生态环境学报 ›› 2026, Vol. 35 ›› Issue (10): 1600-1610.DOI: 10.16258/j.cnki.1674-5906.2026.10.010

• 研究论文【环境科学】 • 上一篇    下一篇

涉重金属企业周边土壤不同营养级生物响应阈值受体分异特征

江落雁1(), 钟嘉劲1, 黎铭洋1, 张池1, 曾曙才2,3, 仇荣亮1,3, 蒋成爱1,3,*()   

  1. 1 华南农业大学资源环境学院， 广东 广州 510642
    2 华南农业大学林学与风景园林学院， 广东 广州 510642
    3 广东省农业农村污染治理与环境安全重点实验室， 广东 广州 510642
  • 收稿日期:2026-04-16 修回日期:2026-09-03 接受日期:2026-09-20 出版日期:2026-10-18 发布日期:2026-10-21
  • 通讯作者: E-mail: chaijiang@scau.edu.cn
  • 作者简介:江落雁（1998年生），女，硕士，研究方向为农业环境保护。E-mail: 1356343114@qq.com
  • 基金资助:
    国家重点研发计划项目(2022YFC3703102)

Receptor-Specific Differences in Response Thresholds among Organisms at Different Trophic Levels in Soils Surrounding Heavy Metal-Related Enterprises

Jiang Luoyan1(), Zhong Jiajing1, Li Mingyang1, Zhang Chi1, Zeng Shucai2,3, Qiu Rongliang1,3, Jiang Chengai1,3,*()   

  1. 1 College of Resources and Environment, South China Agricultural University, Guangzhou 510642, P. R. China
    2 College of Forestry and Landscape Architecture, South China Agricultural University, Guangzhou 510642, P. R. China
    3 Guangdong Provincial Key Laboratory of Agricultural and Rural Pollution Control and Environmental Safety, Guangzhou 510642, P. R. China
  • Received:2026-04-16 Revised:2026-09-03 Accepted:2026-09-20 Online:2026-10-18 Published:2026-10-21

RichHTML

PDF

摘要：

【目的】土壤重金属复合污染下不同营养级生物响应分异及其驱动机制尚不清晰。阐明关键驱动因子及响应节点，对涉重企业周边土壤生态风险识别、早期预警和差异化管控具有重要意义。【方法】整合典型涉重企业周边61个土壤样点的理化性质、重金属总量及植物、蚯蚓和微生物响应数据，采用随机森林和SHAP值分解识别不同受体的主控因子及响应节点，构建多受体生态响应解释框架。【结果】植物毒性指数和微生物响应指数均以土壤性质为主导，贡献率分别为61.42%和56.31%，其中pH、有效磷作用突出；蚯蚓生存指数则受重金属主导，Cr、As和Cd贡献较高。全生物综合指数中，土壤性质贡献率为62.53%，pH、有效磷和阳离子交换量为主要因子。SHAP依赖关系显示，pH响应节点集中于6.6－7.0，Cd、As和Cr在特定受体中出现候选响应区间。【结论】复合污染土壤生态响应体现为土壤性质调控下的受体分异，植物和微生物更偏向指示生境扰动，蚯蚓更直接反映金属暴露风险。研究结果可为复合污染生态风险识别、响应节点筛选和差异化管控提供定量科学依据。

关键词: 土壤重金属污染, 生态风险, 机器学习, 受体分异, 响应节点

Abstract:

[Objective] The differentiated responses of organisms at different trophic levels to combined heavy metal contamination in soil, as well as their underlying drivers, remain insufficiently understood. Identifying the key driving factors and response nodes is essential for ecological risk assessment, early warning, and differentiated management of soils surrounding heavy metal-related enterprises. [Methods] In this study, soils collected around typical heavy metal-related enterprises were investigated by integrating soil physicochemical properties, total heavy metal concentrations, and response data from three receptor groups: plants, earthworms, and microorganisms. Random forest modelling and SHAP value decomposition were used to identify receptor-specific driving factors and response nodes. [Results] The results showed that the plant toxicity index was mainly regulated by soil properties, which contributed 61.42%, with pH and available phosphorus contributing 19.93% and 19.65%, respectively. The microbial response index was also dominated by soil properties, with an overall contribution of 56.31%; available phosphorus and pH contributed 30.99% and 14.24%, respectively, while Cd, As, and Pb also made evident contributions. In contrast, the earthworm survival index was more strongly driven by heavy metals, which contributed 59.50%; Cr, As, and Cd contributed 23.00%, 14.07%, and 6.36%, respectively, indicating that soil animals were more sensitive to direct metal exposure. At the integrated multi-receptor scale, soil properties contributed 62.53%, with pH, available phosphorus, and cation exchange capacity contributing 21.90%, 10.32%, and 10.14%, respectively. These results suggest that soil acidity-alkalinity, nutrient supply, and buffering capacity are important regulators of ecological stress under combined contamination. SHAP dependence analysis further revealed nonlinear responses to key factors across receptors. The response nodes for pH were concentrated within 6.6-7.0 across multiple receptors; those for Cd in microbial and earthworm responses were approximately 2.2-2.8 mg·kg−1; those for As were approximately 76-84 mg·kg−1; and those for Cr in earthworm and integrated responses were approximately 82-94 mg·kg−1. [Conclusion] These findings indicate that ecological responses in soils surrounding heavy metal-related enterprises are characterized by receptor-specific differentiation under the regulation of soil properties. Plant and microbial responses mainly reflect soil habitat suitability and microhabitat disturbance, whereas earthworm responses more directly indicate toxicity associated with metal exposure. This study provides quantitative support for ecological risk identification, response-node screening, and differentiated remediation management of soils contaminated by multiple heavy metals.

Key words: soil heavy metal pollution, ecological risk assessment, machine learning, receptor differentiation, response nodes

中图分类号: