Abstract
The monitoring of heavy metals in the marine environment is often carried out by using bioindicator organisms. In most cases metal concentrations in an area are evaluated using pooled samples from a single sampling site. Thus, a large number of individuals are analysed per site, but neither the statistical distribution of the data nor the intra-site variability are known. In order to optimise the monitoring of heavy metals using this kind of samples, some authors have suggested that the variability among individuals should be studied at least in one site. This work was designed to know the frequency distribution and the inter-individual variability of Hg, Cd, Pb, Cr, Ni, As, Ag, Cu and Zn in four bioindicator organisms (the blue mussel Mytilus galloprovincialis, the clam Venerupis pullastra, the king scallop Pecten maximus and the cockle Cerastoderma edule). In most cases metals in one-individual samples were shown to follow a log-normal distribution. As the pooled samples included more individuals they approached the normal distribution but still being closer to the log-normal one, suggesting that, in all cases, a logarithmic transformation should be used to normalise the data. The inter-individual variability observed indicated that at least two pooled samples of 30 individuals (a hundred in few cases) must be analysed to detect differences of 25% (both between sites and with time).