Monitoring contaminants in wheat grain (2002–06)


Cereals & Oilseeds
Project code:
01 October 2002 - 30 April 2006
AHDB Cereals & Oilseeds.
AHDB sector cost:
HGCA (£121,479, Project No. 2747), HGCA levy payers (£45,000, in-kind)
Total project value:
Project leader:
Sue Salmon at CCFRA (Camdpen BRI)


PR386 Final Project Report

About this project


The overall aim of this project was to monitor the safety and quality of wheat entering UK flour mills. This project covered three calendar years i.e. from January 2003 to December 2005. However, due to the importance of climate conditions during crop growth, results are expressed in terms of harvest years. For simplicity, August intakes are considered to be new crop wheat and as a result, this report includes data for the following periods: 2002 crop (January-July 2003); 2003 crop (August 2003-July 2004); 2004 crop (August 2004-July 2005); and 2005 crop (August-December 2005).

The project involved collection of representative sample sets of wheat entering UK flour mills. Most wheat samples were UK-grown (93%), some were imported wheat and included samples of Canadian, French, German, American and Spanish origin.

Samples were analysed routinely for the mycotoxin ochratoxin A (OTA) and organophosphorus storage pesticides. Specific analyses were carried out to investigate the levels of the following contaminants:

  • Fusarium mycotoxin analyses, for tricothecenes plus zearalenone, were performed on freshly harvested wheat taken from the first 4-5 months of the new crop year (August - December) for the 2003, 2004 and 2005 harvest material.
  • Analysis, after harvest, of UK-grown wheat only for the growth regulator chlormequat was carried out on samples from the 2004 and 2005 harvests.
  • Analysis, after harvest, of UK-grown wheat only for the broad leaf weedkiller glyphosate was carried out on samples from the 2004 harvest.
  • Samples of UK-grown wheat from the 2003 harvest were subjected to analysis for the heavy metals arsenic, cadmium, lead and mercury.
  • In summary, the results indicated that the overwhelming majority of wheat accepted by UK flour millers was legally compliant and safe for use.

Over the period January 2003-December 2005, only 18.9% of samples contained detectable levels of OTA, i.e. at or above the Limit of Quantification or LOQ of 0.1ppb. This means that more than 80% of samples surveyed contained no detectable OTA. In harvest years 2002 and 2004, where higher percentages of positive OTA were recorded, 25-50% of positive samples were reported as being on the LOQ. 

Low mean OTA values of between 0.27 and 0.44ppb were recorded across four very different harvest years. A small percentage of wheat samples (less than 2%) were found to have OTA levels above the EU Action Limit of 5ppb. Regional biases in OTA level tend to be related to growing and harvesting conditions in specific crop years and the data suggested that millers tended to adjust their buying decisions to take account of known regional issues.  Length of storage only appeared to have an impact on the occurrence and concentration of OTA in the 2002 and 2004 harvest material.

Surveillance data indicated that nearly 90% of samples surveyed contained no detectable pesticide residues and that no sample contained residues of any OP pesticide above the Maximum Residue Level (MRL) Pirimiphos-methyl was consistently identified as the main OP pesticide detected. Chlorpyrifos-methyl occurred infrequently.  Malathion was detected in two samples from the 2002 and 2003 harvests only and in both cases, the samples originated from France. The survey data shows that approximately one third of the samples that were declared as treated with post-harvest pesticides contained detectable pesticide residues.

In terms of the trichothecene mycotoxins, deoxynivalenol (DON) was fairly ubiquitous in UK wheat being detected at levels above the LOQ in 88-91% of samples from all three harvest years (2003 - 2005). However, in only one sample out of the 155 tested was DON detected at a level that would exceed the forthcoming legislative limit for wheat grain of 1250ppb. Interestingly, DON levels did not increase in samples from the difficult 2004 harvest suggesting that, even under these difficult conditions, UK millers were able to source wheat of suitable quality for the production of bread, biscuit and household flour.

The second most frequently detected Fusarium mycotoxin was nivalenol (NIV), but the data suggest that the occurrence of this trichothecene is declining in UK wheat destined for the flour milling market. The incidence pattern for T-2 and HT-2 toxins differed between the three crop years. T-2 toxin was most prevalent in 2004 when it was detected in 70% of samples whilst the highest incidence of the HT-2 toxin occurred in 2003 where this toxin was detected in 75% of the samples tested.

Zearalenone was not detected in the majority of wheat destined for use by the UK milling industry. Incidence levels varied between 26.7% and 34% over the three very different crop years (2003-2005).

Chlormequat residues were detected widely in UK-grown wheat samples: it was found in between 80 and 95.6% of samples from the 2004 and 2005 crops respectively. However, the levels detected were very low and significantly below the MRL for wheat for this product.

Only 32% of the samples under test contained glyphosate residues and all samples contained significantly less than the MRL.

Arsenic and heavy metal levels were found to be low in all samples of UK wheat destined for the human food chain from the 2003 harvest.

This surveillance work has been replaced by a new project, (HGCA 3100), which covers wheat and barley entering both the human food chain and the animal feed sector. This has the advantage of facilitating cross sector comparison and discussion in relation to common contaminant issues.