Introduction Agriculture is a process of artificialization of nature. In general, modern agriculture has implied the simplification of the structure of the environment over vast areas, replacing nature's diversity with a small number of cultivated plants and domesticated animals. In fact, the world's agricultural landscapes are planted with only some 12 species of grain crops, 23 vegetable crop species, and about 35 fruit and nut type species, that is no more than 70 plant species spread over approximately 1,440 million hectares of presently cultivated land in the world, a sharp contrast with the diversity of plant species found within one hectare of a tropical rainforest which typically contains over 100 species of trees (Thrupp l998).

But not all forms of agriculture have followed the classic path of artificialization and intensification. In Latin America, systems range from "low intensity" long-fallow swidden to "high intensity" permanent cultivation wherein large areas have been greatly modified from their natural state and are dominated by monocultures. In commercial agricultural areas, natural habitats are lost through expansion of agricultural production, especially of cattle, sugarcane, cotton, soybean, coffee, and (recently) non-traditional export crops. Highly capitalized farms tend either to be on high-quality lands where profitability is contingent on low wages and large landholdings. By contrast, farms of resource-poor peasants tend to be on ecologically marginal lands or on lands recently opened to agriculture. Thus, impoverished farmers lack access to good farmland and capital and are forced by necessity onto remnants of natural areas, which generally occur on steep slopes, along rivers, and in other fragile environments such as forest margins.

In the midst of these extreme types of agriculture, there are, in the region microcosms of traditional farming systems, (i.e., in Mesoamerica, the Andean region, and the Amazon Basin) that have emerged over centuries of cultural and biological evolution and represent accumulated experiences of peasants interacting with the environment without access to external inputs, capital, or scientific knowledge (Chang, 1977; Wilken, 1987). Using inventive self-reliance, experiential knowledge, and locally available resources, indigenous farmers have often developed farming systems with sustained yields (Harwood, 1979; Reinjtes et al., 1992). These agroecosystems, based on the cultivation of a diversity of crops and varieties in time and space, have allowed traditional farmers to maximize harvest security under low levels of technology and with limited environmental impact (Clawson, 1985). There are also several examples of grass-roots rural development programs in Latin America aimed at the maintenance and/or enhancement of biodiversity in traditional agroecosystems, and which represent a strategy which ensures diverse diets and income sources, stable production, minimum risk, efficient use of land resources, and enhanced ecological integrity (Altieri, 1995; Pretty, 1995).

Increasingly, evidence emerging from analysis of traditional agriculture and NGO-led agroecological projects, shows that the combination of stable and diverse production, internally generated and maintainable inputs, favorable energy input/output ratios, and articulation with both subsistence and market needs, comprises an effective approach to achieve food security, income generation, and environmental conservation (Pretty, 1997; Altieri et al., 1998). As it will be argued in this paper, these approaches represent multiple use strategies that enhance the multifunctional nature of agriculture.

The Multifunctional Nature of Traditional Agriculture

Despite the increasing industrialization of agriculture, the great majority of the farmers in the developing world are peasants, or small producers, who still farm the valleys and slopes of rural landscapes with traditional and subsistence methods. It is estimated that in Latin America there are about 16 million peasant units occupying close to 160 million hectares and involving 75 million people, representing two-thirds of the regions total rural population (Ortega 1986).

Many of these agroecosystems are small-scale, geographically discontinuous, and located on a multitude of slopes, aspects, microclimates, elevational zones, and soil types. They also are surrounded by many different vegetation associations,. The combinations of diverse physical factors therefore are numerous and are reflected in the diverse cropping patterns chosen by farmers to exploit site-specific characteristics. Many of the systems are surrounded by physical barriers (e.g., forests, rivers, mountains) and therefore are relatively isolated from other areas where the same crops are grown in large scale. Descriptions of the species and structural diversity and management of these traditional systems are discussed elsewhere (Alcorn, 1984; Altieri et al.,1987; Chang, 1977; Clawson, 1988; Denevan, 1995; Francis, 1986; Toledo et al. 1985).

In many areas, traditional farmers have developed and/or inherited complex farming systems, adapted to the local conditions helping them to sustainably manage harsh environments and to meet their subsistence needs, without depending on mechanization, chemical fertilizers, pesticides or other technologies of modern agricultural science (Altieri, 1995). According to Toledo (1995), indigenous farmers in the hot and humid tropical regions of Latin America tend to combine various production systems as part of a typical household resource management scheme (Figure 1):

1. The milpa system, which may constitute a system of polyuculture including up to 20-25 agricultural and forest species (annual and perennial) and is focused on the cultivation of maize, but in many occasions is combined and even substituted by agricultural market-oriented products (hot pepper, rice, sesame seeds, sugarcane, beans, etc.); 2. The extraction of products from the primary or secondary rainforests of different ages undergoing the succession process; 3. The manipulation of forest-unit sequences at different stages of anthropic disturbance, from which certain marketable products (mainly coffee, vanilla, and cocoa) are obtained; 4. The management of home gardens, which are agroforestry systems located next or close to households. The main features underlying the sustainability of these multiple use peasant systems are (Marten, 1986; Reinjtes et al. 1992): * Farms are small in size with continuous production serving subsistence and market demands * Maximum and effective use of local resources and low dependence on off-farm inputs * High net energy yield because energy inputs are relatively low * Labor is skilled and complementary, drawn largely from the household or community relations. Dependency on traction and manual labor shows favorable energy input/output ratios * Heavy emphasis is on recycling of nutrients and materials * Building on natural ecological processes (e.g., succession) rather than struggling against them * Diversified farm systems based on several cropping systems, featuring mixtures of crops, and crops with varietal and other genetic variability.

* A salient feature of traditional farming systems is their degree of plant diversity, generally in the form of polycultures and/or agroforestry patterns (Clawson, 1985). This peasant strategy of minimizing risk by planting several species and varieties of crops stabilizes yields over the long term, promotes diet diversity, and maximizes returns under low levels of technology and limited resources (Richards, 1985). Traditional multiple cropping systems provide as much as 20 percent of the world food supply (Francis, 1986). Polycultures constitute at least 80 percent of the cultivated area of West Africa, while much of the production of staple crops in the Latin American tropics occurs in polycultures (Table 1). Polycultures produce more combined yield in a given area than could be obtained from monocultures of the component species. Most traditional polycultures exhibit LER values greater than 1.5. Moreover, yield variability of cereal/legume polycultures are much lower than for monocultures of the components (Table 2).

* TABLE 1: Prevalence of Polycultures in Latin American Countries.1 * * Country Dominant Crop Percentage of Crop Grown in Polyculture Brazil Maize 11 Colombia Rice 6 Dominican Republic Maize 40 Guatemala Beans 73 Mexico Maize 20 Paraguay Beans 33 Maize 10 Sweet Potatoes 10 Venezuela Rice 16 Maize 33 Beans 20 Cassava 20 Cotton 50 1Modified after Francis (1986).

 TABLE 2 Coefficient of variability of yields registered in different cropping systems during 3 years in Costa Rica. Cropping system Monoculture (mean of sole crops) Polyculture Cassava/bean 33.04 27.54 Cassava/maize 28.76 18.09 Cassava/sweet potato 23.87 13.42 Cassava/maize/sweet potato 31.05 21.44 Cassava/maize/bean 25.04 14.95 Source: Francis 1986

Many traditional agroecosystems are located in centers of crop diversity, thus containing populations of variable and adapted land races as well as wild and weedy relatives of crops. It is estimated that throughout the Third World more than 3,000 native grains, roots, fruits and other food plants can still be found (Altieri and Merrick, 1987). Thus traditional agroecosystems essentially constitute in-situ repositories of genetic diversity (Altieri et al. 1987). Descriptions abound regarding systems in which tropical farmers plant multiple varieties of each crop, providing both intraspecific and interspecific diversity, thus enhancing harvest security. For example, in the Andes, farmers cultivate as many as 50 potato varieties in their fields (Brush et al. 1981). Similarly, in Thailand and Indonesia, farmers maintain a diversity of rice varieties in their paddies which are adapted to a wide range of environmental conditions, and regularly exchange seeds with neighbors (Grigg, 1974).

Tropical agroecosystems composed of agricultural and fallow fields, complex home gardens, and agroforestry plots, commonly contain well over 100 plant species per field and provide construction materials, firewood, tools, medicines, livestock feed, and human food. Home gardens in Mexico and the Amazon display highly efficient forms of land use, incorporating a variety of crops with different growth habits. The result is a structure similar to a tropical forest, with diverse species and a layered configuration (Brookfield and Padoch, 1994). A list of the most common agroforestry systems prevalent in Latin America is provided in Table 3.

TABLE 3: Principal Agroforestry Systems in Latin America Types of Systems Examples Typical Countries A.Agro-silvicultural systems A.1 Taungya Cordia alliodora + maize, beans or rice Brazilian Amazon Caesalpina velutina + maize Guatemala Gmelina arborea + maize and beans Mexico A.2 Wood-producing trees/ annual crop intercropping Pinus ellioti +soybean or maize Argentina Populus spp. + maize or potato Argentina Inga spp. + rice or banana Brazil Eucalyptus spp. + maize Brazil Cedrela odorata + maize, rice or sugar cane Colombia Spondia mombin or Swietenia macrophylla + maize, beans or rice Mexico A.3 Fruit trees annual crops Citrus, apples, papaya, mangoes, etc. + annual crops Mexico A.4 Shade trees or soil improvers mixed with crop Erythrina spp., Inga sp., Albizzia carbonaria, Cordia alliodora, etc. + coffee, banana Colombia, Costa Rica, Equador A.5 Living fences and/or windbreaks Gliricidia sepium, Erythrina abissinica, Leucaena leucocephala, etc., around crops Colombia, Mexico, Dominican Republic, Cuba, Guatemala Eucalyptus, Populus, Pinus, around crops Chile, Argentina, Uruguay B. Agrosilvopastoral systems B.1 Crops and animals within forest plantations Pinus caribaea + sheep and/or poultry + sorghum, maize, cassava or peanuts Venezuela, Dominican Republic B.2 Living fences around rural communities Casuarina equisetifolia Cedrela odorata, Bromissum alicastrum Cuba, Mexico B.3 Home gardens Several tree, crop, animal mixtures Dominican Republic, Mexico, Cuba, Haiti C. Silvopastoral systems C.1 Animal grazing or forage production under trees Populus sp. + Bromus unioloides or Trifolium sp. Argentina Pinus caribea + Anchrus sp. Brazil Pinus sp. Or Populus sp. + sheep Chile C.2 Animal grazing or forage production within secondary forests Prosopis flexuosa and Aspidosperma sp. with natural pasture Argentina Secondary forests with browsing of Brosimun alicastrum Mexico C.3 Commercial wood- producing trees with pastures Alnus acuminata + Pennisetum clandestinum Costa Rica C.4 Shade trees or soil improvers within pastures Alnus jorullensis + P.clandestinum Colombia Prosopis sp., Parkinsonia microphylla, Cercidium sp. as shade tress in pastures Mexico C.5 Forage trees and shrubs Prosopis spp., Atriplex spp. Chile, Argentina, Peru Lividivia coriari and P. juliflora for goats Colombia Brosium alicastrum for browsing Mexico Source: FAO 1984

Small areas around peasant households commonly average 80-125 useful plant species, many for food and medicinal use (Toledo et al. 1985; Alcorn, 1984). Perennials such as fruit trees are a conspicuous feature of most homegardens (Marten, 1986). In some of the more humid areas, there are so many different kinds of trees and field cops in the homegardens, and they are growing in such abundance that it looks more like a tropical forest than a garden (Clarke and Thaman, 1993). Most diverse homegardens are in reality a collection of domesticated and semi-domesticated plants with a variety of uses including food, fuel, construction materials, herbal medicine, ornamentation, and shade (Table 4). Homegardens are often in continuous production throughout the year and lend themselves to intensive care because they are so conveniently close to the house. They can be fertilized with kitchen wastes, receive supplementary irrigation with well water, and be attended by women and children in their spare time.

TABLE 4: Ecological and cultural functions and uses of trees in Latin America Ecological Shade Soil improvement Animal/plant habitats Erosion control Frost protection Flood/runoff control Wind protection Wild animal food Weed /disease control Cultural/Economic Timber (commercial) Broom Prop or nurse plants Timber (subsistence) Parcelling/wrapping Staple foods Fuelwood Abrasive Supplementary foods Boat building (canoes) Illumination/torches Wild/snack/emergency foods Sails Insulation Tools Decoration Species/sauces Weapons/hunting Body ornamentation Teas/coffee Containers Cordage/lashing Non-alcoholic beverages Woodcarving Glues/adhesives Alcoholic beverages Handicrafts Caulking Stimulants Fishing equipment Fibre/fabric Narcotics Floats Dyes Masticants Toys Plaited ware Meat tenderizer Switch for children/discipline Hats, mats Preservatives, medicines Brush/paint brush Baskets Aphrodisiacs Musical instruments Commercial/export products Fertility control Cages/roosts Abortifacients Tannin Ritual exchange Scents/perfumes Rubber Poisons Recreation Oils Insect repellents Magico-religious Toothbrush Deodorants Totems Toilet paper Embalming corpses Subjects of mythology Fire making Love-making sites Secret meeting sites Source: Clarke and Thaman 1984.

The interface of traditional agroecosystems and natural areas

Most of the above studies of traditional agriculture have focused on the productive units where crops are grown. This limited view of the peasant agroecosystem ignores the fact that many peasants utilize, maintain, and preserve, within or adjacent to their properties, areas of natural ecosystems (forests, hillsides, lakes, grasslands, streamways, swamps etc.) that contribute valuable food supplements, construction materials, medicines, organic fertilizes, fuels, religious items, etc. (Toledo et al. 1985). In fact, the crop-production units and adjacent ecosystems constitute a continuum where plant gathering, fishing, and crop production are actively produced. For many peasant societies, agriculture is considered a part of a bigger system of land use. For example, the P'urhepecha Indians who live in the region of lake Patzcuaro in Michoacan, Mexico, in addition to agriculture, gathering is part of a complex subsistence pattern based on multiple uses of their natural resources (Caballero and Mapes, 1985). These people use more than 224 species of wild native and naturalized vascular plants for dietary, medicinal, household, and fuel needs. Similarly, the Jicaque Indians of central Honduras, who live on the Montana de la Flor reservation, use over 45 plant species from the pine-oak forest, riverine habitat, or dooryard as foods, medicines, fuel, etc. Like their mestizo neighbors, the Jicaque grow corn using slash and burn techniques. The cultivated fields are widely spaced throughout the forest and in travelling from one field to the next, the Jicaque usually collect wild plant food along the way to be added to the cooking pots of the family's compound (Lentz, 1986)

Agriculture- natural ecosystem interfaces are of key significance as it has been shown that farmers accrue general ecological services from natural vegetation growing near their properties. For example, in many highland regions of Central America, the indigenous flora of the higher forests, not only provide valuable native plants for commercial and subsistence products, but also serve as natural barriers to the lowland agricultural crops against the spread of plant diseases and insect pests. Also, clearing comparatively small agricultural plots in a matrix of secondary forest vegetation permits easy emigration of natural enemies of insect pests from the surrounding jungle (Altieri, 1984).

In western Guatemala, small farms depend on nearby forests to manage marginal infertile soils. Leaf litter is carried from nearby forests and spread each year over intensively cropped vegetable plots to improve tilth and water retention. Litter is raked up, placed in bags or nets, and carried to fields by men or horses, or from more distant sources, by trucks. After spreading, the leaf litter is worked into the soil with a broad hoe. In some cases, litter is first placed beneath stable animals, and then, after a week or so the rich mixture of pulverized leaves, manure, and urine is spread over the fields and turned under. Although the quantities applied vary, farmers in Almolonga, Zunil, and Quezaltenango apply as much as 40 metric tons of litter/ha. each year. Rough calculations made in mixed pine-oak stands indicate that one hectare of cropped land requires the litter production from 10 ha. of regularly harvested forest, or less, if harvesting is sporadic (Wilken, 1987).

A case study of a multifunctional traditional farming system

The study conducted in a Totonaca native community of the Papantla region in the state of Veracruz illustrates of a case multiple use peasant management strategy of hot and humid tropical ecosystems. The community entails 166 households totaling a population of 877 and sharing a 15-17 hectare territory. Most households (72%) have between 7 and 9 hectares, while only 9 % own more than nine hectares and 19% less than seven hectares. Most of these households also handle from 3 to 9 ecogeographic or landscape units as resources for production where they implement the multiple-use strategy. The main units that each family manages during production are: milpa (maize fields), pasture ground, home gardens, rainforest for vanilla production, rainforest to extract wood and other products, and cash crop areas (Figure 2).

Using almost exclusively its own physical energy (with scant, almost inexistent use of chemical fertilizers), making little use of outside inputs, and relying on family or community labor, the productive units of this native community are self-sufficient in terms of food, they are energy efficient, they do not generate waste, and they sustain a high level of agrobiodiversity (with 355 species of plants, animals, and fungi). To this should be added the fact the community succeeds in being economically profitable as a result of selling maize, beef, milk, vegetable, fruits, vanilla, brown sugar, palm leaves and other products (Toledo, 1995).

The Nature and Function of Biodiversity in Agriculture

Today, scientists worldwide are increasingly starting to recognize the role and significance of biodiversity in the functioning of agricultural systems (Swift et al., 1996). Research suggests that whereas in natural ecosystems the internal regulation of function is substantially a product of plant biodiversity through flows of energy and nutrients and through biological synergisms, this form of control is progressively lost under agricultural intensification and simplification, so that monocultures, in order to function, must be predominantly subsidized by chemical inputs (Swift et. al. 1996). Commercial seed-bed preparation and mechanized planting replace natural methods of seed dispersal; chemical pesticides replace natural controls on populations of weeds, insects, and pathogens; and genetic manipulation replaces natural processes of plant evolution and selection. Even decomposition is altered since plant growth is harvested and soil fertility maintained, not through nutrient recycling, but with fertilizers.

One of the most important reasons for maintaining and/or encouraging natural biodiversity is that it performs a variety of ecological services (Altieri, 1991). In natural ecosystems, the vegetative cover of a forest or grassland prevents soil erosion, replenishes ground water, and controls flooding by enhancing infiltration and reducing water runoff. In agricultural systems, biodiversity performs ecosystem services beyond production of food, fiber, fuel, and income. Examples include, recycling of nutrients, control of local microclimate, regulation of local hydrological processes, regulation of the abundance of undesirable organisms, and detoxification of noxious chemicals. These renewal processes and ecosystem services are largely biological, therefore their persistence depends upon maintenance of biological diversity. When these natural services are lost due to biological simplification, the economic and environmental costs can be quite significant. Economically in agriculture, the burdens include the need to supply crops with costly external inputs, since agroecosystems deprived of basic regulating functional components lack the capacity to sponsor their own soil fertility and pest regulation. As functional biodiversity decreases, the requirement for higher management intensity increases, thus monocultures must be subsidized with external inputs (Figure 3). Often, the costs involve a reduction in the quality of the food produced and of rural life in general due to decreased soil, water, and food quality when erosion and pesticide and/or nitrate contamination occurs (Altieri, 1995).

Biodiversity refers to all species of plants, animals and microorganisms existing and interacting within an ecosystem. In agroecosystems, pollinators, natural enemies, earthworms, and soil microorganisms are all key biodiversity components that play important ecological roles thus mediating processes such as genetic introgression, natural control, nutrient cycling, decomposition, etc. (Figure 4). The type and abundance of biodiversity in agriculture will differ across agroecosystems which differ in age, diversity, structure, and management. In fact, there is great variability in basic ecological and agronomic patterns among the various dominant agroecosystems. In general, the degree of biodiversity in agroecosystems depends on four main characteristics of the agroecosystems (Southwood and Way, 1970):

1. the diversity of vegetation within and around the agroecosystem 2. the permanence of the various crops within the agroecosystem 3. the intensity of management 4. the extent of the isolation of the agroecosystem from natural vegetation In general, agroecosystems that are more diverse, more permanent, isolated, and managed with low input technology (i.e. agroforestry systems, traditional polycultures) take fuller advantage of work done by ecological processes associated with higher biodiversity than highly simplified, input-driven and disturbed systems (i.e. modern row crops and vegetable monocultures and fruit orchards) (Altieri, 1995).

All agroecosystems are dynamic and subject to different levels of management so that the crop arrangements in time and space are continually changing in the face of biological, cultural, socio-economic, and environmental factors. Such landscape variations determine the degree of spatial and temporal heterogeneity characteristic of agricultural regions, which in turn conditions the type of biodiversity present.

According to Vandermeer and Perfecto (1995), two distinct components of biodiversity can be recognized in agroecosystems. The first component, planned biodiversity, is the biodiversity associated with the crops and livestock purposely included in the agroecosystem by the farmer, and which will vary depending on management inputs and crops spatial/temporal arrangements. The second component, associated biodiversity, includes all soil flora and fauna, herbivores, carnivores, decomposers, etc., that colonize the agroecosystem from surrounding environments and that will thrive in the agroecosystem depending on its management and structure. The relationship of both biodiversity components is illustrated in Figure 3. Planned biodiversity has a direct function, as illustrated by the bold arrow connecting the planned biodiversity box with the ecosystem function box. Associated biodiversity also has a function, but it is mediated through planned biodiversity. Thus, planned biodiversity also has an indirect function, illustrated by the dotted arrow in the figure, which is realized through its influence on the associated biodiversity. For example, the trees in an agroforestry system create shade, which makes it possible to grow only sun-tolerant crops. So the direct function of this second species (the trees) is to create shade. Yet along with the trees might come small wasps that seek out the nectar in the tree's flowers. These wasps may in turn be the natural parasitoids of pests that normally attack the crops. The wasps are part of the associated biodiversity. The trees, then, create shade (direct function) and attract wasps (indirect function) (Vandermeer and Perfecto, 1995).

The key is to identify the type of biodiversity that is desirable to maintain and/or enhance in order to carry out ecological services, and then to determine the best practices that will encourage the desired biodiversity components. As shown in Figure 5, there are many agricultural practices that have the potential to enhance functional biodiversity, and others that negatively affect it. The idea is to apply the best management practices in order to enhance and/or regenerate the kind of biodiversity that can subsidize the sustainability of agroecosystems by providing ecological services such as biological pest control, nutrient cycling, water and soil conservation, etc. The link between agrobiodiversity and multifunctionality

When agricultural development takes place in a natural environment, it tends to result in a heterogeneous mosaic of varying types of habitat patches spread across the landscape. The bulk of the land may be intensely managed and frequently disturbed for the purposes of agricultural production, but certain parts (wetlands, riparian corridors, hillsides) may be left in a relatively natural condition, and other parts (borders and strips between fields, roadsides, and adjacent natural areas) may occasionally be disturbed but not intensely managed. In addition, natural ecosystems may surround or border areas in which agricultural production dominates (Gliessman, 1998).

The heterogeneity of the agricultural landscape varies greatly by region. In some parts of Latin America, where commercial, export agriculture predominates, the heavy use of agricultural chemicals, mechanical technology, narrow genetic lines, and irrigation over large areas have made the landscape relatively homogenous. In such areas, the agricultural landscape is made up mostly of large areas of single crop agricultural production. The expansion of such agricultural landscapes disrupts natural areas in three important ways. First, natural ecosystems become fragmented and important ecological linkages may be changed or uncoupled. For example, the conversion of uplands from native grasslands or deciduous forest to cotton will profoundly affect the nutrient and pesticide inputs into any adjacent wetlands. Second, the fragmentation increases boundary phenomena by increasing the proportion of area that is near a boundary. This results in an exacerbation of the impacts from adjacent agriculture. Third, the absolute loss of natural areas generally means that the remaining patches are increasingly more distant from each other. Thus each remnant takes on more and more the properties of oceanic islands in the sense that source areas for recolonization are often very distant. Thus, local extinction events for both species and genes are unlikely to be balanced by recolonization or gene flow. Unlike real islands, remnant patches of natural ecosystems are highly vulnerable to invasion by weedy plants and animals from surrounding agricultural lands and are vulnerable as well to perturbations created by agricultural production practices (Fry, 1995).

In peasant dominated areas, the use of traditional farming practices with minimal industrial inputs has resulted in a varied, highly heterogeneous landscape-possibly even more heterogeneous than would exist naturally. In such heterogeneous environments, natural and semi-natural ecosystem patches included in the landscape can become a resource for agroecosystems. An area of non-crop habitat adjacent to a crop field, for example, can harbor populations of natural enemies which can move into the field and parasitize or prey upon pest populations. (Altieri, 1994) A riparian corridor vegetated by native plant species can filter out dissolved fertilizer nutrients leaching from crop fields, promote a presence of beneficial species, and allow the movement of native animal species into and through the agricultural components of the landscape.

On the other hand, agroecosystems can begin to assume a positive rather than a negative role in preserving the integrity of natural ecosystems. Many small scale-diversified agroecosystems have been designed and managed in ways that make them more friendly to native species. For example, by encouraging hedgerows, vertebrates can be provided with large habitats, better food sources, and corridors for movement. Native plants can have more suitable habitats and find fewer barriers to dispersal. Smaller organisms, such as below ground microbes and insects, can flourish in organically managed soils and thus benefit other species since they are such important elements in ecosystem structure and function (Glissman, 1998).

By managing agricultural landscapes from the point of view of biodiversity conservation as well as sustainable production, the multiple use capacity of agriculture can be enhanced providing several benefits simultaneously (Thrupp, 1998):

* increase agricultural productivity; * build stability, robustness, and sustainability of farming systems; * contribute to sound pest and disease management; * conserve soil and increase natural soil fertility and soil health * diversify products and income opportunities from farms; * add economic value and increase net returns to farms; * reduce or spread risks to individuals, communities, and nations; * increase efficiency of resource use and restore ecological health; * reduce pressure of agriculture on fragile areas, forests, and endangered species; * reduce dependency on external inputs, and; * increase nutritional values and provide sources of medicines and vitamins.

* The effects of agrobiodiversity in mitigating extreme climatic effects, such as the drought promoted by El Niño's were recently evident in northern Honduras. An agroforestry project reviving the Quezungal method, an ancient agricultural system, speared about 84 farming communities from destruction. Farmers using the method lost only 10 percent of their crops in 1998's severe drought, and actually obtained a grain surplus of 5-6 million pounds in the wake of Hurricane Mitch. On the other hand, , nearby communities which continued the use of slash and burn, were severely affected by El Niño phenomena, which left a legacy of human misery and destruction of vitally important watersheds.

* Such agroforestry programs which reduce deforestation and burning of plant biomass can provide a sink for atmospheric carbon dioxide and also considerably reduce emissions of nitrous oxide. Recent research shows that promoting techniques already familiar to thousands of small farmers in Latin America such as, crop rotation and cutting back on chemical fertilizers through the use of composting can act as important sinks for atmospheric carbon dioxide storing it below the soil surface.

* The benefits of agrobiodiversity in enhancing the multifunctional agriculture extend beyond the above described effects as shown by the impacts of shaded coffee farms in Latin America. Farmers typically integrate into their coffee farms many different leguminous trees, fruit trees, and types of fuel wood and fodder. These trees provide shade, a habitat for birds and animals that benefit the farming system. In Mexico, shade coffee plantations support up to 180 species of birds, including migrating species, some of which play key roles in pest control and seed dispersal.

* Learning how to manage an agriculture that promotes both environmental as well as productive functions will require inputs from disciplines not previously exploited by scientists, including agroecology, ethnoscience, conservation biology, and landscape ecology. The bottom line, however, is that agriculture must adopt ecologically sound management practices, including diversified cropping systems, biological control and organic soil management as replacements for synthetic pesticides, fertilizers, and other chemicals. Only with such foundation can we attain the goal of a multifunctional agriculture.

* Biodiversity and pest management

* Nowhere are the consequences of biodiversity reduction more evident than in the realm of agricultural pest management. The instability of agroecosystems becomes manifest as the worsening of most insect pest problems is increasingly linked to the expansion of crop monocultures at the expense of the natural vegetation, thereby decreasing local habitat diversity (Altieri and Letourneau, 1982). Plant communities that are modified to meet the special needs of humans become subject to heavy pest damage and generally the more intensely such communities are modified, the more abundant and serious the pests. The effects of the reduction of plant diversity on outbreaks of herbivore pests and microbial pathogens is well-documented in the agricultural literature (Andow, 1991; Altieri, 1994). Such drastic reduction in plant biodiversity and the resulting epidemic effects can adversely affect ecosystem function with further consequences on agricultural productivity and sustainability.

* In modern ecosystems, the experimental evidence suggests that biodiversity can be used for improved pest management (Altieri and Letourneau, 1994: Andow, 1991). Several studies have shown that it is possible to stabilize the insect communities of agroecosystems by designing and constructing vegetational architectures that support populations of natural enemies or that have direct deterrent effects on pest herbivores. For example, at the landscape level, data demonstrates that there is enhancement of natural enemies and more effective biological control where wild vegetation remains at field edges and in association with crops (Altieri, 1994). These habitats may be more important as overwintering sites for predators or they may provide increased resources such as pollen and nectar for parasitoids and predators form flowering plants (Landis, 1994). Many studies have documented the movement of beneficial arthropods from margins into crops and higher biological control is usually observed in crop fields close to wild vegetation edges than in fields isolated from such habitats (Altieri, 1994).

* In many cases, weeds and other natural vegetation around crop fields harbor alternate hosts/prey for natural enemies, thus providing seasonal resources to bridge gaps in the life cycles of entomophagous insects and crop pests (Altieri and Letourneau, 1984). A classic case is that of the egg parasitoid wasp Anagrus epos whose effectiveness in regulating the grape leafhopper Erythroneura elegantula was increased greatly in vineyards near areas invaded by wild blackberry (Rubus sp.). This plant supports an alternative leafhopper (Dikrella cruentata) which breeds in its leaves in winter. Recent studies show that French prune orchards adjacent to vineyards provide overwintering refuges for Anagrus and early benefits of parasitism are promoted in vineyards with prune trees plants upwind from the vineyard.

* At the crop field level, most experiments that have mixed other plant species with the primary host of a specialized herbivore show that in comparison with diversified cropping systems, monocultures have greater population densities of specialist herbivores (Andow, 1991). In these monoculture systems, herbivores exhibit greater colonization rates, greater reproduction, higher tenure time, less disruption of host finding and lower mortality by natural enemies (see Table 5 for examples in Latin America). * * TABLE: 5 Selected examples of multiple cropping systems that effectively prevent insect-pest outbreaks in Latin America

* Multiple cropping * Systems Pests (regulated) Factor(s) involved Country Cassava intercropped with cowpeas Whiteflies Aleurotrachelus socialis and Trialeurodes variabilis Changes in plant vigour and increased abundance of natural enemies Colombia Corn intercropped with beans Leafhoppers (Empoasca kraemeri), leaf beetle (Diabrotica balteata) and fall armyworm (Spodoptera frugipedra) Increase in beneficial insects and interference with colonization Colombia Corn intercropped with beans Corn leafhopper (Dalbulus maidis) Interference with leafhopper movement Nicaragua Cucumbers intercropped with maize and brocolli Flea beetles (Acalymma vitata) Lower crop apparency Costa Rica Corn-bean-squash Caterpillar (Diaphania hyalinata) Enhanced parasitization Mexico Corn-beans Stalk borer (Diatraea lineolata) Enhanced predation Nicaragua Source: Altieri 1994.

There are various factors in crop mixtures that help constrain pest attack. A host plant may be protected from insect pests by the physical presence of other plants that may provide a camouflage or a physical barrier. Mixtures of cabbage and tomato reduce colonization by the diamond-back moth, while mixtures of maize, beans, and squash have the same effect on chrysomelid beetles. The odors of some plants can also disrupt the searching behavior of pests. Grass borders repel leafhoppers from beans and the chemical stimuli from onions prevent carrot fly from finding carrots (Altieri, 1994).

Alternatively, one crop in the mixture may act as a trap or decoy- the 'fly-paper-effect'. Strips of alfalfa interspersed in cotton fields in California attract and trap Lygus bugs. There is a loss of alfalfa yield, but this represents less than the cost of alternative control methods for the cotton. Similarly, crucifers interplanted with beans, grass, clover, or spinach are damaged less by cabbage maggot and cabbage aphid. Another factor as predicted by the natural enemies hypothesis is that reduced insect pest incidence in polycultures may be the result of increased predator and parasitoid abundance and efficiency (Altieri, 1994).

Enhancing the productivity and multifunctionality of traditional farming: NGO-led agroecological initiatives In Latin America, economic change, fueled by capital and market penetration, is leading to an ecological breakdown that is starting to destroy the sustainability of traditional agriculture. After creating resource-conserving systems for centuries, traditional cultures in areas such as Mesoamerica, the Amazon, and the Andes are now being undermined by external political and economic forces. Biodivesity is decreasing on farms, soil degradation is accelerating, community and social organizations are breaking down, genetic resources are being eroded and traditions lost. Under this scenario, and given commercial pressures and urban demands, many developers argue that the performance of subsistence agriculture is unsatisfactory, and that intensification of production is essential for the transition from subsistence to commercial production (Blauert and Zadek, 1998). In reality, the challenge is to guide such transition in a way that it yields and income are increased without threatening food security, raising the debt of peasants, and further exacerbating environmental degradation. Many agroecologists contend that this can be done by generating and promoting resource conserving technologies, a source of which are the very traditional systems that modernity is destroying (Altieri, 1991).

Taking traditional farming knowledge as a strategy point, a quest has begun in the developing world for affordable, productive, and ecologically sound small scale agricultural alternatives. In many ways, the emergence of agroecology stimulated a number of non-governmental organizations (NGOs) and other institutions to actively search for new kinds of agricultural development and resource management strategies that, based on local participation, skills and resources, have enhanced small farm productivity while conserving resources (Thrupp, 1996). Today there are hundreds of examples where rural producers in partnership with NGOs and other organizations, have promoted and implemented alternative, agroecological development projects which incorporate elements of both traditional knowledge and modern agricultural science, featuring resource-conserving yet highly productive systems, such as polycultures, agroforestry, and the integration of crops and livestock etc.

Stabilizing the hillside of Central America

Perhaps the major agricultural challenge in Latin America is to design cropping systems for hillside areas, that are both productive and reduce erosion. Several organizations have taken on this challenge with initiatives that emphasize the stewardship of soil resources, utilization of local resources, and inputs produced on farm.

Since the mid 1980s, the private voluntary organization World Neighbors has sponsored an agricultural development and training program in Honduras to control erosion and restore the fertility of degraded soils. Soil conservation practices were introduced-such as drainage and contour ditches, grass barriers, and rock walls-and organic fertilization methods were emphasized, such as chicken manure and intercropping with legumes. Program yields tripled or quadrupled from 400 kilograms per hectare to 1,200-1,600 kilograms, depending on the farmer. This tripling in per-hectare grain production has ensured that the 1,200 families participating in the program have ample grain supplies for the ensuing year. Subsequently, COSECHA, a local NGO promoting farmer-to-farmer methodologies on soil conservation and agroecology, helped some 300 farmers experiment with terracing, cover crops, and other new techniques. Half of these farmers have already tripled their corn and bean yields; 35 have gone beyond staple production and are growing carrots, lettuce, and other vegetables to sell in the local markets. Sixty local villagers are now agricultural extensionists and 50 villages have requested training as a result of hearing of these impacts. The landless and near-landless have benefited with the increase in labor wages from US $2 to $3 per day in the project area. Outmigration has been replaced by inmigration, with many people moving back from the urban slums of Tegucigalpa to occupy farms and houses they had previously abandoned, so increasing the population of Guinope. The main difficulties have been in marketing of new cash crops, as structures do not exist for vegetable storage and transportation to urban areas (Bunch, 1987).

In Cantarranas, the adoption of velvetbean (Mucuna pruriens), which can fix up to 150 kg N/ha as well as produce 35 tones of organic matter per year, has tripled maize yields to 2500 kg/ha. Labor requirements for weeding have been cut by 75 percent and, herbicides eliminated entirely. The focus on village extensionists was not only more efficient and less costly than using professional extensionists, it also helped to build local capacity and provide crucial leadership experience (Bunch, 1990).

Throughout Central America, CIDDICO and other NGOs have promoted the use of grain legumes to be used as green manure, an inexpensive source of organic fertilizer to build up organic matter. Hundreds of farmers in the northern coast of Honduras are using velvet bean (Mucuna pruriens) with excellent results, including corn yields of about 3,000 kg/ha, more than double than national average, erosion control, weed suppression and reduced land preparation costs. The velvet beans produce nearly 30 t/ha of biomass per year, or about 90-100 kg of N/ha per year (Flores, 1989). Taking advantage of well established farmer to farmer networks such as the campesino a campesino movement in Nicaragua and elsewhere, the spread of this simple technology has occurred rapidly. In just one year, more than 1000 peasants recovered degraded land in the Nicaraguan San Juan watershed (Holtz-Gimenez, 1996). Economic analyses of these projects indicate that farmers adopting cover cropping have lowered their utilization of chemical fertilizers (from 1.900 kg/ha to 400 kg/ha) while increasing yields from 700kg to 2,000kg/ha, with production costs about 22 percent lower than farmers using chemical fertilizers and monocultures (Buckles et. al., 1998).

Scientists and NGOs promoting slash/mulch systems based on the traditional "tapado" system, used on the Central American hillsides, have also reported increased bean and maize yields (about 3,000kg/ha) and considerable reduction in labor inputs as cover crops smother aggressive weeds, thus minimizing the need for weeding. Another advantage is that the use of drought resistant mulch legumes such as Dolichos lablab provide good forage for livestock (Thurston et. al., 1994). These kinds of agroecological approaches are currently being used on a relatively small percentage of land, but as their benefits are being recognized by farmers, they are spreading quickly. Such methods have strong potential and offer important advantages for other areas of Central America and beyond.

Soil conservation in the Dominican Republic

Several years ago, Plan Sierra, an ecodevelopment project took on the challenge of breaking the link between rural poverty and environmental degradation. In the central cordillera of the Dominican Republic. The strategy consisted in developing alternative production systems for the highly erosive conucos used by local farmers. Controlling erosion in the Sierra is not only important for the betterment of the life of these farmers but also represents hydroelectric potential as well as an additional 50,000 hectares of irrigated land in the downstream Cibao valley (Altieri, 1990).

The main goal of Plan Sierra is agroecological strategy was the development and diffusion of production systems that provided sustainable yields without degrading the soil thus ensuring the farmers' productivity and food self-sufficiency. More specifically, the objectives were to allow farmers to more efficiently use local resources such as soil moisture and nutrients, crop and animal residue, natural vegetation, genetic diversity, and family labor. In this way it would be possible to satisfy basic family needs for food, firewood, construction materials, medicinals, income, and so on.

From a management point of view the strategy consisted of a series of farming methods integrated in several ways: 1. Soil conservation practices such as terracing, minimum tillage, alley cropping, living barriers, and mulching; 2. Use of leguminous trees and shrubs such as Gliricidia, Calliandra, Canavalia, Cajanus, and Acacia planted in alleys, for nitrogen fixation, biomass production, green manure, forage production, and sediment capture; 3. Use of organic fertilizers based on the optimal use of plant and animal residues; 4. Adequate combination and management of polycultures and/or rotations planted in contour and optimal crop densities and planting dates; 5. Conservation and storage of water through mulching and water harvesting techniques.

In various farms animals, crops, trees, and/or shrubs, are all integrated to result in multiple benefits such as soil protection, diversified food production, firewood, improved soil fertility, and so on. Since more than 2000 farmers have adopted some of the improved practices an important task of Plan Sierra was to determine the erosion reduction potential of the proposed systems. This proved difficult because most of the available methods to estimate erosion are not applicable for measuring soil loss in farming systems managed by resource-poor farmers under marginal conditions. Given the lack of financial resources and research infrastructure at Plan Sierra it was necessary to develop a simple method using measuring sticks to estimate soil loss in a range of concuos including those traditionally managed by farmers and the "improved ones" developed and promoted by Plan Sierra.

Based on field data collected in 1988-1989 on the accumulated erosion rates of three traditional and one improved farming system, the alternative systems recommended by Plan Sierra exhibited substantially less soil loss than the traditional shifting cultivation, cassava and guandul monocultures. The positive performance of the agroecologically improved conuco seemed related to the continuous soil cover provision through intercropping, mulching, and rotations, as well as the shortening of the slope and sediment capture provided by alley cropping and living barriers (Altieri, l985).

Recreating Incan agriculture

Researchers have uncovered remnants of more than 170,000 hectares of "ridged-fields" in Surinam, Venezuela, Colombia, Ecuador, Peru, and Bolivia (Denevan,1995). Many of these systems apparently consisted of raised fields on seasonally-flooded lands in savannas and in highland basins. In Peru, NGO's have studied such pre-Columbian technologies in search of solutions to contemporary problems of high altitude farming. A fascinating example is the revival of an ingenious system of raised fields that evolved on the high plans of the Peruvian Andes about 3000yrs ago. According to archeological evidence these Waru-Warus platforms of soil surrounded by ditches filled with water, were able to produce bumper crops despite floods, droughts, and the killing frost common at altitudes of nearly 4000 meters (Erickson and Chandler,1989).

In 1984 several NGO's and state agencies created the Proyecto Interinstitucional de Rehabilitacion de Waru-Warus (PIWA) to assist local farmers in reconstructing ancient systems. The combination of raised beds and canals has proven to have important temperature moderation effects extending the growing season and leading to higher productivity on the Waru-Warus compared to chemically fertilized normal pampa soils. In the Huatta district, reconstructed raised fields produced impressive harvest, exhibiting a sustained potato yield of 8-14 t/ha/yr. These figures contrast favorably with the average Puno potato yields of 1-4 t/ha/yr. In Camjata the potato fields reached 13 t/ha/yr. and quinoa yields reached 2 t/ha/yr. in Waru-Warus. It is estimated that the initial construction, rebuilding every ten years, and annual planting, weeding, harvest and maintenance of raised fields planted in potatoes requires 270 person-days/ha/yr. Clearly, raised beds require strong social cohesion for the cooperative work needed on beds and canals. For the construction of the fields, NGOs organized labor at the individual, family, multi-family, and communal levels.

Elsewhere in Peru, several NGOs in partnership with local government agencies have engaged in programs to restore abandoned ancient terraces. For example, in Cajamarca, in 1983, EDAC-CIED together with peasant communities initiated an all-encompassing soil conservation project. Over ten years they planted more than 550,000 trees and reconstructed about 850 hectares of terraces and 173 hectares of drainage and infiltration canals. The end result is about 1,124 hectares of land under construction measures (roughly 32% of the total arable land), benefiting 1,247 families (about 52% of the total in the area). Crop yields have improved significantly. For example, potato yields went from 5t/ha to 8t/ha and oca yields jumped from 3 to 8t/ha. Enhanced crop production, fattening of cattle and raising of alpaca for wool, have increased the income of families from an average of $108 per year in 1983 to more than $500 today (Sanchez, 1994).

In the Colca valley of southern Peru, PRAVTIR (Programa de Acondicionamiento Territorial y Vivienda Rural) sponsors terrace reconstruction by offering peasant communities low-interest loans and seeds or other inputs to restore large areas (up to 30 hectares) of abandoned terraces. The advantages of the terraces is that they minimize risks in terms of frost and/or drought, reducing soil loss, broadening cropping options because of the microclimatic and hydraulic advantages of terraces, thus improving productivity. First year yields from new bench terraces showed a 43-65% increase of potatoes, maize, and barley, compared to the crops grown on sloping fields (Table 6). The native legume Lupinus mutabilis is used as a rotational or associated crop on the terraces; it fixes nitrogen, which is available to companion crops, minimizing fertilizer needs and increasing production. One of the main constraints of this technology is that it is highly labor intensive. It is estimated that it would require 2,000 worker-days to complete the reconstruction of 1 hectare, although in other areas reconstruction has proven less labor intensive, requiring only 300-500 worker/day/ha (Treacey, 1989).

TABLE 6: First year per hectare yields of crops on new bench terraces, compared to yields on sloping fields (kg/ha) Cropa Terracedb Non-Terracedc Percent Increase Nd Potatoes 17,206 12,206 43 71 Maize 2,982 1,807 65 18 Barley 1,910 1,333 43 56 Barley (forage) 23,000 25,865 45 159 a All crops treated with chemical fertilizers. b Water absorption terraces with earthen walls and inward platform slope. c Fields sloping between 20 and 50 percent located next to terraced fields for control. d N= number of terrace/field sites. Source: Treacey 1984

 NGOs have also evaluated traditional farming systems above 4000 msnm, where maca (Lepidium meyenii) is the only crop capable of offering farmers secure yields. Research shows that maca grown in virgin soils or fallowed between 5-8 years, exhibited significantly higher yields (11.8 and 14.6 t/ha respectively) than maca grown after bitter potatoes (11.3 t/ha). NGOs now are advising farmers to grow maca in virgin or fallow soils in a rotative pattern, to use areas not suitable for other crops and taking advantage of the local labor and low costs of the maca-based system (UNDP, 1995, Altieri l996).

Organic farming in the Andes

In the Bolivian highlands, average potato production is falling despite a 15 percent annual increase in the use of chemical fertilizers. Due to increases in the cost of fertilizer, potato farmers must produce more than double the amount of potatoes compared with previous years to buy the same quality of imported fertilizer (Augstburger, 1983). Members of the former Proyecto de Agrobiologia de Cochabamba, now called AGRUCO, are attempting to reverse this trend by helping peasants recover their production autonomy. In experiments conducted in neutral soils, higher yields were obtained with manure than with chemical fertilizers. In Bolivia, organic manures are deficient in phosphorous. Therefore, AGRUCO recommends phosphate rock and bone meal, both of which can be obtained locally and inexpensively, to increase the phosphorous content of organic manures. To further replace the use of fertilizers and meet the nitrogen requirements of potatoes and cereals, intercropping and rotational systems have been designed that use the native species Lupinus mutabilis. Experiments have revealed that L. mutabilis can fix 200 kg of nitrogen per hectare per year, which becomes partly available to the associated or subsequent potato crop, thus significantly minimizing the need for fertilizers (Augstburger, 1983). Intercropped potato/lupine overyielded corresponding potato monocultures, and also substantially reduced the incidence of virus diseases.

Other studies in Bolivia, where Lupine has been used as a rotational crop, show that, although yields are greater in chemically fertilized and machinery-prepared potato fields, energy costs are higher and net economic benefits lower than with the agroecological system (Table 7). Surveys indicate that farmers prefer this alternative system because it optimizes the use of scarce resources, labor and available capital, and is available to even poor producers.

TABLE 7: Performance of traditional, modern, and agroecological potato-based production systems in Bolivia Traditional low-input Modern high-input Agroecological system Potato yields (metric tons/ha) 9.2 17.6 11.4 Chemical fertilizer (N + P2O5, kg/ha) 0.0 80 + 120 0.0 Lupine biomass (metric tons/ ha) 0.0 0.0 1.5 Energy efficiency (output/input) 15.7 4.8 30.5 Net income per invested Boliviano 6.2 9.4 9.9 Source Rist 1992

In the Interandean valleys of Cajamarca, near San Marcos traditional farming systems have been drastically modified through elements of conventional farming and urban influences, creating a market-oriented monoculture agriculture which favors cash crops rather than Andean crops. Centro IDEAS, an agricultural NGO, has implemented an organic agriculture proposal in order to revert the above process, supporting a more appropriate rural development strategy that rescues elements of the local traditional agriculture and ensuring food self-sufficiency as well as the preservation of natural resources (Chavez 1989).

The basic aspects of the proposal are:

* Rational use of local resources, conservation of natural resources, and intensive use of human and animal labor. * High diversity of native (Andean) and exotic crops, herbs, shrubs, trees, and animals grown in polycultural and rotational patterns. * Creation of favorable microclimates through the use of shelterbelts, and living fences and reforestation with native and exotic fruit and trees. * Recycling of organic residues and optimal management of small animals. * This proposal was implemented in a 1.9 ha model farm inserted in an area with similar conditions facing the average campesino of the region. The farm was divided into 9 plots, each following a particular rotational design (Table 8). After 3 years of operation, field results showed the following trends: * Organic matter content increased from low to medium and high levels, and N levels increased slightly. Addition of natural fertilizers were necessary to maintain optimum levels of organic matter and nitrogen. * Phosphorous and potassium increased in all plots. * Crop yields varied among plots, however in plots with good soils, (plot 1) high yields of corn and wheat were obtained. * Polycultures overyielded monocultures in all instances * To farm 1 ha of the model farm it was necessary to use 100 man-hours, 15 oxen-hours, and about 100 kgs of seeds.

These preliminary results indicate that the proposed farm design enhances the diversity of food crops available to the family, increases income through higher productivity, and maintains the ecological integrity of the natural resource base.

TABLE 8: Model farm rotational design. Plot Year 1 Year 2 Year 3 1 Maize, beans, quinoa, kiwicha, squash, and chiclayo Wheat Barley 2 Barley Lupinus and lentils Linaza 3 Wheat Favas and oats Maize, beans, quinoa, kiwicha 4 Rye Wheat Lentils 5 Lupinus Maize, beans, quinoa, kiwicha, squash, and chiclayo Wheat 6 Fallow Linaza Barley and Lentils Source: Chavez 1989:

 Since then, this model experience extended to 12 farmers who have undergone conversion to agroecological management in the Peruvian Sierra and Coast. A recent evaluation of the experiences showed that after a 2-5 year conversion process, income increased progressively due to a 20 percent increase in productivity (Alvarado de la Fuente and Wiener Fresco, 1998). Of the thirty three different organic technologies offered by the IDEAS, the 12 case study farmers favored: organic fertilization (11 cases), intercropping (10 cases), animal integration (10 cases), and agroforestry systems (8 cases).

Agroecological approaches in Brazil

The state government extension and research service, EPAGRI (Empresa de Pesquisa Agropecuaria e Difusao de Technologia de Santa Catarina), works with farmers in the southern Brazilian state of Santa Catarina. The technological focus is on soil and water conservation at the micro-watershed level using contour grass barriers, contour ploughing and green manures. Some 60 cover crop species have been tested with farmers, including both leguminous plants such as velvetbean, jackbean, lablab, cowpeas, many vetches and crotalarias, and non-legumes such as oats and turnips. For farmers these involved no cash costs, except for the purchase of seed. These are intercropped or planted during fallow periods, and are used in cropping systems with maize, onions, cassava, wheat, grapes, tomatoes, soybeans, tobacco, and orchards (Pretty, 1995).

The major on-farm impacts of the project have been on crop yields, soil quality and moisture retention, and labor demand. Maize yields have risen since 1987 from 3 to 5t/ha and soybeans from 2.8 to 4.7t/ha. Soils are darker in color, moist and biologically active. The reduced need for most weeding and ploughing has meant significant labor savings for small farmers. From this work, it has become clear that maintaining soil cover is more important in preventing erosion than terraces or conservation barriers. It is also considerably cheaper for farmers to sustain. EPAGRI has reached some 38,000 farmers in 60 micro-watersheds since 1991 (Guijt, 1998). They have helped more than 11,000 farmers develop farm plans and supplied 4300t of green manure seed.

In the hot and dry climate of Ceara, farmers combine production of sheep, goats, maize and beans, but productivity is low and environmental degradation is increasing. In the period between 1986 and 1991, ESPLAR, a local NGO engaged in a broad development program, involving the whole state of Ceara, through a massive training program in agroecology for village leaders. The training spearheaded a series of village-level activities reaching about 600 farmers which resulted in (VonderWeid, 1994):

1. The come back of arboreal cotton cultivation in mixed cropping with leucaena, algarrobo (Prosopis juliflora) and sabia (Mimosa caesalpiniaefdia). A shorter cycle variety was introduced, which together with integrated control of the boll weevil, made it possible to restore cotton fields. 2. The use of small dams for irrigated vegetable production. 3. Enriching the capoeiras (areas with secondary vegetation regrowth) with selected plant species made it possible to support 50 percent more goats per land unit. 4. Introduction of herbaceous legumes for fodder (especially cunha [Bradburya sagittata] ), in crop mixtures or rotated with maize and beans. 5. Planting along contour lines to reduce runoff.

In a similar semi-arid environment, as part of its research for alternatives to slash and burn, the Center for Alternative Technologies of Ouricouri developed a three year experiment to demonstrate the viability of land clearing without burning. The strategy had four components: the rationalized use of labor; the use of crops that compete with natural vegetation regrowth; efficient soil protection; and the harvesting and retention of rainwater. The work reaches at least 500 farmers in 30 communities (Guijt, 1998). The no-burning alternative involved cutting and clearing bush and tree vegetation, sowing crops more densely, and using cattle and horse manure. The first-year results indicated that reasonable production was possible and that tree and bush regrowth can be controlled. One negative aspect, however was the need to use over one-sixth of the available area for the storage of trunks and branches. In the second year bean output increased by over 100 percent relative to the historical average, though the low productivity of maize raised doubts as to its suitability under semi-arid agroecological conditions. Sorghum exhibited a better performance. The accumulation of plant material by the third year was enough to use as mulch. Unfortunately, the initial rains were followed by prolonged drought, and bean output fell sharply because of fungal disease. Nevertheless, the maize yield (552 kg/ha) was above the regional average of 500 kg/ha. (VonderWeid, 1994)

Integrated Production Systems

A number of NGOs promote the integrated use of a variety of management technologies and practices. The emphasis is on diversified farms in which each component of the farming system biologically reinforces the other components; for instance, where wastes from one component become inputs to another. Since 1980, CET, a Chilean NGO has engaged in a rural development program aimed at helping peasants reach year-round food self sufficiency while rebuilding the productive capacity of their small land holdings (Altieri, 1995). The approach has been to set up several 0.5 ha model farms, which consist of a spatial and temporal rotational sequence of forage and row crops, vegetables, forest and fruit trees, and animals. Components are chosen according to crop or animal nutritional contributions to subsequent rotational steps, their adaptation to local agroclimatic conditions, local peasant consumption patterns and finally, market opportunities. Most vegetables are grown in heavily composted raised beds located in the garden section, each of which can yield up to 83 kg of fresh vegetables per month, a considerable improvement to the 20-30 kg produced in spontaneous gardens tended around households. The rest of the 200-square meter area surrounding the house is used as an orchard, and for animals, (cows, hens, rabbits, and langstroth behives).

Vegetables, cereals, legumes and forage plants are produced in a six year rotational system within a small area adjacent to the garden. Relatively constant production is achieved (about six tons per year of useful biomass from 13 different crop species) by dividing the land into as many small fields of fairly equal productive capacity as there are years in the rotation. The rotation is designed to produce the maximum variety of basic crops in six plots, taking advantage of the soil-restoring properties and biological control features of the rotation.

Over the years, soil fertility in the original demonstration farm has improved, and no serious pest or disease problems have appeared. Fruit trees in the orchard and fencerows, as well as forage crops are highly productive. Milk and egg production far exceeds that on conventional farms. A nutritional analysis of the system based on its key components shows that for a typical family it produces a 250 % surplus of protein, 80 and 550% surplus of vitamin A and C, respectively, and a 330% surplus of calcium. A household economic analysis indicates that, the balance between selling surpluses and buying preferred items provides a net income beyond consumption of US $ 790. If all of the farm output were sold at whole sale prices, the family could generate a monthly net income 1.5 times greater than the monthly legal minimum wage in Chile, while dedicating only a relatively few hours per week to the farm. The time freed up is used by farmers for other on-farm or off-farm income generating activities.

In Cuba, the Asociacion Cubana de Agricultura Organica (ACAO), a non-governmental organization formed by scientists, farmers, and extension personnel, has played a pioneering role in promoting alternative production modules (Rosset, 1997). In 1995, ACAO helped establish three integrated farming systems called "agroecological light houses" in cooperatives (CPAs) in the province of Havana. After the first six months, all three CPAs had incorporated agroecological innovations (i.e. tree integration, planned crop rotation, polycultures, green manures, etc.) to varying degrees, which, with time, have led to enhancement of production and biodiversity, and improvement in soil quality, especially organic matter content. Several polycultures such as cassava-beans-maize, cassava-tomato-maize, and sweet potato-maize were tested in the CPAs. Productivity evaluation of these polycultures indicates 2.82, 2.17 and 1.45 times greater productivity than monocultures, respectively (Table 9).
 
 

TABLE 9: Performance of designed polycultures in two Cuban cooperatives Yield (t/ha) Polyculture 1 2 3 LER Lighthouse Cassava-beans-maize 15.6 1.34 2.5 2.82 "28 de Septiembre" Cassava-tomato-maize 11.9 21.2 3.7 2.17 "Gilberto Leon" Cassava-maize 13.3 3.39 -- 1.79 "Gilberto Leon" Beans-maize-cabbage 0.77 3.6 2.0 1.77 "28 de Septiembre" Sweet potato-maize 12.6 2.0 -- 1.45 "Gilberto Leon" Sorghum-squash 0.7 5.3 -- 1.01 "28 de Spetiembre" Source: SANE 1998

The use of Crotalaria juncea and Vigna unguiculata as green manure have ensured a production of squash equivalent to that obtainable applying 175 kg/ha of urea. In addition, such legumes improved the physical and chemical characteristics of the soil and effectively broke the life cycles of insect pests such as the sweet potato weevil (SANE, 1998).

At the Cuban Instituto de Investigacion de Pastos, several agroecological modules with various proportions of the farm area devoted to agriculture and animal production were established. Monitoring of production and efficiencies of a 75% pasture /25% crop module, reveals that total production increases over time, and that energy and labor inputs decrease as the biological structuring of the system begins to sponsor the productivity of the agroecosystem. Total biomass production increased from 4.4 to 5.1 t/ha after 3 years of integrated management. Energy inputs decreased, which resulted in enhanced energy efficiency from (4.4 to 9.5) (Table 10). Human labor demands for management also decreased over time from 13 hours of human labor/day to 4-5 hours. Such models have been promoted, extensively in other areas through field days and farmers cross visits (SANE, 1998).
 
 

TABLE 10: Productive and efficiency performance of the 75% animal / 25% crop integrated module in Cuba Productive Parameters 1st Year 3rd Year Area (ha) 1 1 Total Production (t/ha) 4.4 5.1 Energy Produced (Mcal/ha) 3797 4885 Protein Produced (Kg/ha) 168 171 Number of people fed by one ha. 4 4.8 Inputs (energy expenditures, Mcal) - Human Labor 569 359 - Animal Work 16.8 18.8 - Tractor Energy 277.3 138.6 Source: SANE 1998

Conclusions

Most research conducted on traditional and peasant agriculture in Latin America suggests that small holder systems are sustainably productive, biologically regenerative, and energy-efficient, and also tend to be equity enhancing, participatory, and socially just. Besides crop diversity, peasant farmers use a set of practices that cause minimal land degradation. These include the use of terraces and hedgerows in sloping areas, minimal tillage, small field sizes, and long fallow cycles. By concentrating on short rotations and fewer varieties, agricultural modernization in the same areas has caused environmental perturbation and eroded genetic diversity.

By adopting a multiple use strategy, indigenous farmers manage a continuum of agricultural and natural systems, obtaining a variety of products as well as ecological services thus truly enacting a multifunctional agriculture. Diversified cropping systems, such as those used by peasants, based on intercropping and agroforestry have been the target of much research recently. This interest is largely based on the new emerging evidence that these systems are more sustainable and more resource- conserving (Vandermeer, 1995). These attributes are connected to the higher levels of functional biodiversity associated with complex farming systems. In fact, an increasing amount of data reported in the literature, documents the effects that plant biodiversity has on the stabilization of agroecosystem processes.

In a recently conducted, well replicated experiment, where species diversity was directly controlled in grassland systems, it was found that ecosystem productivity was increased and that soil nutrients were utilized more completely when there was a greater diversity of species, leading to lower leaching losses from the ecosystem (Tilman et al. 1996). In agroecosystems this same pattern applies to insects as herbivore regulation increases with increasing plant species richness. Evidence suggests that as plant diversity increases, pest damage seems to reach acceptable levels, thus resulting in more stable crop yields. Apparently, the more diverse the agroecosystem and the longer this diversity remains undisturbed, the more internal links develop to promote greater insect stability. One aspect that is clear is that species composition is more important than species number per se. The challenge is to identify the correct assemblages of species that will provide through their biological synergisms key ecological services such as nutrient cycling, biological pest control, and water and soil conservation.

While it may be argued that peasant agriculture generally lacks the potential of producing meaningful marketable surplus, it does ensure food security. Many scientists wrongly believe that traditional systems do not produce more because hand tools and draft animals put a ceiling on productivity. Productivity may be low but the causes appear to be more social, not technical. When the subsistence farmer succeeds in providing food, there is no pressure to innovate or to enhance yields. Nevertheless, NGO-led agroecological field projects show that traditional crop and animal combinations can often be adapted to increase productivity when the biological structuring of the farm is improved and labor and local resources are efficiently used (Table 10; Altieri, 1995). In fact, most agroecological technologies promoted by NGOs can improve traditional agricultural yields increasing output per area of marginal land, enhancing also the general agrobiodiversity and its associated positive effects on food security and environmental integrity. It is not a matter of romanticizing subsistence agriculture or considering development per se as detrimental. The idea is to stress the value of traditional agriculture in the preservation of native crop diversity and the adjacent vegetation communities as this mode of appropriation of nature enhances the multifunctionality of agriculture (Toledo, 1995). Basing a rural development strategy on traditional farming and ethnobotanical knowledge, combined with elements of modern agroecology, not only assures continual use and maintenance of valuable agrobiodiversity, but also allows for the diversification of agricultural areas ensuring a variety of ecological services vital for food security, natural resource conservation, economic viability, climate amelioration, cultural preservation, and community empowerment. The challenge is now to promote the right policies and institutional partnerships that can scale-up ecologically based agriculture so that its multifunctional impacts are rapidly spread across the rural landscapes of Latin America.

  TABLE 10: Extent and impacts of agroecological technologies and practices implemented by NGOs in peasant farming systems throughout Latin America Country Organization Involved Agroecological Intervention

No. of Farmers or Farming Units Affected/ No. of Hectares Affected/ Dominant Crops Yield Increases (%)
BB EPAGRI AS-PTA Green Manures Cover Crops 38,000 Families 1,330,000 Maize Wheat 198 - 246 %
Guatemala Altertec and others Soil Conservation Green Manures Organic Farming 17,000 Units 17,000 Maize 250 %
Honduras CIDDICO COSECHA Soil Conservation Green Manures 27,000 Units 42,000 Maize 250 %
EL Salvador COAGRES Rotations Green Maures Compost Botanicals >200 Farmers Nd Cereals 40 - 60 %
Mexico Oaxacan Cooperatives Compost Terracing Contour Planting 3,000 Families 23,500 Coffee 140 %
Peru C Rehabilitation of Ancient Terraces  Raised Fields >1250 Families   nd > 1000   250 Andean Crops   Andean Crops 141 - 165 %
  333 %
CIED Watershed Agricultural Rehabilitation >100 Families N/A Andean Crops 30 - 50 %
IDEAS Intercropping Agroforestry Composting 12 Families 25 Several Crops 20 %
Dominican Republic Plan Sierra Swedforest-Fudeco Soil Conservation Dry Forest Mgmt. Silvopastoral Systems >2,500 Families >1,000 Several Crops 50 - 70 %
Chile CET Integrated Farms Organic Farming >1,000 Families >2,250 Several Crops >50 %
Cuba ACAO Integrated Farms 4 Cooperatives 250 Several Crops 50 - 70 %
Nd= no data Source: Browder 1989, Altieri 1995, Pretty 1997

References
Altieri, M.A. and L. C. Merrick. 1987. Peasant Agriculture and the conservation of crop and wild plant resources. J. Cons. Biol. 1: 49-58.
 Altieri, M.A. and L. C. Merrick. 1987. In situ conservation of crop genetic resources through maintenance of traditional farming systems.Economic Botany 4:86-96.
Altieri, M.A. 1991. Traditional Farming in Latin America. The Ecologist 21:93-96.
 Altieri, M.A. and S.B. Hecht 1991. Agroecology and small farm development.CRC Press, Boca Raton.
 Altieri, M.A. and D.L. Letourneau. 1982. Vegetation management and biological control in agroecosystems. Crop Protection 1: 405-430.
 Altieri, M.A. and D.K. Letourneau. 1984. Vegetation diversity and insect pest outbreaks. CRC Critical Reviews in Plant Sciences 2: 131-169.
 Altieri, M.A. and O. Masera 1993. Sustainable rural development in Latin America: building from the bottom up. Ecological Economics 7: 93-121.
 Altieri, M.A. 1994. Biodiversity and pest management in agroecosystems. Haworth Press, New York.
 Altieri, M.A. 1995. Agroecology: the science of sustainable agriculture. Westview Press, Boulder.
 Altieri, M.A. 1996. Enfoque agroecologico para el desarrollo de sistemas de produccion sostenibles en los Andes. Centro de Investigacion, Educaccion y Desarrollo, Lima.
 Altieri, M.A., P. Rosset and L.A. Thrupp 1998. The potential of agroecology to combat hunger in the developing world. 2020 Brief. IFPRI, Washington, DC.
 Alvarado de la Fuente, F. and H. Wiener 1998. Ofertas agroecologicas para pequenos agricultores: doce experiencias exitosas de agricultura ecologica. Centro IDEAS, Lima.
 Andow, D.A. 1991. Vegetational diversity and arthropod population response. Annual Review of Entomology 36: 561-586.
 Augustburger, F. 1983. Agronomic and economic potential of manure in Bolivian valleys and highlands. Agric. Ecosystem Environ. 10: 335-346.
 Beets, W.C. 1982. Multiple Cropping and Tropical Farming Systems. Westview Press, Boulder.
 Beets, W.C. 1990. Raising and sustaining productivity of smallholders farming systems in the Tropics. AgBe Publishing, Holland.
 Blauert, J. and S. Zadek 1998. Mediating sustainability. Kumarian Press, Connecticut.
 Brokenshaw, D.W., D.M. Warren and O. Werner. 1980. Indigenous Knowledge Systems and Development. University Press of America, Lanham.
 Brookfield, H. and C. Padoch 1994. Appreciating agrobiodiversity: a look at the dynamism and diversity of indigenous farming practices. Environment 36: 7-20.
 Browder, J.O. 1989. Fragile Conditions of Latin America: strategies for sustainable development. Westview Press, Boulder.
 Brush, S.B. 1982. The natural and human environment of the central Andes. Mountain Research and Development 2:14-38.
Buckles, D., B. Triophe and G. Sain 1998. Cover crops in hillside agriculture. IDRC-CIMMYJ, Mexico D.F.
Bunch, R. 1987. Case study of the Guinope Integrated Development Program. Proceedings IIED Conference on sustainable Development, London.
Bunch, R. 1990. Low-input soil restoration in Honduras: the Cantarranas farmer-to-farmer extension project. Sustainable Agriculture Gatekeeper Series SA23.IIED, London.
Caballero, J.N. and C. Mapes 1985. Gathering and subsistence patterns among the P'urepecha Indians of Mexico. J. Ethnobiol. 5: 31-47
Chambers, R. 1983. Rural Development: putting the last first. Longman Group Limited, Essex.
Chavez, J. et al. 1989. Propuesta de agricultura organica para la sierra.
IDEAS-CONYCET, Lima Chang, J.H. 1977. Tropical agriculture: crop diversity and crop yields. Econ. Geogr. 53:241-254.
Clarke, W.C. and R.R. Thaman 1993. Agroforestry in the Pacific: systems for sustainability. United Nations university Press, Tokyo.
Clawson, D.L. 1985. Harvest security and intraspecific diversity in traditional tropical agriculture. Econ. Bot. 39:56-67.
Denevan, W.M., J.M. Treace, J.B. Alcorn, C. Padoch, J. Denslow and S.T. Paitan. 1984. Indigenous agroforestry in the Peruvian Amazon: Bora Indian management of swidden Fallows. Interciencia 9:346-357.
Denevan, W.M. 1995. Prehistoric agricultural methods as models for sustainability. Advanced Plant Pathology 11:21-43.
Erickson, C.L. and K.L. Chandler 1989. Raised fields and sustainable agriculture in lake Titicaca Basin of Peru. In: J.O. Browder (ed), Fragile Lands of Latin America. 230-243 Westview Press, Boulder.
Francis, C.A. 1986. Multiple cropping systems. MacMillan, New York.
Fry, G. 1995. Landscape ecology of insect movement in arable ecosystems. IN: Ecology and integrated farming systems. D.M. Glen et al. (eds). John Wiley and Sons, Bristol, UK.
Gladwin, C. anf K. Truman 1989. Food and farm: current debates and policies. University Press of America, Lantham, MD.
Gliessman, S.R., E. Garcia and A. Amador. 1981. The ecological basis for the application of traditional agricultural technology in the management of tropical agro-ecosystems. Agro-Ecosystems 7:173-185.
Gliessman, S.R. 1998. Agroecology: ecological process in sustainable agriculture. Ann Arbor Press, Michigan.
Grigg, D.B. 1974. The Agricultural Systems of the World: an evolutionary approach. Cambridge University Press, Cambridge.
Guijt, I. 1998. Assessing the merits of participatory development of sustainable agriculture: experiences from Brazil and Central America. In J. Blauert and S. Zadek (eds) Mediating Sustainability pp. 100-128. Kumarian Press, Conn.
Harwood, R.R. 1979. Small Farm Development -- Understanding and Improving Farming Systems in the Humid Tropics. Westview Press, Boulder.
Hecht, S.B. 1984. Indigenous soil management in the Amazon basin: some implications for development In J.O. Browder (ed) Fragile Lands of Latin America pp. 166-181. Westview Press, boulder.
Holt-Gimenez, E. 1996. The campesino a campesino movement: farmer-led, sustainable agriculture in Central America and Mexico. Food First Development Report No. 10. Institute of food and Development Policy, Oakland.
Jimenez-Osormio, J. and S. del Amo 1986. An intensive Mexican traditional agroecosystem: the chiampa. Proc. 6th International Scientific Conference IFOAM, Santa Cruz, California.
Lampkin, N. 1992. Organic Farming. Farming Press, Ipswhich, England.
Lentz, D.I. 1986. Ethnobotany of the Jicaque of Honduras. Econ. Bot. 40: 210-219 Marten, G.C. 1986. Traditional agriculture in South-east Asia: a human ecology perspective. Westview Press, Boulder.
Nair, P.K.R. 1984. Soil productivity aspects of agroforestry. ACRAF, Nairobi.
Norman, M.J. T. 1979. Annual Cropping Systems in the Tropics. University Presses of Florida, Gainesville.
Pimentel, D. and M. Pimentel 1979. Food, Energy and Society. Edward Arnold, London.
Pretty, J. 1995. Regenerating agriculture. World Resources Institute. Washington, DC.
Pretty, J. 1997. The sustainable intensification of agriculture. Natural Resources Forum 21: 247- 256.
Reinjtes, C., B. Haverkort and Ann Waters-Bayer 1992. Farming for the future. MacMillan, London.
Richards, P. 1985. Indigenous Agricultural Revolution. Westview Press, Boulder.
Rist, S. 1992. Ecologia, economia y technologias campesina. Ruralter 10: 205-227.
Rosset, P.M. 1997. Alternative agriculture and crisis in cuba. Technology and Society 16: 19-25.
Sanchez, J.B. 1994. A seed for rural development: the experience of EDAC-CIED in the Mashcon watershed of Peru. Journal of Learnings 1: 13-21.
Sanders, W.T. 1957. Tierra y agua: a study of the ecological factors in the development of Meso- American civilizations. PhD Dissertation, Harvard University.
SANE 1998. Farmers, NGOs and Lighthouses: learning from three years of training, networking and field activities. SANE-UNDP, Berkeley.
Swift, M.S., J. Vandermeer, P.S. Ramakrishnan, J.M. Anderson, C.K. Ong and B.A. Hawkins 1996. Biodiversity and agroecosystem function. H.A. Mooney et al. (eds.). Functional roles of biodiversity: a global perspective. J. Wiley and Sons, N.Y., pp.261-298.
Thrupp, L.A. 1996. New Partnerships for sustainable agriculture. World Resources Institute, Washington, DC.
Thrupp, L.A. 1998. Cultivating diversity: agrobiodiversity and food security. World Resources Institute, Washington, DC.
Thurston, H.D. 1991. Sustainable practices for plant disease management in traditional farming systems. Westview Press, Boulder.
Thurston H.D. et al. 1994. Slash/Mulch: How Farmers use It and What Researchers Know About It. CIIFAD-CATIE, Ithaca, NY.
Tilman, D, D. Wedin and J. Knops 1996. Productivity and sustainability influenced by biodiversity in grassland ecosystems. Nature 379: 718-720.
Toledo, V.M., J. Carabias, C. Mapes, and C. Toledo. 1985. Ecologia y Autosuficiencia Alimentaria. Siglo Vientiumo Editors, Mexico City.
Toledo, V.M. 1995. Peasantry, agroindustriality and sustainability. Interamerican Council for Sustainable Agriculture. Working Paper 3. Michoacan, Mexico.
Treacey, J.M. 1989. Agricultural terraces in Peru's Colca Valley: promises and problems of an ancient technology. In Fragile lands of Latin America.
J.O. Browder (ed). Westview Press, Boulder.UNDP 1995. Benefits of Diversity. UNDP, New York.
Vandermeer, J. and I. Perfecto 1995. Breakfast of biodiversity: the truth about rainforest destruction. Food First Books, Oakland.
Vandermeer, J. 1995. The ecological basis of alternative agriculture. Ann. Rev. Ecol. Syst. 26: 201-224. Wilken, G.C. 1987. Good Farmers: traditional agricultural resource management in Mexico and Guatemala. University of California Press, Berkeley.