Intra-individual variability in fecal cortisol metabolites varies with lifetime exploration and reproductive life history in eastern chipmunks (Tamias striatus)

Overcoming the pitfalls of categorizing continuous variables in ecology, evolution and behaviour

Many variables in biological research-from body size to life-history timing to environmental characteristics-are measured continuously (e.g. body mass in kilograms) but analysed as categories (e.g. large versus small), which can lower statistical power and change interpretation. We conducted a mini-review of 72 recent publications in six popular ecology, evolution and behaviour journals to quantify the prevalence of categorization. We then summarized commonly categorized metrics and simulated a dataset to demonstrate the drawbacks of categorization using common variables and realistic examples. We show that categorizing continuous variables is common (31% of publications reviewed). We also underscore that predictor variables can and should be collected and analysed continuously. Finally, we provide recommendations on how to keep variables continuous throughout the entire scientific process. Together, these pieces comprise an actionable guide to increasing statistical power and facilitating large synthesis studies by simply leaving continuous variables alone. Overcoming the pitfalls of categorizing continuous variables will allow ecologists, ethologists and evolutionary biologists to continue making trustworthy conclusions about natural processes, along with predictions about their responses to climate change and other environmental contexts.

Differences in resource acquisition, not allocation, mediate the relationship between behaviour and fitness: a systematic review and meta-analysis

Within populations, individuals often show repeatable variation in behaviour, called 'animal personality'. In the last few decades, numerous empirical studies have attempted to elucidate the mechanisms maintaining this variation, such as life-history trade-offs. Theory predicts that among-individual variation in behavioural traits could be maintained if traits that are positively associated with reproduction are simultaneously associated with decreased survival, such that different levels of behavioural expression lead to the same net fitness outcome. However, variation in resource acquisition may also be important in mediating the relationship between individual behaviour and fitness components (survival and reproduction). For example, if certain phenotypes (e.g. dominance or aggressiveness) are associated with higher resource acquisition, those individuals may have both higher reproduction and higher survival, relative to others in the population. When individuals differ in their ability to acquire resources, trade-offs are only expected to be observed at the within-individual level (i.e. for a given amount of resource, if an individual increases its allocation to reproduction, it comes at the cost of allocation to survival, and vice versa), while among individuals traits that are associated with increased survival may also be associated with increased reproduction. We performed a systematic review and meta-analysis, asking: (i) do among-individual differences in behaviour reflect among-individual differences in resource acquisition and/or allocation, and (ii) is the relationship between behaviour and fitness affected by the type of behaviour and the testing environment? Our meta-analysis consisted of 759 estimates from 193 studies. Our meta-analysis revealed a positive correlation between pairs of estimates using both survival and reproduction as fitness proxies. That is, for a given study, behaviours that were associated with increased reproduction were also associated with increased survival, suggesting that variation in behaviour at the among-individual level largely reflects differences among individuals in resource acquisition. Furthermore, we found the same positive correlation between pairs of estimates using both survival and reproduction as fitness proxies at the phenotypic level. This is significant because we also demonstrated that these phenotypic correlations primarily reflect within-individual correlations. Thus, even when accounting for among-individual differences in resource acquisition, we did not find evidence of trade-offs at the within-individual level. Overall, the relationship between behaviour and fitness proxies was not statistically different from zero at the among-individual, phenotypic, and within-individual levels; this relationship was not affected by behavioural category nor by the testing condition. Our meta-analysis highlights that variation in resource acquisition may be more important in driving the relationship between behaviour and fitness than previously thought, including at the within-individual level. We suggest that this may come about via heterogeneity in resource availability or age-related effects, with higher resource availability and/or age leading to state-dependent shifts in behaviour that simultaneously increase both survival and reproduction. We emphasize that future studies examining the mechanisms maintaining behavioural variation in populations should test the link between behavioural expression and resource acquisition - both within and among individuals. Such work will allow the field of animal personality to develop specific predictions regarding the mediating effect of resource acquisition on the fitness consequences of individual behaviour.

Quantifying Glucocorticoid Plasticity Using Reaction Norm Approaches: There Still is So Much to Discover!

Hormones are highly responsive internal signals that help organisms adjust their phenotype to fluctuations in environmental and internal conditions. Our knowledge of the causes and consequences of variation in circulating hormone concentrations has improved greatly in the past. However, this knowledge often comes from population-level studies, which generally tend to make the flawed assumption that all individuals respond in the same way to environmental changes. Here, we advocate that we can vastly expand our understanding of the ecology and evolution of hormonal traits once we acknowledge the existence of individual differences by quantifying hormonal plasticity at the individual level, where selection acts. In this review, we use glucocorticoid (GC) hormones as examples of highly plastic endocrine traits that interact intimately with energy metabolism but also with other organismal traits like behavior and physiology. First, we highlight the insights gained by repeatedly assessing an individual's GC concentrations along a gradient of environmental or internal conditions using a “reaction norm approach.” This study design should be followed by a hierarchical statistical partitioning of the total endocrine variance into the among-individual component (individual differences in average hormone concentrations, i.e., in the intercept of the reaction norm) and the residual (within-individual) component. The latter is ideally further partitioned by estimating more precisely hormonal plasticity (i.e., the slope of the reaction norm), which allows to test whether individuals differ in the degree of hormonal change along the gradient. Second, we critically review the published evidence for GC variation, focusing mostly on among- and within-individual levels, finding only a good handful of studies that used repeated-measures designs and random regression statistics to investigate GC plasticity. These studies indicate that individuals can differ in both the intercept and the slope of their GC reaction norm to a known gradient. Third, we suggest rewarding avenues for future work on hormonal reaction norms, for example to uncover potential costs and trade-offs associated with GC plasticity, to test whether GC plasticity varies when an individual's reaction norm is repeatedly assessed along the same gradient, whether reaction norms in GCs covary with those in other traits like behavior and fitness (generating multivariate plasticity), or to quantify GC reaction norms along multiple external and internal gradients that act simultaneously (leading to multidimensional plasticity). Throughout this review, we emphasize the power that reaction norm approaches offer for resolving unanswered questions in ecological and evolutionary endocrinology.

Albatrosses respond adaptively to climate variability by changing variance in a foraging trait

The ability of individuals and populations to adapt to a changing climate is a key determinant of population dynamics. While changes in mean behaviour are well studied, changes in trait variance have been largely ignored, despite being assumed to be crucial for adapting to a changing environment. As the ability to acquire resources is essential to both reproduction and survival, changes in behaviours that maximize resource acquisition should be under selection. Here, using foraging trip duration data collected over 7 years on black-browed albatrosses (Thalassarche melanophris) on the Kerguelen Islands in the southern Indian Ocean, we examined the importance of changes in the mean and variance in foraging behaviour, and the associated effects on fitness, in response to the El Niño Southern Oscillation (ENSO). Using double hierarchical models, we found no evidence that individuals change their mean foraging trip duration in response to a changing environment, but found strong evidence of changes in variance. Younger birds showed greater variability in foraging trip duration in poor conditions as did birds with higher fitness. However, during brooding, birds showed greater variability in foraging behaviour under good conditions, suggesting that optimal conditions allow the alteration between chick provisioning and self-maintenance trips. We found weak correlations between sea surface temperature and the ENSO, but stronger links with sea-level pressure. We suggest that variability in behavioural traits affecting resource acquisition is under selection and offers a mechanism by which individuals can adapt to a changing climate. Studies which look only at effects on mean behaviour may underestimate the effects of climate change and fail to consider variance in traits as a key evolutionary force.

Natural and human-induced environmental changes and their effects on adaptive potential of wild animal populations

A major challenge of evolutionary ecology over the next decades is to understand and predict the consequences of the current rapid and important environmental changes on wild populations. Extinction risk of species is linked to populations' evolutionary potential and to their ability to express adaptive phenotypic plasticity. There is thus a vital need to quantify how selective pressures, quantitative genetics parameters, and phenotypic plasticity, for multiple traits in wild animal populations, may vary with changes in the environment. Here I review our previous research that integrated ecological and evolutionary theories with molecular ecology, quantitative genetics, and long-term monitoring of individually marked wild animals. Our results showed that assessing evolutionary and plastic changes over time and space, using multi-trait approaches, under a realistic range of environmental conditions are crucial steps toward improving our understanding of the evolution and adaptation of natural populations. Our current and future work focusses on assessing the limits of adaptive potential by determining the factors constraining the evolvability of plasticity, those generating covariation among genetic variance and selection, as well as indirect genetic effects, which can affect population's capacity to adjust to environmental changes. In doing so, we aim to provide an improved assessment of the spatial and temporal scale of evolutionary processes in wild animal populations.

Stressed by Maternity: Changes of Cortisol Level in Lactating Domestic Cats

Lactation is the most energetically expensive component of maternal care in mammals. Increased reproductive investment can lead to physiological stress for the mothers, based on the exhaustion of energy resources and increase in glucocorticoids level. This study aimed to estimate the changes in cortisol concentrations during lactation in domestic cats and compared the differences among litter sizes. Eleven females gave birth to 27 litters, which were divided in two groups-small (1-3 kittens) and large (4-7 kittens) litters. Blood samples were collected from each female before mating, after parturition, at 4 and 8 weeks of lactation. We showed that the cortisol level in females changed significantly during lactation-the highest concentrations were observed at the peak of lactation at 4 weeks. Cortisol levels varied significantly among females but did not depend on their maternal experience. We also revealed that there were no differences in cortisol levels between females with small and large litters, but at 4 weeks of lactation, the hormone concentrations were higher in females with small litters. It is likely that these females initially invested less in reproduction, giving birth to fewer offspring.

Physical and Physiological Indicators of Welfare in Guinea Pigs (Cavia porcellus) Serving as Ambassador Animals

Special encounters that allow contact between animals and guests are common in zoos and aquariums. Visitors to the Saint Louis Zoo may touch guinea pigs serving as ambassador animals. We evaluated two welfare indicators in ambassador guinea pigs by comparing glucocorticoid levels and body weights between periods when guinea pigs lived in a habitat accessible to the public and while off exhibit. Mean glucocorticoid levels did not differ between sexes or between on- and off-exhibit periods. There was significant individual variation, and females demonstrated greater variation than males. While on exhibit, glucocorticoid levels slightly but significantly increased in males and decreased in females. Moving guinea pigs between habitats only elicited a significant glucocorticoid response when females were moved off exhibit. Temporary closures of the exhibit had no effect on glucocorticoid levels in either sex. Analyses of the impact of handling rates on males found no impact on glucocorticoid levels. Guinea pigs' body weights were lower while on exhibit. We conclude that guinea pigs serving as ambassador animals at the Saint Louis Zoo demonstrate comparable physiological profiles while on and off exhibit and, when used in a rotational schedule, are a suitable species for animal encounters involving contact with the public.

Repeatable glucocorticoid expression is associated with behavioural syndromes in males but not females in a wild primate

Behavioural syndromes are a well-established phenomenon in human and non-human animal behavioural ecology. However, the mechanisms that lead to correlations among behaviours and individual consistency in their expression at the apparent expense of behavioural plasticity remain unclear. The 'state-dependent' hypothesis posits that inter-individual variation in behaviour arises from inter-individual variation in state and that the relative stability of these states within an individual leads to consistency of behaviour. The endocrine stress response, in part mediated by glucocorticoids (GCs), is a proposed behavioural syndrome-associated state as GC levels are linked to an individual's behavioural responses to stressors. In this study, in wild Barbary macaques (Macaca sylvanus), consistent inter-individual differences were observed in both sexes for GC activity (faecal glucocorticoid, fGC concentrations), but not GC variation (coefficient of variation in fGC concentrations). The expression of the behavioural syndrome 'Excitability' (characterized by the frequencies of brief affiliation or aggressive interactions) was related to GC activity in males but not in females; more 'excitable' males had lower GC activity. There was no relationship in females between any of the behavioural syndromes and GC activity, nor in either sex with GC variation. The negative relationship between GC activity and Excitability in males provides some support for GC expression as a behavioural syndrome-generating state under the state-dependent framework. The absence of this relationship in females highlights that state-behavioural syndrome associations may not be generalizable within a species and that broader sex differences in state need to be considered for understanding the emergence and maintenance of behavioural syndromes.

Relationships between personality traits and the physiological stress response in a wild mammal

Glucocorticoids (GCs) are involved in the regulation of an animal's energetic state. Under stressful situations, they are part of the neuroendocrine response to cope with environmental challenges. Animals react to aversive stimuli also through behavioral responses, defined as coping styles. Both in captive and wild populations, individuals differ in their behavior along a proactive-reactive continuum. Proactive animals exhibit a bold, active-explorative and social personality, whereas reactive ones are shy, less active-explorative and less social. Here, we test the hypothesis that personality traits and physiological responses to stressors covary, with more proactive individuals having a less pronounced GC stress response. In wild populations of invasive gray squirrels Sciurus carolinensis, we measured fecal glucocorticoid metabolites (FGMs), an integrated measure of circulating GCs, and 3 personality traits (activity, sociability, and exploration) derived from open field test (OFT) and mirror image stimulation (MIS) test. Gray squirrels had higher FGMs in Autumn than in Winter and males with scrotal testes had higher FGMs than nonbreeding males. Personality varied with body mass and population density. Squirrels expressed more activity-exploration at higher than at lower density and heavier squirrels had higher scores for activity-exploration than animals that weighed less. Variation in FGM concentrations was not correlated with the expression of the 3 personality traits. Hence, our results do not support a strong association between the behavioral and physiological stress responses but show that in wild populations, where animals experience varying environmental conditions, the GC endocrine response and the expression of personality are uncorrelated traits among individuals.

Among-population divergence in personality is linked to altitude in plateau pikas (Ochotona curzoniae

BACKGROUND

Animals inhabiting high altitudes consistently show slow life-histories. The pace-of-life syndrome (POLS) hypothesis posits behavioural, physiological and/or morphological traits that mediate the trade-off between current and future reproduction or survival, which have coevolved along a slow-fast life history continuum. Previous studies have shown that the life histories of plateau pikas varied across altitude, high-altitude individuals showed slow pace of life which were characterized by few litters per year with small litter sizes. Thus, we hypothesized that pikas populations at higher altitudes would also express personalities characteristic associated with slow life history, such as high sociability, low activity or aggressiveness. We tested this hypothesis by comparing the activity and docility of three plateau pika (Ochotona curzoniae) populations distributed along an altitudinal gradient of the Tibetan Plateau. We predicted that high-altitude pika would be more docile and less active.

RESULTS

The behaviour of 556 pikas, from which 120 individuals were measured at least twice, was quantified. We observed that plateau pikas at high altitudes were less active and more docile than pika at lower altitudes. Activity and docility were significantly and negatively correlated in populations from high altitudes but not in populations from low altitudes.

CONCLUSIONS

Our results support the POLS hypothesis, highlight the existence of personality variation among populations distributed along an altitudinal gradient and emphasise the importance of environmental selection on personality divergence.

Independence between coping style and stress reactivity in plateau pika

The concept of coping style represents the way individual animals react to a stressful situation, both behaviourally and neurophysiologically. Over the last decades coping style has been linked to the development of research on animal personality. Based on this concept, we should find a proactive-reactive continuum in animal populations, with proactive individuals being fast explorer, bold, aggressive, and show high sympathetic reactivity (higher heart rate), as well as low hypothalamus-pituitary-adrenocortical (PHA) axis reactivity to external stressor (higher plasma glucocorticoid level). At the other extreme, shy, lowly aggressive, reactive individuals should be slow in their exploration, and show a low sympathetic reactivity and a high HPA axis reactivity. However, a recent two-tier model proposed that coping style and stress reactivity should be independent of each other. In this study, we tested the two-tier model in a wild plateau pika (Ochotona curzoniae) population on the Tibetan Plateau, by quantifying the associations between several behavioural and physiological traits at the among- and within-individual levels. We repeatedly measured exploration, docility, boldness, heart rate and plasma cortisol concentration in individuals between April and September of 2013. All traits tested were repeatable. At the among-individual level, all behavioural traits were correlated with each other and with heart rate, but were independent of both basal level and variation of plasma cortisol concentration. Most correlations were negligible at the within-individual level. In support of the two-tier model, these results suggest that coping style (i.e. behaviour and heart rate associations) is independent of stress reactivity (i.e. glucocorticoid reactivity) in that species.

Associations between glucocorticoids and sociality across a continuum of vertebrate social behavior

The causes and consequences of individual differences in animal behavior and stress physiology are increasingly studied in wild animals, yet the possibility that stress physiology underlies individual variation in social behavior has received less attention. In this review, we bring together these study areas and focus on understanding how the activity of the vertebrate neuroendocrine stress axis (HPA-axis) may underlie individual differences in social behavior in wild animals. We first describe a continuum of vertebrate social behaviors spanning from initial social tendencies (proactive behavior) to social behavior occurring in reproductive contexts (parental care, sexual pair-bonding) and lastly to social behavior occurring in nonreproductive contexts (nonsexual bonding, group-level cooperation). We then perform a qualitative review of existing literature to address the correlative and causal association between measures of HPA-axis activity (glucocorticoid levels or GCs) and each of these types of social behavior. As expected, elevated HPA-axis activity can inhibit social behavior associated with initial social tendencies (approaching conspecifics) and reproduction. However, elevated HPA-axis activity may also enhance more elaborate social behavior outside of reproductive contexts, such as alloparental care behavior. In addition, the effect of GCs on social behavior can depend upon the sociality of the stressor (cause of increase in GCs) and the severity of stress (extent of increase in GCs). Our review shows that the while the associations between stress responses and sociality are diverse, the role of HPA-axis activity behind social behavior may shift toward more facilitating and less inhibiting in more social species, providing insight into how stress physiology and social systems may co-evolve.

Sex differences in life history, behavior, and physiology along a slow-fast continuum: a meta-analysis

The pace-of-life syndrome (POLS) hypothesis predicts that behavior and physiology covary with life history. Evidence for such covariation is contradictory, possibly because systematic sources of variation (e.g. sex) have been neglected. Sexes often experience different selection pressures leading to sex-specific allocation between reproduction and self-maintenance, facilitating divergence in life-history. Sex-specific differences in means and possibly variances may therefore play a key role in the POLS framework. We investigate whether sexes differ in means and variances along the fast-slow pace-of-life continuum for life history and physiological and behavioral traits. In addition, we test whether social and environmental characteristics such as breeding strategy, mating system, and study environment explain heterogeneity between the sexes. Using meta-analytic methods, we found that populations with a polygynous mating system or for studies conducted on wild populations, males had a faster pace-of-life for developmental life-history traits (e.g., growth rate), behavior, and physiology. In contrast, adult life-history traits (e.g., lifespan) were shifted towards faster pace-of-life in females, deviating from the other trait categories. Phenotypic variances were similar between the sexes across trait categories and were not affected by mating system or study environment. Breeding strategy did not influence sex differences in variances or means. We discuss our results in the light of sex-specific selection that might drive sex-specific differences in pace-of-life and ultimately POLS.

Hormonal and fitness consequences of behavioral assortative mating in the convict cichlid (Amatitlania siquia)

In monogamous species, partner compatibility is a key factor influencing pairing and reproductive success. In pairs with biparental care, studies have mostly focused on behavioral compatibility because it is likely to encourage the coordination of parental care within pairs, leading to a better reproductive success. Behavior modulation, throughout the breeding season and as a function of the social context, is under the regulatory feedback control of endocrine mechanisms. From this link, the idea of hormonal partner compatibility as a key component of pair cohesion and maintenance has recently emerged. Here, we investigated the link between partner behavioral assortment and their hormonal response to the pairing context. We formed assortative and disassortative pairs of convict cichlids based on their behavioral type (proactive or reactive) and took hormone and fitness measurements. Testosterone, 11-ketotestosterone, 17β-estradiol and cortisol levels were measured from fish-holding water before and after pair formation. We found no relationship between the behavioral type of individuals and their pre-pairing hormone levels. Only the level of cortisol was affected by the partner but independently of its behavioral type. Reproductive success was not affected by the level of hormonal similarity within pairs, but we found that the variation in 11-ketotestosterone similarity between the isolated context and the pairing context was related to spawning size, and the variation in cortisol similarity to the number of fry. Behavioral compatibility does not reflect hormonal compatibility in the convict cichlid, but the relationship between reproductive success and the flexibility of hormonal similarity between partners suggests hormonal adjustment within pairs in this species.

Towards powerful experimental and statistical approaches to study intraindividual variability in labile traits

There is a long-standing interest in behavioural ecology, exploring the causes and correlates of consistent individual differences in mean behavioural traits ('personality') and the response to the environment ('plasticity'). Recently, it has been observed that individuals also consistently differ in their residual intraindividual variability (rIIV). This variation will probably have broad biological and methodological implications to the study of trait variation in labile traits, such as behaviour and physiology, though we currently need studies to quantify variation in rIIV, using more standardized and powerful methodology. Focusing on activity rates in guppies (Poecilia reticulata), we provide a model example, from sampling design to data analysis, in how to quantify rIIV in labile traits. Building on the doubly hierarchical generalized linear model recently used to quantify individual differences in rIIV, we extend the model to evaluate the covariance between individual mean values and their rIIV. After accounting for time-related change in behaviour, our guppies substantially differed in rIIV, and it was the active individuals that tended to be more consistent (lower rIIV). We provide annotated data analysis code to implement these complex models, and discuss how to further generalize the model to evaluate covariances with other aspects of phenotypic variation.