Modélisation de la bioaccumulation de métaux traces (Hg, Cd, Pb, Cu et Zn) chez la moule, Mytilus galloprovincialis, en milieu méditerranéen

Monitoring coastal contamination of trace metals using bivalve molluscs as quantitative bioindicators is widely used in many international monitoring programs. The bioaccumulation of metal within an organism results from the interaction between physiological (growth, loss of weight, absorption, accumulation), chemical (metal concentration, speciation, bioavailability) and environmental factors (temperature, SPM, chlorophyll). By utilizing the parameters above in this study, it appears that modeling metal dynamics in marine mussels is a reliable tool. The model allows understanding the bioaccumulation process which results from the interactions between biological, chemical and environmental factors. One of the objective of this study was to couple a bioaccumulation model with a growth model (Dynamic Energy Budget model) in order to consider the changing physiological conditions of the mussel and their impact on metal concentration in tissue. To build and calibrate such a model, in situ experiments to measure uptake (6 months) and elimination (3 months) kinetics were conducted for five metals (Hg, Cd, Pb, Cu and Zn), by transplantation, at three Mediterranean sites (Lazaret bay, Bages pond, and Port-Cros island). These sites were chosen on the basis of their contamination levels. A pharmaco-kinetic compartment model has been used to describe metal kinetics. It has been designed to account for changes in the physiological conditions of the organism, in relation to environmental conditions. Metal uptake from water and food were considered seperately. Metal elimination results from reproduction and/or from direct excretion. The contribution of physiologically determined variables, such as body size and tissue composition, influencing the pharmaco-kinetics of the metals, were quantified. By combining environmental (chemical and abiotic factors) and biological (physiological, growth and bioaccumulation) data, the model constitutes an optimised biomonitoring tool which can be applied to various coastal environments. An application on French biointegrator network (RINBIO) has been carried by inverse analyse and allows us to assess the real contamination