The Great Gene Escape

The seed companies say the plants they've created are safe. But who's to know what will come from a romp in the field with an untamed weed?

By Josie Glausiusz|Friday, May 01, 1998
Contemplating world hunger from the vantage point of a well-laden breakfast table is certainly comfortable, if odd. One morning last January, executives of Iowa-based Pioneer Hi-Bred International, the world’s largest developer, producer, and marketer of genetically improved seed, gathered at the Friend of a Farmer café in downtown Manhattan for a discussion about global food security. Amid the restaurant’s rustic decor—dried hydrangeas in earthenware pots, autumn gourds tumbling from rush baskets, exposed brickwork—the three officials and a group of journalists sat dining on maple syrup–soaked buttermilk pancakes, muffins, corn bread, omelettes, and apple butter as Pioneer’s chairman and ceo, Chuck Johnson, outlined his vision of the future. The business we’re in is ensuring that the world has the capacity to have the food it needs to survive, he explained. That future capacity, he is convinced, can come only from the crops that companies such as Pioneer are producing: high-yield, insect-resistant breeds of corn, soybeans, sorghum, and sunflowers.

Pioneer makes some of its seeds conventionally, by creating hybrids. Back in the 1920s, though, the conventional was radical, and the typical farmer looked upon the newfangled seeds, in Johnson’s words, as witchcraft and Satanism—until he got his first taste of the yield. For the past few years, however, Pioneer has been offering genetically engineered seeds, which have genes spliced into their chromosomes that make them more resistant to insects and weed killers. Johnson told the journalists about herbicide-resistant soybeans and a variety of corn that produces a toxin normally made by a bacterium known as Bacillus thuringiensis, or Bt. Last year, he said, a million acres of the Bt corn were planted in the Midwest, with an increased yield of 10 to 15 percent, thanks to the way the Bt toxin discourages corn-eating insects.

Pioneer’s vice president for marketing, Mary McBride, then chimed in, claiming that these transgenic crops have the power to increase food production in the developing world with minimal environmental impact. The world’s population, she noted, is continuing to rise and must somehow be fed. And with the growing affluence of Asia, much of that increasing population will be eating more meat—thus demanding even more crops to feed the pigs and cows they will consume. By using high-yield transgenic crops, farmers will be able to harvest so much food that they won’t try to cultivate fragile, marginal lands. Pioneer, as McBride put it, is creating virtual acres.

Outside the comfortable confines of the Pioneer breakfast, this sort of unmitigated optimism is harder to find. The public is generally wary of the transgenic crops that are landing in American fields, and there are many vocal critics. As of last October, 24 genetically engineered crops had been approved by the Food and Drug Administration for sale in the United States, a further 8 are awaiting approval, and thousands more are being tested. Many are similar to Pioneer’s crops, engineered to carry Bt toxin or to survive dousing by herbicides that kill the weeds infesting their fields. Others have been made resistant to various viruses, while still others have genes that delay their ripening or thicken their skin.

Opponents of transgenic crops claim that ecological and evolutionary forces could turn these crops into disasters. Perhaps the plants will prove so robust that they will grow aggressively, like weeds, and invade other environments—including a neighboring farmer’s fields. Virus-resistance genes could escape into weeds and make them so hardy they’d outcompete endangered plants in the wild. Antibiotic-resistance genes (which botanists insert into transgenic crops as supposedly harmless markers) might escape into soil bacteria and from there into those that infect humans. Crops engineered to carry Bt-toxin genes might trigger the evolution of ever adaptive Bt-resistant bugs.

Is all this worry just more witchcraft and Satanism? The only way to know how seriously to take such doomsday scenarios is to run experiments. Researchers have only begun to do this work, setting up experiments to see how readily transgenic genes and proteins can escape the crops they were meant to help. The results thus far are proving that the doomsday scenarios are not pure fiction. But the researchers are split over whether the results should be cause for anxiety.

Much of the concern over transgenic crops stems from the promiscuous sexual habits of plants. Sperm are found within pollen grains released by the stamens of flowers. The grains are carried by wind or by insect. If the pollen should land on another flower’s female organ, or carpel, it delivers its sperm to the egg hidden inside. Once the sperm fertilizes the egg, an embryo forms and a seed is produced. Not only can pollen from one breed of plant fertilize another, but different species can sometimes mate and produce hybrids that can reproduce. Genes in one population of plants (crops, for example) can thus seep into another population (neighboring weeds). In the late 1980s geneticist Norman Ellstrand at the University of California at Riverside began warning of the dangers of this genetic escape. One could, for example, imagine an herbicide- resistance gene getting into weeds and making superweeds that could take over a field. Yet this possibility hinged on how likely it was for crops and weeds to hybridize, and for the transgenic genes to establish themselves in the wild population. Ellstrand decided therefore to measure the likelihood, and in 1996 he reported that domesticated sorghum, Sorghum bicolor, could readily form hybrids with a weed called johnsongrass, Sorghum halepense. (Domesticated crops are often surrounded by their close weedy relatives, since both flourish under the same conditions.) Using harmless gene markers rather than actual transgenes, Ellstrand found that wind-carried pollen could create hybrid seeds over 300 feet away from the original crop. These hybrids produced pollen and seeds as viable as the johnsongrass, meaning that they could spread just as aggressively.

Ellstrand thinks that the implications for transgenic crops are quite disturbing. The take-home story is, if you engineer herbicide resistance into sorghum, and johnsongrass is growing within a couple of hundred meters, then you’re really asking for trouble, because then the genes will get into one of the world’s ten worst weeds—johnsongrass—and as soon as you apply herbicide, you’re going to be favoring it, says Ellstrand. Here in the United States, where we use sorghum largely as a forage crop, the worst scenario would be a few million dollars’ worth of damage. But in a place like Africa, where sorghum is a staple crop for humans, an escaped transgene could be disastrous. In Africa, the wrong genes falling into weeds could actually end up creating massive crop failure. There are so many weed relatives in Africa because that’s where sorghum was domesticated.

More recent experiments with actual transgenic crops also show that inserted genes can move between species. Plant geneticist Rikke Bagger Jørgensen of Denmark’s Risø National Laboratory in Roskilde studied the yellow-flowered crop called oilseed rape, known in the United States as canola and in Latin as Brassica napus. Oilseed rape is a cultivated cross between a weed called wild mustard, or Brassica campestris, and Brassica oleracea, the cabbage plant.

Jørgensen planted a version of oilseed rape engineered to survive a weed killer called Basta alongside its wild ancestor (and weedy neighbor) B. campestris. Fertile hybrids easily formed, and when Jørgensen sowed the hybrids together with the original weed, a second generation of seeds was produced. These seeds grew to adulthood without any fuss and turned out to be impervious to Basta as well. Jørgensen returned to her fields the following spring and discovered that this second generation had produced offspring of their own, which continued to be herbicide resistant.

These same genes for Basta resistance, it turns out, can also hop into more distantly related plants. French cytogeneticist Anne-Marie Chèvre of the National Institute of Agronomic Research at Le Rheu, found that these transgenic oilseed rape plants could donate their genes to wild radish (Raphanus raphanistrum). But their effects on the radish are not clear; the genes were carried into the wild radish population over the course of four generations, yet by that point only a quarter of the plants descended from the hybrids were resistant to the herbicide. The problem seems to be that the herbicide-resistance gene wasn’t firmly integrated into the genome of the wild radish. Chèvre, who doubts that the plants will be able to maintain their resistance, is watching to see whether a stable integration may come about in a future generation. If it does, she says, it will be very difficult to manage because the transgene will be spreading in the wild population.

Yet despite these results, Jørgensen and Chèvre remain sanguine about the prospects of transgenic crops. If you can put in genes that give the plant itself a better resistance, for instance, to fungal pathogens or to insect pests, then you can minimize your use of pesticides, and that would be beneficial to the environment, says Jørgensen. And she believes that as long as transgenic oilseed rape is carefully managed, it can be safe. If you spray very early, before the campestris flowers, you minimize its potential to hybridize, she explains. But it would be unwise to grow Basta-resistant oilseed rape alongside a crop resistant to a different herbicide. Then what you’ll have is Brassica campestris plants with multiresistance in very few generations, she says. A weed with only one herbicide-resistance gene would, however, still be manageable. According to Chèvre, You can always destroy the plants with another herbicide.

The prospect of herbicide-resistant crops creating the need for spraying still more herbicides doesn’t fit well with the environmentally friendly image offered by companies like Pioneer. Yet some critics think that biotech corporations are actually comfortable with that prospect because they can make transgenic crops as well as herbicides. (Monsanto, for example, makes Roundup Ready cotton, which is resistant only to the herbicide Roundup—also made by Monsanto.) The biotech companies, since they make the herbicides, don’t see it as a big problem, because it forces them to make a new herbicide, says botanist Hugh Wilson of Texas A&M; University.

Wilson has been studying transgenic gene flow and its possible effects not on the struggle between weeds and crops but between weeds and rare or fragile wild plant species. Herbicide resistance isn’t so much of a problem in this regard, since weed killers are found only on farms. Of far more pressing concern to him is the possibility that genes for resistance to insects, viruses, and fungi can be just as important in the wild. It’s conceivable that a spread of genes from transgenic crops into wild plants could allow them to outcompete other species. Transgenic crops could do the most damage, according to Wilson, in places where crops originated and where there are many wild relatives still thriving. For corn, the center of diversity is Mexico; for potatoes, it is Peru; for sunflowers, it is the United States.

We’ve got to retain genetic diversity, says Wilson. You can look at the potato blight, a situation where you take a subset of genetic diversity, put it in Ireland—boom—it’s hit by something and it’s obliterated immediately. The only way to resolve the problem is to go back to the point of origin, find a gene in wild potato that’s resistant, and fix it by conventional plant breeding. But if that wild potato’s not there, or if that wild potato is genetically uniform because of a weird transgenic interaction, then you’re a loser.

Researchers have indeed shown that virus-resistance genes can escape from some crops into wild rela- tives. But whether this newly resistant wild relative can outcompete other native wild plants is still an open question because research has been so sparse. The lack of work is not for lack of interest. Plant ecologist Allison Snow of Ohio State University in Columbus is trying to start an investigation into whether virus-resistance genes inserted by the biotech firm Asgrow into a squash called Freedom II can persist in the wild and provide a competitive edge. But she’s having trouble getting the necessary funding for the experiment from the U.S. Department of Agriculture. I put in a proposal twice to study this, and both times I was turned down, says Snow. It could be because my proposal had some scientific flaws in it, but I think part of it could be—possibly—political. People don’t want to study this thing. The squash is already deregulated. So the usda has already said that this is safe. The usda claimed it was safe because a different company has used conventional breeding to create a resistant hybrid squash. They didn’t use genetic engineering, so the usda could say that this is really not very different from what’s happened in the past.

Defenders of transgenic crops frequently argue that genetic engineering is in essence no different from the hybrid breeding that farmers have conducted for decades, with no ecological calamity. For 50 years they have been breeding virus-resistant plants, and they behave the same as these transgenic plants, maintains plant pathologist Dennis Gonsalves of Cornell. The wild relatives have the same ability to pick up resistance genes whether they came from natural breeding or whether they came from genetically engineered squash. Yet apparently in all this time wild relatives still haven’t become resistant to viruses (although no one has carefully studied this interaction between weeds and crops).

Unlike Snow, Gonsalves has been able to study the Freedom II with a usda grant. He hand-pollinated transgenic virus-resistant Freedom II squash with pollen from wild Texas gourd, producing hybrids which he then planted in a field three feet apart from nonengineered wild gourd. The experiment produced a mix of results. When he inoculated the plants with viruses, only the transgenic squash managed to produce viable fruit with viable seeds. Elsewhere in the field, however, a different result occurred. Where the virus was scarce—and the wild plants could thus thrive—the transgenic hybrids bred with wild Texas gourd. A small proportion of the offspring carried the transgenes and were resistant to the virus.

But Gonsalves isn’t too worried by his results. You’ve got to be careful to look at the big picture, he says. Among the wild gourd, the virus isn’t much of a menace, while it is a major problem for the cultivated gourd. This presumably is because of the way squash is grown close together, making it easier for the virus to spread from plant to plant, while wild gourd is far more scattered. So even if the virus-resistance gene were to get into the wild gourd, Gonsalves contends that it would hardly make a difference since the weed is unaffected by the virus.

Snow is familiar with this argument but not persuaded. The usda, she says, thinks that these diseases are not really that common in the wild, and they’ve never seen a wild plant with viral disease, so they think maybe that’s not having any effect on the wild population. But no one knows how many diseases are regulating wild and weedy plants. It’s a very difficult thing to study, and there hasn’t been much effort in that area.

These questions are moot when a crop plant has no weedy relatives in its vicinity. One possible way to contain the threat of transgene escape might be to bar certain genetically engineered crops when weedy relatives already exist in a given place. There are no weeds related to maize in Europe, says Chèvre. But we have a lot of wild species more or less related to oilseed rape in the field everywhere. Therefore France has permitted transgenic corn to be grown on its soil. The United States could similarly allow transgenic maize, soybeans, and potatoes to be farmed, since they have no wild relatives with which they are sexually compatible here. On the other hand, squashes and sunflowers do.

There are ways that this policy might go wrong, however. A desperate farmer might ignore the law and plant a transgenic crop that can breed with local weeds. And crop-to-weed gene exchanges are only one kind of change that transgenic crops can bring. Researchers have been developing a transgenic potato, for example, that can fight off the aphids that feed on it. The new potato produces a protein called lectin that ruins the aphids’ digestion. Greenhouse tests have shown that this transgenic potato can reduce populations of the peach-potato aphid by half. That’s impressive but not quite good enough to allow the potatoes to survive on their own. To fully protect their crop, farmers would need to introduce aphid-devouring ladybugs.

But as entomologist Nick Birch of the Scottish Crop Research Institute in Dundee has shown, the lectin in the potato makes ladybugs ill: after eating transgenic potato-glutted aphids, ladybugs produce far fewer offspring and live much shorter lives. Yet even though he has shown how transgenic crops can have harmful effects that spread through a food chain, Birch doesn’t think his results are cause for alarm. If ladybugs can also find aphids in the wild that are unaffected by transgenic potatoes, the plant’s harmful effects will be diluted. In general, Birch thinks that with careful tests of their potential effects, transgenic crops can prove safe—and useful in reducing our dependence on pesticides.

To critics, this sort of cautious optimism is not yet warranted. They view what’s happening now as a vast uncontrolled experiment with consequences we cannot predict—and promises that may never be met.

When the California-based biotech firm Calgene began selling the slow-ripening Flavr-Savr tomato—the first transgenic crop to be introduced in the United States—in 1994, it promoted the launch with a flurry of shiny tomato-shaped leaflets boasting Summertime Taste . . . Year-round! For more information, the public was urged to dial a handy number: 1-800-34tomato.

Call the number now and you’ll hear an anonymous voice telling you it’s been disconnected. Alas, the Flavr-Savr tomato—which incorporates a transgene that allows it to grow red on the vine without getting squashy—has been withdrawn from sale. Monsanto, which bought Calgene last May, cites production and distribution problems. Apparently the tomato just wasn’t tough enough to survive a bumpy ride down a conveyor belt.

The failure of the Flavr-Savr highlights a problem that has nothing to do with safety or gene escape: it’s not clear if transgenic crops will actually live up to corporate claims. Some crops have done modestly well, while the performance of two closely watched transgenics—both produced by Monsanto—have proved embarrassing. One crop, Roundup Ready cotton, was designed by Monsanto to hold up against the company’s herbicide Roundup. Last fall, in its first season, it ignominiously dropped its bolls all over the fields of some Mississippi farmers who had paid to try it out. In February the company began compensating them for their losses. Another kind of cotton, called Bollgard, was designed to ward off bollworms by producing Bt, the insecticidal bacterial toxin. In its trial season in 1996, the Bollgard plants did produce Bt as promised—but not enough Bt to fight off that year’s particularly bad outbreak of bollworms. Some disgruntled farmers had to spray their transgenic crops with old-fashioned pesticides.

Even if Bollgard should be able to produce higher levels of Bt, some critics still think it is doomed to eventual failure thanks to the evolution of resistance. Often a conventional pesticide kills all but a few insects that by chance carry a gene for resistance to the toxin. The survivors then reproduce quickly until they reach former levels, and most of them are now impervious to the pesticide. Some farmers have sprayed Bt on their crops in the past, but insects weren’t able to evolve resistance to it because the chemical broke down rapidly in sunlight. But if you put Bt into the crop, then the pest will be exposed to it from the moment the seed comes up until the plant dies, says Margaret Mellon, director of the Union of Concerned Scientists’ Agriculture and Biotechnology Program. That will create a powerful force for the selection of resistant insects, and Mellon suspects that it would make Bt a useless pesticide in less than five years.

Monsanto counters that resistance can be avoided by preserving refuges of plants that lack Bt. These islands will allow susceptible insects to thrive, and by breeding with the insects exposed to the Bt-engineered cotton, they will dilute any growing resistance out of the gene pool. But Mellon questions whether every farmer would voluntarily set up these refuges, which would presumably be devastated by the pests and produce no profits. If the insects should evolve resistance, crops such as Bollgard, despite all their high-tech armor, will be useless.
Comment on this article

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar

Log in to your account

Email address:
Remember me
Forgot your password?
No problem. Click here to have it emailed to you.

Not registered yet?

Register now for FREE. It takes only a few seconds to complete. Register now »