State dependence explains individual variation in nest defence behaviour in a long-lived bird.

Parental care, such as nest or offspring defence, is crucial for offspring survival in many species. Yet, despite its obvious fitness benefits, the level of defence can consistently vary between individuals of the same species. One prominent adaptive explanation for consistent individual differences in behaviours involves state dependency: relatively stable differences in individual state should lead to the emergence of repeatable behavioural variation whereas changes in state should lead to a readjustment of behaviour. Therefore, empirical testing of adaptive state dependence requires longitudinal data where behaviour and state of individuals of the same population are repeatedly measured. Here, we test if variation in states predicts nest defence behaviour (a 'risky' behaviour) in a long-lived species, the barnacle goose Branta leucopsis. Adaptive models have predicted that an individual's residual reproductive value or 'asset' is an important state variable underlying variation in risk-taking behaviour. Hence, we investigate how nest defence varies as a function of time of the season and individual age, two state variables that can vary between and within individuals and determine asset. Repeated measures of nest defence towards a human intruder (flight initiation distance or FID) of females of known age were collected during 15 breeding seasons. Increasing values of FID represent increasing shyness. We found that females strongly and consistently differed in FID within- and between-years. As predicted by theory, females adjusted their behaviour to state by decreasing their FID with season and age. Decomposing these population patterns into within- and between-individual effects showed that the state-dependent change in FID was driven by individual plasticity in FID and that bolder females were more plastic than shyer females. This study shows that nest defence behaviour differs consistently among individuals and is adjusted to individual state in a direction predicted by adaptive personality theory.

Social competition as a driver of phenotype-environment correlations: implications for ecology and evolution.

While it is universally recognised that environmental factors can cause phenotypic trait variation via phenotypic plasticity, the extent to which causal processes operate in the reverse direction has received less consideration. In fact individuals are often active agents in determining the environments, and hence the selective regimes, they experience. There are several important mechanisms by which this can occur, including habitat selection and niche construction, that are expected to result in phenotype-environment correlations (i.e. non-random assortment of phenotypes across heterogeneous environments). Here we highlight an additional mechanism - intraspecific competition for preferred environments - that may be widespread, and has implications for phenotypic evolution that are currently underappreciated. Under this mechanism, variation among individuals in traits determining their competitive ability leads to phenotype-environment correlation; more competitive phenotypes are able to acquire better patches. Based on a concise review of the empirical evidence we argue that competition-induced phenotype-environment correlations are likely to be common in natural populations before highlighting the major implications of this for studies of natural selection and microevolution. We focus particularly on two central issues. First, competition-induced phenotype-environment correlation leads to the expectation that positive feedback loops will amplify phenotypic and fitness variation among competing individuals. As a result of being able to acquire a better environment, winners gain more resources and even better phenotypes - at the expense of losers. The distinction between individual quality and environmental quality that is commonly made by researchers in evolutionary ecology thus becomes untenable. Second, if differences among individuals in competitive ability are underpinned by heritable traits, competition results in both genotype-environment correlations and an expectation of indirect genetic effects (IGEs) on resource-dependent life-history traits. Theory tells us that these IGEs will act as (partial) constraints, reducing the amount of genetic variance available to facilitate evolutionary adaptation. Failure to recognise this will lead to systematic overestimation of the adaptive potential of populations. To understand the importance of these issues for ecological and evolutionary processes in natural populations we therefore need to identify and quantify competition-induced phenotype-environment correlations in our study systems. We conclude that both fundamental and applied research will benefit from an improved understanding of when and how social competition causes non-random distribution of phenotypes, and genotypes, across heterogeneous environments.

Reproductive effort and future parental competitive ability: A nest box removal experiment.

The life history trade-off between current and future reproduction is a theoretically well-established concept. However, empirical evidence for the occurrence of a fitness cost of reproduction is mixed. Evidence indicates that parents only pay a cost of reproduction when local competition is high. In line with this, recent experimental work on a small passerine bird, the Great tit (Parus major) showed that reproductive effort negatively affected the competitive ability of parents, estimated through competition for high quality breeding sites in spring. In the current study, we further investigate the negative causal relationship between reproductive effort and future parental competitive ability, with the aim to quantify the consequences for parental fitness, when breeding sites are scarce. To this end, we (a) manipulated the family size of Great tit parents and (b) induced severe competition for nest boxes among the parents just before the following breeding season by means of a large-scale nest box removal experiment. Parents increased their feeding effort in response to our family size manipulation and we successfully induced competition among the parents the following spring. Against our expectation, we found no effect of last season's family size on the ability of parents to secure a scarce nest box for breeding. In previous years, if detected, the survival cost of reproduction was always paid after midwinter. In this year, parents did pay a survival cost of reproduction before midwinter and thus before the onset of the experiment in early spring. Winter food availability during our study year was exceptionally low, and thus, competition in early winter may have been extraordinarily high. We hypothesize that differences in parental competitive ability due to their previous reproductive effort might have played a role, but before the onset of our experiment and resulted in the payment of the survival cost of reproduction.

Is parental competitive ability in winter negatively affected by previous springs' family size?

Reproductive behavior cannot be understood without taking the local level of competition into account. Experimental work in great tits (Parus major) showed that (1) a survival cost of reproduction was paid in environments with high levels of competition during the winter period and (2) experimentally manipulated family size negatively affected the ability of parents to compete for preferred breeding boxes in the next spring. The fact that survival was affected in winter suggests that the competitive ability of parents in winter may also be affected by previous reproductive effort. In this study, we aim to investigate whether (1) such carryover effects of family size on the ability of parents to compete for resources in the winter period occurred and (2) this could explain the occurrence of a survival cost of reproduction under increased competition. During two study years, we manipulated the size of in total 168 great tit broods. Next, in winter, we induced competition among the parents by drastically reducing the availability of roosting boxes in their local environment for one week. Contrary to our expectation, we found no negative effect of family size manipulation on the probability of parents to obtain a roosting box. In line with previous work, we did find that a survival cost of reproduction was paid only in plots in which competition for roosting boxes was shortly increased. Our findings thus add to the scarce experimental evidence that survival cost of reproduction are paid under higher levels of local competition but this could not be linked to a reduced competitive ability of parents in winter.