The assessment of the combined effect of eutrophication and pollution on ecosystems is not traightforward due to complex interactions and feedbacks. Eutrophication may increase the primary roduction, dilute the contaminant in the biomass and furthermore increase the scavenging with the rganic matter (Koelmans et al. 2001). On the other hand contaminants can have direct and indirect ffects on the ecosystem balance and the growth of populations (Fleeger et al. 2003). Direct effects are aused by the toxicity of contaminants, which increase the mortality of the affected population; onversely indirect effects are the consequence of reduced food availability or reduced grazing. Thus, hile nutrient acts on the bottom level of the food chain, contaminants may affect higher trophic levels nd the correct understanding of the relative importance of top-down versus bottom-up controls is ssential to evaluate the system.
The traditional approach for the modeling of contaminants in the water column is to consider two wellmixed oxes during stratification periods and one well-mixed the rest of the time (Schwarzenbach et l., 2003; Meijer et al., 2006). The extensive number of 0D models for hydrophobic organic ompounds (Wania and Mackay, 1996; Scheringer et al., 2000; Dalla Valle et al., 2003; Dueri et al., 005) contrasts with the lack of spatially and temporally resolved models, with the exception of the ecently developed coastal lagoon model for herbicides (Carafa et al., 2006) and the one for HCH by lyina et al. (2006). A 1D dynamic hydrodynamic-contaminant model has been developed to analyze he influence of vertical mixing on the distribution of POPs in the water column (Jurado et al., 2007; arinov et al., 2007). The model was applied to the organic contaminants families selected in hresholds, i.e. PCBs, PAHs, and PBDEs, plus dioxins and furans, PCDD/Fs and details are presented n the Deliverable 2.6.2. Recently this 1D dynamic hydrodynamic-contaminant model was coupled ith a food-web ecological model that considers phytoplankton, zooplankton, bacteria and detritus Deliverable 2.6.3). The newly developed model uses nutrient concentrations as forcing and considers he bioaccumulation of POPs in all the ecological compartments. A first validation of the coupled odel has been achieved using experimental data on PAHs obtained at the Finokalia Station, Island of rete, Greece (Tsapakis et al., 2005 and 2006). The results showed that the model is able to reproduce he experimental concentrations as well as the measured fluxes. However, due to the low concentrations of PAHs in the onsidered remote area environment, the toxic effect model could not be alidated with those data.
Within the framework of the Threshold project, mesocosm experiments have been carried out in the sefjord (Denmark) by NERI (Deliverable 4.3.3) to elucidate the combined effects of nutrient and yrene on an ecosystem composed of phytoplankton, zooplankton and bacteria (Hjorth et al. 2007). he experimental results are used here to validate the model presented in Deliverable 2.6.3 on a maller scale, including the toxic effect model and at the same time investigate the direct and indirect ffects observed in the system.