; Waiblinger et al., 2006;
Hausberger et al., 2008; Hemsworth and Boivin, 2011; Hemsworth and Coleman, 2011).
In dogs and cats, a good relationship – based on adequate socialization as well as
adequate interactive behaviour of the owner – helps to prevent behavioural problems
including aggression; both the human and the animal benefit from this. Behavioural problems
are often the reason for placing animals in a shelter or even euthanasia. In both dogs and
horses, the use of punishment for training can cause stress and aggression towards humans,
while animals whose owners use positive reinforcement (in a correct way) show better
compliance.
A special aspect is relevant especially for students of animal behaviour: the potential
influence of the animals’ relationship with humans on outcomes of experimental studies
caused by observer or caregiver influences. The presence of an observer may be sufficient to
markedly influence animal behaviour in observational studies (Davis and Balfour, 1992).
Behavioural tests often require animal handling before or even during tests and, besides the
behaviour of experimenters, the existing human–animal relationship due to former
experiences with caretakers may have considerable effects on the perception of the whole
situation and thus on test outcomes. Testing and assessing the human–animal relationship
itself is a special challenge (reviewed by Waiblinger et al., 2006).
In conclusion, the existing knowledge of the possibilities for achieving a good quality of
relationship between animals and humans should be further implemented to improve animal
and human welfare. However, many questions are still unanswered: for example, questions
about the most effective ways of communication, the perception and importance of special
interactive behaviours and the role and potential of social learning for improved relationships
in farm animals.
Acknowledgements
Thanks to Stephanie Lürzel for her help in language revision.
References
Adachi, I., Kuwahata, H. and Fujita, K. (2007) Dogs recall their owner’s face upon hearing the owner’s voice. Animal
Albuquerque, N., Guo, K., Wilkinson, A., Savalli, C., Otta, E. and Mills, D. (2016) Dogs recognize dog and human
emotions. Biology Letters 12(1), 20150883.
Bernstein, P.L. (2005) The human–cat relationship. In: Rochlitz, I. (ed.) The Welfare of Cats. Springer, Dordrecht, The
Netherlands, pp. 47–90.
Boissy, A., Fisher, A.D., Bouix, J., Hinch, G.N. and Le Neindre, P. (2005) Genetics of fear in ruminant livestock. Livestock
Production Science 93, 23–32.
Boivin, X., Le Neindre, P., Garel, J.P. and Chupin, J.M. (1994) Influence of breed and rearing management on cattle
reactions during human handling. Applied Animal Behaviour Science 39, 115–122.
Boivin, X., Lensink, J., Tallet, C. and Veissier, I. (2003) Stockmanship and farm animal welfare. Animal Welfare 12, 479–
492.
Bradshaw, J.W.S. (1992) The Behaviour of the Domestic Cat. CAB International, Wallingford, UK.
Bradshaw, J.W.S., Casey, R.A. and Brown, S.L. (2012) The Behaviour of the Domestic Cat, 2nd edn. CAB International,
Wallingford, UK.
Brubaker, L. and Udell, M.A.R. (2016) Cognition and learning in horses (Equus caballus): what we know and why we
should ask more. Behavioural Processes 126, 121–131.
Davis, H. and Balfour, A.D. (1992) The Inevitable Bond: Examining Scientist–Animal Interactions. Cambridge University
Press, Cambridge.
Eibl-Eibesfeldt, I. (1999) Grundriss der vergleichenden Verhaltensforschung, 8th edn. Piper, Munich, Germany/Zurich,
Switzerland.
Estep, D.Q. and Hetts, S. (1992) Interactions, relationships, and bonds: the conceptual basis for scientist-animal relations. In:
Davis, H. and Balfour, A.D. (eds) The Inevitable Bond: Examining Scientist–Animal Interactions. Cambridge University
Press, Cambridge, pp. 6–26.
Fleming, P.A., Clarke, T., Wickham, S.L., Stockman, C.A., Barnes, A.L., Collins, T. and Miller, D.W. (2016) The
contribution of qualitative behavioural assessment to appraisal of livestock welfare. Animal Production Science 56,
1569–1578.
Galvan, M. and Vonk, J. (2016) Man’s other best friend: domestic cats (F. silvestris catus) and their discrimination of human
emotion cues. Animal Cognition 19, 193–205.
Gaunet, F. (2008) How do guide dogs of blind owners and pet dogs of sighted owners (Canis familiaris) ask their owners for
food? Animal Cognition 11, 475–483.
Hausberger, M., Roche, H., Henry, S. and Visser, E.K. (2008) A review of the human–horse relationship. Applied Animal
Hediger, H. (1965) Mensch und Tier im Zoo. Albert-Müller Verlag, Rüschlikon–Zürich, Switzerland.
Hemsworth, P.H. (2003) Human–animal interactions in livestock production. Applied Animal Behaviour Science 81, 185–
198.
Hemsworth, P.H. (2004) Human–animal interactions. In: Perry, G.C. (ed.) Welfare of the Laying Hen. CAB International,
Wallingford, UK, pp. 329–343.
Hemsworth, P.H. and Boivin, X. (2011) Human contact. In: Appleby, M.C., Mench, J.A., Olsson, J.A. and Hughes, B.O.
(eds) Animal Welfare. CAB International, Wallingford, UK, pp. 246–262.
Hemsworth, P.H. and Coleman, G.J. (2011) Human–Livestock Interactions: The Stockperson and the Productivity of
Intensively Farmed Animals, 2nd edn. CAB International, Wallingford, UK.

Hennessy, M.B., Williams, M.T., Miller, D.D., Douglas, C.W. and Voith, V.L. (1998) Influence of male and female petters on
plasma cortisol and behaviour: can human interaction reduce the stress of dogs in a public animal shelter? Applied
Hinrichsen, J.K. (1979) Mensch-Tier-Beziehung bei afrikanischen Rindernomaden. KTBL-Schrift 254, 103–110.
Huber, L., Racca, A., Scaf, B., Virányi, Z. and Range, F. (2013) Discrimination of familiar human faces in dogs (Canis
familiaris). Learning and Motivation 44, 258–269.
Karsh, E.B. and Turner, D.C. (1988) The human–cat relationship. In: Turner, D.C. and Bateson, P. (eds) The Domestic Cat:
The Biology of Its Behaviour. Cambridge University Press, Cambridge, pp. 159–177.
Kendrick, K.M. (1998) Intelligent perception. Applied Animal Behaviour Science 57, 213–231.
Kendrick, K.M. (2006) Brain asymmetries for face recognition and emotion control in sheep. Cortex 42, 96–98.
Ligout, S., Bouissou, M.F. and Boivin, X. (2008) Comparison of the effects of two different handling methods on the
subsequent behaviour of Anglo-Arabian foals toward humans and handling. Applied Animal Behaviour Science 113,
175–188.
Lott, D.F. and Hart, B.L. (1977) Aggressive domination of cattle by Fulani herdsmen and its relation to aggression in Fulani
culture and personality. Ethos 2, 174–186.
Lott, D.F. and Hart, B.L. (1979) Applied ethology in a nomadic cattle culture. Applied Animal Ethology 5, 309–319.
Lürzel, S., Windschnurer, I., Futschik, A. and Waiblinger, S. (2016) Gentle interactions decrease the fear of humans in dairy
heifers independently of early experience of stroking. Applied Animal Behaviour Science 178, 16–22.
McConnell, P.B. (1991) Lessons from animal trainers: the effect of acoustic structure on an animal’s response. In: Bateson,
P. and Klopfer, P. (eds) Perspectives in Ethology: Human Understanding and Animal Awareness. Plenum Press, New
York/London, pp. 165–187.
McMillan, F.D. (1999) Effects of human contact on animal health and well-being. Journal of the American Veterinary
Medical Association 215, 1592–1598.
Miklósi, Á. (2007) Dog Behaviour, Evolution, and Cognition. Oxford University Press, Oxford.
Miklósi, Á. (2015) Dog Behaviour, Evolution, and Cognition, 2nd edn. Oxford University Press, Oxford.
Miklósi, Á. and Soproni, K. (2006) A comparative analysis of animals’ understanding of the human pointing gesture. Animal
Nowak, R. and Boivin, X. (2015) Filial attachment in sheep: similarities and differences between ewe–lamb and human–
lamb relationships. Applied Animal Behaviour Science 164, 12–28.
Podberscek, A.L., Paul, E.S. and Serpell, J.A. (2000) Companion Animals and Us: Exploring the Relationships between
People and Pets. Cambridge University Press, Cambridge.
Price, E.O. (2002) Animal Domestication and Behavior. CAB International, Wallingford, UK.
Probst, J.K., Spengler Neff, A., Leiber, F., Kreuzer, M. and Hillmann, E. (2012) Gentle touching in early life reduces
avoidance distance and slaughter stress in beef cattle. Applied Animal Behaviour Science 139, 42–49.
Rushen, J., Taylor, A.A. and de Passillé, A.M. (1999) Domestic animals fear of humans and its effect on their welfare.
Rushen, J., de Passillé, A.M., Munksgaard, L. and Tanida, H. (2001) People as social actors in the world of farm animals. In:
Keeling, L.J. and Gonyou, H.W. (eds) Social Behaviour of Farm Animals. CAB International, Wallingford, UK, pp. 353–
372.
Sambraus, H.H. and Sambraus, D. (1975) Prägung von Nutztieren auf Menschen. Zeitschrift für Tierpsychologie 38, 1–17.
Schmied, C., Boivin, X. and Waiblinger, S. (2008a) Stroking different body regions of dairy cows: effects on avoidance and
approach behavior towards a human. Journal of Dairy Science 91, 596–605.
Schmied, C., Waiblinger, S., Scharl, T., Leisch, F. and Boivin, X. (2008b) Stroking of different body regions by a human:
effects on behaviour and heart rate of dairy cows. Applied Animal Behaviour Science 109, 25–38.
Seabrook, M.F. and Bartle, N.C. (1992) Human factors. In: Phillips, C. and Piggins, D. (eds) Farm Animals and The
Environment. CAB International, Wallingford, UK, pp. 111–125.
Serpell, J. and Jagoe, J.A. (1995) Early experience and the development of behaviour. In: Serpell, J. (ed.) The Domestic Dog.
Its Evolution, Behaviour and Interactions with People. Cambridge University Press, Cambridge, pp. 79–102.
Serpell, J. and Paul, E. (1994) Pets and the development of positive attitudes to animals. In: Manning, A. and Serpell, J.
(eds) Animals and Human Society: Changing Perspectives. Routledge, New York, pp. 127–144.
Spoolder, H.A.M. and Waiblinger, S. (2008) Pigs and humans. In: Marchant-Forde, J.N. (ed.) The Welfare of Pigs. Springer,
New York, pp. 211–236.
Tallet, C., Spinka, M., Maruscáková, I. and Simecek, P. (2010) Human perception of vocalizations of domestic piglets and
modulation by experience with domestic pigs (Sus scrofa). Journal of Comparative Psychology 124, 81–91.
Turner, D.C. (2000) Human–cat interactions: relationships with, and breed differences between, non-pedigree, Persian and
Siamese cats. In: Podberscek, A.L., Paul, E.S. and Serpell, J.A. (eds) Companion Animals and Us: Exploring the
Relationships Between People and Pets. Cambridge University Press, Cambridge, pp. 257–271.
Uvnäs-Moberg, K. (1997) Physiological and endocrine effects of social contact. In: Carter, S., Lederhendler, I.I. and
Kirkpatrick, B. (eds) The Integrative Neurobiology of Affiliation. New York Academy of Science, New York, pp. 146–
163.
Uvnäs-Moberg, K. and Petersson, M. (2005) Oxytocin, ein Vermittler von Antistress, Wohlbefinden, sozialer Interaktion,
Wachstum und Heilung (Oxytocin, a mediator of anti-stress, well-being, social interaction, growth and healing.)
Zeitschrift für Psychosomatiche Medizin und Psychotherapie 51, 57–80.
Vitale Shreve, K.R. and Udell, M.A.R. (2015) What’s inside your cat’s head? A review of cat (Felis silvestris catus)
cognition research past, present and future. Animal Cognition 18, 1195–1206.
Waiblinger, S., Boivin, X., Pedersen, V., Tosi, M., Janczak, A.M., Visser, E.K. and Jones, R.B. (2006) Assessing the human–
animal relationship in farmed species: a critical review. Applied Animal Behaviour Science 101, 185–242.
Wemelsfelder, F. and Mullan, S. (2014) Applying ethological and health indicators to practical animal welfare assessment.
In: Mellor, D.J. and Bayvel, A.C.D. (eds) Animal Welfare: Focusing on The Future. OIE, Paris, pp. 111–120.
Part B
Part B
Species-Specific Behaviour of Some

Important Domestic Animals
Editor’s Introduction
For the second part of the book, we turn our attention to the actual behaviour seen in some of
the more common and economically important domestic animals. Based on the theoretical
approach in the first part, it is now possible to go into detail about what animals actually do.
The species treated in the book have been selected because they are the most common
domestic animals and some are the focus of much ethical concern. They are also those where
there has been most research performed. For most of the species we will meet, there is
considerable knowledge of the behaviour of the ancestors, and of the ways in which these
animals react and behave under different husbandry conditions.
For each species, the authors have attempted to provide information on a number of
different aspects. The origin and domestication history is given some emphasis in all
chapters. Social behaviour and communication, foraging and feeding habits, mating
behaviour and other aspects of reproduction are then treated, although the exact order and
content of the sections may vary somewhat between chapters, dependent on the relevance for
the species concerned. Behavioural ontogeny is a further subject which is treated as far as
information is available.
Most of the text is devoted to the normal behaviour of the species, but in several chapters
we also meet a section of applied problems. This covers typical issues raised by housing and
husbandry conditions, including common behavioural disorders. Where information is
available, the authors have attempted to provide some ideas about possible prevention or
treatment of the types of problem described.
With the information provided in this second part of the book, it is hoped that students of
animal behaviour will be well equipped to differentiate between normal and abnormal
behaviour when animals are observed under practical husbandry conditions.
11 Behaviour of Domesticated Birds: Chickens,
Turkeys and Ducks
J.A. Mench
11.1 Origins
Many avian species have been domesticated for use by humans (Appleby et al., 2004),
including ground-dwelling fowl (chickens, turkeys, guinea fowl, quail and pheasants),
waterfowl (ducks and geese), the flightless ratites (ostrich, emu and rhea) and pigeons.
Chickens (Gallus gallus domesticus), ducks (especially Pekin ducks, Anas platyrhynchos)
and turkeys (Melagris gallopavo) are the most commonly used species worldwide, with an
estimated 56 billion chickens, 2.7 billion ducks and 650 million turkeys raised commercially
for meat or eggs annually.
Chickens were derived mainly from the red junglefowl (Gallus gallus), modern forms of
which are found in India, Burma, Malaysia, Thailand and Cambodia. Junglefowl were
domesticated over 8000 years ago in South-East Asia, and were then taken by humans to
China and from there to Europe. There were domestic fowl in many European countries by
100 BCE, which were brought to the Americas around 1500 CE. During the initial stages of
domestication the fowl was probably valued mainly as a sacrificial or religious bird, or for
cockfighting. However the Romans created specialized breeds, including highly productive
egg-layers, and formed a complex poultry industry. This industry collapsed with the fall of
the Roman Empire and did not resume on a large scale until the 19th century, when there was
an emphasis on breeding fowl for both ornamental and production traits. Modern breeds are
derived mainly from the two types of fowl developed during this period, Asiatic and
Mediterranean. These were then hybridized in the 20th century for commercial use to create
egg-laying and meat strains, referred to as laying hens and broiler (meat) chickens. Laying
strains have been intensely selected for traits such as early onset of egg-laying, a high rate of
egg production, egg quality characteristics (shell strength, egg size) and food conversion
efficiency. Broiler chickens have similarly been selected for food conversion efficiency, but
in addition intense selection pressure has been applied for rapid growth, a high ratio of white
to dark meat (larger breast size) and meat yield, resulting in a chicken that reaches full adult
body size in only 6 weeks, as compared with 17 weeks for an egg-laying chicken.
Turkeys are native to the Americas and wild populations are still widespread throughout
North and Central America. They were domesticated about 2000 years ago in Mexico. The
early Spanish explorers brought turkeys back to Spain, and they were quickly distributed
throughout Europe. The domesticated form of the turkey was then taken to North America by
the European settlers in the 17th century. Breeding focused originally on plumage
characteristics for show purposes. By the 20th century, however, selection programmes had
begun to focus on meat production, and as a consequence turkeys have been selected for
many of the same traits as broiler chickens. In addition, there was selection against the dark
plumage of the wild turkey in order to remove the hormone melanin from the feather
follicles, which improved the appearance of the carcass.
The Pekin duck was derived from the mallard, a species widely distributed in North
America and Eurasia. Mallards were probably domesticated in Asia about 4000 years ago,
and were farmed by the Romans for meat. The mallard has given rise to a large number of
domesticated breeds of duck, including breeds that have a high rate of egg-laying. However,
the most common breed is the Pekin from China, which is used for meat and feather (down)
production. Like broilers and turkeys, commercial Pekin ducks have been selected for rapid
growth and good feed conversion.
These species are raised on a limited scale for local consumption around the world.
Under these conditions they are often kept outdoors in small flocks, sometimes with indoor
shelter provided for inclement weather. For large-scale commercial production, however,
poultry are typically housed indoors to provide better environmental control, allow
automation of feeding, watering, egg collection and manure collection and disposal, and to
protect them from diseases transmitted by wild birds and rodents.
There are two basic types of commercial housing: floor systems and cages (Fig. 11.1).
Cages are the primary type of commercial housing used worldwide for mature egg-laying
hens. Meat-type birds (broilers, turkeys, ducks) and breeding flocks are typically housed in
floor systems, with the birds being housed on flooring made of wire, slats, litter (e.g. wood
shavings) or some combination of these. Because of concerns about the restriction of natural
behaviour in some developed countries, an increasing number of egg-laying hens are being
housed in so-called ‘furnished cages’, which contain perches, nest boxes and dust baths, in
floor systems, or outdoors in free-range or pasture-based systems. A more detailed review of
the different kinds of housing systems in commercial use for poultry can be found in Appleby
et al. (2004).
Fig. 11.1. The conventional (battery) cage is the most common system worldwide for
housing hens (a), although newer designs of cages (furnished cages) containing nest boxes,
perches and dust baths are becoming increasingly common in some developed countries, as
are free-range (b) production systems. Meat-type birds like ducks (c), turkeys and broiler
chickens are usually housed in floor systems containing thousands or tens of thousands of
birds. (Photo courtesy of Darrin Karcher.)
This chapter will focus mainly on the behaviour of chickens, not only because they are
the most commonly used species but also because their behaviour has been best-studied.
Information about the behaviour of turkeys and ducks will also be provided where available.
As discussed in Chapter 2, many domesticated species share behavioural traits that made
them suitable for domestication, so many of the topics covered in this chapter will also apply
to other avian species that are used for egg or meat production.

All poultry species are highly social, although their wild ancestors show different forms of
social organization (Mench and Keeling, 2001). Some live in small relatively stable groups.
Junglefowl, for example, typically live in groups comprised of one male and several females,
with other males being solitary or living in small all-male groups. Turkeys may also live in
mixed-sex groups during the breeding season, but the rest of the year most commonly stay in
all-male or all-female groups. The social groups of mallard ducks are less stable, with the
composition of social groups changing after the breeding season and after migration to
breeding and feeding grounds.
As in all social species, there is extensive communication among group members.
Poultry species have excellent colour vision and acute hearing, and communication occurs
via visual and vocal signals. There is also increasing evidence that olfaction might play a role
in social recognition and communication, since chickens have individual body odours which
originate from a gland at the base of the tail (known as the preen, or uropygial, gland), and
birds can detect one another’s odours (Nicol, 2015).
Postures and displays are used to signal threats and social submission, and particularly
elaborate displays are given during courtship (see below). Features of the head and neck are
particularly important in some species for social recognition and communication. In
chickens, comb size and colour are affected by sex hormone levels and are indicators of
social status. Turkeys have a pale featherless neck area that changes colour to red and blue
during social interactions, and the fleshy, pendant snood of male turkeys becomes engorged
and enlarged during aggression and courtship (Fig. 11.2).
Fig. 11.2. Male turkeys (toms) have brightly coloured faces that play a role in social
communication. During mating, the pendant snood of the tom becomes engorged with blood
and enlarged, and the pattern of colour on the tom’s neck area also changes during social
interactions. (Photo courtesy of Butterball Turkeys, LLC.)
Most poultry species also have an extensive vocal repertoire. Calls can serve a variety of
functions, including warning about approaching predators, decreasing the distance between
flockmates (contact calls), signalling threat or submission, or attracting offspring or other
flock members to food (Fig. 11.3). The most striking vocalizations are the ones that males
make during territorial defence, like the crows of roosters. These calls carry long distances
and thus are an effective way for a male to defend a territory without having to directly
confront the males on neighbouring territories.
Fig. 11.3. Poultry have an extensive vocal repertoire that they use to communicate with
flockmates. This picture shows a hen that has been attracted to a patch of food by another
bird that has given a food call, called the ‘tidbitting’ call. (Photo courtesy of Richard
Blatchford.)
Calls can also be used to signal social dominance. The crow is individually distinctive
and roosters use the quality and rate of crowing to assess the dominance status of other
roosters. Social dominance hierarchies (peck orders) are a common feature of social
organization in poultry. As the name implies, these orders are formed and maintained by
aggressive pecking directed towards the head region of more subordinate birds, which in turn
show submissive behaviour (Fig. 11.4). In mixed-sex flocks, males and females generally
develop separate dominance hierarchies and rarely show aggression towards one another. In
established flocks, aggressive and submissive behaviours are usually subtle and difficult to
observe, although there sometimes can be chasing, pecking and even fighting, particularly
among males.
Fig. 11.4. Social hierarchies in poultry are maintained through aggressive and submissive
interactions. The bird on the left is adopting an aggressive posture, while the bird on the right
is showing submission. (From Appleby et al., 2004.)
Despite periodic aggressive interactions, even when poultry are kept in large areas where
they could avoid one another they tend to cluster together. The tendency to form groups
rather than to live independently evolved primarily for protection against predators. Even in
the absence of predators, however, poultry often move together as a flock. They also tend to
synchronize their behaviours, either because of circadian effects (see below) or due to social
facilitation, with birds ‘copying’ one another’s behaviour. Observing and copying other birds
may allow a bird to exploit new resources (e.g. new food types or sources) or learn new
behaviours. For example, chickens that have an opportunity to observe a trained hen peck a
key to obtain food learn this task much more quickly than chickens that do not have this
experience (Johnson et al., 1986).
In commercial conditions poultry may be kept in extremely large flocks containing
thousands or even tens of thousands of birds (see Fig. 11.1), or in small groups but under
crowded conditions where the birds cannot easily avoid one another (like cages). Except for
breeding flocks, birds are also typically kept commercially in single-age, single-sex flocks
instead of the mixed-age and -sex flocks that would be the norm in the wild. As is true of
many domesticated animals, poultry are relatively tolerant of these variations in social
groupings. However, social behaviour problems can arise in commercial settings (Mench and
Keeling, 2001). Turkeys, for example, can be particularly aggressive towards one another in
commercial flocks, sometimes pecking each other so severely that the head wounds result in
death (Sherwin and Kelland, 1998). They are therefore usually kept in dim lighting to reduce
pecking behaviour.
Social problems are also sometimes seen in female flocks. Subordinate laying hens may
be pecked continuously by other birds – they have heads and combs scarred from pecking,
poor body condition and spend much of their time trying to avoid other birds. This problem
is particularly noticeable in moderate-sized groups, since in small groups the hens know one
another and have a stable dominance hierarchy, whereas in large groups they seem to develop
non-aggressive social strategies for establishing dominance. Even in the small groups typical
of cages, however, more dominant hens may sometimes prevent subordinate hens from
accessing the feeder to such an extent that the subordinate hens lose body condition and stop
producing eggs (Cunningham and van Tienhoven, 1983). In non-cage systems, housing
roosters with the hens can help decrease problems with bullying of subordinates (Odén et al.,
2000), since roosters suppress aggression among hens.
The most problematic social behaviours seen in commercial flocks are severe feather-
pecking and cannibalism (Nicol et al., 2013). These are abnormal behaviours and are more
common in large than small flocks. Severe feather-pecking is the pecking and pulling of
feathers from another hen. Unlike aggressive pecks, which are directed towards the head,
feather-pecks are directed towards body regions like the areas near the vent and preen gland,
and the wings, back and tail. The pecking movements involved resemble feeding pecks rather
than aggressive pecks, and there is evidence that feather-pecking is redirected foraging
behaviour and can be reduced by providing foraging materials. For the recipient, having
feathers pulled out is painful, and birds with large areas of exposed skin have more difficulty
regulating their body temperature.
Cannibalism, which involves the pecking and tearing of the skin and underlying tissues
of another bird, is an even more serious problem and one that can result in extremely high
mortality in flocks. Cannibalism sometimes follows from feather-pecking, but can also arise
independently. One situation in which it starts is when a hen has just laid an egg and her
cloaca is still partly everted. Other hens are attracted to peck at this area, especially if the skin
becomes broken and bleeds, leading to further pecking and consumption of the flesh.
Despite a great deal of research, the exact causes of feather-pecking and cannibalism are
still not completely understood. The extent of these problems can vary enormously from one
flock to another even when the management of the flocks is very similar, and flocks do not
necessarily experience both problems at the same time. Factors affecting the incidence of
both types of behaviour include stocking density, genetics, lighting, whether or not foraging
materials are provided, and feed composition and form. One of the most important factors for
fowl is group size, and these behaviours are more problematic in non-cage systems where
large numbers of hens are housed together. This might be because cannibalism is socially
learned, which leads the problem to spread quickly through the flock.
Once an outbreak of feather-pecking or cannibalism starts it can be difficult to stop it,
which is why poultry producers often use preventive methods. These methods include
keeping the birds in dim or red light (which makes it more difficult for them to see wounds
that could trigger pecking), fitting them with devices (like contact lenses or ‘spectacles’) that
reduce their ability to see, and beak- or bill-trimming. This last is the most common method.
It has traditionally involved using a heated blade to remove about one-third to one-half of the
upper beak or bill, and is effective in reducing the damage that the birds can inflict upon one
another. This practice has been criticized, however, because it causes short-term (and
sometimes long-term) pain, and as a consequence it has been banned in some countries.
There is a newer method for reducing beak length which involves using an infrared beam to
make a small hole in the beak; the beak tip then falls off several days later. This method
appears to cause fewer welfare problems for the birds in terms of long-term pain and beak
malformation, but it does still cause short-term pain (Janczak and Riber, 2015). A promising
alternative is genetic selection to reduce the propensity for birds to show feather-pecking and
cannibalistic behaviour. This approach has been quite successful in reducing these problems
in caged laying hens (Muir, 1996), but it is presently unclear how well it will work for non-
cage flocks, where these problems are most severe.
11.3 Foraging, Feeding and Drinking

Under natural conditions, poultry and their wild ancestors have a varied diet. All species feed
on plants and, depending upon the species, may consume grasses, shrubs, roots, leaves and
berries. They may also eat invertebrates, and some species even eat small vertebrates like
lizards and mice.
Foraging for food is a very important part of the natural behaviour of poultry; for
example, under semi-natural conditions fowl devote a large proportion of their day to
foraging activities (Fig. 11.5). Components of foraging behaviour include ground-pecking,
ground-scratching and grazing. Ducks also filter edible items out of the water by dabbling
using their bills.
Fig. 11.5. A flock of chickens foraging on the range. The rooster is searching the ground and
pecking at edible food items. It is common for the birds to flock together during foraging,
which is a form of protection against predators.
Under commercial conditions, concentrated feed is provided in troughs and is usually

readily available throughout the day. However, poultry still spend considerable time foraging
if an appropriate foraging substrate, like loose litter, is provided. This indicates that they are
motivated to forage even in the absence of a need to do so.
In housing environments without a foraging substrate, such as conventional cages for
laying hens, birds do not have access to such material and may instead spend a substantial
portion of time manipulating the food in the feed trough by flicking it back and forth using
movements of their beak. This can lead to the food being tossed out of the trough and wasted,
which is an economic issue for egg producers.
There has been a considerable amount of research on food selection by chickens
(Appleby et al., 2004). Chickens have a well-developed sense of taste and reject potential
food items that are acidic, bitter or extremely salty. Visual and tactile cues are also important
– both chicks and adults preferentially peck at items of particular colours and have a
preference for pecking at and ingesting small, rounded particles. These preferences are,
however, affected by their experience with particular food types and may also vary
depending upon the birds’ nutritional status. Odour also plays some role in food selection,
since chickens can learn to associate odours with particular food types (Nicol, 2015). In the
wild, these visual and olfactory aspects of food selection would help the birds avoid
consuming toxic plants and insects.
Under natural conditions, wild birds are faced with an array of food items that differ in
nutritional quality, from which they must select a diet to meet their nutritional needs.
Domestic birds show a similar ability to select a nutritionally adequate diet when given a
choice of different feedstuffs, and will adjust their intake of protein, energy, minerals and
vitamins (Hughes, 1984). Poultry do, however, develop a preference for the food to which
they are accustomed and become reluctant to consume new foods that differ in colour, taste,
odour or texture from their typical food (food neophobia). A major change in diet can thus
cause them to reduce their feed intake and consequently result in a reduction in growth or egg
production.
Social factors have an important influence on feeding behaviour. Poultry show a
propensity to feed as a group (i.e. to feed synchronously) and feeding behaviour can be
triggered by the sight or sound of other birds feeding. Individually caged hens will even
synchronize their feeding behaviour with that of the hens in the neighbouring cages. If
insufficient feeder space is provided for all birds to feed simultaneously, they may need to
desynchronize their behaviour in order for all birds to consume enough feed. One
conspicuous feeding behaviour that has a social component is food running. This behaviour
is most commonly performed by younger chickens after they pick up a large food item, and
involves them running rapidly while holding the food item in their beak. This probably
attracts the attention of other birds in the flock so that they help break the food item into
smaller edible pieces.
Bouts of feeding are alternated with brief bouts of drinking. Chicks are attracted to peck
at shiny items, which in a natural setting would likely result in them finding, pecking at and
ingesting water. Poultry drink by scooping up water in their beaks or bills, and then raising
their heads so that the water runs down their oesophagus (Fig. 11.6). In commercial settings
poultry may be offered water in trough or bell-type drinkers or cups, which allows them to
perform this natural drinking motion. However, to prevent water spillage which can lead to
the litter and manure becoming wet, it is now more common for birds to be given nipple
drinkers instead of troughs or cups. Nipple drinkers require the birds to drink in an unnatural
way, although they can learn to do so and in fact develop various strategies for activating the
nipple and consuming water. Poultry do become accustomed to using particular drinker types
and may have difficulty drinking if they are moved to a house where there is a type of drinker
that they have not used before – in this case, they may have to be shown where the drinkers
are located or be taught to activate them.
Fig. 11.6. Birds normally drink by scooping up water and then elevating their heads slightly
so that the water moves down their oesophagus by gravity (a). Because of concerns about
water leaking and making the litter or manure wet, it is common to provide the birds with
nipple waterers, which require them to learn to drink in an unnatural position (b). (From
Houldcroft et al., 2008.)
11.4 Body Maintenance

It is important for birds to keep their feathers clean and in good condition. The feathers
provide a covering and insulating layer that helps to maintain body temperature and prevent
injury to the skin. Birds perform two primary behaviours to maintain plumage condition:
preening and dust- or water-bathing. Chickens preen frequently throughout the day; during a
preening bout, the bird uses its beak and face to distribute oil from the uropygial gland
through its feathers (Fig. 11.7a). The beak is also used at this time to align the barbs of the
feathers and to dislodge external parasites like mites and ticks.
Fig. 11.7. Ground-dwelling poultry maintain their feathers in good condition by preening to
distribute oil through their feathers (a), and by dust-bathing to work loose material through
their feathers, which removes any excess oil. Dust-bathing is often a highly social activity, as
can be seen here (b). (Photos courtesy of Richard Blatchford.)
Water-bathing by ducks, and dust-bathing by other poultry species, serves to remove dirt
and excess oil from the feathers and improves feather structure. Dust-bathing has been
studied more than water-bathing (Olsson and Keeling, 2005). A dust-bathing bout begins
with the bird lying down and pulling loose substrate close to its body (Fig. 11.7b). The bird
then rubs itself on the substrate and shakes its wings and body to toss the material on its back
and work it through the feathers. The dust bath ends with the bird standing up and shaking its
body to remove the excess loose material and to realign its feathers. Fine materials such as
sand or peat are preferred for dust-bathing, probably because they are superior at penetrating
the feathers to reach the downy portions.
To fully perform dust- or water-bathing birds must be provided with either loose material
or water that is deep enough for them to immerse themselves. However, waterfowl raised
commercially are rarely given bathing water due to cleanliness and disease problems, and
poultry housed in cages are generally not provided with loose material. This is a welfare
concern and has led to a great deal of research on the motivational aspects of dust-bathing
behaviour. Laying hens housed in cages without litter material will carry out dust-bathing
movements on the wire floor of their cages, so-called ‘sham’ dust-bathing, although it is
unclear whether this behaviour actually satisfies their motivation to dust-bathe.
11.5 Diurnal Rhythm

Most behaviours in poultry do not occur at random but instead show distinct daily rhythms.
All poultry species are diurnal, and hence sleep at night and are active during the daylight
hours. The behaviours that show the strongest rhythms are those related to feeding, egg-
laying, mating, grooming and sleeping. Feeding behaviour occurs in bouts, or meals, with
most bouts occurring in the first few hours after the lights go on (or dawn, when there is
natural lighting) and then again late in the day before the lights go off (or dusk, when there is
natural lighting). Because poultry do not typically feed at night, the morning feeding peak
allows them to refill their food storage organ (the crop), which has become depleted
overnight. The crop is then filled again for the night during the feeding peak late in the day.
In laying hens, this late-afternoon feeding peak also correlates roughly with the start of
eggshell calcification, at which time there is an increase in the hen’s calcium requirements.
Egg-laying in many poultry species occurs at about the time the lights go on. This is
approximately 24 h after ovulation and reflects the period of time required for the egg to
form as it moves through the oviduct. This means that the pre-laying behaviours associated
with selecting a nest site and building a nest also show a strong diurnal rhythm, since they
occur shortly before the egg is laid. The daily cycle of mating in breeding birds is also related
to the egg-laying cycle because the hen’s fertility is lower around the time of oviposition.
Mating is therefore most frequent in fowl in the afternoon because their eggs are laid in the
morning.
Like feeding, preening behaviour occurs primarily in the morning and late afternoon.
However, dust-bathing behaviour takes place early in the afternoon. The strongly diurnal
rhythm of this behaviour can be used to commercial advantage – to prevent hens from laying
their eggs in the dust bath or on the litter floor which makes the eggs difficult to collect, hens
can be given access to dust-bathing material only in the afternoons, after the peak of egg-
laying.
Although periods of rest may occur throughout the day, most resting and sleeping occurs
at night. During periods of deep sleep poultry tuck their heads under their wings, although
they may also doze with their heads upright or on the ground and their eyes closed or
partially open. The smaller poultry species are very vulnerable to ground and aerial predators
when asleep, so they prefer to roost in areas that offer some protection, for example on the
water (waterfowl) or on elevated roosting areas (chickens and turkeys). Roosting is a highly
motivated behaviour – hens will push through a weighted door to gain access to a perch at
night (Olsson and Keeling, 2002) and 100% of hens in commercial houses will perch on
elevated perches at night if they are given the opportunity to do so. Chickens will also stand
and rest on perches during the day, but this behaviour is more variable than night-time
perching in terms of the percentage of birds in the flock perching and the amount of time the
perches are occupied.
A number of factors can affect the diurnal rhythm of behaviour (Appleby et al., 2004).
For example, feeding rhythms are affected by the form and density of the diet because these
influence how long it takes the bird to consume an adequate meal. They are also affected by
genetics, with birds selected for a high rate of feed intake and weight gain, like broilers and
turkeys, more likely to eat throughout the day and even at night. Egg-laying can be delayed
by factors such as human disturbance or social interference from other birds (e.g. a dominant
hen preventing a subordinate hen from entering a nest box) – in this case, laying may not be
accompanied by pre-laying behaviour and the egg may just be laid while the hen is
performing other activities during the day.
The most important factor affecting all diurnal rhythms, however, is the light cycle.
Poultry have photoreceptors not only in their eyes but also in the pineal gland in their brain,
and thus are very sensitive to light stimulation. Light acts as a ‘time keeper’ (a zeitgeiber)
controlling the circadian rhythms of behaviour. The commercial poultry industry raises birds
under many different kinds of light cycle. Some of these provide only very brief periods of
either light or darkness, or levels of illumination that are so low that there is no distinct light–
dark cycle. Broiler chickens and turkeys, for example, are often kept in very dim lighting to
decrease activity and hence promote more rapid growth. This can lead to a marked change in
their diurnal rhythms of behaviour. Figure 11.8 shows the patterns of foraging and preening
behaviour in broiler chickens given 16 h of light and 8 h of darkness per day, but kept in
either dim (5 lux) or brighter (50 lux) lighting during the light period. The broilers kept in
dim lighting show a much less distinct rhythm of behaviour and tend to distribute their
foraging and preening behaviour more evenly throughout the day and night.
Fig. 11.8. Diurnal rhythms of behaviour are strongly influenced by the light cycle. Broiler
chickens reared with a light cycle that has a strong light–dark contrast show a much more
pronounced rhythm of foraging (b) and preening (d) behaviour than do broilers reared in dim
lighting with little light–dark contrast (a, c), as shown by the average numbers of broilers
performing those behaviours at different times of day over a 48 h period. (From Alvino et al.,
2009.)

The ancestors of poultry species show a variety of different types of mating system, ranging
from promiscuity (polyandry, polygyny or both) to monogamous pair bonds that last for one
or more seasons (Mench and Keeling, 2001). Individual males may set up territories to which
females are attracted during the breeding season, or instead may associate with a harem of
females year-round. Alternatively, males and females may congregate during the breeding
season at special breeding grounds, called leks, where the females select the males with
which they will mate.
Fowl have a harem polygynous mating system, with a dominant male maintaining a
territory and mating with the females that live in his territory. Turkeys may also form harems
under some circumstances, but usually mate in leks. Related male turkeys occupy a breeding
site on the lek and attract females to their group by displaying; the dominant male turkey at a
particular site usually secures most of the matings during the height of the breeding season.
Ducks are more variable in their behaviour. They sometimes form monogamous pair bonds,
but also may mate promiscuously.
Even in apparently promiscuous mating systems, birds (and particularly female birds) do
not mate randomly but are selective about their choice of mates. Mate selection has been
best-studied in junglefowl. Female junglefowl use a variety of physical characteristics to
assess the suitability of an unfamiliar male, including comb colour, eye colour, spur length
and comb size. Comb size is one of the most important cues and, since males that are less
healthy have smaller combs, this may be one method females have for assessing the male’s
fitness (Zuk et al., 1990). If the hen is familiar with the male, his dominance status and
courtship behaviour are more important selection factors than his physical features.
Once potential mates are selected, courtship consists of a chain of stimulus–response
patterns between the male and the female (Fig. 11.9). The male usually is the obvious
initiator of the sequence, but females do encourage courtship by approaching males. Male
courtship displays are often very elaborate, involving noises, vocalizations, conspicuous
postures and spreading the feathers or wings to make the body look larger. Males sometimes
also show colour changes or enlargement of certain body parts, such the snood of male
turkeys mentioned previously. The female signals her receptiveness to mating by crouching
and everting her cloaca, which allows the male to mount her and mate.
Fig. 11.9. Mating behaviour involves an often elaborate sequence of behaviours between the
male and female. When males, like this tom, display by posturing, spreading their tail and
showing neck colour changes, the female may avoid the male or crouch to indicate that she is
receptive. If she crouches, the turkeys then engage in a copulatory sequence that ends with
the tom inseminating the hen. (From Hale et al., 1969.)
In commercial systems, mating patterns may change because of the ways in which the
birds are managed (Appleby et al., 2004). Poultry are typically allowed to mate naturally, but
are often kept in large groups. Under these circumstances, factors such as social dominance
are much less likely to affect mate selection, since the birds probably cannot easily recognize
one another. The ratio of males to females in commercial settings is designed to encourage
promiscuous mating to ensure the maximum fertilization of eggs. In such settings both males
and females may mate a number of times during the day even though such frequent mating is
not necessary to produce fertilized eggs, since hens can store viable sperm in a specialized
gland for days to weeks.
Mating behaviours can be a source of injury in commercial flocks. Males in broiler
breeder flocks are sometimes abnormally aggressive to females, attempting to force
copulations and injuring or even killing the females in the process. These males show little
courtship behaviour, which could be the reason why females are less receptive to mating with
them (Millman et al., 2000). Roosters can also severely scratch females even during normal
mating and it is therefore common for the toes or toe nails of roosters to be trimmed at the
hatchery using a microwave toe trimmer. This procedure reduces injuries to females, but can
cause short-term pain for the roosters and result in them walking less when they are older
(Fournier et al., 2015).
The exception to natural mating under commercial conditions is in turkeys, which have
been selected for such large breast size that the males can no longer get close enough to the
females to copulate. For this reason, turkey hens are artificially inseminated with semen
collected from breeding toms kept in a separate flock.
11.7 Egg-laying, Incubation and Behaviour at Hatching

The behaviour associated with egg-laying is expressed relatively inflexibly, probably because
it is under genetic and physiological control. In fowl, pre-laying behaviour is triggered by the
release of oestrogen and progesterone from the follicle after it is ovulated. Once it is
triggered, the hen is very strongly motivated to find an appropriate nest site in which to lay
her egg. A hen examines different potential sites, selects one and then commences to make a
rudimentary nest by using her feet and rotating her body to create a hollow. During this time
the egg has continued to develop as it moves down the oviduct, and it is laid after nesting
behaviour is complete. The entire sequence, from nest searching to oviposition, takes about
1–2 h. The searching phase of sequence may be extended when hens do not have access to a
suitable nesting site, as in conventional cages.
Eggs laid outside the nest can be a significant problem in commercial production, since
such eggs are difficult to collect and can become cracked and dirty, affecting their
hatchability and economic value. So-called ‘floor laying’ is variable both within and between
systems, but there is increasing understanding of the factors that affect this problem (Appleby
et al., 2004). Nest-box design is important to hens. They tend to prefer dark secluded areas as
they would in the wild, as these areas are better protected from predators, although their
preference is affected by early experience. However, nest-box design probably plays more of
a role in determining which box a hen selects to lay her eggs in, rather than whether she lays
them in a nest box or on the floor. What may be more important to the latter is how
accessible the nest boxes are: most nest boxes in commercial houses are above ground level,
and hens reared with no experience of perching may not be able to reach them easily and
hence instead lay their eggs on the floor (Appleby et al., 1986). In addition, given that egg-
laying in the flock is synchronized, birds may have to lay their eggs on the floor if there are
not enough nests available for all birds that are laying at the same time (Makagon and
Mench, 2011), although they may instead choose to nest gregariously (Fig. 11.10).
Fig. 11.10. Floor laying by poultry can indicate that the nest boxes are inadequate in some
way (e.g. not providing sufficient space for all birds to use the boxes during the peak laying
period; see Makagon and Mench, 2011). However, birds may choose to nest gregariously
even if there are plenty of nest boxes available. (Photo courtesy of Maja Makagon.)
Under natural circumstances, once a clutch of eggs is laid the hen sits on the eggs nearly
continuously to keep them warm until they hatch. Incubation behaviour is triggered by
release of the hormone prolactin, which also causes the cessation of egg-laying. However,
genetically selecting commercial hens to produce a large number of eggs also inadvertently
resulted in selection against incubation behaviour. Therefore most commercial strains of hens
simply leave the egg after oviposition, and the pre-laying and laying cycle then begins again
the next time an egg is about to be laid.
Social interactions between parents and offspring occur in birds even before hatching
(Rogers, 1995). Calls made by the developing embryos stimulate the incubating parent to
turn the eggs or to return to the nest to resume incubation. Embryos also respond to the calls
and behaviour of the parent with calls that further influence the parent’s behaviour. Since the
incubation of eggs is automated in commercial poultry production, these parent–offspring
interactions are absent. However, the embryos also vocalize to one another, which causes the
development of the less advanced embryos to accelerate and thus ensures that all of the eggs
hatch at around the same time.
Environmental factors during incubation can influence later behaviour. Maternal
androgens and oestrogens deposited into the egg, for example, affect brain development and
the sexual behaviour of the offspring (Nicol, 2015). Light, which penetrates the egg when the
hen stands to turn the eggs or when she leaves the nest, is another important factor. The
embryo becomes light-sensitive at about 3 days of incubation, and light stimulation of the
pineal gland and developing eyes results in the two hemispheres of the brain becoming
specialized for different functions (Rogers, 1995). In commercial hatcheries poultry eggs are
typically incubated in complete darkness, but this practice can have negative effects. For
example, chicks that are incubated in complete darkness are more fearful than those that
receive light stimulation during embryonic life (Archer and Mench, 2017).
Domestic poultry are precocial when they hatch, which means that their eyes are open,
they are covered with down, they are mobile and they require little parental care. Under
natural conditions, however, chicks, poults and ducklings would stay close to the hen for
several weeks after hatching. This closeness is maintained through a process called
imprinting. Imprinting is a special form of learning that occurs during a sensitive period
shortly after hatching, where the newly hatched bird instinctively follows the first moving
objects it sees and thus learns the characteristics of its parent. In the absence of a parent or
other adult bird, as is the case in commercial poultry production where typically only birds of
the same age are housed together, chicks can imprint on other chicks, humans or even
objects.
11.8 Care of Offspring

During the first few weeks after hatching chicks, ducklings and poults are not able to
maintain their body temperature well. The hen therefore ‘broods’ them to keep them warm by
covering them with her body and wings. Although precocial young otherwise require little
parental care, the parents do play a role in protecting young chicks from danger and teaching
them about various aspects of the environment. The mother hen helps chicks to learn how to
go up to roost on branches at night. Domestic fowl chicks and turkey poults are attracted to
edible foods (and steered away from inedible or toxic foods) by the pecking and
vocalizations of the hen. In commercial flocks young birds are kept in single-age groups and
thus there are no parental influences on their behaviour. While fowl chicks usually readily
explore and find food and water themselves, or by copying the behaviour of other chicks that
have already found food and water, turkey poults sometimes ‘starve out’, meaning that they
fail to start eating, presumably because of the lack of maternal stimulation.

Poultry become fully independent of their parents a few weeks or months after they hatch.
During this time, their behaviours are developing into the forms seen in adult birds. As they
explore and learn about foods, pecking at inedible items like sand decreases and pecking at
edible items increases. They learn to recognize suitable dust-bathing substrates and the
different elements that make up a dust-bathing bout begin to appear in their behavioural
repertoire, with full dust baths finally being performed when they are several weeks of age.
They also begin to refine their movement and spatial skills, for example learning how to
navigate elevated roosting surfaces.
Social behaviour is also developing at this time. Young chicks frolic and spar with one
another – play behaviour that resembles adult chasing and fighting. Chickens begin to peck
others aggressively when they are as young as 2 weeks of age, although submissive
behaviours are not shown until a few weeks later. Aggression is not apparent in turkeys until
they are much older, about 3 months of age. In either case, these interactions lead to the
formation of male and female dominance hierarchies in the flock, although these may not
become stable until the birds reach sexual maturity.
In addition, developing birds learn the characteristics of appropriate mates for normal
sexual activity through a process called sexual imprinting, which occurs during a sensitive
period prior to sexual maturity. If males and females are reared separately, as sometimes
occurs in commercial production, a lack of ability to undergo sexual imprinting can cause
problems with mating later on in breeding flocks, since the birds may form strong
homosexual pair bonds or show reduced mating behaviour (Appleby et al., 2004). However,
unlike other some other bird species, in poultry it is not necessary for males to be present in
order for the females to become sexually mature – the reason why hens can be kept in all-
female groups for egg production.
References
Alvino, G.S., Archer, G.S. and Mench, J.A. (2009) Behavioural time budgets of broiler chickens reared in varying light
intensities. Applied Animal Behaviour Science 118, 54–61.
Appleby, M.C., Maguire, S.N. and McRae, H.E. (1986) Nesting and floor laying by domestic hens in a commercial flock.
British Poultry Science 27, 75–82.
Appleby, M.C., Mench, J.A. and Hughes, B.O. (2004) Poultry Behaviour and Welfare. CAB International, Wallingford, UK.
Archer, G.S. and Mench, J.A. (2017) Exposing avian embyros to light affects post-hatch anti-predator fear responses.
Cunningham, D.L. and van Tienhoven, A. (1983) Relationship between production factors and dominance in White Leghorn
hens in a study of social rank and cage design. Applied Animal Ethology 11, 33–44.
Fournier, J., Schwean-Lardner, K., Knezacek, T.D., Gomis, S. and Classen, H.L. (2015) The effect of toe trimming on
behaviour, mobility, toe length and other indicators of welfare in tom turkeys. Poultry Science 94, 1446–1453.
Hale, E.B., Schleit, W.M. and Schein, M.W. (1969) The behaviour of turkeys. In: Hafez, E.S.E. (ed.) The Behaviour of
Domestic Animals, 2nd edn. Williams and Wilkins, Baltimore, Maryland, pp. 22–24.
Houldcroft, E., Smith, C., Mrowicki, R., Headland, L., Grieveson, S., Jones, T.A. and Dawkins, M.S. (2008) Welfare
implications of nipple drinkers for broiler chickens. Animal Welfare 17, 1–10.
Hughes, B.O. (1984) The principles underlying choice of feeding behaviour in fowls – with special reference to production
aspects. World’s Poultry Science Journal 39, 218–228.
Janczak, A.M. and Riber, A.B. (2015) Review of rearing-related factors affecting the welfare of laying hens. Poultry Science
94, 1454–1469.
Johnson, S.B., Hamm, R.J. and Leahey, T.H. (1986) Observational learning in Gallus gallus domesticus with and without a
conspecific model. Bulletin of the Psychonomic Society 24, 237–239.
Makagon, M.M. and Mench, J.A. (2011) Floor laying by Pekin ducks: effects of nest box ratio and design. Poultry Science
90, 1179–1184.
Mench, J.A. and Keeling, L.J. (2001) The social behaviour of domestic birds. In: Keeling, L.J. and Gonyou, H.W. (eds)
Social Behaviour in Farm Animals. CAB International, Wallingford, UK, pp. 177–210.
Millman, S.T., Duncan, I.J.H. and Widowski, T.M. (2000) Effect of male-to-male aggressiveness and feed- restriction during
rearing on sexual behaviour and aggressiveness towards females by male domestic fowl. Applied Animal Behaviour
Science 70, 63–82.
Muir, W.M. (1996) Group selection for adaptation to multi-hen cages: selection program and direct responses. Poultry
Science 15, 349–359.
Nicol, C.J. (2015) The Behavioural Biology of Chickens. CAB International, Wallingford, UK.
Nicol, C.J., Bestman, M., Gilani, A.-M. and De Haas, E.N. (2013) The prevention and control of feather pecking:
application to commercial systems. World’s Poultry Science Journal 69, 775–788.
Odén, K., Vestergaard, K.S. and Algers, B. (2000) Space use and agonistic behaviour in relation to sex composition in large
flocks of laying hens. Applied Animal Behaviour Science 62, 219–231.
Olsson, I.A.S. and Keeling, L.J. (2002) The push-door for measuring motivation in hens: laying hens are motivated to perch
at night. Animal Welfare 11, 11–19.
Olsson, I.A.S. and Keeling, L.J. (2005) Why in earth? Dustbathing behaviour in jungle and domestic fowl reviewed from a
Tinbergian and animal welfare perspective. Applied Animal Behaviour Science 93, 259–282.
Rogers, L.J. (1995) The Development of Brain and Behaviour in the Chicken. CAB International, Wallingford, UK.
Sherwin, C.M. and Kelland, A. (1998) Time budgets, comfort behaviours, and injurious pecking of turkeys housed in pairs.
British Poultry Science 39, 325–332.
Zuk, M., Thornhill, R., Ligon, J.D., Johnson, K., Austad, S., Ligon, S.H., Thornhill, N. and Costin, C. (1990) The role of
male ornaments and courtship behaviour in female choice of red junglefowl. American Naturalist 136, 459–473.
12 Behaviour of Horses
D. Mills and S. Redgate
12.1 Origin
The ancestral wild horse was largely a social herbivore that depended on grassland, vigilance,
size and speed for survival. Through domestication the horse has been successfully exploited
to fulfil a wide variety of roles in society (Clutton-Brock, 1992; Hall, 2005). However,
adaptability, within the constraints of good welfare, is not reflected by just physical health
and long-term survival in captivity.
In the industrialized world the horse is frequently kept in social isolation, often fed a diet
low in fibre, mated unnaturally and weaned at an early age. However, the ancestral
phylogeny of the horse still underpins the essence of its behaviour and failure to meet the
horse’s behavioural needs has led to some of the most common problems of horse
management in the industrialized world. Many of these behavioural and welfare problems are
associated with chronic frustration and can be managed with a greater sensitivity to the
nature of the horse (Waran, 2002). However, it is equally important to recognize the ways in
which a horse’s behaviour can be adapted to situations which may conflict with its
phylogeny. If this were not possible, then the success and popularity of the horse as a
domestic ridden animal would not be achievable.
There are an estimated 62 million horses worldwide distributed among 682 breeds (Hall,
2005), suggesting that the horse is one of the most differentiated domestic species. Its
variation in size is well known, with some large Percherons and Shires reaching 76 inches
(190 cm) at the withers in contrast to one dwarf miniature (Thumbelina) standing at only 17
inches (40 cm). However, less attention has been paid to breed variation in behaviour, which
is perhaps surprising given the extent of diversity. Some of the features of a breed relate to
environmental factors which are independent of the domestication process, for example
animals in lower environmental temperatures tend to have bigger trunks (Bergman’s rule)
and smaller extremities like ears and legs (Allen’s rule) as a result of selection for efficient
heat conservation or dissipation, and it seems reasonable to assume that there might also be
differences in behavioural predisposition as a result of such environmental factors. Other
features have been specifically bred for as a result of human intervention following
domestication, such as the four-beat gait of the Icelandic pony (tölt) and Missouri Fox
Trotter. In many cases the breed-typical behaviour will reflect a combination of these two
selective forces (the physical environment and man) – for example, differences in
temperament, which is often colloquially associated with blood temperature (hot blood,
warm blood and cold blood). Breed differences in behaviour also extend to differences in
specific learning abilities such as discrimination learning (Mader and Price, 1980) or operant
manipulation tasks (Hausberger et al., 2004) and predisposition towards developing specific
behaviour disorders such as crib-biting, weaving and stall-walking (McGreevy et al., 1995;
Luescher et al., 1998).
This chapter will focus on the species-typical behaviour patterns of the horse and begins
with a brief review of the concept and importance of normal time budgets of the horse.
12.2 Time Budgets

The time budget of maintenance activity in horses and the factors affecting it have been
reviewed extensively by Houpt (2005), so this section will highlight the importance of this
concept. The majority of a horse’s time is naturally spent foraging, during which the horse
moves short distances frequently, sampling from one patch of herbage after another. The
amount of time spent feeding will depend on the forage type and availability, but it has been
estimated that a horse naturally takes around 30,000 bites per day (Mayes and Duncan, 1986)
and perhaps chews nearly 60,000 times (Cuddeford, 1999). A significant reduction in these
activities, for example by feeding a concentrate diet, may have important behavioural and
physiological effects. Not only may there be an intrinsic tendency to chew, but it has been
known for a long time (Colin, 1886 cited by Alexander, 1966) that saliva is produced
primarily only when a horse is chewing. The volume of saliva produced and its associated
mineral content are not insignificant, with a pony producing around 5–6 l saliva/day that is
rich in sodium, chloride, bicarbonate and, relatively speaking, calcium too. Therefore the
level of salivation affects the level of key minerals including bicarbonate entering the gut,
which may have knock-on effects for gut environment, especially pH and the associated
flora. This in turn might predispose the horse to colic and perhaps behavioural problems such
as crib-biting (Nicol, 1999) or wood-chewing.
When grazing a horse typically moves around foraging for between 50% and 75% of its
time and spends between 15% and 35% of its time standing (Houpt, 2005). This means the
horse will typically take over 10,000 paces per day just as part of its natural grazing
behaviour. Once confined the amount of movement is greatly reduced (less than 5% in a
stable and about twice this value in a small corral; Houpt, 2005). Locomotory activity will
differ between individuals; however, the measurement of activity within a stable can provide
useful information as to the physical well-being and mental state of an animal. A distressed
horse may be found circling repeatedly as sudden isolation can increase locomotory activity.
A change in activity level is often attributed to the horse’s state of health, the diet or the
management routine.
The duration and frequency of resting behaviour including sleep have been used as key
welfare indicators in humans and other species. Sleep deprivation is associated with a
reduction in well-being and in particular is thought to impact on learning and memory. Using
electroencephalography and behavioural observations, four stages of equine sleeping patterns
have been recorded and defined: (i) wakefulness; (ii) drowsiness; (iii) slow-wave sleep
(SWS); and (iv) paradoxical sleep, also known as rapid eye movement (REM) sleep
(Williams et al., 2008).
Horses show polyphasic sleeping patterns, meaning they will sleep for short durations
over multiple time points in any one 24 h period. This pattern is adaptive for a prey species
and reduces their vulnerability to predators. For the same reason horses will show a
reluctance to lie down in order to achieve sternal or lateral recumbency unless they are in a
familiar environment, ideally with known social companions to maintain vigilance. SWS can
be achieved while the horse is standing or lying; however, REM sleep can only be achieved
when the horse is in lateral recumbency (Fig. 12.1). This is due to a decrease in muscle tone,
effectively paralysing the body for the duration of the episode. Hence REM sleep is also
known as sleep of the body; it is critical to restoration and potentially may be incorporated in
memory consolidation.
Fig. 12.1. Horse showing the posture for lateral recumbency, filmed using a wall-mounted
camera and infrared. (Image courtesy of S. Mosseri and R. Piercy, Royal Veterinary College,
London.)
Time budgets of resting and sleeping behaviour of feral horses have been difficult to
define given that horses need to be observed throughout the night and disturbance from an
observer is likely. Despite these caveats of data collection, Przwalskii horses that were
habituated to human presence and at pasture have been observed to spend 15.7% of their time
in standing-resting postures, 1.2% lying sternal and 4.1% lying laterally within a 24 h period.
Recumbent postures were most common between midnight and 04.00 hours and standing
resting during daylight hours (Boyd et al., 1988).
There is a myriad of factors that can have a bearing on the occurrence of sleeping
behaviour. For instance, observations of feral Camargue horses in France revealed that the
mares spent more time in sternal and lateral recumbency during the spring than in any other
season (Duncan, 1985). This coincides with foaling and having a young foal at foot, which
likely impacts on the energy demands of the mare, increasing the requirement to lie down.
The requirement to sleep, for significant portions of the day, is particularly crucial for
younger animals, potentially due to the needs of the body to grow and assimilate new
information.
Management of domestic horses will be dependent on the owner’s preferences and
available resources. How housing and management influence the quality of resting behaviour
and how this subsequently impacts on well-being and athletic performance are unknown and
warrant investigation. There is some evidence to suggest that housing and bedding influence
the longevity of resting behaviour. Houpt et al. (1986) found that pregnant mares kept in
stables spent more time resting overnight than those at pasture, although differences could
have been due to increased opportunities for vigilance or the performance of other
behaviours while at pasture. Greening et al. (2013) found that horses spent more time
recumbent when provided with straw bedding rather than wood shavings.
Establishing normal behavioural parameters is not only important in identifying risk
factors for potential problems with adaptation to the captive environment, but can also give
insight into the early stages of other significant changes. For example, McDonnell (2005)
described the normal daily behaviour of horses in the stall (see Table 12.1) and significant
deviations from this can be used to help determine chronic, acute or subacute pain (Gleerup
and Lindegaard, 2016).
In the following sections we consider some of the key normal behaviours of the horse in
more detail.
Table 12.1. Reference ranges for normal behavioural parameters of activity for the horse
housed alone in a stall or small paddock. (From McDonnell, 2005.)
Activity Episodes/day Duration

Major activity changes (eating, standing 30–110 20–60 min per activity
resting, standing alert, resting recumbent) when undisturbed
Standing resting 10–30 5–120 min each; 8–12
h total
Recumbent rest 0–6 10–80 min each; 0–6 h
total
Feeding on hay fed two or three times daily or 10–30 5–30 min each; 4–12 h
continuously total
Drinking 2–8 10–60 s each; 1–8 min
total
Urinating 4–15 (greater for mares in oestrus and stallions
where marking behaviour is elicited)
Defecating 4–15
Rolling 2–8 2–8 rolls/bout
Erection/masturbation 18–36 (stallion)
9–24 (gelding)
The horse is a generalist herbivore and feral populations can be found worldwide in a variety
of different habitats, ranging from savannah to desert (Boyd and Keiper, 2005). Feral horses
have been observed to select an annual diet of mixed plant species; in general, grass species
and grass-likes are selected predominantly along with various proportions of shrubs, leaves
stems, bark and roots. Environmental conditions and plant species availability have a heavy
bearing on selection patterns. In the UK, New Forest ponies will predominantly graze the
grasslands during the growing seasons and will shift their selection in favour of species that
provide shelter and nutrients, such as gorse, holly and deciduous woodland, during the non-
growing season (Tyler, 1972; Putman, 1986; Gill, 1987). This flexibility towards their food
selection ensures that daily intake remains high throughout the year, which is crucial to the
horse’s digestive strategy (Houpt, 2005). Despite this, nutritional needs may not always be
met if resources are limited. Using microhistological examination of faeces, Pratt-Phillips et
al. (2011) noted that feral horses residing on the Shackleford Banks, North Carolina were
deficient in micronutrients at certain times of the year.
To cope with a nutritionally variable environment the horse, like other herbivores, has
adopted a patch-feeding strategy whereby plant communities are regularly visited and
sampled (Prache et al., 1998; Fleurance et al., 2001). Plants are sought and initially
recognized on the basis of familiar sensory characteristics such as, shape, colour, texture,
smell and flavour. Horses have been observed, for example, to select the young, greener parts
of plants that are higher in nutritional value than the rest of the available plant matter
(Duncan, 1992; Menard et al., 2002). These latter observations support experimental work
that suggests that horses can learn to associate a familiar food’s sensory characteristics with
energy- or nutrient-related post-ingestive consequences (Cairns et al., 2002; Redgate et al.,
2014). Conversely, horses can also learn to avoid foods that make them immediately ill
(Pfister et al., 2002), and the association is even more persistent when the food offered is
novel (Houpt et al., 1990). Neophobic responses are observed when horses are presented
with novel feedstuffs; for example, only small amounts may be consumed on initial
presentation. This cautious response is functional given that the horse does not readily
regurgitate ingesta. Even under short-term dietary restriction, when it would be beneficial for
the horse to alter selection and intake patterns, this neophobic response can persist (van den
Berg et al., 2016).
The useable energy gained from plants is far less than that from animal protein. The
horse compensates by having a digestive strategy characterized by a high voluntary intake

and rate of gastrointestinal passage (Illius and Gordon, 1993; Sneddon and Argenzio, 1998).
Thus to achieve a high voluntary intake the horse spends a large proportion of the day
feeding. Free-ranging horses can spend, in total, between 9 and 16 h each day, grazing and
browsing (Ellis, 2010). Daily time spent grazing varies and can be influenced by internal and
external factors; for instance, as ambient temperature increases foraging behaviour decreases
(Holcomb et al., 2015). Grazing and browsing bouts occur regularly throughout each 24 h
period; time of day affects bout length, as longer feeding bouts have been recorded at dawn
and in the late afternoon (Tyler, 1972). Natural breaks between feeding bouts are usually
short and tend to occur to perform other activities such as other maintenance, social and
reproductive behaviours. Horses rarely fast voluntarily for more than 4 h (Davidson and
Harris, 2002), although they will take longer feeding breaks to shelter from inclement
weather or biting flies (Mayes and Duncan, 1986; Gill, 1987).
Stabled horses that are provided with their total ration ad libitum also feed in discrete
bouts, with breaks for rest or monitoring of the local environment. Feeding bouts separated
by intervals of non-feeding of 10–15 min duration have been classified as meals (Ralston and
Baile, 1982; Laut et al., 1985). Laut et al. (1985) noted that when a 10 min feeding interval
was applied to 24 h observations of stalled ponies feeding ad libitum, they ate approximately
17 small meals each day. Feed processing affects the duration and number of feeding bouts.
Argo et al. (2002) demonstrated this with ponies that were presented with two feeds identical
in nutritional components but offered either as pellets or as chaff ad libitum. The ponies had
longer feeding bouts on the chaff and more frequent but shorter feeding bouts when fed the
pellets.
The anatomy and function of the horse’s digestive tract reflect this little-and-often or
‘trickle’ feeding strategy. The stomach is small and has a limited capacity to hold large
amounts of food, although it rarely empties completely and digesta moves rapidly on to the
small intestine within 20 min of ingestion (Harris and Arkell, 2005). As there is a continuous
flow of digesta from the stomach the horse has no need for a gallbladder and bile is secreted
almost continuously into the small intestine. Digesta finally moves on into the large intestine;
as a hindgut fermenter, a large proportion of the horse’s energy needs are obtained through
the caecal and colonic fermentation of dietary fibre (Hoffman et al., 2001). The steady flow
of forage through the digestive system acts as a fluid reservoir and maintains total
gastrointestinal tract motility and health (Sneddon and Argenzio, 1998; Geor, 2005a).
It has been argued that the domestic horse’s dietary regimen should often resemble the
phylogenetic origins and natural feeding patterns of the horse (Mills and Nankervis, 1999);
that is, with free access to species-rich grassland that has a medium- to low-quality energy
content or ad libitum provision of forage and fibre-based feeds. Where horses are bred and
managed specifically for leisure or sporting purposes the owner or carer has the
responsibility for providing a diet that meets each individual horse’s health and behavioural
needs, and some compromise may be necessary. For example, the feeding strategies of
racehorse trainers will be to maximize performance and as a result horses in training are fed
high-energy feeds with little forage (Geor, 2005b), but this is not without a cost to the
animal’s welfare. By contrast, owners of native-type ponies often have to restrict their access
to grazing and other feeds if they are susceptible to laminitis. This is a debilitating and
painful condition that can be triggered by high levels of starch, present in cereals, and sugars,
present in modern horse pastures, reaching the hindgut (Lockyer, 2005).

In captivity, horses may be kept in single- or mixed-sex groups chosen on the basis of
convenience or familiarity, with little attention given to the composition of the group (other
than to avoid potential unnecessary mixing of stallions, with others or each other) (Hartmann
et al., 2012), and the group is frequently disrupted as one or more animals is removed for
riding or other purposes. The frequency of kick injuries, as a sign of unstable social relations,
should not therefore come as any surprise, but this does not indicate that animals should be
kept in isolation. Horses naturally live in social groups (bands) and the benefits of social
living clearly outweigh the costs for this species. These benefits would appear to relate
mainly to predation avoidance, in terms of increased vigilance when grazing and safety in
numbers when fleeing, and so it can be expected that there has been historically strong
selective pressure for gregariousness and against isolationism and persistent aggression
within social groups (Fureix et al., 2012). However, many modern management systems
appear not to take this into account and seek to house horses alone, or at least in physical
detachment from each other. While it may be necessary for health to separate individuals at
times, adopting housing designs that allow social contact are beneficial for quality of life and
well-being (Yarnell et al., 2015). Within groups, stable long-term linear dominance
hierarchies can be discerned among mares in single- and mixed-sex groupings (Clutton-
Brock et al., 1976; Houpt et al., 1978; Sigurjonsdottir et al., 2003). Within a group mare rank
can reflect a range of factors, some of which have been found to be consistently predictive,
such as size, age and length of residence within the group (van Dierendonck et al., 1995;
Sigurjonsdottir et al., 2003; Rho et al., 2004); however, these are not totally independent as
older mares will tend to be larger and where they are living in stable groups the oldest will
tend to have had the longest residence time. Hierarchical position is to some extent
maintained by the outcomes of aggressive encounters involving chasing and physical
shoving, threats to bite or kick, and ultimately actual kicking and biting (Houpt and Keiper,
1982), but this is almost certainly exaggerated in the captive situation where space and the
ability to retreat are more limited. More importantly, living in a social group requires the
formation of affiliative relationships which are frequently reinforced and evident through
mutual grooming, resting together, proximity and following behaviour. It is found that mares
in the wild will typically have only one or two preferred social partners, other than their own
foals, in their lifetime (Feh, 2005), suggesting that social relationships between mares, at
least, can be very strong and enduring. This can lead to problems of isolation stress in
captivity when a mare is separated from a close social partner; however, the more general
potential social stress of allowing only more transient relationships in captivity remains
unexplored. Van Dierendonck et al. (2009) suggest that familiarity may be a better predictor
of affiliation than kinship; they found among both mares and geldings that affiliates were
more likely to intervene when a preferred partner engaged in affiliative behaviour (such as
play or mutual grooming) with another individual, suggesting that horses appear to take
action to protect their relationships. Awareness of such behaviour may be particularly
important in captive systems.
By contrast, protective and interventionist behaviour by stallions towards their breeding
mares is well known. In nature, a breeding stallion tends to live with a group of females
(harem) and their immature offspring, isolated from other adult males (Feist and
McCullough, 1975; Keiper, 1976; Miller, 1981; Berger, 1986), although variations on this
theme are not uncommon. This includes breeding pairs (Feist and McCullough, 1975; Welsh,
1975; Green and Green, 1977) and the tolerance of a younger second stallion in certain
situations to help defend the harem from other males (Miller, 1981; Berger, 1986; Goodloe et
al., 2000). Alternatively, all-male adult bands may form (McCort, 1984). These findings
suggest that stallions too may be more cooperative in their social relationships than is widely
recognized (Freymond et al., 2013), although a clear dominance hierarchy may also be
apparent when resources are limited. All-female adult groups also occasionally arise in the
wild (Goodloe et al., 2000) and are common in captivity. Geldings (castrated males) tend to
form their own social groups even when mares are available without an adult stallion
(Sigurjonsdottir et al., 2003). The geldings were also reported to play much more than the
mares, with play being absent among adult mares. When resources are not limited aggression
is generally rare.
Young males (up to the age of about 3 years) are tolerated in reproductive bands until
they attempt to breed with one or more mature mares, at which time the stallion will drive
them out (Feh, 2005). At this point they will naturally tend to join up with males of a similar
age until socially mature enough to take control of their own harem (typically around the age
of 6 years). Females tend to leave the social group at about the same age when they come
into oestrus. They are not usually driven out (Monard and Duncan, 1996), but they will
usually refuse to mate with the familial stallion of the group and tend to seek out alternative
partners at this time. Reproductive behaviour is discussed in more detail later in the chapter.
Communication is vitally important when establishing and maintaining cohesive social
groups. Horses naturally graze and forage as a group, with socially facilitated changes in
behaviour as the group moves between watering holes, shelter belts and grazing areas. It is
likely that information about distribution of resources including location of preferred or
higher-value species is shared within the group, although there are exceptions to this rule.
Andrieu et al. (2016) found that while some individuals could be relied upon by the wider
group to lead them to food supplies, through passive recruitment, other group members
appeared to show behaviour akin to deception.
The ability to recognize and discriminate between familiar and non-familiar animals is
fundamentally important for an individual so that it can respond appropriately according to
age, sex and dominance status. Horses utilize visual, auditory and olfactory information to
recognize and respond to conspecifics (Proops et al., 2009; Péron et al., 2014). However, our
understanding of social behaviour and communication between horses has focused primarily
on observations of body language, in particular the overt and subtle movements of the head
and the body (McDonnell, 2003). Recent research work has the potential to enhance this
greatly, as Wathan et al. (2015) have developed an equine facial action coding system
(EquiFACS) that uses the underlying facial musculature to objectively identify and record
facial expressions. This assessment of the subtle movements of the horse’s face will have
valuable application to the understanding of positive and negative mental states such as the
detection and alleviation of pain (Dalla Costa et al., 2014; Gleerup et al., 2015).

Mares are seasonally polyoestrus, displaying repeated oestrous periods during the breeding
season with a period of anoestrus in between. The cycle is approximately 3 weeks, with 1
week of oestrus followed by a 2-week dioestrous phase, but the precise cycle length is
variable between breeds and also between individuals. The oestrus period is markedly longer
than in many other ungulates and while it has been suggested (Kiley-Worthington, 1987) that
this may relate to the need to establish a bond between the sexes at this time, this seems
unlikely given the normal social relationship that exists between a mare and stallion. It may
however relate to competition for the stallion between mares within the same social group
(Curry et al., 2007), since dominant mares may interrupt matings of subordinates or generally
try to lower their reproductive fitness (Rutberg and Greenberg, 1990), and dominant mares
are generally more attractive to stallions (Powell, 2000).
Sexual behaviour is a communicative process involving a complex sequence of
interactions between mare and stallion, but with the mare naturally being much more
proactive than is generally recognized. In semi-feral ponies, mares have been reported to
initiate 88% of the pre-copulatory interactions that led to successful mating (McDonnell,
2000). Visual cues used to attract the attention of the stallion include tail-raising, which can
vary in extent from a slight lift away from the body through to an almost vertical flag-like
posture to expose the vulva, which may in itself act as a visual cue to the stallion especially
when combined with clitoral winking (Fig. 12.2). The vulval labia are darkly pigmented and
when drawn back expose the brighter pink lining of the vestibule wall. Olfactory stimuli
might also be released at this time in addition to the pheromones that are present in the urine
of oestrous mares. These increase interest by and arousal in the stallion, as demonstrated by
an increased rate of flehmen response during which the stallion extends his neck, flares his
nostrils and inhales to draw air into the richly innervated vomeronasal organ (Stahlbaum and
Houpt, 1989; Fig. 12.3). A stallion will often let out a long, low nickering sound at this time.
Fig. 12.2. Clitoral winking in a mare; note exposed pink vulval labia and classic wide-legged
stance.
Fig. 12.3. Flehmen response in a stallion.
The oestrous mare adopts a very characteristic stance similar to that assumed for
urination with the hocks and stifles flexed, the rear limbs slightly abducted and the pelvis
lowered, a position sometimes described as squatting or a sawhorse stance. In this position
the mare is braced to take the stallion’s weight on mounting, but in addition the stance itself
is a visual stimulus advertising the mare’s oestrous status. Oestrous mares also show an
increased incidence of urination; this can range from a full evacuation to just a few drops.
If a mare is receptive a stallion will often nudge and nip along her body towards her
neck. Both individuals may appear to behave quite aggressively to each other at this time, but
if the mare responds favourably to a stallion’s advance his penis is extended and will become
erect. He will normally make several partial mounts before attempting a full mount, usually
from behind but occasionally from the side. Intromission is normally achieved after a few
exploratory thrusts. Copulation does not take long and, after a few pelvic thrusts, flagging of
the raised tail signals ejaculation. The head also tends to be lowered and the facial muscles
relax at this time. During copulation the stallion may bite occasionally at the mare’s neck and
mane. Within about 30 s, the mare will then normally step forward to allow the stallion to
dismount. It is not uncommon for a stallion to squeal a little at this time. In the natural state
the stallion does not have to step backwards to dismount, although this is common in ‘in
hand’ matings and may explain the relatively high incidence of spinal problems in breeding
stallions. The stallion may be willing to serve another mare within 10 min or so, but is likely
to show mated mares greater attention over the next few days. A mare in oestrus tends to be
served on five to ten occasions. These natural patterns differ markedly from what is normally
allowed in the breeding of valuable bloodstock, where the stallion is brought to the mare
when she is believed to be receptive, allowed to mount and then required to dismount almost
immediately. Mares are usually heavily restrained at this time, to both enforce compliance
and protect the stallion from damage.
A range of behaviour problems are often attributed to the underlying physiology of the
sexual cycle and arousal of both mares and stallions. In the case of mares, this includes
persistent oestrous behaviour, aggressivity and disobedience. This has led some to suggest
ovariectomy as a potential solution where the association has been confirmed; however, a
recent critical appraisal of this topic (Crabtree, 2016) suggests that this procedure may
exacerbate the problem in the case of persistent oestrous behaviour. Nevertheless,
ovariectomy may help reduce human-directed aggressivity and more general problems
associated with problematic temperament in over 70% of cases; behaviour problems related
to interaction with other horses are less likely to be improved by the procedure. If the
problem relates to ovarian pain associated with the sexual cycle rather than hormonal
changes, it is suggested that chemical suppression of gonadotropin-releasing hormone may
be a preferable course of action, but the efficacy of this procedure for these more specific
problems is largely unknown. In the case of stallions, many problems may actually relate to
poor management (McDonnell, 2016), including the unnecessary social isolation often
imposed on stallions since they can be successfully managed in a group (Freymond et al.,
2013); however, deslorelin may be used to suppress male sexual hormones prior to surgical
castration if this is believed to be the source of problems (Schönert et al., 2012).
12.6 Behaviour at Birth

The mare appears to be able to exert considerable control over the time of parturition, which
is largely under parasympathetic control (Nagel et al., 2014), and while the normal length of
pregnancy is about 340 days (Rossdale, 1967; Jeffcott, 1972), normal parturition may occur
2–3 weeks either side of this time frame. The importance of parasympathetic dominance to
the specific time of delivery explains the apparent control of the process to when the
environment is calm and there is little apparent risk of disturbance. Foaling will therefore
frequently occur at night (Jeffcott, 1972); this is adaptive as in the natural state the risk of
predation is minimal. Given that the newborn is a ‘follower’ (i.e. depends on remaining with
its dam for protection), there is also a biological advantage for the foal in this timing for such
a flight-dependent prey species. The parturition process is usually divided into three stages
(Rossdale, 1967; Jeffcott, 1972).

First-stage labour: immediately before birth
At this time the mare will appear restless, wandering around a given area, and will switch
between one behaviour and another. Pedometers may be used to help remotely monitor the
onset of this stage (Bachmann et al., 2014). She may swish her tail, turn her head towards her
flanks, kick at her belly, frequently urinate, paw at any bedding, crouch, lie down and get up
again. This stage may be short (minutes), prolonged (hours) or intermittent over a period of
days, although the latter may be more common in captivity and result from frequent
disturbance.
Second-stage labour: parturition
Bursting of the allantochorionic sac is often followed by a flehmen reaction to the uterine
fluids and a nickering. The mare will normally lie on her side, with her uppermost hind leg
extended. She strains forcefully and regularly to deliver the foal, but even this behaviour can
be interrupted by the mare if she is disturbed; she may also pause to rest between bouts of
contractions and even feed at this time. This phase usually lasts between 10 and 30 min until
the main trunk of the foal is delivered. The foal’s movements usually rupture the surrounding
sac and it will pull its own hind legs away from its mother. The mare should not stand
immediately as there is significant blood transfer possible from the placenta to the foal at this
time.
Third-stage labour: postpartum pause
Typically the mare may remain lying for 15–20 min, but 25% of mares in captivity stand
within 5 min, snapping the cord early. Such a response may result in a foal being deprived of
up to a third of its blood supply, inadequate inflation of the lungs and oxygen starvation of
the tissues. When the mare licks the foal for the first time, she will often nicker and a
chemical imprint is made of the foal as bonding begins. Occasionally following a particularly
difficult foaling, especially in primiparous mares, the mare may turn aggressively to the foal
and reject it before bonding occurs. The placenta is normally expelled within about 1–2 h of
delivery of the foal and about two-thirds of mares will investigate it, but ingestion is rare
(Virga and Houpt, 2001). Mares are naturally very protective initially of the newborn foal
and will frequently display aggression towards any early approaches to their offspring. In the
case of a stillborn foal, the mare will initially remain close to it but gradually, over a period
of several hours, normally lose interest in the body and resume grazing.

Soon after the foal rises the mare usually stands to allow it to suckle. She may turn to guide
the foal towards the teat and will normally allow it to suckle for up to 20 min. It is suggested
(Fraser, 1992) that the foal learns to suckle as a result of the successful completion of a series
of modal action patterns including coordination for standing, walking, orientation towards a
dark undersurface (facilitating maternal contact), thrusting with the muzzle (resulting in teat
location), sucking (to obtain milk) and suckling (to efficiently extract milk from the udder).
Inappropriate stimuli in the environment may disrupt this process and cause problems; for
example, the foal may head towards a large wall-mounted hay trough in the stable instead of
its mother’s underbelly. The let-down of milk and suckling by the foal are both believed to
play important roles in the bonding of both mother and foal, and so the artificial rearing of
the orphan foal must be done with caution to avoid human imprinting. Within the first 3 h or
so the foal will normally suckle again and pass the fetal faecal plug (meconium). The foal
will feed almost hourly for the first day or so. The mare will terminate about half of the
nursing bouts in the first week (Tyler, 1972; Crowell-Davis, 1985) by shifting her weight
from one hind limb to the other before simply walking away from the suckling foal. Later the
mare may allow the foal to suckle as she forages. Foals are very playful and will begin
solitary play within a couple of days (Boyd, 1988) and social play within the first month
(Crowell-Davis et al., 1987; Boyd, 1988). The foal stays close to its mother, within about 5 m
(Tyler, 1972), for virtually the whole of the first week and is allowed to suckle at will. During
this time the foal actively tries to stay near its mother when it is awake and it tends to be the
mare that instigates any separation.
Mares may be aggressive to their foals if nursing is painful, for example if she develops
mastitis or if the foal is overzealous in its suckling; this is more likely as the foal gets older,
peaking about 4–6 months of age (Barber and Crowell-Davis, 1994). Over time the foal is
increasingly responsible for any separation between the two and the mother becomes
increasingly responsible for bringing them back together (Tyler, 1972). This trend continues
until the time of weaning. If the mare is pregnant weaning will normally occur shortly before
the next birth, when the foal is about 40 weeks old, otherwise it may continue to suckle for a
year or more (Duncan, 1980). Problems with the mare–foal relationship in captivity are
widely recognized and have been reviewed elsewhere alongside related strategies for their
management (e.g. Żurek and Danek, 2012).
12.8 Offspring Development and Management

The mother’s response to a range of stimuli may play a key role in shaping the behaviour of
the foal (Henry et al., 2005; Christensen, 2016) and this, together with the relatively
immature emotional development of the young foal, has been exploited in early handling
procedures such as Miller’s ‘imprint training’ of the foal (Miller, 1991) to habituate the horse
to a range of human handling procedures. Exposure of the foal to the mother in work may
also be beneficial and is not uncommon in some cultures; for example, a foal may be haltered
alongside its mother in harness. Clearly there is the potential to exploit this further in the
captive setting, through exposing foals in a non-threatening way to a range of stimuli that it is
likely to encounter in later life, a process described as ‘maturation training’ (Mills and
Nankervis, 1999). The welfare and behavioural aspects of different training methods and
disciplines in horses is an area of growing concern and interest (McGreevy and McLean,
2011; Hothersall and Casey, 2012; Hall et al., 2013; Hockenhull and Creighton, 2013;
Henshall and McGreevy, 2014; Baragli et al., 2015), but beyond the scope of this chapter.
Weaning is naturally a gradual process, during which the young not only learn to be
independent, but also to cope with significant frustration as they are denied access to their
mother. However, in captivity the process is usually much more abrupt and occurs much
earlier, and so may be very stressful for mare and foal (Merkies et al., 2016), with long-term
consequences. Waters et al. (2002) have reported that weaning is associated with the onset of
a range of abnormal, stereotypic behaviours, and that individual box weaning was more
likely to induce these problems than group- or field-weaning strategies. They also found that
the diet around the time of weaning may influence the type of stereotypic behaviour
developing in later life, with the introduction of concentrates after weaning increasing the
risk of crib-biting. However these problems are not simply management-related, with dam
social dominance and the amount of time spent suckling (Nicol and Badnell-Waters, 2005)
also increasing the risk of problematic oral behaviour. While there has been growing
attention on the housing and management of the adult horse to prevent welfare problems
(Mills and Clarke, 2002), these findings suggest there should be equal if not greater attention
applied to the management of the foal, as this may have greater preventive value.
12.9 Perception
Despite the horse’s long history of domestication and close association with humans,
relatively little is known about the perceptual world of the horse. This section will focus on
certain adaptations of the sensory system and how these relate to the horse’s behaviour.
The majority of research within this subject area has focused on the visual system,
perhaps not surprisingly given the importance humans place on vision within our own
perceptual world. It is important to note that the horse and human perceptual systems may be
quite different and while we will never fully understand what it is like to be a horse, a generic
understanding is important to sensitively handle and train this species (Beaver and Höglund,
2016). With this greater understanding we can design appropriate training programmes that
utilize stimuli that horses find most salient as described by Hall (2007).
Vision
Horses have large eyes, suggesting that vision is important in this species. The location of the
eyes on the side of the head provides a wide field of view, allowing the horse to continuously
monitor its surroundings for changes in movement that may indicate the presence of a
predator. While horses have almost 360° range of vision in the horizontal plane, due to the
location of the eyes the majority of this view is through monocular vision. At the front of the
horse the monocular fields overlap, providing some binocular vision down the horse’s nose.
Within this binocular field, the ability of the horse to perceive detail or visual acuity
improves. This means that when the horse lowers the head to inspect an object lying on the
ground, the image is viewed from both eyes and is projected on to the area of the retina with
the greatest cell density, at the visual streak (Hall, 2007). It must be stressed that the horse’s
ability to discriminate fine detail is not as sensitive as a human’s; for the horse, seeing the
detail is not as important as ensuring that it can detect movement and react, generally by
fleeing (Beaver and Höglund, 2016).
Particular physiological adaptations of note indicate that the horse can see relatively well
in low light levels, as indicated by the presence of light-sensitive rods on the retina and the
light-reflecting tapetum lucidum. The horse is thought to have dichromatic colour vision, and
certainly anatomical investigations confirm the presence of multiple cone types on the retina.
However, behavioural studies using discrimination tests to investigate colour vision differ in
relation to the range of colours that the horse may be able to recognize; these likely
differences between studies are due to the methodologies employed and individual variation
within the small number of horses utilized. Blackmore et al. (2008) suggest that horses can
reliably discriminate yellow and blue from grey but there may be deficiencies with the
discrimination of red and green. The relevance of colour perception on the ability to jump
obstacles is significant, as horses are jumped at high speed over coloured obstacles against a
variety of different backgrounds: for instance, jumping a coloured fence against green grass
or a sandy surface. Misjudging a fence could be costly at the competition level but more
seriously lead to injury for either the horse and or rider. With this in mind, Spaas et al. (2014)
evaluated jumping performance of horses when ridden over green and blue fences. Although
differences were found between the number of touches and fallen poles, this difference
appeared to be due to the background contrast that the poles presented against rather than the
pole colour per se. This study highlights some of the difficulties that are present when
translating controlled behavioural testing into a more realistic, practical setting.
Hearing
Horses can hear a broadly similar frequency of sounds to humans, although humans can hear
slightly softer frequencies than the horse and the horse can detect higher frequencies out of
the range for humans. Determining this range of frequencies is not straightforward, as a horse
cannot self-report what it can and cannot hear. Animals are therefore trained to respond to
sounds in order to receive food or water, which has been withheld to increase motivation
(Heffner and Heffner, 1983). For a young adult human the audible range of frequencies is
approximately 20 Hz to 20 kHz, whereas the horse can hear 55 Hz to 33.5 kHz with peak
sensitivity occurring between 1 and 16 kHz (Heffner and Heffner, 1983). Common equine
vocalizations occur within this frequency range, for instance a whinny has been measured at
frequencies of between 400 and 2000 Hz (Yeon, 2012). This also means that the horse can
hear frequencies in the ultrasonic range.
Non-invasive techniques used with humans have been applied to the horse to help
understand hearing function and loss. Brainstem auditory evoked responses (BAER) have
been used to evaluate hearing in young (5–8 years) and ageing (17–22 years) horse
populations, and it was reported that as individuals age they can experience some partial
hearing loss (Wilson et al., 2011). It is likely that this will differ between individuals but
these horses should be managed appropriately, particularly when exposed to noisy situations
(e.g. clipping).
The ears can rotate independently of each other, and the horse will move the ears to assist
with sound location and also to communicate. Interestingly, although the ears are quite
mobile, they are less mobile compared with a cat’s or a dog’s. The ability to localize the exact
source of a sound is quite poor in the horse; presumably this attention to detail is less relevant
and the horse just requires a general direction of a sound prior to fleeing.
Olfaction and taste

It is thought that the horse has a relatively good sense of smell, although how this compares
with other species has not been established. Different odours are likely to affect the horse’s
positive or negative sensory experience of a substance by influencing the duration and/or
intensity that the odour is investigated. An increase in sniffing behaviour has been taken as a
sign of a positive interest, as it is thought that this repetitive sampling aids odour
identification.
Olfactory ability is relevant to a number of contexts, including the ability to recognize
conspecifics, avoid predators and select between different foodstuffs. Bonde and Goodwin
(1999) presented horses, in their home stables, with 11 different odours including spices,
essential oils and samples of faeces from a pig, tiger and wolf. Horses were presented with
each odour in a two-way preference test and against a control. Investigative behaviour was
shown towards most odours, including sniffing, touching, licking and biting. The longest
duration of interest was directed towards coriander seeds and ground ginger and the shortest
towards aniseed oil.
Studies have also shown that horses can use biologically relevant olfactory cues, for
instance from body odour or urine, to discriminate between familiar and unfamiliar
conspecifics (Hothersall et al., 2010; Krueger and Flauger, 2011; Péron et al., 2014). In
comparison to biologically relevant odours, novel odours can illicit more extreme changes in
behaviour when presented to horses that are visually isolated from conspecifics, in an open
arena. For instance, horses presented with eucalyptus oil while feeding increased their
frequency of vigilant behaviours (Christensen et al., 2005). Similarly, exposure to odour
taken from natural predators, such as a wolf’s urine, also increased vigilance behaviour and,
when paired with a loud noise, heart rate increased dramatically (Christensen and Rundgren,
2008).
The senses of smell and taste are central to food identification and selection, and short-
term preference studies suggest the horse has the ability to discriminate between a wide range
of flavours when added to a carrier foodstuff (Goodwin et al., 2005). Young horses have been
shown to discriminate between four basic taste modalities: salty, sweet, sour and bitter
(Randall et al., 1978). Preference was assessed by recording voluntary intake and ranking
foods accordingly, although behavioural reactions are also observed; for instance, horses
have been shown to produce very different responses when ingesting sweet and bitter foods
(Jankunis and Whishaw, 2013). Knowledge of the sensory preferences displayed by horses is
useful to horse owners, horse caretakers and equine feed manufacturers, as preferable
flavourants may help to stimulate appetite and increase voluntary intake. This is potentially
very relevant to horses that have to be presented with monotonous diets, such as animals that
may be receiving a clinical diet or restricted intake.
Touch
The skin is by far the largest sense organ in the body, although the sensitivity to pressure
from cutaneous stimulation is thought to vary due to the distribution of sensory nerve
receptors at different points in the skin. It is thought that the muzzle, neck, withers,
shoulders, coronets, rear of the pastern and lower flank are well innervated (Mills and
Nankervis, 1999). The teeth and the muzzle are used to scratch and possibly groom the coat.
Mutual grooming, where horses groom each other on different areas of the body, particularly
the back and base of the neck (Fig. 12.4), is used as a method of scratching those hard-to-
reach areas. It is also thought to facilitate and maintain social bonds by comforting or
calming the horse, as grooming at the base of the withers has been observed to slow heart
rate (Feh and de Mazieres, 1993).
Fig. 12.4. Horses standing nose to tail, the horse furthest away can be seen to be using its
muzzle to groom a social companion.
The horse has a high concentration of sensory receptors around the facial area,
particularly around the eyes, lips and nostrils where vibrissae (whiskers) are located. The
vibrissae are long, stiff hairs that have a distinct tapered shape. They are embedded in the
nerve endings and respond quickly to pressure. The vibrissae are sensitive structures and can
be used to provide information on the location of objects, which may serve a protective
function (e.g. the vibrissae located around the eyes) or to explore the environment (e.g.
during food location and manipulation). As discussed earlier, the horse has relatively poor
visual acuity and thus the vibrissae can compensate for this because they provide information
about objects that might be too close to see. Foals at birth possess a full chin of vibrissae
which is thought to help the foal to sense the mother and locate her teats.
The practice of trimming horses’ vibrissae is controversial and the act is banned in some
countries, for instance Germany, on welfare grounds. In the UK trimming is allowed but
attitudes to the practice vary and are dependent upon horse use. Owners who trim often do so
to ensure that the horse ‘looks tidy’ or due to ‘tradition’; within the competitive disciplines
the owners of show horses are more likely to remove the vibrissae (Emerson et al., 2016).
How this practice impacts upon the horse’s behaviour and subsequent welfare is currently
unknown and deserves evaluation.
References
Alexander, F. (1966) A study of parotid salivation in the horse. Journal of Physiology 184, 646–656.
Andrieu, J., Henry, S., Hausberger, M. and Thierry, B. (2016) Informed horses are influential in group movements, but they
may avoid leading. Animal Cognition 19, 451–458.
Argo, C.M., Cox, J.E., Lockyer, C. and Fuller, Z. (2002) Adaptive changes in the appetite, growth and feeding behaviour of
pony mares offered ad libitum access to a complete diet in either a pelleted or chaff-based form. Animal Science 74,
517–528.
Bachmann, M., Wensch-Dorendorf, M., Hoffmann, G., Steinhöfel, I., Bothendorf, S. and Kemper, N. (2014) Pedometers as
supervision tools for mares in the prepartal period. Applied Animal Behaviour Science 151, 51–60.
Baragli, P., Padalino, B. and Telatin, A. (2015) The role of associative and non-associative learning in the training of horses
and implications for the welfare (a review). Annali dell’Istituto Superiore di Sanità 5, 40–51.
Barber, J.A. and Crowell-Davis, S.L. (1994) Maternal behavior of Belgian (Equus caballus) mares. Applied Animal
Beaver, B.V. and Höglund, D.L. (2016) Behavior as it relates to handling. In: Efficient Livestock Handling: The Practical
Application of Animal Welfare and Behavioral Science. Academic Press, San Diego, California, pp. 13–44.
Berger, J. (1986) Wild Horses of the Great Basin: Social Competition and Population Size. University of Chicago Press,
Chicago, Illinois.
Blackmore, T.L., Foster, T.M., Sumpter, C.E. and Temple, W. (2008) An investigation of colour discrimination with horses
(Equus caballus). Behavioural Processes 78, 387–396.
Bonde, M. and Goodwin, D. (1999) Behaviour of stabled horses when presented with different odours. In: Harris, P.,
Goodwin, D. and Green, R. (eds) Proceedings of the Waltham Symposium. Equine Veterinary Journal Supplement 28,
60–61.
Boyd, L.E. (1988) Ontogeny of behavior in Przewalski horses. Applied Animal Behaviour Science 21, 41–69.
Boyd, L.E. and Keiper, R.R. (2005) Behavioural ecology of feral horses. In: Mills, D.S. and McDonnell, S. (eds) The
Domestic Horse. The Evolution, Development and Management of its Behaviour. Cambridge University Press,
Cambridge, pp. 55–82.
Boyd, L.E., Carbonaro, D.A. and Houpt, K.A. (1988) The 24-hour time budget of Przewalski horses. Applied Animal
Cairns, M.C., Cooper, J.J., Davidson, H.P.B. and Mills, D.S. (2002) Association in horses of orosensory characteristics of
foods with their post-ingestive consequences. Animal Science 75, 257–265.
Christensen, J.W. (2016) Early-life object exposure with a habituated mother reduces fear reactions in foals. Animal
Christensen, J.W. and Rundgren, M. (2008) Predator odour per se does not frighten domestic horses. Applied Animal
Christensen, J.W., Keeling, L.J. and Nielsen, B.L. (2005) Responses of horses to novel visual, olfactory and auditory stimuli.
Clutton-Brock, J. (1992) Horse Power. A History of the Horse and the Donkey in Human Societies. Natural History Museum
Publications, London.
Clutton-Brock, T.H., Greenwood, P.J. and Powell, R.P. (1976) Ranks and relationships in highland ponies and highland
cows. Zeitschrift für Tierpsychologie 41, 202–216.
Colin, G. (1886) Traite de Physiologie Comparée, 3rd edn. Baillière, Paris.
Crabtreee, J.R. (2016) Can ovariectomy be justified on grounds of behaviour? Equine Veterinary Education 28, 58–59.
Crowell-Davis, S.L. (1985) Nursing behaviour and maternal aggression among Welsh ponies (Equus caballus). Applied
Crowell-Davis, S.L., Houpt, K.A. and Kane, L.C. (1987) Play development in Welsh pony (Equus caballus) foals. Applied
Cuddeford, D. (1999) Why feed fibre to the performance horse today? In: Proceedings of the BEVA Specialist Days on
Nutrition and Behaviour. EVJ Ltd, Newmarket, UK, pp. 50–54.
Curry, M., Eady, P. and Mills, D. (2007) Reflections on mare behavior: social and sexual perspectives. Journal of Veterinary
Behavior: Clinical Applications and Research 2, 149–157.
Dalla Costa, E., Minero, M., Lebelt, D., Stucke, D., Canali, E. and Leach, M.C. (2014) Development of the Horse Grimace
Scale (HGS) as a pain assessment tool in horses undergoing routine castration. PLoS One 9, e92281.
Davidson, N. and Harris, P. (2002) Nutrition and welfare. In: Waran, N.K. (ed.) The Welfare of Horses. Kluwer Academic
Publishers, Dordrecht, the Netherlands, pp. 45–76.
Duncan, P. (1980) Time-budgets of Camargue horses. Time-budgets of adult horses and weaned sub-adults. Behaviour 72,
26–48.
Duncan, P. (1985) Time budgets of Camargue horses: III. Environmental influences. Behaviour 92, 188–208.
Duncan, P. (1992) Horses and Grasses. The Nutritional Ecology of Equids and Their Impact on the Camargue. Springer,
New York.
Ellis, A.D. (2010) Biological basis of behaviour in relation to nutrition and feed intake in horses. EAAP Publication 128, 53–
74.
Emerson, L., Griffin, K. and Stevenson, A. (2016) Practice and attitudes regarding trimming of equine vibrissae (sensory
whiskers) in the UK and Germany. Journal of Veterinary Behavior: Clinical Applications and Research 15, 92.
Feh, C. (2005) Relationships and communication in socially natural horse herds. In: Mills, D.S. and McDonnell, S. (eds) The
Domestic Horse. The Evolution, Development and Management of its Behaviour. Cambridge University Press,
Feh, C. and de Mazieres, J. (1993) Grooming at a preferred site reduces heart rate in horses. Animal Behaviour 46, 1191–
1194.
Feist, J.D. and McCullough, D.R. (1975) Reproduction in feral horses. Journal of Reproduction and Fertility Supplement 23,
13–18.
Fleurance, G., Duncan, P. and Mallevaud, B. (2001) Daily intake and the selection of feeding sites by horses in
heterogeneous wet grasslands. Animal Research 50, 149–156.
Fraser, A.F. (1992) The Behaviour of the Horse. CAB International, Wallingford, UK.
Freymond, S.B., Briefer, E.F., von Niederhäusern, R. and Bachmann, I. (2013) Pattern of social interactions after group
integration: a possibility to keep stallions in group. PLoS One 8, e54688.
Fureix, C., Bourjade, M., Henry, S., Sankey, C. and Hausberger, M. (2012) Exploring aggression regulation in managed
groups of horses Equus caballus. Applied Animal Behaviour Science 138, 216–228.
Geor, R.J. (2005a) Diet, feeding and gastrointestinal health in horses. In: Harris, P.A., Mair, T.S., Slater, J.D. and Green. R.E.
(eds) The 1st BEVA and WALTHAM Nutrition Symposia: ‘Equine Nutrition for All’, Harrogate International Conference
Centre, Harrogate, UK, 17–18 September 2005. Equine Veterinary Journal Ltd, Newmarket, UK, pp. 89–94.
Geor, R.J. (2005b) Faster, stronger, sounder: the role of nutritional supplements and feeding strategies in equine athletic
performance. In: Harris, P.A., Mair, T.S., Slater, J.D. and Green. R.E. (eds) The 1st BEVA and WALTHAM Nutrition
Symposia: ‘Equine Nutrition for All’, Harrogate International Conference Centre, Harrogate, UK, 17–18 September
2005. Equine Veterinary Journal Ltd, Newmarket, UK, pp. 105–112.
Gill, E.L. (1987) Factors affecting body condition of New Forest ponies. PhD thesis, University of Southampton,
Southampton, UK.
Gleerup, K. and Lindegaard, C. (2016) Recognition and quantification of pain in horses: a tutorial review. Equine Veterinary
Education 28, 47–57.
Gleerup, K.B., Forkman, B., Lindegaard, C. and Andersen, P.H. (2015) An equine pain face. Veterinary Anaesthesia and
Analgesia 42, 103–114.
Goodloe, R.B., Warren, R.J., Osborn, D.A. and Hall, C. (2000) Population characteristics of feral horses on Cumberland
Island, Georgia and their management implications. Journal of Wildlife Management 64, 114–121.
Goodwin, D., Davidson, H.P.B. and Harris, P. (2005) Selection and acceptance of flavours in concentrate diets for stabled
horses. Applied Animal Behaviour Science 95, 223–232.
Green, N.F. and Green, H.D. (1977) The wild horse population of Stone Cabin Valley, Nevada: a preliminary report. In:
Proceedings of the National Wild Horse Forum: vol. 1. Cooperative Extension Service, University of Nevada, Reno,
Nevada, pp. 59–65.
Greening, L., Shenton, V., Wilcockson, K. and Swanson, J. (2013) Investigating duration of nocturnal ingestive and sleep
behaviours of horses bedded on straw versus shavings. Journal of Veterinary Behavior: Clinical Applications and
Research 8, 82–86.
Hall, C. (2007) The impact of visual perception on equine learning. Behavioural Processes 76, 29–33.
Hall, C., Huws, N., White, C., Taylor, E., Owen, H. and McGreevy, P. (2013) Assessment of ridden horse behavior. Journal
of Veterinary Behavior: Clinical Applications and Research 8, 62–73.
Hall, S.J.G. (2005) The horse in human society. In: Mills, D.S. and McDonnell, S. (eds) The Domestic Horse. The Evolution,
Development and Management of its Behaviour. Cambridge University Press, Cambridge, pp. 23–32.
Harris, P.A. and Arkell, K.A. (2005) How understanding the digestive process can help minimise digestive disturbances due
to diet and feeding practices. In: Harris, P.A., Mair, T.S., Slater, J.D. and Green. R.E. (eds) The 1st BEVA and WALTHAM
Nutrition Symposia: ‘Equine Nutrition for All’, Harrogate International Conference Centre, Harrogate, UK, 17–18
September 2005. Equine Veterinary Journal Ltd, Newmarket, UK, pp. 9–14.
Hartmann, E., Søndergaard, E. and Keeling, L.J. (2012) Keeping horses in groups: a review. Applied Animal Behaviour
Science 136, 77–87.
Hausberger M., le Scolan N., Bruderer, C. and Pierre, J.-S. (2004) Interplay between environmental and genetic factors in
temperament/personality traits in horses (Equus caballus). Journal of Comparative Psychology 118, 434–446.
Heffner, R.S. and Heffner, H.E. (1983) Hearing in large mammals: horses (Equus caballus) and cattle (Bos taurus).
Behavioural Neuroscience 97, 299–309.
Henry, S., Hemery, D., Richard, M.-A. and Hausberger, M. (2005) Human–mare relationships and behaviour toward
humans. Applied Animal Behaviour Science 95, 341–362.
Henshall, C. and McGreevy, P.D. (2014) The role of ethology in round pen horse training – a review. Applied Animal
Hockenhull, J. and Creighton, E. (2013) Training horses: positive reinforcement, positive punishment, and ridden behavior
problems. Journal of Veterinary Behavior: Clinical Applications and Research 8, 245–252.
Hoffman, R.M., Wilson, J.A., Kronfeld, D.S., Cooper, W.L., Lawrence, L.A., Sklan, D. and Harris, P.A. (2001)
Hydrolyzable carbohydrates in pasture, hay, and horse feeds: direct assay and seasonal variation. Journal of Animal
Science 79, 500–506.
Holcomb, K.E., Tucker, C.B. and Stull, C.L. (2015) Shade use by small groups of domestic horses in a hot, sunny
environment. Journal of Animal Science 93, 5455–5464.
Hothersall, B. and Casey, R. (2012) Undesired behaviour in horses: a review of their development, prevention, management
and association with welfare. Equine Veterinary Education 24, 479–485.
Hothersall, B., Harris, P., Sörtoft, L. and Nicol, C.J. (2010) Discrimination between conspecific odour samples in the horse
(Equus caballus). Applied Animal Behaviour Science 126, 37–44.
Houpt, K.A. (2005) Maintenance behaviours. In: Mills, D.S. and McDonnell, S. (eds) The Domestic Horse. The Evolution,
Development and Management of its Behaviour. Cambridge University Press, Cambridge, pp. 94–109.
Houpt, K.A. and Keiper, R. (1982) The position of the stallion in the equine dominance hierarchy of feral and domestic
ponies. Journal of Animal Science 54, 945–950.
Houpt, K.A., Law, K. and Martinisi, V. (1978) Dominance hierarchies in horses. Applied Animal Ethology 4, 273–283.
Houpt, K.A., O’Connell, M.F., Houpt, T.A. and Carbonaro, D.A. (1986) Night-time behavior of stabled and pastured peri-
parturient ponies. Applied Animal Behaviour Science 15, 103–111.
Houpt, K.A., Zahorik, D.M. and Swartzman-Andert, J.A. (1990) Taste aversion learning in horses. Journal of Animal
Science 68, 2340–2344.
Illius, A.W. and Gordon, I.J. (1993) Diet selection in mammalian herbivores: constraints and tactics. In: Hughes, R.N. (ed.)
Diet Selection: An Interdisciplinary Approach to Foraging Behaviour. Blackwell Scientific Publications, Oxford, pp.
157–181.
Jankunis, E.S. and Whishaw, I.Q. (2013) Sucrose bobs and quinine gapes: horse (Equus caballus) responses to taste support
phylogenetic similarity in taste reactivity. Behavioural Brain Research 256, 284–290.
Jeffcott, L.B. (1972) Observations on parturition in crossbred pony mares. Equine Veterinary Journal 4, 209–212.
Keiper, R. (1976) Social organization of feral ponies. Proceedings of the Pennsylvania Academy of Science 50, 69–70.
Kiley-Worthington, M. (1987) The Behaviour of Horses. J.A. Allen, London.
Krueger, K. and Flauger, B. (2011) Olfactory recognition of individual competitors by means of faeces in horse (Equus
caballus). Animal Cognition 14, 245–257.
Laut, J.E., Houpt, K.A., Hintz, H.F. and Houpt, T.R. (1985) The effects of caloric dilution on meal patterns and food intake
of ponies. Physiology & Behavior 35, 549–554.
Lockyer, C. (2005) How to reduce the risk of nutritionally associated laminitis. In: Harris, P.A., Mair, T.S., Slater, J.D. and
Green, R.E. (eds) The 1st BEVA and WALTHAM Nutrition Symposia: ‘Equine Nutrition for All’, Harrogate International
Conference Centre, Harrogate, UK, 17–18 September 2005. Equine Veterinary Journal Ltd, Newmarket, UK, pp. 55–61.
Luescher, U.A., McKeown, D.B. and Dean, H. (1998) A cross-sectional study on compulsive behaviour (stable vices) in
horses. Equine Veterinary Journal Supplement 27, 14–18.
Mader, D.R. and Price, E.O. (1980) Discrimination learning in horses: effects of breed, age and social dominance. Journal of
Mayes, E. and Duncan, P. (1986) Temporal patterns of feeding behaviour in free-ranging horses. Behaviour 96, 105–129.
McCort, W.D. (1984) Behavior of feral horses and ponies. Journal of Animal Science 58, 493–499.
McDonnell, S.M. (2000) Reproductive behaviour of stallions and mares: comparison of free-running and domestic in-hand
breeding. Animal Reproduction Science 60–61, 211–219.
McDonnell, S.M. (2003) The Equid Ethogram: A Practical Field Guide to Horse Behavior. The Blood-Horse, Inc.,
Lexington, Kentucky.
McDonnell, S.M. (2005) Is it psychological, physical, or both? Proceedings of the American Association of Equine
Practitioners 51, 231–238.
McDonnell, S.M. (2016) Revisiting clinical stallion sexual behavior: applying ethology in the breeding shed. Journal of
Equine Veterinary Science 43(Suppl.), S18–S22.
McGreevy, P. and McLean, A. (2011) Equitation Science. Wiley, Chichester, UK.
McGreevy, P.D., French, N.P. and Nicol, C.J. (1995) The prevalence of abnormal behaviours in dressage, eventing and
endurance horses in relation to stabling. Veterinary Record 137, 36–37.
Menard, C., Duncan, P., Fleurance, G., Georges, J. and Lila, M. (2002) Comparative foraging and nutrition of horses and
cattle in European wetlands. Journal of Applied Ecology 39, 120–133.
Merkies, K., DuBois, C., Marshall, K., Parois, S., Graham, L. and Haley, D. (2016) A two-stage method to approach
weaning stress in horses using a physical barrier to prevent nursing. Applied Animal Behaviour Science 183, 68–76.
Miller, R. (1981) Male aggression, dominance, and breeding behavior in Red Desert feral horses. Zeitschrift für
Tierpsychologie 57, 340–351.
Miller, R.M. (1991) Imprint Training of the Newborn Foal. Western Horseman, Inc., Colorado Springs, Colorado.
Mills, D.S. and Clarke, A. (2002) Housing management and welfare. In: Waran, N.K. (ed.) The Welfare of the Horse.
Kluwer Academic, Dordrecht, the Netherlands, pp. 77–97.
Mills, D.S. and Nankervis, K. (1999) Equine Behaviour: Principles and Practice. Blackwell Science, Oxford.
Monard, A.-M. and Duncan, P. (1996) Consequences of natal dispersal in female horses. Animal Behaviour 52, 565–579.
Nagel, C., Erber, R., Ille, N., von Lewinski, M., Aurich, J., Möstl, E. and Aurich, C. (2014) Parturition in horses is
dominated by parasympathetic activity of the autonomous nervous system. Theriogenology 82, 160–168.
Nicol, C.J. (1999) Understanding equine stereotypies. Equine Veterinary Journal Supplement 28, 20–25.
Nicol, C.J. and Badnell-Waters, A.J. (2005) Suckling behaviour in domestic foals and the development of abnormal oral
behaviour. Animal Behaviour 70, 21–29.
Péron, F., Ward, R. and Burman, O. (2014) Horses (Equus caballus) discriminate body odour cues from conspecifics.
Animal Cognition 17, 1007–1011.
Pfister, J.A., Stegelmeier, B.L., Cheney, C.D., Ralphs, M.H. and Gardner, D.R. (2002) Conditioning taste aversions to
locoweed (Oxytropis sericea) in horses. Journal of Animal Science 80, 79–83.
Powell, D.M. (2000) Evaluation of effects of contraceptive population control on behavior and the role of social dominance
in female feral horses (Equus caballus). PhD thesis University of Maryland, College Park, Maryland.
Prache, S., Gordon, I.J. and Rook, A.J. (1998) Foraging behaviour and diet selection in domestic herbivores. Annales de
Zootechnie 47, 335–345.
Pratt-Phillips, S.E., Stuska, S., Beveridge, H.L. and Yoder, M. (2011) Nutritional quality of forages consumed by feral
horses: the horses of Shackleford Banks. Journal of Equine Veterinary Science 31, 640–644.
Proops, L., McComb, K. and Reby, D. (2009) Cross-modal individual recognition in domestic horses (Equus caballus).
Proceedings of the National Academy of Sciences USA 106, 947–951.
Putman, R.J. (1986) Grazing in Temperate Ecosystems: Large Herbivores and the Ecology of the New Forest. Croom Helm
Ltd, London.
Ralston, S.L. and Baile, C.A. (1982) Gastrointestinal stimuli in the control of feed intake in ponies. Journal of Animal
Science 55, 243–253.
Randall, R.P., Schurg, W.A. and Church, D.C. (1978) Response of horses to sweet, salty, sour and bitter solutions. Journal of
Redgate, S.E., Cooper, J.J., Hall, S., Eady, P. and Harris, P.A. (2014) Dietary experience modifies horses’ feeding behavior
and selection patterns of three macronutrient rich diets. Journal of Animal Science 92, 1524–1530.
Rho, J.R., Srygley, R.B. and Choe, J.C. (2004) Behavioral ecology of the Jeju pony (Equus caballus): effects of maternal
age, maternal dominance hierarchy and foal age on mare aggression. Ecological Research 19, 55–63.
Rossdale, P.D. (1967) Clinical studies on the newborn Thoroughbred foal I. Perinatal behavior. British Veterinary Journal
123, 470–481.
Rutberg, A.T. and Greenberg, S.A. (1990) Dominance, aggression frequencies and modes of aggressive competition in feral
pony mares. Animal Behaviour 40, 322–331.
Schönert, S., Reher, M., Gruber, A. and Carstanjen, B. (2012) Use of a deslorelin implant for influencing sex hormones and
male behaviour in a stallion – case report. Acta Veterinaria Hungarica 60, 511–519.
Sigurjonsdottir, H., van Dierendonck, M.C., Snorrason, S. and Thorhallsdottir, A.G. (2003) Social relationships in a group of
horses without a mature stallion. Behaviour 140, 783–804.
Sneddon, J.C. and Argenzio, R.A. (1998) Feeding strategy and water homeostasis in equids: the role of the hind gut. Journal
of Arid Environments 38, 493–509.
Spaas, J., Helsen, W.F., Adriaenssens, M., Broeckx, S. and Duchateau, L. (2014) Correlation between dichromatic colour
vision and jumping performance in horses. The Veterinary Journal 202, 166–171.
Stahlbaum, C.C. and Houpt, K.A. (1989) The role of flehmen response in the behavioural repertoire of the stallion.
Physiology & Behavior 45, 1207–1214.
Tyler, S.J. (1972) The behaviour and social organization of the New Forest ponies. Animal Behaviour Monographs 5, 85–
196.
van den Berg, M., Lee, C., Brown, W. and Hinch, G. (2016) Does energy intake influence diet selection of novel forages by
horses? Livestock Science 186, 6–15.
van Dierendonck, M.C., de Vries, H. and Schilder, M.B.H. (1995) An analysis of dominance, its behavioural parameters and
possible determinants in a herd of Icelandic horse. Netherlands Journal of Zoology 45, 363–385.
van Dierendonck, M.C., de Vries, H., Schilder, M.B., Colenbrander, B., Þorhallsdóttir, A.G. and Sigurjónsdóttir, H. (2009)
Interventions in social behaviour in a herd of mares and geldings. Applied Animal Behaviour Science 116, 67–73.
Virga, V. and Houpt, K.A. (2001) Prevalence of placentophagia in horses. Equine Veterinary Journal 33, 208–210.
Waran, N.K. (2002) The Welfare of Horses. Kluwer Academic Publishers, Dordrecht, the Netherlands.
Waters, A.J., Nicol, C.J. and French, N.P. (2002) Factors influencing the development of stereotypic and redirected
behaviours in young horses: findings of a four year prospective epidemiological study. Equine Veterinary Journal 34,
572–579.
Wathan, J., Burrows, A.M., Waller, B.M. and McComb, K. (2015) EquiFACS: the equine facial action coding system. PLoS
One 10, e0131738.
Welsh, D.A. (1975) Population, behavioural and grazing ecology of the horses of Sable Island, Nova Scotia. PhD
dissertation, Dalhousie University, Halifax, Canada.
Williams, D.C., Aleman, M., Holliday, T.A., Fletcher, D.J., Tharp, B., Kass, P.H., Steffey, E.P. and Lecouteur, R.A. (2008)
Qualitative and quantitative characteristics of the electroencephalogram in normal horses during spontaneous drowsiness
and sleep. Journal of Veterinary Internal Medicine 22, 630–638.
Wilson, W.J., Mills, P.C. and Dzulkarnain, A.A. (2011) Use of BAER to identify loss of auditory function in older horses.
Australian Veterinary Journal 89, 73–76.
Yarnell, K., Hall, C., Royle, C. and Walker, S.L. (2015) Domesticated horses differ in their behavioural and physiological
responses to isolated and group housing. Physiology & Behavior 143, 51–57.
Yeon, S.C. (2012) Acoustic communication in the domestic horse (Equus caballus). Journal of Veterinary Behavior:
Clinical Applications and Research 7, 179–185.
Żurek, U. and Danek, J. (2012) Foal rejection – characteristics and therapy of inadequate maternal behaviour in mares.
Annals of Animal Science 12, 141–149.
13 Behaviour of Cattle
C.B. Tucker
13.1 Origins of Cattle

Domestication
There are two types of cattle: zebu (Bos indicus) and taurine (Bos taurus). Zebu cattle have a
hump at the shoulder and are found mainly in eastern Eurasia and eastern Africa. Taurine
cattlesss do not have a hump and are the predominant cattle in the rest of the world. Cattle
were domesticated between 8000 and 10,000 years ago. Until recently, it was thought that
both zebu and taurine cattle were different forms originating from the same domestication
event of aurochs (Bos primigenius) or wild cattle. More recent genetic analysis suggests that
the domestication process was more complex: two domestication events for taurine cattle
(from Bos primigenius primigenius), once in Eurasia, once in Africa; and a separate, third
domestication event for zebu cattle (from Bos primigenius namadicus) (Bruford et al., 2003).
Breeds
Despite the genetic and physical differences between zebu and taurine cattle, the two types
can interbreed and produce fertile offspring. Zebu cattle are more tolerant of heat than taurine
cattle and the two types are intermixed to create a hardy beef animal, common in hot
countries like Australia. Indeed, there are hundreds of breeds of cattle throughout the world,
produced through centuries of selective breeding, both within and between the two types of
cattle.
In the developing world, cattle serve many functions including food production (both
milk and meat), as draft animals and to maintain grassland. Multi-purpose breeds or breeds
adapted to local weather and grazing conditions are popular. In the industrialized world,
specialized breeds dominate milk and meat production.
Specialized breeds, including Holstein-Friesian and Jersey cattle, are used for milk
production. There are several common husbandry systems in the dairy industry. For example,
in tie-stall barns, cows are kept indoors and tied in one location. Cows are milked, fed and lie
down in their individual stall. Other types of barn utilize a central location for milking and
cows are brought to and from the parlour several times per day. Two systems, freestall (or
cubicle) barns and loose housing, utilize a central milking parlour and provide separate
feeding and lying areas. Feed, water and lying areas are connected by walkways and cows
freely go among these resources. In freestall barns (see Fig. 13.1a), the lying area is divided
by partitions while in loose-housing systems the lying or bedded area is open (see Fig.
13.1b). Dairy cows may also be kept on pasture and bred seasonally (see Fig. 13.1c), such
that calving corresponds with grass growth. Finally, some farms utilize automatic milking
systems, where robots milk cows as they visit, either voluntarily or in order to reach food or
other resources. Robots have been used in combination with freestall, loose and pasture-
based systems.
Fig. 13.1. Common housing systems for dairy cattle include (a) freestall barns, (b) loose
housing in barns and (c) pasture.
Specific breeds of cattle are used for meat production and examples include Angus,
Hereford and Charolais. Beef production is often divided into two phases: cow–calf
operations and finishing/growing to market weight. Cow–calf operations breed cows and
raise calves with the dam until approximately 6 months of age (see Fig. 13.2a). Once
weaned, beef cattle move to finishing operations. In North America and parts of Europe, beef
cattle are typically finished on a grain-based diet in a feedlot (see Fig. 13.2b). In other parts
of the world, beef calves are finished on pasture.
Fig. 13.2. Common systems for beef cattle include (a) cow–calf operations and (b) feedlots
or finishing units.

Cattle are gregarious animals. Indeed, if separated from other animals, isolated cattle will
show clear signs of stress including increased heart rate, vocalization and
defecation/urination (Rushen et al., 1999). Feral cattle aggregate in groups of cows and
calves. Bulls form separate groups and may defend specific areas within the larger
environment and intermittently interact with cow–calf groups. Humans determine group size
and composition in farmed cattle and these groups are typically: (i) all-adult or all-juvenile
females (dairy operations); (ii) a mix of cows, calves and a few bulls during the breeding
season (cow–calf operations); or (iii) a mix of both sexes (feedlots, sometimes castrated or
spayed, depending on age and practice within a country).
The structure within groups of cattle is categorized by both aggressive and affiliative
behaviour. Aggressive behaviour includes threats such as lowering the head (as though to
present horns) and can escalate to physical contact: head-butting the head or body of another
individual (see Fig. 13.3) or head-to-head pushing. In bulls, threat displays are more
elaborate and include vocalizations, pawing and rubbing the head on the ground, and
postures that make the bull look larger. The most common affiliative behaviour in cattle is
allogrooming, or social licking. Social licking between adult cattle is often directed at the
neck region of the body and cattle form grooming partnerships with specific individuals
within a group.
Fig. 13.3. Two dairy cows head-butting the individual in the middle; cows are at the feed
bunk.
Dominant–subordinate relationships are established and influenced by both positive and

aggressive social interactions. These relationships can affect access to resources such as
water, food, lying space, shelter and, in the case of bulls, oestrous females. Aggressive
interactions are common when unfamiliar individuals are mixed together but generally
decline over time as animals establish a dominance hierarchy. In addition, recent evidence
suggests that long-term familiarity also plays a role in the stability of relationships among
dairy cows, for example. Cows that have interacted over a long period of time are more likely
to spend time feeding and lying near each other and allogrooming, compared with those that
have relatively only recent experience with one another (Gutmann et al., 2015). Other
factors, such as testosterone levels, also affect the nature of the social relationships.
Aggression is more common in intact males than castrated steers. For example, castrated
steers are less likely to initiate a fight or spar with other animals compared with intact bulls
(Price et al., 2003).
Individual characteristics such as the presence of horns and body size can influence
social success. For example, Bouissou (1972) compared social success of cattle of similar
size and found that cattle with horns were dominant to dehorned cattle 85% of the time.
Additionally, this same study found that when both animals were hornless, heavier animals
tended to dominate lighter animals.
Cattle are both herbivores and ruminants – meaning that their natural diet consists solely of
plant materials, and that their stomach is divided into four separate compartments: the
reticulum, rumen, omasum and abomasum. The rumen is the largest compartment and acts as
a fermentation chamber that breaks down cellulose and allows cattle to eat and digest plant
material such as grasses, grains and husks. Cattle graze by gathering grass into the mouth
with the tongue. By holding the grass between the bottom incisors and/or tongue against the
upper palate, cattle are able to rip the grass. Cattle do not have incisors on the upper palate,
instead having a hard, ridged dental pad that is used as a grinding surface during rumination.
Rumination is distinct behaviour performed by this class of animal and is sometimes called
‘chewing the cud’. Cattle will ingest plant material and then, after a meal, regurgitate one
bolus of partially digested food at a time, re-chew this food and then swallow again. This
additional chewing helps break down cellulose or plant fibre that would otherwise prove
indigestible. Cattle will spend between 6 and 8 h ruminating each day and can ruminate while
either lying or standing. The length of each feeding and rumination bout varies with feed type
and availability.
Indeed, all aspects of feeding behaviour are influenced by both the distribution and the
type of feed. Cows fed in a barn will typically eat for 4 to 6 h/day, while on pasture cattle
spend more time grazing, 6 to 10 h/day. In addition to time spent feeding, the type of
production system also affects patterns of feeding behaviour. For example, dairy cows on
pasture will have more synchronous feeding behaviour and feed less at night compared with
dairy cows in freestall barns (see Fig. 13.4a). Grass in pasture systems is spread out and all
cows can feed at one time, as they do when fresh grass in provided in the morning. In
freestall barns, feed is available only at the feed bunk and access to this bunk may be limited
by space, making it less likely that all cows will feed at the same time.
Fig. 13.4. Pattern of (a) feeding and (b) lying behaviour for dairy cattle kept on pasture (grey
line, from Tucker et al., 2008) or in a freestall barn (black line, from DeVries and von
Keyserlingk, 2005). All cows were milked twice per day. The percentage cows performing
each behaviour is plotted over 24 h. The arrows in (a) indicate when fresh feed was
delivered.
13.4 Diurnal Rhythm

Cattle show distinct diurnal rhythms when on pasture. They spend more time lying down at
night, although there is often a lying bout during the middle of the day. Indeed, cattle spend
between 8 and 12 h lying down per day, and each lying bout usually lasts just over 1 h. The
diurnal patterns of feeding and lying down are highly synchronized when cattle are kept on
pasture. A number of housing and management factors influence these patterns.
Housing system and feed delivery
The rhythms of feeding during the day and lying down at night are less synchronous when
dairy cows are kept in freestall barns (see Fig. 13.4). Cattle kept in barns tend to spread out
both feeding and lying activity throughout the day and night. However, several cues still play
an important part in synchronizing their behaviour in the barn. For example, delivery of fresh
feed brings cows to the feed bunk (see arrows in Fig. 13.4a). On many dairy farms, feeding
and milking take place at the same time. DeVries and others have separated out these two
factors and found that feed delivery plays the key role in stimulating feeding behaviour.
Cows increased feeding by 82% during the first hour immediately following the delivery of
fresh feed compared with a 26% decrease in feeding time during the first hour after returning
to the pen from milking (DeVries and von Keyserlingk, 2005).
Weather and access to shade
Other factors, such as weather and access to shade, can affect the behavioural patterns of
cattle. For example, cattle will spend more time standing and less time lying down in warm
weather (e.g. Tucker et al., 2008). The reason for this response is unclear. Cattle may lose
more heat when standing because of air circulating around the body; or respiration, one of the
main forms of heat loss, may be more effective while upright. Cattle will also make other
changes in order to cool down in warm weather. Dairy cows will spend more time standing in
shade during the day and graze more at night in order to avoid the heat of the day (Kendall et
al., 2006).
13.5 Mating and Sexual Behaviour

Cattle are polyoestrous, meaning they are able to breed year-round. In pasture-based dairy
systems and in beef cow–calf operations, breeding is typically managed such that calves are
born when grass growth begins. On barn-based dairy farms, cows are managed to calve year-
round in order to produce a steady supply of milk for human consumption.
Females usually are 9 months of age at first oestrus and each oestrous cycle lasts 21 days.
The duration of the receptive phases of oestrus, where the cow will stand to be mounted, lasts
approximately 12 h. Receptive females (sometimes called ‘in heat’) will stand to be mounted
by bulls, but cattle also show female–female mounting. In systems with bulls, female–female
mounting can attract the attention of males. In addition to visual stimuli, bulls use olfaction
to detect if a cow is receptive. Bulls will sniff the genital region and urine of cows and may
display a flehmen response in which the upper lip is curled while the head is outstretched.
The flehmen response assists pheromones and scent molecules to reach the vomeronasal
organ at the roof of the mouth. This organ is used to detect specific chemicals typically
associated with oestrus. Bulls will guard receptive females and attempt to prevent mating
with other males although, depending on the sex ratio within the herd, it is likely that a
female will be serviced by more than one male (Petherick, 2005).
In farming systems without males (as in the case in many dairy farms, and in some beef
operations), artificial insemination is used. When a cow is artificially inseminated, previously
collected semen is inserted into her reproductive tract by a farm worker. Successful artificial
insemination requires that humans detect the receptive phase of oestrus. There are various
ways to detect oestrus in cattle and behaviour plays a key role in each method. Cows may be
observed for female–female mounting to detect oestrus. Alternatively, the tail head of the
cow is painted and this paint is ‘read’ for signs of rubbing. If the paint wears away, this
indicates that the cow has stood to be mounted and is ready to be bred. Cows also show
restless behaviour during the receptive phase of oestrus. Automated measures of movement,
such as a sensor on a collar or a pedometer attached to the leg, are used to record behaviour
and detect the increase in activity associated with oestrus.
Bulls begin to mount conspecfics at approximately 2 months of age. By 4 to 6 months of
age, bulls will mount regularly, but do not reach sexual maturity until approximately 1 to 2
years of age. In systems where multiple males compete for access to females, bull age, size
and dominance all affect opportunity to successfully mate with females (Petherick, 2005).
Indeed, in groups of bulls, only a portion successfully sires offspring. For example, in a
group of 27 bulls on a commercial cow–calf operation, five bulls sired 50% of the offspring
while ten bulls sired no offspring at all (Van Eenennaam et al., 2007).
In order to assess which bulls to use for mating, libido (i.e. both the motivation and
ability to mate) can be evaluated with a service capacity test. Groups of bulls are placed into
a pen with restrained females. Different measures of libido include time spent with each
female and/or the number of mounts in a limited period, for example, 20 min (Petherick,
2005). Although libido is important for breeding success, good performance in a service
capacity it does not guarantee successful reproductive performance.
13.6 Parturition
Gestation in cattle lasts approximately 9 months. Cows separate themselves from the herd
before calving, if adequate space is provided. Dairy cattle will also use a secluded area during
parturition: 79% of animals chose this option over a completely open pen (Proudfoot et al.,
2014). Cows also show restless behaviour and spend more time standing the day before
calving. However, cows give birth lying down and stand after parturition. They then begin to
lick the calf. Licking serves several functions: to encourage the calf to stand and suckle, to
clean off amniotic fluid and to facilitate recognition of the individual calf by the dam. Cows
also engage in placentophagia or consumption of some or the entire placenta. This behaviour
may serve several functions. The placenta may have nutritional value. In addition, cattle are
considered ‘hider’ species; that is, cows will leave the calf behind when foraging in the first
few days after parturition. Placentophagia may reduce detection of the calf by predators. In
contrast, the young of ‘follower’ species like sheep immediately follow the dam after birth,
making placentophagia less important for these animals.

Calves usually suckle for the first time within hours after birth. Calves gain access to the
udder from in front of the hind leg of the dam (see Fig. 13.2a). The first milk after birth,
colostrum, contains immunoglobulins that play an important role in the development of the
calf’s immune system. The lining of the gut is initially porous enough to absorb these
immunoglobulins, but this permeability declines rapidly after birth.
In cattle raised for beef, the calf remains with the cow until weaning, at approximately 6
months of age. Initially, cows seem to initiate many of the suckling bouts, but after a few
weeks, calves seek out the cow. Nursing bouts are also most common early on and decline as
the calf ages. On average, calves suckle between four and ten times per day, although
estimates of suckling frequency vary between studies. Initially, calves remain in peer groups
or crèches for the first days or months. During this phase, cows will leave the calves to graze
and will return to suckle them.
Cows will also vocalize in the weeks after birth and may use vocalizations during
reunification with their calf. Quiet grunting or contact calls are common immediately after
partition and decline in the first weeks of the calf’s life. It is thought that these contact calls
aid in auditory recognition between dam and calf (von Keyserlingk and Weary, 2007).
Occasionally, beef calves are fostered by another dam, or cross-fostered, particularly if
the original dam has died. Similarly, one dairy cow may raise three or four calves from a
young age. Several methods are used to establish a maternal bond with alien calves. One
approach is to make the alien calf smell like the dam’s calf by either covering it with
amniotic fluid from her own calf, or placing the coat of her own calf (if dead) on the alien
calf as a jacket. However, this approach is viable only if the cross-fostering takes place near
the time of partition. Cows may show aggression (kicking, butting) towards an alien calf
when it attempts to suckle. In order to successfully cross-foster in days after birth, a cow may
need to be restrained until she accepts the alien calf.
Although dairy calves may be raised with a foster dam, it is more common to separate
them from the dam within hours of birth. Early separation of cows and calves is thought to
prevent transmission of disease and improve ease of milking the cow in a milking parlour.
However, calves that remain with the dam for longer periods (e.g. 2 weeks) gain more weight
than their counterparts reared by humans. The behavioural response of both the cow and calf
at the time of separation is more marked the more time the pair spend together (Flower and
Weary, 2003).
When separated from the dam, dairy calves are reared by humans in isolation or in
groups of same-age peers. Milk, milk replacer (similar to baby formula for humans) or waste
milk (milk from cows treated with antibiotics that is not used for human consumption) is
used to feed dairy calves. Regardless of source of milk or milk product, calves are typically
fed from a bucket or teat. Teat feeding systems range from simple systems with a bucket,
one-way valve and rubber nipple to automated calf feeders that dispense a specified amount
of food to a given individual. The amount of milk provided varies between production
system and individual farms.
Calves are highly motivated to suck. In dairy systems where calves are fed restricted
amounts of milk, this motivation to suck may be unfulfilled. Cross-sucking, or sucking on
another calf, typically on the ears, head and naval region, is an undesirable behaviour
performed in groups. Although rearing calves in isolation is common in some geographic
regions to prevent this sort of problem, there is growing evidence that this practice impairs
the behavioural development of these young animals (Costa et al., 2016). Calves are often
fed 10% of body weight during the milk-fed period and there is clear evidence that this
feeding level is insufficient (Khan et al., 2011). There are numerous benefits associated with
higher feeding levels administered through a teat, including reduced cross-sucking and much
improved growth rates (e.g. De Paula Vieira et al., 2008). Finally, a combination of slower
milk flow through a teat, hay feeding and access to a non-nutritive artificial teat also reduces
cross-sucking, by providing more opportunity for both nutritive sucking and other forms of
oral manipulation (de Passillé, 2001).

Calves change how they spend their time as they age. Young calves spend a considerable
amount of time lying down, but as calves get older they spend less time lying. Similarly,
calves with access to pasture begin to manipulate grass within the first days of life and
increase time spent grazing as they age. Regardless of management system, all calves are
dependent on milk at the beginning of their life. Their digestive system allows them to
maximize the nutritional value of milk by routing the liquid directly to the abomasum or
glandular stomach (similar to monogastric animals). The oesophageal groove allows milk to
bypass fermentation in the rumen. As calves age, they increase the amount of solid food and
time spend grazing, such that they can survive solely on solid food.
Dairy calves are typically weaned from milk at 5 to 12 weeks of age. The behavioural
response to weaning depends on how the process is managed. When weaning is abrupt,
calves vocalize and are more active at the time milk would normally be delivered. It seems
likely that calves are responding to more than the cessation of milk feeding; they also may
take comfort in other aspects of the feeding regime. Indeed, calves will show this same
behavioural response, increased vocalizations and activity, when weaned from warm water
rather than milk. It is unlikely that water-fed calves are dependent on the water for nutritional
reasons at this time, as they consume up to 1 kg of calf starter or grain per day. Together,
these results indicate that hunger is not the only factor driving the behavioural response to
weaning (Jasper et al., 2008).
Beef calves are typically weaned at 6 months of age. Abrupt weaning involves physically
separating cows and calves. This process results in both nutritional and social stress because
both the dam and milk are removed at the same time. Both cows and calves show marked
behavioural response to weaning: vocalization, walking and reduction in time spent eating.
Calves abruptly separated from their dam will walk, on average, 16 km in one day. As with
dairy calves, weaning distress is not solely nutritional. To demonstrate this point, consider an
alternative to abrupt weaning, or two-step weaning. The first step in two-step weaning is to
place a plastic ring in the nose of the calf. This ring prevents the calf from nursing, but allows
the calf to eat solid food. Several days later, the calf is separated from its mother (step two).
Calves weaned in this manner vocalize and walk less and spend more time resting compared
with abruptly weaned calves (Haley et al., 2005).
References
Bouissou, M.F. (1972) Influence of body weight and presence of horns on social rank in domestic cattle. Animal Behaviour
20, 474–477.
Bruford, M.W., Bradley, D.G. and Luikart, G. (2003) DNA markers reveal the complexity of livestock domestication. Nature
Reviews Genetics 4, 900–910.
Costa, J.H.C., von Keyserlingk, M.A.G. and Weary, D.M. (2016) Invited review: Effects of group housing of dairy calves on
behaviour, cognition, performance, and health. Journal of Dairy Science 99, 2453–2467.
de Passillé, A.M. (2001) Sucking motivation and related problems in calves. Applied Animal Behaviour Science 72, 175–
187.
De Paula Vieira, A., Guesdon, V., de Passillé, A.M., von Keyserlingk, M.A.G. and Weary, D.M. (2008) Behavioural
indicators of hunger in dairy calves. Applied Animal Behaviour Science 109, 180–189.
DeVries, T.J. and von Keyserlingk, M.A.G. (2005) Time of feed delivery affects the feeding and lying patterns of dairy
cows. Journal of Dairy Science 88, 625–631.
Flower, F.C. and Weary, D.M. (2003) The effects of early separation on the dairy cow and calf. Animal Welfare 12, 339–348.
Gutmann, A.K., Špinka, M. and Winckler, C. (2015) Long-term familiarity creates preferred social partners in dairy cows.
Haley, D.B., Bailey, D.W. and Stookey, J.M. (2005) The effects of weaning beef calves in two stages on their behavior and
growth rate. Journal of Animal Science 83, 2205–2214.
Jasper, J., Budzynska, M. and Weary, D.M. (2008) Weaning distress in dairy calves: acute behavioural responses by limit-fed
calves. Applied Animal Behaviour Science 110, 136–143.
Kendall, P.E., Nielsen, P.P., Webster, J.R., Verkerk, G.A., Littlejohn, R.P. and Matthews, L.R. (2006) The effects of
providing shade to lactating dairy cows in a temperate climate. Livestock Science 103, 148–157.
Khan, M.A., Weary, D.M. and von Keyserlingk, M.A.G. (2011) Invited review: Effects of milk ration on solid feed intake,
weaning, and performance in dairy heifers. Journal of Dairy Science 94, 1071–1081.
Petherick, J.C. (2005) A review of some factors affecting the expression of libido in beef cattle, and individual bull and herd
fertility. Applied Animal Behaviour Science 90, 185–205.
Price, E.O., Adams, T.E., Huxsoll, C.C. and Borgwardt, R.E. (2003) Aggressive behavior is reduced in bulls actively
immunized against gonadotropin-releasing hormone. Journal of Animal Science 81, 411–415.
Proudfoot, K.L., Jensen, M.B., Weary, D.M. and von Keyserlingk, M.A.G. (2014) Dairy cows seek isolation at calving and
when ill. Journal of Dairy Science 97, 2731–2739.
Rushen, J., Boissy, A., Terlouw, E.M.C. and de Passillé, A.M.B. (1999) Opioid peptides and behavioral and physiological
responses of dairy cows to social isolation in unfamiliar surroundings. Journal of Animal Science 77, 2918–2924.
Tucker, C.B., Rogers, A.R. and Schutz, K.E. (2008) Effect of solar radiation on dairy cattle behaviour, use of shade and
body temperature in a pasture-based system. Applied Animal Behaviour Science 109, 141–154.
Van Eenennaam, A.L., Weaber, R.L., Drake, D.J., Penedo, M.C.T., Quaas, R.L., Garrick, D.J. and Pollak, E.J. (2007) DNA-
based paternity analysis and genetic evaluation in a large, commercial cattle ranch setting. Journal of Animal Science 85,
3159–3169.
von Keyserlingk, M.A.G. and Weary, D.M. (2007) Maternal behavior in cattle. Hormones and Behavior 52, 106–113.
14 The Behaviour of Sheep and Goats
C. Dwyer
14.1 Origins of Sheep and Goats

Sheep and goats are ungulates (or hoofed mammals), belonging to the highly successful order
Artiodactyla, the family Bovidae (including true bovines, buffalo, goats and sheep) and the
Tribe Caprini (comprising sheep and goats). Wild sheep and goats are among the most
successful Pleistocene mammals, with wide geographical distributions extending from
Europe to Siberia and Alaska to South America. Although the taxonomy of wild sheep and
goats is not clear, and many subspecies exist, there are thought to be six wild species of sheep
(Ovis genus) found in Europe, Asia and North America. The six wild goat species (Capra
genus) are all located in Asia and the Mediterranean basin. Although the wild ancestors of
domestic sheep and goats are generally found in hilly and rugged country, they are highly
adaptive and have successfully colonized a variety of terrains, including desert and island
habitats, and are found in the Arctic and sub-Antarctic. In Asia and Europe sheep and goats
have competed for habitat, resulting in sheep occupying lower mountain slopes and hills,
whereas goats are found in steep cliff areas (see Fig. 14.1; Clutton-Brock, 1999). In North
America the absence of competition from goats has influenced the distribution of wild sheep,
for example the mountain Bighorn, which range over the highest peaks. The behaviour and
habitat of some of these wild sheep and goat species have been extensively studied (e.g.
mountain Bighorn by Geist, 1971; Asian sheep and goats by Schaller, 1977) and provides
valuable insights into the behaviour of the domestic animal.
Fig. 14.1. Wild goats (Capra ibex) in Val d’Aosta, Italy, below Mont Blanc. (Photo courtesy
of Tony Waterhouse.)
History of domestication and breed development

Sheep and goats were among the first species to be domesticated around 10,000 years ago in
the so-called ‘Fertile Crescent’ of the Middle East. Domestic sheep (O. aries) may have
arisen from the domestication of more than one Ovis species: the mouflon (O. orientalis)
contributing to domestic sheep in Europe and the argali (O. ammon) to Asiatic breeds.
Domestic goats (C. hircus) are thought to be descended from one main species, the bezoar
goat (C. aegagrus), although there may have been several simultaneous domestication events
in different parts of the world. The first stage of domestication is the formation of loose ties,
for example by the sharing of watering places, but there is evidence of confinement and
breeding control by Neolithic farmers and the presence of ‘breeds’ by the Bronze Age
(Ryder, 1984). Early agricultural settlements and the cultivation of crops meant that animals
could be kept in enclosures at night and some protection from predators provided. In the
writings of the ancient Greeks, there are descriptions of herding sheep to fresh mountain
pastures in spring and penning in winter, where they were offered a range of feedstuffs (from
barley, clover and alfalfa to oak leaves, figs and pressed grape residues from winemaking).
Once breeding was under the control of man, sheep and goat characteristics began to be
shaped by man’s requirements; for example, white-woolled sheep were preferred over the
wild pattern of brown with a white belly. These developments have led to an abundance of
different sheep and goat breeds, specialized for different environments and production
characteristics: there are now in excess of 1000 sheep breeds worldwide and nearly 600
breeds of goat.
Use and economic importance
Although sheep and goats were domesticated at about the same time, sheep began to become
more popular around the first century ad, particularly in the West. There are now 1.1 billion
domestic sheep, in comparison to 700 million domestic goats, with more than three-quarters
of the world’s goat population found in the developing world, where there may be one goat to
every three people in some countries. The ability of the goat to cope with a harsh
environment and poor food quality has led to it being dubbed the ‘poor man’s cow’.
However, both species owe their popularity to their multi-purpose ability to provide milk,
meat, skins, dung for fuel and wool or fibre. They act as ‘walking larders’ and a visible
display of wealth for nomadic peoples. In Tibet, sheep and goats are even used for portage to
carry salt or grain. In India, sheep play an important role in cultural and religious rituals.
Before shearing (which is considered to be a sacred communal function), sheep participate in
ceremonies to Laxmi (the goddess of money) where they are washed, decorated with paint
and fed jaggery and coconut. Lambs born on days with important religious significance
become devotional animals and neither they nor their offspring are sold or slaughtered, but
confer status and respect on their owners.
Sheep and goats are generally found in regions where the climate is harsh and the terrain
unsuitable for other types of agriculture. Sheep are better at coping with cold and wet
climates, compared with goats, and this has aided their spread from the Middle East, with the
largest sheep populations now found in China and Australia. Goats are overwhelmingly
found in Africa and Asia (more than 90% of the goat population), with greatest numbers
found in India and China. Goats generally sustain small communities, in comparison to the
large-scale production of wool and meat from sheep in Australia and New Zealand. Scientific
research into the behaviour of sheep and goats has also predominantly considered sheep,
although there are increasing investigations into the behaviour of domestic goats.

Both wild and domestic sheep and goats are highly social and live in small to moderate group
sizes. Social living provides protection from predators, assists finding a mate and food, and
helps with care and protection of the young. Groups are matrifocal, where females and their
juvenile offspring remain together on a home range and smaller groups of males are
segregated from the female flock but share an overlapping home range. Some producers of
domestic sheep exploit this social group behaviour by removing the male lambs at weaning
but keeping females together on an unfenced pasture or ‘heft’. Cultural knowledge about the
distribution of food, water and shelter sites is passed on from mothers to daughters and
grand-daughters, and flocks do not stray into adjoining home ranges if they are already
occupied by other sheep flocks. The propensity to show home range behaviour, social group
size and the strength of social attachment varies with breed, with an increase in tolerance for
close contact with a larger number of animals occurring in more highly selected and
domesticated breeds (Dwyer, 2004).
Following and synchronous behaviour
Mother goats and ewes form close bonds with their offspring and this encourages following
behaviour where the offspring accompany first their mother and then any adult sheep about
the home range. Movement generally occurs in single file following a leader, which is an
older and more independent animal. The most dominant animal in the group is usually
towards the front of the movement order (as it can obtain the best grazing when arriving at
the new location) but rarely leads as this may be a more vulnerable position. The allomimetic
or synchronous behaviours of sheep and goats may be an anti-predator strategy as individual
animals are inconspicuous when all animals are engaged in the same activity. Following
behaviours and the maintenance of group cohesion are exploited by farmers when moving
animals, and it is well known that moving a single animal is far more difficult than a group of
sheep or goats. In many societies, sheep are trained to follow the shepherd, who leads the
animals to different pastures or to shelter. Some abattoirs make use of ‘Judas’ sheep, which
live in the abattoir and will lead animals through the lairage and races to encourage easy
movement.
Agonistic behaviour
Female sheep and goats do not often engage in agonistic behaviour, except when competing
for limited resources, and dominance is maintained by subtle behaviours such as eye contact
and resting the chin on the back of another animal to displace it. Male animals, however,
engage in impressive displays and fighting, closely linked to the morphology of their horns
and skulls, primarily to gain access to females. Horns act as rank symbols, where horn size is
recognized and fights do not take place between males of unequal horn size (Geist, 1971),
thus reducing the risk of damage and injury. High-intensity fights, between males of equal
status, are relatively rare, but can last for several hours when they do take place. The
dominant feature of fights by goats is the ‘clash’ where opponents rear up on their hind legs
and crash their heads and horns together. Goat skulls are particularly thick to protect the brain
from the massive impacts that clashes cause. Although sheep also fight by clashing, only
mountain sheep are reported to rear up to deliver blows, and physical fights are preceded by
ritualized agonistic displays that resemble courtship behaviours (see below), including
nudging, kicking with the forelegs and making ‘rumbles’ or growling vocalizations.
Social recognition
If two groups of sheep are mixed, they will remain separate for some time, demonstrating
that the sheep recognize unfamiliar animals. Over time, sheep of the same breed will become
integrated into a single group; however, different breeds of sheep remain as separate groups
even after prolonged exposure to one another. Recognition is based on visual and olfactory
(smell) cues, thus animals that appear visually distinct (as with two breeds) do not integrate,
whereas olfactory differences diminish over time (see Fig. 14.2). Sheep have pedal scent
glands on all four feet, as well as inguinal and facial glands near the eye. Goats have pedal
glands only on two feet and have a tail gland, which may explain why goat tails are
constantly raised whereas sheep tails are usually carried low. Although scent marking is used
in both species, its precise role in social recognition is unclear. The ability of sheep to
distinguish familiar sheep from pictures of their faces alone has uncovered a remarkable
ability of sheep to recognize more than 50 different individuals and to retain that memory for
at least 2 years (Kendrick et al., 2001a).
Fig. 14.2. Social interactions between a group of recently mixed sheep following a sale. Note
many sniffing interactions between animals but no overt aggression or fighting, as might be
expected with other animals after mixing, such as pigs. (Photo courtesy of Fritha Langford.)

Sheep and goats are adapted to cope with harsh climatic conditions and this includes an
ability to utilize a wide variety of food sources. Both species lack front teeth in the upper jaw,
and their smaller mouths allow them to forage closer to the ground than other ungulate
species. Sheep can exploit a wide range of food sources and will eat cacti in the desert,
lichens in the Arctic, tree leaves and fruit, although perhaps the most remarkable adaptation
is that of the feral North Ronaldsay sheep in Orkney, UK, which forages almost exclusively
on seaweed. Seaweed is structurally very different from other plants, being low in copper, but
high in salt, iron and arsenic. However, the sheep are well-adapted to their high salt and low
copper intakes, and can efficiently extract copper from the seaweed to such an extent that
they suffer from copper toxicity on normal pasture. The social and foraging behaviour of
these sheep is also determined by the tide, so that they are able to make best use of the forage
when it is uncovered by the receding tide. Similarly, goats are known to be adventurous in
their feeding habits, sometimes being used to clear areas of thick undergrowth, and they may
sample non-food items such as clothes or plastic bags if these are within reach, which can
cause digestive complications.
Temperate sheep show another adaptation to their harsh environment by the seasonal
decline in appetite, voluntary food intake and metabolic rate that occurs in winter. This
appears to be an adaptation to food scarcity as the decrease in metabolic rate precedes the
decline in appetite and food intake. There is breed variation in the extent of these seasonal
changes, with hill breeds of sheep showing greater decreases in winter than lowland breeds.
Digestive physiology and rumination
Sheep and goats are ruminants (as are cattle and deer), which means that they are foregut
fermenters and have a stomach made up of four chambers: the rumen, reticulum, omasum
and abomasum (the true stomach). The first two chambers contain microbes (bacteria,
protozoa and fungi) which ferment food, including fibre. The fermented contents of the
reticulorumen can pass between the two chambers and are regurgitated and re-chewed
(known as rumination, chewing the cud or cudding). The reticulorumen produces volatile
fatty acids, which are absorbed, and microbial protein which, together with unfermented food
residues, passes into the omasum, and then to the abomasum and small intestine. These
chambers function similarly to the stomach of non-ruminant animals, with enzymatic
digestion and absorption. The rumination process allows ruminant animals to obtain nutrients
from fibrous plant matter more efficiently than hindgut fermenters (such as the horse).
Rumination generally occurs when the animal is lying sternally, although it will ruminate
while standing, and typically occurs for about one-third of the day. Both species graze or
forage for about 8 h/day, although this may increase to around 13 h if suitable forage is
sparse. The need to find time to ruminate is a constraint on the ability of the animal to
increase foraging time to maximize intake. During rumination the animal is in a state of
drowsiness and there has been speculation about whether ruminants do actually sleep.
However, although the amount of sleep and rumination are inversely related, ruminants do
show periods of true sleep, often preceded by rumination.
Browsing and grazing
Although both species will eat a wide variety of foodstuffs, in general sheep are grazers
which occasionally browse, and goats are browsers, although they may also graze. This
means that, when given the choice, sheep will preferentially eat grass and herbage, whereas
goats prefer leaves and shoots from trees and bushes. Goats are well-adapted to their browse
diet with mobile lips to select leaves and a digestive system that is more efficient at dealing
with roughage than sheep. Goats also eat more (in terms of number of bites per day)
compared with sheep, and will consume more dry matter for a similar length of foraging
time. In addition, as agile climbers, in some environments goats will even climb into trees to
obtain browse.
Diet selection
Herbivores foraging in a variable environment (as most pastures are) are continually making
decisions about which plant species, individual plants and parts of the plant to eat. The
composition of their intake may vary markedly from the composition of the pasture and
comes about due to specific preferences, which may be learnt, and active avoidance of plants
that are potentially toxic or may be contaminated with parasites (Forbes and Provenza, 2000).
Dietary preferences show diurnal variation, for example sheep prefer clover in the morning
and grass in the evening, and seasonal variation, suggesting that the animals have some
memory of the location of different food types. In specific tests, sheep have been found to
have excellent spatial memory and are able to learn the location of food patches after a single
trial (Edwards et al., 1996). Similarly, conditioned food aversions also occur after a single
exposure. These abilities are clearly important adaptations for an animal foraging in a
variable environment.
Many factors influence the diet selection made by sheep and goats, including breed
differences, level of hunger, previous experience and social factors. Animals reared in a
particular nutritional environment select a different diet from animals newly introduced to the
environment, suggesting a significant learnt component to diet selection. Thus breed
differences are unlikely to be due to innate differences, but most likely arise due to different
nutritional requirements for growth and/or the different management of breeds that will affect
their previous exposure to plant types. Diet selection and foraging efficiency can also be
constrained or facilitated by social factors. Ewes will not leave the social group to forage on
preferred sites unless accompanied by companions, and will forage for longer on preferred
sites when accompanied by familiar animals. Goats have also been shown to use some plant
species as a medicine, ingesting specific browse species when infected with gastrointestinal
parasites (worms) that are known to alleviate the symptoms of infection, which are not
preferred by goats that are not infected.
Water intake
Although some wild sheep and goats have colonized desert areas, both species still need to
drink almost daily, although they are better adapted to coping with periods without water than
other farmed animals. In general, goats are better at conserving water than sheep and,
possibly due to their browse diet, may be able to obtain nearly all their water requirements
from their food. In drought regions, such as parts of Australia, sheep may spend considerable
amounts of time walking to water, will forage only within range of a water source and can
cope with less than daily drinking only when succulent plants are available. Breed
differences in ability to cope with dry environments are known to exist (Terrill and Slee,
1991), with desert breeds of sheep (fat-tailed and fat-rumped breeds) coping better than more
temperate sheep breeds. The fat-tailed and fat-rumped breeds store fat in adipocytes, in the
rump and tail, that is more rapidly mobilized when required than back fat.
14.4 Habitat Use

Predation risks and food availability are the major evolutionary forces shaping habitat use in
sheep and goats. The most important features of the habitat of wild sheep and goats are
proximity to escape terrain (as flight to cover is their main anti-predator response),
availability of suitable forage and a water supply. Goats and sheep are not territorial and the
home range is not defended. In many situations, domestic sheep and goats are not offered a
choice of habitat and may be kept in rather small fenced paddocks. However, they still show
the diurnal shifts in behaviour shown by their wild progenitors. Sheep generally camp in the
hills if available, or on elevated ground, and move down to the lower regions at dawn to
graze. In temperate climates, sheep graze in the morning, rest and ruminate at midday, graze
again in the evening and move uphill in the evening to their campgrounds. In hot weather
sheep spend more time in the shade and change their grazing patterns such that most grazing
occurs in the evening and at night. Environmental complexity also influences behaviour:
Merino sheep kept in relatively flat and featureless terrain spend more time being vigilant
than when the pasture contains more topographical variation.
Sex and age differences in habitat use
In wild sheep, adult rams will forage further from cover and for longer than ewes, are less
vigilant and are more likely to be apart from the social group. Rams have reduced predation
risks, compared with ewes, and are thus able to take advantage of the more abundant forage
out in the open. Juvenile and young sheep are much more vulnerable to predation, are never
outside the social group and always flee when a predator is present. These behaviours are
reflected in the behaviour of domestic sheep, where rams are less fearful than ewes when
tested in social isolation or when there is a potential predator (e.g. man) present, and young
animals express greater fear than older ones.
Responses to seasons and weather patterns
Both wild and domestic sheep reduce dispersal about the home range in winter, migrate to a
smaller region of the home range and decrease activity. Food preferences also show seasonal
shifts reflecting differences in availability and nutrient content of plant species and, in dry
regions, the water content of plants. Sheep and goats make use of shade in hot weather and
shelter in cold and wet weather. Shelter is particularly important for goats, which have a
much lower tolerance for cold, wet weather than sheep, although shelter use by sheep is
affected by fleece length. Sheep, however, are more affected by hot and humid weather than
goats and have a greater need for shade, preferring windy slopes in hot weather.

Sheep and goats are described as seasonally polyoestrous as, except for tropical breeds, they
do not reproduce year-round. Reproductive behaviour is driven by the sexual activity of the
female, which is influenced by day length in temperate animals (shortening days) and by
rainfall in tropical goat breeds. Mating activity occurs in autumn, when males and females
should be in peak condition after abundant food availability through the summer, and lambs
or kids are born in spring when new plant growth occurs. As neither goats nor sheep maintain
harems, mating behaviour is preceded by aggression among males for access to oestrus
females.
Female sexual behaviour
The oestrus cycle of sheep is 17 days on average, and 21 days in goats. Oestrus and ovulation
occur during a 24–48 h period when conception can occur if the ewe or doe mates. Although
the sexual behaviour of the male is more overt, it is the ewe or doe which holds the initiative
for mating. Oestrus behaviour is similar in both sheep and goats although it is more marked
in does. The female becomes restless and vocal, bleating frequently; she may also urinate
often and raise and fan her tail. Does in oestrus have been seen to mount other females, as
cows do, although this is not seen in ewes. Oestrus females will seek out a male and
reproductive success is greater when the female is able to approach the male than when she is
tethered. Ewes and does will show some courtship behaviour towards preferred males,
turning in front of him, rubbing along his chest and flanks, and following him.
Male sexual behaviour
Males seek out oestrus females and assess their reproductive status by sniffing the anogenital
area and urine, followed by flehmen (when the upper lip is curled allowing access of odours
to the vomeronasal organ). Male goats urinate on to their bellies, legs and beards where the
scent may serve to advertise their dominance status. Once an oestrus female has been
detected the male will attend and court the female. Courtship appears to be an important part
of reproductive success, at least in wild sheep, where males that attempt to mount females
without prior courtship are rejected. Courtship typically begins by the male stretching and
twisting the head and neck horizontal to the ground (termed the ‘low-stretch’), making low-
pitched rumbling or grunting vocalizations and licking the ewe at the shoulders or flanks.
The ram or buck then nudges the female with his muzzle and kicks her with his forelegs to
determine if she will stand to be mounted. If the ewe exhibits head-turning (looking back
towards the male) and stands then the male will mount. The male may mount between one
and four times before he ejaculates. Following ejaculation there is a refractory period before
the male will show interest in another or the same female again.
External factors influencing sexual behaviour

In addition to the seasonal factors that control the onset of sexual behaviour described above,
sexual behaviour can be affected by nutrition, social factors and stress. Undernutrition or
overnutrition, particularly of females but also of males if the decrease in food intake is
severe, reduces mating behaviour and undernourished females may not exhibit oestrus
behaviour. Heat stress, or severe climatic conditions, reduces sexual activity as both males
and females seek shade or shelter, and heat stress can reduce sperm quality resulting in
poorer conception. Other forms of stress, such as fear or exposure to rough handling, can also
reduce or delay oestrus in females.
The social environment can both increase and inhibit mating behaviour. The ‘male
effect’, where the presence of the male induces oestrus in females, is known to occur in both
sheep and goats, and the presence of more than one male may both enhance (through
competition) or reduce (through dominance) the mating activity of other males. Early-life
rearing experience is known to have a profound effect on sexual preferences, as demonstrated
by a remarkable experiment where kids and lambs were cross-fostered to ewes and does
(Kendrick et al., 2001b). Rams fostered to does, or bucks to ewes, preferentially mated with
the species that they were raised by, even though they had associated with others of the same
species as themselves as juveniles and adults. Although ewes and does that had been
similarly cross-fostered initially also showed a preference for the species that had raised
them, they changed to preferring their own species when maintained in single-species groups
as adults. In a less dramatic way, the age and previous mating experience of males and
females affect the completeness of their courtship behaviour, the likelihood of females
seeking out males and standing, and thus reproductive success.
14.6 Maternal Behaviour at Birth

Gestation length in both sheep and goats varies with breed, but generally lasts between 145
and 155 days. As with the majority of mammals, and all farmed mammals, in both species’
care of the young is the exclusive responsibility of the mother and there is no paternal
behaviour.
Isolation and shelter-seeking behaviour
In wild sheep and goats, the normally highly gregarious ewe or doe withdraws from the
social group into remote and rugged terrain within the home range during the periparturient
period. This is believed to reduce the risk of predation and to give the mother and young the
opportunity to develop a strong bond uninterrupted by other females. As a ‘hider’ species,
goat mothers in particular may make potentially costly choices about the location of birth
sites, trading off foraging opportunities and social interactions to ensure the protection of
their offspring. The extent to which this still occurs in domestic animals depends on the breed
and the opportunity that females have to express this behaviour (Dwyer and Lawrence,
2005). However, even housed sheep will make some use of lambing cubicles that afford a
degree of isolation. Ewes will also seek shelter at lambing time if it is cold or wet, which can
markedly improve lamb survival.

Behaviour at parturition
Labour in the ewe or doe is characterized by an initial increase in restlessness and pawing,
lip-licking and then straining in both standing and lying positions. Neither ewes nor does
build a nest in which to deliver their young. However, a birth site is selected and offspring
survival is increased with increased duration spent on the birth site. Parturition is short and
the lamb or kid is delivered within 1 to 2 h. Twinning is relatively common in domestic goats
and sheep, and triplet or larger litter sizes can occur in more prolific, highly selected sheep
breeds.
Onset of maternal behaviour and selectivity
Immediately after birth the mother shows intense and focused interest in the newborn and the
amniotic fluids in its coat and on the ground. This is characterized by licking the lamb or kid
(see Fig. 14.3) and emitting frequent low-pitched bleats or ‘rumbles’. These behaviours serve
two main purposes: (i) they help to dry and stimulate the neonate; and (ii) the mother forms
an exclusive and selective attachment to her own offspring. This attachment is initially based
on the smell of the newborn lamb or kid, and mothers that cannot smell are unable to
distinguish their own young from those of others. Mothers that give birth to twin or larger
litters switch their grooming attention from the first lamb or kid to the newly born littermate,
and form a new and additional olfactory attachment to that offspring, although the amount of
grooming attention received by subsequent young is generally less than that given to the first-
born offspring. Selective attachment is formed within an hour or less of birth, particularly in
experienced mothers, and thereafter the mother will actively reject the offspring of other
mothers by butting and will prevent them sucking from her. Maternal recognition of the
young progresses with offspring age and the mother rapidly learns to recognize her own
young from a distance by voice and visual cues.
Fig. 14.3. Maternal licking or grooming of the wet newborn by (a) sheep and (b) goats.
(Photos courtesy of Cathy Dwyer and Marianne Farish.)
The quantity and quality of maternal care shown to the newborn at parturition can be
affected by maternal experience, environmental factors, temperament and breed.
Inexperienced mothers are slower to begin licking their offspring after birth and are more
disturbed by the behaviour of their lamb or kid. They may butt their neonate when it moves,
and move away as it tries to reach the udder, but experience with the behaviour of the
offspring causes these behaviours to disappear and within a few hours of birth the ewe or doe
giving birth for the first time will show maternal behaviour similar to that of more
experienced animals. Expression of maternal behaviour may also be reduced in mothers
underfed in gestation, particularly in late gestation, in ewes of more highly selected breeds
compared with less selected breeds and in ewes of a nervous disposition compared with calm
animals (Dwyer and Lawrence, 2005).
14.7 Maternal Behaviour during Lactation

Suckling behaviour
Licking of the offspring gradually declines over time after birth and within 6 h of birth the
mother no longer licks the lamb or kid, which is now dry. Maintenance of maternal behaviour
relies on behavioural cues from the neonate, particularly sucking interactions, and mothers
whose offspring die lose interest in the body of their young within a few hours. During the
first week of life the mother allows the offspring to access the udder and suck whenever it
approaches. However, she starts to restrict access after this period by walking away when the
lamb or kid tries to suck. During the first month of life the ewe will actively seek out her
lamb when they are separated, and ewe–lamb distances are generally no more than a few
metres. From 4 weeks onwards, responsibility for approaching falls to the lamb, although the
ewe still regulates the ewe–lamb distance. The ewe begins to signal to the lamb when it may
approach by raising the head and bleating, the lamb may then approach the ewe and suck,
although the ewe generally still ends the sucking bout. Lambs that approach when the ewe
has not signalled, and twin lambs that approach without their sibling, are often not allowed to
suck. Ewes or does generally do not defend their young from predators, although ewes may
turn to face and threaten sheepdogs when they have young at foot, and there are reports of
mountain sheep driving off avian predators with their horns (Geist, 1971).
Weaning
In the wild, lambs and kids may not be weaned until their mothers enter the rut when they are
6 months old. Management of domestic sheep and goats generally involves the removal of
offspring at younger ages, around 8–12 weeks, although lambs and kids of dairy sheep and
goats may be weaned within a day or so of birth. Ruminant digestion and gut development
are almost complete at 8 weeks old, so lambs or kids weaned after this age do not need to be
fed milk replacer. However, although able to survive without the mother’s milk supply, the
psychological bond between mother and young is still present, and young will often suckle
for non-nutritive, comfort reasons. Forced weaning disrupts learning and social development
of the young lamb or kid and causes at least short-term stress to both partners.
In natural weaning systems, weaning seems to be related to the maternal milk supply.
Sheep on better diets, because they are able to produce a greater volume of milk for longer,
will wean their lambs later than ewes with poor nutrition, ewes lambing later in the season
and subordinate ewes within a flock. With natural weaning, it is the mother that initiates the
process by increasing her frequency of sucking refusal and she may also start to show some
agonistic behaviours towards the young.

Lambs and kids are precocious and show active behavioural responses immediately after
birth, standing within 30 min of birth and sucking approximately 1 h after birth. Neonates
orient towards large objects and search along the maternal neck and belly to find the udder
(see Fig. 14.4). Tactile and olfactory cues aid the neonates in locating the udder. Sucking
behaviour acts both to provide the offspring with nutrition and facilitates recognition of the
ewe by the lamb or kid (Nowak et al., 1997). Both species are able to recognize their own
mother at close quarters (mainly by olfactory cues) within a few hours of birth and at a
distance (by visual and vocal cues) between 24 and 48 h old. This is important for survival
because, as described above, the ewe or doe will only feed her own offspring, so it is
important that the offspring approach their own dam. Lambs will also associate preferentially
with their twin littermates, rather than non-related lambs, from about 1 week of age and this
association is stronger if twins are of the same sex.
Fig. 14.4. Udder-seeking behaviour by a newly born Merino lamb. The lamb uses tactile cues
from the underside of the ewe to locate the udder, and has a reflex upward ‘bunting’ response
to pressure on the top of the head and face. (Photo courtesy of Raymond Nowak.)
Ungulate offspring can be classed as ‘followers’, which remain close to the mother
immediately after birth, and as ‘hiders’, where the newborn is left concealed for several hours
and the mother may rejoin the social group, returning periodically to feed the young. Lambs
are followers and accompany their mothers closely immediately after birth and when she
rejoins the flock. Kids are usually hiders for a short period of a few days after birth, before
they become followers and are closely attached to their mothers. In the hiding phase, kids lie
flat and immobile until their mother approaches the location where they were left and calls to
them with low-pitched bleats. Despite these different strategies, there are no species
differences in the ability of lambs and kids to distinguish their dams from other ewes or does.
Learning species-specific behaviours
The raised head posture (or ‘head-up’) that the ewe uses to signal to the lamb that it may
approach and suck is similar to the alarm posture used by adult sheep to communicate
potential danger to the rest of the flock. By teaching her lambs to respond to this signal when
young, the ewe passes on to the lamb the need to pay attention to this important signal. The
flight response to potential predators also seems to have a learned component, although this
may be learnt from other adults as well as the mother. By accompanying their mothers about
the home range, lambs and kids also learn about shelter and water sources, and what and
where food sources are. Exposure to novel foods in the presence of their mothers, even at
ages when the offspring is too young to eat the food (see Fig. 14.5), facilitates acceptance of
those foods later by the weaned young.
Fig. 14.5. A 3-day-old Merino lamb investigating lupin grains while its mother eats them.
This behaviour facilitates learning about foods in the young lamb. (Photo courtesy of
Raymond Nowak.)
Play
Locomotor play behaviour (running and jumping) is expressed by young lambs and kids
within a few hours of birth, although there are breed differences in the timing of the onset of
play behaviour. The frequency of playing and complexity of play (e.g. including butting and
mounting behaviours) increase with age to peak around 10–14 days of age. The lambs or kids
form play bands of peers and may spend periods of time distant from their mothers while
playing. Twins preferentially play with one another, and the young will also initiate play with
their mothers. Sex differences in play behaviour emerge, with males engaging in more
butting and mounting play whereas females spend more time in locomotor play. Play is very
infrequent after about 9 weeks of age, although play may still be expressed sporadically in
juvenile and older animals.
14.9 Management and Welfare

Sheep production is very diverse with respect to products produced (meat, milk, wool or dual
purpose), climate in which sheep are farmed (ranging from cold and wet to hot and dry),
cultural issues (e.g. industrial production, family flocks, transhumance) and breed (with over
800 breeds used across the globe). Broadly we can divide sheep production by the degree of
intensification and amount and type of human contacts (EFSA, 2014). Sheep production
systems then range from intensive housing with no pasture access (some dairy sheep, or
finishing of meat sheep), through various patterns of housing for some part of the time, to
extensive or very extensive systems where sheep are never housed and may get all their
nutrition from the environment. Although some welfare issues will be common to all
situations, other issues are more specifically related to management types, such as fear of
humans or exposure to predation. Goat production for meat is generally less common in the
developed world but, along with sheep, is very important to sustain family production in the
developing world, primarily due to the goat’s ability to thrive in harsh environments and
produce meat and milk on a low-quality diet. Goat dairying is, however, becoming
increasingly intensive and, along with intensive sheep dairying, can be associated with
particular welfare issues associated with intensification and selection for high productivity.
Social behaviour and management
The propensity of sheep to group together and follow one another, and their flight reaction, is
exploited in management. Sheep can be trained to follow a shepherd, as may occur in
nomadic management, or are herded and driven by a shepherd and dogs, as occurs
particularly in extensive systems, where animals may be gathered infrequently. As they
respond primarily to visual cues, the movement of a shepherd, or the extension of an arm or
crook, can cause sheep movement in the desired direction.
As the social group is very important to sheep and goats, social isolation is highly
stressful, and is more stressful than capture or restraint within the social group. Isolation
causes agitation (running, rearing against the sides of the pen, escape attempts) and frequent
high-pitched vocalization. Allowing visual contact with other members of the social group
for animals that need to be quarantined, or keeping a few ‘buddy’ animals alongside an
animal needing special attention, can help to reduce the stress associated with isolation for
management reasons.
Housing
The housing needs of sheep and goats has received less attention than other production
animals, perhaps because they are less frequently housed or because they are generally not
subjected to very restrictive confinement. Nevertheless, in countries such as North America
or Scandinavia, sheep may be housed for nearly 6 months of the year due to extreme
temperatures, and in other countries sheep may be housed overnight and let out to graze
during the day. For both sheep and goats, when given a choice between indoor and outdoor
access, both species spent a considerable portion of their daily time budget outdoors, even in
low temperatures. Housing at high stocking density (generally with less than 1 m2 per
animal) leads to an increase in agonistic social behaviours and reduced lying time, suggesting
the welfare of both sheep and goats is impaired when insufficient space is available. There is
evidence that sheep prefer straw bedding over solid or slatted floors, particularly when shorn.
However, goats do not seem to prefer straw, preferring wood floors or mattresses when
temperatures are low (Bøe, 2007).
Neonatal mortality
For both sheep and goats, rates of neonatal mortality are high. Commonly figures of 10–20%
are quoted but variation between farms is high, with figures as diverse as 3–40% pre-
weaning mortality (Dwyer et al., 2016). The causes of mortality are generally well-described
and occur primarily as a result of birth trauma, the formation of a poor bond between ewe
and lamb and infectious disease. A prolonged or difficult birth exposes the neonate to the risk
of hypoxia and pain associated with delivery can impair the behaviour of both mother and
young. The coordinated expression of offspring behaviour is also influenced by prenatal
nutrition (affecting birth weight), which can slow the ability of the neonate to reach the udder
and suckle, and delay the development of maternal recognition abilities in the young.
Maternal undernutrition also reduces the expression of maternal care at birth, impairing the
ability of the mother to recognize her own offspring, as well as reducing udder development,
colostrum production and milk yield. Lambs and kids are therefore more susceptible to
starvation, hypothermia and abandonment when maternal nutrition is limited. This is
particularly prevalent in extensive conditions where maternal nutrition in pregnancy will be
more affected by seasonal variation in food supply and where opportunities for providing
supplementary feeding are restricted. In addition, particularly for goats, their management in
harsh environmental conditions leaves them vulnerable to low nutrition during periods of
drought which may coincide with pregnancy. Vaccination against major diseases of neonatal
small ruminants is available but may not always be used in some countries, and delayed
colostrum intake, due to inadequacies in the behaviour of mother and young, may also leave
the newborn vulnerable to pathogens present in the environment. In intensive systems,
therefore, where animals are housed in close proximity at lambing time, maintaining good
hygiene is essential to prevent the exposure of newborn animals to pathogens.
Human–animal interactions
For sheep in particular, which may be managed in large groups with little contact with man,
human contact and handling can be a significant source of fear and stress. In addition, most
contacts that sheep do have with people are aversive (shearing, dipping, drenching, etc.). As
sheep are very good at learning associations between places, people and negative
experiences, this means that most management actions, even if benign but occurring in a
place where an aversive procedure happened previously, cause sheep stress. This can be
reduced to some extent by quiet, skilled and sympathetic handling and keeping aversive
contacts to a minimum. Domestic goats are often kept in smaller groups with closer
associations with humans and may therefore have a more positive view of human contact,
although this will still be greatly affected by the quality of handling they receive. Negative
handling of both sheep and goats has been shown to increase avoidance and behavioural
stress responses when humans are present.
Dairy animals which have daily forced contacts with humans (in that the animals cannot
avoid contact) are potentially more vulnerable to negative experiences resulting from human
contact. Research in dairy cows has shown that the attitudes of stockworkers to animals
influence their behaviour towards the animals, which in turn affects the behaviour of the
cows and their milk yield (Breuer et al., 2000). Thus positive handling of dairy animals may
increase not only their welfare but also their productivity.
Predation
As small and relatively defenceless animals, extensively managed small ruminants are more
vulnerable to predation than many other production animals. A number of wild carnivores
will prey preferentially on sheep and goats, including foxes, bears, lynx, wolverines, wolves,
coyotes, eagles, mountain lions, baboons, feral pigs and domestic dogs. Neonates and
juveniles are more vulnerable than adults to predation, the risk of which can vary from less
than 1% in countries with few large predators (UK, Australia, New Zealand) to nearly 30% in
areas with a greater predator density (Americas, Africa and parts of Asia). In addition to the
loss of animals, predation can cause long-lasting fear in the rest of flock, and attacked but
surviving animals may suffer significant injuries. Various methods to deter predators have
been attempted, including housing, electric fencing, aversion learning techniques and
trapping, hunting and killing of predators. In several areas, however, the use of guardian
animals (dogs, donkeys and llamas) has been used with some success in sheep production. In
these situations the guardian animal is raised with the flock from a young age and lives in the
flock to protect it from predation. Although this can be very effective at reducing predation
on sheep flocks, the welfare needs of the guardian animal are often not properly considered.
Extensive environments and welfare
Sheep and goats are generally kept in extensive environments where, at least superficially,
the domestic environment is similar to that in which the wild progenitor evolved. Thus, both
species have a greater ability to express natural behaviour patterns than animals kept in more
confined conditions. How well extensive farm environments meet the behavioural needs of
sheep and goats has never been explored in detail; however, they do have greater behavioural
freedom than most other agricultural animals. The extensive environment is not without its
share of welfare problems though, particularly those stemming from climatic variation and
exposure to predation. Thus, extensively managed sheep and goats may be subject to
temperature extremes and periods of low food and water availability. Provision of shade and
shelter, to allow behavioural thermoregulatory mechanisms, is required and access to
supplementary food and water may be needed in some environments. Frequency of
inspection, and diagnosis and treatment for disease, may also be less frequent in extensive
systems.
Intensive environments and welfare
Intensive environments can provide protection from predation and environmental extremes,
and has the capacity to deliver better nutrition to housed sheep and goats. Treatment for
disease and injury can also be provided more rapidly and there is greater potential for
individualized care. However, the close spatial proximity of a large number of animals may
also increase the spread of disease. Intensive systems provide a greater degree of behavioural
restriction than extensive systems, and the opportunities for choice and exposure to variety
are reduced.
Both intensive and extensive systems have welfare advantages and disadvantages for
sheep and goat production. In all systems the humans responsible for the animals’ care are
crucial to the welfare of those animals, although in extensive systems animals have greater
self-sufficiency and are less directly affected by variations in their carer’s behaviour. For
animals in confinement, which have frequent human contact, the quality of that care can have
a very significant impact on animal welfare.
References
Bøe, K.E. (2007) Flooring preferences of domestic goats at moderate and low ambient temperature. Applied Animal
Breuer, K., Hemsworth, P.H., Barnett, J.L., Matthews, L.R. and Coleman, G.J. (2000) Behavioural responses to humans and
the productivity of commercial dairy cows. Applied Animal Behaviour Science 66, 273–288.
Clutton-Brock, J. (1999) The Natural History of Domesticated Mammals. Cambridge University Press and British Museum
(Natural History), Cambridge.
Dwyer, C.M. (2004) How has the risk of predation shaped the behavioural responses of sheep to fear and distress? Animal
Welfare 13, 269–281.
Dwyer, C.M. and Lawrence, A.B. (2005) A review of the behavioural and physiological adaptations of extensively managed
breeds of sheep that favour lamb survival. Applied Animal Behaviour Science 92, 235–260.
Dwyer, C.M., Conington, J., Corbière, F., Holmøy, I.H., Muri, K., Rooke, J., Vipond, J. and Gauthier, J.-M. (2016)
Improving neonatal survival in small ruminants: science into practice. Animal 10, 449–459.
Edwards, G.R., Newman, J.A., Parsons, A.J. and Krebs, J.R. (1996) The use of spatial memory by grazing animals to locate
food patches in spatially heterogeneous environments: an example with sheep. Applied Animal Behaviour Science 50,
147–160.
EFSA (2014) Scientific opinion on the welfare risks related to the farming of sheep for wool, meat, and milk production.
EFSA Journal 12, 3933.
Forbes, J.M. and Provenza, F.D. (2000) Integration of learning and metabolic signals into a theory of dietary choice and food
intake. In: Cronjé, P.B. (ed.) Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction. CAB
International, Wallingford, UK, pp. 3–20.
Geist, V. (1971) Mountain Sheep: A Study in Behaviour and Evolution. University of Chicago Press, Chicago, Illinois and
London.
Kendrick, K.M., da Costa, A.P., Hinton, M.R., Leigh, A.E. and Peirce, J.W. (2001a) Sheep don’t forget a face. Nature 414,
165–166.
Kendrick, K.M., Haupt, M.A., Hinton, M.R., Broad, K.D. and Skinner, J.D. (2001b) Sex differences in the influence of
mothers on the sociosexual preferences of their offspring. Hormones and Behavior 40, 322–338.
Nowak, R., Murphy, T.M., Lindsay, D.R., Alster, P., Andersson, R. and Uvnäs-Moberg, K. (1997) Development of a
preferential relationship with the mother by the newborn lamb: importance of the sucking activity. Physiology &
Behavior 62, 681–688.
Ryder, M.L. (1984) Sheep. In: Mason, I.L. (ed.) Evolution of Domesticated Animals. Longman, London and New York, pp.
63–85.
Schaller, G.B. (1977) Mountain Monarchs: Wild Sheep and Goats of the Himalaya. University of Chicago Press, Chicago,
Illinois and London.
Terrill, C.E. and Slee, J. (1991) Breed differences in adaptation of sheep. In: Majala, K. (ed.) Genetic Resources of Pig,
Sheep and Goat. World Animal Series B8. Elsevier Science Publishers, Amsterdam, pp. 195–233.
15 Behaviour of Pigs
M. Špinka
15.1 Origins of Pigs

All domestic pigs are descendants of the wild boar (Sus scrofa), a geographically widely
distributed artiodactyl species belonging to the family Suidae. Wild boar is a versatile species
capable of adjusting its way of life to different natural conditions. We will review various
aspects of its natural behaviour at the beginning of each section of this chapter as a basis for
understanding the behaviour of domestic pigs.
During the period when pigs started to be domesticated, the wild boar was living from
northern Japan in the north-east and New Guinea in the south-east, and throughout China,
tropical South Asia, India and the Middle East, and up to the Atlantic coast of Europe and
North Africa in the west. A marked east–west split occurred among the wild boar 1 million
years ago, with ‘European’ populations to the west of Iran differing from those of Asia in
morphology, genetics and probably behaviour.
Domestication of pigs started about 9000 to 10,000 years ago, i.e. shortly after humans
switched from hunting to husbandry in the so-called neolithic agricultural transition. Pigs
were domesticated from the wild boar at two locations: in China and in eastern Turkey
(Groenen, 2016). The Chinese, Indian and European traditional pig breeds are still
genetically closer to their regional wild boar counterparts than to each other. There was an
introgression of Chinese pig breeds into the European stock in the late 18th to early 19th
century. The major modern breeds arose from this mixture, some of them as late as the mid-
20th century, and thus the modern European pigs contain a 20–35% Asian admixture. The
interbreeding between domestic stock and the wild boar has continued in many areas
throughout the history and the two forms can, of course, produce fertile offspring even today.
Thus, substantial genetic variety exists within the species of S. scrofa in its wild, domestic
and feral forms.
The domestication process exposed pigs to a different set of selection pressures than
those acting upon their wild ancestors. The need effectively to escape and/or fight off
predators and actively to seek diverse food sources has subsided, while the selection for large
nutrient intake, efficient food utilization, fast growth and high fertility has been strengthened.
However, the direction of selection has not been the same throughout the history of
domestication. For instance, European domestic pigs were smaller than wild boar until the
Middle Ages, while the proportion of body fat in many breeds was much larger. The recent
phase of pig domestication during the last 100 years or so encompasses intense, large-scale
breeding programmes for genotypes capable of turning formula-based high-energy foods into
lean meat quickly and efficiently. As a consequence, current production breeds of domestic
pigs are much larger (including, for instance, a larger number of thoracic and lumbar
vertebrae) and heavier than the wild boar, and are also capable of a higher energy turnover
and of year-round reproduction.
As far as we know, no single behavioural pattern has disappeared from the wild boar
repertoire during the domestication process (Gustafsson et al., 1999). Therefore, behaviour of
wild boar, feral pigs and domestic pigs kept in natural conditions informs us about the
behavioural needs of pigs and also helps us to understand the structure and the function of
behavioural patterns whose purpose is difficult to see within the barren conditions of modern
intensive indoor systems. Nevertheless, the behaviour of pigs has changed quantitatively due
to domestication – domestic pigs are less active, less aggressive and wary towards potential
predators (including humans), and less likely and less able actively to seek variety in their
diet or food sources. The intensive selection for production traits like large litter size, fast
growth and lean body composition has been modifying genetic predispositions for pig
behavioural traits during the last decades. During the last 15 years, studies are accumulating
on the genetic basis of pig behaviour, on the genetic interactions between behavioural,
physiological and production traits, as well as on the genotype–environment interactions
(Desire et al., 2015b). This, it is hoped, may lead to more sustainable breeding programmes
(Rydhmer et al., 2014).
For urgent practical reasons, it is very important to study the behaviour of pigs that have
been allowed to return to more natural living conditions. During the history of their lives with
humans, domesticated pigs (or imported wild boar) escaped or were introduced to nature in
areas previously unoccupied by wild swine. After these invasions, pigs almost invariably
increased in number and spread across available land as far as it provided cover and surface
water sources. These sites inhabited by feral pigs include many small and large Pacific
islands, much of Australia, the south of the USA and regions in South America. Through the
versatility of their feeding and activity habits and due to high fecundity, feral pigs often
became serious pests because of destruction of vegetation, killing of native species, eating of
crops, invading suburbs of cities, spreading of disease (including to domestic swine) and
even preying on young livestock. Control measures include hunting and even large-scale
poisoning campaigns and yet often have produced only temporary effects. Sometimes, an
annual culling rate of 70% is needed just to keep the populations at a stable level.
Nevertheless, eradication has been achieved in several isolated areas, especially islands.
During the last decades, the wild boar is also recolonizing parts of cultural landscape in
Europe where it was missing for several centuries.

Wild and feral pigs most often live in matrilineal groups of a few kin-related females and
their offspring from the previous year (Podgorski et al., 2014). Adult males may be
associated with these herds, but more often live alone or in bachelor groups, joining the
females only for breeding. However, pig groups are dynamic due to a rapid turnover of
individuals and may even merge into larger associations if there are large, concentrated food
resources. The home ranges of groups overlap, with no indication of between-group
territoriality. Within the groups dominance relationships are established yet this does not
preclude effective cooperation, for instance during risky foraging (Focardi et al., 2015).
In domestic pigs, dominance relationships are less stable than in cattle. Therefore,
fighting back, position reversals, harassing of subordinate animals and renewed fighting after
a temporary separation can occur, even within stable groups. Nevertheless, dominance-
related aggression in juvenile pigs and sows most often arises in two situations: (i) when food
is available in limited space and limited time; and (ii) when groups of pigs are mixed or alien
individuals are added to a group. During mixing-induced aggression, a few animals, usually
those to become dominant, account for most of the fighting (Verdon et al., 2016). If pigs are
mixed into groups larger than about 20 animals the amount of aggression per animal is lower,
as many pigs refrain altogether from fighting, while a minority of animals may actually fight
more intensely than in small groups.
Typical elements of pig fighting include sidewise head-knocking, pushing (shovelling),
levering and biting all over the opponent’s body, especially around the neck and ear region.
Biting results in skin lesions whose numbers in the anterior region of the body are indicative
of the reciprocal aggression level, while lesions to the centre and rear of the body are
associated with received non-reciprocal aggression (Turner et al., 2009). The defensive
manoeuvres involve turning the fore body away from the attacking pig, clinching in
‘antiparallel’ position and, above all, fleeing. The attacking animal usually does not pursue its
fleeing opponent for more than about 3 m. Juvenile pigs mostly fight silently, while attacks
by dominant sows are countered by high-pitched protest squealing of the attacked sow.
Attacks at a feeding place are usually brief, since the dominant animal resumes feeding once
the subordinate retreats. Nevertheless, victims of repeated food-related attacks may become
unwilling to attend the feeding place. Fights during establishment of dominance can escalate
into pitched battles or in the more dominant animal attacking and pursuing the other
repeatedly if the latter has no possibility of escape. During the initial phases of an escalated
fight both pigs attack, often circling around each other in attempts to attain a biting position.
As soon as one of the animals starts using predominantly defensive manoeuvres, the outcome
of the confrontation is decided (Rushen and Pajor, 1987; see Fig. 15.1). More intense acute
aggression immediately after mixing may result in less chronic aggression later (Desire et al.,
2015a).
Fig. 15.1. Five different positions during fighting between unacquainted young pigs. The
sections in the pie chart denote the time spent in each position, the bars depict the biting rate
(bites per time unit). In positions a–c, both pigs are attacking and the biting rate is high.
However, these extremely demanding mutual attacks take up only about one-third of the
time. Most of the time, either one pig is defending (the left pig in position d), or both pigs are
defensive (position e). (From Rushen and Pajor, 1987.)
Aggression in group-housed pigs decreases welfare and production in a number of ways:

(i) biting results in numerous, albeit superficial, skin lesions; (ii) aggression during post-
weaning mixing contributes to other stressful aspects of weaning; and (iii) access to food or
resting places is often reduced due to the fear of the defeated animals. The frequency and
intensity of fighting can be ameliorated by mixing pigs as little as possible, designing pens so
that the subordinate animals can flee and hide, and designing feeders and feeding regimes so
that dominant animals cannot harass others. Both sow and weaner piglet aggressiveness at
mixing is mildly to highly heritable (Turner et al., 2009). This opens prospects for selecting
against aggressiveness, especially as there seems to be a minority of extremely aggressive
animals that contribute disproportionately to the overall level of fighting.

Pigs do not engage in allogrooming, and there are no reports of strong individual
affiliations, although this might be due to a lack of research focus on this question. On the
other hand, pigs have a strong tendency for coordination and synchronization of behaviour in
space and time. For instance, one alarm bark by a pig makes the whole group – or even all
pigs in a room – freeze and attend. Also, exploratory behaviour has a strong tendency to be
synchronized, especially in smaller groups of up to six pigs, and therefore enrichment
materials need to be provided in a way that allows many pigs to engage with them at the
same time (Zwicker et al., 2015).
Humans are important components of pigs’ social environment and have been so over the
entire domestication period. Depending on their experience with human behaviour, pigs
develop either trusting or more or less fearful attitudes towards people. Pigs can distinguish
between a familiar and an unfamiliar person by relying on various cues, including the colour
of their clothing. However, in many situations pigs tend to generalize their experience (either
negative or positive) with specific persons to humans at large (Brajon et al., 2015).
Inconsistent treatment, i.e. alternation between rough and kind handling, is perceived
negatively by pigs as consistently rough handling. Thus, even limited bad experience with
humans can influence pigs’ attitudes towards people as a class for a long time. Therefore, it is
important for caretakers deal with pigs calmly, consistently and as positively as possible.
Educating personnel working with pigs on proper handling decreases fearfulness and is
reflected in better performance of the stock. Pigs are more willing to approach a squatting or
sitting person than one standing but, once the first pig in a group begins contact with the
human, others quickly follow. As people are insensitive to odour signalling, and visual
signals are not employed by pigs, acoustic communication is the domain through which pigs
actively engage with humans. Recent use of pigs as pets documents how an individualized
human–pig relationship can develop. On the farm level, tests have been developed to
quantitatively assess the human–pig relationship (Waiblinger et al., 2006).
15.3 Perception and Communication

The scent sensitivity of pigs matches that of dogs. Pigs use their acute sense of smell in
foraging, gathering social information and mutual communication. Wild boars locate below-
ground and hidden food by smell; they can scent humans over hundreds of metres and avoid
traps based on olfactory cues. In the social realm, scent is for pigs the primary way of
distinguishing familiar from unfamiliar pigs, but also individuals among known pigs; pigs
can recognize and remember at least 30 individuals. Other types of information extracted
from the scent of the individual are its sex, reproductive status and probably dominance
status.
Pigs like sweet taste, especially sucrose, but the sets of natural and artificial compounds
that taste sweet to humans and to pigs overlap only partially. Bitter-tasting liquids (e.g.
quinine) are aversive for pigs.
Hearing is well developed in pigs. Lower hearing thresholds are somewhat higher in pigs
than in humans for most frequencies (i.e. pigs do not hear as faint sounds as we do). Pigs’
ability to locate sound sources only 5° apart is almost as good as in humans and much better
than in other domestic ungulates. Unlike humans, pigs do hear some ultrasound, up to about
45 kHz.
Despite their small eyes pigs have relatively good vision, although they cannot see as
accurately as humans or cattle. Similar to other ungulates, pigs have just a dichromatic
vision, making them capable of recognizing blue colour but not red from green. Pigs have a
blind angle of 50–100° at the rear, thus having a wider field of vision than humans but
narrower than cattle.
The pig snout disc is a powerful and yet sensitive tactile organ used in digging up and
examining food and exploring features of the environment.
Pigs use primarily the olfactory and acoustic channels for communication. In the
olfactory realm, adult males actively advertise their status by producing chemical
compounds, two of which are the source of the unpleasant taint sometimes present in pork
from non-castrated postpubertal males: androstenone, which gives the meat a smell of urine,
and skatol, smelling of faeces. Androstenone is produced in the testes and is present in saliva,
urine and the preputial glands. This compound triggers the standing reflex in oestral sows
and can advance the puberty of gilts. The olfactory system of the female is five times more
sensitive to androstenone than the male system. Pigs also communicate danger by using
smell; a stressed pig releases alarm substances in its urine that can be detected by other
individuals as warning signals about a danger.
Pigs use a rich repertoire of vocal expressions. The pig vocal ethogram could be
described as a set of highly flexible call types connected through an acoustic continuum
(Tallet et al., 2013). Pigs almost always emit calls in time-structured series that can carry rich
information about the identity of the sender, his/her location, body size, condition, motivation
and emotion. Several areas of vocal signalling have been investigated in some depth. One
such area comprises the sows’ grunting during nursing episodes, which will be described in
Section 15.9 below. Another example is the emergency screams that piglets emit during, for
example, painful castration, when trapped under the sow’s body or when fighting for a teat.
These vocalizations reflect the urgency of the situation, thus inciting the mother to intervene,
with longer, louder and higher screams signalling higher danger for the pig. Humans can
therefore also use these calls to monitor the pigs’ situation and assess their welfare. Because
of the need to be heard, domestic pigs in intensive units are quite noisy. Distress-induced
vocalizations are not only signs of compromised welfare, but, in themselves, further
negatively affect the quality of the environment both for the animals and their carers.
Tactile communication is used by piglets during nursing when they trigger milk ejection
and communicate their hunger level to the sow through the length and intensity of teat
massaging.
Wild boar use visual signals like bristle-rising, ear position, tail movement and arching of
the back. In the domestic pig, this communication has substantially diminished, probably due
to morphological changes that prevent the signals being clearly displayed.
15.4 Cognition, Emotions and Personalities

The pig’s ability to learn from experience, memorize and combine new memories with
previously available information is outstanding. This holds especially for food-related and
social tasks, and thus corresponds to the fact that pigs are generalists in terms of their diet
and social animals living in highly flexible social groupings. As for food acquisition, pigs can
learn various associations and operational tasks, including being called to the food by
individually specific acoustic signals. Pigs can deduct information about food location or
preference from observing other pigs. Pigs are also able to choose options based not only on
immediate rewards or punishments, but also on delayed consequences. For instance, zoo-kept
wild boars have been observed not to consume sand-soiled apple halves immediately, but
carry them to water and wash them before eating. Furthermore, pigs possess episodic
memory, i.e. they remember what happened where and in which context. In the social realm,
pigs can not only discriminate between familiar and unfamiliar pigs, but also between
individual pigs within a group or litter, not only if the ‘whole’ individual is present, but also
on the basis of unimodal ‘signatures’ (i.e. just the scent or the calls). Nevertheless, pigs do
not always employ their advanced cognitive abilities, but rely rather on simple cues and rules
of thumb in most situations.
Pigs experience a rich and dynamic affective life. Specific negative emotional states can
be induced in pigs by bodily conditions such as pain, hunger and sickness as well as by
psychological causes such as anxiety, fear and frustration. On the other hand, positive
emotions are in action during fitness-promoting situations such as suckling and play in
piglets, as well as successful engagement with the environment in adult pigs. Pig emotions in
various situations have been usefully described along two main dimensions, namely valence
(ranging from positive to negative) and arousal (ranging from high to low). Long-term
housing conditions (e.g. barren versus enriched environment) influence the longer-lasting
affective states (‘moods’) of pigs. Pigs housed in worse conditions become more pessimistic,
i.e. they tend to appraise ambiguous stimuli more negatively than pigs living in better
conditions (Douglas et al., 2012). Furthermore, emotions have an important social dimension
in pigs as they are readily perceived by other pigs and can spread contagiously in groups.
Therefore an intense affective state (positive or negative) in one animal can have behavioural
and welfare implications for the whole group of pigs (Špinka, 2012).
In spite of their tendency for synchronization and social facilitation, pigs differ a lot
individually in their behaviour and their affective inclinations. Some of these differences
could be described along the dimension of behaving passively versus actively in challenging
situations, during aggression and in exploration. However, this dimension explains only a
relatively small part of pig inter-individual behavioural variation (Melotti et al., 2011). Most
probably, the ‘personalities’ of pigs consist of several dimensions, similarly in regard to what
has been learnt about humans and many other mammalian species.

Pigs are the least efficient, but most flexible, foragers among ungulates. In nature, wild boars
live on a rich and varied diet. About 10% of this consists of animal food, including soil-living
invertebrates, carcasses and small ground-dwelling vertebrates, but also larger, less mobile,
vertebrate prey. In some areas of Australia, feral pigs kill and eat up to 40% of newborn
lambs. In terms of plant matter, roots, young shoots, berries and as well as more bulky food
such as green grasses, herb leaves and tree bark are consumed as pigs are able to ferment
some fibre types in their hindgut with the help of internal microflora. In the longer term,
however, pigs’ growth and reproduction depend on a large proportion of high-energy food in
their diet. This can consist of ‘mast’ (acorn, beechnut), other starch- or sugar-rich food (such
as tree-fern trunks in Hawaii) or grain crops. Accordingly, pigs strongly prefer sugary and
starchy food over more fibrous diets. The current increase and expansion in wild boar and
feral pig populations in many areas of the world are fuelled by crop consumption. Due to
their propensity to trample crops as well as consume them, pigs can cause extensive
agricultural damage. Beside energy, pigs also need protein in their diet, as they cannot
synthesize ten essential amino acids. Extensive rooting of soils by wild and feral pigs is often
due to their foraging for earthworms and other sources of protein. Free-ranging pigs actively
move and forage for about half of their time, mostly in the mornings and evenings, although
they are able to adjust their daily feeding regime to local conditions.
The modern pig breeds have been intensely selected for fast growth rate and for the
ability of sows to deliver lot of milk to the litters of up to 14 suckling piglets. For these
enormous energy outputs pigs need a correspondingly massive food intake, and thus a huge
appetite is now genetically encoded in current pig lines. Food mostly comes as a pre-mixed,
energy-rich formula, needing no searching, foraging, extraction or mastication efforts.
However, pigs need to engage in a certain amount of exploratory, foraging and feeding
activity that cannot be satisfied when no bedding or other rooting and chewing material is
available. This is most urgent in pregnant gilts and sows that neither grow fast nor lactate. In
order to prevent them from becoming obese, these animals are kept on just 60% of their ad
libitum intake. Although their nutritional needs are fully covered, these animals feel
permanently hungry and attempt to resolve the problem through repetitive exploratory and
foraging efforts that can lead to the development of stereotypies, especially in the form of
bar-biting. In order to prevent stereotypies and to keep pigs occupied, the best-suited
enrichment includes materials that are complex, can be manipulated or even destroyed, and
contain small pieces of edible matter (Studnitz et al., 2007). Provision of a diet containing a
high amount of fibre, particularly soluble fermentable fibre or resistant starch, increases
short-term satiety and gives the sows an energy supply that is more equally distributed in
time due to hindgut fermentation of the fibre into short-chain fatty acids.
Pigs are flexible in the time structure of their feeding behaviour and, when accessing the
feeder is difficult, they will use fewer but longer visits to achieve a similar feeding time. On
the other hand, pigs have a high tendency to feed in synchrony. Competition at peak feeding
times may reduce feed intake in group-housed growing pigs if the feeder space is limited.
Low-ranking group-housed sows may achieve only 50–80% of the intake of their higher-
ranking group mates, which may negatively affect their reproduction through both lowered
conception rate and reduced litter sizes. Full-body feeding stalls and electronic feeding
stations may provide protection at feeding, but aggression can nevertheless occur for access
to the stalls and during queuing.
15.6 Coping with the Environment

Wild boars spend about the same amount of time resting and being active. The active period
may be concentrated mainly in daylight hours (in winter) or in the mornings and evenings (in
summer), or even at night (when human disturbance or hunting pressure is high). Resting
occurs in well-covered places in ‘daily beds’.
In domestic pigs, resting takes up to 85% of their time. Different phases of sleep
(drowsiness, slow-wave sleep and rapid eye movement sleep) occupy about half of the total
time, with the greatest proportion of sleep occurring during darkness. Pigs prefer to rest on
soft, dry, clean surfaces.
Pigs are naturally very clean animals. Piglets as young as 5 days old already defecate and
urinate in places remote from their lying areas. If that is not possible and the pen is soiled,
their lying time will be reduced. If the pen design allows a functional division of the
available area, adult pigs use specific dunging areas for elimination. Before lying down, pigs
normally check the cleanliness of the bedding and never lie in soiled areas, if they can avoid
them.
For domestic pigs, the thermoneutral zone (i.e. that range of temperatures they can
withstand without elevating their metabolic rate) decreases with age. Small piglets need to be
kept warm, but they can secure this requirement by huddling and by digging into the bedding
substrate (if such is available). In indoor housing, growing and adult pigs often have
problems with overheating, but rarely with being cold. An isolated 90 kg pig on a concrete,
slatted floor has a thermoneutral zone between 17°C and 26°C or, without a wallowing
opportunity, between 17°C and 23°C. In adult pigs, activity decreases already when the
temperature rises above 24°C, followed by reduced food intake and growth in even higher
temperatures. Pigs do not sweat and therefore rely on behaviour to keep body temperature
down in excessively warm environments. They cool down by lying flat on colder substrates,
and especially by wallowing (Bracke and Spoolder, 2011; see Fig. 15.2), i.e. by moistening
the body surface in pools of water, mud or, if nothing else is available, the faeces/urine slurry.
High respiration rate is a last-resort mechanism to dissipate heat but in itself is energy
consuming. Pigs kept outdoors over the summer must be provided with shade; otherwise they
may be overheated and become sunburnt. Thermal comfort can be assessed by resting
behaviour: if pigs huddle intensely, they are cold; if they lie with sporadic body contact,
temperature is agreeable; if all pigs lie flat on their sides and fully separated, the environment
is too hot.
Fig. 15.2. Given the opportunity, pigs create wallowing spots where they cool down in warm
temperatures.
Pigs lack allogrooming (i.e. the propensity to lick each others’ body) and self-grooming
is also rare. Their principal way of keeping their skin clean and free of parasites is by
scratching against hard objects.
In terms of coping with the environment, the welfare of pigs is best secured if the
housing method allows the animals to divide the space into areas with different functions: a
dry, soft place away from the main ‘traffic’ for resting; feeding places with enough space for
unharassed food intake; a cooler part for elimination; and an area with rooting material for
foraging. Additionally, the area should be large enough for pigs to escape attacks from
dominant animals. This ideal is far from reality in most current husbandry systems, but
progress towards it can be made by gradually establishing minimal standards that will
include elements of these requirements.
Pigs are curious animals, prepared to seize an opportunity to explore new objects and
situations; this drive is stronger the more barren the environment in which they live. Housing
pigs in environments that allow them to acquire new information enhances their welfare,
while lack of such opportunity may even lead to underdevelopment of mental capacity (see
Fig. 15.3). Pig exploratory behaviour is closely related to appetitive phases of foraging. In
barren environments, lack of suitable substrate for rooting and chewing leads to increased
risks of ear-chewing and tail-biting while provision of suitable materials such as horizontally
suspended pieces of fresh wood reduces the damaging behaviours.
Fig. 15.3. Duration of interest in a new object (a tyre) by pigs kept in four different
environments. Pigs in barren indoor environments do not have their need to explore satisfied
and therefore show a rebound behaviour by exploration on a novel object. (From Stolba and
Wood-Gush, 1981.)
When pigs become ill, they show a distinct set of behaviours, including decreased food
intake, increased sleepiness and lying in the ventral position, lack of activity and reactivity,
shivering and vomiting. These changes support the parallel physiological and immune
responses in fighting the cause of disease and reinstatement of homeostasis.
15.7 Mating Behaviour

In nature, wild boar females usually breed once a year, usually during early winter in the
temperate regions of the northern hemisphere. Feral pigs can breed twice a year and spread
the reproduction more evenly over the seasons if they have access to high-energy foods. With
up to four piglets per litter surviving until weaning, wild and feral pig population can quickly
recover after trapping, poisoning or hunting actions aimed at reduction of the population.
Oestrus is highly synchronized in wild boar females within a group, but not necessarily
between groups. Domestic sows usually give birth to about two litters per year, with very
little seasonal variation. In intensive systems, where piglets are weaned between 2 and 6
weeks of age, sows usually come into oestrus about 5 days after being separated from their
young. The so-called standing oestrus (the time during which the sow will stand still in
response to boar stimulation) lasts 1–3 days and, about two-thirds of the way into this period,
ovulation occurs. As the fertility of shed eggs declines within 4–8 h, and the fertile life of
boar sperm in the reproductive tract is longer (about 24 h), the optimal time for mating or
insemination is sometime before ovulation, i.e. about 12 h into the first day of oestrus; this
can then be repeated every 12 h. If a sow fails to conceive, new oestrus periods follow every
18–24 days. When sows are allowed to suckle their piglets for many weeks in the presence of
a boar (e.g. in the so-called family systems), they readily show lactational oestrus and
therefore can become pregnant while still nursing.
About 2 days before the start of standing oestrus, sows begin to show proceptive
behaviour, i.e. active displays of the upcoming receptivity (Pedersen, 2007). This proceptive
behaviour includes spending time near the boar or his pen, nosing the flanks of other sows
and mounting. Dominant sows can mount subordinate females, but mounting ‘up the
hierarchy’ is actively repelled until the period of standing oestrus of the dominant animal.
Proceptive behaviours peak in frequency when the receptive behaviour starts. Receptive
behaviour consists of the sow reacting to the boar’s smell and grunting calls by stopping
movement, possibly by squealing and urinating, and, if the boar mounts, by standing still
under his weight. This standing reaction is also used for manual checking of whether a sow is
in oestrus. The test is much more reliable in the presence of the boar, or at least with a
simultaneous application of the boar pheromone.
In the wild, male pigs mark their presence to females by leaving saliva marks on trunks
and possibly through more frequent wallowing. The sexual behaviour of the domestic male
pig starts with courting females that behave proceptively (Hemsworth and Tilbrook, 2007).
During courtship, boars approach females, utter short series of characteristic grunts, champ
their jaws while salivating and may urinate rhythmically. If the female stands firm, the boar
continues with sniffing her head and anogenital region, nosing or nudging her flanks, and
eventually mounting. Ejaculation lasts many minutes. A boar can copulate up to eight times
over a period of several hours. A sow may mate with more than one boar during one oestrus
period and thus litters of mixed paternity may occur.
Boars differ individually in both their sexual motivation (willingness to mate) and their
sexual dexterity (ability to achieve intromission and ejaculation). Isolation rearing, high air
temperature and presence of dominant males all decrease sexual performance, while
observation of other males mating may enhance it.
15.8 Periparturient Behaviour

The potential reproductive rate of the wild boar is the highest among all ungulates. The high
fecundity level has been taken to the extreme during domestication and modern breeding.
This is inherently connected with relatively little care about individual piglets and therefore
high mortality risks for the young, especially during the period around birth.
Free-ranging sows separate themselves from their group 2–3 days before parturition and
seek a partly sheltered place in which to build a nest (Wischner et al., 2009). Wild boar
females usually nest in dense cover, in warm places and near water. If temperature is low,
free-ranging pigs build large, well-insulating nests, whereas in hot temperatures, nesting is
reduced to bedded, shallow hollows in the ground. Nest building starts about 15 h before
parturition possibly under the influence of increased prostaglandin levels (see also Chapter
3). First the sow walks around, sniffs, digs and roots at the ground to produce a bowl that will
retain the piglets. The sow thereafter switches to carrying and arranging of the available nest
material. In nature, the building starts with branches and then proceeds with soft items such
as grass or fern. The more suitable is the material, the sooner the nest building terminates,
possibly under the influence of increasing oxytocin levels. If no suitable material is available,
the nest-building attempts continue after parturition has begun.
During the parturition itself and immediately following, the sow normally remains lying
passively, with her teats exposed, under the influence of endorphins. Piglets are born at
intervals of about 20 min; they immediately struggle to the udder and wander from teat to
teat, sampling the colostrum which is rich in energy and immunoglobulins. Early colostrum
intake is vital for newborn piglets since they have limited body reserves and their immune
system is not yet functional at birth. In large litters, the last-born piglets may suffer from
reduced colostrum intake because most of it has been consumed by their earlier-born
siblings. The sow sniffs the piglets as they pass her snout and learns to recognize their
identity within the first day, thereafter rejecting any other piglets as alien. Therefore, cross-
fostering of piglets between litters should be done as soon as possible after parturition.
Unlike many other mammals, pig mothers do not lick the newborn progeny.
Within hours, each piglet starts to develop an attachment to a specific teat. The piglet
then defends its teat vigorously against its siblings using sharp canines and incisors capable
of inflicting slashes on the opponents’ faces. In large litters, the difficulty to secure a teat
against competition from stronger siblings may decrease the growth and survival chances of
weaker piglets. In less numerous litters, some piglets may use two teats and this creates the
opportunity for later incursion of alien piglets in group-housing systems.
During parturition and the first 2–3 days thereafter, the average cumulative piglet
mortality is at about 15% even in well-managed herds (Kirkden et al., 2013). The two main
causes of death are undernutrition and crushing by the sow. Very often these two factors
combine, as small and hungry piglets stay closer to the sow and thus are more frequently
crushed. Crushings occur when a sow lies down or rolls over. When a piglet gets trapped
under the sow, it starts squealing. If the sow reacts to the squealing and wriggling of the
piglet by standing up within a minute the piglet usually survives, but longer crushing periods
are mostly fatal. Therefore, calmness during parturition, the early ability of the sow to satiate
piglets with colostrum/milk, her careful changes in position and a strong response to piglet
screams are sow characteristics that enhance piglet survival. Plans to select for these
behavioural traits are being developed despite the fact that their heritabilities are below 0.1
(Stratz et al., 2016).
15.9 Behaviour during Lactation

Immediately after birth milk is available almost continuously, but quickly a pattern emerges
of synchronized suckling every 50 min or so (Špinka and Illmann, 2015). Each nursing
episode begins with 1–2 min of teat massage by the piglets. This massaging triggers oxytocin
release from the sow’s pituitary gland, this hormone causing milk ejection that lasts only 20
s. This is the only period when milk is available and it can be recognized by fast, rhythmical
sucking movements by all piglets in the litter (see Fig. 15.4). Thereafter, piglets switch back
to teat massaging and continue with it for several minutes. Nursing is accompanied by
rhythmical grunting of the sow that serves to call the piglets to the udder and to announce,
through an increase in the grunting rate, the upcoming milk ejection. Sows in auditory
contact also use grunting to synchronize their nursings closely in time. About 10–35% of all
nursing episodes do not contain milk ejection and piglets acquire no milk during these. When
lactating sows are kept in groups, suckling of piglets that do not belong to the same litter is
quite common, but does not present a major problem in stabilized groups.
Fig. 15.4. Suckling piglets ingest milk synchronously during the brief periods of milk
ejection. The approaching milk ejection is announced vocally by the sow and therefore
piglets listen attentively in order not to miss the milk flow.
During the first 2 weeks of lactation, the initiative in suckling gradually shifts from the
sow to the piglets (Špinka and Illmann, 2015). Initially, the sow starts almost all nursings by
exposing her udder in a lateral lying position and also allows the piglets to massage the
mammary glands for many minutes after each milk ejection. Gradually, the sow leaves it
more often to the piglets to initiate the nursings by starting to nudge her udder and thus
stimulating her to assume the nursing position. On the other hand, the sow gradually restricts
the duration of post-ejection teat massage and may later even start nursing in the standing
position. In this way both the frequency of nursing and opportunities for the piglets to
massage the udder decrease. There seem to be two behavioural feedback mechanisms
between piglets’ milk intake and milk production by the sow: (i) hungry piglets attempt to
start nursing episodes at shorter intervals, thereby increasing the milk output per time; and
(ii) hungry piglets engage in more intense and longer final teat massage that may, in turn,
increase the milk output on following nursings.
In free-ranging pigs, sows stay around the nest (and the piglets mostly within the nest)
for a couple of days after parturition, whereafter they join the group and mix with its
members. On the majority of intensive pig farms, lactating sows are confined in a narrow
metal crate. Although the crate may somewhat reduce the mortality of piglets by crushing, it
hinders the nest building and, later in lactation, prevents the sow from leaving the piglets and
thus escaping their sucking attempts. Sows housed in individual pens during parturition and
lactation generally enjoy better welfare. Outdoor farrowing huts are a widely used option in
some countries. Group housing of lactating sows can work well under careful management,
especially if combined with individual housing for parturition and the first week of lactation.
15.10 Piglet Behavioural Development

Behavioural phenotype of a piglet begins being shaped by the interaction of genotype with
environment already during the prenatal life, in a phenomenon called early-life programming.
Alternations in sow nutrition level and diet composition during pregnancy affect piglet
cognitive abilities such as learning and memory (Clouard et al., 2016). Also, individual
piglets that suffer from prenatal growth retardation due to intrauterine competition in large
litters show postnatal behavioural and cognitive alterations that may depend on the phase of
pregnancy when the growth was retarded. Furthermore, exposure of pregnant sows to distress
such as repeated mixing shifts the brain development of the offspring towards an anxiety-
prone phenotype and even compromises maternal behaviour in adult female progeny
(Rutherford et al., 2014).
The suckling period is another sensitive period of life during which environments
enriched through enlarged space, provision of straw, free movement for the sow and/or
contact with other litters have various positive effects on later social coping abilities and
stress resilience (Chaloupková et al., 2007).
Wild boar females and domestic sows in semi-natural enclosures wean their litters at
about 4 months of age. In contrast, piglets are usually weaned between 3 and 5 weeks of age
on commercial farms. The most common way of weaning is to separate piglets from their
mother, move them to another room or building, mix several litters together and switch them
abruptly to solid food. This procedure burdens the piglets with several stressors
synchronously at an age when they are ill-prepared to withstand them (Weary et al., 2008).
The cumulative impact is therefore invariably connected with growth check and often also
with increased frequency of diarrhoea. Piglets weaned at 3 weeks or earlier often engage in
belly-nosing, an abnormal nudging of the bellies of their peers (see Fig. 15.5). Weaning can
be made easier for piglets by weaning later, by removing the sow and by keeping piglets in
their home environment, and also by encouraging solid food intake prior to weaning through
social learning from the mother. Post-weaning litter mixing results in intense fighting.
Masking odours, tranquillizers, dim light and maternal pheromones have all been tried in
alleviating fighting, without much success. When piglets from different litters have the
opportunity to mingle with each other before 3 weeks of age, they become much more
tolerant of alien piglets after weaning.
Fig. 15.5. Early-weaned piglets tend to develop a specific type of abnormal behaviour called
belly-nosing.
Piglets play by the first day of life. The frequency of play behaviour increases until about
3 weeks of age, whereafter it declines, although even adult sows can be stimulated to play by
fresh straw. Play is usually quite synchronized within litters, comprising varied sequences of
solitary locomotory elements like scampering, dashing, pivoting, flopping on the belly and
on the side and head-tossing, combined with patterns of play fighting (often performed in
awkward positions, e.g. while sitting) and play chasing. Play is important as a welfare
indicator since it is performed only when animals are not hungry, ill, stressed or fearful. Play
also enhances welfare (Held and Špinka, 2011) through the immediate self-rewarding effect,
and probably also through enhancement of later social and coping abilities. Play is stimulated
by moderately challenging stimuli such as new bedding material, access to a new space or
meeting alien piglets (before 3 weeks of age).
Rearing pigs in barren environments that discourage natural foraging, sow–piglet

interaction and social interactions such as play renders them less able to withstand stresses
such as weaning, being manipulated by humans, mixing and transportation. Minimal
improvements, such as adding a little straw, do not alleviate the situation. However,
alternatives – such as a more spacious pen with enough bedding and some enrichment objects
– have been proved to make a long-term difference to the development of pigs.
Acknowledgement
Writing of this chapter was supported by the grant MZERO0716 from the Czech Ministry of
Agriculture.
References
Bracke, M.B.M. and Spoolder, H.A.M. (2011) Review of wallowing in pigs: implications for animal welfare. Animal
Welfare 20, 347–363.
Brajon, S., Laforest, J.-P., Bergeron, R., Tallet, C., Hoetzel, M.-J. and Devillers, N. (2015) Persistency of the piglet’s
reactivity to the handler following a previous positive or negative experience. Applied Animal Behaviour Science 162, 9–
19.
Chaloupková, H., Illmann, G., Bartoš, L. and Špinka, M. (2007) Effect of the pre-weaning housing system on play and
agonistic behaviour in domestic pigs. Applied Animal Behaviour Science 103, 25–34.
Clouard, C., Kemp, B., Val-Laillet, D., Gerrits, W.J., Bartels, A.C. and Bolhuis, J.E. (2016) Prenatal, but not early postnatal,
exposure to a Western diet improves spatial memory of pigs later in life and is paired with changes in maternal
prepartum blood lipid levels. FASEB Journal 30, 2466–2475.
Desire, S., Turner, S.P., D’Eath, R.B., Doeschl-Wilson, A.B., Lewis, C.R.G. and Roehe, R. (2015a) Analysis of the
phenotypic link between behavioural traits at mixing and increased long-term social stability in group-housed pigs.
Desire, S., Turner, S.P., D’Eath, R.B., Doeschl-Wilson, A.B., Lewis, C.R.G. and Roehe, R. (2015b) Genetic associations of
short- and long-term aggressiveness identified by skin lesion with growth, feed efficiency, and carcass characteristics in
growing pigs. Journal of Animal Science 93, 3303–3312.
Douglas, C., Bateson, M., Walsh, C., Bedue, A. and Edwards, S.A. (2012) Environmental enrichment induces optimistic
cognitive biases in pigs. Applied Animal Behaviour Science 139, 65–73.
Focardi, S., Morimando, F., Capriotti, S., Ahmed, A. and Genov, P. (2015) Cooperation improves the access of wild boars
(Sus scrofa) to food sources. Behavioural Processes 121, 80–86.
Groenen, M.A.M. (2016) A decade of pig genome sequencing: a window on pig domestication and evolution. Genetics,
Selection, Evolution 48, 23.
Gustafsson, M., Jensen, P., De Jonge, F.H., Illmann, G. and Špinka, M. (1999) Maternal behaviour of domestic sows and
crosses between domestic sows and wild boar. Applied Animal Behaviour Science 65, 29–42.
Held, S.D.E. and Špinka, M. (2011) Animal play and animal welfare. Animal Behaviour 81, 891–899.
Hemsworth, P.H. and Tilbrook, A.J. (2007) Sexual behavior of male pigs. Hormones and Behavior 52, 39–44.
Kirkden, R.D., Broom, D.M. and Andersen, I.L. (2013) Invited review: Piglet mortality: management solutions. Journal of
Melotti, L., Oostindjer, M., Bolhuis, J.E., Held, S. and Mendl, M. (2011) Coping personality type and environmental
enrichment affect aggression at weaning in pigs. Applied Animal Behaviour Science 133, 144–153.
Pedersen, L.J. (2007) Sexual behaviour in female pigs. Hormones and Behavior 52, 64–69.
Podgorski, T., Lusseau, D., Scandura, M., Sonnichsen, L. and Jedrzejewska, B. (2014) Long-lasting, kin-directed female
interactions in a spatially structured wild boar social network. PLoS One 9, e99875.
Rushen, J. and Pajor, E. (1987) Offence and defence in fights between young pigs (Sus scrofa). Aggressive Behaviour 13,
329–346.
Rutherford, K.M.D., Piastowska-Ciesielska, A., Donald, R.D., Robson, S.K., Ison, S.H., Jarvis, S., Brunton, P.J., Russell,
J.A. and Lawrence, A.B. (2014) Prenatal stress produces anxiety prone female offspring and impaired maternal
behaviour in the domestic pig. Physiology & Behavior 129, 255–264.
Rydhmer, L., Gourdine, J.L., de Greef, K. and Bonneau, M. (2014) Evaluation of the sustainability of contrasted pig farming
systems: breeding programmes. Animal 8, 2016–2026.
Špinka, M. (2012) Social dimension of emotions and its implication for animal welfare. Applied Animal Behaviour Science
138, 170–181.
Špinka, M. and Illmann, G. (2015) Nursing behavior. In: Farmer, C. (ed.) The Gestating and Lactating Sow. Wageningen
Academic Publishers, Wageningen, the Netherlands, pp. 297–317.
Stolba, A. and Wood-Gush, D.G.M. (1981) The assessment of behavioral needs of pigs under free-range and confined
conditions. Applied Animal Ethology 7, 383–389.
Stratz, P., Just, A., Faber, H. and Bennewitz, J. (2016) Genetic analyses of mothering ability in sows using field- recorded
observations. Livestock Science 191, 1–5.
Studnitz, M., Jensen, M.B. and Pedersen, L.J. (2007) Why do pigs root and in what will they root? A review on the
exploratory behaviour of pigs in relation to environmental enrichment. Applied Animal Behaviour Science 107, 183–197.
Tallet, C., Linhart, P., Policht, R., Hammerschmidt, K., Šimeček, P., Kratinová, P. and Špinka, M. (2013) Encoding of
situations in the vocal repertoire of piglets (Sus scrofa): a comparison of discrete and graded classifications. PLoS One 8,
e71841.
Turner, S.P., Roehe, R., D’Eath, R.B., Ison, S.H., Farish, M., Jack, M.C., Lundeheim, N., Rydhmer, L. and Lawrence, A.B.
(2009) Genetic validation of postmixing skin injuries in pigs as an indicator of aggressiveness and the relationship with
injuries under more stable social conditions. Journal of Animal Science 87, 3076–3082.
Verdon, M., Morrison, R.S., Rice, M. and Hemsworth, P.H. (2016) Individual variation in sow aggressive behavior and its
relationship with sow welfare. Journal of Animal Science 94, 1203–1214.
Waiblinger, S., Boivin, X., Pedersen, V., Tosi, M.V., Janczak, A.M., Visser, E.K. and Jones, R.B. (2006) Assessing the
human–animal relationship in farmed species: a critical review. Applied Animal Behaviour Science 101, 185–242.
Weary, D.M., Jasper, J. and Hotzel, M.J. (2008) Understanding weaning distress. Applied Animal Behaviour Science 110,
24–41.
Wischner, D., Kemper, N. and Krieter, J. (2009) Nest-building behaviour in sows and consequences for pig husbandry.
Livestock Science 124, 1–8.
Zwicker, B., Weber, R., Wechsler, B. and Gygax, L. (2015) Degree of synchrony based on individual observations
underlines the importance of concurrent access to enrichment materials in finishing pigs. Applied Animal Behaviour
Science 172, 26–32.
16 Behaviour of Dogs
D.L. Wells

The domestic dog, Canis familiaris, is one of more than 36 species of canid, a family
belonging to the order Carnivora that includes, among others, wolves, foxes, coyotes and
jackals. Canids are a widely distributed group of terrestrial predators that, with some
exceptions, are relatively lithe and powerful in build and social in nature.
Many theories have been proposed over the years to explain the ancestry of the domestic
dog (for reviews see Coppinger and Coppinger, 2001; Jensen, 2007; Miklosi, 2014). These
range from the idea that the dog arose from a wild type of canid somewhat similar to today’s
Australian dingo or New Guinea singing dog, to the notion that the dog has multiple
ancestors including, for example, the coyote, Canis latrans, and golden jackal, Canis aureus.
Genetic evidence now suggests that the wolf, and perhaps more specifically, the grey wolf,
Canis lupus (Fig. 16.1), is the closest living relative of the domestic dog; whether or not the
modern-day wolf is the true ancestor of the dog, however, is still under scrutiny.
Fig. 16.1. The grey wolf, Canis lupus, is the closest living relative of the domestic dog.
There is also great debate as to where and when the process of domestication actually
took place. Up until recently, the fossil record was fairly consistent in suggesting that the dog
was domesticated around the end of the last Ice Age, somewhere between 12,000 and 14,000
years ago. More recently, however, dog-like fossils excavated from two separate European
sites have been dated to 33,000 years ago, although it is unclear whether these remains are
ancestors of the modern-day dog or perhaps morphologically distinct wolves (Ovodov et al.,
2011). Studies of molecular genetic markers have yielded contrasting results, although by and
large serve to support the archaeological evidence about the domestication of dogs.
Molecular dating of 18 prehistoric canids, for example, suggests that domestication occurred
somewhere between 18,800 and 32,100 years ago (Thalmann et al., 2013). Wang et al.
(2013) similarly calculate that the split between wolves and dogs occurred about 32,000
years ago. A recent genetic analysis involving hundreds of breeds suggests that dogs may
have been domesticated twice from two distinct wolf populations, once in Asia and once in
Europe (Frantz et al., 2016). The debate surrounding the geographical and temporal origins
of the domestic dog is likely to continue, but most of the evidence that is accumulating seems
to favour multiple domestication events that took place 14,000–33,000 years ago.
Domestication, and further artificial selection by humans, has given rise to a wide variety
of dog breeds. Indeed, today, there are over 400 breeds of domestic dogs, with newly
recognized crossbreeds (e.g. labradoodles, schnoodles, cockapoos) gaining in popularity.
Many of these breeds show enormous morphological diversity. Some (e.g. German shepherd
dog, Japanese akita, Siberian husky) have retained their wolf-like appearance. Others (e.g.
poodle, chihuahua, bichon frise), however, have been highly neotenized and now bear little
resemblance to their wild progenitor (Fig. 16.2). The function of dogs in today’s society is as
diverse as their physical appearance. While most are kept exclusively for the purpose of
companionship, many serve as working animals and are involved in, among other duties, the
herding of sheep, the guarding of property, the detection of criminals, drugs and explosives
and the provision of assistance for those with disabilities (for review see Coppinger and
Coppinger, 2001).
Fig. 16.2. Modern-day dogs, although neotenized to different degrees, share the same
ancestor.

By and large, the Canidae are social carnivores, with a propensity to develop and dwell in
groups. The composition of these groups varies significantly across both species and
individual populations. Some attempts have been made to categorize the canids according to
their social organization and behaviour. Michael Fox (1975), for example, has suggested that
these animals can be loosely divided into three types. Type 1 includes solitary hunters such as
the red fox, Vulpes vulpes, which generally develop temporary pair bonds during the breeding
season. The male may stay to assist with the rearing of the pups, but will usually disappear
once the offspring can fend for themselves. Type 2 canids include those species that develop
more of a longstanding pair bond, such as the dingo or coyote, C. latrans. An abundance of
food may keep the unit together, although offspring will usually disperse if and when food
supplies become scarce. Type 3 canids include the wolf. These animals generally live in
packs of between two and 30 members, with group size regulated by the availability of food
and other resources (Mech, 1970). Packs normally develop complex dominance hierarchies,
or ‘pecking orders’, which play an important role in the determination of access to privileges
and initiatives including travelling, hunting and reproduction. While normally monogamous,
with animals developing long-term pair bonds, polyandrous matings in wolves have been
witnessed.
The social behaviour of domestic dogs is as diverse as that of their wild cousins. Feral
dogs provide an interesting link between the social behaviour of the wolf and the pet dog.
True feral dogs (as opposed to free-ranging animals that may have owners or associate
themselves with a particular household) generally dwell in groups of between two and six
animals. These are not as tightly governed by pecking orders or efforts to develop
cohesiveness, however, as the average wolf pack. Although wolf packs are comprised of
genetically related members, feral dog groups are not typically related and group
membership may be more transient in nature. Unless socialized with humans early on in their
development (see later), feral dogs may show fearful reactions towards people, taking great
efforts to avoid them.
Pet dogs, which spend much of their time in the company of humans, have learnt how to
integrate themselves successfully into the family unit and, in most cases, develop complex
social relationships with their caregivers (Fig. 16.3). Studies using Ainsworth’s Strange
Situation Test (SST), a tool traditionally employed to measure child–caregiver attachment,
have led to the general conclusion that the human–dog relationship is functionally analogous
to the human–infant bond (e.g. Topal et al., 2005). Indeed, many owners do claim to regard
their dog as another member of the family and, in most cases, a child. The close bond of
attachment that has evolved between dogs and humans has resulted in the development of an
infant-like social competence in this species. Unlike its predecessor, the wolf, the domestic
dog is highly attuned to the emotional, behavioural and communicative cues presented by its
human caregivers and engages in social referencing; in other words, it synchronizes its
reactions with those of its owners (see Miklosi et al., 2003).
Fig. 16.3. The domestic dog is one of the most popular companion animals in Western
society, and for many people an integral part of family life.
The exact nature of the relationship that develops between a dog and its caregiver is very
much owner–pet specific, and the success of the pairing is dependent upon the type of bond
that is established. Inappropriate bonding (e.g. over-attachments, under-attachments),
coupled with a lack of obedience training, can sometimes lead to the development of
psychological problems which manifest themselves through the animals’ behaviour (e.g.
separation anxiety, aggression). Some of these problems can be rectified by behavioural
modification (for examples of common behaviour problems and appropriate treatments see
Landsberg et al., 2013); many, however, culminate in the abandonment of dogs on to the
streets or relinquishment to animal rescue shelters.
Although the human–canine bond is not always a harmonious one, people can reap
enormous benefits from the intimate relationship they develop with their dogs. There is now
mounting evidence to suggest that dogs are capable of bestowing their owners a wide range
of physical and psychological health benefits, ranging from reduced depression and enhanced
self-esteem through to improved recovery from heart attacks (for review see Wells, 2007).
For this reason, the dog is now widely employed in a therapeutic captivity (e.g. pet-assisted
therapy).

The Canidae are opportunistic scavengers that will ingest a wide range of animal and
vegetative matter. Much of the diet consumed depends upon the species of canid under
scrutiny and the geography of the habitat. Thus, North American grey wolves generally live
on elk, caribou and deer during the winter months, consuming rodents and lagomorphs in the
summer.
Other wild canids may feed less exclusively on prey and subsist on a wider range of
available food types. Some jackals, for instance, have been noted to live primarily on carrion,
supplemented by fruit, grass and insects such as dung beetles. Coyotes, likewise, have been
shown to eat a diverse range of foods including rabbits, hares, rodents, plants and insects.
Arctic foxes, Alopex lagopus, which live in a relatively harsh environment, eat almost any
form of meat readily available. Prey typically consists of lemmings and voles, but can also
include ground squirrels, seal pups, bird eggs and remnant scraps from polar bears. This
species, and other canids that hunt alone, often hoards surplus food in caches, in some cases
large enough to feed an individual for up to a month. Domestic dogs can often be witnessed
doing the same thing with excess food or other resources (e.g. toys). Here, the front paws are
used to dig a hole, the mouth or nose is used to place the relevant item inside, and the nose is
employed to push over the earth and conceal the resource.
Feral dogs, like wild canids, are opportunistic scavengers. Again, much of their diet
depends upon the availability of food items. City-dwelling feral dogs often forage from bins
and waste-sites and may consume a wide variety of foods, much of which has been prepared
for human consumption. Some feral dogs will kill and consume prey specific to the area.
Several authors, for instance, have witnessed feral dogs on the Galapagos Islands feeding on
marine iguanas, Amblyrhynchus cristatus, and others, in Venezuela’s Llanos, preying on
capybara, Hydrochoerus hydrochaeris. Some concern has recently been expressed over
wildlife depletion arising from feral dog predation on livestock populations. Scientific studies
of such animals in North America and Italy, however, have revealed little, if anything, in the
way of such activity, and much of the damage to wildlife is now thought to be due to free-
ranging and stray dog populations.
The feeding behaviour of pet dogs is very much under the control of humans. A diverse
range of dry (generally cereal-based) and wet (meat-based) foods is available on the market
for the pet dog population. Although diet is ultimately controlled by the owner and may be
related to factors including price, smell and ease of preparation, dogs often show preferences
for certain types of food over others. Studies have shown that dogs prefer cooked over raw
meat, meat over cereal-based foods and show a ranked preference for beef over pork,
followed by lamb, chicken and finally horsemeat (for review see Bradshaw and Thorne,
1992). Like humans, some dogs can develop intolerances to certain components of food,
giving rise to physical and, in some cases (and still in need of more scientific evidence),
psychological problems; a wide range of speciality meals (e.g. wheat-free, preservative-free)
is now available on the open market to cater for these animals.
While dogs may have a preference for protein-rich foods and owners certainly veer more
towards feeding their pets meat-based meals, carbohydrates probably featured heavily in the
diet of early modern-day dogs. Recent studies point to mutations in a number of key genes
(notably Amy2B) involved in starch digestion in domestic dogs (Axelsson et al., 2013;
Ollivier et al., 2016). These mutations, which may have happened during the agricultural
revolution, allowed modern dogs to live on the same starch-rich diet as people. The
mutations may have constituted an important step in the domestication process and point to
an interesting co-evolution of adaptive dietary changes in humans and dogs.
Although dogs may have evolved to digest starch-rich diets, they still have a tendency,
like their wild cousins, to scavenge opportunistically. It is not unusual to see some dogs,
notably those bred for hunting purposes (e.g. terriers), stalking and killing small rodents.
Others may ingest faeces (coprophagy) belonging to either themselves or another animal.
The aetiology of this behaviour pattern in dogs remains largely unknown, although several
anecdotal suggestions have been proposed. Some believe that stool-eating indicates an
imbalanced diet or pancreatic enzyme deficiency; faeces are thus ingested in an attempt to
gain more nutrients. Others adhere to the belief that in dogs coprophagy results from
inappropriate operant conditioning, arising either from the attention directed towards the dog
by its owner (e.g. verbal reprimand serving as positive reinforcement) or the taste of the
faeces themselves.
16.4 Communication
Dogs communicate with both their conspecifics and humans using a wide variety of sensory
systems, including vision, olfaction and audition.
Vision
In relation to some animals (e.g. primates, cats), dogs are not a particularly vision-driven
species. Indeed, it has been estimated that the dog’s eye for detail is approximately six times
poorer than the average human’s. Colour perception is also relatively limited in this species.
Although not entirely colour blind, as often assumed, dogs are believed to perceive the world
in much the same way as a red–green colour-blind person. From an evolutionary point of
view, the dog has not needed these types of visual characteristic to survive; instead, however,
it has developed a good ability to see in dim light and detect subtle movements and motion.
Dogs use a range of visual signals to communicate their intentions to others (Fig. 16.4).
The positioning of the body, ears, tail and direction of eye gaze can all convey important
information about status and behavioural intention (Abrantes, 2003). Dominant animals are
more likely to exhibit an upright body posture, erect ears and tail and maintain direct eye
contact. In some cases, an aggressive dominant individual may bare its teeth and raise its
hackles, i.e. the fur on its back (pilo-erection). All of these visual cues are designed to make
the animal look bigger, more threatening and ultimately reduce the need for physical combat.
Subordinate individuals, by contrast, generally make themselves look smaller, thus lowering
the body, ears and tail and breaking eye contact earlier than their more dominant
counterparts. Extremely submissive animals may roll over on their back, present their
inguinal region and, in some cases, even urinate. Aggressive submissive animals may display
a rather confusing mixture of visual signals that, on the one hand, imply confidence and a
desire to fight (pilo-erection, bared teeth) but, on the other, convey a desire to avoid the
situation and take flight (flattened ears, low body posture, low tail).
Fig. 16.4. Body posture can convey important information about status and behavioural
intention. The dog on the left, with its flattened ears and lowered body, is showing signs of
submission; the dog on the right, with its upright tail and firm stance, is showing signs of
dominance.
All of the above visual signals can be conveyed by dogs to other animals and humans.
The latter, however, are not always adept at accurately interpreting many of the more subtle
behavioural cues and this lack of understanding can, in some unfortunate situations,
culminate in bites and other injuries.
Although dogs can and do employ visual signals to communicate their intentions, the
highly neotenized appearance of some breeds, with their floppy ears, docked tails and longer
fur, has led to a reduced ability to accurately deliver and/or interpret visual information.
Domestic dogs are thus probably more reliant on olfactory signals as a method of
communication.
Olfaction
The domestic dog is well renowned for its remarkable sense of smell, with the olfactory
acuity of this species being somewhere between 10,000 and 100,000 times better than that of
humans. Using its sense of smell, the dog can discriminate between the odours of its own
species, other species (e.g. humans) and successfully match odours to sample. Odours can
have an excitatory or relaxant effect on dogs and, as such, can be used (with care) as a
method of environmental enrichment for animals housed in stressful situations (Wells, 2009;
Fig. 16.5).
Fig. 16.5. Odours with relaxant properties, such as lavender and chamomile, can have a
calming effect on dogs and harbour therapeutic value for animals prone to anxiety or stress.
Olfaction plays an important role in canid communication. Three types of olfactory

signal are utilized for this purpose: the deposition of urine, glandular secretions and faeces.
Urine is the most commonly employed olfactory cue, the composition of which can convey
information about individual status and may be utilized to denote territories and home ranges.
The action of urination can also communicate useful visual information. Male dogs in
particular (although not exclusively) generally show raised leg urination (RLU). Sometimes
males will cock their legs without the production of urine; a so-called raised leg display
(RLD). This is particularly common in the presence of other dogs, suggesting a visually
driven display function of the raised leg.
Glandular secretions are also used by dogs for the purpose of communication. There is a
particularly high concentration of sudoriferous glands, which produce watery secretions,
around the head and anal regions. Dogs are prone to sniffing these areas, suggesting that the
secretions play a role in olfactory communication. The release of apocrine and sebaceous
secretions from the anal glands during defecation also provides key cues, and may be
employed in individual and territorial recognition.
Faecal matter is the least likely product of elimination to be utilized by domestic dogs for
communication; the wolf, by contrast, has been noted to defecate at the junctions of trails,
presumably to convey information on territorial boundaries. Some domestic dogs will scratch
the ground with their hind legs after defecation (and sometimes urination). The precise
function of this behaviour pattern is still unclear. One explanation is that the scratching helps
to spread the scent of the faecal material; in most cases, however, the deposition is seldom
touched, weakening the plausibility of this theory. Other types of odour cue, however, may
still be spread via the scratching. The dog’s sweat glands are located on the pads of its paws
and the sebaceous glands are positioned in between the toes; scratching may therefore help to
spread these specific types of glandular secretion. The action of scratching the ground may
also provide visual information to conspecifics, in much the same way as the RLD discussed
above.
The dog’s remarkable sense of smell, and ability to be trained easily, has resulted in it
becoming widely employed by organizations around the world for purposes including
tracking, search and rescue, cadaver recovery and explosive, narcotic, mould and gas
detection. More recently, it has been shown that dogs can even be trained to use their sense of
smell to detect underlying health problems, including certain types of cancer (for review, see
Wells, 2012).
Audition
Dogs have a remarkable sense of hearing. Not only can they perceive sounds across a wide
spectrum of frequencies, they can often pinpoint the exact source of a sound and accurately
discriminate between a variety of auditory signals. Dependent upon breed and age, the
average domestic dog can hear sounds in the spectrum of 67–45,000 Hz; this compares with
a frequency range of 64–23,000 Hz in humans. The mobility of the dog’s ear, which in many
breeds can prick up and move around, in combination with the ‘cupped’ shape, explains, in
part, the dog’s acute sense of hearing.
Auditory signals in the form of vocalizations are often used as a method of
communication. The domestic dog can elicit a wide variety of vocalizations, including
grunts, growls, whines, yelps, coughs and, of course, barks. The wolf seldom barks, with this
type of signal comprising only about 2% of lupine vocalizations. Moreover, the bark tends to
be context-specific in the wolf, used mainly as an alert or challenging threat. The domestic
dog, in contrast, barks very frequently, for extremely lengthy periods of time, and in virtually
every behavioural context.
The divergence in bark behaviour between dogs and wolves has prompted researchers to
suggest that barking in domestic dogs is generally applicable and may lack specific
communication functions. It has been shown, however, that bark structure in dogs varies
predictably with context (Yin, 2002). Barking, and other dog vocalizations, may thus be a
more complex form of communication than previously thought. It is hoped that research will
shed more light on this in the years to come.
16.5 Biological Rhythms

The wild Canidae vary substantially in their biological rhythms. Wolves and foxes are
generally active at night, while dingoes and African wild dogs, Lycaon pictus, show a
crepuscular pattern of diurnal behaviour, with peak periods of activity occurring at dawn and
dusk, when temperatures are cooler and the opportunities for hunting prey are greater.
The domestic pet dog is diurnal in nature, tending to sleep when its owner does.
However, unlike humans, the dog is a polyphasic sleeper, having several bouts of sleep per
night. The average sleep episode lasts for just under an hour and a half, with mean waking
periods being about 40 min in duration. Most dogs also sleep several times during the day;
the frequency and duration of these sleeping bouts vary substantially between animals, being
dependent upon individual dog–owner routines. As in humans, the canid sleep cycle
comprises both periods of quiet (i.e. non-REM) and rapid eye movement (REM) sleep. It is
during this latter stage of sleep, in which animals are frequently witnessed to paddle their
legs and utter vocalizations, that dogs are believed to dream. Dogs may be able to shift to
nocturnal activities quite easily. Studies of urban stray dogs, for instance, have shown that
such animals can be quite active during the night and early morning.
Sleep disturbances can appear with increasing age, with older animals sleeping more
during the day and less at night. This is often a feature of the degenerative disorder cognitive
dysfunction syndrome, a condition that has been likened to Alzheimer’s in humans.

The domestic dog is sexually mature by about 7–8 months of age; this is in stark contrast to
the wolf, which does not reach sexual maturity until approximately 22 months. Male dogs are
capable of producing sperm, and hence of reproducing, at any time of the year, although most
females are dioestrous, coming into season (‘heat’) only twice annually; one exception is the
basenji, a breed that originates from central Africa and comes into season only once a year.
The average female first comes into season between 6 and 12 months of age.
The oestrous cycle of the bitch comprises four stages: pro-oestrus, oestrus, metoestrus
(or dioestrus) and anoestrus. Pro-oestrus lasts for up to 2 weeks and is characterized by a
bloody discharge, restless behaviour and increasing attractiveness to male animals. Most
females are unreceptive to mating during this stage of the cycle, although males may attempt
mounting.
During oestrus, the female is receptive to mating. This typically involves sniffing of the
anogenital region by the male, mounting, penetration and ejaculation of sperm. In most cases,
the penis is held for some time inside the vagina, and the animals are consequently locked
together in a tie position, often facing opposite directions. Ovulation normally takes place,
even in the absence of a mating, on day 2 of the oestrus phase. Oestrus may last for anywhere
between 10 and 21 days, the length of which is dependent upon whether or not the bitch has
been mated.
The metoestrus period lasts for about 2 months and may constitute the period of
pregnancy if mating, and successful fertilization of the egg, has occurred. Pseudo, or false,
pregnancies are relatively common in the dog, with some females producing milk, building
nests and displaying other signs of maternal behaviour. These behaviour patterns tend to
disappear once hormone levels have stabilized. In rare cases, some bitches may have a false
pregnancy with each heat.
The final stage of the oestrous cycle, anoestrus, represents the period of reproductive
inactivity and lasts for an average of 4–5 months.
16.7 Behaviour at Birth and Parental Care

The gestational period of the domestic dog is approximately 63 days. One or two days prior
to parturition, the bitch may become restless, seek seclusion and shred papers, blankets or
bedding in an attempt to build a ‘nest’. Whelping very often takes place overnight and lasts
anywhere between 3 and 6 h. The first signs of labour are panting and vaginal discharge. The
first puppy normally takes the longest to whelp, with an average interval of 45 min between
the births of individuals. Many whelpings proceed without complication or human
intervention; however, some births may need to be closely monitored, and caesarean sections
are not uncommon, particularly among breeds with large heads.
The average litter size is six or seven, although larger breeds tend to have considerably
more (and smaller) puppies than smaller breeds. Puppies are born in individual placental sacs
containing amniotic fluid; these are normally removed by the mother using her teeth and
sometimes ingested, providing a useful source of nutrients. Puppies are licked by the mother
immediately after birth for the purpose of cleaning and to stimulate breathing. Most puppies,
guided by both olfactory cues and maternal nudging, attach themselves to a teat and start
suckling within the first hour of birth (Fig. 16.6). Mothers often continue to clean their pups
whilst suckling, particularly the anogenital area in a bid to stimulate elimination. This
procedure is repeated for the first few days after birth, as puppies are unable to urinate or
defecate without the aid of their mother’s tongue.
Fig. 16.6. Rottweiler puppies suckling from their mother. Puppies suckle exclusively for the
first 3–4 weeks of their lives, following which weaning begins.
The mother focuses almost exclusively on the cleaning and suckling of her puppies for
the first 2–3 weeks of life, following which she starts to leave them alone for maybe an hour
or two a day. Puppies suckle from their mother for the first 3–4 weeks of life, following
which weaning begins and episodes of nursing decrease. Around this time, the mother may
start to push puppies away from her teats and encourage independence. Most puppies are
fully weaned by 8 weeks of age.
16.8 Development of Behaviour

The behavioural development of the domestic dog can be divided into five stages: prenatal,
neonatal, transitional, socialization and juvenile.
Prenatal period
The prenatal phase covers the entire period from conception to parturition and lasts
approximately 63 days. This period of development is still relatively shrouded in mystery. It
has been discovered, however, that the domestic dog, like many other species, can learn
about chemosensory cues (i.e. odours and/or tastes) during the prenatal period, and that such
learning may play an important role in shaping olfactory-guided development and behaviour
following birth (Wells and Hepper, 2006).
Neonatal period (0–14 days)
The canid puppy is relatively helpless for the first 2 weeks of its life, relying heavily upon its
mother for suckling, elimination and comfort. Most of the puppy’s time is dominated by
feeding and sleeping. The eyes and ears are closed and non-functional during this stage, with
the animal relying most heavily upon its senses of smell, taste and touch. While the puppy
can move, motor abilities are limited and often reflex-driven. Vocalizations including whines,
grunts and mews are relatively common, and often used as a method of alert or attention-
seeking.
Transitional period (14–21 days)
As the name suggests, the transitional phase constitutes a period of rapid neurological and
physical change, the most apparent of which is the opening of the eyes around day 13 and ear
canals between days 18 and 20. The puppy is no longer reliant upon its mother for the
stimulation of elimination, and can both urinate and defecate on its own accord, usually
outside the nest. Motor movements are more advanced than during the preceding neonatal
period, with the animal being able to crawl forwards and backwards, stand up on four paws
and walk, albeit clumsily. During the transitional period, puppies start to interact with one
another, displaying their first signs of play fighting and tail wagging. Vocalizations become
louder and are produced in response to a wider variety of stimuli.
Socialization period (3–10 weeks)
This relatively long period of development sees the onset of more adult patterns of behaviour.
The pup’s sensory and motor abilities are fully developed, allowing it to explore and learn
about its surroundings. The puppy gradually spends more and more time away from its
mother, relying less on feeding and sleeping. Teeth begin to erupt at the start of the
socialization period, and weaning is normally complete by 7–8 weeks of age.
The socialization phase was once considered to be a critical period in the development of
social relationships (Scott and Fuller, 1965). Nowadays, it is more generally recognized as a
sensitive period of development, during which the pup develops important social bonds with
its mother, littermates and humans. Inappropriate socialization during this period can
sometimes lead to fear-induced behavioural problems that often persist into adulthood and, in
many cases, are resistant to modification.
Juvenile period (10 weeks–sexual maturity)
The juvenile period of development runs from approximately 10 weeks of age up to sexual
maturity. The pup grows rapidly during this stage, with most breeds reaching their full height
and weight by about 8 months. Adult teeth replace milk teeth at roughly 5 months of age.
Most puppies are removed from their mothers at the start of this period, and continue to
develop social relationships with humans and other animals. General behaviour is similar to
that expressed by puppies during the socialization phase, but is more advanced and controlled
in nature. Male dogs start to show RLU (see earlier), although the age of onset of this varies
between individuals. Puberty is a rather gradual process in male dogs, with animals
expressing an interest in females that are in oestrus as early as 4 months of age, but not
normally attempting to mate until 7–8 months; puberty in females, by contrast, is much more
sudden, occurring at the time of the bitch’s first season, anywhere between 6 and 12 months
of age.
16.9 Canine Cognition

Considerable attention has been devoted of late to the cognitive abilities of the domestic dog.
Early interest in this area focused on generic canine intellect, giving rise to popular books
and owner-administered ‘IQ’ tests. More recently, research has concentrated on specific
facets of dog cognition in an effort to explore how this species perceives the world, solves
problems and makes decisions.
Studies point to advanced memory and learning skills in the dog. For example, a case
study of a border collie, Rico, indicated that the animal was able to learn the names of over
200 simple items and demonstrated retention 4 weeks after his last exposure to them
(Kaminski et al., 2004). It was concluded that Rico was able to ‘fast map’, i.e. infer the
meaning of a new word after a single exposure by exclusion learning.
The close bond that exists between dogs and humans (see earlier) has promoted a focus
on social cognition, for example perspective-taking, cooperation and deception (for review
see Miklosi, 2014). Studies have shown that many, although not all, dogs are sensitive to
people’s attentional states. Dogs look towards the human face in both times of uncertainty
(e.g. in fearful contexts) or when trying to access something out of reach (e.g. when
attempting to solve an unsolvable problem). Dogs will even use visual and acoustic signals
(e.g. barking) designed to attract their owner’s attention if they cannot reach a valuable
resource (Miklosi et al., 2004).
There is now evidence that dogs can discriminate between the emotional expressions of
human faces, presumably by forming memories of different expressions and the context in
which they are presented (Muller et al., 2015). Dogs can also use human communicative
gestures, notably pointing, to locate an object, food or otherwise. As one might expect,
breeds that work more closely with humans (e.g. gundogs, sheepdogs) employ such signals
more effectively than those that work more independently (e.g. sled dogs).
Dogs’ performance on perspective-taking has led to suggestions that dogs, like humans
and some primates, may have a theory of mind; in other words, an ability to predict the
behaviour or attribute mental states to others. It is acknowledged, however, that results may
be influenced by experimental design and individual differences; it may therefore be more
appropriate at this stage to consider dogs to have a ‘rudimentary’ theory of mind (Horowitz,
2011).
References
Abrantes, R.A. (2003) The Evolution of Canine Social Behaviour. Wakan Tanka Publishers, Naperville, Illinois.
Axelsson, E., Ratnakumar, A., Arendt, M., Maqbool, K., Webster, M.T., Perloski, M., Liberg, O., Arnemo, J.M.,
Hedhammar, A. and Lindblad-Toh, K. (2013) The genomic signature of dog domestication reveals adaptation to a starch-
rich diet. Nature 495, 360–365.
Bradshaw, J.W.S. and Thorne, C.J. (1992) Feeding behaviour. In: Thorne, C.J. (ed.) The Waltham Book of Dog and Cat
Behaviour. Pergamon Press, Oxford, pp. 115–129.
Coppinger, R. and Coppinger, L. (2001) Dogs: A Startling New Understanding of Canine Origin, Behavior and Evolution.
Scribner, London.
Fox, M.W. (1975) The Wild Canids: Their Systematics, Behavioral Ecology and Evolution. Van Nostrand Reinhold, New
York.
Frantz, L.A.F., Mullin, V.E., Pionnier-Capitan, M., Lebrasseur, O., Ollivier, M., Perri, A., Linderholm, A., Mattiangeli, V.,
Teasdale, M.D., Dimopoulos, E.A. et al. (2016) Genomic and archaeological evidence suggests a dual origin of domestic
dogs. Science 352, 1228–1231.
Horowitz, A. (2011) Theory of mind in dogs? Examining method and concept. Learning & Behavior 39, 314–317.
Jensen, P. (2007) The Behavioural Biology of Dogs. CAB International, Wallingford, UK.
Kaminski, J., Call, J. and Fischer, J. (2004) Word learning in a domestic dog: evidence for ‘fast mapping’. Science 304,
1682–1683.
Landsberg, G., Hunthausen, W. and Ackerman, L. (2013) Handboook of Behaviour Problems of the Dog and Cat, 3rd edn.
Saunders, London.
Mech, L.D. (1970) The Wolf: The Ecology and Behaviour of an Endangered Species. Natural History Press, New York.
Miklosi, A. (2014) Dog Behaviour, Evolution, and Cognition. Oxford University Press, Oxford.
Miklosi, A., Kubinyi, E., Topal, J., Gacsi, M., Viranyi, Z. and Csányi, V. (2003) A simple reason for a big difference: wolves
do not look back at humans, but dogs do. Current Biology 13, 763–766.
Miklosi, A., Topal, J. and Csanyi, V. (2004) Comparative social cognition: what can dogs teach us? Animal Behaviour 67,
995–1004.
Muller, C.A., Schmitt, K., Barber, A.L.A. and Huber, L. (2015) Dogs can discriminate emotional expressions of human
faces. Current Biology 25, 601–605.
Ollivier, M., Tresset, A., Bastian, F., Lagoutte, L., Axelsson, E., Arendt, M.L., Bălăşescu, A., Marshour, M., Sablin, M.V.,
Salanova, L. et al. (2016) Amy2B copy number variation reveals starch diet adaptations in ancient European dogs. Royal
Society Open Science 3, 160449.
Ovodov, N.D., Crockford, S.J., Kuzmin, Y.V., Higham, T.F.G., Hodgins, G.W.L. and van der Plicht, J. (2011) A 33,000 year-
old incipient dog from the Altai Mountains of Siberia: evidence of the earliest domestication disrupted by the last glacial
maximum. PLoS One 6, e22821.
Scott, J.P. and Fuller, J.L. (1965) Genetics and the Social Behaviour of the Dog. University of Chicago Press, Chicago,
Illinois.
Thalmann, O., Shapiro, B., Cui, P., Schuenemann, V.J., Sawyer, S.K., Greenfield, D.L., Germonpré, M.B., Sablin, M.V.,
López-Giráldez, F., Domingo-Roura, X. et al. (2013) Complete mitochrondrial genomes of ancient canids suggest a
European origin of domestic dogs. Science 342, 871–874.
Topal, J., Gacsi, M., Miklosi, A., Viranyi, Z., Kubinyi, E. and Csanyi, V. (2005) Attachment to humans: a comparative study
on hand-reared wolves and differently socialized dog puppies. Animal Behaviour 70, 1367–1375.
Wang, G.D., Zhai, W.W., Yang, H.C., Fan, R.X., Cao, X., Zhong, L., Wang, L., Liu, F., Wu, H., Cheng, L.G. et al. (2013)
The genomics of selection in dogs and the parallel evolution between dogs and humans. Nature Communications 4,
1860.
Wells, D.L. (2007) Domestic dogs and human health: an overview. British Journal of Health Psychology 12, 145–156.
Wells, D.L. (2009) Sensory stimulation as environmental enrichment for captive animals: a review. Applied Animal
Wells, D.L. (2012) Dogs as a diagnostic tool for ill-health in humans. Alternative Therapies in Health and Medicine 18, 12–
17.
Wells, D.L. and Hepper, P.G. (2006) Prenatal olfactory learning in the domestic dog. Animal Behaviour 72, 681–686.
Yin, S. (2002) A new perspective on barking in dogs (Canis familiaris). Journal of Comparative Psychology 116, 189–193.
17 Behaviour of Cats
J. Bradshaw

The cat was the first species to become domestic solely for the purpose of controlling pests,
namely the rodents exploiting the stores of dried grain that followed the domestication of
cereal crops (Diamond, 1997). The ancestral species is the wildcat Felis silvestris lybica,
which is still widely distributed, though often in marginal habitats, from northern Europe to
southern Africa and western India. The DNA of modern domestic cats suggests that they first
became isolated from their wild ancestors about 10,000 years ago somewhere in the ‘Fertile
Crescent’, which runs from Turkey to Egypt, supporting the theory that they were
domesticated for their ability to control mammalian pests. The precise location remains
unclear; the most recent study (Driscoll et al., 2007) did not sample any cats from North
Africa, and an earlier study (Randi and Ragni, 1991) identified the North African/Arabian
subspecies F. silvestris lybica as the most likely candidate. The scarcity of archaeological
evidence for domesticated cats until about 4000 years ago suggests that during this period
they were semi-domesticated commensals, valued primarily for their vermin-suppressing
abilities. The choice of F. silvestris appears to have been fortuitous, since many of the other
small species of cat appear to be suitable for domestication (Cameron-Beaumont et al.,
2002).
The earliest solid archaeological evidence for domesticated cats comes from the Egyptian
New Kingdom, in about 1550 bc. By 850 bc the cult of the cat-goddess Bastet had given
religious significance to cats, and they were deliberately bred for sacrificial purposes. There
were also periods during which cats were prohibited to be exported from Egypt (Serpell,
2014), although whether this arose from a recognition of their value as pest controllers, or
their sacred status, or the former in the guise of the latter, is unclear. Certainly cats did not
become widespread in Europe until about the fourth century bc. By the 10th century ad they
had become valued contributors to agricultural economies: at that point in Wales they sold
for as much as a full-grown sheep.

The association between cats and paganism did not help their popularity in many parts of
Western Europe from the time of the Reformation onwards, and some ritualistic persecution
of cats persisted into the 20th century. Nevertheless, from the 19th century onwards the cat
increased in popularity, and by the end of the 20th its numbers had surpassed those of the dog
in many Western countries.
The majority of pet cats are of no particular breed and are referred to variously as non-
pedigree cats, mongrel cats, cross-bred cats and ‘moggies’. Apart from the colour and length
of their coats, which are likely to have been subjected to some deliberate selection by man
(Todd, 1977), they vary little in size compared with most other domesticated species and,
apart from ease of socialization to man, retain many of the characteristics of wildcats. The
genetic differences between wild and domestic cats are small, and studies suggest that
differences between the two in behaviour may be accounted for by as few as 13 genes
(Montague et al., 2014). Pedigree cats still make up a minority of the pet cat population and
most were derived within the past 150 years from local mongrel types: their DNA indicates
that they were derived from ordinary cats by selection, rather than by archaic cross-breeding
with other subspecies or species of Felis (Menotti-Raymond et al., 2008). More recently,
breeds have been developed based on specific mutations, such as the Munchkin (dwarfism),
the hairless Sphynx and the behaviourally unreactive Ragdoll. Others have been derived by
hybridizing domestic cats with other felid species: examples include the Chausie (a hybrid
with the jungle cat Felis chaus) and the Bengal, which derives its spotted coat and
characteristic boldness from the Asian leopard cat Prionailurus bengalensis.
17.2 Effects of Domestication

Excluding pedigrees, the majority of cats are, in contrast to most other domesticated species,
the product of unregulated matings between individuals that have selected one another.
Population control is achieved largely by neutering or by confining entire animals indoors;
the latter is becoming more widespread in urban areas of the USA and parts of continental
Europe, but is still uncommon in the UK and in the countryside. Hunting is a second wild-
type behaviour retained by many cats; although few are very efficient predators, cats are
effective enough as killers of birds and small mammals to have a local impact on their
population dynamics (see below). It was not until the 20th century that cats became valued
more for their companionship than for their vermin-suppressing capabilities, but until
recently it would have proved difficult to breed cats that showed little inclination to hunt.
This stems from the cat’s peculiar nutritional requirements (Zoran and Buffington, 2011),
which it shares with the whole of the Felidae. These include: a restricted ability to shut down
protein catabolism, resulting in a requirement for 18% protein in the diet of growing kittens;
high dietary levels of sulfur-containing amino-acids such as cysteine, methionine and taurine;
an inability to synthesize prostaglandins from small precursor molecules; and a high
requirement for niacin and thiamin. These constraints present no problems when the cat’s diet
consists largely of raw carcasses, but they prevent cats from switching to a predominantly
plant-based diet when prey is scarce, something that dogs, for example, are able to do. Until
modern, nutritionally balanced cat foods became available in the 1970s, hunting was
probably the only way that cats could guarantee themselves enough of these and other key
nutrients to reproduce successfully, and has therefore persisted (Bradshaw et al., 1999).
Probably linked to this is the tendency of most outdoor cats to try to maintain an exclusive
home range, an essential source of food for their recent ancestors.
17.3 Hunting Behaviour and Object Play

All cats, with the exception of some pedigrees, therefore have the potential to become
effective predators, and this has not endeared them to people who are concerned about
protecting wildlife. Their ecological impact is probably most serious when they have been
introduced on to islands where there are no other mammalian predators (e.g. Bonnaud et al.,
2007), but the effect they have on urban bird populations, for example, is still debatable
(Baker et al., 2008; Lilith et al., 2010). Even well-fed pet cats still go on hunting expeditions,
thereby disrupting the behaviour of their prey species even if they are not successful at
killing. Because cats do not hunt cooperatively and are therefore restricted to small prey, if
they are relying on prey for food they need to hunt before they become hungry, and hence
feeding, hunger and predation are only loosely connected in the brain (Adamec, 1976).
Hunting is therefore to a large extent a ‘free-running’ behaviour, and the main influence of
hunger appears to be in the size of prey that is approached and the ‘seriousness’ with which it
is attacked (Biben, 1979). Large or dangerous prey is taken only when the cat is especially
hungry; when the cat is not hungry it may attack the prey but not kill it immediately. This
strategy is presumed to have evolved from a trade-off between the immediate benefits of the
food obtained from a kill and the risk of the cat becoming injured in the process. When these
are balanced, the cat appears to ‘play’ with the prey item, although this an anthropomorphic
description for a behaviour more correctly labelled as arising from conflict between
motivations. Conversely, the factors that induce adult cats to ‘play’ with toys suggest that, in
motivational terms, this is a type of predatory behaviour: the most effective toys resemble
typical prey in size, surface texture and the way they move, and the intensity of the play and
the size of toy preferred both increase with hunger (Hall, 1998).
17.4 ‘Umwelt’
The domestic cat’s sensory capacities appear to be essentially unchanged from those of its
wild ancestors. They overlap considerably with our own, but there are also significant
differences which can help to explain some otherwise baffling cat behaviour. Their sense of
smell, while slightly less sensitive than that of the domestic dog, is thousands of times more
responsive than our own. Moreover, like many mammals including the dog, cats possess a
second olfactory system, the vomeronasal organ (VNO), which is a paired structure lying
between the hard palate and the nostrils and is used for the detection of social odours,
particularly scent-marks left by other cats. The number of functional genes for VNO
receptors (30 in the cat, compared with nine in the dog) indicates that this organ is highly
discriminating and may well, for example, be used to distinguish between the scent-marks of
different individuals (Montague et al., 2014). Olfaction is also essential to mother–kitten

bonding and communication (Mermet et al., 2008). Disturbances to the olfactory
environment in and around the home can be the cause of some problematic behaviour in
adult cats, such as indoor urination.
Vision is adapted to low-light conditions: colour perception is rudimentary, and the
monochromatic rods that pack the retina enable cats to see clearly in near darkness. The eyes
themselves are disproportionately large, in order to gather as much light as possible: as a
consequence, the lens cannot be distorted to focus at close range, and the mobile whiskers
can be swivelled forward to form a tactile impression of objects that come within a few
centimetres of the cat’s snout. Cats’ ears are sensitive well into the ultrasound, enabling them
to detect the squeaks of small rodents that are nigh-on inaudible to our own ears. This is
actually unremarkable, given that the hearing range of mammals scales up with the physical
size of the eardrum; it is the cat’s ability to detect the low-pitched frequencies in (for
example) male human voices that is unexpected: the lower end of their ten-and-a-half octave
range, which beats ours by more than an octave, has been ascribed to a resonating chamber
behind the eardrum (Tuck-Lee et al., 2008; for further information on the senses in general,
see Chapter 2 of Bradshaw et al., 2012).
The ways in which cats use the information gathered by their sense organs are still poorly
understood (Shreve and Udell, 2015). Working memory, as measured by how long cats
remember the location of an object that has disappeared from view, is rather short, declining
after 1 s and disappearing by 60 s. Their understanding of physics seems rudimentary, based
on their inability to distinguish which of two strings is attached to an otherwise inaccessible
food treat (Whitt et al., 2009). However, their long-term memory appears to be highly
developed, especially for social cues such as odours (Nakabayashi et al., 2012). The
neurohormonal mechanisms underlying domestic cats’ social plasticity, unexpected in an
ancestrally solitary species (see below), are unknown.
17.5 Social Behaviour and Communication

The main difference between the wild F. silvestris and the domestic Felis catus is in social
behaviour, both towards other species (covered later in this chapter) and towards members of
its own species. F. silvestris is solitary and territorial, apparently even when its food is
sufficiently clumped to permit the formation of a colony; the largest groups are therefore
females with dependent offspring, which persist for only a few months. F. catus is capable of
living at much higher density than its ancestor (Liberg et al., 2000); this was presumably
selected for during domestication, when granaries became large enough to attract more pests
than one or two cats could control (Bradshaw, 2016). This selection must have been more-or-
less complete before the Egyptians were able to breed large numbers of cats in their temples.
Our knowledge of the social behaviour of cats is derived largely from studies of free-
ranging and feral groups formed around predictable and plentiful supplies of food, such as
farmyards and fishing villages (Kerby and Macdonald, 1988). Here, cooperative behaviour is
observed between females, usually close relatives; the most likely explanation for the
establishment of these groups is an enhanced tolerance for female offspring, possibly selected
for during domestication. Males rarely show any cooperative behaviour, and are often
nomadic, but they may remain attached to a colony, avoiding conflict with older individuals
until they become strong and experienced enough to compete for mating opportunities.
The collaboration between females includes cooperative breeding, in which nests are
shared, kittens are nursed and fed indiscriminately (Fig. 17.1), and there is cooperative
defence of the core area against unfamiliar cats of both sexes. Unfamiliar mature males may
pose a risk of infanticide, although the extent to which this is an important cause of mortality
in this species is unclear (Bonanni et al., 2007). Like lions, but unlike wolves, all females
within the group breed more or less simultaneously, without any reproductive suppression by
‘dominant’ individuals. Cooperation may be maintained by two tactile behaviour patterns,
allorubbing (Kerby and Macdonald, 1988; Fig. 17.2) and allogrooming, although neither of
these has been studied as extensively as they have been in primates.
Fig. 17.1. Kittens from three mothers suckling from one.

Fig. 17.2. Feral cats allorubbing. (Copyright Mel Pressey 2013, used with permission.)
The signalling repertoire of cats largely reflects their sensory abilities and their origins as
solitary crepuscular animals (Brown and Bradshaw, 2014). Scent-marks, efficient for durable
marking of territorial boundaries or locations of particular significance, are produced in: (i)
urine, especially where this is spray-marked (Fig. 17.3); (ii) faeces, if left uncovered; and (iii)
cheek-gland secretions deposited at head height on projecting objects such as walls (Fig.
17.4). Reproductively active male cats (‘toms’) scent-mark particularly frequently and
pungently, their urine containing high levels of sulfur-containing compounds that may be an
indicator of quality. Social odours are perceived through both olfactory receptors and those in
the VNO, which cats bring into use through the ‘flehmen’ posture in which the mouth is
opened slightly and the top lip is retracted.
Fig. 17.3. Male cat spray-marking. (Copyright Alan Peters 2013, used with permission. From
Cat Sense, Penguin/Basic Books.)
Fig. 17.4. Cat scent-marking with its cheek glands. (Copyright David Joyce 2009, shared via
Creative Commons.)
While humans have difficulty in appreciating their complexity and richness, cats
undoubtedly rely a great deal on odours to inform themselves about their environment, both
social and physical. Feral and other free-ranging cats are generally silent, restricting their use
of readily audible vocalizations to sexual and aggressive encounters, and mother–kitten
interactions. By contrast, pet cats, especially Siamese and related breeds, are much more
vocal. In particular, they use the miaow much more frequently; this is probably a learned
response, rewarded by humans by feeding, opening doors and so on. Some cats appear to
have a repertoire of different miaows, each of which is reserved for a different context (Ellis
et al., 2015). Purring, the other characteristic vocalization of the pet cat, is a care-eliciting
signal; there is no scientific evidence that it signifies the emotion of ‘contentment’ (Ellis,
2013).
As predicted from its ancestral solitary lifestyle, the domestic cat’s repertoire of visual
signals is restricted compared with that of the dog. The tail held vertically is a precursor to
allorubbing and allogrooming and thus indicates affiliative intent (Cafazzo and Natoli, 2009;
see Fig. 8.3 in Bradshaw et al., 2012). On the head, the pinnae (external ears) are the most
expressive structures, pushed forward and erected to indicate confidence, and backwards to
indicate intention to withdraw. Flattening of the ears indicates that the cat is expecting to
engage in fighting. Pulled back and flattened is therefore a defensive posture, often
accompanied by hissing. In agonistic encounters, cats will stand at full height, turn side-on
and raise the fur on the back, to maximize their apparent size. Lashing of the tail, often
accompanied by growling, is also a component of aggression. Individuals that are less
confident may hunch their heads into their shoulders and flatten their bodies into the ground.
Once two cats have begun to threaten one another, they find it difficult to de-escalate because
they lack signals that indicate the acceptance of defeat: the loser will usually have to creep
away very slowly, avoiding any rapid movement that might trigger a chase on the part of the
more confident individual.
Cats within the same social group often rub on one another (allorubbing; Fig. 17.2), and
the same behaviour pattern is frequently directed towards owners or displaced on to objects
near to people. As well as being a tactile signal, rubbing inevitably results in the exchange of
some scent, although whether this is socially significant is not yet understood. The exchange
of rubbing can be markedly asymmetric within a given pair of cats, leading to the suggestion
that it may signify subordinate status between individuals that know one another well,
although it is not a submissive signal in the sense that it is never performed in aggressive
contexts. Mutual (allo)grooming is, however, sometimes associated with mild aggression
(van den Bos, 1998) and may therefore be an attempt to assert dominance, the aggression
resulting if this is disputed by the recipient of the grooming.
17.6 Courtship and Reproductive Behaviour

Female cats can be affiliative towards males at any time of year (Kerby and Macdonald,
1988), but reject all sexual advances except during the breeding season(s), typically, in the
northern hemisphere, in late winter and mid-summer. In pro-oestrus, females first increase
their home ranges and then begin to increase their rate of rubbing on objects. At this stage
they presumably also emit odours that are attractive to males, which in turn show an abrupt
increase in the size of the area they patrol to encompass the female’s range. As a result,
several males may congregate around a female, but at this point she usually refuses to mate
with any of them. Surprisingly, overt competition between males in such aggregations is
usually low-key or non-existent (Liberg et al., 2000), although on occasion they will be
dominated by a single male. The female performs displays of rolling, purring, stretching and
rhythmic protraction and retraction of the claws; after a day or two she comes into oestrus
and indicates her sexual receptivity by adopting the lordosis posture in between bouts of
rolling. Typically she will be polyandrous, i.e. mate with several of the males that have
aggregated around her, including the ‘dominant’ individual if there is one, but not
exclusively. The attractiveness of the female for several days before she is sexually receptive
suggests the function of assessing male quality, but this is not supported by the rather
indiscriminate mating that follows. It has been suggested that this and some other paradoxes
in cat behaviour may be explained by the relatively short time, in evolutionary terms, that
they have lived at high density, such that the mating system is not yet optimized (Liberg et
al., 2000).
Apart from such details, it is apparent that most cats have retained wild-type sexual
behaviour, enabling them to cross-breed with fully wild F. silvestris (Daniels et al., 1998) and
with feral cats. There may therefore be considerable gene flow between feral and domestic
populations (Fig. 17.5), with different selection pressures operating on each. As neutering of
the pet population becomes more prevalent, so a higher proportion of genes from the largely
un-neutered ferals will enter the pet population; since the adaptations required for survival as
a feral may not be identical to those required to be a successful pet, this trend, if continued,
may have unintended consequences for the pet population (Bradshaw et al., 1999).
Fig. 17.5. Flow of genes between cat populations (arrows) with the main external factors
controlling reproductive success alongside each.
17.7 Dominance
It is sometimes suggested that group-living cats exhibit a ‘dominance hierarchy’, based on
ritualized signals, and including the possibility of coalitions between weaker individuals to
displace dominants from valued resources, such as occurs in primate groups (Crowell-Davis
et al., 2004). It is self-evident that within most multi-cat households or feral groups there is a
loose ‘peck order’ in which some individuals have priority of access to food. However, there
are differences of opinion as to which of the available signals indicate submission, for
example allorubbing (Kerby and Macdonald, 1988) or defensive behaviour (Natoli et al.,
2001). Different ‘hierarchies’ emerge from the same animals depending upon which
behaviour patterns are used to construct them. If a hierarchy does actually exist in the minds
of the cats themselves, it appears (see above) to play little role in sexual or reproductive
behaviour. Moreover, since the cat is derived from a solitary ancestor, the advanced social
cognition needed to maintain sophisticated hierarchical systems would have to have been
entirely selected for by domestication. The ‘peck orders’ that are observed may be more
simply explained at the level of the dyad (pairwise relationship), using Parker and
Rubenstein’s (1974) resource holding potential (RHP) model. In this conceptual framework,
each animal assesses the other’s apparent fighting ability in relation to its own and estimates
the risks of escalating an encounter, counterbalanced by the benefit of winning. Thus the
motivation to gain the resource under dispute plays an important role in determining the
outcome of an encounter, explaining why a well-fed ‘despot’ will often allow a hungrier cat
to feed. This approach can be applied equally to first encounters as to an established
relationship, and does not require any complex cognitive abilities, such as the capacity to
comprehend and remember third-party relationships. It is also widely applied to species
which base their sociality on a territorial system (e.g. Barlow et al. 1986), which domestic
cats appear to retain or attempt to retain even when they are living at high densities, and may
therefore be more explanatory for cats than the ‘dominance hierarchy’ concept.
17.8 Ontogeny of Behaviour

The behavioural development of kittens is conventionally divided into four phases: (i) the
prenatal period; (ii) the neonatal period which lasts until the end of the second week post-
gestation, during which the kitten is relatively helpless and not yet sensorially competent;
(iii) the socialization period from the third to eighth weeks, when the kitten actively engages
in, and learns about, social relationships; and (iv) the juvenile period, when it becomes
increasingly independent, ending with the transition to adulthood at the onset of sexual
maturity. However, learning has already begun in the prenatal period, as shown by kittens’
preferences for food similar in flavour to that fed to the mother during pregnancy (Becques et
al., 2010).
Kittens are born at an early stage of development, blind, virtually deaf and incapable of
thermoregulation, although the senses of touch, taste and smell are quite well developed.
Accordingly, they use mainly odour and tactile cues on the mother’s abdomen to reflexively
locate a nipple and begin to suckle. Suckling is often accompanied by purring, presumably a
signal to the mother that inhibits her from moving off. Initially, kittens cannot walk and pull
themselves along with rowing motions of the forelimbs, orientating themselves largely by
odour. The hindlimbs mature during the third and fourth weeks of life, enabling the kitten to
start walking and then running; full coordination is achieved in the seventh week. Voluntary
urination and defecation are not possible until about the fourth week, prior to which the
mother disposes of all the kittens’ wastes during grooming.
The kitten’s hearing develops gradually, as the ear canals open up and the external ears
(pinnae) unfold and become upright during the second week of life. Vocalizations made by
other cats elicit more interest than other sounds and become integrated with locomotory
responses during the fourth week; maternal calls are approached, agonistic vocalizations are
avoided.
The timing of the opening of the eyes is very variable, occurring almost any time in the
first fortnight of life, depending on genetic factors, the sex of the kitten and possibly the level
of light in the environment. The optic fluid, initially cloudy, clears gradually, allowing
improved image formation. The visual cortex develops in parallel with the eyes; initially
many of the cells respond to both eyes, but rapidly become specialized to one or the other,
although retaining connections to allow binocular vision. This process is not spontaneous, but
requires visual experience, as does the development of specialized outline- and movement-
detection neurons that are thought to be an essential part of the cat’s prey-detection systems.
As for binocular vision, the initial cells are rather non-specific, but these guide their own
replacement by more specialized cells, specific to vertical or horizontal edges, circular areas
of contrast and movement in various dimensions and directions. If the initial non-specific
neurons are not stimulated, they degenerate spontaneously, such that the visual system cannot
develop from scratch after the kitten is about 3 months old.
From the fourth week of life onwards, kittens spend much of their waking time playing.
Initially, much of this is directed at littermates, but this may simply be because these are the
most interesting objects in the den, apart from the mother. By 7 weeks of age, object and
social play have become distinct, not only in terms of the target they are directed at, but also
in the behaviour patterns employed. Play with objects involves many patterns that are
subsequently incorporated into hunting behaviour, while signals such as the open-mouth,
side-step and face-off (West, 1974) are used mainly or exclusively in social play. Subsequent
development of play is complex and varies from one litter and individual to another
(Bateson, 2014), as might be expected for a ‘luxury’ behaviour, opportunities for which will
be affected by the quality of the environment the cat finds itself in.
The behaviour of the mother is crucial to the success of the litter, initially in selecting a
suitable denning site, since kittens have to be kept warm and dry for the first few weeks of
life if they are to survive. Mothers do not build a nest as such, but they usually choose a well-
protected location in which to give birth, and may temporarily become aggressive towards
other cats or household animals, presumably to reduce the risk of predation. After a few
weeks, the kittens may be moved, one at a time, to a new den, presumably for reasons of
hygiene and/or to reduce ectoparasite transmission. The mother takes advantage of the
‘scruff’ reflex: grasping the kittens by the loose skin behind their necks causes them to curl
up but inhibits all other behaviour. The same reflex is used by male cats to subdue females at
intromission, and as a benign method for immobilizing cats of all ages (Pozza et al., 2008).
Throughout the period up to weaning, the mother’s behaviour provides essential elements
for the kittens’ development; without her involvement (or alternatively that of a substitute
human ‘parent’), they are likely to develop a range of physical, emotional and behavioural
abnormalities (Bateson, 2014). For example, in the wild she will begin to bring half-killed
prey back to the nest once the kittens are about 4 weeks old: this is an important component
in their acquisition of hunting skills. Soon after this she will begin to withdraw from the
kittens for increasing periods of time, stimulating weaning; as a result, there may be a
temporary check in the weight gain of the kittens. Lactation is metabolically costly, probably
critically so for a mother that is subsisting by hunting, and so it pays the mother to wean her
kittens as early as possible so that she is not unduly weakened herself. If she has been
underweight during gestation, she may become aggressive towards her kittens at this point.
The optimum strategy for each kitten, on the other hand, is to prolong lactation and thereby
not expose themselves to the risks of leaving the den until they are older and stronger.
However, this ‘maternal–infant conflict’ is apparent only in large litters, and lasts for only a
few days, after which the kittens continue to suckle for several more weeks, albeit at a lower
rate than before.
17.9 Socialization
The term ‘socialization’ is used rather loosely to describe a diverse set of developmental
processes whereby young animals learn how to both identify and interact with members of
their own species and also how to interact with members of other species, almost always
man. Some, but probably not all, of these processes are analogous to ‘imprinting’ and take
place most readily during ‘sensitive periods’ (Turner, 2000). In the domestic cat, the sensitive
period for socialization to humans begins at about 2 weeks of age, as soon as the relevant
sensory systems become competent; domestic kittens which have not been handled by people
until they are 8 or more weeks old are subsequently much less sociable to people, indicating
that this sensitive period may spontaneously come to an end at about that time, probably
because the kitten then becomes fearful of the unfamiliar. Kittens handled from 2 weeks of
age are friendlier towards people at 5 weeks old than kittens not handled until that point, but
if both are handled from 5 weeks onwards they become behaviourally very similar by 6
months of age (Lowe and Bradshaw, 2002). Very early handling is therefore facilitating, but
not essential, for socialization. Quantity and quality of handling also have long-term effects
on temperament (Casey and Bradshaw, 2008). For kittens which have been adequately
socialized before 8 weeks of age, learning about how to interact with people appears to
continue for several more weeks (McCune, 1995; Lowe and Bradshaw, 2001, 2002), into the
juvenile period. The sensitive period itself appears to be terminated by some internal event,
in that there is no evidence that it continues if positive social interaction with people does not
occur by week 8; it could be terminated by a developmental clock, or by the completion of a
sensitive period for intraspecific socialization.
It is generally assumed that intraspecific socialization, i.e. to other cats, occurs
simultaneously with socialization to people (Bateson, 2014). Since the ancestral species is
solitary, intraspecific socialization in the domestic cat presumably evolved to establish social
bonds with the mother and littermates, and to direct subsequent sexual preferences.
Anecdotally, cats vary considerably and persistently in their tolerance of conspecifics
(Mertens and Schär, 1988), arguing that intraspecific sociability is a stable characteristic of
individuals. Possible underlying genetic influences may include the ‘boldness’ trait that
affects socialization to humans (McCune, 1995); ontogeny is a more likely explanation for
the greater tolerance of littermates for close confinement with each other compared with
cohabiting but unrelated cats (Bradshaw and Hall, 1999).
The sensitive period for bonding with the mother (filial imprinting) appears to peak in the
fourth and fifth weeks (Schneirla and Rosenblatt, 1961), which is compatible with
simultaneous intra- and interspecific socialization. However, Guyot et al. (1980) suggested
that interaction with littermates may be more important in the development of cooperative
social interactions than interaction with the mother. Social play usually declines after 16
weeks of age, suggesting that a sensitive period for the formation of relationships with
individuals other than the mother comes to an end at about this age. Subsequent tolerance of
other cats could be based upon the memorization of the characteristics of individuals, as
presumably occurs earlier with the mother, or a general tendency to be tolerant of all
unfamiliar individuals if a wide variety of cats has been encountered during the sensitive
period. Functionally, both of these seem plausible: based on the pattern of cooperation
between related females, individuals encountered in the first few months of life are likely to
be kin and therefore bring mutual benefit to subsequent cooperation; kittens meeting non-
siblings during the same period are likely to have been born in an environment where food is
both plentiful and reliable enough to favour group-living over territoriality in subsequent
years also.
As in the dog, there is no evidence to indicate that ‘socialization’ to humans interferes
with the development of intraspecific social behaviour; kittens raised with both conspecifics
and humans subsequently direct their sexual behaviour only towards conspecifics.
The quality and quantity of socialization received by an individual is a major contributor
to its ‘personality’, a set of behavioural biases which are repeatable over time and across
situations, including the shy/bold continuum that is evident in animals from many taxa.
Genetic influences are also detectable in cats (McCune, 1995); experimentally these are
easiest to detect as paternal effects, because the male plays no part in rearing kittens, but it is
assumed that mothers have an equivalent genetic input, in addition to direct and indirect
influences on what their kittens learn (McCune et al., 1995). One specific gene, which codes
for a receptor for the hormone oxytocin (OXTR), has recently been identified as having an
influence on a cat’s personality (Arahori et al., 2016). Cat ‘personalities’ appear to become
stable after about 2 years of age (Lowe and Bradshaw, 2002), prior to which they remain
open to environmental influences and learning opportunities.
17.10 The Ethology of Cat–Human Relationships

Cat–human relationships, although they have not received the same level of attention as
interactions between humans and dogs, have been studied from a wide range of perspectives,
only some of which have an ethological component. Those that fall outside the scope of this
book include the supposed influences of cat ownership on human health, which now appear
to be largely coincidental rather than causal (Utz, 2014), attachment and social support
(Stammbach and Turner, 1999) and the psychology of feeding stray cats (Wald et al., 2013).
Ethological (observational) methods have occasionally been used to study cat–owner
relationships (summarized in Turner, 2000), but the interpretation has usually been based on
psychological constructs.
Cat behaviour ‘problems’, as presented clinically, can usefully be divided into those
which are species-typical behaviour performed in an inappropriate context and those which
involve some kind of pathology, although there is overlap between these (see Chaptes 11 and
12 of Bradshaw et al., 2012). The most common behaviour presented clinically is
inappropriate elimination (‘house soiling’; Bamberger and Houpt, 2006) and this is often
‘normal’ scent-marking behaviour performed in an unwelcome place. Other cat behaviour
‘problems’ that derive from normal ethology include territorial aggression, aggression
towards cats within an artificially constituted social group, defensive aggression towards
owners, withdrawal and hiding behaviour, scratching household objects and unwanted
predatory behaviour (see Horwitz and Mills, 2010 for further information). Psychological
stress caused by factors such as conflict with other cats can readily lead to or exacerbate
physical illnesses such as cystitis (Amat et al., 2016).
References
Adamec, R.E. (1976) The interaction of hunger and preying in the domestic cat (Felis catus): an adaptive hierarchy?
Behavioural Biology 18, 263–272.
Amat, M., Camps, T. and Manteca, X. (2016) Stress in owned cats: behavioural changes and welfare implications. Journal
of Feline Medicine and Surgery 18, 577–586. Available at: http://dx.doi.org/10.1177/1098612X15590867 (accessed 24
April 2017).
Arahori, M., Hori, Y., Saito, A., Chijiiwa, H., Takagi, S., Ito, Y., Watanabe, A., Inoue-Murayama, M. and Fujita, K. (2016)
The oxytocin receptor gene (OXTR) polymorphism in cats (Felis catus) is associated with ‘Roughness’ assessed by
owners. Journal of Veterinary Behavior: Clinical Applications and Research 11, 109–112. Available at:
http://dx.doi.org/10.1016/j.jveb.2015.07.039 (accessed 24 April 2017).
Baker, P.J., Molony, S.E., Stone, E., Cuthill, I.C. and Harris, S. (2008) Cats about town: is predation by free-ranging pet cats
Felis catus likely to affect urban bird populations? Ibis 150(Suppl. 1), 86–99.
Bamberger, M. and Houpt, K.A. (2006) Signalment factors, comorbidity, and trends in behavior diagnoses in cats: 736 cases
(1991–2001). Journal of the Veterinary Medical Association 229, 1602–1606.
Barlow, G.W., Rogers, W. and Fraley, N. (1986) Do Midas cichlids win through prowess or daring? Behavioral Ecology and
Sociobiology 19, 1–8.
Bateson, P. (2013) Behavioural development in the cat. In: Turner, D.C. and Bateson, P. (eds) The Domestic Cat: The
Biology of its Behaviour, 3rd edn. Cambridge University Press, Cambridge, pp. 11–26.
Becques, A., Larose, C., Gouat, P. and Serra, J. (2010) Effects of pre- and postnatal olfactogustatory experience on early
preferences at birth and dietary selection at weaning in kittens. Chemical Senses 35, 41–45. Available at:
http://dx.doi.org/10.1093/chemse/bjp080 (accessed 24 April 2017).
Biben, M. (1979) Predation and predatory play behaviour of domestic cats. Animal Behaviour 27, 81–94.
Bonanni, R., Cafazzo, S., Fantini, C., Pontier, D. and Natoli, E. (2007) Feeding-order in an urban feral domestic cat colony:
relationship to dominance rank, sex and age. Animal Behaviour 74, 1369–1379.
Bonnaud, E., Bourgeois, K., Vidal, E., Kayser, Y., Tranchant, Y. and Legrand, J. (2007) Feeding ecology of a feral cat
population on a small Mediterranean island. Journal of Mammalogy 88, 1074–1081.
Bradshaw, J.W.S. (2016) Sociality in cats: a comparative review. Journal of Veterinary Behavior: Clinical Applications and
Research 11, 113–124. Available at: http://dx.doi.org/10.1016/j.jveb.2015.09.004 (accessed 24 April 2017).
Bradshaw, J.W.S. and Hall, S.L. (1999) Affiliative behaviour of related and unrelated pairs of cats in catteries: a preliminary
report. Applied Animal Behaviour Science 63, 251–255.
Bradshaw, J.W.S., Horsfield, G.F., Allen, J.A. and Robinson, I.H. (1999) Feral cats: their role in the population dynamics of
Felis catus. Applied Animal Behaviour Science 65, 273–283.
Bradshaw, J.W.S., Casey, R.A. and Brown, S.L. (2012) The Behaviour of the Domestic Cat, 2nd edn. CAB International,
Wallingford, UK.
Brown, S.L. and Bradshaw, J.W.S. (2014) Communication in the domestic cat: within- and between-species. In: Turner, D.C.
and Bateson, P. (eds) The Domestic Cat: The Biology of its Behaviour, 3rd edn. Cambridge University Press, Cambridge,
pp. 37–59.
Cafazzo, S. and Natoli, E. (2009) The social function of tail up in the domestic cat (Felis silvestris catus). Behavioural
Processes 80, 60–66. Available at: http://dx.doi.org/10.1016/j.beproc.2008.09.008 (accessed 24 April 2017).
Cameron-Beaumont, C., Lowe, S.E. and Bradshaw, J.W.S. (2002) Evidence suggesting preadaptation to domestication
throughout the small Felidae. Biological Journal of the Linnean Society 75, 361–366.
Casey, R.A. and Bradshaw, J.W.S. (2008) The effects of additional socialisation for kittens in a rescue centre on their
behaviour and suitability as a pet. Applied Animal Behaviour Science 114, 196–205. Available at:
http://dx.doi.org/10.1016/j.applanim.2008.01.003 (accessed 24 April 2017).
Crowell-Davis, S.L., Curtis, T.M. and Knowles, R.J. (2004) Social organisation in the cat: a modern understanding. Journal
of Feline Medicine and Surgery 6, 19–28.
Daniels, M.J., Balharry, D., Hirst, D., Kitchener, A.C. and Aspinall, R.J. (1998) Morphological and pelage characteristics of
wild living cats in Scotland: implications for defining the ‘wildcat’. Journal of Zoology 244, 231–247.
Diamond, J. (1997) Guns, Germs and Steel: A Short History of Everybody for the Last 13,000 Years. Jonathan Cape Ltd,
London.
Driscoll, C.A., Menotti-Raymond, M., Roca, A.L., Hupe, K., Johnson, W.E., Geffen, E., Harley, E.H., Delibes, M., Pontier,
D., Kitchener, A.C. et al. (2007) The Near Eastern origin of cat domestication. Science 317, 519–523.
Ellis, S. (2013) Purrfect sounds. iCatCare Magazine 3, 44–46.
Ellis, S.L.H., Swindell, V. and Burman, O.H.P. (2015) Human classification of context-related vocalizations emitted by
familiar and unfamiliar domestic cats: an exploratory study. Anthrozoös 28, 625–634. Available at:
http://dx.doi.org/10.1080/08927936.2015.1070005 (accessed 24 April 2017).
Guyot, G.W., Cross, H.A. and Bennett, T.L. (1980) Early social isolation of the domestic cat: responses to separation from
social and nonsocial rearing stimuli. Developmental Psychobiology 13, 309–315.
Hall, S.L. (1998) Object play by adult animals. In: Bekoff, M. and Byers, J.A. (eds) Animal Play: Evolutionary,
Comparative and Ecological Perspectives. Cambridge University Press, Cambridge, pp. 45–60.
Horwitz, D. and Mills, D.S. (eds) (2010) BSAVA Manual of Canine and Feline Behavioural Medicine, 2nd edn. British
Small Animal Veterinary Association, Quedgeley, UK.
Kerby, G. and Macdonald, D.W. (1988) Cat society and the consequences of colony size. In: Turner, D.C. and Bateson, P.
(eds) The Domestic Cat: The Biology of its Behaviour. Cambridge University Press, Cambridge, pp. 67–82.
Liberg, O., Sandell, M., Pontier, D. and Natoli, E. (2000) Density, spatial organisation and reproductive tactics. In: Turner,
D.C. and Bateson, P. (eds) The Domestic Cat: The Biology of its Behaviour, 2nd edn. Cambridge University Press,
Lilith, M., Calver, M. and Garkaklis, M. (2010) Do cat restrictions lead to increased species diversity or abundance of small
and medium-sized mammals in remnant urban bushland? Pacific Conservation Biology 16, 162–172.
Lowe, S.E. and Bradshaw, J.W.S. (2001) Ontogeny of individuality in the domestic cat in the home environment. Animal
Behaviour 61, 231–237.
Lowe, S.E. and Bradshaw, J.W.S. (2002) Responses of pet cats to being held by an unfamiliar person, from weaning to three
years of age. Anthrozoös 15, 69–79.
McCune, S. (1995) The impact of paternity and early socialisation on the development of cats’ behaviour to people and
novel objects. Applied Animal Behaviour Science 45, 109–124.
McCune, S., McPherson, J.A. and Bradshaw, J.W.S. (1995) Avoiding problems: the importance of socialisation. In:
Robinson, I. (ed.) The Waltham Book of Human–Animal Interaction: Benefits and Responsibilities of Pet Ownership.
Elsevier Science Ltd, Oxford, pp. 71–86.
Menotti-Raymond, M., David, V.A., Pflueger, S.M., Lindblad-Toh, K., Wade, C.M., O’Brien, S.J. and Johnson, W.E. (2008)
Patterns of molecular genetic variation among cat breeds. Genomics 91, 1–11. Available at:
http://dx.doi.org/10.1016/j.ygeno.2007.08.008 (accessed 24 April 2017).
Mermet, N., Coureaud, G., McGrane, S. and Schaal, B. (2008) Odour-guided social behaviour in newborn and young cats:
an analytical survey. Chemoecology 17, 187–199. Available at: http://dx.doi.org/10.1007/s00049-007-0384-x (accessed
24 April 2017).
Mertens, C. and Schär, R. (1988) Practical aspects of research on cats. In: Turner, D.C. and Bateson, P. (eds) The Domestic
Cat: The Biology of its Behaviour. Cambridge University Press, Cambridge, pp. 179–190.
Montague, M.J., Li, G., Gandolfi, B., Khan, R., Aken, B.L., Searle, S.M.J., Minx, P., Hillier, L.W., Koboldt, D.C., Davis,
B.W. et al. (2014) Comparative analysis of the domestic cat genome reveals genetic signatures underlying feline biology
and domestication. Proceedings of the National Academy of Sciences USA 111, 17230–17235. Available at:
http://dx.doi.org/10.1073/pnas.1410083111 (accessed 24 April 2017).
Nakabayashi, M., Yamaoka, R. and Nakashima, Y. (2012) Do faecal odours enable domestic cats (Felis catus) to distinguish
familiarity of the donors? Journal of Ethology 30, 325–329. Available at: http://dx.doi.org/10.1007/s10164-011-0321-x
Natoli, E., Baggio, A. and Pontier, D. (2001) Male and female agonistic and affiliative relationships in a social group of farm
cats (Felis catus L.). Behavioural Processes 53, 137–143.
Parker, G.A. and Rubenstein, D.I. (1974) Role assessment, reserve strategy, and acquisition of information in asymmetrical
animal conflicts. Animal Behaviour 29, 221–240.
Pozza, M.E., Stella, J.L., Chappuis-Gagnon, A.-C., Wagner, S.O. and Buffington, C.A.T. (2008) Pinch-induced behavioral
inhibition (‘clipnosis’) in domestic cats. Journal of Feline Medicine and Surgery 10, 82–87.
Randi, E. and Ragni, B. (1991) Genetic variability and biochemical systematics of domestic and wildcat populations (Felis
silvestris: Felidae). Journal of Mammalogy 72, 79–88.
Schneirla, T.C. and Rosenblatt, J.S. (1961) Behavioral organisation and genesis of the social bond in insects and mammals.
Journal of Orthopsychiatry 31, 223–253.
Serpell, J.A. (2014) Domestication and history of the cat. In: Turner, D.C. and Bateson, P. (eds) The Domestic Cat: The
Biology of its Behaviour, 3rd edn. Cambridge University Press, Cambridge, pp. 83–100.
Shreve, K.R.V. and Udell, M.A.R. (2015) What’s inside your cat’s head? A review of cat (Felis silvestris catus) cognition
research past, present and future. Animal Cognition 18, 1195–1206. Available at: http://dx.doi.org/10.1007/s10071-015-
0897-6 (accessed 24 April 2017).
Stammbach, K.B. and Turner, D.C. (1999) Understanding the human–cat relationship: human social support or attachment.
Anthrozoös 12, 162–168.
Todd, N.B. (1977) Cats and commerce. Scientific American 237(November), 100–107.
Tuck-Lee, J.P., Pinsky, P.M., Steele, C.R. and Puria, S. (2008) Finite element modeling of acousto-mechanical coupling in
the cat middle ear. Journal of the Acoustical Society of America 124, 348–362.
Turner, D.C. (2000) The human–cat relationship. In: Turner, D.C. and Bateson, P. (eds) The Domestic Cat: The Biology of its
Behaviour, 2nd edn. Cambridge University Press, Cambridge, pp. 193–206.
Utz, R.L. (2014) Walking the dog: The effect of pet ownership on human health and health behaviours. Social Indicators
Research 116, 327–339. Available at: http://dx.doi.org/10.1007/s11205-013-0299-6 (accessed 24 April 2017).
van den Bos, R. (1998) The function of allogrooming in domestic cats (Felis silvestris catus); a study in a group of cats
living in confinement. Journal of Ethology 16, 1–13.
Wald, D.M., Jacobson, S.K. and Levy, J.K. (2013) Outdoor cats: identifying differences between stakeholder beliefs,
perceived impacts, risk and management. Biological Conservation 167, 414–424. Available at:
http://dx.doi.org/10.1016/j.biocon.2013.07.034 (accessed 24 April 2017).
West, M.J. (1974) Social play in the domestic cat. American Zoologist 14, 427–436.
Whitt, E., Douglas, M., Osthaus, B. and Hocking, I. (2009) Domestic cats (Felis catus) do not show causal understanding in
a string-pulling task. Animal Cognition 12, 739–743. Available at: http://dx.doi.org/10.1007/s10071-009-0228-x
Zoran, D.L. and Buffington, C.A.T. (2011) Effects of nutrition choices and lifestyle changes on the well-being of cats, a
carnivore that has moved indoors. Journal of the American Veterinary Medical Association 239, 596–606. Available at:
http://dx.doi.org/10.2460/javma.239.5.596 (accessed 24 April 2017).
18 Behaviour of Foxes and Mink Kept for Fur
Production
A.L. Hovland , L. Ahola and J. Malmkvist

Foxes
The red fox (Vulpes vulpes) and the arctic fox (Vulpes lagopus, formerly Alopex lagopus) are
closely related, both belonging to the genus Vulpes under the family Canidae (dog family)
within the order Carnivora. The domesticated form of the red fox and the arctic fox is called
the silver fox and the blue fox, respectively. There are three main colour types of red fox,
varying from fiery red to greyish brown and almost black (silver), whereas arctic foxes have
two colour morphs, white and blue. Although crosses between silver and blue foxes are
sometimes produced on farms, these species do not normally interbreed in the wild. Rather,
as they occur in the same habitat red foxes can exclude arctic foxes, as well as kill both adult
and young.
Both fox species are opportunistic predators, scavengers and hoard food for times of
famine. Red foxes’ body weight averages 5 kg for females (vixens) and 6 kg for males. The
smaller arctic foxes weigh between 3 and 5 kg. The red fox is a highly flexible species. It has
the widest geographical distribution of all canids, inhabiting a variety of habitats from city
centres to Arctic tundra and deserts, forests and farmlands. The arctic fox inhabits only the
northern circumpolar area, including the mainland tundra and Arctic islands in Asia and
North America, and is also found on Iceland and along the Fennoscandian mountain ridge. It
typically faces seasonal food scarcity and a harsh, cold climate. Through morphological,
physiological and behavioural adaptations to these environmental conditions it subsists where
red foxes do not survive.
Silver and blue foxes are commercially farmed for the harvest of their thick winter fur.
Annually, the worldwide production of fox fur averages about 7–8 million pelts, of which the
majority originate from blue foxes. Fox pelts are produced in Finland, Norway, Poland and
several other European countries as well as China, Russia, Canada and the USA.
Mink
The American mink (Neovison vison, formerly Mustela vison) belongs to the family of
mustelids within the order Carnivora. Other mustelids count the European mink (Mustela
lutreola), a smaller, more specialized mink, which is native to parts of Europe. We use the
term ‘mink’ in the text for American mink as this is the farmed species. The mink is an
opportunistic predator. The diet changes with the seasonal availability of prey, and mink can
be found in a range of habitats. The territory of mink is typically smaller in areas with high
food abundance, such as around river banks and in estuarine coastal regions. The natural
resources and the yearly reproductive season determine the behavioural activity of mink. The
adult mink lives solitary, except for the mating period and when bringing up the litter. Mink
have one to several dens in their home range. This den is used to store food items, for rest
and during the rearing of the altricial young born in late April/beginning of May in the
northern hemisphere. There are two phases of fur change per year, with the denser winter fur
valued by humans. The American mink is the most common animal farmed for fur
production, with a world production of 70–80 million pelts per year. Today, main mink-
producing countries include China, Denmark, Poland, the USA, Canada and several others;
however, this is subject to changes due to market demand.
18.2 Origin and Domestication

Foxes
The first silver foxes that were kept in ground enclosures for the purpose of producing fur
were bred from wild-caught black foxes on Prince Edward Island, Canada, around 1880
(Forester and Forester, 1982). Descendants of these silver foxes were imported to Norway in
1914 and formed the basis for the silver fox population in Europe and Russia. At the start,
silver foxes were kept under less intensive conditions more resembling their natural settings.
Later, when farms grew bigger, hygiene concerns resulted in the development and use of wire
mesh cages that still are the common housing environment of today’s farmed foxes.
Typically, domesticated foxes have a larger and longer body size and denser fur compared
with their wild counterparts. Silver foxes’ body weight averages 8–9 kg, peaking to a
maximum of 13–14 kg in the largest males before slaughter (pelting) (Fig. 18.1). A variety of
different colour types from red to black (silver), brown and even white can be produced by
directional selection. The domestication has not changed tail or ear length, morphologies
important for social communication for example. Likely, domestication has not altered the
behaviour repertoire of silver foxes but rather modified the animals’ reactivity and stress
susceptibility.
Fig. 18.1. A silver fox with access to a wooden platform. Platforms mounted in the fox cage
come in different shapes and are sometimes used for resting, but also function as an
observation post and a safe place.
The domestication process of silver foxes has taken two main directions: (i) the
domestication occurring during the commercial production of fur; and (ii) through the
scientific domestication experiments initiated by the Russian geneticist Dimitri Belyaev at
the start of the 1960s. For the production of fur, breeding animals are selected mainly for pelt
quality and (body) size as well as litter size at 3 weeks of age. The breeding is usually
controlled by individual farmers as opposed to the more centralized selection programmes
that are common in, for example, cattle and swine. When the size of farms increased with
more animals per unit, the need for safe handling methods that secured the farmers from
being bitten became apparent. Until recently, a metal tong device was therefore used
regularly during handling, thus making selection for confidence and ease of handling less
efficient. Today, increased effort on improving the human–animal relationship by genetic
selection for confidence has been initiated, particularly in the Nordic countries. This work is
motivated by animal welfare concerns and by knowledge from breeding experiments
demonstrating that selection for reduced fear towards humans gave, apart from increased
litter size, more confident foxes that are less stressed by humans (tolerate handling)
compared with standard foxes (Trut, 1999).
Around the same time as the first attempts to farm silver foxes, during the years 1885–
1899, wild-caught arctic foxes were kept on small islands along the coast of Alaska where
they reproduced without human interference. The interest in blue fox production varied in the
beginning, but during 1916–1930 several animals originating from these island foxes were
sold to the USA, Canada, Norway, Finland and Sweden. Later, also foxes from Greenland
and other North Atlantic islands were crossbred with the Alaskan fox (Rochmann, 1969).
Blue foxes were kept in wire mesh cages from the start and their body size and length have
increased gradually during domestication. Different colour morphs vary from pure white to
light grey (blue) and dark grey. The selection criteria for breeding animals are the same as for
silver foxes. The size and shape of blue foxes have, however, been altered significantly more
during domestication. On average, breeding animals weigh between 7 and 10 kg but at
slaughter the heaviest blue foxes can weigh up to 25 kg (of which 40–50% may consist of
adipose tissue) (Fig. 18.2). The large difference in weight change, compared with farmed red
foxes, is probably linked to arctic foxes’ adaptations to survive in marginal habitats, storing
fat as an energy source that it utilizes during food shortage. The housing conditions and
handling procedures of blue foxes, and thus the environmental effects and selective agents
acting on the commercial farm population for about 100 years, have paralleled those for
farmed silver foxes. Breeding for reduced fearfulness towards humans has shown that also
blue foxes’ confidence can be modified by genetic selection (Kenttämies et al., 2002).
Fig. 18.2. A blue fox housed in a traditional cage equipped with a mesh platform, a wooden
stick and straw (top rear left).
Mink
The American mink was first bred in captivity around 1866 in North America according to
the reports cited by Shackelford (1984). The first mink arrived in Europe around 1920 from
the USA and Canada. Since then a range of colour types (more than 30 different, where
approximately 12 main colour types are sold at auction) has been created from the wild type
via directional selection on farms. Besides the fur traits such as colour and quality, selection
for other traits has taken place. The size of modern adult farm mink commonly averages 3–4
kg for males, whereas wild male mink are reported to weigh up to 1.5 kg in the autumn, and
females average half the weight of males both on farms and in nature. Mink breed well in
captivity and those with less good reproduction have probably been excluded from the stock.
Reproductive performance can be affected by the underlying selection that has taken place
over the decades. For instance, the mink has become easier to mate and its fertility has
improved since farming began on a larger scale. Systematic selection for temperament can
lead to higher mating willingness on farms since mink selected for their tendency to approach
humans can be mated a few days earlier than fearful or non-selected farm mink.
The domestication process typically leads to animals with an elevated threshold of fear
towards humans; a change accomplished via both genetic and environmental adaptation
(Price, 1984). Directional selection for more domestic, less fearful mink has proved possible
(Hansen, 1996). The heritability (h2, a measure of the proportion of variance attributable to
genetic variation) for reduced fearfulness and for approaching humans is estimated to be 22–
36% in farm mink. The heritability is highest for males, which generally is the more
explorative sex. Studies have documented generalized effects on fearfulness following
divergent selection for behavioural responses in a simple test of approach-avoidance towards
humans, with concurrent changes in the brain serotonergic system and secretion levels of the
hormones adrenocorticotropin and cortisol (Malmkvist and Hansen, 2002). Thus, directional
selection for less fearful breeding mink may reduce stress and improve welfare on farms. In
some countries farmers are advised to include reduced fearfulness and increased exploration
in their breeding goal. Behavioural selection for temperament is currently enforced by law in
for example Denmark, which accounts for about 25% of the world production. Apart from a
genetic influence, the behaviour of farmed mink can be affected by the housing environment
and various events during the lifetime of the individual.

Foxes
Red foxes are socially flexible animals (facultative social) that also can live in groups of up
to ten adults. The basis for sociality is related both to animal density and resource
availability, and typically occurs in areas where food is abundant and predictable. Social
constellations range from the more common breeding pair to cooperative reproduction and
shared parental duties among several related females (alloparental behaviour). Apart from the
dominant vixen, these females may be offspring from previous years that accept subordinate
social positions; either assisting the nursing of related offspring (helpers) or sometimes also
producing their own cubs. When both dominant males, usually the father of most cubs, and
subordinate males are present during the cub period and onwards, social units consisting of
several adults of both sexes and of varying social status may form (Baker et al., 2004).
Delayed dispersal of young and acceptance of juveniles in the parents’ home area may further
increase the group size. The lack of aggressive encounters and the presence of positive social
interactions, like grooming and play, maintain social bonds between the group members. The
bonds between females are particularly important for the social dynamic, as relatedness to the
dominant female predicts a daughter’s likelihood of staying with the family (Whiteside et al.,
2011). During the autumn, hormonal changes related to maturing combined with sex, status
and affiliation between littermates determine their propensity to disperse (Harris and White,
1992).
Arctic foxes can also form social groups when food is plentiful; a group of 31 foxes
consisting of four adults and three litters was observed on an artificially fed den in Sweden
(Elmhagen et al., 2014). An obvious benefit of sociality is improved guarding that reduces
the overall predation risk (Norén et al., 2012). Typically, these adults are related and may be
offspring from previous litters that never dispersed or later returned home. Older individuals
usually dominate younger ones and males will outcompete same-age females. Both males
and females perform urine/scent-marking at territorial borders. The frequency of this
behaviour seems to be related to sex and dominance status of the individuals.
Foxes have a variety of social communication signals that range from body, tail and ear
postures to gaping signals and various sounds such as the shrieks, whining, gekkering
(stuttering vocalizations) and yelps in red foxes (MacDonald, 1987). Also, olfactory
signalling through urine and faeces, as well as from the anal and precaudal gland, is
important. These signals communicate the motivations and intentions of an individual, and
include a variety of aggressive displays to play postures. Although territorial behaviour may
be peaceful through urine-marking and defecation, fights may occur between unfamiliar
foxes, particularly during the mating season. Body and facial postures used for
communication in arctic foxes resemble those of red foxes (Frafjord, 1993) but the vocal
signals are distinct and different from red fox calls. Both adult and young arctic foxes emit
various types of sound from cackling to serial calls, consisting of a sequence of several short
barks ending with a longer shriek. These barks vary with foxes’ age and sex, serve the
function of establishing and maintaining contact between group members, but variants are
also used to warn about enemies and predators (Kruchenkova et al., 2003). Social displays
between foxes reveal the dominance status of those involved, where for example signals of
submission include low and crawling body postures, flat ears and withdrawn lips, often
accompanied by whining. Mutual friendly contact between animals includes allogrooming
and nibbling of each other’s fur, resting together and play.
Adult foxes kept for fur production are typically housed solitary. Full social contacts
occur during the nursing period between the vixen and her cubs until weaning. Breeding
males are housed solitary throughout their life. Cubs, often littermates, are kept in pairs or
groups during autumn until euthanizing and pelting in November/December. Young cubs and
juveniles are motivated for social contact and interactions are usually peaceful until October
and the onset of puberty, paralleling wild foxes’ natural dispersal time. Then, social tension
increases and agonistic behaviours become more frequent and intense. The development of
social behaviour of young blue foxes parallels that for silver foxes. In adults, studies on silver
foxes have however shown that the motives for seeking full social contact are both
aggressive and amicable (Hovland et al., 2011). When adult silver fox vixens are housed in
triplet groups, aggression during formation of the social hierarchy is common. Group-
housing of adult blue foxes in semi-natural enclosures from autumn throughout the breeding
period has been examined experimentally (Korhonen and Alasuutari, 1995). In these groups,
dominance hierarchies will form, males usually being dominant over females. However, both
age and season affect the dominance relations, and an individual’s social status may change
through the breeding period. For females, timing of oestrus seems to be affected by their
status. Social tension increases during autumn, when aggression between foxes of both sexes
becomes more frequent and serious during the breeding season. When silver fox cubs are
family-housed in earthen enclosures, their activity pattern, dominance relations and social
behaviours parallel those of free-living foxes (Ahola et al., 2000). In general, the
mechanisms related to maturing and dissolution of the social unit (family) are an important
factor affecting the social dynamics in captive fox groups, influenced by how resources such
as resting places and food trays are distributed.
Mink
In nature, the adult mink lives solitarily, except during the mating period and (for dams)
when bringing up the litter. The mink defends its territory against intruders with aggression.
Some tolerance between the male and female mink occurs as the male territory may overlap
that of several females in the wild (Dunstone, 1993). On farms adult breeding mink are kept
individually. After weaning at about 2 months until pelting at 7 months of age, the growing
kits are traditionally kept in pairs consisting of one male and one female. This housing
stimulates the development of natural mating behaviour and reduces abnormal behaviour in a
period where the level of aggression between the two cage-mates is low.
18.4 Sexual Behaviour, Reproduction and Nursing

Foxes
Mating and nursing
Both fox species are monoestrous, seasonal breeders reaching sexual maturity at the age of
8–10 months. As a rule, the breeding pair is considered the basic reproductive unit in foxes.
However, both males and females may seek extra-pair matings and additional females may
give birth in the same denning system, particularly during food surplus and/or low animal
density. In the northern hemisphere, the mating season starts around February for red foxes
and almost 1 month later for arctic foxes, lasting for about 1 month. Gestation lasts 49–55
days, and cubs are born from April to June/July. In the wild, the portion of successfully
reproducing vixens fluctuates from year to year, and under crowded conditions (high fox
density) subordinate vixens do not normally reproduce. In nature, litter size at birth varies,
normally from four to six (up to 12) for the red fox and from six to eight (up to 18) for the
arctic fox, being highly dependent on food availability. Due to differences in diet and litter
size arctic foxes are described as occurring in two ecotypes, coastal foxes and lemming foxes
(Braestrup, 1941). Coastal foxes, living in a more stable coastal environment, typically give
birth to small litters yearly (averaging six cubs). Lemming foxes’ litter size varies from no
reproduction to very large litters, following the cyclic rodent population.
The vixen gives birth in a den providing protection and shelter for both the vixen and her
cubs. This is important as the cubs’ thermoregulation is poor these first weeks. In the wild,
the den may be an underground burrow, sometimes made of clusters of several dens with
numerous entrances and tunnels. Sometimes it is placed in simpler structures; for example,
under boulders of stone, in a single hole in coarse gravel or even under peoples’ patios and
garden sheds. Cubs are born altricial, and start to see and hear from the age of approximately
2 weeks. Most cubs have their eyes completely open 25 days after birth. After emerging from
the den at 3–4 weeks of age, they quickly learn to move around in their living environment.
In wild fox cubs, the sensitive period of primary socialization occurs at the age of 3–7 weeks.
Both fox species move between several dens, potentially as an anti-predator strategy or due
to changes in competition pressure. Although single mothers have been observed, both
parents often take part in cub care. During the 2–3 first weeks after birth when the vixen
spends most of her time with the cubs, provision of food by the male (father), or sometimes
by other foxes, is important. Time spent with the cubs may vary according to food
availability: in years of food scarcity, both parents may spend most of their time away from
the cubs in search for food. In a Canadian study of red foxes (Vergara, 2001) it was observed
that the male spent less than 1 h/day within a 100 m radius from the den. The corresponding
time for the female was around 4 h. The vixen gradually stops nursing her cubs from the age
of 5 weeks onwards, and suckling usually terminates between 7 and 10 weeks after birth. In
some rare cases this process may last until cubs are as much as 14 weeks old.
Like the wild relatives, farmed foxes are still monoestrous seasonal breeders, coming into
heat in late winter. Since a male fox produces enough sperm to fertilize up to 20–30 vixens,
artificial insemination of vixens became common in 1970s. Today, more than 80% of
breeding females are artificially inseminated. Sperm from breeding males are collected on-
farm and fresh samples are used. The vixen is receptive for a few days and detection of top
oestrus is usually aided with a vaginal oestrus detection device. When applying natural
breeding, a male copulates with five to ten vixens. About 2 weeks before birth, the farmer
provides the vixens with a nest box (artificial den); usually a wooden box with an anteroom.
Vixens give birth unassisted and the litter size parallels wild fox populations. However,
occasionally litters with more than 20 cubs are born. Number of successfully weaned cubs
averages three to five in silver foxes and six to eight in blue foxes. The vixen takes care of
her cubs alone. During parturition, the vixen spends 40–60% of the time grooming,
inspecting and congregating her cubs. The first 3 days after birth, about 95% of the time is
spent inside the nest. Of this, between 65% and 80% is allocated for resting and suckling the
cubs (Braastad, 1993; Pyykönen et al., 2005). Fox cubs will whimper loudly if they get
isolated from their littermates. At about 3–4 weeks of age, cubs start to eat solid food
provided outside their nest box. During feeding, the vixen may vocalize to signal that food is
delivered and may carry food to her cubs. The vixens stop nursing their cubs gradually.
Between the age of 7 and 10 weeks, cubs are separated from their mother. The vixen is then
transferred to her own cage and housed solitary through the rest of the year. Often the whole
litter is kept together for some weeks, before separated into sibling pairs or groups.
Reproductive success
In wild red and arctic foxes, the number of cubs surviving to maturation fluctuates from year
to year, depending mostly on food resource competition, and otherwise on disease outbreaks,
hunting intensity and accidents (e.g. road kills). In red foxes, about three out of four juveniles
die before they reach 1 year of age (Macdonald and Reynolds, 2004). For arctic foxes of the
lemming ecotype, cub mortality can reach 100% when the rodent population crashes early in
summer. In dependent cubs, death may sometimes also be caused by attacks from adult foxes
(infanticide) and cannibalism or neglect or inadequate maternal behaviour (Macdonald,
1987). Similarly to wild conditions, some vixens do not reproduce during all years.
Reproductive losses are more common among primiparous vixens (first-time breeders).
About 30% of primiparous vixens do not wean cubs, whereas the corresponding number for
multiparous vixens is about 10%. Depending on several factors, up to 15–30% of cubs may
be lost during the first 2 weeks; after this critical period cub losses are rare. In silver foxes, it
has been observed that vixens may kill or harm one or more of their cubs (maternal
infanticide). This behaviour is more common in primiparous vixens and found to be related
to dominance relations among neighbouring vixens (Bakken, 1993). Infanticidal behaviour is
rarer in blue foxes, occurring in less than 1% of breeding vixens.
Mink
The mink is a seasonal breeder with the dam giving birth to one litter per year. In nature,
males may fight over access to females in heat. Breeding in captivity has increased the litter
size in farmed mink. On farms the average is currently five to six kits weaned per female but
up to 15 kits per litter can occasionally be seen. Mink have induced ovulation, where the
tactile stimulus from the male penis induces the release of eggs for fertilization. In mink the
fertilized eggs implant in the uterus after an embryonic diapause (i.e. delayed implantation)
and the length of gestation is 30 days after implantation (Sundqvist et al., 1989).
Only natural mating is used on mink farms. Typically, farmers keep a ratio of about five
dams per male and aim to get the female mated at least twice during her receptive period,
which in the northern hemisphere is from late February until mid-March. Mate choice
experiments indicate that the female mink can be promiscuous. The act of mating is
characterized by the male holding the female, biting her in the neck and taking up a rounded
body position over her during intromission. The linkage of the pair lasts from a few minutes
to about an hour.
In nature, the female mink does not necessarily complete a nest from scratch prior to
delivery; rather she may locate and take over a nest in dens of suitably sized prey animals,
such as rodents, muskrats and rabbits. Only the female takes care of the litter. On farms, the
birth of the litter takes place inside a nest box late in April/beginning of May. The maternal
behaviour and interaction with kits are extensive in farm mink. If the mink dam is given the
opportunity to nest-build before delivery, this will result in a large nest and reduced stress
and birth problems in combination with increased behaviour directed towards the offspring to
enhance their survival. Maternal care, including nest building, nursing and protection, is
necessary for survival and growth of the altricial offspring (typically 8–10 g at birth) at least
for the first 4–6 weeks of life (Fig. 18.3). Thermoregulatory and motor abilities are poorly
developed during the first weeks (Harjunpää and Rouvinen-Watt, 2004), and eye-opening and
the first signs of hearing occur after postnatal day 28. Kits vocalize intensively which seems
to induce maternal retrieval if the kit comes astray, as well as to increase care within the nest.
Newborn kits can produce complex ultrasonic vocalization (up to 50 kHz), although they
cannot hear early in life (Clausen et al., 2008).
Fig. 18.3. Maternal behaviour in mink includes nest building, licking, grooming and nursing.
The nest box is preferred for rest and nesting during birth of the yearly litter.
The number of weaned young can exceed the number of active teats (usually up to eight)
in the lactating mink dam, thus several young can share the resource. The gradual shift from
mother’s milk to solid food and drinking water begins around 4 weeks of age. The majority
of milk production tissue in the dam decays 6–8 weeks after birth (Møller and Pinkalski,
2015) concurrent with the increasing nutritional independency of the kits. Weaning could be
defined as the stage at which the mother most sharply reduces the time and effort she devotes
to the offspring (Martin, 1984). A study of feral mink in the UK showed that in July (about
10–14 weeks after birth) the kits were alone or in pairs away from the den of birth, and
aggressive maternal interactions were observed. On farms, timing of separation may depend
on litter size and female performance but is generally conducted between 6 and 10 weeks of
age.
18.5 Diurnal Rhythms, Activity Patterns and Foraging

Behaviour
Foxes
Activity, home ranges and movements
In the wild, both fox species are active predominantly during the night (nocturnal) or at dusk
(crepuscular). However, activity occurs also during daylight, especially when cubs are young.
Patterns of activity depend on prey behaviour (i.e. foxes follow their prey’s activity pattern)
and avoidance of competitors and predators (e.g. human presence shifts foxes’ activity
towards nocturnal behaviour) as well as reproductive status, ambient temperature and
photoperiod. Adult foxes spend a considerable amount of time foraging. This is due to the
scattered availability of food and solitary hunting for small prey or food items. Young red fox
cubs that do not yet forage spend their time outside the den on exploration, general moving,
play and rest (Meyer and Weber, 1996). Young arctic fox cubs either play (about 30%) or rest
(about 50%) when parents move away from the den (Frafjord, 1992). In adults, play
behaviour is rarely observed. Both fox species are highly mobile and may thus travel several
thousand kilometres during their lifetime. These movements may include foraging, territorial
marking and matings. Long-range movements are mostly linked to natal dispersal, and
occasionally also to food-related travels. Daily travelling distances in the arctic fox may
range up to a maximum of 90 km with an average travelling rate of 24 km/day (Tarroux et
al., 2010). In the red fox, daily distances travelled are shorter (following from their smaller
home ranges and overall higher population densities), being up to 12 km/day (Meia and
Weber, 1995). For the red fox, the movements are shortest during the cub-rearing period and
longest during mating time and juvenile dispersal (Soulsbury et al., 2011). The longest
dispersal distances documented for arctic fox are more than 2000 km, and red foxes 300 km,
distances being longer for males than for females.
In the wild, the size of foxes’ home ranges varies considerably and may be in the range of
0.45–40 km2 (up to 50 km2) and 4–60 km2 (up to 125 km2) in the red fox and arctic fox,
respectively, depending on food availability, population densities, foxes’ age and
reproductive status. In the red fox, territories (i.e. defended living areas) as well as home
ranges are larger in winter than in summer, and smaller for urban than for rural foxes,
probably reflecting resource availability.
Under farming conditions, the opportunity for activity and movement depends on the
type of housing facilities, animal densities (season) and current laws and regulations.
Traditionally, cage sizes of adult foxes vary between 0.8 and 1.2 m 2, increasing to 2.0–2.4
m2 for breeding vixens. During early winter, foxes are occasionally given access to several
cages when densities are low. Farmed foxes’ activity rhythms have changed to diurnal rather
than nocturnal; this is likely an adjustment to human activity and management routines.
However, silver foxes housed in large enclosures or cages with limited human presence
return to being crepuscular within a few months. Farmed foxes are most active during the
breeding period and near feeding times. Both silver foxes and blue foxes are active for about
20–40% of their daily time, depending on social and physical housing conditions.
Hunting and foraging
Both species are solitary hunters and use different foraging techniques depending on what
food and prey they are after. Red foxes will eat almost anything; various rodents, rabbits,
birds, berries and fruits, earthworms, insects, road kills, carcasses and household wastes. For
red and arctic foxes living near dumpsters, between 40% and 50% of their winter diet may
consist of household offal (Savory et al., 2014). The diet of arctic foxes also reflects what is
available within their living area. Lemming foxes’ diet is dominated by rodents like
lemmings and voles when they are superabundant during rodent peak years. Birds and eggs
are the most important foods for foxes living in coastal habitats; they also feed on foods of
marine origin (fish, mussels, sea urchins and seaweed). When foxes hunt larger prey like
hares or rabbits they stalk, chase, catch and kill their prey with bites and shakings. When
searching for rodents they use a specific and well-known technique called ‘mousing’ where
they jump and efficiently strike the prey from above. Foxes’ extraordinary hearing skills
combined with a newly discovered magnetic sense (Cerveny et al., 2011) inform the fox
about the precise location of its prey. For obtaining immobile food sources like berries, fruits
and insects, foxes explore the area in a systematic, but more relaxed pace. As a significant
adaptation to variable food availability, both fox species store surplus food for later use.
These caches are made by digging shallow holes that are covered by soil using the muzzle
with repetitive and fixed movement patterns. Observations of arctic foxes have shown that
about 80% of eggs gathered during summer were cached rather than being consumed
(Stickney, 1991).
While arctic and red foxes are similar in many ways, they differ in one important aspect:
the arctic fox has unique adaptations to extreme winter conditions and food shortage. The
arctic fox can survive temperatures below –50°C. It has the ability to turn down metabolic
rate and reduce energy costs for movement through immobility, staying buried in the snow
and by running in a more energy-efficient manner (Fuglei and Øritsland, 2003). Arctic foxes
have a well-developed fat deposition capacity and can withstand starvation for several weeks.
They may also adopt a riskier foraging strategy: following polar bears to parasite seal kill
remains.
In fox fur farms, foxes are usually fed once or twice daily, dependent on season and age.
Their food, which is a wet feed with about 30–40% of dry matter, is made of a fine-ground
mixture of fish and slaughter offal including various types of grain as carbohydrate sources.
The feeding pattern differs somewhat between the fox species as blue foxes often consume
their ration faster than silver foxes. A hungry fox may consume its ration within a few
minutes; however, silver foxes may nibble and eat their food over a longer period during the
day. Foxes’ appetite and food consumption may vary with season and temperature, level of
food competition (littermates), feed quality and portion size. Occasionally, captive foxes may
try to cache surplus or leftover food, seen as digging and muzzle movements indicative of
food burying.
Mink
The yearly reproductive season influences the behavioural activity of mink. Observations of
wild mink during day and night in nature are sparse. Mink are most active around dusk and
dawn, but activity does occur during other periods. The passage of food is relatively short,
approximately 3 h (Bleavins and Aulerich, 1981), thus frequent and smaller meals are taken.
Accordingly, mink have 10–11 eating bouts/24 h at farms when food is abundant. Their
activity increases up to expected feeding time at farms, most evident during periods of
restricted feeding. In one study, farm mink stayed inactive – typically in the nest box – for
more than 70% of their time under abundant food conditions (Hansen et al., 1994). Under
natural conditions, feral mink likewise spend most of the 24 h cycle within a den.
In nature, mink patrol within a home range, presumably to forage and scan for intruders
and partners. Mink may travel far (up to above 4 km) daily and the distance covered depends
on sex, competition and the area’s food abundance. The dam’s territory is about 20% smaller
than that of the male mink, and one male territory can overlap that of several females
(Stevens et al., 1977; Birks and Linn, 1982). Mink tend to avoid open spaces and are hunted
by, for example, the North American eagle owl and humans.
The strictly carnivorous mink is an opportunistic predator. In nature, mink eat a wide
range of prey, such as small mammals (rodents and lagomorphs), fish, amphibians,
echinoderms, crustaceans and birds, dependent on local availability, season and competition.
Mink hunt on ground and can also perform short (mostly less than 10 s) dives to forage in
water (Dunstone, 1993). Their senses while submerged, coat protection and swimming
abilities are relatively poor compared with specialized aquatic hunters like the otter (Wise et
al., 1981). Male mink are approximately twice as big as females and can strike on large prey
(e.g. hare or rabbit), while females, in contrast, may access the burrows of smaller rodents
easier (Birks and Dunstone, 1985). Olfaction and hearing are the predominant senses used
during terrestrial hunting; adult mink have a broad hearing range from 1 kHz to above 70
kHz with peak sensitivity at 8–10 kHz (Brandt et al., 2013). The daily activity pattern and
senses of mink make them well adapted to hunt prey such as small rodents that emit high-
frequency vocalizations.
Mink kill the prey with bites, followed by immediate eating or transport back to the den
that can function as a food store. In nature, mink can also be scavengers, and access to live
prey is not mandatory, and is not used for mink on farms. On many farms, mink are fed with
a finely ground high-protein feed paste and they are generally not fussy eaters if the feed is
fresh. Selection for reduced fear has made farm mink even quicker to approach and eat novel
food.
18.6 Management and Welfare of Foxes and Mink

The typical housing environment of farmed foxes and mink consists of cages made of wire
mesh, with the bottom area being 0.225 m2 for mink and 1.2 m2 for foxes. These cages are
placed successively in rows, typically two to ten rows of varying length, usually inside
uninsulated barns. Mink cages are supplied with a nest box year-round, whereas for foxes
artificial dens are not regularly used outside the breeding season. In mink, the nest box
reduces stress and stereotypic behaviour. Mink and foxes are willing to work for access to a
nest box where they may rest together and where the female delivers her yearly litter. Animal
density varies with season, being at its highest during autumn when cubs are housed together.
For many people the barrenness of the cages and the restricted available space stand in
contrast to fur animals’ natural environments. What do we know about the ability of farmed
foxes and mink to cope with their current living conditions?
Housing facilities: the cages
Using wire mesh as cage floor material relates to hygiene demands and the need to keep the
fur clean. Adult mink and foxes perform most behaviours on wire mesh, including also
energetic activities like running and jumping. However, if solid material such as wooden
plates or a sandy surface is simultaneously available, foxes perform more exploration,
digging and play towards solid floors. The barrenness of cages, in addition to restrictions in
for instance feeding management, may in some periods result in thwarted motivations. If
these states are frequently induced, the development of repetitive, invariable movement
patterns like stereotypic behaviours may occur. Both foxes and mink perform various types
of abnormal behaviour like stereotypies (e.g. pacing, head-twirls, circling) and fur-chewing.
Although the prevalence of stereotypic behaviours differs, the presence of these behavioural
patterns indicates that certain environmental restrictions may overtax fur animals’ ability to
cope. In mink, the stereotypic behaviour occurs mainly pre-feeding and increases during
periods of feed restriction. Another type of abnormal behaviour in both foxes and mink is
fur-chewing, involving suckling, chewing and removal of hair, occurring mainly on the tail
tip. There is a genetic component in the predisposition of mink to perform these types of
abnormal behaviour (stereotypies: Hansen et al., 2010; fur-chewing: Malmkvist and Hansen,
2001) and both can be reduced by cage enrichment.
Cage enrichment like balls, chains and tubes may reduce stress hormones and abnormal
behaviour, increase play and improve reproduction in mink (e.g. Hansen et al., 2007;
Meagher et al., 2014) (Fig. 18.4). Gnawing objects improves dental health, reduces
stereotypic behaviours and tends to improve reproduction in foxes (reviewed in Koistinen
and Korhonen, 2013). Likewise, enrichment and avoiding negative handling procedures
reduce fear and increase curiosity in mink (Bak and Malmkvist, 2016). Cage enrichments,
often in the form of using combined resources, are known to improve the welfare of mink.
Thus, enriched housing may induce positive emotional states in fur animals.
Fig. 18.4. Cage enrichment, such as biting ropes, can improve welfare in mink.
Movement and space
Space restriction is an inevitable consequence of captivity and affects most of our production
animals. In free-living carnivores, movement is motivated by foraging, territorial and
reproductive needs. Fur animals’ hunting motivations are modified by selection and regular
feeding routines, and knowledge on the strength of their explorative motivation and need to
perform space-demanding activities such as running is scarce. Foxes are shaped for ‘being on
the move’, which is evident through their energy-efficient trotting pace; thus the motivation
for locomotion and patrolling may be ‘hard-wired’ and less susceptible to genetic
modification. Locomotor motivation may vary with factors like hunger level and season (e.g.
increase during the dispersal and mating period and decrease with feeding level). Yet, these
questions on locomotor motivations in foxes are still unanswered.
Farmed fur animals can obtain basic bodily postures within the cages, such as walking,
sitting and lying stretched out; however, both foxes and mink will use additional space when
given access. Although studies on the welfare effects of extra cage space in foxes are
inconsistent, additional space provides the opportunity for running and exercise that may
positively affect the body condition of future breeding vixens, for example. In mink, studies
have not found that access to larger cages alone improves their welfare (Hansen et al., 2007).
However, mink will work for access to a running wheel and use it for the same amount of
time as mink without access perform stereotypies. Feed satiety and breeding for reduced
stereotypic behaviour decrease minks’ running-wheel use and activity. Thus, running in
wheels appears to substitute cage stereotypies in mink (Hansen and Damgaard, 2009). In
both foxes and mink, the currently used cage height restricts vertical behaviours like upright
stretching and climbing. In mink, the lower height may be preferred when fed on top of the
cage (Díez-León et al., 2016). Foxes prefer elevated locations in the cage; however, the
general impact of cage height on welfare is unstudied.
Sociality
Both foxes and mink have periodic needs to socialize and to withdraw from conspecifics.
Likely, some of the social needs are fulfilled as the housing system allows visual contact
between animals. However, the unrestricted flow of social signals and odours combined with
high animal density also has side-effects, such as social stress due to dominant and
aggressive neighbours. In foxes, social status may affect maternal behaviour and affect cub
survival. Opportunities for concealment, like in a nest box, or increased distance to
neighbours, for example by access to additional space, are some alternative ways to provide
animals the chance to withdraw. Full social contact is important for behavioural development
and welfare of young animals. Also, adult vixens may have positive social interactions, but
aggression and injuries during mixing are common and detrimental to their welfare (Hovland
et al., 2010). Solitary housing may benefit foxes during periods of increased social
competition, like during sexual maturity, as well as during the breeding season and when
there is food competition. Apparently, foxes’ social functioning and the related welfare
effects may vary according to pair or group composition, age and season. Housing of several
(more than two) mink together occurs on some farms during the growth period. Group-
housing increases the risk of aggression, bite marks and wounds, even when families of mink
are kept together. Bite marks are evident as black spots on the leather side of the skin, visible
only after pelting of dark colour types (Møller and Hansen, 2016). Group-housed mink eat
10% to 20% less in late autumn, potentially due to thermoregulatory benefits of being
socially housed (Hänninen et al., 2008). Additional feeding places can reduce aggression in
groups. However, bite marks during social housing still occur above the low level observed
in male–female pairs.
Management procedures
In farmed foxes and mink, invasive management procedures such as castration, ear-tattooing
or clipping are not common. During a production year, fur animals are handled several times,
for example during litter separation, pelt grading, mating and insemination (foxes only). Both
mink and foxes may perceive handling negatively and respond with fear or aggression. These
short-term procedures act as acute stressors (e.g. Moe and Bakken, 1997), which can be
reduced by habituation and use of appropriate handling methods (Fig. 18.5). In addition,
breeding for reduced fearfulness and ease of handling is important for welfare and helps
animals cope with human contact. Fur animals are often killed by gas (mink) or electrical
stunning (foxes) outside their home cage, thus welfare problems caused by transport and
herding during slaughter are absent.
Fig. 18.5. Cubs are socialized to humans through early contact and positive human–animal
interactions, such as during feeding and the provision of tit-bits and activity objects. Early
handling and contact reduce fear and increase confidence in silver foxes.
Separation of mother and young occurs around the natural period of weaning in fur
animals, which simplifies the transition from milk to solid food. In practice, separation time
and procedure may vary with litter size and condition and maternal motivation of the female.
Abrupt weaning by moving the mother visually away from the cubs has been observed to be
stressful for silver fox vixens, eliciting high rates of contact calls and stereotypies, and
elevating cortisol levels. In mink kits, separation also induces increased calling and transient
cortisol increases in dams. This happens to a higher extent when separation is done at 7
weeks compared with 8 weeks after birth (Malmkvist et al., 2016). Scientific studies on
management and weaning procedures that would alleviate acute separation and weaning
stress are scarce in foxes. In mink, keeping the dam in an intermediate body condition,
providing nest boxes and supplementation of nest-building materials, access to additional
water and gradual removal of kits from large-litter dams (cross-fostering) are all factors
shown to improve their welfare.
The current feeding procedures may generate some challenges for animal welfare.
Intensive feeding of growing animals may result in severe obesity complicating the ability to
move, as found in blue foxes. In contrast, slimming of breeding animals through restrictive
feeding increases stereotypic behaviours and fur-chewing. As the food paste allows little time
for oral manipulation, adding chewable objects like meat bones, wooden sticks, balls or ropes
is useful to satisfy the need for chewing and to occupy animals’ time. Daily access to feed
with a coarser texture, providing the mink with larger animal parts such as fish heads for
chewing and eating, can reduce fur-chewing and stereotypic behaviour (Malmkvist et al.,
2013).
18.7 Adaption and Domestication – Concluding Remarks

In both foxes and mink, selection experiments have shown that behavioural changes relevant
for adaptation to a captive environment (e.g. increased confidence) can be achieved within a
short period of time (<10 years). Increased selection intensity for traits that improve animals’
ability to cope may thus produce rapid changes. The reproductive ability and body size of fur
animals have generally increased due to directional selection at farms. With the increase in
body size, preferences and need for housing resources may change. For example, a larger fox
may be more susceptible to leg problems and less able to use an elevated platform, otherwise
a valued resource in the cage. Consequently, larger animals may require changes in the cage
system in order to secure their welfare. Given the potential of genetic selection and the
behavioural flexibility of fur animals, there is a clear opportunity for further development of
housing systems and management procedures that support fur animals’ needs.
Acknowledgements
Thanks to Nina E. Eide and Steen H. Møller for valuable comments on the text.
References
Ahola, L., Harri, M., Kasanen, S., Mononen, J. and Pyykönen, T. (2000) Effect of family housing of farmed silver foxes
(Vulpes vulpes) in outdoor enclosures on some behavioural and physiological parameters. Canadian Journal of Animal
Science 80, 427–434.
Bak, A.S. and Malmkvist, J. (2016) Enrichment and management influence the emotional state of farm mink. In: Dwyer, C.,
Haskell, M. and Sandilands, V. (eds) Proceedings of the 50th Congress of the International Society for Applied Ethology,
Edinburgh, UK, 12–15 July 2016. Wageningen Academic Publishers, Wageningen, the Netherlands, p. 224.
Baker, P.J., Funk, S.M., Bruford, M.W. and Harris, S. (2004) Polygynandry in a red fox population: implications for the
evolution of group living in canids? Behavioral Ecology 15, 766–778.
Bakken, M. (1993) Reproduction in farmed silver fox vixens, Vulpes vulpes, in relation to own competition capacity and that
of neighbouring vixens. Journal of Animal Breeding and Genetics 110, 305–311.
Birks, J.D.S. and Dunstone, N. (1985) Sex related differences in the diet of the mink Mustela vison. Holarctic Ecology 8,
245–252.
Birks, J.D.S. and Linn, I.J. (1982) Studies on the home range of feral mink (Mustela vison). Symposia of the Zoological
Society of London 49, 231–257.
Bleavins, M.R. and Aulerich, R.J. (1981) Feed consumption and food passage time in mink (Mustela vison) and European
ferrets (Mustela putorius furo). Laboratory Animal Science 31, 268–269.
Braastad, B.O. (1993) Periparturient behaviour of successfully reproducing farmed silver-fox vixens. Applied Animal
Braestrup, W. (1941) A study on the arctic fox in Greenland. Meddelelser om Grönland 131, 1–101.
Brandt, C., Malmkvist, J., Nielsen, R.L., Brande-Lavridsen, N. and Surlykke, A. (2013) Development of vocalization and
hearing in American mink (Neovison vison). Journal of Experimental Biology 216, 3542–3550.
Cerveny, J., Begall, S., Koubek, P., Novakova, P. and Burda, H. (2011) Directional preference may enhance hunting
accuracy in foraging foxes. Biology Letters 7, 355–357.
Clausen, K.T., Malmkvist, J. and Surlykke, A. (2008) Ultrasonic vocalisations of kits during maternal kit-retrieval in farmed
mink, Mustela vison. Applied Animal Behaviour Science 114, 582–592.
Díez-León, M., Quinton, M. and Mason, G. (2016) Preferences for different cage height in farmed American mink
(Neovison vison): taller is not better. Proceedings of the XIth International Scientific Congress in Fur Animal Production.
Scientifur 40, 347–349.
Dunstone, N. (1993) The Mink. T & AD Poyser, London.
Elmhagen, B., Hersteinsson, P., Norén, K., Unnsteinsdottir, E.R. and Angerbjörn, A. (2014) From breeding pairs to fox
towns: the social organisation of arctic fox populations with stable and fluctuating availability of food. Polar Biology 37,
111–122.
Forester, J.E. and Forester, A.D. (1982) Silver Fox Odyssey – The History of the Canadian Silver Fox Industry. Canadian
Silver Fox Breeders Association, PEI Department of Agriculture and Forestry and Irwin Printing, Charlottetown,
Canada.
Frafjord, K. (1992) Denning behaviour and activity of arctic fox Alopex lagopus pups: implications of food availability.
Polar Biology 12, 707–712.
Frafjord, K. (1993) Agonistic behaviour and dominance relations of captive arctic foxes (Alopex lagopus) in Svalbard.
Behavioural Processes 29, 239–252.
Fuglei, E. and Øritsland, N.A. (2003) Energy cost of running in an arctic fox, Alopex lagopus. The Canadian Field-
Naturalist 117, 430–435.
Hänninen, S., Ahola, L., Pyykönen, T., Korhonen, H.T. and Mononen, J. (2008) Group housing in row cages: an alternative
housing system for juvenile mink. Animal 2, 1809–1817.
Hansen, B.K., Jeppesen, L.L. and Berg, P. (2010) Stereotypic behaviour in farm mink (Neovison vison) can be reduced by
selection. Journal of Animal Breeding and Genetics 172, 64–73.
Hansen, S.W. (1996) Selection for behavioural traits in mink. Applied Animal Behaviour Science 49, 137–148.
Hansen, S.W. and Damgaard, B.M. (2009) Running in a running wheel substitutes for stereotypies in mink (Mustela vison)
but does it improve their welfare? Applied Animal Behaviour Science 118, 76–83.
Hansen, S.W., Hansen, B.K. and Berg, P. (1994) The effect of cage environment and ad libitum feeding on the circadian
rhythm, behaviour and feed intake in farm mink. Acta Agriculturae Scandinavica Section A, Animal Science 44, 120–
127.
Hansen, S.W., Malmkvist, J., Palme, R. and Damgaard, B.M. (2007) Do double cages and access to occupational materials
improve the welfare of farmed mink? Animal Welfare 16, 63–76.
Harjunpää, S. and Rouvinen-Watt, K. (2004) The development of homeothermy in mink (Mustela vison). Comparative
Biochemistry and Physiology. Part A, Molecular & Integrative Physiology 137, 339–348.
Harris, S. and White, P.C.L. (1992) Is reduced affiliative rather than increased agonistic behaviour associated with dispersal
in red foxes? Animal Behaviour 44, 1085–1089.
Hovland, A.L., Akre, A.K. and Bakken, M. (2010) Group housing of adult silver fox (Vulpes vulpes) vixens in autumn:
agonistic behaviour during the first days subsequent to mixing. Applied Animal Behaviour Science 126, 154–162.
Hovland, A.L., Akre, A.K., Flø, A., Bakken, M., Koistinen, T. and Mason, G.J. (2011) Two’s company? Solitary vixens’
motivations for seeking social contact. Applied Animal Behaviour Science 135, 110–120.
Kenttämies, H., Nordrum, N., Brenoe, U., Johannessen, K.R. and Bakken, M. (2002) Selection for more confident foxes in
Finland and Norway: heritability and selection response for confident behaviour in blue foxes (Alopex lagopus). Applied
Koistinen, T. and Korhonen, H.T. (2013) Complex housing environment for farmed blue foxes (Vulpes lagopus): use of
various resources. Animal 7, 1354–1361.
Korhonen, H.T. and Alasuutari, S. (1995) Dominance relations in captive groups of adult and juvenile arctic blue foxes
(Alopex lagopus). Polar Biology 15, 353–358.
Kruchenkova, E.P., Gol’tsman, M.E. and Frommolt, K.K. (2003) Rhythm structure of Arctic fox (Alopex lagopus
beringensis and A-l. semenovi) serial barking: age, sex, and context determinants (abstract). Zoologichesky Zhurnal 82,
525–533.
Macdonald, D. (1987) Running with the Fox. Unwin Hyman, London.
Macdonald, D.W and Reynolds, J.C. (2004) Red fox (Vulpes vulpes). In: Sillero-Zubiri, C., Hoffmann, M. and Macdonald,
D.W. (eds) Canids: Foxes, Wolves, Jackals and Dogs. Status Survey and Conservation Action Plan. IUCN/SSC Canid
Specialist Group, Gland, Switzerland and Cambridge, pp. 129–136.
Malmkvist, J. and Hansen, S.W. (2001) The welfare of farmed mink (Mustela vison) in relation to behavioural selection: a
review. Animal Welfare 10, 41–52.
Malmkvist, J. and Hansen, S.W. (2002) Generalization of fear in farm mink, Mustela vison, genetically selected for
behaviour towards humans. Animal Behaviour 64, 487–501.
Malmkvist, J., Palme, R., Svendsen, P.M. and Hansen, S.W. (2013) Additional foraging elements reduce abnormal behaviour
– fur-chewing and stereotypic behaviour – in farmed mink (Neovison vison). Applied Animal Behaviour Science 149,
77–86.
Malmkvist, J., Sørensen, D.D., Larsen, T., Palme, R. and Hansen, S.W. (2016) Weaning and separation stress: maternal
motivation decreases with litter age and litter size in farmed mink. Applied Animal Behaviour Science 181, 152–159.
Martin, P. (1984) The meaning of weaning. Animal Behaviour 32, 1257–1259.
Meagher, R.K., Dallaire, J.A., Campbell, D.L.M., Ross, M., Moller, S.H., Hansen, S.W., Diez-Leon, M., Palme, R. and
Mason, G.J. (2014) Benefits of a ball and chain: simple environmental enrichments improve welfare and reproductive
success in farmed American mink (Neovison vison). PLoS One 9(11), e110589.
Meia, J.S. and Weber, J.M. (1995) Home ranges and movement of red foxes in central Europe: stability despite
environmental changes. Canadian Journal of Zoology 73, 1960–1966.
Meyer, S. and Weber, J.M. (1996) Ontogeny of dominance in free-living red foxes. Ethology 102, 1008–1019.
Moe, R.O. and Bakken, M. (1997) Effects of handling and physical restraint on rectal temperature, cortisol, glucose and
leucocyte counts in the silver fox (Vulpes vulpes.) Acta Veterinaria Scandinavica 38, 29–39.
Møller, S.H. and Hansen, S.W. (2016) Insignificant effect of human activity and access between cages on group housed
juvenile mink. Proceedings of the XIth International Scientific Congress in Fur Animal Production. Scientifur 40, 333–
336.
Møller, S.H. and Pinkalski, M.N. (2015) Yearling mink dams fed restricted in early lactation have less mammary gland
tissue six weeks after birth. In: Nordic Association of Agricultural Scientists (NJF) Autumn Meeting in Fur Animal
Research, Turku, Finland, 29 September–1 October 2015. Proceedings of NJF Seminar 485, 104–112.
Norén, K., Hersteinsson, P., Samelius, G., Eide, N.E., Fuglei, E., Elmhagen, B., Dalén, L., Meijer, T. and Angerbjörn, A.
(2012) From monogamy to complexity: social organization of arctic foxes (Vulpes lagopus) in contrasting ecosystems.
Canadian Journal of Zoology 90, 1102–1116.
Price, E.O. (1984) Behavioral aspects of animal domestication. The Quarterly Review of Biology 59, 1–32.
Pyykönen, T., Mononen, J., Ahola, L. and Rekilä, T. (2005) Periparturient behaviour in farmed blue foxes (Alopex lagopus).
Rochmann, R. (1969) Pelsdyrboken (The Fur Animal Book). Norwegian Fur Breeders Association, Johansen & Nielsen,
Oslo.
Savory, G.A., Hunter, C.M., Wooller, M.J. and O’Brien, D.M. (2014) Anthropogenic food use and diet overlap between red
foxes (Vulpes vulpes) and arctic foxes (Vulpes lagopus) in Prudhoe Bay, Alaska. Canadian Journal of Zoology 92, 657–
663.
Shackelford, R.M. (1984) American mink. In: Mason, I.L. (ed.) Evolution of Domesticated Animals. Longman, London and
New York, pp. 229–234.
Soulsbury, C.D., Iossa, G., Baker, P.J., White, P.C.L. and Harris, S. (2011) Behavioral and spatial analysis of extraterritorial
movements in red foxes (Vulpes vulpes). Journal of Mammalogy 92, 190–199.
Stevens, R.T., Ashwood, T.L. and Sleeman, J.M. (1977) Fall–early winter home ranges, movement, and den use of male
mink, Mustela vison in Eastern Tennessee. The Canadian Field-Naturalist 111, 312–314.
Stickney, A. (1991) Seasonal patterns of prey availability and the foraging behavior of arctic foxes (Alopex lagopus) in a
waterfowl nesting area. Canadian Journal of Zoology 69, 2853–2859.
Sundqvist, C., Amador, A.G. and Barke, A. (1989) Reproduction and fertility in the mink (Mustela vison). Journal of
Reproduction and Fertility 85, 413–441.
Tarroux, A., Berteaux, D. and Bêty, J. (2010) Northern nomads: ability for extensive movements in adult arctic foxes. Polar
Biology 33, 1021–1026.
Trut, L. (1999) Early canid domestication: the farm-fox experiment. American Scientist 87, 160–169.
Vergara, V. (2001) Comparison of parental roles in male and female red foxes, Vulpes vulpes, in Southern Ontario. The
Canadian Field-Naturalist 115, 22–33.
Whiteside, H.M., Dawson, D.A., Soulsbury, C.D. and Harris, S. (2011) Mother knows best: dominant females determine
offspring dispersal in red foxes (Vulpes vulpes). PLoS One 6(7), e22145.
Wise, M.H., Linn, I.J. and Kennedy, C.R. (1981) A comparison of the feeding biology of mink Mustela vison and otter Lutra
lutra. Journal of Zoology 195, 181–213.
19 The Behaviour of Laboratory Mice and Rats
H. Würbel , C. Burn and N. Latham

Mice and rats belong to the order Rodentia, which forms the largest and most diverse group
of mammals, comprising approximately 2300 species or 42% of all known mammalian
species (Wilson and Reeder, 2005). Their common feature is a unique gnawing action,
provided by massive masseter jaw muscles and ever-growing, sharp incisor teeth. Gnawing
allows rodents to eat the toughest nuts and seeds and to gnaw through wood and roots in
search of food and shelter. Most rodents are small-bodied, although the largest species, the
capybara (Hydrochoerus hydrochaeris), may get as large as 60 cm high and 130 cm long.
Most rodents are frugivorous/herbivorous, but some are omnivorous (e.g. most of the genus
Rattus) or insectivorous (e.g. desert dormouse, Selevinia betpakdalaensis). Rodents inhabit
virtually every type of terrestrial habitat. Some species are arboreal (e.g. arboreal squirrels,
New World porcupines); others dig extensive burrow systems (e.g. gerbils, mole-rats, ground
squirrels). Most rodents are nocturnal, depending on secluded shelters for protection against
predators (e.g. mice, rats, voles, hamsters, gerbils), while those living in less predated areas
are diurnal and nest above ground (e.g. guinea pigs, chinchillas). However, individual species
vary greatly in the range of habitats they occupy, reflecting different degrees of adaptability
between them. Among the extreme generalists are the house mouse (Mus musculus) and the
Norway rat (Rattus norvegicus). Their adaptability has facilitated adaptation to human
habitation, allowing them to exploit rich food sources. In turn, it predisposed them for use as
laboratory animals. Today mice and rats account for about 85% of all animals used in
research worldwide. With an estimated 17–22 million vertebrates used in the USA, and about
12 million in the EU, each year about 30 million rodents are used in animal experiments in
these two parts of the world alone.
19.2 Origin and Domestication History

The mouse
The laboratory mouse is descended from one of the most widespread and successful mammal
species, the house mouse (M. musculus). House mice have had a close relationship with
humans dating back many thousands of years. They originated on the steppes of Asia and are
thought to have spread around the world with human migration, now being found in almost
every region of the planet. House mouse populations are typically defined in one of two ways
based on their dependence on humans: commensal mice live in man-made environments,
while feral mice live more independently of humans in natural environments. As illustrated
below, this difference has a number of effects on house mouse behaviour.
Mice have been used in research for over 300 years, but the inbred laboratory mouse (as
we know it today) first appeared a century ago, with the first inbred strain (DBA) being
created by Clarence Cook Little at Harvard University in order to study the inheritance of
coat colour. Today, the laboratory mouse is the most widely used vertebrate species in
biomedical research, with over 1000 genetically defined inbred strains (including the widely
used C57BL/6 and Balb/c strains), and increasing numbers of genetically altered strains with
either gene deletions (knockout) or insertions (transgenics). Numerous outbred strains are
also available (including ICR CD-1), which are genetically more heterogeneous than their
inbred counterparts. However, even these strains are derived from very small numbers of
progenitor animals. Besides their adaptability, a short reproductive cycle, short life span,
small body size and low cost of maintenance were the main features predisposing mice for
‘success’ as laboratory animals.
The rat
Having originated in northern China, the Norway rat, like the house mouse, has spread to all
continents except Antarctica, and is now the dominant rat species in Europe and most of
North America. Its name is actually a misnomer, going back to the English naturalist John
Berkenhout, who gave the brown rat the taxonomic name R. norvegicus, believing that it had
migrated to England on Norwegian ships in the early 1700s.
Laboratory rats were first established in 1895 from a population of albino rats at Clark
University in Worcester, Massachusetts. Originally used for research on diet and physiology,
rats quickly became used in all areas of biomedical research and they are the predominant
animal model in experimental psychology.
Domestic rats are calmer and less likely to bite than their wild ancestors. They tolerate
greater crowding, breed earlier, produce more offspring and, like in most domestic species,
their brains are smaller.
Like laboratory mice, laboratory rats are available as outbred and (albeit less frequently
than in mice) inbred strains. Most strains were derived from the Wistar rat. Other popular
strains are Sprague Dawley, Fischer 344, Long-Evans and Lister hooded rats.
It was not until recently that knockout and transgenic rats became available because the
simple techniques that worked in mice did not work in rats. Meanwhile, genetically modified
rats are increasingly popular because it is widely believed that they model human physiology
better than mice.
19.3 Diurnal Rhythm, Sleep and Grooming Behaviour

The mouse
House mice are generally crepuscular or nocturnal, but their activity patterns may be altered
by factors such as the timing of food availability, human activity in man-made environments
and reproductive state such as lactation (Latham and Mason, 2004). Little is known about the
total or individual duration of activity bouts in free-living mice. Observations of laboratory
mice show that they can be continuously active for several hours at a time, but that they are
active for less than 50% of each 24 h period (Baumgardner et al., 1980) – this is probably an
underestimate of activity in free-living mice whose environments are far more demanding.
Inactive periods are spent in sheltered nests which provide protection against predators
and adverse weather. In feral environments nests are usually constructed in underground
burrows, and may be lined with grass, hair or feathers. Mice are efficient tunnellers in soft
earth, and burrows may consist of simple tunnels up to 1 m long or complex tunnels systems
with numerous chambers and exits. In commensal environments, mice build spherical or
bowl-shaped nests consisting of a loose outer paper- or rag-based structure and lined with
finer shredded material (Randall, 1999). Mice typically nest either alone or with their
breeding partner and family group; although under adverse conditions (e.g. extreme cold)
even highly territorial individuals will nest with other mice, including subordinates that they
normally attack (Crowcroft, 1966).
The transition between inactivity and activity usually involves a period of grooming,
starting with the head and progressing caudally. Grooming also occurs sporadically during
activity bouts and, in captivity at least, it can occupy roughly 17% of the time budget
(Baumgardner et al., 1980). Self-grooming is important for hygiene and maintaining the coat
for insulation, while allogrooming helps to maintain social relationships. Grooming can also
occur particularly intensively after eating and may then, during allogrooming, assist the
transfer of information about foodstuffs (Crawley, 2000).
The rat
The circadian rhythm of Norway rats is most accurately described as being crepuscular, with
most activity occurring just after dusk and before dawn. However, although this is certainly
their preference, they can adjust their rhythm relatively flexibly depending on food
availability, predation risk, weather and sexual activity, so it is not uncommon to see rats in
daylight (Calhoun, 1963; see Fig. 19.1). They do not hibernate but, in temperate climates, are
more active in warmer months than during the winter.
Fig. 19.1. Rats are good climbers. Although mostly nocturnal, it is not unusual to see rats
during daylight depending on food availability, predation risk, weather and sexual activity.
(Photo courtesy of Charlotte Burn.)
Rats sleep within underground burrows in nests made of grass or other vegetation, or
shredded paper if available (Calhoun, 1963). The nests are usually flat or cup-shaped pads
within chambers of 15–40 cm diameter, but are rarely as complex as mouse nests. Provided
that fewer than five or six rats share a nest, it is seldom soiled. The burrow itself consists of
tunnels with a mean diameter of 8–10 cm and an entrance, which can be sealed temporarily
with vegetation or more permanently with soil. The straight sections of the tunnels have
median lengths around 30 cm, beyond which they may bend, form cavities or split into two
tunnels. Despite their apparently delicate forepaws, rats are effective burrowers, preferring
loose, well-drained soil that does not crumble, but their burrows are shallower than those of
many subterranean rodent species.
Most grooming takes place in the safety of the burrow. Rats groom daily, presumably
removing dirt, reducing ectoparasite loads and distributing their own scents and conditioning
oils evenly over the coat. Grooming is often associated with resting, but it can also be
provoked by novelty, disturbance or frustration. Laboratory studies show that grooming starts
with the head and progresses caudally, but grooming bouts are shorter and rarely progress to
the caudal regions when rats are more stressed and vigilant.
19.4 General Activity, Foraging and Feeding Behaviour

The mouse
House mice are creatures of habit, especially commensal mice, where territories or home
ranges tend to be smaller than those of their feral counterparts (Latham and Mason, 2004).
When out and about mice use visual landmarks to navigate around their environment and
repeatedly follow familiar routes, leading to well-worn runways (Randall, 1999). Mice are
also extremely agile, being excellent climbers, able to scale vertical brickwork or bark; they
are proficient jumpers, able to leap over 30 cm vertically and 2.5 m downwards without
harm; and they are capable swimmers (Randall, 1999). However, mice are thigmotactic
(staying near walls) and will generally avoid areas that have been cleared of vegetation,
exposed areas or areas regularly disturbed by human activity (Meehan, 1984).
Regular, predictable movement around the environment serves a number of functions. It
allows the investigation of olfactory cues used in communication, which contain information
about age, sexual status, relatedness and individual identity. These cues also play key roles in
territorial, sexual and parental behaviour, as discussed below. Many of these cues are found
in the urine (although important cues are also present in faeces and plantar-gland secretions).
Urinary odour cues may last for up to 2 days because the volatile signalling pheromones are
bound to non-volatile major urinary proteins which act as a slow-release mechanism (Hurst
et al., 2001). Mice also use regular movements around their territories to monitor the
presence of predators, which include owls and weasels in natural environments, and the
domestic cat in urban areas.
Mice also incorporate foraging into their movements around their environment. Mice eat
up to 20% of their body weight daily, consuming about 200 small meals per night from up to
30 food sites (Meehan, 1984). Mice are omnivores, but show preferences for foods high in fat
and protein. The diet of commensal mice is often determined by the nature of their
environment, such as grain in grain stores; while the diet of feral mice can be very varied and
includes cereals, grasses, roots and seeds (Berry, 1970). House mice also exhibit predatory
behaviour, eating live insects and their larvae, and even seabird chicks (Latham and Mason,
2004). House mice acquire food preferences from their mothers and from other mice through
the transfer of odour cues during allogrooming, and are generally cautious about trying novel
foods unless they have been consumed by other individuals. Mice generally meet their water
requirements from their food, if consuming diets with moisture contents of at least 15%, but
they will drink free water if it is available and require additional water if living in hot, arid
environments or feeding on a dry or protein-rich diet (Meehan, 1984).
The rat
Rats are opportunistic omnivores, consuming diverse foods and employing varied methods of
foraging: factors that have led to them becoming major pests. Wild rats choose to inhabit
areas with available running water (their foremost priority), accessible food, exposed soil for
digging a burrow and cover. They avoid areas densely populated with humans, unless food is
particularly plentiful there, such as through inadequate waste disposal. Rats patrol their
territory regularly, travelling along runs that usually follow vertical features, such as walls;
rats are thigmotactic, so they maintain physical contact with vertical surfaces whenever
possible. The range size is usually about 30–45 m in diameter (Davis, 1953), but can be up to
about 150 m.
Because of their opportunistic lifestyle, rats often forage even when satiated, eating little
but learning the whereabouts of food for future reference. Rats locate food mainly through
olfaction which is orders of magnitude more sensitive than our own. They can even locate the
direction of an olfactory source without moving their heads three orders of magnitude more
quickly than humans can. Wild rats forage by digging in loose substrates such as leaf litter,
by climbing trees, and even by diving into streams to collect molluscs. They are central place
foragers, meaning that once food is found they usually carry it in their mouths to the burrow
or a nearby refuge. There they consume the food, holding it in their forepaws. Their teeth
allow them to chew through tough materials, such as nuts and snail shells, and even lead
piping (Barnett, 2005).
Their diet can include fruit, seeds, nuts, molluscs, eggs, small vertebrates and any food
waste left by humans. In fact, they have similar food preferences to humans, finding sweet,
calorific and high-protein foods particularly attractive, consuming salty or acidic foods in
moderation, and avoiding bitter foods (Burn, 2008). They can even acquire learned
preferences for chilli-flavoured foods, despite initially avoiding it. However, some species
differences exist; for example, rats taste the harmless compound denatonium benzoate as
being less bitter than humans and many other animals do. It can therefore be added to
poisoned baits, discouraging non-target species from consuming them, while rats remain
relatively undeterred.
Rats cannot vomit, but laboratory studies show that consuming toxins makes them more
likely to consume non-nutritional substances, such as clay, possibly serving to ‘neutralize’ the
toxins. Wild rats avoid consuming lethal quantities of poisons primarily through their
neophobic behaviour. That is, they avoid eating unfamiliar foods if familiar ones are
available, and will initially eat only very small quantities of unfamiliar foods, if any. They
then wait several hours before eating more of that food and if they suffer any ill effects from
it, they subsequently avoid eating it. This learned avoidance is a very powerful form of one-
trial learning, and rats may continue to avoid the food for months if not years. A lactating
mother rat will even avoid a novel food if her pups become sick from the milk produced after
she ate it. Calhoun (1963) fed rats in a semi-natural enclosure on standard chow, but
occasionally supplied mixed ‘garbage’, which they initially approached with the stretched-
attend posture, consuming little. However, once accustomed to the garbage, they
concentrated on eating and caching it, only returning to eating the chow once it was finished.
As well as learning about foods from their own experiences, rats can learn about it
socially (Burn, 2008). Rats prefer novel foods after smelling them on conspecifics’ breath, an
effect mediated through carbon disulfide (present in rats’ breath) (Galef et al., 1988), so new
food preferences can spread rapidly through rat colonies. Rats also avoid foods that they
themselves have eaten if they encounter the odour of a sick conspecific soon after eating, but
this does not prevent them from eating the food that might have made the conspecific sick; in
fact, carbon disulfide in the sick rats’ breath still induces a preference for that food.
19.5 Social Organization and Social Behaviour

The mouse
The social structure and territorial behaviour of house mouse populations vary according to
the animals’ environment. In feral conditions both male and female mice may hold loosely
defended overlapping territories (or home ranges). Home range sizes often vary with the
density of available food, with reported densities ranging from one mouse per 3–33 m2 in
fields of grain – populations in some agricultural areas increase fivefold in the weeks
immediately following crop sowing (Latham and Mason, 2004). However, in some areas, for
example the wheatlands of Australia, mice may have home ranges up to an incredible 80,000
m2. In contrast, in commensal environments territories are typically held only by males
(although a male’s breeding mate(s) and juvenile offspring may assist in territorial behaviour)
(Crowcroft, 1966). Territories are also much smaller and more strictly defended.
Territorial boundaries often occur at physically and/or visually striking features of the
environment. Territory ownership is signalled, particularly by dominant males, via regularly
refreshed urine marks along the borders of their territory (Hurst et al., 2001). These marks
are also laid down throughout the rest of the territory, especially on conspicuous objects and
favoured nesting and feeding sites. Dominant males may even venture into neighbouring
territories to over-mark the urine marks of their competitors, as fresh urine marks signal
competitive ability. Dominant males will not tolerate marks from other adult males (i.e.
potential competitors), but will generally tolerate marks from juveniles and females. This is
thought to be because juvenile urine lacks the aggression-eliciting properties of adult male
urine and female urine contains an aggression-inhibiting factor.
When male mice meet they use plantar, preputial and salivary gland scents to identify
each other and to assess each other’s ‘maleness’ and sexual state. At high population
densities (usually in commensal populations) territorial males will chase intruders out of their
territories, biting at their rump and tail. The intruder may indicate his subordinancy by
standing, raising his forepaws and exposing his belly (Crawley, 2000). However, attacks can
be intense and sustained, and if unable to escape the intruder is likely to face severe injury
and even death. In feral populations, territory holders are generally more tolerant of known
subordinates and will vigorously defend only favoured nesting and feeding sites.
The rat
Rats are highly social. Colony sizes can vary widely, from one male and female with their
offspring, to colonies of 200 or more individuals, depending on resource availability and
physical boundaries. An 18-month study (Traweger et al., 2006) of wild rats in Salzburg,
Austria, found that densities ranged from 10 rats/km along the banks of standing waters up to
113 rats/km at river banks, but in many areas of the city none were trapped at all.
Rat colonies are stable with low migration rates, provided the environment allows
(Davis, 1953), so overt aggression towards established colony members is relatively rare.
Most aggression occurs between young adult male rats as they establish their place in the
dominance hierarchy and towards intruders. Older (usually larger) rats are dominant over
younger ones. When population densities are low, rats live in territorial groups of one
dominant male, up to about six females and their offspring. When the density is higher, the
dominant male can no longer maintain a monopoly on the females and many subordinate
males can cohabit within the colony. Subordinates of both sexes usually face restricted access
to food and other resources, have significantly more wounds than dominants and have higher
mortality rates (Davis, 1953).
Rats recognize familiar individuals largely through their individual olfactory
characteristics, which can be determined genetically or acquired from the environment (Burn,
2008). Much olfactory information is mediated through urine, but rats have many other
specialized scent glands. They can thus gain information about each other’s gender,
reproductive state, genetic relatedness, dominance, health status and individual identity
through sniffing each other or their urine. Dominant males in particular use urine-marking to
delineate their territory and to maintain their dominant status. Rats communicate with each
other using scents, including alarm or reward odours that they deposit in harmful or
beneficial environments, respectively. They also communicate using vocalizations – both
audible to humans and ultrasonic – and using postures.
Aggression is strongly mediated through scent (Burn, 2008), with males that smell
unfamiliar (i.e. usually intruders) being attacked; sometimes unfamiliar females are also
attacked. Male offensive attack includes lateral attack, bites targeted at the opponent’s rump,
chasing and pinning the opponent down (Blanchard et al., 2001). Defensive attack involves
bites directed at the opponent’s face. Although the injuries themselves are rarely directly
fatal, serious fights can be very costly, with wounds (mostly on the back or rump) potentially
becoming infected, and some defeated rats can even die following a shock or depression-like
state. Therefore, serious fighting between familiar individuals is evaded when possible
through appeasing body postures, such as lying supine, possibly through audible squeaks and
short, variable vocalizations in the 50 kHz range, and probably through scent signals
although little is known about these. Rats emit an alarm pheromone in response to aversive
events (Burn, 2008), such as electric shocks or predator encounters, and it causes avoidance
in conspecifics, but it is not known if they produce it in defensive contexts. They also
produce a porphyrin-rich secretion, chromodacryorrhoea (or ‘red tears’), from the Harderian
glands behind the eyes in response to diverse stressors, but it is unclear whether
chromodacryorrhoea has an appeasing effect in the Norway rat. The vibrissae are very
important for preventing bites to the face and vibrissal contact alone can be sufficient to drive
an opponent away (Blanchard et al., 2001).
More is known about aggression than about other social behaviours, but rats actively
seek contact with conspecifics. They often rest huddled together in a nest cavity (Calhoun,
1963) and, in captivity at least, are among the few species that occasionally engage in play-
like behaviour even as adults. Play superficially resembles aggression, but the bites cause no
injury and are directed at the neck rather than the rump, and the hair does not become pilo-
erected. Play sessions often end with one rat being pinned down by the other, who then
vigorously nibbles and scratches at the pinned rat’s fur: an activity known as allogrooming.
The social implications of allogrooming are not well understood.
19.6 Mate Choice, Mating Behaviour and Parental Care

The mouse
Mice are prolific breeders when food and nesting material are abundant. The pest control
literature estimates that in resource-rich commensal environments females can produce 50 (or
more) pups per year (Randall, 1999): indeed, this high reproductive rate is one of the reasons
for the success of mice in both natural and laboratory environments. Females have a number
of strategies for maximizing their reproductive output. For example, they prefer unfamiliar
males that exhibit dominant behaviours, such as strong over-marking of urine marks and high
rates of ultrasonic calling (Latham and Mason, 2004). Males similarly also show a preference
for unfamiliar females and for more ‘feminine’ characteristics, such as a shorter anogenital
distance – an indication of reduced prenatal exposure to testosterone. Olfactory cues play a
key role in identifying good-quality mates of both sexes, but they also influence reproductive
state in females, for example by suppressing oestrus or by delaying/bringing forward the
onset of oestrus. Females indicate sexual receptivity through hopping and darting
movements, ear-wiggling and ultrasonic vocalizations. The male responds with his own
ultrasonic vocalizations (sometimes called ‘songs’) and then mounts and copulates with the
female – although subordinate males may not progress with copulation if a more dominant
animal appears (Brown, 1953).
Gestation lasts between 19 and 21 days (Berry, 1970), and during the final few days
females are highly motivated to build complex nests (Brown, 1953; Meehan, 1984). The pups
are often born at night, and the female spends time between each birth cleaning the pups and
moving around the nest. Litters consist on average of five to nine individuals, with equal
numbers of males and females. The newborn mice are blind, hairless and dependent upon
their mothers for food and thermoregulation. They elicit parental behaviours such as nursing,
grooming and retrieval through specific vocalizations, including ultrasonic calls. In some
populations, reproductive females may nest and even nurse their young communally –
particularly if the females cannot distinguish their pups from those of other mothers, as may
occur if the pups are born within a few days of each other. It is unknown whether communal
nursing is an adaptive strategy or a response to suboptimal conditions, but studies have
shown it to be beneficial, both in terms of pup survival and pup growth prior to weaning. The
nature and quality of parental care that pups receive, and other aspects of their early (even
prenatal) environment, have a strong influence on the pups’ development. Thus, for example,
maternal stress can slow postnatal growth in the pups, increase their stress responses later in
life and reduce territorial and reproductive success. The amount of milk that pups receive can
influence when they reach sexual maturity and their adult weight. The amount of licking and
grooming that pups receive influences their later stress responses, and this also has a knock-
on effect on subsequent generations because it influences how much females lick their own
pups once they themselves become mothers. This remarkable developmental plasticity during
these early weeks moulds pups so that they are best suited to the environments in which they
will live as adults (see Fig. 19.2).
Fig. 19.2. Both in mice and rats, higher levels of licking/grooming and active nursing result
in lower stress reactivity and fearfulness in the offspring. This also has a knock-on effect on
subsequent generations because it influences how much females lick their own pups once
they themselves become mothers. This is a remarkable example of developmental plasticity
by which the pups are thought to adjust their defensive neural systems to the demands of the
environments in which they will live as adults. (Photo courtesy of by Ivana D’andrea.)
Before weaning (and particularly during the earliest days) the pups are at risk of
infanticide, both by the mother (perhaps to optimize litter size) and by males – although the
risk of males destroying litters is reduced if the male has mated and subsequently cohabited
with a pregnant female. However, those pups that survive grow and develop rapidly, and
within 3 to 4 weeks they have fur and can thermoregulate, their eyes have opened and their
vision is developed, they have started eating solid food and are regularly venturing out of,
and increasingly away from, the nest (Latham and Mason, 2004). By 5 to 6 weeks the young
mice begin to attain sexual maturity, although this may be delayed up to 12 weeks of age if
breeding conditions are unfavourable (Van Zeegeren, 1980). The onset of sexual maturity
signals a change in family relationships, particularly among the males, where aggression
escalates between the breeding male and his sons (Berry, 1970; Bronson, 1979). Most
juvenile offspring leave their natal territory at this point and disperse to find their own.
However, some offspring remain with their parents where they may ‘tough it out’ as
subordinates with the possibility that one day they may replace their parent as the breeding
male/female (Berry, 1970; Van Zeegeren, 1980).
The rat
Puberty occurs at around 50–60 days in laboratory rats, but in wild rats is probably delayed if
resources are scarce. Matings usually occur around dusk. Dominant males mate more often
than subordinates, especially in small rat colonies where it is easier for them to monopolize
females (Hanson, 2003). Females are fertile only when in oestrus, which occurs every 4–5
days (if they do not conceive) for a period of about 6 h, during which time they may mate
repeatedly. Female fecundity is greatest in small single-male colonies. However, in large
colonies many males mate repeatedly with a female with little active competition with each
other, which causes her stress and reduces her individual fecundity. Given the opportunity,
the female will hide in her burrow for some of the oestrus, and she may turn and box with
males to resist the matings.
Nevertheless, oestrous females are proactive participants in sexual encounters, and are
attracted to males at least partly through a sex pheromone from the male preputial gland;
volatile male odours advance the onset of puberty and stimulate oestrus in females (Burn,
2008). Sexual receptivity is also signalled largely through scents. Females signal their
receptivity by presenting to the chosen male(s), hopping, darting and vibrating their ears, as
well as uttering ultrasonic vocalizations. The male also vocalizes and may chase the female.
Finally the male mounts the female, who adopts a lordosis posture. After ejaculation, the
male stands still and emits a 21–30 kHz vocalization for about 3 min; the pitch and temporal
characteristics of this vocalization closely resemble that made in aversive contexts. Gestation
lasts for about 21–25 days in the wild. In larger colonies, pups may be sired by several
different fathers, usually the more dominant males (Hanson, 2003). Male rats seem equally
fecund all year round, but most pregnancies are observed in spring and autumn when
resources are plentiful; only large females can breed year-round (Davis, 1953).
Females sometimes construct new burrows just before giving birth and will construct
more elaborate nests than those that the adults sleep in. In the laboratory at least, parturition
can take around 6 h. Several females may nest cooperatively, which usually enhances pup
survival (Calhoun, 1963). The usual seven or eight newborn pups in a litter (Davis, 1953) are
hairless, blind and ectothermic, relying on their mothers’ body warmth to maintain their
optimal temperature of about 35°C (Alberts, 2005). Olfactory cues on the mother’s belly
attract pups to suckle and pheromones deposited into the nesting material reduce their
activity levels, keeping them in the nest (Burn, 2008). If they stray from the nest, they
therefore become more active; they also emit 40–50 kHz ultrasonic vocalizations,
particularly if they become cooled, and these vocalizations cause their mother to retrieve and
return them to the nest. Dodecyl propionate from the pup preputial gland stimulates
anogenital licking by the mother, which is essential to the pups’ survival.
Pregnant and lactating females deposit scents in the nest that prevent cohabiting males
from killing the offspring, but intruding males may kill the pups. Males may kill pups that are
not their own, but under stable conditions this is rare because in small colonies almost all
pups are sired by the dominant male, and in larger colonies, the males cannot usually
discriminate which pups are or are not their own because of the polygynandrous mating
system (Berdoy, 2002). If resources are scarce or the pups’ survival is threatened, the female
may kill and/or eat the young, particularly up to about 3 days following birth. This is more
likely to occur in subordinate females in poor condition than in healthy dominant females.
The main developmental stages are as follows if resources are plentiful, but if conditions
are suboptimal they may be delayed. The fur appears from about 5 days of age, and pups start
walking from about day 10 (Alberts, 2005). The pups suckle for approximately 10 h/day,
mostly during the light period, and the dam’s milk production reaches a peak around day 15.
While some food preferences are established in utero (with pups preferring novel foods the
mother consumed during pregnancy), the pups also learn food preferences in the nest from
the taste of their mother’s milk, olfactory cues on the mother’s body and from olfactory and
gustatory cues and pheromones in her faeces. From about day 16 onwards the pups may
emerge from the nest, and are therefore attracted to these familiar food types if available
outside the nest.
Wild female rats have approximately five litters per year, successfully weaning 20–30
pups (Davis, 1953). In the wild, weaning is a gradual process, with suckling ceasing any time
between days 16 and 34 or so (Alberts, 2005). As well as being attracted to familiar foods,
the weanlings are also attracted to foods that colony members are currently eating, through
olfactory cues deposited around the food. Prepubertal rats are not attacked by adults so they
can easily access communal food (Calhoun, 1963). Thus, they avoid consuming potentially
toxic substances, while sampling a variety of the available foods.
Play in wild juvenile rats is probably important for learning foraging techniques,
developing normal social behaviour and establishing dominance hierarchies. Laboratory
studies show that social play primarily reflects adult precopulatory behaviours rather than
aggressive behaviours. This is indicated by the target of playful attack being the nape of the
neck, as in precopulatory behaviour, rather than the rump which is the usual target for
aggressive biting. Mounting and lordosis are rarely seen during play until the rats approach
puberty at around 6–8 weeks old.
After puberty males are more likely to disperse than females, but unless the environment
is unfavourable, dispersal is rare, with studies showing that fewer than 10% in a colony
disperse per year (Davis, 1953). In the wild, most rats that do disperse are unlikely to survive
and rarely integrate successfully into existing established colonies. Taken overall, females
live about 20% longer than the average male, but even within stable colonies, only about 5%
of rats live beyond 1 year.
19.7 Housing and Welfare in Laboratory Rodents

Practical and economic considerations mean that housing conditions for laboratory rodents
differ dramatically from the natural environments to which they are adapted. It is often
argued that hundreds of generations of laboratory breeding must have resulted in animals that
are well-adapted to these conditions. However, this is neither supported by studies of feral
animals or laboratory animals released in wild-type environments (e.g. Berdoy, 2002), nor by
studies on the behaviour and welfare of mice or rats under laboratory conditions (Würbel,
2001). Both laboratory mice and rats have retained the behavioural repertoire of their wild
ancestors. Although quantitative behavioural changes have occurred (e.g. laboratory breeds
are generally more docile), no qualitative differences are detectable. Their behavioural needs
seem to have changed little in the laboratory.
This has numerous implications for the welfare of laboratory mice and rats as well as for
the validity of research conducted with them (Würbel, 2001; Latham and Mason, 2004;
Sherwin, 2004; Bayne and Würbel, 2014). However, the relationship between animal welfare
and research validity is much more complex than often suggested. On the one hand, research
outcomes may be directly compromised by adverse effects of housing on the expression of
physiological and behavioural traits (Würbel, 2001), regardless of whether this is associated
with impaired well-being. Conversely, poor well-being may or may not impinge on research
outcomes, depending on whether it is associated with related behavioural or physiological
changes.
In the following sections, we briefly discuss five aspects of laboratory housing which are
likely to account for the majority of welfare problems induced by laboratory housing of mice
and rats. These five aspects are not independent, however, as they partly overlap in their
effects on the animals and may interact with each other to generate further welfare problems.
Sensory and motor deprivation
Standard cages for laboratory rodents are severely restricted in space and impoverished in
social and environmental complexity compared with the animals’ natural habitat. In addition,
the conditions are highly stable, providing little novel stimulation. Laboratory housing
conditions therefore induce sensory and motor deprivation (Würbel, 2001), and this is further
exacerbated by the current trend towards individually ventilated cage (IVC) systems. For
example, mice and rats are fed concentrated food pellets in hoppers eliminating the need for
extensive, time-consuming foraging and feeding behaviour. Sensory and motor deprivation
during development can result in impaired brain and behavioural development (van Praag et
al., 2000). Barren environments mainly affect the hippocampus and neocortex, resulting in
poor learning and memory (van Praag et al., 2000). In contrast, social deprivation, especially
during adolescence, mainly affects frontal-cortico-striatal pathways, resulting in impaired
inhibitory control of behaviour as indicated, for example, by hyperactivity, stereotypy,
compulsive behaviour and inflexibility in higher-order cognitive functions (Würbel, 2001).
Taken together, sensory and motor deprivation during development may adversely affect the
animals’ ability to adjust behaviour rapidly and flexibly to changing conditions. This may
seriously compromise their ability to cope with the demands of life as laboratory animals,
and may be attenuated by adequate environmental enrichment (Würbel, 2001; Bayne and
Würbel, 2014) (see Fig. 19.3).
Fig. 19.3. In the laboratory, rich sensory and motor stimulation and critical resources may be
provided by adequate environmental enrichment even without interfering with the animals’
visibility. (From Marashi et al., 2003, with permission from Elsevier.)
Lack of critical resources

Standard cages for laboratory rodents also lack some essential resources and prevent the
animals from performing species-specific behavioural responses to external stimuli. For
example, both mice and rats are highly thigmotactic and when faced with predator cues
rapidly seek shelter. Although there may be no real predators in a laboratory setting, some
cues may still be perceived as predator cues by the animals (Burn, 2008), but often no shelter
is provided for them to hide in. Also, rodent facilities are commonly maintained at
temperatures preferred by the personnel, while mice prefer much higher temperatures,
especially for resting and maintenance behaviour (Gaskill et al., 2009). Thus, if mice are
housed without adequate nesting material for nest building, they may suffer from cold stress.
Moreover, because temperature preferences are dependent on strain, sex, age, current
behaviour and time of day, no single temperature meets the requirements of all animals at all
times, and so provision of adequate nesting material to allow the animals control over their
microclimate seems to be the only viable solution (Gaskill et al., 2013). Rodents also
perceive visual, auditory and olfactory cues of conspecifics in neighbouring cages, but are
chronically prevented from exploring them properly. In response to the chronic thwarting of
highly motivated behaviours, the animals may get stuck in appetitive loops of behaviour
since the consummatory phase is never reached. Over time, the repeated attempts to perform
particular behaviours or to reach particular goals may develop into abnormal stereotypic
behaviour (Würbel, 2006). Especially mice develop diverse abnormal stereotypic behaviours
(e.g. bar-mouthing, jumping, somersaulting, pacing, circling, etc.; see Fig. 19.4), most of
which seem to originate from attempts to escape from the cage (Würbel, 2006). Such escape
attempts may not simply reflect aversion towards the cage conditions, but may be more
specifically elicited by the lack of some resource (e.g. shelter, nesting material) or attraction
towards social cues from neighbouring cages. However, not all animals are similarly prone to
stereotypy development. Fureix et al. (2016) recently found in C57BL/6 mice that some
individuals exhibit a passive form of coping with the chronic thwarting of highly motivated
behaviours, expressed as inactivity (awake but motionless), which is associated with
indicators of depression-like states. Furthermore, while laboratory mice of most strains
develop stereotypies at high intensities, rats generally show little stereotypy despite being
housed under seemingly similar conditions (Würbel, 2006). The reasons for this species
difference have remained elusive. One possibility is that rat cages are smaller relative to body
size and therefore may provide insufficient space for the animals to develop elaborate
stereotypies. Alternatively, rats may employ more passive coping styles, as indicated by their
high levels of inactivity under standard laboratory housing conditions.
Fig. 19.4. Mice and other rodents readily develop abnormal stereotypic behaviours when
housed in barren laboratory cages. Stereotypic jumping (white mouse to the left and nude
mouse to the right) and bar-mouthing (white mouse in the right-hand corner of the cage to the
left) originate from two different appetitive behaviours to escape from the home cage. (Photo
courtesy of by Hanno Würbel.)
Phenotypic mismatch and developmental disruption

Artificial social and/or environmental conditions may also alter or disrupt adaptive
mechanisms of developmental plasticity. Developmental plasticity provides for the fine-
tuning of genetically determined response rules. The mechanisms underlying developmental
plasticity are generally triggered by early environmental or social cues (e.g. predator cues,
maternal care) that are somehow predictive of the animals’ future environment, allowing
them to adjust their phenotype to the specific demands of that environment. Under laboratory
conditions, however, the outcome may be maladaptive if these early cues do not properly
predict the animals’ future environment, resulting in phenotypic mismatch or developmental
disruptions (Macrì and Würbel, 2006). For example, in rats (and possibly mice),
environment-dependent variations in maternal care play a significant role in ‘programming’
the pups’ later responses to stressors (Meaney, 2001). In the laboratory, mice and rats are
normally bred under conditions of minimal disturbance and interference in view of
maximizing breeding success. However, these conditions are associated with low levels of
active maternal care, resulting in offspring that are highly vulnerable to stressors later in life
(Macrì and Würbel, 2006). Thus, standard laboratory breeding conditions may produce a
phenotypic mismatch, resulting in animals that are poorly equipped to cope with the
challenges of life as a laboratory animal.
Social conflict
Unstable social groups may induce social conflict, leading to aggression, social stress and
injuries that may even cause death, especially in mice. This is further worsened by the fact
that some naturally variable aspects of the social environment, such as individual odour cues,
are unnaturally homogeneous due to inbreeding, while other naturally stable aspects of the
environment, such as the build-up of odour cues, are regularly disrupted (e.g. during cage-
cleaning). Furthermore, in male mice social conflict and overt aggression may be increased
by enrichments, especially when they can be monopolized. Thus, although the addition of
critical resources such as shelters would seem to benefit the animals as discussed above, they
may induce competition leading to aggression. This should not be taken as an argument
against providing these resources. However, it demonstrates how the different aspects listed
here may interact and highlights the difficulties in designing housing conditions in a way that
integrates these different aspects into a functioning system. Although overt aggression
appears to be less of a problem in laboratory rats, social stress as a result of non-harmonious
groups can occur. This suggests that aggression should be monitored and individuals
separated from non-harmonious groups.
Environmental stressors
Finally, laboratory animals are generally exposed to a multitude of environmental stressors,
many of which are unpredictable. The sensory environment itself may cause stress, with the
animals being inescapably exposed to high artificial light intensities, audible and ultrasonic
sounds that can be very loud, vibration (especially in IVCs) and salient odours that can
potentially include other laboratory animals (even non-rodent species), cleaning products and
myriad other chemicals (Latham and Mason, 2004; Burn, 2008). Normal husbandry events,
such as handling and cage-cleaning, and procedures specific to individual experimental
purposes can cause further stress. In addition, the confines of the barren cages provide the
animals with little control over these stressors. Predictability and controllability have long
been shown to be the major determinants of the stress response. Moreover, because many
stressful events occur in the rodents’ resting phase (when it is light), the animals are less able
to cope with them than if they occurred during their active phase. Consequently, it is not
surprising to find evidence for stress and anxiety-related disorders (e.g. compulsive
behaviours such as barbering in mice) and depressive-like states (e.g. immobility in the
forced swim test in mice, anhedonia in rats).
19.8 Conclusions
Despite hundreds of generations of laboratory breeding, mice and rats have retained their
behavioural repertoire. Although quantitative behavioural changes have occurred, the
animals’ basic behavioural needs and their natural response rules seem to have remained
largely unchanged by domestication in the laboratory. Laboratory conditions that ignore the
animals’ ethology are therefore likely to induce poor welfare and compromise the validity of
animal experiments.
By understanding the ethology of laboratory animals, we can gain insight into how the
captive environments may impact upon animal welfare and formulate testable hypotheses
about potential refinements to laboratory housing and husbandry practices (Olsson and
Dahlborn, 2002; Olsson et al., 2003). After all, a laboratory mouse is still a house mouse, and
a laboratory rat is still a Norway rat.
References
Alberts, J.A. (2005) Infancy. In: Whishaw, I.Q. and Kolb, B. (eds) The Behavior of the Laboratory Rat: A Handbook with
Tests. Oxford University Press, Oxford, pp. 266–277.
Barnett, S.A. (2005) Ecology. In: Whishaw, I.Q. and Kolb, B. (eds) The Behavior of the Laboratory Rat: A Handbook with
Tests. Oxford University Press, Oxford, pp. 15–24.
Baumgardner, D., Ward, S. and Dewsbury, D. (1980) Diurnal patterning of eight activities in 14 species of muroid rodent.
Animal Learning and Behaviour 8, 322–330.
Bayne, K. and Würbel, H. (2014) The impact of environmental enrichment on the outcome variability and scientific validity
of laboratory animal studies. Scientific and Technical Review OIE 33, 273–280.
Berdoy, M. (2002) The laboratory rat: a natural history. Available at: http://www.ratlife.org (accessed 24 February 2006).
Berry, R. (1970) The natural history of the house mouse. Field Studies 3, 219–262.
Blanchard, R.J., Dulloog, L., Markham, C., Nishimura, O., Nikulina Compton, J., Jun, A., Han, C. and Blanchard, D.C.
(2001) Sexual and aggressive interactions in a visible burrow system with provisioned burrows. Physiology & Behavior
72, 245–254.
Bronson, F. (1979) The reproductive ecology of the house mouse. The Quarterly Review of Biology 54, 265–299.
Brown, R. (1953) Social behavior, reproduction and population changes in the house mouse (Mus musculus L.). Ecological
Monographs 23, 217–240.
Burn, C.C. (2008) What is it like to be a rat? Rat sensory perception and its implications for experimental design and rat
welfare. Applied Animal Behaviour Science 112, 1–32.
Calhoun, J.B. (1963) The Ecology and Sociology of the Norway Rat . Public Health Service Publication No. 1008. US
Department of Health Education and Welfare, Bethesda, Maryland.
Crawley, J. (2000) What’s Wrong with my Mouse? Behavioural Phenotyping of Transgenic and Knockout Mice. Wiley,
Chichester, UK.
Crowcroft, P. (1966) Mice All Over. Foulis, London.
Davis, D.E. (1953) The characteristics of rat populations. The Quarterly Review of Biology 28, 373–401.
Fureix, C., Walker, M., Harper, L., Reynolds, K., Saldivia-Woo, A. and Mason, G. (2016) Stereotypic behaviour in standard
non-enriched cages is an alternative to depression-like responses in C57BL/6 mice. Behavioural Brain Research 305,
186–190.
Galef, B.G. Jr, Mason, J.R., Preti, G. and Bean, N.J. (1988) Carbon disulfide: a semiochemical mediating socially-induced
diet choice in rats. Physiology & Behavior 42, 119–124.
Gaskill, B.N., Rohr, S., Pajor, E.A., Lucas, J.R. and Garner, J.P. (2009) Some like it hot: mouse temperature preferences in
laboratory housing. Applied Animal Behaviour Science 116, 279–285.
Gaskill, B.N., Gordon, C.J., Davis, J.K., Lucas, J.R., Pajor, E.A. and Garner, J.P. (2013) Impact of nesting material on mouse
body temperature and physiology. Physiology & Behavior 110, 87–95.
Hanson, A. (2003) Rat behaviour and biology. Available at: http://www.ratbehavior.org (accessed 22 March 2007).
Hurst, J., Payne, C., Nevison, C., Marie, A., Humphries, R., Robertson, D., Cavaggioni, A. and Beynon, R. (2001)
Individual recognition in mice mediated by major urinary proteins. Nature 414, 631–634.
Latham, N. and Mason, G. (2004) From house mouse to mouse house: the behavioural biology of free-living Mus musculus
and its implications in the laboratory. Applied Animal Behaviour Science 86, 261–289.
Macrì, S. and Würbel, H. (2006) Developmental plasticity of HPA and fear responses in rats: a critical review of the
maternal mediation hypothesis. Hormones and Behavior 50, 667–680.
Marashi, V., Barnekow, A., Ossendorf, E. and Sachser, N. (2003) Effects of different forms of environmental enrichment on
behavioral, endocrinological, and immunological parameters in male mice. Hormones and Behavior 43, 281–292.
Meaney, M.J. (2001) Maternal care, gene expression, and the transmission of individual differences in stress reactivity
across generations. Annual Review of Neuroscience 24, 1161–1192.
Meehan, A. (1984) Rats and Mice: Their Biology and Control. Brown Knight and Truscott Ltd, Tonbridge, UK.
Olsson, A. and Dahlborn, K. (2002) Improving housing conditions for laboratory mice: a review of ‘environmental
enrichment’. Laboratory Animals 36, 243–270.
Olsson, I.A.S., Nevison, C.M., Patterson-Kane, E.G., Sherwind, C.M., Van de Weerde, H.A. and Würbel, H. (2003)
Understanding behaviour: the relevance of ethological approaches in laboratory animal science. Applied Animal
Randall, C. (1999) Vertebrate Pest Management – A Guide for Commercial Applicators . Extension Bulletin No. E-2050.
Michigan State University, East Lansing, Michigan.
Sherwin, C.M. (2004) The influences of standard laboratory cages on rodents and the scientific validity of research data.
Animal Welfare 13(Suppl. 1), S9–S15.
Traweger, D., Travnitzky, R., Moser, C., Walzer, C. and Bernatzky, G. (2006) Habitat preferences and distribution of the
brown rat (Rattus norvegicus Berk.) in the city of Salzburg (Austria): implications for an urban rat management. Journal
of Pest Science 79, 113–125.
van Praag, H., Kempermann, G. and Gage, F.H. (2000) Neural consequences of environmental enrichment. Nature Reviews
Neuroscience 1, 191–198.
Van Zeegeren, K. (1980) Variation in aggressiveness and the regulation of numbers in house mouse populations.
Netherlands Journal of Zoology 30, 635–770.
Wilson, D.E. and Reeder, D.A.M. (2005) Mammal Species of the World: A Taxonomic and Geographic Reference, 3d edn.
Johns Hopkins University Press, Baltimore, Maryland.
Würbel, H. (2001) Ideal homes? Housing effects on rodent brain and behaviour. Trends in Neuroscience 24, 207–211.
Würbel, H. (2006) The motivational basis of caged rodents’ stereotypies. In: Mason, G.J. and Rushen, J. (eds) Stereotypic
Animal Behaviour: Fundamentals and Applications to Welfare. CAB International, Wallingford, UK, pp. 86–120.
Index
Note: Page numbers in bold type refer to figures

Page numbers in italic type refer to tables
abnormal aggression 122–123

abnormal behaviour 2, 7–8, 24, 55–58, 119–134, 152, 266
repetitive 55
stereotypic 283, 283
see also stereotypies
adaptive behaviour 5
adenohypohyseal hormones 28–30, 36, 123
adipocytes 204
adrenocorticotropin (ACTH) 28–30, 36, 123
affiliative behaviour 135, 174, 192
see also grooming; licking
Africa 23, 189, 200, 214, 235
aggression 121–123, 143
abnormal 122–123
dominance-related 215
mixing-induced 215
see also dominance
aggressive behaviour 17, 147, 156, 192, 192
agreeableness 105
Alaska 199, 257
Allen’s rule 169
allogrooming 192, 244, 246, 259, 275
allomimetic behaviour 201
allorubbing 244–248, 245
allostasis 123
altruistic behaviour 81–83
American mink (Neovison vison) 255, 258
amygdala 28–29
anatomical traits 19, 20
ancestors 8–9, 13, 19, 63, 79, 85, 119, 155
cats 241, 242, 243

dogs 23–24, 91
mice 281
pigs 214
poultry 162
rats 273, 281
sheep and goats 199
androstenone 218
anhedonia 56–57
Antarctica 273
anticipatory behaviour 74, 133
antidiuretic hormone (ADH) 27
appearance, neotenized 228, 229, 229
appetitive behaviour 40, 43–44, 52, 56
applied ethology 6–10
Archaeopteryx 18
Aristotle 3
artic foxes (Alopex lagopus) 231, 255, 259–264
artificial insemination 195
artificial selection 83
Asia 22, 153, 199–200, 228, 255
associative learning 63–67
audition 232–235
aurochs (Bos primigenius) 189
Australia 189, 200–201, 204, 215, 219, 276
Austria 5, 277
avoidance, learning 15, 28, 72–73, 145, 174, 203–204, 212, 276
barns, freestall (cubicle) 189, 190, 193

basenji 235
bathing, dust- 39, 44–45, 50, 51, 132, 160–162, 161, 167
battery cages 6, 6, 23, 154–156, 154
beef calves 196–197
beef production 189
cow–calf operations 189–191, 191
behaviour 2, 3–10
abnormal 2, 7–8, 24, 55–58, 119–134, 152, 266
repetitive 55
stereotypic 283, 283
adaptive 5
affiliative 135, 174, 192
aggressive 17, 147, 156, 192, 192
allomimetic or synchronous 201
altruistic 81–83
anticipatory 74, 133
positive 133
appetitive 40, 43–44, 52, 56
breed-typical 169
change 63
clever 62
complex 11
consummatory 41, 43–45, 52–53
control 7
courtship 43, 44, 164, 205
defensive 147
desired 13, 57–58

disorders 7–10, 119–123
disturbances 119–123
drinking 47
evolution 2
explanation 47–48, 48
exploratory 42, 217
feeding 32–33, 171–173, 193, 194
foraging 54
free-running 242
frustration-related, chickens 130
genetics 11–25
goal-directed 67
habitual 67
highly motivated 39–41
human 7, 139
infanticidal 261
instinctive 3–5, 11
interactive 147
interventionist 174
intriguing 62
investigative 181
locomotor play 209–210
maternal 17, 27, 33–36
mating 164–165, 164, 175–178, 195–196
motivated 46, 53, 58
mutualistic 81
natural 54, 58, 127, 158
normal 24, 55, 119–121, 120, 152
observed 17, 62, 66, 111
oestrus 205
optimal 20, 21
over-motivated 52
parental 86
plasticity 113
play 90–103
positive anticipatory 133
problematic 13
proceptive 222
protective 174
reproductive 2, 33–36, 42, 79–89, 175
resting 171
roosting, chickens 129
selfish 81–82
sexual 84–86, 162–165, 175–178, 195–196, 222
signalling 145
social 2, 27, 79–89, 155–158, 174–175, 192–193, 211
sow nest-building 4, 4, 34, 35
specific natural 54
state-dependent 114
stereotypic 55–56, 120–121, 121, 266, 283, 283
stress-related 36, 133
submissive 156
syndromes 105, 108–111
territorial 259
traits 19
type 105
udder-seeking 208, 209
undesirable 13
variation, observed 104
vigilance 182
behavioural ecology 21
belly-nosing 225, 226
Bergman’s rule 169
birds
chickens 17, 24, 45, 129, 130
domestic 158
doves 45–46
feather-pecking 14, 121, 122, 157–158
fossil 18
nest-building 130
nipple drinkers 160, 160
social problems 157
wild 158
blue fox 257–258, 257, 264, 268
body maintenance 160–161
see also dust-bathing; grooming; preening
body posture 232, 233
Bovidae 199
brain
amygdala 28–29
behaviour regulation 26–30
consciousness 28
decision making 28–29, 33
emotions 28
hypothalamus 26–28, 36
mechanisms 93, 99
neurotransmission 28–29
brainstorm auditory evoked responses (BAER) 181
Brambell Committee 126–127
breed-typical behaviour 169
breeding 153, 195, 223
browsing 173, 203
cross- 247
in 284
seasonality 33–34, 50–51, 85, 162, 189, 204, 221–222
see also reproduction
British Farm Animal Welfare Council 127
bucks 206
bulls see cattle
Burghardt’s Resource Surplus Theory 96
burrows 274
Caenorhabditis elegans 12, 15

cages
battery 6, 6, 23, 154–156, 154
enrichment 266, 266
individually ventilated cage (IVC) systems 282
calves see cattle
Canada 255–258
canids, puppies 229–230, 236–238, 237

canine cognition 238–239
cannibalism 6–7, 76, 121–122, 157–158, 261
see also infanticide
Cannon, W.B. 123
captive animals 39, 58
captivity 2, 116, 119–121, 174, 179–180
capybara (Hydrochoerus hydrochaeris) 231
cat–human relationships 251
categorization 70–72
cats
allogrooming 244
mutual 246
allorubbing 244–248, 245
cheek-gland secretions 244, 247
companionship 242
controlling pests 241
cross-breeding 247
feral 244–247
gene flow, feral and domestic 247, 248
hairless Sphynx 242
handling 141
indiscrimination 244, 244
kittens 242, 248–251
moggies 241
Munchkin (dwarfism) 242
purring 249
Ragdoll 242
scent-marking 251
scruff reflex 249
spray-marking 244, 246
toms 244
vomeronasal organ (VNO) 243–244
wild (Felis silvestris lybica) 241–243
cattle
beef 189–191, 191, 196–197
finishing/growing 191, 191
breeds 189
bulls 192, 195, 196
libido 196
calves 196–197, 196–198
beef 196–197
dairy 197
calving 196
cows 195–198
dairy 189, 193–195, 212
female–female mounting 195
placentophagia 196
cow–calf operations 189–191, 191, 196
feral 192
freestall (cubicle) 189, 190, 193
Holstein-Friesian 189
loose housing 189, 190
polyoestrous 195
teat feeding systems 197
cheek-gland secretions 244, 247

chemical signals 80
chemoreceptors 15
chickens see poultry
China 19, 153–154, 200, 214, 255–256, 273
cholecystokinin (CCK) 32
circadian rhythm 29–30, 274, 274
clever behaviour 62
climate 200, 204, 210, 274
clitoral winking 176, 176
coastal foxes 260
cognition 63, 73–75
associative learning 63–67
canine 238–239
judgement bias task 74
meta- 75
social 72–73
cognitive bias task 74, 74
cognitive maps 68
colonies, rat 277
commercial conditions 116, 156
commercial production 154, 165
communication 72, 80–81
after hatching 166
human–animal 139–141, 142
companionship 242
competition 83–84
competition theory of motivation 45, 46
concept formation 70–72
conditioned response (CR) 63–64
conditioned stimulus (CS) 63–64
conditioning 63–64
conscientiousness 105
consciousness 28
conservation 8–10
consumer demand theory 131
consummation 41, 55
consummatory behaviour 41, 43–45, 52–53
consummatory stimulus 55
control 29–30, 32–33
control systems model 49, 50
cooperation 84, 215, 238, 244, 251
coping styles 105–107
copulation 95, 165, 177, 278
corticosteroids 28
courtship 43–44, 44, 88, 201, 205, 223
displays 19, 43, 84, 164, 247
stimulus–response patterns 164
cows see cattle
coyote (Canis latrans) 228–231
CRIPSR-Cas9 technique 17
cross-breeding 247
cross-fostering piglets 223
cross-sucking 197
cubs see foxes and mink
cues
body posture 133, 145, 232, 259, 277
developmental plasticity 279, 283
fearfulness 107, 114, 143, 217
suckling 236–237
visual 139, 176, 207, 211, 232
cyclic AMP phosphodiesterase 15
dairy calves/cows see cattle

Darwin, C. 4, 11, 18–19
The Expression of the Emotions in Man and Animals 4
decision-making system 29, 29
declarative representations 66
defensive behaviour 147
Denmark 256–258
developed countries 154
developed world 210
developing world 189, 200, 210
discrimination 70–73, 76, 86, 145, 169, 180–182
diurnal rhythms 161–162, 163, 193–195
diurnal variation 204
DNA (deoxyribonucleic acid) 11–12, 15–17, 22–24, 241–242
dogs
audition 232–235
basenji 235
body posture 232, 233
canine cognition 238–239
and coyote (Canis latrans) 228–231
feral 230–231
and foxes
arctic (Alopex lagopus) 231, 255, 259–264
red (Vulpes vulpes) 229, 255, 258–264
and golden jackal (Canis aureus) 228
guide 146
human attentional states 146
human communicative signals 146
neotenized 228, 229
oestrous cycle 235–236
olfaction 232–234
pet 230, 230
puppies 236–238
suckling 236, 237
raised leg display (RLD) 233–234
raised leg urination (RLU) 233, 238
reduced welfare 128
relaxant property odours 233, 234
stool-eating 232
therapeutic captivity (pet-assisted therapy) 231
vision 232–233
whelping 236
and wolves 234–235
grey (Canis lupus) 228, 229
working 146
domestication 2, 11–25, 23–24, 79, 88, 143
dominance
hierarchy 175, 192, 259, 281

peck order 84, 85, 230, 248
dominance-related aggression 215
dominant animals 232
dominant–subordinate relationships 192
dopamine 28, 93, 114
dopaminergic neurons 28
dopaminergic system 33
doves 45–46
drinking behaviour 47
ducks 152–168
dust-bathing 39, 44–45, 132, 160–162, 161, 167
motivational model 50, 51
eating 32–33, 44–45, 121, 167, 215, 218, 265, 274–276, 279
effective learning 72
egg-laying 161–162, 165–167
Egypt 241
embryology 15
emotions
human-animal 145
motivation link 52–53
negative 41, 52–54, 146
pleasurable 99
positive 39–41, 52, 98–99, 219
unpleasant 137
endothermy 96
energy intake 32
environment
barren 221, 226, 282
enrichment 282, 282
influence 13–14
social 63, 72, 88, 139, 143–144
epigenetics 17–18
episodic memory 70, 218
equine facial action coding system (EquiFACS) 175
equine vocalizations 181
Eurasia 189
Europe 4–5, 22, 153, 191, 214–215, 241–242, 273
European mink (Mustela lutreola) 255
European Union (EU) 272
evolution 2, 11–25
evolutionary stable strategy (ESS) 21
ewes see sheep
experimental ethology 5
exploratory behaviour 42, 217
Expression of the Emotions in Man and Animals, The 4
extensive systems 213
extraversion 105
factory farming 6
faeces 232, 233
familiarization 111
farm animals 7, 88, 126, 137–139, 142, 146–147, 204
farmed fur animals 267
fearfulness
reduction 217, 258, 268
feather-pecking 14, 121, 122, 157–158
feed-forward processes 46–47
feeding
behaviour 32–33, 159–161, 171–173, 193, 194
hand- 140
overnutrition 206
teat systems 197
female–female mounting 195
feral animals 119
cats 244–247
cattle 192
dogs 230–231
horses 171
pigs 215, 221
fighting 216, 216
play- 91, 95, 99
Finland 255–257
flehmen response 176, 177, 195, 205
floor housing system 6, 6, 154
floor laying 165–166, 166
foals see horses
follicle-stimulating hormone (FSH) 28, 34
food
intake 32–33
neophobia 159
novel 43, 64, 209, 210, 275, 276, 280
preferences 205, 275–276, 280
foraging 79, 80, 158, 159, 170–173, 193
efficiency 204
habitats 171
natural 54
optimal theory 21
redirected 56
time 203, 263
fostering, cross- 223
foxes and mink
American mink (Neovison vison) 255, 258
appropriate handling methods 267, 268
arctic fox (Vulpes lagopus) 231, 255, 259–264
blue fox 257–258, 257, 264, 268
cages 265, 266, 266
coastal fox 260
cubs 258–261
European mink (Mustela lutreola) 255
fur chewing 269
group housing 259, 267
kits 260, 263
foxes and mink (continued)lemming foxes 260
maternal behaviour 262, 262
mousing 264
red fox (Vulpes vulpes) 229, 255, 258–264
silver fox 256–258, 256, 264, 268
solitary hunters 264
vixens 255, 258–261, 267–268

wild 265
free-living animals 5
free-living mice 273
free-ranging horses 173
free-ranging pigs 219, 225
free-ranging sows 223
free-running behaviour 242
freestall (cubicle) barns 189, 190, 193
fruit flies (Drosophila melanogaster) 12, 12, 15
Fulani herdsmen 144
fur-chewing 269
gene flow, of cats 247, 248
generalization 70–72
genes 2, 11–17
genetic control 12
genetic determinism 13
genetic inheritance 11–13, 18
genetic modification 23
genetics 13–14, 214
coding 16
genome-wide association studies (GWAS) 16
Germany 182
ghrelin 33
glands
Harderian 278
scent 202, 218, 277
glandular secretions 233–234
goats see sheep and goats
golden jackal (Canis aureus) 228
grazers 30, 203
Great Britain 5
great tit (Parus major) 109, 109
greenish warbler (Phylloscopus trochiloides) 19
Greenland 257
grey wolf (Canis lupus) 228, 229
grooming 274
mutual (allo-) 192, 244, 246, 259, 275
self- 274
social 192, 244, 246, 259, 275
group foraging 80
group living 79–80
group selection 83
group-housing 216, 259, 267
growth hormone (GH) 27, 30
guardian animals 212
habitat competition 199, 200

habitual behaviour 67
hairless Sphynx 242
Haldane, J.B.S. 82
Hamilton, W.D. 82
Hamilton’s Rule 82
hand-feeding 140
handling
inconsistent treatment 217

methods 267, 268
response 107
Harlow, H. 88
Heinroth, O. 5
heritability 258
hider species 206
Holland 5
Holstein-Friesian cattle 189
homeostasis 46, 54
thermal 54
Homo sapiens 22
hormonal control 32–33
hormonal regulation 33–36
hormones 35
adenohypohyseal 27, 28–30, 36, 123
behaviour regulation 28
follicle-stimulating (FSH) 28, 34
hunger regulation 32
luteinizing (LH) 28, 34
melanocyte-stimulating (MSH) 27
nest-building 35
reproductive regulation 34
stress 73, 266
suckling influence 36
thyroid 28
horses
clitoral winking 176, 176
equine facial action coding system (EquiFACS) 175
equine vocalizations 181
feral 171
flehmen response 176, 177
foaling 178
foals 71, 92, 140, 144, 179, 182
free-ranging 173
geldings 175
gentle handling 143–144
mares 175–179
maturation training 179
Miller’s imprint training 179
mutual grooming 182, 183
neophobic responses 172
New Forest ponies 172
normal behavioural parameters 171, 172
nursing 179
parasympathetic control 178
parturition 178
postpartum pause 178
stabled 173
stallions 176–178
tail-raising 176
vibrissae (whiskers) 182
wild 169
house mouse (Mus musculus) 272–275, 285
housing
cages 6, 14, 154–159, 161, 165, 256–257, 264–265

floor system 6, 6, 154
group 216, 259, 267
laboratory 281–284, 282
loose 189, 190
human behaviour 7, 139
human facial expressions 145
human intervention 169
human personality 104–106
human–animal interactions 139–142, 142
human–animal relations 135–149
dairy cow farms 135, 136– 137
general relationship 137–139, 138
individual relationship 137–139
mutual behaviour 137
pleasant or unpleasant emotions 137, 138
social partner 144
human–dog relationship 230–231
human–pig relationship 217
husbandry 7, 152
(5)-hydroxytryptamine (5-HT) 29
hypophysis 26–30
hypothalamic–pituitary–adrenal (HPA) 106
axis 36
hypothalamus 26–30
Iceland 255
iguanas, marine (Amblyrhynchus cristatus) 231
imprint training 179
imprinting 167
inbreeding 284
incubation 165–167
India 200, 214, 241
indiscrimination 244, 244
industrialized world 169, 189
infanticidal behaviour 261
infanticide 122, 244, 261, 279
inheritance, genetic 18
innate behaviour 5
instinctive behaviour 3–5, 11
insulin 33–36
intensive systems 213
interactions
human–animal 139–142, 142
parent–offspring 86–87, 166
interactive behaviour 147
interbreeding 214
interventionist behaviour 174
intriguing behaviour 62
investigative behaviour 181
Iran 214
Italy 231
Japan 214
Jersey cattle 189

junglefowl
intercrossing of 16, 16
red (Gallus gallus) 153
kids see sheep and goats

kin selection 82–83
kits see foxes and mink
kittens see cats
Kluger Hans 146
knockout animals 17
laboratory housing 281–284, 282
lambs see sheep
lateral recumbency 170, 171
learning
ability 18
associative 62–67
avoidance 15, 28, 72–73, 145, 174, 203–204, 212, 276
effective 72
imprinting 88, 167, 179, 250
rats 14, 14
social 72, 76, 143–144, 147
leptin 32
Leroy, C.G. 4
libido 196
licking 196, 207
maternal 207, 207, 208, 279, 279, 280
social 192, 244, 259, 275
locomotor play behaviour 209–210
locomotor-rotational (LR) play 91–93, 92, 97–99
locomotory activity 170
locomotory stereotypies 121
Lorenz, K. 5
psychohydraulic model 48–50, 49
lovebirds (Agapornis spp.) 11
lupine vocalizations 235
luteinizing hormone (LH) 28, 34
male tree shrews (Tupaia belangeri) 125

mares see horses
marine iguanas (Amblyrhynchus cristatus) 231
maternal behaviour 17, 27, 33–36, 262, 262
nursing 80, 81, 87, 87, 179, 224
maternal licking 207, 207, 208, 279, 279, 280
mating
behaviour 164–165, 164, 175–178, 195–196
optimal time 222
patterns 115, 165
promiscuity 85, 162
puberty 218, 238, 259, 279–281
season 97
selection 84, 163–165
systems 85, 163
unregulated 242
matrilineal groups 215
maturation training 179
melanocyte-stimulating hormone (MSH) 27
memory 67–70, 69, 70
spatial 68
Mendel, G. 11
mental capacity 221, 222
Merino sheep 204
metacognition 75
Mexico 153
mice
abnormal stereotypic behaviour 283, 283
allogrooming 275
burrows 274
free-living 273
house (Mus musculus) 272–275, 285
laboratory housing 281–284
long attack latency 125
pups 278–280, 284
short attack latency 125, 125
urinary odour cues 275
Middle East 200, 214
mink see foxes and mink
mitochondria (mtDNA) 22
mixing-induced aggression 215
models
control systems 49, 50
dust-bathing motivational 50, 51
Lorenz’s psychohydraulic 48–50, 49
Rescorla–Wagner 64
Sollwert–Istwert 49
standard stress 123, 124
thermostat-like 49
moggies 241
monogamy 85, 163, 230
Morgan, M. 47
mortality 211
motivated behaviour 46, 53, 58
motivation
feeding 32, 42
maternal 86
model 50, 51
nursing 36
sexual 223
motivational model of dust-bathing 50, 51
motivational states 38–39
mounting, female–female 195
mousing 264
Munchkin (dwarfism) 242
mutual grooming 182, 183
mutualistic behaviour 81
natural selection 18, 81–83, 95–96, 129

neonatal mortality 211

neophobic responses 172
neotenized appearance 228, 229, 229
nest-box design 165
nests, building 13, 54, 130, 223, 262, 268, 282
neurohypophysis 26
neuropeptides 29
neuroticism 105
neurotransmission 28–29
New Forest ponies 172
New Guinea 214
New Zealand 201
no cues treatment 42, 43
non-rapid movement sleep (NREM) 30–31, 31
noradrenaline systems 93
North America 211, 231, 255, 273
North American whooping crane 8, 9
Norway 40, 255–257, 273
Norway rats (Rattus norvegicus) 272–273, 278, 285
circadian rhythm 274, 274
novel food 43, 64, 209, 210, 275, 276, 280
novelty 97
nursing 80, 81, 87
communal 278
frequency 87
grunting 224
painful 179
nuzzling 80–81,
object play 91
observed behaviour 17, 62, 66, 111
variation 104
oestrus
behaviour 205
cycle 235–236
standing 222
offspring 167, 179, 195
olfaction 232–234, 243
ability 181–182
signals 233–234
stimuli 176
ontogeny 5, 8, 15, 105
openness 105
opioids 132
optimal foraging theory 21
Orkney 202
ovariectomy 177
overnutrition 206
oxytocin 17, 35, 38, 251
pain 127
parasympathetic control 178
parental behaviour 86
parent–offspring interactions 86–87, 166
Pavlov, I. 63, 107
peacock 19
peck order 84, 85, 230, 248
Pekin duck 153–154
personality 104–118, 113, 143
assays 110–113, 111
traits 105
pest control 241
pets 135–137, 146
dogs 230, 230
phylogeny 5, 8, 169
physiological change 3
physiology 26–37
pigs
androstenone 218
belly-nosing 225, 226
brain cross-section 26–29, 27
different commodities 132, 132
distress-induced vocalizations 218
episodic memory 218
feral 215, 221
fighting positions 216, 216
free-ranging 219, 223, 225
group-housed 216
inconsistent treatment 217
interbreeding 214
matrilineal groups 215
mental capacity 221, 222
piglets 223, 224–225
suckling 224, 224
sows 4, 4, 34, 35, 215–216, 222–225
free-ranging 223
nest-building behaviour 4, 4, 34, 35
standing oestrus 222
tail-biting 121
training 67, 68, 73
wallowing 220, 221
and wild boar (Sus scrofa) 214–215, 218–223
placentophagia 196
plasticity development 279, 279, 283
play 87–88, 90–91, 96–99
behaviour 90–103, 107–108, 107
functional benefits 94–96
locomotor-rotational (LR) 91–93, 92, 97–99
object 91
rough-and-tumble (R-T) 93–95
social 91–99
play-bow 93, 93
play-fighting 91, 95, 99
playfulness 93
pleiotropic effects 114
Poland 255–257
polyandry 85, 230, 247, 280
polygyny 85
polymorphism 114–115
polyoestrous 195, 205
post-weaning 225
postpartum period 34
poultry
chickens 17, 24, 45, 153–168
behaviour 129, 130
roosting 129
ducks 152–168
social hierarchy 156, 156
turkeys 155, 155
vocal repertoire 155, 156
predators 79
preening 46, 130, 160, 161, 162
preference test 131, 131
prepartum period 34
problematic behaviour 13
procedural representations 66
proceptive behaviour 222
see also courtship
production 34, 135
beef 189–191, 191
commercial 154, 165
conditions 116
progesterone 34
prolactin (PL) 27, 34
promiscuity 85
protective behaviour 174
proto-domestication 22
psychohydraulic model (Lorenz) 48–50, 49
punishment 64–66
puppies, canids 229–230, 236–238, 237
pups 236, 278–280, 284
cannibalism (rodents) 122
dog 145–146
fur seal 95
mice/rats 31, 35, 122, 278–281, 284
purring 249
quantitative trait loci (QTL) 16
Ragdoll 242
raised leg display (RLD) 233–234
raised leg urination (RLU) 233, 238
rams 205
rapid eye movement (REM) 30–31, 31, 170, 235
rats
abnormal stereotypic behaviour 283, 283
bright 13–14
colonies 277
dull 13–14
laboratory housing 281–284
maze-learning 14, 14
Norway (Rattus norvegicus) 272–273, 274, 274, 278, 285
pups 31, 35, 122, 278–281
urine marking 277
wild 275, 281

Ray, J. 3
recognition
ability 211
cross-modal 139
glands 234
individual 137–139
maternal 207
social 155, 202
recolonization 215
red fox (Vulpes vulpes) 229, 255, 258–264
red junglefowl (Gallus gallus) 153
reinforcement 64–66
relationships
cat-human 251
human–dog 230–231
human–pig 217
see also human–animal relations
repetitive behaviour, abnormal 55
reproduction 20, 38
behaviour 2, 33–36, 42, 79–89, 175
oestrus 205, 222
cycle 235–236
polyoestrous 195, 205
see also courtship; mating
Rescorla–Wagner model 64
resource holding potential (RHP) 248
resting behaviour 171
reticulorumen 203
retinohypothalamic tract (RHT) 29
rhythms
circadian 30, 162
diurnal 162, 193
feeding 162
risk dilution 79
ritualization 19
rodents
pup-cannibalism 122
see also mice; rats
roosting 162
rough-and-tumble (R-T) play 93–95
ruminants 203
see also cattle; sheep and goats
rumination 193
Russia 19, 255–256
salivations levels 170

Scandinavia 211
scent
sensitivity in pigs 217
scent-marking 243–244, 251, 259
behaviour 251
scruff reflex 249
season (heat) 235, 238
seasonality 33–34, 50–51, 85, 162, 189, 204, 221–222
seaweed 202
secretions
cheek-gland 244, 247
glandular 233–234
hormones 27, 27
selection
group 83
natural 18, 81–83, 95–96, 129
self-grooming 274
self-rewarding 98–99
serotonin 29
sexual behaviour 84–86, 162–165, 175–178, 195–196, 222
Seyle, H. 123
sheep and goats
allomimetic or synchronous behaviour 201
Artiodactyla 199
Bovidae 199
bucks 206
conserving water 204
does 206, 209
ewes 201, 205–211
extensive 213
grazers 203
habitat competition 199, 200
hider species 206
intensive 213
kids 208–211
lambs 208–211
locomotor play behaviour 209–210
maternal licking 207, 207
maternal recognition 207
Merino sheep 204
negative handling 212
neonatal mortality 211
novel food exposure 209, 210
overnutrition 206
rams 205–206
reticulorumen 203
ruminants 203
seasonally polyoestrous 205
selective attachment 207
sheltered areas 205
social interactions 202, 202
Tribe Caprini 199
udder-seeking behaviour 208, 209
undernutrition 206, 211
wild 199–201, 205–206
shrews, tree (Tupaia belangeri) 125
Siberia 19, 199
signalling behaviour 145
signals
calls 238
chemical 80
olfactory 233–234
tactile 80
visual 80
silver fox 256–258, 256, 264, 268
single-gene influence 14–15
Skinner, B.F. 4, 62
sleep 30–31
non-rapid eye movement (NREM) 30–31, 31
rapid eye movement (REM) 30–31, 31
slow-wave sleep (SWS) 170
sociability 91, 94, 106, 110–112
social animals 7
social behaviour 2, 27, 79–89, 155–158, 174–175, 192–193, 211
competition 83–84
scramble 83
social cognition 72–73
social grooming 192, 244, 246, 259, 275
social interactions 2, 93, 166
of mixed sheep 202, 202
social learning 72, 76, 143–144, 147
social licking 192
social life 22
social play 91–99
social relationships 142–143
sociality 258–259, 267
socialization 140–147, 237–238, 250–251
interspecific 250
intraspecific 250
period 141–142, 238
solitary hunters 264
Sollwert–Istwert model 49
South America 199, 215
sows see pigs
Spain 153
Spalding, D. 4
sporting animals 135
spray-marking 244, 246
stallions 176–178
standard stress model 123, 124
standing oestrus 222
stereotypic behaviour 55–56, 120–121, 121, 266
abnormal 283, 283
stereotypies 120–121, 266–267
locomotory 121
prevention 220
stickleback, three-spined (Gasterosteus aculeatus) 108–109, 108
stimuli 3, 14, 28–30, 39–41, 64–66, 70–71, 71, 80, 121, 123, 179–180, 237
conditioned (CS) 63–64
consummatory 55
different effects 65, 65
eliciting 42, 51
external 38–43, 56, 126, 129
internal 41–43, 126
valuation 51
stool-eating 232
Strange Situation Test (Ainsworth) (SST) 230
stress 119–134
reactivity 29
reduction 258, 262, 265–266
responses 123–124
stress-related behaviour 36, 133
submissive animals 232
submissive behaviour 145, 156
suckling 179–180, 249
piglets 224, 224
puppies 236, 237
tactile signals 80
sucrose consumption 56–57
suprachiasmatic nucleus (SCN) 29
Sweden 257–259
switching costs 45–46
synchrony, feeding 220
tactile signals 80
tail-raising 176
taurine cattle (Bos taurus) 189
teat feeding systems 197
temperament 105–108
territorial behaviour 259
therapeutic captivity 231
thermal homeostasis 54
thermoneutral zone 220
thermoregulation 278
thermostat-like models 49
threatening behaviour 145
three-spined stickleback (Gasterosteus aculeatus) 108–109, 108
thyrotropin (TSH) 28
Tibet 200
Tinbergen, N. 5
four questions 5–8, 15, 58
TNT (explosive 2-4-6-trinitrotoluene) 66
toms see cats; turkeys
training
imprint 179
maturation 179
pigs 67, 68, 73
programmes 180
reward systems 98
traits
adaptive 11
anatomical 19, 20
fitness-related 115
genetic 14
production 128, 153, 215
tree shrews (Tupaia belangeri) 125
Tribe Caprini 199
Turkey 214, 241
turkeys, male (toms) 155, 155
udder-seeking behaviour 208, 209

ulceration 123, 124, 124
unconditioned response (UR) 63–64
unconditioned stimulus (US) 63–66

undernutrition 206, 211, 223
ungulates see cattle; pigs; sheep and goats
United Kingdom (UK) 172, 182, 202, 242, 263
United States of America (USA) 4, 199, 215, 242, 255–258, 272
urine 233, 275
urine-marking 277
variation principle 18
vasoactive intestinal polypeptide (VIP) 36
vibrissae 182
vigilance behaviour 182
vision 232–233
visual signals 80
vixens see foxes
vocalization 146, 164, 181, 196–198, 218, 235–237, 249, 277
barking 93
distress-induced 218
ultrasonic 278–280
vomeronasal organ (VNO) 243–244
von Frisch, K. 5
von Haartman, L. 86–87
Wales 241
wallowing 220, 221
water, conserving 204
water-bathing 160
Watson, J.B. 4, 62
weaning 142, 179–180, 197–198, 225–226, 250, 263, 281
abrupt 198
forced 208
natural 208
post 225
two-step 198
weather effect 97, 195, 204, 205, 273
see also climate
Weiss, J.M. 124–126
welfare 2, 13, 39, 52–55, 75–76, 119–134
animal-based measures 128
assessment 6–7
behaviour 129
health indicators 128
management-based measures 128
physiological indicators 128–129
poor 56–57
production indicators 128
resource-based measures 128
well-being 38, 170–171
West 200
whelping 236
whiskers 182
Wiepkema, P. 45
wild animals 5, 119
birds 158
cattle 189
mink 265
rats 275, 281
sheep and goats 199–201, 205–206
wild boar (Sus scrofa) 214–215, 218–223
wildcat (Felis silvestris lybica) 241–243
wolves 234–235
working animals 135
zebu cattle (Bos indicus) 189

zoo-kept wild boar 218

The Ethology of Domestic Animals An Introductory T...

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Ethology of Domestic Animals An Introductory T...

Uploaded by

Copyright:

Available Formats

Copyright © 2017. CAB International. All rights reserved.

The Ethology of Domestic Animals, 3rd Edition

Tel: +44 (0)1491 832111 Tel: +1 (617)682-9015

Library of Congress Cataloging-in-Publication Data

Names: Jensen, Per, editor.

Commissioning editor: Caroline Makepeace

Typeset by SPi, Pondicherry, India

2.11 The Function of Behaviour

6.4 Natural Selection

11.2 Social Behaviour

18.3 Social Behaviour

Leena Ahola, Kuopio, Finland. Email: l.k.ahola@dnainternet.net

Basic Elements of Animal Behaviour

1.2 The History of Animal Behaviour Studies

1.3 The Schools of the 20th Century

1.4 Modern Approaches to Ethology

1.5 Applied Ethology

work better for the animals (Broom et al., 1995).

A fine example of successful contributions from ethology to animal conservation is the

1.6 The Field of Applied Ethology

2.2 To What Degree is Behaviour Genetically Inherited?

2.3 Genetic Versus Environmental Influence on Behaviour

determinants of behaviour, we should consider genetic traits to be predispositions, which bias

2.4 Single-gene Influences

2.5 How Do Genes Affect Behaviour?

are regulated by genes.

2.6 Localizing the Genes

Genetic association analyses have identified chromosomal regions influencing such

2.8 Evolution of Behaviour

2.9 Tracing the Evolutionary History of Behaviour

2.10 Modification and Ritualization

2.11 The Function of Behaviour

2.12 Domestication – An Evolutionary Case History

2.13 Which Animals Were Domesticated?

mates would make selective breeding much easier.

2.14 Behavioural Effects of Domestication

3.2 Central Regulation of Behaviour

RH, releasing hormone; RIH, release-inhibiting hormone.

The release of the remaining four adenohypohyseal hormones, adrenocorticotropin

Emotion and consciousness

3.4 Regulation of Food Intake and Eating

3.5 Reproductive and Maternal Behaviour

3.6 Stress-related Behaviour

4.1 What Are Motivational States and Why Are They

motivations to interact with peers or environmental enrichments. Here, the underlying

4.2 Motivational States and the Organization of Behaviour

The organization of behaviour

Behavioural sequencing II: ‘feed-forward’ processes

4.3 Conceptualizing Motivational Systems and Investigating

Motivational states as ‘intervening variables’

The psychohydraulic model

A real example of the value of black box models: modelling dust-bathing

lower stopping threshold that varies in parallel to the starting threshold.

4.4 Motivation and Emotion

4.5 Motivation and Applied Ethology

Anhedonia and poor animal welfare

4.6 Conclusions and Links to Other Chapters

Toates, F.M. (1986) Motivational Systems. Cambridge University Press, Cambridge.

influences our attitudes towards and understanding of animal welfare.

5.2 Associative Learning: Pavlovian and Instrumental

5.3 Associative Learning: Reinforcement, Punishment,

5.4 Associative Learning: What is Learned?

the food is foraging with them (Held et al., 2010).