Agrobiodiversity and technology in resource-poor farms

Daniel M. Cáceres

Daniel M. Cáceres. Agronomist, Universidad Nacional de Córdoba (UNC), Argentina. M.Ph. in Development Studies, University of Manchester, UK. Ph.D., UNC, Argentina. Professor, UNC, Argentina. Address: Facultad de Ciencias Agropecuarias, Departamento de Desarrrollo Rural, CC 509, 5000, Córdoba, Argentina. e-mail: dcaceres@agro.uncor.edu

SUMMARY

Farms managed by peasants are traditionally considered as containing a high level of agrobiodiversity. However, the internal heterogeneity of their farming systems and the links between agrobiodiversity, technology and peasant livelihood strategies have been much less explored. In order to explore these relationships, a comparison was carried out between two groups of resource-poor farmers in Northeastern Argentina: agroecological farmers and tobacco growers. The results suggest that the high agrobiodiversity observed in these farms rests on four main diversification strategies: genetic, spatial, temporal, and management diversification. Agrobiodiversity is also the result of the type of technologies used within these farms and the conditions in which the productive processes take place. Despite the fact that both groups of farmers have a very similar farm structure, a shared technological matrix and the same fine-grain logic underlying their approach to farming, their farms showed markedly different levels of agrobiodiversity. Agroecological farmers managed more than three times as many species as did tobacco growers. They also devoted significantly more species to self-consumption and self-input. The findings described here have implications for rural development and policymaking, since the embracement of different approaches to farming can produce largely different impacts on both peasant livelihoods and the environment.

AGROBIODIVERSIDAD Y TECNOLOGÍA EN SISTEMAS PRODUCTIVOS CAMPESINOS

RESUMEN

Generalmente se asume que los sistemas productivos manejados por campesinos contienen una alta agrobiodiversidad. No obstante, la alta heterogeneidad de sus explotaciones, y las vinculaciones existentes entre agrobiodiversidad, tecnología y estrategias de reproducción campesina ha sido poco analizada. Para explorar esta relación, se realizó una comparación entre dos grupos de campesinos del Noroeste Argentino: productores agroecológicos y cultivadores de tabaco. Los resultados sugieren que la elevada agrobiodiversidad observada en estos sistemas productivos descansa sobre cuatro estrategias principales de diversificación: genética, espacial, temporal y de manejo. La agrobiodiversidad es también el resultado del tipo de tecnologías usadas en estas explotaciones y de las condiciones en las cuales tiene lugar el proceso productivo. A pesar de que ambos grupos campesinos tienen una similar dotación estructural, comparten la misma matriz tecnológica, y la misma lógica de "grano fino" orientando su abordaje productivo, sus explotaciones muestran niveles muy diferentes de agrobiodiversidad. El número de especies que manejan los campesinos agroecológicos es más de tres veces mayor al manejado por los que cultivan tabaco. Son también quienes dedican una cantidad significativamente mayor de especies destinadas para autoconsumo y autoinsumo. Los resultados de la investigación tienen implicancias para el desarrollo rural y la generación de políticas agropecuarias, ya que optar por uno u otro abordaje productivo puede producir impactos muy diferentes tanto en las estrategias de reproducción social campesinas como en el ambiente.

AGROBIODIVERSIDADE E TECNOLOGIA EM SISTEMAS PRODUTIVOS CAMPONES

RESUMO

Geralmente assume-se que os sistemas produtivos manejados por camponeses contêm uma alta agrobiodiversidade. No entanto, a alta heterogeneidade de suas explorações, e as vinculações existentes entre agrobiodiversidade, tecnologia e estratégias de reprodução camponesa têm sido pouco analisada. Para explorar esta relação, se realizou uma comparação entre dois grupos de camponeses do Noroeste Argentino: produtores agroecológicos e cultivadores de tabaco. Os resultados sugerem que a elevada agrobiodiversidade observada nestes sistemas produtivos descansa sobre quatro estratégias principais de diversificação: genética, espacial, temporal e de manejo. A agrobiodiversidade é também o resultado do tipo de tecnologias usadas nestas explorações e das condições nas quais têm lugar o processo produtivo. Apesar de que ambos grupos camponeses têm uma similar dotação estrutural, compartem a mesma matriz tecnológica, e a mesma lógica de "grão fino" orientando sua abordagem produtiva, suas explorações mostra níveis muito diferentes de agrobiodiversidade. O número de espécies que manejam os camponeses agroecológicos é mais de tres vezes maior a manejado pelos que cultivam tabaco. São também quem dedicam uma quantidade significativamente maior de espécies destinadas para autoconsumo e autoinsumo. Os resultados da investigação têm implicações para o desenvolvimento rural e a generação de políticas agropecuarias, já que optar por uma ou outra abordagem produtiva pode produzir impactos muito diferentes tanto nas estratégias de reprodução social camponesas como no ambiente.

KEYWORDS / Agrobiodiversity / Agroecology / Farming Technology / Food Security / Industrial Agriculture /

Received. 03/28/2005. Modified: 03/27/2006. Accepted: 04/17/2006.

Agrobiodiversity, defined as the number and composition of species cultivated and raised by farmers, is considered an important factor in food production and environmental conservation (Thrupp, 2000). Different authors have reported positive impacts of agrobiodiversity on productive risk reduction (Schejtman, 1975; Ellis, 1992; Reijntjes et al., 1992; Cáceres, 1994; Pretty et al., 1995), productivity (Rosset, 1999; Shiva, 2001; Coghlam et al., 2002), stability and resilience (Altieri, 1995; Reijntjes et al., 1992; Mannion, 1995; Pretty, 1996; Chapin et al., 2000; Tilman, 2000; Shiva, 2001) and food security (Thrupp, 2000). In the past few decades, however, agrobiodiversity has declined dramatically, paralleling the accelerated loss of biodiversity observed at the global scale (Fowler and Mooney, 1990; Heywood, 1995; Mannion, 1995; Madeley, 1999; Shiva, 2000; Teubal, 2001; Grimble and Laidlaw, 2002). This decline in agrobiodiversity is mainly a consequence of the replacement of grazing lands with croplands, the intensification of agriculture, and the widespread use of transgenic seeds (Pretty et al., 2000; Thrupp, 2000; Wood et al., 2000; Benton et al., 2003). Strong genetic erosion has been observed in farming systems in most regions of the world. The number of rice varieties cultivated in China has declined from ~10000 in the ‘70s to ~1000 at present, and Mexican farmers grow just 20% of the corn varieties that they used to cultivate in the past (Shiva, 2000). For certain crops only 1 or 2 varieties are cultivated, instead of the tens or hundreds cultivated before (Pretty, 1996).

This process of agrobiodiversity reduction and homogenization has been often associated to farming systems managed by market-oriented farmers and especially large-scale industrial agriculture. However, there is strong evidence that it is also occurring on farms managed by resource-poor farmers, who have been traditionally considered as keeping much more diverse agroecosystems (Shiva, 2000). In addition to its consequences for the environment and for the future of agriculture in general, this accelerated agrobiodiversity loss in small farms could seriously impact the livelihoods and food security of resource-poor farmers (Cáceres, 2003).

The links between peasant rationale and agrobiodiversity in traditional societies has been well explored (Schejtman, 1975). However, the reasons why some peasants keep high agrobiodiversity systems while others, who share the same logic, do not, have received much less attention. A better understanding of this process and how it is linked with technology and food security may help in slowing down, or even reversing, the loss of agrobiodiversity in resource-poor farming systems.

Two contrasting theoretical approaches have oriented interventions aimed to improve peasant livelihoods, that may have different consequences for agrobiodiversity. On the one hand, the traditional-neoclassical approach argues that farmers can improve their livelihoods by adopting modern, capital intensive, green-revolution-like technologies (Avery, 1995). On the other hand, the agroecological approach couples peasant knowledge and experience together with modern science in the search for a more appropriate and sustainable path to development (Altieri, 1995, 2002). In order to analyze the consequences of the application of these two contrasting development models for agrobiodiversity, two peasant groups settled in Northeastern Argentina were selected. These represent a good model system because the two groups belong to the same socio-cultural tradition, are settled on similar ecosystems, but have developed different productive strategies. Peasants belonging to the first group are agroecological farmers who use few external inputs and rely mostly on local markets. The second group consists of tobacco growers, who depend mostly on modern technology and on the dynamics of agroindustry and external markets.

On these bases, the present article is aimed to answer the following questions: 1) which are the main strategies followed by resource-poor farmers in order to enhance the agrobiodiversity of their farming systems?, 2) what is the relationship between agrobiodiversity and farm technology?, and 3) how do different farming approaches affect agrobiodiversity?. In answering these questions, special attention was paid to farmers’ perceptions of agrobiodiversity, as well as direct observation of their farms.

Study Area

The study was carried out in the Province of Misiones, Northeastern Argentina (Figure 1). The topography is undulated and the climate is subtropical, without dry season, with an annual rainfall of 1900mm and a mean temperature of 20.8ºC, with frosts occurring from May to August (SAGyP, 1995). Soils are fragile with low aptitude for continuous cropping practices and with tendency to nutrient depletion and hydric erosion The potential vegetation is the Paranaense subtropical forest (Cabrera and Willink, 1973). However, only 40% of the original vegetation remains and the rest has been replaced by secondary forest, grazing savannahs and crops (Rosenfeld, 1998).

Agriculture is the main productive activity in Misiones, with mate, tea, tobacco, sugar cane, tung and cattle being the most important items produced. Forestry exploitation, both native and planted tree species, is also very important. According to the last census available there are 27955 farms in the province (INDEC, 2002). The available data shows a noticeable land concentration, with only 0.57% of the farms having more than 1000ha, but accounting for 44.3% of the available farmland (INDEC, 2002). Therefore, the existence of small holders is widespread. Living conditions for these peasants are very basic and do not fulfill their most urgent needs (Cáceres, 2003).

Methodology

This study was focused on two groups of resource-poor farmers living in the Departments of Leandro N. Alem and San Pedro (Figure 1): agroecological farmers and tobacco growers (hereafter AF and TG, respectively). The selection of these two departments had direct relation with the comparative objectives pursued. Leandro N. Alem hosts the four main tobacco companies established in the Province of Misiones, therefore there are many TG in the area. Conversely, San Pedro has witnessed a growing commitment towards organic farming among resource-poor farmers, and several NGOs, peasant organizations and governmental programs have been fostering agroecological practices in the region during the last years.

Both quantitative and qualitative data were gathered in the field in 30 farms (15 TG, 15 AF). Basic information about the farm and the organization of the farming system was collected through a semi-structured survey. This was complemented with in-depth interviews (Woodhouse, 1998), which helped to understand the farmers’ main productive strategies, and the way they were articulated to the context (e.g., agroindustry, markets, and farmers’ organizations).

The number of farmers interviewed was determined using the concept of theoretical saturation (Glaser and Strauss, 1967; Valles, 1997). The interviews were fully recorded with the consent of the interviewees. In addition to the interviews, non-participant observations of the farms belonging to the 30 interviewees were also carried out. They consisted of a thorough in-farm observation of the productive approach and technological management applied by each farmer. These visits allowed a better understanding of the productive strategy followed by each farmer type, and also allowed comparison between the information provided by the interviewees and the actual productive practices developed in their farms. The fieldwork was carried out between June 1999 and January 2000, which is the period of the year with the most intense productive activity. All the data gathered correspond to a single productive cycle.

Data gathered through the semi-structured interviews were organized into quantitative and qualitative variables. After appropriate codification of non-continuous variables, a Mann-Witney U test for independent samples (Sokal and Rohlf, 1995) was applied to each of the variables considered, to test for significant differences between AF and TG. A wealth of information gathered during conversations with the interviewees, but not included in the semi-structured survey, was managed in two ways. Firstly, the most significant parts of farmers’ speech were organized into categories, such as perception of agrobiodiversity, technological approach to farming, etc., following Cáceres et al. (1999). Secondly, the information collected through non-participant observation was used to reach a better understanding of the productive processes. Both types of qualitative data are also used along the paper in order to complement and illustrate the quantitative findings.

Results and Discussion

Agroecological farmers (AF) and tobacco growers (TG) put into effect two contrasting approaches to farming, drawing upon two different models of rural development. This relates not only with the technologies that they use within their farms, but also with the productive strategies developed by each group, and their different socio-economic articulations with the context within which their economic activities take place.

The farming style followed by AF is relatively new in Misiones. During the 80s a few NGOs fostered this more eco-friendly approach to farming. During the 90s it was strongly supported by the main organization representing small farmers in Misiones, the Movimiento Agrario Misionero. It was also promoted by other NGO's and several governmental development programs (e.g. Programa Social Agropecuario and ProHuerta). Besides, a network of approximately 40 grassroots organizations, the Comisión Interferias, was created in order to promote agroecology and to sell organic products in local markets (Schiavoni, 2001; Cáceres, 2003).

On the other hand, TG have been cultivating tobacco for almost a century. Together with mate, tea and tung, tobacco is one of the main crops cultivated in Misiones. Currently, 47% of the farms growing tobacco in Argentina are located in this province. Unlike in other regions of the country, tobacco has been traditionally linked to small farmers. At present, 95% of the land occupied with tobacco in Misiones is cultivated by them (SAGPyA, 2003). TG have a highly asymmetrical relationship with the agroindustry. Farmers are not able to negotiate tobacco prices, nor to decide whom to sell their production, nor to make the main decisions related to the crop (e.g., how many plants to have, which varieties to plant, or what kind of productive inputs to use along the productive process). Thus, this represents a typical contract farming relationship (Watts, 1990).

Keeping a high in-farm agrobiodiversity: how and why?

A close observation of the farming practices carried out by both AF and TG, as well as their interview responses, suggest that they identify four main strategies for diversification, which in turn play a fundamental role in the maintenance of the agrobiodiversity of their farms: genetic, spatial, temporal, and management diversification. What follows is a description of each of these diversification strategies, illustrated with some comments made by the farmers.

Genetic diversification. It consists of the inclusion of a broad gene pool in the design of the farming system. This involves not only several species, but also different varieties and breeds of farm plants and animals. The main explicit aim pursued by farmers through this type of diversification was to spread risks and thus increase the resilience of their farming systems. For example, one of the interviewees highlighted the importance of drawing upon broad genetic pools in order to minimize the risk of total crop loss as follows.

"...It’s important to have several varieties. Because of the resistance. Because not all animals are the same. I look at plants the same way... for example, the gold banana can get sick and another variety may not. With chickens… the same thing. The ecological breed can get sick on you while the common breed won’t, or the common breed could get sick while the ecological one is fine… sickness doesn’t just hit everybody. That’s why you’ve got to have variety… one kind can be a fallback in case the other is wiped out… there’s less weakness".

Spatial diversification. It is defined as the combination of different species within the same farm at the same time. Intercropping, strip cultivation and patchy use of the farmland are the main strategies of diversification of productive space followed by Misiones farmers. This inclination towards spatial diversification mirrors their continuous search for the most suitable associations between the requirements of each species or variety and the conditions of the environment ("fine-grain" logic, see below). The consequence of this approach is the combination of many different plant and animal species per unit area. This was observed in all the farms visited regardless of farmer type, with the sole exception of plots allocated to tobacco cultivation. Due to its high demand of external inputs, and the recommendations made by the tobacco companies, this is usually cultivated as a single crop. In the interviews, farmers stressed the importance of this kind of diversification in keeping plants healthier and protected from insect damage.

"...What we have here makes sense, you know? Everything mixed together (intercropped)...one plant helps the other, they defend each other… There are insects that might like eating lettuce, or that don’t like it. Say you have a bug out there that doesn’t like Swiss chard and stays away from it another bug that doesn’t like rosemary and doesn’t come around here anymore… the plants are protected…"

Temporal diversification. This type of diversification refers to those farming strategies addressed to obtain the production from a certain species during the longest possible period within a year. This is achieved by sowing or planting the same crop variety at appropriate time intervals and/or by choosing varieties with different ripening or harvesting periods. This type of diversification is related to genetic diversification, for instance, when farmers purposely combine different varieties of the same species in order to obtain the output over a longer harvesting period. However, unlike with genetic diversification, the aim of temporal diversification is not to buffer production against climatic events, pest outbreaks, or disease, but rather to spread the output throughout the longest possible period of time.

Temporal diversification can also be applied to livestock production. For instance, the peasants interviewed do not manage bulls separately from cows, which graze together all year round. Against the advice of extension officers, who recommend concentrating calf births to the season of highest forage availability, these farmers prefer not to introduce any restrictions to the reproduction process. Although this approach may not achieve the highest possible productivity, it seems quite functional from the peasants’ perspective: it spreads milk and beef production along the year, avoids sharp productivity fluctuations, and adequately meets both family and market demands. The importance of temporal diversification was highlighted by one of the interviewed farmers:

"...Of fruit trees, we have a diversity… almost all the kinds that are grown in Misiones. We have two varieties of medlars, the big and the small… four varieties of mandarins, two of oranges, two of lemons, 33 types of avocados... distinguished by their size and quality and flavor, and the season when they ripen. We have avocados almost all year long..."

Management diversification. This form of diversification is rarely considered by those who study productive strategies. It refers to the lack of standardization intrinsic to the technological practices that characterize peasant farming systems. In contrast to wealthier, more market-oriented farmers who use mechanized-modern technology, uniformity, homogeneity, or consistency are not typical features of the technology used by Misiones peasants. Drawing upon traditional technologies, such as the use of fire and draft power, peasant farming practices are mostly not mechanized and of an artisanal nature (see discussion on technological matrix below). Consequently, their productive results are highly heterogeneous and show higher variability than those yielded by mechanized and more standardized modern technologies.

At first inspection, this might appear as hardly relevant or even counter-productive to the dynamics of these farming systems. However, this non-standardized and mostly artisanal technological management generates a highly heterogeneous productive output, which fits peasants’ logics and needs very well. It is yet another way to spread risks, maximize the resilience of the systems, and minimize the likelihood of total production loss. Visiting a small greenhouse where a late frost had severely damaged most tomato plants, one of the farmers commented

"...You should never be too tidy... neatness is also punished. The side shoots that I overlooked grew out, while the ones I properly pinched off, didn’t... you have to learn to overlook things and not worry about it… shouldn’t be too perfectionist, because if you are, nature will punish you..."

With this remark, the farmer pointed out that he had not pruned the tomato plants in the way suggested by agronomic science (extensionists advise farmers to prune off all side shoots to keep the plant from getting too bushy). However, the plants that had been "correctly" pruned did not resprout, and those whose side shoots had not been trimmed did. This can be extended to many other frequent practices in peasant farms, such as the pruning of fruit trees, seedling selection, sowing depth and density, manual harvesting, cattle grazing, and livestock health management. The "untidiness" referred to by the farmer is widespread in peasant farms, and it is not only the result of the technology used, but also the consequence of a highly diverse environment and an artisanal approach to farming.

Agrobiodiversity and technology

Some of the arguments above suggest that high agrobiodiversity is the result of a strategy aimed at supplying essential goods, minimizing risks, adjusting to the local environment, and in general fostering the resilience of the farming systems. This is in accordance to the views of several authors (Reijntjes et al., 1992; Altieri, 1995, 1999; Mannion, 1995; Pretty, 1996; Shiva, 2001).

However, a closer observation of the farms suggests that agrobiodiversity is not only the result of a carefully devised productive strategy. The farming technology used by Misiones peasants, and the logic and basic technological matrix underlying it, also play a crucial role in determining the high agrobiodiversity of their farms. These elements are described in the following section.

Agroecological and tobacco farms show strong structural similarities

Farm structure. AF and TG in the study area have the same historical and cultural background. They also share some important features related not only to the structure of their farms (i.e. land, productive capital and labor), but also to the rationale underlying their approach to farming and the technological bases from which very different technological management styles have been developed.

The average land area occupied by both AF and TG was around 25ha. Both types allocated a similar proportion of farmland to crops and neither was using the entire farmland available (Table I). The latter was mostly related to the amount of labor available within the farm and the use of a low-productivity technology (see below).

In terms of productive assets (e.g. machinery, facilities, tools), there was no significant difference between AF and TG (Table I). On average, the productive capital of these peasants includes an oxteam, a cart used to carry their products and family, a plough and a few hand tools (shovels, hoes). TG also had a barn to store and cure tobacco leaves. In both cases farm infrastructure (pens, fences) is scarce and very basic. Only one interviewee owned an old tractor, used to do part of the tillage. For small farmers, cattle availability can also be seen as a crucial productive asset. It plays an important role because it is used as a capitalization strategy and a sort of "saving account" for periods of economic hardship, or when they face extraordinary expenditures (Silvetti, 2003). In the present case, there was no significant difference in number of cattle owned by AF and TG.

Both farmer types relied on family members as the main source of labor. Depending on the period of the year, and in-farm availability, they sometimes either used hired labor or sold their labor outside the farm. The average number of family members working within the farm was 3.47 (TG) and 4.67 (AF), but this difference was not significant (Table I). The only variable that showed borderline significant differences between AF and TG was use of hired labor (Table I). This is consistent with the fact that tobacco cultivation imposes strong peaks of labor demand that cannot be met solely using family labor.

Shared technological "matrix". The two groups of farmers analyzed correspond to two contrasting farming approaches. As a consequence, there are important differences in the "final" technologies used by AF and TG. Whereas AF try to mimic nature, use environmentally-friendly farming practices, and rely mostly on internal inputs (biological control, intercropping, crop rotation and biological fertilizers), TG are committed to industrial agriculture and heavily rely on modern technologies and external inputs (profuse use of inorganic pesticides and fertilisers, monocropping and use of hydroponic seedlings).

However different these "final" technologies may appear, they both rest on two technological cornerstones: the use of fire as a part of shifting cultivation cycles and the use of draft animals as the main source of power within the farm. For a long time, Misiones peasants have been using fire as an effective and affordable way to clear up new plots of land for agriculture (see FAO, 1984; Bandy et al., 1994; Dixon, 1995; Tomich et al., 1998; and O’Brien, 2002 for a debate on the environmental consequences of slash-and-burn practices in tropical and subtropical regions). The use of draft animals is also a common feature in the farming systems of both AF and TG. This is not only because these peasants cannot afford a tractor. It is also the most appropriate technology, considering the difficult conditions in which they have to till the land, which include broken topography, steep slopes, and the presence of numerous obstacles such as unburnt trees, fallen trunks, partially uprooted roots, and rocks. These two technologies (fire and draft animals) embody the technological matrix upon which the whole productive structure of these farms rests. This technological matrix fits what some authors conceptualize as "paleotechnologies" (Wolf, 1966) that sharply contrasts with the "neotechnologies" that are characteristic of industrial agriculture. Paleotechnologies mostly rely on human and animal labor, have a low labor productivity and are not capital-intensive technologies. Therefore, they are most likely available to poor farmers.

"Fine-grain" logic. Both AF and TG share a common logic that underlies their approach to farming. This rationale is described as a "fine grain" logic, which is mostly related with the use they make of the productive space.

The concept of "fine-" or "coarse-grain" logic comes from an analogy with photography: the finer the grain, the sharper the picture will be. Rather than looking at their productive resources in a "coarse-grain" way, both AF and TG seem to carefully scrutinize every portion of their farm in order to choose the most appropriate combination of productive resources for each species or genotype. They are therefore considered to operate within a "fine-grain" logic. This behavior contrasts with that followed by wealthier commercial farmers and also other small farmers in Central and Western Argentina who are mostly extensive livestock raisers (Silvetti, 2003) and tend to follow a "coarse-grain" logic, not paying such a close attention to the environment, or having such a specialized productive response to it. The following commentary an interviewee illustrates this point.

"...A little more over here I had a hard time preparing the soil because the earth here is acidic. The corn doesn’t grow even. It’s patchy soil, some plots are good… you have to choose. The red earth is only useful for growing manioc and the black earth for corn and tobacco..."

This differential response is probably related with the high number of species and varieties that these peasants are able to grow or raise under Misiones’ benign climatic conditions. But also, and mainly, it is related to the high spatial heterogeneity of the farm environment (soil conditions, water availability, topography and aspect). Peasants have developed a special ability to detect subtle environmental variations and try to make the most appropriate decision for every single productive niche within the farm. This fine-grain perception of the conditions in which the productive process takes place permeates the entire approach to farming of both AF and TG. It is at the basis of their spatial diversification strategy and results in a patchy farm landscape with many different kinds of intercropped species.

The common technological matrix and fine grain logic shared by all peasants, inevitably lead to high agrobiodiversity. On the one hand, its low scale makes it possible to combine many different species in small portions of land, and to make available for cultivation tiny patches that could not be taken advantage of by using modern technology. The use of man-powered sowing machines and draft-powered ploughs that work around obstacles and adjust to the irregularities of the soil, and the development of many different manual weeding and harvest techniques, are just some examples that illustrate this point.

On the other hand, high agrobiodiversity is also a product of the lack of standardization of the productive process, which in turn stems from the technological matrix underlying the farming practices developed by peasants. Thus, even if these farmers would prefer to develop low-biodiversity systems, and for instance focus on monocropping, or sow all the crops at the same time, they would find it hard due to the technologies available to them, the drudgery of most farming practices and the artisanal nature of the productive processes in which they are involved.

But AF and TG support very different agrobiodiversity levels

Despite their similarities in structure, basic technological matrix and underlying logic, agroecological and tobacco farms showed important differences in terms of agrobiodiversity. On average, AF managed three times as many species per farm as TG (Tables II and III). Tobacco farms were fairly diverse, but one of the agroecological farms studied contained as many as 54 managed species. Even the least diverse agroecological farm contained more managed species than did the most diverse tobacco farm (21 vs. 18; Table II). AF cultivated significantly more species of annual crops, vegetables, fruit and non-fruit trees, and pasture than TG. Only timber trees and production and work animals were similar in both types of farms.

There were also important differences in the total number of species managed, taking into account all farms within each type (Table IV). Considered together, TG managed a total of 41 species, whereas AF managed 97.

In addition, there were sharp differences between agroecological and tobacco farms in terms of the role played by agrobiodiversity within the farm (Table IV). AF managed almost three times as many species for self-consumption, and more than twice as many species for self-input as TG. However, and consistent with their common technological matrix, both types of farmers raised the same three species of work animals for self-service.

Why do TG and AF show different agrobiodiversity levels?

In order to explain the differences in agrobiodiversity observed between TG and AF, it is necessary to consider both internal and external causes. Firstly, it relates to the design of the farming systems and the productive strategies followed by each farmer type. While TG aim at obtaining the highest possible monetary income (cash crops), AF focus mainly on global income (cash + food + self-input + self-service crops). Secondly, TG devote most of their farm labor force to tobacco cultivation, because the management fostered by tobacco companies heavily relies on the use of external inputs. AF, on the contrary, prefer to allocate their labor to a wider range of crops, because by doing this they not only produce more food, and for longer periods, but also they lower productive and economic risks. And thirdly, TG have a less diversified articulation with the wider context. Unlike TG, who relate with the market almost exclusively through its main product, AF offer many organic products to the local markets. Thus, market demand feeds back onto farming systems’ design, promoting agrobiodiversity.

In a wider context, the sharp differences in farm-level biodiversity can be linked to two contrasting theoretical approaches to farming. On the one hand, the strategy followed by TG relies on the production of a cash crop, uses modern technology, aims to generate monetary income, and shows a strong subordination to the agroindustry. On the other hand, the strategy of AF does not rely on any specific crop, draws upon the use of ecologically sound technology, focuses on the generation of a global income, and has developed a less asymmetrical relationship with the social actors with which they interact. In the middle run, these differences may have a direct impact on the vulnerability level of each social type (Kasperson et al., 2005), the protection of the environment and the generation of new policies for rural development.

Conclusions

The high agrobiodiversity observed in Misiones resource-poor farming systems rest upon four main productive strategies: genetic, spatial, temporal, and management diversification. In the farmers’ perception, these strategies contribute to spread productive risks and develop more stable farming systems. This is in accordance with the views of other authors, reporting on case studies from different regions of the world (e.g. Altieri, 1995; Shiva, 2000), who have stressed that peasants are more interested in maximizing the productivity of mixed species per unit of land, even when this may imply lower yields per specie, than in obtaining the highest possible yield for a single crop, as most capitalized farmers do.

However, high farm agrobiodiversity is not only the result of peasant positive perception of it, and carefully devised strategies aiming to reach their social reproduction threshold ("fine-grain" logic, need to spread risks, fulfill food needs and search for resilience). It is also the inevitable outcome of the technologies used within these farms and the conditions in which the productive processes take place (low labor productivity technology, artisanal approach to farming, labor drudgery and environmental heterogeneity).

The results also indicate that farmers with the same socio-cultural background and similar perception of agrobiodiversity, who operate their farms drawing on the same technological matrix, in the same environment and with a very similar land, labor and productive capital endowment, do manage systems with markedly different agrobiodiversity. This is directly linked with the productive strategies embraced by different farmers, and has relevant consequences for their livelihoods. In the case under study, the fact that AF managed considerably more species in general, and species devoted to self-consumption and self-input in particular, than did TG, resulted in higher autonomy and self reliance.

The findings have implications for rural development and policy making. The embracement of different theoretical approaches to development can produce very different impacts on both farmers and environment. In this study, the agroecological approach appears to be more appropriate than the traditional-neoclassical one, since it better helps to meet some of farmers’ most important needs, keeps agrobiodiversity higher and fosters the design of seemingly more stable farming systems.

Acknowledgements

This paper was written while I was a visiting scholar at the Center for Latin American Studies (Stanford University). I am very grateful to the Instituto de Desarrollo Social y Promoción Humana (INDES), to the Movimiento Agrario Misionero (MAM), to the National University of Córdoba, and to CONICET. I also thank Fernando Casanoves, Dana Gundling, and to two tobacco companies operating in Misiones, Argentina. Despite their kind collaboration, INDES, MAM and the tobacco companies provided no financial support to this research.

References

1. Altieri MA (1995) Agroecology. The Science of Sustainable Agriculture. Westview. Boulder, CO, USA. 433 pp.        [ Links ]

2. Altieri MA (1999) The ecological role of biodiversity in agroecosystems. Agric. Ecosyst. Envir. 74: 19-31.        [ Links ]

3. Altieri MA (2002) Agroecology: the science of natural resource management for poor farmers in marginal environments. Agric. Ecosyst. Envir. 93: 1-24.        [ Links ]

4. Avery D (1995) Saving the Planet with Pesticides and Plastic. The Hudson Institute. Indianapolis, IN, USA. 432 pp.        [ Links ]

5. Bandy D, Garrity D, Sánchez P (1994) El problema mundial de la agricultura de tala y quema. Agroforestería en las Américas 1: 14-20.        [ Links ]

6. Benton TG, Vickery JA, Wilson JD (2003) Farmland biodiversity: is habitat heterogeneity the key?. Trends Ecol. Evol. 18: 182-188.        [ Links ]

7. Cabrera AL, Willink A (1973) Biogeografía de América Latina. OEA. Washington DC, USA. 120 pp.        [ Links ]

8. Cáceres D (1994) Estrategias campesinos y riesgo. Desarrollo Agroforestal y Comunidad Campesina 3: 2-6.        [ Links ]

9. Cáceres D (2003) Los sistemas productivos de pequeños productores tabacaleros y orgánicos de la Provincia de Misiones. Estudios Regionales 4: 13-29.        [ Links ]

10. Cáceres D, Silvetti F, Soto G, Ferrer G (1999) Las Representaciones Tecnológicas de Pequeños Productores Agropecuarios de Argentina Central. Desarrollo Rural y Cooperativismo Agrario 3: 57-79.        [ Links ]

11. Chapin FS, Zavaleta ES, Eviner VT, Naylor RL, Vitousek PM, Reynolds HL, Hooper DU, Lavorel S, Sala OE, Hobbie SE, Mack MC, Díaz S (2000) Consequences of Changing Biodiversity. Nature 405: 234-242.        [ Links ]

12. Coghlam A, Cohen P, Holmes B, Kleiner K, MacKenzie D, Nowak R, Pearce F (2002) Beyond organics. New Scientist 2343: 32-41.        [ Links ]

13. Dixon RK (1995) Sistemas agroforestales y gases de invernadero. Agroforestería en las Américas 2: 22-26.        [ Links ]

14. Ellis F (1992) Peasant Economics - Farm Households and Agrarian Development. Cambridge University Press. Cambridge, UK. 257 pp.        [ Links ]

15. FAO (1984) Improved Production Systems as an Alternative to Shifting cultivation. Soils Bulletin 53. FAO. Rome, Italy.        [ Links ]

16. Fowler C, Mooney P (1990) The Threatened Gene: Food Policies and the Loss of Genetic Diversity. Lutterworth. Cambridge, UK.        [ Links ]

17. Glaser BG, Strauss AL (1967) The Discovery of Grounded Theory. Aldine. Chicago, IL, USA. 281 pp.        [ Links ]

18. Grimble R, Laidlaw M (2002) Biodiversity management and local livelihood: Rio plus 10. Nat. Resour. Persp. 73: 1-4.        [ Links ]

19. Heywood VH (Ed.) (1995) Global Biodiversity Assessment. UNEP / Cambridge University Press. Cambridge, UK. 1140 pp.        [ Links ]

20. INDEC (2002) Censo Nacional Agropecuario. Instituto Nacional de Estadísticas y Censos. Buenos Aires, Argentina. www.indec.mecon.ar        [ Links ]

21. Kasperson RE, Archer E, Cáceres D, Dow K, Downing T, Elmqvist T, Folke C, Han G, Iyengar K, Vogel C, Wilson K, Ziervogel G (2005) Vulnerable Peoples and Places. In Hassan R, Scholes R, Ash N (Eds.) Ecosystems and Human Well-Being. Current State and Trends. Island Press. Washington-Covelo-London. pp. 143-164.        [ Links ]

22. Madeley J (1999) Big Business, Poor People: The Impact of Transnational Corporations on the World’s Poor. Zed Books. London, UK. 192 pp.        [ Links ]

23. Mannion AM (1995) Agriculture and Environmental Change. Temporal and spatial Dimensions. Wiley. New York, NY, USA. 405 pp.        [ Links ]

24. O’Brien WE (2002) The nature of shifting cultivation: stories of harmony, degradation and redemption. Human Ecol. 30: 483-502.        [ Links ]

25. Pretty JN (1996) Regenerating Agriculture. Earthscan. London, UK. 320 pp.        [ Links ]

26. Pretty J, Guijt I, Scoones I, Thompson J (1995) Regenerating agriculture: the agroecology of low-external input and community-based development. In Kirkby J, O’Keefe P, Timberlake L (Eds.) The Earthscan Reader in Sustainable Development. Earthscan. London, UK. pp. 125-145.        [ Links ]

27. Pretty JN, Brett C, Gee D, Hine RE, Mason CF, Morison JIL, Raven H, Rayment MD, van der Bijl G (2000) An assessment of the total external costs of UK agriculture. Agric. Syst. 65: 113-136.        [ Links ]

28. Reijntjes C, Haverkort B, Waters Bayer A (1992) Farming for the Future. An Introduction to Low-External-Input and Sustainable Agriculture. Macmillan. London, UK. 250 pp.        [ Links ]

29. Rosenfeld A (1998) Evaluación de Sostenibilidad Agroecológica de Pequeños Productores (Misiones-Argentina). Thesis. Universidad Internacional de Andalucía, Spain. 195 pp.        [ Links ]

30. Rosset P (1999) Small is Bountiful. Ecologist 29: 452-456.        [ Links ]

31. SAGyP (1995) El Deterioro de la Tierra en la República Argentina. Alerta Amarillo. Secretaría de Agricultura Ganadería y Pesca. Buenos Aires, Argentina. 287 pp.        [ Links ]

32. SAGPyA (2003) Informe de Productos Regionales - Tabaco. Secretaría de Agricultura Ganadería y Pesca. Buenos Aires, Argentina. www.sagpya. mecon.gov.ar        [ Links ]

33. Schejtman A (1975) Elementos para una teoría de la economía campesina: pequeños productores campesinos de hacienda. El Trimestre Económico 42: 487-509.        [ Links ]

34. Schiavoni G (2001) Organizaciones agrarias y constitución de categorías sociales. Plantadores y campesinos en el nordeste de Misiones (Argentina). Estudios Regionales 20: 7-22.         [ Links ]

35. Shiva V (2000) Stolen Harvest. The Hijacking of the Global Food Supply. South End. London, UK. 146 pp.        [ Links ]

36. Shiva V (2001) Globalización y pobreza. LEISA Revista de Agroecología 17: 7-9.        [ Links ]

37. Silvetti F (2003) La cabra es la vaca de los pobres. Los campesinos capricultores del noroeste de Córdoba desde una perspectiva socioantropológica. Ciencias Sociales 2-3: 47-59.        [ Links ]

38. Sokal RR, Rohlf FJ (1995) Biometry. The principles and Practice of Statistics in Biological Research. Freeman. New York, NY, USA. 887 pp.        [ Links ]

39. Teubal M (2001) Globalización y nueva ruralidad en América Latina. In Giarraca N (Ed.) ¿Una Nueva Ruralidad en América Latina? CLACSO/Asdi. Buenos Aires, Argentina. pp. 45-65.        [ Links ]

40. Thrupp LA (2000) Linking agricultural biodiversity and food security: the valuable role of agrobiodiversity for sustainable agriculture. Int. Affairs 76: 265-281.        [ Links ]

41. Tilman D (2000) Causes, consequences and ethics of biodiversity. Nature 405: 208-211.        [ Links ]

42. Tomich TP, Fagi AM, de Foresta H, Michon G, Muryidarso D, Stolle F, van Noordwijk M (1998) Indonesia’s fires: smoke as a problem, smoke as a symptom. Agrofor. Today 10: 4-7.        [ Links ]

43. Valles MS (1997) Técnicas Cualitativas de Investigación Social. Reflexión Metodológica y Práctica Profesional. Síntesis. Madrid, España. 430 pp.        [ Links ]

44. Watts M (1990) Peasants under contract: Agro-food complexes in the Third World. In Bernstein H, Crow B, Mackintosh M, Martin C (Eds.) The Food Question. Profits Versus People? Earthscan. London, UK. pp. 149-162.        [ Links ]

45. Wolf E (1966) Peasants. Prentice Hall. Englewood Cliffs, NJ, USA. 116 pp.        [ Links ]

46. Wood S, Sebastian K, Scherr SJ (2000) Pilot Analysis of Global Ecosystems. Agroecosystems. International Food Policy Research Institute / World Research Institute. Washington DC, USA. 110 pp.        [ Links ]

47. Woodhouse P (1998) People as informants. In Thomas A, Chataway J, Wuyts M (Eds.) Finding Out Fast. Investigative Skills for Policy and Development. Sage/Open University. London, UK. pp. 127-146.        [ Links ]