Bioassessment is the science of using aquatic organisms as indicators of ecological condition in streams in rivers. Many types of organisms can be used as indicators, for example fish or algae, but bioassessment is most frequently based on benthic macroinvertebrates (BMIs), which are small but visible bottom-dwelling organisms such as insects. BMI data sets typically consist of long lists of species (or taxa) found in a sample and their relative abundances. These data can be simplified into measures of biological condition such as indices of biotic integrity (IBIs) that are designed to be sensitive to human-caused alterations to the landscape, to stream channels and riparian zones, and to water chemistry. IBIs function much like economic indicators: high IBI scores reflect good ecological conditions while low IBI scores reflect poor ecological conditions.
Bioassessment is increasingly used throughout California by water quality monitoring programs, but in the Central Valley bioassessment is more challenging than in other regions of the state because the entire landscape and most streams are highly altered by human activities such as urbanization, agriculture and water diversions. This makes it impossible to evaluate how BMIs respond across a complete gradient of human disturbance within the region, that is, from minimally disturbed reference sites where human activity is absent or minimal and which therefore set the benchmark for biological expectations, to the most altered sites with degraded biology. In the Central Valley, minimally disturbed reference sites are no longer available. Even the ‘least-disturbed’ sites, which represent the best-available chemical, physical and biological habitat conditions given the current state of the landscape, are markedly disturbed. Reference sites in other parts of California, such as the Sierra Nevada or the Sierra foothills, may be significantly less disturbed than Central Valley reference sites.
In this study, BMI data sets from 11 studies conducted at various intervals over the last 14 years were compiled to build an IBI for Central Valley streams. Data were not collected consistently by the different studies, and many gaps were present in associated physical habitat and water chemistry data sets. This could be corrected for BMI samples by standardizing to a consistent level of taxonomic effort. Gaps in other data could not be addressed. Criteria for defining ‘best-available’ reference sites were established as data allowed and were based on local urban and agricultural intensity, stream channel and riparian condition, and stream substrate composition. Eighty BMI metrics were evaluated for inclusion in the IBI based on 4 criteria: 1) sufficient range for scoring; 2) responsiveness to land use and/or local disturbance variables measured at the 150-meter sampling reach (as data allowed); 3) good discrimination between reference and test sites; 4) lack of correlation with other responsive metrics. Five final metrics were selected and scored for inclusion in the IBI: collector richness (number of taxa that are collector- feeders), predator richness (number of taxa that are predators), percent EPT taxa (percent of taxa that are mayflies, stoneflies, or caddisflies), percent clinger taxa (percent of taxa that cling to vegetation) and Shannon diversity (a composite measure of taxonomic richness and evenness of abundance). The final IBI showed good discrimination between reference and test sites, and was validated with an independent data set. BMI metrics and the final IBI were more strongly related to reach-scale physical habitat variables than to water chemistry or land use variables, but detailed water chemistry was lacking for many sites, and some studies have shown that response to land use diminishes when more than 10% of a watershed is degraded by human activities.
Despite data gaps that were less than ideal for indicator development, this study is the first to set expectations for Central Valley BMI assemblages based on best-available reference sites. The Central Valley IBI can be used as a general interpretive framework for benthic samples collected from perennial streams on the valley floor and provides an objective means for rating biological condition in a region with high urban and agricultural intensity. The ability to rank sampling sites relative to explicitly defined biological expectations is essential to any biological monitoring program. Therefore, this index may prove useful in several monitoring applications, including California’s non- point source CMAP program where sampling was stratified to assess and compare stream condition in urban, agricultural and forested watersheds, in stormwater monitoring programs, in point-source pollution investigations, and in stream restoration monitoring. Key recommendations include: 1) that all future bioassessment projects in the Central Valley should collect quantitative physical habitat and water chemistry with consistent protocols at all sites; in situ chemistry and rapid (qualitative) physical habitat are not sufficient for screening reference sites or evaluating BMI responses to stressor gradients; 2) that bioassessment should be added to NPS monitoring where programs are already collecting more intensive stressor data such as pesticides, nutrients and metals. This will provide California’s monitoring programs with better datasets to support future analyses.