Arpad Ambrus
National Food Chain Safety Office
Hungary
Title: Factors affecting the accuracy of measurements of pesticide residues and food contaminants
Biography
Biography: Arpad Ambrus
Abstract
Unit to unit (beetween oranges) and within unit (peel and pulp) variability of pesticide residues result in inevitable variation of average residues in test portions taken from the comminuted laboratory sample for extraction. Sub-sampling of large crops (e.g. watermelon, jackfruit) will further increase the variability. The patchy distribution of mycotoxins in contaminated crops has the same effect at even larger magnitude. To obtain representative test portions of 2-25 g from 1 to 80 kg bulk laboratory sample is a very difficult and challanging task. Nonetheless, it is very rarely tested, though. Neither recovery nor proficiency tests provide information on the efficiency of sample size reduction and comminution of the laboratory sample. Particle size of comminuted material is critical for reducing the relative variability of analytes, expressed as coefficient of variation (CVSp), in the test portion. The relationship of CVSp, the mass of test portion, mTp and the the upper 95th percentile of the diameters of the particles, d, in the comminuted matrix is described by Gy’s sampling theory.
The efficiency of comminution of laboratory sample depends on the equipment, temperature of processing, type and maturity of sample materials, therefore it shall be verified as part of validation of methods and regularly tested during internal quality control of the determination of food contaminants. Test portion size can only be reduced if the combined uncertainty of the results remain acceptable. Examples for testing the efficiency of sub-sampling and comminution will be presented in the lecture.rn
