Behavioural syndromes, syndrome deviation and the within- and between-individual components of phenotypic correlations: when reality does not meet statistics

Evolutionary mechanisms leading to correlations across different behaviours, called behavioural syndromes, are hard to study, mostly because behavioural syndromes are group/population level phenomena. Recently (Herczeg and Garamszegi Behav Ecol Sociobiol 66:161–169, 2012), we introduced the concept of syndrome deviation that allows the study of behavioural syndromes at the individual level by focusing on the individual deviation from the hypothetical perfect group-level behavioural correlation. Subsequently, Dingemanse et al. (Behav Ecol Sociobiol 66:1543–1548, 2012) emphasized that behavioural syndromes refer to the between-individual component of phenotypic correlations, and only this component is relevant for syndrome deviation. They also recommended mixed models to decompose the between- and within-individual correlations. We agree that separating these components is important, but the proposed approach is impractical to apply for functionally different behaviours because (1) the assumption of constant within-individual correlations is unjustified and (2) different behaviours cannot be measured at the same time. Further, our simulations based on mixed models show that the statistical differentiation between the within- and between-individual components is inefficient when using realistic sample sizes. Until the separation of between- and within-individual correlations is resolved, we recommend alternative approaches for empirical behavioural syndrome research that consider the repeatability of the behaviours and the optimal balance between within- and between-individual sample sizes. Syndrome deviation calculated from phenotypic correlations of traits that are proven to be individual specific, or from the between-individual correlations if possible, is a meaningful metric to describe behavioural consistency and to explain its evolutionary significance.     

Personality and parasites: sex-dependent associations between avian malaria infection and multiple behavioural traits

The evolution and ecology of consistent behavioural variation, or personality, is currently the focus of much attention in natural populations. Associations between personality traits and parasite infections are increasingly being reported, but the extent to which multiple behavioural traits might be associated with parasitism at the same time is largely unknown. Here, we use a population of great tits, Parus major, to examine whether infection by avian malaria (Plasmodium and Leucocytozoon) is associated with three behavioural traits assayed under standardized conditions. All of these traits are of broad ecological significance and two of them are repeatable or heritable in our population. Here, we show weak correlations between some but not all of these behavioural traits, and sex-dependent associations between all three behavioural traits and parasite infection. Infected males showed increased problem-solving performance whereas infected females showed reduced performance; furthermore, uninfected females were four times more likely to solve problems than uninfected males. Infected females were more exploratory than uninfected females, but infection had no effect on males. Finally, infected males were more risk-averse than uninfected males but females were unaffected. Our results demonstrate the potential for complex interactions between consistent personality variation and parasite infection, though we discuss the difficulty of attributing causality in these associations. Accounting for complex parasite-behaviour associations may prove essential in understanding the evolutionary ecology of behavioural variation and the dynamics of host–parasite interactions.     

Prospecting at conspecific nests and exploration in a novel environment are associated with reproductive success in the jackdaw

Consistent behavioural differences between individuals of the same population (“personality” variation) might arise if individuals follow different life-history strategies. Thus, it would be important to determine how personality variation relates to behaviours potentially associated with life-history strategies, such as those related to the use of information about the state of the environment. Little is, however, known about how personality is associated with information use and reproductive success. We tested whether wild social jackdaws, Corvus monedula, show consistent behavioural differences in their exploratory behaviour (in a novel environment in the lab and in their reaction towards a novel object in the wild) and prospecting behaviour (number of visits to conspecific nests). We furthermore examined whether these behavioural traits are linked with each other and predictors of reproductive success. Breeding jackdaws were consistent in their exploratory behaviour within, and in their prospecting behaviour between, years. Exploratory behaviour in the novel environment was correlated with the latency to approach a novel object in the wild but not with the frequency of prospecting at conspecific nests. Highly exploratory males and females and frequently prospecting males produced fewer fledglings than less exploratory individuals or less prospecting males, respectively. We discuss the importance of consistent individual differences in exploration and information sampling on individual fitness.     

Individual deviation from behavioural correlations: a simple approach to study the evolution of behavioural syndromes

The study of correlations between different behaviours in a population—referred to as behavioural syndromes—has begun to flourish during recent years. However, the evolutionary mechanisms that cause behavioural traits to vary non-independently from each other are still poorly understood. Here, we bring behavioural syndromes into a new perspective, in which the phenomenon is regarded at the individual level and on a continuous scale instead of as a population-level presence/absence trait. As the correlation between behaviours is never perfect (i.e. r < 1), individuals are likely to vary in how consistently they behave. Therefore, we can predict that if behavioural syndromes at the population level are results of natural selection, the consistency in a suite of behaviours—and not the behavioural configuration per se—should be heritable and involve fitness advantages at the individual level. We define a variable that describes the individual deviation from the hypothetical perfect correlation predicted by the syndrome. The use of such a variable depicting the consistency of behaviours of individuals allows us to make solid evolutionary inferences about correlated behaviours from patterns of individual instead of population variation. We suggest that, by adopting the concept of syndrome deviation, understanding the evolution of behavioural syndromes and, in particular, testing competing evolutionary hypotheses about the origin of behavioural syndromes becomes possible in a more rigorous manner than before.     

How stable are personalities? A multivariate view of behavioural variation over long and short timescales in the sheepshead swordtail,Xiphophorus birchmanni

Many studies have revealed repeatable (among-individual) variance in behavioural traits consistent with variation in animal personality; however, these studies are often conducted using data collected over single sampling periods, most commonly with short time intervals between observations. Consequently, it is not clear whether population-level patterns of behavioural variation are stable across longer timescales and/or multiple sampling periods or whether individuals maintain consistent ranking of behaviours (and/or personality) over their lifetimes. Here, we address these questions in a captive-bred population of a tropical freshwater poeciliid fish, Xiphophorus birchmanni. Using a multivariate approach, we estimate the among-individual variance-covariance matrix (I), for a set of behavioural traits repeatedly assayed in two different experimental contexts (open-field trials, emergence and exploration trials) over long-term (56 days between observations) and short-term (4-day observation interval) time periods. In both long- and short-term data sets, we find that traits are repeatable and the correlation structure of I is consistent with a latent axis of variation in boldness. While there are some qualitative differences in the way individual traits contribute to boldness and a tendency towards higher repeatabilities in the short-term study, overall, we find that population-level patterns of among-individual behavioural (co)variance to be broadly similar over both time frames. At the individual level, we find evidence that short-term studies can be informative for an individual’s behavioural phenotype over longer (e.g. lifetime) periods. However, statistical support is somewhat mixed and, at least for some observed behaviours, relative rankings of individual performance change significantly between data sets.     

The link between social network density and rank-order consistency of aggressiveness in juvenile eels

The emergence of an animal’s personality is the result of interactions between genetics, environment and experience. It is known that individuals are able to modulate their behaviour according to the context or the social environment. Many studies have shown for example, that familiarity among conspecifics diminishes aggressiveness, although little is known about the underlying processes. Nevertheless, personality traits have long been determined while ignoring the social context, especially in lower vertebrates such as fish. In the present experiment, we hypothesize that group connections (network density) may be positively correlated to consistency of aggressiveness by avoiding over-aggressive acts in further encounters. To test this hypothesis, we used eels (Anguilla anguilla) as a model species and monitored both aggressiveness and sociability in 64 individuals over their first 7 months of growth from the glass eel stage. As expected, social fish were less aggressive than their non-social counterparts at all times, highlighting the existence of a behavioural syndrome in eels. Additionally, rank-order consistency of aggressiveness was higher in groups of fish with high social connectivity, compared to those in less-connected fish groups. While aggressiveness must be consistent to be considered a personality trait, our results suggest that both aggressiveness and its consistency are influenced by initial social context.     

Variation in plasticity of personality traits implies that the ranking of personality measures changes between environmental contexts: calculating the cross-environmental correlation

Studies increasingly explore whether there is variation between individuals in how they adjust their behavior to different environmental contexts using random regression analysis (RR). RR estimates the between-individual variance in elevation (expected behavior in the mean environment) and variance in plasticity (individual-specific adjustment of behavior to the environment) and the covariance between these properties. These (co)variances allow deriving both environment-specific repeatability estimates and the individual-level correlation in behavior between environmental conditions, both of which are of key interest for understanding consistency in behavior over space and time, a core issue in animal personality research. However, very few studies use the RR estimates to produce this information. In this paper, I outline how to extract this information (including example code in R) and apply the approach to 16 estimates based on nine published RR studies. Despite the fact that the behaviors have a decent repeatability under different environmental conditions, I find that in half the cases, the cross-environmental correlation is low (<0.4). This implies that individuals differ in their behavioral trait values but that the ranking of these individual values tends to change between environmental conditions. This level of inconsistency in animal personality traits appears not to have been fully appreciated, and future studies on plasticity using RR should explicitly include it.     

Consistent crustaceans: the identification of stable behavioural syndromes in hermit crabs

Consistent individual variation in behaviour has been termed ‘animal personality’ and has been identified across a range of behavioural contexts including aggression, boldness in response to a threatening stimulus and exploration. When behaviours are correlated across multiple functional behavioural categories or ‘contexts’, ‘behavioural syndromes’ are said to be present. It is possible, however, that behavioural syndromes may also show consistencies. Here we investigated the presence of behavioural syndromes linking startle responses, exploration and aggression in hermit crabs and assessed their stability across two situations (low versus high predation risk). Correlation analyses detected behavioural syndromes between startle responses, a measure of ‘boldness’, and the latency to investigate a novel object, as well as the latency to attack an opponent in an aggressive context. The startle response–investigation and startle response–aggression syndromes were stable between situations, whilst there was a lack of relationship between investigation and aggression in each situation. Here we propose that these consistent individual differences in the expression of behavioural syndromes reveal the presence of animal personality, manifesting in not just one, but a suite of interacting traits.     

Latitude or biogeographic breaks? Determinants of phenotypic (co)variation in fitness-related traits in Betaeus truncatus along the Chilean coast

Ectothermal organisms distributed along environmental gradients in a wide geographical distribution display extensive phenotypic variation. This is particularly pervasive along latitudinal clines, which are linked to gradual changes in environmental factors. Widespread species may also be distributed among biogeographic breaks, which in contrast to smooth clines, often show abrupt changes in phenotypic traits. In species with widespread latitudinal distribution that also encompass important biogeographical breaks, it is not clear which of those factors prevails on shaping the phenotypic variation or if some traits are particularly more sensitive to one or the other. To evaluate this, we measured 4 fitness-related traits in 6 populations of the intertidal snapping shrimp Betaeus truncatus, as its distribution along Chile expands over 40° in latitude and three major biogeographical provinces. Here, we statistically evaluated the role of both, latitude and biogeographic breaks, on mean population values of fitness-related traits but also on the variances and covariances (i.e., P-matrix) between them. Overall, our results (1) indicate that latitude is more important than breaks in shaping the phenotypic variation of most of these fitness-related traits, (2) show that the differences in the variance–covariance relationship among traits between the extremes of the gradient arises from gradual increases in variance and rather sharp changes in covariance at mid-latitudes and (3) show that at present, it is difficult to unambiguously determine whether natural selection or plasticity is responsible for the observed pattern in means, variances and covariances and only further work might disentangle these possibilities.     

Heterospecific song matching in two closely related songbirds (Parus major and P. caeruleus): great tits match blue tits but not vice versa

To date, song research has focused primarily on the interactions of conspecifics. However, frequent interactions of songbirds with heterospecifics may necessitate adequate communication outside the species boundary. In this study, we focus on heterospecific communication behaviour of two small sympatric congeneric passerines, great and blue tits (Parus major and Parus caeruleus), which breed in overlapping territories and compete for food and nesting cavities. By means of a first playback experiment, we show that (1) heterospecific matching (imitating songs of the other species) is a strategy frequently used by great tits but not by blue tits, (2) both blue tit trilled and untrilled song can be accurately matched by great tits and that (3) almost half of the great tits in our study population match at least one blue tit song across all studied breeding stages, indicating that this heterospecific matching behaviour is a common feature in this population. A second playback experiment showed that these great tit imitations of blue tit songs do not function in intraspecific communication between male great tits. Hence, these heterospecific imitations appear to be designed for interspecific communication with blue tits. These findings suggest a strong heterospecific influence on the vocal learning process, repertoire composition and repertoire use in great tits and provide a possible mechanism that can drive song convergence in songbirds.     

Size-based predation by kookaburras (Dacelo novaeguineae) on lizards (Eulamprus tympanum: Scincidae): what determines prey vulnerability?

Lizards and birds are both popular ”model organisms” in behavioural ecology, but the interactions between them have attracted little study. Given the putative importance of birds as predators of diurnal lizards, it is of considerable interest to know which traits (of lizards as well as birds) influence the outcome of a predatory attempt. We studied predation by giant terrestrial kingfishers (kookaburras, Dacelo novaeguineae: Alcedinidae) on heliothermic diurnal lizards (highland water skinks, Eulamprus tympanum: Scincidae), with particular reference to the role of prey (lizard) size. Our approach was twofold: to gather direct evidence (sizes of lizards consumed in the field, compared to those available) and indirect evidence (size-related shifts in lizard behaviour). We quantified the size structure of a natural population of skinks (determined by an extensive mark-recapture program), and compared it to the sizes of wild lizards taken by kookaburras (determined by analysis of prey remains left at the birds’ nests). Kookaburras showed size-based predation: they preyed mainly on small and medium-sized rather than large lizards in the field. However, the mechanism producing this bias remains elusive. It is not due to any distinctive behavioural attributes (locomotor ability, activity level, habitat usage) of the lizards of the size class disproportionately taken by the kookaburras. The greater vulnerability of subadult lizards may reflect subtle ontogenetic shifts in ecological and behavioural traits, but our data suggest that great caution is needed in inferring patterns of vulnerability to predation from indirect measures based on either the prey or the predator alone. Instead, we need direct observations on the interaction between the two. Received: 30 May 2000 / Revised: 29 July 2000 / Accepted: 26 August 2000     

Network position: a key component in the characterization of social personality types

In recent years, animal social interactions have received much attention in terms of personality research (e.g. aggressive or cooperative interactions). However, other components of social behaviour such as those describing the intensity, frequency, directedness and individual repeatability of interactions in animal groups have largely been neglected. Network analysis offers a valuable opportunity to characterize individual consistency of traits in labile social groups and therein provide novel insights to personality research in ways previously not possible using traditional techniques. Should individual network positions be consistently different between individuals under changing conditions, they might reflect expressions of an individual's personality. Here, we discuss a conceptual framework for using network analyses to infer the presence of individual differences and present a statistical test based on randomization techniques for testing the consistency of network positions in individuals. The statistical tools presented are useful because if particular individuals consistently occupy key positions in social networks, then this is also likely to have consequences for their fitness as well as for that of others in the population. These consequences may be particularly significant since individual network position has been shown to be important for the transmission of diseases, socially learnt information and genetic material between individuals and populations.     

Neophobia and social tolerance are related to breeding group size in a semi-colonial bird

In semi-colonial species, some individuals choose to breed in isolation while others aggregate in breeding colonies. The origin and the maintenance of this pattern have been questioned, and inherited phenotype dependency of group breeding benefits has been invoked as one of the possible mechanisms for the evolution of semi-coloniality. Using field observations and behavioural tests in the semi-colonial barn swallow (Hirundo rustica), we tested the hypothesis that breeding group size is related to personality. We measured neophobia (the fear and avoidance of new things) and social tolerance of adults and showed that these two independent traits of personality are strongly related to breeding group size. The biggest colonies hosted birds with higher neophobia, and larger groups also hosted females with higher social tolerance. This parallel between group size and group composition in terms of individual personality offers a better understanding of the observed diversity in breeding group size in this species. Further studies are, however, needed to better understand the origin of the link between individual personality and group breeding strategies.     

Social personality traits in chimpanzees: temporal stability and structure of behaviourally assessed personality traits in three captive populations

Animals of many species show consistency in behaviour across time and contexts that differs from other individuals' behaviour in the same population. Such ‘personality’ affects fitness and has therefore become an increasingly relevant research topic in biology. However, consistent variation in social behaviour is understudied. In socially living species, behaviour occurs in a social environment and social interactions have a significant influence on individual fitness. This study addressed personality in social behaviour of 75 captive chimpanzees in three zoos by coding observed behaviour. Fifteen behavioural variables were significantly repeatable (range 0.21–0.93) in at least two of the three zoos. The behaviours showed considerable long-term stability across 3 years, which did not differ from the short-term repeatability. The repeatable behaviours were then analysed with factor analyses. They formed five independent factors, three of which consisted of social traits and were labelled ‘sociability’, ‘positive affect’ and ‘equitability’. The two non-social behaviour factors were labelled ‘anxiety’ and ‘activity’. The factor scores were analysed for sex and population differences. Males had higher factor scores in all traits except ‘sociability’. The factor scores differed also between the zoos, implying considerable external effects in trait expression. The results show that chimpanzees show personality in a broad range of social and non-social behaviours. The study highlights the importance of assessing personality in the social behaviour, especially in cohesive social species, as only then can we understand the consequences of personality in socially living species.     

Do great tits rely on inadvertent social information from blue tits? A habitat selection experiment

Sympatric species sharing requirements are competitors, but recent evidence suggests that heterospecifics may also be used as a source of information. The heterospecific habitat copying hypothesis proposes that individuals of one species might use information inadvertently produced by the breeding performance of individuals of other species to assess habitat quality whenever the two species share needs. In this study, we provide the first experimental test of this hypothesis by examining whether the manipulated reproductive success of blue tits (Cyanistes caeruleus) is used as heterospecific inadvertent social information (ISI) in breeding-habitat selection by sympatric great tits (Parus major). The reproductive success of blue tits was manipulated 1year at the scale of patches by transferring nestlings from decreased to increased patches. No evidence was found of great tits using the reproductive success of blue tits as a source of heterospecific ISI. However, dispersal decisions by adult great tits correlated with information on con- and heterospecific densities, which constitute other sources of ISI. As density and breeding performance are tightly intertwined forms of information, the difficulty in distinguishing between them might lead great tits to use heterospecific ISI more in the form of density than breeding performance when making dispersal decisions.     

Are behavioral syndromes invariant? Spatiotemporal variation in shy/bold behavior in squid

Behavioral syndromes are correlated suites of behavior, analogous to human personality traits. Most work to date has been taken from limited “snapshots” in space and time, with the implicit assumption that a behavioral syndrome is an invariant property, fixed by evolutionary constraints or adaptations. However, directional selection on two mechanistically independent traits (selective covariance) could also result in correlated behaviors. Previously, we have shown that shy/bold behavior in Southern dumpling squid (Euprymna tasmanica) across predator encounter and feeding risk contexts is genetically and phenotypically uncoupled, and hence potentially free to vary independently. Here, we collected data on shy/bold behaviors from two independent wild populations of squid in two different years to test whether behavioral correlations across these same two functional contexts vary through time and space. We detected significant influences of population, sex, and body size on the expression of boldness in squid within each functional context, and this was coupled with significant differences in relative population density and adult sex ratio. Despite these changes in behavior and demographic parameters, we found that correlations between boldness scores across the two functional contexts were largely absent in both wild populations of squid in both years. Our work suggests that some animal groups may be largely characterized by context-specific behavioral expression. A theoretical framework which conceptualizes behavioral syndromes resulting from context-specific behavioral rules may be needed to fully understand why behaviors are sometimes correlated, and why sometimes they are not.     

Genetic variation and covariation during larval and juvenile growth in Mercenaria mercenaria

Quantitative genetic variances and covariances were estimated for shell length of the hard clam Mercenaria mercenaria (L.) at three larval stages (prodissoconch I, 2 d and 10 d post-fertilization) in 1987 and in 1988 after ca. 9 mo of growth. At each sample interval additive genetic variance was a highly significant component of the total size variation. Narrow sense heritability estimates for shell length ranged between 0.58 (±0.10) for prodissoconch I and 1.08 (±0.29) for 2-d-old larvae. There was significant and positive genetic covariance in prodissoconch I and 2-d larval shell length which resulted in a highly significant genetic correlation (r _g=0.74) between these two traits. This covariance is not surprising since the prodissoconch I comprises the majority of the larval shell of a 2-d-old larvae. The genetic covariances between 2-d-old and 10-d-old larvae and between 10-d-old larvae and 9-mo-old juveniles were low and not significantly different from zero. These results indicate that there is substantial genetic variation for shell growth in M. mercenaria but this variation is not stable during development; the genetic variation in shell growth at one stage of development is not strongly related to the genetic variation in growth during other ontogenetic periods. In this study there were no evident constraints to natural selection for increased shell growth rate during development, which coupled with the high levels of genetic variation may suggest that in nature high rates of larval growth may not be normally subject to significant selective pressure.     

Using average autonomy to test whether behavioral syndromes constrain evolution

Genetic correlations between behaviors underlying a behavioral syndrome may constrain the capacity of a population to respond to selection on these behaviors. Average autonomy quantifies the extent in which estimated genetic (co)variances constrain the rate of evolutionary change of behavioral traits forming a syndrome when these traits are under selections in all possible directions of multivariate trait-space. However, it is not clear whether a calculated average autonomy value of an observed syndrome constitutes a significant evolutionary constraint or not. I here outline an approach for testing evolutionary constraint in a syndrome, which is based on comparing the observed genetic (co)variance structure to the one where the genetic covariances are assumed to be zero and taking onboard the uncertainty in the (co)variances between behaviors into the calculations of average autonomy. The approach can be implemented in the context of parametric bootstrap or Bayesian statistics, and I provide a worked example of the latter. I further highlight that when genetic (co)variances are unattainable, the between-individual (co)variances act as an interesting proxy, which is within reach for many behavioral studies. I provide R code for all calculations.     

Behavioural trait assortment in a social network: patterns and implications

The social fine structure of a population plays a central role in ecological and evolutionary processes. Whilst many studies have investigated how morphological traits such as size affect social structure of populations, comparatively little is known about the influence of behaviours such as boldness and shyness. Using information on social interactions in a wild population of Trinidadian guppies (Poecilia reticulata), we construct a social network. For each individual in the network, we quantify its behavioural phenotype using two measures of boldness, predator inspection tendency, a repeatable and reliably measured behaviour well studied in the context of co-operation, and shoaling tendency. We observe striking heterogeneity in contact patterns, with strong ties being positively assorted and weak ties negatively assorted by our measured behavioural traits. Moreover, shy fish had more network connections than bold fish and these were on average stronger. In other words, social fine structure is strongly influenced by behavioural trait. We assert that such structure will have implications for the outcome of selection on behavioural traits and we speculate that the observed positive assortment may act as an amplifier of selection contributing to the maintenance of co-operation during predator inspection.     

Effect of blue tit song syntax on great tit territorial responsiveness – an experimental test of the character shift hypothesis

A previous study of geographic variation in blue tit (Parus caeruleus) song structure showed that changes in blue tit song syntax (i.e. presence/absence of a trill) are correlated with the breeding density of a close competitor, the great tit (P. major), and are not correlated with other environmental factors such as vegetation structure or blue tit breeding density. We tested the hypothesis that blue tit trilled song represents a character shift that evolved because it reduced territorial interactions with more dominant great tits. We conducted five sets of playback trials in three study populations (mainland southern France, Corsica and Denmark) presenting male great tits with blue tit trilled and untrilled songs and great tit songs. We found that great tits respond equally strongly to both blue tit untrilled songs and to great tit songs, but show a significantly weaker response to blue tit trilled songs. These findings are the first experimental evidence that interspecific competition may play an important role in macrogeographic variation of bird song. Received: 10 October 1999 / Received in revised form: 2 April 2000 / Accepted: 15 April 2000