Genesis of Eden Diversity Encyclopedia

Get the Genesis of Eden AV-CD by secure internet order >> CLICK_HERE
Windows / Mac Compatible. Includes live video seminars, enchanting renewal songs and a thousand page illustrated codex.

Join  SAKINA-Weave A transformative network reflowering Earth's living diversity in gender reunion.

Return to Genesis of Eden?

Living in a GM World NS 31 Oct 98 29

Even as you read this page, genetically engineered crops of soya bean, maize, oilseed rape and potatoes are growing in fields dotted around the US, Canada, Argentina and elsewhere. An area of land the size of Great Britain is now home to these transgenic plants. And more are on the way. Suddenly, plant science is no longer a quiet backwater for genial professors and their cuttings. lt is the stuff of big business, patent rivalries and closely guarded technical tricks. If you believe biotech's gainsayers, this brave new plant science is also ushering in a dark age in which all genes will bear a 'no trespassing" sign, and the companies that own them will move them from species to species like Lego bricks, to the detriment of what's left of the natural world and our respect for it. We've seen lesser versions of the shock before. When scientists began manipulating the genes of bacteria in the early 1970s, environmentalists and politicians, especially in the US, voiced many of the some apocalyptic fears about scientists playing God and the risks of genetically manipulated organisms escaping. A quarter of a century later, chemicals and p harmaceuticals plants are home to great stainless steel vats of GM bacteria producing scores of proteins and enzymes for medicine and industry-and nobody minds.

But complacency would be unwise, for biotechnology clearly has entered a new phase. Many of the organisms researchers are manipulating are more complex than bacte. rid and have greater emotional resonance for humans, either because they are mammals or part of our food supply, or both. Even with GM bacteria, researchers' ambitions have grown. Many no longer want merely to keep them in vats behind closed doors. They want to set them to work in the wider world, doing jobs such as neutralising toxic waste. For now, though, the debate about the pros and cons of living in a GM world rages around transgenic plants. And rightly so, because their impact will be felt soonest. In the following articles, we look at the issues behind the controversy. Some problems seem to have been wildly exaggerated-the idea of 'Frankenfoods"' being inherently unsafe, for example (see p 42). Others, notably the risk of transgenes escaping to create 'genetic pollution', may have technical fixes (p 38). But still others are more complex (ps 46 and 50). Even though GM crops do not spell disaster for wildlife and the developing world, their impact is unlikely to be wholly benign. But firstl the quest for the blue rose ...

Brave New Rose

TURNING a rose blue. In an era when researchers can clone mammals and insert genes into plants to ward off cropdevouring insects, you would think this WOL]ld be easy. But it isn't.

Ask Edwina Cornish. Years ago, this Australian biotechnologist and her colleagues began a quest to create in the lab what cannot be created by breeding. They founded a company, Florigene in Collingwood, Victoria. They raised money for the research. They cloned the gene that enables petunias to produce the blue pigment that roses lack. But when they inserted the gene into rose cells, the resulting flower was no bluer than, well, a rose.

Then there is the mysterious case of the mutant loblolly pine. Another dream of plant engineers is to create easy-to-pulp trees. For years, researchers believed the key in all species was an enzyme called cinnamyl alcohol dehydrogenase, or CAD. This, after all, was what the textbooks said all woody plants used to synthesise the lignin polymers that make cell walls sturdy and the extraction of cellulose costly. But then last year, unexpectedly, Ronald Sederoff of North Carolina State University in Raleigh and his colleagues uncovered some mutant pines that broke the rules. The trees had a mutation that blocked all production of the CAD enzyme-yet they still made plenty of lignin. In pine trees at least, genetically manipulating levels of this enzyme would not dramatically help the pulp extractors. You get the picture. Plant biochemistry is turning out to be more unpredictableand harder to tame-than researchers had thought. Even the researchers say so. "In the early days it was easy to be optimistic," says Cornish. "We might have underestimated how long things would take and the complexity of the pathways we were trying to manipulate." But here's the rub. Hard to tame does not mean impossible to tame. Slowly but surely, researchers like Cornish and Sederoff are getting to grips with the complexities of engineering plants. Slowly they are laying the foundations for a world where the initials "GM" will come to prefix far more than just genetically manipulated tomato puree and soya beans. It might take five years, it might take twenty, but we will have genetically modified roses thit are blue, along perhaps with GM geraniums that smell of roses, GM orchids that glow when thev need watering, GM leylandii hedges that stop growing at a reasonable height, GM lawns that (almost) never need mowina", and GM bin liners made from plastics synthesised in plants. Not to mention GM newspapers and wallpaper. if this sounds silly, think what has been achieved so far. Fifteen years ago, there was just one technique, based on the grown gall bacterium, for ferrying genes across the thick walls of plant cells. Now there are several, including two types of gun for propelling DNA into cells at high speed. A decade ago, researchers knew almost nothing about the genes that control the shapes, sizes and flowering characteristics of plants. Now dozens of such genes have been identified and a project to sequence the entire genome of a flowering plant, a weed called Arabidopsis thaliana, is nearing completion. Already efforts are well under way to engineer potatoes to double up as ,,accines; to create transgenic "smart" plants that will use a fluorescent "SOS" protein to give farmers or growers early warning of drought or disease; to equip oilseed rape with bacterial genes for producing biodegradable plastic; and to engineer cotton plants to produce wrinkle-free fibres.

Forests of clones

One by one, even trees, which are notoriously tricky to grow from tissue cultures, as genetic engineering demands, are falling under the spell of biotechnology. As a result, timber and pulp will increasingly come from high-tech plantations where the trees are all clones, engineered to carry new genes for pest and disease resistance, and perhaps made sterile to prevent these transgenes escaping via pollen, says David Ellis of the BCRI Forest Biotechnology Centre in Vancouver. Some tree plantations, mostly in the southern hemisphere, already consist of genetically identical trees, mostlv produced the way gardeners and farmers have cloned plants for millennia-with vegetative cuttings. But as genetic engineering takes off, more and more forestry plantations will begin life as so many cloned tree embryos, frozen until they are needed and then cultivated in vast hydroponic vats. Granted, cloning can be labour intensive and genetic uniformity is not always desirable. But many growers are keen on it because it enables them to raise the quality of all their trees to that of the best. The availability of "elite" genetically manipulated trees will make them even keener. "Biotechnology will accelerate the trend toward clonal forestry," predicts Martin Maunders of Cambridge-based biotech company ATS. One "elite" trait would be easy-to-extract pulp. Ellis points out that despite last year's mutant pine surprise, the synthesis of lignin in other commercial tree species is actually "very well characterised". "We know and have isolated every gene in the biosynthetic path," says Ellis. And in eucalyptus and poplar trees at least, engineering levels of the CAD enzyme does make pulp easier to extract. Researchers elsewhere are experimenting with genes that may boost the growth of trees during winter months or curb the height of fruit trees so they take up less space and their fruit is easier to harvest. Mini cherry trees small enough for the tiniest city garden could be just a few years away. In labs like Sederoff's, meanwhile, efforts are under way to identify genes that affect wood strength and density. Where will it all lead? "A short, fat, fast-growing tree" might be the thing of the future, says Ellis, only half in jest. "With no taper so you don't waste space on the logging trucks." And with technicoloured timber perhaps. For there's nothing about the biology of plant pigments that means grass has to be green or that wood has to be a yellowy brown. Polka dot button holes to match your tie or scarf are some way off, but already a couple of transgenic carnations that are mauve rather than the usual pink, yellow, white or red are being sold by florists in Australia, Japan and the US. The carnations owe their strange hue to the pigment gene Cornish and her colleagues cloned from petunias-the one that has so far failed to turn roses blue. Why the gene works in carnations (up to a point) but not in roses isn't entirely clear. But the researchers suspect petal acidity is a major factor. The gene encodes an enzyme needed to synthesise blue pigment molecules called delphinidins, which are lacking in both roses and carnations. The problem for roses is that these molecules are only blue at high pH, and the cellular cavities, or vacuoles, that hold petal pigments in roses are acidic. To solve the problem, Cornish and her colleagues are pinning their hopes on one of two options-finding a conventional rose variety that is less acidic, or cloning the genes that control petal pH so that they can alter conditions in the vacuoles. Even then, there remains a risk that the rose's natural pigment molecules, the red cyanadins and orangy perlagonidins, will drown out the added blue. One reason why turning grass blue or red might be easier than it sounds is that any biochemical changes might only need to be skin deep. For instance, genetic engineers could use a pigment gene hooked up to a piece of DNA that keeps the gene switched off in all but the outer layer of cells. Another approach might be to make use of silent and unused pigment genes. After all, the green stems and leaves of ornamental flowers have the same genes as the petals. "The genes are there," explains Cornish, "but they are expressed in the flowers, not the leaves." In theory, genetic engineers could rouse these pigment genes from their slumber, producing leaves and stems awash with floral pigments. So, when blue roses do finally begin to emerge from labs, perhaps some them will have the chance to express their native pinkness in their leaves. Some might also have the chance to smell of lemons. "Some people find sweet roses overwhelming, and most cut roses have almost no odour at all," says Michael Dobres. Two years ago, in Philadelphia, Dobres helped found a company called NovaFlora that aims to remedy this sorry state of affairs. One of their projects involves inserting a gene into roses that would enable their petals to produce lemon fragrance molecules. The gene encodes an enzyme called limonene synthase, which citrus plants use to synthesise scent molecules known as monoterpenes. The researchers have already given the gene to petunias and are waiting for their first crop of what they hope will be a lemon scented transgenic flower. Limonene synthase is only one way to perk up scentless plants. "There are hundreds of different monoterpenes out there, synthesised by different enyzymes," says Dobres. Not to mention two other major types of plant fragrance molecule. In future, predicts Dobres, genetic engineers will be able to create finely-tuned fragrances to order in almost any plant. Among the many possibilities would be lemon scented golf courses and GM camomile lawns that are much easier to maintain than the traditional kind. And as for the idea of Calvin Klein scented GM roses. "That would be dynamite," says Dobres. "That's something we definitely aspire to." Of course, achieving all this won't be easy. The scent molecules that transgenic plants make will be produced in vain if they remain trapped inside their tissues. One reason many commercial cut flowers are so odourless in the first place is that breeders select for tough petals with waxy coats. Then again, perhaps genetic engineering could be used to make these coats permeable to scent molecules. It can certainly be used to alter the shape, form and number of flowers that a plant produces. Knowledge of the gene code which specifies the physical arrangement of a flower's sepals, petals, stamens and carpels is so advanced that it is already possible to design "fantasy flowers" that have any of these organs in any position in the flower. And genetic engineers can also alter when a plant flowers. At the University of Leicester, Garry Whitelam and his colleagues have engineered asters so that they flower in the middle of winter, not just in summer. Growing conventional cut flowers in greenhouses in winter is expensive because of the extra lighting needed to make them flower. In a bid to cut costs, the researchers manipulated an aster gene so that it would produce higher than normal levels of a phytochrome protein that enables plants to sense changes in dayiength. The GM asters required only 6 hours of daylight to flower compared with the usual 14.

And when it comes to manipulating the sex lives of plants, this is only the tip of the iceberg, thanks in no small part to Arabidopsis tiialiana. In less than two decades this unprepossessing weed with white flowers has risen from obscurity to become the megastar of plant science. The attraction for researchers is that it has an unusually small genome and grows to maturity in just six weeks. And in the 1980s, they decided to make it their fruit fly-the model organism they would mine for important genes involved in plant development. In the past few years, such genes have been tumbling out of Arabidopsis labs, turning the heads of plant biotechnologists everywhere. Three years ago for instance, Detlef Weigel of the Sail, tute in La jolla and Ove Nilsson at the Swedish University of Agricultural Sciences in UmeA identified two genes in Arabidopsis that act as master switches for triggering flower formation at the ends of shoots. When the researchers engineered Arabidopsis so that the genes would be active all over the plant, every shoot produced a flower. And when they inserted one of the two genes, leafy, into aspen, a tree that normally takes up to two decades to flower was fertile after two months. A spectacular result given that slow sexual development is the bugbear of tree breeding.

Death signal

Other researchers are exploring ways of using Arabidopsis genes to do the exact opposite-prevent flowering. And not just to prevent transgenes spreading into wild relatives. For annual crops such as lettuce and potato plants, flowering is a prelude to death. It sends a signal to the leaves telling them to shut down photosynthesis. Blocking that signal might mean farmers could grow the crops for longer and perhaps get bigger yields because the plants would no longer need to invest resources in making flowers. Nobody has engineered crops this way yet, but the discovery of an Arabidopsis gene called Frigida could encourage researchers to try. In the weed, the gene seems to function as its name suggeststo prevent flowering, or at least to delay it until winter is over. "It would be nice to stick the gene into sugar beet and see what happens," says Caroline Dean, at the John Innes Centre in Norwich. In future, farmers and growers may even use chemical sprays to make their genetically engineered plants flower on cue. Earlier this year, Brian Tomsett and Mark Caddick at the University of Liverpool used an alcohol-sensitive gene from a fungi to make the activities of plant genes controllable from the outside. Simply drenching the roots of the engineered plant with alcohol was sufficient to switch on a gene that stunted growth. And why stop there? Why not manipulate plants so that they can change, on cue, their colour or fragrance? Why not engineer fast growing hedges whose growth can be "switched off " once they reach the required height? Why not... create a GM world? David Concar

Apocalypse When?

Jeremy Rifkin may look more like a prosperous banker than an environmental activist, but he has long been the biotechnology industry's most famous critic. He has written fourteen books in his role as a roving guru of economic and trends, including Algeny, The End of Work and Declaration of a Heretic. During the 1980s, he was one of the key figures in a successful worldwide movement to limit the use of bovine growth hormone to boost milk yields. That time, says Rifkin, "we won, but as we were overproducing milk, the drug was a loser to start with-we were handed a soft ball-. At that time, too, the biotechnology industry was really in its infancy. In his new book, The Biotech Century, Rifkin explains that the industry is fast becoming a gi that the need for public debate is greater than ever. New Scientist met Rifkin and asked him to explain his heretical view of how the industry is developing, where he believes the dangers lie in genetic engineering of crops, why he thinks that the current style of agricultural biotechnology is doomed, and how he would like to see genetic science used. You may disagree strongly with his views but read on ...

"THE public still thinks of biotechnology as producing medical breakthroughs and some new crops. They might worry about the need for labelling of foods, but the industry is really looking at a lot more than that. Genetic engineering is going to produce chemicals, building materials, plastics, fibres, new sources of energy as well as drugs and vaccines. Watch the industry It is very impressive what has happened. In the past year, Monsanto, Novartis, Hoescht Chemical in Germany and Du Pont have all made a decision to shed or sell parts of their chemical divisions. They are all life science companies now. For them, the metamorphosis is to genetic commerce. That signals the passing of one era and the beginning of another. One of the primary purposes of my new book is to explain that the new economic era is going to bring about changes at least as significant as the Industrial Revolution, and it is going to be based on genetic commerce. I use the new term "genetic pollution" in this book. I think it is going to gain currency very quickly. We need to look at the scale of what is happening. Right now, the life science companies are introducing scores of genetically engineered organisms into experimental field tests and several into commercial production. But in the next ten years, the major companies hope to introduce thousands of gene-spliced organisms over millions of acres of land and even into the water. They will reseed the planet with a second genesis. We are not talking about just genetic food crops, but also microbes that eat up landfills and radioactive wastes, and plants designed to secrete chemicals, pharmaceuticals and plastics in every cell. This is an extraordinary change. The life science companies will argue that all they are doing is a more sophisticated form of breeding, and we have had breeding since the Neolithic revolution. That is their argument and it is wrong. Let me go through it systematically This is not an extension of classical breeding. In classical breeding, it is possible to cross relatives to create hybrids-you can cross a donkey and a horse and get a mule-but you can't cross a donkey and an oak tree. But with genetic engineering technology you can cross all the biological boundaries: you can make mice with human growth genes and you can have firefly genes lighting up tobacco plants. Once you can cross all biological boundaries, you begin to see a species as simply genetic information that is fluid. That brings us into a whole new way to conceptualise not only our relationship with nature, but how we use it. Let us take some examples. Suppose a company wants to put a plant into the environment that produces biodegradable plastic in every cell. Its executives come to you and say: "Gee, this is sustainable, this is what you want-it is biodegradable." But what they don't say is what happens when this plant is producing plastic in every cell over millions of acres. What happens when foraging birds and insects and microorganisms and animals come in contact with that plant? There are plants in research and development that have genes that code for specific pharmaceuticals, chemicals and vaccines. It is the scale that is important. Thousands of introductions propagating in land, water and air over the next twenty years: the life sciences companies want to have the land full of genetically engineered crops, both food and pharmaceuticals in less than 10 years. But have they addressed the issue of whether genetic pollution can spread? These companies want to place plants in the environment with herbicide, pesticide and viral resistance genes. They say that this is good for the environment because we will use less herbicide and pesticide. Well, let's look at what could happen. The obvious first problem is build-up of herbicide resistance in weeds. Instead of spraying here and there infrequently, you actually have that herbicide-resistant gene in your crop so you can spray and kill everything in sight without killing your crop. If you are putting a herbicideresistant gene in every cell of every plant over millions of acres, you certainly up the ante for the emergence of resistant strains of weed. That is the most simple, the build-up of resistance. But what if genes jump between species? The scientific community once said that was absurd. Now we are beginning to weigh the evidence. Danish studies of a herbicide-resistant gene show that during pollination it will jump easily over long distances. It is fixed in the genetic code of weedy relatives and is passed on to future generations. If you pass herbicide resistance to weeds, how do you recall that to the laboratory? Look at the landscape. There will be plants over millions of acres producing biodegradable plastics, chemicals and vaccines, all encoding for specific genes that can jump and fix for herbicide, pesticide and viral resistance in weedy relatives. I don't think you even have to be an alarmist. If ust a small fraction of these introductions turn out to be long-term pests, then we have irreversible damage to ecosystems.It could be devastating.

Who will be liable?

And who would pay if there is a catastrophe? No insurance company will provide cover. The insurance industry quietly let it be known early on that it would insure only for short-term crop damage and negligence. There is no long-term insurance. If one of their genes jumps-and it is easy to spot because if you have weeds fixing for herbicide and pesticide resistance and they proliferate, you will be able to identify whose gene that was-you will have a problem that could last for generations. There is no insurance company that will touch it. I've said that companies should go to Congress for a Price Anderson Act [the Act protecting the nuclear industry from catastrophic liability] but they would never do that because if they went for a Price Anderson Act then everyone would be alarmed. So they made the decision to go out without long-term catastrophic insurance. Why aren't members of Parliament and Congress asking who is going to be liable for losses? Will it be the government, home-owners, farmers? The reason the insurance industry and the reinsurance industry will not touch this is that they say we have no way of assigning risk. There is no predictive ecology. There is no ecological risk assessment science. I have been saying this for 15 years. Every government says it is regulating scientifically the introduction of genetically engineered organisms into the environment. But all the players know that there is absolutely no risk-assessment science by which to do it. The liability issue is the industry's Achilles heel. My own bet is that agricultural biotechnology is going to be one of the great disasters of corporate capitalist history. There are two reasons I think that. First, the life sciences industry has misjudged where the consumers are moving in terms of their food preferences. The middle class sets the trends in Europe, japan and North America. I shuttle back and forth every three weeks, and I can tell you that the middle class is moving towards organic foods. Secondly, liability is going to kill this industry. I really don't believe that it will be able to deal with the liability issue over genes jumping. It says it will take all these elaborate precautions and that every farmer will build hedges and fortresseswhat farmer is going do that? I think that it is likely that we will be plagued by genetic pollution, and that we will look back and see chemical and nuclear pollution as not as significanteven though one brought us global warming and the other waste that we cannot deal with for thousands of years. Genetic pollution could be more serious because of the volume of the introductions and because the introductions are so novel. We are placing living things in the environment where there is no evolutionary history on how to accommodate them. I think if you brought 10 ecologists into the room they might say the same thing. The life sciences industry may find it difficult to believe that its agenda could be rejected but remember that the babyboom generation rejected nuclear power, the crown jewel of 20th-century physics. When I was growing up in the 1950s that would have been unthinkable. There is a naive idea that there is only one agenda-are you for this science or are you a Luddite? Are you an advocate of change or are you scared to death of the future? That's not the debate. The issue is not whether you are for or against the science, the issue is what are the technologies we wish to use that hamess that science commercially and socially

Greatest opportunity

I am in favour of genetic science but I believe that there is a hard path and a soft path to using this science. Are we going to put our government research and development money into genetically engineered food which is really a primitive form of applied science? It's a form of science that started with the Baconian tradition. Find ways to engineer your plants so that they are in isolation from ecosystems and can defend themselves and ward off potential incursions from local ecosystems. It is a defensive, Baconian approach to applied science that we have been using for several hundred years. That is hard-path technology. But this same genetic science could be used for developing a sophisticated organic-based agricultural production system in the 21st century. That would be the soft path. With the new science we are learning all sorts of things about how genes are turned on and off and how they are expressed, about the relationship between genotype and phenotype. We have a lot of data on ecological dynamics, we could begin to use that science to create a very sophisticated organic-based agriculture. Instead of engineering our plants in isolation from ecosystems in a defensive way, we could find ways to better understand how traditional varieties fit into local ecosystems so that they are compatible and integrated. From an intellectual point of view, I really think that the shift into the age of genetic science is the greatest opportunity in history. It should force us all, this generation and the next generation, to ask all the big questions. VVhat is the value of a human being? Does life have intrinsic or just utility value? What is our obligation to future generations? What is our sense of responsibility to the creatures with which we coexist? It really does force us, if we are willing and open to have the debate, to rethink our humanity and the meaning of existence. I mean that in a very sincere way. I don't know if we will have the debate, but if we do, it will be a great opportunity and in the final analysis genetic science will be well served." Fl

Running Wild

THE NIGHTMARE scenario goes like this. Farmers blanket their fields with crops engineered to carry genes designed to defeat pests and disease. The crops pass these genes on to wild relatives, turning them into supercompetitive weeds that rampage through the countryside, wiping out rare and vulnerable plant species. The public is baying for blood. Conservationists are firing on all fronts. Faced with crippling lawsuits, biotechnology companies file for bankruptcy Colin Merritt sighs. He's heard it all a thousand times before and he isn't impressed. "Superweed is such an emotive word," he says. "It very often hasn't been thought through." Of course, as a scientist who works for Monsanto, the US biotech giant that stands to make billions from transgenic crops, he would say that. It's his job, just as surely as it's the job of the anti-biotech brigade to do the opposite, picketing transgenic crop fields dressed in full-body protective suits and muttering apocalyptic tales of "genetic pollution".

Not surprisingly, opinions are less polarised among independent researchers. Most agree that the possibility of transgenes being passed from crops to weeds cannot be ignored for the simple reason that plants often hybridise, exchanging their genes in the process, and sometimes producing offspring that are fertile. Indeed, crops such as oilseed rape and oats are themselves the descendants of hybrids. At the same time, however, calls for a blanket ban on the use of high-tech crops miss a vital point4. conventional crops can also pass genes to wild relatives and even become weeds in their own right. Whether a crop is or isn't engineered doesn't tell you whether it poses a threat to the environment, says Alan Gray, an ecologist at the Institute of Terrestrial Ecology at Furzebrook in Dorset "We need to concentrate on the specifics and look at the biology of individual genes in individual crops in individual locations," he stresses. And that, according to government officials in the US who have blazed a trail in approving the commercial growing of transgenic crops, is precisely what has been happening. Take soya beans and cotton, says Arnold Foudin, assistant director of scientific services at the US Department of Agriculture (USDA). Together these account for 90 per cent by area of all transgenic crops grown in the US. And the reason officials are confident the novel genes they carry won't spread and create superweeds is simple: there is nowhere for the genes to go. "The vast majority of crops we have are exotic and have no wild relatives," says Foudin.

Embarrassing relatives

If a crop has nothing to mate with but itself, then transgenes are unlikely to spread. But plenty of crops are grown in places where wild relatives abound, including oilseed rape, rice, maize and sugar beet. And grasses have an embarrassment of relatives with which they could mate. How do regulators deal with these plants? One option is to ban farmers from growing transgenic crops in areas where they have wild relatives. France has imposed a moratorium designed to do just that. Canada has forbidden the growing of transgenic oats because weedy relatives of oats are so widespread, while in the east of the country, farmers are banned from growing transgenic oilseed rape, again because of the presence of weedy relatives with which the crop could hybridise. Over the border, things are a little different. ln the US, farmers are growing high-tech crops such as transgenic oilseed rape despite the presence of weedy relatives in some areas. Regulators there have been less Draconian-or more reckless, depending on your point of view-putting more faith in experiments designed to measure the likelihood of genes jumping from crops to weeds, and their likely consequences. About I per cent of the USDA!s annual budget-a figure of 1 to 2 million dollars goes towards funding research on the environmental impact of transgenic crops. Andrew Paterson, an ecologist from the

Texas A&M University in College Station is looking at gene movement between cultivated sorghum and Johnsongrass, a weedy wild relative. "They can cross with one another," says Paterson. "The question is how often in nature do they do so?" To find out, Paterson's team is using DNA fingerprinting to sample the effects on Johnsongrass of the movement of billions of sorghum pollen grains over vast areas, a painstaking task. The biotech industry's own assessments of gene movement are usually less ambitious. In oilseed rape as in other crops, the industry's favourite modification to date has been to insert genes for herbicide tolerance. To discover if such genes can spread to weedy relatives, companies typically plant a mix of engineered crops and weeds in the same field. After pollination, seed set and the following season's growth, the whole field is sprayed with herbicide. Weeds which remain upright must be carrying the resistance gene. Not everyone is convinced these experiments have been adequate. "Some of the biotechnology companies haven't done the really big-scale work," says Jeremy Sweet from the National Institute of Agricultural Z' Botany in Cambridge. Monsanto, for example, tests engineered crops in plots the size of tennis courts before growing them commercially in fields that stretch beyond the horizon. Sweet is frequently asked by the British government to verify the risk assessments made by biotech companies. "We look at what happens when transgenic crops are grown on the commercial scale," he says.

'If transgenic crops are grown within pollination distance of wild relatives or non-transgenic varieties of the same crop, then at least some movement of their transgones seems inevitable'

And things change. For example, on small plots of oilseed rape, measuring about 20 metres by 20 metres, pollen seldom travels more than 3 metres. On fields of a more commercial scale, up to 50 hectares, that figure rises to about 2 kilometres. "Once we start growing transgenic oilseed rape in the UK on a big scale it will be everywhere," says Sweet. Nor, for oilseed rape at least, is pollen dispersal the only worry. The plant is already a widespread weed throughout Europe and North America. Each time the crop is harvested, moreover, there is a lot of seed spillage in and around the field, and such seed residues can persist for up to a decade. This could be bad news for farmers who switch from growing transgenic to conventional oilseed rape especially if more stringent food labelling rules come into force. Farmers wouldn't be able to guarantee that their conventional crop is free of "hidden" transgenic plants, nor would there be any way of preventing these plants from creating hybrids with the conventional crop. And, of course, any spilt seed can be spread farther by animals or seed lorries. Mike Crawley, an ecologist from Imperial College, London, found that the distribution of feral oilseed rape around London's M25 orbital motorway matches the transit routes of seed lorries.

In other words, agriculture is often messy, and the myth that transgenic crops can somehow be grown in splendid isolation is just that. If transgenic crops grow within pollination distance of wild relatives or non-transgenic varieties of the same crop, then at least some movement of transgenes seems inevitable. "It's obvious that transgenes will get out," says Geoff Squire from the Scottish Crop Research Institute in Dundee. "The big question is, does it matter?" From the segregation and food-labelling perspective, it clearly does. But the likely impact on the environment all depends on which genes jump into which weeds and where, says Joy Bergelson, an evolutionary biologist from the University of Chicago. In 1996, Thomas Mikkelsen and colleagues from the Riso Laboratory in Rofkilde in Denmark showed that herbicide resistance could spread rapidly from transgenic oilseed rape to a wild relative, resulting in a healthy herbicide-resistant weed. If the same happens in farmers' fields the benefits of growing herbicideresistant crops rapidly disappear. But what would happen beyond the confines of the field? In a field sprayed with herbicide, crops and weeds resistant to the chemical will outperform plants lacking the gene. But Bergelson wanted to see how herbicide-resistant plants might fare if they escaped into neighbouring fields and hedgerows where herbicide was never sprayed. The answer, for thale cress weed at least, is "badly". In a field containing a mixture of transgenic and non-transgenic plants, where no herbicide was sprayed, the transgene disappeared after five generations. Bergelson concluded that production of the herbicide-tolerance protein meant the plant had to spend more energy which, in the absence of herbicide, spelt death in the struggle for existence. Studies show that the same is true in many other crops, including tobacco, oilseed rape and rice. If it escapes from farmers' fields, the herbicide-resistance gene is more likely to create punyweeds than superweeds. For other transgenes, however, the picture could be different. Take crops that have been engineered for resistance to a specific insect pest or virus-Bt maize and cotton, for example, which contain a bacterial gene for a toxin that wards off caterpillars. In theory, these genes could confer a competitive edge on plants growing outside as well as inside farmers' fields. If weeds were to acquire them, they could become immune to the pest or to the virus. Would that matter? If the particular pest or virus is the main factor keeping the weed in check, it might. Ask anyone in the US about kudzu, multiflora rose or cheatgrass, and they'll confirm the havoc that exotic species with no natural predators or pathogens can cause. Fortunately for the biotech companies, the innate complexity of biological systems offers some comfort. "In most situations, there are many factors controlling the spread of a species," says Suzanne Warwick of the Eastern Cereal and Oilseed Research Centre in Ontario. So it's unlikely that the spread of a single gene for pest or virus resistance could create superweeds. But unlikely is not the same as impossible, and Bergelson believes that in the absence of any data on the consequences of such transgenes escaping, we would be unwise to relax. Besides in future more and more crops will be engineered to carry resistance genes for several different pests and pathogens. A weed that acquired multiple resistance really might escape natural control. Enter Monsanto's Merritt-armed with more reassurance. This sort of botanical: the nightmare won't happen, he says, because the industry and its regulators will continue to restrict viral and pest resistance genes to crops with no weedy relatives. In the US, he points out, farmers have been permitted to grow a genetically engineered potato that carries genes that make it resistant to both a virus and the Colorado beetlebut only because the potato has no known weedy relatives. If weeds do end up stronger and fitter thanks to an influx of crop genes, it won't be for the first time. Thanks to the efforts of classical breeders, crops resistant to herbicides, insects and viruses have actually been around for decades, and there have been 16 documented cases of genetic exchange between these crops and wild relatives. In Switzerland, for example, cultivated alfalfa, Medicago sativa, has largely displaced pure forms of the yellow flowering sickle medik through a tidal wave of hybridisation.

'There's no denying that engineering sterility would be an effectivee Way Of laying to rest the spectre of the transgenic superweed. Then, what Is good for the company coffers might be good for the environment"

Given that there has been no great uproar over such cases, why the fuss about genes escaping from high-tech crops? Why indeed, asks Merritt. Like others in the industry, he argues that the public has got it all wrong. A new genetically engineered crop is actually less risky than a new variety bred by conventional means. A few years ago, for example, breeders crossed the oat plant with its wild relative in order to introduce mildew resistance. The exercise resulted not just in the transfer of a single resistance gene, but also in the transfer of thousands of other genes, all of which increased the risk of the crop developing weedy characteristics. Genetic engineering is a scalpel to this sledgehammer, says Merritt. "We can be a lot more certain about the changes we are making than in conventional agriculture." Fine, but emotions also come into play.

In terms of cob size, for example, the maize that is grown today is a monster compared with the wild plant from which it was bred, but no matter because its genes all come from plants. In genetic engineering, a crop may be given a gene from a rat or a scorpion or a fish, and, like it or not, that changes the public's perception of risk-even if such transgenic varieties seldom make it from lab to field, and even if the genes in question pose no special safety problems. What's more, the precision argument doesn't always stand up. VVhen Bergelson engineered mustard plants for herbicide resistance, she discovered that they were twenty times more promiscuous than wild mustard or plants conventionally bred for herbicide resistance. Reporting the effect two months ago in Nature, Bergelson speculated that the introduced gene had disrupted genes affecting pollination and fertility. The result highlights the fact that to date geneticists have produced transgenic plants simply by shooting genes into plant cells and hoping for the best. As a result, they have been unable to control where the DNA ends up in the plant's genome. No matter, says Merritt, because "that sort of effect would have been detected long before a crop ever made it to the field". And maybe it would. But until the next wave of more controlled engineering techniques become standard, and as long as engineered plants retain the ability to reproduce and hybridise, we can no more rule out the possibility of a transgene acting unpredictably to harm the environment than we can rule it out when new varieties of conventional crops are planted. "Everything seems all right," says Sweet. "But who knows? Tomorrow an incident may occur."

Suicide Switch

For all these reasons, many biotechnologists increasingly believe that the ultimate answer is to engineer plant infertility. By destroying, or severely curbing, a plant's ability to produce flowers or viable pollen or seed, they believe they can make it all but impossible for a transgene to escape. Hitching a ride on a virus would remain the only way out-and the chance of that happening really would be vanishingly small. A wide range of genetic tricks for neutering crops are already in the pipeline. One of the earliest ideas was to insert genes into chloroplasts of the cell rather than the nucleus. Since most plant species lack chloroplasts in their pollen, this would eliminate the spread of transgenes through pollen dispersal. But not all pollen is chloroplast free and getting genes into each one of the hundreds of tiny chloroplasts is technically very difficult. What's more, chloroplast engineering does not eliminate the problem of pollen coming from outside to fertilse the plant. In the past few years, patents have been issued for techniques linking "suicide" genes to DNA "switches" that can be tripped inside pollen cells, wrecking their development, and also for techniques based on genes that kill off hybrid seeds as they attempt to germinate. Some companies even aim to turn crop fertility into a trait that can be switched on or off with sprays. "We are going to see a lot more of this technology," says Foudin. Not that environmental protection is the engine driving this push. Far from it. Companies see the control of plant fertility as the key to preventing growers from pirating their technology. As long as crops remain fertile, there is a risk that growers will secretly use transgenic seed which they have saved from the last season. If that happens, company profits will take a dive. At present, not all plants are amenable to sterility engineering. But there's no denying that if it is perfected, engineering sterility would be an effective way of laying to rest the spectre of the transgenic superweed. Then, what's good for the company coffers might also be good for the environment. This puts opponents of biotechnology in something of a bind. They see the new sterility technologies as an attack on the age-old rights of farmers to save and store harvested seed. Yet they also complain about the risks of "genetic pollution". The question nobody has begun to ask is what matters more-the pollution or the rights? Martin Brookes

Strange Fruit

CONSIDER the cautionary tale of the celery. In the mid-1980s, celery growers in the US introduced what they thought was a wonderful new strain. Highly resistant to insects, it promised to boost yields dramatically. There was just one small problem. People who handled the celery sticks began complaining of severe skin rashes. Dermatologists discovered that the celery was shedding psoralens, natural chemicals which become irritants and mutagens when exposed to sunlight. Or take the once notorious American Lenape-or rather, don't. All seemed well with this hardy new variety of potato launched in the US and Canada in the 1960s. Then came the bitter truth. Biochemists discovered the source of the tuber's unusual burning flavour: dangerous levels of toxins called glycoalkaloids.

"Many nightmares predicted for genetically engineered crops have already happened," reflects Tony Conner of the New Zealand Institute for Crop and Food Research near Christchurch. It's just that "not many people noticed or cared" because they were the fruits of conventional breeding, not genetic manipulation.

In fact, many biotech insiders and government food regulators, especially in the US, believe that the public has got it all wrong. By the time a "gene food" reaches people's plates it is not merely as safe as a conventional food-in some respects it is actually safer, because of the intensive testing that regulators demand for hightech food crops. By the end of this year millions of Americans will have eaten these foods, says Arnold Foudin of the US Department of Agriculture in Beltsville, Maryland. "And yet you won't be seeing anyone dying in the street."


However, you won't be seeing opponents of gene foods downing their placards, either. In Europe especially, campaigners have been working flat out in recent months to prevent genetically engineered crops being grown on the same scale as in North America. Their tactic has been to play the moral/emotional card for all it's worth and brand all genetically engineered crops "Frankenfoods" regardless of the specifics of each genetic modification. So far, it has worked amazingly well. All hell broke loose in Britain in August when a food scientist appeared on TV claiming-wrongly as it turned out-that a potato he'd engineered was toxic to rats. Inevitably, the questions that really matter have vanished amid the confusion and theatre. How do specific genes and the proteins they encode behave in the body? Do the types of genes and proteins being introduced into high-tech crops raise any new threats to food safety that could go undetected by researchers in companies or goverr-tment labs? The answers reveal that the biotech industry is on solid ground when it claims its products are no riskier than conventional foods. But it strays into some distinctly swampy territory when it claims, as it continues to with some force in the US, that genetically engineered foods need not be routinely labelled. First, there is the issue of food safety. In traditional breeding, scientists often introduce unknown genes into a plant species en masse by hybridising them with a related species with a desirable trait. Genetic engineering, by contrast, involves splicing no more than a few well characterised genes into a plant. That seems less drastic but can still produce unforeseen effects. In either case, the influx of new DNA might end up in critical parts of the genome, altering the behaviour of the plant's normal complement of genes, slashing the production of nutrients or pumping up the level of natural toxins. In many species, plant biochemistry is not just complex and sensitive, it's actually geared up for producing toxins to ward off predatorshence the bitter Lenape and toxic celery. "That's why it's standard to thoroughly analyse these new transgenic plants," says Roy Fuchs of Monsanto in St Louis, Missouri. "We need to see that they are substantially equivalent to commercial plants." To that end, Fuchs and his team run each promising transgenic crop through a battery of biochemical checks. They monitor levels of nutrients, proteins and potential poisons, and, in some cases, feed the crop to livestock to check that the animals gain weight at the normal rate and remain generally healthy. But what about more insidious effects? Some people worry that genetic engineering brings new DNA into the food supply, from microbes, for example. Couldn't this new DNA end up invading our genomes or the genomes of our gut bacteria? Few scientists take this threat seriously. Not even Walter Doerfler, a researcher at the University of Cologne in Germany, whose work has been seized on by opponents of gene foods. Last year, Doerfler's team found that when DNA from a bacterial virus was eaten by a mouse, some snippets of viral genes invaded the animal's bloodstream and cells-and, on rare occasions, even linked itself to mouse DNA (New Scientist, 4 January 1997, p 14; Proceedings of the National Academy of Sciences 94 p 961)). "This generated a lot of hysteria in the genetically engineered food arena," remembers Doerfler. But he believes that mammals have defences against this genetic onslaught. In his experiments, the vast majority of the viral chromosomes were shattered into pieces too small to contain intact genes, And despite scouring tissues throughout the mouse, Doerfler has never found any evidence of active ingested genes-even ones designed to work in human cells.

Shredded genes

Nor are microbes in the human gut likely to pick up genes from food. Most DNA from food will be destroyed well before it reaches the bacteria, with any surviving remnants being shredded again inside the bacteria by so-called restriction enzymes. Even if intact genes were to successfully invade a bacterium or human cell, they're unlikely to spring into action because their activity will be controlled by DNA switches designed to work only in plants. The one exception may tum out to be the antibiotic resistance genes that biotechnologists routinely use as "markers" for handling DNA in bacteria and identifying its presence in plant cells. Despite all the scare stories about these marker genes, those in crops now approved for commercial growth have been genetically scrambled, so there is little chance for their resurrection, or they are of no clinical importance. So it's unlikely that these particular genes could boost the spread of antibiotic resistance in human pathogens. Even so, critics worry that there is nothing to prevent scientists from using different markers in future, and while scientists agree that the chance of one of these genes jumping from food into a new cell is tiny, few will say it is impossible. Technology could soon make it impossible, however. Some years ago, David Ow and his colleagues at the Plant Gene Expression Center in Albany, Califomiaa lab belonging to the US Department of Agriculture-discovered a way of removing marker genes and other extraneous DNA from engineered plant cells. Their approach involved using a pair of molecular scissors called CRE, an enzyme from a bacterial virus, to snip out the antibiotic resistance DNA. Since then, Ow's group has shown the same editing trick also works in an important food crop, wheat. Until now, industry researchers have shown little interest in the work because they insisted that their genes posed no threat. But attitudes seem to be changing. "There is no clinical concern here whatsoever," says Jeff Stein of Novartis in Greensboro, North Carolina. "But we do worry about public perception." While not disclosing too many technical details, Stein says that all future Novartis crop products will be "100 per cent" free of antibiotic resistance genes. Other companies are also investigating ways of cutting out antibiotic resistance genes and surplus DNA. More recently, Ow's team showed that the editing process can run in reverse, enabling researchers to insert foreign genes into plant chromosomes at exact locations (Plant journal, vol 7, p 649)something that has so far been impossible. The method involves the insertion of f DNA "docking sites" into unimportant I areas of a chromosome. In future, researchers will be able to use such sites to slot new genes into plants without disturbing their normal complement of genes. Genetic engineering will finally become the precision tool that the biotech industry claims it to be. Not that this would deal with every worry. In some cases, the transgenic pro i tein encoded by this precision-engineered DNA might itself turn out to be toxic, although detecting this wouldn't be a problem. Unlike conventional breeders, biotechnologists can use the genes that interest them to produce transgenic proteins in bacteria to test on animals. A more subtle effect of proteins is harder to deal with. When molecular biologists shuttle new genes into plants, they might inadvertently introduce proteins capable of triggering respiratory or inflammatory problems in the one to two per cent of people who suffer from food allergies. Scientists at Iowa-based Pioneer Hi-bred, one of the world's largest seed companies, leamt this the hard way. In the early 1990s, its researchers engineered a more nutritious strain of soya bean by adding a gene taken from brazil nuts. The gene encoded a protein rich in methionine, a nutrient that is in short supply in ordinary soya beans. At the company's request, allergy specialist Steve Taylor of the University of Nebraska in Lincoln studied antibodies and immune responses from patients allergic to brazil nuts. Pioneer Hi-bred dropped the soya bean project when Taylor discovered that the hybrid was likely to trigger a major attack in people with brazil nut allergies. To some, it seemed like a narrowly averted disaster. After all, research based on animal experiments published only a few years earlier suggested that the same protein was not an allergen. "Allergy science is in its infancy," says Jane Rissler, a plant pathologist with the Union of Concerned Scientists in Washington DC. "That's a good reason to collect a lot more data before doing these widespread transgenic releases." Taylor himself extracts a different lesson. "It shows you can't be cavalier about allergies," he says. "But it also shows the system is working." The system he refers to is a series of tests that scientists now use to flush out allergens before they are put into crops. If the transgenic protein comes from a known allergenic food, it is subjected to immunological tests. If the protein comes from other sources, researchers study its molecular structure (amino acid sequence), looking for similarities with allergy-triggering proteins in the databases. The protein's chemical hardiness is also scrutinised. In test-tube simulations of the heat, acid and enzymes found in the stomach, most proteins are torn to shreds in seconds. Allergens tend to survive several minutes before they, too, are destroyed.

Mystery ingredients

Even if true allergens do escape detection and make it into transgenic crops, immunologist Yueh-hsui Chien of Stanford University questions whether this represents a new risk to the consumer. "If you regularly eat tomatoes, and then you eat a transgenic one, you know you are eating a few new proteins," she says. "The first time you eat a lobster, you eat several thousand new proteins." But that's a false comparison, argues Rebecca Goldburg, senior scientist at the Environmental Defense Fund, an advocacy group in New York. She points out that someone knows they are eating a lobster. But the new ingredient in the tomato is invisible because transgenic crops are, for the most part, unlabelled and mixed in with the rest of the harvest. "The industry is depriving us of one of our most important natural defence mechanisms," she says. "Reading ingredients." In the US, companies argue that the chance of allergic responses to the current generation of modified crops is too remote to warrant segregation and labelling. And so far, the US Food and Drug Administration has supported this view by introducing rules that require farmers and manufacturers to segregate and label transgenic foods only if there is good reason to suspect they n-dght behave differently in the body than more conventional foods. Officials in Europe made a similar ruling in September, but in Britain and many other countries in the European Union, some manufacturers and retailers have decided to label products voluntarily. Full disclosure may soon be a major fashion. In the industry, the most excited talk is about using molecular biology to lower undesirable chemicals or boost nutrients in food. At Nagoya University in Japan, for example, researchers have managed to slash levels of the major allergenic protein in rice by 70 to 80 per cent by inserting a so-called antisense gene to block the protein's production in the plant. If biotechnology dramatically increases the quality or safety of food, companies on both sides of the Atlantic may soon be falling over each other to market new and improved gene crops-and to provide the public with more information about what they are eating. Then we can decide for ourselves which of the risks-low tech or high tech -we are willing to take when we eat our next meal. Phil Cohen

Live and Let Live

BIG SCIENCE plus agriculture spells bad news for wildlife. This has been the mantra of environmentalists ever since Rachel Carson documented the impact of chemical pesticides on American wildlife in her 1960s classic, Silent Spring. And that is the mantra that Dan Verakis, spokesman for biotech giant Monsanto and defender-in-chief of genetically engineered crops, is keen to turn on its head. "People are saying this technology is the last nail in the coffin of our wildlife," he says. "But we believe it could help wildlife recover." Genetic engineering could slash the volume of herbicides and pesticides that farmers need to dump on the land each year; it could "reverse the Silent Spring scenario". Advocates of the new high-tech crops clearly see this as their strong card, the one that can simultaneously wrong-foot anti-biotech greens and all those who are sceptical simply because they can see no benefits in the technology for anyone other than farmers and shareholders. There's just one small problem. Biotech's biggest money-spinners to date are crops that seem designed to keep farmers hooked on chemicals. Of the 27.8 million hectares of land planted with genetically engineered crops this year in the US, 71 per cent is covered with plants resistant to herbicides, largely soya bean and maize engineered to carry an enzyme that neutralises glyphosate, Monsanto's famous Roundup herbicide. Thanks to this enzyme, the biotech industry gets to sell transgenic seed and a chemical, and farmers can wage war on weeds like never before. In theory, they can spray as much as they want whenever they want and not endanger their crop. Of course, the industry claims these crops are greener than they seem. Farmers planting soya beans resistant to glyphosate have only to apply the herbicide once or twice instead of six or seven times, says Val Giddings of the Biotechnology Industry Organisafion. "The notion held in Europe that farmers want to spray herbicides willy-nilly is nonsense because they want to cut costs." And so they do, but the total volume of chemicals alone doesn't tell the whole story, especially not with herbicide-resistant crops. Whether these are a good thing for wildlife depends not just on how much herbicide farmers spray, but when and where they spray it, and, above all, what their attitude is to weeds. The fact that many take an aesthetic pride in running a farm with "clean fields" means that reversing the Silent Spring scenario might be easier said than done. But first the good news. Even staunch critics of the speed with which biotechnology is revolutionising the world's agriculture acknowledge that transgenic crops engineered to be resistant to specific fungi and insects could make many old-style chemical sprays redundant. In Arkansas alone last year, farmers planted 1 million acres with engineered cotton resistant to the cotton bollworm, and hardly any of I the usual organophosphate pesticides were applied. "It amounts to a saving in pesticide application of 1 litre per acre, enough to fill about 14 large railroad tank cars," says Giddings. "People in the US are more pleased than Europeans when you tell them you don't need chemicals to control insects and nematodes," says Roger Beachy, a plant biologist at the Scripps Institute in La jofla. "The public in Europe don't seem to be aware of the benefits to the environment, and therefore to them personally," he says. How long those benefits will last is not so clear. Introduce a new method of pest or weed control and "you always get a temporary dip in the cost to the farmer", says Margaret Mellon, director of agriculture and biotechnology for the Union of Concerned Scientists, a pressure group in Washington DC. "In the past, those dips have never persisted because resistance has emerged." Sure enough, biotech companies already have to take seriously the problem of pests acquiring resistance to Bt toxin (see "Resistance is useless", p 48). Others, meanwhile, fear that inserting certain pesticide genes into plants could jeopardise the balance between pests and beneficial insects in agricultural ecosystems.

Friendly fire

The aim of such manipulations is to provide the crops with a toxin that can ward off or kill specific pests such as bollworms and aphids, while sparing friendlier insects such as ladybirds. But this year a British team found that a potato engineered to resist an aphid pest also harmed ladybirds, at least in lab conditions. Nick Birch of the Scottish Crop Research Institute in Dundee and his colleagues inserted a gene from the snowdrop into potatoes, enabling the tubers to make a lectin capable of stopping aphids from snacking on them quite so much. When aphids reared on these potatoes were fed to ladybirds, however, the females lived only half as long and produced more than twice the number of unhatched eggs compared with ladybirds fed normal aphids. Still, most ecologists see this as an argument for testing each pesticide gene fully before using it, not for banning them en masse. There are some pesticide genes that shouldn't be inserted into crops until their wider biological effects have been determined, says Brian Johnson, an ecologist with the conservation group English Nature who advises the British government about the release of transgenic organisms. But overall, the environmental benefits of pest-resistant crops could be immense, he says. Herbicide-resistant crops, however, are another matter. English Nature is urging the British government to impose a moratorium on their use until their effects on biodiversity have been explored. Johnson accepts that the crops are not all bad. In the US, for instance, they have led many farmers to switch to Roundup from atrazine, a herbicide chemical that is far more likely to leach into water systems and harm animals and humans. One of Roundup's virtues is that it binds to soil as it breaks down. Another is that it can reduce the need for ploughing, which may contribute to erosion. It is also a very effective herbicide. Perhaps too effective. English Nature fears that armed with Roundup and transgenic crops, farmers will end up wiping out more wild plants than ever before. They may spray less, Johnson says, but because chen-dcals like Roundup kill a broader spectrum of weeds than many other herbicides, their impact would be greater. The result could be weed wipeout In the US, that might not matter so much because less of the nation's biodiversity is locked up in farmland. But, as Johnson says, "Europe doesn't have the luxury of vast wilderness areas like the US". Much more of its wildlife depends on agricultural practices for its survival. In Britain, says Johnson, farmers often permit a modest growth of weeds in and around their fields. These weeds support a variety of insects which provide a lifeline for threatened birds such as linnets and skylarks. In fact, research is just beginning to reveal in detail which plants do and do not harm the yields of different crops, promising a new era of laissez-faire crop management. What will happen to all this if British farmers switch to engineered crops and broad-spectrum herbicides? Not a great deal, according to Alan Dewar at the Institute for Arable Crops Research at Broom's Barn in Suffolk. Dewar says that with conventional herbicides most farmers cleanse their land of weeds anyway. In fact, to avoid damaging their non-resistant crops, most farmers spray their land before the growing season even begins. At least with the transgenic crops, farmers have the option of letting the weeds grow with the crop for part of the season. in experiments carried out this surmner for Monsanto, Dewar and his colleague Mike May found that yields of transgenic sugar beet treated with Roundup remained higher than with conventional herbicides even when Roundup was applied late, letting the weeds survive for much longer than usual in the growing season. "It was obvious to see that the weedy plots were heaving with life," says Dewar. Aphids that normally attacked the beet, for example, colonised the weeds instead. Conservation-minded farmers could use Roundup to allow weeds to grow a bit longer, says Dewar. "But it requires a bit of nerve because it looks a bit of a mess." The major disincentive for letting the weeds grow, however, is that the best yields were in plots sprayed early enough for Roundup to destroy everything. In other words, herbicide-resistant crops could be used in an environmentally friendly way-but only by farmers who are not driven by a "clean field" mentality and who don't mind dropping their yields below the maximurn achievable. These are not the farmers Monsanto seems to have in mind in its brochure for Roundup-ready oilseed rape. One Manitoba farmer endorses the product thus: "The weed control has been exceptional. It [Roundup] annihilated everything." Whether the same would happen in Britain and other countries is hard to say, but one thing at least is clear. The biotech industry is developing two very different sales pitches for its products-one for farmers and one for the rest of us. Martin Brookes and Andy Coghlan

Resistance is Useless

IT SOUNDS like a farmer's dream: crops that produce a steady supply of their own insecticide so there's no need to drench fields periodically with toxic chemicals to kill off the pests. in the US the dream has been a reality for three years now, as farmers enthusiastically plant crops genetically engineered to produce Bt, an insecticidal protein originally found in a soil bacterium, Bacillus thuringiensis. This year Bt cotton engineered by Monsanto accounted for over 20 per cent of all the cotton planted in the US, while Bt maize developed by several companies grew in over 10 per cent of America's cornfields. A third crop, Monsanto's Bt potato, has also begun to hit fields. In total, the USs Bt crops cover an area almost as big as Scotland. Critics fear that in their rush to take advantage of the new technology, however, farmers and the companies that supply them may be killing the goose that lays the golden eggs. Insects have a long history of evolving resistance to chemical insecticides. But scientists worry that Bt crops may provoke resistance even more quickly.

Sprayed insecticides usually degrade rapidly, and if relatively few pests show up in a given year. money-conscious farmers may not spray at all. Bt plants, on the other hand, are equivalent to a continuous deluge of insecticide that keeps the pests under constant pressure to evolve resistance. As Bt crops move into Europe and huge new markets such as China and India in the next few years, the threat of Bt-resistant superbugs seems likely to grow. And that matters because many see Bt as the last line of defence against insect pests. The best response to such a threat, scientists agree, is to stack the cards against resistance by making sure that enough insects survive that are still susceptible to the toxin. Farmers can do that by planting some of each crop without the Bt gene, so that insects can mature without any evolutionary pressur6 to become resistant. If there are enough insects in this refuge, any resistant ones that evolve in the Bt field will mate with susceptible pests. And the Bt-containing plants deliver a toxic dose hefty enough to kill insects with just one resistant parent, this so-called -high doselrefuge' strategy, should keep resistance at bay for many years. But that -ifmay be a big one. So far none of the three commercial St crops has been grown long enough for resistant insects to show up in farmers' fields in the US. Researchem are already seeing ominous hints, however, that for a few pests, St plants aren't delivering the knockout punch. In some Bt corn varieties, for example, toxin levels drop towards the end of the growing season, which allows some la"ae of the Europ4an corn borer to survive even without resistance genes. Even when the toxic dose is high enough, refuges work only if susceptible insects survive on them, and farmers aren't used to encouraging pests. As a result, agricultural companiesand government regulators-tend to recommend refuges that are barely big enough to do the job. If enough susceptible insects manage to survive in the refuge, they will do littie good unless they mate with resistant survivors from the Bt fields. And here, too, some troubling signs are beginning to emerge. At the University of Minnesota, Donald a s colleague David Andow have found that European corn borer adults rarely fly more than 75 metres during their lifetime. That means the refuges must be tucked in very close to the Bt corn, which they seldom are. Entomologists know less about the flight ranges of other pests, but it's clear that some refuge placements are nearly useless. The US Environmental Protection Agency may correct some of these problems when it reconsiders its regulations for Bt plants over the next three years, says Janet Andersen, who directs the agency's biopesticides division. But even if the US plugs the largest loopholes, there's no guarantee that other countries will do the same. In Europe, where Bt crops are starti approval I for commercial planting, regulators are still wrestling with the question of how big the refuges would need to be for European pests and cropping patterns. in poorer countries, where farms are small and farmers uneducated, any regulation may be hard to achieve. Bob Holmes

Food for All

FIVE MILLION Brazilians faced starvation this year. This time it was a drought related to El Nifio that halved grain crops in the northeast of the country, but next year it will be something else. Famine is perennial in Brazil. In September Monsanto, the world's largest supplier of genetically modified seeds, announced it would invest $550 million in Brazil to build a factory producing its herbicide Roundup. Shortly afterwards the Brazilian government made Monsanto's Roundup-resistant soya beans the country's first legally approved, genetically engineered crop. The soya beans will boost profits for the big landowners who grow them to feed beef cattle for export. But most rural Brazilians are subsistence farmers who do not grow soya. No help will trickle down from Monsanto's beans to the starving millions. The story exemplifies the limited contribution genetically modified crops have made so far to eradicating world hunger. It is not that biotech companies are uninterested in the developing world. Far from it: Brazil and other newly industriarising countries are in fact prime targets, with their growing demand for agricultural products, little opposition to biotechnology, and farmers who have risen above hard graft subsistence, but have not yet become customers of the world's seed and agrochemicals conglomerates.

But who will benefit from genetically modified crops in these countries? The companies speak of feeding the starving millions, while conserving the environment. They say that the new technology will have greater benefits in the Third World than anywhere else. "Biotechnology is a key factor in the fight against famine," claims the literature from EuropaBio, the association of European biotechnology companies. "Biotechnology will help increase the yield on limited land." Critics maintain that there is little evidence of this. Instead, they say most of the engineered crops developed or in the pipeline will benefit rich farmers, not the needy. Worse still, they fear the biotech industry's increasing domination of crop research will hurt, not help, the poor. Agriculture does need a new technological saviour. Most of the world's food calories come from grain. A simple redistribution of what we grow now, even if it were possible, will not feed the 10 billion humans expected by 2030. Traditional methods of improving crops seem to have gone about as far as they can. "The fact that we start from the results of more than 5000 years of selective breeding makes further staggering yield increases unlikely," says Lloyd Evans of the CSIRO Division of Plant Industry in Canberra, Australia. "The biggest opportunity for increasing grain yields is to produce varieties more precisely adapted to local conditions." Yet few of these crops have emerged so far. Those that are on or near the market aim to increase farmers' profits by cutting expensive "inputs", such as pesticides. This is little help to farmers who can afford no inputs to begin with, not even the reduced levels needed for these crops, and no help if they cannot aff ord the patented seed. Steven Briggs, head of the Novartis Agricultural Discovery Institute in San Diego, which sequences plant genomes, points to several innovations in the pipeline which might help: fodder crops that contain more calories, so more meat can be produced per hectare of corn or soya; crops that destroy toxins produced by moulds, such as fumonisin, which cause massive crop losses after harvest; and disease-resistant crops, such as sweet potatoes and cassava, staples of the poor, which fend off viruses. Crops that thrive despite drought and salty soils could also let farmers expand production into marginal lands. And the nutritional content of staples could be improved. If maize, for example, can be made to produce more of the amino acids it naturally lacks, the 80 million people who live almost exclusively on maize would get more protein. Ganesh Kishore, head of nutrition at Monsanto, says: "We can make it into a complete balanced meal." Briggs agrees that there are contradictions inherent in bringing high-tech remedies to low-tech farmers. Breeding crops for subsistence, he says, is "emergency aid, not a path to economic growth". Pol Bamelis, from the German giant Bayer and chair of the German and European biotechnology associations, says that the industry "cannot help the fact that there are rich and poor in the world".

Buy out

Biotech companies think genetic engineering will be in the best position to help once farmers everywhere switch from small-scale subsistence to largescale mechanisation. But many activists fear just that process. The high price of the technology could allow the few farmers who can afford it to out-compete their poorer neighbours and eventually buy them out, driving people from the land, says Hope Shand, of the Rural Advancement Fund Intemational in Canada. Monsanto also argues that helping poor farmers would reap another kind of benefit: richer peasants who no longer need to destroy forests to get more land. But this could be simplistic. Steve Vosti, of the International Food Policy Research Institute in Washington DC, has studied poor farmers and deforestation in Amazonia. He says any technology that increases a farmer's profits, or reduces the labour needed per hectare, will cause the farmer to cut down trees to get more land. It is not clear whether the kind of farmer who needs to fell forests to get land, or who eats little but maize meal, will ever be able to afford genetically modified crops. But even if only rich farmers benefit, says Vosti, their expansion would tend to push poorer farmers into forest margins. And there are other disadvantages for the poorest farmers. "New biotechnologies threaten to aggravate problems of genetic uniformity, and increase the dependence of farmers on transnational corporations," says Shand. Even in the industrialised world, people are worried about genetic uniformity arising from the widespread introduction of genetically modified crops. In Missouri this summer, half the soya plants on some farms died of Fusarium mould, after three-quarters of the land was planted with Roundup resistant varieties which turned out not to resist mould. The handful of modified varieties offered by biotech companies will inevi tably be more genetically uniform, hence more susceptible to unfo@ s@, than the plethora of classically bred varieties g;rovm now That problem could be worse in the tropics, where there is more existing crop diversity together with s that seed breeders based in the North may not have anticipated. Tropical countries will also have less money to pay mult'nat'On als for the rights to incorporate proprietary genes into several local varieties.

The last problem stems from the big companies' growing control of both markets and plant genes. Crop scientists must continually breed new crop varieties to meet the ever-evolving threats of pests and disease. In the Third World, this is mainly done by government-funded institutions, and the Consultative Group for International Agricultural Research. But public [email protected] breeders are losing funding, while companies such as Monsanto are rapidly becoming the only source for improved varieties. It already, for example, sells half the maize seed in Argentina.

Losing access

The public breeders are also losing access to plant genes. Last May the CGIAR completed a detailed study of the problems posed by the fact that the genes it needs to do its work are increasingly available only at a price, because companies hold the patents. India recently declined to pay Monsanto $8 million for the use, by its state-owned crop laboratories, of Monsanto's Bt insecticide gene. Those labs will not be able to provide Indian farmers with cheap, locally bred insect-resistant crops. Farmers who can afford to will have to buy whatever Monsanto has to offer. Even if Third World breeders get access to patented genes, they may be forced to protect them in ways that put them out of reach of the poor. Terminator, a gene owned by Monsanto, keeps a plant from producing viable seed. So farmers cannot save seed from patented, genetically mod ified varieties for the next harvest. It also keeps farmers from crossing patented strains with other crops to create new varieties. "Public sector breeders could be under great pressure to use Terminator to protect patented genes in the breeds they produce, in exchange for access to those genes," says Shand The overall effect could be that breeders will not be able to create new varieties to meet evolving threats unless they pay for the genes, and couple them with technologies to prevent the saving of seed. That means fewer, more expensive varieties, plus increased costs for the 1.4 billion poorest farmers who grow 80 per cent of subsistence crops from saved seed. As big northem companies expand their control of crop genes, their choice may be . . Debbie Mack