Contribution of sampling to the variability of pesticide residue data

The uneven distribution of pesticide residues among the treated objects leads to an inevitable variability of pesticide residue levels measured in the samples, which may significantly contribute to the combined uncertainty of the analytical results. A total of 8844 unit-crop residue data derived from 57 lots and 19 field trials were evaluated to determine the characteristic features of residue distribution in unit crops and composite samples. The average residue levels and the corresponding coefficient of variation (CV) values obtained for individual units taken from a given lot showed wide variation from lot to lot. There was no significant difference between the CVs of residue levels in sample sets of various unit crops or composite sample populations of different sizes taken from various crops. The CV values for levels of residues taken from individual lots followed normal distribution. Very good correlation was found between the CVs of the parent and sample populations. The experimentally obtained values were very close to those expected on the basis of the central limit theorem. The estimated typical relative standard uncertainties of sampling medium-size crops for pesticide residue analysis in the cases of sample sizes of 5, 10, and 25 were 37, 25, and 16%, respectively.