Practical strategies for minimising the production of Ochratoxin A in damp cereals

Abstract

Ochratoxin A (OTA) is formed when grain moisture content (m.c.) exceeds 18%.  EU food limits are set at 5ppb with feed limits likely to be 20-fold higher. About 20% of the British crop is harvested at above the m.c. limit for OTA formation and a similar proportion of the crop has detectable levels of OTA although less than 1% exceeds the food limit.

It was considered that the highest risk of OTA formation was in the ambient air drying process when grain above the drying front may remain at its initial m.c. until the drying front passes through, after 10-14 days. This project therefore assessed the risk of OTA formation during drying using the simulation tool 'Storedry', substituting the old biodeterioration criteria of visible mould or germination loss with a new model for OTA formation and fungal growth, based on the time taken before fungal growth enters the logarithmic phase.

Initial comparison indicated shorter safe storage times using the new fungal model.  For instance at 20°C and 20% mc, 10 days were available for drying based on germination loss and visible fungi while there were only 6.5 days based on OTA.  Simulations were run of a 3.0 m deep, 110 m2 bed of 250 t wheat ventilated at 0.05 m3/s.t using the two criteria, 5 locations, 20 years, 4 m.c.s, 2 start dates and 7 drying strategies of fan and heater control.  Of those treatments successfully dried using the old criteria, only 2/3 succeeded using the new spoilage model, so spoilage by OTA in near ambient drying was potentially a serious issue.

A useful measure of drying performance was the maximum bed depth which would allow drying success in all 20 years simulated.  When comparing drying using the new and old criteria, a reduction in bed depth of about 1 m was required to ensure success even in the worst drying year.  For instance when drying grain at 20% mc, running fans continuously and switching heat in at 80% r.h., 3.2 m depth was successful using the old criteria but only 2.3 m was permissible with the new spoilage model.

Many simulations were then run to see how far performance of existing systems could be improved by changing the fan/heater control strategies, with some success.  For example, changing the r.h. at which fans were switched on and off in one of the strategies gave bed depths that were that were 112% higher, costs were 10% lower, drying times were 6% longer.  A larger fan allowed an extra 0.2 m depth.  Even so, a reduction from current bed depth of as much as 1 m would be needed to avoid risk of OTA in poor drying seasons.

The simulation was adapted to explore grain stirring and validated against published results.  Stirring allowed considerably greater depths to be used, ranging from 0.8 m deeper at 22% initial m.c. to 2 m deeper at 20%.  Drying with higher air temperatures gave further gains.