Ecological context and the evolution of mating biases in a freshwater amphipod species complex.

Mating traits are among the most extravagant traits found in nature. As with any trait, understanding the diversification of mating traits requires information on both direct and indirect selection acting on the traits and quantification of the genetic variation available for selection. This dissertation is an attempt to elucidate the possible direct and indirect selection pressures acting on mating traits in the Hyalella azteca species complex.This dissertation raises many questions. (1) What role has sexual conflict played in the evolution of mating traits? In the small species, single females are more susceptible to fish predation while paired than when single. This is not the case for the large species resulting in asymmetries between the two species in the costliness of pairing (Cothran 2004). In chapter 1, I found that the percent increase in pairing duration when females were unable to resist male guarding attempts was greater in the small species than the large species. This result suggests that the degree to which male and female interests differ over guarding duration is greater in the small species. Previous work has shown that paired females are at greater risk of predation in small species populations where they co-occur with positive, size-selective predatory fish. Thus, I argue that sexual conflict may be more important in the small species where the costs to females of early pairing are greatest, thus resulting in greater intersexual asymmetries in optimal pairing durations. (2) What does the genetic architecture of body size and posterior gnathopod size look like, and does heritability of these traits vary across environments? The posterior gnathopod is an anterior appendage that is greatly enlarged in males. In nature, males with larger posterior gnathopods have higher pairing success, which, along with the strong sexual dimorphism, suggests this trait has been shaped by sexual selection. Theory predicts that sexually selected traits will often show heightened condition dependence. An individual's condition has been defined as the total pool of resources acquired for allocation to various traits that impact fitness. This recent emphasis on condition dependency of sexually selected traits highlights the importance of understanding the resources necessary to build sexual traits and how such resources vary over space and time. Specifically, under benign conditions genetic variation associated with mating traits may be masked. Understanding how traits, and their additive genetic variation, vary across environments may shed light on the large disparity among species in within-population trait variation. (3) What are the female fitness consequences of mating with large males in the small species? In chapter 3, I found that in the large species, mating with large males increases female fitness both directly, through decreased predation risk while paired, and indirectly, through the superior mating success of sons. At least the direct benefit is unlikely to manifest in small species populations where fish predators select for small size prey. Therefore, it is possible that the mechanistic basis of size-biased pairing patterns differs between species, potentially explaining species differences in the strength of sexual selection on male traits and sexual size dimorphism. (Abstract shortened by UMI.