By Steve Connor
The Independent
October 10, 2003

21st Century Radio News -Click here for more articles like this.

Cross-pollination between GM plants and their wild relatives is inevitable and could create hybrid superweeds resistant to the most powerful weedkillers, according to the first national study of how genes pass from crops to weeds.

Its findings will raise concerns about the impact of GM crops. Next week the results will be published of farm-scale trials which have studied the impact on the countryside of three types of crop.

The government-funded scientists said the latest findings "contrast" with previous assessments of gene flow between farm crops and weeds. They had suggested that the danger of hybridisation -- where two types of plant cross-pollinate to create another, for example a superweed -- was limited. Superweeds are considered to be a threat because, in some cases, they might absorb resistance to weedkillers from GM crops engineered to be herbicide-tolerant.

But the results of the research, which involved analysing satellite images of the British countryside and patrolling 180 miles of river banks, reveal that hybridisation is both more widespread and frequent than previously anticipated.

Mike Wilkinson of Reading University, who led the study published today in the journal Science, said physical barriers such as isolation distances -- buffer zones designed to stop pollen spreading from GM crops into the wild -- would have only a limited impact on preventing hybridisation.

"This [study] shows that isolation distances will reduce hybrid numbers but not prevent hybridisation. It depends on what level of hybridisation you deem acceptable but if you want to absolutely prevent hybrids then isolation distances will not do so," Dr Wilkinson said. "Hybridisation is more or less inevitable in the UK context," he added.

The study concentrated on non-GM oilseed rape and assessed how easily it cross-bred with a near-relative in the wild called bargeman's cabbage, also known as wild turnip, which typically grows along river banks. Although the research was based on conventional oilseed rape, Dr Wilkinson said the conclusions applied to any flow of genes that could be expected from the GM varieties of oilseed rape that were undergoing farm-scale trials.

"Our findings are directly transferable to almost all sorts of genetically modified oilseed rape," he said. "The only exceptions will be ones where there is male sterility introduced into the crop."

Researchers scoured the countryside for sites where bargeman's cabbage grew near to oilseed rape fields and they used DNA techniques to assess whether any hybrids between the crop and the wildflower had been produced as a result of pollen transfer.

The scientists, from the Natural Environment Research Council and the Centre for Ecology and Hydrology in Dorset, calculated the frequency of hybridisation and used it to estimate the number of hybrids that would form each year across the UK.

They concluded that typically there would be 32,000 hybrids produced annually in wild riverside populations of bargeman's cabbage, and a further 17,000 hybrids growing among a weedier variety of the wildflower which tends to infest farmland. This represents a relatively small fraction of the 88 million wild bargeman's cabbage plants estimated to grow along British riverbanks, but if the hybridisation involved a GM gene that conferred a significant advantage on the weed, the hybrid could quickly spread to pose a superweed threat.

An important outcome of the work is that it will allow scientists to assess what needs to be done to limit the spread of genes and pollen from GM crops. One possibility is to make the male plants sterile so they do not produce pollen.

"If we know how many hybrids to expect then we can test methods that people put forward hoping to prevent hybrid formation. In order to prevent hybrid formation you need to know how many to expect in the first place," Dr Wilkinson said.

"One of the main reasons for doing the work is that this sort of data represents a starting point for us to do predictive modelling, to predict how particular different sorts of genes will behave across the country.

"It's important to know how many hybrids to expect, to know how efficient it has to be to prevent hybrids. The key question is whether the gene that they contain is going to cause a change [to the countryside] or not," he said.

Although the latest study stands in contrast to previous work attempting to predict gene flow between farm crops and wild flowers, Dr Wilkinson said the findings were not totally surprising. "The level of hybrid formation is more or less in keeping with what we expected on a national level," he said. "What's surprised us slightly is the variability between the regions."

21st Century Radio News -Click here for more articles like this.

  21st Century Radio-HOME

Hieronimus & Co., Inc., P.O. Box 648, Owings Mills, MD 21117 USA
Voice Mail: (410) 356-4852 Fax: (410) 356-6229