Significant changes have been made to the HGCA mycotoxin risk assessment to rebuild industry confidence in its assessment of levels in wheat.

The changes will allow greater scores to be attached when heavy rain falls pre-harvest or during flowering. It should prevent high mycotoxin grain being categorised as low to moderate risk, but might mean more low mycotoxin grain being categorised as high risk, according to Simon Edwards of Harper Adams University College.

The risk assessment did not work well last season, he acknowledged at a BASF press briefing last week – something highlighted in a BASF-commissioned farmer survey. Growers found it easy to use, but had found reliability and trust of its results an issue for millers and merchants, the firm’s Graham Hartwell said.

Dr Edwards put forward two possible reasons for its failure: One, the wet, delayed harvest had a major impact on mycotoxin levels, and that such a delay had not been experienced from 2001 to 2008 when the data used to produce the assessment was obtained; two, that the assessment was completed incorrectly.

“I’m fairly sure both occurred, but the wet, delayed harvest was a major issue. We hadn’t experienced a wet harvest like that before.”

The result was too many samples giving a false negative result – loads the risk assessment said were low or moderate risk actually having high mycotoxin levels, he explained.

To counter that additional risk levels have been added to take account of pre-harvest rainfall. Previously the risk assessment had just two levels for the factor – less than 20mm rain, or more than 20mm rain, he said. “Last year the average was close to 100mm. The highest was 340mm, which makes 20mm look rather inadequate.”

In addition, an additional risk level has been added for rainfall during flowering. “It is to future-proof the risk assessment. We haven’t really had a really wet flowering period either, and I would hate for it to happen this year [and have more problems].”

Risk scores for northern England and Scotland have also been reduced as mycotoxin levels remain low even when high rainfall occurs. “I put that down to fewer cereals in the rotation and lower temperatures.”

The new risk assessment had been tested with samples from last season, he said. “It resulted in no false negatives, but did mean there were 36% false positives – samples that had low mycotoxin levels but the risk assessment said were high risk and needed testing.

“It has made the risk assessment more risk-averse, but feedback from NFU, AIC and Nabim is that they would prefer that the risk assessment worked than having to test every consignment.”

The new risk assessment will be detailed in a HGCA topic sheet in June, although the risk assessment is already available on its website.

“I hope the industry will revert to using the risk assessment, which will reduce the amount of testing required,” Dr Edwards said. “The NFU and AIC are keen to do so. Nabim and the breakfast cereal manufacturers are more in a wait-and-see mode. They will want to do some testing early in harvest, but hopefully then they will see the risk assessment is working and will start using it again.”