Wiley, Environmental Toxicology and Chemistry, 5(27), p. 1217
DOI: 10.1897/07-162.1
Wiley, Environmental Toxicology and Chemistry, 2008(preprint), p. 1, 2007
DOI: 10.1897/07-162
Full text: Download
The ecological risk posed by 2,3,7,8-polychlorodibenzo-p-dioxins and furans (PCDD/Fs) and dioxin-like polychlorobiphenyl (PCB) congeners to five edible fish species of the aquatic food web of Venice Lagoon, Italy, was estimated by applying a state-of the-art kinetic bioaccumulation model. Site-specific data were used to define a representative food web. The experimental data set for model validation and application included PCB and PCDD/F congener concentrations in sediments, in water, and in five organisms (both invertebrates and fish). The spatial distribution of risk was evaluated by dividing the lagoon into six homogeneous areas, and for each area, sediment, water, and organism dioxins concentrations were calculated. The bioaccumulation model was calibrated for both nonmetabolizing and metabolizing congeners, the metabolic elimination rates of which were estimated. The model validation showed an acceptable bioaccumulation estimation, evaluated using the model bias parameter. The calibrated model was applied to the six areas of the lagoon to estimate the fish predicted exposure concentration as 2,3,7,8-tetrachlorodibenzo-p-dioxin toxicity equivalents from sediment concentration. Internal no-effect concentrations were calculated for each fish species from literature data. Risk was estimated by applying the hazard quotient (HQ) approach, obtaining the ecological risk for each fish species on the basis of 90 and 99% protection levels, in each of the six lagoon areas. The sediment dioxins concentration does not pose a significant risk to the selected fish species at the 90% protection target (HQ<1), whereas risk is significant (HQ>1) at the 99% protection target. Risk results were higher near the Porto Marghera industrial district, Italy, and in lagoon zones characterized by a low water-exchange rate and freshwater basin inputs.