Linking forest harvest and landscape factors to benthic macroinvertebrate communities in the interior of British Columbia

Detecting the magnitude of human-induced disturbance events, such as forest harvest, on biological communities is often confounded by other environmental gradients and scales at which these effects are examined. In this study, benthic invertebrates were collected from 43 streams across four basins and two geographic regions to (1) determine whether invertebrate abundance and community structure are best explained by historic forest harvest, landscape variables or a combination of both, and (2) evaluate associations among harvest, landscape variables, in-stream physical habitat, and invertebrates. Nonmetric multidimensional scaling showed that invertebrate community structure was primarily explained by watershed area and elevation, and basin and region but not by measures of forest harvest. Model selection using an information-theoretic approach and Akaike's information criterion indicated that watershed area was the most important variable explaining clinger and long-lived taxa richness, while basin was the most important variable explaining total abundance, and total, Ephemeroptera, Plecoptera, and Trichoptera taxa richness. Forest harvest ranked lower than landscape variables in relative importance in all models. These results suggest that landscape characteristics were relatively more important in predicting invertebrate community structure than forest harvest, and should therefore be considered when assessing the impacts of both reach and watershed scale forest harvest on benthic communities. Perhaps, the levels of forest harvest examined in this study had only marginal effects on benthic invertebrates because these ecosystems are naturally resilient as a result of frequent disturbance from forest fires.