91av

TRAITS such as herbicide resistance could spread from genetically modified
oilseed rape to its wild relatives if the crop were widely grown in Britain. The
first large-scale study of gene flow suggests that hybridisation between crops
and weeds is rare—but does occur.

Using satellite images covering 15 000 square kilometres of southern England,
Mike Wilkinson and his colleagues at the University of Reading identified fields
where normal oilseed rape or canola (Brassica napus) was being grown. A close
relative of oilseed rape, the wild turnip Brassica rapa, grows mostly along
river banks, so by looking at maps of waterways the Reading team found two sites
where the plants were likely to be growing next to each other. After testing
leaf samples from 505 plants, Wilkinson found one hybrid formed by
cross-pollination between the oilseed rape and wild turnip.

Wilkinson is wary of reading too much into a single instance. “You’ve got to
be really careful interpreting one,” he says. “Next year, we might get five or
ten. Or we might only get one in ten years.”

He suggests that GM crops are grown at set distances from weedy relatives to
reduce the risk of gene transfer. “I would think this is a very sensible
precaution,” says Alan Gray, chairman of Britain’s Advisory Committee on
Releases to the Environment.

But such precautions couldn’t stop gene movement altogether, says Gavin
Ramsay of the Scottish Crop Research Institute in Dundee. “If you are aiming for
zero transfer, it’s hard to say you’d achieve it with any of these measures.”