VARIABILITY IN MACROINVERTEBRATE RAPID-BIOASSESSMENT SURVEYS AND HABITAT ASSESSMENTS IN A NORTHERN CALIFORNIA STREAM

Variability of the United States Environmental Protection Agency's Rapid-Bioassessment Protocol III was examined by comparing the results of benthic macroinvertebrate surveys and habitat assessments conducted by 14 student groups on each of three distinct 2nd-order reaches of Strawberry Creek, on the Berkeley campus of the University of California: a reach that was recently restored to improve water quality and aquatic habitat; an unrestored tributary, which receives occasional inputs of organic pollution; and an unmodified reference site upstream of the restored site. Impact was assessed by comparison with the reference site using three different impairment thresholds: 83% similarity; 65% similarity; and the interquartile range of the reference-site data. A randomized block analysis of variance showed that taxa richness, the Family Biotic Index (FBI), % Ephemeroptera-Plecoptera-Trichoptera (EPT) abundance, EPT/Chironomidae abundance ratio, EPT/[Chironomidae+EPT], % contribution of dominant taxon, and % shredder abundance discriminated among the restored, unrestored, and reference sites. The metrics of % scraper abundance and ratio of scraper/filterer abundance showed no significant site differences. All 14 sets of results identified the unrestored site as impaired; however, characterization of the restored site ranged from no-impact to various degrees of impact depending on the impairment threshold used. Results of the habitat assessments followed neither predicted site rankings nor benthic survey results. Although the observed variability could have resulted from errors in any of several steps in the protocol (e.g., identification, calculations, analysis), likely influences were the selection of collection point and non-random sub-sampling of macroinvertebrates.