| Summary: | Abstract 1. Studies on biodiversity patterns should optimally relate different scales of temporal community variability to spatial variability. Although temporal biodiversity variability is often negligible compared to spatial variation, it may still constitute a substantial source of overall community variability in stream ecosystems. Boreal streams exhibit seasonally recurring environmental periodicity, which can be expected to induce synchronous dynamics of abiotic variables among sites, and consequently, to produce spatial synchrony of deterministically controlled biological communities with higher intra- than inter-annual community variability. 2. We sampled benthic macroinvertebrates in 10 near-pristine boreal streams on three different seasons (spring, summer, autumn) across 4 consecutive years in northern Finland. We aimed to identify the relative contributions of spatial, inter-annual, and seasonal variability to overall benthic biodiversity; and relate variation in benthic invertebrate communities to key environmental factors, particularly in-stream habitat diversity. 3. Among-site spatial variability was clearly the most important source of variation for both species richness and community dissimilarity. Of the two temporal scales, inter-annual variability contributed more to variation in taxonomic richness and seasonal variability slightly more to variability in community composition. 4. Only inter-annual variation differed systematically from random expectation, indicating strong stability (low variability) of stream macroinvertebrate communities across years, with less variation at sites with higher substrate heterogeneity. Considering the distinct seasonality of the boreal stream environment, seasonal variability accounted for an unexpectedly low amount of total community variability. 5. Although differences between seasons were small, autumnal sampling is likely to be the least susceptible to climatic vagaries, thus providing the most consistent and predictable conditions for benthic sampling in ...
|
|---|