Document(s) 6 de 38

Abstract: On-farm conservation of plant genetic resources can be defined as the continued cultivation and management of a diverse set of crop populations by farmers in the agroecosystems where a crop has evolved. This set may include weedy and wild relatives of the crop, that though not purposely cultivated, may be present together with it, and in many instances tolerated. On-farm conservation is dynamic and is aimed at maintaining the evolutionary processes that continue to shape this diversity. This paper examines and discusses some of these processes that are common to several major crops, such as maize, rice, potatoes and beans. These are: (1) seed flows; (2) variety selection; (3) variety adaptation; and (4) seed selection and storage. It presents some of the methods and variables that can be used to elicit, describe and measure them.

Introduction

There is a worldwide concern over the loss of the diversity of plant genetic resources. A particular worry is the substitution of a diverse set of genetically variable crop landraces by a few genetically uniform modern varieties (Brush, 1991; Harlan, 1992; Hawkes, 1983; National Research Council, 1993; Plucknett et al., 1987). The need to conserve this diversity has been recognized as important for many decades. It has translated into the creation of gene banks around the world, i.e., ex situ conservation (Hawkes, 1983; Plucknett et al., 1987). Lately, on-farm conservation has been advocated as a complementary method to ex-situ conservation (Altieri and Merrick, 1987; Brush, 1991; International Plant Genetic Resource Institute, 1993; Oldfield and Alcorn, 1987).

On-farm conservation of plant genetic resources can be defined as the continued cultivation and management of a diverse set of crop populations by farmers in the agroecosystems where a crop has evolved. This set may include the weedy and wild relatives of the crop that may be present together with it, and in many instances tolerated. It is dynamic and is aimed at maintaining the evolutionary processes that continue to shape this diversity. It is based on the recognition that, historically, farmers have developed and nurtured crop genetic diversity and that this process still continues among many farmers in spite of socioeconomic and technological changes. It emphasizes the role of farmers for two reasons: (1) crops are not only the result of natural factors, such as mutation and natural selection, but also and particularly, of human selection and management; and (2) in the last instance, farmers' decisions define whether these populations are maintained or disappear.

In spite of the increasing interest in on-farm conservation, we have a very limited knowledge and understanding of it. There have been only a few studies aimed at studying the conservation and management of crop genetic resources among small farmers, e.g., potatoes (Brush, 1992; Brush et al., 1981; Brush et al., 1992; Quiros et al., 1990, 1992; Zimmerer and Douches, 1991), maize (Bellon, 1991; Bellon and Brush, 1994; Brush et al., 1988; Louette, 1994), and in rice (Dennis, 1987; Lambert, 1985; Lando and Mak, 1994; Richards, 1986). Table 1 shows some examples of the numbers of varieties maintained for different crops in several parts of the world.

On-farm conservation is carried out by farmers who are interested and willing to do so. It cannot be imposed on them. Therefore the basis of on-farm conservation should be the farmers that maintain crop infra-specific diversity. This requires a solid understanding of what they do, how they do it and why they do it. Therefore, outside agents, such as scientists, development workers, activists, research institutions, government and non-governmental organizations do not carry out on-farm conservation per se, but can identify opportunities and assist farmers in continuing their efforts at conserving crop diversity.

The purpose of this paper is to address what the farmers do. I argue that the diversity maintained by farmers is not just the set of varieties that they keep, but also the management processes these varieties are subject to and the knowledge that guides these processes. In fact, the specific varieties in the set may change through time. Hence farmers' diversity is a process rather than a state. This process can be referred to as 'farmers management of diversity'. Below I define this term and its components, and present the methods that I used to measure some of these components in a case study among small maize farmers in Mexico.

Farmers' diversity management

Farmers' management of diversity refers to the cultivation of a diverse set of more or less specialized crop populations. These populations are named and recognized as units by the farmers: they are "farmers' varieties" as opposed to the improved varieties. The usually are segregated in space, time and/or use. The set of varieties is formed though a constant process of experimentation, evaluation and selection of existing and new varieties. There are two levels of selection: (a) choosing the varieties to be maintained; and then (b) for each one, choosing the seed stock that will be planted the next season. The selection process is dynamic and is influenced by the supply of populations from other farmers, villages, regions or even countries. This supply may involve new populations, as well as existing ones that a particular farmer may have lost and want to replant. Four components of farmers' management of diversity can be identified: (1) seed flows; (2) variety selection; (3) variety adaptation; and (4) seed selection and storage.

Seed flows

The exchange and transport of germplasm are a common historical pattern all over the world, that currently continues, particularly with the introduction of modern varieties. Several studies have documented the flow of seed of different varieties among small farmers (Cromwell, 1990; Dennis, 1987; Louette, 1994; Sperling and Loevinsohn, 1993). These flows can happen within a village, a region, a country or even among countries. These flows may happen as farmers exchange or market seed among themselves, purchase seed from commercial or government outlets, receive seed as a gift, or collect it from other farmers while traveling. The increasing importance of migration as an economic activity for may farmers may foster these flows. These flows include both traditional and modern varieties.

The collection of landraces and their use for the development of modern varieties, as well as the introduction of these varieties in the farming systems of many farmers has expanded the scope of these flows. Modern varieties incorporate germplasm originated in many different countries. It is common to observe modern and traditional varieties being grown by the same farmers (Bellon, 1991; Brush et al., 1992). Provided that a diverse array of crop populations are maintained, the incorporation of modern varieties in a farming system can increase substantially the diversity present in it (Dennis, 1987).

These flows are important to understanding the diversity present in a location because they are the basis for incorporating new varieties and obtaining materials that have been lost but are desirable. It is not uncommon that a farmer may lose a desired variety by accident, or even if purposely discarded may want to recuperate it (Dennis, 1987; Sperling and Loevinsohn, 1993). Furthermore, these flows may have important genetic implications. They may be an important mechanism for the migration of genes. They may counter genetic drift in small populations, for example, for varieties planted in very small areas (Louette, 1994). In a study of 542 potato tubers collected in 18 markets in Peru, isozyme analysis showed that all genotypes belonged to a single large gene pool with considerable gene flow between cultivars of different groups. The transportation and local commercialization of potatoes in the market served to pool together varieties from different regions (Quiros et al., 1992). In rice, Dennis (1987) has documented for Northern Thailand an active exchange of rice germplasm among Thai farmers across village, district, and provincial lines. This meant that a variety need not stay in the same village to persist successfully within a region. In theory a network of seed exchange coupled with a rigorous and consistent seed selection method that produces high quality seed, may allow farmers to abandon poorer lines whenever there is access to better ones, with the eventual cumulative effect of generating and maintaining highly adapted and productive cultivars (Lambert, 1985).

Variety selection

The process of variety selection can be seen as the farmer's decision to maintain, incorporate or discard a variety to be planted in a particular growing season. The diversity of varieties present in the farmer' fields is the outcome of this decision. Diversity should increase, if the number of varieties incorporated and maintained is larger than the ones discarded, and vice versa. The varieties maintained or incorporated are either kept from the previous agricultural cycle or obtained through exchange or purchase.

Farmers continually evaluate each variety they have. This process has two components. One is to find out how a variety performs with respect to each concern or selection criterion, such as its performance under drought or flood conditions. The second is to rank the performance of the varieties in terms of each of the different selection criterion, i.e., which has the best drought resistance and which the worst. Farmers are in a constant process of trying to match their crop populations or varieties to their concerns, which in turn reflect the conditions under which they farm and their goals. In describing the management of traditional rice varieties in Pesagi, a Malay village, Lambert (1985) observed that farmers constantly experiment with rice cultivars. Even with well-known cultivars, individual households test one variety against another, a process of matching varietal performance to small but significant differences in localized habitats.

The fact that farmers have multiple criteria to select what varieties to plant, as well as where, when and how to do it, has been well established and reflect their concerns (e.g., Bellon, 1991; Brush et al., 1981; Brush, 1992; Lando and Mak, 1994; Lambert, 1985; Sperling et al., 1993). These concerns can be grouped into three major types (Bellon, 1991): (a) agroecological, which refers to the performance of a variety with respect to agroecological conditions, such as rainfall, temperature, soil quality, topography, etc.: (b) use, which refers to the performance of a variety with respect to the destination and uses of the output, such as production for subsistence or for the market, production of straw for fodder, taste, texture, yield, etc.; (c) technological, which refers to the performance of a variety with respect to management and inputs, such as the amount of fertilizer applied, delays in weedings, fitting with other crops, etc.. Table 2 provides examples of different selection concerns for several crops.

Farmers' selection concerns are not homogenous and may vary with different agroecological, socioeconomic and cultural conditions. Rich and poor farmers in a productive region probably have very different concerns, as may poor farmers located in productive versus marginal areas. Even with a farming household, there may be differences between male and female concerns. In many crops, such as rice, there may be a clear sexual division of labor that underlines the possibility not only of different concerns, but also of conflicting ones. This area merits further research, given the increasing recognition of the role of women in farming.

Since farmers' concerns are varied, and a good performance with respect to certain concerns often implies poorer performance with respect to others, several varieties are maintained. It is important to underline that in order to explain the development and maintenance of diversity¹ there should be trade-offs among the varieties. Otherwise there would be no need to have several different varieties. For example, Harlan (1992) points out that alleles for disease resistance generally have negative effects on yield in the absence of the disease and sometimes even in its presence. Hence, there are costs associated to resistance. Therefore, it is important to know and understand not only the positive traits of a variety, but also its negative ones. The combination of these two types of traits defines the opportunities for complementarity among varieties.

Variety selection is a process of continual experimentation and evaluation. Much of the new information acquired is transmitted from farmer to farmer. Experimentation and communication have two important roles in the management of diversity. They are the basis of the development of farmers' crop knowledge, and they allow farmers to know and evaluate new and unproved germplasm--in both cases, without jeopardizing their livelihood or scarce resources.

The fact that many small farmers have a well developed knowledge of their crops and crop varieties has been well documented by human ecologists, anthropologists and ethnobiologists (Bellon, 1991; Berlin et al., 1974; Brush et al., 1981; Boster, 1983; Conklin, 1957; Hames, 1983). This knowledge includes ecological, agronomic and consumption characteristics of the crops and crop varieties they plant. In many instances this knowledge is systematized in a regular system of nomenclature, organized in a taxonomic manner, i.e., folk taxonomies (Brush et al., 1981); it may be used to make decisions regarding management, use, storage, culinary aspects and rituals (Bellon, 1991; Boster, 1983; Hames, 1983; Sutlive, 1978).

Variety adaptation

Whenever a farmer finds a variety that is superior for whatever reason, it will be cultivated under the conditions or for the purposes for which is superior. This process contributes to the development of increasingly adapted crop populations. The stronger and more distinctive the selective pressures, the more specialized populations are likely to be. It has been observed that traditional and modern varieties usually are segregated in different areas of the farm, subject to different management and aimed at different uses (Brush, 1991). Bellon (1991) has shown how small maize farmers in Mexico recognize a differential performance of their varieties to soil quality, fertilizer rates and timing of weeding, and they actually manage them accordingly. In rice, the fact that many farmers match different varieties to different field levels that, in turn, reflect different regimes of water availability, is well documented (Lambert, 1985; Lando and Mak, 1994). Certain varieties have been maintained only for very specialized uses such as making rice-starch cosmetics, medicinal preparations, or traditional snack foods and cakes (Lambert, 1985).

Seed selection and storage

Farmers not only choose what varieties to plant or not to plant, or where and how to manage them, but the also the seed that will be the basis for the next season. Variety selection and management are reinforced by a careful and rigorous selection of the seed. Seed selection procedures vary by crop and its reproductive system. In open-pollinated crops, such as maize, seed selection may be fundamental to maintain the integrity of a variety (at least from the point of view of the farmer), which can be easily lost due to hybridization (Bellon and Brush, 1994; Louette, 1994). This may not be a problem in the case of self-pollinated crop, such as rice or beans, or in vegetatively propagated crop, such as potatoes. It may also be important for a farmer to keep varieties separate, in order to facilitate their identification, and for allocation to specific niches. Even if mixtures are planted, in general they are not a random collection of varieties, but specific combinations. For example, in Uttar Pradesh, India, a popular variety in drought prone areas, called gora, is a mixture of brown, black and straw goras, which differ in drought resistance and grain quality (Vaughan and Chang, 1992).

Seed selection may be also important to identify a new population or variety that may arise due to hybridization or mutation. A harvester may single out seed from one or more plants that are perceived as being entirely different from a known cultivar, in an attempt to originate a new strain (Lambert, 1985). This process may be very important to increase diversity in self-pollinated crops, where hybrids between varieties occur at low rates. Nevertheless, the introduction of modern varieties may modify the systems of seed selection because farmers may purchase seed instead. Dennis (1987) noted that it appeared that the practice of on-farm seed selection was declining as improved seed supplies became more available from government agencies. It is also important to emphasize the role of women as seed selectors in many crops, such as beans and rice. Their knowledge and expertise in this respect are increasingly being documented (Conklin, 1986; Sperling et al., 1993).

The processes identified above, can be conceptualized as the dependent variables that one may want to explain in a study of the bases of on-farm conservation or may want to influence in a project to foster it; they can be described and measured, with some being qualitative and some quantitative (Table 3). The studies reviewed show that there is variation in the number and types of variables maintained, the rates of variety replacement, the directions and types of seed flows, the variety selection criteria and the seed selection methods. This variation may be the result of a number of independent variables. These variables are the environmental, socioeconomic and cultural factors that influence the farmers' decision-making. These variables and the mechanisms by which they act are beyond the scope of this paper.

A case study from Mexico

To illustrate the methods to get and measure some of the variables presented above, I will present my work in Mexico. This work was done in Vicente Guerrero, a community of small peasant maize farmers in central Chiapas. The research site, the methods and results have been extensively presented and discussed elsewhere (Bellon, 1991; Bellon and Taylor, 1993). The methods presented here focus on variety selection and variety adaptation.

These farmers produce maize for self-consumption, as well as important surpluses for the national market. They practice plow and swidden agriculture, in relatively good soils, with favorable weather for maize production. Traditional and modern technologies coexist, with varying rates of fertilizers, herbicides, pesticides and an improved variety in both systems. Nevertheless, several characteristics associated with a more traditional agriculture are still common, such as the use of teams of oxen, wooden plows and the dibbling stick, and the utilization of several maize varieties, including landraces. These farmers have benefitted from the development policies of the Mexican government. They have access to government-provided credit, crop insurance, grain storage and marketing facilities, and produce important maize surpluses, mainly sold to the government marketing facility. Another advantage of this community is that it is located close to the state capital and is connected by a good dirt road.

Although improved varieties have been available for the last thirty years, and some have been adopted, these farmers continue to plant up to 15 different varieties, belonging to six different races. However, there have also been changes in the diversity of the varieties maintained there. New varieties have been introduced and become very important, while others, have decreased in importance, although they have not necessarily been eliminated (Bellon, 1991; Bellon and Brush, 1994). Clearly, maize varietal diversity has been dynamic in this community.

For this study I used participatory observation and questionnaires. Of particular importance were two questionnaires: (1) a variety inventory and (2) a landholding inventory. These two questionnaires allowed me to know the level of diversity maintained by farmers, their variety selection criteria, and the association of particular varieties with particular environments.

Variety inventory

In the variety inventory, I asked each farmer² in my sample to list the names of all varieties he knew of. For each name on the list, I asked: whether he had ever planted it, had planted it and no longer did, or continued to plant it; several plant characteristics such as plant height, days to flowering and color of the grain (during my informal conversations with farmers these characteristics were mentioned as important); whether the variety was planted only to maintain seed (a few farmers had told me that they maintained small plots of varieties with the exclusive purpose of keeping seed of a variety); and finally what were the advantages and the disadvantages of each variety. The answers to the last two questions were open-ended, hence the farmers were able to freely express their thoughts. During my previous conversations with farmers, it became clear that they were aware that each variety had both advantages and disadvantages, and that in their decisions of what variety to plant and how to manage it, they were continually weighing these considerations. The farmers' answers to the last two questions were of two types: they referred to a plant trait, e.g., stature, growing cycle, yield; or to a plant response, e.g., drought resistance, lodging proneness, etc. In most cases, the answers were highly interrelated and referred to the same concern. For example, tall stature was related to lodging proneness. Also, while discussing a specific variety, it was clear that some of the considerations were relative to other varieties. For example, the variety 'A' was considered to be more resistant to lodging that the variety 'B', but less than the variety 'C'.

Then I grouped the farmers' answers in two different ways. First, in order to assess the importance of the traits or responses, the answers referring to a particular trait or response, mentioned for at least one variety, were added together. This measures how many farmers referred to a particular trait and are expressed as a percentage of the total sample size (n = 93)³. Second, each answer involving a particular trait given by each farmer for each variety was counted. This number counts how many of the farmers' answers associated a particular trait with a specific variety. For each variety, this number was expressed as a percentage of all the answers that referred to such a trait. The analysis was limited to the three most important varieties, that represented a modern improved variety (MV), a traditional variety (TV) and an a variety that originally had been improved but that had been in this place for over 30 years, mixing with the local germplasm (IMV).

The farmers' answers reflected concerns with different aspects of maize production and use. These concerns can be grouped into the three major types referred to earlier; agroecological, use and technological. Table 4 presents these concerns, their importance and association with each variety. The farmers believe that each variety performs differently with respect to one or several of these concerns. Hence, they fit different needs or they allow for different management. For example, farmers believe that there are differences in investment and management among varieties. This is described in the folk categories 'aguantador' (sturdy) and 'delicada' (delicate). A sturdy variety was one that could withstand delays in weeding and the application of fertilizer without substantial losses in yield, while a delicate one would suffer important losses under those delays.

The variety inventory provided information on several variables: the varieties known and planted, the selection criteria, and the association between the selection criteria and the varieties. It was shown that for the three most important varieties there was consistency between what farmers said and the way they managed them (Bellon, 1991). There are two central findings of my work with these farmers. First is that they have a set of concerns associated with the performance of their maize varieties, as well as a knowledge base of how they perform for each concern in relationship with each other, and can rank them accordingly. The second is that no variety alone seems to address all of the farmers' concerns, and the process of varietal adoption is more complex than a simple dichotomous decision on adopt/do not adopt. Planting one variety did not prevent a farmer from planting another one (Table 5). Farmers are interested in varieties with contrasting traits that fit different needs and constraints, rather than a single plant with a particular trait, such as high yield.

Landholding inventory

In the landholding inventory I asked each farmer in my sample to list the plots of land he had. For each plot, I also asked: the area, both in hectares and in the local units ('litros'); the land tenure; the types of soils present according to their folk soil taxonomy; and the area occupied by each soil type declared. Then for each soil type in each plot, I asked what crop was planted, and if maize was, what variety was planted that year and the previous year, and finally, if two varieties were planted, what were the reasons.

This information allowed me to cross-check the variety inventory (because the varieties planted declared in both inventories should be the same). It provided me with the area planted per variety, as well as the number of farmers planting each variety. Figure 1 presents these results. Two interesting findings are that: while the traditional variety was only planted in 12% of the area, it was planted by almost 40% of the farmer; and that, while the modern variety was planted by approximately 80% of the farmers, it was only in less than 50% of the area. This also shows that most varieties were individually maintained by few farmers in small areas.

Figure 1. Relative importance of maize varieties by percentage of area and farmers. Source: Bellon and Brush (1994).

This information also allowed me to test the degree of association among varieties and soil types. A colleague and I were able to show statistically that, on average, the improved variety was planted more frequently in the better soils, while the traditional variety was in the worse soil (Bellon and Taylor, 1993).

This information also allowed me to measure the degree of land fragmentation present in farmers' fields, which is related to variety diversity (Bellon and Taylor, 1993; Brush, 1992). This is also related to how different soils are distributed among the farmers. I found that, although there was socioeconomic stratification among farmers, there was no marked concentration of any soil type by any specific socioeconomic group, and soil types are distributed among farmers by their abundance (Bellon, 1994).

Finally I should point out that seed selection, management and storage are extensively presented and discussed elsewhere (Bellon and Brush, 1994).

Conclusions

In this paper I have argued that the focus of on-farm conservation is the farmers' management of diversity, which is not just the set of varieties that they keep but also the management processes these varieties are subject to and the knowledge that guides these processes. Farmers' management of diversity has four components: seed flows, variety selection, variety adaptation and seed selection and storage. These components can be conceptualized as the dependent variables that one may want to explain in a study of the bases of on-farm conservation or may want to influence in a project to foster it. I presented some methods to acquire information on two of these components, variety selection and variety adaptation, through the use of variety inventories and landholding inventories.

References

Altieri, M.A. and L.C. Merrick, 1987. In situ conservation of crop genetic resources through maintenance of traditional farming systems. Economic Botany, 41:86-96.

Bellon, M.R., 1991. The ethnoecology of maize variety management: a case study from Mexico. Human Ecology, 19:389-418.

Bellon, M.R., 1994. Landholding fragmentation: are soil folk taxonomy and equity important? Paper presented at the 1994 Meetings of the Society of Applied Anthropology, April 13-17, Cancun, Quintana Roo, Mexico.

Bellon, M.R. and S.B. Brush, 1994. Keepers of maize in Chiapas, Mexico. Economic Botany, 48: 196-209.

Bellon, M.R. and J.E. Taylor, 1993. Farmer soil taxonomy and technology adoption. Economic Development and Cultural Change, 41:764-786.

Berlin, B., D.E.Breedlove and R.H. Raven, 1974. Principles of Tzeltal plant classification: an introduction to botanical ethnography of a Mayan-speaking community in highland Chiapas. New York, NY:Academic Press.

Boster, J., 1983. A comparison of the diversity of Jivaroan gardens with that of the tropical forest. Human Ecology, 11:47-68.

Brush, S.B., 1991. A farmer-based approach to conserving crop germplasm. Economic Botany, 45:153-165.

Brush, S.B., 1992. Ethnoecology, biodiversity and modernization in Andean potato agriculture. Journal of Ethnobiology, 12:161-185.

Brush, S.B., H.J. Carney and Z.Huaman, 1981. Dynamics of Andean potato agriculture. Economic Botany, 35:70-88.

Brush, S.B., M.Bellon Corrales, and E.Schmidt, 1988. Agricultural development and maize diversity in Mexico. Human Ecology, 16:307-328.

Brush, S.B., J.E.Taylor and M.R.Bellon, 1992. Biological diversity and technology adoption in Andean potato agriculture. Journal of Development Economics, 39:365-387.

Conklin, H.C., 1957. Hanunóo agriculture: a report on an integral system of shifting cultivation in the Philippines. FAO Forestry Development Paper No.12. Rome:FAO.

Conklin, H.C., 1986. Des orientements, de vents, de riz... pour une étude lexicologique des savoirs traditionnels. Journal d'Agriculture Traditionnelle et de Botanique Appliquée, 33:3-10.

Cromwell, E., 1990. Seed diffusion mechanisms in small farmer communities: lessons from Asia, Africa and Latin America. Network Paper 21, Agricultural Administration (Research and Extension) Network. London:Overseas Development Institute.

Dennis, J.V., 1987. Farmer management of rice variety diversity in northern Thailand. Unpublished Ph.D. dissertation, Cornell University. Michigan University Microfilms, Ann Arbor.

Hames, R., 1983. Monoculture, polyculture and polyvariety in tropical forest swidden cultivation. Human Ecology, 11: 16-34.

Harlan, J.R., 1992. Crops and Man. Madison, WI:American Society of Agronomy.

Hawkes, J.G., 1983. The Diversity of Crop Plants. Cambridge, MA:Harvard University Press

International Plant Genetic Resources Institute, 1993. Diversity for development: the strategy of the International Plant Genetic Resources Institute. Rome:International Plant Genetic Resources Institute.

Lambert, D.H., 1985. Swamp Rice Farming: The Indigenous Pahang Malay Agricultural System. Boulder and London:Westview Press.

Lando, R.P.and S. Mak, 1994. Cambodian farmers decisionmaking in the choice of traditional rainfed lowland rice varieties. IRRI Research Paper Series 154. Los Banos:IRRI.

Louette, D., 1994. Gestion traditionnelle de variétés de maïs dans la Réserve de la Biosphère Sierra de Manatlan (RBSM, états de Jalisco et Colima, Mexique) et conservation in situ des ressources génétiques de plantes cultivées. Thèse de l'Ecole National Supérieure Agronomique de Montpellier.

National Research Council, 1993. Managing global genetic resources: agricultural crop issues and policies. National Academy Press, Washington, D.C.

Oldfield, M.L. and J.B. Alcorn, 1987. Conservation of traditional agroecosystems. Bioscience, 37:199-208.

Plucknett, D.L., N.H.J. Smith, J.T. Williams and N.M. Anishetty, 1987. Gene banks and the world's food. Princeton, NJ:Princeton University Press.

Quiros, C.F, S.B. Brush, D.S. Douches, K.S. Zimmerer and G.Huestis, 1990. Biochemical and folk assessment of variability of Andean cultivated potatoes. Economic Botany, 44:254-266.

Quiros, C.C., R.Ortega, L. van Raamsdonk, M.Herrera-Montoya, P,.Cisneros, E. Schmidt and S.B. Brush, 1992. Increase of potato genetic resources in their center of diversity: the role of natural outcrossing and selection by the Andean farmer. Genetic Resources and Crop Evolution, 39:107-113.

Richards, P., 1986. Indigenous Agricultural Revolution. London:Hutchison.

Sperling, L. and M.E.Loevinsohn, 1993. The dynamics of adoption: distribution and mortality of bean varieties among small farmers in Rwanda. Agricultural Systems, 41:441-453.

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.

Sutlive, V.H., 1978. The Iban of Sarawak. Arlington Heights, VA:AHM Publishing Co.

Vaughan, D.A. and T.T. Chang, 1992. In situ conservation of rice genetic resources. Economic Botany, 46:368-383.

Zimmerer, K.S. and D.S.Douches, 1991. Geographical approaches to native crop research and conservation: the partitioning of allelic diversity in Andean potatoes. Economic Botany, 45:176-189.

Footnotes:

1. This refers to diversity that is directly useful. Nevertheless, diversity may be maintained also as an option, because farmers may not know the future benefit or availability of particular varieties, or because humans can value diversity for its own sake, with no ulterior purpose. (BACK)

2. Maize agriculture was done by male farmers exclusively in this community. (BACK)

3. The total sample was 97, but there were four missing farmers because they could not be reached for further interviews. (BACK)

Document(s) 6 de 38