Human health risk apportionment from potential sources of heavy metals in agricultural soils and associated uncertainty analysis.

Abstract

Evaluating heavy metal pollution level in the soils and apportioning the source-specific health risk of heavy metals are critical for proposing environmental protection and remediation strategies to protection human health. This study explored heavy metal pollution and associated source-specific health risks in a typical rural industrial area, southwestern China. A total of 105 topsoil samples were collected and the concentrations of heavy metals, including As, Cd, Cr, Cu, Ni, Pb and Zn, were determined. Pollution load index and enrichment factors were used to evaluate the pollution level of heavy metals. Positive matrix factorization (PMF) model was applied to apportion the heavy metals and the associated source-specific health risks to adults and children were estimated via combining the PMF model with the health risk assessment. The results indicated that the soils were highly polluted by multiple heavy metals, especially for Cd, with the EF values of 24.94 and 22.55 in the upstream and downstream areas, respectively. Source apportionment results showed that atmospheric deposition, smelting activities, fertilizer and sewage application, and agrochemical utilization were the main anthropogenic sources, with the contributions of 37.11%, 23.69%, 19.69%, and 19.51%, respectively. Source-specific risk assessment identified atmospheric deposition as the priority source for the non-carcinogenic (NCR) and carcinogenic risks (CR) in the study area, with the contribution of 43.71% and 52.52% for adults, and 44.29% and 52.58% for children, respectively. Moreover, non-carcinogenic and carcinogenic risks posed to children (NCR: 2.84; CR: 1.31 × 10-4) from four sources was higher than those posed to adults (NCR: 0.29; CR: 5.86 × 10-5). The results of source-specific health risk assessment provided the valuable information on the priority sources for pollution preventing and risk controlling.