Abstract
During the past decades, agricultural soil heavy metal pollution has been becoming increasingly severe due to urbanization and industrialization. However, the impact of externally input heavy metals on deep soils remains unclear because most previous relevant research only focused on surface soils. In the present study, Concentrations of eight heavy metals (Cu, Zn, Ni, Pb, Cr, Cd, As, and Hg) were determined for 72 pairs of surface and deep soil samples collected from an agricultural region close to the Pearl River estuary. Subsequently, heavy metal pollution and potential health risks were assessed using the Geo-accumulation Index and Potential Ecological Risk Index, a dose response model and Monte Carlo simulation, respectively. Principal component analysis (PCA) and the positive matrix factorization (PMF) receptor model were combined to analyze heavy metal sources. The results indicated that average concentrations of all heavy metals exceeded their corresponding background values. Cd was identified as the main pollutant due to its extremely high values of I(geo) and Er. Unacceptable potential heavy metal non-carcinogenic and carcinogenic risks indicated by respectively calculated HI and TCR, higher than thresholds 1.0 and 1.0 × 10(-4), mainly arose from heavy metals As, Cd, Cr, and Ni through food ingestion and dermal absorption. Anthropogenic sources respectively contributed 19.7% and 38.9% for soil As and accounted for the main contributions to Cd, Cu, and Hg (Surface: 90.2%, 65.4%, 67.3%; Deep: 53.8%, 54.6%, 56.2%) within surface and deep layers. These results indicate that soil heavy metal contents with deep layers were also significantly influenced by anthropogenic input. Therefore, we suggest that both surface and deep soils should be investigated simultaneously to gain relatively accurate results for soil heavy metal pollution and source apportionments.