Spatial and temporal patterns of diatom assemblages, and their drivers, in four US streams: evidence from a long-term dataset

Bioassessment to evaluate stream integrity and determine changes related to point-source discharges is typically focused in wadeable streams, with limited understanding of seasonal and annual variation. We used a multi-year (n = 13), multi-site (n = 5–7), seasonally (spring and fall) sampled dataset to evaluate spatial and temporal patterns in diatom assemblages relative to measured environmental variables, land use, and pulp and paper mill discharges in a wadeable stream (Codorus Creek, PA) and three non-wadeable rivers (Leaf River, MS; McKenzie and Willamette rivers, OR). Analysis of variance (ANOVA) and permutational ANOVA (PERMANOVA) showed that significant spatial differences in commonly used diatom biotic integrity/diagnostic metrics and assemblage structure were common in the wadeable stream, but rare in the non-wadeable rivers. Season-specific diatom patterns were observed in all streams regardless of size, but annual variation was more prevalent in the non-wadeable rivers. Environmental variables explained 35–58% of the variability in diatoms in the spring and 33–50% in the fall, with physical habitat characteristics associated with stream morphology and seasonality more important than those associated with anthropogenic inputs such as land use and point sources. Findings from this study highlight the value of spatially and temporally comprehensive datasets in understanding and interpreting diatom assemblage patterns.