Cover Image
close this bookBiotechnology and the Future of World Agriculture (GRAIN, 1991)
close this folderProviding the inputs
View the document(introduction...)
View the documentThe biased focus: herbicide tolerance
View the documentEroding the options
View the documentArtificial solutions?

(introduction...)

'In two decades, we won't be spraying crap on plants any more. '
(Sam Dryden, then President of Agrigenetics, a US biotechnology company)'

'Screening of cultivars for genetic resistance to new, highly potent herbicides, is becoming as important as screening the same cultivars for genetic resistance to prevalent disease and insect pests.'
(Don Duvick, former director of research of Pioneer Hi-Bred seed company) (2)

Perhaps one of the most exciting and promising possibilities of agricultural biotechnology is to decrease the need for chemical inputs in crop production. Virtually every article on this issue starts off saying that biotechnology has unlimited possibilities in this direction. Newsweek promises its readers that biotechnology will produce plants which 'can destroy plant and insect attackers with little or no help from people'. (3) Howard Schneiderman, R&D Director of Monsanto, also paints a bright future:
I believe . . . that with the new biotechnology almost anything that can be thought of can ultimately be achieved. [He refers specifically to] new treatments for disease, new ways of controlling pests, crops which produce their own pesticides. (4)

This euphoria about the possible impact of biotechnology on agriculture is easy to understand. Biotechnology, at least in theory, can provide the tools for increased pest resistance in crops and for the reduction of dependence on chemical nitrogen fertilizers. Although the work is not as easy as it might seem, it is possible to transfer the genes responsible for pest resistance to crop-plants. Also, research is being carried out to genetically engineer micro-organisms that attack pests and diseases, the so-called biological pesticides.

The breeding of pest resistance into crops has always been a painstaking and expensive job and certainly has not received the attention it deserves. The US Office of Technology Assessment (OTA) believes that in the past decades there was less resistance-breeding because of the availability of cheap pesticides. (5) The main focus of plant-breeding has always been to increase yields rather than to reduce inputs. Private breeding programmes especially lack emphasis on pest-resistance breeding, according to the OTA. (6) In many ways, chemical pesticides were used to compensate for the lack of genetic resistance that might have been bred into crops. Increased emphasis on mono-cropping, based on a few very vulnerable varieties, has served to encourage an agricultural system that needs enormous amounts of pesticides but still loses 20 to 50% of the harvest to pests and diseases (7) Contrary to the impression generally given, crop losses due to pests and diseases have actually increased during the past 30 years. For example, US farmers lost some seven per cent of their crop to insect pests in the 1940s, a figure that had increased to 13% by 1974. (8)

Will biotechnology reverse this trend toward increased crop vulnerability and associated increased pesticide use? It might and it might not. To a large extent it very much depends on whether research priorities are sufficiently directed towards it. Biotechnology provides some very powerful tools to increase pest resistance in agricultural crops, but it certainly does not automatically cause a major shift to pest-resistance breeding. As pointed out earlier, biotechnology research is heavily dominated by the private industry which might have its own agenda. The OTA clearly had this in mind when it stated that:

Much of the agricultural research effort is being made by the agricultural chemical industry, and this industry may see the early opportunity of developing pesticide-resistant plants rather than
undertaking the longer term effort of developing pest-resistant plants. (9)

In this context, the optimistic expectations of Howard Schneiderman, quoted at the outset of this chapter, should be viewed with some scepticism. Schneiderman's company is among the largest pesticide producers in the world.

The biased focus: herbicide tolerance

Nowhere does the immense discrepancy between potential and actual developments in biotechnology become clearer than with current biotech research on herbicide tolerance. Over the years, the use of herbicides has grown dramatically, as a result of changing agricultural techniques: monocropping, mechanization and non-tillage farming. World sales of herbicides amount to almost $5 billion annually, representing some 40% of total pesticides sales in the world.'ø Although the industry often claims that the newly developed herbicides harm neither humans nor the environment, recent research has detected several cases of carcinogenicity caused by herbicides and toxic herbicide residues in groundwater. In general, very little is known about the long-term effects of herbicide residues in the environment.

One problem that limits the use of herbicides is the fact that many herbicides not only attack the weeds they are supposed to kill, but also harm the crop they are supposed to protect. This restricts the farmer in the amount of herbicide she or he can use. Also, some herbicides might not harm a specific crop, but linger too long in the soil and damage the crop that is planted the next season.

The first efforts to reduce the damage that herbicides can cause to crops, were undertaken by Ciba-Geigy. Ciba, which had already bought up several seed companies in the 1970s, developed a chemical 'coat' for seeds to protect them against the herbicides produced by them. This 'herbishield' was wrapped around Ciba-Geigy seeds, thus providing the company with a double profit: the farmer buys the Ciba-Geigy seeds packaged with the Ciba-Geigy herbicides. After successfully introducing the package in industrialized countries, Ciba is now trying to penetrate the markets of the South.

With biotechnology, this process is being further sophisticated. Millions of dollars are being pumped into research to genetically alter crops in order to resist higher doses of herbicides. In Table 5.1 some of the current research on herbicide tolerance is listed. The main source of this listing is based on 'Derwent Biotechnology Abstracts'. (11) Derwent scans more than 1,000 scientific publications and patents related to biotechnology and provides indexed abstracts each fortnight, thus providing probably one of the most complete sources on biotech research. Added to the listing were the results of a few recent studies on the matter. At least 93 institutions have been involved in research on herbicide tolerance since the mid-1980s. All major crops are subject to the search for tolerance to a whole range of different weed killers. The substantial involvement of universities and other public institutions seems at a first sight surprising (48 public institutions in total). One reason for this might be that public institutions tend to publish more freely, and are thus over-represented in the table.

What the table does not show are the dollar figures attached to the projects. The US teased 'Biotechnology Working Group' estimates that publicly funded institutions in the United States alone spent $10.5 million on herbicide tolerance in the past few years. (12) But the major work is done in the corporate laboratories. The public institutions listed in the table generally limit their work to one or two herbicides and only a few crops. Rather than aiming for the commercialization of herbicide tolerant varieties, several of them perform this research to upgrade their knowledge on gene transfer technology in general. In contrast, major pesticide producers such as Monsanto and Du Pont each have a whole range of different projects involving many crops and different chemicals and do focus on the commercialization of the technology. Du Pont alone is investing $13 million in research on tolerance to its new sulfonylurea herbicides. (13) Additionally, most of the work by the smaller biotechnology companies such as Calgene and Plant Genetic Systems is under contract to the TNCs. Keith Pike, Marketing and Sales Director at ICI Seeds, thinks that herbicide tolerance technology is concentrated within a dozen corporations. (14)

No biotechnologically engineered herbicide-tolerant crop has yet reached the farmer's field, but many are now being field tested. The early 1990s are generally seen as the time when the first tolerant crops will become commercially available. If these predictions are confirmed, herbicide tolerance is likely to be the first major result of agricultural biotechnology made available to farmers on a large scale. Opinions differ as to how large. Early estimates talk of an annual value of herbicide-tolerant crops of $2.1 billion by the turn of the century, while other - probably more realistic projections range from a low $75 million to a high $320 million annually. (15) But apart from the profit from the seeds themselves, the chemical TNCs are also interested in the resulting increased use of the chemicals that are sold with them. Teweless reports that a main reason for being involved in this field is the 'hope of selling the seed and the chemical as a pair' thus creasing 'e complementary demand for both chemical and seed'. (16)

A case in point is the work on atrazine, an already widely used herbicide in maize, a crop which is naturally tolerant to it. But soybean, which is often sown in rotation with maize, is very sensitive to the herbicide. As atrazine is persistent and lingers long in the soil, its residues can damage such crops as soybean that are planted the year after. Du Pont has now isolated a gene that enables mutant pigweed to withstand atrazine. According to Charles Arntzen, Du Pont's Associate Director for plant science and microbiology, these mutants 'have a trait that says: "I don't care how much chemical you throw at me, it doesn't faze me". (17) Such perspectives don't faze the chemical companies either. Teweless calculates that with the development of atrazine-tolerant soybeans, atrazine sales would increase by about $120 million annually. (18) A study prepared for the European Commission concurs: 'If the predominant varieties of soya bean were resistant to atrazine, about two or three times more atrazine would be used on related crop land."9

Proponents of herbicide tolerance often point out that this would help to eliminate the older persistent and more dangerous class of herbicides in favour of the new and environmentally safe ones. Graph 5.1, which breaks down the data of Table 5.1 by herbicide, shows that this argument does not reflect reality. The old and persistent triazines (including atrazine) are by far the most researched herbicide group: 30 of the 82 groups for which herbicide tolerance is sought. Triazines have been linked to chronic health effects, such as central nervous system disorders. (20) Dangerous concentrations of atrazine were found in 29% of the samples in a US survey on surface-water quality. (21) Another six institutions are working on the old 2,4-D herbicides, suspected of causing cancer, birth defects and mutations. (22) Paraquat, probably the most toxic herbicide around for humans, is being researched for tolerance by five groups. In all, more than half of the groups listed in Table 5.1 have the old class herbicides included as targets for tolerance.


Table 5.1 Research on herbicide tolerance by private and public institutions - Part 1: Industry


Table 5.1 Research on herbicide tolerance by private and public institutions - Part 1: Industry - continued


Part 2: Public institutions


Part 2: Public institutions


Global research on herbicide tolerance - Number of institutions by herbicide type

This does not mean that the newer herbicides are safe for humans and the environment. Du Pont's low-dose wheat herbicide 'Glean' is a chlorosulfuron, belonging to the group of sulfonylureas which are being researched for tolerance by at least 13 groups. Wheat is naturally immune to Glean, but other crops are not. North Dakota farmer John Leppert has experienced the consequences: 'It would be at least four years in North Dakota before a field treated with Glean could be used for some broadleaf crops.' (23) Also, American Cyanamid's new Imidazolines, used on soybean, are persistent and their residues harm other crops that follow in the cultivation cycle. Cyanamid, Du Pont and others are trying to do something about it, not by developing non-chemical weed control strategies, but by developing crops that tolerate the chemicals. Even glyphosate, which according to one Monsanto executive is so kind to the environment that it could have been 'designed by God', (24) may not be completely safe. Sold by Monsanto as 'Roundup', many of the required safety tests on this herbicide in the USA have been invalidated because of submission of misleading data on the results. Consequently, some health and safety data of this chemical are still under review. It might cause health problems due to its formulation with other ingredients, and because of possibly harmful chemical reactions in the human stomach. (25)

Apart from the obvious negative impact on the environment and the risk to human health, the increased use of herbicides can have serious indirect consequences. Research shows that some herbicides can make crops more susceptible to insect pests and diseases by altering the plant's physiology. Several crops are researched by Du Pont for tolerance to the herbicide picloram, which increases the sugar output in the roots of wheat and corn, encouraging sugar-loving fungal pathogens. (26) Another experiment showed that when maize was treated with the recommended doses of herbicide 2,4-D, for which tolerance is also being sought, it became infested with three times as many corn-leaf aphids. The maize also became more susceptible to European corn borers, corn smut disease and Southern corn leaf blight. (27) Herbicide-resistant crop lines could end up requiring more insecticides and fungicides as well, thus binding farmers even more firmly on to the pesticide treadmill.

Extended use of herbicide-tolerant crops themselves is not without risk either. Crop relatives tend to cross with each other, but also with weedy relatives that grow close by. According to crop scientist Jack Harlan, most crops have one or more sexually compatible weed relative with which they can exchange genetic information. (28) Should genetically engineered crops start handing over their herbicide-tolerant genes to weeds, the farmer is in trouble. The same gene that made it possible to use more of the same herbicide on a particular crop would then be the reason for the farmer to use even greater quantities on it, as the weeds also become tolerant.

From the TNC perspective, it is not hard to understand this heavy research emphasis on herbicide resistance. The use of herbicide-resistant crops will substantially increase the total global herbicide market, and thus the total revenues of the TNCs involved. Also an attractive prospect of herbicide-tolerance engineering is that it offers companies the possibility to bind older herbicides that go off-patent to a specific crop, thus extending the time frame of their use. (29) Another reason emerges when the costs of developing seeds and pesticides are compared. It is simply cheaper to adapt a crop to a herbicide than to develop a new herbicide. A report issued by the European Parliament puts it this way:

From the point of view of the industry, herbicide-resistant varieties are, above all, developed for economic reasons, since the development costs of a new herbicide are up to 20 times higher than those for a new variety. (30)

With both sectors often in the hands of the same TNC, the company can choose; and the choice does not seem to be difficult.

From a socio-economic and agronomic perspective, however, it is difficult to understand why scarce human resources and finance are devoted to make crops resistant to pesticides rather than to pests. With biotechnology, a further development of plant sciences could help to design herbicide-free weed control strategies. These could include better crop rotation techniques, mixed cropping systems repressing the growth of weeds, and, possibly, the use of allelopathic crops that produce natural herbicides. Biotechnology could be used together with traditional plant-breeding to help develop crop varieties that cover the soil at an early stage, thus preventing weeds from becoming a major problem. Rather than totally eliminating the weeds, which is the aim of chemical weed control, such integrated strategies focus on weed management where the farmer uses methods at his disposal to keep damage by weeds at acceptable levels.

Especially for developing countries that so desperately need low-input and locally adapted technologies for their farmers, the present priorities for biotechnology do not make much sense. As with Ciba-Geigy's 'herbishield', herbicide-resistant varieties will find their way to the Third World through the extensive distribution infrastructure of governments and TNCs. This Northern technology will, as with the Green Revolution varieties, primarily be adopted by the large farmers, resulting in a further dependence of the Third World on the North for chemical inputs. It will marginalize the rural poor who need a very different type of technology. In chapter 2, it was already explained how increased herbicide use destroys farming practices in which associated weeds are actually useful plants and form an important source of protein in the local diet and provide extra income for the village people.

Eroding the options

The focus on herbicide-tolerance does not mean that nothing is being done in other areas. Companies and public institutions are using biotechnology on different fronts in order to modify agriculture's input requirements. One of them is breeding for pest resistance. As such, this breeding objective is nothing new. Farmers have been selecting crops with resistance to insects, fungi and viruses for centuries, some of them with remarkable success. More recently, plant-breeders have been doing the same, although as noted earlier, the availability of cheap pesticides did not add to the incentives to focus on this type of research. Biotechnology now comes in as a potentially very powerful tool. Genetic engineering and tissue culture techniques allow scientists to enormously broaden the available gene pool in the search for resistance genes. No longer limited to the germplasm of sexually compatible relatives of the same species, genes can be cloned into crops from theoretically any host, be it other plants, micro-organisms or animals.

While herbicide-tolerant crops are likely to be the first genetically engineered crops available to the farmers, crops with built-in pest resistance might not take long to follow. Much of this research is being done in the public sector, but the chemical giants are also getting involved. Monsanto has already field tested a tomato incorporating genes to ward off tomato hornworms and fruit worms, as well as an engineered tomato with built-in resistance to a virus disease. (31) ICI is focusing on insect-resistant maize and virus-resistant sugar-beet, (32) and Sandoz works with PGS on virus-resistance for several crops. (33) 'Of course [the pesticides industry] is going to lose business as a result of introducing plants that are more resistant to insects or pests', says Riley from Shell Chemicals. But by investing in genetic engineering, 'we're likely to win more than we lose', he asserts. (34)

Although scientists warn that it will take several years before such new plants are available to farmers, the prospect is appealing: plants that themselves fight their enemies without need for external weapons. While the potentials are indeed enormous, the limitations are just as impressive. The resistance that biotechnology might breed into crops in the near future will be based on one or a few genes. The manipulation of entire gene complexes is still far too difficult to handle. Ed Dart, Research Director with ICI Seeds, thinks that crops with single gene traits such as insect-resistance are likely to become seed company standards during the coming decade. (35) This 'onegene/one-pest' resistance is relatively easy to overcome by pests, which are continuously adapting themselves to new situations. Just as pests can develop resistance to pesticides, they are also able to find a way round pestresistance in crops, especially when this resistance is provided by only one gene.

Current biotech research is heavily focused on an extremely narrow range of organisms and genes in the search for new instruments to combat pests and diseases. Graph 5.2 shows a breakdown of research reports on biological control agents as listed in Derwent Biotechnology Abstracts in 1989. The work on the bacteria Bacillus thuringiensis (Bt) is a case in point. A gene from Bt is responsible for the production of a protein that kills certain insects if they ingest it. This microbe and its 'miracle genes is covered in over 37% of all research papers on biological control agents listed by Derwent in 1989. Another six per cent of the papers report on other Bacitlus species, while another ten per cent focus on Pseudomonas, a microbe that promises to be of help in combating diseases caused by fungi. Taken together, these data suggest that over half of all listed biotech research on biological control agents focus on only two bacterial genera. And even this is probably an underestimate. As noted earlier, the Derwent Abstracts tend to under-represent the research in the private sector due to its secrecy. The private sector is likely to focus more strongly on those applications with clear commercial possibilities, of which Bt is the most important. Bt is, after all, responsible for 95% of all current big-pesticide sales. (36)


Research on biological control agents - Reports in 1989, by micro-organims

Engineering plants with Bt genes in them is just one of the different R&D strategies. Other strategies are the encapsulation of the toxin itself for direct use on the crop or the incorporation of the Bt gene into other microbes that naturally colonize the roots of crops or live inside the crop itself. (37) The vehicle might be different, but the 'cure' is each time the same: Bt's 'killer-gene' and the protein that it produces. Here lies exactly the danger of the cure. Researchers from the US Plant Genetic Institute have already found that several insect species can develop resistance to the Bt protein. (38) Research by Monsanto points in the same direction. With research efforts so much directed towards a single microbe, and a specific toxin of that microbe, farmers using the new crops with Bt genes in it might soon face the old problems again. Entomologist Fred Gould puts it this way:

If pesticidal plants are developed and used in a way that leads to rapid pest adaptation, the efficacy of these plants will be lost and agriculture will be pushed back to reliance on conventional pesticides with their inherent problem. (39)

All this is not to say that the use of biotechnology to produce pest-resistant crops or 'bio-pesticides' cannot be beneficial to the farmer and to agriculture in general. We desperately need an agriculture which uses fewer harmful chemicals and other external inputs. The question is whether the current reductionist, biotech pest control approach, almost entirely focusing on single genes 'that work', is one that will help to solve the problems. The view of the biotechnologist is narrow in the sense that it focuses on solutions at the molecular level only. It is also narrow in the sense that most biotech research is dictated by commercial interests. This normally means that the solutions sought must have a global character: TNCs do not tend to work for small market niches, but aim at large market shares. The molecular mind, together with the global market share concern, leads to what is probably the most fundamental threat to sustainable low-input agriculture: uniformity.

Genetic uniformity leads to crop vulnerability. The Irish potato farmers know what that means. In the 1840s, when their staple but uniform potato crop was wiped out by blight, more than one million people died of starvation. One and a half centuries later, the US maize farmers have also learnt the hard way. In 1970, 15% of the US maize harvest was lost to the devastating effect of a little fungus-causing blight, with some states losing over half of their harvest. The cause: genetic uniformity. South-East Asian rice farmers too, using the uniform varieties from the International Rice Research Institute, can sustain their yields only with massive use of pesticides. Many ask for how long.

Whether biotechnology will help to reduce the need for chemical inputs depends, then, to a large extent on whether it will contribute to introducing more genetic diversity in the farmers' fields; in principle it can. Biotechnology opens up a tremendous pool of germplasm which earlier was not feasible for plant- and animal-breeders. Although not creating new genes as such, it provides techniques to move germplasm between organisms that do not exchange genes naturally. It also can provide for quicker germplasm identification, better storage of genetic resources and the speeding-up of plant- and animal-breeding in general.

But in the same way as biotechnology can help broaden the genetic base of our agriculture, it also has the frightening means dramatically to reduce the diversity that still remains in the farmers' fields. With the industry avid for global sales as the driving force behind the big-revolution, increased uniformity is likely to be the dominating trend.

Though the capacity to move genetic material between species is a means for introducing additional variation, it is also a means for engineering genetic uniformity across species writes Jack Kloppenburg in his excellent study on the political economy of plant-breeding. (40) Another case in point is the likely widespread use of tissue culture. Through tissue culture, mass production of genetically identical plants is possible. Such cloned plants are exact copies of each other, and massive recourse to them in a certain crop would seriously increase the vulnerability of that crop. Some estimate that clonally propagated crops are six times more vulnerable to pests than their seed-bred counterparts. (41) The wide use of cloned crops will undoubtedly lead to the increased use of pesticides.

Artificial solutions?

A look beyond the horizon of the year 2000 might show an agriculture without any seed at all. 'We want something that has the ease of handling and high germination efficiency of a seed, but has the genetic uniformity of a clone', says Dennis Gray of the University of Florida. (42) Gray is working on 'artificial seeds'. Normal seed consists of an embryo (resulting from fertilization) surrounded by a reserve of the necessary starch and nutrients for germination and initial growth. Using what scientists call 'somatic embryogenesis', artificial seed technology consists of the mass multiplication of plant embryos in fermentation tanks, each of which is then encapsulated in a jelly-like coat. To some extent it is a sophisticated form of tissue culture resulting in a manageable end-product that can be stockpiled, sold and sown. The California-based biotechnology firm Plant Genetics Inc. (PGI) is generally acclaimed as the front runner in this field as it is involved both in the growing and the encapsulation of the embryos. But others are following fast (see Table 5.2). Supported with funds from the EEC Eureka project, Rhone Poulenc, Nestle and Limagrain are involved in a joint research project on artificial tomato seed, and the Japanese giant food processor Kirin Brewery is building a special research laboratory for the purpose.

The implications are, at least in principle, enormous. The French biotechnology magazine Biofutur calculates that ten fermentation tanks of ten litres each can provide the whole of France with the 'seed' it needs for its entire carrot production. Some figures indicate a production of 80,000 embryos per litre per day. (43) A few tanks more, and the rest of the world is provided for as well. This might sound like science fiction, and to some extent it is. There are still formidable technical hurdles to be overcome. At the current state of the technology, production costs are a problem too. An average artificial seed now costs about the same as a hybrid tomato seed. Translated to major crops, however, the costs to the farmer would be exorbitant: to sow a hectare of sugar-beet, soybean or wheat with artificial seed would currently cost $4000, $13,000 and $50,000 respectively/44 This is the main reason that much of the research is currently focused on the high-value vegetable seed, such as carrot and celery. But several companies are now working on automating the methods of mass propagation, which would bring the costs down further. Encapsulated artificial seeds also provide for the opportunity further to enhance the chemical connection: Plant Genetics Inc. is working with Ciba-Geigy to encapsulate a fungicide together with the somatic embryos.


Table 5.2 The work on artificial seeds


Table 5.2 The work on artificial seeds - continued

The question is not so much whether artificial seed technology will reach the farmers' fields, but when. The stakes for the industry are high. Perhaps the greatest danger of mass introduction of artificial seeds technology is, again, the further narrowing-down of genetic diversity, and its accompanying increase in crop vulnerability. This will undoubtedly lead to a further increase in the use of pesticides, be it of biological origin or not.

Whatever happens, use of the 'good old chemicals' in agriculture will persist for at least the foreseeable future. The chemical industries are the first to admit it. 'There certainly would not be enough food produced to feed the world without pesticides', says Reg Norman, Managing Director of Ciba-Geigy Agrochemicals. 'Without plant protection chemicals, cereal yields would drop in pans of Europe by one-quarter in the first year and by almost one half in the second', threatens the European Chemical Industry in its advertising. But the Third World needs the chemicals most, according to the industry. With stagnating sales and tighter environmental control in the North, it is the developing countries where growing markets will be found. The International Union for the Conservation of Nature (IUCN) calculates that developing countries will increase their pesticide use from some $2 billion in 1980 to over $5 billion in the year 2000. (45) It might be that resistant crops and big-pesticides, generally designed for growing conditions in the North, manage to find their way to the farmers' plots in the foreseeable future. But the production facilities of the 'ordinary chemicals' will still be there. They might have the same fate as DDT and other chemicals: largely banned in the North but massively used in the South. An ICI spokesman puts it this way: 'Where large numbers of people are undernourished or even starving, use of plant protection chemicals can make the difference between food and starvation.' (46) It can also mean the difference between starvation and sickness, as the two million people poisoned by pesticides each year (47) might argue.

Notes and references
1. Quoted by Jack Doyle in Altered Harvest, Viking Press, New York, 1985, p.90.

2. Quoted by Jack Doyle,'Biotechnology's Harvest of Herbicides', in Genewatch, Vol.2, Nos.4-6, Boston, 1985, p.19.

3. Schulman et al, in Newsweek, 18 February 1985.

4.Quoted by Jack Doyle, 1985, op.cit., pp.109-10.

5. OTA, Pest Management Strategies in Crop Protection, Vol. l, Washington, 1979.

6. OTA, quoted in Jack Doyle, 1985, op. cit., p. 190.

7. FAO, quoted in F. Wengemayer: 'Biotechnik far die Landwirtschaft aus der Sicht der Industrie', in Entwicklung + Landlicher Raum! Vol.20, No.5/85, 1985.

8. Pablo Bifani, NewBiotechnologies for Rural Development, ILO, World Employment Programme Research, Geneva, 1989, pp.43-4.

9. Office of Technology Assessment, Commercial Biotechnology: an International Analysis, OTA, Washington, 1984, p.177.

10. Wood Mackenzie &c Co., Agrochemical Overview, 1983.

11. Derwent Biotechnology Abstracts, Derwent Publications Limited, London. Issues from 1986 to 1989 were scanned.

12. Rebecca Goldburg et al, Biotechnology's Bitter Harvest, A Report of the Biotechnology Working Group, USA, 1990, p.21.

13. Agricultural Genetics Report, Du Pont and AGS TransferResistance to Sulfonylurea Herbicides, Vol. 6, No. 2, April 1987, p.1 (cited in RAFl Backgrounder on Herbicide Tolerance, RAFI, March 1989).

14. Personal communication to author during a biotechnology conference in Norwich, UK, December 1989.

15. Rebecca Goldburgetal, 1990, op.cit., p.17.

16.Quoted in Jack Doyle, 1985, op.cit., p.15.

17. Quoted in 'Agrichemical Firms Turn to Genetic Engineering',Chemica/ Week, 3 April 1985, p.36.

18. Quoted in Jack Doyle, 1985, op. cit., p.15.

19. M. Chiara Mantegazzini, The Environmental Risks from Biotechnology, Frances Pinter Publishers, London, 1986, p.74.

20. Rebecca Goldburget al, 1990, op. cit., p.31.

21. Survey of US Environmental Protection Agency between 1977 and 1981, quoted in Jack Doyle, 'Herbicides of potential interest to biotechnology'. Unpublished manuscript.

22. Jack Doyle, 'Herbicides of potential interest to biotechnology', op. cit., p.33.

23. Jack Doyle,'Herbicides and Biotechnology: Extending the Pesticide Era.' Paper presensed to a NOAH conference on biotechnology, Copenhagen, l November 1988, p.9.

24. úShow Me: Monsanto's Marketing Woes', in The Economist, 10 March 1990.
25. Jack Doyle, 1988, op. cit., p.18.

26. Ibid., p.20.

27. David Pimentel, 'Down on the Farm: Genetic Engineering meets Ecology', in Technology Review, 24 January 1987.

28. Jack Harlan, cited in Jack Doyle,'Biotechnology's Harvest of Herbicides', 1985, op. cit.

29. P. Niemann, 'Herbizidresistenz als Zuchtziel', in NachtrichenbL Deutschen Pflanzenschutzdienst, No.41, Braunschweig, 1989, p.38.

30. European Parliament, Commission on Agriculture, Fisheries and Food,'Draft Report on the effects of the use of biotechnology', Brussels, September 1986. (Doe. PE 107.429/ rev.)

31. 'Insect Resistance in Third Monsanto Field Test', in AgBiotechnology News, July/ August 1987,p.5.

32. 'ICl's Agbiotech Goals for the 1990s and Beyond', in AGROW, No. 96, 6 October 1989, pp.5-6.

33. 'Plant Defence', Project Brief, in The Economist Development Report. October 1985.

34. Quoted in 'Agrochemical Firms turn to Genetic Engineering', Chemical Week, 3 April 1985,pp. 36-40.

35. Cited in 'ICl's Agbiotech Goals for the 1990s and Beyond', in AGROW, No.96, 6 October 1989, pp.5-6.

36. Genetic Engineering and Biotechnology Monitor, UNIDO, July-September 1986, p. 40.

37. 'Microbial Insecticides, Special Focus on BT', in RAFI Communiqu,, Pittsboro, January 1989, p.3.

38. 'Des resistance a la toxine de Bacillus', Biofutur, No.89, April 1990,p.12(insects referred to are Plodia interpunctella , and Heliothis virescens) .

39. Fred Gould, quoted in: 'Microbial Insecticides, Special Focus on BT 1989, op. cit.

40. Jack Kloppenburg, First the Seed, Cambridge University Press, New York, 1988, p. 244.

41. Gordon Conway (ed.), 'Pesticide Resistance and World Food Production', cited by Pat Mooney,'Impact on the Farm', in UNCSTD, ATAS Bulletin, Vol. 1, No. l, New York, November 1984, p.46.

42. Quoted in 'Artificial Seeds made from Clones', in AgBiotechnology News, 1987, p. 10.

43. C. Nouaille, V. Petiard, 'Semences Artificielles: Reves et Realites', inBiofutur, No. 67, April 1988, pp.33-8.

44. Ibid.

45. Cited in Pablo Bifani, 1989, op. cit., p.42. Figures refer to 90 developing countries.

46. All quotes in this paragraph are from 'Chemicals Help Feed the World'. Special advertising section of the European Chemical Industry, in Time Magazine, 16 October 1989.

47. Estimate for 1983 of UN Economic and Social Commission for Asia and the Pacific. Cited in Omar Sattaur, 'A New Crop of Pest Controls', in New Scientist, 14 July 1988, p.49.