IAG ring test feed composition 2016

Abstract

The analysis of composition in terms of ingredients is important for detecting economic fraud and for monitoring feed safety. Composition analysis and label control of feed is regulated in Regulation (EC) 767/2009. In a broader view, composition analysis in the entire food chain can improve the effect of monitoring actions. The new legislation on food labelling (Regulation (EC) 1169/2011), effective from December 13th 2014, obliges to provide more detailed information to customers on composition and related topics. A ring test was organized for the microscopic determination of botanic composition in animal feed in the framework of the annual ring tests of the IAG - International Association for Feeding stuff Analysis, Section Feeding stuff Microscopy. The organizer of the ring test was RIKILT - Wageningen UR, The Netherlands. The aim of the ring study was to provide the participants information on the performance of the local implementation of the method for composition analysis of feed. The sample was based on an artificially produced feed mimicking a ruminant feed, and distributed without label information. The participants were requested to produce a correct declaration of the ingredients of the sample. The results were analysed using the IAG model for uncertainty limits. Shares of ingredients in the feed formulation outside the limits of the model were indicated as underor over-estimations. A total of 25 sets of results was returned. The percentage of under- or over-estimations was 28.6% for the seven main ingredients. In the overview of results, the two declared wheat ingredients and the two declared corn products were pooled to one ingredient each. This was necessary since some participants declared a general ingredient (“wheat” and “corn”) and others a specific type (gluten or bran). The use of the original declarations would result in an extra number of non-matching estimations without precise justification. The share of the citrus pulp, in the presence of an equal amount of beet pulp, was underestimated or not detected in 44% of the results. Citrus pulp as such is recognisable as feed ingredient. Still almost three quarter of all estimations appeared to be correct in the ranges of the uncertainty model. This means that visual inspection of the composition of a sample can be used for label control and this method can support traceability of ingredients in case of an incidence. The current results indicate that specific formulations can influence the precision of the estimation of the composition of the feed. The current lack of a complementary system for (chemical) proximate analysis could be a drawback for the overall approach of supporting traceability, necessary for fighting food fraud and for supporting feed safety. Besides a proper method description and up-to-date descriptions of ingredients, well developed skills of technicians are vital for a good performance.