Heavy metal contamination in agricultural soils threatens ecosystem safety and sustainable land use, particularly in geologically sensitive areas. This study aimed to assess the pollution status, ecological risks and source contributions of eight heavy metals (Hg, Cd, Pb, As, Cr, Cu, Ni and Zn) in soils from a dry-hot agricultural region of central Yunnan, China. To improve source apportionment, this study applied and compared three models: APCS-MLR, PMF, and Random Forest. Analysis of 1790 soil samples showed mean concentrations (mg/kg) of 0.03 for Hg, 0.17 for Cd, 25.01 for Pb, 7.46 for As, 85.91 for Cr, 36.20 for Cu, 31.75 for Ni, and 69.24 for Zn. Pollution assessment indicated that Cu and Cd were the main pollutants, while ecological risk assessment identified Cd and Hg as the dominant ecological risk factors. Four major sources were identified: industrial hybrid sources, natural background, atmospheric deposition and agricultural activities, with industrial hybrid sources contributing the largest share. These results indicate that integrating APCS-MLR, PMF, and Random Forest provides a more reliable framework for source identification and supports targeted soil pollution control in regions affected by both natural and anthropogenic inputs.
Song et al. (Fri,) studied this question.