and Human Values 14: 303±310, 1997.

c 1997 Kluwer Academic Publishers. Printed in the Netherlands.

Biological control and agricultural modernization: Towards resolution of some contradictions

Miguel A. Altieri, Peter M. Rosset and Clara I. Nicholls ESPM-Division of Biology, University of , Berkeley, USA

Accepted in revised form June 12, 1997

Abstract. An emergent contradiction in the contemporary development of biological control is that of the prevalence of the substitution of periodic releases of natural enemies for chemical and the dominance of biotechnologically developed transgenic crops. Input substitution leaves in place the nature of agroecosystems, which in itself is a key factor in encouraging problems. Biotechnology, now under corporate control, creates more dependency and can potentially lead to Bt resistance, thus excluding from the market a key . Approaches for putting back biological control into the hands of (from artesanal biotechnology for grassroots biopesticide production Cuban style to -to-farmer IPM networks, etc.) have been developed as a way to create a farmer centered approach to biological control

Key words: Biological control, Environmental policy, IPM programs Miguel A. Altieri is Associate Professor, University of California, Berkeley. General Coordinator of the UNDP sponsored program Networking and Extension (SANE) promoting capacity building in agro- ecology in Asia, Africa and Latin America. Chairman of the CGiAR-NGO committee to encourage partnerships among NGOs and International Agricultural Research Centers to scale-up agroecological strategies. Peter M. Rosset is the Executive Director of Food First ± The Institute for Food and Development Policy, based in Oakland, California. Among the books he has published are Agroecology (McGraw-Hill, 1990), The Greening of the Revolution: Cuba's Experiment with Organic Agriculture (Ocean Press, 1994), and A Cautionary Tale: Failed US Development Policy in Central America (Lynne Rienner Publishers/Food First Development Series, 1996). The Institute for Food and Development Policy has an on-going US-Cuba exchange program on , coordinated with the non-governmental Cuban Organic Farming Association (ACAO). Clara I. Nicholls is a Colombian entomologist, technical advisor of the Consorcio LatinoAmericano sobre Agroecologia y Desarrollo (CLADES), and a graduate student of at UC Davis, conducting research on diversi®cation strategies of large scale vineyards for sustainable pest management.

Introduction management (Altieri, 1994). As envisioned by the early proponents and practitioners, biological control Biological control was originally de®ned, as ªthe action of was to become a self-sustaining strategy, through which parasites, predators, or in maintaining another farmers relied for on the ecological services organism's population density at a lower average than provided by the ªrestoredº functional , thus would occur in their absenceº (De Bach, 1964). As such, avoiding dependence on costly . Historically biological control distinguishes itself from all other forms this has been the case worldwide, with a number of pests of pest control by acting in a density-dependent manner, brought under permanent control, since the successful that is: natural enemies increase in intensity and destroy a control of the cottony-cushion scale in larger portion of the population as the density of that pop- California with the vedalia , Rodolia cardinalis, ulation increases, and vice-versa (De Bach and Rosen, which was imported from in 1888 (van Driesche 1991). In a strict ecological sense, applied biological and Bellows, 1996). control can be considered a strategy to restore functional The combined savings attributed to the California agri- biodiversity in agroecosystems by adding, through classi- cultural industry from seven major classical biological cal and/or augmentative biocontrol techniques, ªmissingº control programs conducted between 1928 and 1979 was entomophagous or by enhancing naturally occur- estimated at about US$ 20 million, not accounting for ring predators and through conservation and in¯ation and/or discount (Hagen and Franz, 1973; van den 304 MIGUEL A. ALTIERI,PETER M. ROSSER AND CLARA I. NICOLLS

Bosch et al., 1982). Such environmental bene®ts, how- without challenging the monoculture structure of agricul- ever, have rarely been accounted for in normal economic tural systems, greatly diminishes the potential to develop analysis of . a more sustainable agriculture. By only addressing envi- This natural phenomenon of pest population regulation ronmental concerns, biological control as an input substi- created by human management through the enhancement tution approach offers little hope of either reversing the of the interaction of biological control agents, plants, and rapid degradation of the resource base for future produc- provided the ecological basis for what in the tion, or of resolving the current pro®t squeeze and debt 1970s became known as insect pest management (IPM) trap in which the world's farmers are caught (Rosset and strategies (van den Bosch et al., 1982). Under the original Altieri, 1997). IPM vision, agroecosystems were to be diversi®ed and We further argue that biological control (especially managed to enhance natural control and pesticides would classical biological control) is not free of issues relating only be used in ªemergency situations.º Unfortunately, to social equity. For over two centuries, industrialized such a vision slowly eroded and IPM, increasingly under countries have freely appropriated genetic and other pressures from the agrochemical and now the biotech- biological resources from developing countries for agri- nological industry, came to mean something more like cultural production without compensating them for such Integrated Management; that is, the justi®ca- services (Kloppenberg and Kleinman, 1987). In our view, tion of pesticide use only when pest populations surpass a the exchange and importation of natural enemies is liable predetermined economic threshold. The problem is that, to similar equity considerations as has been the case for as the monocultural structure of agroecosystems, which crop germplasm. We also question as ªfalse promisesº lacks ecological defense mechanisms, remains unchal- corporate claims that genetic engineering is the wave of lenged, pest problems continually surpass tolerable levels, the future for biological control (Hindmarsh, 1991). We and thus require constant control interventions. Even suggest that this form of biotechnology may exacerbate when natural enemies are featured in IPM programs in pest and other problems of conventional agriculture, and biologically impoverished , biological con- undermine ecological methods of farming such as biolog- trol has tended to function more as a strategy to ªpatch upº ical control itself. Furthermore, numerous large-scale monocultures. A totally different outcome is observed in releases of genetically engineered organisms risk erod- diversi®ed agroecosystems, where pests are more likely ing genetic diversity and distorting ecological processes to remain below economic thresholds when natural enemy such as natural control in agroecosystems (Rissler and biodiversity is high (Andow, 1991). Mellon, 1996). In fact, such biotechnology may seriously In a way, classical and augmentative biological control threaten the food systems of Third World countries. has had an ªecologically naiveº view of modern agricul- As an alternative, we discuss remarkable advances ture. By assuming and accepting the persistence of large made in developing countries on farmer-led biolog- scale monoculture , most of the time biological ical control efforts that constitute ªground-upº farmer-to control specialists seek only to ªbalanceº monocultures farmer approaches to technology transfer and develop- through the addition of natural enemies that may be key ment and that represent potential avenues for reaching to controlling a speci®c pest. This narrow acceptance of production autonomy. We also elaborate on how Cuba's the present structure of agriculture (i.e., as monocultures unique and amazing advances in the artesanal and decen- that are at the very root of most pest problems) as a given tralized development and application of are condition, restricts the real possibility of implement- slowly serving to demystify the concept that biotech- ing alternatives that challenge such a structure (Levins, nology can only be high-tech and executed in sophisti- 1973). cated laboratories under private corporate control. It could thus be argued that while often painted as a radical alternative to chemically based control in industrial farming, biological control does not actually Classical biological control and social equity challenge the more fundamental bases of industrial agriculture. Biological control has often been wrongly Classical biological control involves the introduction of promoted as an appropriate technology of low envi- natural enemies from the center of origin of an insect ronmental impact whose diffusion per se could initiate that has become an exotic pest elsewhere (van broader social change. This technological determinism Driesche and Bellows, 1996). Throughout history this has has to a signi®cant extent prevented biological control involved hundreds of exchanges between various world specialists from understandingthe structural roots of envi- regions of natural enemies used against agricultural insect ronmental degradation and the larger crisis of industrial pests. The analysis conducted by Altieri (1991), provides agriculture. a measure of the ªbiological control contributionº of each In this paper, we argue that to de®ne biological control of six regions to world agriculture and also a measure as only a technical substitute for agrochemicals inputs, of the ªbiological control dependenceº of each region on BIOLOGICAL CONTROL AND AGRICULTURAL MODERNIZATION 305 non-indigenous sources of natural enemies. It is clear that culture have been dealt with mainly by pesticides,many of the six regions are interdependent in terms of biological which have been restricted or banned in the industrialized control agents. It is also obvious that there are countries countries (Conway and Pretty, 1991). that are disproportionately more dependent than others for A closer look at biological control programs judged to natural enemies (USA and ) and that other regions be ªagronomically successfulº may show that they are not have made more signi®cant contributions(Asia, including so successful in social terms. For example, the vast major- India). ity of the classical biological control efforts conducted If the regions are clustered into industrialized coun- in developing countries (many of which were sponsored tries (USA, Canada, Western Europe, and Australia) and by the governments of industrial countries) have been developing countries (Latin America, Asia, and Africa), primarily directed at commercial, industrial and export it is observed that the industrialized countries have signif- tree crops such as coffee, coconut, citrus, cocoa, and icantly bene®ted from the natural enemy richness of the banana and not local food crops (Hansen, 1987). This developing countries. Up to 1965, industrialized countries trend was particularly notorious in British Commonwealth had received 49 species of natural enemies from develop- sponsored projects during colonial times. Given the struc- ing countries for the control of various agricultural insect tural realities of developing countries, it is obvious that pests, whereas the developing countries received only 30 these efforts were mostly for the bene®t of large-scale species from the industrialized countries. Asia (including commercial farmers, and not for the large masses of peas- India) has provided 21 different species of natural enemies and the rural poor in these countries (Murray, 1994). to the USA and Canada, but in return have only received Notable exceptions are the biological control programs one species from . against wheat in Brazil and Chile, against rice When more recent data on the inter-regional intro- pests in southeast Asia and against pests in Africa. ductions of parasitic insects against of agri- These projects, in addition to targeting crucial food crops, cultural, forestry, and medical importance are analyzed, also emphasize building indigenous capabilities to imple- it reinforces the point about the world's interdependence ment pest management programs, encouraging the use of on biological control agents (Luck, 1981). When exam- simple and low-cost techniques easily adaptable by small ining data from the 1970s and 1980s, the industrialized farmers (Hansen, 1987; Thrupp, 1996). countries-developing countries dependency relationship Another issue that illustrates inequities in the inter- still holds and again appears highly dependent on regional exchange of biological resources and that is foreign natural enemy sources. Through 1981, developing compounded by the contradictory nature of the contempo- countries donated 353 species of natural enemies to the rary structure of the world economy, is the fact that while industrialized countries, whereas the developing coun- developing countries were supplying biological control tries only received 260 natural enemy species from the agents to industrialized countries, chemical companies industrialized countries. Again, Africa, Asia, and Latin from the industrialized countries were engaged in a America stand out as net contributors. massive export of pesticides to developing countries. The data on inter-regional exchanges suggests that From 1974 to 1978, imports of pesticides by develop- because there has been more transfer of natural enemies ing countries increased from $641 million to almost $1 species from developing countries to industrialized coun- billion. Up to the late 1970s, 38 percent of the interna- tries than vice-versa, it could be argued that industrialized tional trade in pesticides occurred in developing countries countries have accrued a ªbiological control debtº with (Weir and Shapiro, 1981). In a period of just two years, developing countries. It could also be argued that such US companies increased their pesticide exports from $615 debt is related to the fact that the agriculture of the indus- million to $1 billion. Tragically, 30 percent of all pesti- trialized countries is based on introduced plant material cides exported from the USA were unregistered, that is, and therefore vulnerable to exotic pests amenable for not approved for use in the USA by the Environmen- classical biological control (i.e., by 1970 there were 212 tal Protection Agency (EPA). In other words, develop- insect pests of foreign origin in the USA) or that inten- ing countries became a kind of dumping ground for the sive transfers were the result of a much greater ®nancial USA and other industrialized countries (Murray, 1994). and scienti®c capacity of the industrialized countries to Similarly, UK pesticide exports (mostly to developing do so. Another argument could be that given the ecolog- countries) grew by 211 percent in value over the 1975± ical vulnerability of high-input agricultural monocultures, 79 period, reaching about 66,000 tons by 1979 (Conway industrialized countries have a greater need to utilize and Pretty, 1991). Latin America's share of the global natural enemies to patch up unstable agroecosystems than pesticide market, currently around 10 percent, is steadily developing countries. It is possible that with the expansion increasing. Brazil alone accounts for nearly 50 percent of of monoculture-based agroexports in developing coun- the total sales in the region, followed by Mexico, Argen- tries, this need may also increase in all such countries. tina, and Colombia. From 1980 to 1986, pesticide sales Thus far in these countries, pest problems in export agri- rose dramatically in Brazil and Argentina.If current trends 306 MIGUEL A. ALTIERI,PETER M. ROSSER AND CLARA I. NICOLLS continue, the cost to Latin America of chemical pest con- ªinput substitutionº (Rosset and Altieri, 1997). The thrust trol is expected to reach $3.97 billion by the year 2000 is highly technological, with the ªlimiting factorº men- (Belloti et al., 1990). tality that has driven conventional agricultural research In many developing countries, governments until in the past. Agronomists and other agricultural scientists recently subsidized pesticide production and sales. The have for generations been taught the ªlaw of the mini- median level of subsidy was about 44 percent of total mumº as a central dogma. According to this dogma, at retail costs. Such subsidies make pesticides considerably any given moment there is a single factor limiting yield cheaper, thus encouraging farmers to use more chemicals increases, and that factor can be overcome with an appro- than they would if they had to pay the full costs (Murray, priate external input such as a chemical or biological 1994). These subsidies undermine efforts to promote more (Rosset and Altieri, 1997). ecologically-sound pest control methods such as biolog- There are several problems with this approach. It ical control. International assistance agencies based in focuses on the most super®cial level of integration in the industrialized countries, including the World Bank and US agroecosystem, that of a single species, the crop, with a Agency for International Development (USAID), have in single limiting factor, such as an insect pest. It denies the the past been involved in promoting pesticide use in devel- rich, scienti®c basis provided by the science of ecology for oping countries, either directly through agricultural devel- the importance of higher levels of interaction, including opment loans, or indirectly though support for local agri- synergism, antagonism, and multiple-species interactions cultural credit programs or technical assistance programs such as those of herbivores and their natural enemies. (Repetto, 1985). Although international agencies have From a practical standpoint, the outcome of the ªlimiting announced new policy guidelines governing pesticide use factorº approach inevitably is that as a farmer ªsolvesº in developmentprojects, such guidelines have been unjus- the problem of one symptom, he or she is confronted with ti®ably slowly implemented. Such sponsored assistance another, ªunexpectedº problem. If he or she uses urea to hinders biological control in developing countries and overcome nitrogen as a limiting factor, for example, they promotes use of pesticides, while these countries contin- are all to often then confronted with an outbreak of insect ue supplying bene®cial organisms to industrialized coun- pests such as aphids or white¯ies, whose numbers are tries. Such a situation is unethical and suggests a type of dramatically increased by the greater availability of free ªecological imperialism.º It should be noted, however, nitrogen in the plants' sap upon which they feed (Altieri that many biological control workers in the industrialized and Rosset, 1996). This perpetuates a process of treat- countries actively oppose such policies, and are working ing symptoms rather than the real causes that evoked the hard to develop and promote more equitable alternatives. ecological imbalance. Further inequities may arise with the emergence of In this context, we ®nd the prevalence of input substi- biotechnology, ®nanced mostly by private interests in tution in alternative or ªsustainableº agriculture to be the industrialized countries. As interest in genetically- alarming. Essentially, the capital-intensive, monoculture- engineered biological control agents increases, it is based system of conventional agriculture is left intact. All possible that developing countries may be caught in changes are relatively minor. A toxic pesticide is removed purchasing ªpatented natural enemiesº at a high cost. and a biological product is substituted. While these The ®nal irony is that such novel biotic agents be based changes may suggest a more environmentally benign on genetic resources originally obtained at no cost from direction, they leave in place the key forces that are driving developing countries (Kloppenburgand Kleinman, 1987). the agricultural crisis; extensive monoculture, excessive use of machinery, input control by , depen- Biological control and input substitution dence of fossil fuels, and very high capital requirements. This approach neither addresses the debt trap that farmers The goal of the sustainable agriculture movement is to are caught in because of high costs of machinery and generate major technological adjustments in conventional inputs, nor the ecological basis of declining yields ± the agriculture to make it more environmentally, socially, reduction of functional biodiversity of agroecosystems and economically viable (Vandermeer, 1995). The main (Altieri and Rosset, 1996). focus has been to substitute less noxious inputs for the Evidence for the increasing dominance of this faux- agrochemicals that are blamed for so many of the prob- sustainable approach is everywhere. Organic farming, lems associated with conventional agriculture. Emphasis commonly viewed as a holistic concept, is now heavily is now placed on purchased biological inputs such as commodity and capital oriented. Publications directed thuringiensis, a microbial pesticide that is now at organic farmers are ®lled with advertisements for widely applied in place of chemical insecticides, and expensive biological pesticides, commercial , is marketed by major chemical companies under brand insectary-produced natural enemies, botanical extracts, names like Dipel and Javelin. This type of technology microbial and other soil amendments, etc., increasing pertains to a dominant technical approach we have called their dependence on ªgreenº suppliers (Lampkin, 1990). BIOLOGICAL CONTROL AND AGRICULTURAL MODERNIZATION 307

Throughout the world, an input substitution industry bene®t of reduced synthetic insecticide use that Bt-crops is emerging and internationally funded IPM programs, and sprays could bring. With the loss of Bt's ef®cacy, government extension agents, and commercial sales farmers ± particularly organic farmers who depend on Bt representatives urge farmers to use new, safe, and effec- ± will be faced with decisions on how to reduce resurgent tive biological products, like Javelin, which may cost damage by insect pests. Also, the widespread use of trans- as much as $150 a liter, or even Avermec, which may genic varieties will further decrease within and between cost more than $400. These products are indeed safer, ®eld genetic variability, already the source of the great and in many cases more effective, than methyl parathion. susceptibility of ªmodernº agriculture to novel pest and Nevertheless, a question must be asked.In its crudest form disease outbreaks (NAS, 1972). this question is, ªWhat is more injurious to the health of An alternative to the input substitution approach a family whose annual income may be well under and future dependence on biotechnology packages is an $1000 per year ± exposure to the occasional whiff of agroecological strategy to achieve sustained agricultural methyl parathion or having to pay an additional $393 for productivity by breaking the monoculture structure and an essential production input?º More generally, if alter- dependence on off-farm inputs through the design of inte- native products raise production costs for First and Third grated agroecosystems (Altieri, 1995; Altieri and Rosset, World farmers already caught in a cost/price squeeze, and 1996). Agroecology states that the health and performance increase their already excessive dependence on off-farm of agroecosystems depends on how well established is a suppliers of inputs, then input substitution biological con- diverse assemblage of natural enemies and antagonists. trol does not offer a way out of the crisis (Rosset and By assembling a functional biodiversity, it is possible Altieri, 1997). to potentiate synergisms that subsidize agroecosystem The dependence of farmers is further exacerbated processes by providing ecological services such as the as multinational companies attracted by the commer- activation of soil biology, the recycling of nutrients, the cial opportunities presented by agricultural biotech- enhancement of bene®cial arthropods and antagonists, nology develop genetically engineered plants that exhibit and so on. Today, there is a spectrum of biodiversi®cation ªbiological controlº properties. A case in point is the practices and technologies available that are of a preven- development of transgenic Bt-containing crops, which tative nature and act by reinforcing the robustness of the corporations claim will be environmentally clean and agroecosystem. The impacts of many of these practices more effective than existing insect control strategies. Not have been scienti®cally documented (Vandermeer, 1995). only will these transgenic crops be more expensive, but based crop rotations, one of the simplest forms just like chemical pesticides, transgenic crops can be of biodiversi®cation, improve yields by the known action expected to exert strong selection pressure in favor of of interrupting , disease and insect life cycles. They pests with a resistance to the natural biotoxins that are also can have subtle effects such as enhancing the growth used (Rissler and Mellon, 1996). and activity of soil biology, including vesicular arbuscular There is general concern that widespread use of Bt- mycorrhizae (VAM), which allow crops to more ef®ci- containing crops could accelerate the development of ently use soil, water, and nutrients and also to remain insect pest resistance to Bt. Already, eight species of healthy. insects have developed resistance to Bt toxins,either in the On the other hand, ®eld plant diversi®cation of agro- ®eld or laboratory (Hruska and Pavon, 1997). The prob- ecosystems can result in increased environmental oppor- lem with transgenic Bt plants is that they will increase tunities for natural enemies and, consequently, improved the exposure of pests to Bt, especially where the Bt is biological pest control. The various vegetational designs continuously expressed in a plant throughout its growing available in the form of , weed diversi- season. This is signi®cant because long-term exposure to ®ed crop systems, cover crops, and living mulches and Bt toxins promotes development of resistance in insect their effects on pest populations and associated natural populations. This kind of exposure could lead to selec- enemies have been extensively reviewed (Altieri, 1994 tion for resistance in all life stages of the insect pest on and references therein). Factors involved in pest regula- all parts of the plant for the entire season (Rissler and tion in diversi®ed agroecosystems include among others: Mellon, 1996). Such concerns have recently materialized increased /predator populations, available alter- according to the Union of Concerned Scientists (1996), native prey/hosts for natural enemies, and optimum who af®rm that during the 1996 growing season, Mon- synchrony between pests and natural enemies (Altieri, santo's failed to control cotton bollworm on 1994). thousands of acres, raising questions about the adequacy Research has shown that by adding plant diversity of the resistance management plans that the US EPA had to existing monocultures, it is possible to exert changes approved for Bt cotton. in habitat diversity that in turn favor natural enemy Unless effective resistance management strategies are abundance and effectiveness. This information can be developed and implemented, growers may soon lose the used to design intercropping and cover cropping that 308 MIGUEL A. ALTIERI,PETER M. ROSSER AND CLARA I. NICOLLS enhance predator and parasitoid diversity and abundance, government sponsorship, Bt is now widely used in 30 thus resulting in lower pest loads than in monocultures provinces for the control of various pests of agriculture (Andow, 1991). Manipulation of wild vegetation adja- and forests (Entwistle, 1993). cent to crop ®elds can also be used to promote biological Since Cuba's trade relations with the socialist bloc col- control, since the survival and activity of natural enemies lapsed in 1990, pesticide imports dropped by more than often depends upon the resources provided by the vege- 60 percent, fertilizers by 77 percent, and petroleum for tation around crop ®elds. Hedgerows and other landscape agriculture dropped by 50 percent. In order to deal with features have received signi®cant attention in Europe such shortages, massive efforts were initiated to ®nd ways regarding their effects on bene®cial distribu- to reduce chemical use and to develop alternatives for tion and abundance in adjacent crop ®elds (Fry, 1995). In management of plant diseases, insect pests, and . California, the parasite was effective in The production of biopesticides and biological control controlling the Erythroneura elegantula agents are at the heart of this new quest with the creation in vineyards adjacent to wild blackberries, which harbor a of about 220 Centers for the Production of Entomophages non-economic leafhopper Dikrella cruentata, whose and Entomopathogens (CREEs) where decentralized, serve as the only overwintering resource for Anagrus. ªartesanalº production of biocontrol agents takes place Recent studies have shown that prune trees planted next (Rosset and Benjamin, 1994). The centers produce to vineyards can also allow early-season buildup of Ana- a number of entomopathogens (Bacillus thuringien- grus epos and early bene®ts of are promoted in sis, , anisopliae,and vineyards with prune trees plants upwind from the vine- Verticillium lecanii), as well as one or more species of yard (Flint and Roberts, 1989). Researchers now recom- , depending on the crops grown in each mend that as many prune trees as possible should always area. In 1994, production levels of B. thuringiensis and B. be planted upwind from the vineyard. bassiana reached 1300 and 780 metric tons respectively. From the perspective of farmers' independence, the CREEs are maintained and operated by local technicians, advantage of the agroecological approach, which empha- many of them sons and daughters of owners of com- sizes the use of biodiversity, is that corporations cannot panies, which produce and distribute these products to appropriate the workings of a balanced agroecosystem. local, state, cooperative, and private farmers. CREEs are In other words, the agroecological approach offers mini- thus biofactories that produce low priced microbial prod- mal pro®t to agribusiness corporations; on the contrary, it ucts for local use (Rosset and Benjamin, 1994). Opening encourages farmers' productive self-reliance. Latin America and other developing countries' markets to Cuba's biotechnology products and expertise can pro- vide poor and dependent countries access to alternative Putting biological control back in the hands of and cheaper technologies. In fact, Cubans are willing farmers to train people from Lesser Developed Countries (LDC) in biotechnology, thus enabling them to develop their Throughout the world, more and more farmers upset with own appropriate biotechnology and to escape the techno- the high economic, environmental, and health costs of logical control and treadmill imposed by multinationals. the pesticide treadmill, are engaging in efforts to replace As rural communities within LDCs bene®t from Cuban chemical-intensive pest control methods with alternative technological advances, a parallel technological path to agricultural approaches (Thrupp, 1996). The emergence the prevailing corporate model can be developed, thus of such local initiatives has taken many forms: from providing farmers with more options, and even with community organizations to farmer-to-farmer networks, the possibility of becoming technologically independent at times supported by non-governmental organizations through the creation of simple community managed insec- (NGO) or even local governments. In many cases, IPM taries, microbial insecticide, and biofertilizer manufactur- technologies developed or diffused by NGO technicians ing facilities. have taken advantage of farmer-initiated activities to fur- In another political context,in southern Brazil, farmers ther scale-up the spread of IPM technologies useful to quickly took advantage of the identi®cation by Empresa small farmers (Altieri, 1993). Brasileira de Pesquisa Agropecuaria (EMBRAPA) scien- Historically in , large-scale production of the tists of a highly speci®c Nuclear Polyhedrosis (NPV) microbial insecticide occurred in to control the velveteen in soybean. Farmers communes through solid or liquid fermentation in tanks. realized that they could readily mass-produce the virus by Wheat bran, corn meal, soybean, defatted cottonseed themselves, by just collecting infected , grind- cake, and peanut bran are the main media components ing them up, and then mixing them with water to produce used in Bt production. In the pilot plant at Hubei Academy a suspension that could be sprayed. If the suspension of Agricultural Sciences, production grew from 26 tons in is kept frozen, it will retain its viral activity for longer 1983 to 90 tons in 1984, and to 900 tons in 1990. Under than a season. In essence, farmers can ªmanufactureº BIOLOGICAL CONTROL AND AGRICULTURAL MODERNIZATION 309 their own pesticide, thus effectively bypassing the need Industrialized countries should accept that a system of to depend on government or industry for the supply of the compensation may provide the basis for a stable mecha- virus (Hansen, 1987). In a nearby zone in Parana, organic nism for future exchange of biological control agents, and farmers have noticed the disease suppressing effect of if properly administered by developing country govern- a fermented compost preparation that contains naturally ments or other international institutions such as UNFAO, occurring and other antagonists, which they those ®nancial resources could be used to expand research spray on citrus and other fruit trees to control Oidium and and development in biological control that truly bene®ts Botrytis, thus obviating the use of chemical all farmers, while simultaneously helping to regenerate (personal observation). and preserve the already degraded environments of devel- The FAO-initiated IPM program for rice in south oping countries (Altieri, 1991). and southeast Asia has become a major model for how It will be of utmost importance for biological control to establish farmers' networks to implement participa- scientists and practitioners to oppose the expansion of tory IPM and it is touted as one of the most sustain- corporate hegemony in crop protection and production. able crop protection options for the future. The program As corporations make capital available, universities and emphasizes an innovative approach of farmers' learn- research institutions are exposed to new and increased ing about IPM, natural enemies, and rice agroecology pressures to serve private interests at the expense of through practical experience and in ªFarm Field Schoolsº the public good. The danger inherent in this issue is that enhance farmers' knowledge about bene®cial biodi- that publicly funded agricultural research, like corporate versity and scienti®c crop-management skills. By 1986, research, is already being targeted toward the develop- about 17,000 farmers had been trained per season in ment of pro®table products to sell to farmers, rather than Sri Lanka. In Kasakolikasan, Philippines, 3,800 farmers toward research to help farmers cut inputs and costs, and have been trained, and their use of pesticides dropped thus to decrease their overall dependency. between 60±98 percent and rice yields had increased This is why an agroecological approach that encour- between 5±15 percent (Thrupp,1996). The BIOS (Biolog- ages biointensive IPM must be emphasized. By taking ically Intregrated Systems) program in Califor- advantage of the integration of plant and animal biodi- nia implemented by dozens of almond and walnut grow- versity to enhance complex interactions and synergisms, ers, demonstrates that biologically integrated systems it is possible to design agroecosystems that sponsor their ( with an undergrowth of selected cover crops), own pest control without the need for external chem- encourage natural control and thus reduce the reliance on ical or biological inputs. Diversi®ed cropping systems, pesticides and can be pro®table (Thrupp, 1996). such as those based on intercropping or cover cropping of orchards, have been the target of much research recently. Conclusions This is largely based on the newly emerging evidence that these systems are more sustainable and more resource- There is no question that biological control can reduce conserving (Vandermeer, 1995). Much of these attributes production costs and can ameliorate some direct environ- are connected to the higher levels of functional biodi- mental impacts of agriculture, such as pesticide residues versity (including natural enemies) associated with such and resistance, but it does not reduce the fundamental complex farming systems. The key is to identify the type vulnerability of monocultures. Furthermore, when used in of biodiversity that is desirable to maintain or enhance in the mode of ªinput substitutionº (i.e., inundative releases order to sponsor ecological services, and then to deter- or biopesticides), it replaces cheap, ecologically harmful mine the best practices that will encourage the desired inputs with expensive and benign ones, thus increasing biodiversity components. The idea is to apply the best costs and failing to address the economic crisis faced by management practices in order to enhance or regenerate the world's farmers. the kind of biodiversity that can subsidize the sustainabil- We have also argued that the disproportionate quan- ity of agroecosystems through enhanced biological pest tity of natural enemies that industrialized countries have control, thus freeing farmers from the heavy burden of appropriated from developing countries raises questions external input dependency (Altieri and Rosset, 1996). of equity in biological control inherent to the global If input substitution is required, as is often the case patterns of exchange of and access to natural enemies. At during the transition toward organic production, then the issue is the substantial ecological debt that industrialized local, decentralized, and low cost Cuban-style biofacto- countries accrued with developing countries in the form ries offer the best hope to farmers who are able to organize of , and which has remained largely and cut the umbilical cord to the external suppliers of uncompensated. With increasing involvement of private technology. Also the creation of farmer-to-farmer net- corporations in bioprospecting, this question may become works that implement participatory IPM programs are the even more critical. most appropriate strategy to achieve an environmentally 310 MIGUEL A. ALTIERI,PETER M. ROSSER AND CLARA I. NICOLLS friendly, socially sound, and economically viable agricul- Hagen, K. S., and J. M. Franz (1973), ªA history of biological tural production. control,º in R. F. Smith et al. (eds.), History of entomology Recent experiences in the Third World suggest that an (pp. 433±476). Palo Alto, CA: Annual Reviews Inc. agroecological-grassroots development is possible, espe- Hansen, M. (1987), Escape from the pesticide treadmill.New cially when implemented at the local level. A number of York: Consumers Union. grassroots rural development programs in Latin Ameri- Hindmarsh, R. (1991), ªThe ¯awed `sustainable' promise of genetic engineering,º The Ecologist 21: 196±205. ca working in peasant communities are broadening the Hruska, A. J., and M. L. Pavon, eds. (1997), Trangenic base of participation in the production process under an plants. Bacillus thuringiensis in Mesoamerican agriculture. agroecological paradigm, and daring to create equitable Honduras: Zamorano Academic Press. arrangements for distributing the bene®ts from sustain- Huffaker, C. B., and P. S. Messenger (1976), Theory and practice able production increases (Altieri, 1995). Through local of biological control. : Academic Press. participation, local technologies are developed based on Kloppenberg, J., and D. C. Kleinman (1987), ªThe plant local knowledge and resources, and used as instruments of germplasm controversy,º BioScience 37: 190±198. empowerment and social organization. This allows peas- Laing, J. E., and J. Hamai (1976), ªBiological control of ants to meet their basic needs without increasing depen- insect pests and weeds by imported parasites, predators and dence on external assistance. The objective is to build pathogens,º in C. B. Huffaker, and P. S. Messenger (eds.), upon the best of the local initiatives to ensure and support Theory and practice of biological control (pp. 686±744). New York: Academic Press. self-reliance and local adaptation. Lampkin, N. (1990), Organic farming. Ipswich: Farming Press. Levins, R. (1973), ªFundamental and applied research in agri- culture,º Science 181: 523±524. References Luck, R. F. (1981), ªParasitic insects introduced as biolog- ical control agents for arthropod pests,º in D. Pimentel Altieri, M. A. (1991), ªClassical biological control and social (ed.), Handbook of pest management in agriculture,Vol. equity,º Bulletin of Entomological Research 81: 365±369. II (pp. 125±284). Boca Raton, FL: CRC Press. Altieri, M. A. (1993), Crop protection strategies for small Murray, D. L. (1994). Cultivating crisis: The human costs of pes- farmers. Boulder, CO: Westview Press. ticides in Latin America. Austin: University of Texas Press. Altieri, M. A. (1994), Biodiversity and pest management in National Academy of Sciences (1972). Genetic vulnerability of agroecosystems. New York: Haworth Press. major crops. Washington, DC: NAS. Altieri, M. A. (1995), Agroecology: The science of sustainable Repetto, R. (1985), Paying the price: Pesticide subsidies in agriculture. Boulder, CO: Westview Press. developing countries. Washington, DC: World Resources Altieri, M. A., and R. Rosset (1996), ªAgroecology and the Institute. conversion of large-scale conventional systems to conven- Rissler, Jr., and M. Mellon (1996), The ecological risks of engi- tional management,º International Journal of Environmen- neered crops. Cambridge, MA: MIT Press. tal Studies 50: 165±185. Rosset, P., and M. A. Altieri (1997), ªAgroecology versus input Andow, D. A. (1991), ªVegetational diveristy and arthropod substitution: a fundamental contradiction of sustainable agri- population response,º Annual Review of Entomology 35: culture,º Society and Natural Resources (in press). 561±568. Rosset, P., and M. Benjamin (1994), The greening of the Belloti, A. C., C. Cardona, and S. L. Lapointe (1990), ªTrends revolution: Cuba's experiment with organic agriculture. in pesticide use in Colombia and Brazil,º Journal of Agri- Melbourne (Australia): Ocean Press. cultural Entomology 7: 191±201. Thrupp, L. A. (1996), New partnerships for sustainable agri- Conway, G. R., and J. N. Pretty (1991), Unwelcome harvest: culture. Washington, DC: World Resources Institute. Agriculture and pollution. London: Earthscan Publishers. Union of Concerned Scientists (1996), ªBt cotton fails to control DeBach, P. (1964), Biological control of insect pests and weeds. bollworm,º The Exchange 7(1): 1, 8. New York: Reihold. Van den Bosch, R., P. S. Messenger, and A. P. Gutierrez (1982), DeBach, P., and D. Rosen (1991), Biological control by natural An introduction to biological control. New York: Plenum enemies, 2nd edn. Cambridge: Cambridge University Press. Press. Entwistle, P. F. (1993), Bacillus thuringiensis: An environmental Vandermeer, J. (1995), ªThe ecological basis of alternative agri- biopesticide. Chichester: John Wiley and Sons. culture,º Annual Review Ecological Systems 26: 201±224. Flint, M. L., and P. A. Roberts (1989), ªUsing to Van Driesche, R. G., and T. S. Bellows (1996), Biological con- manage pest problems: Some California examples,º Amer- trol. New York: Chapman and Hall. ican Journal of Alternative Agriculture 3: 164±167. Weir, D., and M. Shapiro (1981), Circle of . San Fran- Fry, G. (1995), ªLandscape ecology of insect movement in cisco: Institute for Food and Development. arable ecosystems,º in: Ecology and Integrated Farming Systems. Proceedings of the 13th Long Ashton International Address for correspondence: Miguel A. Altieri, University of Symposium, Bristol, UK. New York: John Wiley and Sons. California, Berkeley, ESPM-Division of Insect Biology, 201 Fowler, C., and P. Mooney (1990), Shattering: Food, politics, Wellman-3112, Berkeley, CA 94720-3112, USA and the loss of genetic diversity. Tucson: University of Ari- Phone: (510) 642-9802 zona Press. E-mail: [email protected]