The integrative phenotype of moor frogs (Rana arvalis) - Evolutionary physiology and adaptation to multidimensional selection

In nature, populations can be affected by multiple chronic and short-term stressors simultaneously, which can interact with each other both by influencing phenotypic expression and as agents of natural selection. These interactions can shape adaptive processes at the level of populations and ultimately influence phenotypic and genetic variation in nature. Adaptive processes can be characterized by multiple selective forces acting on multidimensional phenotypes, which can lead to trade-offs in trait evolution. The integrative phenotype, characterized by all traits of an individual, can be split into genetic (G), environmental (E) and G x E interaction components. Hence, adaptation to environmental stressors can be driven simultaneously by genetic variation, phenotypic plasticity and a combination of both. Physiological traits, such as glucocorticoids, are often linked to many other traits of the composite phenotype. Their expression is influenced by abiotic as well as biotic environmental stressors and they can facilitate adaptive responses of organisms through plasticity or genetic variation. However, the exact role of physiological processes in adaptation of natural populations to stress is poorly understood. In this thesis, I aimed to understand the role of the integrative phenotype, and specifically of corticosterone, as a mediator of adaptation to environmental stress using individual level studies. I investigated the integrative phenotype of Rana arvalis tadpoles in five populations (two acid [AOP], one intermediate [IOP] and one neutral pH origin [NOP] population) along an acidification gradient in Sweden. These populations were previously found to show adaptive divergence and dynamic plastic responses to acid and predator stress in tadpole life history and morphological traits. I used a series of common garden laboratory experiments, where individuals from different populations were reared in different combinations of two pH (acid or neutral) and two predator stressor (predator cue or no cue) treatments. In chapter I, I reared individuals from three populations (AOP, IOP, NOP) under chronically acid or neutral pH conditions to study the links between corticosterone and mid-larval stage traits (life history and morphology). I found divergence between populations in corticosterone levels, with AOP tadpoles having lower and NOP tadpoles relatively higher corticosterone, with IOP being intermediate. Higher corticosterone levels were correlated to faster developmental time and shorter tails and bodies, hinting to a mediator effect of corticosterone. In chapter II, I investigated short-term behavioural and hormonal responses to acute acid and predator stress using one AOP and one NOP from the extreme ends of the acidification gradient. In this experiment, tadpoles from both populations reduced their activity in acidic pH, but under predator stress NOP tadpoles decreased their activity, whereas AOP tadpoles increased their activity. Moreover, AOP tadpoles increased their corticosterone levels under stress, whereas NOP tadpoles did not. In chapter III and IV, I studied tadpoles from two AOP and two NOP and reared them under a combination of chronic acid and/or predator stress. In chapter III, I studied the linkage between metabolic activity (oxygen consumption) and life history traits of tadpoles. AOP tadpoles had lower oxygen consumption and slower development overall and tadpoles reduced their oxygen consumption when exposed to predator cues. In chapter IV, I studied the integrative phenotype of tadpoles and its linkage to corticosterone (including corticosterone manipulation) and metamorphic fitness traits. The common garden experiment revealed that AOP tadpoles were larger and needed longer to reach metamorphosis, had deeper tail muscles and shorter tails, and exhibited mostly lower corticosterone levels. Acid stress led to higher corticosterone levels, and tadpoles with smaller and shallower bodies, deeper tail muscles, and slower development to metamorphosis. Tadpoles exposed to predator stress had deeper tail muscles and shorter tails, were bigger at metamorphosis, needed longer until reaching metamorphosis, and responded to freshly added predator cues by reduced behavioural activity. Moreover, AOP tadpoles reared under predator stress reduced their general behavioural activity, whereas NOP tadpoles did not. The causal role of corticosterone was supported by a corticosterone manipulation experiment: experimental increase in corticosterone levels increased tail muscle depth and body depth and resulted in smaller size of tadpoles. Taken together, I conclude that a combination of G (population), E (acid and predator stress), and GxE (population specific responses to acid and predator stress) effects are at play in the responses of R. arvalis to environmental acidification. Importantly, my research indicates that corticosterone is a potential mediator of the observed differences in morphology, life history and metabolic activity between populations. Together with previous work, my studies indicate that adaptive divergence of R. arvalis along the acidification gradient is most likely driven by selection via both predator and acidity stress on the integrative phenotype. My thesis sheds light on the composite phenotype of an amphibian and the value of individual level studies in unravelling eco-evolutionary processes in organismal stress responses