Investigating the possible relationship between pink grains and fusarium mycotoxins in 2004 harvest feed wheat

Abstract

In certain regions of the UK the 2004 wheat harvest produced an unusually high number of lots containing pink grains.  As some feed compounding companies were rejecting wheat that contained more than five visible pink grains per kilogram, unusually high levels of feed wheat were being rejected in autumn 2004. 

Pink grains are not necessarily associated with fusarium infections or with the mycotoxins that can be produced by some Fusarium species.  Consequently, with the higher than usual level of rejections of feed wheat due to the presence of pink grains, it was questioned whether it was appropriate to exclude 'pink grain' lots on the basis of the risk of mycotoxin presence?  The subject was discussed at a meeting called by AIC on 3rd September 2004.  The outcome was to request the HGCA to undertake fusarium mycotoxinanalysis of feed wheat samples containing pink grains to see if a relationship between pink grain number and deoxynivalenol (DON) existed.

Sixty-eight feed wheat samples with visible pink grains were submitted by AIC member companies and tested for both trichothecene mycotoxins and Fusarium species.  The aim was to obtain samples from around the UK roughly in the proportion to wheat acreage by region.  In the event, sample numbers were biased toward eastern regions where the problem was perceived to be worst.

The analyses showed that there was no relationship between pink grain number and either DON or Fusarium species.  The number of pink grains present in a sample does not allow any prediction of mycotoxin levels.  

DON levels ranged from 20 to 11,500 parts per billion (ppb).  Only 1.5% of samples exceeded 5,000 ppb.  The sample with the highest value was submitted as having too many pink grains to count.  When the results are considered in relation to the maximum levels for feed discussed at a Commission Working Group meeting in Brussels in mid September then only the single, highly contaminated, sample (11,500ppb) was above the maximum level discussed of 6,000-8,000ppb.

Other work, not focussed specifically on rejected feed wheat samples shows that Fusarium mycotoxin levels in 2004 were generally low and similar to previous years.  The level of fusarium ear blight (FEB) in 2004 was also normal when compared with previous years in the Defra winter wheat disease survey.

This short project must be considered with reference to the full R & D programme funded by HGCA, Food Standards Agency, Defra and LINK that is presently active in the area of FEB and associated mycotoxins.