Ecological Niche Modeling of Equids in the Great Plains during the Miocene

Geographic distributions of Miocene species in the subfamily Equinae are predicted from the Great Plains using an integrative approach, ecological niche modeling (ENM). The Equinae underwent a dramatic radiation as climate changed from warm and humid in the middle Miocene to cooler and more arid conditions during the late Miocene. Predicted distributions of individual species from ENM are analyzed in relation to this climate change. The ENM modeling system employed is GARP (Genetic Algorithm using Rule-set Prediction). This method predicts the geographic extent of a species' fundamental niche based on environmental variables coupled with known species occurrence points and provides a means to quantify a species' geographic range. ENM in the fossil record combines environmental data from a variety of sedimentological resources. In this study, data from stable carbon isotopes, paleosols, paleofloral analyses, and paleofaunal assemblages are utilized in combination to interpolate environmental coverages. Species ranges are predicted for two successive time slices during the Miocene that span from the Mid-Miocene Climatic Optimum into the early Pliocene (14.5-8.5 Ma and 8.5-2.0 Ma). During the initial cooling phase, species with larger predicted ranges were more likely to survive between the Barstovian and Clarendonian than species with smaller ranges. As climate continued to become more cool and arid in the late Miocene, range size became irrelevant to survival from one North America Land Mammal Age to the next and extinction rates increased. In addtion, patchy distributions are more common than continuous distributions in the middle Miocene when speciation rates are high. During the late Miocene, when speciation rates are lower, continuous ranges are more common. This is the first use of ENM and GARP in the continental fossil record. Its use here demonstrates the advantages of integrating paleontological, sedimentological, and geochemical data in a quantitative spatial analysis across multiple temporal intervals.