Characteristics and quantitative model of heavy metal transfer in soil-rice systems in typical rice production areas of Zhejiang Province

The understanding of characteristics of heavy metal transfer in soil-rice systems can improve soil quality in production areas and guide the safe production of rice. We collected soil and rice samples from three typical rice production areas (Nanxun, Shengzhou and Wenling) located in the northern, central and southern parts of Zhejiang Province. The controlling factors of heavy metal transfer were studied based on a transfer model set up for hybrid rice and japonica rice. The objective of the study was to identify transfer traits of heavy metals in soil-rice systems in typical rice production areas in Zhejiang Province and to guide safe agricultural production. The results suggested that the physico-chemical properties were different in the three areas. pH (mean value of 5.52), organic matter (mean value of 39.4 gkg^-1), EC and heavy metal fractions contents in soil in Shengzhou area were lower than those in the other two production areas. Sand content of soil in Shengzhou area was higher than that in the other two areas. Heavy metals in soils and rice were significantly different from each other of rice production areas. Heavy metals (Cd, Cu and Zn) contents in soil in Wenling area were significantly higher than those in the other two areas. Then heavy metals contents in rice in Shengzhou area were significantly higher than those in Nanxun area (P < 0.05). No carbonate bound fraction of heavy metals was detected in the study. The corresponding contents of exchangeable, Fe-Mn oxide bound, organic bound, and residual fractions of heavy metals in Shengzhou were lower than those in the other two production areas due to the lowest total heavy metals contents in Shengzhou soil. The enrichment indexes (EI) of heavy metals were different in the three production areas. Generally, EIs of Cd and Zn were higher than those of Cu and Ni. Also EIs in Shengzhou area (range of 0.018–0.521) were significantly higher than those in the other two areas (range of 0.004–0.143 for Nanxun area and of 0.007–0.269 for Wenling area). Both soil physico-chemical properties and heavy metals fractions were important factors influencing heavy metal enrichment indexes. Compared with heavy metals fractions, soil physico-chemical properties contributed more to the movement of heavy metals in soil-rice systems. A log-linear model of heavy metals combined with the physico-chemical properties and heavy metal fractions well predicted the availability of heavy metals in soil-rice systems under practical production conditions. The accuracy of the model prediction for hybrid rice was better than that for japonica rice. The Ni (regression coefficient r was 0.61 and 0.70 at P < 0.01 for hybrid and japonica rice, respectively) model was better than that of other heavy metals. However, the accuracy of the model prediction of hybrid rice Cd content (r = 0.21 at P > 0.05) was poor. In that case, it was necessary to conduct further research in order to improve the accuracy of the model by either using more of the environmental variables or adjusting the variables.