基于自动机器学习模型预测作物籽粒重金属浓度

doi:10.13287/j.1001-9332.202506.018

摘要/Abstract

摘要： 随着工业化进程的加速和农业活动的频繁,作物重金属污染已成为当前农业生产中一个不容忽视的问题。本研究基于54篇文献的791组数据,利用自动机器学习模型对作物籽粒重金属浓度进行预测。研究选取有机肥施用量、有机肥重金属浓度、土壤重金属浓度、有机质、酸碱度、阳离子交换量、黏粒含量、砂粒含量、粉粒含量和作物类型10种影响因素作为输入变量,选取铬(Cr)、镉(Cd)、铅(Pb)、砷(As)和汞(Hg)在作物籽粒中的浓度作为输出变量,评估深度学习(DL)、分布式随机森林(DRF)、极度随机树(XRT)、堆栈集合(SE)、梯度提升机(GBM)和广义线性模型(GLM)6种模型的模拟预测效果,并分析影响作物籽粒重金属累积的关键因素。结果表明: 不同重金属的最佳预测模型存在差异。DL模型对Cr、Pb、As和Hg的预测效果最优,而GBM模型对Cd的预测精度最高。特征重要性和SHAP分析显示,有机肥施用量与作物类型是影响作物籽粒重金属累积的关键因素,有机肥施用量、土壤重金属浓度、有机肥重金属浓度、砂粒含量与作物籽粒重金属浓度呈正相关,阳离子交换量、酸碱度、有机质、黏粒含量与作物籽粒重金属浓度呈负相关。综上,DL和GBM模型在预测作物籽粒重金属浓度中具有优势,生产中需严格控制有机肥施用带来的重金属输入风险。

关键词: 机器学习, 重金属, 籽粒, 有机肥, 预测

Abstract: With the acceleration of industrialization and the intensification of agricultural activities, heavy metals (HMs) pollution in crops has become an issue that can not be ignored in current agricultural production. Based on 791 data sets from 54 publications, we predicted HMs concentrations in crop grains by using automated machine learning (AutoML) models. Ten factors were used as input variables: organic fertilizer application, HMs concentration in organic fertilizer, soil HMs concentration, soil organic matter, pH, cation exchange capacity, clay content, silt content, sand content and plant types. The concentrations of chromium (Cr), cadmium (Cd), lead (Pb), arsenic (As) and mercury (Hg) in crop grains were set as output variables. We evaluated the simulation and prediction performance of six models: deep learning (DL), distributed random forest (DRF), extremely randomized trees (XRT), stacked ensemble (SE), gradient boosting machine (GBM) and generalized linear model (GLM), with which we analyzed the key factors driving heavy metal accumulation in crop grains. The results showed that the optimal prediction model differed for different HMs. The DL model provided the best prediction for Cr, Pb, As and Hg, while the GBM model achieved the highest prediction accuracy for Cd. Feature importance and SHAP analysis revealed that the application of organic fertilizer and plant type were the key factors influencing HMs accumulation in crop grains. Organic fertilizer application, soil HMs concentration, organic fertilizer HMs concentration, and sand content were positively correlated with HMs concentration in crop grains, while cation exchange capacity, pH, organic matter, and clay content were negatively correlated with heavy metal concentration in crop grains. In summary, the DL and GBM models performed better in predicting heavy metal concentrations in crop grains. The input risk of heavy metals during organic fertilizer application must be strictly controlled.

Key words: machine learning, heavy metal, grain, organic fertilizer, prediction

张业翔, 陈奉献, 张宇红, 陈希娟. 基于自动机器学习模型预测作物籽粒重金属浓度[J]. 应用生态学报, 2025, 36(6): 1889-1897.

ZHANG Yexiang, CHEN Fengxian, ZHANG Yuhong, CHEN Xijuan. Predicting heavy metal concentration in crop grain using automated machine learning models[J]. Chinese Journal of Applied Ecology, 2025, 36(6): 1889-1897.

参考文献

[1] 岳聪, 黄顺寅, 屠春宝, 等. 植物修复重金属镉污染土壤研究进展. 现代农业科技, 2024(17): 129-135
[2] Hou SN, Zheng N, Tang L, et al. Pollution characteristics, sources and health risk assessment of human exposure to Cu, Zn, Cd and Pb pollution in urban street dust across China between 2009 and 2018. Environment International, 2019, 128: 430-437
[3] 陈印军, 方琳娜, 杨俊彦. 我国农田土壤污染状况及防治对策. 中国农业资源与区划, 2014, 35(4): 1-5
[4] Agnihotri A, Praveen G, Anuj D, et al. Counteractive mechanism(s) of salicylic acid in response to lead toxicity in Brassica juncea (L) Czern. cv. Varuna. Planta, 2018, 248: 49-68
[5] Rashid A, Schutte JB, Ulery A, et al. Heavy metal contamination in agricultural soil: Environmental pollutants affecting crop health. Agronomy, 2023, 13: 1521
[6] Liu WR, Zeng D, She L, et al. Comparisons of pollution characteristics, emission situations, and mass loads for heavy metals in the manures of different livestock and poultry in China. Science of the Total Environment, 2020, 734: 139023
[7] Onwosi OC, Igbokwe CV, Odimba NJ, et al. Composting technology in waste stabilization: On the methods, challenges and future prospects. Journal of Environmental Management, 2017, 190: 140-157
[8] 张彦, 张惠文, 苏振成, 等. 长期重金属胁迫对农田土壤微生物生物量、活性和种群的影响. 应用生态学报, 2007, 18(7): 1491-1497
[9] Imoto Y, Yasutaka T. Comparison of the impacts of the experimental parameters and soil properties on the prediction of the soil sorption of Cd and Pb. Geoderma, 2020, 376: 114538
[10] Naz M, Dai ZC, Hussain S, et al. The soil pH and heavy metals revealed their impact on soil microbial community. Journal of Environmental Management, 2022, 321: 115770
[11] Liu SH, Ji XH, Chen Z, et al. Silicon facilitated the physical barrier and adsorption of cadmium of iron plaque by changing the biochemical composition to reduce cadmium absorption of rice roots. Ecotoxicology and Environmental Safety, 2023, 256: 114879
[12] 陈文哲, 黄秋香, 孟凡德, 等. 草酸和酒石酸对稻田土壤中砷解吸行为的影响研究. 生态环境学报, 2024, 33(8): 1298-1305
[13] 王泓博, 苟文贤, 吴玉清, 等. 重金属污染土壤修复研究进展: 原理与技术. 生态学杂志, 2021, 40(8): 2277-2288
[14] 刘彤, 杨立琼, 石亚楠, 等. 我国场地土壤修复技术综合分析[EB/OL]. (2024-07-02) [2024-10-23]. 生态学杂志. https://link.cnki.net/urlid/21.1148.Q.20240701.1512.002
[15] Huang BY, Lu QX, Tang ZX, et al. Machine learning methods to predict cadmium (Cd) concentration in rice grain and support soil management at a regional scale. Fundamental Research, 2024, 4: 1196-1205
[16] Bagheri M, Mirbagheri SA, Ehteshami M, et al. Analysis of variables affecting mixed liquor volatile suspended solids and prediction of effluent quality parameters in a real wastewater treatment plant. Desalination Water Treatment, 2016, 57: 21377-21390
[17] 李香伶, 陈奉献, 陈希娟. 基于XGBoost模型的土壤中阿特拉津降解预测. 应用生态学报, 2024, 35(3): 789-796
[18] 中华人民共和国国家统计局. 中国统计年鉴. 北京: 中国统计出版社, 2023
[19] Zobeiri A, Rezaee A, Hajati F, et al. Post-Cardiac arrest outcome prediction using machine learning: A systematic review and meta-analysis. International Journal of Medical Informatics, 2025, 193: 105659
[20] Zhong SF, Zhang K, Bagheri M, et al. Machine learning: New ideas and tools in environmental science and engineering. Environmental Science and Technology, 2021, 55: 12741-12754
[21] Flynn R, Olukoya O. Using approximate matching and machine learning to uncover malicious activity in logs. Computers and Security, 2025, 151: 104312
[22] Wu J, Huang CM. Machine learning-supported determination for site-specific natural background values of soil heavy metals. Journal of Hazardous Materials, 2025, 487: 137276
[23] Solayman M, Islam AM, Paul S, et al. Physicochemical properties, minerals, trace elements, and heavy metals in honey of different origins: A comprehensive review. Comprehensive Reviews in Food Science and Food Safety, 2016, 15: 219-233
[24] 国家地球系统科学数据中心. 中国高分辨率国家土壤信息格网基本属性数据集(2010—2018)[EB/OL]. (2021-07-08) [2024-04-26]. https://www.geodata.cn/
[25] 中华人民共和国农业农村部. 中华人民共和国农业行业标准有机肥NY/T 525—2021. 北京: 中国标准出版社, 2021
[26] Ghosh S, Ojha KP, Carnesecchi E, et al. Exploring QSAR modeling of toxicity of chemicals on earthworm. Ecotoxicology and Environmental Safety, 2020, 190: 110067
[27] Bian L, Wang Z, White DL, et al. Machine learning-assisted calibration of Hg. Biosensors and Bioelectronics, 2021, 180: 113085
[28] Sebastian D, Sebastian C, Jakob H, et al. Deep learning based communication over the air. IEEE Journal of Selected Topics in Signal Processing, 2017, 12: 132-143
[29] Mahmoud SE, Li J, Nourhan K, et al. Assessing and predicting soil quality in heavy metal-contaminated soils: Statistical and ANN-based techniques. Journal of Soil Science and Plant Nutrition, 2023, 23: 6510-6526
[30] Liu T, Wang MS, Wang MY, et al. Identification of the primary pollution sources and dominant influencing factors of soil heavy metals using a random forest model optimized by genetic algorithm coupled with geodetector. Ecotoxicology and Environmental Safety, 2025, 290: 117731
[31] Cushman S, Kilshaw K, Campbell R, et al. Comparing the performance of global, geographically weighted and ecologically weighted species distribution models for Scottish wildcats using GLM and random forest predictive modeling. Ecological Modelling, 2024, 492: 110691
[32] 路港滨, 俄胜哲, 袁金华, 等. 石灰性土壤与作物系统中重金属的迁移运转及安全阈值研究. 土壤与作物, 2025, 14(1): 84-94
[33] 王柳茜, 余丹, 王冬艳, 等. 吉林省黑土区土壤重金属元素的生物有效性转化效率特征及相互关系. 江苏农业科学, 2017, 45(8): 274-278
[34] 马阳阳, 廉梅花, 王鹏. 超富集植物修复土壤重金属污染及强化措施研究进展. 环境保护与循环经济,2024, 44(1): 25-32
[35] Ghori NH, Ghori T, Hayat MQ, et al. Heavy metal stress and responses in plants. International Journal of Environmental Science and Technology, 2019, 16: 1807-1828
[36] Gao C, El-Sawah AM, Ali DFI, et al. The integration of bio and organic fertilizers improve plant growth, grain yield, quality and metabolism of hybrid maize (Zea mays L.). Agronomy, 2020, 10: 319
[37] Sadeghi S, Nafez AH, Nikaeen M, et al. Microbial indicators in municipal solid waste compost and their fate after land application of compost. Journal of Environmental Health Science and Engineering, 2022, 21: 85-92
[38] Singer AC, Shaw H, Rhodes V, et al. Review of antimicrobial resistance in the environment and its relevance to environmental regulators. Frontiers in Microbiology, 2016, 7: 1728
[39] Gutiérrez C, Fernández C, Escuer M, et al. Effect of soil properties, heavy metals and emerging contaminants in the soil nematodes diversity. Environmental Pollution, 2016, 213: 184-194
[40] Mukhopadhyay R, Sarkar B, Palansooriya KN, et al. Natural and engineered clays and clay minerals for the removal of poly- and perfluoroalkyl substances from water: State-of-the-art and future perspectives. Advances in Colloid and Interface Science, 2021, 297: 102537
[41] 王翠红, 周清, 黄启为, 等. 不同类型水稻土上外源镉对玉米生长发育影响的研究. 农业环境保护, 2001, 20(5): 293-296