Invertebrate traits for the biomonitoring of large European rivers:: an initial assessment of alternative metrics

1. The application of environmental policy and legislation across large-scale administrative units creates a growing need for standard tools to assess and monitor the 'ecological health' of rivers, a requirement that can be achieved through the description of ecological functions of lotic invertebrate species in river communities. 2. To assess alternative metrics, we tested how the functional structure (described by 14 biological traits) of invertebrate communities in 190 large river reaches differed with respect to differences in taxonomic resolution (species, genus, family), taxa weighting of traits (raw abundance, ln-transformed abundance, presence-absence data) and consideration of alien species (inclusion or exclusion), and how these differences influenced the potential of functional descriptions to discriminate river reaches across a gradient of multiple human impacts. 3. Functional descriptions derived at the level of species, genera and families were very similar, whereas functional descriptions derived from raw abundances differed significantly from those derived from both ln-transformed abundances and presence-absence data. Functional descriptions after the exclusion of alien species differed considerably from those including alien species. 4. Generally, the functional descriptions significantly discriminated river reaches according to the level of human impact. Taxonomic resolution scarcely influenced the discrimination of impact levels, whereas the use of raw abundances decreased impact discrimination in comparison with ln-transformed abundances and presence-absence data. Exclusion of alien species also decreased discrimination of impact levels. 5. When considered separately, individual biological traits describing maximal size, number of descendants per reproductive cycle, number of reproductive cycles per individual, life duration of adults, reproductive method, parental care, body form and feeding habits had the highest potential to discriminate the level of human impact. 6. Our findings indicate that genus or perhaps family identifications are sufficient for large-river biomonitoring using invertebrate traits. Although raw abundances could provide a better discrimination of low levels of human impact, presence-absence data should be sufficient to discriminate functional community changes caused by elevated levels of human impact across Europe.