9652599 Cane As natural habitats become increasingly isolated and fragmented, species composition and abundance on these "islands' change. We presently lack the biological information needed to predict which species will go extinct or remain unaffected, and the influence that these changes have on other species and their interactions in an ecosystem. This project examines whether common features of life history and ecology of bee species can be used to predict their persistence or extinction in habitat fragments of different sizes and degrees of isolation. Most previous work in this area has been done with birds and mammals. Bees are crucial pollinators in many plant communities. We focus on the bees that visit creosote bush. No plant species attracts more bee species (120+) than creosote bush. This bee fauna exhibits a diversity of ecological and life history features. The plant dominates the southwestern deserts of the United States; its 20,000 year advance with desertification is clearly documented by the fossil record. These attributes together provide a model for other communities that are less diverse, less understood, or more difficult to study. Because of this shrub's broad distribution, the results of this habitat fragmentation study for a desert basin in southern Arizona can be applied to regional and subcontinental processes. Combined with its historic spread throughout much of the western United States, the creosote bush and its bee guild can be used as an exemplary model for ecosystem response to global change.
