Improving risk assessment to minimise fusarium mycotoxins in harvested wheat grain


Cereals & Oilseeds
Project code:
01 July 2006 - 30 June 2009
AHDB Cereals & Oilseeds.
AHDB sector cost:
£173,733 from HGCA project (RD-2006-3288)
Project leader:
S. G. Edwards Harper Adams University College, Newport, Shropshire TF10 8NB


pr477-final-project-report pr477-abstract-and-summary

About this project


European legislative limits for Fusarium mycotoxins, deoxynivalenol (DON) and zearalenone (ZON) were introduced in 2006 for cereals and cereal products for human consumption. Fusarium mycotoxins were monitored in UK wheat from 2006-2008.

There was a large variation in levels detected each year. Lowest levels were detected in 2006, which had a dry summer. Highest levels were detected in 2008, which had a wet summer and, in particular, a delayed wet harvest. ZON was particularly high in 2008 with 29% of wheat grain samples at harvest exceeding legal limits.

This caused major issues for the cereal processing industry, particularly for the supply of bran for human consumption, as ZON concentrations were highest in this mill fraction.

Modelling the mycotoxin concentrations against the known agronomy of each sample produced similar models for DON and ZON. The majority of the variation between samples was explained by year and region, indicating that weather is the main factor affecting the concentration of these toxins.

Previous crop, cultivation and variety were also significant factors within the models. Additional factors of cereal intensity within the rotation and crop debris management were found to be not significant.

The HGCA fusarium mycotoxin risk assessment was validated/amended based on each year's results. The main modification to the risk assessment was the increase in the number of risk categories for rainfall at flowering and pre-harvest, in light of the high mycotoxin levels experienced in 2008.

To improve the predictive ability of the mycotoxin models further would require more precise weather date. The variance accounted for by year and region is largely attributable to differences in climate at specific crop growth stages.

The inclusion of weather parameters from pre-flowering to harvest is likely to account for much of this variance. The access to meteorological data, the necessary model development and software development and maintenance will need to be considered by HGCA and other relevant government and industry bodies.