Spatial dependence and bioavailability of metal fractions in paddy fields on metal concentrations in rice grain at a regional scale

Although the bioavailability of heavy metals has been widely investigated, little information is available on the spatial correlations of heavy metals in soil-rice systems at a regional scale. A study of heavy metals in soil-rice systems at a present rice production area could provide valuable information on the safety of rice production and provide guidelines beneficial to agriculture management and strategic sustainable agriculture in China and other rapidly developing regions in the world. The overall goals of this study were to identify the characteristics of metal fractions and their bioavailability to rice plants in the paddy fields of a present rice production region. In the rice harvest season (October 2006), 96 pairs of rice grain and rooted soil samples were collected from rice production area of Wenling in southeast Zhejiang province, China, which is one of the well-known electronic and electric waste (E-waste) recycling centers. Soil samples were analyzed for total heavy metal concentrations, metal fraction concentrations, and soil properties. Soil properties analyzed in this study included soil pH, electrical conductivity, organic matter, Fe oxides, and soil particle size distribution. Rice grain samples were analyzed for heavy metal concentrations. Multivariate statistical and geostatistical methods were applied to study the spatial dependence characteristics of metal fractions and their spatial correlation with uptake by rice plants in the rice production area and to identify the bioavailability of metal fractions to rice plants. The paddy soils of the studied area showed Cd contamination and some paddy soils presented a potential Cu, Ni, and Zn risk. The elevated levels of Cd were predominantly in non-residual (extractable) fractions. The percentage of Cd in fractions decreased in the order of exchangeable > Fe-Mn oxide bound > residual > organic bound fraction. In contrast, soil Cu, Ni, Pb, and Zn were mainly in the residual (non-extractable) fractions. The fractions of Ni, Pb, and Zn followed the order of residual > Fe-Mn oxide bound > organic bound > exchangeable fraction; the fractions of Cu decreased in the order of residual > organic bound > Fe-Mn oxide bound fraction. The spatial distribution patterns of non-residual fractions exhibited similarities with the highest metal concentrations in the northwest area owing to the industries and E-waste recycling activities. Most metals in rice grain were the strongest spatially correlated with the exchangeable fraction, followed by the organic bound fraction, indicating that exchangeable and organic bound fractions had the highest bioavailability. Rice Cd and Zn were strong spatially correlated with exchangeable, Fe-Mn oxide, and organic bound fractions; rice Ni and Cu were strongly spatially correlated with the exchangeable and organic bound fractions, respectively. The principal component analysis results also confirmed that exchangeable, Fe-Mn oxide, and organic bound soil fractions can be considered as bioavailable fractions to rice for Cd and Zn, while exchangeable and organic fractions were more important sinks for Ni and Cu, respectively. Due to a comparatively high input of Cd in the paddy soils, soil Cd was predominantly associated with non-residual fractions, especially with the exchangeable fraction. The soil Cu, Ni, Pb, and Zn were largely associated with the residual fraction while little associated with the exchangeable fraction. The bioavailability of the fractions to rice varied with metal fractions. In general, the exchangeable fraction had the highest bioavailability to rice plants, followed by the organic bound fraction. The bioavailability of the fractions to rice also varied with heavy metals. The exchangeable, Fe-Mn oxide, and organic bound fractions had high bioavailability to rice for Cd and Zn; the exchangeable and organic bound fractions had highest bioavailability for Ni and Cu, respectively.