Effects of long-term storage on Fusarium toxin concentrations in wheat--sources of error of the analytical results.

The concentrations of the Fusarium toxins deoxynivalenol (DON) and zearalenone (ZON) of a heavily contaminated wheat grain batch were followed over a period of 1 year by taking samples 15 times every 28 days. The air temperature and relative humidity at the top of the wheat batch ranged between 7 and 22 degrees C and 44 and 55%, respectively, and corresponded to a variation in the moisture content of the wheat grain between 11.5 and 12.3%. None of these fluctuations were related to ZON and DON concentrations, which varied between 0.46 and 0.66 and 15.0 and 19.5 mg/kg DM. Therefore, the data were used to analyse the error sources for the analytical results. It was found that the variance proportions due to sampling and sample preparation plus analysis were not similar for DON and ZON. The variance proportion due to sampling was found to be 0.62 for ZON, which corresponded to a variance proportion of 0.38 due to sample preparation plus analysis. In contrast, the latter variance proportion for DON was estimated to be 1.0 and consequently completely superimposed the sampling error. It is concluded that long-term storage of contaminated wheat grain does not affect the concentrations of DON and ZON considering the measured fluctuations in ambient temperature, relative humidity and moisture content of the grain. Therefore, no degradation of DON and ZON occurred during the storage of wheat for a period of one year under ambient conditions. The effects of sampling and sample preparation plus analysis on the final analytical results are different for DON and ZON and require further consideration.