Biotechnology has been described as a tsunami washing over U.S. agriculture. The industry, which has been transformed during this century by hybrid seed, synthetic fertilizers, and chemical pesticides, faces swift and broad change. The potential of biotechnology to increase food production in the U.S. and around the globe has received wide exposure in the media.

The effects of biotechnology, however, extend well beyond food production to food safety and the environment. My remarks focus on the environmental issues. Advocates of genetically modified (GM) crops assert that large environmental benefits will result from their adoption, including less use of toxic pesticides. Opponents argue, with equal vigor, that the new technology holds the potential for significant and perhaps irreversible environmental damages, including unintended effects on species diversity.

In truth, the environmental legacy of agricultural biotechnology cannot yet be reliably predicted. The lack of precision should not come as a surprise. The commercialization and adoption of genetically modified crops by farmers are occurring at unprecedented rates for a new technology. Not surprisingly, sparse science and little experience are available to forecast the effects of these organisms on diverse environmental systems.¹ Moreover, in contrast to new knowledge that raises yields or lowers cost, the science to understand natural ecosystem effects is not pulled by potential profits. The weak environmental science base and the strong private incentives for selling genetically-modified (GM) crops could lead to the premature approval of some products.

The environmental impacts of the widespread use of agricultural biotechnology merit further study, given their substantial potential benefits and downside risks. In my comments today, I first review the major environmental effects linked to GM crops. The review is based largely on a statement by the Royal Society, which I highly recommend (Royal Society, 1998). Based on the pace and scope of potential effects induced by GM crops, I argue for caution and more research to avoid excessive environmental risks, and to capture the full benefits of the technology.

Potential Environmental Benefits

The environmental damages of farm production increasingly appear in the media. Water pollution by fertilizers and pesticides, wetland habitat loss, and air pollution have become issues of public concern in many areas (Ervin et al., 1998). The public that formerly granted agriculture an exemption from most environmental regulations seems increasingly willing to impose controls on offending farm practices. Thus, new technologies that have the potential to improve environmental quality and increase profits hold appeal among farmers, agribusiness firms, environmental regulators, and the public.

Four types of environmental benefits may flow from the use of GM crops.

1. Reductions in chemical pesticides or use of more environmentally-benign pesticides

The first generation of crop biotechnologies has emphasized improved pest protection. New varieties of transgenic soybeans, corn, and cotton were marketed for the first time in the 1996 crop year. For example, Bt² corn, Bt cotton, and "Roundup Ready" soybeans are genetically engineered with traits that are resistant to pests, or that are tolerant to pesticides with lower environmental risks, e.g., glyphosate. Sales of the new seeds have been brisk, growing from a few percentage points of total acres in 1996 to 44 percent of all cotton acres, 38 percent of soybeans, and 24 percent of corn acres in 1998 (Union of Concerned Scientists, 1998). Part of the reason for their rapid introduction is that about one-half the time and one-fifth to one-seventh the cost are required to approve a new biotechnology product, compared to a new chemical pesticide compound (Ollinger and Fernandez-Cornejo, 1995).

In theory, the transgenic crops will reduce levels of chemical pesticides and lower the environmental risks of pesticides still in use. In practice, it is too early to tell how pesticide uses will change. The final outcome hinges on uncertain farmer behavior and pest population dynamics, such as the development of resistance to the pesticide (discussed below). We know that farmers do not automatically decrease fertilizer applications that are unnecessary to reach their yield goals, because they fear losses from adverse weather. A similar response could apply to pesticides, resulting in less reduction in pesticides than predicted by biotechnology proponents.

2. Savings in energy and air emissions from more efficient transport of less-perishable products

One of the first bioengineered products approved for commercial use in agriculture was the "Flavr-Savr" tomato. A principal motive for breeding the new tomato was to enhance its hardiness during shipping, resulting in less spoilage and more efficient transport. If less damage is incurred for this and other biotechnology foods, less energy will be used to ship a given amount of product, and air pollution will decline. Little evidence exists on this conceptual effect, primarily because the changes in transport under the new GM crops are not yet clear. The ultimate impact depends on the adjustments in transport patterns and technology and on consumer acceptance of the products.

3. Less irrigation water use for drought-resistant varieties

Water is a critical input for crop production. Many semi-arid and arid regions must irrigate to produce agricultural crops. Even many temperate areas rely on supplemental irrigation to avoid yield reductions during low rainfall and high temperature conditions. If the predictions of climate change causing more variability in weather conditions are accurate, even greater reliance on irrigation may be necessary. An early promise of biotechnology was to deliver more drought-tolerant crop varieties that would minimize yield loss and quality damage during such climate stress. Any savings in water use could maintain in-stream water uses, e.g., fish habitat, or reduce the withdrawals from depletable aquifers. The promise of drought-tolerant GM crops is appealing, but as yet unfulfilled.

4. Reduced conversion of habitat-rich grassland and forests, from higher-yielding varieties

A central premise of using GM crops for environmental improvement is that higher yields will reduce the need to convert pasture or forest lands into production as the world population grows. The grasslands and forests spared from conversion can remain as habitat for wildlife and other uses. The validity of this conceptual argument hinges on the answers to several linked questions. Will the actual yields be significantly higher? If so, will the decreased pressure for conversions occur where habitat preservation is needed to maintain species? If yes, will the habitat-rich lands remain out of production during periodic rises in crop prices? A lesson from the U.S. experience with the Soil Bank Program of the 1960s is that the retired lands soon return to production under high prices, as happened in the early 1970s. Hence, yield increases are a necessary but insufficient step to assure protection of this type of habitat.

Potential Environmental Costs

Unexpected costs often accompany the benefits of new technologies. The annals of science and industry are full of such surprises, such as the waste disposal problems for nuclear reactors. In the case of GM crops, several risks have been identified. The Royal Society statement covers five types of negative environmental effects.

1. Increased rate of pest-resistance

We know from decades of experience that pests become increasingly resistant to chemical pesticides via natural selection. That is, the pests whose genetic makeup makes them less susceptible to control by the pesticide survive at greater rates and then multiply to become larger parts of the population. This process of increasing pest resistance over time seemingly applies to all pest controls. The key to limiting the growth in resistance, and subsequent decline in pesticide efficacy, is to maintain a large pool of pests with susceptible genes. In short, limit the application of the pesticide. If the increased resistance is not managed properly, the rate of pesticide application must be increased, thus, ironically, hastening resistance, or new compounds must be invented. This process has been aptly described as the "pesticide treadmill."

Increased pest resistance is most worrisome for the introductions of the Bt cotton and Bt corn. If the susceptible gene pool decreases too far it may be impossible to resurrect the original population, adding an element of irreversibility. Bt, a biological control, is an important pesticide for organic agriculture. If the efficacy of a Bt strain is lost, it may be necessary to replace it with a synthetic pesticide that has higher environmental toxicity. Organic agriculture would be left without a key pest control option in this case. To counter the natural erosion in efficacy, the manufacturers of Bt crops require that a minimum portion of fields planted with their GM seeds remain in non-Bt varieties, to allow cross breeding and sustain the susceptible gene pool. The necessary size of this "refugia" to assure sustained efficacy is a good example of immature science on these effects. Manufacturers of Bt crops have gradually increased refugia size as they have discovered higher rates of resistance than originally expected. The Royal Society statement characterizes this environmental risk as significant and urges regulation to counter the potential damages.

2. Adverse effects on non-target species

Three types of non-target species effects exist. First, certain GM crops can decrease the population of pests upon which natural predators rely. The reduced numbers of pests diminish prey numbers and thereby threaten the populations of these natural enemies that live exclusively on the pests. This type of effect can apply to all pesticides, not just GM crops. According to the Royal Society review, the majority of studies to date have not shown significant adverse effects of this type. Two studies have detected effects on certain beneficial species, such as lacewings and ladybirds (Birch et al., 1997; Hillbeck et al., 1998). These findings underscore the need for more advanced experimentation to detect and understand the full range of non-target species effects. In particular, the effects on soil biodiversity merit study because we have only a partial inventory of soil microbiology, such as organisms living in the root systems of weed species that are eliminated by herbicide-tolerant GM crops. Second, GM crops may affect insect pollinators, such as bumblebees and honeybees. Toxins from Bt crops have thus far shown no adverse effects on these bees, but further studies are being conducted. Third, the Bt crops may kill other non-target insects through pollen drift, as shown in the experiment in which Monarch butterfly larvae died (Loosey et al., 1999). The field significance of this laboratory finding is as yet unclear.

3. Transfer of genes to wild relatives

An issue raised about herbicide-tolerant GM crops is the transfer of genes to wild relatives, creating weeds resistant to the herbicides. Some fear that transgenes from several herbicide-tolerant varieties could become concentrated in one weed species, resulting in a "super weed" resistant to all known forms of control. These risks exist also for traditional breeding of herbicide-tolerant crops. However, some GM plants appear to have an abnormally high chance of outcrossing compared to conventionally-bred varieties (Bergleson et al., 1998). The authors explain that their findings do not prove that the transgene itself causes the enhanced outcrossing, but rather that a difference in outcrossing between transgenic and mutant plants exists. The likelihood of such a gene transfer depends on the particular species and the location; hence, generalizations are unwise. The probability of gene transfer is greater if the crop variety is an outbreeding species for which crosses occur between distantly-related parent plants, such as canola, rather than inbreeding species, such as soybeans. The Royal Society concluded, "[I]t is inevitable that some gene transfer will occur from certain crops, but the level of gene transfer to wild relatives from GM crops is likely to be the same as from non-GM crops [p.6]." Nonetheless, they recommend that a responsible public body address the question of GM and non-GM crop gene transfer.

4. Transfer of GM genes to other commercial crops

Potential gene transfer between GM and non-GM commercial species is of concern also. The inserted genes may have harmful effects on other crops, such as when two herbicide-tolerant crops cross and present pest control problems for volunteer plants in the next generation. GM and non-GM plants of the same species may cross if the crops are adjacent. It is this risk that is most worrisome to organic farmers. If their crops are found to contain a significant portion of GM genes, then they cannot sell them for premiums in the organic market. The cost also would apply to growing organic seed crops. Again, crops that outbreed, such as maize and canola, pose the largest risks. The principal way to control these risks is by ensuring adequate distance between fields of GM crops and crops susceptible to such gene transfer. Evidence on the frequency and extent of gene transfer is sparse for GM crops, as might be expected. The Royal Society recommends more investigation and stronger guidelines for isolation distances in the U.K.

5. Transfer or combination of virus-resistant crop genes

The fifth set of possible environmental risks relates to the development of virus-resistant crops. Biotechnology scientists focused on virus resistance for crops first because viruses have a simple genetic makeup that is relatively well understood. Three possible environmental risks have been identified for those plants in which viral DNA sequences are inserted in the plant to interfere with the infecting virus and give "pathogen-derived protection" (Royal Society, 1998, p.11). First, viral "coat protein" genes in certain plants may be taken up by unrelated viruses infecting the plants, and change the method of virus transmission between plants. The second risk is that an inserted gene will recombine with unrelated viruses infecting the GM plants and thereby create a new, unknown virus. Third, it is possible that switching on the inserted gene at the same time as the plant is infected with an unrelated virus will aggravate symptoms. Each of these risks must be compared to the common natural situation of joint infections by two or more viruses, and evaluated on a case-by-case basis. A summary judgment of the seriousness of this environmental risk was not made.

Interpreting Environmental Risks

The Royal Society statement emphasizes repeatedly that all environmental risks of GM crops must be evaluated relative to environmental risks for non-GM crop situations. Parenthetically, this comparison should not be restricted to current, conventional production technologies. It should include alternative agricultural systems that are technically and economically viable, such as organic and integrated pest management (IPM) approaches. In economic analysis, this means that we must compare the with GM crop situation to the most likely situation without GM crops. To illustrate, if GM crops are not used in some situations, higher levels of synthetic pesticides may be applied. In other cases, GM crops may displace production that employs few or no pesticide applications (e.g., organic).

What a reader is struck by in reviewing the Royal Society statement are the few studies that have documented environmental risks of individual GM crops. Given that most GM crops have been grown commercially only since 1996, there are few research studies on possible environmental effects in the laboratory and especially in the field. In other words, we are very early on the learning curve regarding GM crops, as emphasized at the outset. Under these conditions, there is a strong need for post-release monitoring of the natural environment affected by all GM products. Those monitoring systems are generally not in place in the U.S. For example, we need to inventory the numbers and variety of species of nematodes or bacteria, the populations of insects, and numbers of birds and animals. Given the rapid pace of adoption, we should have early warning systems to avert large and possibly irreversible effects.

A strong concluding message in the Royal Society statement is the need for evaluating the environmental effects of GM crops as a whole rather than as one-by-one regulatory reviews. If potential GM crops are screened on an individual risk-benefit basis, the decisions may result in a lack of analysis of the overall impact of the set of technologies on whole ecosystems. Such an individual crop approach will not likely capture the full set of long-term effects of GM crops, whether positive or negative. Moreover, the case-by-case analysis will miss possible interaction (synergistic) and scale effects, as we have learned with pesticide reviews. This may occur, for example, if different GM crops exert common environmental influences. The analogy in pesticide regulation is the cumulative effects of a family of compounds that may affect endocrine function. Similarly, individual reviews will miss threshold issues that may occur if common environmental effects of GM crops are aggregated across the countryside, e.g., predator-prey relationships.

Why Caution about Agricultural Biotechnology and the Environment is Needed

Given the small science and experiential bases for evaluating the short-term effects of GM crops on our natural ecosystems, not to mention the long-term, whole-system effects, more research and monitoring seem prudent strategies. As mentioned repeatedly, the pace and scale of change that may cause large and irreversible environmental effects reinforces this conclusion. Further argument for caution is the risk that early domination by industry leaders may concentrate risks and limit innovation and diversity in biotechnologies. Diversity can protect against unexpected environmental risks from widespread adoption of a few dominant products.

We have time to be cautious. World food supplies are plentiful, and productivity advancements apart from biotechnology continue. Indeed the real price of food has fallen over the past 30 years, and world population estimates continue to fall, now down to approximately 7.9 billion. The most serious obstacles to feeding developing country populations are generally caused by low incomes, inadequate human capital, and poor distribution systems, not production shortfalls. GM crop technology can come on stream very rapidly once adequate reviews have been conducted to assure that excessive environmental and human health risks will not arise.

Expert reviews conclude that the most serious constraint on future global food production is not lack of technology (Crosson and Anderson, 1992). The real shortage is in developing human capital and research to manage agricultural systems in ways that will provide abundant and low-cost food, while also conserving natural resources, protecting the indigenous environment, and preserving cultural values. Human capital is essential to devising agricultural systems that fit the physical, biological, economic, social, and cultural bases that govern food production in those countries. In some cases, GM crops may be quite appropriate. In others, they will not be, as evidenced by the decision by the Consultative Group on International Agricultural Research (CGIAR) to reject the "terminator" technology at this point.

Biotechnology's potential environmental effects mostly affect public good resources, such as pest-resistance gene pools and biodiversity. Hence, regulation is necessary to ensure full consideration of these broad public values. However, the existing regulatory system is segmented, resembling a "patchwork quilt" of oversight, with potential gaps due to immature science. This incomplete regulatory system, much of it adapted from pesticide reviews, may miss serious short-term risks. More importantly, it does not send appropriate signals to the public and private research systems about which paths for agricultural biotechnologies will likely offer the greatest social net benefits over the long run (Ervin and Schmitz, 1996).

There are divergent views on the adequacy of the existing regulatory system for agricultural biotechnologies (Environmental Law Institute, 1999). These differences should be expected for such a volatile technological area. We have substantial theory and evidence to suggest that agencies responsible for regulating certain industries can be captured by the regulated industries. Given the vast importance of the agricultural biotechnological revolution for food, energy, human health, and the environment, an independent over-arching body commissioned by the government may be needed to evaluate the full sweep of issues emerging from the GM crops. The Royal Society Committee endorsed the creation of such a body, but the U.K. has a different regulatory institutional structure. The stakes for assuring sound oversight and decisions about GM crops and livestock are large. Getting the regulatory institutions right is not only in the interests of those concerned about negative environmental effects, but for the industry as well. If a large human or environmental health catastrophe emerges due to poor oversight, it might cause not only a short-term setback for the industry, but also jeopardize the entire future of biotechnology and its considerable potential. Hence, sound approaches that minimize the maximum potential loss for industry and the environment may be most prudent at this stage.

References

Bergleson, J., C. Purrington, and G. Wichmann. 1998. "Promiscuity in transgenic plants." Nature 395(Sept. 3): 25.

Birch, A.N.E., I.E. Geoghegan, M.E.N. Majerus, C. Hackett, and J. Allen. 1997. "Interactions between plant resistance genes, pest aphid populations and beneficial aphid predators." Scottish Crop Research Institute Annual Report. Dundee, Scotland. p. 68-72.

Crosson, P., and J. Anderson. 1992. Resources and Global Food Prospects – Supply and Demand for Cereals to 2030. World Bank Technical Paper no. 184. World Bank, Washington, DC.

Environmental Law Institute. 1999. "Forum: Regulating Biotechnology." Environmental Forum 16(2): 46-56.

Ervin, D.E., C.F. Runge, E. Graffy, W. Anthony, S.S. Batie, P. Faeth, T. Penny, and T. Warman. 1998. "Agriculture and the environment: A new strategic vision." Environment 40(6): 8-15, 35-40.

Ervin, D.E., and A. Schmitz. 1996. "A new era of environmental management in agriculture?" American Journal of Agricultural Economics 78:198-206.

Hillbeck, A., M. Baumgartner, P.M. Fried, and F. Bigler. 1998. "Effects of transgenic Bacillus thuringiensis corn-fed prey on mortality and development time of immature Chrysoperla carnea. (Neuroptera: Chrysopidae)." Environmental Entomology 27(2): 480-487.

Hubbell, B., and R. Welsh. 1998. "Transgenic crops: Engineering a more sustainable agriculture?" Agriculture and Human Values 15:43-56.

Loosey, J., L. Rayor, and M. Carter. 1999. "Transgenic pollen harms Monarch larvae." Nature 399: 214.

Ollinger, M., and J. Fernandez-Cornejo. 1995. Regulation, Innovation and Market Structure in the U.S. Pesticide Industry. AER-719. U.S. Dept. of Agriculture, Economic Research Service, Washington, DC.

Royal Society. 1998. Genetically Modified Plants for Food Use. London.

Union of Concerned Scientists. 1998. "A surge in commercial transgenic crops." Gene Exchange (Fall/Wtr.): 7,13.

Williams, N. 1998. "Agricultural biotech faces backlash in Europe." Science 281(Aug. 7): 768-771.

Endnotes

1. The development of genetically altered animals is on a slower track due to a variety of technical, social and ethical concerns that arise with animals but not with plants. Nonetheless, there is every reason to suspect that these animal biotechnologies will gather momentum as the lure of improved quality, lower costs, and higher profits drive their development and adoption.

2. Bt stands for Bacillus thuringiensis, a bacterium present in soils that acts as a natural insecticide. Gene sequences controlling the production of crystal endotoxins in B. thuringiensis have been genetically engineered into the corn plant itself, creating resistance to the European corn borer.

All information contained on this Web page may be used without permission, provided credit is given.