Spatial Prediction,Environmental Risk And Safe Utilization Zoning Of Soil Heavy Metals In Cultivated Land In Lanxi City,Zhejiang Province

The concentrations of heavy metals in cultivated soil directly affect crop growth and food production,which further affect food quality and human health.It is of great significance for regional agricultural production and food safety to study the spatial distribution,environmental risks and safe utilization zoning of soil heavy metals in cultivated soil.In this study,Lanxi City,Zhejiang Province was selected as the research area,and 777 soil samples from the surface layer(0-20 cm)of cultivated land were collected based on the grid sampling method.Geostatistical methods were used to analyze the spatial variation and spatial distribution characteristics of soil heavy metals(As,Cd,Cr,Cu,Hg,Ni,Pb,Zn).Multiple linear regression,ordinary kriging,regression kriging and random forest regression were all used to predict the concentrations of heavy metals in cultivated land in Lanxi City,and the prediction accuracy of the four model methods was compared.Combining hot spot analysis and indicator kriging technology to further explore the environmental risks of soil heavy metals.Finally,the soil heavy metal pollution was evaluated by the influence index of comprehensive quality of soil,and safe utilization zoning of soil heavy metals in cultivated land was based on the soil pollution evaluation.In addition,suggestions were made for different grades of utilization zones to ensure safe utilization of cultivated land.The main conclusions of this study are as follows:(1)The mean concentrations of soil heavy metals were,in descending order:Zn(74.69 mg/kg)>Cr(41.83 mg/kg)>Pb(31.89 mg/kg)>Cu(20.80 mg/kg)>Ni(15.44mg/kg)>As(7.53 mg/kg)>Cd(0.22 mg/kg)>Hg(0.12 mg/kg).The mean concentrations of the eight heavy metals all did not exceed the risk screening values for soil contamination of agricultural land of China,but the mean concentrations of Cd,Cu,Hg,Ni and Zn were higher than the corresponding soil background values of Jinqu Basin and Zhejiang Province,indicating that different levels of enrichment appeared.As,Cd,Cu,Hg,Pb,and Zn had moderate spatial autocorrelation in the cultivated soil in the study area,while Cr and Ni had weak spatial autocorrelation.The high values of As,Cr,Cu,Ni,and Hg were mainly distributed in the middle and west of Lanxi City.The high values of Cd,Pb,and Zn were mainly distributed in Lingdong Township,Shanghua Street and Yunshan Street.(2)The prediction results of different models showed that ordinary kriging and multiple linear regression models could only reflect the overall trend of heavy metals distribution,and it was hard to reflect local variation,so the prediction accuracy was relatively low.Regression kriging and random forest models could better reflect the local spatial variation,and the random forest prediction results were more precise.By comparing the prediction accuracy of different models,the model with the highest accuracy for As,Cd,Cr,Hg,Pb,and Zn was the random forest model,with R~2 of 0.59,0.52,0.58,0.47,0.48,and 0.41,respectively.The model with the highest accuracy for Cu was the ordinary kriging model,with an R~2 of 0.37.The model with the highest accuracy for Ni was regression kriging model,with an R~2 of 0.48.(3)Hot spot analysis and indicated kriging interpolation results showed that the distribution of hot spots was consistent with the environmental risk areas above medium risk.The hotspots of As elements were mainly distributed in the south of Shanghua Street,the south of Nubu Street,and the south of Xiangxi Town;The hotspots of Cd,Pb and Zn were mainly distributed in Lingdong Township,the northern part of Shanghua Street and Yunshan Street.The hotspots of Cr,Cu and Ni were mainly distributed in the western part of the study area and the northern part of Xiangxi Town.The hotspots of Hg were distributed in Chixi Street,Youbu Town,Yongchang Street.(4)About 75%of the cultivated land area in the study area had utilization risks,mainly low and medium-level risks,and the high-risk areas were small.For safe and basically safe utilization zones,a protection strategy should be implemented.The low-risk utilization zone and the medium-risk utilization zone should adopt dynamic monitoring and early-warning prevention and control strategies respectively.High-risk utilization zone should limit agricultural utilization.