Sowing Technology


Do We Really Want To Pit Agriculture Against Nature?


Craig Holdrege and Steve Talbott



Drive the Nebraskan backroads in July and you will encounter one of the great technological wonders of the modern world:  thousands of acres of corn extending to the vanishing point in all directions across the table-flat landscape.  It appears as lush and perfect a stand of vegetation as you will find anywhere on earth -- almost every plant, millions of them, the same, uniform height, the same deep shade of green, free of blemish, emerging straight and strong from clean, weed-free soil, with every cell of every plant bearing genetically engineered doom for the over-adventurous worm.


If you reflect on the sophisticated tools and techniques lying behind this achievement, you will likely feel some of the same awe that seizes so many people when they see a jet airliner taking off.  There can be no doubt about the magnitude of the technical accomplishment on those prairie expanses.  And yet, the question we face with increasing urgency today is whether this remarkable cornucopia presents a picture of health and lawful bounty, or instead the hellish image of nature betrayed.


Actually, it is difficult to find much of nature in those corn fields. While nature manifests itself ecologically, contextually, today's advanced crop production uproots the plant from anything like a natural, ecological setting.  This, in fact, is the whole intention.  Agricultural technology delivers, along with the seed, an entire artificial production environment designed to render the crop independent of local conditions. Commercial fertilizer substitutes for the natural fertility of the soil. Irrigation makes the plants relatively independent of the local climate. Insecticides prevent undesirable contact with local insects. Herbicides discourage social mixing with unsavory elements in the local plant population. And the crop itself is bred to be less sensitive to the local light rhythm.  Where, on the farm shaped by such technologies, do we find any recognition of the fundamental principle of ecology -- namely, that every habitat is an intricately woven whole resisting overly ambitious efforts to carve it into separately disposable pieces?


But all this represents only one aspect of agriculture's abandonment of supporting environments.  The modern, agribusiness operation in its entirety has wrenched itself free from the rural economic and social milieu that once sustained it.  The farm itself is run more and more like a self-contained factory operation.  And the trend toward vast monocultures -- where entire ecologies of interrelated organisms are stripped down to a few, discrete elements -- has become more radical step by step:  first a single crop replacing a diversity of crops; then a single variety replacing a diversity of varieties; and now, monocultures erected upon single, genetically engineered traits.


As the whole process drives relentlessly forward, the organism itself becomes the denatured field in which genes are moved to and fro without regard to their jarring effect upon the living things that must endure them.  Want to make a tobacco plant glow in the dark?  Easy -- inject a firefly gene!  Want a frost-resistant strawberry?  Try a gene or two from a cold-water flounder.


Yet, despite such freakish prodigies, the overriding question about biotechnology is not whether we are for or against this or that technical achievement, but whether the debate will be carried out in just such fragmented terms.  In focusing on technological wonders to improve agriculture, are we losing sight of the things that matter most -- the diverse, healthy, and complex communities and habitats we would like to live in?  The question to ask of every technology is how it serves, or disrupts, the environment into which we import it.


Is Genetic Engineering New?


The natural setting whose integrity we need to consider first of all is that of the individual organism.  The challenge we're up against here emerges in the frequently heard argument that genetic engineers are only doing what we've always done, but more efficiently.  Writing in the New York Times, Carl B. Feldbaum, president of the Biotechnology Industry Organization, objected to the claim by critics that "what [traditional breeders] do is `natural' while modern biology is not":


Archaeologists have documented twelve thousand years of agriculture throughout which farmers have genetically altered crops by selecting certain seeds from one harvest and using them to plant the next, a process that has led to enormous changes in the crops we grow and the food we eat.  It is only in the past thirty years that we have become able to do it through biotechnology at high levels of predictability, precision and safety.


But the concern about genetic engineering today isn't that it enables us to commit altogether new mistakes.  Rather, it's that it perfects our ability to commit old ones.  No one is suggesting that the abuse of our technical powers began with the discovery of the double helix.  Using conventional techniques, breeders have, for example, produced Belgian cattle with such overgrown muscles that they cannot be delivered naturally; birth requires Caesarian section.  Likewise, there are hobbyist chicken breeders who -- to judge from the pictures in their magazines -- are more interested in bizarre effects that tickle human fancies than in the welfare of the chickens themselves.


The difference is that with genetic engineering we can now manipulate living organisms much more efficiently and more casually than ever before. The technician need scarcely be distracted by the animal itself. There's none of the Frankenstein drama and messiness. We can construct our monsters in a clean and well-lit place.


Moreover, Feldbaum's claim completely glosses over what is unprecedented about genetic engineering:  that it selects isolated genes, not entire healthy organisms.  Writing in Science (March 26, 1999), geneticist Jon W. Gordon assesses the failed attempts to create heavier farm animals by inserting appropriate genes.  In pigs, the addition of growth hormone- producing genes did not result in greater growth, but unexpectedly lowered body-fat levels.  In cattle, a gene introduced to increase muscle mass "succeeded," but the growth was quickly followed by muscle degeneration and wasting.  Unable to stand up, the experimental animal had to be killed.


Such results are hardly surprising when you consider the isolated and arbitrary intrusion represented by single-gene changes. By contrast -- and this is what Feldbaum ignores -- traditional breeding allows everything within the organism to change together in a coordinated way.

As Gordon writes,


Swine selected [by traditional methods] for rapid growth may consume more food, produce more growth hormone, respond more briskly to endogenous growth hormone, divert proteins toward somatic growth, and possess skeletal anatomy that allows the animal to tolerate increased weight.  Dozens or perhaps hundreds of genes may influence these traits. If there's a logic to ecological relationships that says, "Change one thing and you change everything," the same applies to the interior ecology of the organism.  Responsible traditional breeding is a way of letting everything change without violating the whole -- because it is the organism as a coherent and healthy whole that manages the change.



Do Organisms Need Preserving?


This points to another consideration as well.  In traditional breeding the integrity of the organisms themselves places limits upon what can be done -- limits you could reasonably call "natural."  For example, you could not cross a strawberry with a cold-water fish in order to obtain strawberries with "anti-freeze" genes.


The problem now is that we can break through these limits, but we have not replaced the safeguard they represented.  Today, such a safeguard can come only from our own, intimate, respectful understanding of the organism as a whole and of the ecological setting in which it exists.


This is the decisive question:  does the organism possess a wholeness, an integrity, that demands our respect?  And can we gain a deep enough understanding of it to say, "This change is a further expression of the organism's governing unity, and that change is a violation of it"?  A difficult challenge, and not one we have trained ourselves to meet.  You have to see a plant or animal in its own right and in its natural environment in order to begin grasping who or what it is.  But given what ecologists David S. Wilcove at Environmental Defense and Thomas Eisner at Cornell University have called the "demise of natural history" in our time, there is not much hope of greater familiarity with the organisms whose natures we manipulate -- certainly not by those laboratory- and test tube-bound researchers who are doing the manipulating.


Nevertheless, some things are fairly obvious.  It's hard to understand how the Mad Cow debacle could have occurred if anyone had bothered to notice the cow.  How could we possibly have fed animal parts to ruminants? Everything about the cow, from its teeth to its ruminating habits to its four-chambered stomach, fairly shouts at us, herbivore!  Can we violate an organism's integrity in such a wholesale manner without producing disasters -- for the organism, if not also for ourselves?


What the Mad Cow episode illustrates is that our notions of safety are relative to our understanding of the organism.  And nothing has tended to fragment our view of the organism as powerfully as genetic engineering. Instead of a coherent whole expressing an organic unity through every aspect of its being, the engineers hand us a bag of separate traits and molecular instrumentation.


Are Bioengineered Products Adequately Tested?


Only such a fragmenting mentality could suggest (in the words of former U.S. Secretary of Agriculture, Dan Glickman) that "test after rigorous scientific test has proven these [genetically engineered] products to be safe."  This suggestion is simply false on its face.  The application to cows of bovine growth hormone (rBGH) produced by genetically engineered bacteria was approved primarily on the basis of tests with rats -- not cows, and not people who consume cow products.  Genetically altered Bt corn was approved without being tested for its effects on beneficial species such as green lacewings or on "incidental" species such as the Monarch butterfly.  (Subsequent research has suggested the possibility of harm to both Monarchs and lacewings.)


But the more fundamental problem is that, because the organism is an organic unity, its assimilation of foreign DNA potentially changes everything.  Gene expression and protein levels are altered in ways that have proven consistently unpredictable.  About one percent of genetic transfers yield the looked-for result; the other ninety-nine percent are all over the map.  For example, when scientists engineered tomatoes for increased carotene production, they indeed got some plants with more carotene -- but those plants were unexpectedly dwarfed.  No one expected this experiment to yield dwarfed plants.


So even the one percent statistic paints too optimistic a picture.  This "success" rate reflects a focus on the particular trait that was looked for; but even when this trait is obtained and the resulting organism is used as the founding ancestor of a new, genetically altered line, it remains to ask:  what about the subtle changes throughout the rest of the organism -- changes not directly related to the researcher's intent?  If there can be immediately obvious changes such as dwarfing, there can be many more unobvious ones.  It's hard to test for changes when anything can happen and you don't know what you're looking for.  In actual practice, almost no such testing is done.


Is Biotechnology Good for the Environment?


Against this backdrop, the biotech companies' promotion of genetically altered crops as the Great Green Hope of the environment due to the promise of reduced pesticide applications is puzzling at best.  After all, the entire thrust of the factory-farmed monocultures encouraged by these companies is to eliminate across huge acreages all traces of any environmental richness that might have been worth preserving in the first place.  And now the corporate research laboratories are poised to release into this devastated landscape a continuing stream of alien genes that, in their own right, promise to become the ultimate, uncontrollable pollutants.  Chemical spills can eventually be cleaned up, but there is no recalling the replicating genes we have loosed upon the natural world.


If there's any claim that must be evaluated ecologically, it's the claim of environmental benefit.  Yet, as Michael Pollan remarks in a New York Times Magazine piece on genetically engineered potatoes:  those who simply take vast monocultures for granted will always think they have, say, a Colorado potato beetle problem -- rather than the total environmental problem of potato monoculture.


Certainly there are silver bullets to be had, even if their unfortunate tendency is to rip crudely through the delicate, ecological fabric they are aimed at.  Perhaps the most obvious silver bullet is Bt cotton.  The relatively mild Bt toxin engineered into the crop is highly effective against the bollworm and substitutes for an extraordinarily nasty series of sprayings in conventional cotton fields.  Yet, to leave the matter there is to accept the conventional approach as the only alternative.  And it is also, as Charles Benbrook points out, extremely irresponsible.  Benbrook is former executive director of the National Academy of Sciences Board on Agriculture and now an agricultural consultant in Sandpoint, Idaho.  He sees Bt, in its normal, externally applied form, as perhaps the most valuable pesticide ever developed.  It is approved for organic as well as conventional use, and controls many serious pests not otherwise easily controlled.  He calls it a "public good," and suggests that engineering it into crops on a massive scale is the moral equivalent of loading everyone's toothpaste with antibiotics.  Yes, the antibiotics would yield an immediate "benefit" in terms of reduced incidence of certain diseases.  But the consequences for both immediate and long-term health would be ugly indeed, since disease microbes would develop resistance much more rapidly than otherwise.  In the case of Bt, the inevitable development of resistance by pests will reduce the useful lifetime of this invaluable pesticide to a small fraction of what it would otherwise be.  Then we'll be off to search for the next silver bullet.


It's a measure of the narrow vision of the biotech industry's environmental assessment that the Bt toxin in the crop itself is never added into the calculations of pesticide use.  Yet, speaking of corn, Benbrook estimates that (depending on how you frame the question) there is 10 to 10,000 times as much Bt toxin produced in the crop as would have been applied in the usual external applications -- and that's assuming a year in which the corn borer needed to be controlled at all.  It can hardly be doubted that the amount of Bt toxin in Bt corn intended for human consumption exceeds any residue on conventional, Bt-sprayed corn.


Moreover, researchers have recently discovered that the Bt toxin released by the crop into the soil binds to soil particles and is then highly resistant to biodegradation.  The implications for beneficial soil organisms are almost completely unknown -- although the researchers found that a high percentage (90 - 95%) of insect larvae exposed to the toxin died.


Crops genetically modified for resistance to herbicides pose similar problems.  Knowing that their crops will more or less tolerate an herbicide, farmers are not likely to reduce their applications. Monsanto has requested and received from the Environmental Protection Agency a threefold increase in allowance for glyphosate residue on Roundup Ready soybeans.  (Glyphosate is the active ingredient in the company's Roundup herbicide.)  The increased residues are hardly an environmental improvement, especially in light of the fact that glyphosate has been linked to non-Hodgkin's lymphoma (a cancer of white blood cells) in a study reported in the journal, Cancer (March 15, 1999).


The vast expansion of acreage in herbicide-resistant crops has led to huge increases in the use of glyphosate -- a 72% increase in 1997 alone, according to the U.S. Department of Agriculture.  This large-scale adoption of single-pronged weed-control strategies is deeply troubling because it encourages herbicide-resistance in weeds (already observed with glyphosate) and wholesale shifts in weed populations.  These shifts require additional herbicides, and the resulting treadmill, as Benbrook puts it, "is on hyperdrive today.  We'll burn up the current generation of herbicides in five, ten, or fifteen years instead of three to five decades."


The alternative to the treadmill is to turn our attention away from silver bullets and look at ecological integrity.  Mary-Howell Martens, who was formerly a genetic engineer and conventional farmer, now farms 1100 acres organically in New York state.  Like many other organic growers, she and her husband, Klaas, grow soybeans without using any herbicides.  They work instead with nature, relying on soil fertility (the calcium-magnesium ratio in particular affects weed vigor); long, diverse rotations, including corn, soybeans, clover, and grains, to disrupt weed cycles; clean seeds; well-timed tillage early on, so that the crop gets ahead of the weeds and tends to smother them; and avoidance of high-salt fertilizers, since salt compounds stimulate weed growth.  Later weed control can be done mechanically, on a spot basis, as needed.


Orchestrating Nature's Complexity


Most people regard genetic engineering as the future of agriculture, if only because it is sophisticated, cutting-edge science.  But impressive procedures in the laboratory do not automatically equate to precise effects upon nature.  Even if it were true that DNA presents us with a kind of master computer program controlling the living organism, every software engineer knows about the unpredictable and sometimes disastrous consequences for massively intricate programs when someone goes in and "twiddles the bits."  Already in 1976, when computer programs were vastly simpler than today, MIT computer scientist Joseph Weizenbaum could write a now-classic chapter entitled "Incomprehensible Programs" where he pointed out that any substantial modification of a large, complex program "is very likely to render the whole system inoperative."


In its application to agriculture, genetic engineering is crude, blindfolded, trial-and-error science -- and not only because the consequences of particular genetic alterations are largely unknown.  The farmer is often prevented from exercising skilled judgment based on the ecological realities of the local environment.


Take, for example, the farmer who plants Bt corn as protection against the European corn borer.  (Bt corn has been engineered so that the Bt toxin -- a pesticide naturally produced by the bacterium, Bacillus thuringiensis -- is manufactured in each cell of the plant.)  Such a farmer commits to round-the-clock, season-long application of a pesticide in his fields before he knows whether the corn borer will even be a problem.  In major parts of the corn belt, the answer is that, during most seasons, it will not.


If you really want technical sophistication, don't look at the latest biotech application, but at the many successes of Integrated Pest Management.  IPM is founded on decades of painstaking investigation into the incredibly complex and subtle weave of natural ecologies.  Where the main trend of today's biotech agriculture is to isolate the farm from its environment, reducing the operation to the simplistic terms of a few manageable variables, IPM at its best tries to work with the environment, penetrating the boundless complexity with an understanding that can turn intricate equilibria to good use.


It's one thing to take the heavy-handed biotech approach and engineer a pesticide into every cell of a crop; it's quite another to manage the ecological interrelationships of the farm so that the offending insect is controlled by the natural balances of the larger context.  Tragically, the more simple-minded, heavy-fisted approach tends to destroy the possibilities inherent in the more subtle practice.  Among other problems, converting an entire crop into a pesticide virtually guarantees the rapid emergence of pest resistance, which IPM has taken such pains to avoid.


Working with natural complexity rather than against it is the aim of a remarkable research organization in Kenya, the International Centre of Insect Physiology and Ecology (ICIPE).  The Centre brings together molecular biologists, entomologists, behavioral scientists, and farmers in an interdisciplinary effort to control the various threats to African crops.


The most important pests of corn and sorghum on that continent are the stemborer and striga (witchweed), which, together, can easily destroy an entire crop.  ICIPE researchers developed a "push-pull" system:  a grass planted outside the cornfield attracts the stemborer; a legume planted within the cornfield repels the insect and also suppresses witchweed by a factor of forty compared to a corn monocrop -- all while adding nitrogen to the soil and preventing erosion; and, finally, an introduced parasite radically reduces the stemborer population.


ICIPE director Hans Herren won the World Food Prize in 1995 after the Centre gained control over the mealy bug that threatened the cassava crop, a staple for 300 million people.  (A small, parasitic wasp was instrumental in the success.)  No chemical applications and no costs to the farmers were involved.  Yet Herren doubts he could obtain funding for such a project now.  "Today," he says, "all funds go into biotechnology and genetic engineering."  Biological pest control "is not as spectacular, not as sexy."


The Real Future of Agriculture


Fortunately, some work on Integrated Pest Management continues, and the results are often so dramatic that one wonders why the genetic engineering labs have secured all the glamour for themselves.  Even the simplest step toward balance sometimes yields striking results.  In what the New York Times called "a stunning new result" from a vast Chinese agricultural experiment, tens of thousands of rice farmers in Yunnan province "have doubled the yields of their most valuable crop and nearly eliminated its most devastating disease -- without using chemical treatments or spending a single extra penny."


The farmers, guided by an international team of scientists, merely interplanted two varieties of rice in their paddies, instead of relying on a single variety.  This minimal step toward biodiversity led to a drastic reduction of rice blast, considered the most important disease of the world's most important staple.  The fungicides previously used to fight rice blast were no longer needed after just two years.


The experiment, covering 100,000 acres, "is a calculated reversal of the extreme monoculture that is spreading throughout agriculture, pushed by new developments in plant genetics," observed Martin S. Wolfe in an August 17, 2000 commentary in Nature. The problem, Wolfe suggests, is that monocultures provide a field of dreams for the development of super pests. The conventional solution -- to breed resistant varieties and develop new fungicides -- leads to rapid pest resistance.  "Continual replacement of crops and fungicides is possible, but only at considerable cost to farmer, consumer, and environment."


These costs make the virtues of the new rice system all the more dramatic. How was rice blast overcome?  Researchers, Wolfe says, have identified several factors in play.  To begin with, a more disease-resistant crop, interplanted with a less resistant crop, can act as a physical barrier to the spread of disease spores.  Second, when you have more than one crop variety, you also have a more balanced array of beneficial and potentially harmful pests that hold each other in check.  A single pathogen, such as the one involved in rice blast, is therefore less likely to gain the upper hand.


Also, of the two varieties of rice used in the Chinese experiment, the taller variety was the one more susceptible to blast.  But, when planted in alternating rows with the shorter variety, the taller rice enjoyed sunnier, warmer, and drier conditions, which appeared to inhibit the fungus.


And, finally, a kind of immunization occurs when crops are exposed to a diversity of pathogens.  Upon being attacked by a less virulent pathogen, a plant's immune system is stimulated, so that it can then resist even a pathogen that it would "normally" (that is, in a monoculture) succumb to.


This last point reminds us that disease susceptibility is not a fixed trait of a crop variety, but relative to the conditions under which the crop is grown.  Many existing susceptibilities reflect the crop's extreme isolation from anything like a natural or supportive environment, with its checks and balances.  This environment includes not only other plants, but also the complex, teeming life of the soil -- life that is badly compromised by "efficient" applications of fertilizers, herbicides, and pesticides.  And, as these new findings indicate, even a "healthy" variety of disease organisms is important.  What biotech company, focused on the latest, profit-promising lethal gene, would encourage such a balanced awareness among farmers?


Should the Students Re-engineer the Teacher?


When biotech proponents say, as they often do, "Prove to us that anyone has died or been made seriously sick by genetically engineered foods," the pathology is in the question itself.  The underlying stance is, "If you can't show us the corpses, where the hell's the problem?"  This suggests a complete unawareness of the ecological, social, economic, and ethical questions posed by the whole trend of technological agriculture.


If the right questions were being asked by those pushing biotech on farmers, they would be saying, "Look, here's why we think this kind of crop -- and farm, and business structure, and community -- is better for society than a highly diversified, local, small farm-based, organic agriculture."


But they do not address this larger picture, continually drawing our attention instead to particular technological achievements.  They offer the farmer specific "solutions," but, as Amory Lovins, co-founder of the Rocky Mountain Institute, has remarked, "If you don't know how things are connected, then often the cause of problems is solutions."  Nor are they quick to mention the one way their systems do surpass all alternatives: they offer more patent opportunities for biotechnology concerns.  It's hard to package all the local variations and contingencies of an environmentally healthy agriculture into a proprietary, uniform, for-all- purposes commercial system.


The question is why we would want such a package.  The assembly-line uniformity and near-sterility of those endless Nebraskan corn fields certainly do appeal to some of our current inclinations, but they are not the inclinations of nature.  It's true that we must work creatively upon nature.  But eliciting the yet-unrealized potentials of an ecosystem is one thing; firing silver bullets at it is quite another.  We have scarcely begun to understand all that nature can teach us about the bounty of the earth, and it would be a shame for the students, having gained a little knowledge, to attempt an ambitious re-engineering of the teacher.


Biologist Craig Holdrege is author of Genetics and the Manipulation of Life: The Forgotten Factor of Context, and director of The Nature Institute in Ghent, New York www.natureinstitute.org. Steve Talbott is a senior researcher at The Nature Institute and editor of its hardcopy newsletter, In Context.


Copyright 2001 by The Nature Institute.



[email protected]