Background/Question/Methods The study of species rarity has long been considered in the ecological literature and is becoming more important as alarming changes in climate and human land uses have resulted in losses of individual species and an overall reduction in biodiversity.  To combat increasing extinction rates, active management of rare, threatened, and endangered species has become the norm, but often biological and ecological knowledge of specific species is unavailable to inform this management.  Characterization and classification of the communities that support rare species may be used to direct management and restoration.  Using a plot-based approach, we intensively studied 3 rare species that occur in the sandhills of North Carolina:  Astragalus michauxii, Pyxidanthera brevifolia, and Stylisma pickeringii.  These species are endemic to frequently burned, upland, longleaf pine-wiregrass savannas, but the 3 species rarely co-occur.  Consequently, we expected that these 3 rare species would all be associated with a consistent suite of more common species.  However, we also expected that relative abundances of the common species and fine-scale environmental factors would differ among the areas supporting the 3 rare species.  We sampled 10 x 10 m plots in at least 20 subpopulations of each species, with plots placed in order to maximize the occurrence of the 3 rare species.  We surveyed each plot for presence and percentage cover of vascular flora.  Interpretation of vegetation patterns was based on multivariate analysis of vegetation and associated environmental data. 

Results/Conclusions Preliminary results indicate that while there are similarities in presence and abundance of associated species but, there are also significant differences.  We also found that all 3 species occur in a single International Vegetation Classification association, the Pinus palustris/Quercus laevis-Quercus (incana/margarettiae)/Gaylussacia dumosa/Aristida stricta Woodland (CEGL003591) in the Pinus palustris/Quercus spp. Woodland Alliance (A.520).  Given the differences in species composition that we observe, it is possible that new community types might be warranted.  Understanding the community context in which rare species occur is critical if we are to be successful at conservation, management, and restoration of natural systems.