Sexual selection and the evolution of sexual dimorphism

In this dissertation, I ask how sexual selection shapes the phenotypes and genetic architectures of sexually dimorphic traits.; I first examined the diversity and interactions of sexual selection mechanisms. Among these, male mate choice is especially poorly understood. By synthesizing relevant theory and data, I found that male choosiness is predicted and observed in a wide variety of taxa. I also investigated the co-occurrence of multiple sexual selection mechanisms in seven species of piophilid flies. Focusing on one species, the sexually dimorphic fly Prochyliza xanthostoma , I investigated how multiple sexual selection mechanisms interact. I found that males with relatively elongated heads were at a disadvantage in combat, but were attractive to females and transferred more sperm. Thus, contrary to long-held assumptions, multiple mechanisms of sexual selection may be in conflict. Moreover, P. xanthostoma females obtained direct fitness benefits by ingesting part of the ejaculate, which suggests that male head shape may be an honest signal of mate quality.; I used both theoretical and empirical approaches to investigate the evolutionary consequences of sexual selection. Existing theory predicts positive allometry in most sexually selected traits. However, I developed new theory (with Troy Day) which suggests that allometric patterns (and associated phenotypic parameters) are difficult to predict without detailed knowledge of fitness functions. I also developed theory on the evolution of genetic architecture in sexually selected traits, predicting that intra-locus sexual conflict may select for genomic imprinting, favouring silencing of trait alleles inherited from the weakly-selected or opposite-sex parent. I used P. xanthostoma to compare the phenotypic expression and genetic architecture of sexually dimorphic traits employed in sexual competition with other traits that are not. Male head shape exhibited strong condition dependence, potentially explaining the correlation between male head shape and ejaculate size. As predicted by theory, condition dependence increased with the degree of trait exaggeration. Results also suggested an inter-sexual genetic correlation for condition dependence. Trait inter-sexual genetic correlation decreased with degree of trait sexual dimorphism, as predicted by evolutionary models. However, two male sexual traits exhibited striking differences between maternal and paternal heritabilities. These patterns contradict ordinary Mendelian assumptions, but support the new imprinting theory.