Variability Associated with Chemically Testing Screened Farmers Stock Peanuts For Aflatoxin1

Abstract

Abstract Forty farmers stock lots of runner peanuts suspected of containing aflatoxin were identified by the Federal State Inspection Service using the visual Aspergillus flavus method. A 227-kg portion was removed from each of the 40 lots. Each 227-kg portion was screened over a belt screening device with 0.953-cm (24/64 inch) spacing to remove loose shelled kernels, foreign material, and small pods. Each screened portion was divided into ten 9.5-kg samples. Each sample was shelled, all kernels in the sample were comminuted in the Federal State subsampling mill, and the aflatoxin in duplicate 356-g subsamples per sample was extracted and quantified using HPLC methods. The total variability among the 10 aflatoxin test results was determined for each lot. The total variability was partitioned into sampling, subsampling, and analytical variability components for each lot. All variance components were shown to be functions of the aflatoxin concentration. Using regression techniques the functional relationship for each variance components and aflatoxin concentration was developed. The total variance associated with a 9.5-kg sample, 356-g subsample, and HPLC quantification when testing a screened farmers stock lot at 20 ppb is 295.2 and the CV is 89.5%.