Evaluation of suitability and comparability of stream assessment indices using macroinvertebrate data sets from the Northern Lakes and Forests Ecoregion

Researchers and managers within the Upper Midwest currently use a variety of sampling methodologies and biological indices to assess ecological condition of stream systems. With multiple entities collecting bioassessment data it is important that we determine the comparability of data and the indices derived from these data for effective assessment of natural systems. In this study we assessed the similarity of data collected by different agencies and we focused on data from one watershed to examine the outputs of different indices for stream assessment, and the temporal variation of index score within sites. We compared duplicate macroinvertebrate community data collected by the Little River Band of Ottawa Indians and the Michigan Department of Environmental Quality for overall community composition and index scores derived from these data. Duplicate samples were similar in composition index scores. Taxonomic resolution was addressed and indicated that genus level resolution gives a more favorable score when using indices. We also evaluated the utility of currently available macroinvertebrate indices of biotic integrity to assess data from the Big Manistee River watershed. The indices evaluated were the Hilsenhoff biotic index, the benthic community index for the Northern Lakes and Forests (NLFBCI), the Great Lakes Environmental Assessment Survey (GLEAS) procedure 51 for macroinvertebrates and a biological condition gradient model for the Upper Midwest. Outputs from the indices were moderately correlated (Spearman rank order correlation, r = 0.35–0.698) though they indicated different assessments of overall site integrity. Compared with larger scale regional indices, locally calibrated indices generally classified sites as having better biological condition. Replicate samples collected within sites indicated the GLEAS had higher levels of variability (0–265%CV) within sites than the other indices (<10%CV). Data from long-term (10 year) monitoring stations were used to evaluate seasonal and long-term index performance. There were differences in index score classifications from spring and fall samples indicating that standardization of sampling time is necessary for comparative analysis. Temporal trends over 10 years reveal natural variation and set the baseline for evaluating the influence of anthropogenic effects. Overall, results indicate that choice of index can alter assessment of site condition. For bioassessment in the Big Manistee River watershed the NLFBCI performs well and accurately reflects site condition.