Skip to Content

Biotech Goes Wild

Genetic engineering will be essential to feed the world’s billions. But could it unleash a race of “superweeds”? No one seems to know. And nobody’s in charge of finding out.

A few miles outside Sacramento, several large greenhouses sit behind a fence. In the summer the familiar heads of sunflowers are visible through the glass and in the fields surrounding the greenhouses. The plants are tall, straight and healthy, with thick leaves that reach for the California sunlight. They look exactly like sunflower plants grown throughout the United States-except for the plastic cages around each flower.

The flowers are covered by biologists at Pioneer Hi-Bred’s research facility in Woodland, Calif., which owns the greenhouses, the fields around them, and the sunflowers in both. The plants are transgenic-that is, genes from other organisms have been inserted into their chromosomes. Caging the sunflower heads helps prevent the breeze from wafting genetically engineered pollen around the area, which would violate federal laws banning release of unapproved transgenic organisms.

To protect Pioneer’s trade secrets, the researchers are chary of discussing their work, but government permits suggest that the sunflowers in Woodland have been subjected to the full armamentarium of contemporary biotechnology. Pumped up by genes from as many as a half a dozen other species, the plants repel moths and viruses, fight off fungus diseases, and produce seed with a shelf life beyond that of their nonengineered cousins. To Pioneer, these super-sunflowers, as they are sometimes called, will be a small but significant step forward in the struggle to feed the world’s exploding population, which is projected to level off at 10 billion or so. But to critics, they-and the agricultural biotechnology that created them-are an ecological menace that will wreck the natural systems on which human life depends.

The battle between these entrenched views is fierce. In the last year, farmers and activists ruined five metric tons of transgenic seed in France, trashed fields of genetically altered crops in Germany, and convinced seven European supermarket chains to stop selling store-brand goods containing bioengineered products. This February, a coalition of 70 groups and individuals sued the U.S. Food and Drug Administration to block the use of a dozen transgenic crops as an “imminent” threat to the environment.

Even as the U.S. government promotes agricultural biotechnology, European countries are backing away from what activists call “Frankenfoods.” Austria and Luxembourg have banned genetically modified corn; Norway has also outlawed the corn as well as five other biotech crops; France has prohibited all transgenic plants. To push the British government to enact a moratorium, Greenpeace dumped four tons of genetically modified soybeans outside 10 Downing Street in February.

Biotech’s supporters, on the other hand, argue that it will create nothing less than a second Green Revolution. In the first, agricultural scientists used conventional breeding techniques to create the high-yielding strains of wheat and rice that have doubled world grain harvests since the 1950s. During that time the number of hungry people fell by three-quarters, according to the U.N. Food and Agricultural Organization, despite a huge population increase. But global population numbers continue to rise, and researchers now must do it all over again. According to a projection released last August by the International Food Policy Research Institute, a think tank in Washington, D.C., world demand for rice, wheat and maize will increase 40 percent by 2020-and the only way to feed those mouths is through biotechnology. If activists succeed in banning transgenic crops, argues Robert L. Evenson, an agricultural economist at Yale University, they will end up “hurting the poor of three continents.”

Caught between these extremes is a group of agricultural ecologists and plant geneticists who are trying to understand the implications of the new technology. Although some activists claim genetically altered crops are a direct threat to human health, researchers generally dismiss such fears: There is little evidence that transgenic genes, in and of themselves, are likely to be toxic or promote disease. However, biologists do believe that in some cases foreign genes in crops can pass into other, nonagricultural species, with potentially dangerous effects. “It’s inevitable that they will get out,” says ecologist Joy Bergelson of the University of Chicago. “That doesn’t necessarily mean that there will be negative repercussions. But there could be some. And right now we don’t know enough about what they could be and when they could occur.”

“The technology is brilliant,” says Paul Arriola, a plant geneticist at Elmhurst College, in Elmhurst, Ill. “In many respects, it’s a godsend.” Nonetheless, Arriola believes biotech is outpacing both the scientific understanding of its risks and the development of a regulatory apparatus to supervise its use. Because, in Arriola’s view, “we don’t really know what to regulate, or how to do it,” the world is in the middle of “a huge, ongoing experiment. We could create a real environmental mess. And that could stop this technology from doing some real good.”

Superweeds

The fight over transgenic farming is anything but academic. In 1996, the first year transgenic seed was widely available, farmers planted 1.74 million hectares (4.3 million acres) of the new varieties. This year, according to Clive James, head of the nonprofit International Service for the Acquisition of Agribiotech Applications, as many as 50 million hectares worldwide-an area bigger than Germany-are planted with genetically modified crops. “It’s one of the fastest adoptions of technology I’ve ever seen,” James says.

About three-quarters of that land is in the United States, most of it planted in bioengineered corn and soybeans. But the technology is growing even faster in Argentina-the area the country devoted to transgenic soybeans tripled between 1997 and 1998. Although exact figures are not available, China, the world’s biggest producer of cotton and tobacco, is, according to James, “aggressively increasing” the land planted with genetically altered versions of both crops.

By far the most important bioengineered trait today is herbicide tolerance, which accounts for two-thirds of all transgenic crops. A technology dominated by Monsanto, it lets plants withstand the use of selected weed-killing chemicals, so that farmers can apply them without fear of destroying their crops. Monsanto’s “Roundup Ready” soybeans, which resist the company’s Roundup herbicide, were introduced in 1996; last year, they covered an estimated 10 million hectares-a third of the U.S. farmland devoted to that crop. Next in importance is insect-resistant corn, including DekalBt corn, modified by Monsanto’s recently acquired Dekalb subsidiary to produce a bacterial insecticide, and StarLink corn, produced by AgrEvo, a joint venture of German chemical giants Hoechst and Schering. Principally aimed at fighting off the European corn borer, transgenic corn last year occupied 6.5 million hectares in the United States-a fifth of the nation’s total corn crop.

More-much more-is on the way. As sales of bioengineered seeds rose from $75 million in 1995 to more than $1.5 billion last year, half a dozen huge companies in Europe and the United States positioned themselves to exploit a market that is widely believed to be on the verge of exploding. According to U.S. Department of Agriculture records, some 4,500 genetically altered plant varieties have been tested in this country, more than 1,000 in the last year alone. About 50 have already been approved for unlimited release, including 13 varieties of corn, 11 tomatoes, four soybeans, two squashes, and even a type of radicchio. Hundreds more are in the pipeline, among them plants that will produce industrial and pharmaceutical chemicals (see past issue “The Next Biotech Harvest”).

This rush to market alarms some biologists, who believe transgenic crops are being released before the environmental implications are understood. The most immediate worry is whether genetically engineered crops will spontaneously breed with their wild relatives, creating hybrid “superweeds.” Just as a single Brazilian bee researcher created a continent-wide nuisance by accidentally letting aggressive African bees hybridize with gentle domestic bees, the release of alien genes could, in theory, produce noxious “killer-bee” plants.

Surprisingly little is known about such natural hybridization, explains plant geneticist Norman C. Ellstrand of the University of California at Riverside. Until recently, agricultural scientists focused on protecting farmers; the small amount of hybridization research done in the past primarily concerned the introgression of genes from the wild into cultivated species, rather than the other way around. “People had the idea that [crop-weed hybridization] wasn’t a very common or interesting phenomenon,” Ellstrand says. “But when they finally got around to looking at it, they basically spent a lot of time being surprised at what could happen.”

Initially, scientists thought genes were unlikely to flow from transgenic crops to weeds, because known crop-weed hybrids are often sterile. But last September, Bergelson and two Chicago colleagues startled researchers with a study of Arabidopsis thaliana, a mustard species often used as a test organism by plant geneticists. Usually, the plant pollinates itself, implying to scientists that foreign genes in transgenic A. thaliana would not escape by hybridization. But after the researchers planted ordinary A. thaliana, transgenic herbicide-resistant A. thaliana, and a naturally occurring, herbicide-resistant mutant variety, they learned that the transgenic plants were 20 times more likely to outcross than the mutants-they were “promiscuous,” as a headline in the journal Nature put it. “Nobody knows why,” Bergelson says. “We’re still trying to find the mechanism that drives the pattern we saw. There’s a lot we don’t understand, including how common it is.”

The implications are ominous. A decade ago, for instance, European sugar beets spontaneously mixed with a wild relative, creating a hybrid species that is now a continent-wide problem. Whereas the sugar beet is biannual-the root is harvested at the end of the second year -the new weed is an annual. At the end of the year, Ellstrand says, “the root turns into a chunk of wood that damages farm equipment or gets into the sugar-beet processing plant and screws up the machinery. You can’t kill it with an herbicide because any herbicide that gets the weed hits its relative. It’s not until the thing blooms and flowers that you see it, and by that time it has set seed that gets into the beet field forever.”

Transgenic crops have already shown the potential to create similar problems. The prospect of herbicide- or insect-resistant superweeds is particularly dismaying. In 1995, Monsanto and AgrEvo introduced herbicide-tolerant oilseed rape (Brassica napus), the plant that is the source of canola oil. One year later, an 11-member team from the Scottish Crop Research Institute reported, to scientists’ surprise, that pollen from oilseed rape fields can travel as much as two kilometers. At almost the same time, three Danish geneticists discovered that transgenic Brassica napus readily breeds with a weedy relative, Brassica campestris. The resulting plants look much like B. campestris-but are unaffected by herbicides. Taken together, says Dean Chamberlain of the University of North Carolina at Greensboro, the two reports “showed that hybridization is a real concern and that you need a very large buffer area around your plot to control it.”

When Ellstrand reviewed the literature on the 30 most agriculturally important plant species, most scientists he consulted believed few hybridize easily. In fact, he found evidence that more than 25 of the crops can break the species barrier, sometimes with unrelated species. Included in that list is wheat, which Robert S. Zemetra and his colleagues at the University of Idaho reported in April can outcross with bearded goatgrass, a problem weed in the western United States.

“What really shocks me as a biologist is that you have two species with different numbers of chromosomes hybridizing,” says Allison Snow, a botanist at Ohio State. “Goatgrass has 28 chromosomes and wheat has 42, but they can cross.” Biologists have regarded viable offspring from such mismatches as almost impossible. As a result, they thought the range of species that could hybridize was limited. The goatgrass-wheat hybridization suggests that the range is bigger than had been thought.

“You get very low rates of reproduction,” Snow says. “But when you’re talking about acres and acres of wheat with goatgrass all around them, even a very low probability event can occur.” If hybridization created insect-resistant goatgrass in areas where the weed’s spread is naturally controlled by insects, she says, “that could end up being the only kind of goatgrass you have, and then you might end up with even larger infestations of it than we already have.” Such fears are one reason that insect-resistant Bt crops-which contain genes from the bacterium Bacillus thuringiensis-have been targeted by activists.

In the United States, transgenic corn is unlikely to pose much risk of hybridization because it has no close relatives. But Mexico has teocinte, the wild plant that may be the ancestor of modern corn. What would happen if Mexican farmers planted bioengineered corn? Could the new genes affect the fitness of teocinte, which some agricultural ecologists view as a potential storehouse of valuable genes for future corn breeders? “With the information we have now,” Snow says, “it’s hard to tell when the long-term risks are serious enough to ban certain crops.”

Looming behind the ecologists’ fears is the belief that molecular biologists who work with DNA on the laboratory bench don’t understand fully how it behaves in the field. According to Rosemary S. Hails of the British National Environmental Research Council’s Institute of Virology and Environmental Microbiology, “The risk assessment of transgenic organisms is a multidisciplinary subject, which should include ecologists, molecular biologists, agronomists and sociologists.” Instead, companies tend to delegate decisions about the release of transgenic crops to molecular biologists-who are not trained to appreciate the full complexity of how the genetic code interacts with environmental factors.

“How fast would a new weed get around?” Snow asks. “Nobody really knows. I’m sort of assuming that most of these crops will be approved eventually and people like me will study what the consequences are. Then, after the cat is out of the bag, we may figure out how to regulate this technology.”

A Hungry World

Given these risks, why do so many of these scientists support the continued development of agricultural biotechnology? One answer is witchweed. Witchweed, the common name for three species in the genus Striga, is a parasitic plant that feeds on the roots of cereals and legumes in much of Africa. Attacking maize, sorghum and millet-the continent’s three most important cereal crops-Striga, in the view of Gebisa Ejeta, an agronomist at Purdue University, is a “scourge” of African agriculture. It has been estimated that the weed destroys 40 percent of the continent’s total cereals harvest-a staggering loss in the world’s hungriest places.

From a biological perspective, Striga is fascinating. Its seeds, smaller than grains of sand, lie dormant for as long as 20 years, waking only when aroused by a chemical emitted by the roots of the host plant. While still underground, the parasite plants develop root-like organs called haustoriums, which penetrate the host roots and siphon nutrients. Scores or hundreds of Striga plants can attack the same host. Witchweed eventually grows into fields of five-foot-tall plants with pretty pink flowers, but by that time it has long destroyed the crops it feeds on. Because each plant produces as much as 100,000 seeds, witchweed is almost impossible to eradicate-the United States spent four decades wiping out a single small outbreak in the Carolinas.

Because witchweed rapidly adapts to new hosts, losses in Africa keep growing. When the parasite made it impossible to grow sorghum in eastern Sudan, desperate farmers tried to grow pearl millet. At first millet was immune. But within a few years witchweed was wreaking havoc on the new crop, too. “People are literally starving because of Striga,” says Ejeta.

Who’s Watching the Greenhouse?

The uncertainty is due, in part, to the lack of a rigorous regulatory framework to sort out the risks inherent in agricultural biotech. The plastic cages covering the heads of the sunflowers help keep the transgenic pollen out of the environment, a general requirement for obtaining a federal permit to grow a test crop of bioengineered plants. But other than monitoring the plots, the government imposes few conditions on biotech tests. The main reason is that Congress has not passed any specific environmental law for genetically engineered agriculture. Instead, transgenic crops are evaluated by three overlapping federal agencies: the Food and Drug Administration, the Environmental Protection Agency, and the Department of Agriculture.

Each government agency has a different statutory responsibility, which sometimes leads to anomalies-and gaps in regulations. The FDA, for example, doesn’t look at the safety of foods that have been engineered to express pesticides, because pesticides are by law exempt from the agency’s purview. Nor does the EPA, which is required to treat such foods as pesticides. Because pesticides, of course, are toxic substances, the agency only establishes human “tolerances” for each compound. (Responding to critics’ concerns, the agency announced this spring that it may rethink its approach.) For its part, the USDA simply tries to make sure that the crop grows in the way that the manufacturer says it will. The disjointed legal mandates, observes EPA biotechnology adviser Elizabeth Milewski, “make life interesting.”

One worrying consequence of this patchwork of regulations is that no one has direct responsibility for looking at long-term effects on the environment. “We have a first-approximation understanding of the population biology of these plants and the insects, microbes and virus populations,” says Neal Stewart, a biologist at the University of North Carolina at Greensboro. “But we know very little about the community ecology and virtually nothing about the ecosystem ecology of what these genes will do. And we are not pursuing this knowledge actively.” Stewart’s concerns bore fruit in May, when Cornell scientists reported that pollen from Bt corn can kill the caterpillars of monarch butterflies.

According to Sally McCammon, science adviser to the USDA Animal and Plant Health Inspection Service, biotech field trials can be of any size and last for any length of time, though one or two years is the standard. From the companies’ point of view, the tests are efforts to learn whether new crop varieties will perform as intended. The government’s main job, McCammon says, “is to certify that the test is biologically contained.” Transgenic plants must be kept apart from plants they might cross-pollinate. “Afterwards you have to account for it,” McCammon says. “We make sure that you bag what you take out and that the plant material is plowed under.”

These measures are necessary, to Snow’s way of thinking. But by ensuring that transgenic genes won’t escape into the environment, they also make it impossible to learn what will happen if they do. “The ecological questions don’t even get touched,” she says. “In fact, it’s illegal to touch them.” She believes that the environment and industry would be better served by introducing a second level of testing devoted to ecological questions. Another step, in her view, would be to fund academic research into the ecological hazards-currently the sole source of federal funds, the biotechnology-risks panel of the USDA, has a budget of less than $2 million.

Technical controls may also be possible, says Gressel of the Weizmann Institute. Most transgenic crops today have a single foreign gene. But companies are already working on inserting several genes simultaneously into the plant’s genome. In a May article in the journal Trends in Biotechnology, Gressel argues that if these multiple genes were inserted in close proximity to each other on the chromosome, potential hybrids would inherit all of them at once. And if the secondary genes coded for traits such as preventing dormancy, the hybrids would be less, not more, dangerous than their wild parents. For crops, the inability to lie dormant doesn’t matter, because the seed is harvested and replanted each year. But a weed that is unable to produce seed that can remain dormant until an opportune time to germinate is at a significant disadvantage. “The hybrid weed will be weaker, not stronger,” Gressel says.

“I’m more worried about the future than the present,” Ellstrand says. “So far it’s okay-we don’t have killer tomatoes flying through the air. But we need to be thoughtful and careful about what we’re doing, and there are some people and some portions of the industry where they have a better tradition of that than others. People who have worked with plants outside in real life seem to have a better handle on it than people who have worked with chemicals all their life. If we keep paying attention to what’s happening in the field, we might be able to make this technology realize its promise.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.