SELECTIVE EXTINCTION AND RANDOM ORIGINATION DROVE THE SHORT-LIVED NATURE OF LARGE AND MARINE CROCODYLOMORPH CLADES

Life history traits such as body size, habitat, and geographic range have been shown to affect the probabilities of species and genera going extinct and/or originating, although most of these predictors have dampened or entirely absent effects during periods of mass extinctions. While some work has been conducted on how vertebrate extinction and origination is influenced by such traits during mass extinction events, little work has been done on how this compares to background selectivity. Crocodylomorpha, despite its depauperate modern diversity, has a long and diverse fossil record which includes species of varying sizes and habitats, making it an ideal clade for such an analysis. To accomplish this, I supplemented a previously compiled dataset of crocodylomorph body size and habitat data and timescaled the most recent supertree of 375 crocodylomorph species. Using this phylogeny and data, I then performed cluster analyses and phylogenetic logistic regressions for each geological stage to test whether extinction and origination have been phylogenetically clustered and/or selective at any point over the last 230 million years. I find that extinction and origination are usually no more clustered than would be expected by chance, although notable exceptions to this are clustered extinctions at the end of the Jurassic and Cretaceous periods and clustered originations at the beginning of the Jurassic and Cretaceous periods. Several stages exhibit extinction selectivity against marine species, while several other stages show extinction selectivity against large body sizes. As with previous analyses, extinction selectivity appears neutral during periods of mass extinctions. Finally, origination exhibits little to no selectivity with regards to habitat or body size. Overall, these results demonstrate that the evolution of large sizes and/or marine lifestyles in crocodylomorphs often resulted in higher extinction risk; however, origination appears to have been nearly random. The combination of these two trends could help explain why large and/or marine clades were often short-lived. Beyond crocodylomorphs, these results indicate that investigations of longer time intervals may reveal variation in selectivity that may otherwise be missed when analyzing a mass extinction in isolation.