Accumulation of heavy metals in soils poses a potential risk to plant production, which is related to availability of the metals in soil. The phytoavailability of metals is usually evaluated using extracting solutions...Accumulation of heavy metals in soils poses a potential risk to plant production, which is related to availability of the metals in soil. The phytoavailability of metals is usually evaluated using extracting solutions such as salts, acids or chelates. The purpose of this study was to identify the most significant soil parameters that can be used to predict the concentrations of acetic and citric acidextractable cadmium(Cd), lead(Pb) and zinc(Zn) in contaminated woody habitat topsoils. Multiple linear regression models were established using two analysis strategies and three sets of variables based on a dataset of 260 soil samples. The performance of these models was evaluated using statistical parameters. Cation exchange capacity, CaCO_3, organic matter, assimilated P, free Al oxide,sand and the total metal concentrations appeared to be the main soil parameters governing the solubility of Cd, Pb and Zn in acetic and citric acid solutions. The results strongly suggest that the metal solubility in extracting solutions is extractable concentrationdependent since models were overall improved by incorporating a change point. This change point detection method was a powerful tool for predicting extractable Cd, Pb and Zn. Suitable predictions of extractable Cd, Pb and Zn concentrations were obtained, with correlation coefficient(adjusted r) ranging from 0.80 to 0.99, given the high complexity of the woody habitat soils studied. Therefore,the predictive models can constitute a decision-making support tool for managing phytoremediation of contaminated soils, making recommendations to control the potential bioavailability of metals. The relationships between acetic and/or citric acid-extractable concentrations and the concentrations of metals into the aboveground parts of plants need to be predicted, in order to make their temporal monitoring easier.展开更多
基金“Agence de l’Environnement et de la Maitrise de l’Energie”(ADEME)“Agence Nationale pour la Recherche”(ANR)for their financial support of the STARTT programmeADEME for the financial support of the PHYTENER programme
文摘Accumulation of heavy metals in soils poses a potential risk to plant production, which is related to availability of the metals in soil. The phytoavailability of metals is usually evaluated using extracting solutions such as salts, acids or chelates. The purpose of this study was to identify the most significant soil parameters that can be used to predict the concentrations of acetic and citric acidextractable cadmium(Cd), lead(Pb) and zinc(Zn) in contaminated woody habitat topsoils. Multiple linear regression models were established using two analysis strategies and three sets of variables based on a dataset of 260 soil samples. The performance of these models was evaluated using statistical parameters. Cation exchange capacity, CaCO_3, organic matter, assimilated P, free Al oxide,sand and the total metal concentrations appeared to be the main soil parameters governing the solubility of Cd, Pb and Zn in acetic and citric acid solutions. The results strongly suggest that the metal solubility in extracting solutions is extractable concentrationdependent since models were overall improved by incorporating a change point. This change point detection method was a powerful tool for predicting extractable Cd, Pb and Zn. Suitable predictions of extractable Cd, Pb and Zn concentrations were obtained, with correlation coefficient(adjusted r) ranging from 0.80 to 0.99, given the high complexity of the woody habitat soils studied. Therefore,the predictive models can constitute a decision-making support tool for managing phytoremediation of contaminated soils, making recommendations to control the potential bioavailability of metals. The relationships between acetic and/or citric acid-extractable concentrations and the concentrations of metals into the aboveground parts of plants need to be predicted, in order to make their temporal monitoring easier.
