1. IAG proficiency test feed composition 2019
- Author
-
van Raamsdonk, L.W.D., Smits, C.P.A.F., Hedemann, B., Vliege, J.J.M., van Raamsdonk, L.W.D., Smits, C.P.A.F., Hedemann, B., and Vliege, J.J.M.
- Abstract
A proficiency test was organized for the microscopic determination and semi-quantification of botanic ingredients in the formulation of an animal feed, in the framework of the annual proficiency tests of the IAG - International Association for Feeding stuff Analysis, Section Feeding stuff Microscopy. The organizer of the proficiency test was Wageningen Food Safety Research, The Netherlands. The aim of the proficiency study was to provide the participants information on the performance of the local implementation of the method for composition analysis of feed. The current proficiency test was focusing on the control of the label declaration of the botanic composition of a ruminant feed. Participants received a sample of the ruminant feed and were asked to check the correctness of the label information. Therefore, the label declaration was provided together with the sample. The formulation as declared by the label showed the correct composition. Results should show the share of the different ingredients in percentages. Indicated shares were considered under- or overestimations when exceeding the limits of the IAG uncertainty interval model.A total of 22 sets of results was returned. Six of the nine ingredients had shares with intervals of which the lower limits were still higher than zero percent in the IAG estimation model. Correct estimation of these shares needs more precision than in a situation where only an upper limit applies. The reported shares of these six ingredients were within the limits of the uncertainty model in 87.9% of the total number of estimations. The correct estimations of all reports of the shares of nine ingredients is 80.3%. Six out of 22 participants delivered an errorless composition, which is 27%. Besides this, six participants made one error, three made two errors and two participants made three errors. There is no clear correlation with the method applied. Only three participants reported an indication of the correctness of the label. Th
- Published
- 2020