Environmental predictability of taxonomic and functional community composition in high‐latitude streams

Abstract

Summary In stream ecology, it has traditionally been proposed that local habitat conditions pose a strong environmental filter resulting in predictable community types. However, recent studies have demonstrated continuous rather than discrete community variation along environmental gradients across streams. Our aim was to examine the continuity of stream communities using functional traits of benthic macroinvertebrates. Using previous studies and the environmental filtering paradigm as a background, we hypothesised that the taxonomic composition of communities should vary continuously over environmental gradients, whereas the functional composition of communities should form discrete community types according to specific habitat conditions. As stream environmental characteristics vary between seasons, we also asked if this had an effect on the characteristics of community variation in time. Our data were collected from 50 streams draining into the River Tenojoki (70°N, 27°E) in Finland and Norway. The first data set was collected during late spring 2013 and the second data set during early autumn of the same year. Using multivariate regression trees and constrained ordination methods, we analysed the relationships of taxonomic and functional composition of macroinvertebrate communities to various in‐stream environmental factors. Our results suggest that the variation in the importance of environmental factors between seasons combined with the high heterogeneity of habitats precludes the formation of clear taxonomic and functional community types in northern streams. Conservation efforts should hence focus on including variability in environmental conditions so as to encompass all components of biological communities. However, as there were some noticeable responses of individual traits to different environmental factors, discrete functional communities may be found if key traits of the organisms can be identified.