Genes From Engineered Grass Spread for Miles, Study Finds
By Andrew Pollack
September 21, 2004
A new study shows that genes from genetically engineered grass can spread much farther than previously known, a finding that raises questions about the straying of other plants altered through biotechnology and that could hurt the efforts of two companies to win approval for the first bioengineered grass.
The two companies, Monsanto and Scotts, have developed a strain of creeping bentgrass for use on golf courses that is resistant to the widely used herbicide Roundup. The altered plants would allow groundskeepers to spray the herbicide on their greens and fairways to kill weeds while leaving the grass unscathed.
But the companies' plans have been opposed by some environmental groups as well as by the federal Forest Service and the Bureau of Land Management. Critics worry that the grass could spread to areas where it is not wanted or transfer its herbicide resistance to weedy relatives, creating superweeds that would be immune to the most widely used weed killer. The Forest Service said earlier this year that the grass "has the potential to adversely impact all 175 national forests and grasslands."
Some scientists said the new results, to be published online this week by the journal Proceedings of the National Academy of Sciences, did not necessarily raise alarms about existing genetically modified crops like soybeans, corn, cotton and canola. There are special circumstances, they say, that make the creeping bentgrass more environmentally worrisome, like its extraordinarily light pollen.
Because Scotts has plans to develop other varieties of bioengineered grasses for use on household lawns, the new findings could have implications well beyond the golf course. And the study suggests that some previous studies of the environmental impact of genetically modified plants have been too small to capture the full spread of altered genes.
Scotts says that because naturally occurring bentgrass has not caused major weed problems, the bioengineered version would pose no new hazards. And any Roundup-resistant strains that might somehow develop outside of intentionally planted areas could be treated with other weed killers, the company said.
In the new study, scientists with the Environmental Protection Agency found that the genetically engineered bentgrass pollinated test plants of the same species as far away as they measured -about 13 miles downwind from a test farm in Oregon. Natural growths of wild grass of a different species were pollinated by the gene-modified grass nearly nine miles away.
Previous studies had measured pollination between various types of genetically modified plants and wild relatives at no more than about one mile, according to the paper.
"It's the longest distance gene-flow study that I know of," said Norman C. Ellstrand, an expert on this subject at the University of California, Riverside, who was not involved in the study but read the paper.
"The gene really is essentially going to get out," he added. "What this study shows is it's going to get out a lot faster and a lot further than people anticipated."
One reason the grass pollen was detected so far downwind was the size of the farm - 400 acres with thousands of plants. Most previous studies of gene flow have been done on far smaller fields, meaning there was less pollen and a lower chance that some would travel long distances. Those small studies, the new findings suggest, might not accurately reflect what would happen once a plant covers a large area.
"This is one of the first really realistic studies that has been done," said Joseph K. Wipff, an Oregon grass breeder. Dr. Wipff was not involved in the latest study but had conducted an earlier one that found pollen from genetically engineered grass traveling only about 1,400 feet. That test, though, used less than 300 plants covering one-tenth of an acre.
The effort to commercialize the bentgrass has attracted attention because it raises issues somewhat different from those surrounding the existing genetically modified crops.
It would be the first real use of genetic engineering in a suburban setting, for example, rather than on farms. And the grass is perennial, while corn, soybeans, cotton and canola are planted anew each year, making them easier to control.
Bentgrass can also cross-pollinate with at least 12 other species of grass, while the existing crops, except for canola, have no wild relatives in the places they are grown in the United States. And crops like corn and soybeans have trouble surviving off the farm, while grass can easily survive in the wild.
The bentgrass, moreover, besides having very light pollen - a cloud can be seen rising from grass farms - has very light seeds that disperse readily in the wind. It can also reproduce asexually using stems that creep along the ground and establish new roots, giving rise to its name.
Because of the environmental questions, the application for approval of the bioengineered bentgrass is encountering delays at the Department of Agriculture, which must decide whether to allow the plant to be commercialized.
After hearing public comments earlier this year, the department has now decided to produce a full environmental impact statement, which could take a year or more, according to Cindy Smith, who is in charge of biotech regulation.
Ms. Smith, in an interview yesterday, said the new study "gives some preliminary information that's different from previous studies that we're aware of." But more conclusive research is needed, she said.
Bentgrass is already widely used in its nonengineered form by golf course operators, mainly for greens but also for fairways and tee areas, in part because it is sturdy even when closely mown. It is rarely used on home lawns because it must be cared for intensively. And creeping bentgrass does not cross-pollinate with the types of grass typically used on lawns, scientists said.
Executives at Scotts, a major producer of lawn and turf products based in Marysville, Ohio, said the genetically engineered bentgrass would be sold only for golf courses. They said golf courses cut their grass so often that the pollen-producing part of the plants would never develop.
And because nonengineered creeping bentgrass has not caused weed problems despite being used on golf courses for decades, they said, the genetically modified version would pose no new problems.
"There has been pollen flow but it has not created weeds," Michael P. Kelty, the executive vice president and vice chairman of Scotts, said in an interview yesterday. He said Scotts and Monsanto, the world's largest developer of genetically modified crops, had spent tens of millions of dollars since 1998 developing the bioengineered bentgrass.
The questions about the grass come after Monsanto, which is based in St. Louis, said earlier this year that it was dropping its effort to introduce the world's first genetically engineered wheat, citing concerns by farmers that its use in foods might face market opposition.
Scotts is also developing genetically modified grass for home lawns, like herbicide-tolerant and slow-growing types that would need less mowing. But those products still need several more years of testing, Dr. Kelty said, adding that the company would avoid types of grass that could become weeds. "We don't want to put a product out there that is going to be a threat," he said.
Scotts and Monsanto have received some support for their argument from the Weed Science Society of America, a professional group, which conducted a review of the weed tendencies of creeping bentgrass and its close relatives at the request of the Department of Agriculture.
"In the majority of the country these species have not presented themselves as a significant weed problem, historically," said Rob Hedberg, director of science policy for the society, summarizing the conclusions of the review. He said that because people have generally not tried to control bentgrass and similar species with Roundup, known generically as glyphosate, "the inability to control them with this herbicide is a less significant issue."
Still, the society's report noted that bentgrass could be considered a weed by farms that are trying to grow other grass seeds. And the Forest Service, in comments to the Agriculture Department earlier this year, said that bentgrass has threatened to displace native species in some national forests.
John M. Randall, acting director of the Invasive Species Initiative at the Nature Conservancy, said bentgrass and related species had been a threat to native grasses in certain preserves that the group helps manage, including a couple near Montauk Point on eastern Long Island.
Other opponents of the genetically modified grass seized on the results. "This does confirm what a lot of people feared - expected, really," said Margaret Mellon, director of the food and environment program for the Union of Concerned Scientists in Washington. "These kinds of distances are eye-popping."
The new study was done by Lidia S. Watrud and colleagues at an E.P.A. research center in Corvallis, Ore., who were trying to develop new methods to assess gene flow, not specifically to study the bentgrass.
They put out 178 potted and unmodified creeping bentgrass plants, which they called sentinel plants, at various distances around the test farm. They also surveyed wild bentgrass and other grasses. They collected more than a million seeds from the plants, growing them into seedlings to test for herbicide resistance and doing genetic tests.
The number of seeds found to be genetically engineered was only 2 percent for the sentinel plants, 0.03 percent for wild creeping bentgrass and 0.04 percent for another wild grass. Most of those seeds were found in the first two miles or so, with the number dropping sharply after that. Still, said Anne Fairbrother, one of the authors of the report, finding even some cross pollination at 13 miles "is a paradigm shift in how far pollen might move."