The legacy of macroevolution: understanding and predicting evolutionary responses to novel environments

Abstract

Understanding how species will respond to global change is a major challenge for biologists. Two factors that play an important role are evolutionary constraints and contingencies. Evolutionary constraints are limitations on how species can adapt and change imposed by past evolutionary history, which could be driven by genetic, developmental, or functional variation. Evolutionary contingencies are chance events or evolutionary pathways that affect how species will respond to challenges and can lead to differences in species responses even when exposed to the same environmental pressures. Studying macroevolution of morphological traits and changes in microhabitat usage allows us to investigate the roles of these processes in generating variation in response to different habitats, and accounting for scale allows us to gain a more holistic picture of evolution. Understanding the role of evolutionary history is necessary to predict species responses to global change. In this work I focus on evolution within lizards as these species use a variety of different habitats and the relationship between morphology and microhabitat use has been studied extensively. We investigated the evolution of hindlimb allometry across lizards and identified that the majority of variation in lizards can be explained under a Brownian motion model of evolution, although we did identify contingent evolution in skinks. We then investigated the evolution of hindlimb length and adhesive toepad size in two "model clades" for studying urban tolerance and invasion biology. We identified evidence for evolutionary contingencies in the two clades with opposite patterns of trait evolution for each clade. We then used phylogenetic information, species traits, climate, and geographic data to predict species invasion probabilities across 486 lizard species. We found that incorporating phylogenetic information allowed us to ac- count for complex or difficult to measure traits and improved our models performance. We then developed a feature selection procedure to compare species traits with randomly simulated phylogenetic traits in order to identify if these traits contained additional predictive power beyond phylogenetic information. Taken together these results highlight the importance of evolutionary constraint and contingency in the predictability of evolution and the utility of phylogenetic information for predicting future responses to environmental change.