Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations In Conjunction with the IUFRO Conference on Tree Biotechnology in the New Millennium

July 22–24, 2001

Steven H. Strauss ([email protected]), Oregon State University and H.D. (Toby) Bradshaw ([email protected]), Conference Organizers

Skamania Lodge, Columbia River Gorge, Stevenson, Washington, USA

See the IUFRO meeting web site, to download a copy of this proceedings, and to view all poster and paper abstracts: http://www.fsl.orst.edu/tgerc/iufro2001/ PREFACE

ore than 200 economists, ecologists, environmentalists, ethicists, molecular biologists, industry representatives, and government regulators from 23 countries con- M vened at Skamania Lodge, along the scenic Columbia River Gorge between Wash- ington and Oregon, for a 2-day symposium on the ecological and societal aspects of transgenic forest plantations (http://www.fsl.orst.edu/tgerc/iufro2001/eco_symp_iufro.htm). Of the 28 invited lectures given at the symposium, 14 were from scholars who presented a broad environmental, ecological, or ethical view. The symposium was held in conjunction with the biennial meeting (22–27 July 2001) of the International Union of Forestry Research Organizations Unit on the Molecular Biology of Forest Trees (Vienna, Austria: http://iufro.boku.ac.at/). This proceedings attempts to capture the main issues raised in the lectures, breakout sessions, and summary statements. Among the widely accepted conclusions from the conference are the following:

•A great deal more scientific research is the most glaring need to help answer questions of benefit and safety, and thus of social acceptability. The question of “do we really need it?” cannot be answered until much more is learned from laboratory and field research. •Plantation forests have the potential to concentrate industrial wood production on a small land base, and thus spare wild forests from intensive harvest. Whether this actually occurs depends on social mechanisms for protection as well as on technological innovation. • The long lifespan of trees, and the common presence of wild or feral relatives, are particularly troublesome for benefit/safety assessment, which is therefore likely to require a combination of modeling, monitoring, and adaptive management. Genetically engineered flowering control was considered critical for some traits to restrict transgene dispersal. However, the notion that trees engineered with one or a few genes are functionally analogous to exotic invasive organisms, and thus are likely to “threaten” wild forests, was widely rejected. •Research to date has demonstrated a high degree of health and stability of performance of genetically engineered trees in field trials oriented toward possible commercial applications. •None of the attendees called for a moratorium on all field research with genetically engineered trees, as Greenpeace has demanded. But Faith Campbell (American Lands Alliance) and Sue Mayer (GeneWatch UK) called for a moratorium on the commercial release of genetically engineered trees until further research indicates that large-scale plantings would be environmentally safe. •Participants expressed a high level of optimism about the potential for research to enable beneficial applications of GM trees. A survey taken at the end of the meeting (Appendix 1) revealed that 88% agreed or strongly agreed with the statement “Sufficient basic and applied research upon the genetic stability and ecosystem behavior of GM trees, and upon the design of biological safety mechanisms, can create environmentally safe and societally beneficial trees and outcomes.” This was the first international symposium to attempt to forge a consensus on how to move forward in research and public debate. It is abundantly clear that substantial progress was made in many directions, however it is also clear that many additional steps will be needed for the potential benefits and acceptability of genetic engineering in forestry to be fully explored.

H.D. Bradshaw, Jr. and Steven H. Strauss, 13 November 2001 TABLE OF CONTENTS

Welcome to Symposia Steve Strauss and Toby Bradshaw ...... 7 Introduction of Keynote Speaker Toby Bradshaw ...... 9 Keynote Address: Future Forests: Environmental and Social Contexts for Forest Biotechnologies Hal Salwasser ...... 10 Context and Goals for Symposium Steve Strauss ...... 20 Welcome to the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, in Conjunction with the IUFRO Conference on Tree Biotechnology in the New Millennium Paul G. Risser ...... 23 Moderator Introductions Steve Strauss ...... 25

POLICY AND ECONOMICS ...... 27 The Economic Contribution of Biotechnology and Forest Plantations in Global Wood Supply and Forest Conservation Roger A. Sedjo ...... 29 Restoring the Forests David G. Victor and Jesse H. Ausubel ...... 47 Biotechnology and the Forest Products Industry Alan A. Lucier, Maud Hinchee, and Rex B. McCullough ...... 57 Responding to New Trees and to the Issues at Hand: The Institute of Forest Biotechnology W. Steven Burke ...... 62 Will the Marketplace See the Sustainable Forest for the Transgenic Trees? Don S. Doering ...... 70

ETHICS ...... 83 The Ethics of Molecular Silviculture Paul B. Thompson ...... 85 Ethical and Social Considerations in Commercial Uses of Food and Fiber Crops Sandy Thomas ...... 92

REGULATORY POLICIES ...... 99 Regulation of Transgenic Plants in the United States David S. Heron and John L. Kough ...... 101 International Regulation and Public Acceptance of GM Trees: Demanding a New Approach to Risk Evaluation Sue Mayer ...... 105

FOREST SCIENCE ...... 111 Transgenic Trees: Where are We Now? Dave Ellis, Rick Meilan, Gilles Pilate, and Jeffrey S. Skinner ...... 113 Perspectives on Risk in Transgenic Forest Plantations in Relation to Conventional Breeding and Use of Exotic Pines and Eucalypts: Viewpoints of Practicing Breeding and Transformation Scientists R.D. Burdon and C. Walter ...... 124 Use of Transgenic Resistance in Plantation Trees: Efficacy, Risk, and Integration with Other Pest Management Tactics Kenneth F. Raffa ...... 138

ECOLOGICAL SCIENCE ...... 149 Ecological Considerations for Potentially Sustainable Plantation Forests James R. Boyle, Heléne Lundkvist, and C.T. Smith ...... 151 GE Trees: Proceed only with Caution Faith T. Campbell and Rachel Asante-Owusu ...... 158 Biodiversity Implications of Transgenic Plantations John P. Hayes ...... 168 Potential Impacts of Genetically Modified Trees on Biodiversity of Forestry Plantations: a Global Perspective Brian Johnson and Keith Kirby ...... 176 Invasiveness of Transgenic vs. Exotic Plant Species: How Useful is the Analogy? James F. Hancock and Karen E. Hokanson ...... 187

BREAKOUT SESSIONS ...... 193 Instructions to Leaders...... 195 1A — Social / Global Context Daniel B. Botkin and Jarmo Schrader ...... 197 1B — Social /Global Context Trevor Fenning ...... 199 2A — Biological Context /Biodiversity Alvin Yanchuk ...... 201 2B — Biological Context /Biodiversity Gerald Tuskan and John Hayes ...... 203 3A — Invasiveness / Weediness Jim Hancock ...... 205 3B — Invasiveness / Weediness Don Doering ...... 206 4A — Wood Modification Chung-jui Tsai ...... 208 4B — Wood Modification Malcolm Campbell ...... 210 5A — Benefits and Safety of Pest Management Applications Armand Seguin ...... 213 5B — Benefits and Safety of Pest Management Applications Sarah Covert ...... 214

SUMMARY V IEWS ...... 217 Policy Perspective Roger A. Sedjo ...... 219 Ethics Perspective Paul B. Thompson ...... 220 Industry Perspective Alan A. Lucier, Maud Hinchee, and Rex B. McCullough ...... 221 Ecological Science Perspective Kenneth F. Raffa ...... 222 Forest Biotechnology Perspective Steve Strauss and H.D. (Toby) Bradshaw ...... 223 Environmental NGO Perspective Faith Campbell, Sue Mayer, and Mario Rautner ...... 225

MODERATOR PERSPECTIVES ...... 227 Lessons from Twenty-five Years of Debate on the Risks and Benefits of Biotechnology Michael T. Clegg ...... 228 The Urgent Need for Field and Laboratory Science to Inform Ethical and Policy Debates Hal Salwasser...... 231

APPENDICES ...... 233 Appendix 1 — Exit Survey of Symposium Participants ...... 235 Appendix 2 — Participants, Tree Biotechnology Meetings, July 22–27, 2001...... 238 Welcome to EcoSocial and Molecular Biology Symposia

Steve Strauss Dear Colleagues,

n behalf of my co-organizer Toby Bradshaw, my associates at Oregon State University, and the staff of the conference office at the College O of Forestry at OSU, I want to officially welcome all of you to this conference. We have been planning it for two years, from logistics, to complex issues of scientific coverage and speaker selections, to obtaining grant and sponsor support to make it possible. It has been an amazing labor of love. This is by far the biggest party we have ever thrown. My sincere thanks to the many sponsors who have made it possible, which in addition to those listed at the front of our binders, include CSIRO Australia and some others who, for fear of terrorism, elected not to have their names listed. What a pity that in this great democracy we live in, an organization feels that it cannot support an open scientific conference without retribution. While we welcome peaceful protests, like the one we witnessed earlier today, there is no place or need for violence in the already vigorous world debate on biotechnology. It may be a big party, but it will also be the most demanding and exhaust- ing party I have ever attended. The first two days, starting this evening with our keynote lecture, will be an intensive and controversial journey, as we explore the many and diverse ecological and social issues that surround genetic engineering in forestry. We will need to listen carefully not only to the technical views, but to the ethical and personal attitudes that underlie them. This will be a great challenge for all of us, and with the strong media presence here, I can tell you that the world is watching. Luckily, we have a brilliant, thoughtful, and responsible set of speakers to help guide us. Then with little break, we will begin the hard biotechnology science, which will continue all day in close succession, including two concurrent evening sessions, through mid-day Friday. As always, this is complex stuff—even for us experts. Please pace yourself. However, we are fortunate to have a truly outstanding slate of scholars that it is no stretch to say represent the best that the science of forest biotechnology has to offer anywhere on planet Earth. Also note that there is a survey in the front binder. We will ask you to fill this out, on the associated electronically scored sheet, at the end of the EcoSocial symposium. Please wait until then so you will have had the benefit of the symposium to inform you. This will help us to record the views of the attendees with respect to a number of issues. Finally, I want to thank all of the attendees for making this conference a priority in their professional and personal lives. As many of you have pointed out, the cost is by no means insignificant. The time away from your families is always hard. And many of you have traveled across many time zones to get here. The great turnout—by far the highest yet for an IUFRO forest biotechnology meeting—shows that the agenda of topics and speakers are of interest to many. I would now like to introduce my colleague and co-organizer Toby Bradshaw, who many of you know well. He will introduce our keynote speaker. Toby is a professor at the University of Washington in nearby Seattle, and has conducted

7 pioneering studies on the genes that control adaptive traits in plants, including poplars, for many years.

8 Introduction of Keynote Speaker

Toby Bradshaw

t is my privilege tonight to introduce the keynote speaker, Dr. Hal Salwasser, from Oregon State University. Hal is the Dean of the College of Forestry at I Oregon State, as well as the Director of the Forest Research Laboratory there. Hal came to OSU from a career at the U.S. Forest Service, where his last post was as the Director of the Pacific Southwest Research Station at Berkeley, Cali- fornia. He was the first person to hold the Boone and Crockett Endowed Profes- sorship at the University of Montana, where his work on wildlife conservation is well known. His undergraduate work was done at Cal State-Fresno, and he received his PhD in Wildland Resource Science from Berkeley. Hal’s talk tonight will be on the subject of future forests: environmental and social context for forest biotechnology.

9 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 10-19. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Keynote Address

Future Forests: Environmental and Social Contexts for Forest Biotechnologies

Hal Salwasser t is a privilege to be asked to share some thoughts with a group of scholars and interested citizens on such an important topic as Future Forests: Environ- I mental and Social Contexts for Forest Biotechnologies. Its significance is signaled by how many people have come from so many different countries to participate in the symposium. I am most pleased that two great public universities have hosted an open, public, scientific forum on such an important matter. This is the right thing for leading institutions of learning to do, to engage people with diverse perspectives and experiences to talk about, not just the science, but the social and ethical implications of biotechnology in general, and genetically modified organisms in particular. To set the stage for this meeting, I am going share some perspectives on forests, forest management, forest conservation, and the role that biotechnology might play in future forests. I generally begin all of my presentations with a question, a fairly simple question: Why are we talking about this stuff anyway? I have a number of answers to that question, and I encourage you to think about why we should be talking about forests and biotechnology as well. The order of my answers will probably disclose some of my philosophical biases because these are in the order that I tend to think of them. Fundamentally, we talk about forests because forest ecosystems are vital for life on earth. They form the headwaters of our major river systems. They sustain biological diversity and wildlife habitats in extraordinarily rich ecosystems. They are, obviously, the sources of wood, which is an environmentally superior raw material, when you stack it up against the other things that people might use as alternatives. For example, if we compare the energy and water use for building a 10 x 100 ft wall out of steel versus out of wood, we find that the wood wall uses much less energy and water than the steel wall. In addition to their environmental superiority, wood products meet many essential needs. Every day of our lives we encounter the benefits of wood. You can just look around this room and this marvelous hotel and get a first hand feel for that. Globally, wood is also an in- credibly important material for energy. In some developing countries of the world as much as 70% of the energy for cooking and heating in rural areas still comes from wood. There is also the role of wood in carbon sequestration. The best estimates I’ve been able to find show that about 40% of the carbon that is stored on the land is stored in forests or forest soils. Now, granted most of the carbon stored in the world is in the oceans, but forests are the big players on land. Forests also provide a multitude of recreational and spiritual values. Many of our cultures across the globe view forests as a major part of their identities. As an angler and a sometimes-hunter—and a sometimes-hiker, although not as often as my wife would like me to be—forests are an important part of my identity. So forests have all of these incredible values, and are just really remarkable places. 10 Fortunately, forests still cover a fairly large area of the land surface of the world, about a quarter globally, about a third here in the United States (Figure 1). Here in the Pacific Rim states, forest cover ranges from just under 40% in California to just about 50% up in Washington. Globally and in the United States, this is not as much forest as we used to have. That’s because people transform forests through a wide variety of activities. This is not a recent occurrence. People have been transforming forests probably for as long as people have been around, at least for as long as they knew how to pick up a stick that had a fire burning on the end of it. Of these transformations, the most significant by far globally has been conversion of forests for agriculture and human dwellings. How we manage forests also transforms them. It typically does not change them from forests to something else, but it changes the character of the forests, their structure and composition. Examples of this are livestock grazing, recreation, and climate change—which of course, is a natural force of change, but the degree to which we’ve exacerbated climate change, that’s a human effect. And I would encourage you to think about the degree to which water diversions and dams have transformed forests. We must have hundreds of thousands of hectares of riparian forests now sitting underneath reservoirs in the United States alone.

1996 60 50 8000 bp cent 40 Per 30 as

st 20 Land Area re of 10 Fo 0 World United States

Figure 1. Forests cover a lot of land—but less than they used to. (Sources: World Resources 2000–2001; USDA FS RPA Assessment 2000.)

World population growth is the major force behind all this forest transformation, and we are still on a trajectory to increase our numbers by one-third to one-half by mid-century. So, we are not, by any means, free from the effects of human population growth. Some of the transformations that I’ve talked about tend to be relatively permanent, at least when we think of them in human lifetimes. Of course they are not permanent in geological times. Over very long periods of time, nature has a way of setting things back and erasing whatever im- print we’re probably going to have. But we generally consider changes such as urban sprawl and agricultural transformations to be forest lost. Not all the transformations are losses though; forests are restored by a lot of the things that we do. We plant forests; sometimes we plant them back on abandoned agricultural lands. We certainly plant them back in places where we have harvested trees. To me, those kinds of transformation are net benefits to the environment. Even though we can and do make beneficial changes in forests, there is still a problem: Globally, the losses are outstripping the gains. Since somewhere around the start of the agricultural revolution/industrial revolution, we’ve lost about 20%–50% of the original forested cover of the world. The 20% is an estimate on the low end, the 50% perhaps an estimate on the high end. It’s very hard to be precise about this. During this same period, we’ve gained about 1000%

11 in numbers of people. So when you compare the losses and the gains, we end up with a lot less forest area per person to provide all of those benefits we expect out of forests: the water, the carbon sequestration, the wood, the biological diversity, the recreational access, the cultural identities, and so forth. That creates the future that we are heading into. We are going to have a smaller forested area, and we are going to expect it to serve more people. As those people gain in knowledge and affluence, they are going to expect the forest to serve them in more ways. We have seen a 40% increase in wood use in the last three decades. We are anticipating seeing at least a 20% increase in the next two decades. We’ve been on a fairly steady trajectory of about 1% increase in global wood use per year on average (Figure 2).

35 30 25 20 15 10 Acres per Person 5 0 8,000 bp 1850 1930 1975 2000 2050

Figure 2. The amount of global forest per person has declined precipitously over the past 250 years. (Source: Source: World Resources 2000–2001, plus interpolations and projections.)

Meanwhile the demand for everything else we want from forests in addition to wood continues to grow. Water, I am convinced, is the ultimate resource of value to come from forests. When wars are fought over forest resources, they’re going to be over water. We are seeing that happen in Oregon right now this year. Biological diversity conservation has also grown in importance over the last couple of decades and is now recognized as a major benefit that we expect from forests. But biodiversity is a very complex issue without universally clear strategies on how best to conserve it; too often protecting it in the short run in one place merely shifts human impacts to later dates or to other places. We are just on the verge of understanding the role of forests as carbon stores. We will eventually recognize that forests and wood held in permanent status and not burned up in some way are going to play a major role in carbon sequestration. Now, the good news for future forests is that according to the best estimates— and some of those come from speakers here this week—we are within a decade or two of reaching a point at which approximately 40% of the industrial wood fiber produced in the world is going to come from planted forests. It is the hope and expectation of all foresters and conservationists that planted forests will relieve some of the pressure to take wood from native forests, at least in the developed part of the world. An important part of my message in all this, is that forests can and will be sustained through management. Without management, our experience is that people either convert forests to some non-forest use or they over-run the forests. They over-run it with agriculture, or urban growth, or livestock grazing, or just harvest too much stuff, and the stuff ends up being not just the trees, but the 12 plants and the forage and the wildlife and the water. With sustainable management, though, we have learned in many parts of the world that forests can be restored and can be protected. I don’t want to have you perceive that I am saying that through management we can restore, and protect, and sustain the same kind of native forest that was there once upon a time before there were people, but we can still sustain and protect some pretty diverse, pretty productive forests. So, let’s talk about what it will take to sustain forests for all their values and uses. We, in the state of Oregon, have been blessed with some visionary leadership over the years. And just recently we experienced another pulse in that leadership. Our governor, John Kitzhaber, and our legislature just passed a new law that is called the Oregon Sustainability Act of 2001. That law recognizes sustainability as a goal for all programs of state agencies in Oregon, and adopts the Brundtland Commission and the United Nations definitions of sustainable development as the core goal. The governor added a few words about using, developing, and protecting resources, and then made sure that people understood that this is all done from the joint perspective of meeting environmental, economic, and community objectives. This is now the law and policy of the state of Oregon, and it gives us a tremendous new tool to work with. Sustainability—this concept that is somewhat elusive, but big enough to provide a place for a lot of people to stand—is going to shape the future of forest management in Oregon, as it is in many other parts of the world. It already is shaping the forest management of today in many places, through both governmental programs and the programs of the marketplace and the private sector. It’s going to focus on meeting both current and future needs, so it has intergenerational equity built into it. It seeks a balance between environmental protection, economic development, and the perpetuation of diverse and vital communities. So it has a good balance. It views people as part of nature and as part of ecosystems, not as a separate entity or as something to be dealt with after you figure out how to take care of the non-people parts. It forces us to consider all the forests in our decision, not just the ones in our own backyard. If done well, we hope it will enable us to keep forest ecosystems healthy and productive. Sustainability is a concept that rests in a global context; it is not something we can accomplish at a local level just by focusing on local matters, not even if local means regional or national. Let me give you a couple of statistics here that illustrate why we must view forest sustainability in a global context. These percentages are accurate to within a few percentage points. I’m sure they vary from year to year, based on some market conditions. A year or so ago, I read a paper that said somewhere around 30%–35% of all the industrial wood—that’s construction lumber, wood panels, furniture, paper, and packaging—that is consumed in the world has crossed at least one international border from the time it was a tree until the time it was used. One third of the world’s wood products is moving across international boundaries. At least in recent years, the United States has imported between 35% and 40% of its softwood lumber from another country. Most of it comes from Canada, but increasingly it also comes from the southern hemisphere—from some of the fast-growing plantations in the southern hemisphere. More than wood is in the global marketplace. All over the world, we see forest enterprises that started out as local or regional companies expand to become national corporations and are now globally integrated corporations. Compa- 13 nies such as Weyerhaeuser in the United States, or UPM-Kymmene and Stora- Enso from Finland and Sweden. You used to be able to name the country that housed these firms, but no longer. They are not just U.S. companies or Scandina- vian companies. They have lands in different continents, they have mills in different continents, and they are marketing their products in a global marketplace. Carbon, wood, and biodiversity are all recognized as global issues now. People are actually selling carbon credits in one hemisphere to countries in another hemisphere. If that isn’t enough evidence, consider this: about one-third of the gradu- ate students in forestry at Oregon State University have come to school here from another country. The globalization of everything in forestry is just astounding. Let’s move on to forest management. Just about anywhere in the world, you’ll find that forests are managed for many different purposes. Some are managed to produce resources that people need, while others are managed for recreational purposes; some are managed for national parks, wilderness areas, and so forth, and yet others as nature reserves. Sustainable forestry must be as broad as those many different purposes because it is those purposes that we wish to sustain. So, sustainable forestry involves diverse forest types, it treats each of them differently, and it focuses on trying to match the goals and capabilities and needs with the kind of management. Now, what I’m saying to you here is that sustainable forestry is not defined by any single particular approach, and that it can be applied to the management and protection of a national park just as well as it can be applied to what you might call a tree farm or a fiber farm. Consider, three major points in this spectrum: industrial- strength forestry, integrated multi-benefit forestry, and wilderness or nature preservation forestry. I think of the most intensive types of forestry—almost on an agricultural mode of trying to put as much of the solar energy and the site’s productive capability into the fiber—as industrial-strength forestry. Most of the world’s industrial wood is going to come from these kinds of forest uses eventually. We are well on the way to a transition from extracting much of our wood from natural or semi- natural forests to getting most of it from this type of managed forest. There is a lot of potential for biotechnology and genetically modified organisms in industrial- strength forestry. Certainly if you increase productivity, equally important is to do that in ways that reduce environmental impact, such as by reducing the amount of chemicals, fertilizers, pesticides, maybe even water, used in production, while improving product quality and consistency. So as we consider the topics this week, it is appropriate to have a focus on the utility of industrial strength forestry. If predictions by some of the leading scientists are correct, we may end up with 10%–15%, perhaps even a little less, of the world’s forested area in industrial-strength forestry. We probably will have a similar percentage in parks and wilderness areas, also vitally important for the values they sustain. What that means is the remainder—the vast majority of the world’s forests— will be in some intermediate kind of stewardship, management that is integrated for multiple benefits. These integrated multi-benefit forests also hold potentials for biotechnology. The goal here is to do a better job of optimizing joint production. A major role of integrated, multi-benefit forestry will be to protect vulnerable endangered species, including species that are vulnerable to exotic pests or diseases. If we can figure out how to put resistant genes into those native strains, we may be able to reduce the potential for the next white pine blister rust, or chestnut blight, those kinds of diseases where the native species and the host country of the disease organism have some kind of resistance. 14 The third major type of forestry, we might call it nature preservation or reserve forestry, is what you might think of being practiced in parks and wilderness areas—an extremely important part of the whole landscape mosaic of forests in providing all the things that we need. But it would be a mistake to think that these places are not managed. I know of only a few places in the world, and they’re not in the United States, where these kinds of places are not managed. Here in the United States, we manage them very actively to reduce human activities and their impact, and to prevent exotic species from coming in or to try to get rid of them when they are there. We even have some national parks holding commercial timber sales to get the forest structure back to the conditions they want to sustain there. The key point is that these forests are not managed for economic gain. They are managed in ways that perpetuate their natural values. I think there will be some great potentials for genetically modified organisms here, especially if we can restore species that are endangered by exotic diseases and pests. And the great indirect benefit is that if we can figure out how to meet most of the world’s wood needs from industrial-strength and integrated multi-benefit forests, it should allow us to put more of the forest land into this more protected classification. Let us consider the roles of different owners for forest sustainability. Much of our focus in the past two decades has been on federal forestlands. That is the wrong place to put the focus for sustainable forestry. The most productive forests, and the largest land areas are held in private ownerships. These are not just industrial ownerships, they tend to be family ownerships—small tracts of land. There will be roles for national forests and national parks to play in sustainable forestry; they will largely be on the nature preservation, reserve forestry end of the spectrum. There will be roles for industry to play; they will largely be in the industrial-strength forestry end of the spectrum. But family forests, which make up about 60% of the forested area of the United States, are going to be major players. The percentages of how ownerships will contribute to the full spectrum of sustainable forestry will differ by countries of the world based on the kinds of tenured ownership they have. But just addressing the forest management parts of sustainability will not be sufficient. This is because of one fact: the managers and the forest industry that produces the wood products that we use, are not who create the demand. The demand for forest uses, products, and benefits—whether it’s the water, or the recreation, or the wood or the biodiversity—comes from everyone who uses or wants those products. So, the future for sustainability means that all people must be involved, including the forest managers, the manufacturers, and the end users. What we choose to use, how we choose to use it, where and how we decide to produce it and through what technologies, and what we decide to do with it when we are through—all of these are important points in reaching our goal of sustainability. We face a lot of challenges in aspiring to this goal in a world that is filling up with people. To date, we have been partially successful in using laws and regulations to prevent people from doing undesired things to forests and waterways. Now, we need some innovative policies that will entice people to do the right thing instead of just preventing them from doing the wrong thing. We need to recognize the trade-offs in the choices we make. In our world full of people, there are no easy choices. There are no choices without trade-offs in either the economic, or environmental, or social sectors. We need to continually invest in new knowledge and technologies, even if we just hope to keep up. We have to learn 15 to protect water, and fish, and wildlife more effectively in our managed forests; to extend lifelong learning so all people engaged in forest conservation and the use of forests will understand what it takes to provide those uses; and to create a common ground on sustainability. Now I’m going to take a little bit of a risk here and offer to you what I think is a simple five-step framework to organize our thinking and our dialog on sustainability. I don’t mean to imply that any of this is going to be easy. 1. We must focus ecosystem transformations so that overall sustainability is enhanced. We cannot stop ecosystem transformations but we can make them more conducive to environmental health, economic vitality and community livability.

2. We must begin to focus on renewable natural resources, and use and conserve them wisely. Shift as quickly as we can, as much as we can, to solar-powered resources. This is certainly important for the United States, because we are such huge consumers of non-renewable resources. 3. We must develop knowledge, technology, and systems for sustaining desired social, environmental, and economic conditions, while approaching these desired conditions simultaneously. This is enormously difficult to do. We’ve tended to go after them one at a time. We have economic development schemes, and then we have regulations to stop the economic development schemes to protect the environment, and somehow in all of this, the communities get lost in the process. We have examples showing up on the front pages of our newspapers every day about what is happening to local people as these single-dimension agendas are being worked out.

4. We must manage ecosystems, and especially the human enterprises based on our knowledge and technologies to meet these combined social, environmental, and economic goals.

5. We need to pay attention to the fact that not everybody is benefitting at the same rate and to the same degree in the economic development of this world. It greatly saddens me that the gap between the really well off and the very poor keeps getting wider and wider. That just doesn’t strike me as a sustainable proposition at all. It is inevitable that technology advances will occur that can help us meet human needs and goals for quality of life and equity. But if they don’t occur in places like this, in academic settings, in open, public, democratic, rigorous scientific forums, then they are going to occur behind closed doors, in the dark of night, without the safeguards, without the public dialogue. It’s not a question about whether genetically modified organisms are going to pop on the scene, and get into the environment. It’s a question of who is going to set the ground rules for how that is done, and to what degree will we use this technology in managing nature and in equitably meeting human needs. Those are the issues we must address this week. We also need to address concerns about transgenic trees themselves, issues about safety and security: • Whether, and the degree to which, they will enhance productivity

16 • Whether, and to the degree to which, they are going to help us reduce the environmental impact of meeting people’s needs for resources

• Whether we can use them to mitigate some of the undesired consequences of global change •How to deal with intellectual property rights

• The ethics of intervening with creation or nature •How are we going to make decisions on all this? Who gets to make the decisions? Who’s going to be participating and who might be left out?

These are not the only issues we need to deal with in the area of sustainable forestry. There are many places where we can and will do better. We can have better harvest practices; we can improve our pro- Extreme Views ductivity practices; we can do better on maintaining biodiversity. But Nature in many parts of the world, people are squabbling over forests, taking Knows divisive polar “all-or-none” positions. So, one of our most important Best Markets Govt. tasks is to create common ground (Figure 3). I see this as one of the Know Knows biggest challenges that we face, in general, and specifically in the area Best Best of biotechnology. You, I am sure have encountered one or more of the extreme ideologies in your work. There are people who think that markets know best, and you just turn it over to the marketplace and Common Ground everything will be fine. Or, that scientists know the answers, let’s just ask the scientists and let them tell us what to do. Some people think that nature knows best, just leave nature alone and let it do its thing, Science Locals and on and on. The point here is that when you take these kinds of Knows Know Best Best fierce ideologies, and you take them to the extreme, they just rip apart the common ground. It leaves no room for people to come together I Know Best Laws are Clear and have a dialogue. Figure 3. The challenge of creating common ground amid As I think through these though, it strikes me that you can take extreme ideologies is crucial for the future of biotechnology. every one of these philosophical points and just turn its direction around (Figure 4). For example, if you know what you want to achieve, Constructive Views markets are a great way to get there. Everybody has some ideas; we Learn from have to have science to inform our choices so we know what the pos- Nature sibilities are as well as the consequences. The government does have a role; it’s to set the standards. And locals know a lot, and we can prob- Markets Govt. are a sets ably reach better policies if we involve them in the deliberations. If Means Standards we can take those ideologies and turn them into constructive views, we can probably rebuild the common ground that once existed, that Common we will certainly need to have to find prudent positions and choices Science Locals on how to go forward on the concept of sustainability, and what role Informs Ground Know Choices a Lot biotechnology and genetic engineering will play. So it is time to talk about all the consequences of our choices. Will genetically engineered trees be part of our path to sustainability? That is the question for this symposium. If not, then how are we Everyone Laws Give has Ideas Direction going to meet the needs of these billions and billions of people, and Figure 4. Ideologies can yield constructive views, which can not just their material needs? If yes, then what will be the rules of in turn help rebuild common ground and inform choices engagement? The choices are up to us, and there are consequences to that will guide us, not only in biotechnology and genetic every choice we could possibly make. engineering, but toward sustainability.

17 QUESTIONS FROM THE AUDIENCE

Toby Bradshaw, ‘terrorist target’

Hal, I’d like to ask you, you mentioned in your talk that you don’t see us returning to the forest that we had once upon a time. Could you elaborate on that? What do you mean? Do you mean that we’ll never have any forests that are like they were once upon a time, or that we just don’t have as much of that kind of forest as we once had?

Well, I think the answer to that depends on how precisely you want to define the forests as they were once upon a time. We will never have the conditions of forests we had ‘once upon a time’, because we won’t be returning to ‘once upon a time’. Our climate is changing, our population is growing, our technology is different, we have put all sorts of stuff into the water and the air that never existed once upon a time. So, we’re not going to go back and be able to overcome all of that and recreate a past that will never be again. But does that mean we can’t have diverse, productive, healthy forests, and lots of them? No, I think we can have that. We can even manage to perpetuate forests that are pretty close to the kind of natural diversity and function that they might have had if people weren’t around. We won’t be able to ever get there completely. It’s just amazing to me—this morning there’s an article in The Oregonian about the effects of air pollution coming out of Southern California on the national parks down there. We just simply aren’t going to be able to overcome that. So the forests that we have in the future are going to be impacted by human enterprise. We can do the best we can to reduce that human enterprise, the effects of that enterprise as best we can, in the places that we want to maintain as naturally as is possible to maintain them. We can put our energy into producing some of the things we need in as small an area as we possibly can. I think you’ll hear some good presentations on that this week. I’m not discouraged or pessimis- tic about future forests. The fact that we can’t go back and make things like they were two or three hundred years ago doesn’t trouble me.

Steve Strauss, Oregon State University

Hal, . . . I’ve heard comments from demonstrators that we’ve just messed with nature too much, and that biotechnology is obviously going to produce new kinds of genes that will enter the environment, genetic contamination, as it were. So let’s just stop doing it. Let’s just not mess with things anymore. What would you say to someone who has that point of view?

Oh, boy. At what point do you want to start reversing the clock? Do you want to start with the things that we’ve done with humans, with crops? I don’t know how to have a dialogue with somebody who wants to stop the world while they get off. The world keeps going forward. My biggest concern on genetically modified organisms is that if it’s not done in places where we can have an open, scientific, democratic process of figuring out how to set the rules of engagement, then we are going to have to live with the consequences of somebody doing it somewhere else where they didn’t have the benefit of that open democratic process. I would prefer that we didn’t have to have the situation where we put this 18 much intensive intrusion into nature, but I think the human population hasn’t give us much option.

Hal Salwasser, final comments

You know, I want to make a comment here. I hope you all are taking a good lesson from Steve Strauss and Toby Bradshaw. They’ve got me giving a keynote talk and moderating a panel and then sharing perspectives at the end of the first two-day session. Now, I’ve got to figure that what they had in mind is that they want their Dean to understand what they are doing, and this is a way to absolutely guarantee that he keeps paying attention; it’s really sneaky. I am a wildlife biologist by background, by the way, who focused on forest wildlife habitat. I don’t know much, or I didn’t know much about this area. I know a lot more now than I did a few weeks ago. I am really pleased that Steve and Toby have set me up for paying attention here, I expect to become relatively knowledgeable on this topic and understand what some of our faculty are up to. Good strategy! Thank you very much for your attention, and I look forward to the next couple of days.

19 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 20-22. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Context and Goals for Ecosocial Symposium

Steve Strauss

t is my pleasure to begin this Symposium on Ecological and Societal Aspects of Transgenic Plantations, a part of the 10th international meeting of the In- I ternational Union of Forestry Research Organizations, better known as IUFRO, Section on Molecular Biology of Forest Trees. The first IUFRO molecular biology meeting was organized by Howard Kriebel, a true pioneer on DNA studies of trees, in Ohio, USA, in 1985. Subsequent meetings of this group have been noteworthy for their high quality science and lovely venues, which have included Ontario, Canada; Lappland, Sweden; Lake Tahoe, California; Bordeaux, France; Scarborough, Maine, USA; Gent, Belgium, Quebec City, Canada; and Oxford, England. It appears from the registration list that the trend of increasing attendance at these IUFRO meetings is continuing, and in fact may be accelerating. With more than 230 registered, representing 17 different countries, the attendance at this meeting substantially exceeds the number at the previous meeting (approximately 170). Despite the world controversy on plant biotechnology, and perhaps in part because of the world controversy, scientific interest in forest biotechnology continues to increase. The large professional and media turnout at our meeting demonstrates that there is great scientific and social interest in forest biotechnology. From a scientist’s view, particularly considering the obscurity of our field just a few years ago, this might seem surprising and inappropriate. However, on deeper reflection I think you will find it highly appropriate. Human populations and resource consumption continue to grow, generating an increasing demand for wood and for the many other products and services of forests. Stresses on forests from humans, direct and indirect, also continue to mount. The world is searching for ways in which to both conserve and protect forests, while providing for a growing stream of forest products, with as little ecological disruption as possible. It is no surprise that biotechnology, with its scientific depth and technological novelty, is viewed as holding considerable promise for sustainable resource production. But what kinds of biotechnologies do we want, and how should they be developed and agreed upon? These are difficult and complex questions, often with political, moral, and ethical dimensions beyond the reach of science. Genomes, including those of trees, are being discovered and studied in detail for the first time in history. And methods for direct use of that knowledge, both via genetic engineering and DNA markers, have been developed that make it possible to act on this knowledge in the near term, and thus influence the genetic composition and management of forests. It should be no surprise that those with concerns about forest biotechnology are alarmed. Our field has deep and broad scientific power, and the potential to apply it. Many who do not understand the scientific issues fully, or do not trust the social institutions that regulate the science and technology, are apprehensive or even frightened. We need to accept this, and to honestly and openly provide reliable information and accurate research results in order to help society make its choices. 20 One important aspect of this meeting—technical considerations and conclusions aside—is that this very endeavor demonstrates a sincere attempt on the part of our scientific community to analyze, in an integrated, diverse, and open manner, the consequences of the technology we are developing. Toby and I strongly believe that it is the ethical responsibility of this community to do this, and we hope that the media, and even those opposed to biotechnology in forestry, even if they do not agree with our views, at least recognize the sincere effort we are making toward this end. This is not a community with its head stuck in the sand. It is a community that is reaching out and reflecting in the finest tradition of scientific reason and skepticism. Toby and I chose the speakers because they possess diverse but thoughtful views. For many we know little more than this about their perspectives on forest biotechnology. I think you will see that this is not a highly pre-selected group chosen for a robotic love of forest biotechnology. It is a vigorous, wide-ranging debate that we seek. This meeting focuses on biological science, with economics, business, and ethics as frameworks to help understand the motivation and context of the science and technology. We could have an entire meeting devoted to social and cultural issues of GM trees, however, that is not our current goal. In the spirit of good science, we therefore urge participants to avoid discussing GM trees as though they constituted a vague or generic set of concerns. The benefits and risks depend on the genes, how they are modified, the method of gene transfer, the intensity of research and safety evaluation, and the social as well as ecological context in which they might be deployed. There are myriad details to consider in biotechnology and they all matter. Thus, we ask that both speakers and the audience be as specific as possible, and whenever possible explain what benefits or risks you see for specific kinds of GM trees, and why. Social views about resources, the roles of humans, ethical behaviors, and what constitutes sustainable development vary widely. The only certainty appears to be that the technological options that result from the rapidly growing science of biotechnology will neither be simple nor without controversy. As a biotechnology scientist, I find it hard to imagine a more exciting or challenging time to be living in.

INTRODUCTION TO PRESIDENT RISSER

To help set the stage for this conference, I would like to introduce Dr. Paul Risser, an internationally known ecologist and the President of Oregon State University. Dr. Risser was appointed the 13th president of OSU in 1996. He was awarded his PhD in 1967 from the University of Wisconsin, and has served in diverse faculty and administrative positions during his career. Dr. Risser’s professional interests include grassland and forest ecosystems, environmental planning and management, landscape ecology, and global change. He has led several multi-institutional and international scientific studies, and wrote and edited several books and over 90 invited chapters and scientific papers for refereed journals. He served as director of ecosystem studies at the U.S. National Science Foundation in 1975–1976, and is a fellow of the American Association for the Advancement of Science. He is a past president of the Ecological Society 21 of America and the American Institute of Biological Sciences, and has consulted for the National Academy of Sciences, the Smithsonian Institution, the U.S. En- vironmental Protection Agency, and many other public and private organizations. In Oregon, he has been appointed by the governor to serve as chair of the Sci- ence Panel for Oregon’s Environmental Stewardship Plan, and he chairs the Willamette River Restoration Initiative Board. Dr. Risser’s deep and broad background in ecology and biological sciences, including environmental policy issues, makes him uniquely well suited to launch this symposium on ecological and societal dimensions of transgenic forestry.

22 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 23-24. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Welcome to the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, in Conjunction with the IUFRO Conference on Tree Biotechnology in the New Millennium

Paul G. Risser President, Oregon State University

or decades, breeding systems have been used to change, relatively slowly, the characteristics of plants and animals. Today we are rapidly learning F much more about genes and genomes. Based on this new information, we now have the technology to more directly modify the properties of plants and animals. This technological revolution is already underway, most strikingly in medicine, where an increasing number of pharmaceuticals are produced in genetically engineered organisms. Similarly, many different genetically modified crops have been produced or are in production, others are awaiting commercial acceptance, and still more are in the stage of advanced development. These manipulations have led to significant benefits in health care, food production, and environmental protection. They are not without controversy, however, as we all know. From a strictly scientific viewpoint, the question is not whether these technologies are feasible, but rather, in what ways and under what conditions can they benefit humans and the environment? And at what point in the accumulation of this extraordinary new knowledge are we confident enough to go ahead with the technology, and to do so with a high probability of net social and environmental good? Resolving these issues will never be easy because they confront the most fundamental of our beliefs and values. They intersect with human health, ecological integrity, privacy, assumption of risk, democratic process, and economic well being. And because there are concerns about threats to our very basic human values, people will disagree about these technologies, sometimes very strongly. Over the centuries, technologies have caused controversy, but this biotechnology is much more poignant because it seems more powerful and it affects the very essence of plants and animals. This conference has been organized because the controversy has started to play itself out in forestry. Here it is complicated because in several ways, forestry itself can be controversial. We frequently use the words “forest” or “forestry” in imprecise ways. In some applications, we practice forestry with great intensity, where trees are clearly parts of wood farms—crops to be tended with one undisputed dominant product. In other places we manage forests with diverse ecological and social products in mind, of which wood may be only a modest output. In still other places, we harvest trees solely for the purpose of ecological management, if we harvest them at all. Depending on the state of genetic knowledge and potential uses, biotechnology will fit in very different places along this management spectrum. At least for the 23 immediate future, this genetic technology is likely to be confined to intensive plantation systems. These intensively managed systems occupy an extremely small portion of the world’s forestry and agricultural landscape, yet they can produce enough wood to satisfy a disproportionate amount of the world’s need. At least theoretically, high productivity from these plantations can relieve pressure for harvesting wild forests. In addition, with proper management, some of the new plant traits could have environmental as well as production benefits within plantation systems. Consideration of the use of transgenic trees is focused on these plantation forests. Despite the relatively small area of these forests, concerns about the use of transgenic trees have been voiced from many quarters. Some raise concerns because they see a real threat to all wild forests; others fear a slippery technology slope from which there is no realistic return. Some of the strongest voices of concern are from genetic practitioners themselves—who, for more than a decade, have called for strict measures to mitigate ecological risks. I am proud that an Oregon State University scientist, and the co-convener of this conference, Dr. Steve Strauss, has been at the forefront internationally in stimulating an open and vigorous debate. He and Toby Bradshaw have now put together this absolutely first-rate international symposium, the first of its kind in forestry. As speakers and participants, you will examine the social context and the ecological safety of genetically engineered forest species from virtually all major perspectives. As I conclude these welcoming remarks, let me speak as a university president. Among the most significant values of great universities is the ability to bring together the best minds from many disciplines, and to focus these intellectual abilities on complex topics of particular importance to society. This intellectual pursuit must be accompanied by great attention to the ethical dimensions of the issue, and must encompass multiple perspectives supported by careful and thoughtful analyses. Moreover, these deliberations must be tested by peers and communi- cated to interested and affected constituencies. Your challenge here will require rising to the standards and expectations of great universities. You must consider both the potential benefits and risks to our forests from this technology, you must do so in the context of a world that is growing hungrier for resources, and you must consider our shared responsibility for the health of our biosphere. You will learn from each other, you will integrate ideas and information, and with a little luck, you will be able to forge a collective vision for moving forward on the most productive research agenda, and for constructing guidelines for the application of this technology. If it can be done, this carefully constructed conference, with the best experts from around the world, will certainly be successful. Please accept my best wishes. Thank you.

24 Moderator Introductions

Steve Strauss

would now like to introduce our competent moderators and then turn the symposium over to their care. I Last night you heard about the background of Dr. Hal Salwasser, Dean of the College of Forestry at Oregon State University and our keynote lecturer. There- fore, briefly, let me remind you that before coming to OSU, Dr. Salwasser was a Regional Forester and Research Station Director for the U.S. Forest Service, the Boone and Crockett Chair of Wildlife and Conservation at the University of Montana, and has been an active member of the Society of American Foresters, the Society for Conservation Biology, and the Wildlife Society. Dr. Salwasser brings a broad forest policy and ecological perspective to the symposium. Dr. Clegg is a population geneticist who got his BS and PhD degrees from the University of California at Davis, working there under the eminent geneticist Dr. Robert Allard. He was a Professor at Brown University, the University of Georgia, and is presently a Distinguished Professor of Genetics at the University of California at Riverside—where he also served as Dean for six years. Dr. Clegg has published over 130 peer-reviewed publications, all at the cutting edge of population and molecular genetics. He is a past president of the American Genetic Association and the Society for Molecular Biology and Evolution, and was elected to the National Academy of Sciences in 1990—where he has participated on a number of committees and boards, including as chairman, from 1992 to 1995, of the National Research Council Committee on Scientific Issues in the Endan- gered Species Act. Dr. Clegg has also had editorial responsibilities for seven top journals in biology and population genetics. Dr. Clegg brings a strong plant evolutionary and molecular genetic perspective to the symposium.

POLICY AND ECONOMICS

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 29-46. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf The Economic Contribution of Biotechnology and Forest Plantations in Global Wood Supply and Forest Conservation

Roger A. Sedjo

ABSTRACT

Over the past 30 years industrial plantation forests have become a major supplier of industrial wood. The reasons for this change are several and include the improved economics of planted forests due to technological innovations, the increases in natural forest wood costs due to increasing inaccessibility and rising wood costs from natural forests due to various pressures from environmentalists to reduce harvesting in old-growth forests. Forestry today is on the threshold of the widespread introduction of biotechnology into its operational practices in the form of sophisticated tissue cultures, which produce clonal seedlings, and through the use of genetically modified organisms (GMOs), which produce desired tree and wood traits. As more of the world’s industrial wood is being produced on planted forests, the potential to introduce genetic alterations into the germ plasm utilized in planting is obvious. In many cases the biotechnology likely to be introduced in forestry is simply an extension of that being utilized in agriculture, e.g., herbicide-tolerant genes. However, biotechnology in forestry is also developing applications unique to forestry, e.g., genes for fiber modification, lignin reduction and extraction, and to promote straight stems and reduced branching. This paper discusses the growing role of plantation forests and the potential impacts of biotechnology on forestry. Traditional breeding and some aspects of biotechnology are discussed briefly and some of the various types of biotechnological innovations in progress in forestry and that may be forthcoming over the next decade or two are identified. A quantitative estimate is made of the potential economic impact of one transgenic application—that of the herbicide-resistant gene in forestry—and some of the potential environmental benefits associated with various types of biotechnology innovations are discussed. The potential benefits from the introduction of biotechnology to forestry promise to be large. For example, the widespread use of the herbicide-resistant gene for planted forest establishment is estimated to have potential cost-savings approaching $1 billion annually. The economic benefits will be found in the form of lower costs and increased long-term availability to consumers of wood and wood products. Roger A. Sedjo is Additionally, there is the potential for substantial environmental benefits from bio- Senior Fellow and direc- technology in forestry. An environmental implication of the increased productivity of planted forests due to biotechnology is likely to be that large areas of natural forest might tor of the Forest Econom- be free from pressures to produce industrial wood, thereby being better able to provide ics and Policy Program, biodiversity habitat. The shift away from harvesting natural forests to alternative plantation wood sources is already well underway. Also, other environmental benefits from for- Resources for the Future, est biotechnology are likely. Through biotechnology, trees can be modified so as to allow Washington, DC. them to grow in previously unsuited areas, e.g., arid and saline areas. This characteristic could not only increase wood outputs, but might be appropriate for promoting increased carbon sequestration, which could contribute to the mitigation of the global warming [email protected] problem, or through the provision of other environmental functions, such as enhanced watershed protection. Additionally, biotechnological innovation can be used in the resto- 29 ration and rehabilitation of badly disturbed species or habitats. For example, biotechnology gives promise for the restoration of the almost extinct American chestnut tree in the United States. Finally, biotechnology gives promise of providing enhanced potential for carbon sequestration from more rapidly growing planted forests. This could result in a greater contribution of forest sinks to addressing the global warming problem. The health, safety, environmental and ownership dimensions of biotechnology sometimes raise concerns, although in forestry, these differ in some important ways from the concerns of agriculture. Wood is rarely ingested directly by humans and thus, food health or safety is generally not an issue, although cellulose is sometimes used as a “filler” in food products. Also, ownership and property rights issues related to biotechnological innovations appear to be more tractable in the longer harvest rotation of forestry than in typical seasonal agriculture. In most cases the concerns associated with forestry related to the possibility of genetic escape from transgenic to wild trees. Although many of these risks appear to be negligible, transgenic trees that involve significant risks could be avoided while society still provides for the introduction of negligible risk plant genetic alterations.

ver the past 30 years industrial plantation forests have become a major supplier of industrial wood, gradually displacing wood from natural O forests. The reasons for this change include the improved economics of planted forests due to technological innovations, relative increases in wood costs from natural forests due to rising extraction costs, and pressures by environmental activists to reduce harvesting in old-growth forests. Forestry is currently undergoing an important transition from a wild resource, which had typically been foraged, to a planted agricultural crop, which is harvested periodically, as are other agricultural commodities—only the time scale for forestry is longer. The transition of forestry from foraging to an agricultural cropping mode has been underway on a significant scale only within the past half century or less (Sedjo 1999). Planted forests benefit from the same types of innovations that are common in other agriculture. As with other agriculture, economic incentives for investments in plant domestication, breeding and plant improvement activities will occur when the investor can capture the benefits of the improvements and innovations. As in other types of agriculture, early plant improvements involved identification of trees with desired traits and attempts to capture offspring that had the desired traits through the identification of superior trees. In recent decades traditional breeding techniques have been practiced in forestry as they have been in other agriculture. In the 1990s, however, modern biotechnology, including tissue culture, began to be undertaken in earnest in forestry. Additionally, a relatively large number (124) of confined traits of transgenic trees have been undertaken in the U.S., but only one transgenic tree species (papaya) has been authorized for release (McLean and Charest 2000). The benefits from the introduction of biotechnology to forestry have the potential to be large. The economic benefits will be found in the form of lower costs and increased availability to consumers of wood and wood products. Addi- tionally, biotechnological innovation has the potential to beneficially address a number of important environmental issues. Biotechnology can be used in the rehabilitation of habitats under pressure either from an exotic disease, as with the American chestnut tree (Castenea dentate) in the United States (Bailey 1997), or 30 from invasive exotics. Additionally, an This paper is organized as follows. Recent decades have seen continu- implication of the increased productiv- The general introduction of plantation ing increases in biological productivity, ity of planted forests due to biotech- forestry biotechnology is followed by especially in agriculture. This has been nology may be that large areas of natu- a discussion of the application of tra- driven largely by technological innova- ral forest might be free from pressures dition breeding and modern biotech- tions that have generated continuous to produce industrial wood, perhaps nology to tree improvements. The next improvements in the genetics of prima- thereby being better able to provide section presents a broad overview of rily domesticated plants and animals. biodiversity habitat. Also, through bio- the application of traditional breeding Much of this improvement has been technological improvements trees can and modern biotechnology, including the result of plant improvements that be modified so as to allow them to genetic modification, to trees. The sec- have been accomplished by traditional grow in previously unsuited areas, e.g., tion also discusses the various types of breeding techniques through which arid lands, saline areas and so forth, biotechnological innovations in for- desired characteristics of plants and thereby providing missing environmen- estry that could be forthcoming in the animals, e.g., growth rates or disease tal functions, such as watershed protec- next decade or so. The third section resistance, can be incorporated into the tion. Such uses could not only increase undertakes a case study that estimates cultivated varieties of the species in wood outputs, but might be appropri- the potential benefits associated with question. ate for promoting increased carbon se- the use of a herbicide resistant gene in Changes driven by technology, questration in forest sinks and thereby forestry and discusses broadly the types however, are not new. Hayami and contributing to the mitigation of the of potential economic benefits that so- Ruttan (1985) have pointed out that global warming problem (IPCC 2001). ciety could realize from biotechnology. in the United States, most of the in- The ownership and environmental This is followed by a discussion of po- creased agricultural production that dimensions of biotechnology in forestry tential environmental benefits and an- occurred in the two centuries before differ in some ways from agriculture and other section on concerns associated 1930 was the result of increases in the so raise somewhat different questions. with biotechnology. Finally, the paper amount of land placed in agriculture, Ownership and property rights issues presents a summary of the implications and most of the increased production related to biotechnological innovations of biotechnology to forestry. reflected increased inputs in the form appear to be more tractable in the longer of labor saving technology—either ani- harvest rotation of forestry than in typi- mal or mechanical. In Japan, however, cal seasonal agriculture. This is because OVERVIEW where land was limited, substantial it usually takes several years before a tree improvements in rice productivity were will flower and the seed is available; by The domestication of a small made by careful selection of superior, that time the seed technology may have number of plants, particularly wheat, yield-increasing seed. Land productiv- become obsolete. On the environmen- rice, and maize, is among the most sig- ity in grain production in the United tal side, unlike most agriculture there are nificant accomplishments in the hu- States showed little increase until the few major concerns for direct health or man era. Modern civilization would be 1930s, as most of the gains in produc- safety from the consumption of geneti- impossible without this innovation. tion were due to innovations that al- cally modified wood products, although Common features associated with plant lowed more land to come into produc- cellulose is sometimes used as filler in domestication include high yields, large tion, e.g., new equipment and mecha- food products. There are, however, con- seeds, soft seed coats, non-shattering nization. By contrast, land productiv- cerns related to genetic transfers that seed heads that prevent seed dispersal ity in Japan was a function of biotech- might occur between transgenic and and thus facilitate harvesting, and a nological improvements in the form of wild trees, and the potential implications flowering time that is determined by improved seed and increased yields. for the natural environment. planting date rather than by natural However, in the United States after the day length (Bradshaw 1999). 1930s, when most of the highly pro-

31 ductive agricultural land was in the dertaken in forestry in earnest. As more forest were often established on grass- U.S., the focus of innovation was re- of the world’s industrial wood is being lands. directed to plant improvement, which produced on planted forests, the poten- It was soon recognized that if the increased land productivity through tial to introduce genetic alterations into costs of planting were to be under- higher yields. Until fairly recently these the germ plasma utilized in planting is taken, the effect would be enhanced to improvements were achieved through obvious. Commercial forestry today is the extent that improved seed or tree the use of traditional plant breeding on the threshold of the widespread in- seedlings could be used. Thus, the de- techniques, which gradually increased troduction of biotechnology in the cision to plant also provided incentive agricultural yields. form of sophisticated tissue cultures for for tree improvement. Initially, tree cloning seedlings, and in the form of improvement was accomplished genetically modified organisms. through traditional breeding tech- Plantation Forestry Early tree planting activities typi- niques. cally consisted of replanting seedlings Planted forests for timber began after timber harvest. Factors important in earnest in the 19th century in Eu- in the decision to replant included The Effects of rope and about the middle of the 20th property rights—so that those who Plantation Forests century in North America. Over the bore the costs of replanting would be past 30 years industrial plantation for- able to capture the benefits of the fu- Figure 1 provides a simple sche- ests have become a major supplier of ture harvest—and protection capacity, matic that illustrates the effects asso- industrial wood. The reasons for this which helped ensure that the tree crop ciated with the lowering of costs pro- change are several. These include the would not be destroyed prematurely by vided by planted forests. In the ab- improved economics of planted forests pest or fire. It is not a coincidence that sence of forest plantations the volume vis a vis natural forests, due in large widespread tree planting occurred only of industrial wood harvested in a pe- part to technological innovations that after forest control had reduced sub- riod is determined by the intersection increased planted forest productivity as stantially the incidence of forest wild- of supply, S. and demand, D, at eo. well as to the relative increases in wood fire (Sedjo 1991). Much of the early In this situation price is Po and the costs from natural forests due to ris- planting in the United States took quantity harvested is Qo. The intro- ing extraction costs and pressures by place on lands that once had been duction of relatively low cost planta- environmental activists to provide naturally forested; but in more recent tion forestry is represented by the line more stringent harvesting standards decades, it has occurred on land that segment aS’. At price P1 plantations thereby reducing harvesting in old- had previously been used for agricul- provide cheaper source of industrial growth forests. ture. In the South, for In recent decades traditional example, such land breeding techniques have been prac- had often been in cot- P S ticed in forestry as they were in other ton or tobacco. A e0 agriculture. Early improvements in similar phenomenon P0 trees involved identification of “supe- a e1 was seen in newly es- P1 S' rior” trees with desired traits and at- tablished planted for- tempts to capture offspring having the ests overseas. In New desired traits. The planting of geneti- Zealand, forests were D cally improved stock began about planted on sheep pas- 1970. In the 1990s, modern biotech- 0 Q ' Q Q Q ture, in Chile, on mar- 1 0 1 nology, including tissue culture and ginal grain lands, in genetic modification, began to be un- Argentina and Brazil, Figure 1. Industrial wood.

32 wood than do natural forests. This Forest management new source of timber results in a new surely began in part of Table 2. Global harvests by forest management condition, equilibrium, e1, with a lower price, P1 the world more than circa 1995. and a higher harvest volume, Q1. 2000 years ago. For ex- Percent of Global Notice, however, that the volume har- ample, written manage- Forest Situation Harvest Industrial Wood vested from natural forests in reduced ment directives appeared Old-growth 30 from Q0 to Q1'. This reflects that fact in China as early as 100 Second-growth, minimal management 14 that the low-cost plantation wood is BC (Menzies 1985). Indigenous second-growth, managed 22 displacing wood from natural forests. However, significant ar- The effects of biotechnology are to eas of managed forest Industrial plantations, indigenous 24 future reduce the costs of production probably were not com- Industrial plantations, exotic 10 thereby shifting down even further the mon in Europe until the Source: Sedjo 1999. aS’ portion of the supply curve (not Middle Ages. Planted Notes: Old-growth includes Canada, Russia, Indonesia/Malaysia. shown in Figure 1). forests began in earnest Second-growth, minimal management includes parts of the in the 19th century in U.S. and Canada, Russia. Europe, but not until Indigenous second growth, managed: residual. Industrial plantations, indigenous: Nordic, most of Europe, Impacts of the middle of the 20th a large but minor portion of U.S., Japan, and some from century in North Biotechnologically China and India. Induced Changes in America. The planting Forestry of genetically superior natural forests, become increasingly at- stock began about 1970, and the seri- tractive as an investment for produc- ous planting of genetically modified Currently, most of the world’s ining future industrial wood. The plan- trees is just now beginning in parts of dustrial wood is drawn from natural tation manager can control some of the the subtropics, such as New Zealand forests in what is essentially a foraging important variables, such as choosing and South America. operation. In the past harvests occurred a location for the planted forest and the As Table 2 indicates, even today a from forests created by nature as hu- species. Former agricultural sites often large portion of the world’s industrial mans simply collected the bounty of are desirable locations for planted for- wood supply originates in natural, non- nature. Table 1 indicates how this pro- ests because they are usually accessible managed forests. In recent decades, cess has changed over time as humans and reasonably flat, thereby lending however, the widespread introduction gradually developed silvicultural tech- themselves to both planting and har- of tree planting worldwide for indus- nology. vesting. Often, acceptable access exists trial wood production has via the former agricultural transport in- Table 1. Transitions in forest management and harvests. resulted in most of the in- frastructure. The planted forest can also creases in global harvests be- be located in proximity to important Type Period ing drawn from planted for- markets. Within limits, the manager Wild forests 10,000 BC - present ests. can choose a species appropriate to the Managed forests 100 BC - present The potential of the site, which may also have good mar- Planted forests 1800 - present widespread introduction of ket access and a reasonably short har- genetically improved trees Planted, intensively managed 1960 - present vest rotation. can have important environ- The economic advantages of Planted, superior trees, mental and economic ef- traditional breeding techniques 1970 - present planted forests have led to their wide- fects. With increasing yields spread adoption in a number of regions Planted, superior trees, and shortened rotations, genetic modification 2000 - future throughout the globe; they are having planted forests, rather than an important influence on global tim-

33 ber supply. Over time, a greater share of the forest. In the Table 3. Gains in loblolly pine from various traditional breeding of the world’s industrial wood supply past, operational approaches. has been and will be coming from quantities of seed planted forests. Planted forests today from production Technique Increase in yields (%) account for most of the increased glo- seed orchards were Orchard mix, open pollination, first generation 8 bal output and their production is re- derived from open Family block, best mothers 11 placing the timber formerly provided pollination. Today, Mass pollination (control for both male and female) 21 by native and old-growth forests that however, more so- Source: Personal communication with researchers, Westvaco Corpora- are no longer available for harvest due phisticated large- tion, Summerville, SC. to political changes (e.g., Russia) or scale, controlled-pol- policy changes (e.g., within the U.S. lination techniques are in place that agricultural products, tree hybrids are National Forest System). offer the potential of further improve- often a means to improve growth and ment of the offspring of two superior other desired characteristics. Hybridiza- parents. tion crosses trees that are unlikely to TRADITIONAL The results of traditional breeding breed in nature, often where parents do approaches to improve tree yields are BREEDING not occur together in sympatric popu- instructive to illustrate the possibilities lations. These crosses often exhibit of traditional breeding (Table 3). For growth and other characteristics that most tree species, the typical approach neither of the parent species alone can Selection involves the selection of superior trees for match. In the United States, for example, establishment in seed orchards. Experi- several hybrid poplars have shown re- ence has shown that an orchard mix of Tree improvement most often has markable growth rates, which exceed first-generation, open-pollinated seed relied on traditional breeding tech- those found in parent populations.1 The can be expected to generate an 8% per niques like selection of superior (plus same is true for the Eucalyptus grandis generation improvement in the desired candidate) trees for volume and stem and urophylla hybrids in many parts of characteristic, e.g., yield. More sophisti- straightness, and grafting these into the tropics and subtropics. cated seed collection and deployment breeding orchards and producing seed techniques, such as collecting seed from orchards. When breeding orchards be- the best mothers (family block), can re- gin to flower, pollination of selections BIOTECHNOLOGY sult in an 11% increase in yield, while is artificially controlled, seeds are col- mass-controlled pollination techniques, lected, progeny tests are established, Biotechnologies used in forestry fall which control for both male and female and the best offspring are chosen for into three main areas: the use of vegeta- genes (full sibling), have increased yield the next cycle of breeding. By identi- tive reproduction methods, the use of up to 21%. fying and selecting for desired traits, genetic markers, and the production of breeding can select for a set of traits genetically modified organisms (GMOs) that can improve wood and fiber char- Hybridization or transgenic trees. Most of the biotech- acteristics, improve the form of the nologies used in forestry today are in the tree, provide other desired characteris- A variant of the traditional breed- category of tissue culture and molecular tics, and improve growth. These traits ing techniques is that of hybridization, marker applications (Yanchuk 2001). are introduced into the genetic base which has provided robust offspring by that is used for a planted forest. This bringing together populations that do contributes to the more efficient pro- not normally mix in nature. This ap- 1 Growth in hybrid poplar stands is 5-10 times the duction of industrial wood and to an proach is widely used in forestry. As in rate of native forest (Toby Bradshaw, University of improved quality of the wood output Washington, personal communication). 34 Cloning and Vegetative modest, and therefore may be finan- species, however, the process is more Reproduction cially justified. Additionally, because difficult, as simple vegetative propaga- the clone provides the vehicle through tion does not normally occur or occurs Vegetative reproduction comprises which desired foreign or artificial genes only infrequently. Here, “tissue culture” a broad range of techniques involving are transferred, cloning techniques techniques provide the tools to quickly the manipulation of plant tissue that must be developed in order for genetic produce genetically engineered plants ultimately allows for vegetative repro- engineering in forestry to be viable. and clones to regenerate trees with de- duction of the whole plant. Tissue cul- The ability to use inexpensive clon- sired traits (Westvaco 1996, pp. 8–9). ture broadly refers to clonal techniques ing techniques varies with species and of growing plant tissue or parts in a genus. For some species, typically hard- nutrient medium containing minerals, woods, cloning can be as simple as us- Genetic Markers sugars, vitamins, and plant hormones ing the vegetative propagation properties Genetic markers are used to try to under sterile conditions. However, for inherent in the species to accomplish the find a relationship between the mark- some tree species, cloning approaches genetic replication. This might involve ers and certain characteristics of the have been limited thus far (Pullman et simply taking a portion of a small branch tree. A major approach to genetic ma- al. 1998). In general, there has been from a desired superior tree and putting nipulation of trees utilizes molecular greater success cloning hardwoods, e.g., it into the ground, where it will quickly biology. Molecular biology has two fac- poplar and some species of eucalyptus, take root (rooted cuttings). Where veg- ets. The first facet is that which may than conifers. etative propagation is part of the natu- aid the efficiency of traditional breed- The development of cloning tech- ral process, large amounts of “clonal” ing programs. One problem with tra- niques in forestry is important for a material can be propagated via rooted ditional approaches in tree breeding is number of reasons. First, if superior cuttings, the cuttings of which come the long growth cycles generally re- trees are available, an approach must from “hedge beds.” Here the process quired by trees, which make this probe developed to allow for the propa- continues until sufficient volumes of veg- cess very time consuming. Techniques gation of large numbers of seedlings etative materials with the desired genes such as molecular biology and molecu- with the desired characteristics if these are available to meet the planting re- lar markers, which identify areas on the traits are to be transferred into a quirements. chromosome where genes that control planted forest. With tree planting of- Eucalyptus, poplar, and acacia the desired traits occur, can accelerate ten involving more than 500 seedlings tend to be effective propagators. Other the process and enhance the produc- per acre,2 large-scale planting of im- genera propagate less readily. Many tivity of the traditional approach. The proved stock would require some species in the pine family, e.g., loblolly, second facet is where specific genes are method of generating literally millions and to a lesser extent, slash pine, are identified and modified to affect bio- of genetically upgraded seedlings at a difficult propagators. Radiata pine, chemical pathways and the resulting relatively low cost. The costs of the common in plantations in New phenotypes. For example, lignin genes improved seedlings are important, since Zealand and Chile, appears to have the can alter the amount, type, and form the benefits of improved genetics are best record on this account. Propaga- of lignin that is produced. delayed until the harvest. With harvests tion improves when certain procedures In recent years, molecular ap- often occurring 20 years or more after are undertaken. For example, using the proaches to tree selection and breeding planting, large costs for improved seed shoots emerging from newly trimmed have shown significant promise. The may seem difficult to justify financially. clonal hedges increases the probability molecular approach, although limited However, if the costs of plantings are of successful regeneration. For many in application by its expense, involves going to be incurred, the incremental genetic material being identified, col- costs associated with planting improved 2 It is estimated that 4 to 5 million trees are lected, bred, and tested over a wide genetic stock are likely to be quite planted in the U.S. every day. range of sites. Rather than simply 35 choosing specific tree phenotypes on have more than doubled those of ear- achievable through traditional tree the basis of their outward appearance, lier plantings. breeding. It also allows species to have the molecular approach identifies the attributes that would not be possible areas of the chromosomes that are as- through natural processes. For ex- sociated with the desired traits. “Mark- Genetically Modified ample, in concept, frost-resistant genes ers” are used to identify the relative Organisms (GMOs) could be transferred from plants or position of genes on the chromosome other organisms found in cold north- that control expression of a trait. This The term biotechnology is often erly regions to tropical plants, thereby approach exploits the genetic variation, associated with generic transformations increasing their ability to survive in which is often abundant, found in as it involves the introduction of se- cooler climates. natural populations. Molecular mark- lected foreign genes into the plant ge- These attributes or traits can be ers and screening techniques can be nome. In this approach, specific genes characterized as silvicultural, adaptabil- used to examine the DNA of thou- are identified and modified to affect ity, and wood quality (Table 4). Silvi- sands of individual trees to identify the biochemical pathways and the result- cultural traits would include growth few, perhaps less than a dozen, with the ing phenotypes. Thus far, transgenic rate, nutrient uptake, crown and stem optimal mix of genes for the desired trees have not been used commercially form, plant reproduction (flowering), outputs. These techniques are currently for wood production (McLean and and herbicide tolerance. Growth po- being applied to the development of Charest 2000). However, the promise tential, for example, has a substantial improved poplar in the United States is substantial, as has been demonstrated genetic component, with rates differ- and eucalyptus in Brazil.3 in agriculture. Potential applications ing by 50% between families or differ- Recent work on hybrid poplar in include herbicide-resistant genes, pest- ent clonal lines. Traditional breeding the Pacific Northwest has shown a resistant genes (Bt), and genetic alter- approaches are steadily improving elite- 20% increase in yields in plantations ation that would provide certain de- line yield potentials. A subset of these and an additional 20% on dry sites sired wood characteristics—e.g., the traits is found in Table 5. These traits where irrigation can be applied (east of promise of controlling the lignin in include those that are most likely to use the Cascade Mountains).4 Growth rates trees is dependent on the ability to biotechnology for further commercial with these plantations are impressive. identify and modify lignin genes, development. The first three traits of Yields are about 7 tons per acre, or thereby altering the amount, type, and the list in Table 5 are traits that, in the about 50 cubic meters per hectare and form of lignin that is produced in the judgment of many experts, could be improvements in the yield continue.5 tree (Hu et al. 1999). As noted, the featured prominently in biotechnologi- These growth rates are approximately ease of gene introduction (transforma- cal innovations in forestry over the next three times the growth rates of typical tion) varies with different tree species decade. pine plantations in the southern and genus, and is generally more diffi- Planted trees typically require her- United States. Elsewhere in the world, cult in conifers than in hardwoods. bicide and, in some cases, pesticide ap- for example, Aracruz in Brazil, yields plications for one or two years after of hybrid eucalyptus are reported to planting. The introduction of a herbi- FUTURE cide-resistant gene can reduce the costs 3 of herbicide applications by allowing Toby Bradshaw, Director of the Poplar IOTECHNOLOGICAL Molecular Genetic Cooperative at the University B fewer, but more effective applications of Washington, Seattle, personal communication. without concern over damage to the Also see Westvaco 1997. INNOVATIONS IN seedlings. The use of a pest-resistant gene 4 Toby Bradshaw, University of Washington, FORESTRY can eliminate the requirement to apply personal communication. the pesticide altogether. Flowering con- 5 Withrow-Robinson et al. (1995), p 13. Gene alteration can result in trol allows a delay of several years in unique gene combinations that are not 36 ging pressure on natural forests, Table 4. Forest traits that can be improved through biotechnology. thereby reducing pressures on certain Silviculture Adaptability Wood quality traits biodiversity and habitat (Sedjo and Growth rate Drought tolerance Wood density Botkin 1997). Modified tree species Nutrient uptake Cold tolerance Lignin reduction also show promise of being useful in Crown/stem Fungal resistance Lignin extraction providing environmental services in Flowering control Insect resistance Juvenile fiber areas where trees now may have diffi- Herbicide Branching culty surviving—for example, in arid Source: Context Consulting provided information on potential innovations and their likely cost implication or drought-prone areas, areas with sa- based on the best judgment of a panel of experts. line conditions or frost zones. Also, given the potential of biological sinks priority. Trials with low lignin trees as a tool to mitigate the build-up of Table 5. Traits of interest in forestry. have already been undertaken in greenhouse gases associated with global warming, the ability to establish • Herbicide tolerance Aracruz Cellulose in Brazil (Claes Hall, carbon sequestering plantations in re- • Flowering control personal communication, 20 January • Fiber/lignin modification 2000). gions not currently forested could become a very important tool in mitigat- • Insect tolerance ing climate change (IPCC 2001). • Disease tolerance •Wood density BENEFITS OF • Growth • Stem straightness BIOTECHNOLOGY Productivity • Nutrient uptake • Cold, wet, drought tolerance Benefits come in different forms. A distinguishing feature of the in- The economic benefits can be realized troduction of technology is increased in the form of lower market costs for productivity, e.g., in output per unit flower initiation, non-flowering habit, or producing products. This typically con- input. Alternatively stated, technology sterility. This control may be useful in verts into lower prices for consumers can be viewed as either cost reducing preventing certain transgenic plants from of those products. Some of these cost or yield (output) enhancing. From a transmitting genetically modified matter reductions are examined in detail later societal point of view, this implies that to other plants and/or from migrating in this paper. Additionally, benefits can society gets more output for its expen- into the wild. be realized through the development of diture of inputs, i.e., a societal increase As with pest resistance, disease re- increased quality and/or new products. in efficiency. For the consumer, the sistance is also important, and the tech- These benefits are typically recognized implication typically is that relative nology for genetic modification for dis- within the market and are reflected by prices of the desired good fall com- ease resistance is fairly well developed. cost or price changes. pared with what they would have been In New Zealand, for example, the first Benefits can also be realized out- in the absence of the innovation. Plan- applications of genetically modified side the market. In agriculture, for ex- tation forestry has enjoyed success in pine (Pinus radiata) are likely to in- ample, benefits can accrue due to in- recent decades, in part, because it has volve “stacking”, that is, combining creased protein content in genetically experienced cost-reducing technology several genetically modified genes, per- modified rice. One important set of thereby giving planted forests a com- haps including those of pest- and dis- nonmarket benefits in forestry has been petitive advantage over natural old- ease resistance and flowering control, the substitution of plantation grown growth forests (Sedjo 1999). Further- in the seedling. Lignin control is wood for the wood of primary forests. more, the opportunities with the ap- viewed by the industry as an important This has reduced the commercial log- plication of biotechnology to forestry appear substantial. 37 Tree Improvements survival, and rapid growth of raw wood Other wood characteristics relate material, tree improvement programs to utility in producing the final prod- With the planting of trees for in- can also focus on wood quality. Wood uct. The absence of large or excessive dustrial wood production, there is an quality includes a variety of character- branching, for example, influences the inherent incentive to improve the qual- istics, including tree form, fiber qual- size and incidence of knots, thereby ity of the germ plasma so as to gener- ity, extent of lignin, improved lignin allowing for fuller utilization of the ate tree improvements that can be cap- extractability, and so forth. Further- tree’s wood volume. Desired character- tured at harvest. Tree improvements more, the desired traits vary by end istics or properties of final paper prod- can take many forms (Table 6). Thus product. Wood quality may involve ucts include paper tear strength, sur- far, the most common emphases of tree one set of fiber characteristics for face texture, and brightness; these are improvement programs are increased pulping and paper production and an- all properties that relate in part to the growth rates, stem form, and disease other set of characteristics for milling nature of the wood fiber used. Some resistance. Growth typically refers to and carpentry. Wood desired for fur- characteristics relate to wood used in wood volume growth or yields. Dis- niture is different from that desired for final wood products, for example, ease- and pest-resistance traits are also framing lumber. In addition, some straightness facilitates production of desired to promote or insure the characteristics are valued not for their boards or veneer in solid wood prod- growth of the tree. Resistance traits utility in the final product, but for their ucts. Other examples are related to may be oriented to specific problems ease of incorporation into the produc- milling and use in carpentry, such as common in the growth of particular tion process. wood color, strength, and surface char- species or to extending the climatic For pulp and paper production, acteristics. In addition, wood fiber is range of certain species. For example, there are certain characteristics desired increasingly being processed into struc- the development of frost-resistant eu- to facilitate wood handling in the early tural products such as strand board, calyptus would allow for a much stages of pulp production. For ex- fiberboard, and engineered wood prod- broader planting range for this desired ample, the straightness of the trunk has ucts, which have their own unique set commercial genus. Other improvement value for improving pulp and paper of desired fiber characteristics. possibilities include, as in agriculture, products, in that less compression as- In recent years pulp producers the introduction of a herbicide-resis- sociated with straight trees generates have begun to move away from sim- tant gene to allow for more efficient preferred fibers. A straight trunk is also ply producing standardized “commod- use of effective herbicides, especially in important in pulp production, since it ity” pulp and toward the production of the establishment phases of the planted allows ease of handling and feeding specialized pulp for targeted markets. forest. Besides ensuring establishment, into the production system. Paper pro- For example, Aracruz, a Brazilian pulp duction requires fiber company, has asserted that it can cus- with adequate strength to tomize its tree fibers to the require- Table 6. Tree improvement programs. allow paper sheets to be ments of individual customers. This Important attributes produced on high-speed requires increased control over the mix machines. Ease in pro- and types of wood fibers used. Cus- • Growth rates cessing includes the tomized products require customized • Disease and pest resistance • Climate range and adaptability breakdown of wood fi- raw materials. However, in the case of •Tree form and wood fiber quality, e.g., straightness bers in processing and Aracruz, thus far the control has been of the trunk, the absence of large or excessive the removal of lignin, a provided through cloning, but not branching, the amount of taper in the trunk. compound found in the transgenic plants. • Desired fiber characteristics that may relate to ease in tree that is removed in processing, e.g., the break-down of wood fibers in the pulp-making process. chemical processing.

38 Anticipated Cost Saving per ton of pulp Table 7. Possible financial gains from future biotech innovations. Innovations output. If these costs are reduced Innovation Benefits* Additional 3 A recent study (Table 7) identified $10 per m , this operating costs a number of innovations in forest bio- provides a surplus Clone superior pine 20% yield increase $40/acre or 15%– technology believed to be feasible (or effective cost after 20 yr 20% increase within the next decade or two and es- reduction) of Wood density gene Improved lumber strength None about $47 per ton timated the possible financial benefits Herbicide tolerance Reduce herbicide and None 6 of wood pulp (as- of their introduction. The develop- gene in eucalyptus weeding costs potentially ment costs of the innovations are not suming 4.7 cubic (Brazil) saving $350 or 45% per ha 7 meters per tonne considered. The innovations noted in Improve fiber Reduce digester cost None of pulp), assuming Table 7 suggest a potential decrease in characteristic potential savings costs and/or an increase in wood vol- wood prices are of $10 per m3 not affected. This ume or quality. Rates of return have Reduced amount Increase value $15 per m3 None been estimated from many of them. type of innovation of juvenile wood (more useable wood) would be impor- For example, the 20% increased vol- Reduce lignin Reduce pulping costs None tant to the forest ume due to the cloning of superior potential of $15 per m3 pine is estimated to provide a financial sector, since a mill would be willing * The actual cost savings experienced by the tree planter will depend return of about 15%–20% on the in- importantly on the pricing strategy used by the gene developer and the cremental investment cost of $40 per to offer a pre- portion of the savings to be captured by the developer and that passed on acre. This assumes initial yields of 15 mium for wood to the grower. m3 per ha per year and a stumpage fiber that had a Source: Context Consulting. price of $20 per m3. Similarly, cost sav- low processing ings should be realized for improved cost. If the improved fiber is common, A CRUDE ESTIMATE OF innovations that reduce the amount of then it would be expected to create THE GLOBAL IMPACT: low value juvenile wood or reduce the processing cost savings that would amount or difficulty of extracting lig- eventually be passed on to the con- A CASE STUDY OF sumer. Thus, substantial cost savings nin in the pulping process. HERBICIDE RESISTANCE In another example given in Table could be generated. 7, the herbicide and weeding potential This section examines the poten- cost savings in a Brazilian planted for- tial costs savings of a specific biotech- est due to the herbicide tolerance trait nological innovation—the introduction is estimated to generate an immediate of a herbicide-resistant gene—on the reduction of $350 per ha in the estab- 6 costs of establishing future commercial lishment costs in the two- to three-year The distribution of the benefits of a patented innovation is complex. Initially, one would expect forests and thus on the potential future establishment period. Obviously, this most of the benefits of the innovation to be timber supply. By inference, the likely potential degree of financial benefit, captured by the price charged for the improved effect on harvests from natural forests product. Subsequently, however, the price charged which reduces initial establishment is also examined. The approach used is costs on the order of 40%, is substan- for the new technology typically declines. At the end of the patent period, the technology becomes that of a crude partial equilibrium ap- tial. Biotechnological innovations that part of the public domain. proach,8 which estimates the cost sav- modify wood fiber characteristics so as 7 As is well known, once the investment is made ings associated with the development of to reduce pulping costs have also be in innovation, it is a fixed cost and unrelated to a specific innovation as applied to for- estimated. The value added from the marginal cost associated with the distribution estry—the herbicide-resistant gene. The pulping is about $60 per m3 or $275 of the product.

39 savings in plantation establishment resistance would re- Cost costs are estimated on the basis of the duce the costs of plan- data presented above. These savings are tation establishment translated into the lowering of the sup- by an average of about Cost S ply curve for planting activity. This re- $35/acre for fast-grow- 0 sults in an incremental addition to ing softwoods (reduced Cost1 plantings. Due to the delay between costs of 15%) and an D planting and harvest, the direct impact average of $160/acre on harvests is delayed to the future tim- for fast-growing hard- 0 Q0 Q1 Q ber supply.9 woods (reduced costs Figure 2 provides a schematic of of 30%) through the Figure 2. New plantation starts. the demand and supply for plantation elimination of the forests. As the diagram shows, if the costs of other pest mitigation activities.10 Another issue is the extent to which costs of plantation establishment de- In North America about 4 million acres lower establishment costs would increase crease from Cost0 to Cost1, this is re- are planted annually. If 98% (3.9 mil- total plantation establishment. Of the 10 flected in a downward shift of the sup- lion) are softwood and 2.0% (0.1 mil- million ha of forest planted annually, we ply curve from S to S’. other things lion) hardwood, the potential cost reduc- assume that about 1 million ha repre- constant, and the quantity of plantation potential at current rates of plant- sents new industrial plantations.11 As- tions increases from Q0 to Q1. The ing would be $136.5 million for soft- sume that the actual costs to the indus- economic benefits are the cost savings, woods and $16 million for hardwoods try were reduced by the full amount of which is represented by the area be- or a total savings of $152.5 million an- the cost reduction realized through the tween the two cost curves and bounded nually. innovation, for example, that the inno- by the demand curve on the right and Worldwide about 10 million acres vation was priced at marginal cost. This the vertical axes on the left. of plantation forest are planted per year. would be an average reduction of 22.5% Table 8 presents estimates of the If the plantings are roughly 50–50 co- in plantation establishment costs. Under cost reduction in plantation establish- nifer and hardwood and the plantings these circumstances what increase would ment for the herbicide-resistant gene remain unchanged, the potential saving be expected in the annual rate of plan- used in this study. Forest plantation es- from the introduction of the herbicide tation establishment? The expected tablishment involves incurring substan- resistant gene is $175 million for soft- amount would depend, in part, on the tial costs in an early period in order to woods and $800 million for hardwoods, responsiveness of demand to price generate larger, but discounted, benefits where the development of the clonal changes. This responsiveness is captured at some future time. High-yield planta- prerequisite is largely developed (Table in the economist’s use of price elastici- tion forestry involves plantations with 9). Thus the potential global cost sav- ties.12 To examine this question, we de- harvest rotations of from 6 to 30 years. ings is about $800 million annually, with To the extent that costs of establishment enabling technology that is essentially 11 Sedjo (1999) estimated this to be about can be reduced, net benefits can be available today for hardwoods, and 600,000 ha for the tropics and subtropics, while achieved. Experts estimate that herbicide roughly $975 million annually, once the model of Sohngen et al. (1999) estimated new plantations to be about 850,000 ha low-cost conifer cloning has been per- annually. The somewhat higher figure used in 8 A more sophisticated modeling approach would fected. Thus, the near-term potential this study reflects the inclusion of new plantation involve integrating estimates into a forest sector benefits are quite large, even if softwoods establishment in the temperate regions and anecdotal evidence suggesting that these earlier systems model (e.g. see Sohngen et al. 1999). are not considered. estimates were on the modest side. 9 It should be noted, however, that the 10 12 anticipation of greater future supplies will effect The percentages are based on an updating of Price elasticity is simply the percentage change current actions, including current harvests (see plantation establishments costs as found in Sedjo in quantity divided by the percentage change in Sohngen et al. 1999). (1983). price. 40 velop and estimate the impacts from additional planting would be divided the demand elasticity is –0.7.14 In this three scenarios: the maximum impact, evenly between conifer and hardwood. case we estimate a total of 78,750 ad- an intermediate impact, and a low im- Furthermore, if we assume that growth ditional ha planted per year with an pact (Table 10). rates on plantation forests would aver- increase in total production at harvest Scenario A: Maximum Impact— age 20 m3 per ha per year for softwoods of 1.969 million m3 per year. After 20 Given an initial total annual rate of glo- and 30 m3 per ha per year for hard- years of planting at this rate the addi- bal planting of 1.0 million ha and as- woods, the result of the additional tional continuous wood production suming an infinite supply elasticity and plantings would result in a future addi- would be about 39.375 million m3 per a unitary demand elasticity for forest tion to total annual production at har- year. plantation plantings (a derived demand), vest of 2.5 million m3/yr. If these in- the estimated impact would be the es- creases in plantings were realized each ENEFITS OF OREST tablishment of an additional total plant- year for a 20-year period, about 100 mil- B F ing area of 225,000 ha per year. This as- lion m3/yr of additional industrial wood BIOTECHNOLOGY: A sumes that the additional planting would production would be generated annually SUMMARY reflect current mix of planting. i.e., the after 20 years.13 Scenario B: Economic Benefits Intermediate Im- pact—Suppose the As noted, a distinguishing feature Table 8. Herbicide resistance benefits. same conditions of the introduction of technology is • $35/acre ($87/ha) cost reduction for fast-growing softwoods* obtained as in Sce- increased productivity, for example, in • $160/acre ($400/ha) cost reduction for fast-growing hardwoods nario A, except output per unit input. From a societal point of view, this implies that society *It should be noted that for many conifers low-cost clonal forestry is not that the supply well developed. Thus, the wide-spread application of GMOs to conifers is elasticity was 1.0. gets more output for its expenditure of not feasible at this time. However, New Zealand appears to have a In this case a total inputs; there is a societal increase in workable system for Pinus radiata. of 112,500 addi- efficiency. The above analysis suggests Source: Context Consulting. tional ha planted that the annual economic benefits in per year would re- reduced costs associated with the intro- Table 9. Potential cost saving from herbicide sult in a total increased produc- duction of only one transgenic gene, resistant gene (in millions of U.S. dollars). tion at future harvest of 2.5 mil- the herbicide-resistant gene, could re- lion m3/year. After 20 years of duce the global costs of the establish- North America Total global planting this would generate ment of planted forests by as much as Hardwood 136.5 800.0 about 50 million m3/yr of addi- one billion dollars annually. This cost Conifer* 16.0 175.0 tional continuous production. reduction implies an increased rate of Total 152.5 975.0 Scenario C: Estimated Mini- tree plantation establishment into the * Assumes successful development of enabling mum Impact—The assumption indefinite future and more industrial commercial clonal technology. is that supply elasticity remains wood at lower prices in the future. Of a +1.0, as in Scenario B, but that course, substantial additional economic Table 10. Scenario summaries. benefits could be derived from the host of other biotechnological innovations, Scenario Additional 1-year 20-year plantings additional m3 additional m3 13 At the 0.5% annual increase consumption, on a 1997 production/ Scenario A 225,000 5.00 million 100.0 million consumption base of 1.5 billion m3, 14 This is approximately the recent FAO estimate Scenario B 112,500 2.50 million 50.0 million global industrial wood consumption of -0.67 for the elasticity of demand for Scenario C 78,750 1.97 million 39.4 million would be expected to increase about 7.5 industrial roundwood. million m3 annually.

41 including a variety of additional Environmental Benefits forests by the substitution of low-cost transgenic trees with various other eco- of Forest Biotechnology wood from plantations, biotechnology nomic advantages. improved trees could be modified to Furthermore, the increased bio- The above discussion has focused specifically provide certain desired en- logical and economic productivity of on the economic or financial benefits vironmental services. Modifications planted forests has important positive of biotechnology to forestry. These fi- that would allow trees to grow in pre- spillovers to the environment. In- nancial benefits are manifest through viously unsuitable areas, such as arid creased planted-forest productivity im- reduced costs and/or higher production and degraded lands, could enable trees plies the creation of more low-cost of wood, and through enhanced qual- to provide restoration benefits, as well plantation forests and lower-cost indus- ity through improved traits and wood as traditional ecosystem services such trial wood associated with those plan- characteristics, suitable for both solid as erosion control and watershed pro- tations. Wood from planted forests de- wood products and pulp and paper tection. Additionally, certain desirable velops a greater comparative cost ad- products. Additionally, as discussed species could be modified to allow vantage over wood harvested from below and summarized in Table 11, them to grow in areas that were previ- natural forests. Thus, while harvests biotechnology in forestry can be used ously unsuitable because of frosts or a from planted forests increase, produc- to achieve a number of environmental cold climate. This modification could tion from natural forests declines. In outputs and generate improvement in not only increase wood outputs, but short, plantation wood is substituted various environmental objectives. In might be appropriate for environmen- for natural forest wood, thereby leav- addition to the protection from har- tal objectives. ing the natural forests for other uses, vests afforded natural and old-growth Additionally, biotechnology pro- including ecosystem and biodiversity vides the potential of restoring species preservation.15 severely damaged through pests and 16 The American Chestnut was decimated around disease, such as the American chest- 15 The argument that plantation wood substitutes the turn of the 20th century by a introduced nut.16 for wood from natural forests is substantially fungus. However, the fungus acted only on the different from the issue of land involved in grain above ground portions of the tree. Thus, live roots And finally, forestry has been shown production in that forestry compares a foraging remain and could provide the bases for a to have substantial potential for mitigat- with a cropping activity. A recent FAO study restoration should the fungus be controlled, ing the build-up of atmospheric green (1996) estimated the global demand elasticity of through genetic modification. Appropriate genes industrial wood at –0.67. appear to be available in the Chinese Chestnut. house gases, including carbon, believed to be the cause of anticipated global warming (IPCC 2001). Biotechnology Table 11. Environmental benefits. applied to forestry could assist in en-

Environmental outputs Biotechnological innovations hancing the carbon sequestration ability of forests, and thereby provide additional Reduced pressure to log Plantation wood from more productive carbon mitigation possibilities. natural and old-growth forests. forests will substitute for wood from natural forests at lower costs. To summarize, the benefits of biotechnology in forestry can be viewed as Protection forests can be Genetically improved trees with land established on degraded protection and land restoration coming in two groupings. First, biotech- or arid lands. capabilities suited to poor sites. nology has generated a number of innovations that will significantly reduce costs Carbon sequestrating forest Genetically improved trees capable of and/or enhance the quality of the for- can be established on sites substantial carbon sequestration estry outputs, thereby enhancing society’s previously not suitable to forestry. suited to biologically poor sites. efficiency in resource use. Some portion Species restoration. The potential species restoration of these benefits is likely to be transferred of the chestnut. to the consumer through lower prices,

42 and we would expect the transfer to in- controversy. The nature of the contro- populations. In many cases plantation crease over time. Additionally, biotech- versy in agriculture has developed tree species would be exotic and thus nology has the potential to generate a around at least five issues. exchange would not be a factor. In number of environmental benefits First, is the issue of ownership of cases where genetic exchange could be through its effect on the competitive the modified genes and the question of a problem, a method to prevent or re- structure of the forest industry. In gen- how much ownership/control the bio- duce their “escape” would be to pro- eral, this will be through decreasing the technology companies have over their mote sterility or reduce or delay flow- competitive advantage of the harvest and transgenic products after they have ering (see DiFazio et al. 1999). The use of natural and old-growth timber been distributed. An important ele- implications of gene escape are likely and increasing the substitution and use ment in the discussion relates to the to differ depending on whether the of plantation wood—thereby imparting ongoing controversy regarding the gene would confer a selection advan- a degree of protection from commercial broader philosophical issue of the own- tage to the wild plants. This is likely logging to the natural and old-growth ership of biodiversity and improved to depend upon the nature of the ge- forests, which are viewed as having products. Are wild genetic resources netic alteration. greater environmental value. Finally, the the property of all of humanity or of A fourth issue relates to the im- biotechnological modification of a tree the country in which they reside? And pact of the biotechnology on the resis- can allow it to perform a broader and are developed biotechnology products tance of the targeted pest population. more useful set of both economic and the property of the developer or should It is well known that pests adapt environmental functions and services. they be available without royalty pay- through natural selection to the intro- These include, for example, enhanced ment to all of humanity? (For example, duction of pest-controlling chemicals. carbon sequestration generally, and its see, Kloppenburg, Jr. 1988; Sedjo The same response would be expected potential in regions that have been de- 1992.) This controversy continues to to attempts at genetic pest control. As graded and are currently difficult for for- be manifest in the difficulties in inter- in agriculture (Laxminarayan and estry. Biotechnology can also enhance preting and finalizing the “biodiversity Brown 2000), in forestry there could other desired environmental objectives, treaty” coming out of the UNCED be a problem of the pest population such as restoration, watershed enhance- “Earth Summit” meeting in Rio de adapting to the modified gene and ment, and erosion control in areas typi- Janeiro in 1992. thereby undermining its longer-term cally not suitable to forests and/or areas The second issue in the overall effectiveness. The long period of for- subject to cold, frost, and drought. controversy relates to the health, safety est growth would seem to exacerbate and environmental aspects of the problem, as it would allow insect transgenic products. Although there is populations many generations to de- POTENTIAL COSTS OF little or no evidence that transgenic velop a resistance mechanism. Various foods are unsafe, health concerns are approaches are being considered to BIOTECHNOLOGY IN raised due to the lack of long-term ex- overcome this problem including the FORESTRY: SOME posure and experience with such prod- continuing development of new pesti- ucts. The health issue is not generally cides in agriculture and the use of refu- CONCERNS raised for trees as they are not usually gia to dilute the development of resis- viewed as a human or animal food tance in the pest population. Transgenic biotechnology has be- source. Finally, there is the issue of come quite controversial when applied A third issue with transgenic whether biotechnology applied to ag- to agriculture (e.g., see Science 1998). plants is the question of genetic trans- riculture will increase the demand for However, in drugs, medicines, and fer to nearby domestic or wild popu- land, thereby putting increased pressure pharmaceutical applications, transgenic lations. For forestry the concern is on natural habitats. Some recent work biotechnology is essentially without largely with genetic transfer to wild suggests this is likely to be the case if

43 the demand for agricultural products fer of genes from exotic conifer to in- serious problem if it results in the al- is elastic (e.g., Angelsen and Kaimowitz digenous conifer trees is precluded. tering of the comparative competitive 1998).17 However, this is unlikely to be Where the species is indigenous, an position in dealing with the pests of a problem in forestry where demand is approach may be the introduction of various types similar natural vegetation. almost always estimated to be inelastic sterility as a vehicle for preventing the Ultimately, the seriousness of this prob- and productivity of planted forests con- release of genes that might transfer to lem depends importantly on the prob- siderably greater than that of natural the natural environment. Note that the ability of the transfer of a survival gene forests. major reason for introducing a steril- into the wild, on the scale of the trans- In some ways the biotech issues in ity gene into trees is not, as in agricul- fer, and on the comparative change in forestry appear to be modest compared ture, to retain control over future seed the competitive balance within the with those in food. Since wood prod- sources, but rather to prevent the es- natural habitat. This becomes an argu- ucts are not ingested they are unlikely cape of genes into the natural environment for the introduction of controlled to have any direct human health or ment through the tree-flowering pro- exotics. safety effects, either in the short- or cess. long run. The ownership issue associ- Finally, if modified genes do es- ated with the use of seeds from cape, how serious are the “expected” SOME IMPLICATIONS transgenic plants to create subsequent consequences or the “worst case” con- crops is likely to be less important as sequences? In the case of the herbicide- OF BIOTECHNOLOGY well, due to the long periods required tolerant gene, the consequences of re- FOR FORESTRY: for flowering in trees. lease into the wild are probably small. A more pressing concern, however, Herbicides are unlikely to be applied WEIGHTING BENEFITS relates to the potential for genetic to most of the natural environment. If AND COSTS transfer from the transgenic tree to the herbicides are to be applied, types can 18 surrounding natural environment. As be used to which the escaped genes do The benefits of applying biotech- noted, this is not a problem in cases not confer tolerance. In the interme- nology to forestry are potentially huge. where the tree is an exotic and there- diate and longer term, the herbicide in The estimates above suggests that the fore no similar species of trees are question will almost surely be replaced introduction of only one type of bio- found in the natural environment, e.g., periodically in the normal course of technological innovation, a herbicide- since conifer species are not indigenous product change and development. resistant gene, could generate benefits to South America, the accidental trans- Thus, the presence of that modified estimated at up to $1 billion annually gene in the natural environment ap- in reduced forest plantation establish- pears unlikely to constitute any serious 17 It has been noted that since cattle are ment costs and an expansion in the rate increasingly being placed in feedlots where they short- or long-term environmental of plantation establishment by up to consume grains, the total demand for grain, problem. Similarly for genes that affect 225,000 additional ha per year. The human and animal, may be elastic. This implies tree form or fiber characteristics, the that if grain prices fall, e.g., due to biotechnology, increased production would not only the total area of land in grains could increase. release of this gene into the natural generate increased social welfare However, it should be noted that where both environment is unlikely to provide a through lower commodity prices, but grain and cattle are part of society’s diet, the competitive advantage in survival and would also generate environmental feeding of grain to cattle has resulted in a decline therefore unlikely to have significant or in pasture area. Thus, total agricultural land, benefits in the form of decreased har- grain plus pasture, may have decreased even if the adverse consequences. vesting pressure on natural forests. area in grains increased. However, this situation could Furthermore, it is well docu- change if a survival gene is involved. mented that there has been a gradual 18 See Mullin and Bertrand (1998) for a detailed For example, the release of a Bt gene discussion of many of these issues in a Canadian worldwide shift in industrial wood pro- into the wild could constitute a more context. duction from natural forests to plan- 44 tations. Such a trend could have advan- SUMMARY AND AAEA International Conference, tageous effects on native forests and Agricultural Intensification, Eco- nomic Development and Environ- biodiversity in that as harvest pressures CONCLUSIONS ment, Salt Lake City, Utah, 31 are relieved and native forests can be July–5 August, 1998. devoted to other purposes, including The benefits of biotechnology in Bailey, R. 1997. American Chestnut conservation. The more productive are forestry, both economic and ecological, Foundation. Center for Private Con- are potentially enormous. The wide- forest plantations, the more they can servation. CEI, Washington, DC. deflect harvesting pressures away from spread use of a herbicide-resistant gene Bradshaw, T. 1999. A Blueprint for For- natural forests. in forestry could result in a savings of est Tree Domestication. College of up to $1 billion annually. However, the Additionally, biotechnology ap- Forest Resources, University of plied to trees offers an additional tool benefits must be compared with the Washington, Seattle, WA. Down- in dealing with specific environmental costs. Recently, biotechnology in agri- loaded from http:// problems, including land and water culture has come under attack for its poplar2.cfr.washington.edu/toby/ protection, as well as presenting the potential health, safety, and environ- treedom.htm. July 2001. potential to deal more effectively with mental risks. The application of bio- Context Consulting. N.D. West Des global warming and atmospheric car- technology to forestry, however, poses Moines, IA 50266. bon mitigation. somewhat different considerations than DiFazio, SP, S Leonardi, S Cheng, and The costs of biotechnology in for- biotechnology’s applications elsewhere. SH Strauss. 1999. Assessing Poten- estry are much more problematic. In For example, direct health and safety tial Risks of Transgenic Escape many cases the potential costs of the risks appear nonexistent or negligible. from Fiber Plantations. Gene Flow and Agriculture: Relevance for introduction of biotechnology in for- The environmental risks that exist ap- Transgenic Crops 72:171–176. estry appear to be negligible or mod- pear to relate largely to the potential est at best. Herbicide resistance and for altered genes to move out of FAO. 1996. Long-term Trends and Pros- pects in the Supply and Demand for form and fiber modification appear to transgenic trees into the natural envi- Timber and Implications for Sus- ronment. The damages associated with offer minimal potential damages. The tainable Forest, by David Brooks, greatest concern is probably related to the escape of many types of these ac- Heikki Pajulja, Tim Peck, Birger the escape of modified genes into the tivities are negligible and probably can Solberg, and Philip Wardle. FAO, natural environment. The costs associ- be reduced substantially by the delay Rome. ated with this are unclear, but in many or elimination of flowering and/or by Hayami, Y, and VW Ruttan. 1985. Ag- cases would be negligible. Furthermore, introducing the species into foreign ricultural Development: An Interna- most could probably be reduced sub- environments where similar species are tional Perspective. Johns Hopkins stantially by the delay or elimination not found in the wild and gene trans- Press, Baltimore. See p. 115. of flowering and/or by introducing the fer is highly improbable. Where the Hu, W-J, SA Harding, J L, JL Popko, J species into foreign environments risks cannot be adequately mitigated, Ralph, DD Stokke, CJ Tsai, and where similar species are not found in certain selected types of biotechnologi- VL Chiang. 1999. Repression of Lignin Biosyntheses Promotes Cel- the wild and gene transfer is highly cal activities could be precluded. lulose Accumulation and Growth improbable. in Transgenic Trees. Natural Bio- Finally, biotechnology in forestry technology 17(August): 808–812. takes many forms. Even if certain EFERENCES R IPCC. 2001. Technical and Economic transgenic trees are viewed as poten- Potential of options to Enhance, tially risky, there are a host of genetic Angelsen, A, and D Kaimowitz. 1998. Maintain and Manage Biological modifications that offer negligible so- When Does Technological Change Carbon Reservoirs and Geo-engi- cial risk. in Agricultural Promote Defores- neering, Chapter 4 in Working tation? Paper presented at the Group III, Third Assessment Report

45 on Climate Change, Intergovern- Sedjo, RA. 1992. Property Rights, Ge- mental Panel on Climate Change, netic Resources, and Biotechno- Pekka Kauppi and Roger Sedjo, logical Change. Journal of Law and convening lead authors. Economics XXXV:199–213. Kloppenburg, JR, Jr., ed. 1988. Seeds Sedjo, RA. 1999. The Potential of and Sovereignty: The Use and Con- High-yield Plantation Forestry for trol of Plant Genetic Resources. Duke Meeting Timber Needs. New For- University Press, Durham, NC. ests 17:339–359. Laxminarayan, R, and GM Brown. Sedjo, RA, and D Botkin. 1997. Using 2000. Economic of Antibiotic Re- Forest Plantations to Spare Natu- sistance. RFF Discussion Paper ral Forests. Environment 00–36. 39(10):15–20, 30. McLean, MA, and PJ Charest. The Sedjo, RA, A Goetzl, and M Steverson. Regulation of Transgenic Trees in 1997. Sustainability of Temperate North America.” Silvae Genetica Forests. RFF, Washington. 49, 6:233–239 (in press). Sohngen, B, R Mendelson, and RA Menzies, N. 1985. Land Tenure and Sedjo. 1999. Forest Management, Resource Utilization in China: A Conservation, and Global Timber Historical Perspective. Paper pre- Markets. American Journal of Ag- sented at the conference, Manag- ricultural Economics 81(1). ing Renewable Resources in Asia, Westvaco. 1996. Fiber Farm Expansion. Sapporo, Japan, June 24–28, 1985. Forest Focus 20(4). Westvaco, Tim- Mullin, TJ, and S Bertrand. 1998. En- berlands Division, Summerville, vironmental release of transgenic SC. trees in Canada—potential ben- Westvaco. 1997. Forest Focus 21(1). efits and assessment of biosafety. Westvaco, Timberlands Division, The Forestry Chronicle 74, No. 2: Summerville, SC. 203–219. Williams, N. 1998. Agricultural biotech Pullman, GS, J Cairney, and G Peter. faces backlash in Europe. Science 1998. Clonal Forestry and Genetic 281 (Aug 7): 768–771. Engineering: Where We Stand, Future Prospects, and Potential Withrow-Robinson, B, D Hibbs, and J Impacts on Mill Operations. Beuter. 1995. Poplar Chip Produc- TAPPI Journal 81(2). tion for Willamette Valley Grass Seed Sites. Research Contribution 11, Sedjo, RA. 1983. The Comparative Eco- College of Forestry, Forest Re- nomics of Plantation Forests. Re- search Laboratory, Oregon State sources for the Future, Washing- University. ton, DC. Yanchuk, AD. 2001. The Role and Sedjo, RA. 1991. Forest Resources: Re- Implication of Biotechnological silient and Serviceable. Chapter 3, Tools in Forestry. Unasylva 204 pp. 81– 122, in America’s Renew- (vol.52):53–61. able Resources, ed. K. D. Frederick and R.A. Sedjo, Resources for the Future, Washington, DC.

46 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 47-56. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Restoring the Forests*

David G. Victor Jesse H. Ausubel ABSTRACT

The 20th century witnessed the start of a “Great Restoration” of the world’s forests. Efficient farmers and foresters are learning to spare forestland by growing more food and fiber in ever-smaller areas. Meanwhile, increased use of metals, plastics, and electricity has eased the need for timber. And recycling has cut the amount of virgin wood pulped into paper. Although the size and wealth of the human population has shot up, the area of farm and forestland that must be dedicated to feed, heat, and house this population is shrinking. Slowly, trees can return to the liberated land. We develop a plausible and attractive scenario for how far this “great restoration” can proceed by 2050—if farmers lift yields at about 2% per year and thus grow ever more food on smaller areas of land, and if foresters continue the shift to high yield plantation forests, which reduces the wooded area that must be devoted to timber supply. The average timber yield needed to achieve David G. Victor is Robert our “great restoration” scenario is about 5 cubic meters per hectare per year. That can be W. Johnson, Jr., Senior attained without genetically modified (GM) trees, but insofar as GM trees allow for even higher yields, they make it feasible to shrink even further the area of production forests. Fellow for Science and Hectares freed from timbering can be available for other purposes such as protection of biological diversity, watershed protection and nature’s intrinsic beauty. Technology at the Council on Foreign Relations. SKINHEAD EARTH? [email protected] ight thousand years ago, when humans played only bit parts in the world Jesse H. Ausubel is Direc- ecosystem, trees covered two-fifths of the land. Since then, humans have tor of the Program for the E grown in number while thinning and shaving the forests to cook, keep warm, grow crops, plank ships, frame houses, and make paper. Fires, saws, and Human Environment at axes have cleared about half of the original forestland, and some analysts warn Rockefeller University, that within decades, the remaining natural forests will disappear altogether. But forests matter. A good deal of the planet’s biological diversity lives in New York. forests (mostly in the tropics), and this diversity diminishes as trees fall. Healthy forests protect watersheds and generate clean drinking water; they remove car- [email protected] bon dioxide (a greenhouse gas that traps heat in the atmosphere) from the air and thus help maintain the climate. Forests count—not just for their ecological This essay is based on the and industrial services but also for the sake of order and beauty. findings of a Council on Fortunately, the 20 century witnessed the start of a “Great Restoration” of the world’s forests. Efficient farmers and foresters are learning to spare forestland Foreign Relations study by growing more food and fiber in ever-smaller areas. Meanwhile, increased use group. For more informa- of metals, plastics, and electricity has eased the need for timber. And recycling has cut the amount of virgin wood pulped into paper. Although the size and tion on the studies under- wealth of the human population has shot up, the area of farm and forestland lying this article see * This paper was first published in Foreign Affairs (November/December 2000, vol. 79, Number 6, pp. http://greatrestoration.rockefeller.edu. 127–144) and is reprinted with permission. More information on Foreign Affairs is available at www.foreignaffairs.org. 47 that must be dedicated to feed, heat, New studies in tropical western Africa ests”—have not vanished. Earth’s trees and house this population is shrinking. reveal that deforestation in that region therefore need a comprehensive and Slowly, trees can return to the liberated is only one-third the rate previously durable solution: to expand and accel- land. believed, and in some areas forests are erate the Great Restoration worldwide. In the United States, this Great rebounding. Brazil, for its part, is of- Agriculture and logging—the two Restoration began with a big stick. ten in the forest press. Farmers’ fires, main threats to natural forests—must Horrified that farmers and loggers were cattle ranching, and timber cutting continue their transformation into stripping America of its trees five times denude the Brazilian Amazon by per- modern, ultra-efficient industries. faster than they were growing, and haps half a percent each year, and the The seedlings and saplings of this worried about the economic conse- government seems powerless to stop it. transformation have already been quences of a “timber famine,” President By some estimates, four-fifths of planted. But the progress and poten- Theodore Roosevelt created the federal Brazil’s local wood consumption is il- tial of modern agriculture and forestry Forest Service and pushed landowners legally felled. Yet at the same time, Bra- remain little known to many to start sustaining timber resources. zil has become a powerhouse in forest policymakers, and requisite techniques Since about 1950, U.S. forest cover has planting. Established on already de- are reviled by others who prefer “natu- increased—despite the country’s emer- graded and abandoned land, eucalyp- ral,” low-intensity production. And in gence as the world’s bread and wood tus and pine stands in Brazil supply a much of the world, the conditions nec- basket. Geographers have observed a rising fraction of the world’s lumber essary for these new methods, such as transition from deforestation to refor- and paper and relieve the pressure on affordable commercial energy and ef- estation in countries as distant as natural forests. fective land-use regulation, remain elu- France and New Zealand, where new Yet still the world’s forest estate sive. production methods have spared for- dwindles. Even in countries where The chart illustrates the immense ests and regulation has locked the gains woody areas are expanding, threats to areas at stake. Two paths now stand in place. Studies by forest experts in the remaining uninterrupted original open. Along one, leading to the Finland reveal that by the 1980s, tracts of trees—what the World Re- “Skinhead Earth” scenario, quaint and wooded areas were increasing in all sources Institute calls “frontier for- major temperate and boreal forests. These mid- and high-latitude forests account for half the world’s total and span some 60 countries. Such forests today are also healthier: the biomass (or total amount of living matter) per hectare (100 meters square, or about 2.5 acres) has increased even more rapidly than the size of the forests themselves. But the Great Restoration is far from complete. Despite major gains in some areas, the world’s sylvan balance sheet still bleeds trees, owing to widespread deforestation in the tropics. Yet even there, progress has begun to peek through. Preliminary satellite data sug- Figure 1. Sources (rounded estimates): 6000 B.C., World Conservation Monitoring Centre, gest that the rate of tropical deforesta- World Resources Institute, and World Commission on Forests and Sustainable Development; 1990s, U.N. Food and Agriculture Organization Global Fibre Supply Model data; 2050, tion has slowed 10% in the last decade. authors’ projections.

48 inefficient agriculture and forestry will than 2% around 1970. Still, by 2050, ply to nearly double, while cropland persevere. By 2050, forests will dwindle the total population will have in- expanded less than 10%. In India, ris- by 200 million hectares—about five creased, perhaps to as much as 8 or 10 ing yields almost entirely offset increas- times the area of California—and lum- billion. Taming population growth fur- ing demand for cropland, so the area berjacks will regularly shave about 40% ther will likely lessen the threat to for- under cultivation barely changed. of forests. Along the other, however, ests, but protecting the forests seems The conventional wisdom, the farmers and foresters will intensify pro- only a marginal addition to the impe- “Skinhead Earth” scenario, holds that duction and shrink their footprint. tus for population reduction. as much as 200 million hectares of for- Forests will spread anew to more than Rising income, meanwhile, has est will be lost in the next decades as 200 million hectares, and only 12% of raised the population’s demand for agriculture extends to feed larger and forestlands will hear cries of “timber.” food, multiplying the effect of its grow- richer populations. Current trends, This vision for a Great Restoration is ing numbers. The rich eat more than however, suggest not balding but re- realistic—one that the right domestic do the poor. But the main effect of in- growth. If farmers sustain the 1.8% and foreign policies can secure. The come growth has been to add meat to annual yield improvement they have focus is on the year 2050. That may many diets. And in terms of land used, achieved in recent decades, they could seem distant, but trees grow slowly, and eating animals that eat plants is less ef- meet the growing demand for primary capital-intensive logging firms adjust ficient than eating plants directly. As a calories while releasing 200 million their practices gradually. In one de- rule of thumb, a vegetarian diet re- hectares of cropland. cade—the time frame for most foreign quires about 3,000 primary calories But farmers can do even better policies—little change can appear. But daily. Meat-eaters consume twice that than that and offer even more land to five decades’ work, with steady guid- amount. Vegetarian diets could there- the trees. The authors’ research with ance, will make the restoration of the fore markedly reduce the land required Paul Waggoner of the Connecticut forests truly great. to grow food. But secretaries of state Agriculture Experiment Station has are unlikely to convince carnivores to shown that, with some extra effort, an switch from T-bones to tofu. increase in yield of 2% per year—a SMART FOOD Given the difficulty in changing plausible goal—could spare a total of population and diet, the best way to 400 million hectares. In other words, Many different forces, including reduce food’s impact on forests will be today’s farmland could be cut by more urban sprawl, pollution, and fire, can to change the fourth factor: how farm- than a quarter through smarter agricul- diminish forests. But around the world, ers grow crops. Yield—the amount of tural techniques. Sustaining a 2% rate agriculture and timber cutting do crops produced per hectare of land— of increase will not be easy, but history much of the clearing. Farmers are usu- is the key indicator. Over the last quar- and technology suggest it can be done. ally cited as forests’ primary foes. As ter-century, average yields of cereal Since sustained efforts to raise U.S. Time’s millennial Earth Day issue la- grains, including maize, rice, and yields began in the 1940s, average mented, “agriculture is the world’s big- wheat, rose 1.8% each year worldwide. yields for wheat and soybeans have al- gest cause of deforestation.” Some countries achieved dismal re- most tripled and corn yields have more Just how much land is actually sults—yields rose only 0.8% per year than quadrupled. And farmers have needed for agriculture integrates several in developing Africa and actually de- hardly tapped the full potential. Cham- variables: the size of the population, its clined in Angola, Malawi, and Zimba- pion American corn growers have lifted income and diet, and the yield of crops bwe. Other countries, big ones, out- yields well above 20 tons per hectare grown. Already, growth in human paced the pack. Yields rose an average without irrigation. Meanwhile, average numbers is slowing—the present popu- of 2.5% annually in Indonesia and U.S. corn yields stand at only 8 tons lation growth rate of 1.3% per year has more than 3% yearly in China. These per hectare, and average world corn declined steadily from a peak of more gains allowed the worldwide food sup- yields are a meager 4 tons.

49 How many of the hundreds of expand, not shrink, their incomes. States, for example, leftovers from lum- millions of hectares that farmers can That leaves “diet” and yields. The ber mills account for more than a third spare will revert to trees? The amount wood “diet” required to nourish an of the wood chips that are turned into depends on where cropland is aban- economy is determined by the tastes and pulp and paper; what is still left after doned and how people choose to use actions of consumers and by the effi- that is burned for power. And further it. One and a half centuries ago, farm- ciency with which millers transform vir- improvements in management and ers had deforested two-thirds of Con- gin wood into useful products. Chang- technology will squeeze even higher necticut. Once they abandoned their ing tastes and technological advances are value out of products and spare more farms to build guns and aircraft engines already lightening pressure on forests. virgin wood. In British Columbia, and sell insurance, however, the forests Concrete, steel, and plastics have re- since the mid 1980s, sawmills have gradually recovered the landscape. But placed much of the wood once used in lifted the lumber obtained per cubic free land does not always become for- railroad ties, house walls, and flooring. meter of log at an average rate of 1.2% est. In South Dakota, abandoned farms Genes, silicon, and even ceramics—not per year. Worldwide, the pulp and pa- become grass prairies, not woodlands. boards—are the growth materials for the per industry is shifting a significant Worldwide, no sure equation links the new economy. Demand for lumber has share of production from chemical to liberation of cropland to the return of become sluggish, and in the last decade, mechanical pulping, which cuts the trees. Guessing moderately, however, with the implosion of the wood-inten- wood required for a ton of useful pulp about half the land freed might even- sive Russian economy, world consump- by half. And recycling has helped close tually revert to forest—say, 200 million tion of boards and plywood has actually leaks in the paper cycle. In 1970, con- hectares, or three times the size of Texas declined. sumers recycled less than one-fifth of and four times the size of Spain. But the appetite for “pulp- their paper; today, the world average is wood”—logs that are chipped, softened double that. into pulp, and then drawn into sheets New engineering has also helped FAST FORESTS of paper and board—is still climbing, decouple demand for virgin wood from driven by the 5% annual rise in pulp the swelling population and economy. Farmers may no longer pose much consumption in developing countries. For example, floor systems built from threat to forests. But what about lum- Pulpwood accounts for more than a engineered wooden I-beams use about berjacks? As with food, the area of land quarter of industrial wood consump- one-quarter less fiber than traditional needed for wood is a multiple of popu- tion. Paperwork proliferates in devel- construction with solid rectangular lation, income, “diet,” and yield. The oping countries, and inside the glass ribs. And as a substitute for plywood, appropriate focus is on industrial and steel shells of the new economy, millers make oriented strand board wood—logs cut for lumber, plywood, information machines still consume (OSB) by gluing wood flakes in per- and pulp for paper. Although trees are paper voraciously. Reliable electronic pendicular layers. OSB can be manu- also cut for fuel, most fuel wood is archives and electronic books will even- factured from small trees, and it con- thinned from hedgerows, shrubs, and tually quiet the taste for paper. So far, sumes the whole tree, except for bark other open sources—not forests. however, life still requires hard copy. and limbs. By contrast, plywood Again, of the relevant factors, Meanwhile, more efficient lumber mills—which peel timber into sheets strategies to save the forests should not and paper milling is already carving and glue them together like cream emphasize limiting population and in- more value from the trees we cut. Be- cookies—work only with larger trees come. Those government agencies and cause waste is costly, the best mills— and leave an unpeeled core at the cen- nongovernmental organizations operating under tight environmental ter of every log. (NGOS) most concerned with forests regulations and the gaze of demanding As this suggests, the wood prod- have little leverage over the number of shareholders—already make use of ucts industry has learned to increase its people, and societies should aim to nearly the entire log. In the United revenue while moderating its consump-

50 tion of trees. This is not surprising, for moved on, returning only if trees re- ply from “old-growth” forests—mature efforts to lower trade barriers and im- generated on their own. Most of the natural forests dominated by large, old prove management of forest resources world’s forests still deliver wood this trees—is tightening while the relative are increasingly exposing millers world- way, with an average annual yield of costs of trees from plantations are fall- wide to prices, competition, and con- perhaps two cubic meters of wood per ing. In Oregon, for example, public sumer requirements that are spreading hectare. If yield remains at that rate, pressure and laws to protect endan- innovation and efficiency more widely. as illustrated, by 2050 lumberjacks will gered species have reduced felling on Large, capital-intensive pulp and paper regularly saw nearly half the world’s federal lands by four-fifths since the mills are already responding—their in- forests. That is a dismal vision—a mid-1980s. Offsetting that shrinking vestors demand it. But in much of the chainsaw every other hectare. supply is rising production on private world, sawmills thrive on remoteness, Lifting yields, however, will spare land in the southern United States— trade barriers, and artificially cheap logs more forests. Raising average yields 2% where sunlight, moisture, and good that shield them from competition. By per year would lift growth over 5 cu- soils for forests abound. Today, the one estimate, 3,000 sawmills in Argen- bic meters per hectare by 2050 and American South—which Bruce Zobel tina function with an average input of shrink production forests to just about of North Carolina State University only 1,000 cubic meters of wood per 12% of all woodlands—the Great Res- called the “wood basket of the year. At such small scales—less than toration. world”—supplies 15% of the world’s one-hundredth the size of the most Industry has already taken big industrial timber, at a sustainable av- modern sawmills—millers can hardly steps along the restoration path by sow- erage yield of about 5 cubic meters per implement the most efficient practices. ing intensively managed “plantation” hectare. Demand for industrial wood, now forests that act as wood farms. Accord- Outside the United States, dimin- about 1.5 billion cubic meters per year, ing to the U.N. Food and Agriculture ished access to traditional sources of has risen only 1% annually since 1960 Organization (FAO), one-quarter of virgin wood and the need to control while the world economy has multiplied industrial wood already comes from wood costs are also concentrating pro- at nearly four times that rate. Conven- such farms, and the share is poised to duction. In British Columbia, where tional wisdom predicts that the total soar once recently planted forests ma- most forests are old growth, regulators amount of wood harvested will reach 2.5 ture. At likely planting rates, at least have reduced the allowable cut by billion cubic meters in 2050. But the one billion cubic meters of wood—half nearly a third over the last two decades, figure could be much lower if millers the world’s supply—could come from and more restrictions are likely. Clark improve their efficiency, manufacturers plantations by the year 2050. Semi- Binkley, former dean of the University deliver higher value through the better natural forests—for example, those that of British Columbia’s School of For- engineering of wood products, and con- regenerate naturally but are thinned for estry, has argued that the province’s sumers recycle and replace more. To- higher yield—could supply most of the logging can remain competitive only by gether, these steps could shrink demand rest. Small-scale traditional “commu- shrinking its footprint and raising to about 2 billion cubic meters per year nity forestry” could also deliver a small yields to twice or three times the cur- and thus reduce the area of forests cut fraction of industrial wood. Such ar- rent average annual yield of 2.2 cubic for lumber and paper. rangements, in which forest dwellers, meters per hectare. In Brazil last year, As with agriculture, yield—cubic often indigenous peoples, earn revenue the government and a coalition of 189 meters of wood grown per hectare of from commercial timber, can provide environmental groups scuttled a plan forest each year—provides the largest essential protection to woodlands and to open half the Amazon forest for leverage for change. Historically, for- their inhabitants. potential clearing. Meanwhile, nearly estry has been a classic primary indus- Changes in both markets and all new Brazilian industrial wood try; like fishers and hunters, foresters regulation explain the shift toward comes from high-yielding plantations have exhausted local resources and then high-yield, land-sparing forestry. Sup- in the country’s southeast, outside the

51 Amazon region. China has reduced plentiful rainfall, hybrid poplars deliver Internet. Three-quarters of South cutting of natural forests by a fifth 50 cubic meters per hectare. And un- American plantations were planted af- since 1995. Malaysia and Indonesia, der extreme conditions—with irriga- ter countries adopted incentive dominant exporters of tropical old- tion, fertilization, and intensive pest schemes, usually subsidies. Yet today, growth logs, have both announced re- controls—eucalyptus has been clocked the private establishment of new plan- ductions that could halve felling in at 100 cubic meters per hectare (or 20 tations is continuing despite the fact their ancient forests by 2010. New times the goal of 5 cubic meters by that governments are scaling back in- plantations in those countries will not 2050). centive programs. mature in time to fill the gap, but Foresters can push trees even Another source of concern has planted forests in New Zealand, Chile, faster. Today, the most advanced tree- been the profitability of private invest- and elsewhere stand ready to deliver. breeding programs are only in their ment in these industries. A recent Chile alone will earn $3 billion in for- second, third, or fourth generations, PricewaterhouseCoopers study found eign exchange this year from forest since trees, unlike annual wheat and that the 50 largest global forestry com- products, nearly all grown on planta- maize, are slow to reach sexual matu- panies earned, on average, a paltry tions that cover only 3% of Chilean rity. Modern biology can already speed 4.1% return on capital investments. territory. Trade is rationalizing world breeding, however, by spotting the Over-capacity in the industry and vast wood production toward the highest— genes for superior performance early potential supplies of wood from poorly and most land-sparing—yields. and then growing plants with those regulated forests have undercut prices With economics already favoring traits through traditional methods. and hurt the performance of even the intensive production, foresters should Genetic engineering, now in its in- best-run firms. A history of poor re- be able to lift the average world yield fancy, will be able to insert or delete turns makes it hard for the forest in- in lumbered forests to 5 cubic meters selected genes directly and should dustry to raise still more money to con- per hectare by 2050. A recent study gradually gain acceptance. Big tree tinue the shift to high-yield wood pro- compiled by Wood Resources Interna- planters—such as Westvaco Corpora- duction. The current consolidation of tional, the World Bank, and the World tion—are already placing large bets on the timber industry, however, will help Wildlife Fund (WWF) suggests that biotechnology, which promises to surviving firms win new investors. more than a fifth of the world’s virgin boost the economic advantage of plan- Government efforts to improvement wood is already produced from forests tation forestry. Having spent heavily on management and restrict cutting of with yields above 7 cubic meters per state-of-the-art mills and to select and natural forests will also favor modern hectare. And foresters have only begun rejigger tree genes, the forest industry industry, which has a smaller footprint. to tap the potential for high growth. has come to prefer planted forests, Environmentalists nevertheless Roger Sedjo at Resources for the Fu- which let it control what stock grows worry that industrial plantations will ture has documented that economically where. deplete nutrients and water in the soil competitive plantations in Brazil, Economists, environmentalists, and produce a vulnerable monoculture Chile, and New Zealand can sustain and people who live in the woods have of trees where a rich diversity of spe- yearly growth of more than 20 cubic all raised warning flags about intensive cies should prevail. Meanwhile, advo- meters per hectare with pine trees. industrial forestry. Some worry that cates for indigenous peoples, who have Aracruz Cellulose, Brazil’s top planter plantation forestry is prone to fail be- witnessed the harm caused by crude of eucalyptus—a hardwood good for cause much of it depends on wasteful industrial logging of natural forests, some papers—has invested heavily in government subsidies. Indeed, public warn that plantations will dislocate for- forestry research that now delivers an funds have helped establish viable land- est dwellers and upset local economies. extraordinary average of 43 cubic sparing plantations—just as they Pressure from these groups helps ex- meters per hectare. In the Pacific helped initiate other new waves of in- plain why the best practices in planta- Northwest and British Columbia, with dustry, including jet travel and the tion forestry now stress the protection

52 of environmental quality and human growing wood farms. Together, these of commas and clauses. Worse, since rights—and why large firms, with the measures can increase to 3 billion hect- Rio, the central debate has been most exposure to pressure, are gener- ares the area of forests that are left for whether and how to negotiate a legally ally the most scrupulous. In Sweden, nature, the protection of watersheds binding forest treaty. Experience in for example, large industrial forest and indigenous peoples, and other managing other international environ- owners aim to follow strict codes of non-industrial uses. In contrast, the mental problems shows that binding conduct that respect the traditional “Skinhead Earth” scenario will shrink treaties work best when they include practices of indigenous peoples, these non-industrial forests to 1.8 bil- detailed commitments with which gov- whereas smaller landowners still tend lion hectares. This difference—1.2 bil- ernments can comply. A binding in- to fence the reindeer-herding Saami lion hectares—is almost twice the area strument is ill suited to forests, how- people out of their traditional grazing of the Amazon Basin. One central ever, because governments—and the grounds. question remains, however: How can people they represent—do not yet As with most innovations, achiev- foreign policy help farmers, foresters, share a vision for how to protect the ing the promise of high-yield forestry millers, and consumers do their part? world’s woodlands. Moreover, detailed will require feedback from a watchful Much useful activity is already actions would necessarily vary by coun- public. Public scrutiny will help indus- under way. Environmental NGOs try and be extremely difficult to codify try to make the new technologies so- around the globe have organized be- into a single international law. Key el- cially acceptable. The main benefit of hind forest protection. All major for- ements of a sensible coherent vision— the new approach to forests will not estry firms now participate in various such as lifting grain and forest yields— reside within the planted woods, how- activities to lessen the environmental are impossible to plan top-down by ever. It will lie elsewhere: in the trees harms of forestry. Multilateral develop- regulatory treaty. spared by more efficient forestry. An ment funders such as the World Bank A better approach would begin by industry that draws from planted for- have added the protection of forests adopting a nonbinding but clear, quan- ests rather than cutting from the wild and their role in alleviating human titative, measurable goal: namely, a for- will disturb only one-fifth or less of the poverty to their agendas. The United est estate expanded by 200 million area for the same volume of wood. In- Nations engages forestry issues through hectares in 2050 and in which a smart, stead of logging half the world’s forests, the FAO and the ongoing effort to sustainable forestry industry concen- humanity can leave almost 90% of implement commitments made at the trates on little more than 10% of the them minimally disturbed. And nearly 1992 Earth Summit in Rio de Janeiro forested area. This “90:10” vision all new tree plantations are established (at which forestry policies were hotly would serve to anchor and focus a bot- on abandoned croplands, which are contested). Since Rio, an alphabet soup tom-up process through which govern- already abundant and accessible. of panels, forums, and task forces on ments and stakeholders—individually forests have filled U.N. meeting rooms. and collectively—would explore the This year, the U.N. launched an an- actions they must take to achieve their FOREST-FRIENDLY nual Forum on Forests to provide an goal by 2050. Responses could then outlet for the many clamoring voices. vary as necessary. Some countries, such FOREIGN POLICY Forests do not suffer from a lack of as Brazil and Indonesia, could conclude attention in international politics. that the best way they can contribute Actors in the wood drama can The problem is the absence of a trees to the world balance sheet is by thus take three basic approaches to pre- clear and widely shared goal to guide improving the regulation of their pub- serving and restoring the world’s for- policy. Because the U.N. framework lic lands. Others, such as Chile and ests: lifting crop yields, choosing value includes all nations, forest agendas are New Zealand, could do their part by over volume in making wood products, confused and exceedingly complex, and striving to become industrial wood and concentrating forestry in fast- progress is measured by the placement baskets. Still others, such as Russia,

53 could focus on improving forest insti- eas, is fitfully reported in many places. the United States) to large old-growth tutions. Sten Nilsson of the Interna- All but a few countries lack data and harvesters (Indonesia and Russia). They tional Institute for Applied Systems analysis of milling efficiency. Private include major exporters (Canada and Analysis has shown that Russia has groups, especially commercial firms, Malaysia), the world’s largest importer great potential to spare trees by expos- could fill the gaps. But so far they have of forest products (Japan), and a vari- ing the forest sector to modern mar- had little incentive to do so because no ety of consumer needs and preferences. ket discipline and regulation. guiding forest vision has informed and The list encompasses forest hegemons A bottom-up process is needed focused the policy debate. of every region, and the behavior of because no single set of policy instru- In other examples of international governments, firms, and NGOs in ments is appropriate to all settings. environmental cooperation—such as these nations sets world standards in Factors such as land ownership vary cleaning up the North Sea or combat- forestry. widely. In the United States and most ing acid rain in Europe—clear, ambi- The Forest 14 do not correspond of western Europe, for example, forests tious, and achievable visions backed by to any existing and effective interna- are held mainly in private hands. The data systems have proven to be key to tional institution, so one question will United States alone has ten million success. In those cases, as in forestry be how to convene them. The Group forest owners. Most U.S. industrial today, governments were at first uncer- of 8 (G-8) might act as a catalyst. It wood comes from private land, and tain what they could achieve but were includes 4 of the Forest 14 (Canada, ownership fragments when inheritance keen to make an effort. Nonbinding Japan, Russia, and the United States), splits wood tracts among offspring. In legal frameworks, along with periodic and its other members (France, Ger- this setting, improving environmental performance reviews, facilitated action many, Italy, and the United Kingdom) standards in wood production has re- and learning. Only when governments feel strong public pressure to protect quired certification schemes that are had come to understand what commit- forests. Already, the organization has compatible with private land owner- ments they could realistically imple- focused on forest topics such as illegal ship. Programs such as the voluntary ment did they establish binding trea- logging and counterproductive subsi- “Tree Farm” system of standards have ties to lock in progress. dies. Moreover, the G-8 is the only succeeded in engaging owners of small high-profile international forum— forest parcels who are wary of costly other than the more inclusive Interna- production standards that only large THE FOREST 14 tional Monetary Fund (IMF), World landowners can afford. By contrast, in Bank, and U.N.—that engages Russia, Canada and many developing coun- An effective diplomatic strategy the world’s most forested nation, on tries, governments own forests and use for restoring forests will require adjust- topics important to Moscow. And the concessions to control cutting. In such ing conventional wisdom and updat- G-8 also has experience engaging de- settings, policies should focus on set- ing existing institutions. Leadership by veloping countries—as became evident ting the right standards for granting a set of key countries could substan- last year with the creation of the larger concessions and on the firms that do tially ease the task: Australia, Brazil, G-20 to discuss key global financial the cutting. Canada, China, Finland, India, Indo- and economic issues. The G-8 does not Measuring progress will require a nesia, Japan, Malaysia, New Zealand, have the built-in means to analyze for- better system for tracking and assess- Russia, South Africa, Sweden, and the est issues, but the Forest 14 could en- ment. Data on forest cover already United States. These “Forest 14” con- list its members and other partners abound, but reliability varies by coun- trol two-thirds of the world’s wood- such as the World Bank-WWF Forest try, as do definitions of terms as fun- lands and span diverse forest types and Alliance to sponsor studies in their ar- damental as “forest.” Information on management strategies, from intense eas of comparative advantage—a prac- key elements, such as changes in crop plantations (New Zealand and South tice used effectively for other kinds of and timber yields and production ar- Africa) to mixed use (China, India, and international environmental coopera-

54 tion. Topics would include lifting grain frontier forest is still protected by eco- is produced “sustainably.” So far, only yields, setting goals and requirements nomic factors—remote locations and a tiny fraction of production forests for high-yield forest plantations, craft- unfavorable terrain keep farmers and have been so certified, and most con- ing strategies for increasing the effi- lumberjacks at a distance. But threats sumers have refused to pay extra for ciency of milling, examining the poten- multiply where roads and rails pen- “green” wood. But certification is gath- tial for recycling and substituting other etrate, bringing saws to trees and tim- ering force; standards established over materials for wood, creating programs ber to markets. Revenue from the next few years may lock in forest to raise the regulatory capacity needed ecotourism may help preserve forests, practices for decades. These standards to stem illegal logging, and eliminat- as might schemes to value forests’ con- should be set with the path to long- ing subsidies that perversely effect tribution to the ecosystem (such as term restoration in mind. In principle, wood production and use. their climate-cooling sequestration of the leading certification system—the As the stakeholders debate the vi- carbon). Forest Stewardship Council—is com- sion of a Great Restoration, they will patible with such a goal, but efforts are clarify the needed complementary poli- needed to demonstrate that economi- cies and programs. One such require- COMMON CAUSE cally feasible certification can favor ment is better strategies for dealing high-yield growth. Certification that with the vast areas that lie “in the For the great restoration to suc- favors low-yield strategies may produce middle”—lands that are not under in- ceed, farmers, foresters, and environ- a happy tree but lead to a small forest. tensive cultivation or wood production mentalists must recognize their com- The certification debate under- but are also not formal, strictly pro- mon interest in high-yield production. scores the fact that no single approach tected nature areas. To date, much of Those concerned with forests have tra- is enough for achieving the Restoration the debate over protecting forests and ditionally viewed farmers as part of the by 2050. Policy must exert leverage in wilderness has focused on formally de- problem. But by lifting yields, farmers all areas: adopting new technologies marcated and legally protected areas. can be part of the solution. Brussels and practices to improve forestry and Such protection rightly safeguards and Washington can help matters by agriculture, building a better informa- Earth’s greatest forest treasures, but for- paying farmers to grow forests instead tion system, and launching a bottom- mal protection holds little promise for of paying them not to grow food. up process for translating the grand most of the world’s woodlands. Today, Meanwhile, foresters are wary of envi- vision of the Great Restoration into only about 8% of world’s forests are ronmentalists who, they fear, seek to detailed strategies. Realistically, one formally protected in parks. Many gov- make forestry unprofitable and to fence cannot expect all nations to come on ernments hesitate to expand formal off every parcel of land that can be board at once. But surely 14 countries protection, for fear of locking away freed from production. Environmental- can take the process seriously. With land that might serve other purposes. ists, in turn, accuse foresters of destroy- them in the lead, the rest will follow. In many settings, forest dwellers also ing diversity, polluting the land, and Although 2050 remains distant, resist “protecting” their forests because displacing local people. But Big Tim- most elements of the plan need to be well-meaning but ham-fisted govern- ber and Big Green can and must learn put in place in half that time—by ments have tried to secure forests in to meet each other’s core concerns. 2025. Trees are slow growers, and so their natural state by banning long- The conflict between these groups the saplings that will deliver nearly all standing local practices such as hunt- is especially evident in the effort the 2 billion cubic meters of wood ing and small-scale forestry. launched by the environmental com- needed in 2050 must start growing 20 Another critical need is to find munity—and by some forest-products to 25 years earlier. The year 2040 ways to assign economic value to companies, mainly in Sweden, that al- might suffice as a start date for some standing forests (other than as cut tim- ready meet extremely tight environ- fast-growing trees (such as eucalyptus ber). Most of the world’s untouched mental standards—to certify wood that and poplar), but even plantations of

55 those trees will require investments in achieve no net loss in their forests. mills and other assets that are best Some cutting of natural woods may planned and built gradually and well continue, but it will be offset by resur- in advance. gent forests growing on liberated farm To achieve all of this by 2025 will and timber lands. By 2025, the Forest require meeting even more immediate 14 can promise that there will be no goals. Over the next five years, the For- more loss of natural forests, including est 14 should adopt a draft strategy the large tracts of frontier forests that along the lines laid out above, which are nature’s vital legacy. will help focus subsequent debates over Neither feeding the world popu- policy. And they must start the decade- lation nor supplying timber and pulp long process of building the data col- requires the world forest estate to lection and analysis system necessary shrink, as it has ever since ancient civi- for bottom-up assessments of national lizations felled their forests to smelt, forest policies. In parallel, they should build, heat, and cook. Rather, while start measuring overall progress. Will profitably meeting growing demand for demand for cut wood really reach 2 wood products, humanity can vastly billion cubic meters by 2050? If wood increase the area of forests and simul- consumption does not level out at 2 taneously reduce the amount of those billion cubic meters per year—perhaps forests that is disturbed. Such a Great because of rising demand for paper— Restoration is truly a worthy goal for can foresters lift yields more rapidly to the landscape of the new millennium. compensate? Are crop yields rising at the 2% per year needed to liberate 200 million hectares of agricultural land for forests? Are wood yields rising rapidly enough so that the planted forests of 2025 will average 5 cubic meters’ growth per hectare? Are forestry firms expanding plantations at about 2% per year—a rate consistent with historical patterns and sufficiently rapid to deliver enough planted wood by 2050? Are countries implementing policies to help the liberated land recover and to protect the forests still not cut? News reports and publicity along the way can help realize the vision. Benchmarks set and accomplishments achieved should be well publicized to make the reality and significance of the Great Restoration apparent to all. Within the next decade, the 14 nations that lead the effort should manage to

56 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 57-61. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Biotechnology and the Forest Products Industry

Alan A. Lucier Maud Hinchee Rex B. McCullough

ABSTRACT

Forest biotechnology has great potential to produce important benefits for the forest products industry and the general public. Benefits to industry may include higher value raw materials, lower manufacturing costs, and further improvements in environmental performance. Potential public benefits include new supplies of renewable energy and materials; effective new options for solving difficult problems in environmental management and ecological restoration; and new opportunities for employment and sustainable development in an industry based on renewable resources. The potential benefits of forest biotechnology justify accelerated efforts to advance the underlying science; develop and test promising applications; evaluate ecological risks and social concerns; and develop appropriate policy frameworks. Alan Lucier is Senior Vice he forest products industry is large and complex. It employs millions of President with the Na- people with diverse skills at locations around the world. The industry’s tional Council for Air T products (Table 1) help meet important human needs for such things as housing, information, packaging, and personal hygiene. Improvement, Inc., Re- There are good reasons for optimism about the future of the forest products industry. World demand for forest products will increase substantially with in- search Triangle Park, NC. creases in population and economic prosperity. Moreover, wood has inherent [email protected] environmental advantages relative to other raw materials. For example:

•Economic demand for wood provides important incentives for afforestation, Maud Hinchee is Chief reforestation, and sustainable forest management. Technology Officer with •Most products made from wood are renewable, recyclable, and require less ArborGen, Summerville, fossil energy to manufacture than competing materials. SC. •Residuals from wood processing are important sources of renewable energy. The forest products industry is already the world leader in biomass energy [email protected] production, and will probably increase its production substantially if new technologies (e.g., biomass gasification) are successful. Rex McCullough will retire Optimism about the industry’s future is tempered by serious challenges. The in October 2001 from the land base for future wood production will be constrained severely by competing land uses (e.g., agriculture, residential development, and wilderness). In addition, position of Vice President the industry is contending with dynamic and difficult market conditions; a large of Forestry Research with and growing number of government regulations with major impacts on the industry; and important stakeholder initiatives such as forest certification. Weyerhaeuser Company, The industry’s wood supply challenge can be overcome by increasing production on lands well suited to intensive silviculture and by developing land- Tacoma, WA. scape management strategies that improve the overall condition of forest ecosys- [email protected] tems. Forest managers are making substantial progress in these directions by implementing technologies such as tree improvement, weed control, wildlife manage-

57 ment, landscape design, and many others. Effective integration Table 1. Examples of forest products that help meet important of technology options with economic, ecological, and social ob- human needs. jectives is one of the forest industry’s top priorities and the es- Forest Resources sence of sustainable forestry. • sawlogs, pulpwood, fuel wood Manufacturing facilities in the forest products industry are • recreation opportunities technologically diverse and operate in many different countries, • ecosystem services such as water purification, carbon climates, and markets. General concerns include high capital sequestration, wildlife habitat costs, low commodity prices, and a complex array of environ- Building Materials mental and energy issues. The industry has made substantial • framing lumber, structural panels, siding, beams, floor progress in environmental and energy performance, but still faces joists, roof trusses, interior paneling major challenges in these areas. • paper components of wall board, counters, and insulation Biotechnology can help the forest products industry over- Communication Papers come some of its most important challenges. In this paper, we • books, newspapers, magazines •office papers, stationary, school and note pads, drawing describe benefits of potential biotech applications in the industry’s paper forestry and manufacturing operations. We also discuss obstacles • greeting cards, poster and display boards to progress in forest biotechnology generally and some promis- Packaging ing paths forward. • boxes, bags, drums, tubes, spools, cores • paperboard for food packaging, milk cartons, juice cartons • pallets, wood shipping containers BIOTECHNOLOGY AND TREE Tissue and Absorbent Fibers MPROVEMENT • personal hygiene products I • paper towels • diapers The productivity and quality of agricultural crops have been • convalescent bed pads greatly improved by centuries of breeding, testing, and genetic Specialty Cellulose selection. Modern crop varieties are much better sources of food • acetate textile fibers and fiber than their wild ancestors, and they greatly reduce the • photographic films amount of land that must be cultivated to meet human needs. • plastics, pharmaceuticals, food products In comparison to agricultural crops, trees planted for wood • thickeners for oil drilling muds production are wild, undomesticated plants. Most efforts to • rayon for tires and industrial hoses improve trees for wood production have been underway for less Other Forest Products than 50 years. Initial results are scientifically and economically • steam & electricity from biomass fuels important, and they confirm expectations based on agricultural • Christmas trees experience that tree species have enormous genetic potential that • envelopes, labels, file folders could be expressed in valuable new varieties. • toys, decorations, sporting goods • mulch, compost, wood ash, and other soil amendments Progress in forest tree improvement has been constrained • railroad ties, utility poles by various difficulties inherent in tree breeding and propagation. • landscaping timbers, fence posts These include (a) need for multi-year progeny tests; (b) multi- • disposable cups & plates, take-out food containers year delays from seed germination to flowering; (c) self incom- • furniture, tool handles, musical instruments patibility in important species (no inbred lines); and (d) biological • specialty chemicals, fragrances and economic obstacles to large-scale vegetative propagation of superior lines (especially in conifers). Biotechnology has great potential to accelerate tree improvement and enable production of higher-value raw materials for the forest products industry. Key technologies include (a) ad- 58 vanced breeding strategies based on Many other benefits from biotech- ers have been disappointing for various marker-aided selection; (b) improve- nology and tree breeding are possible. reasons—most notably high capital ments in vegetative propagation based For example: costs (especially in the pulp and paper on somatic embryogenesis and/or or- sector) and intense price competition •Pest management strategies based ganogenesis; and (c) rapid introduction in the industry’s commodity markets. on improvements in the genetic re- of valuable traits into superior Disappointing financial returns, sistance of trees and reduced quan- germplasm by genetic engineering. coupled with general economic global- tities of insecticides and fungicides. Acceleration of tree improvement ization, are driving a dramatic restruc- through biotechnology will enable sub- •Ecological restoration strategies en- turing of the industry. Mergers and stantial increases in tree growth rates abled by genetic engineering of tree acquisitions are producing a few glo- on sites close to mills. When ready for species devastated by exotic diseases bal-scale competitors and creating harvest, these sites will yield large num- (e.g., American chestnut). niches for new smaller-scale companies. bers of uniform stems per hectare. Although financial and market is- High yields will reduce harvesting costs • Carbon sequestration, soil reclama- sues are dominant near-term concerns, and the area of forest land required to tion, and bioremediation strategies industry leaders have keen interests in meet mill demands for raw material. enabled by new trees capable of tol- the potential of technology to reduce Short-haul distances to mills will re- erating poor soil conditions such as manufacturing costs and create new duce log transportation costs. Efficien- drought and chemical contamina- products. Biotechnology in particular cies in harvesting and transportation tion. has enormous potential. In addition to will reduce fossil fuel consumption and •New strategies for sustainable pro- improving the quality and quantity of CO emissions associated with raw raw material supplies, biotechnology 2 duction of valuable chemicals in material acquisition. trees based on genetic engineering could have radical impacts on pulping Faster growth is important, but is of secondary metabolic pathways. processes, waste-to-energy systems, and only one of the potential benefits of other aspects of forest products manu- accelerated tree improvement via bio- facturing. For example: technology. For example, new tree va- BIOTECHNOLOGY AND •Biotechnology could enable the de- rieties with special wood properties will velopment of new pulping processes enable more rapid development of raw FOREST PRODUCTS based on selective enzymatic cleav- material supplies tailored to the re- MANUFACTURING age of lignin polymers. Potential quirements of manufacturing processes. benefits include lower capital costs, Improvements in raw material quality The forest products industry is higher product quality, and lower will allow mills to reduce manufactur- under great financial pressure at consumption of both chemicals and ing costs and improve product quality. present. Overall returns to sharehold- energy. See Table 2 for examples. •Biotechnology could enable the de- Table 2. Examples of wood quality improvements and benefits that might be achieved velopment of new systems for con- through biotechnology. verting organic residuals into Wood quality improvements Potential benefits bioenergy. Potential benefits include lower costs for solid waste manage- Smaller core of juvenile wood Greater lumber strength and stability ment and reduced need for fossil Higher specific gravity Higher pulp yields relative to inputs of energy energy. and chemicals in the pulp mill Lower lignin content Reduced inputs of chemicals and energy in pulp bleaching

59 EALIZING THE R Table 3. Forest biotechnology projects supported through Agenda 2020. POTENTIAL OF FOREST Principal Project title Lead institution BIOTECHNOLOGY Investigator Brunner Dominant negative mutations Oregon State University During the past century, the for- of floral genes for engineering sterility est sector has made great progress in Chang Exploiting genetic variation of fiber North Carolina State University developing better systems for growing components and morphology in and harvesting trees, making and dis- juvenile loblolly pine tributing products, and reducing envi- Davis Molecular physiology University of Florida ronmental impacts. Progress has been of nitrogen allocation in poplar enabled by research, development, and Davis Molecular determinants of University of Florida integration of technologies as diverse as carbon sink strength in wood forest regeneration, landscape manage- Li Search for major genes using North Carolina State ment, chemical and material recycling, progeny test data to accelerate University and biological treatment of wastewater. development of superior loblolly Biotechnology is poised to make pine plantations significant contributions in various sys- Neale Genetic marker and quantitative USDA Forest Service trait loci mapping for wood quality tems in the forest sector. Realizing the traits in loblolly pine and hybrid poplars great potential of forest biotechnology Peter Accelerated stem growth rates and Institute of Paper will be an enormous and exciting chal- improved fiber properties of loblolly pine Science & Technology lenge. The rate of progress will depend Pullman Trees containing built-in pulping catalysts Institute of Paper on science and technology factors in- Science & Technology teracting with social, economic, and Tsai Genetic augmentation of syringyl Michigan Technological University political issues. lignin in low-lignin aspen trees Inadequate government support Tschaplinski Biochemical and molecular Oak Ridge National Lab for pre-competitive research is an im- regulation of crown architecture portant obstacle to progress in forest Tuskan Marker-aided selection for wood Oak Ridge National Lab biotechnology. Through its Agenda properties in loblolly & hybrid poplar 2020 program, the forest products in- Whetten Pine gene discovery project North Carolina State University dustry has suggested priorities for pre- Williams QTL and candidate genes for Texas A&M University competitive research and provided growth traits in Pinus taeda L. funding for several projects in partnership with the U.S. Department of En- ergy and the Forest Service (Table 3). trols of key processes such as wood for- tant. Informed discussions, research, [MS 9 (Lucier) table 3 near here] mation are formidable tasks that will and collaborations involving diverse Agenda 2020 and other programs take decades at current rates of parties are needed to better define is- are supporting valuable projects, but progress. A major initiative is needed sues and potential solutions. The new the low overall level of funding for pre- to accelerate pre-competitive research Institute of Forest Biotechnology competitive research is a critical limit- in these areas. (www.forestbiotech.org) will have an ing factor in forest biotechnology. The ecological, social, and policy important role in bringing diverse par- Mapping the genomes of model tree issues associated with forest biotechnol- ties together and organizing necessary species and discovering molecular con- ogy are complex and extremely impor- activities.

60 CONCLUSIONS ACKNOWLEDGEMENTS

Forest biotechnology holds impor- The authors thank Dawn Parks at tant opportunities and challenges for Westvaco Corporation for helpful com- the forest products industry. The ments and suggestions. industry’s technology leaders appreciate the economic potential of forest biotechnology and have diverse views on critical issues such as time to commercialization and risk management strategies. The future of biotechnology and its value to the forest products industry will be affected greatly by public perceptions of social and ecological issues. We believe the potential benefits of forest biotechnology justify greater public support for pre-competitive research to advance the science; develop and test promising applications; evaluate ecological risks and social concerns; and develop appropriate policy frameworks. In this paper, we have outlined some promising applications of forest biotechnology with emphasis on their possible value to the forest products industry. Potential benefits to the public are also substantial. They include (a) new supplies of renewable energy and materials; (b) effective new options for solving difficult problems in environmental management and ecological restoration; and (c) new opportunities for employment and sustainable development in an industry based on renewable resources.

61 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 62-69. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Responding to New Trees and to the Issues at Hand: The Institute of Forest Biotechnology

W. Steven Burke

ABSTRACT

Effective and thoughtful development, application, and acceptance of complex new technologies require many steps and participants. The movement from science and research to products and public impels attention to three areas simultaneously: to science and research; to industry and products; and to a richly complicated mixture of societal, ethical, environmental, regulatory, and public issues. The academic and industrial forest endeavor, historically not technology-intensive, is brought to new challenges by the process and issues of biotechnology development. Application of biotechnology to trees and forests is, moreover, particularly challenging because of their extraordinary importance to human development, culture, values, and economies; trees are, after all, the only plant or crop to which large numbers of people have routinely ascribed moral value. As a result, a full and rigorous societal dialogue involving all parties attentive from whatever vantage point to forest biotechnology—scientists and researchers, industry, public interest groups and ethicists, consumers and policy-makers—is requisite for considered and effective application of the technology to trees. Bringing about such engagement will require new strategies, engagement among parties varying in values and agenda, consensus and shared ground if possible, and sustained effort. Gaining such outcomes will be as important as demanding. The Institute of Forest Biotechnology, a private non-profit corporation, was established in 2000 in Research Triangle Park, North Carolina, to work for this engagement. Not a site for research, the Institute will bring diverse parties together to address—through projects, meetings, and publications—the research, scientific, industry, societal, and economic issues of forest biotechnology worldwide.

hree reasonable assertions offer a framework for thought: •Trees are the only plant routinely ascribed intrinsic moral value by large T numbers of people. •Biotechnology will change some trees, and be applied worldwide in coming years. •Trees from technology will seem, to many, manifestly different from W. Steven Burke is trees of tradition. Senior Vice President, How can we characterize an endeavor characterized by these—and other, equally complicated—assertions? Why is forest biotechnology so very rich in is- Corporate Affairs and sues as well as potential, in implications as well as complexity? External Relations North How do we feel about trees? How do we feel about transgenic trees? We know how we feel about trees. Quite sensibly and understandably, we Carolina Biotechnology love them, with an atavistic fervor rooted in something not easily defined. Center, Research Tri- We are perhaps uncertain of our responses to transgenic trees and forest biotechnology, for the endeavor comes new to our attention. angle Park, NC. How can we prepare our work, activities, and policies for the increasing development and use of forest biotechnology over coming decades? Doing so is req- [email protected] uisite, as forest biotechnology worldwide will prove a complicated combination 62 of technology, imperative, and societal of societal and economic return from juncture, for sustained usually large issues. Are we ready? How can we further exploration and eventual devel- investment. In stage 7, testing and tri- think about trees, about forest biotech- opment of promising research? Return als establish the safety and effectiveness nology, and about strategies of re- of both sorts is of course requisite, in of the new product. Stage 8, is manu- sponse? some balance determined appropriate facturing or growing. Stage 9 proves I offer a framework for our think- by varied parties. The forest industry increasingly important: adoption by ing, under four necessarily interrelated worldwide is, at present, very much existing companies or intermediates of headings: engaged in determining whether suffi- the new product or application, a de- cient return can over time be gained cision of use by the food processing or 1. Thinking about Technology to merit substantial investment in for- healthcare or pulping industries. Stage 2. Thinking about Trees est biotechnology. 10 reveals final acceptance and use, at The continuum from discovery to the consumer as well as societal levels. 3. Representative Forest Biotechnology application is a combination of stages, Stage 11 is the key juncture that we Issues participants and resources, and issues. It seem too seldom to address, compelled 4. The Newly Established Institute of also requires attention to the context as we are by the demands of the im- Forest Biotechnology. and imperative of biotechnology devel- mediate: charting the future. opment. It is difficult to imagine an area of biotechnology as demanding of future THINKING ABOUT Eleven Broad Stages charting—of a thoughtful and pur- poseful analysis of the horizon—as for- TECHNOLOGY The process involves many differ- estry and trees. ent steps, each different in require- To consider the trees of technol- ments, participants, length, and out- ogy, we must have a reasonable under- Participants and comes. Eleven broad stages are key, standing of technology development, largely constant and common, and Resources the context in which forest biotechnol- must be addressed by a combination of ogy grows. Technology development is The process and stages of technol- vision, funding, activities, and policy. a process—a continuum—a movement ogy development are shaped and moved They can be laid out in a suggestive, from societal idea and need to prod- along by varied persons and entities, pri- but not exact, schema. uct and societal impact. The movement vate and public. Some participants are Science and research, stage 1, pro- is deliberate, long-term, sequential, and directly and always required to move vide the requisite foundation. Stage 2, complicated. It can be easily broken biotechnology from science to product: policy and impetus, impels the decision down into three main phases: researchers; universities and laboratories; to commit to further development. small and large companies; entrepreneurs • Discovery, a matter largely of science Technology transfer proves an increas- and risk-takers; investors; regulators; in- and research ingly crucial stage 3, enabling the termediate users or adopters; trained movement of promising ideas from the •Development, in the main the realm workers; manufacturers or growers; ethi- research to the private sectors. In stage of industry in free market societies cists; educators; catalysts, policy-makers 4, application and product possibilities and committed governments; and ac- •Application, eventual societal utili- are actively explored, as foundational cepting users. zation and, we hope, beneficial im- impetus for stage 5: involvement or Other participants support and pact. formation of companies working to facilitate the process. While not re- A prime question must be very move the possibility to commercial required, their presence and thoughtful soon addressed towards the end of the ality. Stage 6 manifests the ongoing participation often accelerates results. discovery phase: what is the likelihood imperative, particularly key at this Among such entities are administra-

63 tors; technology transfer officers; bio- manifestly changing living organisms Lack of required resources and par- technology centers or initiatives; incu- would not yield issues? ticipants at any stage can slow or even bators and research parks; varied gov- The issues of biotechnology are stop the process. Without, for instance, ernment agencies; policy-makers with societal or policy, ethical or personal— trained researchers, or risk-taking com- a long-term view; experienced manag- or, in most cases, an overlapping mix- panies, or accepting users, technology ers able to bring experience to a sec- ture. They are generally interesting and development is less likely to come about. ond or third project, company, or tech- usually consequential as well as numer- Failure to address key issues or challenges nological challenge; and good critics. ous. A short representative list of the at any stage can also slow or even stop Other participants are indirectly broad issues at hand can reasonably the process, lessening acceptance as well involved, or skeptical, or even possibly include questions about the very rea- as economic and societal gain. Here are hostile, but they also shape the process son for undertaking biotechnology; two easy examples: First, insufficient of technology development, for they policy and commitment; technology funding for science at the beginning of shape the nature and terms of the transfer; safety and risk; who benefits; the process can truncate a vital move- movement from science to public. regulations; labeling; public acceptance; ment. Is funding for biotechnology in These parties are varied in agenda and the morality of it all; use of technol- trees sufficient to ensure good research approach, and include researchers in ogy in general, particularly that based and good analysis of outcomes? Second, other fields; ethicists and philosophers; on living organisms; trade and inter- insufficient preparation for public re- government agencies; investors and national implications; sources of capi- sponse can curtail the final stage of the funders varied projects, agencies, and tal; process versus product, unexpected process. Was sufficient attention paid institutions; informed thinkers and consequences; evaluation of outcomes; early to the issues of food and agricul- uninformed thinkers; questioners; non- and fear of the new. More specific is- tural biotechnology? Are we preparing government organizations; legislators sues are those of: cloning; stem cells; early and thoughtfully enough for the and policy-makers; other professions; bioterrorism; globalization; altered ecological, societal, and policy issues to public interest groups; users; and—to landscapes; new food; biodiversity; ge- inevitably accompany forest biotechnol- our recent consternation, here in the netic privacy; reducing life to genes; ogy? American Northwest, directed to rights of animals; rights of plants; con- The reality and implications of the trees—terrorists. trol over nature; genetic transference; process of technology development are These participants vary enor- ownership of germplasm; and increasingly apparent for the forestry mously in training, values, and expec- xenotransplantation. endeavor. Development and explora- tations—and, as such, in their response While this process—combining tion of biotechnology make the forestry to stages, outcomes, and the issues at- stages, participants, and issues—is endeavor now more a technology-di- tendant to the process. complicated, the key point is simple: rected enterprise than one shaped by for effective, thoughtful, and appropri- its traditional approach and slower- Intrinsic Issues ate biotechnology development, in any moving results. Traditionally and his- sector, this process and its attendant torically not technology intensive, the Issues are intrinsic to the process imperatives must always be understood forestry endeavor has brought biotech- of biotechnology development. Some and always addressed. The reason is nology to the challenges of this process are expected and follow logically from clear. For movement along the con- and of its issues. earlier experience; others are new and tinuum to be successful and acceptable, Understanding the process of largely without precedent. They vary a sector—such as forestry—or even a technology development is thus in- by stage and by participant point of single company must gain or involve creasingly required for participants in view. They demand our considered at- all participants, must move through all forest biotechnology. The trees of tech- tention and should not occasion our steps, and must address all appropri- nology are not—in genesis, develop- surprise. Is it likely that a technology ate issues. ment, and attendant issues—exactly

64 like the trees of tradition. Reflecting industry, and society in probably address all of some of the issues. the complicated process from which roughly equal measure. Worldwide, in fact, we see substantial they spring, they are also more diffi- No earlier technology has from its differences in underlying values, in is- cult to bring about. onset so required this imperative; none, sues judged important, in ethical certainly, has yielded comparable delib- frameworks, and in commitment to The Context of Technology erate attention in all three areas simul- measured public discourse. taneously. Such attention is as much The imperative to effectively, re- Development societal responsibility as necessary strat- alistically, and credibly anticipate and The process does not take place in egy. Experience reveals that we have too address the issues of biotechnology is a void. Technologies do develop within often addressed the ethical and societal enormous and cannot be questioned. the life, culture, and values of a soci- implications of technology too late, at Doing so is neither academic nor a ety . . . within the zeitgeist, as the Ger- the end of the process, as a last thought luxury, but is instead the sine qua non mans so nicely characterize this com- if not as an after-thought. Sometimes of movement from science to public. bination of time and spirit. This sur- this has been done with admirable in- Analysis and resolution of key issues rounding context is shaped by history, tentions, and sometimes more to in- will be required in some countries if tradition, relation, values, and expec- duce public acceptance. certain research and applications are to tations, of individuals as well as of so- We must do better with biotech- move forward. It is quite possible that ciety. This cultural and societal context nology, giving attention as appropriate certain products and applications will directly or indirectly affects our ap- to numerous, difficult, and often very be philosophically vetted, rather than proach to technology development new issues at all points along the con- more traditionally evaluated largely in and, simultaneously if not always tinuum of technology development. terms of feasibility and safety. It is, clearly, our responses to it. This con- Identifying, understanding, and ad- moreover, possible that certain out- text shapes research and product pri- dressing different issues at different comes, results, or applications will ul- orities, attitudes to risk taking, polices stags is enormously challenging. Our timately not be developed, acceptable, and issues, public and institutional re- intentions are good; the community or used. sponses, and funding decisions. working for biotechnology has proven Such considerations will apply, to Thoughtful awareness of this con- remarkably aware of its full societal re- as yet undetermined degrees, to forest text, and realistic attention to it, is responsibilities. However, practical reali- biotechnology, as to food biotechnol- quired for effective, long-term biotech- ties often make difficult our attention ogy, use of stem cells, and human clon- nology development and use. Trees in to the issues at hand. First, participants ing. Accordingly, the imperative to particular demand such awareness and vary greatly along the continuum, in thoughtfully prepare for the issues at attention, for our responses to trees are their tasks, vantage points, expecta- hand is strong in particular for all per- strongly shaped by intricate cultural, tions, and values. Second, few partici- sons and institutions applying new bio- personal, visual, and historical factors. pants are appropriately trained, par- logical science to trees and forests. ticularly in ethical evaluations. Third, The Imperative of stages and results are usually separated by time, place, and participants. Re- THINKING ABOUT Biotechnology sults or implications later in the pro- TREES Development cess cannot be easily anticipated or controlled in earlier stages. Finally— We probably know more about Because shaped by issues and po- and perhaps most important in general trees than about the unfolding process sitioned fully within society, biotech- as well as specifically in relation to for- of biotechnology development. nology development thus requires at- est biotechnology—there is worldwide Trees are profoundly important. tention—from the beginning of the no common imperative to identify and They are requisite for life on this process and continually—to science, 65 planet, key to environment and ecol- endeavor, and somehow shaped to safe, The ambiguity often implicit in ogy. They are, and always have been, appropriate, and acceptable ends. ethical decision-making about tech- requisite for civilization, key to human Forest biotechnology is barely ex- nology might prove particularly vex- and societal development. They create plored, largely just beginning its move- ing in forest biotechnology. How is one of the world’s largest and most ment along the continuum from re- a determination made between (a) important economic sectors. They have search and science to applications and the undisputed need to grow more extraordinarily wide mythic, symbolic, society. As a result, issues, questions, trees on less land and (b) possible religious, and historical resonance. and implications can be better identi- displeasure and risks attendant to Trees have greater impact on cul- fied and addressed early; the varied widespread plantations of transgenic ture and consciousness than any other participants can, and should, be trees? crop or plant. As noted earlier, most brought early to the requirement do- • The environmental and eco- people give an intrinsic moral value as ing so. Doing so will be demanding, nomic value of trees altered to well as actual value to trees. Behaviors for the intrinsic uncertainties of this have less lignin content and pos- as well as policy reflect our value-based new technology in general are further sible related stresses and out- responses to trees worldwide. Human affected by the compelling importance comes to trees so altered. responses to threats or loss are passion- of trees to our consciousness and our ate, emotional, and often shaped by planet. • The need to grow altered trees ef- barely conscious imperatives. Forests The environmental, societal, ficiently in controlled settings are preserved by policy in richer or policy, and ethical questions arising and the diminishment or loss of more enlightened countries. Protecting from forest biotechnology will be con- some forests as rich ecological en- the landscape, in which trees are key, sequential. Each is complicated and vironments. has a moral imperative in a growing not easily addressed, combining scien- number of places. tific, industry, and public imperatives. • The imperative that transgenic It is therefore difficult to imagine a A few can be listed as representative; trees do not flower or reproduce more societally challenging global issue there are more. (for improved productivity or to than genetic engineering of forest trees. prevent genetic transference) and • Concerning the initial technological the ecological benefits of flower- imperative: Why undertake forest ing to the environmental and to REPRESENTATIVE biotechnology at all? other organisms. FOREST • Concerning a more realistic impera- • Concerning the human imperative to tive: How can forest biotechnol- BIOTECHNOLOGY work for improvement and survival: In ogy be thoughtfully developed and the face of varied undisputed factors ISSUES appropriately applied? How can (including land limitations and the careful attention to questions and ever-present need for wood prod- This challenge is understandable. issues at all stages be assured? ucts), is it unethical not to develop Merging the process of biotechnology and apply forest biotechnology? development with trees and forests • Concerning the derived benefit: Who worldwide will yield a rich mixture of will predominantly benefit from for- • Concerning the state of trees: Largely questions, implications, and issues—as est biotechnology? Will different ben- non-altered, the wildness of trees is much societal and ethical as scientific efits be gained by different parties, in remarkable, contrasting with centu- or economic. Science and research, a reasonable balance? ries of deliberate alteration of other regulations and risk analysis, industrial key species, and conveys a large part •Concerning the inevitable tension be- and societal priorities, environment of their appeal. It also suggests pos- tween perceived benefits and perceived and ecology, must all be brought to the sibly easier genetic transference be- liabilities: 66 tween altered and non-altered forest biotechnology gain germ- firmly held often inchoate passion stands. plasm and economic benefits for a about “natural” trees exist with a few nations, at the expense of those realistic recognition that they are •Concerning the long life and large pres- less sophisticated or less economi- necessarily, in some cases, resources ence of trees: How can potential envi- cally developed? to be altered? How can understand- ronmental outcomes be anticipated able emotion find balance with over many years? How might genetic • Concerning the landscape: Honoring practical technology? alternations prove unstable or yield and preserving the landscape is a unexpected changes over time? moral imperative. So is improving it, Persons and places attentive to but agreement is less clear on accept- trees—and forest biotechnology— • Concerning the tree versus the forest: able means. Do genetically altered worldwide must begin to address such How do quantity and clustering trees violate this imperative, in large questions. How can this be effectively matter? Are fewer genetically altered or small numbers? Other alternations done? What framework or resources trees (here and there, in parks and to trees do not seem to do so. can assist? orchards, in your back garden) more acceptable than large numbers • Concerning the many, varied, and com- neatly arrayed? Why? plicated environmental implications to HE NSTITUTE OF which attention must be paid: Envi- T I • Concerning the type of altered tree: ronmental and ecological questions FOREST Are apple trees immune to fungus are inherently as much ethical as sci- in northern Europe, or regained IOTECHNOLOGY entific, and should be judged as such. B American elms, more acceptable Avoidable harm to the environment than pines altered for quicker tim- Answers, strategies, and assistance and living organisms is a prime moral ber production? Why? will be required for many years. Over failing. coming decades, without question, for- • Concerning managed tree plantations: • Concerning the ethical slippery slope: est biotechnology worldwide will mix Intensively managed plantations Threatened or diminished tree spe- science, industry, society, technological increase steadily worldwide, with cies can realistically be regained in process, participants different in tasks good reasons, but often yield re- time through biotechnology. Does and agenda, and layered issues. This sponses different than for other this prepare somehow for regaining complexity must be granted, discussed, large crop plantings. Will forest bio- other, non-crop, species? and somehow addressed by the widely technology plantations yield even divergent parties attentive to forest bio- more acute responses? • Concerning dialogue and engagement: technology. Ethical standards are needed for de- • Concerning variable regulatory and The task is demanding and con- velopment and application, but also policy frameworks worldwide: Coun- sequential. Remarkably, no entity had for discourse and opposition. Parties tries vary in their attention to test- until recently been established to ad- reflexively polar (on either side) can ing and trials, as well as their atten- dress the task and the challenges. This probably be discounted. tion to ethical guidelines, the envi- void was seen as surprising by persons ronment, and public discussion. • Concerning the forests on which life attentive to forest biotechnology, and Forest biotechnology will likely be depends: What do we expect of our also as a liability. The absence of a early applied in countries with less forests? How do we define—or re- strong central voice for forest biotech- strong imperatives or experience in fine—the natures, outcomes, and nology lessened the likelihood that these areas. uses of a forest? policy and issues worldwide will be addressed with appropriate strengthen, • Concerning the status of countries in- • Concerning the passion that trees so credibility, and thoughtfulness. volved in forest biotechnology: Will understandably induce: How can 67 Responding to this absence, a di- dard problem-solving, and innovative Governance verse committee of over 25 persons was partnerships. Accordingly, the Institute brought together by the North Caro- will, through activities, governance, Twenty board members will mani- lina Biotechnology Center, catalyzed by and philosophy, fest the imperative for varied voices, a reasonable premise: forest biotechnol- representing three main groups in •Serve varied parties attentive to for- ogy could be assisted from the onset roughly equal balance: public interest est biotechnology, within the pro- by an organization directed to partner- and non-governmental, academic and cess and within the larger societal ship, the issues at hand, and multiple governmental, and industry and indus- environment, as convener, problem- vantage points. The group—represent- try-related. The first 10 board members solver, common ground, and part- ing research, policy, academic, public, have been determined: Christine Dean, ner. and corporate interests— worked over Weyerhaeuser; Robert Friedman, the an 18-month period, from 1999 until • Assist existing organizations and H. John Heinz III Center for Science, early 2001. Merging imagination, activities rather than unnecessarily Economics, and the Environment; long-term vision, and common sense, duplicate efforts. Robert Kellison; Lori Knowles, the the group crafted the philosophy, ap- Hastings Center; Dennis LeMaster, proach, and governance of a new en- •Work for activities and decision- Purdue University; Alan Lucier, Na- tity: The Institute of Forest Biotechnol- making informed and balanced by tional Council for Air and Stream Im- ogy. The mission of the newly-estab- diverse voices. provement; John Pait, The Timber lished Institute is bold in nature and Company; Ronald Sederoff, North large in intent: To work for societal, Emphases Carolina State University; Ben Sutton, ecological, and economic benefits from CellFor; and myself. The Institute will direct attention appropriate uses of biotechnology in and activities to three main areas. In forestry worldwide. Science and Research, the Institute will The North Carolina Biotechnol- Expected Key Initial ogy Center has committed initial fund- •Identify key topics for societal, eco- Activities, 2001–2002 ing of over $300,000 to the Institute, logical, and genetic research. which will be housed administratively Administrative goals are to at the Center until resources are avail- •Work for partnerships and funding. •Gain an exceptional Executive able for an independent site. Addi- In the area of Policy the Institute will Director tional funding is sought, and over time must come in appropriate balance from •Identify areas in which forestry, •Move to full 20-member board project, industry, government, and regulatory, technology, or public •Work for short- and long-term foundation sources. The Institute’s first policy is required. funding. employee, Ms. Susan McCord, has ini- • Coalesce partnerships and projects. tiated activities. Communicational activities will initially The sensibility and approach, em- Responding to Societal Imperatives, the phases, governance, and expected key ini- Institute will •Inform parties worldwide about tial activities of the Institute are out- the Institute •Identify key areas of societal, envi- lined below. ronmental, policy, and ethical is- •Gain responses about activities sues. Sensibility and Approach and approach. The challenges of forest biotech- •Develop educational materials, Addressing ecological and ethical is- projects, and multi-party meetings. nology demand imagination, non-stan- sues, the Institute plans to

68 • Commission a scientifically based also valuable in due proportion to the study of the ecological risks as- challenges and issues to be raised in sociated with forest biotechnol- coming decades by the application of ogy biotechnology to trees and forests.

•Sponsor a workshop bringing together diverse parties to shape and address these issues.

The Heritage Trees Program, the Institute’s first program, has a direct premise: the tools of biotechnology can be used to regain or strengthen threatened or diminished species, yielding the best possible combination of scientific, ecological, and societal outcomes. The Program will

• Commission a report to identify key species and what tools of molecular biology can be appropriately brought to bear upon them.

•Sponsor a workshop to coalesce partnerships and to focus efforts of varied groups.

•Develop a peer-reviewed grants program to help fund targeted research.

Information and Technology Transfer activities will

•Develop a short publication introducing forest biotechnology.

•Develop the Institute as a site for information and resources on forest biotechnology.

•Respond to initial requests from institutions and industry for project assistance.

Development of the Institute of Forest Biotechnology is enormously demanding and challenging . . . but

69 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 70-81. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Will the Marketplace See the Sustainable Forest for the Transgenic Trees?

Don S. Doering ABSTRACT

A public or privately financed market for genetically engineered trees depends upon how the technology is applied, who participates in the decisions of commercialization, and society’s general acceptance of the release of genetically engineered organisms into the environment. The international debacle of the introduction of genetically engineered food and fiber crops provides valuable lessons to the nascent tree biotechnology industry that products must have broad social utility, be designed for environmental safety, and be tested for ecological impacts. The introduction of genetically engineered trees can occur via an Don S. Doering is on the appropriate regulatory framework and a collaborative effort of the public and private sector as well as stakeholders from civil society. Meeting the future global needs for fresh senior staff of the World water, biodiversity, materials, energy, habit, and paper will require keeping sight of the Resources Institute’s sustainable forestry goals beyond the transgenic trees.

Sustainable Enterprise orests and tree plantations can sustainably provide all the necessary goods Program, Washington, and services—from timber to protected habitat—that the world wants. F Today we are far from that sustainable ideal. The genetic engineering of DC. trees has the potential to contribute to both sustainable and unsustainable forestry practices. That contribution of genetically engineered trees depends on the [email protected]. private and public use of the technology and how economic markets develop for wild and domesticated trees. Will that marketplace develop and who will shape He is an expert on corpo- its path? rate strategy for sustain- This essay is based on the assumption that genetically engineered trees will be commercialized, but the timing and trajectory of commercial introduction are able business and on the far from certain. Will a market develop in some countries, while intense societal role of biotechnology in opposition closes markets in other regions, as has been the case for genetically engineering food crops? Can we learn from the experience with other genetically sustainable agriculture. modified organisms to improve the chances that public and private genetic engi- He has served as a Senior neering of trees will safely and fairly serve the needs of society? The view presented here, based on observation of crop biotechnology, is that there may be Fellow at the Wharton conditions under which transgenic trees reach the marketplace. But the needed pre-conditions of stakeholder engagement, increasing the social utility of prod- School of Business and ucts, novel partnerships, and design-for-environment call for a new approach to holds a Ph.D. in Biology commercial genetic engineering of plants and a break with the example of first- generation genetically engineered crops. from M.I.T. The opinions expressed herein are GM TREES: HOW AND WHEN, NOT IF those of the author and do not represent the Genetically engineered trees are in the environment. In hundreds of experimental field trials of dozens of species and in dozens of countries, private and opinions of the World public sector scientists are conducting research on transgenic trees. None of these Resources Institute. trials are pre-commercialization, however many may be viewed as commercial prototypes. The cautious but obvious interest of the forest industry and interest 70 among academic scientists in geneti- ogy by its proponents share many simi- comes an advocate of ecosystem resto- cally engineered trees is reminiscent of larities with the aims of crop ration. the early moments of agricultural bio- biotechnologists. As we contemplate Were Dr. Seuss to invent this story technology. As described below, a set increased corporate and public invest- today, he might incorporate powerful of forces lend an air of inevitability to ment in tree biotechnology, it is fair to features of the 21st century market- commercial introduction of genetically wonder how will society benefit and place. This is still a time of powerful engineered trees. how many corporate casualties will multi-national corporations in which Even as genetically engineered there be along the way? supply and the openness of markets are trees are intriguing, it is hard to imag- strongly influenced by corporate inten- ine that a forest industry executive or tions and political influence; technolo- tree biotechnologist would want to see LISTENING TO THE gies are still pushed to the marketplace. product introduction of trees follow Product performance now includes the the example of GE food crops. At first LORAX: CREATING A consumer demand, regulations for glance, introduction of herbicide resis- 21ST CENTURY product lifecycle stewardship, recycling, tant and pesticidal crops is a story of and environmentally friendly design. fantastically fast technology adoption MARKET The ability to sell to society is now and market penetration. Only five governed by demand and the societally years after their introduction, almost The Lorax is a clear and compel- granted license-to-operate. As is so three-quarters of U.S. cotton, over half ling story of unsustainable forest man- richly illustrated by European rejection of U.S. soybean and a fifth of the U.S. agement and unsustainable business of America’s genetically engineered corn crop is planted in GE varieties strategy (Seuss 1971). The book pre- crops, stakeholders exclusive of custom- (Carpenter and Gianessi 2001). With sents a 20th-century view that a mar- ers and shareholders now have the another look, introduction of GE food ket is created by product performance, power to close global markets through crops is a product introduction night- supply, and demand. In the classic protest, political power, and boycott. mare. Agricultural biotechnology children’s story by Dr. Seuss, the Once- Corporations (including some of the CEOs have been fired or replaced, cor- ler uses the fiber of the Truffula Tree largest timber companies and largest porate reputations have been shaken, (species unknown) to knit thneeds, a wood-buyers) have been forced to billions of dollars of shareholder value multi-purpose textile that is “a-fine- change business practice and strategy has been lost, and ag-biotech divisions something-that-all-people-need.” The because of civil society activism. Social of large companies have been swapped, demand is immediate, voracious, and license-to-operate, societal acceptance, closed, and expelled from their parent is served by the Once-ler’s innovation stakeholder engagement, and corporate companies. The societal unity in Eu- that automates thneed production and social responsibility are now central rope in rejection of biotechnology Tr uffala harvesting—over the repeated strategic issues of business leadership crops has slammed closed valuable and furious objections of the Lorax. and not just public relations functions. grain markets and ‘frankenfoods’ have The Lorax is himself a forest resident In today’s update of The Lorax, the become an icon of the global anti-cor- and the spokesperson for other Truffula Lorax would be supported by outside porate and anti-globalization move- ecosystem stakeholders such as the Bar- activists in his thneed opposition, bar- ments. Some of the same companies ba-loots, Swomee Swans, and Hum- ba-loots would be found chained to involved in food crop biotechnology ming Fish. As the thneed industry and trees and to thneed shop shelves, leg- are involved in GE trees, many field tree harvesting degrade the basis of the islation would protect the Swomee trials are testing the same traits of ecosystem, the native species migrate, Swans, and perhaps the story would glyphosate tolerance and Bacillus the resource collapses, and the thneed have a different outcome. thuringiensis (Bt) endotoxin expression, business collapses. The Once-ler is left and the aspirations for tree biotechnol- in financial ruin and—too late—be-

71 THEMES OF THE SOCIAL seen by many) as being potentially environmental record and political in- unsafe for humans, animals, and the fluence; images of clearcuts on public CONTROVERSIES OF environment. The utility of the prod- lands and protesters blocking logging CROP BIOTECHNOLOGY ucts to the consumer and even to the trucks have been potent symbols farmer are questioned. Not only is the throughout the last 30 years of the transparency of companies under fire, environmental movement. A union of Genetically engineered crops uni- but so has been the pro-biotechnology the timber industry and the plant bio- fied many social movements that had stance of the U.S. and European gov- technology industry comes poorly never been so united and so armed ernments and the lack of transparency armed to a battle based upon public with potent symbols such as baby food, in the regulatory process for GE crop trust, and is a dream marriage to an monarch butterflies, and the small- approvals. But perhaps of greatest so- anti-corporate activist (Lenzner and scale farmer. A common cause was cietal distaste to opponents of biotech- Kellner 2000, Sampson and Lohmann found that united individuals and nology is the lack of personal choice 2000). NGOs with interests in food safety and in choosing foods derived from geneti- The agricultural biotechnology consumer choice, farmers’ rights and cally engineered commodities, coupled industry used, and still uses, the prom- property rights, spirituality in a tech- with the perceived unethical and ised benefit of biotechnology for sus- nological world, economic justice and opaque actions of the biotech compa- tainable food supply to earn its social global trade, corporate influence and nies in seemingly imposing the prod- license to operate. In doing so, they economic justice, hunger and the en- ucts upon the market. may have picked the wrong arena for vironment.1 Recognizing this conflu- taking on opposition interests. First- ence of interests is not to deny their generation biotechnology products did validity but it helps to see underlying RUST not address causes of food insecurity themes that merit deep consideration T and were not designed to be grown in by a nascent forest biotechnology in- The benefits of transparency, re- climates or designed for agricultural dustry: individual choice, product spected ethics, and the support of so- systems where there is food insecurity. safety, social utility, transparency, and cial values are captured in the word The arena of food security and sustain- ethics. trust. The ag-biotechnology companies, able agriculture put the industry in Genetic modification itself is con- and by association many scientists who debates with opponents with superior troversial yet we have examples that worked with those companies, have knowledge of global food needs and society weighs its risks and controver- largely lost the trust of the politically with harsh critiques of industrialized sies against benefits. The genetic engi- active public, particularly the ‘anti- agriculture.2 The companies’ relative neering of microbes to produce vital biotech’ activist organizations. The loss financial investments in ‘public good’ medical therapeutics such as insulin or of trust has many origins, some of projects vs. industrial agriculture erythropoietin is acceptable because the which hold lessons for commercializa- projects are in no way proportional to organisms are under high containment tion of GE trees. The public’s experi- their treatment in the industry’s pub- and unable to replicate in the environ- ence with automotive, tobacco, phar- lic relations material. Such dissonance ment, the product has a clear utility in maceutical, chemical, and oil compa- saving lives, and the medial consumer nies has created suspicion when large 2 has information and choice. It is fair to note that many biotechnology and apparently powerful companies opponents continue to make a similar strategy The first products of the biotech- claim to act in the public good and mistake in arguing against biotechnology on nology industry were seen (and still are scientific grounds when actually their objections make loud protestations of public are rooted in much more complex social and 1 See for example: The Genetically Engineered safety and benefit. The logging and economic issues. As the scientific community Food Alert (www.gefoodalert.org), The Organic timber products industry has similar inexorably and effectively addresses those issues, Consumer Association (www.purefood.org) and the chance for valuable and productive societal reputational liabilities for its poor past The Five Year Freeze (www.fiveyearfreeze.org). debate is missed. 72 between message, image, and action MAXIMIZE SOCIAL unlikely that products of genetically fuels distrust. engineered trees will be cheaper. So Forest biotechnology will not save UTILITY how can forest biotechnology create the world forests in the coming decades social benefit for those that may bear any more than crop biotechnology will Sustainable business is business unknown environmental impacts? solve problems of food insecurity. Tech- that raises its social utility by creating nology is only a small component to environmental and social value in ad- solve problems rooted in long histori- dition to economic value, while not HIGH IMPACT PRIVATE cal economic and political inequity, depleting resources. Good environ- environmental mismanagement, and mental performance in the past meant FOREST traditional harvest practices. There has doing “less harm.” Today’s stakeholder BIOTECHNOLOGY also been little or no detailed analysis demands performance beyond regula- today to say that genetically engineered tory compliance and favors companies Among the traits and benefits un- trees might relieve pressures on threat- and products that do “more good” der consideration for forest biotechnol- ened biodiverse forests or might signifi- rather than less harm. Many have ob- ogy that may have the greatest social cantly impact pulp and paper supply served that there would have been so- utility are the reduced-lignin designs in the regions of highest future de- cietal acceptance of biotechnology if that may lower the chemical, water and mand. only the first products had yielded di- energy use, and pollution created by The lesson for business is to trans- rect consumer benefit such as better- the pulp and paper industry. Though parently test, quantify, and communi- tasting, more convenient, safer, more the cost savings may not reach far cate the economic benefit of tree bio- nutritious, or cheaper foods. The same down the value chain, the benefit will technology to themselves and to soci- will be true of forest products. An of- occur at the site of milling. The com- ety. The typical consumer understands ten associated observation is that the munities at or near the plantations and the desire of businesses to grow, to re- agricultural biotechnology industry has the paper mills may receive a net envi- duce costs and to eliminate environ- asked the public to bear unknown risks ronmental benefit of cleaner water and mental liabilities. Genetically engi- of genetic engineering with no per- air in their communities. These prop- neered tree plantations in regulated and ceived direct benefits. erties may even allow branding of an industrialized markets may well im- A public invited to live next to otherwise undifferentiated product. prove corporate profitability and lower genetically engineered forests or or- But there is a catch. Selling cleaner a company’s net environmental im- chards will ask “are we getting wood processes involves admitting current pacts. “We’re doing this for our busi- or paper that is better, stronger, longer- environmental liabilities and dirty pro- ness” will resonate more truly with a lasting, more appealing or cheaper? Or cesses. How much of a paper skeptical public than will claims that “are we getting fruit that is better-tast- company’s resource base has to be from massive investments in new industrial ing, longer-lasting, more nutritious, or reduced-lignin trees to make a measur- technologies are motivated for the lo- cheaper?” Considering the lack of dif- able and transparent environmental cal public good or are intended to di- ferentiation and low cost among so impact and to outweigh subjecting all rectly save distant threatened forests. A many timber and paper products, the its production to scrutiny? As air and transgenic pine in Georgia will no products of genetically engineered trees water regulations and energy costs put more save the forests of Indonesia than may not deliver these kinds of direct increasing pressure on the industry, the will an improved soybean grown in consumer benefits. With farm costs day may come surprisingly soon. Iowa benefit the food-insecure peoples (plantations in this case) as a small A second intriguing trait for social of Africa and Asia. Business messages component of retail costs combined utility is fast growth. In this case the are more effective when true and with the high costs of regulatory ap- goal would be to develop a fast-grow- simple, rather than simplistic. proval and compliance, it seems equally ing pulpwood or hardwood with a

73 short enough rotation time to change strict sense, but might be developed a large positive impact on ecosystem the economics of logging in biodiverse through the public investment in tree restoration and upon the tourism, frontier or secondary forests. Like in biotechnology. What if publicly funded landscaping, timber and forest-product the lignin example, benefits would agricultural biotechnology had pre- industries. need to be regional, if not local. A ceded the private sector’s rush to mar- There are a wide variety of company may convince the public that ket? It is easy to imagine a public more projects under way in the public re- its fast growing GE tree plantation re- receptive and a market more open to search sector on reduced or increased moves its need to log on public and genetic engineering if the first we heard lignin content, increased cellulose con- private lands. If that claim was proved of genetically engineered crops was tent, faster growth, more uniform by selling or giving private forest lands Vitamin A-enhanced rice, a sweet po- growth, growth in marginal or arid to the state for public use, it would be tato to feed Central America’s hungry, soils, and other projects. The appar- that much more convincing. It will be or a high-protein cassava that grew in ently small investment in these projects hard for a forest biotechnology indus- the depleted soils of East Africa. and their use of the plantation species try to justify genetically engineered This “public-first” scenario is still of the developed world may not pro- trees in places such as New Zealand, possible with tree biotechnology, duce the near-term and high-impact Canada, Europe, or the United States though it will require large investment “icon” products described above that on claims that it relieves pressure in and careful choice of target species and would shape societal opinion. Other Russia, Indonesia, the Amazon, or preferred traits. The first public prior- alternatives that seem much more fan- Gabon. Agricultural biotechnology has ity may be rapid reforestation of aban- ciful in their benefit and technical re- revealed the different cost-benefit equa- doned and degraded agricultural lands alizations include engineered control of tions in different parts of the world. A to create measurable benefits of soil stress response and adaptation to allow fast-growing tree plantation in Asia stabilization, watershed protection, adaptation to climate change, produc- that saves adjacent forests and meets habitat restoration, and timber production of bio-based fuels, and trees de- local pulp demands is a different tion. Fast-growing plantation trees designed for carbon sequestration. proposition. Environmental safety is signed for tropical zones might also be Public sector efforts in tree bio- most like to be met by multiple, be- used to create plantation buffers technology face the same challenge as nign mechanisms for tree sterility and around threatened tropical forests to the application of agricultural biotech- plantations managed for biodiversity supply pulp, timber, fuel, and forest nology to global food needs: lack of and ecosystem services. In either the products to local communities. Fast- scientific knowledge of tropical species, case of reduced-lignin or fast-growth, growing fuel woods that grow on mar- lack of scientific and regulatory capac- a private company will have to meet ginal soils might also help protect for- ity in developing countries, diversity of the highest standards of environmen- est frontiers, raise living standards, and species and culture methods, and the tal safety, ethics, and transparency to support economic development. concentration of R&D dollars and in- win the public trust. Another possible development tellectual property in the private sec- that would facilitate public acceptance tor. The research agenda today is not of genetically engineered trees would driven by a global analysis of needs and HIGH IMPACT PUBLIC be specific disease resistance that saves the functions of trees and forests. Cre- a tree of high environmental, eco- ating a genetically engineered tree to FOREST nomic, or symbolic value. In America, deliver measurable public benefit BIOTECHNOLOGY genetic engineering for fungal resis- would call for tens of millions of dol- tance that would allow the restoration lars in public and private scientific in- A market for tree biotechnology of the American elm and American vestment that is guided by a deep needs need not be an economic market in the chestnut to Eastern forests could have analysis and public participation.

74 DESIGN FOR THE vention at the moment of design. The should not be confused with product same is true of genetically engineered prototypes or with products engineered ENVIRONMENT crops and the same will be true of ge- for the market. netically engineered trees. An informal mapping performed The agricultural biotechnology at WRI of the environmental issues of industry is just coming to appreciate THE IRONY OF INPUT genetic engineering of food crops sorts the implications of design and the anal- issues into those of direct impact, such ogy of front-of-pipe designs to reduce TRAITS as human food safety, ecosystem harm, cost and risk. Consider as an example, animal safety, loss of genetic diversity, that the design of Bt corn was simply Environmental impacts, commer- resource depletion, and unknown im- to achieve gene expression in corn. The cial benefit, and social acceptance are pacts and the indirect impacts that ge- accomplished goal of constitutive ex- specific to the engineered trait and to netically engineered crops may have on pression of Bt toxin in all corn tissues, the physical and cultural context of the the intensification and spread of indus- among them the corn pollen, has lead silvicultural system. This point cannot trialized, chemical-intensive monocul- to the high costs of testing on pollen be over-emphasized; the engineered ture. At least four primary mechanisms flow, the need for extensive refugia, trait (i.e., the modification and its ex- may mediate most of the direct envi- complex grower contracts and compli- pression in the plantation context) is ronmental threats such as toxin pro- ance schemes, and the persistent con- the determinant of direct and perceived duction, gene disruption, weed cre- troversy of impacts on non-target lepi- social utility. The first crop biotechnol- ation, and genesis of new pathogens. dopterans such as the monarch butter- ogy products to see large-scale plant- At the root of almost all these poten- fly. Another example of design failure ing all featured “input” traits. The in- tial risks are three core issues: the con- is the need to eliminate antibiotic re- put traits act as production inputs or trol of gene expression, the potential of sistance markers and to develop alter- work in conjunction with production gene transfer, and the intended design native selectable markers. These cases inputs to the agricultural system and of the engineered organism. suggest principles for design, such as their benefits accrue to the supplier of Most agricultural molecular biolo- that (1) the introduced gene should the input and to the farmer; the mar- gists don’t label themselves as genetic only be specifically released into the keted product has no new functional engineers, and the language of engi- environment; and (2) there be no func- characteristics. neering and design is not used to de- tional open reading frames in Only a small volume of crops such scribe genetically modified crops. transformants, except the gene of in- as soy, canola, corn (and even cotton However, these crops are engineered terest. for fiber use) are consumed in their products, and an engineering mindset Had such principles guided the pure form, and they are chiefly the would serve the industry and society. priorities of basic and applied research, low-cost ingredients to value-added Engineers have spent the last several the risks and benefit of first-generation food products. Farm-gate prices of decades learning and proving that en- biotechnology products may have been most commodities are at historic lows vironmental benefit is best achieved by very different. Tree biotechnologists in the United States, and the farmer’s design, and that approximately 80% of can adopt the mindset of green prod- share of the consumer dollar spent on the environmental impact and costs of uct designers and use design principles grains and vegetables is roughly $0.04– a product is determined at the point for environmental safety to drive their $0.07 (National Agricultural Statistics of design (Tischner and Charter 2001). product development agenda and to Service 2001). Small improvements in The end-of-pipe solutions of scrubbers, identify frontiers of basic research. The farm productivity or reductions in in- waste treatment, and toxic disposal are transgenic trees planted to date for re- put and labor costs are imperceptible far more costly to society, business, and search purposes should be recognized to the final supermarket customer. The the environment than pollution-pre- as the experiments that they are and impact of the input traits such as pest

75 and herbicide tolerance have no direct herbicide resistance in food crops has humans. The environmental benefits cost benefit to the end consumer of the not been transparently shared by the were also not transparently communi- engineered crops, and it is worth cal- sponsoring companies and has been cated to the customers of foods that culating whether any input trait could publicly questioned by critics of bio- contain biotech-derived ingredients, have a direct consumer price benefit. technology. and there were charges from environ- For herbicide tolerance, the envi- Third, glyphosate or Bt toxin and mental activists of threats to nature and ronmental benefits are the replacement the chemicals it replaces are produced beneficial insects, as well as threats to of more toxic herbicides by glyphosate and sold by the same set of agrochemi- the purity of organic crops. Drawing and the adoption of no-till farming cal companies and are used by the same consumer attention to the EPA and methods that save labor, fuel, soil, and customers and are all approved by the FDA’s findings that either genetically water. For the Bt crops, the benefits are same regulatory authorities who also engineered crops or chemical pesticides reduced dependence upon more toxic sell, use, and approve the genetically can be used safely is not a consoling pesticides. The input traits of herbicide engineered seeds. Drawing too much thought to today’s consumer. Perhaps and pest resistance for food crops de- attention to the chemicals and their another incongruity in communicating livered little perceived social utility for relative food and environmental safety the benefit of genetically engineered their claimed impact upon the environ- (which should be comparable when crops was the resistance of the indus- ment and food safety; the reasons bear used within approved limits) might try to label consumer products in a lessons for tree biotechnology. also draw criticism to the agrochemi- consumer society where advantages are First, the general public is unaware cal industry, the regulatory authorities, so prominently emblazoned on prod- of and may not want to know the and the farmers—a no-win situation uct labels. A citizen logically wonders, quantity and nature of chemicals used for everyone. The last reason why her- “if this is so good for me and the envi- on crops in industrial agriculture or of bicide resistance was not marketable to ronment, why isn’t it advertised on the the negative impacts of modern farm- consumers returns to the idea of label?” ing. The benefit of “less herbicide” trust—and that those who promoted To a world that does not perceive draws attention to the use of chemi- the benefits are also those who would trees as crops and perceives forests as cals and associates the consumer prod- profit the most by selling the seed and symbols of nature, trees that produce uct with chemical intensive and “non- the herbicide. Balanced against no di- bacterial insecticides or are made to be natural” farming. Although no-till rect cost benefit and unappreciated in- sprayed with chemicals are not likely farming is an important advance, com- direct benefit are fears of environmen- to be accepted if there are perceived plex environmental issues of destruc- tal risk and human health risk. “Why environmental risks. Transgenic trees tive farming, non-point source water should I bear even remote or unknown designed to be herbicide tolerant for pollution, and soil loss are distant from risk, if others profit and I don’t ben- the benefit of survival at the seedling the decisions about food purchase. The efit?” demands the concerned con- stage or long-term plantations of trees dramatic growth of the organic foods sumer. expressing Bt toxin irrespective of pest market is largely a testament to fears A similar set of reasons explains levels seem a poor starting point for the of the safety of foods and to a lesser why there is no perceived direct con- industry. The message to forest bio- degree, environmental concern. The sumer benefit to the pest resistant technologies should be very clear: com- creators of food brands want to asso- traits. For the Bt toxin crops, the mercialize output and social utility ciate food with a natural rather than simple description of the crop is that traits well before commercializing in- destructive image of farming. Less it produces its own insecticide instead put traits that might increase chemical harm is not as compelling an associa- of using insecticidal chemicals—draw- use, promote chemical use, or that tion as crops that deliver more societal ing attention to the use of chemicals draw attention to chemical use and the good. Second, the data on the eco- and to the fact that the consumer may “unnaturalness” of tree farms. nomic and environmental benefit of be eating a poison, though harmless to

76 PRESSURES ON THE This creates a conflicted context for term liabilities are only subject to forest biotechnology. The vision of pro- speculation. One thing is certain: that FOREST INDUSTRY prietary and advanced technology and a strategy, culture, and physical method the seductive visions of genetically engi- for long-term product stewardship is Discussion of forest biotechnology neered super-trees must be alluring to particularly important for transgenic often starts with “does the world need leaders of an industry of bulldozers, trees. All these features will raise costs transgenic trees?” This is a very impor- chain saws and pulp mills. Moving out and demand skilled labor and new tant question and may be an impor- of contested forests and into privately management methods. Thus, a tant guide for public sector research owned plantations must also be attrac- transgenic tree needs a new public ori- and development. More immediately, tive. At the same time, product devel- entation to trees as crops, new science, the question might be, “does industry opment costs and regulatory costs and new regulatory systems, and new man- want transgenic trees?” Agricultural the visions of anti-biotechnology protest- agement practices in the industry for biotechnology has shown that when ers destroying test plots of trees and at- product development, product stew- there is a powerful economic motiva- tacking company CEOs must make the ardship, and plantation management. tion for industry, the genetically engi- same business leaders distinctly queasy. Each of these changes lowers the prob- neered products will be developed. The ability that the companies at the front strong financial push and pull on of the learning curve will profitably Monsanto from their huge investments NEW CAPABILITIES execute product introduction. in seed companies and rising valuation The Role of Scientists on Wall Street was a powerful accelera- AND CULTURE Besides the external pressures on tor for GM crop introduction and cre- the industry and the need for new ca- ated a competitive environment that The movement into genetically pabilities, the scientific community it- demanded a similar response from engineered trees also calls for a signifi- self facilitates and confounds good de- Monsanto’s agrochemical rivals, includ- cant cultural and technical change in cision-making by the industry. Biotech- ing Dow, DuPont, Aventis, and the industry. For the agrochemical/ nology has been science-driven, as new Novartis (now Syngenta). pharmaceutical giants, the technology discoveries seek applications and eco- A crude snapshot of the forest prod- and regulatory processes of genetic en- nomic value. In areas of biology less uct industry (timber, pulp, and paper) gineering were not entirely new, and funded than human biomedicine, ge- shows an industry under regulatory pres- played to their competitive strengths. netic engineering raises the possibility sure, rising competition from global The molecular biology, ecological test- of increased funding, scientific interest, competitors and an industry that is striving, compliance issues, intellectual and the potential for riches to scien- ing for modernization, value-added property strategy, and regulatory pro- tists, investors, and research institutions products, and an improved reputation. cesses to commercialize a transgenic from patents and new biotechnology The public pressure from activist orga- tree are not part of the traditional and companies (Smith et al. 1999). For sci- nizations on both ends of the value current capabilities of the forest prod- entists who have spent lifetimes study- chain, from the logging companies, to uct industry. ing forest and tree biology, genetic en- paper mills, to the do-it-yourself chains The first transgenic tree planta- gineering is a powerful tool to unlock has been great. The result of that pres- tions will have measures for biological scientific mysteries. Enthusiasm for the sure is unprecedented demand for prod- and physical containment, intensive science and technology is real and un- ucts from certified forests and a direc- ecological monitoring protocols, and derstandable in the world of science. tive for the industry to transform itself fences or barriers for economic and Whether the motivation of scientists is from one of the last extractive industries physical protection. The long-term the purest interest in discovery, a genu- to a sustainable industry based on renew- impacts of transgenic plants are un- ine hope for sustainable technologies, able resources. known and thus the potential long- or the desire for recognition and fund-

77 tives of civil society that might have neered products. Patents are credited Table 1. The possible pitfalls of forest guided product design and introduc- with being the foundation of the phar- biotech? tion. Today, the agricultural biotech- maceutical industry and with creating • Lack of expertise that bridges sectoral nology companies have put in place the conditions for the birth of the bio- gaps and interdisciplinary gaps high-level stakeholder advisory boards technology industry; the patent race • Lack of analysis of global or local needs from diverse societal arenas, though it accompanying the human genome ef- • Seeking public trust upon altruistic is too early tell how those boards are forts reflects their continued impor- claims of distant environmental benefits impacting company action.3 tance to the industry (Regaldo 2000). • Failure to engage stakeholders in One still hears the mantra of Intellectual property has also been one product and field trial design ‘sound science’ repeated in debates on of the most contentious issues in the • Failure to create public-private partnerships biotechnology and the implication that opposition to biotechnology for the • Commercial pressure to go to market if only the public understood the science, validity of patenting life forms, the use too early the products of biotechnology would be of patents for economic control and • Regulatory costs create pressures for embraced. Sound science does not shape competitive advantage, and the patent- unethical practices the marketplace and is low on the list of ing of species considered to be in the • Imitation of agricultural products; the lure the basis of consumer choice. Fears, de- public domain and natural patrimony of easy input traits sires, and price shape consumer accep- of developed countries. Although the • Science-driven choices rather than tance, and this is obvious from the cars value of patents is a common assump- market-pulled we drive, vitamins we take, clothes we tion, there is also an analytic literature • Over-valuation of patents wear, foods we eat, and the risks we bear to suggest that patents are often over- • Long-term liability and stewardship for pleasure and convenience. The pri- valued, do not create strong competi- issues orities for genetically engineered trees tive barriers, and have lower economic should not be guided solely by ‘sound value and strategic utility than is often ing, there is a powerful confluence of science’ and scientists. The pressures on assumed (Mazzoleni and Nelson 1998; reasons to be excited and to promote industry and the motivations of scientists Cohen et al. 2000). genetic engineering of trees. in regulated and in less-regulated emerg- A tree biotechnology initiative will The basic scientists at the fore- ing economies create a force for the de- have to deal with the large suite of pat- front of exploring genetic engineering velopment of transgenic trees, and scien- ents on molecular methods and genes of trees are the scientists sought as ad- tists can be inspired to serve society and likely to be used to create a transgenic visors and collaborators for the com- be held accountable if other sectors of tree. But should tree technologists seek panies exploring the possible commer- society become engaged in this issue. to patent engineered species and their cialization of genetically engineered underlying technology? If genetically trees. This was also the case for crop engineered trees are owned by the same biotechnology; so why was industry THE ROLE OF companies that will grow and process totally unprepared to address and re- INTELLECTUAL the trees, do they need the same pro- solve so many environmental and so- tection as seeds for crops that may pass cial issues? The reasons are the belief PROPERTY through a complex value chain? Are the system of “sound science” and the ab- costs of the patents in direct terms and sence of other scientific and social sci- Intellectual property often as- in potential societal opposition justified ence viewpoints. The molecular biolo- sumes a central role in the strategies for when weighed against the extremely gists and the corporate strategists the development of genetically engi- long life-cycle of trees, the ease with thought that the other party had a which ownership may be established handle on the potential risks of the and protected, the rapid development products. Missing from the implemen- 3 The author is a member of Monsanto’s Biotechnology Advisory Council. of new technologies, and other means tation were ecologists and representa- 78 to protect property? The case for pat- themselves. The companies sponsoring tree plantation may not seem much enting trees is not obvious, merits research and development in geneti- more unnatural than just the planta- analysis, and may be a weak and in- cally engineered trees are forest land- tion itself, and the view of genetically correct strategic assumption. owners and timber product companies; engineered trees may be different if it the understanding of industry needs occurs as part of a gradual and envi- that guides R&D originates in the in- ronmentally responsible transition to COMPARISON OF dustry itself. Though it is not yet clear, tree plantations. there seems to be no explicit interest The forest product industry does FOREST VS. AG- in the export and sale of transgenic not have the benign image of farming; BIOTECH INDUSTRIES seedlings, though the high cost of it may be possible to sell plantations product development will probably and engineered trees for doing “less This essay has explored the analo- create the pressure for exactly such a harm” than logging in natural or pub- gies between genetically engineered value capture strategy. The molecular lic forests. ‘Naturalness’ is not a con- trees and crops and the system that has biologists involved in engineering trees sumer value of most timber and forest produced commercial products with a were first trained as tree and forest bi- products; we do not seek a natural central thesis that the forest industry ologists, and may be more likely to quality to our lumber or copy paper as can learn a great deal from crop consider the wild and managed biologi- we do to a piece of corn, a vegetable, biotechnology’s failings. There are, cal context and complex forest system or fruit. This is shown by the relative however, important differences be- than their crop science colleagues. objections to genetically engineered tween the two industries, and some The forest industry is already un- cotton in contrast to genetically engi- that may help prevent repetition of the dergoing significant change and mod- neered corn or genetically engineered same mistakes (Table 2). First is that ernization, and genetic engineering, wheat. Cotton is also the one crop the technologists are the customers rather than catalyzing disruptive where there is the clearest data that change—as it did for the old chemical shows lowered use of harmful chemi- and drug companies—exists in cals on the industrial cotton crop (Car- Table 2. Key differences of fiber vs. food biotech- the context of other changes to- penter and Gianessi 2001). An average nology industries. ward greater environmental and consumer may fidget a moment to • The leading technologists are the customers of social responsibility. The nega- think that their jeans or underwear the product. tive ecological impacts of the contain cotton from genetically engi- • No agri-chemical industry equivalent; less timber and pulp industry are al- neered plants, but the response is less financial pressure. ready recognized; they make up visceral than the discovery that their • Limited intention of exporting and selling genetic stocks. a powerful story in the public breakfast or lunch contains ingredients • Understanding of needs originates in the mind. It may take a while for derived from genetically engineered industry. people to see trees as crops, but crops. • The scientists are tree- and forest biologists the acceptance of plantations is The fiber system is simpler than with systems approaches. • Industry is already undergoing change. underway. The tree industry the food system, since transgenic trees • Ecological damage of the forest industry is does not have the legacy of in- may be developed by their primary recognized. frastructure and planting prac- harvesters and processors. There are •Transition to plantations is underway. tices as crops, and there may be fewer players, fewer products, fewer • Long timeframes in tree science and business. some chance of a biodiverse and species and culture methods, and sim- • Naturalness is not a quality of paper and timber. • Fiber system is simpler and more public than ‘eco-silvicultural’ practice devel- pler value chains. This simplicity may the food system. oping with lower barriers, rather make it easier to design products and • Fiber is less contentious than food. than changing agriculture to a develop value chain relationships with • Forests and trees have high symbolic value. different model. A transgenic more aligned interests than the current

79 path of genetically engineered seeds Balanced against all these com- development, conversion to agricul- from biotechnology company, to parisons that make commercialization tural lands, ore and oil extraction, over- farmer, to processors, to traders, to of acceptable genetically engineered hunting and over-logging, global cli- food companies, to supermarkets, and trees more probable, is the symbolic mate change, and destruction of water to consumers and restaurants. value of forests and trees. A treatment resources. Billions of people in the Time may also be on the side of of the symbolic history and value of world have pressing needs for energy, genetically engineered trees. The trees and forests and is beyond the paper, and materials that are needed mindset of the forest product industry scope of this essay. The American bio- features of economic development, is much longer than that of the crop technology companies underestimated improved health, literacy, and com- industry, which is based in annual the cultural symbolism and importance merce. And forests play central roles in cycles, and it is normal for forest com- of agriculture in Europe, and biotech- protecting watersheds, purifying air panies to think in 5-, 10-, and 20-year nology activists underestimated the and water, stabilizing climate, protect- time frames. The slow growth of trees desire many developing countries have ing species diversity, offering cultural ensures there will be no fast product to use biotechnology or to self-deter- and spiritual value, and supporting introduction; and delays of a year to mine their own technological choices. tourism and recreation. In a world get the best transformant or to choose Timber product companies may be in whose population is due to grow by a a proper genetic background will have for a shock at how the public feels third in the next 25 years (and chiefly less of an impact on the rate of tree about its trees, especially if the geneti- in less-developed countries), imagina- commercialization. Tree scientists are cally engineered tree does not directly tive solutions with a place for technol- also a patient lot. There is no time or connect to the protection and renewal ogy will be needed to meet global financial pressure from the public mar- of forests. The typical public opinion needs for water, materials, energy, and kets on the industry to meet or exploit survey sponsored by the biotechnology paper. This will require keeping sight the promises of the technology in any industry asks about possible acceptance of the sustainable forestry goals beyond near time frame. of benefits and not, “Would you like a the transgenic trees. plantation of genetically engi- This author has never seen a deep Table 3. Framework conditions for a GE tree market. neered trees in your back- analysis of the role of tree biotechnol- yard?” Still, the framework ogy that considers the values that for- • Social utility is the foremost concern. conditions for a market for ests and trees deliver, in the context of • Design for the environment is of highest priority. transgenic trees are fairly clear specific nations, social values, regula- • Business motives and plans are transparent. • Business communications and actions are aligned and conceivable when ab- tory structures, and economic sce- with investments. stracted from our experience narios. We cannot question the utility • Stakeholders are engaged in decision-making. with genetically engineered of genetically engineered trees without • Private and public investment are balanced. crops (Table 3). a serious question of “compared to • Research in ecological impacts of transgenic tree what?” There is opportunity to develop plantations. sustainable silviculture that is part of •Value capture and business strategy is not based THE QUALITY OF integrated management of productive in patents. forests, working landscapes, and pro- • The first developed traits are output traits. OUR ANSWERS tected forests to maximize ecosystem • Stakeholders that perceive risks can make EPENDS ON THE choices and perceive benefits. D goods and services for all human uses. •Target markets have appropriate regulatory IMAGINATION OF Tree plantations do not have the legacy capacity. of crop agriculture, such as a history • Region-specific products are developed. OUR QUESTIONS of monoculture or a system of produc- •Technology is applied to serve needy populations tion and production inputs, and they The world’s most biodi- and protect biodiversity. might be designed for biodiversity. verse forests are threatened by 80 During the next five years, the Lenzner, R, T. Kellner. 2000. Corpo- public and private sector will make rate Saboteurs: They wrecked Monsanto, now they’re after the critical decisions about investment in U.S. drug industry. Is your com- genetically engineered trees. Risk or pany next? In Forbes, pp. 156–168. benefit is not intrinsic to genetic engi- Mazzoleni, R, RR Nelson.1998. Eco- neering. The experience with the genomic theories about the benefits netic engineering of crops has proved and costs of patents. Journal of Eco- that cultural, environmental, and economic Issues 32: 1031–1052. nomic risks and benefits are each con- National Agricultural Statistics Service. ceivable and are each achievable with 2001. Agricultural Prices. U.S. De- genetic engineering. Tree biotechnol- partment of Agriculture. ogy has not yet crossed the proof-of- Regaldo, A. 2000. The Great Gene concept threshold for either risk or Grab. Technology Review 49–67. benefit, and the traits and species to be Sampson, V, L. Lohmann. 2000. Brief- chosen for commercial modification ing 21 - Genetic Dialectic: The Bio- remain uncertain. Whether and how logical Politics of Genetically Modi- genetic engineering can equitably and fied Trees. Corner House, London. safely serve the needs of sustainable Seuss, TG. 1971. The Lorax. Random development remains to be seen. We House, New York. 62 p. can make wise choices as citizens, sci- Smith, KR, N Ballenger, K Day- entists, and businesspeople about how Rubenstein, P Heisey, C Klotz- to develop the technology, with what Ingram. 1999. Biotechnology re- safeguards and to what ends. The search: Weighing the options for a thoughtful, creative, and rigorous con- new public-private balance. Agri- sideration of this question should be cultural Outlook October: 22–25. limited only by our knowledge in the Tischner, U, M Charter. 2001. Sustain- moment and not by the imagination able Product Design. In Sustainable and courage to envision and realize fair Solutions: Developing products and processes, shared benefits, and a sus- services for the future, ed. U Tischner and M Charter. Greenleaf tainable future. Publishing Trowbridge, UK, p. 469. REFERENCES Carpenter, JE, LP Gianessi. 2001. Ag- ricultural Biotechnology: Updated Benefit Estimates. National Center for Food and Agricultural Policy, Washington, DC, p. 46. Cohen, WM, RR Nelson, JP Walsh. 2000. Protecting their intellectual assets: Appropriability conditions and why U.S. manufacturing firms patent (or not). National Bureau of Economic Research, Cambridge, MA.

ETHICS

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 85-91. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf The Ethics of Molecular Silviculture

Paul B. Thompson

ABSTRACT

The paper provides a general discussion of ethics as applied to technical practices. Ethics is defined as the explicit articulation of the underlying rationale for engaging in or regulating a technical practice. Recombinant techniques for plant transformation silviculture raise ethical issues largely because they have brought the technical practices of silviculture and plant development before the public eye in a manner that is unprecedented in Paul B. Thompson holds recent memory. This has placed practitioners in the position of needing to make an articulate and non-technical statement of the rationale—the ethic—that guides the use and the Joyce and Edward E. development of science and technology within plant and animal sciences. Too often they Brewer Chair in Applied have been unable to do this. The unfortunate result can be an erosion of confidence and trust in the technical competence of specialists and a ratcheting effect that links ethical Ethics at Purdue Univer- issues with the perception of elevated risk. Although it is difficult to propose measures that would constitute a rapid response to this situation, the longer-term need is to de- sity. His book, Food velop an ongoing effort to increase the capacity for articulation and communication of Biotechnology in Ethical the professional ethic that guides the technical practices of silviculture, and to ensure that technical professionals are receptive to constructive criticisms of their prevailing practices. Perspective, was pub- This paper provides an overview of normative issues associated with molecular silvi- lished in 1997. He has culture. Following a brief clarification of terminology, ethical issues are broken down into four categories: religious or metaphysical concerns, environmental ethics, social ethics and also served on several professional ethics. Contested issues in each category are briefly reviewed, followed by a succinct statement of the author’s considered views. national committees including the USDA Agricultural Biotechnol- TERMINOLOGICAL INTRODUCTION ogy Research Advisory he relevant practices of molecular silviculture include especially the use Committee (ABRAC) of recombinant methods for plant transformation within research, pro and the National T duction, and conservation contexts, but also the development and application of genetic techniques such as genomics and informatics that need not in- Reaseach Council Sub- volve plant transformation. This definition leaves open some amount of ambiguity regarding the scope of the practices under review. However, it is ambiguity committee on Environ- over the term ethics that is far more likely to create confusion or misunderstand- mental Risks from Com- ing. As such, some pains will be taken to clarify the intended scope of ethics, and the role that the academic discipline of philosophy can play in elucidating the mercialization of ethics of any technical practice. Transgenic Plants. Practices involve ethics insofar as they are understood to serve larger purposes or to be valuable in themselves. To examine the ethics of a practice is to investigate how the practice can be understood to be done well or poorly, to in- [email protected] quire into the purposes or value of the practice, and to articulate standards for performance, justification, or evaluation of the practice. Some people reserve the word ‘ethics’ for issues involving conflict of interest or sexual misadventure, and

85 the recent spate of interest in research Both these points of view involve ethi- The balance of the paper describes ethics has focused attention on issues cal principles, and one role of philoso- the four previously noted domains of of scientific misconduct. The term is phy is to spell out the manner in which ethics in which issues arise in connec- also associated with cultural mores, re- conflicting ethical principles contribute tion with molecular silviculture. Each ligion, and even irrationality in some to each of these contrasting points of section includes a review of the ongo- quarters. Philosophers have developed view. ing debate. In each section, the review a somewhat more specific interpreta- There has arguably been a one- is followed a very concise statement of tion of ethics that stresses the explicit sidedness to press coverage about the my own views. Space constraints do formulation of justifications, ethics of biotechnology. The word ‘eth- not permit extended arguments on the desiderata, and codes of conduct. Each ics’ is generally associated with view- topics in question. involves a set of claims intended for use points that are critical of using biotech- in deriving statements that specify a nology, and often unilaterally opposed particular policy or a particular course to all uses of recombinant techniques RELIGIOUS AND of action ought to be followed. Philo- for plant and animal transformation. sophical ethics (or moral theory) is the Although these critical viewpoints are METAPHYSICAL ETHICS study of how ethical principles can be often countered by sources who cite systematically used to develop logically potential benefits from recombinant The first question that leaps for- consistent and conceptually coherent techniques, these voices advocating the ward when the subject of ethics is ethical arguments. The claim that for- weighing of risk and benefit are not broached in connection with molecu- est policy should promote efficient use represented as expressing an ethical lar genetics is whether this whole area of resources is, thus, an ethical prin- perspective. Yet from the standpoint of of science and technology doesn’t trans- ciple because it advocates the norm of philosophical ethics, arguing for a prac- gress some sort of boundary or abso- efficiency as a criterion for the forma- tice by citing the relative value of its lute prohibition. And even if simply tion and justification of management costs and benefits is a time honored learning about the genes is permitted, plans and forest policies. and logically coherent approach to eth- some clearly believe that moving genes Philosophers interpret disputes ics. through recombinant techniques is over the legitimacy or justifiability of Although philosophers can be ex- not. This is, in other words, the ‘play- a given practice as involving compet- pected to use a common vocabulary in ing God’ domain of ethics. There is ing or incompatible ethical principles. discussing ethical issues, it is still the neither doubt that many people react This is, perhaps, contrary to those who case that the judgments and opinions to the prospect of altering the genetic presume that the word ‘ethics’ signals that individual philosophers develop make-up of living things with repug- a particular class of special consider- are somewhat personal. As such, philo- nance, nor difficulty in understanding ations, distinct from those that would sophical literature in ethics often con- why they might tend to express their be characterized as ‘social’ or ‘eco- sists in the statement of a particular reaction by questioning the ethics of nomic’. For example, consider a hypo- viewpoint or evaluation, followed by an such practices. In the case of issues thetical dispute between someone argument intended to support the con- made familiar by press coverage of hu- holding the view that tree biotechnol- clusions expressed therein. In this con- man cloning and stem-cell research, ogy is justified because it promotes ef- text it is important to cover a wide grounding these reactions in terms of ficient use of natural resources and range of issues in a succinct fashion, specific religious norms is a fairly someone holding the view that tree but it is also important for a philo- straightforward process. However, it is biotechnology is unacceptable because sophical author to be as unambiguous surprisingly difficult to articulate why it is unnatural. As philosophers see it, as possible in communicating the judg- the alteration of plants, including trees, this is not a dispute in which only one ments that he or she has reached on would transgress generally recognized person is making an ethical argument. the issues in question. ethical boundaries, or how it would

86 relate to well-established religious tra- While we should be cautious about risks of transgenic plants, and I would ditions. discarding that wisdom, it is very dif- think that given the lengthy reproduc- Although I can certainly imagine ficult to see how it translates to a world tive cycle of trees, these questions are a theological/metaphysical conception of DNA, coding and non-coding se- particularly vexing in the area of silvi- of nature that ascribes certain inten- quences, and micro-cassettes. Those culture. The main focus of environ- tions or purposes to the fabric of real- who presume that phenomenally de- mental risk from transgenic plants has ity, and I can imagine that those in- rived norms bearing on topics such as been the potential for unintended im- tentions and purposes might be inimi- heredity or living appropriately in na- pacts on so-called non-target organ- cal to biotechnology, I am frankly not ture can be transcribed literally into isms: gene flow to close relatives, and the philosopher to offer a sympathetic talk about genes, proteins, and molecu- inadvertent effects on habitat that af- portrayal of such a viewpoint. In fact, lar life processes are guilty of naive ge- fect other forms of plant and animal my reading on the ethics of biotech- netic determinism. Unfortunately, too life. Although these are inherently em- nology suggests that most people who many scientifically trained people are pirical questions, the framing and are inclined to worry about this possi- guilty of this—but that is an issue for analysis of environmental risks involve bility are actually somewhat reluctant professional ethics, and I must not get a number of value judgments that re- to make bald statements about God’s ahead of myself. quire a sophisticated mix of science and intentions or purposes. Instead, they Before leaving the religious and ethics. The question of whether to refer ambiguously to the sanctity of metaphysical domain, I want to stress minimize Type I or Type II statistical life, or defend the repugnance that that I am not dismissing these issues. I errors is one example. The question of many people feel on first learning am not saying that the biologists’ lan- which populations to specify when for- about the new genetics (see Thompson guage is a true description of reality, mulating relative probabilities is an- 1997, 2000). while ordinary or religious descriptions other. Are we interested in transgenic I do not want to imply anything of the phenomenal world are false. trees as a class? Should they be com- but respect for people who offer these Rather, I am saying that I do not how pared with all non-transgenic trees, or points of view. In fact, I take great to build the bridge between these two should we be making a comparison comfort in the fact that they do not kinds of language. As such, I do not between trees that are genetically simi- profess to be on the hotline to God in know how to apply norms of humility lar, save for the transgene of interest? deriving their concerns about genetic and respect for life at the molecular Do impacts on land use count in the technology. Nevertheless, my own view level. As will become evident shortly, I environmental risk assessment, even on these issues is that the humility and do think that we can build bridges that though human decision making would cautiousness endorsed by those who relate to some specific environmental be involved in bringing them about? take such perspectives is more appro- and social concerns. I am not sure These are not purely scientific ques- priately expressed as a component of whether it is important to build bridges tions, and there should be a more ex- environmental or social ethics rather in the domain of religious ethics, but plicit and conscientious effort to ad- than as a specific reaction to the fact if it is, that work is surely in its infancy dress these ethical questions in techni- that recombinant techniques are being so far. cal debates about environmental risk. used. There is, in my view, a large and And then, of course, we get to the growing gap between the language that question of whether these risks are ac- we use to make sense of the phenom- ENVIRONMENTAL ceptable. At present, the debate has enal world of daily life and the lan- stressed uncertainty. Does the open- guage of molecular biology. The moral ETHICS ended nature of these questions provide wisdom that we derive from our reli- a reason to block either research or gious and cultural traditions is fitted to There are, I think, a lot of open commercialization of transgenic trees? a world of rocks, trees, and flowers. questions about the environmental There are environmental activists who

87 think that uncertainty provides the involve such extensive costs that it is Muir himself would have gone, argues basis for sweeping argument against not actually a feasible alternative at all. against all human use of trees would transgenic silviculture. They often link So the alternative to tree biotechnol- conclude that absolutely every conceiv- their argument to the precautionary ogy may actually be an unacceptably able application of tree biotechnology principle. Although the precautionary exploitative expansion of current prac- is impermissible. Only someone taking principle can be formulated in various tices in industrial forestry. If this is the Pinchot much further than Pinchot ways, its ethical importance consists in case, then precaution may actually himself would have gone could think the way that it offers an alternative to weigh in favor of transgenic trees. My that the impact of tree biotechnology norms or decision rules that promote viewpoint on the environmental eth- on protected wilderness areas is ethi- risk-taking whenever expected benefits ics of molecular silviculture is that it cally irrelevant. This brings us back to exceed probable losses. A precaution- depends on some background and con- the issues we started with: the environ- ary approach would differ in that losses textual elements of forest policy, and I mental risks of genetic engineering on associated with environmental damage need to hear more about it before I non-target organisms. It is certainly are treated as a special case on any of could form a firm opinion. possible that different environmental several grounds. For example, one There is also an even more gen- values will lead people to frame ques- might argue that ecological complex- eral set of issues in environmental eth- tions of risk assessment in different ity or the relative weak predictive ics. There has been a tremendous ways, and that is one reason why non- power of ecological models provides a amount of ink spilled over the debate scientists need to be included in the reason to expect that environmental between anthropocentrism and process of environmental risk assess- consequences may be much worse than ecocentrism in environmental ethics, ment. predicted. The irreversibility of envi- and this debate is often traced back to ronmental outcomes are also cited as the philosophical conflict between reasons to weigh possible losses much Gifford Pinchot and John Muir over SOCIAL AND POLITICAL more heavily than expected benefits. In the future of American forests. Pinchot both cases, advocates of a precaution- is portrayed as a figure who saw wil- ETHICS ary principle would demand a higher derness as deriving all its value from the standard of evidence for expected ben- various uses—including recreational Many of the activists who have efits than for possible environmental uses—that humans make of it. Muir is opposed biotechnology in agriculture hazards (see Raffensperger and Ticknor portrayed as someone who believed ground their opposition in a socio-po- 1999). that forests, trees, and wilderness were litical argument. I will sketch the terms Although I endorse a precaution- intrinsically valuable, irrespective of of this argument briefly, though I will ary approach in environmental policy, any use that humans made of them. say at the outset that, in my view, the I do not think that it entails a sweep- Clearly, this debate continues to reso- considerable merits of this argument as ing indictment against biotechnology. nate throughout forest policy (see a case for the reform of our technol- The key to my judgment is that even Norton 1991; Callicott and Nelson ogy policy do not translate into persua- a precautionary approach requires one 1998). sive reasons for singling out genetic to evaluate a proposed course of action Does this debate have any bearing technologies. Critics of biotechnology in comparison with its alternatives. If on tree biotechnology? My own view in agriculture allege that it is a tool for the alternative to tree biotechnology is is that its bearing is rather indirect, and increasing the control that a few cor- that the human species will desist from that certainly the anthropocentrist/ porations hold over the entire food sys- all use of forest products, precaution ecocentrist divide does not translate tem. They see biotechnology as a might weigh in against biotech. The directly into pro and con positions on weapon being wielded against poor problem, of course, is that abolishing tree biotechnology. Only someone farmers in the developing world, and all human use of forest products would who, taking Muir much further than as a token in a process of globalization

88 that is intensifying the economic power was simply a result of the first two fac- applied to some of the early products of multinational companies and inter- tors. With consolidation among their of biotechnology, and this considerably national capital. industrial partners, academic research- undercut support for them, particularly This is a complex argument in its ers would have found themselves work- among advocates of poor and small details, but there are at least four im- ing with a smaller number of firms, farmers. The analysis was also applied portant components to it. One is that even if their actual collaborations with retroactively to ‘Green Revolution’ the period of the late 1980s/early industry had remained unchanged. technologies of the 1960s, resulting in 1990s clearly did see a considerable Academic researchers also found them- a considerable cooling of enthusiasm consolidation within major seed, agri- selves needing industry partners in or- for productivity enhancing technolo- chemical, and forest products compa- der to have freedom to operate within gies in the developing world. Again, nies, as well as the pharmaceutical in- the new era of industrial patents. Some biotechnology just happened to come dustry. This consolidation was sparked have also argued that the nature of bio- along at a time when social scientists by industry’s judgment that genetic technology has tended to blur the dis- were applying these methods to ex ante technologies would be key sources of tinction between research and develop- case studies (see Kalter 1985). Al- profitability in the future, and that cap- ment. The need to acquire patents en- though forestry is different in some turing these profits would depend tered life science departments at uni- respects, it is not wholly different, and upon vertical integration of technology versities in a new way, as well, making the timing of new biotechnologies co- discovery and delivery processes. Al- academic departments seem like private incided with a new level of conscious- though economists debate whether this companies to outsiders. The net result ness among economically disadvan- consolidation has created monopoly was at least the perception that pub- taged producers about the effects of power within the life-science industry, licly funded, putatively not-for-profit technology on their interests. the sheer volume of activity in merg- academic science was pretty much in- The combined upshot of these ers and acquisitions throughout this distinguishable from profit-seeking in- four factors was that biotechnology period could not fail to have captured dustrial product development (see became the poster child for those who the attention of anyone interested in Kenney 1986; Kloppenburg 1988). see technology as a force driving mod- economic inequality (Teitelman 1989). Finally, these events were occur- ern societies toward economic and po- Second, U.S. law for intellectual ring at a time when economists and litical inequality. My own view on the property changed in the 1980s, allow- sociologists had recently completed social ethics of technical change is ac- ing for an expansion of patent rights new analyses of the way that increases tually very close to that of Andrew over genes, gene processes, and even in the efficiency of production technol- Feenberg, who describes himself as a whole organisms. This displaced the ogy were linked to socioeconomic left-leaning pro-socialist critic of capi- Plant Variety Protection Act, which changes in rural areas. The so-called talism (Feenberg 1999). Feenberg be- was weaker both in the sense that it technology treadmill was a staple of lieves that technological innovations provided exemptions for researchers social science analysis throughout the have indeed tended to serve the inter- and for growers propagating plants for 1970s. This analysis showed how more ests of capital throughout history, but their own use. Thus, in addition to efficient production technologies fu- he also believes that this has primarily industry consolidation, the bigger, con- eled a process of change in the struc- been because of the way that owners solidated life-science companies had ture of farming, leading to fewer and of capital have been linked to the de- new legal tools at their disposal for larger farms. This transition was velopers of technology through social concentrating economic power and coupled with a decline in the need for networks. In most cases, it would be exerting control (Fowler 1995). rural service industries and a gradual possible to have the benefits of new Third, there was a simultaneous but inexorable economic decline in technology without the socially desta- shift in the relationship between indus- rural areas. The theoretical techniques bilizing inequalities, if only the devel- try and academic research. In part, this for predicting structural change were opers of technology could be linked in

89 networks with comparatively disadvan- good reason. The ideal of scientific implicit in the question, “How can taged people. Thus, what is needed is objectivity became crucial to the estab- people who are so closely allied with a political reform of the social infra- lishment of rigor and credibility in sci- corporate interests when it comes to structure for developing technology, entific disciplines. Entanglement in social issues be trusted when it comes not opposition to any particular form religious debates over evolution and to environment?” And then there is of technology itself. Although I’ve abortion came to seem less and less one more inference that gets made: never thought of myself as pro-social- relevant to the conduct of science on “Isn’t it dangerous to leave the future ist, and though I think the details of a day-to-day basis. Nevertheless, by in the hands of such people?” In this any reform will prove to pretty com- absenting themselves from any discus- way, the silence of life scientists on so- plex, I find myself in substantial agree- sion of the social networks in which cial issues are translated into positive ment with Feenburg’s social ethic for their work is applied and the technolo- allegations of environmental risk (Th- technology. gies that are adapted from it, life sci- ompson and Strauss 2000). Yet none of this really provides an entists have adopted an ethic that per- Now I must be as clear as possible. argument against tree biotechnology. mits powerful actors to use science si- I do not believe that life scientists as a The implication is that those who are lently in the extension and exertion of whole are in league with corporate in- opposing biotechnology for reasons of their power. The uses to which science terests, nor do I believe that corpora- social ethics are chasing the cape, when has thus been put are often quite de- tions are evil or even that their inter- they should be after the bullfighter. fensible, and have in many cases been ests are antithetical to the social ends I Now, I must qualify my remarks by progressive. Yet even in these cases, I support. I certainly do not believe that saying that I don’t want to ban corpo- submit that the quietness of the alli- life scientists’ failure to become in- rations or the profit motive, nor do I ance between science and economic volved in debates over the social con- want government ownership and con- power is distressing. trol and socioeconomic impact of bio- trol of technology development. As I Ironically, the very quietness of the technology contributes to environmen- said already, the details will turn out life science community is coming to tal risk. What I am saying is that these to be pretty complex. But in each of undermine the very objectivity and are fairly natural inferences for people the four elements described above, it credibility that a previous generation’s to make. I in fact believe that these is the social networking far more than professional ethic was designed to en- inferences have substantially frustrated the use of gene-based or recombinant sure. There is, I submit, a feedback the accomplishment of both environ- techniques that leads to the unfortu- mechanism between the quietude of mental and social causes that I strongly nate social results. We could put an end life scientists with respect to the social support. The expenditure of goodwill to biotechnology tomorrow and it implications of technology and the and intellectual resources in opposing would not improve the situation with public’s willingness to place confidence biotechnology is, in my view, a perfect respect to economic inequality one in their opinions with respect to envi- waste of energy and money by people iota. ronmental risk. Environmental activists whose general values and aims I en- are networked with social activists in dorse. I wish that I were more articu- attempting to constrain the growth of late and effective in making the case for PROFESSIONAL ETHICS global corporate power. They are my view among environmentalists and bound to overhear some of the indict- social activists. But I am also saying And this brings me to my final ments raised against biotechnology in that an ethical failure among life sci- domain. For almost a hundred years, these quarters. The silence of the life entists has contributed to this unfor- the professional ethic in the life sci- science community with respect to tunate result, as well. ences has been to avoid ethical issues such issues can be deafening. The next I do not believe that science and when at all possible. Life scientists inference is unfortunate, in my view, technology automatically translate into came by this ethic honestly and for but not altogether unexpected. It is socially beneficial outcomes. If the out-

90 comes are to be beneficial, there must Raffensperger C and J A Tickner, eds. be conscious and deliberate work at 1999. Protecting Public Health and the Environment: Implementing the building the social networks and think- Precautionary Principle. Island ing through the environmental im- Press, Washington, DC. pacts. Undertaking such conscious and Teitelman, R. 1989. Gene Dreams: Wall deliberate work is, in my view, a Street, Academia and the Rise of Bio- needed and too often lacking element technology. Basic Books, New York. in the professional ethic for the life sci- Thompson, PB. 1997. Food Biotechnol- ences. The symposium we are currently ogy in Ethical Perspective. Chapman involved with is a notable and impor- and Hall, London. tant exception to this general trend, Thompson. PB. 2000. Food and Agri- and I hope that my remarks will be cultural Biotechnology: Incorporat- taken as encouragement to follow ing Ethical Considerations. Cana- through on the work that has begun dian Biotechnology Advisory here. Committee, Ottowa. 40 p. Also available in French: Intégration de facteurs d’éthique à la biotechnologie alimentaire et agricole. Ottowa: REFERENCES 2000, Comité Consultatif Canadien de la Biotechnologie. Callicott, JB, and MP Nelson, eds. 1998. The Great New Wilderness Thompson, PB and SA Strauss. 2000. Debate. University of Georgia Research ethics for molecular sil- Press, Athens. viculture. In Molecular Biology of Woody Plants, Vol. 2, ed. S. M. Jain Feenberg, A. 1999. Questioning Technol- and S. C. Minocha. Kluwer Aca- ogy. Routledge, London. demic Publishers, Dordrecht, pp. Fowler, C. 1995. Biotechnology, patents 485–511. and the third world. In Biopolitics: A Feminist Reader on Biotechnology, ed. V. Shiva and I. Moser. Zed Books, London. Kalter, RJ. 1985. The new biotech agriculture: unforeseen economic consequences. Issues in Science and Technology 13:125–133. Kenney, M. 1986. Biotechnology: The University-Industrial Complex. Yale University Press, New Haven, CT. Kloppenburg, J. 1988. First the Seed: The Political Economy of Plant Tech- nology: 1492-2000. Cambridge University Press, Cambridge. Norton, BG. 1991. Toward Unity among Environmentalists. Oxford University Press, New York.

91 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 92-98. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Ethical and Social Considerations in Commercial Uses of Food and Fiber Crops

Sandy Thomas

ABSTRACT

The introduction of genetically modified (GM) crops into the environment and food chain in Europe has been highly controversial. The prospect of the predicted growth in GM crops over the next 10 years was greeted with outrage and unease, especially compared with the near-indifference shown by most consumers in Canada and the USA. Un- derlying this reaction in Europe have been concerns about possible harm to human health, damage to the environment, and unease about the ‘unnatural’. The introduction of GM crops was perceived by many as the imposition of new and uncertain technology, which did not offer any obvious benefits. In particular, there have been fears over the use of antibiotic-resistance marker genes and the possibility of increasing and unpredictable exposure to allergens. Concerns that herbicide-tolerant crops might encourage use of broad spec- Sandy Thomas is Director trum herbicides and the emergence of herbicide-tolerant weeds, and that insect-resistant of Nuffield Council on crops might damage non-target species, were also widespread. There has also been unease about the commercial exploitation of the technology, particularly in relation to intellectual Bioethics and Senior property issues. It is not yet clear whether the extensive patenting that has taken place in Fellow, Science Policy plant technology has had a restrictive effect on research. In Europe, the debate has been focused almost entirely on GM food crops, and there has been relatively little discussion and Technology Research on the issues raised by the development of forest biotechnology. Most concerns mentioned Unit (SPRU), University above are ethically based and concern principles of rights and general welfare, as well as unease about our relationship with the natural world. There has been broad acceptance of of Sussex, and member of living with a considerable amount of human intervention, but GM technology is perceived the Economic and Social by some as a ‘step too far’. These reactions raise important public policy issues. One objective of public policy is to understand these concerns more fully and to take account of Committee of the Euro- them when regulatory guidelines are being drawn up. The outcomes of GM debate in pean Communities (Sec- Europe have been profound: a de facto moratorium, a decline of commercial investment, increased distrust of scientific advice, and increased anxiety about new technologies. This tions: External Relations paper will consider the implications of the GM crop experience for the development of and Agriculture and Rural forest biotechnology. Development and the Environment). Dr. Tho- he introduction of genetically modified (GM) food crops into the environment and food chain has become highly controversial in the UK, much mas was recently ap- T of Europe, and other parts of the world. The idea that GM crops for pointed to the UK Gov- food and fiber will form a large proportion of the plants grown by farmers in the UK over the next 10 years has been met with a wide range of reactions, from ernment Commission on outrage and unease to acceptance. By contrast, consumers in the United States Intellectual Property and Canada have greeted their introduction with near-indifference. The principal objections to GM crops and the food products made from them have concerned Rights. possible harm to human health, damage to the environment, and unease about the ‘unnatural’ status of the new technology. Concerns over human health have [email protected] focused primarily on the use of antibiotic-resistance marker genes, the possibility of increasing and unpredictable exposure to allergens, and the uncertainty about 92 the long-term effects of novel combi- GM technology. In doing this, there be wrong to carry out a certain genetic nations on human health. Alarmist me- will be opportunities to achieve breed- modification because to do so would dia reports surrounding Pusztai’s ex- ing goals more rapidly and the prom- threaten human health or harm the periments, which claimed to demon- ise of genuinely novel varieties that environment. Because it may be scien- strate immunological damage to rats could not be produced by other means. tifically feasible to undertake a particu- when fed GM potatoes, also served to Whether GM trees are successfully lar experiment or introduce a new type heighten concerns. Environmental adopted or not, will depend in part on of crop for commercial planting, it concerns have arisen on a number of how they are perceived by the public. does not follow that it would be ethi- counts. There have been fears that GM There are, without doubt, lessons to be cally right to do so. To decide what it herbicide-tolerant crops might encour- drawn from the wide range of ethical is right or permissible to do involves, age farmers to use more broad spec- and social issues raised by GM food therefore, bringing together our scien- trum herbicides with a negative impact crops. In this paper, these issues are tific understanding with our ethical on insect and bird life. The risk of considered within the context of the principles to decide what we should do transfer of transgenes to wild popula- UK debate. given the capacities for genetic modi- tions and the possibility that insect-re- fication that have been developed.1 sistant crops might cause damage to Practical reasoning involves weigh- non-target species has also been raised. THE ETHICAL CONTEXT ing up or balancing the benefits of a Many consumers in the UK have also technology like genetic modification objected to what they see as the impo- FOR GENETIC with its potential harms or disadvan- sition of a new and uncertain technol- MODIFICATION tages. However, few questions of prac- ogy that does not appear to offer clear tical reasoning about policy or practice benefits. can be dealt with in a simple form. In 1997, the Nuffield Council on Supporters of GM plant technology Bioethics decided to set up a Working Ethical Principles claim that it will raise agricultural pro- Party to consider the ethical and social ductivity, assist the development of The development of GM plant issues raised by the anticipated intro- safer, more nutritious foods with a technology broadly raises two kinds of duction of GM crops. At that time, longer shelf-life, and contribute to the issues: the scientific and the ethical. there was relatively little public inter- goal of increased food security for the Science is concerned with understand- est in the subject. By the time the poor in developing countries. Against ing the world in which we live and in Council published its report in 1999, these claims, we must set the claims of particular the causal relationships that the public debate in the UK was at its those who assert that not only is GM shape that world: for example, the as- height. Since then, the concerns of food technology a threat to human sociation between genes as a molecu- consumers about the safety of GM health and the environment, but that lar sequence and the characteristics, crops have led to a de facto morato- its introduction will raise the profits of such as resistance to disease, that genes rium on their use in the European the private sector, whilst at the same express. If we are to alter or change the Union (EU) and their disappearance time depriving poor producers of pri- characteristics of plants in an informed from the food chain. The fact that mary commodities access to markets way, then an understanding of such these events have occurred in the ab- and to the new varieties of seed. causal patterns is necessary. By contrast, sence of evidence implicating GM Whether GM technology is morally ethics is concerned with what we ought crops as a risk to human health or to acceptable is a matter of the plausibil- or ought not to do. Ethical principles the environment illustrates the impor- provide standards for the evaluation of 1 tance and the potency of public opin- Nuffield Council on Bioethics. 1999. policies or practices. For example, they Genetically Modified Crops: The Ethical and ion when introducing new technolo- may guide us to decide that it would Social Issues. Nuffield Council on Bioethics, gies. Today, forestry is poised to adopt London.

93 ity of these factual claims and their a series of questions derived from a prompt questions about the rights or importance in the light of moral prin- concern with the principle of justice. welfare of plants, in the way that ani- ciples. Who will be the principal beneficiaries mal experimentation raises questions from the introduction of genetic modi- about the rights of animals. They have, fication and what obligations do they however, provoked a reaction, particu- Welfare, Rights, and have to compensate the losers? larly in Europe, that is difficult to place Justice In its report on GM crops, the within the ethical framework of wel- Nuffield Council did not draw a sharp fare, rights, and justice. Some people There are three main types of ethi- distinction between ethical concerns perceive GM crops as ‘unnatural’ and cal principle that are relevant to the and social issues. On the one hand, those who disapprove of their introduc- evaluation of policies or practices. The ethical principles concern the social tion for this reason are among the first is a principle of general welfare, framework within which society lives. strongest critics. For all the decline in which requires governments and other On the other, there is a need to be formal religion, there remains a deep- powerful institutions to promote and aware of the social and technological rooted belief that society ‘tinkers’ with protect the interests of citizens. The background against which ethical issues nature at our peril. second principle is the maintenance of are discussed. Scientific, ethical, and Others have argued that it is un- people’s rights, such as their rights to social issues cannot be wholly separated ethical to treat nature in an ‘industrial’ freedom of choice as consumers. The from each other, nor should they be. fashion, not merely because of the un- third is the principle of justice, and it It is, broadly speaking, an ethical fortunate consequences of doing so, requires that the burdens and benefits choice to apply scientific knowledge in but because they believe it to be inher- of policies and practices be fairly shared the hope of improving the human con- ently wrong. Whereas the first of these among those who are affected by them. dition. Different societies have set dif- concerns can be accommodated under When we consider the introduction of ferent values on the acquisition and use the principle of the general welfare, the a new technology, such as genetic of scientific information; trying to use second makes ‘the environment’ an modification, we can ask a series of scientific knowledge for what Francis object of ethical concern, regardless of questions in the light of these three Bacon called “the relief of man’s estate” how the environment affects the inter- general principles. Will the technology may seem an obvious choice, but it is ests of human and other animals. GM promote the general welfare by improv- not an inevitable one. It is the ethical crops thus raise ethical issues about the ing food safety or by reducing the use basis of the regulation of commercial rights and wrongs of the ways we af- of chemical pesticides in the environ- development and production of GM fect the environment that are especially ment? Or will the technology pose crops and the promotion of genuinely difficult to analyze and resolve. The unknown risks for consumers and the useful research by government action government of a modern democratic environment that we would be wise that mostly concerns us. society is required not merely to ac- not to take if we are concerned about commodate the deeply held moral con- general welfare? What implications victions of its citizens, but to treat them does the technology have for the rights Natural versus with respect. However, such convic- of consumers, for example, the right to Unnatural tions, on difficult issues such as re- be informed about the food one is eat- search on embryos, are usually held by ing or the environmental impact of a In setting out the three main types minorities no more numerous than particular product that one is buying? of ethical principles that are relevant to those who hold the opposite convic- What implications are there for the the evaluation of GM technology, there tion. Governments cannot legislate or rights of scientists to be free to conduct is a need to consider one substantive regulate by making these convictions their research in ways that protect their issue, namely the ethical status of the the basis of law, but have to pursue intellectual integrity? Finally, we pose natural world itself. GM crops do not policies that can command something

94 close to a reflective consensus. Thus, the interests of well-off consumers or Welfare safety, health, economic well-being, the environment. However, critics ar- and the avoidance of environmental gue that GM crops will bring benefits Another type of welfare concerns degradation are commonly the goals of only to the producer, not to the con- the pleasure that people derive from government policy. sumer, and that any risk of harm can- living alongside the natural world, even not therefore be justified. Both views when they may not have the opportu- imply that it is right to balance the nity to visit it. The pleasure may ex- The Precautionary good achieved against the harm im- tend to simply ‘knowing that it is there’ Principle posed. A stringent interpretation of the and could be visited. There are also is- precautionary principle would preclude sues of rights here. One particular con- The concern of government with such balancing. It may, however, be flict is between the rights, for example, the welfare of its citizens underlies best interpreted, not as part of a cost/ of forestry companies to exploit the much current regulatory practice. One benefit calculation, but as a principle environment and the rights of others, of the duties of forestry companies in- governing how such calculations such as environmentalists, who might troducing GM trees, whether in experi- should be made. The principle does wish to preserve the environment as an mental trials or for commercial use, not yield very definite prescriptions, amenity. Other welfare considerations would be to ensure that they do no but does indicate caution for those who that relate to the environment in a harm or, that any harm is so slight as would introduce new technologies. more straightforward way are con- to be broadly acceptable. The regula- cerned with the direct consequences of tory system for GM crops and their genetic modification. The main con- products in the EU is predicated on ETHICAL ISSUES AND cern here is that we risk damaging the this basic proposition. The prevention economic and amenity resources of the THE NVIRONMENT of harm is sometimes extended to pro- E environment. These concerns are closer mote the adoption of the precaution- to the traditional role of public policy Public concerns about the envi- ary principle. This principle sets the in ensuring safety.3 However, one of the ronment have been increasing over the avoidance of harm to consumers and difficulties in assessing the potential past 40 years. Most of these concerns the environment at the head of the list risks of new technologies is that there are ethically based and the majority of regulatory goals. However, the uni- is often an absence of agreed measures center on welfare, that is, the welfare versal adoption of the precautionary of the relative seriousness of different of present as well as future generations. principle risks an imbalance between kinds of potential harms. the avoidance of harm and the achieve- Clearly if the resources within our en- ment of a positive good. vironment became so badly damaged The precautionary principle can that it was not possible to sustain hu- Rights be viewed as a simple, welfare-based man, animal, and plant life, the loss of principle.2 As such, it raises familiar welfare would be infinite and the moral The genetic modification of plants problems, of which the most important responsibility of those who brought also raises questions of rights. For ex- is to define the conditions under which about such environmental disasters ample, do forestry companies have a the avoidance of harm should take pri- undeniable. right to pose environmental risks, how- ority over the attempt to do good. ever small, in pursuit of benefits, Common sense suggests that the devel- whether these are profits, consumer opment of crops that substantially re- benefits, or both? In general, individu- 2 duce hunger or improve nutrition in Nuffield Council on Bioethics. 1999. Genetically Modified Crops: The Ethical and the developing world would justify Social Issues. Nuffield Council on Bioethics, 3 Krebs, JR, et al. 1999. The second Silent running the risk of modest damage to London. Spring? Nature 400: 611– 612.

95 als and others have the right to risk debate some of the NGOs lost public non-segregation of GM from non-GM their own well being, but not that of support through their repeated destruc- foodstuffs was a critical and costly er- others. A balance has to be struck be- tion of field trials, their impact in the ror leading to a lack of consumer tween, for example, the legitimate de- absence of specific evidence about the choice, which mattered to the con- sire of forestry companies to create possible risks of GM crops was sub- sumer who was concerned about safety. wealth, the needs of consumers, and stantial. The fact that the main culprit was a the continuing need to maintain a sus- ubiquitous food ingredient in the form tainable environment for future genera- of GM soya meant that the effects were tions. The Consumers far reaching. The multinationals were moreover portrayed by some as being Consumers in the UK were par- in pursuit of profit, despite uncertain- THE CONTEXT FOR THE ticularly influential in the debate be- ties over the safety of the technology, cause the food retailers were particu- both for human health and the envi- UK GM DEBATE larly responsive to their views. Once ronment. There were also concerns doubts were raised about the safety of over issues relating to ownership of and In the UK debate on genetic GM food in the UK supermarkets, fu- access to the technology. modification of crops, four constituen- eled by alarmist media reports and a In the case of patents for GM cies proved to be of particular signifi- lack of discriminatory food labeling, crops, two public concerns have been cance. These were the non-governmen- consumers were quick to voice their visible.4 One has been with the legiti- tal organizations (NGOs), consumers, concerns. Their lack of confidence in macy of ‘owning life’. Various interest the public sector researchers, and the government advisory bodies after the groups have been campaigning, on private sector in the form of the seed BSE fiasco, the perceivedlack of trans- ethical grounds, against the concept and agrochemical industry. parency in labeling GM food products, that property rights can exist in genetic and the lack of clear benefits were po- material or activities associated with it.5 tent factors, which combined to erode The NGOs While some of these objections can be public confidence at a rapid rate. Re- attributed to deeply held beliefs, in tailers effectively removed approved others, a misunderstanding of the The NGOs who were actively GM food ingredients and products patent system may play a part. The campaigning against the introduction from the supermarket shelves and re- other concern has been with the pat- of the new technology had, in some placed them with non-GM supplies. enting of GMs and the research tech- cases, a wider environmental agenda, niques associated with the development of which GM crops were but a part. of GM crops. Patent holders may be At a time when public confidence in The Private Sector reluctant to license patents with broad the government scientific advisory sys- claims to key technologies to their tem was at an all-time low, some of the The multinational companies in- competitors or to public-sector research NGOs, particularly Greenpeace and evitably played a significant role in the Friends of the Earth, were able to cap- debate. As agents for the development ture considerable public sympathy. of the new technology, they were vili- 4 Nuffield Council on Bioethics. 1999. Both of these NGOs and others ran fied by the majority of NGOs and Genetically Modified Crops: The Ethical and Social Issues. Nuffield Council on Bioethics, highly professional campaigns, which viewed with distrust by much of the London. were highlighted by the media at a time public. The role of Monsanto in alien- 5 when the UK Government was rela- ating the UK consumers, a key factor See Dworkin, G. 1997. Property rights in tively silent on some of the issues. Al- genes. Philosophical Transactions of the Royal in sparking the controversy, has been Society of London Series B (Biological Sciences) though at a later stage in the ongoing widely discussed elsewhere. The initial 352: 1077–1086.

96 institutions. Companies may seek pat- The Scientists not mirrored in the UK. As a result, ents that will not advance research or the UK regulations concerning GMOs production, but that deter competitors Finally, the fourth grouping that were not excessively stringent, and re- and prevent research in areas that has been important in the UK contro- search, development, and production threaten their monopolies. In addition, versy consists of the scientists, namely of new GMO-related products was al- public laboratories are increasingly de- those in the public sector. At a time of lowed to progress. At this stage, these manding a royalty on future commer- uncertainty, many looked to the products were associated with the phar- cial developments from their publicly country’s scientists for reassurance and maceutical sector, which was at an ear- funded colleagues in their terms for li- impartiality. However, it proved diffi- lier stage of development in the appli- censing access to research tools. Con- cult for researchers working in the area cation of GMOs than was the agro- solidation in the agrochemical and seed of GM to make much of an impact on chemical industry. industry has continued to reduce the the debate for two main reasons. First, That situation has changed pro- list of owners of the important ‘en- in the UK (unlike in the USA), the foundly. Recent public reaction against abling’ intellectual property for plant majority of scientists have little or no the introduction of GM crops into the genetic modification and plant molecu- experience in dealing with the public. environment and the inclusion of GM lar genetics.6 There are now around six Many were reluctant to be drawn into ingredients in food has been more in- major industrial groups who between a debate in an area where the public tense in the UK than in any other Eu- them control most of the technology was disaffected and substantially preju- ropean country and shows little sign of that gives freedom to undertake com- diced against the technology. Second, abating. Despite the lack of evidence mercial R&D in the area of GM a number of scientists who had exper- to show that GM crops threaten our crops.7 There has also been concern tise in GM technology were perceived environment or health in new and sig- over the patenting of DNA sequences as non-independent because they were nificant ways, a wide range of organi- and particularly ESTs (expressed se- participating in research partnerships zations have disassociated themselves quence tags).8 The question of patent with industry. The fact that research from GM crops and their products or dependency9 for partial sequences or collaboration with the private sector is are considering doing so—major food where the gene function is unknown increasingly the norm today in many retailers, agricultural landlords (such as is important not only in the pharma- public-sector institutions around the the Duchy of Cornwall), local council ceutical industry, in relation to human world did not serve to diminish pub- authorities, restaurants, and schools. genes, but also in the plant science sec- lic skepticism. Over the past 24 months, the UK pub- tor, both public and private. lic has not been involved in an informed debate, but rather has been the 6 Nuffield Council on Bioethics. 1999. THE OUTCOMES OF target of a stream of relentless and Genetically Modified Crops: The Ethical and 10 Social Issues. Nuffield Council on Bioethics, negative propaganda. However, the London. THE UK GM DEBATE recent decision by Greenpeace to join

7 some UK fringe groups and destroy Nuffield Council on Bioethics. 1999. The scope and intensity of the Genetically Modified Crops: The Ethical and GM maize trials may prove to be some- Social Issues. Nuffield Council on Bioethics, controversy over GM crops in the UK thing of a turning point. The UK pub- London. took most people by surprise. In gen- lic is wary of GM food but, one sus- 8 ESTs are partial DNA sequences which eral, over the past decade, the UK pub- pects, would not support the wholesale represent genes that are turned on in a particular lic has not been strongly opposed to abandonment of the very research that tissue type or organism. the development of biotechnology. The will provide answers to some of the 9 There is concern over the extent to which strong reactions against recombinant patents on ESTs may impose dependency or ‘reach DNA technologies in Germany and 10 through’ to subsequent patent applications with Beringer, J, and H Wallace. 1999. Natural full-length sequences. Denmark during the early 1990s were justice? New Scientist 14 August, p.47.

97 questions the technology poses. The ered. There is an ongoing need to build Wambagu, F. 1999. Why Africa needs future of GM crops in Europe has in public confidence and trust in the new agricultural biotech. Nature 400: 15–16. a number of ways been postponed, GM technologies. Progress will only be which some interpret as a de facto made if there is transparency and open- Krebs, JR, et al. 1999. The second Si- moratorium.11 This has resulted in a ness in the scientific advisory process lent Spring? Nature 400: 611–612. decline of commercial investment in for regulating and approving the intro- GM crops. The role of plant biotech- duction of the new trees. At the out- nology research in the public sector set, an open dialogue between the vari- too, may need to shift in emphasis. ous stakeholders to identify and discuss There is a legacy of increased distrust the range of issues raised will help to of scientific advice, already weakened prevent the development of a pro- at the outset, and increased anxiety longed and unbalanced debate in about new technologies. Elsewhere in which scientific evidence played a the world, GM crops continue to be minimal role. It is worth noting that researched and grown. The United the initial U.S, GM field tests in 1986 States, Argentina, and China are sub- and 1987 took place after open discus- stantially committed, and a number of sions among scientists, regulators, developing countries see a role for GM farmers, and environmentalists.13 technologies in improving their food security.12 It is estimated that U.S. citizens eat GM food products at every REFERENCES meal and as yet, after over eight years, no adverse events to health have been Beachy, RN. 1999. Facing fear of bio- reported. technology. Science 285: 335. So in conclusion, what are the les- Beringer, J, and H Wallace. 1999. Natu- sons that the forest biotechnology sec- ral justice? New Scientist 14 (Au- tor might draw from this protracted gust): 47. and difficult debate? First and fore- Dworkin, G. 1997. Property rights in most, the issues that will be raised by genes. Philosophical Transactions of the introduction of GM trees will not the Royal Society of London Series B be confined to the scientific. Careful (Biological Sciences) 352: 1077– 1086. thought must be given to how members of the public perceive the impact McCabe, H, and D Butler. 1999. Eu- of existing forestry practices on the ropean Union tightens GMO regulations. Nature 400:7. environment and how GM technology might change this perception. Issues of Nuffield Council on Bioethics. 1999. rights and welfare for both consumers Genetically Modified Crops: The Ethical and Social Issues. Nuffield and producers will need to be consid- Council on Bioethics, London.

11 McCabe, H, and D Butler. 1999. European Union tightens GMO regulations, Nature 400:7.

12 Wambagu, F. 1999. Why Africa needs 13 Beachy, RN. 1999. Facing Fear of Biotechnol- agricultural biotech. Nature 400: 15–16. ogy. Science 285: 335.

98 REGULATORY POLICIES

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 101-104. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Regulation of Transgenic Plants in the United States

David S. Heron John L. Kough

ABSTRACT

The Agencies primarily responsible for regulating biotechnology in the United States are the US Department of Agriculture (USDA), Environmental Protection Agency (EPA), and the Food and Drug Administration (FDA). Products are regulated according to their intended use, with some products being regulated under more than one agency. Geneti- cally engineered plants must conform with standards set by State and Federal marketing statutes such as State seed certification laws, the Federal Food, Drug, and Cosmetic Act (FFDCA), the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), the Toxic Substances Control Act (TSCA), and the Plant Protection Act (PPA). There are no national requirements for varietal registration of new plant varieties. The U.S. unified home page for biotechnology has further information (www.aphis.usda.gov/biotech/OECD/ usregs.htm). Within USDA, the Animal and Plant Health Inspection Service (APHIS) is responsible for protecting U.S. agriculture from pests and diseases. APHIS regulations provide procedures for obtaining authorization prior to importation, interstate movement, or field testing of a regulated article in the United States. The regulations also provide for a peti- David S. Heron works for tion process for the determination of nonregulated status. Once a determination of nonregulated status has been made, the product (and its offspring) no longer require APHIS the USDA Animal and authorization for movement or release in the US. EPA sets standards for the safe use of Plant Heath Inspection pesticides, both chemical and those that are produced biologically. The authority of FIFRA is used to regulate the distribution, sale, use and testing of plants and microbes producing Service, Plant Protection pesticidal substances. Under FFDCA, EPA sets tolerance limits for substances used as pes- and Quarantine, Permits ticides on and in food and feed, or establishes an exemption from the requirement of a tolerance. EPA also establishes tolerances for residues of herbicides used on novel herbi- and Risk Assessments, cide-tolerant crops. FDA regulates foods and feed derived from new plant varieties under the authority Riverdale, MD. of FFDCA. FDA policy is based on existing food law, and requires that genetically engi- [email protected] neered foods meet the same rigorous safety standards as is required of all other foods. Consistent with its 1992 policy, FDA expects developers to consult with the agency on John L. Kough works for safety and regulatory questions. the Biopesticides and ur joint paper today will be a brief overview of the Federal regu- Pollution Prevention latory framework for the regulation of transgenic plants in the United Division in EPA’s Office O States. We will also discuss some information that may be relevant for those interested in the development of transgenic forest trees, and we will of Pesticide Programs, provide some sources for additional information, much of which is available at Washington, DC. agency web sites. The United States unified home page for biotechnology has further infor- [email protected] mation and links to the agency sites (See http://www.aphis.usda.gov/biotech/ OECD/usregs.htm).

101 COORDINATED which Agency has regulatory responsi- stances currently found in food. Many bility. This information is available at of the food crops currently being de- FRAMEWORK FOR THE US Unified Home page for biotechnol- veloped using biotechnology do not REGULATION OF ogy at http://www.aphis.usda.gov/ contain substances that are significantly biotech/OECD/usregs.htm. different from those already in the diet BIOTECHNOLOGY and thus do not require pre-market approval. Consistent with its 1992 In 1986 the Coordinated Frame- FDA Role policy, FDA expects developers to con- work for the Regulation of Biotechnol- sult with the agency on safety and ogy was adopted by federal agencies As a part of the Department of regulatory questions (for further infor- (see 51 Fed. Reg. 23302; June 26, Health and Human Services, FDA mation see http://vm.cfsan.fda.gov/ 1986) to provide a coordinated regu- regulates foods and feed derived from ~lrd/biotechm.html). Because of this latory approach intended to ensure the new plant varieties under the author- meeting is focusing on forest trees, we safety of biotechnology research and ity of the Federal Food, Drug, and will not discuss food safety reviews in products by using existing statutory Cosmetic Act. FDA policy is based on greater detail, but refer interested par- authority and building upon agency existing food law, and requires that ties to the citations above. experience with agricultural, pharma- genetically engineered foods meet the ceutical, and other products developed same rigorous safety standards as is re- through traditional genetic modifica- quired of all other foods. FDA’s bio- APHIS Role tion techniques. The Coordinated technology policy treats substances in- Framework is consistent with the con- tentionally added to food through ge- Within USDA, the Animal and clusion of the 1987 report of the Na- netic engineering as food additives if Plant Health Inspection Service tional Academy of Sciences that found they are significantly different in struc- (APHIS) is responsible for protecting that the potential risks associated with ture, function, or amount than sub- U.S. agriculture from pests and dis- genetically engineered organisms eases. Under the authority of the Plant Protection Act, should be similar in kind to those as- Table 1. Regulatory responsibility for common types of sociated with traditionally bred organ- engineered plants. APHIS oversees the impor- isms. tation, interstate move- Regulatory The Coordinated Framework ment, and the field testing, New trait/organism review by Reviewed for identified three main agencies: U.S. under controlled condi- Department of Agriculture, Environ- Viral resistance in USDA Safe to grow tions, of most genetically food crop EPA Safe for the mental Protection Agency, and Food engineered organisms environment and Drug Administration. The Coor- (termed “regulated articles” FDA Safe to eat in the regulations), particu- dinated Framework further emphasized Herbicide tolerance USDA Safe to grow larly most new plant variet- the need for ongoing interagency co- in food crop EPA New use of ordination mechanisms to ensure that companion herbicide ies, and assures that these policy and scientific questions would FDA Safe to eat new varieties are as safe to be addressed across agencies. Products Herbicide tolerance USDA Safe to grow use in agriculture as tradi- of biotechnology are regulated accord- in ornamental crop EPA New use of tional varieties. ing to their intended use, with some companion herbicide All of those actions re- Modified oil content USDA Safe to grow products being regulated under more quire permission from in food crop FDA Safe to eat APHIS, in essence a certi- than one agency. Modified flower color USDA Safe to grow fication that the action will Table 1 contains some examples of in an ornamental crop common types engineered plants and be performed in a safe man-

102 ner. APHIS authorizes field tests un- rectly or indirectly cause disease or dam- garding the petition process at http:// der either permitting or notification age to plant, plant parts, or processed www.aphis.usda.gov/biotech/ procedures that limit the persistence of products of plants. These include find- petguide.html. Further information on the viable test plants in the environ- ings that the new plant variety other aspects of the APHIS oversight ment at the conclusion of the test. of biotechnology can be found on the 1. Exhibits no plant pathogenic prop- Both procedures require the applicants at http://www.aphis.usda.gov/biotech/ erties to meet the same standard of safety in Once a Determination of the conduct of the trials. 2. Is no more likely to become a weed Nonregulated Status is issued, the new To date, APHIS has authorized than the non-engineered plant variety may be treated, from USDA’s thousands of field tests for more than perspective, like any other variety of 50 plant species, but relatively few of 3. Is not likely to increase the weedi- the crop, i.e., it may be grown, tested, these have been tree species. However, ness of any other plant with which or enter traditional crop breeding pro- APHIS has authorized field tests for it is sexually compatible grams without any other special over- transgenic spruce, pine, poplar, walnut, 4. Will not cause damage to processed sight with respect to the APHIS regu- citrus, cherry, apple, pear, plum, and agricultural commodities lations. Once any other requirements persimmon. The public database for all from other agencies are satisfied, the APHIS authorizations can be accessed 5. Is not likely to harm other organ- plant can enter into commerce and be at http://www.nbiap.vt.edu/cfdocs/ isms that are beneficial to agricul- sold. All the normal phytosanitary con- fieldtests1.cfm ture. trols apply to these varieties just as they As testing of a “regulated article” Depending on the nature of the do to traditional varieties. proceeds, the developer gathers infor- modified plant, other relevant issues mation to confirm that the product has may be considered, also. Electronic the new intended property, and is as copies of past APHIS assessments (de- EPA’s Role in Assessing safe to grow in the environment as tra- cision documents) can be obtained at ditional varieties. Under the APHIS Pest-resistant Trees the APHIS web site regulations, when enough information http://www.aphis.usda.gov/biotech/. When a tree species is intention- is gathered, the developer can petition As part of the petition process, ally modified to enhance its resistance APHIS to make a Determination of APHIS also conducts an environmen- to pests, these plants are expressing a Nonregulated Status. tal assessment to ensure that any envi- new substance termed a plant-incorpo- When APHIS receives a petition, ronmental impacts are not likely to be rated protectant (PIP). A PIP is defined a team of agency scientists begins the significant as a consequence of the review, and the agency announces to as both the pesticidal substance and the agency determining that the plant va- the public that the petition has been genetic material necessary for the pro- riety or line will no longer be consid- received. If the review team decides duction of that substance. A PIP ex- ered a regulated article (nonregulated that the petition is complete and ready pressed in this tree must be registered status). This assessment includes a con- for full review, the agency makes the with the Environmental Protection petition available for public review and sideration of the potential effects on Agency (EPA) before the tree can be comment. As part of the review pro- the “wider” environmental. This is grown commercially. Under the re- cess, APHIS considers all available mandated under the National Environ- quirements of the Federal Insecticide, information and public comments be- mental Policy Act of 1969, that ad- Fungicide and Rodenticide Act fore making a determination that a dresses the general decision-making (FIFRA), EPA must determine that use plant will no longer be considered a process for all government actions. of the PIP does not present an unrea- regulated article. All of these petitions require case- sonable adverse effect to the environ- In these reviews, the APHIS stan- by-case review. APHIS provides users’s ment, taking into account the eco- dard is that an organism must not di- guides and additional information re- nomic, social and environmental costs

103 and benefits of that use. If the tree also possible. The biological fate is the February 1998—Insect Resistance has dietary uses (e.g., walnut, pinyon movement of the PIP gene into related Management pine), EPA must also make a finding plant species and potential for PIP ex- December 8, 1999—Product Char- that there is a reasonable certainty that pression in outside its intended plant acterization and Non-Target Effects no harm will result from the aggregate host. exposure to that PIP. It is important to In addition to the risks and ben- October 18-20, 2000—Re-Evalua- realize that plants modified to be re- efits of PIPs, EPA has also been in- tion of PIPs Expresssing Bacillus sistant to a herbicide are not consid- volved in examining the risks of wide- thuringiensis Toxin ered to express a PIP, but rather may scale use of PIPs selecting for resistance have a new use for the herbicide. EPA in targeted pest populations. EPA be- February 29, 2000—Cry9C and may have to consider the risks associ- lieves that many of the PIPs express- other Non-Digestible Proteins ated with the new exposures to that ing Bacillus thuringiensis toxins provide June 7, 2000—Protein Toxicity As- herbicide’s residues, but not the herbi- a safer means of controlling insect sessment cide resistant plant itself. pests. Therefore, it has been deter- EPA must examine the use of the mined to be in the public interest to November 28, 2000—StarLink Di- PIP for its environmental safety in that insure the long-term efficacy of these etary Exposure Assessment tree species. To do this EPA makes use PIP products, as well as protecting the July 17-18, 2001—CDC/FDA of the detailed description of the PIP usefulness of pesticides based on Bacil- Analysis of StarLink Corn and Re- as expressed in the plant provided as lus thuringiensis bacterium as an alter- fined Dietary Exposure Assessment. part of the USDA-APHIS review. In native to some of the more toxic addition EPA reviews data generated chemical alternatives. Many work- In addition, the website for EPA’s specifically to address potential envi- shops, public meetings, and SAPs have Biopesticides and Pollution Division in ronmental effects and their associated been devoted to the topic of insect re- the Office of Pesticide Programs pro- risks. The environmental safety exami- sistance management, and EPA contin- vides up to date information on regis- nation is based on two aspects: the fact ues to be actively involved in insuring tered PIPs as well as any recent scien- that the PIP has been introduced to responsible use of this pest-control tific or regulatory information that is control a pest and therefore has some technology. cogent to PIPs. The website is also a level of toxicity to a pest species, and As is true for the development of good source on microbial and bio- the potentially new exposures provided any risk assessment guidelines, EPA has chemical pesticides that are available by the PIP’s expression in the modified consulted with scientific experts in for use: plant. The toxicity examination for public meetings called Scientific Advi- www.epa.gov/pesticides/biopesticides species other than the target pest is sory Panels (SAP) to receive input on similar to that utilized for microbial EPA’s approach for assessing the envi- pesticides. In addition to the direct ronmental safety of PIPs. For a discus- toxicity issue, there is the consideration sion of the data necessary for an envi- of both the biological and chemical fate ronmental assessment, please visit the of the pesticidal substance of the PIP. following website summarizing the re- The chemical fate is the off-site move- sults of recent SAPs: ment and environmental persistence of www.epa.gov/scipoly/sap/index.htm the PIP. This would include the poten- The website is organized by the tial for long-term exposure in the soil date of the SAP so the following dates and possible effects to sensitive non- and subject matter serve as handy ref- target species if exposure to significant erences for these reports: amounts of PIP expressing pollen were

104 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 105-110. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf International Regulation and Public Acceptance of GM Trees: Demanding a New Approach to Risk Evaluation

Sue Mayer

ABSTRACT

Establishing agreement about the environmental safety of releasing GM (genetically modified) trees to the environment will pose more challenges than for GM crops. The data considered necessary to determine genetic stability, the extent and rate of gene flow, persistence, and invasiveness of a GM food crop typically involves experiments lasting over several generations conducted under different environmental conditions. The characteristics that make trees so attractive to genetic engineers, namely their long generation times and slow growth, mean that collecting similar data about their environmental performance will require much long periods if it is to match that considered acceptable for GM crops. However, having to conduct ecological research over many years would compromise the economic viability of GM trees and conflict with the claimed benefits of speeding up tree domestication and improvement. Reconciling these issues in a manner that commands public confidence will be a particular challenge for the regulation of GM trees. Judgements will have to be made much more explicitly, given the lack of data, revealing the inevitably subjective nature of risk assessment. Even more demandingly, the approach that is taken will either have to satisfy or be sensitive to different social, economic, and regulatory regimes in different countries to avoid acrimonious trade disputes. Research on public attitudes to GM crops in Europe and the United States indicates that there is little confidence in the institutions dealing with safety and that public concerns are not captured by the present framing of the regulatory system. Therefore, new approaches that combine deliberative methods involving the public in defining criteria and their relative importance with expert knowledge will be required. In addition, examining different options across a range of different criteria—physical, social, economic, and ethical—will also be needed, as claims of safety will be more difficult to sustain. In this situation, rather than giving the false illusion that science can determine whether GM trees are safe or dangerous, science contributes and is necessary for decision making, but does not form a sufficient basis alone.

rees are long lived, large, and have long reproductive cycles. These char Sue Mayer is the Director acteristics have contributed to their relative lack of domestication and of GeneWatch UK, The T ‘improvement’ through conventional breeding techniques. In contrast to, say, maize or rice where one, two, or more (with judicious use of both hemi- Mill House, Manchester spheres) generations can be obtained in a single year, even the most rapidly growing Road, Tideswell, trees take at least four or five years and often much longer to produce the next generation. However, genetic engineering brings the prospect of speeding up the Derbyshire, SK17 8LN, domestication of trees to fit with 21st-century forestry demand for more effi- UK. cient pulp and paper production. Speed, uniformity, and ease of processing and management are the driving forces behind tree breeding for this sector (Tzfira et [email protected] al. 1998; Merkle and Dean 2000).

105 In parallel with excitement about in Europe show that a conventional judgements in deciding what the rel- the potential to use genetic modifica- risk-assessment approach will not suc- evant criteria are (the framing of the tion (GM) comes concern about its ceed in reconciling these fundamental risk assessment), their relative impor- environmental, social, economic, and tensions. Unless a more comprehensive tance, and deciding upon the accept- political implications. Exploitation of and inclusive approach to assessing the ability of risks given the inevitable sci- GM trees is linked with ‘intensive, future of GM trees is adopted, polar- entific uncertainty that exists (Stirling short-rotation (e.g., 3–25 years) plan- ized debate and conflict will be the 1998; National Research Council tations’ (Strauss et al. 1999), which outcome. 1996). These value judgements may raises questions about environmental vary according to the social, political, sustainability, equity, and aesthetics and cultural context in which they are (Sampson and Lohman 2000). In ad- VALUE JUDGEMENTS IN made. For example, until the recent dition, gene movement from GM trees debate about whether Bt toxin in GM to wild relatives or the establishment RISK ASSESSMENT OF maize can affect monarch butterflies, of GM trees outside a plantation is GMOS regulatory concerns and conflict in the possible, perhaps inevitable, as is con- United States focused on whether re- sidered to be the case for GM food The conventional approach to risk sistance would emerge in target organ- crops (Moyes and Dale 1999). Because assessment of GM organisms focuses isms, compromising future use of Bt as even those trees grown in plantations, on the genetically modified trait and an insecticide, and strategies to avoid such as eucalyptus and aspen, are rela- how it affects the phenotype of the this. In Europe, the emphasis in the tively undomesticated, related native organism. In the United States, the debate about environmental impacts of trees—which they can cross fertilize— claim is that the regulatory risk assess- GM Bt maize has been on the poten- often surround them. Tree pollen and ment of a GM organism does not con- tial for tri-trophic effects on non-tar- seed can travel long distances and some sider the process of genetic modifica- get species (Levidow and Carr 2000). trees can spread widely via suckers, so tion, but rather looks at the product. Even inside the European Union, there if GM trees became more invasive they In Europe, in contrast, the process of have been disputes between Member could spread widely. Furthermore, be- genetic modification is the trigger for States over the boundaries of regula- cause GM trees are being developed for regulation. However, a comparison of tion, causality, and acceptability a particular type of forestry, their de- the data requirements for assessment (Levidow et al. 1997), which, with the velopment (or not) will also have eco- shows little difference with the process changing political situation in Europe, nomic, social, and political implica- of genetic modification being included have led to the revision of the Direc- tions. (seen in, for example, evaluation of the tive covering the environmental safety These far-reaching questions stability of the transferred gene(s)) on of releasing GMOs (2001/18/EC). about environmental safety will be ex- both sides of the Atlantic. The claimed Whilst value judgements may be tremely difficult to answer. How will differences are more to do with the inevitable in making assessments about the regulatory system cope with the political demands of the industry in the the desirability or otherwise of a par- competing demands of the industry to United States, that gene modification ticular course of action, problems may capitalize on the potential for speed, be constructed as being no different arise when the values and judgements with the natural constraints of the bio- than conventional breeding, and the of those taking the decision do not logical cycles of trees and increasing different way in which it is viewed in coincide with those held by the pub- public concern about the safety of GM Europe (Mayer 2001). lic. In the UK, there is evidence of a organisms? This paper examines the As well as the characterization of dislocation between public attitudes to debate over the safety of GM trees and the regulatory process, it is well recog- GM crops and foods and those of the its social and political shaping, and ar- nized that risk assessments of GMOs Government and its advisors. This mis- gues that lessons from the GM debate and other technologies involve value match extends to the scope and fram-

106 ing of regulations and policy and is adopted a position statement in Sep- technology by industry, lead to an ex- thought to underlie the present bitter tember 1999 that laid out their vision pansion of research on benefits and controversy about GM crops (Grove- of the future of GM trees and their risks, and help to instill confidence in White et al. 1997). The problems sur- regulation (for the full text, see http:// the public that environmental integrity rounding GM crops in Europe are of- fsl.orst.edu/tgerc/iufro_pos- is being safeguarded. ten put down to a post-BSE (mad cow) statm.htm). Prefaced by an explanation An ideological commitment to atmosphere of lack of trust in institu- of a “position statement”, it states that GM trees and intensive plantation ap- tions. This lack of trust in institutions it comes from “a society of profession- proaches to forestry is seen in the first managing safety in Europe is related to als who know the scientific aspects of paragraph calling, among other things, their (in)ability to deal with complex an area of technology controversy in on the highly political and scientifically situations bringing together the nature depth”, thereby positioning the authors questionable justification of mitigation of the risks, their irreversibility, and as authoritative by using the social sta- of greenhouse gases. Science is then their potential for surprises, with tus of science. However, as the follow- enlisted in the second paragraph to re- whether there is a need for the prod- ing extract from the conclusion shows, strict the terms of regulatory controls uct and who bears the risk. All of these the statement not only moves outside as it has been elsewhere (e.g., James et factors have combined to bring the call the expertise of tree molecular biolo- al. 1998), with the potential benefits for a choice about whether to eat GM gists, but ventures into the political of GM trees being used as a justifica- foods or not. In the United States, the domain in its efforts to shape the scope tion for a narrow framing of the risk apparent lack of vociferous public re- of risk evaluation and how uncertain- assessment. As with the GM crops de- jection of GM foods is characterized as ties should be viewed: bate (Levidow and Carr 2000), a sci- there being more confidence in tech- Tree plantations are expected to con- entific base (or ‘sound science’) is be- nological developments in the United tinue to expand as a result of increasing used as an ideology to stifle debate States compared to Europe (e.g., Dale ing demand for their many renewable and to give control of determining 1999). However, recent research shows products, their importance to mitiga- what are the credible risks and benefits that concerns about GM foods are in- tion of greenhouse gases, and the envi- to a certain interest group. Here a com- creasing in the United States, and, as ronmental protection afforded to large munity is acting politically to further in Europe, it is the regulatory process areas of native forest. It is therefore im- its own interests, whilst making claims that is being questioned (Horning portant that rates of plantation produc- to be ‘scientifically’ justified. Priest 2000). tivity be made as high as possible within There are plenty of questions to the context of good environmental ask about whether a focus on specific stewardship. Transgenic technology, trait(s) is a sufficient to basis upon wisely used, promises significant eco- SHAPING THE which to address environmental con- nomic and environmental benefits. REGULATION OF GM cerns about GM trees. It clearly pre- It is critical that regulatory and certifi- cludes any broader assessment of for- cation systems have a strong scientific TREES estry and whether the intensification base that focuses on specific transgenic heralded by GM is to be welcomed. traits and their deployment, and that How well will the evolving regu- Furthermore, the indirect impacts of the demands of compliance are reason- lation of GM trees learn from the able in light of the credible risks and GM trees include changes to forestry problems encountered during the in- benefits expected. These systems should practices which demand attention. For troduction of GM crops and foods? also attempt to minimize the burdens example, the introduction of herbicide- One influential body, the International required for conduct of field trials while tolerant crops would change the pat- Union of Forestry Research mandating research on significant envi- tern of herbicide use with knock on Organisations (IUFRO) Working Party ronmental impacts expected. Such sys- implications for biodiversity. Similarly, on Molecular Biology of Forest Trees, tems would foster investment in bio- Bt-toxin-based approaches to insect

107 control would alter exposure to insec- scale experimental releases do not in- new European approach takes account ticides. In contrast to food crop farm- vestigate environmental impacts of indirect effects, socioeconomic im- ing, insecticide use in forestry is rather (Mellon and Rissler 1995). The un- pacts are excluded because of a long- restricted because of the scale of for- knowns and uncertainties in the eco- standing policy commitment to bio- ests and costs involved. However, ex- logical impacts of GM trees are enor- technology as a driver of industrial posure to Bt continuously over many mous, and any effects are likely to re- competitiveness, and thus, European years could have serious impacts on main undetected for many years and decisions may also fall foul of public insect populations that will be ex- be irreversible. Even if male-sterile trees opinion if this is not agreed for GM tremely difficult to predict. In Europe, are produced, how confident can we be trees (Mayer and Stirling 2001). the relevance of indirect impacts on that the transformation will be stable? biodiversity has been recognized, and Studies of non-GM hybrid trees may experiments are beginning to consider show that although gene transfer can FINDING MORE them (Firbank et al. 1999), but a trait- take place over long distances, offspring based approach to risk tends to do not compete well against native ROBUST APPROACHES marginalize these indirect impacts and trees; how confident can we be that the TO RISK EVALUATION preclude research. same will hold true for the traits intro- In addition, there is a question duced through genetic modification The main justification for inten- about whether safeguards put in place and that there have been no unex- sifying forestry practices via GM is to will be effective or practicable. The pected changes as a result of the gene meet the increasing demand for pulp Starlink contamination of maize for introduction? It is clear that the level and paper and the need to make its human consumption in the United of confidence that is expected in order production more efficient (Tzfira et al States, and the nearly one-third of U.S. to make judgements about GM crops 1998; Sederoff 1999). However, this farmers using GM Bt crops who may safety (which itself is contested as in- begs the question as to whether this have failed to fully comply with refuge adequate, e.g., Melon and Rissler 1995; demand should be met or needs to be strategies to prevent the development Rissler and Mellon 1996) simply will met. If demand can be reduced, this of Bt resistant insect populations not be available for GM trees. The only might obviate the demand for GM (Coghlan 2001), shows how important reason for accepting a lower standard trees. In other words, GM trees and this dimension can be in determining can be if there has been a policy com- intensification of forestry are unlikely whether a hazard is realized or not. mitment to GM trees and plantation to be the only solution to the problem Again, however, such issues are forestry, and a belief that the benefits of pulp and paper supply. Similarly, if marginalized in a risk-assessment sys- will outweigh any risks. we were to accept as non-contentious tem that focuses on the transgenic trait. Therefore, if the United States fol- the aims of mitigating greenhouse gases And why should the burdens on lows the advice of the IUFRO Work- and protecting native forests, would field trials be minimized? Asking ap- ing Party on Molecular Biology of For- GM trees be the only or best solution? propriate questions that are rigorous est Trees. it will politically restrict the A problem of conventional risk as- and justifiable is the most important framework of the assessment and lower sessment is that it never addresses ques- issue, and proper scrutiny of this is the demands for data to verify assumptions like these openly. Only one option needed. Easier, less-regulated experi- tions, in the interests of the industry is compared to an ill-defined yardstick ments do not mean the best data will and tree molecular biologists. If these of harm (usually conventional practice) be produced upon which to base a risk value judgements and policy commit- and a single answer is given—safe or evaluation. It may simply be that num- ments are widely shared in society, then dangerous—which conceals a host of bers of trials are used as a spurious it will be an approach that carries pub- scientific uncertainties, ignorance, and claim of safety, as they have been with lic confidence; if not, conflict is the value judgements. Conventional risk as- GM crops, even though most small- more likely outcome. Even though the sessment also relies only on a narrow

108 construction of expert knowledge and desirable than regulation for a variety term research in forestry is small com- the moral, social, political, and economic of different reasons, including that they pared with crop production (Robinson values the chosen experts bring to the are more likely to be followed and that 1999), and trees have proved more tech- judgements they make. The public is consumer confidence will be increased, nologically demanding to manipulate effectively excluded from the process, as thus improving market potential. Key than other plants (Tzfira et al. 1998). is expertise from outside the officially areas of disagreement, such as the rela- However, there is considerable interest mandated boundaries. tive human health effects of foods pro- in promoting their use, and controversy However, there are alternatives that duced under different systems and the has already surrounded trials with GM can bring a more robust approach to risk rigor of regulatory controls were also trees in both the United States and Eu- evaluation by examining the relative per- highlighted, proving opportunities for rope. In deciding upon experimental tri- formance of options across a variety of further research that might resolve the als and commercialization, regulators different criteria. One of these ap- disputes. What was quite clear was that will inevitably have to make their under- proaches is multi-criteria mapping, the framing of the assessment and the lying political, social, and economic jus- which has been used to examine options judgements made in scoring—which tifications more explicit than for GM for the growing of GM oilseed by 12 criteria were included and excluded food crops, because the scientific uncer- participants with very different perspec- and the factors influencing scoring— tainties and unknowns are so great and tives (Stirling and Mayer 2000; Stirling drove the final outcome. Altering the physical containment impossible. A and Mayer, in press). The options in- weighting placed on the criteria made claim to risk assessments being scientifi- cluded three with varying controls on relatively little difference to the out- cally driven alone will not be effective GM crop production, and three that come, highlighting how value laden at concealing the value judgements. excluded GM: conventional, low input, and important framing and scoring are. The imposition of GM trees in and organic systems. In this study, the Whilst this use of multi-criteria forestry and the denial of the impor- outcome was driven by the participants mapping showed the potential for tance or relevance of the wider dimen- who chose the criteria for evaluation, comparing different options, it relied sions of risk, including socioeconomic scored the performance of each option on people who would widely be re- impacts, is likely to fuel controversy under each criterion (including an esti- garded as ‘experts’. Therefore, to be and engender a polarized and non-pro- mate of uncertainty), and attributed robust it must be brought together ductive debate. Past experiences with weightings to the criteria. Criteria cho- with public input to inform the choice GM crops have demonstrated that sen ranged from those concerning food of criteria, weightings, and options. such conflict will not be contained in- and environmental safety, to social, eco- Without this, the approach could suf- side national boundaries. Given the nomic, and agronomic issues. fer from a dislocation from public val- cultural, social, and symbolic impor- The outcome was not a single an- ues as conventional approaches do. tance of forests, there is likely to be a swer, but a map of the debate in the However, the exercise showed that such fierce battle fought at all levels, from form of relative rankings of the perfor- an approach was practical and could be the local to the intergovernmental, mance of the options from a range of used productively in the early stages of unless steps are taken to reevaluate the perspectives, which is then available as technology assessment to assist in de- approach being taken to risk assess- a heuristic device for decision makers. termining policy and how judgements ment. New methods will be needed It illuminated areas of agreement, such should be made, if a case-by-case ap- that allow for public input and thus as the finding that an organic system proach is subsequently adopted. have the potential for promoting more performed better than any other option socially resilient decision making, but in environmental terms, whatever the there will have to be openness to dif- starting position of the participant. DISCUSSION ferent outcomes by all sides. Voluntary controls on GM crops and No GM trees are currently in com- Whilst science will be necessary in foods were also seen as generally less mercial production. Investment in long- informing decisions, it is not a suffi-

109 cient basis upon which to do so (Na- Farm-scale evaluation of GM crops Opinion in Biotechnology 11: 298– tional Research Council 1996). The explained. Nature 399: 727–728. 302. EU-US Biotechnology Consultative Grove-White, R, P Macnaghten, S Moyes, C, and P Dale. 1999 Organic Forum (2000) recognized the legiti- Mayer, and B Wynne. 1997. Un- Farming and Gene Transfer from certain World: Genetically Modified Genetically Modified Crops. John macy of the increasing demands to Organisms, Food and Public Atti- Innes Institute, Norwich, UK. democratize the decision-making pro- tudes in Britain. Centre for the National Research Council. 1996. Un- cesses around GM foods, and this les- Study of Environmental Change, derstanding Risk: Informing Deci- Lancaster University, UK. son should be taken on board by the sions in a Democratic Society. Na- GM tree community. Hornig Priest, S. 2000. US public opin- tional Academy Press, Washington, Multi-criteria mapping is not the only ion divided over biotechnology? DC. Nature Biotechnology 18: 939–942. approach that could be used to evaluate the Rissler, J, and M Mellon. 1996. The use of a new technology in a particular set- James, RR, SP Difazio, AM Brunner, Ecological Risks of Engineered Crops. ting. There are many participatory tech- and SH Strauss. 1998 Environ- MIT Press, Cambridge, MA. mental effects of genetically engi- niques, such as citizen juries and consen- Sampson, V, and L Lohman. 2000. neered woody biomass crops. Bio- Genetic Dialectic. The Biological sus conferences, which could also be ap- mass and Bioenergy 14: 403–414. plied, and a combination of approaches Politics of Genetically Modified Trees. Levidow, L, and S Carr. 2000. Unsound The Corner House Briefing 21. may prove the best solution. It may seem science? Trans-Atlantic regulatory The Corner House, Sturminster that techniques like these will be cumber- disputes over GM crops. Interna- Newton, Dorset, UK. some and will delay technological progress. tional Journal of Biotechnology 2: Sederoff, R. 1999. Building better trees 257–273. However, the intense controversy surround- with antisense. Nature Biotechnol- ing GM crops and the damage that has Levidow, L, S Carr, D Wield, R von ogy 17:750–751. Schomberg. 1997. European bio- been done to an industry and to confidence Stirling, A. 1998. Risk at a turning technology regulation: framing the in public institutions demonstrates that point? Journal of Risk Research 1: risk assessment of a herbicide tol- shortcuts may not, in the long run, be a 97–110. erant crop. Science, Technology and productive approach. If the trajectory of Human Values 22: 472–505. Stirling, A, and S Mayer. 2000. Precau- GM crop development and its regulation tionary risk appraisal of a geneti- Mayer, S. 2001. The regulation of ge- cally modified crop. International had been allowed to have greater public netically modified food. In The Journal of Occupational Health and input in the 1980s and early 1990s, the Encyclopaedia of Life Support Sys- Environmental Medicine 6(4): 296– tems, Theme: Social, Educational situation now might be very different. 311. and Political Aspects of Biotechnol- ogy F Dolberg (ed) EOLSS, Cam- Stirling, A, and S Mayer. Multi-criteria REFERENCES CITED bridge, UK (in press). http:// mapping the genetically modified eolss.co.uk crop debate: A pilot study of a genetically modified crop in the Coghlan, A. 2001. Almost a third of Mayer, S, and A Stirling. 2001. Find- UK. Environment and Planning C US farmers broke rules for planting a precautionary approach to (in press). ing GM maize last year. New Sci- technological developments - les- entist 5 February 2001. sons for the evaluation of GM Strauss, S, W Boerjan, J Cairney, M Dale, P. 1999. Public concerns over crops. Journal of Agricultural and Campbell, J Dean, D Ellis, L transgenic crops. Genome Research Environmental Ethics (in press). Jounanin, and B Sundberg. 1999. Forest biotechnology makes its 9:1159–1162. Mellon, M, and J Rissler. 1995. position known. Nature Biotechnol- Transgenic crops: USDA data on EU-US Biotechnology Consultative ogy 17:1145. Forum. December 2000. Final small-scale tests contribute little to Report. commercial risk assessment. Bio/ Tzfira, T, A Zuker, and A Altman. 1998. technology 13:96. Forest-tree biotechnology: Genetic Firbank, LG, AM Dewar, MO Hill, MJ transformation and its application Merkle, SA, and JFD Dean. 2000. For- May, JN Perry, P Rothery, GR to future forests. Trends in Biotech- est tree biotechnology. Current Squire, and IP Woiwood. 1999. nology 16: 439–446. 110 FOREST SCIENCE

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 113-123. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Transgenic Trees: Where are We Now?

Dave Ellis Rick Meilan Gilles Pilate Jeffrey S. Skinner ABSTRACT Dave Ellis is the Director of Numerous field tests throughout the world clearly demonstrate that it is possible to Molecular Biology for genetically engineer diverse tree species. Despite these results, the main challenge with CELLFOR, Inc., based in trees continues to be the development of efficient transformation systems for the most desired genotypes. In this regard, it is not transformation per se that is limiting, but rather Vancouver, British Columbia. the tissue culture system needed to regenerate whole trees from single cells containing the [email protected] inserted genes (transgenes). Although more research is needed, results from field tests have been invaluable for improving our understanding of how inserted genes will be expressed Rick Meilan is an Assistant in long-lived perennials. Available evidence suggests that the issues associated with transgene expression in Professor in the Forest trees will be the same as those observed in agronomic crops. To date we have seen that 1) Science Department at transgene expression levels vary between transformation events/lines, species, transgenes, and traits being modified; 2) targeted, tissue-specific expression of transgenes is possible; Oregon State University, 3) transformation events that give consistent levels of gene expression in primary and the Associate Director transformants are the rule, not the exception; and 4) it is possible to find whatever is being sought, whether it be gene silencing or commercially useful expression. These trends of the Tree Genetic Engi- suggest that although additional work is needed in the area of temporal and developmen- neering Research Coopera- tal regulation, stable, long-term expression of transgenes in trees is readily achievable. Similarly, somaclonal variation does not appear to be a widespread phenomenon when tive. regenerating transgenic trees. The use of embryogenesis helps reduce the incidence of this [email protected] with gymnosperms, and with available practices and care in tissue culture, somaclonal variation is not a limitation with angiosperms. Gilles Pilate is a research Another factor that will play a role in the use of genetically modified trees is the possibility of undesirable levels of transgene spread. Environmental risks associated with scientist in the Unité transgene movement are specific to the gene(s) being inserted, the tree species, and the Amélioration, Génétique et environment within which the transgenic tree is planted. Clearly, pollen and seed have the greatest potential to facilitate transgene movement. Whether this occurs depends on Physiologie Forestières at the reproductive biology of the species in question, the management scheme being used, INRA, Orléans, France. and the proximity of transgenic trees to sexually compatible relatives. Although cultural and deployment strategies could be used in some locations to minimize the risk of transgene pilate@orleans_inra.fr spread, another approach is to link genes that inhibit floral development with the transgene of interest. Several labs are using this approach to develop methods for transgene contain- Jeff Skinner is a Faculty ment. Research Associate in the

Department of Horticul- n many respects, the process of stably integrating a piece of DNA into the ture at Oregon State Uni- nuclear genome is similar in both trees and agricultural crops. There are nu- I merous examples of success with transformation in both systems. The first versity. commercial release of a transgenic tree occurred several years ago in Hawaii with papaya. This release, the first and only commercial release, to our knowledge, of [email protected] a long-lived perennial crop, not only demonstrates the efficacy of this approach

113 for the genetic improvement of trees, The ability to introgress a Generalities in the literature can, but also serves to answer many of the transgene through breeding is key to and have, led researchers to assume that questions regarding the stability of the very rapid development of engi- numerous tree species can be trans- transgene expression in trees. As with neered maize varieties. A single transformed. It should be emphasized, how- other agricultural crops, papaya grow- formation event can theoretically be ever, that in many cases a single geno- ers have enthusiastically embraced bio- introgressed from a breeding line into type of a species has been transformed. technology, as evidenced by the rapid an elite line in nine generations, or in Poplar, which has been touted as the adoption rates and continued satisfac- a few as three years. This enables the “model tree” for molecular biology tion with the product. Although nu- various maize cultivars, each specifically studies, provides a good example. In- merous trans-genic crops have been adapted to different locations or envi- deed, it has a small genome and, by all commercialized (see: http:// ronments, to all have the same reliable indications in the literature, it can be www.aphis.usda.gov/biotech/), papaya expression of a trait imparted by a transformed at a high frequency. Yet a represents the only commercial release single insertion event. In addition, the careful look at the reports on field tri- of a transgenic perennial. resulting varieties can be readily ampli- als of transgenic poplars reveals that a At present there is neither a short- fied for rapid deployment. Equally im- majority of lines described have been age of genes nor methods for incorpo- portant is that due to the ease with produced in just a handful of geno- rating them into trees, and the genes which a transgene can be introgressed, types. Further, only a few if any, of that are inserted are expressed ad- tissue culture and transformation sys- these genotypes are grown commer- equately. What we currently lack is the tems are not needed for each commercially. Thus, the manipulation of com- ability to coax individual cells from cial variety of maize. mercially important genotypes in tis- commercially important genotypes to This is in stark contrast to trees, sue culture remains an obstacle to gain- grow into whole plants. This is not an where introgressing a transgene into su- ing the full economic benefit from issue with transgenics per se, but is in- perior genotypes through breeding is transgenics in forestry. stead a tissue culture problem. It is also not currently possible. Even if the long Despite this limitation, there are not a problem that is unique to trees, regeneration cycle in trees could be over 100 reports of trees being trans- as it exists with agricultural crops as overcome, inbred lines or breeding formed with valuable traits, including well. lines required for such an introgression herbicide and insect resistance (Fillatti Why then are there far more com- program have not been developed. et al. 1987; De Block 1990; McCown mercial releases in agricultural crops? Right now, the only option for trees is et al. 1991; Brasileiro et al. 1992; Economics play a large role, but fun- to develop robust transformation sys- Devillard et al. 1992; Chupeau et al. damental biology also plays a role. In tems that are adaptable to the many 1994; Donahue et al. 1994; Wang et annual agricultural crops, tissue culture genotypes used in clonal plantation al. 1996; Cornu et al. 1996; Leplé et systems need to only be developed for forestry. al. 1995; Heuchelin et al. 1997; Meilan a very few breeding lines and then the However, tissue culture systems et al. 2000); modification of lignin transgene can be rapidly incorporated for all commercially important geno- (Van Doorsselaere et al. 1995; Baucher into elite lines through traditional types in forestry are not available at et al. 1996; Pilate et al. 1997; Hawkins breeding. For example, in maize, even present. Each genotype requires a et al. 1997; Hu et al. 1997; Franke et though hundreds of different slightly different set of conditions to al. 2000; Li et al. 2001); glutathione transgenic varieties are available and induce individual cells, containing the metabolism (Foyer et al. 1995; Strohm transgenics account for 26% of the inserted DNA, to differentiate into et al. 1995); modification of cellulose acreage planted in 2001, all of these whole plants. Our ability to identify (Shani et al. 2001); bioremediation varieties were derived from only five the specialized conditions needed for (Rugh et al. 1998; Gordon et al. 2001) transgenic events in as few as two individual genotypes is currently a and altered hormone biosynthesis breeding lines. major limitation. (Eriksson et al. 2000). In this paper, we

114 discuss some of the challenges, both quirements for complete molecular transgenes are expressed faithfully over perceived and real, for using transgenes characterization, including sequence time. In agricultural crops, there is no in long-lived perennial crops. information, for each insert and its question that such stability in transgene flanking genomic DNA (i.e., character- expression exists; markets would not izing a single site requires less effort have increased if transgene expression GENE EXPRESSION than doing multiple sites). levels could not be trusted. In any case, It should also be noted that the all that is important is to have the A question continually asked re- authors have found no reproducible transgene expressed sufficiently to im- garding transgene expression in long- evidence to support the dogma that part the desired trait (e.g., herbicide lived perennial crops, such as trees, is there is increased variation in expres- tolerance, or insect resistance, etc.) at how can one be sure that transgenes sion patterns with multiple insertion useful levels. Presently, we do not fully will continue to be expressed over the sites, and we do not want to imply that understand how transgene expression years or decades required for a rotation? this helps explain the rationale for se- levels vary over time or in different To answer this question, one must look lecting simple insertions events. In dis- environments. at the data available from transgenic cussions with colleagues working in There is evidence that transgene crops, both annual and perennial. agricultural biotechnology, it has been expression is very stable in trees over It is important to start any discus- repeatedly confirmed that there is little time. In an early study, Ellis et al. (un- sion on the expression of transgenes support for the notion that a greater published) measured the expression of with the realization that variation and number of insertions translates into GUS (a visual reporter gene) in field- instability in transgene expression does increased variation in transgene expres- grown transgenic poplar and spruce exist, just as variation in gene expression. Rather, the available data suggest over a three-year period. In this study, sion exists in breeding. Indeed, if varia- that the desired level of transgene ex- they observed that variation in expres- tion in gene expression did not exist, pression occurs frequently enough to sion levels between individuals within breeding programs would be severely obviate the need to use insert complex- a single line were greater in the field hampered. Further, just as in breeding, ity as the primary criterion for select- than in tissue culture. They speculated if one looks hard enough, it is possible ing transgenic events. that similar trends in gene expression to find virtually any pattern of gene The foregoing discussion addresses would also be noted if one were to expression desired, from suppression to variation in transgene expression, but measure a native gene, and that envi- high-level expression. Fortunately, in the most important question, asked in ronmental factors affecting gene ex- both transgenics and breeding, the se- the first paragraph of this section, con- pression were minor in tissue culture lection for stable gene expression is a cerns the stability of transgene expres- and far greater in a field setting. In frequent event and is very reliable. If sion over time in long-lived crops. As addition to this within-line variation, this were not true, there would no with expression level, it is easy to differences between lines were signifi- commercial transgenic crops and the speculate that biotechnology compa- cant and these differences were consis- industry would not have advanced to nies select for simple insert patterns in tent throughout the year. Of greater its present state. hopes of achieving more stable expres- importance was the observation that Historically, primary selections sion. Again, we find no reliable docu- during the three-year study, the level were done based on the desired level mentation to support the doctrine that of GUS expression was not significantly or location of transgene expression. simple transgene insertion patterns lead different (P < 0.05) from year to year Recently there has been a shift toward to more or less stable expression. for ~85% of the 15 yearly sample selecting for simpler insertion patterns It is of paramount importance to dates. and agricultural crops are now selected have transgenes expressed in the ex- In another aspect of this same study, for single events. This change has been pected way throughout the life of the transgene expression was measured over made mainly to satisfy regulatory re- tree. There is abundant evidence that several years in poplars that contained

115 GUS under the control of the promoter examined transgene expression in field- dence for expression stability is main- from a wound-inducible gene, PINII. grown transgenic poplar for a decade. tenance through meiosis. While using GUS expression was found to be con- In this study, they also observed very primary transformants in proven elite sistent over the two-year field study. Fur- stable and dependable GUS expression genotypes is the current focus of ther, the level of GUS expression was from year to year. More importantly, transgenic trees, future breeding pro- faithfully induced in a very strict devel- they found no evidence for transgene grams will benefit from the inclusion opmental pattern throughout both years. rearrangements, multiplication, loss, or of transgenic traits. Therefore, the These data demonstrate that transgenes other modifications. These data are the question of transgene expression in need not be expressed continuously in strongest indication that transgene ex- progeny is of practical significance. In order to be reliably expressed in trees at pression is stable in long-lived peren- the only known example of transgenic specific developmental stages. This issue nial crops such as forest trees. inheritance in forest trees, Pilate et al. is crucial in the later discussion on the Han et al. (1997) evaluated the (2001) observed the expected Mende- disruption of reproductive structures in use of a matrix-attachment region lian segregation ratio for progeny from trees. (MAR) fragments derived from a to- a line containing a single-copy insert. Meilan et al. (2001a) assessed the bacco gene for increasing the frequency Moreover, the authors detected a frag- stability of transgene expression in 40 of Agrobacterium-mediated transforma- ment of the expected size hybridizing transgenic lines (i.e., independent tion. MARs are elements of DNA that to a GUS probe in all kanamycin-re- events) of hybrid cottonwood (Populus can enhance and stabilize transgene sistant lines. In contrast, a line contain- trichocarpa x P. deltoides) grown at three expression (and Thompson 1996). A ing four inserts produced progeny con- field sites during four years of field tri- binary vector that carried the GUS re- taining 1, 2, 3, and 4 copies of the seg- als. All lines were transformed with a porter gene containing an intron and regated transgene. Transgene expression binary vector that included two genes an NPTII gene was modified to con- level between the progeny from both conferring tolerance to glyphosate tain flanking MAR elements within the transformed lines was highly variable, (GOX and CP4), a gene encoding re- T-DNA borders. The modified and but this is to be expected from het- sistance to the antibiotic kanamycin MAR-containing vectors were used to erozygous offspring. (NPTll), and GUS. Agrobacterium transform tobacco, a readily transform- In addition to variation in the tumefaciens was used for transforma- able poplar clone (P. tremula x P. alba), level or timing of transgene expression, tion; callogenesis and organogenesis and a recalcitrant poplar clone (P. there have also been numerous reports occurred under kanamycin selection. trichocarpa x P. deltoides). MARs signifi- of transgene silencing. This phenom- To test the stability of transgene expres- cantly enhanced transgene expression enon has been reviewed extensively sion, they repeatedly applied herbicide and transformation efficiency, but the (e.g., Finnegan and McElroy 1994; to all lines after outplanting, challeng- effects varied widely in magnitude Stam et al. 1997; Matzke and Matzke ing ramets from previously untreated among genotypes. MARs increased 1998). One form of silencing occurs at lines during their fourth season of veg- GUS gene expression approximately the level of transcription, and usually etative growth. They used maintenance 10-fold in the two hybrid poplar clones involves methylation of promoter re- of herbicide tolerance and GUS expres- and two-fold in tobacco one month gions, but may also involve chromatin sion as indicators of transgene stabil- after co-cultivation with Agrobacterium, modification (van Blokland et al. 1997; ity. Their data show that all lines that and increased the frequency of kana- Finnegan et al. 1998). These reports were highly tolerant in year one con- mycin-resistant poplar shoots recovery have erroneously led many to conclude tinued to be highly tolerant in year more than eight-fold. Thus, MARs that silencing is a major problem con- four. hold considerable promise for use with fronting genetic engineering. As men- In what is perhaps the most am- poplar, if necessary. tioned in the beginning of this section, bitious study done to date with Despite these encouraging results, it is possible to find whatever is desired, transgenic trees, Pilate et al. (2001) it has been suggested that the best evi- and gene silencing is no exception.

116 Because it is scientifically interesting, their growth and herbicide tolerance In one study on the induction of there are many reports of transgene si- for several years. somaclonal variation in poplar, Ostry lencing in the literature. However, it is Contrary to what we have collec- et al. (1994) started with the same a rare occurrence in our hands and tively seen in our work, there is recent mother plant, and compared regener- definitely not a problem at present evidence for a high level of co- or ated shoots derived from embryogenic with the promoters, genes, or con- sense-suppression in poplar with con- cell suspensions, protoplasts, leaf mi- structs currently being tested in trees. structs used to over-express genes in- cro-cross sections, callus, and shoot or Pilate et al. (unpublished data) volved in lignin biosynthesis. Examples root cultures. Somaclonal variation was observed only one case of transgene of genes that have been co-suppressed scored using morphological characters silencing in their work. This silencing include COMT, CAD, 4CL, and C4H and increased resistance to Septoria, event occurred after selection but prior (Jouanin et al. 2001; Tsai et al. 2001). based on a leaf disc assay. Morphologi- to any analysis of the transformed lines. It is not known whether this suppres- cal variants were noted in adventitious Interestingly, this is similar to the only sion is the result of using homologous shoots regenerated from roots, callus, case of transgene silencing seen by Ellis genes or will become more frequent as and protoplasts, with shoots from stem et al. in poplar, where again the puta- more tree genes are used. It is feasible callus yielding the highest level of tive silencing event was found early in that this silencing phenomenon could Septoria-resistant plants. These data tissue culture. In both cases, 5- be unique to the lignin biosynthetic confirm earlier work suggesting that azacytidine had no effect on the silenc- pathway. Clearly, it will be interesting the longer cells are maintained in an ing, indicating that either methylation to see if silencing is a function of the unorganized culture, the greater the was not involved, or that the silencing biochemical pathway being manipu- likelihood of somaclonal variation oc- was not due to an azacytidine-revers- lated or the source of the genes used curring (Larkin and Scowcroft 1981; ible methylation event. These two in- to manipulate the pathway or both. Bajaj 1990; Deverno 1995). cidents of gene silencing, while of aca- In a related study (Serres and demic interest, occurred in less than McCown, unpublished), plants from 1% of all transgenic lines analyzed. SOMACLONAL approximately 400 colonies derived In addition, Meilan et al. (2001b) VARIATION from individual protoplasts were regen- conducted a two-year field test in erated and grown ex vitro. After 10 which they evaluated 80 transgenic Much like transgene silencing, weeks in a greenhouse, the plants were lines for genetically engineered herbi- somaclonal variation, a stable genetic scored for 17 different parameters rang- cide tolerance. They observed an change in somatic cells of a plant, does ing from leaf size and shape to intern- abrupt increase in mean herbicide occur, but can be minimized or ode length. In all, nearly 50% of the damage for two of these lines from one avoided with careful control over cul- regenerated plants had some morpho- year to the next. One possible expla- tural conditions. Somaclonal variation logical differences from the control, nation for loss of tolerance is transgene tends to more common with unstable with the greatest number of variants silencing. A recent report described genomes, such as with tetraploids, al- being in height, leaf length/width ra- how cold-induced dormancy led to el- though we know of no tetraploid trees tio, stem diameter, and number of evated methylation, which in turn, led currently in commercial use. However, nodes. These changes were stable in the to partial transgene silencing (Callahan with the increased use of hybrids or greenhouse and most remained et al. 2000). The possibility of year-to- triploids, the occurrence of somaclonal throughout the first season when the year variation in transgene expression variation may increase. Despite this, it plants were grown in the field. The highlights the importance of conduct- should be noted that, to date, the au- plants were allowed to over-winter out- ing multi-year trials. All 80 lines were thors have rarely seen evidence for of-doors and were scored again the fol- vegetatively propagated and have been somaclonal variation in any of their lowing spring, after leaf flush. In the planted on another site to monitor transgenic poplars. second year, only three lines continued

117 to display abnormalities; two lines somaclonal variation will not be a able for several reasons (Strauss et al. showed sectoring in the leaves, perhaps problem with transgenic plants. 1995; Brunner et al. 1998; Skinner et caused by a transposition event, and al. 2000). First, it will enable the de- the remaining line had the appearance velopment of trees that are incapable of a diploid (Serres et al. 1991). Clearly TRANSGENE of producing sexual propagules. This the majority of the variations were epi- CONTAINMENT would limit gene flow into the wild, genetic changes and not true genetic or helping to mitigate ecological concerns somaclonal changes. It must be understood from the over establishment of transgenic plan- Meilan et al. (unpublished) have outset that no one can guarantee ab- tations. Second, it will likely prevent produced >3,200 independent trans- solute sterility in trees with the tools the growth reduction associated with genic lines in 16 different poplar geno- that are currently available. However, the onset of maturation (Eis et al. types (14 of cottonwood, sections from a scientific, or even a purely risk 1965, Tappeiner 1969; Teich 1975). Tacamahaca and Aigeiros, and two of assessment standpoint, flowering con- Third, it could eliminate the produc- aspen, section Populus or Leuce). They trol may not always be needed before tion of pollen and other nuisance re- field-tested 557 of those lines, and have transgenic trees can or should be grown productive structures. grown most of the remainder in the commercially. There also may be cases One common way to engineer ste- greenhouse for extended periods. Of where it would be desirable to incor- rility is to ablate cells by expressing a porate transgenes into a conventional the total, they have observed morpho- deleterious gene in a tissue-specific breeding system. Thus, the need for logical abnormalities that were not in- fashion (Brunner et al. 1998). Floral sterility will depend on the trait; the duced by transgene expression in only tissue-specific promoters are fused to environment within which the three lines (0.1%). None of these vari- one of a variety of cytotoxin genes that transgenics will be grown; the species; ant lines were in hybrid aspen. The lead to rapid and early death of the tis- and various social, political, and ethi- three lines in which they did observe sues within which the gene product is cal considerations. Each case must be putative somaclonal variants were all expressed (Skinner et al. 2000). One considered individually. hybrid cottonwoods, whose transfor- of the more popular ways to engineer That being said, virtually everyone mation protocol requires that they that sterility in herbaceous plants employs who is working with transgenic trees spend nearly twice as long in an un- an RNAse gene that, although isolated for fiber production is also interested differentiated state as do the aspens from a bacterium, encodes an enzyme in engineered sterility. Controlling the (Han et al. 2000). that is common in plants and animals potential for transgene spread will fa- These studies are consistent with (Mariani et al. 1990). cilitate commercial release and may the view that the avoidance of an un- A second way to genetically engi- help to moderate the perceived nega- organized callus stage, and the mini- neer flowering control is through the use tive environmental impacts of geneti- mized use of hormones to maintain or of dominant negative mutations cally engineered trees. It is also likely regenerate shoots are important mea- (DNMs). DNMs suppress the function that the trees engineered to be repro- sures for reducing somaclonal varia- of a gene at the protein level by ductively sterile will grow faster and tion. It should be noted, however, that overexpression of a mutant version of a prevent unwanted genetic pollution. even in the case of protoplasts, where protein (Espeseth et al. 1993). Inhibition thiadiazuron and both an auxin and a is thought to occur by a variety of ONTROL OF cytokinin were used for regeneration, C means, including formation of an inac- less than 2% of the regenerated shoots REPRODUCTIVE tive heterodimer, sequestration of protein exhibited stable and confirmed STRUCTURES cofactors, sequestration of metabolites, somaclonal variation. or stable binding to a DNA regulatory We maintain that with careful at- Being able to genetically engineer motif. The usefulness of this approach tention to culture conditions, reproductive sterility in trees is desir- for floral control was demonstrated in

118 Arabidopsis with DNM versions of the Now that floral homeotic genes from induce early flowering in poplar (Weigel AGAMOUS (AG) gene (Mizukami et al. poplar have been cloned and characterized and Nilsson 1995). When the PTD pro- 1996). Expression of a truncated AG (Brunner et al. 2000, Sheppard et al. 2000; moter was used to drive the expression of protein in which the C-terminal region Rottmann et al. 2000), work has begun on a cytotoxin gene (ribosome inactivating was deleted resulted in flowers pheno- experimenting with promoters from pop- protein, RIP), petals and stamens were ab- typically similar to those observed in ag lar genes. The promoter from PTD (the P. lated in Arabidopsis; petals, stamens, and mutants, suggesting that the truncated trichocarpa homolog of DEFICIENS) ap- carpels were absent in transgenic tobacco. version of AG was inhibiting endog- pears to be the most floral-specific in its PTD::RIP also appears to prevent flowers enous AG function. expression pattern (Sheppard et al. 2000). from forming on poplar co-transformed A third technique to control flow- We have shown that PTD promoter directs with 35S::LFY (Figure 2). Expression of ering involves post-transcriptional gene expression of the GUS gene early in the PTD::RIP had no significant effects on silencing (PTGS). Recent studies in a development of floral organs in Arabidopsis growth in tobacco. Based on these results, variety of eukaryotic organisms have and poplar (Figure 1). The latter was co- it appears that the PTD promoter may be shown that double-stranded RNA is a transformed with the LEAFY (LFY) gene useful for engineering sterility in a variety potent inducer of PTGS. This ap- from Arabidopsis under the control of the of species. proach to induced silencing has been 35S promoter, which has been shown to termed RNA interference (RNAi) (Fire 1999; Bosher and Labouesse 2000). Recent work in plants using inverted- repeat transgenes showed that RNAi could provide a reliable means for engineering stable suppression of gene activity in plants (Waterhouse et al. 1998; Chuang and Meyerowitz 2000). Below we briefly discuss recent progress made with one approach to flowering control in trees.

CELL ABLATION

In early attempts at utilizing cell ablation with poplar, heterologous promoters, which had shown floral-spe- Figure 1. GUS expression under the control of the PTD promoter in Arabidopsis (A) and poplar (B) flowers. cific expression in tobacco and 1.0 Arabidopsis (Koltunow et al. 1990; n=16 n=15 n=20 0.9 Hackett et al. 1992; Wang et al. 1993), Flowering frequency were used to drive the expression of 0.8 in background 0.7 Reporter activity in two cytotoxin genes, DTA (Greenfield induced flowers Figure 2. Flowering 0.6 et al. 1983) and barnase (Hartley et al. frequency (Freq.) of 0.5

1988). When introduced into poplar transformed with Frequency 0.4 transgenic poplars, these fusions re- 35S::LFY alone (which 0.3 induces flowering) or in sulted in decreased vegetative growth, 0.2 conjunction with a fusion n=10 suggesting leaky expression in non-tar- between either a reporter 0.1 get tissues (Meilan et al. 2001c). gene (GUS) or a cytotoxin 0 35S::LFY 35S::LFY 35S::LFY gene (RIP). PTD::GUS PTD::RIP 119 CONCLUSION tude to Amy Brunner, Caiping Ma, Brunner, AM, R Mohamad, R Meilan, Pearce Smithwick, Vicky Hollenbeck, LA Sheppard, WH Rottmann, and SH Strauss. 1998. Genetic engi- Sarah Dye, Céline Pugieux, and Simon We have shown that significant neering of sexual sterility in shade Hawkins for their help with the work; amounts of data have been gathered for trees. Journal of Arboriculture to Wayne Parrot and Janet Carpenter transgenic trees over the past several 24(5):263–273. for valuable information and discus- years in numerous small-scale, short- Callahan, AM, R Scorza, L Levy, VD sion; and to Margarita Gilbert for com- term field trials conducted in three Damsteegt, and M Ravelonandro. ments on the manuscript. countries and on two continents. These 2000. Cold-induced dormancy af- data show that, contrary to what our fects methylation and post-transcriptional gene silencing in critics may believe, we can produce REFERENCES transgenic plums containing plum transgenic trees with almost no evi- pox potyvirus coat protein gene. dence of collateral genetic damage to International Society of Plant Mo- Baucher, M, B Chabbert, G Pilate, J van the tree (i.e., somaclonal variation), and lecular Biology Congress, 18–24 Doorsselaere, MT Tollier, M Petit- June 2000, Québec, Canada. that inserted genes are expressed stably Conil, D Cornu, B Monties, M from year to year, after vegetative van Montagu, D Inze. 1996. Red Chuang, C-F, and EM Meyerowitz. propagation, and in a variety of envi- xylem and higher lignin extract- 2000. Specific and heritable ge- ronments. We have also shown that it ability by down-regulating a netic interference by double- should soon be possible to genetically cinnamyl alcohol dehydrogenase in stranded RNA in Arabidopsis poplar. Plant Physiology 112:1479– thaliana. Proceedings National engineer flowering control in trees. 1490. Academy of Sciences USA 97 Finally, we recognize the need for (9):4985–4990. larger-scale, long-term ecological stud- Bajaj, YPS. 1990. Somaclonal variation- origin, induction, Chupeau, M-C, V Pautot, Y Chupeau. ies in order to evaluate potential risks cryopreservation, and implications 1994. Recovery of transgenic trees af- to the environment. in plant breeding. In: Somaclonal ter electroporation of poplar proto- Variation in Crop Improvement. plasts. Transgenic Research 3:13–19. Vol. I, YPS Bajaj (ed.), Springer- Cornu, D, J-C Leplé, M Bonnadé- ACKNOWLEDGMENTS Verlag, Berlin, pp. 3–48. Bottino, A Ross, S Augustin, A Bosher, JM and M Labouesse. 2000. Delplanque, L Jouanin, G Pilate. The work described herein was RNA interference: Genetic wand 1996. Expression of a proteinase funded in part by the Tree Genetic and genetic watchdog. Nature Cell inhibitor and a Bacillus thuringiensis d-endotoxin in Engineering Research Cooperative Biology 2:E31–E36. transgenic poplars. In: MR Ahuja, (http://www.fsl.orst.edu/tgerc/ Brasileiro, ACM, C Tourneur, J-C W Boerjan, DB Neal (eds.) Somatic index.htm), U.S. Department of En- Leplé, V Combes, L Jouanin. Cell Genetics and Molecular Genet- ergy (Bioenergy Feedstock Develop- 1992. Expression of the mutant ics of Trees. Kluwer Academic Pub- Arabidopsis thaliana acetolactate ment Program, grant # 85X-ST807V), lishers, pp. 131–136. synthase gene confers National Science Foundation I/UCRC chlorsulfuron resistance to DeBlock, MD. 1990. Factors influenc- Program (grant # 9980423-EEC), transgenic poplar plants. Transgenic ing the tissue culture and the Consortium for Plant Biotechnology Research 1:133–141. Agrobacterium tumefaciens-medi- Research (grant # OR22072-78), ated transformation of hybrid as- Brunner, AM, WH Rottmann, LA pen and poplar clones. Plant Physi- Agenda2020 (grant # Sheppard, K Krutovskii, SP ology 93:1110–1116. FC0797ID13552), a grant from the DiFazio, S Leonardi, and SH Monsanto Company, and by EU Strauss. 2000. Structure and ex- Deverno, LL. 1995. An evaluation of somaclonal variation during so- project AIR2-CT94-1571. The authors pression of duplicate AGAMOUS orthologs in poplar. Plant Molecu- matic embryognesis. In: Somatic would also like to express their grati- lar Biology 44:619–634. Embryogenesis in Woody Plants, S

120 Jain, P Bupta, and R Newton thylation in plants. Annual Review pression in both pollen and pistil (eds.), Vol. 1, Kluwer Academic of Plant Physiology and Plant Mo- of tobacco. Plant Journal 2:613– Publishers, Dordrecht the Nether- lecular Biology 49:223–247. 617. lands, pp. 361–377. Finnegan, J and D McElroy. 1994. Han, K-H, R Meilan, C Ma, and SH Devillard, C. 1992. Genetic transfor- Transgene inactivation: plants fight Strauss. 2000. An Agrobacterium mation of aspen (Populus tremula back. Bio/technology 12:883–888. transformation protocol effective x Populus alba) by Agrobacterium in a variety of cottonwood hybrids Fire, A. 1999. RNA-triggered gene si- rhizogenes and regeneration of (genus Populus). Plant Cell Reports lencing. Trends in Genetics 15:358– plants tolerant to herbicide. C.R. 19:315–320. 363. Acad. Sci. 314:291–298. Han, K-H, C Ma, and SH Strauss. Foyer, CH, N Souriau, S Perret, M Donahue, RA, TD Davis, CH Michler, 1997. Matrix attachment regions Lelandais, KJ Kunert, C Pruvost, DE Riemenschneider, DR Carter, (MARs) enhance transformation L Jouanin. 1995. Overexpression PE Marquardt, N Sankhla, D frequency and transgene expression of glutathione reductase but not Sankhla, BE Haissig, JG Isebrands. in poplar. Transgenic Research glutathione synthetase leads to in- 1994. Growth, photosynthesis, 6:415–420. creases in antioxidant capacity and and herbicide tolerance of geneti- resistance to photoinhibition in Hartley, RW. 1988. Barnase and barstar: cally modified hybrid poplar. Ca- poplar trees. Plant Physiology Expression of its cloned inhibitor nadian Journal of Forest Research 24 109:1047–1057 permits expression of a cloned ri- (12):2377–2383. bonuclease. Journal of Molecular Franke, R, CM McMichael, K Meyer, Eis, S, EH Garman, and LF Ebell. Biology 202:913–915. AM Shirley, JC Cusumano, C 1965. Relation between cone pro- Chapple. 2000. Modified lignin in Hawkins, S, J Samaj, V Lauvergeat, A duction and diameter increment of tobacco and poplar plants over-ex- Boudet, and J Girma-Pettenati. Douglas-fir [Pseudotsuga menziesii pressing the Arabidopsis gene en- 1997. Cinnamyl alcohol dehydro- (Mirb.) Franco], Grand fir [Abies coding ferulate 5-hydroxylase. genase: Identification of new sites grandis (Dougl.) Lindl.], and west- Plant Journal 22:223–234. of promoter activity in transgenic ern white pine [Pinus monticola poplar. Plant Physiology 113:321– (Dougl.)]. Canadian Journal of Gordon, MP, S Doty, L Newman, T 325. Botany 43:1553–1559. Shang, S Strand, X Wang. 2001. Using trees in phytoremediation. Heuchelin, SA, L Jouanin, NB Eriksson, ME, A Israelsson, O Olsson Paper presented at the Trees Bio- Klopfenstein. 1997. Potential of and T Moritz. 2000. Increased technology in the New Millen- proteinase inhibitors for enhanced giberellin biosynthesis in nium IUFRO/Molecular Biology resistance to insect and pathogen transgenic trees promotes growth, of Trees Meeting, July 22–27, pest of Populus. In: NB biomass production and xylem fi- Stevenson, Washington, USA. Klopfenstein, YW Chun, M-S ber length. Nature Biotechnology Kim, MR Ahuja, eds., 18:784–788. Greenfield, L, MJ Bjorn, G Horn, D Micropropagation, Genetic Engi- Fong, GA Buck, RJ Collier, and Espeseth, AS, AL Darrow, and E neering, and Molecular Biology of DA Kaplan. 1983. Nucleotide se- Linney. 1993. Signal transduction Populus. USDA Forest Service, quence of the structural gene for systems: Dominant negative strat- General Technical Report RM- diphtheria toxin carried by egies and mechanisms. Molecular GTR-297, pp 173–177. corynebacteriophage. Proceedings of Cell. Diff. 1:111–161. the National Academy of Sciences Hu, W-J, SA Harding, J Lung, JL Fillatti, JJ, J Sellmer, B McCown, B USA 80:6853–6857. Popko, J Ralph, DD Stokke, CJ Haissig, L Comai. 1987. Tsai, VL Chiang. 1999. Repression Grishok, A, H Tabara, and CC Mello. Agrobacterium-mediated transfor- of lignin biosynthesis promotes 2000. Genetic requirements for mation and regeneration of cellulose accumulation and growth inheritance of RNAi in C. elegans. Populus. Molecular and General in transgenic trees. Nature Biotech- Science 287:2494–2497. Genetics 206:192–199. nology 17:808–812. Hackett, RM, MJ Lawrence, and CH Finnegan, EJ, RK Genger, WJ Peacock, Jouanin, L, T Goujon, R Sibout, B Franklin. 1992. A Brassica S-locus and ES Dennis. 1998. DNA me- Pollet, I Mila, JC Leplé, G Pilate, related gene promoter directs ex- 121 M Petit-Conil, J Ralph, and C McCown, BH, DE McCabe, DR AGAMOUS: Characterization of Lapierre. 2001. Lignification in Russell, DJ Robison, KA Barton, the DNA binding domain and poplar and arabidopsis plants with KF Raffa. 1991. Stable transforma- analysis of dominant negative mu- depressed COMT and CAD activ- tion of Populus and incorporation tations. Plant Cell 8:831–845. ity. Paper presented at the Trees of pest resistance by electric dis- Pilate, G, E Guiney, M Petit-Conil, C Biotechnology in the New Millen- charge particule acceleration. Plant Lapierre, J-C Leplé, B Pollet, I nium IUFRO/Molecular Biology Cell Reports 9:590–594 Mila, C Halpin, L Jouanin, W of Trees Meeting, July 22–27, Meilan, R, DJ Auerbach, C Ma, SP Boerjan, W Schuch, and D Cornu. Stevenson, Washington, USA DiFazio, and SH Strauss. 2001a. 2001. Field and Pulping Perfor- Koltunow, AM, J Truettner, KH Cox, Stability of herbicide resistance and mance of Transgenic Trees with M Wallroth, and RB Goldberg. GUS expression in transgenic hy- Altered Lignification. Submitted to 1990. Different temporal and spa- brid poplars (Populus sp.) during Nature Biotechnology. tial gene expression patterns occur several years of field trials and veg- Pilate, G, J-C Leplé, C Lapierre, B during anther development. Plant etative propagation. HortScience Pollet, M Baucher, J van Cell 2:1201–1224. (in press). Doorsselaere, W Boerjan, M Petit- Larkin, PJ and WR Scowcroft. 1981. Meilan, R, K-H Han, C Ma, SP Conil, V de Nadaï, and L Jouanin. Somaclonal variation—a novel DiFazio, JA Eaton, E Hoien, BJ 1997. Transgenic poplar trees with source of variability from cell cul- Stanton, RP Crockett, ML Taylor, altered lignin and improved ture for plant improvement. Theo- RR James, JS Skinner, L Jouanin, pulping characteristics. In: Proceed- retical and Applied Genetics G Pilate, and SH Strauss. 2001b. ings of the 1997 TAPPI Biological 60:197–214. Growth and glyphosate tolerance Sciences Symposium, October 19– of transgenic poplars. I. Two-year 23, San Francisco, CA, pp. 507– Leplé, JC, M Bonade-Bottino, S field performance. Submitted to 514. Augustin, G Pilate, VD Letan, A Canadian Journal of Forest Research. Delplanque, D Cornu, L Jouanin. Rottmann, WH, R Meilan, LA 1995. Toxicity to Chrysomela Meilan, R, A Brunner, J Skinner, and S Sheppard, AM Brunner, JS Skin- tremulae (coleoptera: Strauss. 2001c. Modification of ner, C Ma, S Cheng, L Jouanin, G chrysomelidae) of transgenic pop- Flowering in Transgenic Trees. In: Pilate and SH Strauss. 2000. Di- lars expressing a cysteine protein- Molecular Breeding of Woody Plants. verse effects of overexpression of ase inhibitor. Molecular Breeding Progress in Biotechnology series. A LEAFY and PTLF, a poplar 1:319–328. Komamine and N Morohoshi, edi- (Populus) homolog of LEAFY/ tors. Elsevier Science BV, FLORICAULA, in transgenic pop- Li et al. 2001. The last step of syringyl Amsterdam (in press). lar and Arabidopsis. Plant Journal monolignol biosynthesis in an- 22: 235–246. giosperms is regulated by a novel Meilan, R, Ma, C, Cheng, S, Eaton, JA, gene encoding sinapyl alcohol de- Miller, LK, Crockett, RP, DiFazio, Rugh, CL, RB Senecoff, RB Meagher, hydrogenase. Plant Cell 13:1567– SP, and Strauss, SH. 2000. High and SA Merkle. 1998. Develop- 1585. levels of Roundup™ and leaf- ment of transgenic yellow poplar beetle resistance in genetically en- for mercury phytoremediation. Mariani, C, M DeBeuckeleer, J gineered hybrid cottonwoods. In: Nature Biotechnology 16:925–928. Tr uettner, J Leemans, and RB KA Blatner, JD Johnson, and DM Goldberg. 1990. Induction of male Serres, R, M Ostry, B McCown, D Baumgartner, eds., Hybrid Poplars sterility in plants by a chimaeric ri- Skilling. 1991. Somaclonal varia- in the Pacific Northwest: Culture, bonuclease gene. Nature 347:737– tion in Populus hybrids regener- Commerce and Capability. Wash- 741. ated from protoplast culture. In: ington State University Coopera- Ahuja, ed., Woody Plant Biotechnol- Matzke, M and AJ Matzke. 1998. Epi- tive Extension Bulletin ogy, Plenum Press, NY, pp. 59–62 genetic silencing of plant MISC0272, Pullman, WA, pp. transgenes as a consequence of di- 29–38. Sheppard, LA, AM Brunner, KV verse cellular defence responses. Krutovskii, WH Rottmann, JS Mizukami, Y, H Huang, M Tudor, Y Cellular and Molecular Life Sciences Skinner, SS Vollmer, and SH Hu, and H Ma. 1996. Functional 54:94–103. Strauss. 2000. A DEFICIENS ho- domains of the floral regulator molog from the dioecious tree 122 Populus trichocarpa is expressed in Teich, AH. 1975. Growth reduction Zamore, PD, T Tuschl, PA Sharp, and both female and male floral mer- due to cone crops on precocious DP Bartel. 2000. RNAi: Double- istems of its two-whorled, uni- white spruce provenances. Envi- stranded RNA directs the ATP-de- sexual flowers. Plant Physiology ronment Canada Bi-monthly Re- pendent cleavage of mRNA at 21 124:627–639. search Notes 31:6. to 23 nucleotide intervals. Cell 101:25–33. Shani, Z, I Levi, S Mansfield, L Roiz, Tsai, C-J, SA Harding, P Pechter, RJ M Dekel, and O Shoseyov. 2001. Steinbrecher, D Richter, and V Modulation of wood fibers and Chiang. 2001. Transgenic aspen studies on the mode of action of with altered lignin. Paper presented cellulose-binding domains at the Trees Biotechnology in the (CBDs). Paper presented at the New Millennium IUFRO/Mo- Trees Biotechnology in the New lecular Biology of Trees Meeting, Millennium IUFRO/Molecular July 22–27, Stevenson, Washing- Biology of Trees Meeting, July 22– ton, USA. 27, Stevenson, Washington, USA. van Blokland, R, M Ten Lohuis, and P Skinner, JS, R Meilan, AM Brunner, Meyer. 1997. Condensation of and SH Strauss. 2000. Options for chromatin in transcriptional re- Genetic Engineering of Floral Ste- gions of an inactivated plant rility in Forest Trees. In: Molecular transgene: Evidence for an active Biology of Woody Plants, SM Jain role of transcription in gene silenc- and SC Minocha (eds), Kluwer ing. Molecular and General Genet- Academic Publishers, Dordrecht, ics 257:1–13. The Netherlands, pp. 135–153. Van Doorsselaere, J, M Baucher, E Spiker, S and WF Thompson. 1996. Chognot, B Chabbert, MT Tollier, Nuclear matrix attachment regions M Petitconil, JC Leple, G Pilate, and transgene expression in plants. D Cornu, B Monties, M Van Plant Physiology 110:15–21. Montagu, D Inze, W Boerjan, L Jouanin. 1995. A novel lignin in Stam, M, JNM Mol, and JM Kooter. poplar trees with a reduced caffeic 1997. The silence of genes in acid 5-hydroxyferulic acid O- transgenic plants. Annals of Botany methyltransferase activity. Plant 79:3–12. Journal 8:855–864. Strauss, SH, WH Rottmann, AM Wang, H, HM Wu, and AY Cheng. Brunner, and LA Sheppard. 1995. 1993. Development and pollina- Genetic engineering of reproduction regulated accumulation and tive sterility in forest trees. Molecu- glycosylation of a stylar transmit- lar Breeding 1:5–26. ting tissue-specific proline-rich Strohm, M, L Jouanin, KJ Kunert, C protein. Plant Cell 5:1639–1650. Pruvost, A Polle, CH Foyer, H Waterhouse, PM, MW Graham, and Rennenberg. 1995. Regulation of M-B Wang. 1998. Virus resistance glutathione synthesis in leaves of and gene silencing in plants can be transgenic poplar (Populus tremula induced by simultaneous expres- x P. alba) overexpressing glu- sion of sense and antisense RNA. tathione synthetase. Plant Journal Proceedings of the National Academy 7:141–145. of Sciences USA 95:13959–13964. Tappeiner, JC. 1969. Effect of cone produc- Weigel, D and O Nilsson. 1995. A de- tion on branch, needle and xylem ring velopmental switch sufficient for growth of Sierra Nevada Douglas-fir. flowering initiation in diverse Forest Science 15:171–74. plants. Nature 12:495–500. 123 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 124-137. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Perspectives on Risk in Transgenic Forest Plantations in Relation to Conventional Breeding and Use of Exotic Pines and Eucalypts: Viewpoints of Practicing Breeding and Transformation Scientists

R.D. Burdon C. Walter

ABSTRACT Rowland Burdon FRSNZ [Fellow of Risk factors associated with transgenic material of exotic pines and the Royal Society of New Zealand] eucalypts reflect (1) special features of pines and eucalypts as such, com- has worked at the New Zealand pared with other forest tree crops; (2) the technical and commercial risks that may be specific to transgenic crops rather than the products of con- Forest Research Institute since 1964, ventional breeding; and (3) the mainly perceived risks of transgenic plants mainly in the areas of the genetics to the environment or human health. The features that are in varying and gene resources of Monterey pine degree specific to pines include crop lifespan (relating to both the potential magnitude of certain risks and the need for stable, long-term ex- (Pinus radiata), quantitative breeding pression of transgenes); natural silvics (largely favoring species monocul- methodology, and tree breeding tures); soil and climatic tolerances (effectively restricting the feasible con- strategy. In recent years Dr. Burdon trol over growing environments); wind pollination (causing problems of containment); and the associated outbreeding system. All these features has focused significantly on risk may accentuate the risks associated with operational deployment of management. He assisted in develop- transgenic material. Comparisons between pines and eucalypts, and the ing the IUFRO Position Statement on implications of the main differences, are outlined. Exotic status tends to reduce some risks, but accentuate others. Among the risks, those involv- Genetic Modification. ing crop vulnerability are seen as especially relevant, unlike many of the

[email protected] commonly perceived risks associated with transgenic crops. Such risks are potentially of great economic and therefore social significance. Use Christian Walter has worked in Ger- of transgenics may often contrast with conventional breeding, and espe- many and New Zealand on genetic cially evolution, in potentially skipping the typical testing of low-frequency alleles by natural selection. However, the introduction of an al- engineering in agricultural plants and ready tested gene into an organism may effectively mimic accelerated forest trees. His current focus is on evolution. Field testing can help to identify and minimize various risks, but can in some cases significantly increase time required to commer- gene expression in conifers, and the cialize transgenics. Spread of risk, among transformation events and long- term effect of genetically engi- transgenes used, in addition to among recipient genotypes, seems highly neered material in field trials. Dr. prudent, but regulatory protocols may create obstacles. However, strategies may need to be developed that stop the spread of transgenic mate- Walter has been heavily involved in rial through sexual reproduction, without incurring new and significant making submissions to New risks. Further, the use of multiple transgenes in any recipient genotype Zealand’s recent Royal Commission may increase risks, from possible interactions among transgenes as well as simply summing the specific risks for individual transgenes. Risks of on Genetic Modification. using transgenics must be evaluated against those of not using a tech- [email protected] nology that has the potential to greatly benefit humankind.

124 o begin, we define the general cations within the genome. If such a a specific environment, this may not be concept of risk, as a function sequence is already present in nature in so in a different environment. T (roughly speaking, a product) one way or another, we have something In any event, certain risks of crop of the probability of a negative effect very much like the possible outcome failure can arise in plantations whether occurring and its magnitude (or seri- of natural gene transfer (also termed they result from conventional breeding ousness). In turn, the magnitude can Horizontal Gene Transfer or HGT). or genetic engineering. Further, a spe- be viewed as roughly the product of the Although many opponents of genetic cific new genotype produced either severity of the event and the scale on engineering regard the former as a risk way must respond to the specific envi- which it occurs (Burdon 1999); for and ignore the latter, no potential or ronment in which it is grown. The re- instance, 100% mortality of a cultivar perceived effect can actually be re- sultant of genotype and environment may be a severe event in relation to that garded as a risk without some actual will finally decide the success or fail- cultivar, but not a serious one if only a analysis. A new sequence of DNA in ure of the crop (genetically modified few trees of it have been planted. an organism will often have an effect or not), and the potential risks. Where the probability is unknown, the such as a new protein produced, if it To address our specific brief, we term uncertainty is widely used. Yet, codes for a structural gene, or simply will first identify key features that dif- even if the probability is of a low but the ‘wastage’ of a small amount of re- ferentiate, in connection with risk fac- uncertain value, it cannot be readily sources to replicate the new DNA. The tors, exotic pine plantations from tra- discounted if the potential magnitude latter type of effect can be regarded as ditional crop plants and/or other for- is extreme. In that situation, it becomes neutral with regard to the environment est trees, and eucalypt plantations from necessary to make judgments as to just or human health. In specific circum- exotic pine plantations. In this context how low the probability is, and there stances, the effect will be beneficial to we will address the various categories will be cases where there are good rea- the organism, for instance, when the of risk. We will then compare briefly sons for believing that it is essentially new gene provides selective advantage. the use of genetic transformation with zero. The genetic engineering debate The effect may be negative if the new the processes of both natural evolution worldwide has become characterized by sequence of DNA reduces fitness. Also, and classical breeding, and consider confusion over the term risk. While an effect that is beneficial to the organ- horizontal gene transfer. We will fol- there are attempts to define and quan- ism might be a risk for the environ- low by addressing risk profiles and risk tify risk, there is growing concern that ment. The same, however, is true for management, before our final com- the term is used misleadingly, with a the outcome of a conventional breed- ments. readiness to view any non-zero prob- ing experiment, where new gene com- ability as being effectively high, and to binations may be ‘beneficial’ for some ignore risks that arise without genetic applications, but a risk for the environ- DISTINGUISHING engineering (Bazin and Lynch 1994). ment. The generation of an organism Further, the occurrence of an effect as with potentially unknown characteris- FEATURES OF EXOTIC such is often not clearly defined. The tics is a potential threat in relation to PINE CROPS risk discussion is in many cases char- all efforts of humankind to modify acterized by claims that a specific ef- existing planting stock according to The crucial features that often dis- fect will occur, without reference to, their needs. tinguish pine plantations from other and analysis of, the individual steps Any potential risk should have to types of crop and other forest trees may that have to happen for an effect to be assessed in the context of the par- be listed as crop lifespan (i.e., rotation occur. ticular organism, and in context of the age), natural silvics, soil tolerances, cli- The result of a genetic transforma- particular environment the organism is matic tolerances, wind pollination, and tion is an organism that carries a new living in. While the engineered or bred the associated outbreeding system. Ex- sequence of DNA at one or more lo- product may have low risk potential in

125 otic status is a less consistent feature of salvage is possible, will carry the com- tion to counter some risk factors (e.g.. pine plantations (given, for example, pounded costs of crop establishment, certain diseases) may not be economi- the huge areas of plantations of loblolly tending, and protection. Furthermore, cally feasible. The climatic tolerances of pine within its natural range), and one a delayed failure may involve plantings many pines will also favor their use on of more ambiguous significance (in re- made over a number of years, which sites imposing such constraints. spect of biotic risks that may interact can be especially damaging in itself and with the risks of genetic transforma- preclude any rapid recovery of the for- tion) (Burdon 1999). For example, an est-growing enterprise. Wind Pollination exotic crop may be growing in the absence of a disease, but if the disease The wind pollination has twofold arrives, and conventional breeding or Natural Silvics significance, in its potential for the transformation have accidentally (and long-distance spread of genetic mate- perhaps unknowingly) increased a sus- Pines are typically strongly light- rial into exotic and native stands (e.g., ceptibility that is already high in the demanding, like many species that are Sedgley and Griffin 1989: DiFazio et exotic environment, the outcome could either pioneers in an ecological succes- al. 1999), and its likely impact on the be very serious. sion or else are involved in fire-induced level of diversion of resources into re- climaxes. As such, they are usually far production. While most of the pollen more conveniently grown in pure, settles within short distances from Crop Lifespan even-aged stands (Burdon, in press). points of release, dense pollen clouds Thus growing pines for convenience can still occur at considerable distances Even though pines are often fast- and high economic returns entails the from large areas of the stand (Lindgren growing they are seldom grown on very risks that are inherent in growing pure, et al. 1995). However, if the species are short rotations. Contributing factors even-aged stands, although the major exotics, the ecological and political are, part that pure, even-aged stands often concerns over pollen contamination are play in the natural ecology is likely to reduced, compared with species that •Not having the extreme relative reduce the inherent riskiness of the sys- are native to the growing region, espe- growth rates that allow mean annual tem. cially if none of the native species are increments to culminate at very interfertile with the plantation species. early ages In fact, there are relatively few situa- • Their wood being more valuable for Soil and Climatic tions where such interfertility is likely solid-wood products than for Tolerances to arise with pines. pulping or board products, also The diversion of resources into militating against extremely short Pines are generally adapted to rela- reproduction can be major (e.g., Field- rotations tively low soil fertility (Burdon in ing 1960; Cremer 1992), and is all the press), which has various implications. more significant because the pollen •Wood quality tending to be low, Large areas of land can thus be avail- component represents ‘high-grade’ bio- especially for the more lucrative cur- able for growing pines, which affects mass in having a high content of nu- rent uses, in very young trees. the potential scale of risk. The low trients. Suppression of reproduction is The relatively long lifespan can fertility demands are likely to be related an attractive goal, offering up to 100% accentuate risks related to any tree-im- to the limitations in early growth po- reduction in pollen flow and redirec- provement technologies in several tential, which affect crop lifespan. tion of energy flows potentially result- ways. A crop can be exposed to risk Moreover, the land that will be avail- ing in increased timber production over a longer period. A delayed failure, able for pine plantations will often rep- (Ledig and Linzer 1978). Genetic en- which may still occur before profitable resent sites where intensive interven- gineering strategies exist to achieve this

126 goal to a high degree, including the normal food chain in a given ecosys- ms) If it comes from transgenic mate- expression of cytotoxic genes in repro- tem to remain intact. rial it is likely to be even more so, if ductive tissue, influencing reproductive only because it could raise a whole new pathways through homeotic genes, and area of public concerns. However, sup- the suppression of specific genes in- Outbreeding Behavior pression of pollen production would be volved in reproductive development sought in the interests of improving net (Strauss et al. 1995). None of these Pines, with very few exceptions, production as well as allaying concerns technologies however can guarantee a are natural outbreeders (Richardson over containment. complete barrier to reproduction when 1998; Burdon, in press), which has used on their own, and a pyramiding some important implications. The tree- strategy may be more successful in to-tree genetic variation can be impor- COMPARISONS OF completely controlling reproduction in tant for population resistance to dis- a given stand. This is of particular im- eases (Thielges 1982), which may not EUCALYPTS WITH PINES portance in areas where interfertility always fit conveniently with the clonal with native species can be expected. systems that may be needed in order to use genetic transformation. The tree- Other, more subtle considerations Call for Genetic can arise. For instance, elimination of to-tree genetic variation that is typically pollen cone formation in pines may associated with the outbreeding behav- Transformation change crown configuration by elimi- ior is likely to make genetic gains Eucalypts can be highly vulnerable nating zones of shoot without foliage. readily available from recurrent selec- to weed competition and very subject Thus, the genotypes with the best tion within existing populations. At the to defoliation by insects (Cromer and crown configuration for crop produc- same time, such genetic improvement Eldridge 2000). In that resistance to tivity before suppression of pollen cone can be a crucial platform for optimiz- herbicides and insect attack are both formation may not be so after this has ing on the improvements achievable being widely pursued through genetic been achieved. Hence there can be through genetic modification, because modification, we have two areas where complex interactions between genetic the merit of a transformant will inevi- the use of transgenics will tend to be engineering and classical breeding, tably be limited by the general merit particularly attractive for eucalypts. which could lead the unwary into un- of the recipient genotype. Interest in the respective areas is exem- expected losses. Complex interactions plified by Edwards et al. (1995) and may arise among traits being addressed Harcourt et al. (1995). by conventional breeding, and some Exotic Status argue that they may be far less impor- Where the pine species is grown tant because of the less radical changes as an exotic, genetic contamination of Factors Mitigating Risks that it will bring within a short time its natural stands is not usually an is- frame. This conclusion however, still of Using Transgenics sue, although hybridization with some lacks supporting data and more re- other pine species might very occasion- The insect pollination of search is required to fully understand ally be. If ex-situ gene resources are to eucalypts, as opposed to wind pollina- this issue. be maintained, as potential sources of tion of pines, will tend to reduce the Another environmental aspect renew germplasm that is unrelated to risks of long-distance gene flow, mean- lated to sterility considerations is the existing material in breeding popula- ing that containment is inherently importance of pollen-feeding native tions, any pollen contamination from easier, creating less call for suppression insects and birds. It may be appropri- commercial stands is generally very of reproduction. Eucalypt plantations ate to design strategies that still allow unwelcome. (Burdon and Kumar, in have, until quite recently, been almost inviable pollen production and the

127 entirely exotic, unlike huge areas of crop failure could be very high, unless Development-related intensively cultivated pine plantations alternative pulpwood supplies are Risks (e.g., Pinus taeda and P. elliottii) within readily available. This is, however, a their native ranges. This has made po- management problem that is associated As with any new or very imma- tential gene flow into natural popula- with any use of biological systems for ture technology genetic transformation tions far less of an issue, but a recent production processes and risks are as- will have its technological risks upsurge of establishment of eucalypt sociated with conventionally produced (Burdon 1992), relating to whether it plantations in Australia is changing the material and transgenic material alike. will actually succeed, or at least do so picture (Cromer and Eldridge 2000). Moreover, in such cases the plantings without prohibitive development costs. The very short rotations on which made in just a single year, which might A spread of risks, in respect of the ge- eucalypts are often grown, usually for all be exposed to some unsuspected risk netic modifications that are pursued, pulpwood or fuelwood, should in some factor, would represent a considerable may be indicated. While this may en- ways mitigate the impact of crop fail- fraction of a plantation estate. tail considerable dispersal of effort, risk ure. Moreover, the flat or easy terrain management is an expected compo- on which eucalypts are very often nent of applying any technology (Anon grown, and the associated site cultiva- RISK CATEGORIES 1999). Consequently, risk related to tion, should help facilitate protective other forest plantation technologies intervention against, say, a disease that The risks may be classified in vari- should be reconsidered and analyzed in may flare up, whether or not it arises ous ways. The first breakdown that we context and in comparison with new specifically in transgenics. are adopting is into technologies. This area has indeed been neglected considerably over the last de- •Development-related risks, involved cades, and despite the fact that both in pursuing and developing genetic Factors Accentuating clonal propagation and conventional modification (GM) or genetic en- Risks of Using breeding have generated examples of gineering Transgenic Eucalypts non-desirable effects (consider for in- •Deployment-related risks, associated stance somaclonal variation). Interfertility among eucalypt spe- with the operational use of geneti- A related area of risk, which is at cies is common, which favors cross- cally modified crops. the fringes of our brief but can have pollination among species. This may be major economic implications, is pos- A pervasive problem in evaluating accentuated by the way in which flow- sible misallocation of resources between risk is that many of the risks men- ering seasons can be greatly altered by GM and conventional breeding tioned in relation to genetic engineer- exotic environments, thereby creating (Burdon 1992; 1994; 1998). overlaps between species in flowering ing are perceived and to a great extent times, which can break down natural unsubstantiated, and that at least a reproductive isolation. However, any considerable proportion of them will Deployment-related impact of interspecific pollination is not be real. Much has been written Risks likely to be confined to ex-situ genetic speculatively about potentially devas- tating effects related to GM. However resources growing very close to Briefly, these risks may be grouped we wish to point out that some of the transgenic crops. The short rotations into some partly overlapping categories: that may appear superficially to be a risks associated to the products of ge- •Ecological mitigating factor, may not work en- netic engineering are largely shared tirely this way; where eucalypt planta- with the products of conventional •Human health breeding or even of natural evolution. tions feed highly capitalized pulp mills •Cultural objections the costs of supply disruption through •Crop vulnerability. 128 Ecological risks ible if it occurs. It is a frequent phe- HGT occurs regularly but, even in nomenon in nature, particularly be- conjunction with ubiquitous mutation Ecological risks involve, in prin- tween bacterial species (Lorenz and in microorganisms, seldom if ever does ciple, two avenues of gene flow: Wackernagel 1994; Eisen 2000), but any ecological harm. Thus if there a HGT into and between higher organ- high probability of HGT occurring, it •Pollen flow into natural popula- isms has also been postulated and oc- seems very unlikely to be ecologically tions, invasion of natural ecosystems casionally demonstrated (Nielsen et al. harmful if it does occur. through seed produced by 1998; Kado 1998). The transfer of spe- transgenic crops, and transgenic cific DNA sequences from cultivars developing a weed poten- Agrobacterium to cells of plant species Human health risks tial in their own right during an infection process, can also be This category is technically prob- •Spontaneous HGT in the field regarded as HGT. It is currently the lematic, in the sense of generally rep- (Syvanen et al. 1994; Dale 1999; only known example of a type of HGT resenting tenuous possibilities, even Mullin and Bertrand 1998). that not only occurs frequently in nature, but also involves the transfer of with most food crops, but vehement All such risks will depend strongly DNA from a microorganism to a perceptions are now a fact of life (Ho on the introduced gene(s) conferring a higher organism. The transfer of DNA 1998; Antoniou 1996). The only food- material fitness advantage in the wild. from gut bacteria to cells lining the stuffs provided directly by pines are Often there is no reason to expect any mammalian gut has occasionally been seeds, and the species concerned are such advantage, and any effect will al- postulated, but solid scientific data to not major plantation crops. Indirect ways have to be evaluated in context support this hypothesis has not been food production occurs through collec- with the particular environment the produced. The unraveling and study of tion of fruiting bodies of edible sym- organism is placed in. Also, any effect HGT has so far has led to the notion biont fungi, which again raises the is- and its magnitude will have to be com- that the total genome, including all sue of HGT (Droege et al. 1998). pared with those arising from the use organisms, may in fact be highly flex- Other, very hypothetical possibilities of already accepted and practiced tree- ible (Jain et al. 1999; Lawrence 1999), arise through the role that pine mate- improvement technologies. Further, and that horizontal gene transfer is a rial might play in natural food chains, the issue of HGT is far from trivial in common tool of evolution (de la Cruz or in incidental contamination pro- that if it occurs at reasonable frequency, and Davies 2000; Woese 2000). It is cesses (e.g., through deposition of pol- its impact may not be any greater than therefore highly questionable whether len on food crops). Consider however, the simultaneously occurring horizon- GM would add materially to the effects that a gene transformed into pine can tal gene transfer of identical genes in a of natural mutation combined with contaminate food resources, but it can natural environment. Horizontal gene any natural capacity for HGT, given also flow from its original host organ- transfer may very well exist—and that that mutation surely occurs frequently ism to potential foods. The difference would be an indication that the “total among the vast numbers of individual between the two sources appears sig- genome” of all organisms taken to- microorganisms. nificant to opponents of genetic engi- gether is indeed more flexible than pre- Looking at the situation slightly neering, but the data to substantiate viously expected (Jain et al. 1999). differently, the probability of harm aris- their claims are not forthcoming. Among the perceived risk catego- ing from HGT represents the probabil- With issues of human health and ries for transgenic forest plantations, ity that HGT will occur, multiplied by ecological side effects many parts of HGT is perhaps the most difficult to the probability that, if it has occurred, society call for applying the precaution- discuss on a rational, properly in- harm will result. If the probability of ary principle. On the other hand, formed basis. Also it poses the prob- occurrence is high, that will almost many of the undesirable possibilities lem of not being amenable to risk certainly be part of a situation whereby invoked seem remote indeed and sci- spread and being potentially irrevers- entific data to substantiate claims are 129 still not produced. Concerning foods matically seal people off from the pro- tially important, and will give it spe- and food chains, even secondary prod- cesses of education and open-minded cial attention. The economic signifi- ucts of the very small number of new discussion of facts and foreseeable con- cance, and thence the social signifi- genes that would be used would arise sequences of using transgenics. Further, cance, are potentially enormous. in the context of many thousands of it is suspected that if such religious and Categories of technical risk, pri- natural products that are routinely cultural concerns were applied even- marily associated with deployment of detoxified in the small amounts that handedly, there would be objections to transgenics, are summarized in Table 1, occur (unless food spoilage is involved, many other modern-day human activi- along with risk properties, predispos- such as the production of large ties that currently pass almost without ing risk factors, and appropriate types amounts of aflatoxin). thought. On a narrower front, genetic of countermeasures. Putative risk fac- engineering also must be reviewed criti- tors relating to possible cultivar decline Cultural issues cally in comparison with older technolo- associated with genetic transformation gies, such as breeding and the mass are summarized in Table 2. These two It has become evident that signifi- clonal propagation used to deliver ge- Tables provide a backdrop for much of cant parts of society regard the transfer netic gains. the ensuing discussion. of genes from one organism to another Crop vulnerability can possibly as against their religious or cultural be- Crop vulnerability arise from genetic modification in vari- liefs. Such beliefs, whatever their origins, ous ways (Table 2), some direct and are often extremely difficult to address While this issue has been little some indirect. The more direct ways and overcome, because they can auto- publicized, we see it as being poten-

Table 1. Summary of categories of technical risk, putative risk properties and risk factors, and potential approaches to counter the risks.

Risk category Type(s) of prime General Specific Likelihood Potential severity Predisposing factor(s) countermeasures Related to technology Transient gene High Troublesome, but in Transformation technique, Careful testing development expression some situations could gene(s) concerned? even be advantageous Ecological Direct Largely precluded Problem in managing Wind pollination, [species Conferring sterility contamination of by exotic status of ex-situ gene resources native to area, presence of ecosystems genera in question interfertile relatives], efficient seed dispersal Horizontal transfer Frequent Most unlikely to be Highly dependent of selective HGT mechanism needs to amongst significant advantage of transgene(s) be better understood. microorganisms, in field Need for more research. unlikely for higher organisms Human health Allergenicity Extremely low for Most unlikely to be Wind pollination, quest for No specific measures transgenes serious durable heartwood Food contamination Extremely low Most unlikely to be Conceivably applicable to No specific measures serious nectar (eucalypts) or edible envisaged fungi (pines) Crop vulnerability Cultivar General hazard, Very high at top of Various (see Table 2) Various (see Table 2) decline/failure widely variable range (see Table 2) Non-durability of Potentially Potentially troublesome. Relying on single genes of Using multiple resistance resistance significant large effect for resistance. factors (with its own risks)

Note: Risk categories need not be mutually exclusive. For example, horizontal transfer could theoretically lead to food contamination, while non-durability of resistance could cause cultivar decline/failure. 130 Table 2. Summary of factors believed to generate risks of cultivar decline/failure, with suggested ranking (in descending order) of risk potential for different categories within each factor, and preferred countermeasure(s) for each factor. Note that many of the conclusions here must be based on opinion, for lack of solid data.

Factor Ranking of categories within factor Remarks

Transgene source Synthetic genes > genes from distant taxa > genes from Risk spread a potential defense close relatives > genes from within species* apart from choice of source

Transformation method Biolistics > Agrobacterium Agrobacterium poses greater technical difficulty with taxa concerned. Risk spread potentially very effective defense. Role of transgenes Structural genes > regulators or anti-sense sequences Lower rankings unclear Type of genes Homeotic genes (e.g flowering) > other genes May constitute a key hurdle in suppressing reproduction Magnitude of gene effect Large > small Major genes preferred nonetheless, for various reasons Number of gene insertions Multiple > few > single Multiple insertions may be needed on regulatory grounds, or to confer durability of resistance In-vitro culture technique(i) Single-cell lines > organogenetic cultures (ii) Adventitious shoots > axillary shoots

* Condition favoring use of traditional breeding.

involve side-effects of the action of new What may be the classic object genetic transformations or specifically structural genes or of modification (ei- lesson is the case of the corn blight to ones involving flowering. One ther over-expression or down-regula- epidemic in the United States in 1970 breeding scientist’s view is that it pru- tion) of the action of existing genes; (Levings 1989). It resulted from a com- dent to assume the former, even side- effects of the process of inserting bination of massive reliance on the though the homeotic nature of various the new DNA sequences; or silencing Texas cytoplasmic male-sterility factor flowering genes gives reason to believe of newly introduced genes (Finnegan in order to produce the hybrid maize, that risks might be higher when sup- and McElroy 1994; Matzke and an unsuspected side-effect whereby that pression of flowering is involved. How- Matzke 1995). Similar effects can be factor created extreme susceptibility to ever, this case can also be seen, at least expected from the combination of sets a new strain of the pathogen, and the by one transformation scientist, as an of new genes that have the potential to eventual appearance of that pathogen illustration of how unwanted and un- influence each other, creating undesir- strain. Admittedly that involved a mu- expected (and sometimes hazardous) able effects. Indirect effects can arise, tant gene in an organelle genome and effects are not restricted to the use of for instance, from the impacts on de- it did not involve genetic modification, material produced by GM. Classical ployment practices resulting from the so it is not strictly parallel to what breeding efforts can lead to exactly that use of transgenic material, such as gen- would be achieved by genetic transfor- as well, although the case in point, erating a shift towards use of clonal mation involving the nuclear genome. while arising in the era of classical material, or from extending the envi- It is a matter of opinion as to how rel- breeding, was achieving something that ronmental range of a species after in- evant this case is to genetic transfor- is now being pursued by GM. Any serting a disease resistance gene only for mation in general. Also in the realm of transgenes that get used, and their the resistance to break down through opinion is whether, even if it is rel- ‘downstream’ products, will admittedly pathogen mutation (Burdon 1999). evant, it is so to a broad spectrum of be much better characterized than this 131 male-sterility factor, but that will not ably alarmist in the context of all avail- Low-risk Applications eliminate all possibility of nasty sur- able evidence. It is observable that prises. many claims made are based on lack Uses of genetic transformation will of scientific understanding or even a not be confined to operational use in The precautionary refusal to accept a scientific approach commercial crops. It has great prom- principle to the issue. Very often doomsday sce- ise as a research tool. Transgenes can narios are presented, with little or no be used to study pathways of gene ac- All told, this is an area of consid- scientific evidence to back up those tion and their effects on phenotype, as erable uncertainty, where there is scope statements. This has become particu- an aid to identifying appropriate breed- for enormous variation in subjective larly obvious in submissions to the ing goals, and/or helping identify de- assessments of the hazards, particularly Royal Commission on Genetic Engi- sirable genes that might be exploited those relating to ecological side effects neering in New Zealand (RC), where by enhancements of classical breeding. and human health. At one extreme proponents and opponents of genetic With purely research-oriented applica- there will be those who see virtually engineering were asked to present their tions some may still see risks, notably limitless need to apply the precaution- cases (www.gmcommission.govt.nz). A in the area of containment, but the ary principle. Among them there will witness brief by Dr. Elaine Ingham of magnitude of such risks should be only be those who may see individual haz- Oregon State University affiliations, a small fraction of that of the poten- ards as being slight, but the different who testified for the Green Party of tial risks associated with operational use hazards as being so numerous as to New Zealand, cited to the RC a non- and deployment. generate a significant total hazard; they existent paper in support of a claim will tend to be exasperated by what that genetically modified Klebsiella they see as the uncritical enthusiasm of planticola bacteria had, if released, the GENETIC the optimists. At the other extreme, potential to devastate plant life on the there will be those who see the hazards planet. When the non-existence of the MODIFICATION as being very minor, and they may in- paper was exposed, Dr. Ingham and COMPARED WITH clude devotees of Lovelock’s Gaia prin- the Green Party of New Zealand had ciple, whereby a biota, and its interde- to apologize to the RC for misleading EVOLUTION AND pendent environment, have an enor- them (Walter et al. 2001; Fletcher CLASSICAL BREEDING mous inherent resilience. For many of 2001). Furthermore, the use of data by the optimists the precautionary prin- Dr. Ingham to support her case in New genes have appeared and ciple might be seen, with much irrita- front of the RC has proved to be sci- spread through populations through- tion, as something like allegations of entifically unsustainable. Her conclu- out evolutionary history, even though child abuse in an acrimonious custody sions in relation to the potential harm- certain genes that perform basic func- battle, easily made but once made are ful nature of the bacterium were made tions may be highly conserved. It is inherently very hard to disprove. It is on the basis of a single experiment in evident that various processes that al- also interesting to note that there is no the laboratory where controls were not low the transfer of DNA sequences commonly agreed ‘ precautionary prin- properly applied. Also very important from one organism to another, such as ciple’, as more than 35 different ver- is that a naturally existing variant of HGT between bacteria, Agrobacterium sions exist. Klebsiella planticola shares the key, al- gene transfer to plant cells, and virus A major issue related to genetic cohol-producing characteristic of the infections, have all contributed to engineering of all crops is the lengths genetically engineered bacterium variation and long-term evolution (see to which some parts of society try to (Jarvis et al 1997), without showing above). It is instructive, though, to impose viewpoints that, while they any harmful effect on plant life in its compare the process of introducing a may dearly held, seem most improb- environment. transgene with these natural processes 132 whereby a gene can spread through a constructs that modify the expression RISK PROFILES AND population. In nature, a new allele will of existing structural genes, which are typically appear as a rare variant. It can likely to present relatively low risks, to RISK MANAGEMENT have various fates, which range from structural genes of totally novel func- rapid, often random extinction, tion within the species, for which the Our focus here is on deployment- through persistence at low, if variable attendant risks may be much higher. related risks (Table 1). For any risk frequencies, to spreading through the Although HGT may lead to rapid there will be a profile of probability in population more or less slowly and ulti- spread of genes through populations of relation to severity. In principle, this mately becoming fixed. The exact fate some organisms, it appears most im- will conform to a mathematical func- will depend whether it is selectively probable that it could have any serious tion, but one that is seldom closely neutral or nearly so, on the one hand, impact on a commercial plantation or defined; typically there will be a high or significantly advantageous, on the its associated biota. probability of minor losses trailing other, plus large elements of chance. Artificial selection is in several re- away to much lower probabilities of Alternatively, new alleles can enter the spects intermediate between genetic much higher percentage losses. The population by migration (usually pol- modification and natural evolution. In distinction between seriousness and len and/or seed dispersal). Either way, that it depends strictly on genes that severity can be illustrated by how a a new allele will tend to enter the have occurred naturally within living total loss, if it involves only a small, population on a ‘toe-in-the-water’ ba- populations of a given organism, it isolated plantation, may not be serious, sis. If such a gene proves deleterious, stands close to evolution. However, the provided the owner has a good risk even with some time lag, it will be intensive and highly directional selec- spread. In fact we are typically look- eliminated or remain at a low fre- tion that it can entail can generate ing at low probabilities of ‘disaster’ out- quency with minimal cost in popula- closer parallels with use of transgenics, comes, albeit with almost no idea of tion fitness. Population bottlenecks, of especially if technology can be used to exactly how low they are, although one sort or another, can accelerate the select for specific genes that have ma- some transformations will be seen as process and even set evolution on new jor phenotypic effects. less risky than others (Table 2). On the courses, but will be accompanied by Operational use of transgenics can other hand, the length of rotation can risks of extinction. Yet even a popula- thus lead to more rapid and complete make such disasters very serious, de- tion bottleneck would seldom, by it- exposure to risks of adverse side-effects pending on the exposure to specific self, bring a gene to a frequency that of specific genes than can be expected risks. The potential seriousness, even if would immediately allow it to pervade in natural evolution or even classical the probability is low, demands some the population. Viruses have the poten- artificial breeding, in which massive form of risk management (Burdon tial to spread genes rapidly through substitutions of particular alleles will 1999; 2001). If active countermeasures populations of many organisms and tend to be far slower (except for high- against risks cannot be well targeted, even across species barriers. However, risk cases of some complete clonal yet risk spread is feasible, then risk viruses for pines or eucalypts have not monocultures). This does not mean spread is strongly indicated. yet been described. that undue risk exposure cannot be In practice, we will be working Use of a transgene, by contrast, incurred with conventional breeding, with very little quantitative informa- can immediately raise the effective fre- particularly when combined with mass tion or, in the jargon, with great un- quency of a new gene to 100% in a vegetative propagation for clonal for- certainty. True, there are many cases portion of the crop’s range, if its action estry (e.g., Libby 1982; Burdon 2001), where transformation has been associ- is fully dominant, which would create but it does mean an essentially new ated, directly or indirectly, with disas- 100% exposure to any adverse side ef- area of risk which poses its own man- trous effects on fitness, but they will fects. Clearly, though, there are various agement challenges. often be manifested in cell lines that categories of transgenes, ranging from are obviously not worth committing to

133 field trials although the exact reasons be used. This is true however for any less stringent requirements of suppres- for their state are seldom identified. For technology in use, and people are al- sion, a number of independent sup- the material that has sufficient prom- ways prepared to accept some risk, as pressor genes that could be available in ise to be tested in the field, the risks, long as sufficient benefits are realized various combinations. in terms of the probability function in or anticipated. The amount and level Meeting this requirement may, relation to severity, remain very uncer- of field testing of transgenic material ironically, be impeded by regulatory tain. However, the potential serious- before commercial release has to be mechanisms that require separate pro- ness, which may be extreme in the case guided by scientific argument, but also cessing of all elements of risk spread, of delayed manifestation of induced by the level of residual risk a society is proliferating costs of compliance, disease susceptibility, may dominate prepared to take. rather than addressing an appropriate the appropriate risk management. A In the present context of possible spread of risks as a single ‘package’. significant period of field testing use of transgenic plantations, risk For very low-probability events (which can be preceded by laboratory spread is seen as a crucial component such risk spread, where it is applicable, testing and physiological studies) seems of risk management. It appears espe- can reduce the probability of cata- essential in some, but not all cases, to cially appropriate for addressing poten- strophic crop failure by orders of mag- eliminate effects such as greatly en- tially very serious events of low, but nitude. To express this in another way, hanced susceptibility to climatic dam- very uncertain probabilities. An ideal the probability of simultaneous ‘disas- age (say, in the case of a transformant risk-spread strategy for GM or conven- ters’, each of low probability, involving for wood chemistry that might ad- tional products should be based on the an unacceptable proportion of a set versely affect the mechanical stability following planks (cf Burdon 1999): independent transformations is should of the tree). However, such testing will become almost vanishingly low. On the 1. As a starting point, using a diver- only reduce the risks, rather than to- other hand, such use of multiple sity of recipient genotypes, tally eliminate them, and in the short transgenes will increase the theoretical term, is liable to greatly erode the po- 2. It should involve different, and as risks of HGT, but by no more than a tential time savings in using GM com- precise as possible, gene-insertion factor of the number of risk-spread el- pared with conventional breeding. The events, ements—if no such element can be transformation scientist, however, looks identified as incurring elevated HGT- to the future when, with the risks of 3. Use of a number of different related risks. Thus, the relative increase transgenes being much better known, transgenes or regulator constructs in in HGT-related risks should be far less the field testing can be foreshortened order to achieve a particular objec- than the relative reduction in risks of to the point of allowing major time tive, GM-related crop failure. savings using GM compared with con- 4. It should attempt to avoid the co- For the longer term, leading into ventional breeding. transfer of unwanted DNA se- active countermeasures against the In general, society has to make quences. risks, more research into gene integra- decisions on the use of this new tech- tion and expression characteristics, and nology. This will involve weighing up Requirement 3, however, may be research on the influence of transgenes the prospective benefits against the difficult to meet with genetic modifi- on natural ecosystems, should help to risks. In weighing up the risks there cation designed to eliminate all repro- define constructs that carry a signifi- will be considerations of how the risks ductive activity, which would be de- cantly lower risk potential. Examples can be addressed by risk spread and/or sired for reliable genetic containment are new strategies for the selection of active countermeasures against the risk as well as optimizing resource alloca- transgenes avoiding antibiotic selection (Table 1; see also Burdon 2001), and tion for crop production. However, or the complete elimination of selec- what remaining level of risk may have opinions may vary on the need for tive markers (Joersbo and Okkels 1996; to be accepted if the technology is to complete suppression and, for meeting Haldrup et al. 1998; Sugita et al.

134 1999). Although a perceived negative e.g., conventional breeding, which will for combating some of the most seri- effect of antibiotic-resistance markers is have its own risks, although these are ous diseases from this planet. The ben- not substantiated by scientific data, and much more readily accepted by soci- efits of genetic engineering are becom- it is possible to involve antibiotics that ety at this point in time. Nor should ing obvious in many areas. In medicine are of no further clinical interest, the it be forgotten that GM can be used they are huge, and in many areas they avoidance of such markers may in- as a research tool, largely without in- are very widely accepted. In agriculture crease acceptance of genetic engineer- curring the risks attendant upon opera- they are also great, but are now beset ing by the general public. tional use in commercial crops. by problems of public acceptance. The There may be a call to use mul- A breeding scientist may argue example of Golden Rice, a genetically tiple transgenes simultaneously in cul- that the risks associated with conven- modified and vitamin A enhanced rice tivar genotypes, based on technical tional breeding may be more easily that has the potential to save hundreds needs and/or regulatory requirements managed than those of transgenics, of thousands from blindness, is just for genetic containment in field de- which typically involve new genes of one example. In forestry the prospec- ployment. This has a potential, which large effect. However, a transformation tive benefits are also great (Strauss et is admittedly very speculative, to in- scientist may consider the less predict- al. 1999), but the ecological ramifica- crease greatly the inherent risks associ- able risks of conventional breeding that tions pose problems of acceptance ated with deployment of transgenic are associated with many unidentified while the time frames for plantation material (Burdon 1999). Not only will genes as greater than GM of one char- crops both create attractions for use of there be a straightforward accumula- acterized gene. Thus the comparative GM and accentuate certain risks. The tion of the risks associated with each risk potential for the alternative tech- lesson is surely that those who manage individual transgene, but increasing nologies remain arguable. Possibly the applications of technologies like numbers of transgenes will exist in much more important is that GM can genetic modification need to have good greatly increasing numbers of combi- in some cases lead to more rapid and advance preparations for surprises and nations that might generate adverse complete exposure to certain risks. For they need to have management tools interactions between different some conventional breeding products, to minimize any impact on the envi- transgenes—if such transgenes involve however, this problem can also arise. ronment, human health, or economic significant risk factors (Table 2). Experience with other technolo- prosperity. gies, notably biotechnology, is that na- We can make our own technical ture often springs nasty surprises, just judgements of risks, and enhance them CONCLUDING when one imagines that problems are with quantitative risk analysis, and de- solved. Resistance to antibiotics, the vise risk-management strategies. Yet we COMMENTS dangers associated with blood transfu- cannot ignore public perceptions, how- sions, and the corn blight epidemic of ever exasperated we may become at The focus has been largely on the 1970 are just three of the most obvi- times. Moreover, even the occasional risks associated with genetic transfor- ous examples. They illustrate, however, scientific fiasco can have strong and mation. However, they need always to the need for a comprehensive risk/ben- enduring effects on public confidence. be balanced against the prospective efit analysis and an informed decision Subjective judgements on the part of benefits, noting that the benefits will on what level of risk we are prepared the public may be perceived vividly by require critical examination in respect to accept for obtaining a certain level scientists, but scientists can never avoid of both their validity and their depen- of benefit, in a context of an unavoid- areas where they must be guided in dence on the use of genetic transfor- able element of uncertainty. Notwith- some degree by their own subjective mation. The risks must also be placed standing the problems of antibiotic re- judgements. in perspective with those of alternative sistance, no one would seriously de- methods of achieving the same goals, mand that we stop using a key weapon

135 Cremer, KW. 1992. Relations between Fielding, JM. 1960. Branching and REFERENCES reproductive growth and vegetative flowering characteristics of growth in Pinus radiata. Forest Ecol- Monterey pine. Bulletin 37. For- Anon. 1999. Australian/New Zealand ogy and Management 32: 179–199. estry and Timber Bureau, Standard: Risk Management AS/ Canberra, NZS 4360:1999. Standards Austra- Cromer, RN, KG Eldridge. 2000. The lia. ISBN 0 7337 2647X. eucalypts as tree crops. Pp. 226– Finnegan, J, D McElroy. 1994. 269 in Last, FT (ed.) Ecosystems of Transgene inactivation: Plants fight Antoniou, M. 1996. Genetic pollution. the World, Vol. 19: Tree Crop Eco- back! Bio/technology 12: 883–888. Nutritional Therapy Today 6(4): 8– systems. Elsevier, Amsterdam, The Fletcher, L. 2001. New Zealand GMO 11. Netherlands. debacle undermines green lobby. Bazin, MJ, JM Lynch, eds. 1994. Envi- Dale, PJ. 1999. Public concerns over Nature Biotechnology 19: 292. ronmental gene release. Models, ex- transgenic crops. Genome Research Haldrup, A, SG Petersen, FT Okkels. periments and risk assessment. 9:1159–1162. Chapman & Hall, Suffolk, En- 1998. Positive selection: a plant gland. De la Cruz, I, I Davies. 2000. Hori- selection principle based on xylose zontal gene transfer and the origin isomerase, an enzyme used in the Burdon, RD. 1992. Tree breeding and of species: lessons from bacteria. food industry. Plant Cell Reports the new biotechnology — in dam- Trends in Microbiology 8(3): 128– 18: 76–81. aging conflict or constructive syn- 133. ergism? Pp. 1–7 in Proceedings of Harcourt, R, X Zhu, D Llewellyn, E IUFRO Meeting on Breeding Tropi- DiFazio, S, F Leonardi, WT Adams, SH Dennis, J Peacock. 1995. Genetic cal trees, Cartagena & Cali, Colom- Strauss. 1999. Potential impacts of engineering for insect resistance in bia, October. CAMCORE, North hybrid poplar plantations on black temperate plantation eucalypts. Pp. Carolina State University, Raleigh, cottonwood plantations (Abstract). 406–408, in Eucalypt Plantations: North Carolina, USA. Proceedings of the Western Forest Improving Fibre Yield and Quality. Genetics Association Meeting, Flag- Proceedings CRCTHF - IUFRO Burdon, RD. 1994. The place of bio- staff, Arizona, July 1999. Conference, Hobart, Australia, technology in forest tree improve- 19–24 February. CRC for Temper- Droege, W; A Puehler, W Selbitschka. ment. Forest Genetic Resources. ate Hardwood Forestry, Hobart. FAO, Rome, No. 22. Pp. 2–5. 1998. Horizontal Gene Transfer as a biosafety issue: a natural phe- Ho, MW. 1998. Genetic engineering: Burdon, RD. 1999. Risk-management nomenon of public concern. Jour- Dream or Nightmare? Gateway issues for genetically engineered nal of Biotechnology 64(1): 75–90. Books, Bath, UK. forest trees. New Zealand Journal of Forestry Science 29: 375–390. Edwards, GA, NW Fish, KJ Fuell, M Jain, R, MC Rivera, JA Lake. 1999. Keil, JG Purse, TA Wignall. 1995. Horizontal gene transfer among Burdon, RD. 2001. Genetic aspects of Genetic modification of eucalypts genomes: the complexity hypoth- risk - species diversification, genetic - objectives, strategies and progress. esis. Proceedings of the National management and genetic engineer- Pp. 389–391, in Eucalypt Planta- Academy of Sciences USA 96(7): ing. New Zealand Journal of For- tions: Improving Fibre Yield and 3801–3806. estry 45(4): 20–25. Quality. Proceedings, CRCTHF - Jarvis, GN, ERB Moore, JH Thiele. IUFRO Conference, Hobart, Aus- Burdon, RD. (in press). Introduction 1997. Formate and ethanol are the tralia, 19–24 February, CRC for (Chapter 1), in Pines of Silvicultural major products of glycerol fermen- Temperate Hardwood Forestry, Importance (ed. R Stubbs). CAB tation produced by Klebsiella Hobart. International, Wallingford, UK. planticola strain isolated from red Burdon, RD, S Kumar. (in ms). Sto- Eisen, JA. 2000. Horizontal gene trans- deer. Journal of Applied Microbiol- chastic modelling of the impacts of fer among microbial genomes: new ogy 83, 166–174. insights from complete genome four generations of pollen contami- Joersbo, M, FT Okkels. 1996. A novel analysis. Current Opinion in Genet- nation in unpedigreed gene re- principle for selection of transgenic ics and Development 10(6): 606– sources. Submitted to Silvae plant cells: positive selection. Plant 611. Genetica. Cell Reports 16: 219–221.

136 Kado, CI. 1998. Agrobacterium-medi- Richardson, DM, ed. 1998. Ecology and ated horizontal gene transfer. Biogeography of Pinus. Cambridge Genet. Eng. 20: 1–24. University Press. Lawrence, JG. 1999. Gene transfer, spe- Sedgley, M, AR Griffin. 1989. Sexual ciation, and the evolution of bac- Reproduction of Tree Crops. Aca- terial genomes. Current Opinions in demic Press. Microbiology 2(5): 519–523 Strauss, SH, WH Rottmann, AM Levings, CS. 1989. The Texas cytoplasm Brunner, AM, LA Sheppard. 1995. if maize: Cytoplasmic male steril- Genetic engineering of reproduc- ity and disease susceptibility. Sci- tive sterility in forest trees. Molecu- ence 250: 940–947. lar Breeding 1: 5–26. Ledig, FT, DIH Linzer. 1978. Fuel crop Strauss, SH, et al. 1999. IUFRO Posi- breeding. Chemtech. 8: 18–27. tion Statement on Biotechnology. http://www.fsl.orst.edu/tgerc/ Libby, WJ. 1982. What is a safe num- iufro_pos-statm.htm. ber of clones per plantation? Pp. 342–360 in Heybroek, HM et al. Sugita, K, E Matsunaga, H Ebinuma. (eds.) Resistance to Diseases and Pests 1999. Effective selection system for in Forest Trees. Proceedings of the generating marker-free transgenic 3rd International Workshop on plants independent of sexual cross- Genetics of Host-Pathogen Inter- ing. Plant Cell Reports 18: 941– actions in Forestry, Septem- 947. ber1980, Wageningen, The Neth- Syvanen, M. 1994. Horizontal gene erlands. Pudoc, Wageningen. transfer: Evidence and possible Lindgren, D, L Paule, XH Shen, R consequences. Annu. Rev. Genet. Yazdani, U Segerstron, J Wallin, 28: 237–261. ML Lejdebro. 1985. Can viable Thielges, BA. 1982. A strategy for pollen carry Scots pine genes over breeding population-based disease long distances? Grana 34: 64–69. resistance in forest trees. Pp 194– Lorenz, MG, and W Wackernagel. 215 in: Proceedings of the Joint 1994. Bacterial gene transfer by IUFRO Meeting of Working Parties natural genetic transformation in in Genetics about Breeding Strate- the environment. Microbiol. Re- gies Including Multiclonal Varieties, views 58(3): 563–602. Sensenstein, September 1982. Lower Saxony Forest Research In- Matzke, MA; AJM Matzke. 1995. How stitute, Staufenberg-Escherode, and why do plants inactivate Federal Republic of Germany. transgenes? Plant Physiology 107: 679–685. Walter, C, MV Berridge, D Tribe. 2001. Genetically engineered Klebsiella Mullin, TJ and S Bertrand. 1998. En- planticola: A threat to terrestrial vironmental release of transgenic plant life? trees in Canada - potential benefits www.forestresearch.co.nz/pdf/ and assessment of biosafety. The InghamRebuttal.pdf. Forestry Chronicle 74 (2): 203–219. Woese, CR. 2000. Interpreting the uni- Nielsen, KM, AM Bones, K Smalla, JD van versal phylogenetic tree. Proceed- Elsas. 1998. Horizontal gene transfer ings of the National Academy of Sci- from transgenic plants to terrestrial ences USA 97(15): 8392–8396. bacteria—a rare event? FEMS Microbiol. Rev. 22(2): 79–103.

137 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 138-147. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Use of Transgenic Resistance in Plantation Trees: Efficacy, Risk, and Integration with Other Pest Management Tactics

Kenneth F. Raffa

ABSTRACT

Transgenic expression provides some opportunities for pest resistance in trees that cannot be achieved by other means. However, environmental risk assessment poses some special difficulties with trees. These relate to the long time scale of host relative to pest generation times, the large spatial scale of some plantation systems, and the multiple uses and ecological functions of adjoining conspecific plantation and native trees. Several issues are addressed: 1) How does the scale at which we evaluate GMO’s affect our estimate benefits and risks? 2) What risks can be identified, and how can they best be addressed? 3) What experiences provide the most sound precedents upon which to evaluate transgenic trees: Should we evaluate them as pesticides deployed by another means, traditionally bred resistant varieties, or planned introductions of biocontrol agents? The integration of multiple pest management tactics, implementation of biotype management strategies, examples where general principles can be more valuable than specific details, and weaknesses in current funding approaches to risk assessment are discussed.

early 20 years have passed since the first report of transgenic expres sion in plants (Meeusen and Warren 1989). It’s been over 10 years since Nthe development of transgenic Bt resistance in hybrid poplar against tent caterpillars and gypsy moths (McCown et al. 1991; Robison and Raffa 1994). Yet we still have only a limited ability to assess the long-term implications of genetically modified organisms on forest ecosystems. I’d like to suggest four reasons for keeping the issue of environmental safety of genetically engineered trees at the forefront: First, trees pose some inherently distinct problems. Unlike agricultural crops they have generation times much longer than those of their major pests. Also, trees are both commercial plantings and components of native ecosystems. The same species grow adjacent to each Kenneth F. Raffa is with other, and provide multiple ecological functions and human resources. Second, the Department of the track record of agricultural transgenic plants is too brief a time scale to allow full evaluation (Parker and Karieva 1996; van Embden 1999). By comparison, Entomology and the problems arising from injudiciously selected biological control agents, calendar Department of Forest application of pesticides, vertebrate predator elimination, and fire suppression often were not detectable until many decades after implementation. In general, the longer Ecology and Manage- the time lag between implementation and the appearance of a problem, the greater the difficulty in addressing it. Third, pest managers and seed producers of agri- ment, University of cultural crops have implemented specific practices aimed at preventing biotype Wisconsin - Madison. evolution (Roush 1997), but parallel practices have not yet been developed for trees. This reflects both the greater emphasis placed on agriculture and the diffi- [email protected] culties of working with trees. Fourth, government-sponsored research on Risk 138 Assessment of GMOs has fostered a against all of them. It is especially dif- into a specific mechanism. That insight false sense of assurance. As anyone who ficult when one insect prefers the very is often achieved by removing extrane- vies for competitive grants can attest, group of phytochemicals that confers ous noise from the system. In the case a key to success is defining tight, nar- resistance against others (Raffa et al. of ecologists, our ability to resolve a rowly focused experiments that can be 1997). question is usually improved once we conducted under controlled conditions Deploying pest-resistant trees may identify additional factors that are im- over short time frames, using organ- allow for reduced input of insecticides. pacting on our unit of study. Our pre- isms that lend themselves to simple This has already been demonstrated in vious lack of understanding is over- bioassays. Unfortunately, the key ques- some agricultural crops. Moreover, en- come when we recognize how to intro- tions about environmental safety are hancing productivity on intensively duce complex feedback into what we long-term, complex, and operative managed lands may alleviate demands previously thought was a simpler sys- across multiple trophic levels. on the land base, thus allowing more tem. My discussion will focus on regions to be set aside for wilderness Let me provide some illustrations. transgenic resistance against insects in and biodiversity objectives. Transgenic Sheep production is an important in- plantation trees. However, I hope that capabilities may also help us respond dustry throughout much of western some of the general principles I discuss more quickly to invasive species, which North America. Coyotes can pose sig- can be applied to other types of are an increasing threat associated with nificant problems, by feeding on transgenic properties. I also need to heightened global commerce. lambs. So range managers set out to emphasize that this discussion cannot be A number of potential risks have eradicate coyotes. Early attempts were extrapolated to self-regenerating trees, also been identified. These include evo- unsuccessful: the coyotes were too dissemination by insect vectors, etc. lution of insect biotypes resistant to the clever and avoided the traps. Eventu- transgene, adverse direct and indirect ally trap and bait technologies im- effects on nontarget organisms, and proved, and coyote populations de- POTENTIAL AND RISKS alteration of ecosystem processes such creased. Thereafter sheep populations as nutrient cycling and pollination. crashed too. The reason was that range OF GENETIC Molecular biologists and ecologists biologists underestimated the complex- ENGINEERING agree on both the potential benefits ity of the system. Coyotes also eat rab- and risks of transgenic plants, but dif- bits, which soared after coyotes were The advantages provided by ge- fer on the degree to which they em- eliminated, devoured the grass, and netically engineered pest resistance are phasize each. I think this reflects dif- starved the sheep. Similar experiences significant. Insects and pathogens are ferences in the level of biological orga- occurred with attempts to remove major limitations to tree productivity nization at which they work. Ecology wolves to favor beaver fur production, (Can. FS 1994). Moreover, the condi- is the study of interactions. Ecologists with fire suppression, and with DDT tions under which trees are grown deal with systems that are open, com- applications against spruce budworm. commercially, usually even-aged, evenly plex, and to some extent undefined. The general lesson is that when you spaced monocultures, make them more Molecular biologists deal with systems transfer technology from closed, con- manageable for the producer, but also that are closed, controlled, and simpli- trolled conditions to open complex much more susceptible to the insects fied to the best of their ability. There systems, there are almost always un- that consume them. Breeding for re- is an inherent tension between the re- foreseen parameters that affect system sistance is far more difficult in trees ality of the field vs. the precision of the behavior. than inannual crops, for obvious rea- laboratory. When a molecular biologist Based on their different experi- sons of size and time. Moreover, trees encounters a difficult problem, resolu- ences, molecular biologists and ecolo- are faced with complexes of insects, tion often comes through either a tech- gists tend to ask the question differ- and it is difficult to select for resistance nical advance or an improved insight ently. Molecular biologists frequently

139 ask, “What are the specific risks that starting with a point of common agree- ity if does not proceed to a critical can be anticipated based on the prop- ment. This is provided by the first evaluation of each potential harm, or erties of a specific product?” This re- National Academy of Sciences (1987) if it extrapolates uncritically from arti- flects a belief that assuming a risk report that has shown a remarkable ficial conditions to ecosystem levels. I might exist without the ability to char- ability to stand the test of time. This see three disadvantages to this ap- acterize or quantify it is unscientific. In report emphasized that the “product proach. First, it can lead to lost oppor- contrast, ecologists ask: “What is the not the process” should be the focus. tunities. No undertaking is without evidence that specific information de- This report stated, “Assessment of risk risk, and refusal to manage it can pre- rived from closed systems over short should be based on the organism, not clude realization of the benefits de- time intervals can be extrapolated to the method of engineering.” Remark- scribed above. Second, well-intended multiple interactions in open systems ably, a white paper by the Ecological environmental policies, like well-in- over longtime periods?” This reflects a Society of America independently tended technologies, can have unin- belief that assuming extrapolation in reached an identical conclusion (Tiedje tended consequences. An example is the absence of proof is as unscientific et al. 1989): “transgenic organisms the Delaney Clause, which was in- as performing experiments without should be evaluated and regulated ac- tended to prevent registration of car- testing whether Heisenberg’s Uncer- cording to their biological properties, cinogenic pesticides, and in the process tainty Principle is satisfied. To ecolo- rather than according to the genetic favored more toxic compounds. A third gists, deployment of GMOs is a clas- techniques used to produce them.” disadvantage is that it distracts atten- sic scaling problem, and there are not The product not process guideline tion from more meaningful environ- many examples where extrapolating is sometimes misinterpreted to mean mental risks, both unrelated and re- across scales yields consistently predict- that environmental concerns with lated to GMOs. able relationships. GMOs are unwarranted. Nothing At the other end of the spectrum Resolving such dichotomies be- could be further from the truth. This is the view that deployment should tween the experiences and assumptions principle dismisses scientifically un- only be delayed where adverse conse- of molecular biologists and ecologists founded concerns that a particular quences have been demonstrated. This is one of the most important steps in product may exert an effect simply be- approach provides the advantage of both maximizing the benefits and cause it was genetically engineered. But insisting on scientifically based criteria. minimizing the risks of genetic engi- it also focuses attention on interactions: However, it has some significant dis- neering. Ecological understanding can How will this organism interact with advantages. First, it seriously underes- be useful both in identifying risks in- other organisms in complex, open eco- timates the complexity of scaling across herent in extrapolating from closed to systems, over long periods of time? levels of biological organization, from open systems, and in integrating mul- Moreover, the product not process per- small areas to landscapes, and from tiple sources of insect mortality to im- spective yields a corollary: “If the prod- time scales based on rapid bioassays prove efficacy. Likewise, molecular uct not the process is critical, then ex- within grant cycles, to times scales methods will provide some of the most pertise in the methods of genetic engi- based on ecological and evolutionary valuable tools for ameliorating risks neering is not directly relevant to pre- processes. There is nothing scientific identified by ecologists (Raffa 1989). dicting how novel organisms will in- about ignoring how system processes teract with ecosystems” (Raffa et al. change as you scale up, as there are 1997). many precedents in meteorology, soil DEFINING RISK There is no uniform agreement on erosion, engineering, and biology. Sec- ASSESSMENT how to approach risk. One approach ond, the argument that risks should be is to list every conceivable adverse con- considered scientifically based only How can we integrate molecular sequence. Perhaps this isn’t a bad start- once they have been ‘demonstrated’ and ecological approaches? I suggest ing point, but it loses scientific valid- removes the proactive element of risk

140 assessment, and shifts the emphasis to evolution is not a rare phenomenon, Gould 1992, Rebollartellez et al. 1994). remediation, which is far more diffi- but is rather an inevitable consequence The microbial insecticide Bacillus cult. This reactive approach has cost of any selection pressure that is applied thuringiensis is currently the most widely agrichemical companies severely, which continuously, uniformly, and at suffi- used insect control agent in forestry. It is why they now invest substantial re- cient intensity to cause high mortality is also widely used in tree fruit produc- sources to prevent pesticide resistance. (Gould 1988; Raffa 1989; Tiedje et al. tion. Its acceptance arises from a com- Finally, the argument that all risks can 1989; Abbott 1994; Seidler and Levin bination of efficacy, economic, and en- be classified as either ‘demonstrated’ or 1994; Timmons et al. 1995;1996; van vironmental attributes. Bt is sprayed on ‘conjectural’ creates a false dichotomy. Embden 1999). Biotype evolution has an as-needed basis only, which greatly di- It ignores the broad range of antici- occurred against all categories of pesti- minishes the likelihood of biotype evo- pated problems that may be considered cides, and also against resistant culti- lution. In contrast, uniform and con- ‘ realistic’. I believe we can define ‘re- vars, biological control agents, and cul- tinuous expression in trees could more alistic risk’ on the basis of two criteria tural manipulations (Raffa 1989). It is rapidly select for resistant insects, which (Raffa et al. 1997): 1) The existence of prevalent among all taxa of insects, in turn could reduce the efficacy of an relevant precedents that suggest likely pathogens, and weeds. The physiologi- environmentally compatible tool used in outcomes; 2) A proposed mechanism cal and biochemical mechanisms are other cropping systems. There is sub- that can be delineated on the basis of likewise well understood. In the case of stantial overlap in the species of insects biologically verified processes. I will insects, these mechanisms include al- that feed in these systems. An additional focus on two anticipated responses to tered behavior, detoxification, excre- effect that needs to be considered is the pest resistant transgenes in tree plan- tion, impermeability, and target site possibility that resistant biotypes will tations, evolution of resistant biotypes, insensitivity. have more damaging behavior than the and alteration of complex ecosystem The consequences of biotype evo- original genotypes. Examples of altered processes. lution range from loss of efficacy to behavior come from agroecosystems to additional secondary effects. In most which pesticides have been applied. An- cases, loss of efficacy is the major and other significant threat of biotype evo- BIOTYPE EVOLUTION: perhaps only significant consequence. lution is cross-resistance. Resistance to Lost efficacy poses a substantial risk to one chemical class often confers resis- RISKS AND the producer. Discovery, development, tance to widely unrelated compounds. MANAGEMENT registration, and marketing of new pest An example occurred with the introduc- control chemicals and transgenes are tion of the synthetic pyrethroids. Al- Biotype evolution is the emer- expensive processes. Corporations now though these materials were highly ef- gence of new pest races that are no invest heavily in biotype prevention fective against lab and most field popu- longer susceptible to a previously effec- tactics throughout all stages of lations, they were almost immediately tive control tactic (Roush 1987; 1997; agrichemical development to avoid ineffective in certain regions. The pat- Tabashnik 1994). The underlying these losses. tern soon became obvious. Areas in mechanisms are well understood, at Effects that extend beyond reduced which high levels of the organochlorine several levels of biological organization. efficacy are of more concern from an DDT had been applied harbored bio- At the population level, biotype evo- environmental perspective. These in- types resistant to pyrethroids. These lution is simply natural selection pro- clude impacts on other resource manag- compounds are structurally dissimilar, ceeding in an accelerated and directed ers, altered insect behavior, cross resis- but insect alteration of sodium channels fashion. A few individuals possessing tance to other pesticides, and cross re- conferred resistance against DDT, and fortuitous mutations survive, repro- sistance to natural plant defenses did likewise against pyrethrum. Pyre- duce, and ultimately occur at dispro- (Heinrichs and Mochida 1984, Fry thrum is a naturally occurring botanical, portionately high frequencies. Biotype 1989, Hilbeck et al. 1998, Johnson and so from the insect’s standpoint, DDT is

141 just a recent human variation on an an- cally modified plants. These tactics arose sity onto transgene expression adds ancient angiosperm theme. This raises the both from agriculture and our under- other layer of protection. Hybrid pop- possibility that genetically engineered standing of naturally coevolved stable lars seem particularly amenable to this products might not only select for cross plant – insect interactions. First, we approach. Whether or not clones are resistance to other pesticides, but to should identify those systems in which transgenic, this is a valuable safeguard naturally occurring plant defenses as biotype evolution appears particularly against new insect and disease races and well. This has already been observed with likely, and avoid them. Expression of species. Fourth, the transgene should be synthetic pesticides (Fry 1989). transgenic traits is more problematic compatible with other control methods. The conditions that foster evolu- with trees than annual crops, because of This helps maintain a diverse and option of resistance biotypes are relatively their long rotation times relative to in- posing array of selective pressures. Natu- well understood. In particular, the rate sects. Many tree production systems pro- ral systems often pose insects with con- of evolution of pesticide resistance de- vide crop to pest generation ratios an or- flicting alternatives. For example, aphids pends more on the pattern of applica- der-of-magnitude above that required to escape predation by releasing alarm tion than the mode of action. Patterns elicit pesticide resistance (Georghiou and pheromones when predators are near. of expression that are uniform, large- Saito 1983). Thus, rapid rotation sys- Some wild plants react to aphid feeding scale, unidirectional, and continuous tems such as Populus and Eucalyptus are by producing aphid alarm pheromone, select more rapidly for resistance than less at risk than are slower systems such which causes the aphids to drop from do intermittent patterns. If there has as Douglas-fir. Second, biotypes are less the plant. Imagine insects that became been one achievement of pest manage- likely to evolve when expression is lim- immune to this defense: if they ignored ment that ranks above all others over ited in time. This is a key lesson obtained alarm pheromones they would become the last 30 years, it is the replacement of the transition from calendar to den- more susceptible to predators. Fifth, ap- of calendar applications with targeted sity-triggered applications. One application of potentiators that specifically applications triggered by specific pest proach to simulating this is through interfere with resistance mechanisms can densities that surpass carefully defined wound-inducible expression. This is one greatly prolong activity. For example, economic thresholds. This transition of the most important means by which insecticide resistance is often due to el- has provided an overall reduction in long-lived trees maintain stable defenses evated detoxification enzyme titers. In- pesticides, increased profits to the against herbivores and pathogens. Other hibitors of P450’s, such as piperonyl grower, improved pest control, and approaches include limiting expression butoxide, interfere with detoxification, greater compatibility with diverse man- to certain periods of the growing season and so can render resistant individuals agement tactics such as biological con- and to certain age categories. Various susceptible. Again turning to coevolved trol, mating disruption, and cultural insect species show clear patterns of sea- systems as a model, some plants produce manipulations. Thus, a particular tech- sonal abundance, as well as association both insecticides and synergists. We have nology, such as constitutive transgene with particular age categories. Trees dif- recently reported synergy of Bacillus expression in plants, may simulta- ferentially allocate defenses according thuringiensis by linear aminopolyol neously represent a modern innovation these abundance patterns. Third, expres- zwittermicin A, an antibiotic from soil at one level of scale, and an archaic sion is most stable when it is limited in bacteria. A sixth strategy is continued throwback at another. Conversely, space. Spatial expression can vary among monitoring of resistance. As incipient many of the solutions to problems plant tissues, as is common in tree-re- resistance is identified, its mechanism identified by ecologists can best be sistance mechanisms, and also among and mode of inheritance are character- solved by molecular biologists, or by trees. This tactic has already been em- ized, and specific counter measures can molecular biologist – ecologist teams. ployed with transgenic cotton and corn, be employed. A number of tactics are available to in which fixed percentages of the Experience with pesticides also manage biotype evolution, and many of ‘transgenic’ seed are in fact not suggests what is unlikely to work. First, these can be readily transferred to geneti- transgenes. Superimposing genetic diver- it seems reasonable that interrupted use

142 of a pesticide will allow resistant popu- introgression are employed, then ad- where prior experience with pesticides lations to return to their original gene verse environmental effects are likely to and new cultivars have resulted in ex- frequencies. The underlying assump- be limited and subject to remediation. acerbated pest problems. These include tion is that resistance characters are However, if gene flow is likely, risk as- direct and indirect effects on natural inherently disadvantageous and will be sessment needs to be tailored accord- enemies, and release of competitors. selected against in the absence of the ingly. The extent to which transgenic pesticide. Unfortunately, this has The issue of potential gene escape trees will directly affect predators de- proven to not always be the case has generated much debate among ge- pends on the gene product. In particu- (Brattsten et al. 1986; Roush 1987, neticists (e.g., Regal 1993; Linder and lar stable materials are more likely to 1997; Gould 1988). Second, the strat- Schmitt 1995; Strauss et al. 1995; be biomagnified across trophic levels. egy that multiple genes of Bt are avail- Timmons et al. 1995; Karieva et al. For example, biomagnifcation is not able, and so we can stay ahead of re- 1996; Paoletti and Pimentel 1996; known to occur with Bt. As we project sistance by deploying new ones as re- Parker and Karieva 1996). However, into the future, three considerations sistances evolve is not supported by there is general agreement there is some should be considered. First, herbivores experience with other pesticides. Thus, gene flow from cultivated to native and usually evolve resistance to xenobiotics it is no surprise that pest management feral plants. Some potential mecha- more rapidly than do predators. This specialists are not enthused about this nisms include hybridization, dissemi- appears to relate to their coevolution- approach. Insects have shown remark- nation of vegetative material, and vec- ary history with plant defense chemi- able ability to develop cross resistance tors. Strauss et al. (1995) concluded cals. Second, even if a xenobiotic has against very distant chemical groups “gene flow within and among tree equivalent toxicity to an herbivore and (Georghiou and Saito1983; Brattsten populations is usually extensive, which its predator, it will have a greater ef- et al. 1986; Mullercohn et al. 1996). makes the probability of transgene es- fect on predator than herbivore popu- Cross resistances have occurred among cape from plantations high (Adams lations. This has been demonstrated in many classes of pesticides, and by sev- 1992, Raybould and Gray 1993)”. numerous mathematical models, and eral mechanisms. Given the evolution- They based their conclusions on “high relates to issues of prey finding, prey ary history of insects in contending rates of gene dispersal by pollen and handling, and differences in reproduc- with millions of phytochemical com- seed, and proximity of engineered trees tive capacity. Third, most predators are binations over hundreds of millions of to natural or feral stands on interfer- generalists, so control methods that years, placing our confidence in a few tile species.” reduce predators of target pests can varieties of Bt does not seem the best Proposed adverse environmental cause population increases of non-pest option. effects include enhanced weediness, species. Relationships become more reduced biodiversity, and alteration of complicated when parasitic insects are ecosystem processes. Moreover, Parker considered. In general, parasites de- ALTERATION OF and Karieva (1996) identified pest re- velop better in healthy than weakened sistance as the form of transgene prop- host insects (Visser 1994; Havill and COMPLEX ECOSYSTEM erty most likely to cause such prob- Raffa 2000). So anything that reduces PROCESSES lems. The issue of enhanced weediness herbivore vigor is likely to result in has received much attention, and so parasite death prior to its completing The issue of whether transgenic will not be addressed here. Likewise, development. In this regard, transgenic plants could exert serious environmen- the issue of biodiversity has been dis- plantings can become sinks for para- tal effects hinges greatly on whether cussed in detail elsewhere, and so time site populations (Johnson and Gould there is gene flow into native plants. If does not allow me to add to this. I will 1992). Parasites tend to be have more gene expression is limited to planted focus instead on ecosystem processes. narrow host ranges than predators, but trees and mechanisms for preventing In particular, I will focus on examples few are true specialists. Thus the pos-

143 sibility of secondary pest flare-up re- described several putative mechanisms, vegetative parts are contained, there are mains, as is commonly seen with based on known features of insect no apparent risks beyond those associ- sprayed toxins. physiology, behavior, and population ated with any environmental input. If Emergence of secondary pests is a ecology. I have also used experiences those conditions are not met, however, common and serious problem when with pesticides as a precedent. But not the pesticide analogy is applicable but insecticides or resistant cultivars are all biologists agree that this is an ad- inadequate. Pesticide inputs can be deployed. For example, mite outbreaks equate precedent, and so competing halted at any time, and materials with often follow applications of pyre- models have been proposed for risk excessively long residual periods are throids, and historically followed appli- assessment. I would like to comment banned. This is a valuable aspect of cation of DDT. The most serious out- on some of the most common models pesticides, as compounds that had been breaks of spruce spider mite followed proposed for evaluating pest-resistant approved for many decades commonly aerial application of DDT against plants. lose their certification as new toxico- spruce budworm. Likewise, introduc- Pesticide registration as a model logical problems are discovered. tions of new cultivars have resulted in for pest resistant transgenes has several Transgenic plants are often com- the emergence of previously unimpor- advantages. First it focuses attention on pared to plants bred by traditional tant pests. A new cultivar that was bred the gene product. It can be tested like methods. This analogy has some value for pest resistance sometimes reduces other compounds for acute and in that it recognizes that environmen- the pest species but frees others from chronic toxicity to humans, for effects tal risks of transgenes apply to all mo- competition. In other cases, breeding on beneficial organisms, and alone or nocultures, regardless of how the ge- for unrelated properties inadvertently in combination with phytochemicals. nome was derived. But the comparison enhanced the plant’s susceptibility or Such tests are expensive, but the meth- is sometimes taken further to empha- nutritional suitability to another insect. ods are straightforward, experimental size the added advantage of genes be- Again, natural systems demonstrate conditions can be shielded from inves- ing introduced in a targeted rather than that these are not rare examples. For tigator bias, and the results can be in- random fashion. From an efficacy almost every phytochemical that has terpreted relatively easily. I do not standpoint that is indeed a major ben- been shown to confer resistance against mean to minimize the difficulty of con- efit to genetic engineering. But when most insects, there are also other spe- ducting toxicological studies, and rec- it’s extrapolated to environmental risk cies that use this so-called defensive ognize that there are sometimes pecu- it reverses the product not process prin- compound for nutrition, defense, or liar dose – response relationships, com- ciple—suddenly the process is the ba- communication. The seriousness of plex immunological interactions, and sis for justification. At the scale of pest this problem again depends on whether unpredictable multi-chemical interac- management, the only thing transgenic there is gene flow from transgenic tions. But corporations welcome this and traditionally bred crops have in plants into wild and feral populations. approach because it provides them with common is the process, i.e., deploying a clear target, and they can draw on the toxin through plants. It is also years of experience with agrichemicals, sometimes argued that a single gene RISK A SSESSMENT: pharmaceuticals, industrial reagents, would have less effect than the multiple and cosmetics. Researchers in all sec- rearrangements that take place through MECHANISMS AND tors want to be sure that materials they traditional breeding. However there is PRECEDENTS introduce into the environment are little evidence to support this assump- safe. Whether this approach is adequate tion, and certainly biological bases for As stated above, risk assessment depends largely on whether transgenes being skeptical. First, single genes of- can be both proactive and realistic have the potential to become estab- ten have dramatic effects, as in the when it is founded on verified mecha- lished in wild populations. If sterility cases of many human diseases, pesti- nisms and relevant precedent. I have is incorporated into the genome and cide resistant insects and pathogens,

144 and cultivar-resistant pathogens. Sec- are intended to become self sustaining. tion of the product not process principle; ond, the genes that can be achieved In the case of transgenic trees, this b) extrapolation across multiple levels through traditional breeding are lim- would be an unintended consequence. of biological organization, spatial scales ited to the pool of genes from the Treating GMOs as we do putative and time frames without verifying the plant’s evolutionary history. This is a biocontrol agents does not pose insur- underlying assumption that system strong argument in favor of the efficacy mountable obstacles. Literally hun- processes remain uniform; c) insuffi- and utility of genetic engineering, but dreds of biocontrol agents have been cient attention to general principles not anticipated environmental safety. approved for release. Moreover it is where specific information is not yet Our most severe pest problems have disingenuous for proponents of GMOs available; and d) systematic biases in occurred in noncoadapted systems. to argue that they have been singled research support that favor a risk analy- Third, this analogy does not consider out for unprecedented scrutiny, and at sis process based on precise but sim- either pleiotropic effects, or gene by the same time object to risk assessment plified conditions over complex but environment interactions. Yet we know standards applied to other sectors for more realistic interactions. these can be extremely important in many decades. It might be argued that determining how insect – plant inter- the standards applied to biological con- actions and insect populations behave. trol agents are too lax, as some adverse REFERENCES So resistance breeding is a useful but effects have occurred. However, most inadequate precedent. such problems have arisen either from Abbott, RJ. 1994. Ecological risks of A third regulatory area with appli- organisms that were released privately, transgenic crops. Trends in Ecology cability to transgenic pest resistance is or were released many years ago under and Evolution. 9: 280–282. the planned introduction of beneficial standards less well informed than those Adams, WT. 1992. Gene dispersal organisms. First it is consistent with the applied at present. within forest tree populations. New product not process philosophy. The Forests 6: 217–240. 1987 National Academy of Sciences Brattsten, LB, CW. Holyoke, Jr., JR states introducing genetically modified CONCLUSIONS Leeper, and KF Raffa. 1986. In- organisms “poses no risks different secticide resistance: Challenge to pest management and basic re- from the introduction of unmodified Risk assessment of pest resistant search. Science 231: 1255–1260. organisms.” Note they did not say “less trees is still in its infancy. We can iden- Canadian Forest Service. 1994. Forest than”; they said not “different from.” tify certain risks, some of which would depletions caused by insects and To argue that GMOs should be treated exert environmental harm, others diseases in Canada 1982 – 1987. differently from other introduced or- which would squander valuable Forest Insect and Disease Survey. ganisms would be a radical departure transgenes. The extent to which vari- Canadian Forest Service, Natural from the product not process guideline. ous risk-assessment approaches apply Resources Canada. Second, GMOs and biological control depends on the extent to which gene Fry, JD. 1989. Evolutionary adaptation agents are both selected because of spe- flow can be eliminated. Where steril- to host plants in a laboratory popu- cific desirable properties. Third, risk ity is introduced and vegetative mate- lation of the phytophagous mite assessment of biocontrol agents places rials are contained, procedures for pes- Tetranychus urticae Koch. Oecologia 81: 559–565. a premium on ecological consider- ticide evaluation are most applicable. ations, specifically on how these agents Where either ingredient is missing, the Georghiou, GP and T Saito. 1983. Pest will affect other components of the most applicable standard is the planned Resistance to Pesticides. Plenum Press. NY. ecosystem. This is precisely the area introduction of putatively beneficial where risk assessment of GMOs has organisms. Several obstacles impede Gould, F. 1988. Evolutionary Biology lagged. The biocontrol analogy also has scientifically based risk assessment. and Genetically Engineered Crops. BioScience 38: 26–33. a limitation. Biological control agents These include 1) inconsistent applica-

145 Havill, NP, and KF Raffa. 2000. Com- Mullercohn, J, J Chaufaux, C Buisson, Regal PJ. 1993. The true meaning of pound effects of induced plant re- N Gilois, V Sanchis, and D “exotic species” as a model of ge- sponses on insect herbivores and Lereclus. 1996. Spodoptera littoralis netically engineered organisms. parasitoids: Implications for (Lepidoptera, Noctuidae) resis- Experientia 49: 225–234. tritrophic interactions. Ecological tance to CRYic and cross-resistance Robison, DJ, and KF Raffa. 1994. Entomology 25: 171–179. to other Bacillus thuringiensis crys- Characterization of hybrid poplar Heinrichs, EA and O Mochida. 1984. tal toxins. J. Econ. Entomol. 89: clones for resistance to the forest From secondary to major pest sta- 791–797. tent caterpillar. Forest Science 40: tus: The case of insecticide-induced Paoletti, MG and D Pimentel. 1996. 686–714. rice brown planthopper, Genetic engineering in agriculture Roush, RT. 1987. Ecological genetics Nilaparvata lugens, resurgence. and the environment. Bioscience of insecticide and acaricide resis- Protection Ecology 7: 201–218. 46: 665–673. tance. Ann. Rev. Entomol. 32: 361– Hilbeck, A, M Baumgartner, PM Fried, Parker IM and Kareiva P. 1996. Assess- 380. and F Bigler. 1998. Effects of ing the risks of invasion for geneti- Roush, R. 1997. Managing resistance transgenic Bacillus thuringiensis cally engineered plants: Acceptable to transgenic crops. Advances in corn-fed prey on mortality and de- evidence and reasonable doubt. insect control: The role of velopment time of immature Biol Conserv. 78: 193–203. transgenic crop plants. 271–294. Chrysoperla carnea (Neuroptera, Chrysopidae). Environmental En- Raffa, KF. 1989. Genetic engineering Seidler, RJ, and M Levin. 1994. Poten- tomology 27: 480–487. of trees to enhance resistance to tial ecological and nontarget effects insects: Evaluating the risks of bio- of transgenic plant gene products Johnson, MT and F Gould. 1992. In- type evolution and secondary pest on agriculture, silviculture, and teraction of genetically engineered outbreak. BioScience 39: 524–534. natural ecosystems: general intro- host plant resistance and natural duction. Molecular Ecology 3: 1–3. enemies of Heliothis virescens (Lepi- Raffa, K.F., K.W. Kleiner, D.D. doptera, Noctuidae) in Tobacco. Ellis,and B.H. McCown. 1997. Strauss SH, Rottmann WH, Brunner Environmental Entomology 21: Environmental risk assessment and AM and Sheppard LA. 1995. Ge- 586–597. deployment strategies in geneti- netic engineering of reproductive cally engineered insect-resistant sterility in trees. Molecular Breed- Linder CR and J Schmitt. 1995. Po- Populus. pp. 249, In: NB ing 1: 5–26. tential persistence of escaped Klopfenstein, YW Chun, M-S transgenes: performance in Strauss, SH, SA Knowe and J Jenkins. Kim, and MR Ahiya (eds.) transgenic, oil-modified Brassica 1997. Benefits and risks of Micropropagation, Genetic Engi- seeds and seedlings. Ecological Ap- transgenic trees. Journal of Forestry neering, and Molecular Biology of plications 5: 1056–1068. 95: 12–19. Populus. Kareiva, P, IM Parker and M Pascual. Tabashnik, BE. 1994. Evolution of re- Raybould, AF, and AJ Gray. 1993. Ge- 1996. Can we use experiments and sistance to Bacillus thuringiensis. netically modified crops and hy- models in predicting the invasive- Annual Rev. Entomol. 39: 47–79. ness of genetically engineered or- bridization with wild relatives: a Tiedje, JM, RK Colwell, YL. Crossman, ganisms. Ecology 77: 1670–1675. UK perspective. Journal of Applied Ecology 30: 199–219. RE Hodson, RE Lenski, RN Mack McCown, BH, DE McCabe, DR and PJ Regal. 1989. The planned Russell, DJ Robison, KA Barton, Rebollartellez, EA, N introduction of genetically engi- and KF Raffa. 1991. Stable trans- Gorrochoteguiescalante, M neered organisms: Ecological con- formation of Populus and incorpo- Reynanava and A Solissantamaria. siderations and recommendations. ration of pest resistance by electri- 1994. Effect of Bacillus Ecology 70: 298–315. cal discharge particle acceleration. thuringiensis var israelensis upon the Timmons, AM, YM Charters, SJ Plant Cell Reports 5: 590–594. predatory capacity of Buenoa sp. (Hemiptera: Notonectidae) against Dubbels and MJ Wilkinson. 1995. Meeusen, RL and G Warren. 1989. In- Culex pipiens quinquefasciatus Assessing the risks of wind polli- sect control with genetically engi- (Diptera: Culicidae) larvae. Ento- nation from fields of genetically neered crops. Ann. Rev. Entomol. mology News 105: 295–298. modified Brassica napus ssp. 34: 373–381. oleifera. Euphytica 85: 417–423. 146 Timmons, AM, YM Charters, JW Crawford, D Burn, SE Scott, SJ Dubbels, NJ Wilson, A Robertson, ET Obrien, GR Squire and MJ Wilkinson. 1996. Risks from transgenic crops. Nature 380: 487. van Emden, HF. 1999. Transgenic host plant resistance to insects — Some reservations. Annals of the Entomo- logical Society of America 92: 788– 797. Visser, ME. 1994. The importance of being large — the relationship between size and fitness in females of the parasitoid Aphaereta minuta (Hymenoptera, Braconidae). Jour- nal of Animal Ecology 63: 963–978.

147

ECOLOGICAL SCIENCE

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 151-157. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Ecological Considerations for Potentially Sustainable Plantation Forests

James R. Boyle Heléne Lundkvist C. T. Smith

ABSTRACT

Essential forest ecosystem components and processes must be maintained at all levels of forest management intensity—from ‘wilderness’ and ecological reserves, from which no trees are removed for commercial purposes, to ‘fiber farm’ plantations that are managed as intensively as cropping systems. Minimum requirements for sustainable forest ecosystems include viable tree propagules; system capacity to provide aeration and support for roots; water supply; and nutrient supply for flora and fauna, including functional ecosystem components and processes required for energy and nutrient dynamics; propagules and habitats for mycor- rhiza-forming fungi; and ecosystem resilience, resistance or protection from epidemic or damaging tree pathogens—insects, animals, or microbes. On watershed and landscape scales sustaining viable populations of native organisms is important. For some ecosystems, protection from, or resilience to, wind, fire, or other disturbances may be needed. James R. Boyle is with This framework for sustainable forests must include definitions of system bound- the College of Forestry, aries and interchanges and interactions with adjacent systems, system components and processes, inputs to the system, and outputs. Inherent in such a view would be assump- Oregon State University, tions about availability of energy sources for system dynamics and disturbances, and cli- Corvallis, Oregon. matic setting of each system. And, to be truly inclusive, frameworks for sustainability must include considerations of interactions between forests and human communities within and [email protected] dependent on them. To be relevant to assessing ecosystem sustainability, ecological research must include humans and human systems along with forest ecosystems at all scales. Heléne Lundkvist is with These criteria of sustainable forest ecosystems provide a convenient conceptual frame- the Department of Ecol- work for examining the potential for introduced species and genetically modified organisms (GMOs) to have unanticipated ecological impacts at stand or landscape levels of reso- ogy and Environmental lution. While there is considerable experience with introductions of exotic tree plantations (e.g., radiata pine), inadequate research has been conducted to test long-term hy- Research, The Swedish potheses about the impacts of introducing GMOs into forest ecosystems. However, this University of Agricultural conceptual framework suggests that critical analysis might be focused on such areas as interspecific organism interactions (e.g., plant–plant, plant–animal, plant–insect); pollen Sciences. drift contributing to sexually mediated gene modification and distribution; vegetative or

[email protected] clonal population expansion of GMOs; GMO-mediated effects on ecosystem components and processes that alter energy or nutrient dynamics; and GMO-related changes in eco- C. T. Smith is with the system resilience and resistance. All considerations must be made at stand, forest, watershed, landscape, and regional scales. Department of Forest

Sciences, Texas A&M e humans manage many types of forests for many products and University. values, ranging from firewood and structural materials to recreational and spiritual values. For this discussion we’re considering [email protected] W forests that, in general, are forest tree plantations that are much simpler and ‘more organized’ than are traditionally wildland forests. We focus attention and examples 151 and discussion on temperate and bo- managed tree plantations. Consider- protection. Among considerations of real forests. And, we include emphasis ations of using intensively managed such plantations, at least in some on potentials of transgenic trees as trees, in whatever form, raise, in the places, are the visual quality values of components of these forest plantations. minds of some, concerns about ‘alien’ geometric shapes of plantations, with For many, considerations of forest organisms in our forests. But, of numerous attendant aesthetic and eco- management begin at and focus on the course, we already have many aliens in logical implications. In England and stand level, a level beyond individual our forests, e.g., many weeds, some Scotland landscape architects have trees but less encompassing than an birds, insects, and other organisms. worked with foresters to ‘reshape’ entire “forest”. Most of the exotic species in our planted forests to fit more aesthetically In many regions of the world for- forests today, with some exceptions into landscapes. Plantation edges that ests and tree plantations exist as tessera— (kudzu, gorse, Scotch broom) are not are more amoeba-like than linear may small parts—in landscapes that are mo- so very visually intrusive. Transgenic be not only more aesthetically pleasing, saics of many land uses—human settle- plantations amongst forests might be but will provide more total edge and ments, roads, farmlands, plantations, otherwise. Many forests and forest tree more potentials for ecological interac- and wildland forests. This is the case in plantations are intensively managed, tions with adjacent forests or other the McKenzie River Valley in the Cas- with thinning, competition control and land cover. cades Mountains east of Eugene, Or- other stand tending procedures. Will In contrast to relatively small plan- egon, USA, in Queensland, Australia, intensive tree plantations in the forests tations fitted into landscapes, there are, and in southern Sweden (Figure 1). of today be much different than what of course, vast landscapes of planted Our main focus for this brief dis- we now have? Those are the issues we and intensively managed forests, such cussion is single-species, intensively confront here. as those of Pinus radiata in New Intensively man- Zealand, and Pinus patula in aged plantations al- Swaziland. Some of these planted for- ready exist in many ests dominate landscapes for thousands places in the world, of hectares and provide ecological con- e.g., pine plantations ditions that are quite different from as parts of mosaics of native forests or from previous agricul- deciduous forest land- tural land uses. scapes in the south- And, there are also vast expanses eastern United States, of mostly single-species native/natural parts of landscapes in forests, such as those of Douglas-fir and Sweden, New lodgepole pine that originated after Zealand, Australia, large fires in the Pacific Northwest and and Chile. In Scot- intermountain regions of the United land, New Zealand, States and Canada. Foresters and for- Australia, and else- est scientists have long worked in such where, there are vast forests in Europe, Russia and North areas of exotic planta- America, so we have significant knowl- tions, intensively edge of the ecology of large, single-spe- managed and used for cies forests, some managed and some a variety of purposes. relatively wild. These plantations Ecological considerations of plan- provide wildlife habitations are important at several scales, Figure 1. McKenzie River Valley, Oregon (above), and from stand, to watershed, landscape Queensland, Australia, landscapes. tats and watershed 152 and region to country and, perhaps, ough and careful considerations of this continental expanses. For considering potential. Use of infertile trees, as is ecological implications of plantation being done now in some experimental forests, including those of transgenic and production plantations, is one trees, the ecosystem concept provides method of limiting interactions. In a useful framework of reference. This addition to potential gene flows, the concept has been used to assist in un- issues of adjacency and edges have nu- derstanding much of our current merous implications for movements of knowledge of forest ecology, at many organisms and propagules, plant, ani- scales, from stand to landscape. This mal and microbial, between plantation concept is oft depicted as some sort of trees and forests or agricultural lands. model, whether as in a simple sketch In arid central Oregon, USA, thou- or as some relatively complex math- sands of hectares of hybrid poplars grow ematical set of equations and relation- in a landscape where limited populations ships in a computer-based model. of native cottonwoods grow in moist As Hamish Kimmins has written stream courses, so interchanges of organ- in his Forest Ecology textbook, it’s use- isms and organic matter are likely. In the ful to consider ecosystems, whatever lower Columbia River Basin, northwest the scale and setting, with these ele- of Portland, Oregon, USA, there have ments: boundaries, components (physi- for decades been hybrid poplar planta- Figure 2. Pinus contorta, Cragieburn, New cal and biological), structure, processes tions on former agricultural lands, inter- Zealand. of transfers of energy and matter, in- mixed with numerous other land covers teractions and interdependencies, and, and uses. Island, New Zealand, is in sharp con- dynamics over time. In other settings, such as the tus- trast to the relatively simple structure As considerations of ecosystems re- sock grass covered highlands of the of a 42-year-old Pinus radiata forest quire more details, models, of whatever Clarence River Valley and Cragieburn form, become more complex. Defini- regions of South Island, New Zealand, (Figure 3). tions, evaluations, and estimates of all reforestation/afforestation has been Diseases of trees and forests, fun- model elements add complexity and de- implemented with a variety of exotic gal or otherwise (see Figure 4, mistle- tails, for example, of transfers of nitro- conifer species, including, inter alia, toe on Pinus ponderosa in central Or- gen from soils to tree roots, or move- Pinus contorta (Figure 2). One issue egon), must be considered in potential ments of nitrate ions into stream waters. that New Zealand foresters and other interactions among plantations and ad- In context with a larger forest and land- land managers have confronted is the jacent forests. Insects and other scape, there will be many exchanges of “weediness” of P. contorta, seeds of arthropods also interact in many ways organisms, matter and energy among which have spread beyond intended with trees and forests, and perform es- planted forests and adjacent ecosystems, planted areas and have grown into sential, as well as detrimental functions. including aquatic systems. troublesome trees in places where they For example, early stages of disintegra- For example, where transgenic are not wanted. tion and decomposition of many plant trees grow in close proximity to native At another scale of forest ecosys- residues are caused by arthropods. populations of the same genus or spe- tem consideration, forest structure is Scales for considerations of struc- cies, such as some plantings of hybrid very important, as it creates and influ- ture, diseases and insects, and other poplars, there exists potential for gene ences habitats for numerous organisms, forest ecosystem components and pro- flow to and from adjacent trees. Ge- plant, animal, and microbial. The com- cesses may range from landscape and neticists working with potential plex structure of a native podocarp for- watershed to stand, individual tree transgenic trees have made many thor- est with understory tree ferns on North rooting zone, rhizospheres and mycor-

153 composition processes. However, it is also most likely that these modified plant tissues will fall within the range of existing substrate for decomposer organisms—they are not likely to be ‘alien’ (b) to the systems (Figure 7). Mixed species for- troduced fungus-resistant trees (Figure ests, such as those in Sweden with 5). Of course, such resistance would/ birches and pines, could be models for (a) could likely exist in tissues of residues considering using mixtures of species in forest floors and in roots—which in plantations of transgenic trees. The Figure 3. Forest structure: (a) podocarps and might influence decomposition pro- relative merits of mixed species plan- (b) 42-yr-old Pinus radiata plantation. cesses both qualitatively and quantita- tations have been thoroughly discussed tively, and, also, might in the forestry literature. Such planta- affect mycorrhizae. tions provide diversity in many ecosys- For these de- tem characteristics e.g., niches for epi- tailed, process-level phytes; ground flora, and for fauna of considerations, con- all sizes; microclimate; hydrology; bio- ceptual models can also be useful. Here’s simple ‘model’ of in- Gene "fdr" (fungal disease resistance) introduced to tree. teractions of forest vegetation with soils Increased fungal disease resistance in trees (Figure 6). Of course, the Effects on Secondary effects on mycorrhizae decomposers depicted relationships Figure 4. Mistletoe in Pinus ponderosa, are much more complex and “chaotic” central Oregon, USA. Reduced ability to Faster Slower than represented, and are very difficult sorb water on decomposition decomposition rhizae, and microscopic soil organic to study and define individually. xeric sites matter and clay particles. Reduced Forest floor Soil biota and the processes and Following are examples of some forest floor thickens ecosystem properties that they control "Dry site" species more detailed scales of considerations favored and influence are examples of the ex- Altered micro Slower from our experiences with forest soils. & meso fauna nutrient cycling treme complexity of soil-vegetation sys- Similar examples should be developed tems and of other sub systems of for- Reduced tree for all important forest components Biodiversity = more ests and plantations. Modifications in xerophilous organisms growth rates and processes. For example, Tommerås the chemical or structural composition et al. in discussing potential implica- ?? ?? of wood or other plant tissues will tions of transgenic Picea abies in Nor- C-sepuestration, inter alia… likely affect the decomposer commu- wegian forests, created this general Figure 5. Potential effects on soils of disease- nities and subsequently also the de- scheme of possible implications of in- resistant trees (from Tommerås et al.).

154 geochemistry; soil fauna and microbes. patterns; patterns of open spaces; and, 1. Recognition of scale—stand-water- Considerations of plantations and spe- site properties and processes. shed/landscape-region—is impor- cies homogeneity discussed by Peterken tant; and, scale “down” from stand include physiognomy; composition; to tree to soil particle and root tip structure; age structure; community CONCLUSIONS and microorganism. 2. An “ecosystem concept”, at numer- Our cultures and daily lives in- Forest Productivity– ous levels of detail is useful to con- A View from the Bottom clude dependence on millions of sider dynamics of matter, energy hectares of intensively managed and and organisms, and to, for example manipulated lands with highly sim- Vegetation-Root System help frame and develop working, plified and cultured vegetation, in- testable hypotheses about, e.g., cluding agricultural food crops, Woody Residues & Forest Floor grazing lands, vineyards, fruit and • interspecific organism interac- nut orchards, and forest tree plan- tions (plant-plant, plant-animal, Soil Fauna, Flora, Microbes tations grown for wood and fiber. plant-insect Intensively managed plantation for- Organic Matter & Humus • effects of single-species systems ests and agricultural-like tree farms on ecosystem components and "Mineral Soil" are currently providing wood fiber, Topsoil, Structure, Subsoil processes, which alter energy or and other forest ecosystem ‘services’, Aeration Water Nutrient Rhizosphere nutrient dynamics, e.g., alter- Rooting Storage Supply & e.g., soil protection from erosive Drainage Symbiosis ation of microorganism-arthro- forces in many places, soil building pod-bird food webs with Pinus radiata on sand dunes in Soil/Site/ Forest Productivity New Zealand, and on abandoned, • changes in resilience and resis- Figure 6. Soils and forest productivity. degraded agricultural and grazing tance of ecosystems in relation to lands in central disturbances. Chile. For generations Finally, in all these relatively nar- intensively managed, rowly focused technical and ecological cultured crops and considerations, we must remember the lands have coexisted larger context in which we think about with native vegeta- trees and forests. We assert that our tion ecosystems and overall goal as forest scientists is to con- human habitations, tribute to potential sustainability of as exemplified by the human communities in the face of ex- historic landscapes of panding populations and limits of eco- Sweden and else- systems’ processes, services and resil- where. iences. We’re working to somehow “de- We suggest the sign the human enterprise to fit nature” following for further (as David Orr has written). We have thinking about eco- many current examples of forestry as a logical implications human enterprise fitting rather well of plantations of for- into “nature”, for example, landscapes est trees, including of agricultural, forested and human- occupied lands in Queensland, Austra- Figure 7. Roles of soil fauna (by Dr. Andy Moldenke). transgenic trees:

155 lia, many parts of Sweden, Norway, New Forests 17: 239–262. mass and Bioenergy 14: 403–414. and Finland, and Minnesota, USA. We DiFazio, SP, S Leonardi, S Cheng and Joffre, R., Ågren, G.I., Gillon, D., must remember that our ultimate con- SH Strauss. 1999. Assessing poten- Bosatta, E. 2001. Organic matter cerns are not for today’s genetic or eco- tial risks of transgene escape from quality in ecological studies : system science, but for our grandchil- fiber plantations. BCPC Sympo- theory meets experiment. Oikos sium Proceedings No. 72: Gene Flow 93:451-458. dren and great-grandchildren and the and Agriculture: Relevance for Jukes, MA, J Peace, R Feraris. 2001. next seven generations. Transgenic Crops. British Crop Pro- Carabid beetle communities asso- tection Council. ciated with coniferous plantations Duncan, AJ, SE Hartley, M Thurlow, in Britain: the influence of site, ACKNOWLEDGEMENTS S Young, B Staines. 2001. Clonal ground vegetation and stand variation in monoterpene concen- strucrture. Forest Ecology and Man- We thank Steve DiFazio, Jace trations in Sitka spruce (Picea agement 148:271 – 286. sitchensis) saplings and its effect on Carson, Steve Strauss, Kermit Kavanagh, K, G Stankey and J Boyle. their susceptibility to browsing Cromack, Richard Waring, Steve 1999. The integration of planted damage by red deer (Cervus Garman, and others for stimulating and natural forests in a regional elaphus). Forest Ecology and Man- landscape. New Forests 18: 97–109. discussions and inspiration. agement 148: 259–269. Kirk, DA, K Hobson. 2001. Bird-habi- Dyck, WJ 2000: Nature conservation tat relationships in jack pine bo- in New Zealand plantation for- real forests. Forest Ecology and Man- REFERENCES estry. Chapter 4 in Craig, JL, agement 147: 217–243. Mitchell, N.Saunders, D.A. (eds.). Axelsson, AL, and L Östlund. 2001. Nature Conservation 5: Conserva- Lilienfein, J. et al. 2001. Effects of Pinus Retrospective gap analysis in a tion in Production Environments: caribaea forests on the C, N, P and Swedish boreal forest landscape Managing the Matrix. Surrey S status of Brazilian savanna using historical data. Forest Ecology Beatty & Sons, Chipping Norton, oxisols. Forest Ecology and Manage- and Management 147: 109–122. New South Wales, Australia. Pp. ment 147: 171–182. 35–43. Binkley, CS. 1999. Ecosystem manage- Mullin, TJ, S Bertrand. 1998. Environ- ment and plantation forestry: new Elkie, PC, and RS Rempel. 2001. De- mental release of transgenic trees directions in British Columbia. tecting scales of pattern in boreal in Canada - potential benefits and New Forests 18: 75–88. forest landscapes. Forest Ecology and assessment of biosafety. The Forestry Management 147: 253–261 Chronicle 74: 203–219. Boyle, JR, JK Winjum, K Kavanagh and EC Jensen, eds. 1999. Planted For- Euskirchen, ES, J Chen and R Bi. 2001. Nambiar, EKS. 1999. Pursuit of sus- ests: Contributions to the Quest for Effects of edges on plant commu- tainable plantation forestry. South Sustainable Societies. Kluwer Aca- nities in a managed landscape in African Forestry Journal 184: 45– demic Publishers, Dordrecht, The northern Wisconsin. Forest Ecology 62. and Management 148: 93–108. Netherlands. Peterken, GF. 2001. Ecological effects Brooks, DJ. 1993. U.S. Forests in a Glo- Hodge, SJ, G Patterson and R McIn- of introduced tree species in Brit- bal Context. General Technical Re- tosh. 1997. The approach of the ain. Forest Ecology and Management port. RM-228. USDA Forest Ser- British Forestry Commission to the 141: 31–42. conservation of forest biodiversity. vice, Rocky Mountain Forest and Powers, RF. 2001. Assessing potential Manuscript report, The British Range Experiment Station, Fort sustainable wood yield. Pp. 105– Forestry Commission, Roslin, Collins, Colorado, USA. 128. In: The Forests Handbook. Midlothian, Scotland. Cannell, MGR. 1999. Environmental Julian Evans (Ed.) Vol. 2. Blackwell impacts of forest monocultures: James, RR, SP DiFazio, AM Brunner Science, Ltd., Oxford, U.K. and SH Strauss. 1998. Environ- water use, acidification, wildlife South, DB. 1999. How can we feign mental effects of genetically engi- conservation, and carbon storage. sustainability with an increasing neered woody biomass crops. Bio-

156 population. New Forests 17: 193– 212. Spears, J. 1988. Preserving biological diversity in the tropical forests of the Asian region. In: Biodiversity. E. O. Wilson, editor. National Academy Press, Washington DC. Strauss, SH, SP DiFazio, R Meilan. 2001. Genetically modified poplars in context. The Forestry Chronicle 77: 271–279. Sutton, WRJ (Wink). 1999. The need for planted forests and the example of radiata pine. 1999. New Forests 17: 95–109. Tommerås, BÅ, et al. 1996. Long-term environmental impacts of release of transgenic Norway spruce (Picea abies). NINA-NIKU Project Report 003. 48 pages. Foundation for Nature Research and Cultural Heritage, Trondheim, Norway. Trevors, JT, P Kuikman, B Watson. 1994. Transgenic plants and bio- geochemical cycles. Molecular Ecol- ogy 3: 57–64. Yin, R, LV Pienaar, ME Arnow. 1998. The productivity and profitability of fiber farming. Journal of Forestry 96: 13–18.

157 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 158-167. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf GE Trees: Proceed only with Caution

Faith T. Campbell Rachel Asante-Owusu

ABSTRACT

Environmentalists are concerned about potential negative impacts of genetic engineering of trees. Among our concerns are 1) novel genetic material escaping into wild relatives; 2) impacts of inserted pesticidal properties on forest food webs and ecosystem processes; 3) repercussions of pests’ developing resistance to pesticidal properties; 4) enhanced ‘weediness’ of the transgenic organism or its relatives in natural systems; 5) impacts of altered lignin content on forest food webs and ecosystems; and 6) negative environmental impacts from application of technologies intended to manage the GE organism—including induced sterility and increased use of herbicides. The proponents and regulators bear the responsibility for carrying out empirical research to minimize uncertainty surrounding these concerns. While no one can characterize the risks with precision, their potential magnitude and irreversibility argue for applying the Precautionary Principle. Current regulatory programs lack a true scientific foundation because they rely too much on assumptions and subjective judgements rather than experimental data. These faults are exacerbated by the absence of true peer review and apparent conflicts of interest. GE organisms should be studied intensively by independent researchers before they are approved for commercial use. Government, academia, and industry should devote equal effort to exploring alternative, more environmentally benign approaches to achieving the economic, social, and environmental goals that motivate genetic engineering experiments.

Faith Campbell is a staff enetic modification offers the industrial forest sector potential for developments undreamed of 20 years ago. One of the main advantages is person with the Ameri- Gthe cheaper production of wood fiber. However, the advantages of new technologies are often easily conceived while the costs are difficult to appreciate can Lands Alliance, 726 due to uncertainty, risk, and imperfect knowledge. 7th Street, S.E. Washing- Many environmentalists are concerned about a wide range of potential negative impacts that could arise from genetic engineering of trees. At present, scien- ton, D.C. 20003. tific knowledge does not allow precise definition of these risks. However, the his-

[email protected]. tory of introductions of exotic species demonstrates that novel phenotypes often behave in ways that were not predicted and that serious ecological effects can arise Rachel Asante-Owusu is following very rare events (Royal Society of Canada 2001). While some environmentalists might argue that the difficulty of predicting an independent environ- impacts warrants a prohibition on all genetic engineering of trees, we advance a mental consultant based less Draconian version of the Precautionary Principle. First, forestalling possible environmental damage is far superior to attempting a ‘cleanup’ after harm has in Switzerland. occurred. Furthermore, it is irresponsible to proceed with commercial deployment of GE technology until scientific understanding of both forest ecosystems and [email protected]. the behavior of the genetically engineered organisms in nature is considerably improved. Finally, those proposing use of genetically engineered tree—and regu- 158 lators charged with protecting the en- have sought to dampen these concerns Questions about the vironment and public welfare—bear by putting forward reasoned interpre- Underlying Rationales the responsibility for carrying out the tations of scientific theory. Rarely, how- empirical research needed to minimize ever, can either side ground its position Research should also examine the uncertainties about the nature and ex- in empirical research. The best way to underlying rationales for utilizing ge- tent of the possible risks and to iden- resolve the disputes over the risks is to netic engineering technology in for- tify effective ways to manage those risks conduct research on the points of con- estry. For example, will there be short- that society decides to accept. tention. ages in the future for the types of wood Building a truly scientific founda- products supplied by fast-growing tion for managing the risks associated plantations? New Zealand expects to with genetic engineering will require Possible Risks Requiring double its harvest from plantations by reform of both the regulatory structure Study 2020 (ForestPacRim: NZ Forest Min- and the science community. We need ister addresses Forestry conference: to provide greatly increased funds to Environmentalists want further Date: 10/5/2000 7:02:23 AM Eastern independent scientists studying rel- study of Daylight Time). In Brazil, some plan- evant aspects of forest ecology. Agency • the extent to which inserted genetic tations of improved Eucalyptus already responsibilities must be changed to 3 material might escape from the en- produce 90 to 100 m per hectare per eliminate conflicts of interest. Regula- gineered organism into wild rela- year. A subsidiary of Champion Inter- tors need to insist on peer-reviewed tives and the repercussions of such national expects to begin producing experimental data rather than assump- introgression 2.6 million tons of chips a year by tions and subjective judgements. 2004 on new plantations totaling We call on all players—including • impacts on forest food webs and 100,000 ha. Fast-growing plantations corporations, academic scientists, pub- ecosystem processes resulting from of improved Eucalyptus are already in lic officials, and interested citizens—to inserted pesticidal properties the ground or planned for other South put in the sustained effort needed to American countries (USDA Forest Ser- •repercussions of pests’ developing bring about this restructuring. We be- vice 2001). Recent analysis by World resistance to inserted pesticidal lieve that all will benefit from enhanced Wildlife Fund and the World Bank properties protection of the environment, im- Alliance has shown that 20% of the proved credibility, and a better quality • the risk that transgenic organism or world’s total forest estates is sufficient of life. their relatives might be more to supply current and future genera- “weedy”, especially in natural sys- tions’ industrial roundwood needs if tems proper management systems are put in ENVIRONMENTALISTS’ place [The Forest Industry in the 21st • potential impacts on forest food Century. WWF International March CONCERNS webs and ecosystem processes re- 2001]. sulting from other types of genetic Second, can alternative technolo- Environmentalists worry about manipulation, for example, of lig- gies reduce demand for GE wood fi- possible negative impacts that could nin content ber? Victor and Ausubel (2000) point arise from genetic engineering of trees. out that advances in milling and pro- Our concerns arise in part analogies • possible negative environmental im- cessing technologies could result in glo- with invasive species and a growing pacts from application of various bal demand being stabilized at around skepticism about regulators’ foresight technologies intended to manage 2 billion cubic meters per year by 2050 and willingness to act. Scientists work- the GE organism—including in- (current demand is about 1.6 billion ing with the technology and regulators duced sterility and increased use of herbicides. cubic meters). A number of annual 159 plants, including hemp Cannabis sativa deforestation requires complex strate- mercial crop raises many red flags. This L. and kenaf Hibiscus cannabinus L., gies in which more efficient production is true even if the changes are subtle, have proven qualities as paper crops; of fiber plays only a small role. or take many decades or a century or among other advantages, they lack Finally plantations of fast-growing more to reach a substantial proportion lignins. Various agricultural wastes can trees are not an effective way to miti- of the wild population. be utilized in making paper. An in- gate global warming. If natural forests It appears to us from the literature creased commitment to recycling and are cut down to make room for the that there is a high likelihood that an

demand reduction programs could also plantations, a huge pulse of CO2 is inserted gene will escape. Scientists have an impact. One expert, Rick Meis emitted when the forest is cleared. Sci- have measured substantial gene flow of TreeCycle Inc. (personal communi- entists agree that less than 30% of the from a wide variety of wind-pollinated cation) estimates that utilization of ag- carbon removed from a forest by log- tree species (CEQ/OSTP). Studies of ricultural wastes and demand reduction ging actually ends up stored long term poplars by the TGERC at Oregon State could together cut U.S. demand for in wood products. The other 70% is University, reported by DiFazio et al.

virgin fiber by 20%. Finally, innovative oxidized into CO2 and emitted, partly (1999), found 18% of seeds caught in use of enzymes in wood-fiber process- through the burning of logging slash traps were of hybrid origin. Up to ing (ForestPacRim: NZ Forest Minis- and mill wastes, and partly through the 3.8% of seedlings sprouting in cleared ter addresses Forestry conference: Date: decay of the slash that does not burn. plots had been fathered by male trees 10/5/2000 7:02:23 AM Eastern Day- The time required for a fast-growing in the plantations. Other sources not- light Time) might better solve the en- replacement plantation to recapture the ing a likelihood of genetic leaks include vironmental and energy costs of pro- lost carbon ranges from perhaps 30 a USDA-convened panel of experts on cessing lignin from tree fibers. years, when the original forest was a poplars (Strauss 1999), Raffa (1997), Third, does intensive management young, natural southeastern grove, to and Charest (1995). of plantations—whether composed of over 300 years when it was Pacific There has been considerable dis- genetically engineered or convention- Northwest old growth. cussion as to whether novel genes ally bred trees—actually result in in- would survive in wild populations. creased protection for natural forests? Many scientists believe that to persist, At a global level, forest loss and degra- SPECIFIC ISSUES the novel genes must convey traits that dation are being driven by two failures: give trees a competitive advantage in of the market and policy makers to Escape the wild. Several scientists doubt her- value properly the whole range of for- bicide tolerance or insect resistance will est goods and services; and of authori- The chance that any inserted ge- convey such benefits; according to the ties to curb illegal logging (up to 50% netic material from the engineered or- theory, these genes would probably die of the world’s timber is harvested ille- ganism might escape into wild relatives out quickly. Other types of traits—e.g., gally). Other factors include forest fires and change the genome of native trees height, diameter growth, and disease and conversion to agriculture. In the troubles ‘purist’ environmentalists. The resistance—are thought more likely to United States, it is questionable authors can envision good reasons for convey fitness and persist (CEQ/ whether wood products corporations deliberately altering the genome of a OSTP). DiFazio et al. (1999) found would retain ownership of large land- native tree species, primarily to help that the seedlings bearing genetic ma- holdings from which they were not restore a species endangered by exotic terial from the plantation trees estab- harvesting timber. If they sold the land, pests. Other environmentalists might lished in greater numbers than ex- to what use would buyers put it? Al- not accept even this rationale. How- pected, grew faster, and had higher sur- ternatively, corporations might expand ever, the prospect of altering wild trees’ vival over the winter. The authors con- profitable plantations to control a genomes as an unintended conse- clude that the major factors that would greater share of the market. Curbing quence of the development of a com- limit the spread of human-altered ge-

160 netic material in poplars are the fact webs and ecosystem processes. Accord- scientists rule out impacts from wind- that a very low percentage of poplar ing to Raffa (1989), all tissues of all blown pollen—if it is ingested. As to seedlings can establish in the wild (be- tree species are exploited by a variety soil organisms, the Royal Society cause of the seedlings’ inability to com- of insects in nature. These forest insects (2001) expected Bacillus thuringiensis pete with most vegetation) and the di- are important components of complex (Bt) exuded from roots or plant mate- lution of the novel genetic material by food webs that are poorly understood rial in soil to elicit some response from the sheer quantity of heavily flowering by scientists. These gaps in understand- the rhizozphere and soil microbial wild trees. However, the extent to ing, especially those involving interac- community—but it did not know which progeny containing GE traits tions between trees and herbivorous whether these changes would be sig- might establish and spread needs to be insects, impede predictions of these nificant. In May 2001, the Ecological resolved by long-term empirical stud- effects. Society of America published its new ies before commercial use of GE trees Certainly, spraying of pesticides to position on genetically engineered or- can be approved (CEQ/OSTP). These manage insects in tree plantations has ganisms; one of the Society’s principal experiments must examine each spe- major impacts on non-target insect concerns was the possibility of harm to cific engineered line, not just rely on populations and on the many insect non-target species, such as soil organ- generalizations drawn from study of predators. However, changing to trees isms, non-pest insects, birds and other traditionally bred hybrids (Royal Soci- that exude their own pesticides might animals (http://esa.sdsc.edu/ ety 2001). not be the only, or even best, solution statement0601.htm). Tr uly preventing genetic leaks re- to this problem. quires the reproductive isolation of the Many scientists believe that GE GE plant. Achieving this isolation has trees with the ability to express pesti- Lignin Manipulation been challenging for agronomic crops, cidal properties will cause less damage and we believe it will be more difficult to nontarget organisms than does Lignin is essential to trees and with regard to trees—especially when spraying of insecticides. However, this their structure, although it must be the species being modified is native. All belief does not appear to be supported removed from the cellulose for produc- possible isolation techniques, ranging by empirical research. Furthermore, tion of high quality paper. On the from the practices now used to control Raffa et al. (1997) say that alteration other hand, producers of biofuels wish pollination of hybrid poplars to such of multi-trophic processes is a ‘realis- to increase the lignin content. Little new techniques as genetically induced tic’ risk from use of pesticide-express- information is currently available as to sterility, must be tested—for the full ing GE trees. The results could include how alterations in lignin content might length of the proposed growing pe- reduced populations of predators, para- affect feeding and populations of de- riod—to ensure they meet the stringent sites, scavengers, pollinators, and en- foliators (lignin reduces the ability of standards appropriate for managing dangered or valued species. The Royal herbivores to digest plant material) GE trees (CEQ/OSTP). Achieving re- Society of Canada (2001) found stud- (Barriere and Argillier 1993). Such productive isolation might be more ies to be inconclusive as to the impacts changes could also affect soil structure difficult as regards other types of trees, of genetically engineered pesticidal and fertility as lignin slows down plant such as pines (Charest 1995). properties on natural enemies. One decomposition and degradation by complication focused on by the Royal microbes (Reddy 1984). Society is that some parasites rely on Pesticidal Properties pollen or nectar when they are adults, raising the importance of evaluating Resistance Environmentalists worry that in- impacts of the pesticidal agent in pol- If insects develop resistance to inserted pesticidal properties could have len. There has been little study of pol- serted pesticidal properties, the reper- highly damaging effects on forest food linators other than honeybees. Nor can

161 cussions could be profound (Raffa A second problem is determining enhanced invasiveness has received 1989). Of course, insects develop re- how large such refuges should be and considerable attention. However, the sistance to pesticides that are sprayed where they should be located relative studies to date have been quite re- as well as—presumably—those exuded to the Bt crop (e.g., embedded within stricted. We suggest that a variety of from plants. As far as we can deter- the planting, or immediately alongside, factors warrant reconsideration of the mine, scientific data are currently in- or at some distance away). The EPA ‘weediness’ issue. adequate to support an evaluation of and USDA Animal and Plant Health First, most scientists think that whether resistance is likely to occur Inspection Service (APHIS) are cur- introducing herbicide tolerance per se more or less rapidly in response to ge- rently revising upward their standards is not likely to exacerbate plants’ inva- netically expressed pesticides or simi- for corn and cotton (see FIFRA SAP sive potential. However, empirical re- lar chemicals in sprays. The Royal So- 2001)—demonstrating that the agen- search on each ‘line’ of GE trees is ciety of Canada (2001) called for im- cies’ earlier rules were based more on needed to test this theory (Royal Soci- mediate establishment of meaningful assumptions than on empirical data. ety 2001, CEQ/OSTP 2001). The re- resistance monitoring guidelines. Third is the difficulty of ensuring cently detected spread of herbicide-re- In general, the more widely a that these refuges are maintained by the sistant canola also raises questions that toxin is used, the greater the possibil- forester in the face of countervailing deserve exploration (Royal Society ity that insects will develop resistance. pressures to maximize return. At 2001). present, the EPA is not certain that Bt is already in widespread use across Second, nearly all the studies of corn farmers actually maintain the North America—as sprays and in GE weediness to date have been done in pesticide-free refuges mandated by ex- crops. The U.S. Environmental Protec- managed ecosystems. Regulators also isting regulations governing Bt corn tion Agency (EPA) is responsible for rely on the weed science literature— (EPA/USDA Workshop 1999). Until balancing use of Bt in various forms which rarely discusses whole categories the EPA demonstrates an increased ca- and targeting various insect/crop com- of plants demonstrated to be invasive pacity to enforce its own rules, we binations to minimize development of in natural systems. Among these over- think it is foolish for the agency to add resistance. Might the agency decide at looked natural area weeds are many the further complication that would be some point to impose a ceiling on uses forestry, horticultural, and pasture posed by Bt trees. of Bt in order to slow development of plants. [For lists of plants considered Raffa et al. (1997) found that resistance? invasive in natural systems in the some tree/insect relationships created A favored strategy to slow or mini- United States, contact the lead author situations in which it was too likely mize insects’ developing resistance is cre- that insects would evolve resistance in or consult web sites maintained by the ation of refuges—breeding sites for in- response to a tree that exudes its own Plant Conservation Alliance sects that have not been exposed to the pesticide. These “too risky” groups in- (http://www.nps.gov/plants/alien) specific chemical and thus have not faced cluded bark beetles and wood borers. or The Nature Conservancy evolutionary pressures to develop resis- Have other scientists studied this ques- (http://tncweeds.ucdavis.edu)]. tance. However, planting GE trees ex- tion and highlighted other relation- Genetic changes that increase their re- pressing Bt close to non-Bt trees in ref- ships that might deserve special care? sistance to herbicides now used to con- uges would reduce the reproductive iso- Are the financial backers and scientists trol them would exacerbate problems lation of these separate genotypes and working on GE manipulation of trees arising from their invasiveness. force greater dependence on other ap- paying heed to such warnings? Furthermore, managers of natural proaches to minimize genetic leaks. If systems face pressures that circumscribe the refuges were planted with poplars their options for controlling invasive engineered to tolerate herbicides, that Weediness plants. The public often demands that trait would further complicate manage- The question of whether GE they avoid using certain chemicals or ment to prevent persistence and spread. plants—or their relatives—will exhibit biological control. They must mini-

162 mize damage to non-target vegetation. induced sterility and increased use of answers to the question of whether Finally, they cannot pass on the costs herbicides—might have their own genetically engineered organisms differ of control. Plants that pose minimal negative environmental impacts. in substantial ways from their tradi- problem in managed systems can be- Regarding herbicides, the focus tionally bred counterparts. come difficult-to-manage, damaging has tended to be on ‘herbicide ready’ In presenting the Ecological Soci- invaders in natural systems. trees. If regulators and concerned citi- ety of America’s May 2001 statement Third, many scientists believe that zens are to understand the relative risk on genetic engineering, Diana Wall, a more invasive or weedy plant might posed by use of herbicides in planta- Director of the Natural Resource Ecol- result if an introduced gene conveys tions of ‘herbicide ready’ trees, we need ogy Laboratory at Colorado State Uni- some increased “fitness”—that is, make better information on the extent to versity, said “It’s important to recognize the plant more resistant to one or more which herbicides are now used in plan- that some GMOs can possess genu- environmental stresses (Mellon 1993; tation forestry and dependable projec- inely new characteristics that may re- Louda 1999; Marvier and Kareiva tions of likely practices using the GE quire much greater scrutiny in terms 1999; Royal Society 2001; all wrote organisms. Society then needs to con- of scientific research than organisms about herbaceous plants)—or enables sider whether there are better ways to produced by traditional techniques of it to grow faster (CEQ/OSTP 2001). solve the weed problem in plantations. plant and animal breeding. In particu- Again, proponents and regulators must We note, however, that herbicides lar, we really need more peer-reviewed take greater care to ensure that they are are expected to be important tools for research on the potential environmen- not increasing the fitness of any plant preventing escape of trees containing a tal effects of GMOs.” The ESA’s con- that is already invasive in natural sys- variety of engineered traits. As should cern focused on organisms that can tems. be clear from our earlier statements, we persist without human intervention Finally, genetically engineered concur that it is vitally important to and the exchange of genetic material trees pose special challenges because prevent genetic leaks. However, the between GMOs and unaltered organ- plants that have undergone a shorter environmental costs of the manage- isms within the environment. Some period of breeding by people are more ment technique need to be evaluated GMOs may also be given traits which likely to retain characteristics that in determining whether to move for- would provide an advantage over na- would allow them to be ‘weedy’ than ward with this technology. tive species in some environments are such crops as maize (Mellon 1993; No information is currently avail- (http://esa.sdsc.edu/statement0601.htm). Royal Society 2001). (Again, horticul- able as to whether trees with an altered The Royal Society (2001) empha- tural plants and range or pasture grasses lignin profile will react differently to sized that the question of whether GE have similar short breeding histories). pest attack and other stresses. Studies techniques are no different in charac- Evaluating whether a genetically with tobacco genetically engineered for ter or consequences than traditional engineered plant might become inva- reduced lignin content found the techniques must be decided by empiri- sive will not be an easy task. Studies plants to have weaker resistance to vi- cal investigation. The Canadian panel have shown that weediness can mani- ral attack (Maury et al. 1999). stressed that the assumption that shuf- fest as late as 150 years after introduc- fling a crop plant’s genome—as hap- tion (Marvier and Kareiva 1999). pens in traditional breeding—is not ARE GE PLANTS likely to create harmful progeny is no longer valid with GE plants, when the Environmental Costs of ‘DIFFERENT’? novel genes are derived almost exclu- Managing GE Trees sively from non-plant sources. The To a great extent, conflict over the risks are particularly great when deal- Techniques adopted to manage extent of risks posed by genetically en- ing with plants that have not under- genetically engineered trees—including gineered trees stems from opposing gone millennia of selection to enhance

163 desirable traits and expunge undesir- potential risks from pleiotropic effects. (National Plant Board 1999; Campbell able properties. These plants are more According to the panel, when a plant 2000), these same programs must also likely to retain a capacity to compete is genetically engineered, the result is be re-evaluated as they are applied to successfully outside of a managed not the precise placement of a new genetically engineered organisms. Both agroecosystem. piece of genetic code into a carefully types of organisms can behave in un- The committee that prepared the selected section of the new host’s ge- expected ways and cause serious eco- study by the U.S. Council on Environ- nome. Rather, each insertion occurs at logical impacts; and these impacts will mental Quality and Office of Science a nearly random location—resulting in be hard to predict from data collected and Technology Policy (2001) was ap- potential differences in the way the in conventional ecological experiments parently divided on the question. Three gene functions. Furthermore, the re- conducted at restricted spatial and tem- of four statements in the report lean mainder of the host’s genome is also poral scales (Royal Society 2001). toward recognizing substantial differ- affected. Insertion of a single gene will ences between conventionally bred and be accompanied by a range of changes genetically engineered organisms that that will, in turn, be affected by the TYPES OF RESEARCH at least warrant careful study. The first genome of the host, the host plant’s statement notes that, while cross-breed- developmental and physiological status, NEEDED AS A BASIS ing of organisms has occurred for more and environmental pressures. Conse- FOR ANALYZING RISK than 10,000 years, rDNA technology quently, regulators cannot limit their “vastly expands the potential to intro- evaluation of a transgenic variety’s po- Among others, the Royal Society duce new genetic material, [requiring] tential impacts to those that might arise of Canada (2001) and Raffa et al. scientists, regulators, and the public to from the predicted phenotypic charac- (1997) have outlined research pro- rethink the adequacy of these existing teristics conferred by the transgene grams that would better support analy- oversight mechanisms.” Later, the re- chosen for insertion. Instead, officials sis of the potential risks associated with port says that the fact that a greater must empirically assess each genetic genetic engineering. They agree on the variety of genetic constructs can now line for the potential consequences of need for broad ecological research con- be incorporated more quickly into these pleiotropic effects. Again, the risk ducted by independent scientists. The organisms with different genetic back- of unanticipated and unwanted Royal Society specifically stated that grounds requires regulatory agencies to changes is greater in plant types with independent ecological research that develop specific regulations and guid- a short history of domestication. looks at the potential impacts of GE ance on their use. Finally, when dis- We find interesting the analogies organisms is more likely to provide cussing poplars specifically, the report some—including the U.S. National novel insights than is research con- states, “changes possible by genetic en- Academy of Sciences—draw between strained by the regulatory framework. gineering can be different in kind and genetically engineered organisms and The Royal Society panel stressed degree. . . .” On the other hand, the invasive exotic species. Over the past that studies must be of sufficient du- study cites the National Academy of decade, scientists and the U.S. govern- ration to incorporate the complete life- Sciences as twice finding that there is ment have increasingly recognized that time of the GE organism. It recom- no evidence of unique hazards in us- efforts to protect the environment from mended a staged approach, involving ing rDNA techniques or in moving damage by exotic species are inad- a series of experimental comparisons genes between unrelated organisms. equate. APHIS regulates genetically with conventional varieties, conducted The panel of the Royal Society of engineered organisms under the same under ever more realistic conditions. Canada (2001) justifies its insistence legal authority as it manages exotic The goal would be to determine on empirical testing of the similarities “plant pests”. To the extent that whether the GE crop differs from its and differences between traditionally agency’s programs have been found to conventional counterpart in any life- bred and engineered organisms on the be deficient as regards exotic species history attribute likely to have fitness

164 implications for survival in the wild. entists, economists, and people con- In particular, we endorse the panel’s call Among the specific studies called for cerned about social implications, for relying on empirical data, establish- were evaluation of empirical data on worker health, economic viability ing a mechanism to obtain input from the fitness costs and benefits of inserted should also be consulted. independent scientists throughout the GE traits in non-crop species taken As stated at the onset, we believe regulatory process, and increasing pub- from diverse ecological contexts; ex- corporations and research consortia lic access to the scientific data and as- periments to determine the likelihood involved in developing GE trees should sumptions used in reaching decisions of pollen-mediated gene flow to related be major contributors to effort to im- (Royal Society 2001). Finally, we urge species; and rigorous experimentation prove scientific knowledge for manage- expanding the agencies’ regulatory au- testing the impact of GM plants on ment of these organisms. This is prob- thority to ensure that all types of GE both target and non-target insect spe- ably best done by establishing a sepa- organisms are subject to scrutiny and cies. Because of pleiotropic effects, rate, independent foundation through regulation. these studies cannot be limited to which could fund independent scien- evaluation of the predicted phenotypic tists’ work. The current program man- characteristics; they must also assess aged by the USDA Cooperative State THE PRECAUTIONARY empirically the unanticipated changes’ Research, Education, and Extension consequences across time and environ- Service (CSREES) is not adequate. APPROACH ments. Funding is too low—just $1.3–$1.5 The Royal Society panel’s report million per year. Furthermore, the re- There is the potential that at least contained many warnings about biases search is tied too closely to both regu- some of the plausible damaging im- in results that might arise from small latory agencies and immediate regula- pacts of genetic engineering might be scale, short-term studies. The poplar tory concerns to provide the longer- irreversible, irremediable, and of cata- experts convened by the USDA term perspective and interdisciplinary strophic proportions. All parties are (Strauss 1999) also said that field tri- approach needed. It is essential that currently operating without sufficient als of trees are less likely to provide re- whatever mechanism is established be scientific data to characterize these risks liable data because costs dictate that the seen as independent. (Figure 1). Regulators should not as- tests will be too small and too brief to Of course, regulatory agencies sume that the present “absence of evi- allow assessment of both commercial continue to need scientific assistance in dence” equals an “evidence of absence” viability and ecological impacts. conducting their work. We concur of risk. The Canadian panel urged Raffa et al. (1997) suggested rank- with the Royal ing risks according to whether they Society that the Entry Point would be localized within the planta- agencies need to

Is there a level of scientific tion or might affect adjacent ecosys- be more transpar- certainty and consensus based on independently, Yes tems; whether the environmental harm ent, to involve a Implement trials, submit peer-reviewed empirical data Quantify risk via assessment results to independent peer that can provide a reliable methods agreed by review and report conclusion basis to assess (and quantily) scientists, government would be self-perpetuating; and wider variety of to government and the risk of adverse impacts & civil society other stakeholders on the environment, human whether potential mitigating tactics are expertise, and to health and productivity? available. They called for proactive re- reduce the appar- No Can the risk be managed to Yes safeguard the environment search on the likelihood of various en- ent conflicts of Invoke the and human health? precautionary principle Put in place policies, vironmental hazards (identified in the interest. Cana- guidelines and procedures No that effectively address, monitor article) as well as methods for offset- dian panel pro- and mitigate risk to environment & human health Do not commercially develp this technology ting them. They recommended using posed many spe- Moratorium on commercial application and agree an interdisciplinary approach through- cific actions that gridelines for comprehensive flield trials out the discovery and development we believe could process. We would add that social sci- be applied here. Figure 1. The precautionary principle.

165 regulators to delay commercial use of from GE organisms. Further, scientists Campbell, FT. 2000. Curbing Invasions any GE organism until the specific ge- have undertaken specific challenges, by Exotic Species: a “report card” evaluating the USDA Animal and netically engineered line had been thor- including efforts to develop sterile or- Plant Health Inspection Service. oughly studied at six relevant levels: ganisms, in order to avert possible risks. American Lands Alliance March genome, transcript, protein, metabo- However, these programs are not as 2000. lite, health impacts, and environmen- effective as they need to be—largely Charest, PJ. 1995. Genetic Engineer- tal impacts (Royal Society 2001). because of how regulators and propo- ing of Tree Species: The Canadian We endorse the Royal Society’s nents of the technology respond to Experience. Proceedings and Papers (2001) call to invoke the Precaution- data gaps and uncertainties. from the 1995 Risk Assessment Re- ary Principle in those cases in which We therefore ask that scientists, search Symposium, June 6–8, 1995. there are some scientific data (although institutions, corporations, and regula- [CEQ/OSTP] U.S. Council on Envi- incomplete, contested, or preliminary) tors adopt a broader conception of the ronmental Quality and Office of or plausible scientific hypotheses or Precautionary Principle—the charac- Science and Technology Policy. models (even though contested) that terization discussed in the report. The 2001. Assessment of Environmen- tal Regulation of Genetically En- establish a reasonable prima facie case Precautionary Principle is not a license gineered Organisms. for the possibility of serious harm, and to block a decision; as illustrated here there is significant uncertainty. and discussed in the report by the DiFazio SP. , S Leonardi, S Cheng, SH Strauss. 1999. Assessing potential Royal Society of Canada, it is a risk- risks of transgene escape from fi- management strategy for use when the ber plantations (can be obtained CONCLUSIONS paucity of information precludes quan- from the TGERC web site at http:/ tifying the risks (Asante-Owusu 2001). /www.fsl.orst.edu/tgerc/ The United States lacks a coher- It shifts the burden to the proponents, genfor2.htm) ent, independent regulatory program in ways which we hope will result in FIFRA Scientific Advisory Panel Meet- for GE organisms that is based on the greatly increased funding for indepen- ing, October 18 – 20, 2000, findings of empirical study. Neither the dent research into the potential risks Marriott Crystal City Hotel. Sets research establishment nor the regula- and possible management approaches. of Scientific Issues Being Consid- ered by the Environment Protec- tory agencies has an adequate process We close by reiterating our belief tion Agency Regarding: Bt Plant- for obtaining the data needed for im- that there are probably alternative, less Pesticide Risk and Benefit Assess- proving our understanding of the pro- risky, ways to ‘solve’ some of the prob- ments. SAP Report No. 2000–07 cesses, risks, and management options lems GE trees are intended to address. March 12, 2001 for GE organisms. Controversy about We urge more funding for research Louda, SM. Insect Limitation of Weedy use of this technology will only grow into them, as well. Plants and Its Ecological Implica- as long as these fundamental flaws not tions. In Traynor, P.L, Westwood, addressed. J.H. Proceedings of a Workshop on: To correct these gaps and restore REFERENCES CITED Ecological Effects of Pest Resistance Genes in Managed Ecosystems. Janu- confidence in the U.S. environmental ary 31 – February 3, 1999. protection regulatory structures, we Asante-Owusu, R. 2001. World Wild- recommend adoption of the Precau- life Fund UK and World Wildlife Marvier, M. and P Kareiva. Extrapolat- Fund International May 2001. ing from Field Experiments that tionary Principle for management of www.panda.org/forests4life. Remove Herbivores to Population- genetically engineered organisms. We Level Effects of Herbivore Resis- recognize that the regulatory program Barriere, Y. and O Argillier. 1993. tance Transgenes. In Traynor, P.L, Brown-midrib genes of maize - a is already “precautionary” in the sense Westwood, J.H. Proceedings of a review. Agronomie, Vol. 13, No. 10, Workshop on: Ecological Effects of that it is intended to prevent or avoid pp. 865–876. the potential dangers that might arise Pest Resistance Genes in Managed

166 Ecosystems. January 31 – February Biotechnology prepared by The 3, 1999. Royal Society of Canada at the re- quest of Health Canada, Canadian Maury, S., L Hoffman, G Pincon P Food Inspection Agency and En- Geoffroy, R Bronner, C Lapierre, vironment Canada. January 2001. B Fritig, and M Legrand. 1999. Www. rsc.ca/foodbiotechnology/ Transgenic tobaccos with depressed indexEn.html] OMTs have reduced lignin content, modified vascular organiza- Strauss, S. Report of the Poplar Work- tion and weakened anti-viral resis- ing Group. In Traynor, PL, tance. Seminar 36 Biotechnology Westwood, JH Proceedings of a and Forestry Conference, Oxford, Workshop on: Ecological Effects of United Kingdom. Pest Resistance Genes in Managed Ecosystems. January 31 – February Mellon, M. Perils Amidst the Promise: 3, 1999. Ecological Risks of Transgenic Crops in a Global Market. 1993 Union United States Department of Agricul- of Concerned Scientists. ture. Forest Service. Forest Prod- ucts Laboratory. 2001. Pest Risk As- National Plant Board. 1999. Safeguard- sessment of the Importation into the ing American Plant Resources: A United States of Unprocessed Euca- Stakeholder Review of the APHIS- lyptus Logs and Chips from South PPQ Safeguarding System. July 1, America. General Technical Report 1999. FP.-GTR-124. April 2001. Raffa, KF. 1989. Genetic Engineering Victor DG and JH Ausubel. 2000 Re- of Trees to Enhance Resistance to storing the Forests. Foreign Affairs Insects. Bioscience Vol. 39, No. 8. 79:127–144. Raffa, KF, KW Kleiner, DD Ellis, and (http://phe.rockefeller.edu/ BH McCown. 1997. Environmen- restorinforests/). tal Risk Assessment and Deploy- World Wildlife Fund International. ment Strategies for Genetically 2001. The Forest Industry in the Engineered Insect-Resistant 21st Century. WWF International Populus. In Klopfenstein, Woo March 2001. www.panda.org/ Chun, Kim, Ajuja. forests4life. Micropropagation, Genetic Engi- neering, and Molecular Biology of Populus. 1997. Reddy, CA. 1984. Physiology and bio- chemistry of lignin degradation, In: Current Perspectives in Microbial Ecology - Proceedings of the third international symposium on microbial ecology, eds. Klug MJ and CA Reddy. American Society for Mi- crobiology, Washington, D.C. 1984, pp. 558–571. Royal Society of Canada. 2001. Ele- ments of Precaution: Recommenda- tions for the Regulation of Food Bio- technology in Canada. An Expert Panel Report on the Future of Food

167 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 168-175. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Biodiversity Implications of Transgenic Plantations

John P. Hayes

ABSTRACT

Implications of transgenic plantations on biodiversity vary dramatically depending on context, such as whether the transgenic plantation is replacing an existing tree plantation, a native forest, an agricultural field, or native non-forested vegetation. When transgenic plantations are established in areas not previously occupied by a tree plantation, implications of ‘plantation effects’ often will be more profound than influences specifically related to the transgenic nature of the trees. Potential ‘special transgenic effects’ also vary substan- John P. Hayes is an tially depending on the specific transgenic trait under consideration, and most traits have associate professor and potential positive and negative implications to biodiversity in an area. Direct off-site effects of transgenic plantations include influences on biotic interactions and landscape struc- wildlife ecologist with ture, and potential escape of transgenic organisms or genes to non-transgenic populations. the Department of Indirect off-site effects are primarily related to potential land-use changes stemming from use of transgenic plantations. Although likely indirect off-site effects are generally hypoth- Forest Science, College esized to be positive (decreased pressure for wood fiber production on native forest lands resulting in increased area in reserves), I argue that there are a number of equally likely of Forestry, Oregon scenarios that suggest indirect off-site effects would have negative implications for conser- State University, vation of biodiversity. I discuss a number of lessons from traditional forest conservation issues that may be applied to the debate surrounding transgenic plantations and propose Corvallis, Oregon five hypotheses concerning the implications of transgenic plantations to biodiversity. 97331. He also serves as the program director se of genetically modified organisms is controversial. Transgenic trees present a complex set of potential societal advantages, ethical issues, for the Cooperative Uand hypothesized environmental risks (Mullin and Bertrand 1998; Forest Ecosystem Re- Matthews and Campbell 2000). Some of the controversy surrounding transgenic plantations concern potential implications to conservation of biodiversity. In this search (CFER) program, paper, I present an overview of potential biodiversity implications of transgenic a cooperative program plantations. I discuss the likely effects of use of transgenics on biodiversity at different spatial scales, present some pertinent lessons from more traditional forest between Oregon State conservation issues, and propose five general hypotheses concerning the influences of transgenic plantations on biodiversity. Some of the conservation issues related University, the USGS to transgenic plantations are unique to transgenics, others are relevant to inten- Forest and Rangeland sively managed plantations in general. The potential conservation implications of use of transgenic trees extend beyond transgenic plantations. For example, Ecosystem Science transgenic disease resistence may facilitate restoration of rare species, such as the Center, the US Bureau American chestnut (Castanea dentata). However, this paper focuses solely on transgenic plantations. of Land Management, and the Oregon Depart- SPATIAL SCALE AND BIODIVERSITY ment of Forestry. Ecological processes and response of organisms to habitat and environmental [email protected] condition differ with spatial scale. Similarly, the implications of transgenic plan- 168 tations to biodiversity also differ with plantation. Biodiversity implications of shrews, spiders, and other animals that spatial scale. For the purposes of this conversion of an existing plantation to rely on insects for forage could be af- discussion, I discuss potential effects of a transgenic plantation of the same fected by changes in the forage base. transgenic plantations at two spatial species are restricted to ‘special However, while use of insect-resistant scales, local and off-site. Local effects transgenic effects’, whereas conversion trees would certainly have implications are here defined as influences of of some other type of vegetative com- to the biodiversity of the stand, the net transgenic plantations on biodiversity munity to a transgenic plantation in- impact of use of transgenics in this case that act at the scale of the forest stand clude influences related to special is not necessarily negative. Use of in- or individual tree. Off-site effects are transgenic effects and ‘plantation ef- sect-resistant trees would likely result influences realized at the scale of the fects’. in reduced use of insecticides in the landscape or greater. Although pre- plantation. Reduced insecticide use sented here as a dichotomy, many of could result in decreased impacts on the potential implications actually span Special transgenic effects non-target species, such as those feed- multiple spatial scales and it is not al- ing on ground vegetation. ways possible to clearly distinguish be- Special transgenic effects are the Actual or functional sterility is of- tween local and off-site effects. unique influences of transgenic organ- ten considered to be a desirable trait isms that manifest because of conse- in transgenic plants, as this would quences related to expression of the eliminate or greatly reduce potential of Local Effects transgenic trait. Special transgenic ef- flow of transgenic genes (via pollen) or fects vary greatly depending on the propagules (via seed) outside of the Establishment of a transgenic specific transgenic trait. Most plantation (Strauss et al. 1995) (al- plantation can alter ecological charac- transgenic traits can have positive or though sterility may not eliminate teristics of an area and potential man- negative influences on local biodiversity movement of transgenic material in agement practices of a site. Vegetative (Table 1). species with active vegetative reproduc- composition and structure, use of in- For example, transgenic plantation). Reduced flow of transgenes or secticides or other chemicals, and sev- tions of trees with insect resistance propagules presumably has positive eral characteristics of the trees them- would certainly impact abundance, implications for biodiversity. However, selves, including resistance to insects, species richness, and diversity of insects in cases where other species depend on fungi, or bacteria, lignin content of the in transgenic plantations. Indeed, re- the pollen, nectar, seeds, or fruit of wood, and flowering and fruiting can duction of certain populations of in- plantation trees, sterility could result in all be altered through establishment of sects is the very outcome desired by use cascading ecological effects, whereby a transgenic plantation (Matthews and of transgenic, insect-resistant trees. the species dependent on the flowers Campbell 2000). At the local scale, the Reductions in insect populations could or fruit would be negatively impacted. implications of establishing a have cascading ecological effects Clearly these effects are dependent on transgenic plantation to the through the food web. Birds, bats, the tree species of interest and the com- biodiversity of an area vary substantially with the context in which the Table 1. Some potential influences of special transgenic effects on biodiversity at the local plantation occurs. For example, the scale. biodiversity implications of converting Transgenic trait Positive effect Negative effect an existing non-transgenic hybrid poplar plantation to a plantation of Insect tolerance Reduced use of insecticides Impacts on trophic relationships transgenic poplars is very different Fungal resistance Reduced use of fungicides Impacts on trophic relationships Reduced lignin content ? Impacts on decay organisms from the implications of converting a Plant sterility Reduced gene transfer Impacts on trophic relationships native forest to a transgenic poplar

169 munity of organisms present in the diversity of species. Similarly, conver- Movement of propagules plantation and surrounding area. sion of other native vegetative commu- and transfer of genetic Likely biodiversity effects of some nities, such as native grasslands or sa- material transgenic modifications, such as alter- vannah, to transgenic plantations gen- ation of lignin content of trees, are erally would result in negative impli- There has been considerable atten- unclear. There are potential implica- cations to biodiversity. Frequently, ar- tion to issues related to escape of tions of altered lignin content for de- eas that could be converted to transgenic plants to other environ- composers or for species that forage on transgenic plantations do not currently ments and introgression of transgenic organisms that live in the wood, but I support native vegetation. Local con- genes to non-transgenic plants. Al- suspect that these implications are sequences of conversion of these lands, though often considered together, these likely to be relatively small. such as agricultural fields, to two issues actually have very different biodiversity in transgenic plantations potential implications to biodiversity could have negative, neutral, or posi- and the contexts in which they are Plantation effects tive effects, depending on the charac- important also differ. teristics and management of the Transfer of transgenes can only Plantation effects are the conse- transgenic plantation and the agricul- result in significant biodiversity impli- quences of the vegetative species com- tural field. cations where there are native or intro- position and stand structure in a tree duced non-transgenic species that are plantation. As organisms generally re- genetically similar enough to the spond strongly to vegetative composi- Direct Off-site Effects transgenic species to allow introgres- tion, stand structure, and aspects of the sion and that are geographically close physical environment mediated by veg- The influences of transgenic plan- enough to allow the transfer of genetic etative communities, plantation effects tations often may extend beyond the material. Probability of spread is influ- can have a profound influence on the immediate plantation and influence enced by the phenotypic expression biodiversity of an area. Indeed, when biodiversity in surrounding areas. Di- and selective advantage imparted by transgenic plantations are planted in rect off-site effects are the influences of the trait (James et al. 1998). Introgres- sites not previously occupied by a plan- the presence of transgenic plantations sion of transgenes presents reduced risk tation of the same species, implications on biodiversity of areas outside of the when transgenic plantations are com- of plantation effects are likely to greatly plantation. Three potentially important posed of exotic species with no geneti- outweigh implications of special implications are: introduction of cally similar species nearby. In addition, transgenic effects at the local scale. Just propagules and transfer of genetic ma- use of sterile transgenic plants would as the influences of special transgenic terial to other sites, influences on bi- effects vary among transgenic traits, otic interactions, and influences on plantation effects vary with the type of landscape structure. Forested Non-forested vegetative community present prior to Initial establishment of the transgenic planta- conditions Existing Native Agricultural Native plantation forest field vegetation tion (Figure 1). In general, biodiversity implications will be negative when native forests are converted to plantations. Relative to native forests, transgenic plantations are generally structurally Potential Special Plantation effects effects on transgenic + simple, have minimal vegetative species biodiversity effects Special transgenic effects diversity, and lack many of the specialized habitats that are relied upon by a Figure 1. The influences of transgenic plantations on biodiversity at the local scale depends on context. 170 also eliminate or greatly reduce poten- pecially trophic relationships, at the sequence of this, a larger proportion of tial for transgenic introgression. When local scale, transgenic plantations could forested lands would be available for introgression of transgenes from a also influence biotic interactions at biodiversity reserves. However, in- transgenic plantation is possible, it larger spatial scales. Influences are po- creased area in biodiversity reserves is could have implications to biodiversity tentially greatest for trophic relation- only one potential outcome of wide- if one of two conditions are met. First, ships of wide-ranging species, such as spread use of transgenic plantations, if the special transgenic effects of the birds and bats. Locally induced changes and I contend that it is not necessarily trait pose significant implications to in species composition within a plan- the most likely outcome. I argue that biodiversity, introgression could be an tation could influence competitive re- the assumption that decreased pressure important issue. Conversely, if the spe- lationships or food webs at larger spa- on native forests for production of cial transgenic effects of the transgene tial scales. wood fiber equates to positive benefits do not have significant implications to for biodiversity is naive. Just as imple- biodiversity, then its introgression is of mentation of intensive agriculture and lesser importance. Second, even if the Influences on landscape increased productivity of agricultural transgenic trait has no evident direct structure lands did not result in large areas de- implications to biodiversity, introgres- voted to conservation of native grass- sion could be important if it influences Intensively managed plantations lands in the United States, increased genetic diversity within the species in can influence the composition and productivity of tree plantations would some important way. configurations of landscapes. not necessarily result in increased area Movements of propagules of Transgenic plantations can play a role allocated for conservation of native for- transgenics out of transgenic planta- in increasing or decreasing fragmenta- ests. Although establishment of tions can also be reduced through use tion and connectivity in an area. The biodiversity reserves for some of the of sterile transgenic plants, but even influence of a transgenic plantation on lands under public ownership is con- when using sterile plants escape of landscape structure, and the resulting ceivable under this scenario (although some transgenics is possible for species implications to biodiversity, are highly even this is highly dependent on the with vegetative reproduction. It is pos- context dependent. social, political, and economic environ- sible, but highly unlikely, that many of ment in which the public lands occur), the transgenic traits under consider- it seems highly unlikely that native for- ation today would increase the prob- Indirect Off-site Effects est lands that are currently under pri- ability of escape of propagules vate ownership would be allocated to Use of transgenics may have a (Hancock and Hokanson 2001). In biodiversity reserves if pressure for number of indirect, but highly signifi- contrast to issues related to introgres- wood fiber production on those lands cant, effects on biodiversity. Indirect sion, escape of exotic species decreased. On private lands, sustainable off-site effects are the consequences of (transgenic or otherwise) generally pose management of native forests for both changes in land use precipitated by use greater potential impacts to biodiversity commodity and non-commodity val- of transgenic plantations. One possible than would escape of transgenic native ues may be the key to conservation of indirect effect, often cited by propo- species, especially if the transgenic biodiversity. Highly productive nents of use of transgenic plantations, plants are sterile. transgenic plantations could decrease is that high-yield transgenic plantations the economic value or perceived social will decrease the pressure for wood fi- value of maintaining native forests in Influences on biotic ber production on other forest lands. a forested condition. This could result interactions The potential for this outcome is well in increased conversion of native for- documented through modeling (Sedjo Just as special transgenic effects ests to transgenic plantations or con- 2001; Victor and Ausubel 2001). It is could influence biotic interactions, es- version of native forests to non-forest often further postulated that, as a con- 171 use, such as agriculture or housing. Many of the questions concerning the to enhance their apparent credibility. Conversion of native forests to planta- potential biodiversity implications of Perhaps this is a reflection of an un- tions or a non-forested condition transgenic plantations to biodiversity fortunate under-valuation of ethical would generally have negative implica- are yet to be fully developed. Many of issues and public opinion relative to tions for conservation of biodiversity. the potential biodiversity issues raised scientific understanding in the deci- High economic value of transgenic to date may prove to be unimportant, sion-making process. Regardless of the plantations also could spawn conversion and other issues that currently have yet reason, confusing ethical and scientific of lands under some other form of land to be hypothesized may emerge. Al- implications often muddles the debate use to transgenic tree plantations. If though some of the issues related to and the decision-making process. In- transgenic traits are developed that en- biodiversity and transgenics could be formed decision-making and meaning- able establishment of plantations in ar- unique, many of the issues parallel ful debate best results when social pref- eas that currently have unsuitable grow- more traditional issues in conservation erences and ethical implications are ing conditions, resulting changes in land- of biodiversity typical of other forest given strong weight, and when scien- use patterns have potential biodiversity systems. Consequently, many of the tific issues are clearly separated from implications. As with local effects, the issues concerning conservation of non-scientific issues. exact nature of these implications are biodiversity in more traditionally man- dependent on context, and could be ei- aged forests are likely to have relevance Biodiversity may be ther positive or negative. to transgenic plantations. Here I Indirect off-site effects of use of present four lessons relevant to conceptually too broad transgenic plantations are likely to be transgenic plantations based on my to provide criteria for highly significant. Actual implications personal experiences working with for- decision-making. could either be negative or positive de- est conservation issues. Attempting to manage for pending on the way that lands freed biodiversity is a noble goal, but is from pressures of production of wood Separating scientific nearly impossible to implement as a fiber are managed. Future patterns of management objective. Because the land use are somewhat unpredictable issues from social and concept of biodiversity is so all-encom- and unraveling the interactions among ethical issues is critical. passing, managing for biodiversity is economic patterns, population growth, In recent years, conservation of probably too broad a concept to pro- consumption, and patterns of land use biodiversity has become a powerful is- vide a meaningful metric for decision- is complex. However, more realistic as- sue influencing land management de- making and for evaluation of manage- sessment of likely implications of in- cisions, and scientific knowledge of ment alternatives. Persistence or viabil- creased productivity of transgenic plan- biodiversity implications has become ity of key species or guilds of interest, tations could be modeled under a range an important metric for evaluation of maintenance of specific ecological of social, political, and economic as- management approaches. As a result, functions, and maintenance of genetic sumptions. arguments that are not scientific in resources in an area are examples of nature or that only tangentially pertain characteristics that, although some- SOME LESSONS FROM to biodiversity are sometimes pre- times difficult to measure, are gener- sented under the guise of science and ally more tractable than broad goals TRADITIONAL FOREST conservation of biodiversity. In some related to biodiversity. Refinement of CONSERVATION ISSUES cases this results from an honest mis- the biodiversity objectives for areas understanding of the science underly- considered for establishment of Our understanding of the impli- ing an issue. In other cases, public transgenic plantations is a necessary cations of transgenic plantations on opinion and social and ethical issues step to evaluation of the conservation biodiversity remains in its infancy. are dressed up in the lexicon of science implications of a management scenario. 172 All resources and values controversial and subject to strong species of tree, size of plantation, man- can not be maximized emotion on all sides of an issue. A agement approach, landscape context, on every parcel of land. strong scientific foundation is necessary and geographic area. Are there any gen- to evaluate the ecological implications eral principles? In the absence of em- of management activities. Such a foun- pirical data, the answer to this question It is not possible to maximize out- dation is entirely lacking for issues re- is debatable. Here I propose five hy- put of all resources and forest values lated to biodiversity and transgenic potheses concerning likely implications on every parcel of land. Clear identi- plantations. In the absence of such of transgenic plantations to fication of the goods and services desired from lands under a given land al- data, arguments and debates concern- biodiversity. I present these hypotheses location is a necessary step to evalua- ing the influences of transgenic plan- to stimulate debate and research, rather tion of the societal acceptability of dif- tations on biodiversity are based on than as a set of incontrovertible prin- ferent management alternatives. What speculation and logical argument, and ciples, although with evaluation these ecological goods and services are de- are often shaped by personal values and hypotheses may form the basis for a set sired from lands managed intensively beliefs. Some questions and issues can of general principles. These hypotheses for wood fiber production? How im- be addressed through modeling stud- also could provide the foundation for portant is conservation of biodiversity ies and small-scale experiments, but a program of research to investigate the on these lands? Are the societal goals many important questions can only be influences of transgenic plantations on the same for transgenic plantations answered from field-based, stand-scale biodiversity. that occupy 20 hectares as for planta- studies. tions that occupy 1000 hectares? Al- Furthermore, some questions con- though scientists can help elucidate the cerning the implications of special Hypothesis 1: implications of establishment of transgenic effects can only be answered Biodiversity transgenic plantations on different as- in transgenic plantations. However, implications will be pects of biodiversity, these questions substantial progress on many relevant greatest when are inherently non-scientific in nature. ecological questions can be made If there is minimal expectation for through studies in existing non- transgenic plantations transgenic plantations to play an im- transgenic plantations. Ecological re- replace or impact the portant role in conservation of search in non-transgenic plantations amount of native forest. biodiversity at the local scale, then the could answer a subset of the questions focus shifts to the likelihood and mag- and help refine research questions to be This hypothesis is likely to be nitude of off-site effects. The answers asked in transgenic plantations. equally true for transgenic and non- to these questions may differ among transgenic plantations, but because of individuals and societies. In any case, the potential for substantial increases more clear resolution of these ques- ARE THERE ANY in productivity and economic effi- tions would be highly valuable to help ciency in transgenic plantations, the focus the issues of concern. GENERAL PRINCIPLES? potential of transgenic plantations to influence the amount of native forest A strong information There are numerous uncertainties may be greater than for non-transgenic concerning the biodiversity implica- base is essential to plantations. As noted above, transgenic tions of transgenic plantations. The plantations have the potential to relieve facilitate intelligent uncertainties are further exacerbated by some of the pressure on native forests decisions. the fact that the implications are highly to produce wood fiber, and this could dependent on context. Implications are either increase area of biodiversity re- Land management and environ- likely to differ with transgenic trait, mental issues often are inherently serves or increase conversion of native 173 forests to non-forested land. The extent through use of sterile organisms. In simple environments. Forest planta- to which use of transgenic plantations cases where fertile or partially sterile tions are often structurally simple, al- influences area of native forest could plants are used, use of exotics with no though the complexity of plantations have great implications to conservation genetically similar species nearby would can vary dramatically with species and of biodiversity. further minimize potential conse- management approach. Depending on quences of introgression of transgenes. how employed, use of transgenics could result in an increase or decrease Hypothesis 2: in structural complexity of plantations. Biodiversity Hypothesis 4: For example, transgenic herbicide re- implications of special Ecological implications sistance would likely alter patterns of transgenic effects will to biodiversity are herbicide use in plantations. Altered greatest when exotic use of herbicides in a manner that in- be greatest when traits creases amount of understory vegeta- are have cascading species are used in tion would likely have positive effects ecological effects transgenic plantations. on biodiversity, whereas decreased un- manifested at multiple derstory vegetation would likely have Although genetic implications are spatial scales. negative effects on biodiversity. Cre- generally reduced through use of exot- ative use of transgenics in plantations ics, ecological implications of using could increase the structural complex- Use of transgenics may alter food exotic transgenics are greater than for ity of the plantation, if this were a goal webs, and thus transgenic plantations use of native transgenics. Ecological of plantation managers. could have cascading ecological effects consequences of escape will generally on biodiversity. Alteration of foliage be greater for exotics than for native characteristics and production of flow- species. In addition, plantations of na- ers, fruit, and seeds could have impli- ACKNOWLEDGMENTS tive species are more likely to provide cations that resonate throughout the similar ecological niches as found in food chain. Effects that are exhibited This paper benefitted from a native forests. at multiple spatial scales and influence number of interesting and provocative trophic relationships outside the plan- discussions with Steve Strauss and from tation are likely to be the most signifi- Hypothesis 5: comments, ideas, and criticisms from cant. several of the participants at the Biodiversity IUFRO International Symposium on implications of Ecological and Societal Aspects of Hypothesis 3: Genetic transgenic plantations Transgenic Forest Plantations. implications to are a function of biodiversity are greatest structural complexity of REFERENCES CITED when transgenic plants the plantation. are established near Hancock, J, and K Hokanson. 2001. enough to genetically The number and types of ecologi- Invasiveness of transgenic vs. Ex- cal niches available for species often is otic plant species: how useful is the similar species to allow a function of the structural complex- analogy? Web page. introgression. ity of an area. As a result, structurally James, RR, SP Difazio, AM Brunner, complex environments are generally and SH Strauss. 1988. Environ- Issues related to introgression can more species rich than are structurally mental effects of genetically engi- be eliminated or greatly reduced 174 neered woody biomass crops. Bio- mass and Bioenergy 14:403–414. Mathews, JH, and MM Campbell. 2000. The advantages and disadvantages of the application of genetic engineering to forest trees: a discussion. Forestry 73:371–380. Mullin, TJ, and S Bertrand. 1998. En- vironmental release of transgenic trees in Canada - potential benefits and assessment of biosafety. For- estry Chronicle 74:203–219. Sedjo, RA. 2001. The economic contribution of biotechnology and forest plantations in global wood supply and forest conservation. Web page. Strauss, SH, WH Rottman, AM Brunner, and LA Sheppard. 1995. Genetic engineering of reproductive sterility in forest trees. Molecu- lar Breeding 1:5–26. Victor, DG, and JH Ausubel. 2001. The great restoration of the world’s forests. Web page, http:// greatrestoration.rockefeller.edu.

175 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 176-186. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Potential Impacts of Genetically Modified Trees on Biodiversity of Forestry Plantations: a Global Perspective

Brian Johnson Keith Kirby

ABSTRACT

Contrary to popular belief, plantations can harbor a significant part of woodland natural biodiversity, and some plantations are crucial to the survival of certain specialized birds and mammals, especially in areas where plantations are managed specifically to enhance wildlife. There may be compelling environmental reasons for increasing plantation areas in the future, including introducing GM plantations, if adequate biosafety can be achieved. If these new GM plantations replace farmland or low-density plantations, this could alleviate commercial pressure on old-growth forest, with considerable benefits to global biodiversity. Improvements to forest trees via conventional selective breeding are severely limited, but transgenic technology offers the opportunity to domesticate trees, to tailor their characteristics more closely to the requirements of commercial forestry and the end-user of forest products. Transgenic techniques can produce varieties that enable different management regimes to be used to grow them. This is especially important in agricultural crops and trees, where the use of agrochemicals can be changed by transgenic traits such as herbicide tolerance, and pest and disease resistance. Such changes can have both adverse and beneficial impacts on native biodiversity in contact with agriculture and forestry. The impact of GM varieties on biodiversity is much more likely to depend on the traits they possess, and not the process by which such traits are achieved. The potential impact on biodiversity of transgenic herbicide tolerant forest trees lies in their ability to be able to withstand the application of broad spectrum herbicides used to control competing vegetation especially in the early stages of plantation establishment. Brian Johnson is Head of The early stages of plantations are known to be important for woodland biodiversity, whether English Nature Advisory plantations are being newly established on open land, or are replacing felled trees. It might be expected that the effects of herbicides used on GMHT trees would lead to widespread Unit, English Nature, weed kill, which in turn would give improved tree growth and quicker and more complete Taunton, UK. canopy closure. If herbicide tolerant trees are ever to be used in GM plantations, there would need to be management schemes to counter any adverse effects on natural biodiversity. [email protected] These could include factors such as leaving areas unsprayed to act as the equivalent of ‘rides’ and ‘glades’ in non-GM forests. Keith Kirby is forestry Changing the quality of timber in growing trees has long been a goal of plant breed- and woodlands officer ers. Transgenic technology offers a way of achieving radical changes in factors such as the lignin-to-cellulose ratio of both conifers and deciduous hardwoods. In doing so, there is for English Nature, the potential to alter the palatability of such trees to phytophagous animals, with consequent damage to the trees. This could lead to the need for more agrochemical manage- Peterborough, UK. ment of such plantations, or for the insertion of multiple traits such as insect, fungal, and [email protected] viral resistance to combat such damage. Concerns about gene transfer to and from transgenic trees has led to increased research into mechanisms for engineering sexual incompetence into transgenic trees. The 176 result is likely to be trees that produce little pollen, and few flowers, seeds, and fruits. Not only are these food resources important for supporting biodiversity in coniferous and deciduous hardwood forests, they are also important economically in some parts of the world, especially where plantations produce copious nectar that can support bee-keeping enterprises. If sexual incompetence is necessary for biosafety reasons, then single variety plantations may be undesirable, and mixed varieties (GM and/or conventional) stands would be preferable. There may also be biosafety advantages in growing mixed stands of GM and conventional trees, specially in reducing selection pressures for the evolution of pest resistance. Mixed stands could also be designed to further reduce gene flow in situations where sexual incompetence mechanisms in GM trees are not totally effective. In assessing risks from annual GM crops, it is possible to carry out comparative ecological studies, such as the farm-scale evaluations of GMHT crops in the UK. With long- lived perennials such as trees, however, a predictive modeling approach is the only realistic option if regulatory authorities are to make decisions on consent for release within a reasonable time scale.

ebate about the use of biotechnology to create transgenic plants and animals has largely centered on the perceived direct risks and benefits Dof introducing new genes into crops, fish, and trees. Concerns about food safety and gene transfer have dominated the debate, but there has been relatively little discussion about how plants and animals possessing transgenic traits might enable new forms of agriculture, aquaculture, and forestry management with consequent indirect risks and benefits. Historically the impacts of new management regimes in agriculture, fish farming, and forestry have been profound, changing whole landscapes and their associated biodiversity. This paper focuses on assessing the possible effects of genetically modified (GM) trees on the biodiversity of forest plantations, and does not address in detail how the use of transgenic trees might impact on forestry strategy, such as the possibility of high- yielding GM forests replacing old-growth cropping as a prime source of high- quality timber, pulpwood and other forest products. Plantations have been a major part of European forestry for at least 300 years, and are an increasingly important feature of commercial forestry throughout the globe. Research has revealed much about how plantation composition and management affects the ecology of the natural biodiversity found in and around plantations. The ecological principles revealed by this research effort, centered in the conifer plantations of Europe and the pine plantations of the southern United States, are applicable to plantations worldwide.

PLANTATION ECOLOGY

Large-scale afforestation of agricultural and marginal land has been practiced in Europe since the 18th century and in some areas, old-growth forests have been replanted either in part or by complete replacement of the ancient trees. Planta- tions now make up over 50% of all forests and woodlands in Europe and supply a higher proportion of timber and other forest products than old-growth forests (Peterken 1993). It is estimated that the area of plantations worldwide is over 185 million ha, of which around 60% are in temperate regions and around 40% in the tropics (FAO 2000). 177 Plantations are not only the most servationists and foresters have argued flowering) and genomics research. productive and profitable forest areas, that substantial areas of new planta- There is also increasing research inter- but can also make a significant contri- tions on open land will be necessary to est in genetically modifying trees to bution to general biodiversity; in some alleviate commercial pressure on natu- enable them to be grown in saline and regions of Europe they harbor most of ral forests, especially those still retain- arid soils. the forest biodiversity. However in ing large areas of primary old-growth There has been little consideration other parts of the world plantations forest ecosystems (Shepherd 1993). of the effects that plantations of GM make a relatively minor contribution to These arguments are being in- trees might have on forest ecology and forest biodiversity, often replacing creasingly applied to tropical areas of biodiversity, but generally, in inten- highly diverse ecosystems with species- Africa, Asia, and the Pacific Rim, sively managed landscapes, the estab- poor monocultures. However, some where commercial pressure on rain for- lishment of plantations on land that tropical plantations on low-grade farm- ests and other natural ecosystems is has been previously managed as arable land can be ecologically diverse. For leading to severe degradation of natu- farmland and sown grassland results in example, in Malaysian plantations of ral biodiversity, climate, water re- a net gain for biodiversity (Nature Eucalyptus moth diversities can be as sources, and soils. Plantations of com- Conservancy Council 1991), whereas high as those in natural secondary for- mercially valuable trees such as teak, replacement of old-growth forest and est (Chey et al.1997). mahogany, and eucalyptus are increas- semi-natural vegetation by plantations In Europe and North America, the ing in area as logging from natural for- often lowers biodiversity, favoring colo- planting dates and management history ests becomes more difficult both physi- nizing, ruderal flora and fauna over the of new and replacement woodlands are cally and politically. diverse wildlife of natural ecosystems sometimes well documented, allowing If plantations are to replace a sig- (Nature Conservancy Council 1986), research to be carried out comparing nificant proportion of present-day but this is not always the case for all plantations of different ages and tree commercial production from old- taxa. For example, recent research in species in a large number of different growth forests, there is a strong argu- the UK reveals that species richness of locations with differing management ment to make them as productive as macrofungi in conifer plantations is as regimes. The biodiversity of American possible in terms of both yield and high as that of semi-natural pine and and European plantations has therefore quality, and also to lower management oak woodlands (Humphrey et al. been the subject of much research over inputs. This ‘intensive silviculture’ 2000). the past 40 years, with the results of (Moore and Allen 1999) is being rap- The biodiversity of forest planta- research being used as a basis for plant- idly adopted throughout the world as tions depends on a number of factors, ing schemes and management regimes a way of increasing production per unit of which the following are the most that favor high biodiversity in planta- area whilst simultaneously reducing important. tions that also give high timber-crop unit production costs. In common yields. with development of transgenic agri- Size In North America, Asia, and parts cultural crops, most research on GM of Central and Eastern Europe, plan- trees has concentrated on traits aimed Larger plantations generally sup- tations on previously afforested areas at securing these goals. Pest and disease port more biodiversity (Peterken are increasingly replacing old-growth resistance, modified quality traits such 1993). This is mainly because in large forests that have been logged, leading as altered lignin content, and tolerance forests there are likely to be more to increasing concern about the sur- to herbicides have been the main traits sources from which propagules of vival of species associated with ancient for transgenic trees, with current re- woodland species can colonize the forest areas, some of which are still search also focusing on domestication plantations, and larger forests are more being logged. In order to maintain tim- of trees through transgenic traits for likely to contain a wider variety of geo- ber production in these regions, con- accelerated breeding (e.g., precocious morphological features than small

178 woodlands. This means that larger higher in deciduous stands, especially managers practice localized weed con- plantations are more likely to contain those where abundant tree flowers and trol around the growing trees, leaving more biological and physical ‘niches’ seeds are available. Deciduous stands areas between the crops to undergo that can be exploited by woodland usually have a greater diversity of physi- natural succession. In many parts of the wildlife. cal structure than conifers, especially if world, tolerating weeds (including na- the latter are even-aged single species tive trees and shrubs) in young planta- Sources of woodland stands (Moss 1979). Tree species diver- tions could have a severely damaging sity within plantations is also important effect on establishment of the planted species already on site because stands that have a mixture of trees, especially in some parts of the or nearby tree species can be expected to have a tropics where colonization by grasses Peterken and Game (1984), in an wide range of tree architecture and food and native woodland shrubs can easily extensive review of the biodiversity of resources, not only from the foliage, out-compete the crop. This may be less plantations in the UK, found that flowers, and seeds of living trees, but also of a problem in temperate regions, woods that had been planted adjacent in the range of saprophytic fungi and where tree growth can overcome com- to ancient woodland and old wood- other species feeding on deadwood. peting flora, often with little interven- land/hedge features were much richer tion other than pre-planting weed con- in woodland biodiversity, and so were Management trol. Single-application weed control those that included water features such measures may not have long-lasting Biodiversity is higher in stands as streams and rivers. effects on non-crop plant and insect that are managed for diverse forest ar- diversity, and if forest managers use chitecture. Small coups areas and short herbicides carefully, leaving some areas Age of Forest rotations maintain higher natural untreated, biodiversity can be main- Studies (e.g., Peterken 1993) com- biodiversity because they result in vari- tained (Morrison and Meslow 1984; paring vascular plant colonization of able age structure between coups, and Santillo et al. 1989; Mellin 1995). plantations of different ages have found more regular opening up of the forest The sporadic use of insecticides in that older plantations are slightly richer floor, allowing flora and fauna to ben- forestry management can have adverse than recently planted stands, but after efit. Smaller coup areas also provide impacts on non-target animals, although around a century of establishment, more forest rides and firebreaks, which these effects are usually confined to short plantations do not gain in vascular are also very important, both as sources periods after application, with plant richness, presumably because as of colonization and for the provision recolonization rapidly replacing the new species colonize, early colonizers of ‘forest-edge’ habitats, which favor fauna. Stribling and Smith (1987) die out. Although rates of colonization high insect and bird diversity. Studies showed that in oak/maple forests in the of specialist woodland plants are often in the United States and Europe United States, spraying insecticides for very slow, especially into closed-canopy (Mitchell and Kirby 1989) indicate the control of gypsy moth did not ap- mature stands, some insects, birds, and that the more diverse the spatial and pear to damage bird populations. In con- mammals rapidly colonize new stands, temporal physical architecture, the trast, pine beauty moth control pro- taking advantage of the physical and higher the biodiversity of the forest, no grams in Canada had short-term adverse biological resources provided by the matter which tree species are planted. effects on birds (Spray et al 1987). trees themselves. Generally, high plant abundance and diversity (“forest weeds”) in early Nutrient Levels and stages of plantation growth are impor- Tree Species tant for the survival of insects and Geology of Soils Within the temperate zone, insect breeding birds (Moss 1979). The high- High nutrient levels on clay soils and bird diversity is generally much est natural biodiversity is found where give an impoverished ground flora

179 composed of a few very competitive from their ancestral form, which in and other products that are closer to species, whereas the ground flora of turn enable different management re- market needs, and of producing trees plantations on sands, peat, and loams gimes to be used to grow them. This that can be grown more easily and with lower nutrient levels are generally is especially important in agricultural quickly in plantations. more diverse (Ferris et al. 2000). Soil crops and trees, where the use of agro- The impact of such trees on types also affect the capacity of plan- chemicals can be changed by transgenic biodiversity is much more likely to tations to acidify watercourses within traits such as herbicide tolerance, and depend on the traits they possess, them. Peat soils in particular can exac- pest and disease resistance. Such rather than the process by which such erbate acidification, adversely affecting changes can have both adverse and traits are achieved. Currently the traits the ecology of forest streams and the beneficial impacts on native discussed below are more easily lakes and rivers into which they flow biodiversity in contact with agriculture achieved using transgenic techniques, (Stoner et al. 1984). The replacement and forestry (Johnson 2000). Conven- but in the near future an increasing of deciduous forest by conifers can also tional breeding can produce agricul- knowledge of tree genomics, coupled significantly reduce stream flow by in- tural crops with similar traits (e.g., her- with marker-assisted breeding, may be creasing transpiration and intercepting bicide tolerance and pest resistance), capable of producing similar results. rainfall, damaging the ecology of but selecting for such traits is very dif- woodland streams (Swank and ficult to achieve in forestry tree species, Douglass 1974). where plant breeders have to contend GM Herbicide with long generation times and traits Tolerance (GMHT) that may only manifest themselves at POTENTIAL EFFECTS OF maturity. With the advent of biotech- Table 1 shows that herbicide tol- nology and cloning techniques, there erance is one of the main areas of re- GM TREES ON is the real possibility of developing search and development in transgenic PLANTATION domesticated trees that produce timber trees. This research effort has been fo- IODIVERSITY B Table 1. Number of GM tree field trials, 1988–2000.

Biotechnology is an important Genus HT IR VR Lignin Markers Others Total extension of conventional breeding Betula 1 1 2 techniques, not least because the tech- Castanea 1 1 nology enables radical changes in phe- Corcia 10 2 12 notypes through the insertion of gene Eucalyptus 4 1 2 3 2 12 cassettes that confer new traits. These Juglans 7 1 7 15 changes are often not achievable using Liquidambar 3 3 Malus 5 6 2 16 29 conventional breeding techniques, Olea 2 2 4 partly because the time taken to find Picea 3 3 6 or construct suitable genes would be Pinus 1 1 11 2 15 inordinately long, but mostly because Populus 41 36 10 14 42 143 the genes necessary to produce desired Prunus 3 3 1 7 traits are not present in the gene pool Pyrus 3 3 of the species and its ancestors. Total 55 53 15 13 41 75 252 Transgenic techniques, like other forms of plant breeding, produce vari- Note: HT = herbicide tolerance; IR = insect resistance; VR = virus resistance. “Other” traits include fungal resistance, salt tolerance, altered flowering, and faster growth. eties that exhibit different phenotypes Source: Rautner, M., 2001. Biotech and Development Monitor: 44/45, pp. 2–7, Amsterdam.

180 cused on achieving tolerance to broad- young planted trees and diverse wild on agriculturally degraded soils, provid- spectrum herbicides such as glyphosate plants that support the food webs on ing soil stability, soil refurbishment, and glufosinate-ammonium. These which they rely. Santillo et al. (1989) and carbon sequestration. herbicides are commonly used in con- and Morrison and Meslow (1984), for However these developments also ventional forestry to control vigorous example, found that herbicide treat- raise the possibility of plantations species of grasses, herbs, and shrubs ments of felled and replanted forest sig- growing on soils that do not naturally that compete with the planted crop. nificantly reduced the abundance and support a characteristic woodland flora The main effect of using herbi- diversity of phytophagous arthropods. and fauna. Plantations on saline soils cides in forestry plantations, whether Studies in Sweden have linked the di- may develop a ground flora and under- conventional or GMHT, is the poten- versity and abundance of insectivorous story of species found in shaded gul- tial for destruction of native woodland birds with the availability of herbivo- lies in saline habitats, such as salt flats flora and dependent fauna already rous insect larvae in Picea forests and salt marshes. In the tropics, salt- present on site. Drift to watercourses (Atlegrim and Sjöberg 1996). tolerant plantations might support as- and hedge banks within and on the Widespread and routine applica- sociated wildlife characteristic of man- margins of plantation sites can result tion of broad-spectrum herbicides to grove swamps, especially in coastal and in destruction of flora that are poten- new plantations might prevent the river floodplain locations. However, tial colonizers of the developing wood- establishment of woodland understory there may be relatively few plant spe- land. Pre-planting destruction of these species that rely for germination on the cies capable of tolerating the double flora and their associated fauna could light and moisture of newly planted stress of shade plus high salt levels. result in impoverished forest ecosys- stands. Woody shrubs generally cannot The implications for biodiversity tems in GMHT plantations in later germinate in closed-canopy forests, and of these developments are difficult to years. Pre-planting herbicide regimes establish themselves either early in the predict from ecological theory. Only by for GMHT plantations would be simi- plantation successional stages or only establishing pilot-scale plantations can lar to those already being used, and can very much later after the stand has ecologists begin to understand the im- be expected to have similar impacts on been thinned. In the absence of these pacts of afforesting saline and arid ar- native biodiversity. At present, herbi- shrubs, plantations are poorer in wild- eas with their established and highly cide application after planting often life because they are not only simpler adapted ecosystems. risks damaging conventional trees, es- in species composition but also in pecially where competing wild plants terms of structural diversity. require high concentrations of herbi- PEST AND DISEASE cides for effective control. The introduction of herbicide tolerance would Tolerance to Adverse RESISTANCE allow broad-spectrum weed control at Soils any stage in plantation development, GM Insect Resistance but would be most likely to be used Drought resistance and tolerance (GMIR) in the early stages of tree growth, when to acidic and saline soils are active ar- a wide range of forest and forest-edge eas of research into transgenic trees. It has been argued by some con- species tend to be present. These characteristics may eventually tributors to the GM debate that the If GM trees tolerant to broad- enable afforestation of areas having introduction of insect resistance into spectrum herbicides were to be widely soils where commercially valuable trees forest trees would render plantations used, GMHT plantations could be less are currently unable to grow success- devoid of phytophagous and pollen- attractive for species of birds and in- fully. This could be beneficial to feeding insects (Owusu 1999 and vertebrates that rely on young planta- biodiversity and general environmen- Tickell 1999). Whilst this may be a risk tion habitat, with its combination of tal health if it allowed trees to thrive where broad-spectrum anti-feedants

181 and insecticides (such as lectins and Current proposals for managing such sition then leaf litter and brash could protease inhibitors) are produced by risks usually assume that resistance in break down more slowly, with unpre- GM trees, more specific transgenic in- insects would effectively be a recessive dictable ecological consequences for secticide resistance such as Bt is only trait if high doses of Bt toxin were used arthropod and fungal components of likely to have significant adverse effects in the transgenic plants (Andow et al. the forest ecosystem. on woodland ecosystems where the IR 1998). This is a rather speculative as- trait impacts on keystone phytopha- sumption that, if incorrect, could prove gous species, such as Lepidoptera and fatal to resistance management schemes Transgenic Fungal and Coleoptera, that are themselves endan- relying on maintaining refugia of sus- Virus Resistance gered or are crucial parts of food webs ceptible insects (Gould 1998; Andow supporting rare or endangered species et al. 1998). If pest populations and Traits conferring general fungal of birds or mammals. This could be a non-target insects develop resistance to and viral resistance have been inserted greater risk in tropical areas, where rare the GMIR trait, and this is likely where into tree species, mostly into fruit-pro- species may be more specialized in their the trait is inserted into long-lived ducing varieties such as papaya, apple, food and habitat requirements than in trees, then there could be a long-term and cherries (Table 1). Whilst virus temperate zones. The introduction of management trend towards increased resistance is unlikely to have an adverse trees containing Bt toxins in these situ- insecticide use. Although target pest impact on the biodiversity of planta- ations could have a significant adverse populations may be reduced by an IR tions, it is possible that the introduc- effect on these rich ecosystems. Lectins trait, this could allow other, previously tion of generalized fungal resistance and protease inhibitors may be more rare, ‘secondary’ pest species to flour- could affect decomposer ecosystems in generalized in their effects, and risk ish, leading to increased need for con- plantations, although there is a large destroying or deterring most ventional chemical control (Sharma number of fungi involved in such pro- arthropods, and probably mammals and Ortiz 2000). Ashouri et al. (2001) cesses in woodlands and it is likely that and birds feeding on the trees, but ex- have shown that GMIR potatoes de- the traits would be overcome by at least posure to the toxins would depend on signed to resist Colorado potato beetle some of the fungi present. However levels of gene expression in various (Leptinotarsa decemlineata) had unex- many saprophytic fungi are quite spe- parts of the GM trees. pected effects on populations of potato cialized in their choice of substrate, so These risks must be assessed in aphids feeding on the transgenic plants. there could be adverse effects on the relation to the risks posed by the use In one transgenic line they studied, diversity of macrofungi in GM fungal- of insecticides to control pests in con- these effects could have had a tertiary resistant plantations. ventional plantation management. impact on the spread of potato viruses. Concern about the conservation Risks from conventional pesticide use This work shows that even in a simple of fungi has increased in recent years are relatively low due to the sporadic agricultural ecosystem, it is difficult to because there has been a significant nature of insecticide use in conven- predict ecological perturbations caused decline in the abundance of many fun- tional forestry, and the capacity of by GM plants with simple monogenic gal species throughout Europe. woodland fauna to recolonize after in- insect resistance. Humphrey et al. (2000) found that secticide use. Without comparative re- IR traits might affect the soil and mature stands of pine and spruce held search it is difficult to estimate relative wood decomposer cycle within planta- the greatest diversity of fungi, and that risks to biodiversity of introducing tions because the initial stages of de- clear felling with removal of deadwood GMIR traits into trees. This is com- composition in forest ecosystems often was associated with significant reduc- pounded by the difficulty of predict- involve comminution of plant material tions in fungal species. Of the 419 spe- ing long-term effects, especially where by arthropods, especially mites (Edlin cies they recorded in plantations, 157 both target and non-target insects 1970). If the GM trait were adversely were litter saprophytes. As the manage- could develop resistance to the trait. to affect this initial stage of decompo- ment of GM plantations might differ

182 from that of conventional stands, im- of trees which could affect their wind The production of pollen, nectar, pacts on fungal diversity may become firmness in plantations. This could lead seeds, and fruits is an important fac- an important aspect of monitoring. to changes in how and where they are tor in maintaining natural biodiversity grown, or in their patterns of response in uniform single-species plantations of to extreme wind conditions. Such conventional trees, whether coniferous QUALITY TRAITS changes could impact forest strategy by, or deciduous (Palik and Engstrom for instance, encouraging more plan- 1999). In temperate regions, for ex- So far most research on changing tations in less exposed situations. There ample, coniferous plantations are cru- the quality of timber has focused on may implications for biodiversity in- cial to the survival of several species of varying lignin/cellulose ratios in de- herent in such changes. seed-eating birds (e.g., crossbills in ciduous woods such as poplars (Populus Europe and Clark’s nutcracker in the spp.) Wood from these GM trees U.S.) and some mammals, such as red would be better suited to the industrial STERILITY AND OTHER squirrels in Britain and the red tree- processes used to recover cellulose for mouse in the United States. Specialist IOSAFETY RAITS paper and board manufacture, poten- B T birds, such as the yellow-bellied sap- tially leading to reductions in the sucker, depend on a range of conifer Genetic modification is often used chemicals and energy needed for and deciduous tree saps and nectars at to manipulate tree species that are ei- pulping, which in turn should reduce critical points in their breeding cycles ther native to the region in which they pollution from mills (Petit-Conil et al. (Tate 1973). In tropical regions, de- are to be grown or sexually compatible 2001). ciduous plantations often produce co- with native trees that are closely related. Little is known about possible ef- pious quantities of nectar and pollen Trees are by nature long-lived peren- fects of altering lignin and cellulose that support a wide range of insects nials, and regulatory authorities are contents on palatability of GM trees to and the food webs associated with rightly concerned about genetic stabil- phytophagous animals, although ex- them. Honey bees often feed on de- ity and gene flow from the transgenic perimental releases of reduced-lignin ciduous hardwood plantation species, varieties to native species. In some spe- poplars appear to be no more suscep- and in some areas provide a valuable cies, notably poplars and aspens tible to insect attack than are conven- source of income. There is therefore a (Populus spp.) and willows (Salix spp.), tional poplars. Other trees with altered conflict between the desire to enhance there is concern that transgenics may quality traits may, however, be found biosafety by suppressing sexual repro- be able to propagate vegetatively, but to behave in similar ways to hybrid duction in transgenic trees, and the most risk assessment has centered on agricultural crops, requiring more de- need to maintain biodiversity, and in gene transfer via pollen. fense against phytophagous arthropods some areas economically important There has been a trend toward try- and fungus and virus attack. If such activities such as plantation-based bee- ing to engineer sexual incompetence defense were to be in the form of in- keeping. into transgenic trees, either by disrupt- creased agrochemical use, then the The most obvious solution to ing pollen production mechanisms or biodiversity value of these GM plan- maintaining biodiversity among sterile by suppressing flowering, and therefore tations might be lower than that of trees would be to plant mixed stands the production of seeds and fruits. If conventional plantations. Alternatively, of transgenic and conventional trees, these measures were adopted commer- it might be necessary to introduce either in direct admixture or by plant- cially, plantations would be of trees multiple GM traits, such as insect, fun- ing blocks of different trees. Mixed- producing neither pollen nor fertile gal, and viral resistance to combat dam- species stands are known to increase ovaries. A major element of the wood- age. forest biodiversity by increasing struc- land food web would be either unavail- Changing lignin:cellulose ratios tural complexity, but as Larsen (1995) able or greatly reduced in these plan- may affect the strength characteristics points out, tree species for admixture tations. 183 should be chosen to support food web could be offset by simplification of ing fitness lies in identifying the prin- interactions by taking account of plantation food webs caused by high- ciple components of fitness relative to known co-evolutionary relationships. growth GM trees that included pest- the organism in question and the habi- Plantations of this type should be more and disease-resistant traits. tats within which it lives. Muir and resistant to physical or biotic stress Howard (2001) have modeled impacts (Larsen 1995). on fitness of a theoretical transgene Because genetically engineered RISK ASSESSMENT that influences several fitness param- sexual incompetence will occasionally eters simultaneously in a fish, Japanese fail, there could be biosafety implica- There are serious concerns about medaka (Oryzias latipes). They show tions for mixed-species plantations if introducing high growth traits and pest that, for a wide range of fitness values, the species admixture included trees and disease resistance into transgenic transgenes conferring quite small in- that were sexually compatible to trees, especially where the tree species creases in fitness could spread in transgenics. If the mixed species were already shows invasive tendencies. medaka populations. Using a similar sexually incompatible, risks could be There are examples where non- methodology, models could be devel- acceptably low, and such mixtures transgenic conifers and deciduous trees oped for estimating impacts of could add to biosafety by further re- have invaded natural ecosystems (for transgenes (such as increased growth) ducing vegetative propagation and pol- example Hippophae rhamnoides, Pinus on tree fitness, providing there are suf- len flow from GM trees that sporadi- sylvatica, Rhododendron ponticum, ficient data available for identifying cally regain fertility. Robinia pseudacacia, and Quercus ilex and estimating fitness components. A in Europe, and Syringa in the United major problem with the modeling ap- States). These invasions have often proach to fitness assessment of GROWTH TRAITS been from plantations into surround- transgenic perennials lies in the diffi- ing natural woodland. Research has culty of predicting the environments Transgenic traits designed to give shown that predicting invasiveness is into which such trees might spread. faster growth have been suggested as a very difficult, especially for long-lived Unless this can be done with some cer- possible solution to improving produc- perennial species like trees (Williamson tainty, it will not be possible to esti- tivity of biomass plantations and those and Fitter 1996; Manchester and Bul- mate fitness components to populate where pulpwood is the primary goal. lock 2000). a model. Such plantations could be advanta- Risk assessment processes world- If GM trees are ever to be used in geous to native biodiversity, especially wide have not yet been able to predict plantation forests, it is likely that they if they were subjected to regular short- with any certainty what impacts on will contain combined GM traits, such rotation coppicing favoring woodland native biodiversity might result from as altered timber quality with insect flora and bird species that require a releasing transgenic perennial plants resistance. It could be very difficult to range of ecological successions within such as trees. The issue of gene flow model the direct and indirect impacts forests. has dominated the debate and has on biodiversity of these GM multiple- Even if biomass and pulpwood largely focused on attempts to measure trait phenotypes, even if the potential coppicing with regular fertilizer use rates of gene transfer (the question of ecological impacts of each trait in iso- were to result in species-poor ground “Does it happen?”), with little research lation were known. In assessing risks flora, the temporal and spatial range of on the impacts transgenes might have from annual GM agricultural crops, it vertical structure produced could favor on fitness of recipient organisms is possible to carry out comparative woodland flora and fauna, including (“Does it matter?”). This is especially ecological studies over three or four bird species that require a range of difficult to predict in the longer term years, such as the ‘farm scale evalua- structural successions within forests. where long-lived perennials such as tions’ of GMHT crops in the UK However, these potential advantages trees are involved. The key to predict- (Firbank et al. 1999; Firbank and

184 Forcella 2000), but with long-lived ACKNOWLEDGMENTS tween vegetation, site type and perennials such as trees, a predictive stand structure in coniferous plantations in Britain. Forest Ecology modeling approach may be the only The authors thank Anna Hope, and Management 136 (1–3): 35– realistic option if regulatory authorities English Nature for reviewing an early 51. are to make decisions on consent for draft of the manuscript. Firbank, LG, AM Dewar, MO Hill, MJ release within a reasonable time-scale. May, JN Perry, P Rothery, G Where the ecological relationships be- Squire, IP Woiwod. 1999. Farm tween trees and ‘target’ conservation LITERATURE CITED scale evaluation of GM crops ex- species are known it may be possible plained. Nature 339: 727–728. to model impacts. An example can be Andow, DA, DN Alstad, Y-H Pang, PD Firbank, LG and F Forcella. 2000. Ge- found in the use of hybrid poplars by Bolin, and WD Hutchison. 1998. netically modified crops and farm- breeding golden orioles (Oriolus oriolus) Using an F2 screen to search for land biodiversity. Science 289: in the UK, where the physical struc- resistance alleles to Bacillus 1481–1482. ture and phenology of the trees is cru- thuringiensis toxin in European Gould, F. 1998. Sustainability of cial to breeding success. The critical corn borer (Lepidoptera: transgenic insecticidal cultivars: in- Crambidae). Journal of Economic factors in this ecological relationship tegrating pest genetics and ecology. Entomology 91: 579–584. are known (Milwright 1998) so Ann. Rev. Entomol. 43: 701–726. transgenic poplars could be assessed for Ashouri, A, D Michaud, and C Humphrey, JW, AC Newton, AJ Peace. Cloutier. 2001. Unexpected effects their suitability as golden oriole habi- 2000 The importance of planta- of different potato resistance fac- tat. tions in northern Britain as a habi- tors to the Colorado potato beetle tat for native fungi. Biol. Conserv. The use of ecological modeling (Coleoptera:Chrysomelidae) on 96 (2): 241–252. may improve the capacity of risk asses- the potato aphid sors to be able to predict biodiversity (Homoptera:Aphididae). Environ- Johnson, BR. 2000. Genetically modified crops and other organisms: impacts, but comparison between mental Entomology 30(3): 524– 532. Implications for agricultural transgenic and conventional planta- sustainability and biodiversity. In: tions can only be made if models can Atlegrim, O and K Sjöberg. 1996. Ef- GJ Persley and MM Lantin, eds., be populated with quantitative data. To fects of clear-cutting and single-tree Agricultural Biotechnology and the selection harvests on herbivorous provide adequate data for risk assess- Poor: Proceedings of an International insect larvae feeding on bilberry ment there is clearly a need for more Conference, Washington, D.C., (Vaccinium myrtilis) in even-aged 21–22 October 1999, pp. 131– research on the population dynamics of boreal Picea abies forests. Forest 138. Consultative Group on Inter- transgenic trees and the general ecol- Ecology and Management 87: 139– national Agricultural Research, ogy of plantations, especially forests 148. Washington, DC. populated by tree species that are cur- Chey, VK, JD Holloway, MR Speight. Kirby, K. 1992. Woodland and Wildlife. rently the focus of transgenic and ge- 1997. Diversity of moths in forest Whittet Books, London nomic development. There is a particu- plantations and natural forests in Larsen, JB. 1995. Ecological stability of lar need for more research into the Sabah. Bulletin Entomol Res 87 (4): 371–385. forests and sustainable silviculture. ecology of plantations in the tropics, Forest Ecology and Management 73: where the relationship between forest Edlin, HL. 1970. Trees, Woods and Man. 85–96. trees and non-crop biodiversity is im- Collins, London. Manchester, SJ, and JM Bullock. 2000. portant for nature conservation and FAO. 2000. Forest Resources Assess- The impacts of non-native species may be critical to the stability and pro- ment 2000 www.fao.org/forestry/ on UK biodiversity and the effec- ductivity of future plantations, whether fo/fra/index.jsp. tiveness of control. Journal of Ap- transgenic or conventional (Larsen Ferris, R, AJ Peace, JW Humphrey, AC plied Ecology 37, 845–864. 1995). Broome. 2000. Relationships be-

185 Mellin, TC. 1995. The effects of inten- Lowland Forests. Nature Conser- strategy. Australian Forestry 56 (4): sive forest management practices vancy Council. Peterborough, UK. 237–335. on the natural vegetative commu- Owusu, RA. 1999. Gm Technology in Spray, CJ, HQP Crick, ADM Hart. nities of loblolly pine plantations the Forest Sector - a Scoping Study 1987. Effects of aerial applications in North Carolina. Thesis. North for WWF. World Wide Fund for of fenitrothion on bird populations Carolina State University, Raleigh, Nature, WWF-UK. of a Scottish pine plantation. Jour- North Carolina. nal of Applied Ecology 24: 29–47. Palik, B and RT Engstrom. 1999. Milwright, RDP. 1998. Breeding biol- Chapter 3: Species Composition. Stoner ,JH, AS Gee, KR Wade. 1984. ogy of the Golden Oriole Oriolus Pp. 63–94. In: Maintaining The effects of acidification on the oriolus in the fenland basin of east- Biodiversity in Forest Ecosystems. ecology of streams in the upper ern England. Bird Study 45: (3) Malcolm L Hunter, ed. Cambridge Tywi catchment in West Wales. En- 320 – 330. University Press, Cambridge, UK. vironmental Pollution A, 35: 125– Mitchell, PL and KJ Kirby. 1989. Eco- 157. Peterken, GF. 1993. Long-term floris- logical effects of Forestry Practices in tic development of woodland on Stribling, HL and HR Smith. 1987. long-established woodland and their former agricultural land in Effects of Dimilin on diversity and implications for nature conservation. Lincolnshire, England. Pp. 31–45 abundance of forest birds. North- Oxford Forestry Institute, Occa- In: Ecological Effects of Afforestation. ern Journal of Applied Forestry 4: sional Paper No 39, Oxford. C Watkins, ed. CAB International, 37–38. Moore, SE and HL Allen. 1999. Chap- Oxon, UK. Swank, WT, and JE Douglass. 1974. ter 12: Plantation Forestry. Pp400– Peterken, GF, and M Game. 1984. His- Streamflow greatly reduced by con- 433. In: Maintaining Biodiversity torical factors affecting the num- verting deciduous hardwood in Forest Ecosystems. Malcolm L. ber and distribution of vascular stands to pine. Science 185: 857– Hunter, ed. Cambridge University plant species in the woodlands of 859. Press, Cambridge, UK. central Lincolnshire. Journal of Tate, J. 1973. Methods and annual se- Morrison, ML and EC Meslow. 1984. Ecology 72, 155–182. quence of foraging by the sap- Effects of the herbicide glyphosate Petit-Conil, ML, CPGL, C Jean, E sucker. Auk 90: 840–856. on bird community structure, Guiney, W Boerjan, I Jouanin,C western Oregon. Forest Science 30, Tickell, O. 1999. GM trees — Flower- Chirat, and B Pollet. 2000. De- 95–106. less Forests. Forestry and British signing lignin structure for modi- Timber September 1999: 36–39. Moss, D. 1979. Even-aged plantations fying pulping performances: Appli- as a habitat for birds. In: The ecol- cation to poplar trees. Tappi Williamson, M and A Fitter. 1996. The ogy of even-aged plantations. ED Pulping/Process and Product characters of successful invaders. Ford, DC Malcolm, J Atterson, Quality Conference, Boston, Pro- Biological Conservation 78: 163– eds. Institute of Terrestrial Ecology, ceedings (Session 12): paper 12– 170b. Cambridge, UK. 1. Muir, WM, and RD Howard. 2001. Santillo, DJ, DM Leslie, PW Brown. Fitness components and ecological 1989. Responses of small mam- risk of transgenic release: a model mals and habitat to glyphosate ap- using Japanese Medaka (Oryzias plication on clearcuts. Journal of latipes). The American Naturalist Wildlife Management 53, 164– 158: 1: 1–15. 172. Nature Conservancy Council 1986. Sharma, HC and R Ortiz. 2000. Nature Conservation and Afforesta- Transgenics, pest management, tion in Britain. Nature Conser- and the environment Current Sci- vancy Council. Peterborough, UK. ence 79 (4): 421–437. Nature Conservancy Council. 1991. Shepherd, KR. 1993. Significance of Nature Conservation and the New plantations in a global forestry

186 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 187-192. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Invasiveness of Transgenic vs. Exotic Plant Species: How Useful is the Analogy?

James F. Hancock Karen E. Hokanson

ABSTRACT

Numerous ecologists and evolutionary biologists have incorrectly suggested that genetically engineered crops are analogous to exotic introductions. A growing body of evidence indicates that exotic species become invasive when they are introduced into a new area, where there are few to none of the natural constraints with which they evolved, and so they fill a new niche and their numbers explode. Most of the successful exotics are already good colonizers somewhere else and carry a whole syndrome of traits associated with weediness. This is very different from the situation facing transgenic forestry and agronomic crops. The crop antecedents are generally poor competitors outside the agroecosystem and carry few weediness traits. After the crop is engineered, it will not be removed from the complex array of natural constraints that currently faces it, and in most cases only one of those constraints will be removed by the addition of a new trait. In fact, it is much easier to predict the environmental risk of transgenic crops than an exotic introduction, as the level of risk in transgenics can be measured by evaluating the fitness impact of a single engineered trait, rather than a whole syndrome of potentially invasive traits. The risk of most transgene deployments can be effectively predicted by considering the phenotype of the transgene and the overall invasiveness of the crop itself.

t has commonly been suggested that invasive, exotic species can be used as models for evaluating the risk of release of transgenic crops (NAS 1987; Tiedje Jim Hancock is with the I et al. 1989; Parker and Kareiva 1996; Marvier 2001). For example, Keeler (1998) states, “one set of data that can be used to understand what engineered Department of Horticul- organisms are likely to do is derived from the literature on introduced organisms. They are not genetically engineered, but they represent organisms that were ture, Michigan State introduced into communities of organisms which they had no previous experi- University, East Lansing, ence.” We are all familiar with the ‘environmental disasters’ associated with the in- MI 48824. troduction of exotic species. In many instances, these species were intentionally

[email protected] introduced, such as Rhododendron in the U.K., pine in Australia, kudzu in the southeastern U.S., and purple loosestrife in eastern North America (Keeler 1988; Karen Hokanson is with Mooney and Drake 1986; Crowley 1997). Others arrived on their own, such as the Dutch elm disease and corn leaf blight in North America. The vast majority the Department of Horti- of introduced organisms perish or don’t establish self-sustaining populations cultural Science, Univer- (Pimentel et al. 1989), but we keep being drawn to those that do. sity of Minnesota, St. HARACTERISTICS OF NVASIVE PECIES Paul, MN 55108. C I S

[email protected] So, what does make a species invasive? To answer this question, we first need to define what we mean by invasive. Probably the most common definition given 187 is the ability to increase when rare; introduced from other parts of North INVASIVENESS OF however, all successful species meet this America and 1% are interspecific hy- criteria. Crowley (1997) suggests that brids. Based on these results, Reichard AGRONOMIC AND invasive species should really be called developed a decision tree for accep- FORESTRY SPECIES “problem plants,” where the species has tance of exotic woody species into passed some threshold of abundance North America, which begins with the In fact, only a small percentage of and someone is concerned. He suggests question “Does the species invade else- agronomic and forestry crops are im- that to understand the population bi- where, outside of North America?” portant weeds outside of agro-environ- ology of an invading plant genotype, Two other important questions in the ments (Table 1). They rely on human we need knowledge of the following: decision tree are “Is the species in a disturbances to become established and 1) abiotic environment, 2) the biotic family or genus with species that are rarely persist outside of specific habi- environment, 3) interaction between already strongly invasive in North tats. Clearly exceptions exist, such as biotic and abiotic environment, and 4) America?” And, “Is the species native barley, rapeseed, and rice, but over the year. In other words, the nature of to parts of North America other than 80% of all crop species do not persist invasiveness is a very complex situation. the region of the proposed introduc- Sarah Reichard recently published tion?” Other questions in the decision a series of papers presenting a frame- tree concern whether or not the spe- work for evaluating plant invasivenes cies is a sterile interspecific hybrid, rate Table 1. Survival of North American crops in native environments (Reichard and Cambell 1996; Reichard of vegetative reproduction, the length and Hamilton 1997; Reichard 1999). of the juvenile period, and germination Non-persistent Persistent/ Persistent/ Her decision tree is based on a predic- requirements. non-invasive invasive tive model derived from discriminant What her analysis indicates is that Beet Apple Barley and regression analysis of a number of a high percentage of the exotic species Broccoli Asparagus Rapeseed structural, life history, and biogeo- that become invasive are already excel- (Canola) graphical characteristics of introduced lent colonizers somewhere else and Carrot Blueberry Rice woody plants. The characteristics ana- their population size explodes when Cauliflower Cranberry Sorghum lyzed, most of which have routinely they are introduced into a new area Celery Pear Sunflower been implicated in association with where there are few to none of the Citrus Poplar Wheat Cucumber Spruce invasiveness, included: native range, natural constraints with which they Cotton Strawberry whether or not the species invades else- evolved. This is very different from the Eggplant where, leaf longevity, polyploidy, repro- situation facing transgenic forestry and Lettuce ductive system, vegetative reproduc- agronomic crops. They will not be re- Maize tion, minimum juvenile period, length moved from the complex array of Melon of flowering period, flowering season, natural constraints that currently face Onion length of the fruiting period, fruiting them, and only a very limited number Pea season, dispersal mechanism, seed size of these constraints will be removed by Peanut Pepper and seed germination requirements. the addition of a new trait through Potato Of the woody plants that invaded genetic engineering. The array of fac- Soybean the United States, Reichard found that tors regulating natural populations Squash 54% invade other parts of the world, must be complex, as the introduction Sugarcane 44% spread by vegetative means, most of single biological control agents have Sunflower have shorter juvenile periods, and 51% rarely had much of an impact on in- Tobacco have seeds that germinate without pre- vasive, exotic species (Pimentel et al. Watermelon treatment, while only 3% have been 1984). (Source - Hancock et al., 1996)

188 in native environments. Crawley et al. and long seed dispersal, discontinuous unspecialized pollinators, variable seed (2001) have generated some excellent germination, long lived seed, vigorous dispersal distance, high seed produc- evidence of how poorly crop genotypes vegetative reproduction, rapid growth tion, seed production in many environ- do in native environments whether to flowering, brittle propagules, con- ments, vigorous vegetative propagation, they were genetically modified or not. tinuous seed production, vigorous brittle propagules, and polyploidy. When they compared the performance competitors, self-compatible, However, they are outcrossing, have of transgenic and non-transgenic rape, unspecialized pollinators, very high discontinuous seed production, short maize, beet and potato in 12 native seed output, plastic seed production seed longevity, narrow germination re- environments, the genetically modified and polyploidy. When Keeler (1989) quirements, discontinuous germina- plants were never found to be more took Baker’s weediness traits and com- tion, are weak competitors, and grow invasive or persistent than their ante- pared the worst weeds to agronomic slowly. cedents. In fact, all populations of crops she found that serious weeds This suggests that in most agro- maize, rape, and beet were extinct af- possessed an average of 81% of these nomic and forestry crops, a whole syn- ter 4 years, and only conventionally traits, while random non-weeds had drome of traits would need to be al- bred potatoes were left after 10 years 59% and crop plants had 42%. tered through genetic engineering to (and only at one site). The transgenic To date, eleven tree crops have make them invasive; and Baker’s list rape and maize expressed tolerance to been genetically engineered in the excludes most biotic controls. Because the herbicide glufosinate, the geneti- United States and tested in the field: agronomic crops are often poor com- cally modified sugar beet were resistant apple, papaya, citrus, persimmon, pear, petitors in nature, their impact on na- to glyphosate and the transgenic pota- plum, pine, poplar, sweetgum, spruce, tive populations has also been gener- toes expressed either the insecticidal Bt and walnut. When they are rated ac- ally limited due to introgression. There toxin or pea lectin. cording to Baker’s characteristics, they are numerous instances where hybrid- In his classic work, Baker (1965; all fall well below the random non- ization with wild relatives has increased 1974) associated a complex array of weeds, ranging from 21 to 50% (Table the weediness of the native species in traits with colonizing ability including: 2). Poplar has the highest average of agronomic fields through crop mimicry broad germination requirements, short 50%, possessing the weediness traits (Ellstrand et al. 2000), but there is little

Table 2. Weediness traits in transgenic trees that have been field tested in the United States.

Weediness trait Apple Papaya Citrus Persimmon Pear Plum Pine Poplar Sweetgum Spruce Walnut

Broad germination requirements no no no no no no no no no no no Discontinuous germination no no no no no no no no no no no Long lived seeds (>5 years) no no no yes no no yes no yes yes yes Rapid growth no yes no no no no no no no no no Continuous seed production no no no no no no no no no no no Self pollinated no no yes no no no no no no no no Unspecialized pollinators no yes no no no no yes yes yes yes no High seed output yes yes yes yes yes yes yes yes yes yes yes Seeds produced in many habitats yes no no yes yes yes yes yes no yes no Short and distant seed dispersal yes yes yes yes yes yes yes yes yes no no Vigorous vegetative reproduction no no no no no no no yes yes no no Brittle propogules no no no no no no no yes no no no Vigorous competitors no no no no no no no no no no no Polyploid (2n > 28) yes no no yes yes no no yes yes no yes % Weedy traits 28 28 21 36 28 21 36 50 43 28 21

189 evidence of crop genes effecting the native populations should it escape. neered counterparts. Familiarity allows overall fitness of a native species. Even In fact, it is much easier to pre- regulators to efficiently assign levels of though crop species have been planted dict the environmental risk of risk, without doing any additional ex- among their progenitors for thousands transgenic trees than an exotic intro- periments, when the phenotypic effects of years, we are not aware of any reduction, as the level of risk in of transgenes closely mimic conven- port where the native fitness of the wild transgenics can be measured by evalu- tionally deployed or native genes. species was noticeably changed. When ating the fitness impact of a single en- Hokanson et al. (2000), outline a David Duvick (2000) ask a group of gineered trait, rather than a whole syn- number of examples where transgenic 20 experienced plant breeders if the drome of potentially invasive traits. A genotypes have similar non-transgenic introduction of conventional resistance unique genotype is not being intro- phenotypes such as insect and virus genes has led to undesirable conse- duced into an environment where its resistance. quences with respect to the weediness native constraints are removed. The This approach was what was rec- of a crop or its relatives, the breeders species is already in that environment ommended by the first group of sci- knew of no example. and we know how invasive it is. What entists who evaluated the environmen- we need to worry about is whether the tal risks of transgenic crops. In the of- addition of a single gene will increase ten cited paper of Tiege et al. (1989), Predicting the its existing level of invasiveness to prob- they state “transgenic organisms should Environmental Risk of lem levels. An increase in vegetative be evaluated and regulated according GMOs reproduction, a decrease in the need for to their biological properties (pheno- pretreatment requirements, or a short- types), rather than according to the It has been suggested that geneti- ened juvenile period could certainly genetic techniques used to produce cally engineered trees pose significantly raise red flags concerning invasive po- them . . .” and “Long term experience greater environmental risks than do tential. But these alterations are cur- derived from traditional plant breeding genetically engineered food crops, be- rently no more likely to be accom- provides useful information for the cause the genes inserted into trees are plished through genetic engineering evaluation of genetic alterations simi- more likely to ‘escape’ into the wider than they are through traditional ge- lar to those that might have been pro- environment (Campbell 2000). Plan- netic improvements. If these character- duced by traditional means, and such tation trees have been altered through istics are the subject of any research alterations are likely to pose few eco- breeding far less than have most agro- efforts toward genetic improvement, logical problems.” One of the major nomic crops, and as a result, are much they should bear close scrutiny for their conclusions of the National Academy more closely adapted to native habitats effects on invasiveness of the species. of Sciences report on “Field testing than are most crop species. However, In some cases, the risk involved in genetically modified organisms: Frame- most are not highly invasive in their the deployment of these transgenes can work for decisions” was that crops native geographic range, and the be efficiently evaluated through the modified by genetic engineering will transgenic derivatives and any native/ concept of familiarity (Hokanson et al. pose risks that are no different from engineered hybrids will be subjected to 2000). APHIS now assesses risk based those modified by classical genetic the complex array of factors that nor- on the biology of the crop, the nature methods. mally regulate the native populations. of the introduced trait, the receiving The problem with using the con- The bottom line in assessing the envi- environment and the interaction be- cept of familiarity is finding genes of ronmental risk of both transgenic trees tween these. Knowledge of these fac- equivalent effect and strength in natu- and herbaceous crops is the nature of tors provides familiarity, which allows ral populations. Reasonable arguments the transgene, i.e., how significant an decision makers to compare genetically can be made for many of the trangenes impact will it have on the fitness of engineered plants to their non-engi- that are similar to conventionally deployed resistance genes, but numerous

190 other engineered genes will produce grown in their original environment Hancock, JF, R Grumet and SC phenotypes that are unique to the spe- with its complex array of natural con- Hokanson. 1996. The opportunity for escape of engineered genes from cies or have broader effects than the straints. Normally, only one of these transgenic crops. HortScience 31: native genes. Some of these transgenes constraints will be removed by the ad- 1080–1085. are likely to be effectively neutral in the dition of a new trait by genetic engi- Hokanson, K, D Heron, S Gupta, S native environment, such as herbicide neering. The risk of most transgenes Koehler, C Roseland, S resistance, but others that alter repro- deployment can often be effectively Shantharam, J Turner, J White, M ductive potential and physiological tol- predicted by considering the pheno- Schechtman, S McCammon, and erances may have much more signifi- type of the transgene and the overall R Bech. 2000. The concept of fa- cant impacts. Regardless, it is much invasiveness of the crop itself. miliarity and pest resistant plants. easier to assign risk to transgenic crops Pp.15–20 in: PL Traynor and JH Westwood, eds. Ecological Effects of than exotic species, as we can restrict Pest Resistance Genes in Managed our worry to the effect of one gene on REFERENCES CITED Ecosystems. Information Systems the fitness of a species in the place it for Biotechnology, Blacksburg, VA. is already grown, rather than making Baker, HG. 1965. Characteristics and Kareiva, P, IM Parker, and M Pascual. modes of origin of weeds. In: HG guesses about the fitness of a whole 1996. Can we use experiments and Baker and GL Stebbins, eds. The species genome in a unique environ- models in predicting the invasive- Genetics of Colonizing Species. Aca- ment. ness of GMOs. Ecology 77:1670– demic Press, New York. 1675. Baker, HG. 1974. The evolution of Keeler, KH. 1988. Can we guarantee weeds. Ann. Rev. Ecol. Syst. 5:1–23. CONCLUSIONS the safety of genetically engineered Campbell, FT. 2000. Genetically Engi- organisms in the environment. neered Trees: Questions Without The patterns of spread of invasive, CRC Critical Reviews in Biotechnol- Answers. American Lands Alliance, ogy 8:85–97. exotic plant species cannot be used to Washington DC. predict the environmental impact of Keeler, KH. 1989. Can genetically en- Crowley, MJ. 1997. Chapter 19: transgenic trees and agronomic crops. gineered crops become weeds. Bio/ Biodiversity. In: MJ Crawley, ed. technology 7:1134–1139 While it is true that some transgenes Plant Ecology, Blackwell Science Marvier, M. 2001. Ecology of will influence individual traits associ- Ltd., Oxford, UK. ated with invasiveness, numerous other transgenic crops. Amer. Scientist 89: Crowley, MJ, SL Brown, RS Hails, DD 160–167. natural characteristics of these species Kohn, and M Rees. 2001. Mooney, HA, and JA Drake. 1986. make single changes unlikely to sub- Transgenic crops in natural habi- Ecology of Biological Invasions of stantially alter their competitiveness. tats. Nature 409: 682–683. North America and Hawaii. Invasive species have almost always Duvick, DN. 2000. Consequences of Springer-Verlag, New York. been introduced somewhere where classical breeding for pest resis- National Academy of Science. 1987. they have few to none of the natural tance. Pp. 37–42 in: PL Traynor Introduction of Recombinant Dna- constraints with which they evolved, and JH Westwood, eds. Ecological engineered Organisms into the En- Effects of Pest Resistance Genes in and so they fill new niches and their vironment: Key Issues. National Managed Ecosystems. Information population numbers explode. In many Academy Press, Washington D.C. Systems for Biotechnology, cases, these species were already inva- Blacksburg, VA. Parker, P, and IM Kareiva. 1996. As- sive in their original habitats. This is sessing the risks of invasion for ge- Ellstrand, NC, HC Prentice, and JF very different than making a single netically engineered plants: Accept- Hancock. 1999. Gene flow and in- able evidence and reasonable change in a species already with mul- trogression from domesticated doubt. Biological Conservation tiple controls. Most engineered species plants into their wild relatives. 78:193–203. are poor colonizers and they will be Annu. Rev. Ecol. Syst. 30: 539–563.

191 Pimentel, DC, C Glenister, S Fast, and D Gallahan. 1984. Environmen- tal risks of biological pest control. Oikos 42: 283–290. Pimentel, D, MS Hunter, JA LaGro, RA Efroymson, JC Landers, FT Mervis, CA McCarthy, and AE Boyd. 1989. Benefits and risks of genetic engineering in agriculture. BioScience 39: 606–614. Reichard, SH. 1999. A method for evaluating plant invasiveness. Pub- lic Garden April, 1999. Reichard, S, and F Cambell. 1996. In- vited but unwanted. American Nurseryman 184: 39–45. Reichard, SH, and CW Hamilton.1997. Predicting invasions of woody plants introduced into North America. Conservation Biology 11:193–203. Strauss, S. 2000. Report of the poplar working group. Pp. 105–112 In: PL Traynor and JH Westwood, eds. Ecological Effects of Pest Resistance Genes in Managed Ecosystems. In- formation Systems for Biotechnol- ogy, Blacksburg, VA. Tiege, JM, RK Colwell, YL Grossman, RE Hodson, RE Lenski, RN Mack, and PJ Regal. 1989. The planned introduction of genetically engineered organisms: Ecological considerations and recommendations. Ecology 70: 298–315.

192 BREAKOUT SESSIONS

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 195-196. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Instructions to Leaders

Each of five breakout topics will be discussed by a subset of meeting participants in two independent one-hour sessions, A and B. These questions are proposed as guides. Focus on the questions your group considers most important. Please also identify additional issues and priority areas for research. Time is short. Please manage your time carefully so the group does not spend all its time on a small technical issue. Work closely with your recorder to take notes on important points. You will be asked to present the results of your session to the main group, and summarize them in written form for the proceedings (1–2 pages).

BREAKOUT 1 — SOCIAL/GLOBAL CONTEXT

A) How can plantation forests play a role in the sustainable production of wood and fiber to meet human needs? How can the use of GM trees in plantation forests contribute toward this goal or detract from this goal?

B) How is it ethical or unethical to use GM in plantation forests? If unethical, why (relative to conventional breeding, hybridization, and exotic tree species)?

C) What are the most urgent research needs to address scientific and public concerns? What are the pros and cons, goals or limitations of a moratorium on field and/or laboratory research?

D)In what specific ways are regulations for research and deployment of GM trees in the USA adequate, excessive, or in need of fundamental redesign with respect to the goals of both protecting the environment and enabling socially desired technical progress? What changes would you recommend?

BREAKOUT 2 — BIOLOGICAL CONTEXT/ BIODIVERSITY

A) How might GM trees that are herbicide resistant, insect resistant, sterile, or have modified wood have negative or positive net impacts on biodiversity at the stand level, landscape level, and/or global level? Is a broad conclusion possible and how is it useful? How do biodiversity implications differ for native vs. exotic species? What kinds of research are needed?

B) Assuming that GM trees in forestry will be used largely in intensive plantation systems, usually in conjunction with clonal propagation, how are the ecological issues that GM plantations present significantly different in magnitude than those already inherent in plantation systems (e.g., exotics, hybrids, intensive breeding, short rotations, intensive silviculture)?

195 C) What represents adequate testing and deployment for enances)? What represents adequate testing and deploy- GM trees grown on a 5–15 year rotation? How does this ment for GM trees with modified wood grown on a 5– depend on species, growth rate, propagation method, and 15 year rotation? culture environment? C) What are expected changes to nutrient cycles or trophic interactions with trees possessing GM wood, either in- BREAKOUT 3 — INVASIVENESS/ side or outside of plantations, and are they greater than those normally associated with variations in intensive EEDINESS W plantation silviculture?

A) How is the commonly made analogy between GMOs and intentionally introduced, invasive exotic species a BREAKOUT 5 — BENEFITS AND useful one, or a misleading one, considering the relative SAFETY OF PEST MANAGEMENT magnitude of their biological risks and potential for unpredictable impacts on the environment? APPLICATIONS B) What kinds of GM traits, if any, are expected to lead to A) Under what circumstances can insect resistance that results increased weediness, invasiveness, or dysgenesis? What from one or few types of transgenes (e.g., Bt), be consid- are the pros and cons that all GM trees, because of their ered sustainable in plantation forestry? Can transgenic sys- method of creation, be considered potentially invasive tems be considered sustainable, even if specific genes are until proven otherwise via experimentation? not (e.g., sequential uses of different transgenes)? In what C) Can the potential ecological risks for invasiveness be pre- kinds of plantation systems and species, if any, is major dicted with high reliability from knowledge of the ge- gene insect resistance an acceptable option? netic changes imparted without the need for long-term B) What non-target effects of pest resistance transgenes on (multigeneration?) field trials to assess invasiveness (e.g., insects and soil organisms, within or outside plantations, for domestication traits like short stature or reproduc- are likely to be great or modest (assuming tissue-specific tive sterility). promoters are employed)? How will they compare in D)For what kinds of GM traits, in what species and envi- magnitude to other impacts from intensive silviculture ronments, are fertility reduction systems highly desirable? of plantation forests (variation among species and geno- How effective must fertility reduction systems be (sta- types, weed control, density control, fertilization)? bility, efficiency), when needed, to give acceptably low environmental risks, and how should this be established? C) Are herbicide resistant trees expected to have significant What constitutes acceptably low risk? economic or environmental benefits? How can they be managed to provide healthier soil and aquatic systems than available with alternative means for weed control, BREAKOUT 4 — WOOD as well as provide economic benefits? MODIFICATION

A) What are the goals of wood modification? How do GM goals differ from those of conventional breeding?

B) What concerns posed by trees with GM wood are greater than those of genetically novel trees produced during conventional breeding (families, clones, hybrids, prov-

196 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 197-198. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Summary of Breakout Session 1A — Social / Global Context

Daniel B. Botkin [email protected] Jarmo Schrader [email protected] We began by considering question (C), what are the most urgent research needs to address scientific and public concerns? Next, we considered the question, what are the general goals, transgenic or not, related to forestry, from a societal point of view? The group listed the following as included in society’s goals for forests: more environmentally friendly ways to produce wood and fiber; increase and provide more efficient and environmental friendly ways to produce biopolymers; human welfare, including economic and social well being; pursuit of fundamental scientific knowledge; and enhancing the opportunities for people to contact nature. More specification, in terms of more environmentally friendly ways to produce wood and fiber, one would focus on: faster breeding; increased carbon sequestration in forests as part of the production of wood and fiber; improvement in wood quality; phyto-remediation; and understanding supply and demand for timber products. Societal considerations arise from these goals, and include: learning what are present public concerns; understanding what the public ought to know (creating an informed public); learning to what extent the public is informed; determining who benefits from the use of biotechnology for trees, who owns the inven- tions and property, and who bears the burdens; how to deal with public concerns about uncertainty. Societal uncertainty is of three kinds: (1) uncertainty arising from how much one can trust what scientists, corporations, government agencies and non-governmental organizations say; (2) uncertainty due to lack of scientific understanding; and (3) intrinsic stochastic properties and variation in trees and forest and plantation ecosystems. These societal considerations lead to a suggestion for specific research programs that include: ethical, legal and social issues (ELSI) programs for forest biotechnology—a funded research program based on prior ELSI programs. This would be a standard research program, with requests for proposals, and investigator-driven research. There would be government funding of research by scientific and other experts. ELSI also include public involvement and the development of educational, curriculum materials. In addition, research ought to be done to establish guidelines to determine risk and to determine how to do risk assessment specifically for biotechnological modifications of trees. And there is a need to understand human aspirations regarding forests, and to understand these beyond what one learns typical from public opinion polls as what the public believes. An understanding of human aspirations for forests also goes beyond but encompasses the specifics of forest biotechnology.

197 Third, research is needed on how to involve public decision-making related to biotechnology. Fourth, ecological research is needed to understand forest plantations as ecosystems, in regard to such qualities as biological diversity within plantations, the potential and limits of plantations to support specific endangered or threatened species, and to serve as habitat for biodiversity characteristic of a forest type.

198 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 199-200. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Breakout Session 1B — Social /Global Context

Trevor Fenning [email protected] Starting note: It was decided not to address question ‘D’ in relation to the regulation of transgenic trees in the United States, as this was a very detailed point. The bullet points indicate the main discussion points raised in relation to each question. Question A. How can plantation forests play a role in the sustainable production of wood and fiber to meet human needs? How can the use of GM trees in plantation forests contribute towards this goal, or detract from this goal? Where will the wood come from if there are no plantations?

•Existing (wild) forests will be spared.

•Plantations contribute to global sustainability.

•Is demand growing? (There was some disagreement about how much.)

•Plantations will allow increased production, if it is needed.

• GM needed? (No consensus). The distinction of ‘GM’ from other forms of tree biotechnology is probably artificial in relation to the major issue of plantations in general.

• Ownership of GM trees will be the big issue, especially in developing countries, or on plantations not owned by big corporations, i.e., who will plant them?

• Who benefits will also depend upon the traits the trees are modified with, as well as other details e.g., the locality of where they are planted.

• There will probably be less opposition to GM plantations if it is not only big corporations who benefit. For GM plantations to be accepted, the benefits must be seen to accrue to local populations, ‘consumers’ and the environment.

Value of remaining forests? That is, merely because they are no longer needed for wood supply, does not mean they will automatically be protected, especially if they lose value. If the issue of GM trees adversely affects public attitudes to plantations in general, that could be a big set-back to global sustainability. There is huge variation in trees in general, which can be utilized, but not in all species all of the time. To extend the use of trees there is no alternative in some cases other than to genetically modify them, especially in extreme situations—such as reducing desertification, or in restoring salty or polluted soils. Question B. How is it ethical or unethical to use GM in plantation forests? If unethical, why (relative to conventional breeding, hybridization, and exotic tree species)? Whose ethics? Definitions?

•If the GM trees are deemed ‘safe’ and bring public or environmental ‘good’, then they are by definition ‘ethical’.

199 • Long-term benefits are ethical. tent would be especially valuable, especially in developing countries, where fuel wood is a major need. • Ownership is again a critical issue, especially if the benefits accrue only to big corporations or remote govern- More industry-orientated goals (below) were also con- ment agencies, etc. sidered as offering legitimate benefits, but might be a ‘harder sell’ if proceeded with in isolation to projects of more obvi- •Global and local public benefits must be widely per- ous social benefits, as listed previously. ceived. •BioPharming with GM trees, e.g., production of phar- • Risks and Benefits—the situation must be avoided where maceuticals in rubber latex. profits and benefits are privatized, while the risks are socialized, i.e., society has to carry to costs of any errors •More rot-resistant trees (during growth or as timber?). or problems, while the benefits are not widely felt. •Trees resistant to more environmentally benign herbi- • The focus should be on forestry needs / needs which trees cides. can help, then identifying the most appropriate solutions, •Trees with reduced lignin content, or other properties i.e., GM should not be the primary issue. which make processing easier, or improve/maintain wood Question C. What are the most urgent research needs quality while simultaneously improving productivity. to address scientific and public concerns? What are the pros A global and local view of the problems, alternatives, and cons, goals or limitations of a moratorium on field and/ and costs is needed in each case: or laboratory research? Many valuable R+D projects were identified, which •Marketing may be needed. would be widely appreciated as in the public good (as below), and no fundamental objections were raised: •Scientists should show leadership, independence, and clarify choices on the matter of GM trees. If we feel as a • GM trees with improved productivity probably offer the scientific body that the benefits of GM trees can be most important overall social and environmental ben- gained safely, then we should say so, but care must be efits at the global level. e.g., improving the photosyn- taken not to slip into giving merely commercial opin- thetic efficiency of selected tree species. ions [the boundary is probably a personal one].

•Use GM to produce trees for combating environmental The IUFRO Tree Biotechnology group could organize degradation, especially desertification, growing soil sa- a number of working/advisory sub-groups to explore more linity in many areas and detoxifying contaminated waste carefully the requirements and feasibility of particular project sites. The benefits of GM trees do not have to be re- areas (as previously) or on particular issues. stricted to plantations! But . . . some problems and public concerns cannot be addressed by experiment and research, however! •Trees resistant to the impacts of global warming are needed, e.g., drought-resistant trees—even out of desert boundaries, drought problems are increasing in many areas.

•Resistance to exotic diseases is needed in many cases. There are some well-known examples, but there may be others in future.

•Produce GM trees better suited as alternative energy sources to fossil fuels, e.g., trees with increased lignin con-

200 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 201-202. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Summary of Breakout Session 2A — Biological Context /Biodiversity

Alvin Yanchuk The question of how might GM trees for herbicide or insect resistance, or modi- [email protected] fied wood characteristics, etc., have negative or positive net impacts on biodiversity at the stand, landscape, and global level, was the primary question put to the group. This is a large and perhaps all-encompassing question, and the group had three main discussion points. First, the improved productivity added by GM trees/ traits, if significant, could reduce local pressure on native wild forests. However, this may not be generally true, as wild forests may be providing other commodity values and may not be accessible for such forest management options. Sec- ond, it was clear that we need to focus in on the specific trait / scenario of the GM stand to evaluate what level or function of biodiversity we are affecting. Ef- fect of any type of GM plant/tree or stand on local ‘biodiversity’ is too large to be helpful on the specific organism interactions that may be under question. Fur- thermore, it is likely that most of the interest will be on the effects of GM trees and stands on local surrounding ecosystems (i.e., populations of related or unrelated species), rather than within GM stands. Biotechnologists and tree breeders will of course be interested on the performance of the trait and the stand, but this was considered of less interest compared to the larger question on the ecological and social concerns being expressed. The question of how do biodiversity implications differ for native vs. exotic species that may undergo GM transformation, was also put to the group. In general, it was expected to be that the use of exotics would be of less concern or impact, on biodiversity, than with native species. Reasons for this, were due to the likely reduction of cross pollination/contamination with related wild species. What kinds of research are needed to address more of these concerns, before there may be more scientific and public acceptance? The group thought it is important, again, to be specific on what types of GM traits are being proposed for study or release. For example, those that kill organisms, modify internal structures/chemistry, etc, will all require much different types of investigation. GM ecological impact studies, which need to be in place prior to deployment on commercial scales, should consider the many of the protocols currently present in crop research. How are the ecological issues that GM plantations present significantly different in magnitude than those already inherent in plantation systems? The group thought that there is not much difference and it is still primarily a question of how biodiversity is changed at the landscape level by how we manage plantations at this level. Although the specific GM trait in question may have a relatively smaller impact, there are nevertheless ecological issues that need to be considered. Research, therefore, could examine the effects of non-GM plantations on current landscape level biodiversity. Specific traits and specific organisms and populations could be targeted to evaluate evaluate effects (e.g., gene flow) of non- GM traits (or some markers). This may allow us to make inferences to GM tree stand impacts, without larger scale GM trait research. What represents adequate testing and deployment for GM trees grown on a 5–15 year rotation? There appears to be two issues: 1) testing the internal perfor-

201 mance of the GM traits in various lines, and 2) ecological impacts of the GM populations on adjacent wild populations. No single time frame was deemed appropriate as results will be evaluated, refined and re-evaluated over time and probably by a larger groups of people (i.e., an adequate outcome rather than length of time). Again, this will be trait and situation specific. In terms of ecosystem evaluation, may need to look at entire rotation time, plus more! Continued monitoring will likely be necessary until proven no longer needed.

202 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 203-204. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Summary of Breakout Session 2B — Biological Context /Biodiversity

Gerald Tuskan [email protected] John Hayes Question A. How might GM trees that are herbicide resistant, insect resistant, sterile, or have modified wood have negative or positive net impacts on biodiversity [email protected] at the stand level, landscape level, and/or global level? Is a broad conclusion possible and how is it useful? How do biodiversity implications differ for native vs. exotic species? What kinds of research are needed?

Positive:

•Species restoration

•Positive influences on biodiversity and environmental integrity, if replacing annual crops with perennial crops.

Negative:

•Introduction of exotic species may (or may not?) be accelerated by use of GM trees.

•For biosafety reasons, there may be reasons that deployment in exotics may be greater.

• GM trees may mitigate or increase potential for invasiveness.

It depends:

•If perception is that it benefits the public, concerns about environmental impact and unknown will be reduced.

Research needed:

•Baseline characterization (ecology of non-GM plantations, indicators, conservation targets and their food webs, outcrossing):

• (Alternative) Direct experimentation on research plots

•How many clones needed to be deployed?

•How do we deploy different types of management under different conditions (e.g., riparian, soil characteristics)?

•Research to support policy question regarding where could/should GM be deployed

•Monitoring for movement of gene flow, movement of exotics, diversity, etc.

•Quantify net diversity change resulting in change in forest condition from GM trees. 203 What characteristics could lead to:

• Catastrophic outcome (e.g., influence on endangered species)

•Population effects on characteristics of juvenile plants.

What are the implications of large areas in monotypic, genetically simple forests? How is natural gene flow influenced by use of transgenic plantations? How do GMOs influence ecosystem processes? Question B. Assuming that GM trees in forestry will be used largely in intensive plantation systems, usually in conjunction with clonal propagation, how are the ecological issues that GM plantations present significantly different in magnitude than those already inherent in plantation systems (e.g., exotics, hybrids, intensive breeding, short rotations, intensive silviculture)? In many ways the issue is not GM trees but clonal forestry. Question C. What represents adequate testing and deployment for GM trees grown on a 5–15 year rotation? How does this depend on species, growth rate, propagation method, and culture environment? It depends on what level of risk the society is willing to accept (distinguish between risk and meaningful risk)— needs to consider both the magnitude of the consequence and the probability of its occurrence. Risk will be higher for traits that are favored by natural selection (especially of juvenile plants), possibility of introgression into wild population, characters that influences multiple traits. Testing could be relatively short if effects of traits are likely to be reversible, assuming mechanisms to withdraw use at a later time.

204 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 205. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Summary of Breakout Session 3A — Invasiveness / Weediness

Jim Hancock [email protected] A) Is the commonly made analogy between GMOs and intentionally introduced, invasive exotic species a useful one, or a misleading one, considering the relative magnitude of their biological risks and potential for unpredictable impacts on the environment? The general consensus of the group was that the analogy is misleading. A single gene addition to a species already growing in an area is not the same as the introduction of an exotic whole organism. However, the group felt that the analogy is useful in identifying those traits that make species invasive.

B) What kinds of GM traits, if any, are expected to lead to increased weediness, invasiveness, or dysgenesis? The group felt that dramatic alterations in physiological tolerances such as drought or stress tolerance could lead to increased weediness, along with large changes in germinability and dormancy of seeds. But, it was pointed out that physiological trade-offs (pleiotropy) often occur in conjunction with these types of alterations that negatively impact on other fitness traits. The group was split on the risk of GM pest resistance. The environmental effects of GM resistance genes depend on how many species are controlled, whether phenotypically similar types of native resistance exist and how fast the pest evolves.

C) Can the potential risks for invasiveness be predicted with high reliability from knowledge of the genetic changes imparted without the need for long-term (multigeneration?) field trials to assess invasiveness. The group consensus was that long term trials can often be avoided if we have thorough knowledge of the biology of the species (its invasiveness) and the nature of the inserted gene. The group also felt that critical information on establishment and growth characteristics is often provided in the initial short-term trials.

D)For what kinds of GM traits, in what species and environments, are fertility reduction systems highly desirable? The group felt that it depends on the characteristics of the species, but fertility reduction should be incorporated along with any trait that will significantly impact on a species invasiveness, particularly if plantations are adjacent to native strands. The group was most concerned about dramatic changes in stress tolerance and the stacking of resistance genes in already invasive species. The risk of single gene GM pest resistance was thought to be dependent on the engineered species and its environment. Most alterations of growth and development were not considered to be great risks, as they were thought to have negative fitness tradeoffs (as do many of the resistance genes). Herbicide resistance was considered to be neutral in the natural environment. The point was made that we need to think globally; decisions on fertility reduction should be made by considering all the places in the world a GM species might be introduced.

205 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 206-207. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Summary of Breakout Session 3B — Invasiveness / Weediness

Don Doering [email protected]

A) Value of the Analogy The session attendees agreed that the analogy between GM trees and introduced invasives was a poor one and that an a priori assumption of treating a GM tree as an invasive is not warranted for many reasons. Chief among those reasons that was cited is the knowledge about and behavior of the genetic background of engineered trees and the low likelihood of the modification initi- ating all the features that confer weediness. It was also agreed, however, that the analogy does guide important questions that are valuable to ask of any GM tree in a regulatory context, in design of the transgenic tree, and in field trial design. The traits that confer weediness that were listed in the presentation earlier in the day provide a checklist or a decision tree with which to assess if the modification may alter key traits and if action must be taken.

B) Which traits? The attendees agreed that traits that might alter sexual or vegetative propagation in any way are of primary concern as were any traits such as growth and resistance that may provide a selective and competitive advantage to the modified tree. There was discussion if one can assume that pest or disease resistance could confer sufficient advantage that a tree might be invasive. There was some disagreement about whether there were cases in which a single pest or class of pest provided the constraint against weediness. This question appears to be one for future investigation or consideration. While it was broadly agreed that GM trees should not bear a presumption of weediness, whether a specific trait allows a tree to occupy a new ecological niche is highly dependent on context. It was repeatedly raised that what was true of a particular tree in one ecosystem or culture context may not be true in different environment. Sterility was a major focus of discussion with most participants in clear agreement that if a tree is sterile and incapable of independent vegetative propagation, invasiveness is highly unlikely.

C) Risk Prediction and Testing. Participants agreed that there is considerable risk in extrapolating the results of experimentation in closed systems with the transfer of those trees to open environments and even among different open environments. Experiments and trials must be explicitly designed and monitored to assess increased invasiveness or competitive advantage of the GM tree. Though there was agreement that risk could not be accurately predicted, the participants expressed that the combination of many risk mitigation strategies could bring risk to not only acceptable levels but to levels that would require only short- term testing. Those strategies include: 206 •Engineered reduced fitness (e.g., short height)

•Engineered sterility

•Redundant engineered mechanisms for sterility

•Active and adaptive management

•Monitoring protocols and metrics for invasiveness.

D) Fertility reduction and risk. This question did not stimulate a great deal of discussion. There was the widely held view that there can not be such a thing as zero risk. To the question of how effective must fertility control be, the answer seemed to be “as effective as it can be.” As with other issues, the question is highly dependent upon trait and context. It was pointed out that control of weediness via outcrossing could also be through the design of complex constructs that reduced the chance of productive transfer and recombination. The discussion of acceptable risk appeared to be beyond the scope of the dwindling time of the breakout session. Participants agreed that risk is a function of the probability of invasiveness as well as its consequence. Some participants agreed that the more important question is not to determine the consequences at a time of such little data but to ask what are the standards that we use to describe those consequences, e.g., environmental standards as well as social and economic descriptions of consequences.

207 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 208-209. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Summary of Breakout 4A — Wood Modification

Chung-Jui Tsai [email protected] A) What are the goals of wood modification? How do GM goals differ from those of conventional breeding? Discussion: The goals for wood modification through genetic engineering include the following traits:

• lignin content

• cellulose content and quality

• fiber length

• fibril angle

• wood density

•rot resistence

• number of knots

• wood extractives

General consensus of the group is that the goals of wood modification through GM are no different from those of conventional breeding. GM and conventional breeding work toward the same goals which are tailored to product needs (e.g., pulp and paper or solid wood). Major differences, however, are as follows:

• GM can provide faster gains (as opposed to long breeding cycles)

• GM can increase precision of the modification (as opposed to conventional crossings)

• GM can create higher degrees of variation, exceeding those achieved by breeding

• GM may be the only means for wood modification when conventional breeding is not feasible (e.g., in certain parts of the world, for certain species, or for certain traits that are naturally present, such as engineering S lignin in conifers)

• GM is an important research tool to advance our fundamental knowledge. Transgenic trees have proven to be a powerful tool in recent years to address basic questions.

The group feels that it is inappropriate to separate GM and conventional breeding approaches for comparison in considering tree improvement programs. The group agrees that GM and conventional breeding approaches complement each other and should be used together. This has been the case in many studies where conventionally bred elite clones are used for further improvement by GM.

B) What concerns posed by trees with GM wood are greater than those of genetically novel trees produced during conventional breeding (families, clones, hy-

208 brids, provenances)? What represents adequate testing C) What are expected changes to nutrient cycles or trophic and deployment for GM trees with modified wood interactions with trees possessing GM wood, either in- grown on a 5–15 year rotation? side or outside of plantations, and are they greater than those normally associated with variations in intensive Although GM offers many advantages over conventional plantation silviculture? breeding (see discussion in A), there are also concerns sur- Changes to nutrient cycles or trophic interaction cer- rounding GM wood. The major issue is risk of side effects, tainly are of concern with plantation trees possessing GM . including unknown public hazards, unknown trophic effects, However, when compared with intensively managed clonal unknown below ground effects on nutrient cycles and carbon plantations, the group failed to identify any significant dif- sequestration etc. However, it was pointed out that none of ference. The GM plantations are likely to be managed the these concerns is unique to GM trees. It was suggested that same way as clonal plantations (i.e., fertilization, irrigation GM trees with modified wood properties might gain easier etc). However, the issue bears complexity associated with public acceptance as the process would mostly likely involve the target traits, degree of variation, and species, and should plant genes. Using non-plant genes seems to cause uneasiness be dealt with on a case-by-case basis. Considering that these among the public. The group feels that whether using non- are not issues unique to GM trees but are also common to plant genes in GM research bears greater risk than using plant field GM crops, the group feels that GM researchers work- genes is unknown, and was not convinced that GM wood ing on tree improvement are held to a higher standard of would gain more public acceptance. Moreover, engineering accountability than agonomists using GMO. of rot resistance or cellulose biosynthesis will still likely involve the use of non-plant genes. Discussion regarding adequate testing and deployment of GM trees with altered wood quality suggested that the testing clearly has to go beyond evaluation of the targeted wood properties. Adequate testing may need to include evaluations of whole tree growth and development, crop vulnerability and ecological safety. Environmental factors also need to be taken into consideration, hence GxE testing may need to be conducted at multiple sites and times. There also needs to be an adaptive management plan allowing flexibility to deal with new information or unforeseen developments. Again, we felt that these issues are no different from the clonal plantation practices. As to the duration of the testing, it is an important and practical issue facing researchers, tree growers (forest industry), regulatory agents and public (with different risk consideration). Industry may not invest or be interested in GM trees at all if full rotation testing is required (not time-saving). However, short range, such as 5-year testing may not be acceptable to the public and regulatory agents. The appropriate length of testing may need to be determined according to the end use of the GM wood. For example, half- or quarter-rotation testing might be adequate for short-rotation wood crops for pulp and paper production, but testing until full rotation age might be necessary for solid wood applications.

209 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 210-212. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Summary of Breakout Session 4B — Wood Modification

Malcolm Campbell [email protected] This session set about to address the following three questions, and to iden tify additional issues and priority areas for research.

A) What are the goals of wood modification? How do GM goals differ from those of conventional breeding?

The goals of wood modification (be it by GM or conventional breeding) fell into two categories:

•Basic science goals—a tool for research into wood properties. Examples include

• candidate gene testing for QTL association studies

• determine effect of modified wood properties on end uses (e.g., pulping)

• determine impact of changes on one wood component on the quantity/properties of another wood component (e.g., lignin and cellulose).

•Applied goals—end-use properties are the drivers here. Examples include

• modification of the usual suspects: lignin, cellulose, chemistry (extractives / secondary metabolites)

• modification of morphology and development

•production of novel products.

The goals of GM and conventional breeding were viewed as being the same, but the means to the end, the rate of change, and the end product may be radi- cally different. In general, it was agreed that directed modification was a greater possibility with GM and that the rate of change could be faster with GM (a short- cut to modification relative to conventional breeding). Furthermore, it was agreed that GM has the potential to extend the range of modification beyond the ‘natural range’ of variation. For example, it is theoretically possible to modify gymnosperms by GM such that they produce syringyl lignin, which is not normally found in gymnosperms. Alternatively, GM angiosperm trees could be modified to produce entirely guaiacyl lignin, which has not been reported in natural populations of these trees. Research priorities: identification of genes involved in wood formation, identification of promoters to drive gene expression during specific times in wood formation, association studies to identify loci important in wood formation/properties.

B) What concerns posed by trees with GM wood are greater than those of genetically novel trees produced during conventional breeding (families, clones, hybrids, provenances)? What represents adequate testing and deployment for GM trees with modified wood grown on a 5–15 year rotation? 210 The short answer to former question is: “It depends”. production to interaction with the environment, which The potential problems associated with the growth, devel- is considered below) and industrial. It is clear that trade- opment or perception of GM trees that were identified in offs will be needed on both sides—as is the case for phar- this session were as follows: maceuticals, for example. This is best dealt with using current risk assessment protocols and incorporating them •Trees have characteristics that extend beyond the “nor- with “adaptive management” approaches. mal” range of natural variation; although, it remains to be determined what the “normal” range is for many traits. •Research priorities: epistasis analysis, development of risk assessment protocols for GM trees, field trials. •Unforeseeable problems in dealing with trees with “extreme wood” - difficult to deal with the unexpected. C) What are expected changes to nutrient cycles or trophic •Impact of pleiotropic effects—what else might be affected interactions with tree possessing GM wood, either in- by the modification? side or outside of plantations, and are they greater than those normally associated with variations in intensive It is important to point out that it was generally agreed plantation silviculture? that the concerns with respect to GM trees might be less As for question B above, the answer to this question is than those associated with conventionally bred trees, for the “It depends”. The supposition is that wood modification following reasons: would impact the following aspects of tree biology: •One can ‘track’ the introduced gene easily. •resource allocation might change •Single gene introductions, or, in the instance where • post-harvest biotic interactions might change multiple genes are introduced, there is a limit to the number of genes, and the genes are known. • inadvertent susceptibility to certain pests or pathogens may arise due to the modification. • The introduced genes and their direct effects are well- characterized and known. However, it was noted that these changes would not be unique to GM trees relative to trees with wood modifi- In this session, the following points were made rela- cations through conventional breeding, but the extent to tive to what constitutes adequate testing of GM trees: which these changes may occur might differ, particularly as •Epistasis analysis is required, but this might be facilitated the changes to wood properties become more and more “ex- by GM as one can ‘track’ the introduced allele easily. treme”. It was felt that increased invasiveness of trees with • The space and time required for testing are unlikely to modified wood properties was highly unlikely. The reason be different from those required for conventional breed- for this is that trees that are investing resources into wood ing, but, GM testing is likely to be more extensive. production are likely to do so with a trade-off to reproduc- NOTE: Following from this point, it was asked if it was tive development. That is, from an evolutionary perspec- really necessary to have more extensive testing of GM tive (i.e., not a tree health perspective), trees with modified trees. General consensus was that it was not really nec- wood are likely to have a lower fitness and, if anything, be essary and that tests equivalent to those used for con- less invasive. This is a point that needs to be emphasized to ventionally bred trees were likely to be substantial those concerned with the invasiveness of forest trees. Again, enough—at least from a scientific and industrial perspec- the issue here is not unique to GM trees and also applies to tive. trees which have been generated through conventional • There are two imperatives that were identified that drove breeding. testing: biological (including everything from plant re- While horizontal gene transfer is frequently raised as a

211 concern in some quarters, it is not realistic to expect that modifying wood properties would increase the likelihood of this already rare occurrence. There is a concern in some quarters that generating trees with improved wood properties/production will lead to intensification of clonal forestry and a change in management practices. It was felt that this was an issue that was not unique to GM trees and needed to be dealt with more broadly under the category of “plantation issues”. It is clear that empirical data are needed in order to address question C. It remains to be determined how these data are to be collected for GM trees, but it was felt that the collection methods should no different from those already used in traditional breeding efforts. Furthermore, it was agreed that data collection should deal with GM trees on a trait-by-trait basis, and not generically. It is important to note that one has greater control over the modified material when using GM. There is control during the process of producing the trees, which is laboratory-based. There is also excellent control at the level of the end-product, which can be readily and easily tracked. The features of GM are considered a substantial benefit when it comes to assessing and mitigating any risks that may arise with trees that have modified wood properties/production.

212 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 213. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Summary of Breakout 5A — Benefits and Safety of Pest Management Applications

Armand Seguin [email protected] Tree genetic engineering for increased pest resistance represents a topic with a significant list of concerns from the public point of view. This field of research has also been the subject of numerous discussions among scientists. The questions provided in the book of abstracts were not answered specifically; instead, the group discussed a few ideas provided by the chairperson. The first question was, Why use genetic engineering to control pests? The group discussed a number of reasons, and then selected the following as the most important:

1. Increased productivity—in the context of tree plantations, reducing losses will provide greater returns and increased quality.

2. Reduction of chemical uses—reduction of chemicals could also have a positive impact on biodiversity by preventing the destruction of non-target species.

3. Increased capability to respond to a specific pest—this aspect is important for maintaining biodiversity in the context of a major infestation by a specific pest.

4. Maintaining soil conservation—in this case herbicide resistance could provide an efficient way to prevent soil erosion.

The second question focused on the trait to introduce. Although this represents a case-by-case approach, some general conclusions can also be drawn (see ‘to do list’ below).

1. Could be simple and well known (e.g. Bt endotoxin)

2. Could be complex (metabolic engineering)

3. Materials that have a high turnover should be preferred.

Finally, the group discussed and created a ‘to do list’ of suggestions: In all cases the issues of environmental and economic viability should be prioritized. Once the economic needs are well established, solutions with current and biotechnical approaches should be evaluated. For each case, research on the impact on non-targeted species and potential effects on plant chemistry should be evaluated. It is also important to test the four reasons (proposed above) for using genetic engineering to control pests. In many cases, field trials are also needed to address specific questions related to issues such as recombinant protein turnover and weediness. Sterility should be viewed as an efficient approach to reduce the risk of gene movement from plantations with consequent risks for weediness or expanded non-target effects.

213 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 214-215. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Breakout Session 5B — Benefits and Safety of Pest Management Applications

Sarah F. Covert [email protected] We began by discussing two questions that were considered central to the controversy over transgenic plants and relevant to potential pest management approaches in particular:

1. Is the use of non-plant genes to modify trees inherently problematic?

The participants in the break out session expressed differing views in response to this question. Those who thought that this technology is inherently problematic cited its novelty and its difference from what happens in nature as the reason for their concern. Those who thought the use of non-plant genes in trees is not problematic pointed out that the occurrence of horizontal gene transfer across species boundaries is an analogous, naturally-occurring process, albeit it appears to occur at a very low frequency.

2. Are the risks of tree transformation with plant genes different than the risks of tree transformation with non-plant genes?

The participants believed that the perception of the non-scientific public is that transformation with non-plant genes is more risky than transformation with plant genes. If this is in fact true, then scientists and biotechnologists need to be cognizant of this fact, especially when explaining their work and discussing its perceived risks with the public. In general, the scientists in the room thought that the actual risks associated with either type of transformation are dependent upon the specific function of the transgene, or are unknown.

The second topic we considered was the sustainability of single-gene transgenic insect resistance strategies (e.g., the expression of Bt toxin). We identified the nondurability of this strategy in the field as the major challenge faced by it, although this problem is not exclusive to the transgenic deployment of pesticides. Potential solutions to this problem include the pyramiding of multiple pesticide genes in individual plants, the inclusion of refugia in the field, and the precise timing of pesticide expression (e.g., during the months of greatest insect attack and/or in the most susceptible tissues). It was pointed out that in trees, the pyramiding of resistance genes could only be achieved via transgenic approaches because the long generation times in trees would make pyramiding via repeated breeding cycles impractical. Transformation, therefore, provides a special advantage over breeding for improving pest or pathogen resistance in trees. Additional challenges faced by pesticide-expressing trees include unintended effects on the trees themselves and non-target organisms in the environment, and the fact that the controversy over transgenics has created a particularly high burden of proof for the value of such projects. The sequential use of pesticidal transgenes was considered a sustainable approach for long-term management, particularly because this does not differ significantly from more conventional pest and pathogen management schemes. Transgenic plants expressing pesticides were considered most

214 appropriate for species grown in short rotation plantations because the risk of transgene escape would be minimized in short rotations and because there would be an opportunity to deploy new genotypes at the end of the rotation if insect resistance had failed. Potential non-target effects on insects and soil microorganisms are likely to depend upon the characteristics of the transgene and are likely to be minimized as advances in engineering technology make it possible to target gene expression to specific tissues and time periods. It is currently difficult to say how transgenic insect resistance is likely to compare to intensive management practices in terms of non- target effects because we have a poor understanding of the effects of intensive management or conventional breeding on microbes and non-target insects. As a consequence this was identified as an area in need of further research. A potential unintended risk associated only with transgenic plantations is transgene escape, although breeding projects founded on crosses between distant relatives certainly could introduce novel genes into wild relatives when the hybrid progeny are planted in new locations. When considering the potential benefits of trees expressing herbicide resistance, we used poplar as an example case. In terms of economic benefits, herbicide resistant poplar would result in labor savings as well as cultivation savings (e.g., fewer herbicide applications and a reduced need for soil disturbance). Potential environmental benefits of herbicide resistant poplar include soil conservation, reduced water use due to reduced weed competition, and the use of environmentally benign herbicides. The only perceived risk associated with this technology was transgene escape. This was viewed as problematic because it might make it difficult to control wild plants with environmentally benign herbicides. Research needs in this area include assessment of the natural invasive tendencies of any species to be transformed, mechanisms for fertility control, and the development of a greater variety of benign herbicides. Appropriate regulations will help reduce fears about transgenics and will guide environmentally sound deployment, but it is unlikely that such regulations will be adopted worldwide. In addition, during the course of the discussion the point was made that the current scrutiny applied to transgenic plants might prompt a similar scrutiny of conventional breeding for plant improvement.

215

SUMMARY VIEWS

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 219. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Policy Perspective

Roger A. Sedjo [email protected] The objective of enlightened policies is to promote objectives that, in the broad context, should lead to improvement of the generate welfare, in cluding, but not limited to economic welfare. These objectives should are also appropriate for policies related to innovations in forest biotechnology. In the economic realm, policies are usually designed to promote income, employment and economic growth. Policies focusing on promoting innovation, investments, the efficient use of resources and an equitable income distribution would generally be regarded as contributing to the general welfare. However, attention is given to activities and/or products that may generate negative externalities, e.g., air or water pollution, or unrecognized health and safety risks. These include negative externalities that may be associated with the introduction of biotechnological innovations, including transgenic trees. Biotechnological innovations in forestry clearly have the potential over the long term of promoting enhanced economic well-being. They can lower costs, improve quality and make goods and services more accessible to humankind. Additionally, they can provide positive environmental services, including the provision of land restoration services to the rehabilitation of almost extinct species, such as the American chestnut. Furthermore, the evidence over the past several decades that wood harvested from plantation forests substitutes for wood that would have been harvested from natural and old-growth forests is compelling, despite the reluctance of some to acknowledge the evidence. Biotechnology can enhance this shift to planted forests. However, innovations may also involve risks and uncertainties. In forestry, major concerns with transgenics tend to focus on the possibility of unplanned and negative impacts on the natural environment. In order to better understand the nature and magnitudes of these potential risks, a regulatory approach has been created for forestry in which USDA APHIDS bears the responsibility for determining the extent of the externalities that may be present and ultimately the acceptability of a forest transgenic innovation. Under this system various types of tests and trials are undertaken, including field trials, to determine whether the innovation should be allowed to move to commercial applications. The procedure accepts, rejects, or requires resubmission of the innovation for commercial use. It could well be that some transgenic innovations are essentially riskless, and should readily be accepted for commercialization, while others involve risks so great that prudence would suggest delaying the innovation until more information is available and, perhaps, ultimately rejecting indefinitely the commercialization of the innovation. Finally, one may always ask if the procedures are adequate. However, it must be recognized that no amount of testing will remove all uncertainty. Ultimately the testing procedure requires an informed judgment of whether the testing is adequate and appropriate.

219 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 220. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Ethics Perspective

Paul B. Thompson [email protected] This symposium is evidence that scientists working on tree biotechnology are off to a very good start in addressing controversial issues and ethical responsibilities. The papers and working group sessions have brought a number of key topics to the forefront. It is really quite remarkable that these topics should be addressed so thoroughly at this early stage in the development of the science. I would like to offer just a few concluding comments in reaction to what I have seen here. First, scientists seem prone to two tendencies that cause problems both with respect to public receptivity toward technology and to the discharge of ethical responsibilities. One is a tendency to evaluate technology solely in terms of net outcomes. Clearly outcomes—costs, benefits and risks—matter a great deal. Yet we do not look with favor on others who are too quick to conclude, “the end justifies the means.” Even when the end does justify the means, we like to believe that others have given due consideration to values not easily characterized as subject to “trade-offs.” For example, there should be due consideration given to the intrinsic value of natural ecosystems, and affected parties should have an opportunity to participate in decision-making, and to give or withhold consent. It was distressing to see workshop groups moving toward trade-off rationalization after only a few minutes of discussion. The other tendency is to analyze controversy in terms of a distinction between real and perceived risk. While it is true that people can and do misjudge either the likelihood or the degree of hazard that is associated with the use of biotechnology, it is also true that the source of controversy can lie elsewhere. There may be different value judgments being made about how to understand the normative importance of uncertainty, for example, or about the socio-economic consequences of using technology. One should be sure that others share one’s values before presuming that different judgments of risk can be attributed to a mistake about probability or hazard. Although the capabilities of bioethics should not be overstated, including philosophers or others with training in bioethics throughout the research and development process for technology is one way to hold such tendencies in check. One would hope that at least four or five bioethicists attach themselves to the emerging field of tree biotechnology early on, and that they are welcomed and included both at scientific meetings and at fora, such as this one, where social, ethical and public issues are the primary topic of discussion. A good way to make this happen would be for a few far-sighted deans to create positions in bioethics within forestry or environmental science programs at their universities, and to provide support for publication and teaching on the ethical issues of tree biotechnology. And may I conclude with the hope that this remarkable beginning becomes a standard practice for the emerging field of tree biotechnology.

220 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 221. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Industry Perspective

Alan A. Lucier [email protected] Maud Hinchee We extend sincere thanks to Steve Strauss and Toby Bradshaw for or [email protected] ganizing this important symposium and inviting our participation. Rex B. McCullough The potential benefits of forest biotechnology have been clarified and re-affirmed. [email protected] Good progress has been made in developing a more integrated community view of ecological and economic opportunities. The symposium has outlined important uncertainties and challenges in the future of forest biotechnology, with considerable emphasis on ecological concerns. Rather general discussions of ecological risks indicate a need for research focused on specific technology applications with careful attention to characteristics of the transgenic trees themselves; characteristics of environments in which transgenic trees might be deployed; and the design and expected effectiveness of risk reduction measures. Several speakers have provided useful insights into social, economic, ethical, and regulatory issues associated with forest biotechnology. These issues require substantial and sustained attention even though the path forward is often unclear and potentially treacherous. The complex implications of forest biotechnology seem to require new approaches and unconventional partnerships such as those envisioned by the Institute of Forest Biotechnology. We and others at the symposium have discussed the potential value of biotechnology to the forest products industry. It is clear that strategies and perceptions vary greatly among companies and stakeholders. In general, requirements for commercialization of forest biotechnology will include:

(a) Expectations of superior returns to shareholders with acceptable risk relative to alternative uses of capital

(b)Environmental performance will be maintained or enhanced

Further investments in R&D are critical to realizing the potential of forest biotechnology. An important economic hurdle is reducing costs associated with vegetative propagation of important softwood species such as loblolly pine. Ac- celerated research on ecological concerns and risk management options (e.g., flowering control) will be needed to satisfy environmental and social requirements. Research organizations, both public and private, have essential roles and enormous opportunities in forest biotechnology. In the United States, inadequate government support for pre-competitive research is a significant obstacle to progress.

221 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 222. www.fsl.orst.edu/ tgerc/iufro2001/eprocd.pdf Ecological Science Perspective

Kenneth Raffa [email protected]

It was widely accepted by presenters from a broad range of backgrounds that genetic engineering of plantation trees can offer some significant environmental benefits. These include reductions of pesticidal inputs, reduced pressures on wilderness areas arising from increased productivity of intensively managed plantations, and increased response capability to biological invasions. Whether these potential benefits can be realized will depend largely on whether potential risks can be managed. A common theme throughout many of the talks concerned the issue of scale: How can we extrapolate from short term experiments under controlled conditions to scientifically reasonable projections of long term consequences at the landscape level? This issue remains unresolved, and should be a major area of focus. Each person’s approach to this question reflects, to some extent, the level of biological organization at which they commonly work. Mo- lecular biologists often find that resolution of difficult scientific problems is achieved by deeper understanding of specific mechanisms, and by improved techniques for approaching intractable questions. Ecologists, in contrast, often find that resolution of difficult scientific problems is achieved by recognizing which factors originally perceived as outside their unit of study are in fact exerting strong feedback on the system. These experiences color the extent to which biologists working at different scales trust that extrapolations from laboratory and small field studies to long term and landscape projections can be made. In some ways, however, these differences offer an opportunity, by identifying how differing approaches can best be integrated. For example, ecological approaches can help identify what types of feedback processes might yield negative unintended consequences (biotype evolution, alteration of ecosystem processes, gene escape). But in many cases possible remedies to these concerns can be substantially improved by molecular methods (plant sterility, localized expression, exogenously triggered expression, etc.). Some immediate suggestions for improving the environmental safety of genetically engineered trees include:

1. Limit deployment to plantation trees, as opposed to suggestions (not made at this meeting) of using insect vectors or other means for naturally regenerated trees.

2. Limit deployment to sterile trees, and conditions under which spread of vegetative material can be prevented.

3. Employ biotype prevention tactics when pest resistance genes are employed.

4. Employ risk assessment procedures used to evaluate planned releases of biological control agents as a template.

5. Recognize that short-term risk assessment programs favored by current funding approaches bias our understanding in a direction that underestimates ecological risk.

222 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 223-224. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Forest Biotechnology Perspective Steve Strauss [email protected] H.D. (Toby) Bradshaw [email protected] Based on plenary lectures and discussions, we believe there was strong support from most meeting participants for the following conclusions: 1. Increasing human demand for wood and fiber will be increasingly met from intensively managed plantation forests. As the Earth’s human population grows by 50% (to 9 billion in 2050) and standards of living increase, there will be a commensurate rise in demand for the forest products. This wood and fiber must be produced in a manner that is economically and ecologically sustainable. Plantation forests, intensively managed with the best tools of modern agriculture—irrigation, fertilization, weed control, and genetic improvement—will supply much of the world’s wood needs and spare native forests by concentrating wood and fiber production, particularly that for industrial uses, on just 1%–10% of the land area now used for timber harvest. The growing science of genomics— where the structure and function of large numbers of genes are analyzed and compared across species—will provide many new opportunities for the use of genetic engineering to aid in the rapid domestication of trees, including increasing yield and the customization of woody feedstock qualities for various fiber and energy uses. 2. Novel aspects of genetically engineered plantation forests require long-term multidisciplinary field research. Intensive plantation forestry, while making use of many methods derived from agriculture, differs from agriculture in several important ways; the longevity of trees, their lack of domestication, and the frequent proximity of wild relatives are three such differences. There is a great deal of knowledge and experience to be gained from starting “medium-scale” experiments (tens to thousands of hectares) with GM trees in plantation forests. Such field trials would allow issues such as stability of trait expression, tree health, degree of genetic containment, and non-target effects to be monitored on ecologically relevant temporal and spatial scales. Risks and benefits could therefore be quanti- fied. As in many other forest research areas, the paradigm of “adaptive management,” where economic and ecological issues are examined during initial stages of use, and adjustments made to management based on results, will be important for GM trees if economic and ecological issues are to be studied adequately. 3. Fertility reduction will be important for many applications. Systems for fertility reduction (“sterility”) will be critical for many commercial uses of GM trees, to minimize gene flow into natural ecosystems. Because several options exist for achieving fertility control, mounting an aggressive research program, including long-term field trials with carefully chosen species, genes, and environments, seems warranted. Developing a partnership with ecologists, population geneticists, evolutionary biologists, regulators, companies, and interested environmental NGOs to assist in study design will be necessary and desirable. 4. Domestication traits pose less environmental risk. There are clear biological distinctions among most traits being considered for genetic engineering with respect to risk assessment. Some are clearly domestication traits, in the sense that they may improve productivity within tree farms but are highly likely to enfeeble

223 trees in the face of natural selection (and thus pose no risk of increased invasiveness). Sterility, dwarfism, and lignin modification are examples. Other traits may have benefits in wild populations, or reduce efficiency of human control; examples are insect resistance based on novel toxins, or herbicide resistance, respectively. For situations where significant wild populations are adjacent to plantation forests, gene flow of domestication transgenes pose little ecological concern, and thus do not warrant the same degree of empirical scrutiny as possible fitness- or weediness-related genes. For these genes, uncertainties can likely be resolved via adaptive management (see above). 5. Biological analogy between invasive exotics and transgenics is specious. There was strong consensus that the analogy between invasive introduced organisms and transgenic organisms is of little biological merit. Because of the vast differences in the degree of genetic and ecological novelty between novel species and transgenics with one or a few novel genes, the ecological risk from invasive exotics is much larger and less predictable than for transgenics with well-characterized genes. The well-established methodology for assessing risks of exotics, however, can be useful for helping to guide risk assessments of transgenic trees.

224 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 225. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Environmental NGO Perspective

Faith Campbell [email protected] Sue Mayer [email protected] Environmentalists hold a great diversity of views about GE trees; we are Mario Rautner presenting our personal views, not the “NGO position”. In this context, [email protected] we note the small number of environmentalist participants in this conference. The potential role of GE trees is part of a larger discussion about forestry and resource use issues. There are likely to be alternatives to the GE paradigm— alternatives that we urge be explored thoroughly. This conference represents a small step toward both reaching out to various viewpoints and exploring the wider resource use issues and alternative approaches. However, the statements this morning indicate that few have absorbed the qualms raised by ecologists speaking here. We hope that people will continue to think about the issues we and others have raised.

225

MODERATOR PERSPECTIVES Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 228-230. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Lessons from Twenty-five Years of Debate on the Risks and Benefits of Biotechnology

Michael T. Clegg [email protected] In the early 1970s it became apparent to the molecular biology community that it was soon going to be possible to put foreign (eukaryotic) genes into viruses or into bacterial plasmids to produce so called “recombinant DNA” mol- ecules. The significance of this feat was that it promised to provide a means to amplify genes from any source of interest (for example, humans) to greatly facilitate the study of gene structure and function. (Prior to this achievement it was virtually impossible to study gene structure in organisms more complex than viruses.) In the early - mid 1970s several laboratories were engaged in a race to succeed in ““gene splicing”” research. Some labs were trying to use fairly complicated chemical splicing approaches to introduce genes into viruses. However, the real breakthrough came with the discovery of restriction endonucleases (restriction enzymes) which made it possible to easily splice genes into bacterial plasmids. The bacterium of choice for this research was Escherichia. coli because much was known about the genetics of E. coli and its constituent plasmids. Very quickly plasmids of E.coli were engineered with antibiotic resistance genes as selectable markers to use in recovering plasmids with novel gene inserts. For example, one of the first such plasmids, pBR322, was constructed to carry resistance for tetra- cycline (tet) and ampicillin (amp). Thus insertion of a recombinant DNA mol- ecule into the tet gene yielded an E. coli colony that was tet sensitive and amp resistant, providing a quick screening method for recombinant plasmids. The irony was that all of this depended on exploiting years of research on E. coli, a bacterium that is a resident of the human gut. So, there was a lot of obvious concern about what risks this might pose for human health. Could recombinant DNA experiments inadvertently produce new pathogenic strains of E. coli that might pose a human health risk? Were there potential risks to the environment? Some even raised the question should we be creating entirely novel organisms that were outside the bounds of natural processes? In 1975 a major conference was organized under the auspices of the National Research Council at Asilomar in California. The organizers of that meeting were people like Paul Berg, Maxine Singer and others prominent in the research community. The question addressed by the conference was how to proceed with recombinant DNA research safely. The participants at the Asilomar conference identified measures to minimize risk by (1) focusing on ways to create attenuated strains of E. coli which could not live outside the laboratory. (The K12 strain of E. coli was quickly developed for this purpose.) And (2) agreeing on a regulatory framework to govern Federally funded research in the area of recombinant DNA technology. (There was very little private sector recombinant DNA research to be concerned about at that time.) As a direct consequence of the conference, a process was developed to regulate publicly funded research in recombinant DNA technology so as to prevent research that might pose threats to human health or to the environment. For example, certain kinds of experiments with recombinant DNA in viruses were

228 identified as too risky and it was agreed that these would not be pursued. Most publicly funded recombinant DNA research was supported by the National Insti- tutes of Health (NIH) and so the initial responsibility for organizing a regulatory system fell on NIH. The NIH created the RAC (Recombinant DNA Advisory) Committee to create levels of containment appropriate for different categories of experiments and to review the progress of recombinant DNA research with respect to safety. The RAC committee experience was successful and over the ensuing decade or so, the containment prescriptions which the original RAC committee developed for different kinds of experiments, were progressively relaxed as more was learned about the technology and its associated risks. For a young scientist like myself, the debate of the 1970s was very stimulating and intellectually exciting to observe, albeit from a distance. The debate raised many important questions about scientific responsibility and about how to anticipate some of the risks and impacts of new technologies, while at the same time allowing research to progress. Relatively quickly recombinant DNA research began to pay off with new approaches to human health. One of the first break- throughs was the engineering of genes for the production of human insulin, thus assuring a safe and adequate supply of this essential product for diabetics. Many other novel pharmaceutical products have since been developed and these have increased the standard of human health. Virtually from the start it was clear that recombinant DNA methods and their associated technical innovations might have a large potential impact in agriculture and forestry. Beginning in the mid-late 1980s, the USDA and other organizations began a series of workshops focused on the risks associated with the field testing and commercialization of transgenic crops. I had the opportunity to participate in four such workshops during this period. All of the potential risks associated with the applications of these technologies in agriculture were identified during this process and to my knowledge no new credible risks have been identified in the ensuing decade of discussion. Moreover, the National Research Council produced a series of influential reports on different aspects of the use and deployment of transgenic organisms beginning in the late 1980s and continuing to the present (yet another NRC report on the regulation of transgenic crops is due out in October of 2001). Despite this long history of study and debate, and despite the clear successes of biotechnology in pharmaceutical development, a segment of the public still harbors serious concerns about the impacts of these technologies in food production and in forestry. Somehow we have failed to convince a segment of the public that we can be trusted to manage these new technologies in ways that avoid harm to the environment or to human welfare. In short there appears to be a credibility gap. In seeking reasons for the persistence of the credibility gap it is instructive to compare some essential differences between the systems of science in biomedicine and in agriculture/forestry. To begin, there is a healthy balance in the biomedical area between public good research investment and private investment, which is markedly different from that in agriculture/forestry. The National Institutes of Health has a research budget of roughly 16 billion dollars a year largely devoted to investments in public good research. In contrast, investments in public good research in agriculture/forestry are an order of magnitude smaller and are dwarfed by private sector research investments in this sector. As a consequence, it may be more difficult for the public to accept that we as scientists are acting in the public 229 good rather than in the interests of private gain because we are more dependent on private investment for our research agenda. What are some of the other reasons that we are still debating these issues today, 25 years later? Perhaps another reason is that we have a regulatory appara- tus for the monitoring and development of transgenic crops and transgenic forestry products that is cumbersome. It is a three-agency agreement that fragments responsibilities and is difficult to understand. It is not transparent and hence it is not clear to the public that the regulatory system necessarily acts in the public interest. Finally, there is the big question that has come up repeatedly in this meeting, who benefits and who bears the risk? It is clear to all of us that we individually benefit from research in biomedicine; our lives have been extended, the range of diseases that pose threats to us and to our families has been diminished. We are generally a more healthy society today than we were a quarter century ago. Whereas it is not clear in the agriculture/forestry sector that we individually benefit from the research agenda. We have done a poor job of convincing the public that they will benefit from a wider range of useful products based on biotechnological innovation in the future. The challenge for the future is to address this fundamental question of public benefit.

230 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 231-232. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf The Urgent Need for Field and Laboratory Science to Inform Ethical and Policy Debates

Hal Salwasser [email protected] There is much that could be covered in a summary of this conference; I may miss some things that strike you as vital. But these are some points that resonated with me. I found four take-home points. I think these points could become what I call barrier busters, things needed to help us move to the next level of understanding. Take-home point No. 1: We need field research. We cannot get the next level of much-needed data on genetically modified organisms without empirical research, such as field trials, which must be long term and comprehensive. As several have pointed out at the conference, we thus far have very little data on field performance and safety. Thus, our discussions of the prospects for both benefits and risks are loaded with speculations based on concepts and lab work, too much speculation and not enough articulation as testable hypotheses. At a minimum, we need to begin recasting our speculations so they become testable hypotheses and then describe what we believe would be the evidence that would refute these hypotheses. We must take this fundamental step in the scientific method in order to move beyond a discourse of speculations based on concepts and theories and ideologies. When we do conduct the research, we must show people the data and use the data to continue the discussions. Take-home point No. 2: We need to question the universal appropriateness of the precautionary principle. We heard mention a number of times in the concluding presentations of the precautionary principle; if we do not know all the ramifications of something we should be very cautious in how we proceed. This is not the only option for public policy, though it is certainly an appropriate consideration. Many of the advances in our civilization have come from the opposite of a preoccupation with caution; they have resulted from a bias for boldness. Society must always ask questions such as, in what ways and under what conditions will we use biotechnology and genetically modified trees? At what point do we determine that our knowledge is sufficient, the risks acceptable, and the benefits equitable, such that we should proceed with field use of these organisms? At the same time, however, we face enormous natural resource challenges: how to equitably feed, clothe, and house current and future generations while simultaneously taking care of environmental health. While some will argue that we must wait until all potential outcomes of the potential roles for new technologies are perfectly certain, others must step forward to tackle these challenges boldly, to go where others have never been or are too timid to venture. We need a judicious balance between our tendency for caution and a bias for boldness. Perhaps it is time for the scientific community to open up the precautionary principle to some critical thinking about whether it is in fact the most prudent public policy in all natural resources issues ? does it serve well a dynamic world that is growing in numbers of people, many of whom are living under great inequities in a world that continues to be full of surprises? Take-home point No. 3: When do we go public? Is this the right time for public dialog on GMOs? I raise this issue because I’m not sure from the dialog that we have had this week and from all the other issues that people are talking about in the natural resources policy arena whether the time is really right for a visible, open public dialog on genetically modified organisms. This dialog is cer-

231 tainly appropriate eventually, but if we decide to move now into the public policy arena, then what are the message we are going to use to entice people to engage in a meaningful and constructive dialog? Who will carry the messages, how should we structure the dialog, what do we expect the outcomes to be? And what sort of grand visions do we use to stimulate people within that dialog? Maybe the time is right, but I would much prefer that we be at least to the stage of having testable hypotheses and proposed research trials rather than lots of speculations before opening the dialog. Whenever we open the public dialog, we need to be abundantly open and transparent about our assumptions and our unknowns. Take-home point No. 4: Clarify the role of science in policy. We must get a better grasp of the appropriate roll for science in what is inherently a socio-political process, that is the public policy process. Science informs policy choices; it tells us what’s going on, what’s possible, and what the likely consequences of choices might be. It does not and should not tell us what the choices should be, though individual scientists certainly will have opinions on that. Science leadership does not mean that scientists make the decisions all by themselves; the political process and the markets make the choices. We are, as Dr. Paul Risser said in his opening comments at this conference, at the intersection of many issues relevant to one of the most vital ecosystems for life: forests. Scientists cannot drive or decide the choices that people make about forests. On the other hand, we cannot have rational and prudent choices without science being a key player. So we must create complimentary roles for science and the policy process. Toward that end, I am intrigued by what Sue Mayer described as ‘multi-criteria mapping’. This may be the framework for getting scientific perspectives into the social and political discussion arena. These are main points that I take home from this conference. If we take them on, then maybe they could serve as the bridges to the next levels of understanding on the legitimate roles for GMOs in our future forests. But one final thought: I am really intrigued by what Dr. David Victor talked about, this idea of a great forest restoration. Worldwide, we are restoring forests while we preserve some and manage others to sustain desired conditions of environments, economies, and communities. So far we are doing this with existing and somewhat traditional technologies. Not all nations are on track with this restoration and are still losing forested area. Are genetically modified organisms essential to make this great forest restoration work, or are they still down the road somewhere, or do they fit at all? I think those are legitimate questions for all of us to consider as we leave this conference.

232 APPENDICES

Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 235-237. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Appendix 1 — Exit Survey of “EcoSocial” Symposium Participants

TABLE 1. SUMMARY OF SURVEY RESPONSES (N=136). Responses (%) Survey questions Agree Neutral Disagree No response

1. Intensively managed plantation forests (including hybrids and exotic 91 6 3 0 species) can make a significant contribution to environmental quality and sustainable production of wood and fiber for human use. 2. Sustainably harvested natural forests are not sufficient to meet global 76 15 9 1 wood demands in the next half century. 3. The specific traits of plantation forests, and not the method by which the 77 14 9 1 traits were produced (conventional breeding or GM), are the primary determinants of the risks and benefits of deployment. 4. GM of plantation forests is unethical, regardless of any scientific 1 4 94 0 consensus on risks and benefits. 5. There should be a moratorium on any field research with GM trees, even if 8 7 85 0 reproduction is prevented. 6. The current US regulatory framework for GM trees is adequate to insure 24 53 24 1 environmental safety and enable socially acceptable technical progress. 7. The dominating roles of patents and corporations in GM research and 54 26 20 0 application are major barriers to public acceptance. 8. The ecological issues that GM plantations present are different in detail, 72 13 16 1 but not in magnitude, from those common to intensive plantation systems. 9. Experiments on a large scale (plantation to landscape level) and long time 64 14 21 0 frame (at least equal to the age at harvest) are required for learning about the realistic levels of risks and benefits of GM trees. 10. No amount of research, or containment methods, can reduce the 10 4 86 0 ecological risks of GM plantations to an acceptably low level. 11. The ecological risks of GM trees should be assessed only by scientists 37 16 47 1 with no financial ties to industry or activist organizations. 12. GM trees are a threat to biodiversity at all spatial scales. 11 11 78 0 13. The risks for invasiveness from GM trees can be predicted well from the 36 33 31 0 nature of the traits imposed. 14. Traits that move qualitative aspects of trees out of the range produced by 40 40 21 0 natural selection will tend to domesticate, and thus can be assumed to be of little concern for increasing invasiveness of GM trees or progeny (e.g., short stature, modified wood, non-flowering). 15. Fertility reduction systems are needed to minimize spread of all types 58 11 31 0 of GM trees. 16. Single, exogenous genes for pest resistance, such as Bt, pose significant 27 26 46 0 risks for ecological damage. 17. Single or multiple sterility transgenes, used with the best available 63 28 9 0 transformation, gene expression-stabilizing, and field screening methods, can provide adequate fertility reduction for commercial uses of GM trees. 18. The analogy between GM and intentionally introduced, but invasive and 59 23 18 1 ecologically damaging exotic species, is an inappropriate one, considering the magnitude and uncertainty of their biological risks. 19. GM trees pose a significant threat to wild forests through their increased 7 15 78 0 invasiveness.

235 (Table 1. continued)

Responses (%) Survey questions Agree Neutral Disagree No response

20. GM trees with modified wood pose much greater concerns than trees 15 15 70 0 whose wood has been modified via conventional breeding (after both have performed favorably in half-rotation field trials). 21. Sufficient basic and applied research upon the genetic stability and 88 8 4 0 ecosystem behavior of GM trees, and upon the design of biological safety mechanisms, can create environmentally safe and societally beneficial trees and outcomes. 22. I am actively involved in research that is expected to facilitate some 62 6 32 1 commercial uses of genetically modified trees within a decade.

Yes No

23. I am a molecular biologist or forest biotechnologist. 78 22 24. I am a college student. 24 76 25. I have a Ph.D. 79 20 26. The conference was informative. 96 2 27. The conference focused on the important issues. 93 4 28. The lectures were well delivered and valuable. 93 3 29. The conference was well organized. 97 1

TABLE 2. DETAILED SURVEY RESPONSES.

Responses (%) Survey questions Strongly Strongly No agree Agree Neutral Disagree disagree response

1. Intensively managed plantation forests (including hybrids and exotic 54 37 6 2 1 0 species) can make a significant contribution to environmental quality and sustainable production of wood and fiber for human use. 2. Sustainably harvested natural forests are not sufficient to meet global 44 32 15 7 2 1 wood demands in the next half century. 3. The specific traits of plantation forests, and not the method by which the 52 25 14 7 1 1 traits were produced (conventional breeding or GM), are the primary determinants of the risks and benefits of deployment. 4. GM of plantation forests is unethical, regardless of any scientific 0 1 4 26 68 0 consensus on risks and benefits. 5. There should be a moratorium on any field research with GM trees, even if 3 5 7 22 63 0 reproduction is prevented. 6. The current US regulatory framework for GM trees is adequate to insure 4 19 53 20 4 1 environmental safety and enable socially acceptable technical progress. 7. The dominating roles of patents and corporations in GM research and 13 41 26 18 2 0 application are major barriers to public acceptance.

236 (Table 2. continued)

Responses (%) Survey questions Strongly Strongly No agree Agree Neutral Disagree disagree response

8. The ecological issues that GM plantations present are different in detail, 19 53 13 13 3 1 but not in magnitude, from those common to intensive plantation systems. 9. Experiments on a large scale (plantation to landscape level) and long time 27 37 14 17 4 0 frame (at least equal to the age at harvest) are required for learning about the realistic levels of risks and benefits of GM trees. 10. No amount of research, or containment methods, can reduce the 2 8 4 29 57 0 ecological risks of GM plantations to an acceptably low level. 11. The ecological risks of GM trees should be assessed only by scientists 12 25 16 33 14 1 with no financial ties to industry or activist organizations. 12. GM trees are a threat to biodiversity at all spatial scales. 4 7 11 30 48 0 13. The risks for invasiveness from GM trees can be predicted well from the 5 31 33 25 6 0 nature of the traits imposed. 14. Traits that move qualitative aspects of trees out of the range produced by 4 36 40 18 3 0 natural selection will tend to domesticate, and thus can be assumed to be of little concern for increasing invasiveness of GM trees or progeny (e.g., short stature, modified wood, non-flowering). 15. Fertility reduction systems are needed to minimize spread of all types 25 33 11 25 6 0 of GM trees. 16. Single, exogenous genes for pest resistance, such as Bt, pose significant 6 21 26 35 11 0 risks for ecological damage. 17. Single or multiple sterility transgenes, used with the best available 14 49 28 7 2 0 transformation, gene expression-stabilizing, and field screening methods, can provide adequate fertility reduction for commercial uses of GM trees. 18. The analogy between GM and intentionally introduced, but invasive and 19 40 23 11 7 1 ecologically damaging exotic species, is an inappropriate one, considering the magnitude and uncertainty of their biological risks. 19. GM trees pose a significant threat to wild forests through their increased 1 6 15 44 34 0 invasiveness. 20. GM trees with modified wood pose much greater concerns than trees 5 10 15 35 36 0 whose wood has been modified via conventional breeding (after both have performed favorably in half-rotation field trials). 21. Sufficient basic and applied research upon the genetic stability and 43 45 8 2 2 0 ecosystem behavior of GM trees, and upon the design of biological safety mechanisms, can create environmentally safe and societally beneficial trees and outcomes. 22. I am actively involved in research that is expected to facilitate some 31 31 6 11 21 1 commercial uses of genetically modified trees within a decade. 23. I am a molecular biologist or forest biotechnologist. 57 21 0 5 17 1 24. I am a college student. 3 21 0 11 65 10 25. I have a Ph.D. 69 10 1 5 16 3 26. The conference was informative. 66 30 1 1 1 1 27. The conference focused on the important issues. 53 40 3 1 2 0 28. The lectures were well delivered and valuable. 45 48 4 2 1 0 29. The conference was well organized. 68 29 1 0 1 0

237 Proceedings of the First International Symposium on Ecological and Societal Aspects of Transgenic Plantations, S.H. Strauss and H.D. Bradshaw, eds. College of Forestry, Oregon State University, 2001. pp. 238-250. www.fsl.orst.edu/tgerc/iufro2001/eprocd.pdf Appendix 2 — Registered Participants, Tree Biotechnology Symposia

Markku Aalto Claire Arnold (906) 487-2915 University of Helsinki Zentrum Fur Angewandte [email protected] POB 56 (Viikinkaari 5) Genetik Helsinki Fin-00014 Muthgasse 18 Wout Boerjan Finland A-1190 Vienna Vib Gent Belgium Laboratory of Genetics [email protected] [email protected] Ledeganckstraat 35 Tom Adams Tocee Ateeq-Ur-Rehman Gent 9000 Belgium Oregon State University Shakir Enterprizes [email protected] Forest Science 1-15R-20 Main Bazar Joerg Bohlmann Corvallis OR 97331 Kot Azam Khan (541) 737-6583 Kasur UBC Biotech Laboratory [email protected] Pakistan 6174 Universtiy Blvd. Rm 237 Vancouver BC V6T 1Z3 Waidi Alabi [email protected] Canada Vintage Agricultural Industry Awosanya (604) 822-0282 18A Branco Street 3rd Floor 25 Liwe Ikorodu (604) 822-6097 Lagos Lagos 234-1 [email protected] Nigeria Nigeria [email protected] [email protected] Ben Bonnlander Nature Conservancy Preston Aldrich Aladaba Adolf Kesse Kwasi Ayarkwa 156 Country Club Rd Hardwood Tree Improv. & Regen. Center College of Renewable Street Nat. Albany OR 97321 1159 Forestry Building Purdue Res. (541) 926-9698 University P.O. Box 214 (541) 926-9698 West Lafayette IN 47907 Sunyani B/A [email protected] Ghana (765) 496-7256 Michael Bosela (765) 496-7255 [email protected] Hardwood Tree Improvement [email protected] Suzanne Bertrand Purdue University Francis Alipoe NCSU Forest Biotech Center 1159 Forestry Bldg P.O. Box OS 2776 2500 Partners II Bldg West Lafayette IN 47907 Ghana Raleigh NC 27695 [email protected] West Africa (919) 513-0098 (919) 515-7801 Daniel Botkin Arie Altman [email protected] University of California The Hebrew Univ. Of Jerusalem P.O Box 30700 Faculty of Agriculture Rishikesh Bhalerao Santa Barbara CA 93130 P.O. Box 12 Umea Plant Science Centre (805) 452-3988 Rehovot 76100 Dept of Forest Genetics & Plant (805) 569-9170 Israel Physiology [email protected] [email protected] S-901 83 Umea Sweden Jean Bousquet Sara Andersson [email protected] Forest Biology Research Centre Umea Plant Science Centre Pavillon Marchand Dept. Forest Genetics & Plant Sara Blumer Ste. Foy Quebec Physiology Michigan Technological University Canada S-901 83 Umea 1408 Townsend Dr-SFWP (418) 656-3493 Sweden Houghton MI 49931 (418) 656-7493 [email protected] (906) 487-2912 [email protected] 238 Jim Boyle Steven Burke David Carey Oregon State University North Carolina Biotechnology USDA Forest Service College of Forestry Center 359 Main Rd Corvallis OR 97331 15 T. W. Alexander Drive Delaware OH 43015 [email protected] Research Triangle Park (740) 368-0187 NC 27709 (740) 368-0152 H. D. Toby Bradshaw (919) 549-9366 [email protected] Univ of Washington College of (919) 549-9710 Forest Resources John Carlson [email protected] Box 355325 Pennsylvania State University Seattle WA 98195 Victor Busov Forest Resources (206) 954-4392 North Carolina State University Ferguson Bldg (206) 685-1728 1025 Biltmore Hall University Park PA 16870 [email protected] Raleigh NC 27695 (814) 863-7561 (919) 515-7565 (814) 863-1357 Micheal Adusei Brobbey (919) 515-3169 [email protected] Kwame Nkrumah University of [email protected] Science & Tech. Jace Carson P.O. Box AY 210 Glaucia Barbosa Cabral Oregon State University Asuoyeboah Kumasi Aracruz Celulose S.A. 321 Richardson Hall Ghana Rodovia Aracruz/ Barra Do Corvallis OR 97331 Riacho Km 25 (541) 737-1390 Amy Brunner Aracruz ES [email protected] Oregon State University Brazil Reinhart J. M. Ceulemans Department of Forest Sciences [email protected] Corvallis OR 97331 University of Antwerp Dept of [email protected] John Cairney Bio. Insitute of Paper Science & Universiteitsplein 1 Steven Brunsfeld Wilrijk Be 2610 Technology Belgium University of Idaho 500 10th Street (323) 820-2256 Moscow ID 83844 Atlanta GA 30318 (208) 885-7211 (323) 820-2271 (404) 894-1079 [email protected] (208) 885-6226 (404) 894-4778 [email protected] [email protected] Joshi Chandrashekhar Michigan Tech University Nancy Bryson Faith Campbell Crowell & Moring Llp School of Forestry & Wood American Lands Alliance Products 1400 Townsend Dr 1001 Pennsylvania Ave NW 726 7th Street SE Houghton MI 49931 Washington DC 20004 Washington DC 20003 (202) 624-2529 (906) 487-3480 (202) 547-9120 (906) 487-2915 (202) 628-5116 (202) 547-9213 [email protected] [email protected] [email protected] Shujun Chang Rowland Burdon Malcolm Campbell New Zealand for Research Inst. Westvaco University of Oxford Private Bag 3020 P.O Box 1950 South Parks Rd Summerville SC 29484 Rotorua Oxford OX1 3RB New Zealand (843) 851-4837 United Kingdom [email protected] (843) 871-7185 [email protected] sxchang@westvaco. Com

239 Pierre Charest College Park MD 20742 Athens GA 30602 Canadian Forest Service (301) 405-4371 (706) 542-1205 580 Booth Street (301) 314-9308 (706) 542-8356 Ottawa Ontario K1A 0E4 [email protected] [email protected] Canada Marie Connett John Davis [email protected] Westvaco University of Florida Zenn-zong Chen Po Box 1950 P.O. Box 110410 Taiwan Forestry Research Summerville SC 29485 Gainesville FL 32611 Institute (843) 851-4865 (352) 846-0879 53 Nan-Hai Rd (843) 875-7185 (352) 846-0879 Taipei 100 [email protected] [email protected] Taiwan Bernadette Connors Christopher De Chant [email protected] Suny College of Envir. Science University of Cincinnati Hong Chen & Forestry Dept of Bio Sciences Canadian Food Inspection 1 Forestry Drive P.O. Box 210006 Agency Syracuse, NY 13210 Cincinnati OH 45221 59 Camelot Dr (315) 470-4748 (513) 556-9735 Nepean Ontario K1A 0Y9 (315) 470-6934 (513) 556-5299 Canada [email protected] [email protected] (613) 274-7374 (613) 228-6629 Peter Constabel Jeffrey Dean University of Alberta University of Georgia [email protected] Biological Sciences School of Forest Resources Vincent Chiang Edmonton AB T6G 2E9 Athens GA 30602 Michigan Tech University Canada (706) 542-1710 1400 Townsend Drive (780) 492-0412 (706) 542-8356 Houghton MI 49931 (780) 492-9234 [email protected] (906) 487-2959 [email protected] (906) 487-2915 Sarah Deumling R. James Cook Zena Timber [email protected] Washington State University 4550 Oak Grove Rd Michael Clegg 307 Johnson Hall Rickreall OR 97371 University of California; Pullman WA 99164 (503) 585-6380 Riverside (509) 335-3722 (503) 540-8937 Riverside CA 972521 (509) 335-9581 [email protected] (909) 787-4672 [email protected] [email protected] Linda Deverno Janice Cooke Canadian Forest Service Morris Cline University of Florida 580 Booth Street 12d Ohio State University PO Box 110410 Ottawa Ontario K1A OE4 1735 Neil Ave 134 Newins-Ziegler Hall Canada Columbus OH 43210 Gainsville FL 32611 [email protected] [email protected] [email protected]

Gary Coleman Sarah Covert University of Maryland Forest Resources 2102 Plant Sciences Bldg University of Georgia

240 Narinder Dhir Gary Dunning (206) 685-9995 Alberta Forest Service Yale University (206) 543-3254 9920-108 Street 9th Floor 360 Prospect Street [email protected] Edmonton Alberta T5K 2M4 New Haven CT 06511 Canada (203) 432-5966 Jean-marc Frigerio Inra (780) 422-5245 (203) 432-3809 (780) 427-0084 [email protected] Genetique et Amelioration Bp 45 [email protected] 33611 Gazinet Cedex Sarah Dye France Steve Di Fazio Oregon State University [email protected] Oregon State University 321 Richardson Hall 321 Richardson Hall Corvallis OR 97331 Jean Fruci Corvallis OR 97331 (541) 737-8491 Pew Initiative on Food & Biotechnology (541) 737-8487 [email protected] [email protected] 1331 H Street, NW Suite 900 Hiroyasu Ebinuma Washington DC 20005 Ronald Dinus Nippon Paper Industries [email protected] 2490 Goshen Rd 5-21-1, Oji, Kita-ku Mitsue Fukui Bellingham WA 98226 Tokyo 114-0002 (360) 966-4027 Japan Forest Products Research Inst. (360) 966-4027 [email protected] Kukizaki Tsukuba Science City [email protected] Dave Ellis Ibaraki 305-8687 Don Doering Director of Molecular Biology Japan World Resources Institute 244-8623 Granville St [email protected] 10 G Street Ne 8th Floor Vancouver BC V6P 5A1 Washington DC 20002 Canada Barry Goldfarb (202) 729-7655 (604) 207-0633 North Carolina State University Dept of Forestry Box 8002 (202) 729-7637 (604) 207-0692 [email protected] [email protected] Raleigh NC 27695 (919) 515-4471 Carl Douglas Trevor Fenning (919) 515-3169 University of British Columbia Max Planck Inst. For Chem Ecol [email protected] Dept of Botany 6270 University Carl Zeiss Promenade 10 Blvd. Jena 07745 Sonia Cm Goncalves Vancouver BC B6T 1Z4 Germany IBET / ITQB / UNL Apartado 12 Canada [email protected] (604) 822-3554 Oeiras 2781-901 (605) 882-6089 Metthias Fladung Portugal Institute of Forest Genetics [email protected] [email protected] Sicker Land Str 2 D-22927 Esteban Roberto Gonzales Dale Draper Grosshausdorf BC Ministry of Forests Germany Universidade De Sao Paulo Tree Improvement Branch 3rd mfladung@uni_hamburg.de Av Padua Dias N 11, Cp83 Piracicaba-Sp 13400-970 Floor 712 Yates Street David Ford Brazil Victoria BC College of Forest Resources [email protected] University of Washington Canada [email protected] Seattle WA 98195

241 Milton Gordon College Station TX 77843 [email protected] University of Washington (979) 845-2971 John Hayes Dept of Biochem Box 357350 (979) 845-3272 Seattle WA 98195 [email protected] Oregon State University (206) 543-1769 Richardson Hall 301l Kyung-hwan Han Corvallis OR 97331 (206) 685-8279 [email protected] Michigan State University (541) 737-6589 Dept of Forestry 126 Natural (541) 737-1393 Michael Greenwood Resources [email protected] University of Maine East Lansing MI 48824 Berthold Heinze 5755 Nutting Hall (517) 353-4751 Orono ME 04469 (517) 432-1143 Institute of Forest Genetics [email protected] [email protected] Hauptstrasse 7 A-1140 Vienna Jacqueline Grima-Pettenati Jim Hancock Austria UMR CNRS/UPS 5546 Michigan State University [email protected] Pole De Biotechnologie Vegetale Department of Horticulture Bp 17 East Lansing MI 48824 David Heron USDA-APHIS Auzeville (517) 353-6494 Castanet Tolosan 31326 [email protected] 4700 River Road France Unit 147 [email protected]. Fr Ashraful Haque Riverdale MD 20737 Bangladesh Forest Research (301) 734-5141 John Gripp Institute (301) 734-8669 Li-cor Biosciences Sholoshahar [email protected] 4308 Progressive Ave Chittagong Lincoln NE 68504 Chittagong 4000 Heike Hertel (402) 467-0742 Bangladesh Instit. for Forest Genetics & Forest Tree Breeding (402) 467-0831 [email protected] [email protected] Eberswalder Chaussee 3 Dil Haque Waldsieversdorf Andrew Groover Ministry of Foretsry & Environ- D-15377 USDA Forest Service ment Germany 1100 West Chiles Rd Abdul Hamid Road [email protected] Davis CA 95616 Dhaka (530) 758-1060 Dhaka 1000 Magnus Hertzberg Umea Plant Science Centre (530) 758-1070 Bangladesh [email protected] Dept of Forest Genetics & Plant Scott Harding Physiology Marie-Claude Gros-Louis Michigan Technological Univer- S-901 83 Umea Laval University sity Sweden Forest Biology Research Center SFWP 1400 Townsend Dr [email protected] Sainte Foy Houghton MI 49935 Quebec G1K 7P4 Kathy Hetterick Canada Jeff Harford USDA Forest Service Li-cor Biosciences 1992 Folwell Ave David Gwaze 4308 Progressive Ave St. Paul MN 55108 Texas Forest Service Lincoln NE 68504 (651) 649-5172 TX A&M University Forest (402) 467-0742 (541) 649-5285 Science Lab (402) 467-0831 [email protected] 242 Takashi Hibino (206) 903-0262 Puyallup WA 98371 Forestry Research Institute (206) 903-0263 (253) 445-4522 Nobono 24-9 [email protected] (253) 445-4569 Kameyama, Mie, 519-0212 [email protected] Japan Keith Hutchison University of Maine Helen Jones [email protected] 5735 Hitchner Hall University of Oxford John Hickman Orono ME 04469 South Parks Road Deere & Company (207) 581-2827 Oxford 0x1 3rb 3300 River Drive (207) 581-2827 United Kingdom Moline IL 61265 [email protected] [email protected] (309) 765-3872 (309) 765-3807 Katsuaki Ishii Todd Jones Forest Products Research Insti Weyerhaeuser [email protected] Matsuanosato 1 Kukisaki-machi 32901 Weyerhaeuser Way S. Maud Hinchee Inashiki-gun Federal Way WA 98001 Arborgen Ibaraki-ken 305-8687 (253) 924-6031 180 Westvaco Rd Japan (253) 924-6970 Summerville SC 29484 [email protected] [email protected] (843) 851-4676 (843) 832-2164 Larry Jackson Megan Jones Applied Industry Analysis Ltd Southern Cross University [email protected] Po Box 5611 P.O. Box 157 Stanley Hirsch Wellesley Street Lismore NSW 2480 CBD Technologies Auckland 1036 Australia 2 Pekeris St Park Tamar New Zealand [email protected] Rehovot [email protected] Israel Keinonen Kaija Juan Jarmalillo-Corr University of Joensuu [email protected] Forest Bio Research Center Dept of Biology Yan Hong Pav. Marchand Cite Universitye 80107 Joensuu Institute of Molecular Laval Finland Agrobiology Ste-foy Quebec G1K 7P4 [email protected] 1 Research Link Canada National University of Singapore (418) 656-2131 Mee-Sook Kim Singapore 117604 [email protected] University of Idaho Moscow ID 83844 Republic of Singapore [email protected] Brian Johnson (208) 885-4386 Head of English Nature (208) 885-6226 Jianjun Hu Advisory Unit [email protected] Wan Shou Shan English Nature Yong Kim Beijing 100091 Taunton Ta1 5aa P.R. China United Kingdom Korea Forest Research Institute [email protected] brian.johnson@english- Omockcheundong 44-3 Suwon Kyunggido 441-350 nature.org.uk Susan Hurst Korea Tilligen, Inc Jon Johnson [email protected] 1000 Seneca Street Suite 200 Washington State University Seattle WA 98101 7612 Pioneer Way E

243 Anita Klein Denis Lachance Eleonore Lickl University of New Hampshire Canadian Forestry Service Hblva Rudman Hall 46 1055 Rue De P.e.p.s. Borschkegasse 16/19 College Road Po Box 3800 Sainte-foy Vienna A-1090 Durham NH 03824 Quebec G1V 4C7 Austria (603) 862-2455 Canada [email protected] (603) 862-4013 [email protected] [email protected] Terri Lomax Samuel Land OSU Program for the Analysis Ned Klopfenstein Mississippi State University of Biotech Issues USDA Forest Service Forestry Box 9681 2082 Cordley 1221 South Main Street MS 39762 Corvallis OR 97331 Moscow ID 83843 [email protected] (541) 737-5278 (208) 883-2310 (541) 737-3573 (208) 883-2318 Ma Latif [email protected] [email protected] Bangladesh Forest Research Institute Sholoshahar Terri Long Teiji Kondo Chittagong 4000 University of Georgia Forest Tree Breeding Center Bangladesh UGA Life Sciences Bldg 3809-1 Ishi, Juo [email protected] Dept. of Genetics Taga, Ibaraki 319-1301 Athens GA 30602 Juha Lemmetyinen Japan [email protected] [email protected] Universtiy of Joensuu Po Box 111 Carol Loopstra John Kough 80101 Joensuu Texas A&M University USEPA Finland Dept of Forest Science Ariel Rios Bldg 1200 [email protected] 2135 Tamu Pennsylvania Ave NW College Station TX 77843 Bailian Li Washington DC 20004 (979) 862-2200 (703) 308-8267 North Carolina State Universtiy (979) 862-4790 (703) 308-7026 Dept of Forestry Box 8002 [email protected] Raleigh NC 27695 [email protected] (919) 515-3168 Alan Lucier Sandeep Kumar (919) 515-3169 NCASI Institute for Forest Genetics [email protected] P.O Box 13318 Sieker Landstrasse 2 RTP NC 27709 Laigeng Li Grosshansdorf D-22927 (919) 558-1999 Germany Michigan Tech University (919) 558-1998 [email protected] 1400 Townsend Drive [email protected] Houghton MI 49931 Carlos Alberto Labate (906) 487-2293 Helene Lundkvist Universidade De Sao Paulo (906) 487-2915 The Swedish University of Ag Av Padua Dias N 11 Cp83 [email protected] Science Piracicaba-Sp 13400-970 Sull Dept of Ecology Jingyi Li Brazil Se 750 07 Uppsala [email protected] Oregon State University Sweden 321 Richardson Hall [email protected] Corvallis OR 97331 (541) 737-8491 [email protected] 244 Caiping Ma (325) 470-6956 Charles Michler OSU Dept of Forest Science [email protected] Purdue University 321 Richardson Hall 1159 Forestry Building Corvallis OR 97331 Claire McCallum West Lafayette IN 47907 (541) 737-8486 Tilligen (765) 496-6016 1000 Seneca Street Ste 299 [email protected] (765) 496-7255 Seattle WA 98101 [email protected] Joanne MacDonald (206) 903-0262 Natural Resources Canada (206) 903-0263 Celia Miguel P.O Box 4000 [email protected] Instituto De Biologia Experime Fredencton NB E3B 4R3 Apartado 12 Canada Susan McCord Oeiras 2781-901 (506) 452-4210 Institute of Forest Biotechnology Portugal 15 TW Alexander Dr Research (506) 452-3525 [email protected] [email protected] Triangle Park NC 27709 (919) 549-8889 Lawrence Miller Stephanie Martel (541) 549-9710 Boise Cascade Universtiy Quebec a Montreal [email protected] P.O Box 500 Cp 8888 Succ Wallula WA 99363 Centre-ville, Rex McCullough (509) 544-6535 Montreal QC H3C 3P8 Weyerhaeuser (509) 545-9964 32901 Weyerhaeuser Way S Canada [email protected] (514) 987-3000 Federal Way WA 98001 (253) 924-6115 Mahmoud Mohamed (514) 987-4647 [email protected] (253) 924-6736 National Research Center [email protected] El-Tahrir Street-Dokki Etsuko Matsunaga Cairo Nippon Paper Industries Richard Meilan Egypt Oregon State University 5-21-1, Oji, Kita-ku [email protected] Tokyo 114-0002 321 Richardson Hall Japan Corvallis OR 97331 Rozi Mohamed (541) 737-6097 Oregon State University [email protected] [email protected] Dept of Forest Science Susan Mayer Corvallis OR 97331 Genewatch UK Eduardo Jose Mello (541) 737-8479 The Mill House Manchester Rd Cia Suzano De Papel D Celulose (541) 737-1393 Av. Brigadeiro Faria Lima Tideswellbuxton-Derbyshir Sk17 [email protected] 8ln 1335 6 Andar United Kingdom Sao Paulo, SP 01452-919 Nurul I. Mollah Brazil Government of Bangladesh (129) 887-1898 (129) 887-2531 [email protected] Sherebangla Nagor [email protected] Dhaka Scott Merkle Bangladesh Charles Maynard University of Georgia [email protected] School of Forest Resources Suny College of Envir. Science & Forestry Athens GA 30602 Michele Morgante 1 Forestry Drive (706) 542-6112 Dupont/University of Idine (706) 542-8356 1 Innovation Way Syracuse NY12310 (315) 470-6560 [email protected] Newark DE 19714 [email protected] 245 Thomas Moritz Olga Nicolaou Betty Pelgas Umea Plant Science Centre Forest Genetics-CHH University Laval Dept of Forest Genetics & Plant 1135 Arawa St 990 Bourlamaque #710 Physiology Rotorua Quebec G1R 4W4 S-901 83 Umea New Zealand Canada Sweden [email protected] (418) 648-3105 [email protected] (418) 649-6737 Ove Nilsson [email protected] Alison Morse Swedish University of Ag University of Florida Sciences Gary Peter P.O Box 110410 Umea Plant Science Centre Institute of Paper Science & Gainesville FL 32605 Umea Technology (352) 846-0881 Se-90183 500 10th Street NW (352) 846-1277 Sweden Atlanta GA 30318 [email protected] [email protected] (404) 894-1081 (404) 894-4778 Nicholas Muir Kabwe Nkongolo [email protected] Temple-Inland FPC Laurentian University 229 N Bonovie St 935 Ramsey Lake Rd. Paula Pijut Jasper TX 75951 Sudbury Ontario P3E 2C6 USDA Forest Service (409) 384-3434 Canada 1992 Folwell Ave (409) 384-5394 (705) 675-1151 St. Paul MN 55108 [email protected] (705) 675-4859 (651) 649-5172 (651) 649-5285 David Mycock Stephen J. B. O’Leary [email protected] University of Witwatersrand University of Victoria School of Animal, Plant and P.O. Box 3020 Gilles Pilate Environ. Sciences Victoria BC V8W 3N5 INRA-UAGPF Bag 3 Canada Avenue De La Pomme De Pin Johannesburg 2050 [email protected] Ardon Olivet 45166 South Africa France Shinitiro Oda [email protected] [email protected] Cia Suzano De Papel D Celulose David Neale Av. Brigadeiro Faria Lima Anne Plovanich-Jones Institute of Forest Genetics 1335 6 Andar Michigan State University Dept. Environmental Horticul- Sao Paulo Dept. of Botany & Plant ture Sp 01452-919 Pathology UC Davis Brazil East Lansing MI 48824 Davis, CA 95616 [email protected] (517) 355-5076 [email protected] (517) 432-1090 Bryan Parkes [email protected] Neil Nelson Carter Holt Harvey Forest University of Minnesota Duluth- Genetics Linda Pope NRRI 1135 Arawa Street University of Maryland 5013 Miller Trunk Highway Rotorua Plant Science Building Room Duluth MN 55811 New Zealand 2102 (218) 720-4285 [email protected] College Park MD 20742 (218) 720-4329 (301) 982-2750 [email protected] [email protected]

246 Brett Poulis [email protected] Roland Schafleitner University of Victoria Austrian Research Center Paul Risser, President P.O Box 3020 A-2444 Swibersdorf Victoria BC V8W 3N5 Oregon State University Austria Canada 600 Kerr Ad Building [email protected] Corvallis OR 97331 (250) 721-8926 (250) 721-7122 (541) 737-4133 Konstantin Schestibratov [email protected] (541) 737-3033 Shemyakin & Ovchinnikov [email protected] Institute William Powell Science Ave 6 Antje Rohde Suny College of Environ. 142290 Pushchino Science & Forestry Vib University of Gent Belgium Moscow Region 1 Forestry Drive Dept of Plant Genetics Flanders Russia Interuniversity Syracuse NY13210 [email protected] (315) 470-6744 Gent 9000 (315) 470-6934 Belgium Axel Schmidt [email protected] Max Planck Insititute of [email protected] Chemical Ecology Jeanne Romero-Severson Kenneth Raffa Cal-zeiss-promenade 10 University of Wisconsin Purdue University D-07749 Jena 345 Russell Laboratories 1159 Forestry Building Germany West Lafayette IN 47907 Madison WI 53706 [email protected] (608) 262-1125 (765) 494-6560 (765) 496-7255 Jarmo Schrader (608) 262-1125 [email protected] [email protected] Umea Plant Science Centre Dept Forest Genetics & Plant Mario Rautner Bob Rutledge Physiology Greenpeace International Canadian Forest Service S-901 83 Umea 1055 Rue De Peps 1726 Commercial Drive Sweden Vancouver BC V5N 4A3 Sainte-foy Quebec G1V 4C7 [email protected] Canada Canada (418) 648-5832 Roger Sedjo (604) 253-7701 (604) 253-0114 (418) 648-5849 Resources for the Future [email protected] 1616 P Street NW Mudashiru Salami Saker Washington DC 20036 Mohammad Reza 40 Herbert Rd (202) 328-5065 Portharcourt Riso National Laboratory (202) 939-3460 Plant Bio Dept. PBK 301 Rivers State [email protected] P.O. Box 49 D-4000 Nigeria [email protected] Armand Seguin Roskilde Canadian Forest Service Denmark Hal Salwasser, Dean mohammad.reza.naghavi@riso 1055 Rue De Peps OSU College of Forestry Sainte-foy Quebec G1V 4C7 Bryce Richardson Peavy 150 Canada Corvallis OR 97331 USDA Forest Service (418) 648-5832 1221 S Main St (541) 737-1585 (418) 648-5849 Moscow ID 86843 [email protected] [email protected] (208) 883-2362 (208) 883-2318

247 Lance Sentman Company [email protected] Institute of Human Geography 112 Rue Franqueville-Akwa Rhonda Stokoe 4820 Vesper Drive P.O. Box 4736 Everett WA 98203 Douala Southern Cross University (425) 258-8928 Cameroon P.O Box 157 Lismore NSW 2480 [email protected] [email protected] Austrailia Jerry Settelmeyer Caroline Smith [email protected] Davis High School Oxford University 315 W 14th Street South Parks Road Steven Strauss Oregon State Forest Science Davis CA 95616 Oxford 0X1 3RB (530) 757-5400 United Kingdom Richardson Hall [email protected] Corvallis OR 97331 Nguelou Serge Ngouffo (541) 737-6578 Forum for Biotech Research for Feyisayomi Sodeinde (541) 737-1393 Development Forestry Research Institute of [email protected] Bastos Nigeria Yaounde Forestry Road Idi Ishin Bjorn Sundberg Umea Plant Science Centre [email protected] Jericho Ibadan Oyo State 000234 Dept of Forest Genetics & Plant Ziv Shani Nigeria Physiology CBD Technologies Ltd. S-901 83 Umea [email protected] 2 Pekeris Park Tamer Sweden P.O. Box 199 Kerry Sokol [email protected] Rehovot 76100 University of Maine Ben Sutton Israel 101 Nutting Hall Orono ME 04469 Cellfor Inc. Sandra Sharry (207) 581-2839 Suite 410-355 Burrard Street Universidad Nacional de la Plata Vancouver BC B6C 2G8 (207) 581-4257 Diag. 113 NRO. 469 [email protected] Canada La Plata (604) 602-9229 Buenos Aires Tuomas Sopanen (604) 683-6859 Argentina University of Joensuu [email protected] P.O. Box 111 Oded Shoseyov 80101 Joensuu Makoto Takahashi Inst of Plant Science & Genetics Finland Forest Tree Breeding Center P.O. Box 12 Takizawa [email protected] Rehovot 76100 Iwate 020-0173 Israel Simon Southerton Japan [email protected] Csiro Division of Forestry [email protected] P.O Box E4008 Gancho Slavov Naoki Tani Kingston ACT 2604 Oregon State University Australia Forestry & Forest Products 321 Richardson Hall [email protected] Research Inst. Corvallis OR 97331 1 Mansunosato (541) 737-8479 Brian Stanton Kukizaki 305-8687 [email protected] Greenwood Resources Japan 349 NW 7th Ave [email protected] Sylvianne Sirimuye Camas WA 98607 Environ & Engin Services (360) 833-9922 248 Charles Tauer Houghton MI 49935 Sara Von Arnold Oklahoma State University SLU Dept of Forest Genetics Yoshihiko Tsumura 08C Ag Hall Backlosavagen 2c Stillwater OK 74078 Forest Products Research Insti P.O. Box 7027 (405) 744-5462 Matsunosato 1 Uppsala 75007 Kukizaki (405) 744-9693 Sweden [email protected] Ibaraki 305-8687 (461) 867-3230 Japan (461) 867-2718 Gail Taylor [email protected] [email protected] University of Southampton Hannele Tuominen Bassett Crescent East Christian Walter Southampton Hampshire SO16 University of Helsinki New Zealand Forest Research 7PX POB 56 (Viikinkaari 3d) Inst Ltd Helsinki 14 United Kingdom Sala Street; Private Pag 3020 [email protected] Finland Rotorua [email protected] New Zealand Frank Telewski [email protected] Michigan State University Gerald Tuskan Oak Ridge National Laboratory Wj Beal Botanical Garden Tina Wambach 412 Olds Hall P.O Box 2008 McGill University East Lansing MI 48824 MS 06422 21111 Lakeshore Rd (865) 576-8141 (517) 355-9582 Ste-anee De Bellevue H9x 3v9 (517) 432-1090 (865) 576-8143 Canada [email protected] [email protected] (514) 398-7851 David Victor (514) 398-7897 Sandy Thomas [email protected] Nuffield Council on Bioethics The Council on Foreign Relation 28 Bedford Square 58 E 68th Street Akm Waziullah New York NY10021 London Wc1b 3js Bangladesh Forest Research United Kingdom (212) 434-9621 Institute [email protected] (212) 434-9848 Sholoshahar [email protected] Chittagong 4000 Paul Thompson Akinbobola Victor Bangladesh University of Purdue [email protected] Department of Philosophy Ekiti State West Lafayette IN 47907 P.O. Box 2674 Nicholas Wheeler Surulere Lagos 23401 (765) 494-4295 Weyerhaeuser (765) 496-1616 Nigeria P.O Box 420 [email protected] [email protected] Centralia WA 98531 (360) 330-1738 Jean-francois Trontin Patrick Von Aderkas Univeristy of Victoria (360) 330-1742 Afocel [email protected] Domaine De L’etancon 1700 Finnerty 77370 Nangis Victoria BC V8W 3N5 Lee Wi Young Canada France Korea Forestry Research Institute [email protected] (250) 721-8925 Omockchen-Dong 44-3 (250) 721-7120 Suwon, Kyunggido 441-350 Chung-jui Tsai [email protected] Republic of Korea Michigan Technological University [email protected] SFWP 1400 Townsend Dr 249 G. Douglas Winget Abdul-Hakeem Yesufu University of Cincinnati Midland Integrated Sugar Co. Dept. of Biological Sciences 12 Sura Mogaji Street P.O Box 210006 Ilupeju Lagos Cincinnati OH 45221 Nigeria (513) 556-9735 [email protected] (513) 556-9735 Bundaa Yondon [email protected] Institute of Biotechnology Brenda Wingfield P.O. Box 183 University of Pretoria Ulaanbaatar 48 Dept. of Genetics 210648 Pretoria 2 Mongolia South Africa [email protected] [email protected] Birgit Ziegenhagen William Winner BHF Grosshandorf Germany Oregon State University Institute for Forest Genetics Dept. of Botony & Plant Sieker Landstrasse 2 Pathology Grosshansdorf Corvallis OR 97331 Germany (541) 737-1749 [email protected] (541) 737-3573 [email protected]

Keith Woeste Hardwood Tree Improv. & Regen. Center 1159 Forestry Building Purdue University West Lafayette IN 47907 (765) 496-6808 (765) 496-7255 [email protected]

Alvin Yanchuk BC Forest Service Research Branch 3rd Floor 712 Yates Street Victoria BC V8W 9C2 Canada (250) 387-3338 (250) 387-0046 [email protected]

250