RECONSTRUCTING LATE CENOZOIC PALEOENVIRONMENTS IN THE GREAT PLAINS USING OVERLAPPING GEOGRAPHIC RANGES OF MODERN SPECIES

The geographic distribution of extant mammalian species is determined, at least in part, by the range of environmental conditions that can be tolerated by constituent populations, which is a measure of the ecological niche of the species. Assuming niche conservatism, extinct populations of a species should have had the same environmental tolerance as extant populations, thus past biogeographic distributions for extant species might be used to estimate paleoclimate quantitatively. The Neogene and Quaternary fossil records of the Great Plains provide an excellent test case for this approach as numerous extant species of small mammals have first appearances over the last several million years, allowing past co-occurrence of species to constrain paleoclimate even further based on environmental conditions where species co-occur today. We have developed a GIS database of modern mammalian species ranges in the Great Plains and environmental parameters that may influence each species niche, including mean annual, minimum, and maximum temperature, mean annual precipitation, elevation, soil type, and vegetation type. Geographic ranges for 18 modern species of rodents, lagomorphs, and soricomorphs with good fossil records were used to identify the range of climatic conditions where species ranges overlap. From the Paleobiology Database, we identified 19 fossil collections in the region that contained at least three of the 18 modern species, allowing us to constrain paleoenvironmental conditions for those localities based on the environmental tolerances of those species where they co-occur today. In this initial analysis, we grouped the localities into two time intervals: <0.068 Ma and 0.3-0.4 Ma. Based on the modern tolerances of co-occurring extant species, mean annual temperature increased from 7.8±1.8° C for the older localities to 12.9±3.3° C for the younger localities, which is close to the mean annual temperature in eastern Kansas today. The increase might reflect better species sampling in warmer Holocene localities or a geographic bias in the fossil localities. Estimated mean annual precipitation shows no trend with an average value of 542±114 mm, close to the modern average for central Kansas. This method shows promise and analyses including more taxa likely will yield better-constrained estimates.