16 October 1998

Tests now revealing minutest levels of GM matter in food

By Charles Abel

SCIENTISTS can now routinely check whether food products contain even minute amounts of genetically modified ingredients. But such testing is unlikely to resolve the issue of which foods should be labelled and which should not.

Testing is already so sophisticated it can detect as little as 0.001% contamination in raw ingredients, according to scientist Bert Popping of the CSL in Norwich.

Genetic sources

Accuracy is slightly lower in processed foods, due to the effects of processing. But the test can check the source of any genetic material which is found, he told last weeks AgraEurope Biotechnology in the Food Industry conference in London.

"Differences in the way novel traits are inserted into crops mean we can distinguish between the genes used by different companies for the same characteristic, such as herbicide tolerance," Dr Popping noted.

"Reliability is very good," he continued. "We use three separate laboratories to avoid contamination causing false positives. In ring trials we have had a 100% no mistakes record."

Cost of the test is £135/sample. But although the procedure takes just eight hours, results can take up to 10 working days due to the current workload.

More precise tests to identify smaller doses of GM material could be possible, Dr Popping noted. "But scientists agree that it makes no sense to adopt a lower threshold."

Finland already uses a 2% trigger for GM labelling, he noted. Durum wheat used for pasta making also has a 2% limit for non-Durum wheat content. "The technology is widely available for checking that sort of threshold – it makes sense."

Indeed, with current seed standards requiring 99.5% commercial seed purity it is illogical to insist upon a tighter standard for the produce which results, added John MacLeod director of the National Institute for Agricultural Botany.

A lower threshold would also risk widespread difficulties should low rates of inadvertent mixing occur, Dr Popping noted.

But Geraldine Schofield, head of regulatory affairs for food processing giant Unilever, felt labelling should go beyond thresholds. It should also reflect source.

Failure to do so could leave consumers confused, she said. A lack of clarity and uncertainty from policymakers and industry will slow sales, hinder free trade and inhibit technical development, she forecast.

Consumers may want to choose between food with GM material and food without, she acknowledged. But they may also wish to choose between food which used GM material in its production and food which did not.

Vegetable oil

That could mean labelling for vegetable oil which is extracted from GM rape, soya or maize seed, even though the oil contains no GM material.

"A negative list of food products from GM crops which contain no GM material and which need no labelling may be acceptable legally. But that has to be balanced with accurate labelling and that means giving consumers what they want."

Genetically modified processing enzymes, which are used at very low rates in many food processes, may soon appear on food labels for just that reason, she noted.

Similarly, meat and milk from animals which have been fed with feed containing GM products may also need labelling.

"There must be openness and clarity to achieve consumer confidence now. Only then, in three to five years time, will products with consumer benefits be accepted."

GMFOODTESTING

&#8226 Routine testing possible.

&#8226 0.001% accuracy.

&#8226 2% threshold preferred.

&#8226 Seed purity an issue.

&#8226 GM origin may over-ride GM content when labelling.