A nondescript grass discovered in the Oregon countryside is hardly an alien invasion. Yet the plant - a genetically modified form of a grass commonly grown on golf courses - is worrying the US Department of Agriculture (USDA) enough that it is running its first full environmental impact assessment of a GM plant.
It is the first time a GM plant has escaped into the wild in the US, and it has managed it before securing USDA approval. The plant, creeping bentgrass (Agrostis stolonifera, carries a bacterial gene that makes it immune to the potent herbicide glyphosate, better known as Roundup. The manufacturer, The Scotts Company, Marysville, Ohio, is hoping the grass will provide a turf that makes it easier for golf course owners to manage their fairways and greens by letting them kill competing weedy grasses with glyphosate.
Jay Reichman and colleagues at the US Environmental Protection Agency's labs in Corvallis, Oregon, identified nine escapees out of 20,400 plants of various grass varieties sampled within a 4.8-kilometre radius of the site where the bentgrass is being cultivated, the most distant 3.8 kilometres away. The team showed that the GM grass has spread both by pollinating non-GM plants to form hybrids, and by seed movement.
Bentgrass is a perennial, so once out there it regrows year after year, whereas most GM crops - mainly soybeans, maize and canola (oilseed rape) - are annuals, unable to reproduce, harvested each year and replaced with an entirely new crop the next. Another worry is that unlike the other GM crops, bentgrass has many relatives in the US with which it can cross-breed or hybridise, potentially passing on the glyphosate-resistance gene to other species - with unpredictable results.
"It's a cautionary tale of what could happen with other GM plants that could be of greater concern," says Reichman. "I suspect that more examples of this will show up." His report will appear in the October issue of Molecular Ecology.
If bentgrass is approved by the USDA, it could prove a hit with the thousands of golf course managers throughout the US, making it easy for the crop to spread far and wide. If it reaches environmentally sensitive wildernesses or establishes itself by waterways, removing it could require weedkillers far more harmful than the relatively benign glyphosate.
"It's definitely a new set of variables we've not had to deal with in previous GM crops," says Eric Baack of Indiana University in Bloomington, who comments on Reichman's findings in Current Biology (vol 16, p R1). Still, it isn't clear whether the gene would have much impact in the wild. "You wouldn't expect the weedkiller-resistance gene to be a particular advantage in the wild," says Baack. Also, the USDA doesn't class conventional bentgrass as a "noxious" weed.
There is however the possibility of litigation if the GM grass contaminates other elite grass strains under cultivation. Some 70 per cent of the US's commercial grass seed is grown in Oregon, so there is the potential for accidental adulteration.
The USDA is not taking any chances. "This is a perennial, and has wild and weedy relatives, and it's something we think we need to know the environmental impact of before it's deregulated," says a spokeswoman for the USDA's Biotechnology Regulatory Services in Riverdale, Maryland. "There's no current set date for when [the environmental impact assessment] will be finished," she says.
Whether the US public takes any notice of the furore is another question entirely. "I don't think people will worry about lawns and golf courses if they've not shown any worries already about GM food," says Baack.
From issue 2564 of New Scientist magazine, 09 August 2006, page 9