摘要：
The study investigates the bioavailability of selenium(Se) in potentially Se-enriched soils. A total of 64 paired samples of rice and the corresponding root soils were collected from Lujiang County, Anhui Province, China. A linear regression was used to study the parameters that are best related to the bioavailability of Se, including physical and chemical properties, nutrient abundances, Se concentrations,distribution of Se in soil samples,and Se concentrations in the associated rice. The results indicate that the rice the study area contains 0.037-0.120 mg∕kg (Se),whereas the associated soil samples contains 0.260-1.177 mg∕kg (Se). Se in the soil samples was mainly distributed between humic acid bound phases (15.5%-31.5%), strongly bound organic phases (12.5%-38.6%), and residual phases (8.1%-68.5%). Increasing abundance of effective sulfur (S) and phosphorus (P) positively correlated with the concentrations of Se in rice as a result of an increase in the abundance of Se in strongly bound organic phases,humic acid bound organic phases,and the extractable concentrations of Se in soil. In contrasts,increases cation exchange capacity(CEC) and the concentrations of available K,Fe and Mn in soil inhibited the uptake of Se by rice as a result of the reduction in the concentration of available Se in soil, primarily as this Se was present in soils in the form of humic acid bound Se,strongly bound organic phase Se,extractable Se,and water-soluble Se. Soil Se concentrations have a strong positive correlation with rice Se concentrations (R=0.616, P<0.01). In addition, the concentration ratios of non-residual phase Se and organic matter (NRE-Se∕OM) also strongly positively correlate with rice Se concentrations (R=0.774, P<0.01).This study used a multiple stepwise lineal logarithmic regression approach combined with NRE-Se∕OM values,effective S and Fe soil concentration data,and the total Mn soil concentration data to generate a model for Se uptake by rice,with some 76.0% of the variance in the model being explained by variations in these values. This means that soil NRE-Se∕OM, effective S and Fe,and the total Mn values can be used to effectively evaluate soil Se bioavailability.%为研究潜在富硒土壤硒生物有效性,采集安徽庐江潜在富硒地区64组水稻及对应根系土壤样品,分析土壤理化性质、土壤养分、全硒、不同形态硒及大米硒含量,采用线性回归法探究土壤硒生物有效性的影响因素.结果表明,大米硒含量(记作Serice,下同)为0.037~0.120 mg∕kg,土壤全硒含量(记作Setot,下同)为0.260~1.177 mg∕kg,土壤硒以腐殖酸态硒(15.5%~44.7%)、强有机态硒(12.5%~53.3%)和残渣态硒(8.1%~68.5%)为主.土壤有效硫、有效磷通过提高强有机态、腐殖酸态和浸提性硒含量并促进水稻籽实(大米)对硒的吸收,而土壤阳离子交换量(cation exchange capacity, CEC)、速效钾、全铁和全锰通过降低水稻生物可利用态硒含量(包括水溶态、腐殖酸态、强有机态和浸提性硒)抑制水稻籽实(大米)对硒的吸收.Setot与Serice显著正相关(R=0.616,P<0.01),非残渣态硒含量与有机质含量之比(记作NRE-Se∕OM)与Serice相关性最强(R=0.774,P<0.01).土壤NRE-Se∕OM、有效硫、有效铁、全锰构建的多元逐步线性对数回归模型可解释水稻籽实(大米)吸收硒的76.0%的方差.因此引入变量NRE-Se∕OM,并结合土壤有效硫、有效铁和全锰能有效评价和预测研究区土壤硒生物有效性.
