Source-sink relationships and health risk assessment of heavy metals in farmland soils of a Southwest Karst Industrial cluster: based on PMF and Monte Carlo simulation
摘要
Karst agricultural soils are characterized by inherently elevated heavy metal backgrounds, a condition that is frequently intensified by superimposed industrial activities. However, the pathways linking source contributions to human health risks remain insufficiently resolved in such geochemically complex systems. In this study, an integrated analytical framework combining positive matrix factorization (PMF), Boruta feature selection and Monte Carlo simulation was applied to a representative industrial-agricultural karst region in Southwest China. The results showed that Cd, Cr, and Pb were widely enriched in surface soils. PMF analysis identified six major contributing sources. The natural geological source was the highest proportion, which was 28.4%, followed by traffic and historical sources (19.7%), industrial emissions (18.5%), coal combustion (13.0%), wastewater irrigation (12.0%), and agricultural practices (8.2%). Soil pH, organic matter, vegetation coverage, and related factors were identified as key regulators of crop heavy metal bioaccumulation. The non-carcinogenic risk of adults was acceptable, and the hazard index of children exceeded the safety threshold (average = 1.12). Cr, As and Cd were the main risk drivers. There is an obvious risk amplification effect between sewage irrigation and industrial pollution sources, and the toxic effects and mass concentrations do not appear in proportion. The geological source has an obvious contribution to the risks of Cr and Ni, indicating that the relatively high background value of the karst environment is not necessarily favorable. The findings suggest that soil management should shift from concentration-based control toward a pathway-oriented framework that targets sources and processes driving actual health risks.