Phenotypic Integration of Sexually Selected Traits in a Songbird

Abstract

Natural selection favors traits that fit not only the external environment, but also the internal environment of the organism. As a consequence, traits often show a pattern of correlation, or phenotypic integration. In this dissertation, I examined both the evolutionary processes and the physiological mechanisms that generate phenotypic integration. I studied a natural population of a songbird, the dark-eyed junco (Junco hyemalis), focusing on the male "mating phenotype," the suite of morphology, physiology, and behavior used to attract and compete for mates. In Chapter 1, I review literature suggesting that correlational selection, which occurs when traits interact in their effects on fitness, may have effects on the physiological mechanisms that underlie integrated suites of traits. In Chapter 2, I found that correlational sexual selection favored an association between body size and a white patch on the tail feathers ("tail white"), an ornament used both in courtship and male-male competition. I also found that body size and tail white were genetically correlated. These results suggest that correlational selection may maintain the integration of the two traits. In Chapters 3-5, I focus on the role of the steroid hormone testosterone in the mating phenotype. In Chapter 3, I measured natural variation in testosterone levels and found that more attractive males had higher androgen responsiveness. That is, males with more tail white produced more testosterone in response to an injection of GnRH, a hypothalamic hormone. This suggests that investment in mating behavior (which seems to be controlled by testosterone) may covary with attractiveness. Indeed, in Chapter 4, I found that androgen responsiveness naturally covaries with both mating and parental behavior. Males that produced more testosterone defended their territories more vigorously and fed their offspring less often. Finally, in Chapter 5, I examined how selection acts on androgen responsiveness, and found that males with very high or very low responsiveness were less likely to survive. Combined, these studies suggest that testosterone, on a physiological level, and correlational selection, on an evolutionary level, act as integrators of the male mating phenotype.