Evaluation of various approaches to predict cadmium bioavailability to rice grown in soils with high geochemical background in the karst region, Southwestern China.

Evaluating the bioavailability of Cd to rice (Oryza sativa L.) was essential in the karst region, Southwestern China, where the soils have previously been shown to be anomalously enriched in Cd through geogenic processes. In this research, we examined the bioavailability of Cd to rice samples collected from 278 sites in Guangxi province, where rice is the most widely cultivated cereal crop that is responsible for the largest human dietary exposure to Cd. Both soil chemical extraction and soil-plant transfer modelling approaches were used to predict the bioavailability to rice. Some of the soil types were highly enriched in Cd, but their bioavailability was low, since the soil carbonates raised soil pH and remarkably reduced Cd bioavailability. In contrast, acidic soils (Ca was largely leached) with relatively low total Cd, the grown rice plants accumulated higher Cd in their grains. Results from CaCl2 extraction experiments provided good predictions for Cd in rice grain grown in soils of different types. Stepwise multiple regressions revealed soil pH and soil Ca content were the dominant factors that control the transfer of Cd from soil to rice. An extended Freundlich-type model and a polynomial surface model provided good prediction for Cd in rice grains. The diffusive gradients in thin films (DGT) technique gave the best estimation of soil Cd bioavailability, whereas water-extracted soil solution Cd provided relatively poor fits. Regional soil threshold that derived using the models, can avoid exceedance of Cd in rice and thereby enable local agricultural practitioners or authorities to develop appropriate management for croplands with high Cd background.