Data from: Climate-driven hydrological variability determines inter-annual changes in stream invertebrate community assembly

Although flow regime is one of the major drivers of riverine communities, not much is known about how inter-annual variability and extremes of flow influence community assembly mechanisms. We used data on benthic macroinvertebrates and modelled flow regimes in 23 near-pristine boreal streams to assess how community assembly mechanisms and species occupancy varied in response to inter-annual variability in flow conditions across 11 successive years encompassing extreme (both low and high) flow events. A null model approach was used to test whether deterministic or stochastic processes dominated community assembly and how much regional (among-stream) flow variability contributed to community variability (β-diversity). Mean daily flow and the greatest rate of flow rise were the strongest flow-related descriptors of invertebrate community composition. Communities were differentially assembled depending on the direction of change in flow magnitude: in high-flow years, communities were more similar than expected by chance, while at low flows they tended to be more dissimilar than expected. Beta-diversity of macroinvertebrate communities was related to among-stream flow variability only at high flows. Common species correlated strongly with flow variability and contributed most to variation in β-diversity, suggesting that changes in assembly mechanisms are mainly driven by common species. While homogenization of communities in high-flow years reflected increased species occupancies and environmental sorting, increased turnover during low flows likely resulted from stochastic extinctions and dispersal limitation. Our findings suggest that extreme hydrological events exert a strong control over stream invertebrate community assembly, and their effect may be even more profound in the future as high and low-flow spells are expected to occur more frequently, not allowing time for communities to recover