Sampling and improvement of mycotoxin analysis: Panel discussion

Abstract

In this panel discussion the following topics were discussed: SAMPLINGThe underestimation of analytical results due to sampling errors was found to have a strong influence on the consumer protection and/or industry liability because sampling errors are larger than analytical errors.It was stressed that the analyst should be aware of the sampling problems. These problems are different for various mycotoxins. However, it is possible to learn about the estimation of sampling errors from the approach which has already been used for the sampling of peanuts and corn-case.Sampling of large lots of agricultural products was found to be a basic area of research. Any research work in the area of sampling should therefore lead to good sampling procedures. A good sampling procedure can be characterised by taking enough samples (including retail samples) to allow later confirmation or further control of the sampling process by additional or further measurements after a certain time interval, providing sufficient information on the concept of distribution, demonstrating vigilance, and allowing the monitoring of the end-product.In addition, future sampling procedures need to recognize the fact that naturally contaminated feed is being used in industrial processes such as fermentation, detoxification (with clay, chemicals, etc.), and washing, etc., where highly contaminated materials are being produced and precautions for health care of workers have become extremely important. Sampling protocols need to be developed for contaminants (including mycotoxins) in heterogeneous materials such as patulin in apples, fumonisins in grains, and/or toxic plant components and are therefore important topics for future research activities. The demand of future sampling procedures for fumonisins should be focussed on matrices such as corn grits, flour, polenta, and extruded corn. The co-occurrence of zearalenone and fumonisins should also be taken into account when sampling procedures are developed for fumonisins in corn based products, because both mycotoxins are being produced by the same fungi. The heterogenous occurrence of moulds, sometimes present in "pockets" in bulk lots and the variability in mycotoxin production by moulds resulting in variable mycotoxin contents on individual kernels, forms a necessity to develop sampling procedures. The results obtained for aflatoxins might be a useful model in this case (see FNP 55 from FA0 sampling plans for aflatoxin analysis in peanuts and corn, ISSN 0254-4725). Regulatory activities in the area of sampling and analysis, such as the "code of practice" (for sampling) or a "convenant" (for analytical procedures) containing sampling procedures and analytical criteria are currently in force and/or drafted in The Netherlands to allow control measurements of industrially produced food and feed. This approach is compulsory to manufacturers and enables the controller to trace back the primary source of contaminated materials. ANALYSISThe following future research activities related to the analysis of myc...