New Scientist December 15, 2001, Letters

It is difficult to avoid genetic contamination from genetically modified oilseed rape (24 November, p 14). Is this a surprise?

Three years ago, you printed my letter warning of the danger of transferring herbicide tolerance from GM rape into the closely related weed, charlock (29 August 1998, p 49). I also mentioned that oilseed rape was, by then, one of the commonest feral plants in Britain.

The greatest problem is seed spillage at harvest, plus the long life of seed in soil. Here in west Wales virtually no roadside verge or previously cultivated land is free of viable seed and once GM rape is widely sown it will behave similarly.

Our neighbouring farmer neglected the seedlings from a previous rape crop and by mid-summer almost every plant was seriously infested with cabbage white butterfly caterpillars, which formed a larder for the local insectivorous birds while the flowers were a magnet for hoverflies, honey bees and other insects.

During gales these pollinators can be driven many tens of kilometres. A caterpillar's gut, being a seething microbiological morass, is an ideal incubator for genes that can pass from plant tissue into bacterial species, and foraging finches and chats then pick up the caterpillars. The transfer scenario was completed as we watched a feeding finch explode into a cloud of feathers as it was taken by a peregrine falcon - which no doubt discarded the bird guts and dead caterpillars elsewhere.

There is no practicable separation distance that will prevent the genes from GM oilseed rape mixing with the conventional crop.

John Etherington (Haverfordwest, Pembrokeshire)