Abstract

This study assesses the impact of errors in sorting and identifying macroinvertebrate samples collected and analysed using different protocols (e.g. STAR-AQEM, RIVPACS). The study is based on the auditing scheme implemented in the EU-funded project STAR and presents the first attempt at analysing the audit data. Data from 10 participating countries are analysed with regard to the impact of sorting and identification errors. These differences are measured in the form of gains and losses at each level of audit for 120 samples. Based on gains and losses to the primary results, qualitative binary taxa lists were deducted for each level of audit for a subset of 72 data sets. Between these taxa lists the taxonomic similarity and the impact of differences on selected metrics common to stream assessment were analysed. The results of our study indicate that in all methods used, a considerable amount of sorting and identification error could be detected. This total impact is reflected in most functional metrics. In some metrics indicative of taxonomic richness, the total impact of differences is not directly reflected in differences in metric scores. The results stress the importance of implementing quality control mechanisms in macroinvertebrate assessment schemes.