Food: How to Detect GMO?
    Le Monde

    Friday 13 October 2006

    Does the agro-food industry know how to control GMO? The question presents itself as the General Directory for Competition, Consumption and Fraud Repression (DGCCRF) publicized its annual audits account on Wednesday October 11. In 2005, out of 69 food product samples taken, 17 contained traces of genetically modified organisms. Their quantity was, however, inferior to the regulatory threshold of 0.9%. That publication comes at a time when transgenic rice had been discovered this summer in trade channels, as the DGCCRF confirms. But this episode is only the most recent in a series that began several years ago.

    In 2000, transgenic rapeseed seed that had not been officially recognized was sold in France by Advanta, as was GMO soy seed from Asgrow, a Monsanto subsidiary. In 2001, a corn called Starlink, prohibited for consumption, was found in chips in the United States, which provoked a massive product recall.

    In April 2005, Bt corn (as in Bacillus thurengiensis, the bacteria from which the introduced gene came) not authorized in Europe was imported here; while in China, transgenic rice was detected in foodstuffs.

    Finally, last July, a variety of rice produced by Bayer that is not officially recognized turned out to be mixed in with conventional rice: since then multiple controls have revealed contaminations in several European countries, including France. In the beginning of September, Greenpeace and Friends of the Earth announced in their turn that analyses of rice-based food products imported from China all demonstrated traces of contamination.

    These slip-ups pointedly pose the question about GMO detection. For authorized transgenic plants, things are relatively simple. Since 2004, European regulation obliges GMO producers to communicate elements that allow these products to be identified once they're commercialized. "Before, laboratories had enormous problems overseeing the market," remembers Guy van den Eede, Head of the European Commission's Laboratory, which develops detection protocols in Ispra (Italy). "We had to negotiate with firms for years to obtain that information."

    Witness the confusion between two types of corn with "insecticide" properties that are very close, Syngenta's Bt 10 and Bt 11, one authorized, the other not. Between 2001 and 2005, 150,000 tons of illegal cereals were produced on the other side of the Atlantic, a part of which was exported to Europe before the deception was revealed.

    From now on, every variety must be identifiable. Faced with a product of indeterminate composition, such as flour, the controllers can, through spotting of DNA sequences specific to each species, verify, for example, whether there is only corn or whether traces of soy are also there. They research the presence of GMO in each sample, thanks to specific sequences called "promoter" or "terminator" sequences.

    The investigations can be pushed still further and the variety of GMO itself identified if the sequences furnished by the producers are available. Quantification is the final stage of the tests, since above a 0.9% threshold of fortuitous GMO presence, the food product must be labeled as GMO.

    For highly processed products such as oils and flavors for which there is no DNA, that quantification is impossible. The proof of the absence of GMO then becomes administrative, through the implementation of traceability of the basic products used, that also in the framework of a European Directive.

    In France, several public laboratories are charged with control. That of the DGCCRF, situated in Strasbourg, effects about 350 tests a year. "It's not much, but we essentially target primary ingredients so as to be as much upstream of any potential contamination as possible," explains Patrick Philippe, who is responsible for the biological team at the laboratory. It knows how to identify transgenic soy, ten corn and five rapeseed varieties. But up until then, it had never concerned itself with rice....

    In Orleans, the laboratory for the Food General Directorate (DGAL) analyzes imported seed, while the semi-public laboratory of the Group for Study and Control of Varieties and Seed Crops (Geves) studies those that must be registered in the catalogue of cultivated species.

    On their side, in order to make the quality of their products prevail, the agribusiness industry appeals to private laboratories, like the Eurofins laboratory in Nantes. "The distribution network - seed producers, importers, distributors - also seeks to do upstream controls so as to avoid costly product recalls," testifies Fran ois Vigneau, Executive Director of the laboratory, which effects "several tens of thousands of tests a year." In total, all these tests represent a considerable cost, every operation approaching 10 Euros.

    The enormous complication of the control procedures in less than ten years has, in any case, had an impact on the total food distribution system. As the INRA researcher who coordinates the European study program on the coexistence between GMO and conventional networks, Yves Bertheau, notes, "GMO have served as the model for other distribution channels and the methods for detecting them will serve to detect other contaminants, such as mycotoxins."

    For authorized GMOs, the procedures seem to be well broken-in. On the other hand, for non-identified GMOs, such as American and Chinese rice, the controllers have fewer resources and are constrained to proceed by a process of elimination: any GMO that does not figure among authorized plants is rejected. "But to identify a specific seed, sometimes we need 15 analyses," deplores Yves Bertheau. "We should improve our methods to reduce the process to a single stage, but that is not allowed for in the regulation."

    Consequently, the laboratories have implemented information networks to exchange data that is sometimes confidential. The international protocol on biosecurity expressly provides for the pooling of this data. But the planned data bank is not operational. And the United States, premier producer of GMOs, is not signatory to the protocol.

    "Overall, it is not in the operators' interest to cheat if they want to preserve their markets," Yves Bertheau notes, however. The American authorities have recently demonstrated an interest in European regulation governing traceability. And pressure from American distributors, scalded by several product recalls, could change where these lines are drawn.

    A particular difficulty arises for seed crops: while a 0.9% threshold of fortuitous GMO presence is allowed for food, no European directive has fixed an admissible threshold for non-authorized GMO contaminating crop seed.

    In principle, no trace should be tolerated in crop seed samples analyzed: that's the position taken by Italy and Greece. But inspectors in Great Britain and the Netherlands allow up to 0.5% of fortuitous presence. In France, the DGCCRF aligns itself with the strictest position, but the DGAL allows a level of 0.1%.

    This attitude, which could lead to the pollution of organic seed crops, is criticized by the promoters of that distribution channel. On the other hand, for conventional seed crops (neither GMO nor organic), that tolerance is essential to reduce costs in the face of American competitors who are not subject to such constraints. "It would be indispensable for us to know within which legal framework we work," deems Philippe Gracien, spokesman for the Interprofessional Crop Seed Group (GNIS). This lack of precision only fosters argument and consumer anxiety.