Calculation of contaminant distribution in tree and ground nuts

Abstract

A comparison of two different calculational methods is presented to predict the sample distribution P(C;N) of granular products at any sample size, from an experimental distribution at a single N. The Negative Binomial density distribution can handle any N that is large with respect to the test size, but is limited to the precision of two parameter functions. Its application is simple and rapid, particularly at large λ, and useful for obtaining the lot average concentration, <c>. The non-parametric method can handle any approximation to the experimental data. It is treated as a sum of Poisson (or multinomial) distributions with varying λl =Npl , where pl is the probability of a kernel containing contaminant in a range of c. This method is currently restricted to λl <2, and is more laborious than the Negative Binomial. It is ideal for discerning small changes in P(C;N) at small N, indicative of the effects of production, processing and sorting. These methods were applied to the analysis of contamination in assorted processed and unprocessed pistachios, almonds, and peanuts. For this paper, the method is illustrated using aflatoxin contamination of various commodities, but the method can be applied to any contaminant of a granular commodity. The sparse approximation for distributions (λl <0.1), previously postulated, was verified. P(C;N) at low C was established in all cases, leading to a much clearer understanding of the source of contamination. The overall shapes of P(C;N) at low N were alike among tree nuts and among two very different peanut lots. Where available, test data at widely spaced N were consistent with calculated results. In one pistachio lot, the change of P(C;N) upon image sorting indicated a hitherto unknown contamination source. The methods show great applicability for risk analysis and developing sampling protocols.