A method for apportionment of natural and anthropogenic contributions to heavy metal loadings in the surface soils across large-scale regions.

Quantification of the contributions from anthropogenic sources to soil heavy metal loadings on regional scales is challenging because of the heterogeneity of soil parent materials and high variability of anthropogenic inputs, especially for the species that are primarily of lithogenic origin. To this end, we developed a novel method for apportioning the contributions of natural and anthropogenic sources by combining sequential extraction and stochastic modeling, and applied it to investigate the heavy metal pollution in the surface soils of the Pearl River Delta (PRD) in southern China. On the average, 45-86% of Zn, Cu, Pb, and Cd were present in the acid soluble, reducible, and oxidizable fractions of the surface soils, while only 12-24% of Ni, Cr, and As were partitioned in these fractions. The anthropogenic contributions to the heavy metals in the non-residual fractions, even the ones dominated by natural sources, could be identified and quantified by conditional inference trees. Combination of sequential extraction, Kriging interpolation, and stochastic modeling reveals that approximately 10, 39, 6.2, 28, 7.1, 15, and 46% of the As, Cd, Cr, Cu, Ni, Pb, and Zn, respectively, in the surface soils of the PRD were contributed by anthropogenic sources. These results were in general agreements with those obtained through subtraction of regional soil metal background from total loadings, and the soil metal inputs through atmospheric deposition as well. In the non-residual fractions of the surface soils, the anthropogenic contributions to As, Cd, Cr, Cu, Ni, Pb, and Zn, were 48, 42, 50, 51, 49, 24, and 70%, respectively.