Predicting Stream Fish Communities Across Multiple Drainage Basins of the Great Plains

To evaluate sustainability of fisheries resources, managers must understand how landscape features influence fish communities. Moreover, understanding the generality of such landscape – fish community linkages is important if predictive models are to be extrapolated to other drainage basins. We used constrained ordination to assess the relative importance of stream size, catchment-scale, and local-scale variables in explaining stream fish community structure at two spatial extents (basin and sub-basin), among 3 basins, and among 13 sub-basins in Kansas, USA. Across all (sub-) basins, stream size was consistently the most important predictor of community structure, while catchment- and local-scale variables were of less importance in explaining community structure. The importance of stream size varied the most among basins, while the importance of catchment- and local-scale variables was more consistent among basins. Among-basin differences in the strengths of interactions among environmental features operating at different spatial scales may account for among-basin differences in variable importance. These results indicate that the generality of stream size and catchment-scale variables as predictors of stream fish community structure is relatively low, but the generality of local-scale variables is relatively high. When quantifying the landscape effects, fisheries managers should consider building predictive models at the basin or sub-basin extent.