A multivariate assessment of the trophic state of a man-made reservoir in North Tunisia

In the present work, we attempt to test the applicability of common trophic state indices such as Carlson’s index (1977), probability distribution diagrams of Vollenweider (1982), and fixed boundary systems of the OECD (1982) and of Nurnberg (1996) for evaluating the risk of eutrophication in the Sejnane reservoir in North Tunisia. The environmental parameters considered include monthly values of nutrients, chlorophyll a, Secchi transparency, turbidity and dissolved oxygen over a decade of monitoring from November 2001 through October 2011. To compile data and to identify key aspects of water quality, a multivariate technique as Principal Component Analysis was performed. Results show that productivity is limited by sedimentary turbidity, which contributes to decreasing water clarity and promotes the adsorption of phosphorus to suspended matter that decreases, indirectly, the bioavailability of nutrients. The Nurnberg model provides an accurate assessment of eutrophication with a clear internal coherence between self-indicators, and appears to be the most suitable way to evaluate eutrophication risk in the reservoir