Assessment of metal contamination in dregded sediments using fractionation and Self-Organizing Maps

Although total metal content is frequently the initial approach for measuring pollution, no information is provided about mobility and environmental risk. In this paper, a metal fractionation (sequential extraction) technique and artificial neural networks (Self-Organizing Maps, SOMs) have been used jointly to evaluate the pollution level of the sediments dredged from the dry dock of a former shipyard in the Bilbao estuary (Bizkaia, Spain). The load pollution index (LPI) for the upper, middle and bottom layers of the sediments was 7.65, 8.22 and 10.01, respectively, for six metals (Cu, Mn, Ni, Cr, Pb and Zn). This showed that upper sediments were less polluted than the lower ones. Consequently, a reduction in the pollution level of metal discharged into the river in recent years was confirmed. According to fractionation results, the most mobile minor elements were Cu, Pb and Zn, as they are mainly associated with the non-residual fractions. The statistical approach of Self-Organizing Maps (SOMs) revealed that Ni, Pb and Zn amounts in the residual fraction followed the same pattern associated with simultaneous discharges of slags into the river. However, other hazardous discharge sources are responsible for the high accumulation of those metals in the non-residual fractions.