Interspecific associations of rangeland grasshoppers studied in Gallatin Valley, Mont., during 1988-1990 when regional densities were low, showed a great deal of spatial variation. Although >40 rangeland grasshopper species were collected valley-wide throughout this study, mean species richness at the habitat type level ranged from almost-equal-to 10-17; results suggest that species replacement occurred over the gradient of habitat types sampled. Based on the results of this and previous research, it appears that rangeland grasshopper communities consist of facultative assemblages of species. At the valley level, it was possible to distinguish between species groups in terms of their general distribution characteristics (i.e., broadly, intermediately, or narrowly distributed). Although there appeared to be a maintenance in the proportion of the species in each of the three distribution groups among the six habitat types studied, proportions differed from those found at the valley level. Similar to other studies in plant and insect ecology, a positive relationship between species abundance and the number of sites collected was found at both the habitat type and valley levels. Because of the differences found in species composition at the habitat type level, the generalization of detailed experimental studies, beyond the habitat type level, must be done with caution. Lastly, it is clear that to understand rangeland grasshopper communities in general, large-scale observational studies designed to depict pattern must be conducted together with small-scale studies designed to elucidate process.
