Mesoscale patterns in butterfly communities of the central Great Basin and their implications for conservation

Among the most pressing conservation challenges is conservation of landscapes where major threats to the ecosystem are known, but spatially explicit and species-specific data are few and managerial resources are limited. In such landscapes, researchers can assist managers by collecting baseline data on broad ecosystem patterns; examining whether these patterns are predictable; and determining how to implement regional inventories for selected taxa. This dissertation explored these issues using butterflies of the central Great Basin as a model system. Butterflies have been lauded as potential indicators of ecological phenomena of concern. This dissertation tested whether species richness and elevational ranges of montane butterflies respond predictably to elevational gradients, and whether the effects of climate change on butterfly distribution patterns in different mountain ranges are likely to vary. It also compared the efficacy of butterfly inventory strategies for insular montane landscapes that vary in spatial and temporal scale. Species richness and elevational ranges of montane butterflies in different mountain ranges do not have identical responses to elevational gradients. However, it may be possible to predict the response of butterflies to an elevational gradient if the vegetation and topography of each mountain ranges is considered. Elevational Rapoport effects appear to be widespread. In the Toiyabe Range, butterfly distributions showed a "proper" Rapoport effect (species at higher elevations had greater elevational ranges). Research in the Toquima Range, by contrast, documented a new phenomenon, a reverse Rapoport effect (negative correlation between elevation and elevational range). The effects of climate change are likely to vary among mountain ranges. Perturbation of butterfly distributions probably will depend on range topography as well as species-area relationships. This research suggested that extensive sampling over a broad area rather than intensive sampling in a small area may better assess butterfly species richness and distribution patterns at the regional level. Although long-term inventories ultimately will detect more species and are necessary to quantify temporal variation in species distributions, a rigorous single-year inventory nonetheless may provide a solid baseline assessment. Identification of broad ecosystem patterns is far from futile, and is invaluable for managing for ecological sustainability at the landscape scale.