The ecosystems of the Earth are under siege from a variety of human impacts. Because as impacts accumulate, ecosystems can suddenly cross thresholds that may significantly change them, it is imperative to develop ways to recognize when such thresholds are approached. One early warning sign of approaching thresholds is changes in the number of species -- that is, species richness -- that are interacting in the ecosystem. Therefore, this research focuses on establishing the natural baseline for species richness in the American West, so that it becomes possible to use observed deviations from the baseline to recognize when ecosystems are in peril. To establish the natural baseline, it is necessary to take into account how species richness patterns vary over times much longer than human life spans. This project uses the exceptionally good fossil record of mammals to do this. Information on fossil and modern mammal occurrences will be aggregated into databases that will be made available to the public, and which allow sorting of the mammal species into the geographic area and the time interval in which they lived. The species richness per geographic area per interval will then be statistically analyzed and compared. In this way, modern species richness values can be interpreted in the context of this empirically determined, long-term baseline, with values falling outside the natural range of variability being cause for concern. The work also will train postdoctoral and graduate students in global change issues.