The study combined multiple models to provide a deeper understanding to soil heavy metal contamination and source information, which are essential for controlling pollution and reducing human health risks. In this study, the agricultural soils were collected from the Qingyuan City of China as an example. The multiple models (APCS/MLR, PMF, and GDM) were used to identify and quantitatively apportion the main sources of heavy metal pollution in the area. The results showed that Cu (56.4 %), Ni (70.9 %), B (44.5 %), and Cr (72.8 %) were associated with natural sources, such as soil parent material and soil-forming processes. However, Pb (41.2 %), Zn (61.8 %), Hg (67.0 %), and Cd (69.6 %) were associated with agricultural activities, atmospheric deposition, vehicle exhaust emissions, and vehicle tires, while Mo, Se, and Mn were possibly derived from natural sources, including rock weathering and soil parent materials. Additionally, the network of environmental analysis revealed that soil microbes are far more sensitive to soil heavy metal pollution than herbivores, vegetation, and carnivores. This study can serve as a guideline for reducing the ecological and health risks associated with heavy metals in soil by controlling their preferential sources. Environmental implication Combining multiple models is more effective approach to wide understanding of heavy metal contamination and source information, which is essential for controlling pollution and reducing human health risks. Based on multiple models (APCS/MLR, PMF, and GDM) and network environ analysis, a comprehensive method for apportioning soil heavy metal sources and assessing ecological risk had been provided. Further, the present study can be a guideline for reducing ecological and health risks by heavy metals in soil by controlling preferential sources.
Keywords: Agricultural soils; Heavy metals; Network environ analysis; Risk assessment; Source apportionment.
Copyright © 2023 Elsevier B.V. All rights reserved.