Does sampling method or microhabitat type determine patterns of macroinvertebrate assemblage structure detected across spatial scales in rivers?

Abstract

Patterns in macroinvertebrate assemblage structure detected across spatial scales in rivers vary among studies, and it is not clear whether methodological differences in sampling methods or differences between rivers are responsible. We aimed to separate the roles of microhabitats and sampling methods in determining patterns in macroinvertebrate assemblages across spatial scales in the Johanna River, south-eastern Australia. We hypothesised that less sample-to-sample variability in the structure of macroinvertebrate assemblages would be detected using a Sürber sampler than by scrubbing stones. However, Sürber samples were more variable than individual stones, although invertebrate abundance did not differ. To examine the effect of habitat and substratum, we hypothesised that taxa richness, invertebrate abundance and sample variability would differ among leaf packs and stones. Variation among samples in taxa richness and abundance was higher among leaf packs than among cobbles, whereas variability in taxon composition was much higher among stones. Leaf packs in runs had four times as many taxa as did stones in riffles, and more individuals. Leaf packs may therefore provide a more representative sample of the assemblage than do stones. Sürber samplers may randomly subsample riffle assemblages, inflating the sample-to-sample variability detected.