Species traits for future biomonitoring across ecoregions: patterns along a human-impacted river

1. Current budgets for environmental management are high, tend to increase, and are used to support policy and legislation which is standardized for l arge geographic units. Therefore, the search for tools to monitor the effec ts of this investment is a major issue in applied ecology. Ideally, such a biomonitoring tool should: (1) be as general as possible with respect to it s geographic application; (2) be as specific as possible by separating diff erent types of human impact on a given ecosystem; (3) reliably indicate cha nges in human impact of a particular type; and (4) be derived from a sound theoretical concept in ecology. 2. We developed an approach to biomonitoring which matches these 'ideal' ch aracteristics by focusing on numerous, general biological species traits (e .g. size, number of descendants per reproductive cycle, parental care, mobi lity) and on the habitat templet concept, which relates trends in these gen eral species traits to disturbance patterns. Using the French Rhone River a nd benthic macroinvertebrates as an example, we have used the data to demon strate a general framework and the potential of our approach rather than to produce a ready-made tool. Our data covered a large river and its major tr ibutaries, which has a catchment that crosses ecoregions, and known gradien ts and discontinuities in human impact. 3. We applied multivariate analyses to evaluate how the distribution of spe cies traits in invertebrate communities could discriminate environmental di fferences along the Rhone in comparison to traditionally used approaches (e .g. community structure, based on species abundances, or ecological species traits, such as velocity preferences and pollution tolerance). Invertebrat e community structure expressed in terms either of the abundance or the tra its of species reliably indicated differences in overall human impact. The community structure based on biological traits was less confounded by natur al spatial gradients and reliably indicated human impact, while community s tructure based on ecological traits was the most confounded by natural spat ial gradients and was the poorest indicator of human impact. Community stru cture based on species abundances was an intermediate indicator of human im pact. 4. These results indicate that a revision of biomonitoring approaches which have been based on a single aspect of the biological responses may be warr anted. The biological traits of species could separate the different types of human impact. Therefore, the use of these traits in biomonitoring could improve existing multi-metric approaches. Future research has to show if th e general applicability of species traits allows the development of a uniqu e biomonitoring tool for running waters of the European Union, for running waters in temperate climates on several continents, for freshwater, marine and terrestrial systems, and/or for global biodiversity assessment.