Abstract
Traditionally, patterns of species diversity were investigated exclusively at the local community scale. However, it has since been recognized that processes at regional scales also influence community dynamics. The integration of processes at local and regional scales to better understand species diversity patterns at different scales is an emerging focus in community ecology, and is examined with metacommunity theory. Several paradigms utilized to examine metacommunity theory include the species-sorting, the mass-effect, the patch dynamic, and the neutral models. Research was conducted in several subalpine meadows in the El Dorado National Forest. The compositions of several insect communities were compared, by utilizing species abundances and richness, among microhabitats (i.e. local scale) and meadows (i.e. regional scale). Two types of data were collected including insect community composition (i.e. abundance and richness), and environmental variables (i.e. vegetation abundance and richness, and soil moisture and pH). Insect community composition was categorized into dispersal groups and functional groups, and vegetation community composition was categorized according to growth forms (i.e. graminoid and forb/herb). The entire insect metacommunity was differentially structured according to meadows and microhabitats. Insect communities responded to the soil moisture regimes and other environmental variables exhibited by the various microhabitats. These results suggest that the entire insect metacommunity may be structured according to the species-sorting model of metacommunity theory. Analyzing distinct dispersal and functional groups within the insect metacommunity suggested support for different metacommunity models. Herbivores, and both the good disperser and poor disperser groups structured according to the species-sorting model. The other functional groups (i.e. predators, parasitoids, pollinators, and omnivores) were structured by a combination of the species-sorting, mass-effect, and neutral metacommunity models.