The Chironomidae community response to substrate changes in a lowland river (Balkan Peninsula)

Lowland rivers are exposed to severe anthropogenic degradation, which leads to the alteration of natural substrates. Such degradation in river morphology can significantly influence the community structure of aquatic macroinvertebrates by increasing the quantity of manmade substrates (e.g. riprap). Due to the preferences of specific taxa, a survey of the community found on artificial substrates can significantly bias the result of a community structure analysis, characteristic for a particular habitat. Chironomidae larvae are the dominant group among aquatic macroinvertebrates, and yet their sensitivity to substrate composition changes has only been considered in a few studies. In this work, we wanted to test whether the community structure in the Sava River (Balkan Peninsula) differed significantly on natural and artificial substrate. Also, the intention was to detect which taxa group contributed to the changes in the community structure. The distribution and separation of the sampling sites on a Non Metric Dimensional Scaling (NMDS) plot were mostly influenced by the substrate type. NMDS analysis presented two groups of sites assigned with specific community structure. The indicator value analysis (IndVal) revealed important chironomid representatives for both groups of sites (natural and artificial substrates). The significance of substrate specificity, as an important factor influencing the chironomid community structure, was confirmed by a distance-based permutational multivariate analysis of variance (PERMANOVA), as well as IndVal, since different indicator taxa inhabited different microhabitats in the same sampling sites. This information stresses the importance of the substrate type, pointing out that in determining the representative community for a certain habitat, artificial substrates should be avoided in sampling design. This is especially true for bioassessment and river typology methods where the community structure is used as a multivariate indicator.