How to manage the impact of gene flow on oilseed rape grain quality? Simulation case studies of three contrasted landscapes

Grain admixture due to gene flow of oilseed rape (OSR) could be economically damaging. Different strategies are currently used or proposed to manage grain quality, ranging from homogenisation at the silo level to tactical decision rules at the field level. The relevance of these general strategies was appraised in the case of genetically modified (GM) OSR in three contrasted regions in Europe: Beauce Blésoise (France), Schleswig-Flensburg (Germany), and Fife (UK). Field patterns, crop allocation and agricultural practices were derived from existing datasets and complementary field surveys. Then a gene flow simulator was used to assess how the local contexts influenced the grain admixture between GM and non-GM OSR (without separation measures). The simulations showed that grain admixtures in fields followed a leptokurtic curve. While, however, the worst case was similar in the three regions, the median differed greatly depending on contextual factors such as the size and arrangement of fields. Grain admixture very rarely exceeded the 0.9% threshold for non-GM products if assessed at the level of the silo, at which the grains from all non-GM crops were combined, while maintaining crops below the threshold at the field level required management of a few high risk situations. Analysing grain admixtures and commercial risks at different decision levels (field, silo) demonstrated the efficiency of “flexible” as opposed to “rigid” measures, but this technical analysis did not take other criteria like regulatory issues into account.

► We simulated gene flow between GM and non-GM oilseed rape fields in 3 regions.
► Max GM grains % in non-GM fields were similar between regions unlike medians.
► Distributions of GM grains % in non-GM fields were leptokurtic.
► GM grains % in non-GM silos very rarely exceeded the 0.9% threshold.
► The results suggest flexible measures would be more convenient than rigid ones.