Stable patterns in species distributions in metal-contaminated environments

International audience ; A large database of river diatoms (comprising more than 580 taxa) was constituted based on field surveys carried out in 6 different countries (France, Spain, Switzerland, Canada, Vietnam, China), in rivers exposed to various loads of heavy metals in the water. After taxonomy harmonization, the patterns in diatom community structure were investigated for 187 samples, all collected from hard substrates in rivers of circumneutral waters. As the sites were contaminated by a mixture of different metals (mainly Al, As, Cd, Cr, Cu, Fe, Hg, Ni, Pb, Zn) with various loads, metal concentrations were converted into a single score after Clements et al. (2000) in order to determine 4 categories of metal inputs: background, low, moderate and high. The biotypology (i.e. structuration of the diatom dataset) indicates that the species are influenced by the biogeographical context as well as metal inputs. The most structuring environmental parameters are investigated, and discriminating analyses are used to determine the relevance of some particular species (e.g. Eolimna minima, Nitzschia palea, Surirella angusta) as well as teratological forms for the biomonitoring of heavy metal pollutions. Special attention is also given to the information brought by other traits often cited as reliable for metal assessment, i.e. cell size distribution and diatom growth forms and postures.