Jonathan Latham and Allison Wilson
LL601, the genetically engineered rice variety that has contaminated the US rice supply, is safe. The USDA says so. The UK Food Standards Agency says so. The US Food and Drug Administration, on 12 Sept, stated that: “LL601 rice poses no risk to human health and does not raise any food, feed safety or environmental concerns.” These assertions have been echoed by other organisations such as the USA Rice Federation: “We understand the EU sensibilities are a little different than the United States but nonetheless the product is a safe one”.
These statements, originating from officials of various government organisations, are noteworthy in part because LL601 has not passed through the official scientific and regulatory approval process in any country. Remarkably, these agencies apparently feel able to presume that a positive outcome of the scientific review process can be guaranteed.
Fortunately, not every regulator concurs with this presumptive approach. Based on similar data, the European Food Safety Authority (EFSA) GMO panel has taken a different stance. In a statement released by the EU on Sept 14, they said “According to the Statement of the Panel issued today there is insufficient data to provide a full risk assessment in accordance with EFSA’s GM guidance.” Unremarkable as it may seem, this statement has the exceptional merit of not preempting completion of the scientific review process.
With the honorable exception of the EFSA GMO panel, these official and unqualified assertions of absolute safety are inappropriate for another reason too. No food is absolutely safe for all people, and a GMO crop is at best only as safe as its non-transgenic counterpart. This statement should be scientifically obvious, but it is noteworthy because one of the perennial complaints of the biotechnology industry is that the public demands ‘unreasonably’ that their products be unqualifiedly safe. That public expectation looks a lot more reasonable when that is what the public is offered by their official regulatory bodies.
In reality there is a further constraint on the extent to which a guarantee of absolute safety can be offered by regulatory bodies. It consists of the fact that it is extraordinarily difficult for the risk assessment process to verify the safety of a GMO crop, even to the extent of concluding that it is as safe as its non-GMO parent. As Jeremy Sweet (vice chair of the EFSA GMO panel) has expressed it: “All sorts of things happen that you cannot possibly assess in a risk assessment”, -which is why the EU plans to monitor GMO crops for environmental and health effects. To put it another way, any guarantee of safety can only be as good as the rigour and comprehensiveness of the risk assessment it is based upon.
So how rigorous and comprehensive is the evidence for the safety of LL601? USDA has so far published a draft environmental assessment containing a preliminary decision to grant deregulated status (i.e. approval) to LL601. As explicitly indicated in the summary, approval is based on the finding that “There was no indication from field reports to suggest that LLRICE601 would behave any differently than lines LLRICE62 and LLRICE06 under field conditions.” (EA p3) and the assertion on page 19 that “ Morphological and biochemical data reveal no significant differences between LLRICE601 and Cocodrie [the parent],” (EA p20).
These statements imply that considerable testing has been done and that these tests found no differences between LL601 and Cocodrie. In fact, the number of agronomic and phenotypic measurements taken can be summarised briefly as the parameters required to establish plant protection rights plus some information on dormancy and shattering and a compositional analysis of 9 biochemical parameters.
Overall, the data and the conclusions drawn contain significant inadequacies and discrepancies and the data are inadequate either to support or to refute the finding of no difference. There is not the space here to elaborate on the flaws in the application but readers may do so for themselves at: Bayer CropScience Petition 06-234-01P. In this application readers can find tables summarising the results of experiments whose methods are inadequately described, which contain data with inadequate or non-existent statistical analysis and others in which an inappropriate comparison (control) was used. Some experiments combine all three flaws. Nevertheless, if one accepts the data as presented, there appear to be clear differences between LL601 and Cocodrie (the parent line). Perhaps the clearest example of this is that LL601 has a shattering score of 5 whereas the parent scores 3 on a scale of 1-9. Not only does this indicate a clear difference, which contradicts the conclusions made in the summary, but enhanced shattering is accepted as a weediness character indicative of a plant pest risk. It would be very interesting indeed to know whether this altered trait contributed to the ability of one or a few small experimental trials of LL601 to contaminate a large proportion of the US long grain rice crop.