So Sexy It Hurts: Evolution of Sexually Selected Traits at Broad Phylogenetic Scales

Abstract

Sexual selection has long been regarded as a key evolutionary force. However, in spite of the attention sexual selection has received, there remain many unanswered questions in the field (Wiens and Tuschhoff 2020). Why do many animals lack sexually selected traits, such as weapons or signals? Conversely, why do some clades have so many different traits? At what rates do sexually selected traits evolve, and are traits gained at greater rates than they are lost? Are sexually selected traits positively correlated, such that species “pile on” more and more traits, or do tradeoffs and trait costs result in negative correlations between different traits? How long do sexually selected traits last? The goal of this dissertation is to answer these questions using phylogenetic methods at broad scales. In Appendix A, we examine sexually selected traits across all animals. We find that most animals lack precopulatory sexually selected traits, but that there are hotspots of sexually selected trait evolution, particularly in arthropods and chordates. We find no evidence that the presence of lifestyles thought to reduce the need for sexually selected traits are related to the proportion of sexually selected traits in phyla. We also find little evidence that sexually selected traits impact diversification at large scales. We find a low rate of trait gain and high rate of loss regardless of the phylum, sensory modality, or selection mechanism involved. Finally, we find that traits of different sensory modalities are positively correlated with each other, and that similarly traits of different selection mechanisms are positively correlated with each other. In Appendix B, we focus on visual traits within tetrapods. Similar to our results across all animals, we find that most tetrapods lack visual sexually selected traits. However, among species with visual traits, multiple traits are common. Different types of visual traits, namely behavior, color, enhanced morphology, and specialized morphology, show neither clumped nor random phylogenetic distributions. Visual traits are lost much faster than they are gained across tetrapods, regardless of the type of trait. In most cases, rates are similar across different types of traits within tetrapod groups and across tetrapod groups within trait categories. Finally, types of visual traits are positively correlated with each other. In Appendix C, we focus on how long visual traits last at a macroevolutionary scale across tetrapods. We specifically hypothesized that traits that could be more easily obscured from predator’s view, either because they are hidden or temporary, would last longer than exposed traits. We find that even the oldest visual traits are much younger than the tetrapod groups in which they occur. We also find that hidden traits are significantly older than nonhidden traits, and that temporary traits are significantly older than permanent traits. These findings support the idea that reducing attention from predators may reduce trait loss, and that ecological factors such as predation may influence evolution at the scale of millions of years. Overall, the various parts of the dissertation add to our understanding of sexual selection by helping to address unanswered questions in the field.