Abstract 1. Current views in ecology emphasise that community structure is the sum of multiple processes, with imprints of both regional and local drivers. However, the degree to which stream insect assemblages are structured by spatial configuration (complying with the dispersal-based neutral hypothesis) and local environmental features (complying with the niche-based species sorting hypothesis) has not been rigorously examined based on surveys in multiple years.
2. Stream sites in a boreal drainage system were surveyed during three consecutive years and the relative contribution of spatial configuration and local environmental variables to aquatic insect assemblage structure (characterised by both abundance and presence–absence data) was assessed. Separate analyses were conducted for mayflies (Ephemeroptera), stoneflies (Plecoptera), caddisflies (Trichoptera), and non-biting midges (Diptera: Chironomidae) in each year.
3. There were no relationships between the spatial location and local environmental features of streams in Mantel tests, facilitating exploration of their independent effects on assemblage structure. The study found virtually no effects of spatial location on stream insect assemblages across the study drainage system, as evidenced by Mantel tests and canonical correspondence analyses (CCA). The environmental variables were also rather weakly associated with assemblage structure, with the total amount of explained variation ranging from 9.8% to 31.7% in the CCAs. There were no appreciable differences in the amount of environment-related explained variation in assemblage structure between mayflies, stoneflies, caddisflies, and midges, but some between-year differences were noticeable in most insect groups. The environmental variables that were significantly related to assemblage structure exhibited some between-group and between-year variability. In general, patterns shown by abundance and presence–absence data were highly similar.
4. It appears that stream insect assemblages comply with the niche-based species sorting hypothesis in the context of metacommunity ecology. In contrast, the absence of spatial structuring suggests that stream insect assemblages do not comply with the neutral hypothesis, being not strongly dispersal limited at the within-drainage basin scale.