Two scientists believe that they may have identified a way to allay some of the anxieties of campaigners against genetically-modified organisms (GMOs).

The scientists, whose work is reported in the current issue of Nature Biotechnology, are Susan Scott and Dr Mike Wilkinson, of Reading University.

After a three-year field study they conclude that a new type of genetically-engineered crop will be very unlikely to transfer its modified genes to wild relatives.

If this is correct, it should mean there is much less likelihood of the emergence of so-called "superweeds" - plants which have accidentally acquired new characteristics from a GM crop.

The research concentrated on chloroplast genetic engineering, known also as transplastomic technology.

It involves inserting foreign genes into DNA from the plant's chloroplasts, sub-cellular structures that produce energy by photosynthesis.

Offering some reassurance

Genetic engineering normally inserts the alien genes into DNA from the plant's nucleus.

Transplastomic technology may improve on existing methods

Supporters of chloroplast genetic engineering believe that it may be a safer form than using the nucleus, because in many species chloroplast DNA is thought to be maternally inherited and so is not transmitted in pollen.

Genes from the nucleus, though, are carried in the pollen, and so have the potential to enter populations of related plants.

The authors used the slight size difference between the chloroplast genes of oilseed rape and its close relative, the wild turnip, to track the spread of chloroplast genes in pollen from cultivated fields to neighbouring weeds.

They report that they found negligible evidence that the genes were spreading.

Very infrequent escapes predicted

Dr Wilkinson told BBC News Online: "On the one hand the work suggests that there will be some escape of genes.

"But on the other, it suggests that the scope, range and frequency of the escapes will be very rare indeed.

"Probably, for a modified gene to spread rapidly and cause great environmental damage, it would need to confer on the wild plant a pretty big advantage."

The authors also conducted a two-year study of 18 wild turnip populations in areas close to cultivated oilseed rape fields.

This was to see whether genes from modified seeds spilt in fields containing the wild plant could cross repeatedly from crop to weed to produce hybrid plants.

What they found was that "at least in these environmental conditions, such hybrid oilseed rape is doomed to extinction, for reasons unknown".