Statistical models for estimating seawater metal concentrations from metal concentrations in mussels (Mytilus edulis)

Conclusions

The results of this study have shown that field studies can be used to develop statistical models for estimating the concentration of trace metals in seawater knowing the concentration of metals in the mussels. Although only about 50% of the variance of the soluble metal and about 20% of the metal as SPM can be accounted for by the models there is no reason to doubt that better models can be obtained in the future ff parameters which reflect instantaneous effects are included. Such parameters could be measurements of the freshwater input as reflected by salinity and turbidity measured with sechi disk readings. Obviously one of the important variables to consider is salinity as both DAVENPORT (1977) and PHILLIPS (.1976) have shown that this variable affects copper and zinc uptake. These parameters could help solve the problem encountered in this study in which a constant bias is seen in the results: low concentrations of metals in the seawater are overestimated and high values are underestimated. Nevertheless the models presented in this paper can be considered as a first step in estimating time-integrated concentrations of trace metals using easily measured parameters in mussels collected in a natural environment.