Aspects of the biosafety of transgenic oilseed rape (Brassica napusL.)

Hybridization between transgenic Brassica napus L. and its wild relatives: Brassica rapa L., Raphanus raphanistrum L., Sinapis arvensis L., and Erucastrum gallicum (Willd.) O.E. Schulz

The frequency of gene flow from Brassica napus L. (canola) to four wild relatives, Brassica rapa L., Raphanus raphanistrum L., Sinapis arvensis L. and Erucastrum gallicum (Willd.) O.E. Schulz, was assessed in greenhouse and/or field experiments, and actual rates measured in commercial fields in Canada. Various marker systems were used to detect hybrid individuals: herbicide resistance traits (HR), green fluorescent protein marker (GFP), species-specific amplified fragment length polymorphisms (AFLPs) and ploidy level. Hybridization between B. rapa and B. napus occurred in two field experiments (frequency approximately 7%) and in wild populations in commercial fields (approximately 13.6%). The higher frequency in commercial fields was most likely due to greater distance between B. rapa plants. All F(1) hybrids were morphologically similar to B. rapa, had B. napus- and B. rapa-specific AFLP markers and were triploid (AAC, 2n=29 chromosomes). They had reduced pollen viability (about 55%) and segregated for both self-incompatible and self-compatible individuals (the latter being a B. napus trait). In contrast, gene flow between R. raphanistrum and B. napus was very rare. A single R. raphanistrum x B. napus F1 hybrid was detected in 32,821 seedlings from the HR B. napus field experiment. The hybrid was morphologically similar to R. raphanistrum except for the presence of valves, a B. napus trait, in the distorted seed pods. It had a genomic structure consistent with the fusion of an unreduced gamete of R. raphanistrum and a reduced gamete of B. napus (RrRrAC, 2n=37), both B. napus- and R. raphanistrum-specific AFLP markers, and had <1% pollen viability. No hybrids were detected in the greenhouse experiments (1,534 seedlings), the GFP field experiment (4,059 seedlings) or in commercial fields in Québec and Alberta (22,114 seedlings). No S. arvensis or E. gallicum x B. napus hybrids were detected (42,828 and 21,841 seedlings, respectively) from commercial fields in Saskatchewan. These findings suggest that the probability of gene flow from transgenic B. napus to R. raphanistrum, S. arvensis or E. gallicum is very low (<2-5 x 10(-5)). However, transgenes can disperse in the environment via wild B. rapa in eastern Canada and possibly via commercial B. rapa volunteers in western Canada.