Integrating animal temperament within ecology and evolution.

TL;DR: It is proposed that temperament can and should be studied within an evolutionary ecology framework and provided a terminology that could be used as a working tool for ecological studies of temperament, which includes five major temperament trait categories: shyness‐boldness, exploration‐avoidance, activity, sociability and aggressiveness.

Abstract: Temperament describes the idea that individual behavioural differences are repeatable over time and across situations. This common phenomenon covers numerous traits, such as aggressiveness, avoidance of novelty, willingness to take risks, exploration, and sociality. The study of temperament is central to animal psychology, behavioural genetics, pharmacology, and animal husbandry, but relatively few studies have examined the ecology and evolution of temperament traits. This situation is surprising, given that temperament is likely to exert an important influence on many aspects of animal ecology and evolution, and that individual variation in temperament appears to be pervasive amongst animal species. Possible explanations for this neglect of temperament include a perceived irrelevance, an insufficient understanding of the link between temperament traits and fitness, and a lack of coherence in terminology with similar traits often given different names, or different traits given the same name. We propose that temperament can and should be studied within an evolutionary ecology framework and provide a terminology that could be used as a working tool for ecological studies of temperament. Our terminology includes five major temperament trait categories: shyness-boldness, exploration-avoidance, activity, sociability and aggressiveness. This terminology does not make inferences regarding underlying dispositions or psychological processes, which may have restrained ecologists and evolutionary biologists from working on these traits. We present extensive literature reviews that demonstrate that temperament traits are heritable, and linked to fitness and to several other traits of importance to ecology and evolution. Furthermore, we describe ecologically relevant measurement methods and point to several ecological and evolutionary topics that would benefit from considering temperament, such as phenotypic plasticity, conservation biology, population sampling, and invasion biology.

Summary

Introduction

• Temperament describes the idea that individual behavioural differences are repeatable over time and across situations.
• The authors propose that temperament can and should be studied within an evolutionary ecology framework and provide a terminology that could be used as a working tool for ecological studies of temperament.

II. WHAT IS TEMPERAMENT?

• Defining temperament traits, like traits in general, is not trivial (Wagner, 2001).
• The authors discuss the concept of a trait in an ecological and evolutionary context and extend this concept to provide a terminology for temperament traits.
• An important property of this model is that phenotypic variation of a composite trait (e.g., B) will depend on the cumulative and interactive genetic and environmental effects on the variation at an underlying level.
• Furthermore, the authors avoid confusion between the terms; here shyness-boldness is used when there is no component of novelty associated with the measured behaviour.
• This is the case for examples of parental style, dominance, leadership, foraging style, or dispersal.

III. WHY HAVE ECOLOGISTS AND EVOLUTIONARY BIOLOGISTS NEGLECTED TEMPERAMENT?

• The potential importance of animal temperament to both applied and theoretical studies of ecology and evolution has been widely recognised (e.g., Stamps, 1991; Wilson et al., 1994, Sih et al., 2004a, b), yet relevant empirical studies are generally few and far between (but see Section III.4).
• The correlation between predator-inspection behaviour, aggressiveness and exploratory behaviour in three-spined stickleback (Huntingford, 1976b) may not reflect inter-trait correlations but rather individual consistency in fear towards an unfamiliar aquarium (Maier, Vandenhoff & Crowne, 1988; Budaev, 1997).
• Though seemingly similar to the experimenter, however, successive replications of a test for the same individual may differ due to micro-environmental effects, thereby changing the expression of the behaviour (Henderson, 1990).
• An evolutionary and ecological approach to the study of temperament will thus include estimates of phenotypic and genetic correlations between the traits (Sih et al., 2004b).
• The common-garden approach can also be extended to study evolutionary constraints or trade-offs; one could do so by looking at the correlation between two phenotypic traits measured in different populations (Roff, Crnokrak & Fairbairn, 2003).

V. CONCLUSIONS

• (1) The authors aim was to build a general framework for the ecological and evolutionary study of temperament and to review the evidence that temperament traits are heritable, linked to fitness, and correlate with several other important traits.
• (2) Temperament appears to affect the various ways an individual interacts with its environment, whether in its reactions with predators, food sources, and habitat, or in its social or sexual interactions with conspecifics.
• (3) Temperament phenotypes will be favoured or disfavoured by selection depending on the particular ecological conditions experienced by the population.
• Furthermore, individual differences in temperament may affect variation in habitat use or assortative mating, and thus will create conditions for non-random mating.

Figures

Table 3. Summary of a terminological and methodological framework for the ecological study of temperament

Table 2. The four-step program on the ecological role of temperament

Fig. 1. Conceptual model of the organisation of behavioural traits. F1 and F2 represent two different biological functions. The second level is composed of behavioural traits (Bs) involved in each function. The next level is composed of neurophysiological (Ns; e.g., hormones, neurotransmitters, neuromodulators) and structural (e.g., neuronal structures, muscle characteristics) traits involved in each behavioural trait. The final level is composed of genes (Gs) that are involved in each neuro-physiological trait. For simplicity we do not show environmental effects, interactions among traits within each level, and feedback effects. We also do not include the developmental dimension of behavioural trait construction (see text). The traits B1 and N1 share exactly the same genes, and the genetic correlation estimated between these traits is 1 (Falconer & MacKay, 1996; Roff, 1997; Lynch & Walsh, 1998). Because of environmental effects, the phenotypic correlation should be lower, but still strong and significant. In that case, we could say that these two traits are part of the same behavioural syndrome (Sih et al., 2004b) and are two facets of a coping style (Koolhaas et al., 1999). Similarly, trait N1 and N2 show a genetic correlation of 1. B1 and B3 will be genetically correlated, although to a lesser extent, because some genes influencing B3 (e.g., G5 through N3) do not influence B1. Finally, B1 and B6 belong to two different genes-neurophysiology-behaviour pathways and will therefore show a null genetic correlation. This model could be extended to one function studied in two different environments within the same context (e.g., antipredator defence in F1 ¼ high-risk and F2 ¼ low-risk environment), at different ages (e.g., anti-predator defence in F1 ¼ juvenile and F2 ¼ adult), in the two sexes (e.g., mating behaviour in F1 ¼ males and F2 ¼ females). For instance, some genes influencing a behaviour expressed in a high-risk environment may not influence the behaviour expressed in a low-risk environment. This means that two measures of the same phenotypic trait (i.e., antipredator defence) can be considered as two genetically different traits (Falconer & MacKay 1996).

Fig. 3. A reaction norm approach to the study of temperament that aims to disentangle exploration from boldness, when experience or habituation can affect the individual response to a test. Each line represents the behavioural response of a given genotype, or individuals experiencing the experimental arrangement three consecutive times (three points). With this approach we assume that the animals experience a decline in novelty. In A, genotypes vary in their response to the experimental arrangement, but do not perceive the change in novelty (i.e., the reaction norm is flat and phenotypic plasticity is null); in this case exploration and boldness are identical. In B, behavioural responses decline with time, indicating phenotypic plasticity, but all the genotypes react in the same way (i.e., null GxE interaction). In this case exploration and boldness are different overall; they can be considered as genetically identical (genetic correlation between exploration and boldness equals 1). In C, genotypes differ in how fast they change their response to the experimental arrangement with experience (i.e., GxE occurs). In this case, exploration and boldness are two different traits.

Table 1. A non-exhaustive list of definitions of temperament, personality and coping style

Fig. 2. (A). Flow diagram illustrating the proposed framework for the ecology of temperament traits. Note that the arrows do not represent all possible links between variables. Arrows between tests and temperament traits indicate possible direct measurements. Note that some traits measured with a specific test may be affected by other traits. (B). An example that considers the great tit, Parus major. Single arrows indicate direct relationships between two variables; double arrows indicate correlations between variables. Dotted lines indicate that evidence is lacking for a mediating effect of any of the component traits. * Fitness consequences of temperament trait variation differ with year and ecological conditions. † Given the strong correlation between ‘boldness’ and ‘exploration’, these traits should be regarded here as the same temperament trait. Note that we report the links made by the authors. Note that trait names used in the figure follow the terminology of the authors of the studies, rather than the terminology developed in the present paper. Numbers in figure: (1) Verbeek et al. (1994); (2) Dingemanse et al. (2002); (3) Verbeek et al. (1996); (4) van Oers et al., (2004b); (5) van Oers et al. (2004a); (6) Marchetti & Drent (2000); (7) Verbeek et al. (1999); (8) Dingemanse & de Goede (2004); (9) Dingemanse et al. (2003); (10) Both et al. (2005); (11) Dingemanse et al. (2004).

Table 5. Evidence for genetic effects on the variation in temperament traits in animals. Studies are separated into three categories: (1) repeatability, (2) heritability, and (3) genetic difference between populations. Method: R ¼ repeatability; BD ¼ heritability/ breeding design; AS ¼ artificial selection; POP ¼ population comparisons (common-garden experiment); IL ¼ strain or inbred-line comparison. ‘‘]’’ sign indicates for the presence of significant within-individual consistency, among-line or among-population variation without providing estimates

Table 4. Examples of phenotypic and genetic correlations between temperament traits and other traits in animals. AS ¼ genetic/artificial selection; BD ¼ genetic/breeding design; PC ¼ phenotypic correlation; LC ¼ genetic line comparison (i.e., differences in temperament traits between lines artificially selected for the other trait). Quad] ¼ positive quadratic; PSA ¼ phenotypic selection analysis. ‘‘]’’ and ‘‘-’’ indicate positive and negative correlations between the two traits, respectively

Full text

Biol. Rev. (2007), 82, pp. 291–318. 291
doi:10.1111/j.1469-185X.2007.00010.x
Integrating animal temperament within
ecology and evolution
Denis Reale
1
*, Simon M. Reader
2,3
, Daniel Sol
3,4
, Peter T. McDougall
3
and
Niels J. Dingemanse
5
1
Canada Research Chair in Behavioural Ecology and Groupe de Recherche en Eco logie Comportementale et Animale, Departement des Sciences
Biologiques, UniversiteduQuebec a
`
Montreal, CP-8888, Succursale Centre-ville, Montreal, Quebec, H 3C 3P8, Canada
2
University of Utrecht , Behavioural Biology and Helmholtz Institute, Padualaan 8, PO Box 80086, 3508 TB Utrecht, The Netherlands
3
Department of Biology, McGill University, 1205 Dr. Penfield Ave., Montreal, Quebec, H3A 1B1, Canada
4
CREAF, Center for Ecological Research and Applied Forestries, Universitat Auto
`
noma de Barcelona, E-08193 Bellaterra, Catalonia, Spain
5
Animal Ecology Group, Centre for Evolutionary and Ecologicial Studies, and Department of Behavioural Biology, Centre for Behaviour and
Neurosciences, University of Groningen, PO Box 14, 9750 AA Haren, The Netherlands
(Received 5 October 2005; revised 26 January 2007; accepted 30 January 2007)
ABSTRACT
Temperament describes the idea that individual behavioural differences are repeatable over time and across
situations. This common phenomenon covers numerous traits, such as aggressiveness, avoidance of novelty,
willingness to take risks, exploration, and sociality. The study of temperament is central to animal psychology,
behavioural genetics, pharmacology, and animal husbandry, but relatively few studies have examined the ecology
and evolution of temperament traits. This situation is surprising, given that temperament is likely to exert an
important influence on many aspects of animal ecology and evolution, and that individual variation in
temperament appears to be pervasive amongst animal species. Possible explanations for this neglect of tem-
perament include a perceived irrelevance, an insufficient understanding of the link between temperament traits
and fitness, and a lack of coherence in terminology with similar traits often given different names, or different
traits given the same name. We propose that temperament can and should be studied within an evolutionary
ecology framework and provide a terminology that could be used as a working tool for ecological studies of
temperament. Our terminology includes five major temperament trait categories: shyness-boldness, exploration-
avoidance, activity, sociability and aggressiveness. This terminology does not make inferences regarding
underlying dispositions or psychological processes, which may have restrained ecologists and evolutionary
biologists from working on these traits. We present extensive literature reviews that demonstrate that tem-
perament traits are heritable, and linked to fitness and to several other traits of importance to ecology and
evolution. Furthermore, we describe ecologically relevant measurement methods and point to several ecological
and evolutionary topics that would benefit from considering temperament, such as phenotypic plasticity,
conservation biology, population sampling, and invasion biology.
Key words: temperament, personality, individual differences, behavioural syndromes, coping styles, context-
specificity, shyness-boldness, exploration, aggressiveness, sociability.
CONTENTS
I. Introduction ...................................................................................................................................... 292
II. What is temperament? ...................................................................................................................... 292
(1) Behavioural trait concepts in an ecological and evolutionary context ..................................... 293
(2) Terminology for ecologists .......................................................................................................... 293
* Address for correspondence: E-mail: reale.denis@uqam.ca
Biological Reviews 82 (2007) 291–318 Ó 2007 The Authors Journal compilation Ó 2007 Cambridge Philosophical Society

(3) Properties of temperament traits ................................................................................................ 296
III. Why have ecologists and evolutionary biologists neglected temperament? .................................... 296
(1) Inconsistent terminology, perceived irrelevance and lack of generalization ............................. 296
(2) Reluctance to take an individual-based approach ..................................................................... 297
(3) The lack of a general framework ............................................................................................... 297
(4) Evidence for the ecological importance of temperament ......................................................... 297
IV. Integrating temperament within ecological and evolutionary studies ............................................. 298
(1) Measuring temperament traits ................................................................................................... 298
(a ) Variation ................................................................................................................................ 299
(b ) Repeatability (r) ..................................................................................................................... 302
(c ) Heritability ............................................................................................................................ 303
(2) Validating measures of temperaments traits .............................................................................. 304
(a ) Biological and ecological validity ......................................................................................... 304
(b ) Behavioural syndromes ......................................................................................................... 304
(3) Linking temperament with fitness .............................................................................................. 305
(a ) Measuring selection ............................................................................................................... 308
(b ) Adaptive explanations for the maintenance of variance in temperament traits ................. 308
(4) Temperament traits as adaptation: comparative studies ........................................................... 309
(a ) Intra-species population comparison .................................................................................... 309
(b ) Interspecific comparative studies .......................................................................................... 311
V. Conclusions ....................................................................................................................................... 312
VI. Acknowledgments ............................................................................................................................. 312
VII. References ......................................................................................................................................... 312
I. INTRODUCTION
How and for what reasons individuals - animals or humans -
differ in the way they react to potential risks, handle novelty,
or interact with conspecifics, are questions that fascinate
scientists and the general public alike. The study of
temperament (or personality) differences has proven utility
at many levels of society: from improvements in animal
production, welfare, and conservation, to pharmacology,
and, in humans, to prediction of disease risk, job
satisfaction, risk-taking behaviour, and reaction to social
stress (Boissy & Bouissou, 1995; Le Neindre, Boivin &
Boissy, 1996; Caspi et al., 1997; Grandin, 1998; Martin,
1998; Carlstead, Mellen & Kleiman, 1999; Trut, 1999;
Malmkvist & Hansen, 2001; Boissy et al., 2005; Boyce & Ellis,
2005; Ellis, Jackson & Boyce, 2006; McDougall et al., 2006).
Psychologists have long been interested in the study of
human and animal temperament, which has led to
significant theoretical and empirical developments (Wilson,
1994; Gosling, 2001; Boyce & Ellis, 2005). By contrast,
ecologists and evolutionary biologists have generally shown
little interest in the concept of temperament, although
certain temperament traits perceived to affect fitness dir-
ectly, such as aggressiveness and reaction to predators, have
received substantial scientific attention (Clark & Ehlinger,
1987; Huntingford, Wright & Tierney, 1994; Wilson et al.,
1994; Boissy, 1995; Clarke & Boinski, 1995; Greenberg &
Mettke-Hofmann, 2001; Sih, Bell & Johnson, 2004a;Sihet al.,
2004b; Ellis et al., 2006). In recent years temperament has
begun to receive theoretical and empirical attention from
ecologists (Wilson et al., 1994; Clarke & Boinski, 1995;
Greenberg & Mettke-Hofmann, 2001; Sih et al., 2004a, b;
Dall, Houston & McNamara, 2004). However, ecologists
generally do not perceive temperament as an important
addition to our understanding of the ecology and evolution
of animals. This is surprising given a growing body of
evidence showing that temperament traits affect important
ecological processes such as niche expansion, dispersal or
social organisation.
The reasons why temperament has not yet been inte-
grated into ecological theory are diverse, and include dif-
ficulties in definition, in finding appropriate methods to
quantify temperament, and in testing the significance of
these traits in the field. Our goal here is to build a supporting
conceptual and methodological framework for the ecolog-
ical study of temperament that may help overcome these
difficulties. Moreover, we review extensive evidence for the
genetic basis of temperament traits and links between
temperament and traits of importance to evolutionary
ecology, such as reproductive rate and survival. Obviously,
many important questions regarding temperament remain
to be resolved before developing a general framework.
Nevertheless, we hope that the proposed framework
will serve to guide and encourage future research in the
field.
II. WHAT IS TEMPERAMENT?
A necessary first step is to define temperament traits in
terms relevant to evolutionary ecology. Defining tempera-
ment traits, like traits in general, is not trivial (Wagner,
2001). We discuss the concept of a trait in an ecological and
evolutionary context and extend this concept to provide
a terminology for temperament traits.
Denis Re
´
ale and others292
Biological Reviews 82 (2007) 291–318 Ó 2007 The Authors Journal compilation Ó 2007 Cambridge Philosophical Society

(1) Behavioural trait concepts in an ecological
and evolutionary cont ext
A character (or trait) can be considered as a characteristic of
an organism shared by all or some of the individuals of
a species that can vary, although not necessarily, among
these individuals (we consider character and trait as
synonyms; see Wagner, 2001, on the diversity of the
character concept). Measured individual values for that
character are called phenotypes. A quantitative genetic
framework can provide a biological definition of a trait.
Quantitative genetic models, which have received some
recent support, assume that the variance of phenotypic
quantitative traits (i.e., the trait measured) is influenced by
a relatively large number of genes, each with small
individual effects, and by a series of environmental effects
(Falconer & Mackay, 1996; Lynch & Walsh, 1998; Roff,
1997; Flint, 2003; Reif & Lesch, 2003). An important aspect
of our definition is that two traits can be associated at
the phenotypic level, illustrating their potential genetic
or epigenetic links (Henderson, 1990; Wagner, 1996;
Cheverud, 1996; Sih et al., 2004b).
Fig. 1 shows a conceptual model of the organisation of
behavioural traits derived from Henderson (1990) and
Wagner (1996). The goal of this model is to illustrate the
hierarchical structure of traits within an organism rather
than to describe the precise physiological and develop-
mental machinery underlying the expression of those
traits. Hence it does not include developmental feedback
effects occurring at different levels or environmental
effects. F1 and F2 represent two different biological func-
tions of an animal species, such as anti-predator defence
and mating. Behavioural traits (Bs; note that they could be
other trait categories like life-history or morphological
traits) involved in each function are not directly influenced
by genes, but result from a complex network of neuro-
physiological and structural traits themselves a result of the
indirect effects of genes (see Johnston & Edwards, 2002).
Phenotypic variation of a trait B results from both among-
individual allelic variability at loci with additive and non-
additive (e.g., dominance, epistasis) genetic effects, and
from environmental variation (Falconer & MacKay, 1996;
Roff, 1997; Lynch & Walsh, 1998). An important property
of this model is that phenotypic variation of a composite
trait (e.g., B) will depend on the cumulative and interactive
genetic and environmental effects on the variation at an
underlying level. Note that to simplify we limited the
number of levels to a minimum; a behavioural trait not
shown here could integrate several traits at underlying level
B. For example, maternal behaviour in a mammal is
composed of many other behavioural traits (e.g., nursing,
nest building, grooming, vigilance and defence against
predators). Therefore, the measure of a trait at one level, to
some extent, also represents other traits that are genetically,
developmentally and functionally related to the studied
trait. For instance, B1 and N1 in Fig. 1 can be considered
genetically the same trait since the same genes influence
them, regardless of the fact they are phenotypically
different: one trait is a behaviour pattern, the other a
hormone concentration.
(2) Terminology for ecologists
Table 1 lists several published definitions of temperament
and personality. The distinction between temperament and
personality is often vague, but temperament frequently has
F1
F2
Genetic
Neuroendocrine
and structural
Behaviour
Function
B1 B2 B3 B4 B5 B6
N1 N2 N3 N4 N5 N6
G1 G2 G3 G4 G5 G6
Fig. 1. Conceptual model of the organisation of behavioural
traits. F1 and F2 represent two different biological functions.
The second level is composed of behavioural traits (Bs) involved
in each function. The next level is composed of neuro-
physiological (Ns; e.g., hormones, neurotransmitters, neuro-
modulators) and structural (e.g., neuronal structures, muscle
characteristics) traits involved in each behavioural trait. The
final level is composed of genes (Gs) that are involved in each
neuro-physiological trait. For simplicity we do not show en-
vironmental effects, interactions among traits within each level,
and feedback effects. We also do not include the developmental
dimension of behavioural trait construction (see text). The traits
B1 and N1 share exactly the same genes, and the genetic
correlation estimated between these traits is 1 (Falconer &
MacKay, 1996; Roff, 1997; Lynch & Walsh, 1998). Because of
environmental effects, the phenotypic correlation should be
lower, but still strong and significant. In that case, we could say
that these two traits are part of the same behavioural syndrome
(Sih et al., 2004b) and are two facets of a coping style (Koolhaas
et al., 1999). Similarly, trait N1 and N2 show a genetic
correlation of 1. B1 and B3 will be genetically correlated,
although to a lesser extent, because some genes influencing
B3 (e.g., G5 through N3) do not influence B1. Finally, B1 and
B6 belong to two different genes-neurophysiology-behaviour
pathways and will therefore show a null genetic correlation.
This model could be extended to one function studied in two
different environments within the same context (e.g., anti-
predator defence in F1 ¼ high-risk and F2 ¼ low-risk
environment), at different ages (e.g., anti-predator defence in
F1 ¼ juvenile and F2 ¼ adult), in the two sexes (e.g., mating
behaviour in F1 ¼ males and F2 ¼ females). For instance, some
genes influencing a behaviour expressed in a high-risk
environment may not influence the behaviour expressed in
a low-risk environment. This means that two measures of
the same phenotypic trait (i.e., antipredator defence) can be
considered as two genetically different traits (Falconer &
MacKay 1996).
Animal Temperament 293
Biological Reviews 82 (2007) 291–318 Ó 2007 The Authors Journal compilation Ó 2007 Cambridge Philosophical Society

a more restrictive meaning than personality, often describ-
ing differences in emotionality or describing traits that are
demonstrated very early in life (Budaev, 1997; Box, 1999).
Given that personality and temperament are frequently
distinguished on arbitrary grounds we treat them here as
synonyms.
Many definitions in Table 1 refer to both a measurable
element (i.e., the expression) and ‘‘unobservables’’, or
qualities that are difficult to measure (i.e., the individual
dispositions). Similarly, several temperament traits imply
inference of psychological mechanisms underlying the
expression at the behavioural level. For instance emotion-
ality refers to behavioural and peripheral changes presumed to
accompany high sympathetic nervous activity (Archer, 1973), an-
xiety is the fear of a potential danger (Boissy, 1995; File, 2001),
and some definitions of neophobia focus on the fear of novel
objects (e.g., Beissinger, Donnay & Walton, 1994). We think
that both these ‘‘unobservables’’ and the inference of
psychological properties may restrain ecologists and evolu-
tionary biologists from studying temperament; such inher-
ent dispositions are not implied in the definitions of
morphology (i.e., the form of living organisms; Oxford Dic-
tionary of English, 2005) or life history (the series of changes
undergone by an organism during its lifetime; Oxford Dictionary of
English, 2005), two categories of traits commonly studied by
ecologists and evolutionary biologists. Based on these
considerations, we propose that temperament, personality
and individuality describe the phenomenon that individual
behavioural differences are consistent over time and/or
across situations (Budaev, 1997; Box, 1999; Lowe &
Bradshaw, 2001; Gosling, 2001; Dall et al., 2004). Here
‘‘consistent’’ does not mean that trait values cannot change
with age or environmental conditions, but that differences
between individuals are largely maintained. Although
temperament is considered at the individual level, it can
be extended to the other levels, describing for example
consistent behavioural differences between families, po-
pulations or species.
Table 1. A non-exhaustive list of definitions of temperament, personality and coping style
Definition Source
Temperament: a person’s or animal’s nature, especially as it permanently affects their
behaviour. Personality: the combination of characteristics or qualities that form an
individual’s distinctive character.
The Oxford English Dictionary (2005)
Temperament: relatively consistent, basic dispositions inherent in the person that
underlie and modulate the expression of activity, reactivity, emotionality, and
sociability.
Buss et al. (1987)
Temperament: the characteristic phenomena of an individual’s emotional nature,
including his susceptibility to emotional stimulation, his customary strength and
speed of response, the quality of his prevailing mood, and all peculiarities of
fluctuation and intensity of mood; these phenomena being regarded as dependent on
constitutional makeup and therefore largely hereditary in origin.
Allport (1937), p. 54
In addition to the notion that temperament reflects biologically based individual
differences in emotional responding, modern temperament theories also incorporate
Allport’s idea that these biological differences are innate and form the foundation
upon which mature personality develops.
Clark & Wilson (1999), p. 400
Personality: those characteristics of individuals that describe and account for consistent
patterns in feeling, thinking and behaving. Temperament: in human research... the
inherited, early appearing tendencies that continue throughout life and serve as
foundation to personality.
Gosling (2001), p.46
Temperament: 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. It is the stance that an individual takes towards its
environment across time and situations. It refers to styles of responsiveness and not
to specific acts.
Box (1999), p. 34
Temperaments and personalities: integrated behavioural phenotypes and stable traits
that are consistent over time and across situations; broad and consistent dimensions
of individuality.
Budaev (1997), p. 399
Individual animals often behave in a way that distinguishes them from other members
of their species of the same sex and age class. When such differences are consistent
over time, they can be referred to as ‘personalities’ or ‘behavioural style’.
Lowe & Bradshaw (2001)
Coping style: a coherent set of behavioural and physiological stress responses which is
consistent over time and which is characteristic to a certain group of individuals. It
seems that coping styles have been shaped by evolution and form general adaptive
response patterns in reaction to everyday challenges in the natural habitat.
Koolhaas et al. (1999)
The expression of individual behavioural and physiological phenotypes or ‘coping
styles’ is defined as the way individuals cope behaviourally and physiologically with
environmental and social challenges, irrespective of life history state, sex or
motivational state.
Pfeffer et al. (2002)
Denis Re
´
ale and others294
Biological Reviews 82 (2007) 291–318 Ó 2007 The Authors Journal compilation Ó 2007 Cambridge Philosophical Society

We propose, as a working rule, that each temperament
trait should be defined according to the ecological situation
(sensu Sih et al., 2004a) in which it is measured. The
underlying concept is that each term should be operation-
ally defined and its ecological validity tested (see Section V).
We divide temperament traits into five categories: (1)
shyness-boldness, an individual’s reaction to any risky
situation, but not new situations. This includes reaction to
risky situations, such as predators and humans. Note
that ‘docility’, ‘tameness’ and ‘fearfulness’ have been used
in the specific context of reaction to humans (Boissy,
1995; Grandin, 1998; Boissy et al., 2005); (2) exploration-
avoidance, an individual’s reaction to a new situation. This
includes behaviour towards a new habitat, new food, or
novel objects. This situation can also be considered risky if,
for example, a new object may represent a potential
predator. We have deliberately not included neophobia and
neophilia in our terminology because both are considered
as part of exploration. Furthermore, from an ecological
point of view, we can assume that exploration will be the
main target of selection. Neophobia and neophilia are, on
the other hand, more relevant to those interested in the
mechanisms responsible for individual variation in explo-
ration (see e.g., Greenberg & Mettke-Hofmann, 2001);
(3) activity, the general level of activity of an individual.
Activity can interfere with the measurement of exploration
or of boldness; it has thus been proposed to obtain a
measure of activity in a non-risky and a non-novel en-
vironment (Barnett & Cowan, 1976; Renner, 1990). The
last two trait categories are expressed in a social context; (4)
aggressiveness, an individual’s agonistic reaction towards
conspecifics; (5) sociability, an individual’s reaction to the
presence or absence of conspecifics (excluding aggressive
behaviour). Sociable individuals seek the presence of con-
specifics, while unsociable individuals avoid conspecifics.
We are conscious of the limits created by the simplifi-
cation of our terminology; this terminology should
be regarded as a working tool, and not an exhaustive list.
This simplified terminology is essential to encourage
ecological research into temperament. Once we have ob-
tained sufficient information for a large group of species
from different ecosystems, we can start generalizing about
the ecological and evolutionary role of temperament
traits. At that point, we could refine our terminology for
temperament traits, using more terms and operational
concepts.
This terminology has several advantages:
(1) The five proposed categories do not include the notion
of underlying dispositions. Furthermore, we avoid
confusion between the terms; here shyness-boldness is
used when there is no component of novelty associated
with the measured behaviour. When novelty is as-
sociated, it is preferable to use exploration-avoidance.
(2) Using a hierarchical model (Fig. 1), temperament
traits, like morphological or life-history traits, refer to
a category of traits. Each temperament trait can
potentially be measured by a set of correlated
behavioural or physiological variables. For example,
a bird’s body size could be measured with wing length,
tarsus length, beak size or body mass. Similarly
a rodent’s exploratory phenotype could be measured
by the distance covered and by the frequency of
rearing and sniffing in an open-field. Temperament
can thus be measured using physiological, hormonal,
and/or behavioural indices measured in a specific
ecological situation. Choosing a measure depends on
a study’s goals. Researchers interested in mechanisms
would give priority to endocrine, neurobiological and
behavioural levels. Alternatively, those more interested
in the function of temperament traits would focus on
the behavioural level and the consequences of this
behavioural variation on fitness. Both approaches are
worthwhile and should be integrated for a better
understanding of temperament traits.
(3) According to the hierarchical model, we can consider
integrated behaviour patterns (sensu Henderson, 1990)
as outcomes of the temperament traits categories
defined above. This is the case for examples of
parental style, dominance, leadership, foraging style,
or dispersal. For instance, parental style represents the
reactions of a parent towards its progeny when the
progeny is in interactions with conspecifics or novel or
risky situations; permissive mothers will allow their
progeny to interact with conspecifics, whereas restric-
tive mothers will prevent contact between their
progeny and conspecifics (Maestripieri, 1993, 1998;
Fairbanks, 1996). In this context, it will be possible to
test if one of the five categories of temperament traits
affects parental style. These integrated behaviour traits
are themselves integrated into a higher level function.
For example parental style, fecundity and maternal
investment may be considered as important compo-
nents of lifetime reproductive success. The hierarchical
model proposes to put behaviour traits into a network
of traits with different levels of interactions and
integration. It thus differs from Wilson et al.’s (1994)
shyness-boldness model that proposes a more hori-
zontal structure of behaviour traits.
(4) With this terminology we do not presuppose the
correlation of all the traits together in a whole
‘‘temperament’’, an idea common to concepts such
as personality (Costa & McCrae, 1992), coping style
(Koolhaas, De Boer & Bohus, 1997; Koolhaas et al.,
1999; Pfeffer, Fritz and Kotrschal, 2002) or behav-
ioural syndromes (Clark & Ehlinger, 1987; Sih et al.,
2004a, b). Traits may also be consistent across
functional behavioural categories or contexts (con-
text-generality, the converse of context-specificity; Sih
et al., 2004a, b; Coleman & Wilson, 1998). For
example, an individual may be bold in feeding,
mating and anti-predator contexts, in which case
boldness would appear to be a context-general
temperament trait. The extents to which temperament
traits correlate with one another and are context-
specific are empirical questions rather than defining
features of temperament (Coleman & Wilson, 1998).
Indeed, context-specificity and inter-trait correlation
need not be considered as separate questions, since
two context-specific traits can be usefully considered
as two traits. The evolutionary, functional, proximate,
Animal Temperament 295
Biological Reviews 82 (2007) 291–318 Ó 2007 The Authors Journal compilation Ó 2007 Cambridge Philosophical Society

Frequently asked questions

Q1. What have the authors contributed in "Integrating animal temperament within ecology and evolution" ?

The study of temperament is central to animal psychology, behavioural genetics, pharmacology, and animal husbandry, but relatively few studies have examined the ecology and evolution of temperament traits. The authors propose that temperament can and should be studied within an evolutionary ecology framework and provide a terminology that could be used as a working tool for ecological studies of temperament. This terminology does not make inferences regarding underlying dispositions or psychological processes, which may have restrained ecologists and evolutionary biologists from working on these traits. The authors present extensive literature reviews that demonstrate that temperament traits are heritable, and linked to fitness and to several other traits of importance to ecology and evolution. Furthermore, the authors describe ecologically relevant measurement methods and point to several ecological and evolutionary topics that would benefit from considering temperament, such as phenotypic plasticity, conservation biology, population sampling, and invasion biology. 

Q2. What future works have the authors mentioned in the paper "Integrating animal temperament within ecology and evolution" ?

The authors predict that temperament will form an important part of future research on various ecological topics. ( 4 ) Temperament may have important consequences for several ecological topics, such as: ( a ) population dynamics and genetics ( i. e., dispersal, individual movement, gene flow, and the genetic composition of meta-populations ) ; ( b ) landscape ecology ( i. e., changes in the structure of the landscape will affect the movement of individuals differently according to their temperament ) ; ( c ) community ecology ( i. e., individual variation in some sets of correlated temperament and morphological traits may be viewed as functional sub-categories in the organisation of communities ) ; ( d ) invasion biology ( i. e., could temperament be an important factor in the invasiveness syndrome ? ) ; and ( e ) speciation ( i. e., temperament variation may be responsible for the geographic and reproductive isolation of individuals characterised by particular combinations of behavioural and morphological and life-history traits ). 

Q3. What is the role of the social environment in the maintenance of variance in temperament traits?

In the case of temperament traits, both frequency- and density-dependent processes are likely to contribute to the maintenance of high levels of variation (Wilson et al., 1994; Dall et al., 2004), because the social environment is probably an important determinant of the relation between temperament and fitness. 

Q4. What are the reasons why a large number of neurophysiological systems would be costly?

The authors can advance three reasons: (1) the evolution of a large number of neurophysiological systems would be costly, thereby creating a strong constraint on the diversity of possible behavioural responses and thus a strong integration between temperament traits. 

Q5. What is the significance of the adaptive explanations for the maintenance of variation in temperament traits?

note that the adaptive explanations for the maintenance of variation in temperament traits given above only explain the co-existence of different temperament types given that individuals show consistent behaviour. 

Q6. How can the authors measure the survival and reproductive success of individuals whose phenotype has been?

This could be done by measuring the survival and reproductive success of individuals whose phenotype has been manipulated; such as using agonists or antagonists of some neurotransmitters involved in temperament variation (e.g., tryptophan and fluoxetine increase and cyproheptadine and fenfluramine decrease 5-hydroxytryptamine (5-HT) activity, respectively, affecting dominance acquisition in male vervet monkeys Cercopithecus aethiops; Raleigh et al., 1991). 

Q7. Why are ecologists and evolutionary biologists encouraged to focus on temperament under novel, risky or?

Ecologists and evolutionary biologists may also be encouraged to focus on temperament under novel, risky or challenging conditions for this reason, but also because they are determinant for the differential survival and reproduction of individuals. 

Q8. What is the problem of establishing a causal link between the variables under study?

The problem of establishing a causal link between the variables under study may in part be overcome with methods that allow estimation of ancestral states (Harvey & Pagel, 1991).