Associations between small dams and mollusk assemblages in Alabama streams

Small dams are ubiquitous yet poorly understood features in many streams. Dam removal is being used increasingly in stream restoration projects as a means to enhance habitat connectivity and ecosystem function. However, habitatand assemblage-level effects of small dams on stream mollusk assemblages are poorly documented. We examined associations between stream physicochemical habitat variables and mollusk assemblages at 22 small (,10 m) dams in 3–5 order Alabama streams. We sampled 66 reaches (3 reaches/dam) associated with intact, breached, and relict small dams. For each dam, we designated 3 study reaches: 1) immediately downstream from the dam, 2) 500 to 5000 m downstream, and 3) 500 to 5000 m upstream from the impounded or formerly impounded zone. We used principal components analysis (PCA) to examine variation in physical-habitat conditions across all sites. Four principal components accounted for ,72% of the variation in physical-habitat conditions across sites. One PC score (PC1, corresponding to increased substrate size) was negatively associated with several mollusk metrics including total mussel abundance, taxon richness, catch-per-unit-effort (CPUE), and density. We observed few significant differences between simple habitat variables at sites upand downstream of dams. However, streams with intact dams had significantly higher mussel catch rates (CPUE) and taxon richness than did streams with breached or relict dams. We used forward-stepwise multiple regression to model the effects of habitat variables (as standardized PC scores) on mollusk assemblage metrics. PCs representing substrate composition were the strongest predictors of total mussel abundance and richness. Abundance of other mollusks including deposit-feeding snails, the exotic bivalve Corbicula fluminea, and fingernail clams was correlated with PC scores describing variability in substrate organic matter composition or stream gradient. We think these data indicate that some intact dams enhance mollusk habitat in downstream reaches. Streams with intact dams appear to be more geomorphically stable than streams with breached or relict dams and conditions in the mill reach may reflect preconstruction stream conditions. Breached dams warrant higher prioritization for removal than intact structures because habitat degradation may persist for decades and impede re-establishment of native mollusk populations.