Nov 5, 2003

Archive

Send

GM oilseed rape: Pollen dispersal and outcrossing

“We found hybrids only in brown mustard.”

Oilseed rape is able to outcross to several related species, but fertile progeny are rare. A working group at the Federal Biological Research Centre for Agriculture and Forestry (BBA) in Kleinmachnow led by Bernd Hommel has been working on the outcrossing behaviour of herbicide-resistant oilseed rape - not in the laboratory, but in a field trial. GMO Safety spoke to Dr. Bernd Hommel from the BBA in Kleinmachnow about outcrossing, volunteer rape and the possible accumulation of GM rape on the fields.

GMO Safety: In Central Europe related species can be found to which oilseed rape can outcross. You have studied these processes under field conditions, by creating several habitats with different plant species related to oilseed rape around the trial field. How frequently does oilseed rape outcross and to which plants?

Bernd Hommel: To look more closely at outcrossing in oilseed rape we planted two related species which are common in our region; field mustard and wild radish, and also the less common black and brown mustard, directly adjacent to the oilseed rape plots, ensuring that they would flower at the same time as the rape. Pollination was to take place naturally. We planted the seeds collected from these plants in the greenhouse, and the young plants were treated with the herbicide. We could be certain that any surviving plants were hybrids of the transgenic oilseed rape.

The high outcrossing rate of transgenic oilseed rape to brown mustard was surprising. At 0.26 percent it was in some cases even higher than the outcrossing rate of transgenic to conventional rape, which was around 0.15 to 0.32 percent. The hybrids resembled the oilseed rape not only in terms of herbicide resistance; the leaf shape was also very similar. The hybrids from brown mustard and oilseed rape developed well but produced no viable seed.

We were unable to find any hybrids in any of the other species. The sexual compatibility of oilseed rape with brown mustard was in fact anticipated, since they both stem in part from the same genetic source, which goes back several hundred years.

In natural conditions rape pollen is very rarely compatible with other cruciferous plants. This has also been shown by other studies from Germany and elsewhere. Here in Germany around one million hectares of oilseed rape, which flowers from April to May, has been grown annually for several years. If the outcrossing of oilseed rape to related species were a fairly common occurrence, then hybridisation with field mustard, wild radish, shepherd’s purse, field pennycress and various rocket species would be widespread and clearly visible in rape-growing areas. And yet there are no indications of this. It has not yet been proven that the genetically modified oilseed rape varieties are somehow better equipped to cross species boundaries than conventional rape varieties.

GMO Safety: If genetically modified oilseed rape is cultivated, its pollen will also spread. And outcrossings will also occur, either to other rape plants or to wild species such as brown mustard. How are such processes to be assessed from an ecological point of view?

Bernd Hommel: An outcrossing from cultivated plants to related species does not generally cause problems; if this were the case we would have to stop growing oilseed rape in Central Europe. In Mexico, too, cultivated maize is grown, despite the fact that teosinte, the wild form, is widespread. The ecological relevance of such outcrossings only became an issue there with the emergence of green genetic engineering. Why? Genetic engineering involves traits that are currently not a feature of existing varieties or their wild forms, since they are derived genetically from foreign organisms such as bacteria.

With regards to possible ecological consequences of an outcrossing, each trait must be considered separately. In our view, an outcrossing of herbicide resistance to related species is not serious. We would not expect such hybrids to have a selection advantage since herbicides are not used on a large scale on non-agricultural land. The selection of natural herbicide-resistant biotypes as result of agriculture is far more likely than outcrossing. Acquired herbicide resistance is already widespread today in associate farmland flora. This too could be spread further through outcrossing.

An insect-resistant oilseed rape would of course be assessed differently. In this case hybrids in semi-natural habitats could gain in strength, because they can produce more seeds and so suppress other species. When assessing the ecological relevance, however, it should not matter whether insect resistance is bred classically or genetically.

GMO Safety: There is currently much debate about coexistence and threshold values. The European Commission proposes that conventional rapeseed may contain up to 0.3 percent GMO contamination. On the basis of your research, do you think this value is appropriate and realistic? Do you anticipate that GMO levels will accumulate over time both in oilseed rape products and in ecosystems?

Bernd Hommel: That is a very important question. The rape growers along with the seed growers must ensure that the 0.3 percent contamination level is upheld.

Up to 10 percent of oilseed rape is lost during harvesting. In subsequent years these losses re-emerge as weeds, or so-called volunteers, in the following crops. But we can rule out the possibility of a significant accumulation of GM rape in the soil as a result of this. During the first two years after the rape harvest, a high proportion of the volunteer rape falls victim to chemical and mechanical weed control in the following crops. If the farmer grows oilseed rape on the same field three to five years later, very low levels of volunteer oilseed rape from the previous rape harvest will be present in the soil seed bank. Levels would be even lower if the oilseed rape were cultivated using the ploughless tillage system, which is gaining in popularity. It must also be considered that volunteer rape only very rarely comes into flower on arable land.

Compared with cultivated GM oilseed rape, volunteer rape therefore makes only a very insignificant contribution to pollen transfer to conventional rape or to related species. We believe it is not possible for GM contamination in harvested oilseed rape products to exceed the 0.9% threshold as a result of volunteer rape. Rape volunteers can certainly grow well within the oilseed rape crop, but the three to five-year growing interval should ensure that levels of contamination remain very low.