An alternative to slurry mixing to minimise sample preparation variance for determination of ochratoxin A in wheat

Abstract

Many countries have established maximum limits for ochratoxin A (OTA) in cereal grains and implemented surveillance programs for OTA in wheat shipments. For shipment to some countries, certification for OTA content is mandatory. These control activities require the capability to measure OTA in bulk grain shipments with accuracy and precision. It is known that the nugget effect caused by the heterogeneous nature of mycotoxin contamination in agricultural commodities creates major challenges for generating representative test results due to the potential for high variances in the sampling / sample preparation phases of the analytical process. The water-slurry mixing approach to sample preparation has greatly minimises variances associated with this phase of the analytical process, but this is not a practical technique for all laboratories. The potential magnitude of variances for subsampling raw and ground grain for the dry-milling approach to sample preparation and the means for reducing variances to acceptable levels is not fully understood. We investigated the repeatability of OTA measurements in 2 kg laboratory samples subsampled from 20 kg samples of raw wheat and in 100 g test portions subsampled from 2 kg of ground wheat. In addition, the effect of mixing time on the repeatability of OTA results was investigated prior to subsampling to obtain test portions for analysis. Results show that for subsampling of a primary sample of raw wheat using a conventional sample divider the variability of OTA results decreases with increasing weight of the laboratory sample relative to the weight of the primary sample. In order to improve repeatability, the proportion of primary sample separated out to produce a laboratory sample should be as large as operationally feasible and ideally about 50% of the weight of the primary sample. Four factors are identified for separating out a test sample from a raw wheat laboratory sample.