Environmental variables interact across spatial scales to structure trichopteran assemblages in Ouachita Mountain rivers

An important goal in aquatic ecology is to determine the interacting variables that regulate community structure; however, complex biotic and abiotic interactions coupled with the significance of scale have confounded the interpretation of community data. We evaluated stream and riparian habitat features in southeastern Oklahoma, USA at a range of spatial scales from local, in-stream variables to large-scale, regional characteristics to address the following questions: (1) How much variation in trichopteran community composition can be attributed to local, regional, and spatial variables? and (2) What environmental variables are most important in determining trichopteran community structure? We collected data on caddisfly community structure, local and regional environmental variables, and spatial location on the landscape from 25 sites in four rivers. We analyzed these data using canonical correspondence analysis (CCA) and variation partitioning. Our analysis explained approximately 60% of the variation in caddisfly community composition. We found that local and regional environmental variables were near equal in importance in governing caddisfly communities, with each accounting for approximately a quarter of the explained variation. Although pure spatial variables were less important, the amount of variation shared among spatial variables and local and regional variables was substantial, indicating that biogeographic history is also key to understanding caddisfly distributions. We also found a strong influence of human landuse (i.e., percent of land in agriculture, distance to roads) on caddisfly community composition. Our study indicated that communities are influenced by factors across scales, and that bioassessments should focus on not only local habitat conditions, but also incorporate larger-scale factors.