收集现有文献资料,统计土壤、稻米/小麦籽粒As、Cd、Pb含量的污染数据,将数据分为污染调查数据和添加盐作物栽培试验两类,分别计算其作物富集系数(PUF),并对PUF的分布特征及规律进行分析和总结.结果表明,污染调查数据计算稻米对As、Cd、Pb的PUF中值和全距分别为:0.026(0.004—0.090)、0.150(0.014—1.470)、0.005(0.001—0.031);添加盐试验数据计算稻米对As、cd、Ph的PUF中值和全距分别为:0.010(0.003~0.033)、0.360(0.056—1.700)、0.002(0.001—0.019).污染调查数据计算小麦籽粒对As、cd、Pb的PUF中值和全距分别为0.010(0.001—0.110)、0.190(0.030—2.110)、0.017(0.001~0.075);添加盐试验数据计算小麦籽粒对As、cd、Pb的PUF中值和全距分别为:0.010(0.003~0.028)、0.150(0.055~0.730)、0.001(0.001—0.014).添加盐试验数据与污染调查数据得出的PUF跨幅有明显差异.PUF数值受土壤污染程度、环境条件及作物本身特性等因素的多重影响而不易精准预测,但ln(PUF)符合Gaussian概率分布(R^2在0.38~0.94之间).PUF概率模型用于一般风险评估及土壤环境基准制定的保守风险概率计算的条件是,目标污染场地污染程度等参数取值必须在PUF概率模型推导时对应参数取值范围以内,且土壤基本性质及作物生长的环境条件较为一致.
The statistical characteristics of plant uptake factor (PUF) of As, Cd and Pb for rice/wheat cereal were analyzed and summarized based on collected data in China. The data were selected and distinguished between field investigation and glasshouse experiment in which metal salts were added to clean soils as grown media. Results showed that the PUF median and range of As, Cd and Pb for rice cereal were 0.026 (0.004-0.090), 0.150 (0.014-1.470), 0.005 (0.001-0.031) and 0.010 (0.003-0.033), 0. 360 (0. 056-1. 700), 0. 002 (0. 001-0. 019) for field investigation and salt added experiment respectively. The PUF median and range of As, Cd and Pb for wheat cereal were 0.010 (0.001-0. 110), 0, 190 (0.030-2. 110), 0.017(0.001-0.075) and 0.010 (0. 003-0. 028) , 0. 150 (0. 055-0. 730) , 0. 001 (0. 001-0. 014) for field investigation and salt added experiment respectively. Significant PUF distribution differences were found between field investigation and salt added experiment. PUF provided not exactly predictive ability of metal uptake by crops since the PUF values were influenced by a variety of factors including the soil contamination levels, basic soils characteristics, the cultivation of the crops and other environmental conditions. However, the In(PUF) followed Gaussian distribution (R^2 = 0. 38 - 0. 94) which is useful for general risk assessment and soil benchmark derivation of contaminated croplands. Use of these statistical models is restricted to the range of data they have been derived, as extrapolation outside this range is often unreliable. Soil properties and surrounding environment of the crops should also be in accordance with the conditions PUF derived.