Potential Persistence of Escaped Transgenes: Performance of Transgenic, Oil‐Modified Brassica Seeds and Seedlings

We performed two experiments designed to assess the risk that seed‐oil‐modification transgenes will increase the persistence of feral Brassica napus canola and interspecific hybrids of B. napus canola and wild Brassica rapa, a weedy relative. The first experiment, conducted at field sites in California and Georgia where oil‐modified canola will be grown commercially, tested whether buried seeds of transgenic high‐stearate canola had increased survivorship and dormancy. Performance of the high‐stearate type was compared to nontransgenic null segregant and parental lines. In California, no differences in initial proportions of dormant seeds and rates of exit could be detected between high‐stearate canola and its controls, suggesting low probability that high‐stearate canola will form a larger or more persistent seed bank than its nonpersistent controls. In Georgia, although high‐stearate canola initially had as low or lower proportions of dormant seeds than its controls, high‐stearate seeds exhibited no detectable exit from the seed bank, whereas both controls had significant rates of exit. Hence, escaped high‐stearate seed may persist for a longer period than its controls at this site. Differences between the sites highlight the need to conduct risk assessment over the range where a transgenic crop will be commercialized. The second experiment, a greenhouse study, measured the relative ability of oil‐modified canolas and wild × crop hybrids to emerge from four depths in the soil (0, 0.5, 4, and 10 cm) and their subsequent seedling vigor. We tested lines of B. napus canola carrying either the high‐stearate gene or a transgene for high‐laurate production, using nontransgenic parental types as controls. We also examined the impact of the high‐laurate transgene in interspecific wild B. rapa × B. napus canola hybrids. Performance of the high‐laurate hybrids was compared to nontransgenic hybrids and the B. rapa wild parent. For all seed types, no seedlings emerged from 10 cm, and all seedlings emerged from 0 and 0.5 cm. A higher proportion of high‐stearate canola emerged from 4 cm than its control, but for all depths, high‐stearate canola emerged more slowly and had significantly less biomass than its control 2 and 4 wk following emergence. In contrast, high‐laurate canola's total emergence and timing of emergence could not be distinguished from its control. Although high‐laurate canola's 2‐wk biomass was less than that of its control, by 4 wk, its biomass was equivalent due to its significantly higher relative growth rate during that period. The different results for the two oil‐modification transgenes suggest that even transgenes with similar functions should be considered on a case‐by‐case basis. From 0 and 0.5 cm, high‐laurate wild × canola hybrids' total emergence, timing of emergence, and biomass accumulation were indistinguishable from their wild parent. High‐laurate hybrids emerged more rapidly and had greater biomass at 2 wk than their hybrid controls. Our results indicate that high‐laurate hybrids, emerged from shallow depths, may experience performance advantages that will allow them to perform as well as their persistent, wild parent.