Detecting anthropogenic stress in an ecosystem: 1. Meiofauna in a sewage gradient

Traditional monitoring or source partitioning of the environment is accomplished using chemical analyses. This will only inform the measurer of the current chemical state of that system and will not provide information about the biological availability or effect of these chemicals. As part of a larger study of the Ria Formosa lagoon in Portugal, the meiofauna community was quantified from intertidal sediments in a sewage contamination gradient. Traditional univariate measures of diversity, species richness, and evenness were not statistically able to separate contaminated from uncontaminated regions, although trends were evident. Within the nematode community, there were changes in the feeding mode between sites (diatoms feeders, predators, selective and nonselective feeders), although they were not correlated to the change in organic carbon content nor the silt-clay fraction. Multivariate approaches using principal component analysis (PCA) and partial least squares (PLS) signatures were able to identify and quantify the sites affected by different stressors. In response to sewage stress the community structure changed, but this was not reflected in the k-dominance curves or other univariate measures. Since the PLS method defines a signature based on the species present in the community, it was able to identify other sites impacted by sewage and provide a measure of degree of effect. Those sites identified as having a >10% sewage-related community were revisited, and evidence of sewage discharges were found.