Abstract
In a quantitative genetics framework, evolutionary change in quantitative phenotypes depends upon both selection, the change in the fitness weighted distribution of phenotypes within a generation, and on transmission across generations, which is shaped by the pattern of additive genetic variances and covariances among phenotypes (modeled as the G matrix). Due to their evolutionary importance there has been much work to estimate these parameters in natural populations and to understand what factors shape them. However, manipulative experiments to establish causality are relatively rare, many aspects of the environment that likely shape selection remain unexplored, and the effects of internal physiological mechanisms on G are largely unknown. In this dissertation, I investigate the role of testosterone in structuring G, and the role of parasites in structuring selection on their hosts using both experimental and comparative methods. In Chapter 1, I test the idea that hormones, through their pleiotropic effects on multiple phenotypes, are important in shaping patterns of additive genetic covariance. To this end, I performed a large-scale breeding study in brown anole lizards (Anolis sagrei), paired with the experimental manipulation of testosterone during development. I show that the hormone testosterone structures patterns of additive genetic covariance. Females given testosterone exhibit a G matrix that is statistically indistinguishable from that of control males and statistically distinct from that of control females. This demonstrates that the hormonal environment in which genes are expressed is important for shaping patterns of additive genetic variance and covariance, which themselves are important for the short term response to selection.
In Chapters 2-4, I turn my attention to the role of parasites in structuring variance in fitness and shaping patterns of natural selection in host populations. In Chapter 2, I perform a meta-analysis to determine how costly parasites are to their hosts in terms of survival, and to test whether the survival cost of parasitism is mediated by host mating system and sex. Across a phylogenetically and ecologically diverse set of hosts and parasites, I show that on average, parasitized hosts have 3.5 times greater odds of mortality than unparasitized hosts. By increasing the odds of mortality, parasites increase the opportunity for selection. Further, within promiscuous and polygynous species, males have a greater survival cost of parasitism than females, while in monogamous systems, females suffer greater costs than males. In Chapter 3, I test for costs of parasites in terms of growth, performance, survival and mating success in A. sagrei hosts. I experimentally removed the nematode parasites from A. sagrei using a custom, extended-release formulation of the antiparasite drug ivermectin that I developed. I demonstrate that parasites in this system impose costs to the growth, performance, and mating success of their hosts. In Chapter 4, I expand on the work from Chapter 3 and perform a larger experiment, again using wild A. sagrei, to investigate the impact of parasites on phenotypic selection. I show that parasites decrease survival for adults across the breeding season, thus increasing the variance in relative fitness. Additionally, parasites shape selection in juveniles by changing the correlation between some phenotypes and fitness. The results in adults are ambiguous but suggest that parasites may affect selection both by increasing the opportunity for selection and by altering the correlation between phenotype and fitness.
