Sympatric rattlesnakes with contrasting mating systems show differences in seasonal patterns of plasma sex steroids

Abstract

Long-term field studies conducted in Arizona show that two species of sympatric rattlesnakes differ in the structure of their mating systems, primarily in frequency and timing of mating seasons, despite exposure to identical environmental conditions. The western diamond-backed rattlesnake, Crotalus atrox, has two distinct mating seasons within a single annual spermatogenic cycle. The first mating season occurs from late August to early October. Following a hibernation period of 4 months, the second mating season occurs from mid-March to early May. Because there is a mating season in spring, long-term sperm storage by females during winter is facultative. In contrast, the black-tailed rattlesnake, Crotalus molossus, has a single mating season (mid-July to early September) within a single annual spermatogenic cycle. Due to the absence of a mating season in spring, long-term sperm storage by females during winter is obligatory. In both species, ovulation and fertilization occur in spring, and offspring are produced from mid-July to early September. Based on these robust data, we tested the hypotheses that seasonal patterns of plasma sex steroids (testosterone, 5α-dihydrotestosterone, and 17β-estradiol) differ between males in wild populations of C. atrox and C. molossus, and that peak levels would be coincident with the mating seasons. Specifically, we predicted that there would be two peaks of sex steroids in C. atrox and one peak in C. molossus, and that baseline levels would be detected outside the periods of mating and spermatogenesis. Our results supported these predictions. Furthermore, absolute concentrations of plasma testosterone and 5α-dihydrotestosterone, but not 17β-estradiol, were higher in C. atrox than in C. molossus. We discuss a possible scenario for the evolution of the different mating seasons in these sympatric rattlesnakes, and advocate that comparative approaches to address such questions should integrate proximate and ultimate causation to increase explanatory power.