Tadpoles show various behavioural, morphological and life-history adaptations to evade predators or escape their attacks. These traits are costly and their expression involves trade-offs, resulting in predictable patterns of phenotypic integration. Within individuals, these patterns arise from plastic adjustments in traits in response to extrinsic factors, such as predation risk. In contrast, among individuals, they reflect differences in both extrinsic and intrinsic factors, such as resource availability or genetic quality. Here, we tested the hypothesis that these patterns of integration, both within and among individuals, directly depend on how tadpoles manage their energy budgets. Our approach was both theoretical and empirical. We created an optimality model for tadpole life history. This model examined two key parameters: the time spent feeding, which determined both the tadpole’s daily energy budget and the probability of encountering predators, and the portion of the daily energy budget allocated to developing defensive traits. While this distribution increased the likelihood of avoiding predator attacks, it also decreased growth and development rates. The model also considered two patterns of among-individual variation: variation in metabolic efficiency and metabolic capacity. From this model, we derived predictions about trait covariation within and among individuals. To test these predictions, we conducted an experiment using tadpoles of the brown frog, Rana dalmatina . We raised siblings either in the presence or absence of predators, and we repeatedly tested them in arenas both with and without caged predators. We observed a consistent relationship between theoretical and empirical findings at the individual level. Tadpoles that were raised with predators showed reduced activity, developed larger defensive traits (such as deeper tails and faster movements), grew more slowly and took longer to metamorphose than their control siblings. In both ontogenetic treatments, the observed among-individual covariation pattern supported the idea that tadpoles differed in metabolic efficiency rather than metabolic capacity. • Predators prompt covariation in behaviour, morphology and life history of tadpoles. • We investigate optimal trade-offs with mathematical models and a rearing experiment. • Trait covariation explains trade-off between growth and survival within individuals. • Among individuals, similar trade-offs but different resources explain covariation.

Extreme temperatures and heatwave events present challenging conditions for wildlife and are increasing in frequency and intensity in many regions due to climate change. High daily temperatures increase physiological stress and cause mortality in susceptible individuals (e.g. from poor health or exposure) but may also drive behavioural changes as individuals seek to thermoregulate (e.g. seeking shelter or water). As daily high temperatures accumulate into heatwave events, the ability of wildlife to tolerate conditions can diminish and exacerbate stress. Although climate change is well known to decouple species interactions, here we examined how extreme conditions may intensify interactions between predators and prey. In particular, we explored whether predators can exploit the thermoregulatory requirements of prey as they increasingly require access to water. We presented evidence from the use of artificial waterpoints by dingoes, Canis dingo , and eastern grey kangaroos, Macropus giganteus , on a wildlife property in Australia's semiarid drylands, asking whether high temperatures and heatwaves alter species behaviour and interactions between them. Both species increasingly accessed waterpoints as daily maximum temperatures increased; however, the degree of co-occurrence at waterpoints increased significantly as the temperatures became extreme and resulted in heatwaves. Not only did waterpoints become increasingly important for both species during heatwaves, but dingo hunting attempts on kangaroos at these times were significantly higher than expected based on the number of days when heat anomalies occurred. Additionally, dingoes relaxed their predisposition to hunt kangaroos in poor body condition, making kangaroo hunting opportunistic during these times. Our findings emphasize the potential for climate change to exacerbate predator–prey interactions, which may have an impact on species persistence if prey is unable to adapt to predators that exploit their thermoregulatory demands. • We studied the influence of heatwaves on kangaroo - dingo interactions at waterpoints. • Kangaroo and dingo activity at waterpoints increased significantly during heatwaves. • High activity rates meant kangaroos were unable to avoid interacting with dingoes. • Kangaroos of all body conditions experienced predation during extreme temperatures.

Parasites impose strong selection on the immune genes of their hosts. Much of what we know about the evolutionary ecology of immune genes comes from studies of the major histocompatibility complex (MHC), whose products recognize nonself-antigens and initiate an adaptive immune response. MHC loci are remarkably polymorphic, due in part to widespread nonrandom (generally disassortative) mating. At innate immune loci, however, it remains an open question whether nonrandom mating occurs and, if so, whether mating is assortative or disassortative. We sequenced wild-breeding song sparrows, Melospiza melodia , at part of the extracellular domain of Toll-like receptor 3 ( TLR3 ), an innate immune locus that detects viral infection and varies geographically across song sparrow populations. Social mates were more similar at TLR3 than potential mates (i.e. cross-sex pairs of individuals not mated to one another). By contrast, social mates did not differ from potential mates in neutral-locus (microsatellite) genetic similarity. This is the first evidence, to our knowledge, that innate immune loci may be targets of nonrandom mating. Assortative mating at TLR3 could be adaptive if homozygotes have a fitness advantage, perhaps involving enhanced dimerization. We conclude that (1) nonrandom mating with respect to immune genotype is not restricted to genes of the MHC and (2) individual genetic variation at TLR3 in this population appears to be maintained despite, not because of, patterns of mate choice. • We examined whether nonrandom mating occurs at innate immune loci in song sparrows. • Social pairs were more genetically similar than random male–female pairs at TLR3. • Social pairs were not more similar at neutral loci (microsatellites). • Ours is the first demonstration of nonrandom mating based on non-MHC immune loci. • TLR3 variation is presumably maintained despite, not because of, mate choice patterns.