Data composition and taxonomic resolution in macroinvertebrate stream typology

In the EU water framework directive (WFD) a typological framework is defined for assessing the ecological quality of water bodies in the future. The aim of this study was to test the effect of data composition and taxonomic resolution on this typology. The EU research projects AQEM and STAR provided 1660 samples of 48 stream types sampled all over the major geographical gradients in Europe. These stream types fit the WFD typological demands and fit to the major European geographic regions (ecoregions). The samples included gradients from reference conditions to samples with bad ecological quality. Despite standardisation, there were large differences between the participating countries concerning the number of taxa, the number of specimens, and the taxonomic resolution. The macroinvertebrate data were analysed by using detrended correspondence analysis (DCA). The distribution patterns using all samples, only reference samples, and only degraded samples showed that the use of species-level (or 'best available taxonomic' level) performed better at a practical (fine) scale in comparison to family-level. The analyses further showed that even the use of a standardised protocol can not easily overcome (i) differences in site conditions that force the researcher to deviate from the protocol as well as (ii) the experiences of the researcher(s) and (iii) the available taxonomic knowledge.