USING ISOTOPIC, ECOMORPHOMETRIC, AND BIOGEOGRAPHIC ANALYSES OF MODERN RODENTS AS ANALOGUES FOR FOSSIL RODENT PALEOECOLOGY

Rodents are diverse and abundant in the Cenozoic fossil record but have not been used in terrestrial paleoecology and paleoclimatology as often as large-bodied ungulates. We summarize three approaches we are applying to modern, North American rodents that inform uses of fossil rodent assemblages in paleoecology and paleoclimatology: stable isotope analysis, dental ecomorphometrics, and biogeography. We have developed a dataset of >500 δ¹³C values of hair from 14 common and widespread species across the southern Great Plains that represent a range of trophic categories. Species and trophic categories have statistically distinct δ¹³C values. Regionally, rodent diets are biased towards C₃ biomass, but the sampled species reflect the regional gradient in C₄ grass abundance. At the local scale, we have analyzed hair of >300 individuals of 14 species from Sevilleta LTER for which stomach content analyses have been published. Both C and N isotope ratios are distinct among species but vary by habitat within species and clearly reflect trophic level. The third dataset comprises δ¹³C values of different tissues for multiple individuals of 15 species and shows strong correlations δ¹³C values of different tissues, allowing for conversion of δ¹³C values of hair to corresponding enamel values with reasonable uncertainty. Rodent species occupy a wider range of trophic categories than do primarily herbivorous ungulates, so δ¹³C values for rodents provide a more limited view of diet. Using a dataset of µCT scanned mandibles of >120 species of North American rodents at resolutions of 5-31µm, we can use a small number of 3D shape metrics for whole tooth rows to predict diet category with >90% accuracy. These isotopic and ecomorphometric datasets, when combined with biome type for modern rodent assemblages, provide means to generate detailed dietary and environmental information for fossil rodent assemblages. Finally, we have developed a GIS method that uses the overlap in modern species ranges to estimate quantitatively climate parameters accurately for a locale based on the distribution of values in the area of sympatry even for small numbers of species. Together, these methods provide a basis for developing quantitative reconstructions of ecology and environment from fossil rodent assemblages throughout the Cenozoic.