Screening of extraction methods for Cd and As bioavailability prediction in rhizospheric soil using multivariate analyses

Abstract

Heavy metals in tailings and mining wastes from abandoned mines can be released into adjacent agricultural field and bioaccumulated in crops or vegetables. Therefore, prediction of metal bioavailability has become an important issue to prevent adverse effect of bioaccumulated metals on human health. In this study, single and sequential extraction methods were compared using multivariate analysis to predict the bioavailability of Cd and As in contaminated rhizosphere soils. Single extraction using 0.1 M HCl for Cd and 1.0 M HCl for As had an extraction efficiency of 8–12% for soil Cd and 14–17% for soil As compared to total concentration extracted with aqua regia. Using sequential extraction, Fe–Mn-bound Cd (FR3) and residual Cd (FR5) were the dominant fractions representing 43 and 41% of total Cd concentration. For As, the strongly absorbed form (FR2) was the most abundant chemical fraction showing 45–54% of the total As concentration in soil. Multivariate analyses showed that single extraction with HCl and total concentration of Cd and As in soil were significantly correlated to potato and green onion plant tissue metal concentration. Although little information was obtained with multiple regression analysis because of multicollinearity of variables, the result of principle component analysis (PCA) revealed that the highest positive loading was obtained using total concentration of Cd and As in soil in the first principle component (PC1). In addition, total concentration of Cd and As in soil was independently grouped with other chemical fractions by cluster analysis. Therefore, the overall result of this research indicated that total concentrations of Cd and As in rhizosphere soils were the best predictors of bioavailability of heavy metals in these contaminated soils.