Using Multiple Biological and Habitat Condition Indices for Bioassessment of California Streams

A long-term goal of the SWAMP bioassessment program has been to use multiple indices of ecological condition in conjunction to produce more complete assessments of stream health than provided by any single index alone. Ideally, combined assessments should be based on different taxonomic assemblages, taking advantage of their different responses to various stressors deriving from upstream land use practices. In this study, the combined use of three ecological indices currently used to assess stream condition in California was explored. The indices used were the California Stream Condition Index (CSCI) based on benthic macroinvertebrates, the “H20” index based on diatoms and soft algae, and the California Rapid Assessment Method (CRAM) for riparian habitat condition. First, comparisons of index performance were used to assess whether cases of disagreement among indices indicate moderate levels of stress, to which some taxonomic assemblages (or physical habitat indicator in the case of CRAM) have responded but not others, or whether disagreement among indices was more likely to be “noise” due to poor performance in one or more index. H20 and CRAM did not perform as well as CSCI on a statewide scale for some performance measures, but often performed better than null CSCI, which was used as a benchmark for what would constitute poor index performance. Second, the frequency with which the three indices agreed and disagreed about site condition was assessed to identify whether cases of agreement and disagreement occur in systematic and predictable ways that relate to the particular stressor(s) affecting the site. Patterns of agreement and disagreement among the 3 indices were non-random: the indices frequently agreed that reference sites were not degraded, and that high – activity sites were degraded. Disagreements were most common at sites with moderate amounts of human activity. Where the indices disagreed, CSCI and CRAM were more sensitive to physical habitat stressors, whereas H20 was more sensitive to chemical stressors. The use of multiple indices in conjunction to infer the ecological condition of streams, each based on a different taxonomic assemblage or data type, greatly strengthens confidence in results from bioassessment surveys, reduces the likelihood of incorrect conclusions from sampling error or natural variability, and improves our ability to diagnose causes of degradation.