In recent decades, the study of heavy metal pollution has garnered significant attention owing to the advancement of industrialization. To explore the contamination of heavy metals in an intensive industrial and agricultural region in Shandong province, China, 101 surface soil samples and 80 samples of edible crop parts were collected from the vicinity of an industrial park. A positive matrix factorization-multilayer perceptron model (PMF-MLP) was established to identify the sources of heavy metals and quantify the complex relationships between pollution sources, crop types, and pollution status. Index and human health-risk methods were used to assess the heavy metal pollution. The results show that (1) lead (Pb) and cadmium (Cd) in the soil may originate from industrial-traffic mixed pollution sources in the surrounding industrial park, whereas copper (Cu) is derived from agricultural pollution sources in the southern farmland. (2) Pollution is primarily concentrated in the central and northern regions of the study area. The analysis of the PMF-MLP model indicates that human activities account for the majority (79.6 %) of the risk of associated with heavy metal pollution. Among them, industrial, traffic, and agricultural mixed pollution sources, agricultural pollution sources associated with northern livestock farms, and crop types contribute to 49.3, 24.5, and 5.80 % of the total risk, respectively. (3) The oral intake of heavy metals represents the primary route of entry into the human body. Cd and Cu are the most significant elements associated with adverse human health, with Cd and Cu contributing the most to carcinogenic and non-carcinogenic risks in both adults and children, respectively. The results will provide references for the formulation of control strategies to curb heavy metal pollution.
Keywords: Heavy metal; Positive matrix factorization (PMF); Risk assessment; Source analysis.
Copyright © 2025 The Authors. Published by Elsevier Inc. All rights reserved.