Use of multivariate analyses to investigate the contribution of metal pollution to diatom species composition: search for the most appropriate cases and explanatory variables

Multivariate techniques (ter Braak and Verdonschot, 1995) allow the elucidation of ecological factors, which explain most of the variation in diatom distribution. In particular, variance partitioning (Borcard et al. 1992) may help to separate different sets of environmental factors influencing diatom distribution. In this paper we use multivariate techniques and variance partitioning to elucidate the contribution of metal pollution to the structure of the diatom community. The main aim of this study was to provide analytical and statistical tools to determine the influence of metal pollution on the ecological status of fluvial systems. In order to achieve this main objective the following specific objectives were planned: first to describe the gradient of pollution and the corresponding diatom community at a regional level ; second: to obtain a good characterization of metal loads and metal availability in areas chronically exposed to low metal pollution and third: to investigate the relative contribution of different types of pollution: eutrophication and organic pollution vs. metal pollution to diatom species composition using variance partitioning multivariate analyses. In a first step, multivariate analyses were performed at a regional scale, using data provided by the Catalan Water Agency. The data set aimed to cover a large environmental gradient. Available date included a characterization of water mineralization, nutrient content, water metal concentration and inventories of diatom species from different watersheds influenced mainly by urban, agricultural and industrial activities. In a second step and due to the limited data available concerning metal pollution, a more detailed sampling was planned in the Fluvià watershed slightly impacted by agricultural, urban and small industrial activities. In this case, metal analysis techniques were improved (reducing detection limits from 10 ppb to 0.1 ppb) and the analyses of metal concentrations in biofilms and DGTs were also included.