Assistant Professor New Jersey Institute of Technology Jersey City, New Jersey
Unseen among 14,000 extant ant species are the products of a radiation of the earliest ant lineages. Stem ants – lineages diverging prior to the ancestor of modern ants – dominated the Cretaceous as an abundant and diverse group for over 20 million years, but vanished during the late Cretaceous while modern ants diversified into one of nature’s greatest success stories. What drove this dramatic extinction and turnover? Were stem ants more ecologically vulnerable to extinction pressures, or did modern ants outcompete them? To evaluate potential causes and mechanisms of this stark turnover, we developed a quantitative machine learning framework to predict the ecologies of extinct ants, using known ecomorphological correlates from extant ant species. We find that stem ants occupied a broader range of ecologies than previously hypothesized, comparable to what we see in many lineages today. To broaden our dataset, we reconstructed ancestral morphology and ecology of modern and stem lineages. With our comprehensive dataset incorporating extant, fossil, and reconstructed data, we tested extinction selectivity through linear models and changes in ecomorphospace. Results highlight ecological similarity between modern and extinct ants, suggesting that niche occupation itself was not the deciding factor in the extinction of the earliest ants, and competition may have played an additional role.