Wheat (Triticum aestivum L.) grains uptake of lead (Pb), transfer factors and prediction models for various types of soils from China

Lead (Pb) contaminated in farmlands has become a deep threat to global food security and human health. In this study, the bioavailability of Pb in 18 types of soil to wheat (Triticum aestivum L.) grains were investigated, and reliable empirical models of Pb in wheat grains were established based on soil properties. The results showed that the average bioconcentration factor (BCFgrain/total-Pb) in acidic soils was approximately 3.30 times than that in alkaline soils (ANOVA P < 0.05). Significant positive relationships between wheat grain Pb concentration and soil total Pb or EDTA extractable Pb were presented through the results of simple linear regressions (P < 0.001). The stepwise multiple linear regression models indicated that soil pH and soil total Pb were determined to be the two most reliable and reasonable factors in predicting wheat grain Pb concentration, with 83.8% explanation of variation. Soil total Pb compared with EDTA extractable Pb was applied to better improve prediction models in describing Pb transfer from soils to wheat grains. Furthermore, grouped models divided into two parts with pH of 7.5 also generated well prediction in wheat grain Pb concentration. Our prediction models were successfully verified within 95% prediction intervals for published literature data (including other wheat varieties). Moreover, the results indicated that ungmuped models performed better in predicting accuracy within 400 mg kg(-1) of soil total Pb, and grouped models showed better extrapolation stability when Pb in soil were overly high. Our results in the study were conduce to evaluate food security of Pb in contaminated agricultural soils.