Stressor-response relationships between landscape features and benthic macroinvertebrate indices using the USEPA Wadeable Stream Assessment program data

In 2004 the USEPA conducted the National Wadeable Stream Assessment (WSA) in order to measure and report the state of rivers and streams in the U.S. (USEPA, 2004). A total of 500 sites were surveyed for the nationwide study. Results from the study were aggregated into nine reporting ecoregions in order to provide statistical robustness in comparing sample sites with reference conditions. In the Interior Highlands of Arkansas, Missouri, and Oklahoma, 32 sites were sampled for the WSA project. The purpose of our study was to investigate the utility of this limited data set in determining stressor-response relations in an ecoregion smaller than one of the nine regions designed by the USEPA. A hierarchical approach using multivariate, step-wise, and correlation analyses was developed in order to examine the scalar relations of catchment and reach scale influences on stream conditions, and to evaluate macroinvertebrate assemblages for detection of impact. The strongest model relations (P=0.0001) were the combined effect of land cover, riparian conditions, and geomorphology on percent Wood (R2=0.76), Mean Substrate Diameter (R 2=0.72) and percent Coarse Gravel (R2=0.71). Model predictions of land cover impacts were found to significantly (P < 0.003) explain the variation in response of Thalweg Depth (R2=0.33) and Incision Height (R2=0.30). When coupled with riparian conditions, these relations strengthened with R2=0.58 (P=0.0001) and R 2=42 (P=0.001), respectively. Riparian cover showed a significant relation (P=0.0001) to Bankfull Width (R2=0.51). Direct relations were found between land cover and two macroinvertebrate metrics, Burrower Taxa (R2=0.33 (P=0.0009)) and Hilsenhoff Index (R2=0.22 (P=0.029)). However, the combined with reach scale variables the explanation in variance increased to R2=0.47 (P=0.009) and R2=0.35 (P=0.037), respectively. These results indicate that WSA data set can be used to characterize stressor-response relations among landscape features and macroinvertebrates in the Interior Highlands.