Integrated modelling of eutrophication and organic contaminant fate & effects in aquatic ecosystems. A review.

Eutrophication and contamination with micropollutants have been major problems in water quality management. Both problems have been subjected to extensive research and modelling but traditionally are treated separately. Traditional simulation models for aquatic systems can be categorised as eutrophication models, contaminant fate models, food web models and food chain bioaccumulation models. Because they are single issue models, many interactions and feedbacks between the food web, nutrient and toxicant cycles are missed. Linking these models is essential to evaluate the fate and risks of contaminants in systems with changing nutrient loading, to assess the natural attenuation of contaminants or to understand the selfpurifying capacity of ecosystems. Combination of the single issue models requires inclusion of 'interaction processes' to account for the coupling between the (sub-) model types, such as organic carbon cycling. toxicity, transport and accumulation of organic contaminants in the food chain, and bottom up versus top down control of primary production and nutrient cycling. This review first provides a brief overview of traditional approaches in modelling eutrophication, contaminant fate, food web dynamics and food chain bioaccumulation. Second, five existing integrated eutrophication, fate and/or effects models are reviewed. Third, the gaps and limitations in modelling the four types of interaction processes are discussed. It is concluded that these models are invaluable tools to focus attention to feedback mechanisms that are often overlooked, to identify dominating processes in ecosystems, to formulate counterintuitive hypotheses on ecosystem functioning, or to assess short term risks of acutely toxic stressors. However, the potential of integrated models for long term simulations of contaminant exposure. food chain bioaccumulation and effects to higher trophic levels remains limited, mainly because of principal limitations in food web modelling. In contrast, the potential of integrated models for long term simulations of contaminant fate is better because the environmental distribution of contaminants is mainly determined by the major abiotic compartments and by biotic compartments at the base of the food chain.