摘要
A generally accepted method for predicting the bioavailability and transfer of cadmium(Cd) from soil to plants has not yet been established. In this study, the diffusive gradients in thin films(DGT) and conventional extraction methods for metal fractions were used to investigate how effective these methods were at predicting Cd bioavailability to ryegrass(Lolium perenne) and to assess whether these holistic variables were independent of the soil property variables. The influences of soil properties on the predictive capabilities of the different methods were evaluated using multivariate statistics. The Cd concentrations in the shoots and roots of ryegrass correlated more closely with the effective Cd concentration measured by DGT(CE), compared with those with soil total Cd, pore water Cd, and ethylenediaminetetraacetic acid disodium salt(Na2EDTA)- and acetic acid(HOAc)-extractable Cd. When soil properties were included in a stepwise multiple linear regression, Cd transfer to L. perenne was negatively influenced by p H and positively influenced by organic matter(OM). The multivariate statistics showed that the adjusted correlation coefficients for the plots involving soil total Cd, pore water Cd, and Na2EDTA- and HOAc-extractable Cd had been significantly improved after considering the influences of soil properties, which suggested that these methods were p H and OM dependent. For CE, the relationship was p H independent or only dependent in strongly acidic soils. These results suggested that the main soil factors affecting bioavailability had the lowest impact on the DGT technique. Therefore, the DGT technique provided significant advantages over conventional methods when assessing Cd bioavailability.
A generally accepted method for predicting the bioavailability and transfer of cadmium （Cd） from soil to plants has not yet been established. In this study, the diffusive gradients in thin films （DGT） and conventional extraction methods for metal fractions were used to investigate how effective these methods were at predicting Cd bioavailability to ryegrass （Lolium perenne） and to assess whether these holistic variables were independent of the soil property variables. The influences of soil properties on the predictive capabilities of the different methods were evaluated using multivariate statistics. The Cd concentrations in the shoots and roots of ryegrass correlated more closely with the effective Cd concentration measured by DGT （CE）, compared with those with soil total Cd, pore water Cd, and ethylenediaminetetraacetic acid disodium salt （Na2EDTA）- and acetic acid （HOAc）-extractable Cd. When soil properties were included in a stepwise multiple linear regression, Cd transfer to L. perenne was negatively influenced by pH and positively influenced by organic matter （OM）. The multivariate statistics showed that the adjusted correlation coefficients for the plots involving soil total Cd, pore water Cd, and Na2EDTA- and HOAc-extractable Cd had been significantly improved after considering the influences of soil properties, which suggested that these methods were pH and OM dependent. For CE, the relationship was pH independent or only dependent in strongly acidic soils. These results suggested that the main soil factors affecting bioavailability had the lowest impact on the DGT technique. Therefore, the DGT technique provided significant advantages over conventional methods when assessing Cd bioavailability.
基金
supported by the China Postdoctoral Science Foundation (No. 2013M530783)
the Research Foundation for Advanced Talents of Qingdao Agricultural University, China (No. 6631115029)
