Risk Assessment Of Heavy Metals In Soil-rice System Based On Spatial Analysis

With the rapid development of urbanization and industrialization in recent years, Chinese agricultural soils and cropsare suffering from increasing damage from heavy metals, which are derived from various pollution sources including agriculture, traffic, mining and especially the flourishing private metal recycling industry.In this study,219pairs of rice grain and their corresponding soil samples were collected from Wenling in Zhejiang province. Total concentrations of heavy metals in soil and rice grains, and soil physic-chemical properties were analyzed. Based on Geostatisticsand relevant spatial analysis techniques, we studied the metal pollution level, spatial correlation and variation, prediction accuracy, and the risk assessment of heavy metals in soil-rice system.Local Moran’s I index was employed to identify metal pollution hotspots. The results revealed that the heavy metal pollution hotspots were mostly distributed in northwestern county with more E-waste dismantling workshops there, indicating that E-waste dismantling may be the main sources of these metals, leading an increased heavy metal concentrition in the study area.Accumulation and availability of heavy metals in soil-rice system may be influenced not only by soil heavy metal concentrations, also by soil physico-chemical properties. The higher the content of soil organic matter and sand, the greater the enrichment coefficient of heavy metal is, thus the availabilityof heavy metal is stronger. And when the soil pH, conductivity, silt and clay content are higher, the heavy metal enrichment coefficients become lower. Among all the properties, soil pH and organic matter were the most important factors controlling the uptake of heavy metals in rice.Among variety of spatial analysis methods, kriging method of interpolation maps tended to be smooth with less local details. However, the maps produced by sequential gaussian simulation method had rich spatial structure information and strong volatility. For soil Cu, the prediction accuracy of simulation was higher than that by kriging, but opposite for Cd. In general, Simulation method expanded the prediction range detecting more local details, while kriging was narrowing the scope of the forecast, reflecting a smoothing effect.Ordinary kriging was compared with Bayesian maximum entropy(BME) in spatial analysis. The prediction accuracy and volatility of BME_HS, containing soft data and hard data in BME, was higher than that of Ordinary Kriging did in interpolating of soil Cd and Cu. The prediction accuracy of BME_H was between BME_HS and Ordinary Kriging.The results of soil environmental quality assessment and human health risk assessment in soil-rice system showed that Cd, Cu, Ni and Zn in soil were obviously enriched in the study area. The areas with high risk of soil Cd, Cu, and Zn were mainly distributed in the northwestern part of the study area and Ni in the east. In addition, there were2%and3%of the soil samples belonged to moderate pollution of heavy pollution levels, respectively. At the same time, rice in some parts of the study area had been polluted by Cd seriously. Consumption of rice in the study area would lead to Cd potential health risks especially for children, since individual HQ was greater than1.