Document(s) 21 de 38

Abstract: Centralized plant breeding, producing uniform varieties broadly adapted to huge areas, leads to reduction of farmer-managed crop diversity and increased dependence on genebanks for the survival of the genetic resource base. Much evidence points to decentralized breeding, specific local adaptation, and intra-varietal diversity as advantages from a biological point of view. The exploitation of location-specific adaptation and more heterogeneous varieties require a different organizational framework compared to that of the established formal plant breeding institutions.

The systematic use of improved, but still location-specific and heterogeneous varieties is possible if seed selection is devolved to the community level. Potentials of that approach are discussed with reference to two cases: one, in Ethiopia, which involves traditional seed selectors, and one, in the Philippines, which re-introduces abandoned seed selection practices. Organization of such decentralized seed selection is likely to result in new demands on the genebanks and new challenges to the scientific breeding community. While genebank collections are little used in current plant breeding, local seed selectors will be interested in previously collected landraces from their own or ecologically-similar areas. Plant breeders will be challenged to supply enhanced germplasm or populations for local selection or reselection.

Introduction

Relieving crops of all sorts of environmental stress, leaves light and temperature, together with varietal characteristics, as the only determinants of yield. That achieved, the same high yielding variety can be used all over a climatic zone, and target area for the variety is enormous. If stress can not be eliminated, however, adaptation to a usually site-specific environment becomes necessary. In that case, the target area for each variety will be small.

Do these different perspectives warrant different approaches to plant breeding? Conventional wisdom says no. In experiments where varieties are tested at various levels of inputs, the same high yielding variety usually comes out as top yielder at all levels. Therefore, plant breeders often claim that their varieties not only perform excellently under high-input conditions, but would also be better than traditional varieties in a low-input environment. However, when these trials are taken outside of experimental farms and the varieties are tested under local or farm conditions, researchers discover what is called 'crossover in performance'. At a certain level of stress there is a crossover point beyond which local varieties or landraces perform better than the high yielding varieties (see Evans, 1993:164-168 and Ceccarelli, 1996).

Farmers who experience such situations are not a tiny minority. To some degree, most if not all farmers have to cope with local stress conditions. The stress-free environment is hard to achieve, and even harder to sustain. In many favorable areas, farmers abandon the high-input technology for economic reasons² or because of ecological problems³, thus increasing the need for locally-adapted germplasm. But modern plant breeding, whether public or private, can not supply adapted germplasm everywhere. Only a system of local seed selection can ensure that. And that means devolution of plant breeding.

Can such decentralized breeding be compatible with development needs in a changing world and meet economic aspirations in a poor society? If the answer to these questions is going to be 'yes', the decentralized breeding must be able to take advantage of the power of science as well as of the capacities of local communities. Three aspects should be considered and made mutually compatible: breeding technology, participatory research methods, and organization at community level.

Technology: toward an evolutionary plant breeding

Exploiting heterogeneity and crop evolution in farmers' fields are outside the scope of most plant breeding research. One exceptional experiment, however, has shed some scientific light on the issue. It was started at the University of California (UC) in 1928. Composite cross populations of barley were produced, some of which were extremely diverse in origin of sources. These populations were exposed to continuous natural selection in current modern farming environments (Allard 1988, 1992) and became the subject of studies during the career span of several generations of UC professors. It appears that after low yields in initial years, the composite cross populations gradually improved in performance and eventually became quite good yielders, with excellent yield stability and disease resistance. These results inspired Suneson (1956) to propose an evolutionary plant breeding method. After assessment of later generations of the same material, Soliman and Allard (1991) concluded that such evolutionary breeding "is unwarranted" if yield potential is the major goal. However, if disease resistance and yield stability are two major objectives, "the composite cross approach is an efficient method". This amounts to saying that a major part of world agriculture, many high-input systems included, could be well served by this approach.

In short, this means constructing a body of broadly diversified germplasm and exposing it to natural selection in areas of contemplated use (Suneson 1956). For those who are familiar with traditional farmers' breeding, this may sound like reinventing the wheel. In fact it is an improvement of the old wheel of plant breeding.

The first step, constructing a body of of broadly diversified germplasm, is not all that straightforward. Science has access to world collections and information sources that are unavailable to farmers. A research institute can chose relevant germplasm and make composite cross populations with an evolutionary potential, most probably far beyond that of locally available varieties.

The immediate outcome, the early generation composite cross population, will be unadapted everywhere and is likely to yield poorly. With time, however, recombinations and natural sorting will improve the adaptation, and, according to the Californian experience, narrow the gap with commercial varieties. The long term outcome could be populations that outperform commercial varieties in disease resistance and yield stability and that may be used as a source of artificial selection for high yield.

The disease resistance appears to have evolved through the building up of polygenic complexes. Therefore, it provides a durable resistance (Allard 1990) as opposed to the monogenic and, therefore, mostly non-durable resistance usually bred into commercial pure line varieties. The stability, at least to some degree, depends on the buffering effect of crop heterogeneity. The Californian experiment shows that when the population is propagated in isolation for a very long time, diversity will start declining, resulting eventually in reduced stability (Soliman and Allard 1991).

Taking the lessons from these experimental findings into the context of current development needs, a few conclusions can be drawn. We need breeding populations with a very high evolutionary potential, and these populations must be exposed to the stress conditions of, or similar to, current farm environments. Furthermore, a certain level of diversity within populations must be maintained in order to sustain evolutionary potential and yield stability.

Landraces are usually found to be heterogeneous. But do they have the evolutionary potential required by current development needs? Are they being subjected to a selection pressure that ensures yield stability, and are they as high yielding as they could be?

Scientific evidence may not be available to give direct answers to such questions. Community visits may not be helpful either. A confusing picture of different selection practices and frequent change of seeds (and therefore lack of persistent long term selection), appears in many communities where traditional seed systems prevail. It is also common to see the coexistence of modern and traditional varieties (Brush 1995). More systematic efforts and a certain level of organization are necessary to make full use of knowledge already existing within a culture, to exploit fully the potential within the locally available germplasm, and to take advantage of opportunities provided by science.

Community organization

Widely diverse forms of organizations dealing with seed management have sprouted up at the community level in recent years. I will present an Ethiopian case representing a traditional society, and a Philippino case representing a modern society. Criteria of classifying these societies as traditional and modern are access to external markets and farm inputs. These were absent in the Ethiopian case and present in the Philippino context.

The Ethiopian case is from Tigray, in the middle of the famous Abyssinian gene center (Berg 1992). Renowned for genetic richness and for knowledge and culture related to seed management, the area should be expected to provide an excellent site for community seed banks. But seed banks were organized more because of poor seeds than because of genetic wealth. Poor seeds were seen as one of the reasons for poor agricultural performance. This was in the 1980s and the area was isolated by war. No external support was possible and community leaders had to look for sources of improvement within their own communities. They knew that those sources existed as, in all societies, there were experts known for their skills in traditional seed selection who had fine local seeds. But they needed an organization to extend the benefits of those experts and those seeds to the wider community.

The Philippine case is from Mindanao, from a community where all farmers have formal school education, and where most of them have taken over their farms after the introduction of modern farming. These farmers had no memory of a pre-Green Revolution practices, such as seed selection and traditional seed management. In recent years, however, some of them have switched to organic farming because of declining profit margins in the high-input system. This has created a need for a different type of seed and also an organization to recover and reintroduce the lost traditional seeds as well as to re-establish a local seed system.

These two cases, one isolated from, and the other influenced by, the modern system required different organizational approaches to the seed management problems. In both cases, however, local seeds and on-farm selection were established as the platform on which to build.

In Ethiopia, the felt problem was poverty and recurrent famines. Community leaders saw poor crop performance associated with poor seeds as one of the causes, but also knew about individuals who had good seeds and had a reputation of being excellent seed selectors. They decided to extend the good seeds through a credit scheme. Community seed banks were established and the local experts on traditional seed selection were used to identify good seeds for lending. Unlike genebanks, they were not concerned with conservation, but rather with the circulation of seeds. Like a commercial bank, they put their capital to work. They also had social concerns and gave priority to loan applicants who were poor and had a particular need for access to good seeds (and who needed protection against private moneylenders). The seed banks were owned by the community and controlled by democratically-elected community assemblies.

The first of these seed banks was established in 1988 and, within a few years, was replicated in most local districts of a region of close to four million people. The growth of the community seed banks continued after peace, in 1991. But now the challenge is twofold: the seed banks must develop in order to remain relevant; and they must protect their integrity and independence when government institutions and seed companies start appearing in the area.

During the war, people had no choice. They had to depend on their own resources. And they proved to themselves that the necessary skills and the required good seeds existed within their own community and could be used to solve their immediate problems. Currently, these Ethiopians do not seem to consider their seed effort as an emergency measure that can be phased out when government services and seed companies begin to function and, perhaps, to take over. They want to keep their seed banks as permanent institutions. Their challenge now is to realize the development potential of these institutions and the evolutionary potential of their seeds within the context of an opened-up economy. But the seed banks and their associated seed supply system are also a challenge for the scientific plant breeding system which is now being re-established in the area. That challenge need not render the local seed supply system obsolete by supplying better commercial seeds; scientists could opt to work with local farmer-breeders in order to ensure that seeds offered through the seed banks remain competitive.

The case from the Philippines, the Community Based Native Seed Research Center (CONSERVE) in Mindanao, arose as a response to critical economic problems of the high-input system of rice cultivation. With increasing prices for inputs and decreasing prices for the produce, profit margins were shrinking, and farmers became dangerously dependent on moneylenders. Some individuals saw a switch to organic farming as the only way out. In that situation, farmers needed a new organization. Unlike the Ethiopian case, where community assemblies representing the entire community were the organizers and owners of the seed activities, this Philippine group was a minority, and membership was individual.

Starting in 1992, this group organized a search for local traditional seeds, which were recovered from farms in isolated remote areas. These seeds were multiplied and distributed to the members for on-farm evaluation and screening. After only two years of operation, I visited the project and found farmers discussing seeds with excitement. From an initial challenge of sorting and selecting among a great number of landraces offered to them, some had already started selecting within landraces, and some had started crossing varieties and keeping written records of what they were doing. They involved their wives and children.

More than a change of seeds had occurred; there was a change of mind also. Before, these farmers grew modern varieties. Such varieties are supposed to be pure, and off-types are considered as impurities. In case they were saving seeds, they had been taught to rogue the off-types before harvest to maintain varietal integrity. The change of mind involved seeing diversity in the field as a resource, rather than an impurity, and seeing themselves as active selectors, rather than passive receivers of ready-made varieties.

It is a mistake to consider this as a rejection of science. It is a withdrawal from the commercial seed system. If it is true that scientific plant breeders are working for farmers and not for seed companies, they might find organized farmer groups another outlet for their scientific achievements. That would require the development of participatory plant breeding methods.

Participatory plant breeding

• The shielding of crops from environmental stress in high-input systems is facing increasing economic and ecological problems⁴. Scientists are changing their attitudes, and a new paradigm is being formulated. Instead of modifying the environment to suit the requirement of high yielding varieties, the varieties need to be modified to suit the environment⁵.

• The claim that modern varieties can be made broadly adapted and be superior across most farming environments within an ecogeographic region is being challenged. A number of recently published reports find varieties with specific local adaptation to perform better when varieties are exposed to local stress environments (Evans 1993, Ceccarelli 1996).

• If relevant diversity exists in a locale, the combined action of natural and artificial selection within a local environment may be an efficient breeding method. Experiments show that this may work also in a fertilizer-intensive system (Allard 1988, Soliman and Allard 1991).

• In recent years farmer groups working with local seeds have been organized all over the world. They are not primarily conservers of old seeds. They want their seeds to be improved, in an evolutionary, slow and steady way, and under their own control.

• Finally, participatory methods have been developed in order to facilitate the involvement of farmers together with scientists as active and equal partners in research to generate relevant farm technology (Mettrich 1993). Such methods can be applied also to plant breeding (Sperling et al. 1993).

It is also commonly observed that farmers change and exchange seeds. A traditional farming system rarely functions as an environment for the static preservation of old landraces.

I have never come across a community where seed management is uniform. Often practices range from neglect to very simple mass selection, but with a few scattered individuals who devote an exceptional amount of effort to the maintenance or improvement of seed quality. These exceptional persons, very often women, may be the source of good seeds for others in the community.

Once such a community is organized for seed management and improvement, it becomes possible for it to establish links to scientific institutions. According to what I have gathered in a number of community visits, there is nothing in traditional attitudes that would make people reluctant to join a participatory breeding scheme -- if it provides them with a wider selection base and prospects of progress through on farm evolutionary breeding.

Community resources for participatory approaches to seed management and breeding are not limited to indigenous culture and traditional practices. The educational status and experiences of modern farmers may also be turned into a resource for community action. In the Philippine group, a number of the members were high school graduates and a few had a university degree. And moreover, most of them had a couple of decades' experience with modern input-intensive farming. Seed activities opened their minds towards the traditional societies, towards themselves and towards the modern world. The traditional societies supplied them with their seeds, and through the seeds, they learned to appreciate the values and achievements of these societies. They discovered their own potential, and saw that the outside world could bring more than technology packets: it could bring knowledge and ideas to be exploited and further developed by themselves.

In conclusion: using biodiversity

Genebanks are organized to serve scientific plant breeding. In recent years genebanks are also being used to supply seeds for re-establishment of landraces that have been lost from disaster areas. This is done or planned for areas in Cambodia, Eritrea, Ethiopia, Somalia and Rwanda. But otherwise, local communities are not yet established as bona fide users of genebank materials. If farmers are organized and linked up to scientific institutions, it would be possible to establish a channel for return of relevant germplasm from genebanks to farm communities.

To some degree the direct return of landraces to areas from where they were originally collected and to other areas with similar agroenvironmental conditions may be warranted. However, it might be more useful if genebank materials are made available in the form of enhanced and enriched populations. Scientific institutions can synthesize such populations, but adaptation and selection should take place in farmers' fields.

References

Allard, R.W., 1990. The genetics of host-pathogen coevolution: Implications for genetic resource conservation. Journal of Heredity, 81:1-6.

Allard, R. W., 1992. Reproductive Systems and Dynamic Management of Genetic Resource. In Dattee, Y.C., C. Dumas and A. Gallais, 1992: Reproductive Biology and Plant Breeding, 325-334. Springer Verlag.

Berg, T., 1992. Indigenous knowledge and plant breeding in Tigray, Ethiopia. Forum for Development Studies, No 1(1992):13-22.

Brush, S.B., 1995. In Situ conservation of landraces in centers of crop diversity. Crop Science, 346-354.

Cassman, K.G., M.J. Kropff, and Yan Zhen-De, 1994. A conceptual framework for nitrogen management of irrigated rice in high-yield environments. In, S.S. Virmani, ed., Hybrid rice technology: new developments and future prospects. Selected papers from the International Rice Research Conference, 81-96. Los Banos, The Philippines :IRRI

Ceccarelli, S., 1996. Positive interpretation of genotype by environment interaction in relation to sustainability and biodiversity. In M. Cooper and G. L. Hammers, eds., Plant adaptation and crop improvement. Wallingford, UK, CAB International; Andhra Pradesh, India, ICRISAT; Los Banos, The Philippines:IRRI (In Preparation).

Evans, L.T., 1993. Crop evolution, adaptation and yield. Cambridge, UK: Cambridge University Press.

International Rice Research Institute (IRRI), 1994. Reversing trends of declining productivity. MEGA projects. Leaflet describing new research projects.

Kannenberg, L.W., 1987. Utilization of Genetic Diversity in Crop Breeding. In, C.W. Yeatman, D. Kafton and G. Wilkes, eds., Plant genetic resources: a conservation imperative. Chapter 9:93-109. Colorado, USA: Westview Press.

Kropff, M.J., K.G. Cassman and H. H. van Laer, 1994. Quantitative understanding of the irrigated rice ecosystem and yield potential. In, S.S. Virmani, ed, op.cit., 97-113.

Mettrich, H., 1993. Development oriented research in agriculture: an ICRA textbook. Wageningen, the Netherlands.

Pingali, P.L., 1993. Opportunities for diversification of Asian rice farming systems: A deterministic paradigm. In, S.M. Miranda & A.R. Maglinao, eds. Irrigation management for rice based farming systems in Bangladesh, Indonesia and the Philippines. 291-325. Colombo, Sr Lanka: International Irrigation Management Institute (IMMI).

Sánchez, P.A., 1994. Tropical Soil Fertility Research: Towards the Second Paradigm. Transactions 15th World Congress of Soil Science (Acapulco Mexico), 1: 65-88.

Soliman, K. M. and R. W. Allard, 1991. Grain Yield of Composite Cross Populations of Barley: Effects of Natural Selection. Crop Science, 31:705-708.

Sperling, L., M. E. Loevinsohn and B. Ntabomvura, 1993. Rethinking the Farmer's role in Plant Breeding: Local Bean Experts and On-station Selection in Rwanda. Experimental Agriculture, 29: 509-519.

Suneson, C. A., 1956. An Evolutionary Plant Breeding Method. Agronomy Journal, 56,188-191.

Footnotes:

I also thank farmers organized in the CONSERVE project at Cotabato in Mindanao, the Philippines, the CONSERVE staff, and their supporters in SEARICE, Manila. They, too, received me with never failing rural hospitality and shared their experiences and views during a visit in August 1994. I also owe much to Ms. Mary Lou L. Alcid, Manila, who reviewed the project together with me.

In the recent couple of years, I have also learned from friends within the Community Biodiversity Development and Conservation Programme (the CBDC-programme) whose vast experience from working with communities in Latin America, Africa and South East Asia has been a great inspiration. The viewpoints and reflections in this paper are, however, entirely my own responsibility. (BACK)

2 Pingali (1993:298) summaries the economic situation for irrigated rice production in this way: "Given low prices, declining or stagnant yields and increasing input costs, the profitability of rice production has been steadily declining". (BACK)

3 the International Rice Research Institute (IRRI) has experienced declining yields in long term experiments with high yielding varieties and the same amount and timing of nitrogen inputs. In the early years of the Green Revolution, dry season yields of 9 t/ha and wet season yields of 6-7 t/ha were common. Today the yields are 6,5 t/ha in the dry season and 4-5 t/ha in the wet season (Cassman et al., 1994). (BACK)

4 See Pingali (1993) for review of economic problems of high-input rice farming, and IRRI (1994) for summary of yield decline data and hypotheses of causal soil environment problems. (BACK)

5 This plant breeding perspective of the new paradigm is discussed by Ceccarelli (1996). The soil science perspective of the same paradigm is formulated in this way: "Rely more on biological processes to optimize nutrient cycling, minimize external inputs and maximize the efficiency of their use. Efficient nutrient management, therefore, is the basis for the second paradigm" (BACK)

6 "The wealth of genetic variation in adaptive responses to soil and climatic conditions conserved in the world's gene banks is little known and less used relative to that in resistance to pests and diseases, but it may yet prove to be the most important genetic resource of all" (Evans 1993:168). (BACK)

Document(s) 21 de 38