State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, Beijing, China.
Sci Total Environ. 2012 Jul 15;430:126-31. doi: 10.1016/j.scitotenv.2012.04.071. Epub 2012 May 25.
Antibiotic residues in surface soils can lead to serious health risks and ecological hazards. Spatial mean concentration of antibiotic residues in the soil is the most important indicator of a region's environmental risk to antibiotic residues. Considerable estimation error would lead to an inefficient strategy of pollution control that happens when sample size is small and the estimation model does not match the spatial features of the object to be surveyed. On the basis of the available datasets, it was found that the distribution of antibiotics residue in soil follows a spatial stratification pattern. Accordingly, we used a new spatial estimation method called Mean of Surface with Non-homogeneity (MSN) to estimate antibiotic concentrations in surface soil of the Shandong Province, an important vegetable growing region in China. The standard error of the mean estimates obtained by MSN was significantly smaller (by about 1.02-6.82 μg/kg) than the estimation errors produced by three mainstream methods, simple arithmetic estimation (2.9-11.8 μg/kg), stratified estimation (2.5-10.6 μg/kg) and ordinary kriging estimation (2.2-8.2 μg/kg).
土壤中抗生素残留会导致严重的健康风险和生态危害。土壤中抗生素残留的空间平均值是衡量一个地区抗生素残留环境风险的最重要指标。当样本量较小时，且估计模型与被调查对象的空间特征不匹配，会导致估计误差较大，从而使污染控制策略效率低下。根据现有数据集，发现土壤中抗生素残留的分布呈空间分层模式。因此，我们使用了一种新的空间估计方法，称为非均质性表面均值（MSN），来估计中国重要蔬菜种植区山东省表层土壤中的抗生素浓度。MSN 得到的均值估计的标准误差明显更小（约 1.02-6.82μg/kg），比三种主流方法（简单算术估计（2.9-11.8μg/kg）、分层估计（2.5-10.6μg/kg）和普通克里金估计（2.2-8.2μg/kg）产生的估计误差小。
