Predictive biological indices have transformed the bioassessment landscape by allowing universal indices to be applicable across diverse environments. The successful development of a predictive benthic macroinvertebrate index for California wadeable streams helped to demonstrate the power of these tools in complex geographic settings. However, previous efforts to develop predictive algal indices for California were limited by poor performance and were ultimately unsuccessful. For this study, we leveraged a robust statewide dataset to develop two different types of predictive algal indices for California wadeable streams: an index of observed-to-expected taxa (O/E) to measure taxonomic completeness and a multimetric index (MMI) to evaluate ecological structure. We developed multiple versions of each index, including one for diatoms, one for soft-bodied algae, and a hybrid index using both assemblages. We evaluated index performance using a series of screening criteria for precision, accuracy, responsiveness, and regional bias. We found that final index performance varied among all assemblages: the best performing O/E index was a diatom-only index, whereas the predictive diatom and hybrid MMIs out-performed all other indices with excellent responsiveness and precision. We found that in comparison to benthic macroinvertebrates, algal communities were characterized by high beta diversity across reference sites and low average species richness per site, resulting in disparate algal populations that were challenging to model with predictive approaches, particularly for soft-bodied algae assemblages. While all O/E indices were considered to have weak performance, the predictive diatom and hybrid MMIs are accurate, responsive, and precise indices that will provide a powerful assessment of biological condition for statewide applications.
