Bioavailable arsenic and amorphous iron oxides provide reliable predictions for arsenic transfer in soil-wheat system

The application of current soil quality standards based on total arsenic (As) fails to assess the ecological risks of soil arsenic or to ensure the safety of crops and foods. In this study, bioavailable arsenic instead of total arsenic was applied to improve predictive models for arsenic transfer from soil to wheat (Triticum turgidum L.). The stepwise multiple-linear regression analysis showed that bioavailable arsenic and amorphous iron oxides (Fe_OX) were the two most important factors contributing to arsenic accumulation in wheat grain, with the explained percentage of variation being up to 82%. Compared with the bioavailable arsenic extracted by NH₄H₂PO₄, bioavailable arsenic extracted by HNO₃ from soils generated better predictions of the amount of arsenic in grain. The best reliable model was log[As_grain] = 0.917 log[HNO₃-As] - 0.452 log[Fe_OX] - 1.507 (R² = 0.82, P <  0.001). Consistently, bioavailable arsenic and Fe_OX were also the key factors to predict arsenic accumulation in wheat straw, leaves and spikes. Our prediction models was successfully verified for three independent soils. Our results highlight the role of soil bioavailable heavy metals in predicting their transfer in soil-plant systems and can be used to improve existing Chinese soil quality standards.