Cover Image
close this book The Fragile Tropics of Latin America: Sustainable Management of Changing Environments (1995)
close this folder Part 3 : The Peruvian Amazon
close this folder Subsistence- and market-oriented agroforestry in the Peruvian Amazon
View the document (introductory text)
View the document 1 Introduction
View the document 2 Traditional agroforestry in north-eastern Peru
View the document 3 Swidden-fallow agroforestry among the Bora indians
View the document 4 Market-oriented agroforestry in Tamshiyacu
View the document 5 Conclusions and recommendations
View the document References

Subsistence- and market-oriented agroforestry in the Peruvian Amazon

(introductory text)

Subsistence- and marketoriented agroforestry in the Peruvian Amazon

1 Introduction
2 Traditional agroforestry in north-eastern Peru
3 Swidden-fallow agroforestry among the Bora indians
4 Market-oriented agroforestry in Tamshiyacu
5 Conclusions and recommendations


Christine Padoch and Wil de Jong

1 Introduction

1 Introduction

In recent years, researchers have demonstrated that indigenous Amazonians have a profound knowledge of agricultural soils, useful plants and animals, and productive agricultural and forestry techniques (Boom, 1987; Clay, 1988; Denevan and Padoch, 1988; Dufour, 1990; Posey, 1982, 1984, 1985; Prance, 1984; Vickers and Plowman, 1984; and others). Many of these researchers have suggested that indigenous knowledge and technologies can be incorporated into the planning of contemporary agricultural and forestry efforts in Amazonia. Data showing traditional practices to be economically attractive alternatives for recent colonists and other non-indigenous people who need and want market goods are, however, still very scarce. The paucity of demonstrated economic potential is among the most important factors that have recently prompted some experts to question the usefulness of indigenous patterns in future economic development (Redford, 1990).

Increased interest in studying the resource use patterns of indigenous or long-resident, but non-tribal communities of Amazonia, such as Peru's ribereños (de Jong, 1987; Hiraoka, 1985a, 1985b, 1986, 1989; Padoch and de Jong, 1987, 1989, 1990; Padoch et al.,1985) or the caboclos of Brazil (Anderson et al., 1986; Parker, 1985) may help fill this gap in the discussion of development alternatives for the region. These populations usually participate more actively in local and even export markets than do most tribal peoples, and have needs and expectations that more closely reflect national norms. ribereño and caboclo resource use practices show clearly that these frequently ignored people are inheritors as well as adapters of indigenous traditions. Examination of many traditional swiddenfallow agroforesty practices in the Department of Loreto, Peru, provides several examples of how indigenous patterns have been adapted to the market-oriented economy of ribereño villagers (Denevan and Padoch, 1988; Padoch and de Jong, 1987, 1989).

2 Traditional agroforestry in north-eastern Peru

2 Traditional agroforestry in north-eastern Peru

Amazonian agroforestry systems include a diversity of production patterns, ranging from the most subtle manipulation of largely natural forests to increase their economic yields, to the creation and maintenance of nearmonocultural commercial orchards (Hecht, 1982; Padoch and de Jong, 1989; Padoch et al., 1985). Many, although not all, of these systems are cyclic and are based on an alternation of intensively cultivated swiddens and less intensively managed fallows. Swidden-fallow use by tribal peoples has been studied in several parts of the Amazon Basin, including areas in Ecuador (Irvine, 1985) and Brazil (Dufour, 1990; Posey, 1982).

We conducted studies on agroforestry in the Department of Loreto, Peru, between 1981 and 1983 as part of a cooperative programme of research between the University of Wisconsin (USA) and the Universidad Nacional de la Amazonía Peruana, and in 1984 and 1985 as participants in a research programme of the Instituto Nacional de Investigacíon y Promoción Agrícola. In this paper, we report on two examples of swidden-fallow agroforestry systems: one typical of a community of tribal Bora Indians, the other a pattern found in the ribereño village of Tamshiyacu. These examples share some characteristics. Both begin as swidden fields, both involve quantitative and qualitative changes in management practices, and both exhibit major shifts over time in species composition and production. However, major differences between the two are also notable, particularly in the diversity, quantity, and destination of the species produced, and in the intensity and quality of their management.

The types that we present are only two of a large variety of swid den-fallow agroforestry patterns found in the region and should not be considered typical of the Department of Loreto. After presenting and comparing these examples of traditional resource use practices, we shall attempt to point out several aspects of these systems that may prove useful in planning for enhanced agricultural and forestry production in the humid tropics.

Figure 9.1 Area of study in the Department of Loreto, north-eastern Peru, showing locations of the Bora community of Brillo Nuevo and the ribereño village of Tamshlyacu. (Map drawn by Carol Gracie, adapted from Padoch and de Jong, 1987.)

3 Swidden-fallow agroforestry among the Bora indians

3 Swidden-fallow agroforestry among the Bora indians

The Bora community of Brillo Nuevo is located along the Yaguasyacu River, a tributary of the Ampiyacu River, which itself joins the Amazon around the town of Pebas, 120 km north-east of Iquitos (fig. 9.1). Much of the village territory of Brillo Nuevo is still covered by mature humid tropical forest, although considerable areas of secondary forest, swiddens, and small pastures are found near the settlement. Climatically the area is characterized, by consistently high temperatures (around 26°C throughout the year), high humidity, and rainfall averaging almost 3,000 mm per year. (See Denevan and Padoch, 1988, for a more detailed description of the area.) The residents of Brillo Nuevo were brought to the present site by a patron from neighbouring Colombia in 1934. While contemporary Bora use market products and have become assimilated to varying degrees into Peruvian society, their subsistence activities still largely reflect traditional practices. Bora resource use in their homeland closely resembles Brillo Nuevo patterns (Centlivres et al., 1975; Gasché, 1980). The considerable distance of the village from Iquitos, the major regional market, and the difficulty of using boats in the Yaguasyacu river during low water, makes it difficult for most Brillo Nuevo residents to participate actively in local markets.

Bora farmers make new agricultural fields in old forests as well as in secondary forest. Like most Amazonian terra firme farmers the Bora follow a cycle that begins with the cutting, drying, and burning of a forest plot between 0.5 and 1 ha in size. Any valuable species found in the original forest cover, including fruit trees, palms used for fibre, thatching, or construction, and important timber trees are usually spared.

Initial planting includes a variety of crops, but manioc (Manihot esculenta), the dietary staple, predominates in new swidden fields. The number of crop species planted varies from field to field. Some swiddens are planted with very few species, others with a large variety. Planted species usually include annuals and perennials. The latter often are species of fruit trees, but other woody species such as coca (Erythroxylum coca), or barbasco (Lonchocarpus spp.), which is used as a fish poison, are frequently also present.

Bora farmers distinguish between the agricultural phase, in which their fields yield manioc, and which lasts for about two years, and the following phase, in which fields are less actively managed. This second phase begins the transition from the swidden to the swidden-fallow stage. While this transition is occurring in one holding, the Bora cultivator is often already clearing other sites to plant new annual crops. Thus a farmer manages several plots at the same time: swidden fields producing annual crops, other fields just passing into the fallow stage, and still others in the mature forestfallow phase.

The assemblage of plant species present in a Bora swidden fallow is often very diverse and changes considerably during its life cycle. The pre-existing fruit and other economically important forest trees that were originally spared are often present later in the managed fallow field. Many plants that appear spontaneously from coppicing stumps, or that sprout from seeds, are of use to the owner and may also be left untouched or may even be actively protected. In fact, weeding is often directed toward aiding the growth of valuable "weed" species, as well as the protection of planted crops. Still other individual plants are the result of natural reseeding by planted species. Thus, deliberate planting is only one of many pathways by which a particular individual plant may find itself in a Bora swidden fallow at any particular stage.

The swidden-fallow fields that result from the interaction of these complex processes often include a large variety of useful plants. Denevan and Treacy (1988: 17-18), for instance, found a Bora field, five years after it was first planted, that featured a small, unproductive manioc patch. Six of the other twelve originally planted crops were, however, still yielding well and were frequently harvested. The outer edges of the field were not being actively weeded any more, and secondary vegetation had encroached considerably. This invading vegetation, however, also contained many useful species. Handicraft materials and medicinal species were abundant in this zone, while several trees useful for construction were nearing harvestable size.

Weeding and protection of plants in many Bora swidden fallows continue for at least twelve years after their first planting. One nineyear-old fallow sampled at Brillo Nuevo still included five cultivated fruit species as well as coca; all but the last of these were represented by only a few individuals. Together with these planted trees numerous useful - protected but not planted - species were found, including such valuable construction timbers as tropical cedar (Cedrela odorata).

Bora swidden-fallow management techniques result in fields with a very patchy pattern. While overall management of a plot may decrease with time, limited areas may continue to be actively cleared for many years. In older fields the cultivated area shrinks rather than being totally abandoned. Even when a forest canopy covers the entire plot, old fallow areas still yield fruits or other useful materials. During occasional visits farmers will often clear a few lianas or other plants which threaten to interfere with economically valuable species. Even after twenty years cultigens may still be harvested. One such field that was studied included three cultivated fruit species and an other 53 useful native forest species. These included species used as construction materials, medicines, and foods, and in basket-making, saltmaking, and extraction of pitch.

Swidden-fallow fields located at larger distances from the village attract a variety of forest animals that come to feed on the fruits of the planted and protected trees. In certain seasons animals appear in fallow fields in higher numbers than in the surrounding forest. The swidden-fallow fields then become important hunting sites, especially abundant in fruit-eating monkeys and large rodents.

Although the general pattern we described above is fairly common among Brillo Nuevo farmers, numerous geographical, cultural, ecological, personal, and purely chance factors give rise to variations in the composition of a swidden fallow. Among important variables that contribute to this diversity are: the age of the original vegetation as well as its previous use, the distance of the field from the village, other fields the owner farms, soil conditions, drainage, slope, and aspect of the site.

Although each Bora swidden fallow differs in exact species composition from every other, most older fields are rich in useful plants. In the fallows that were sampled at Brillo Nuevo, a total of 133 useful species were identified. The Bora swidden-fallow agroforestry pattern, characterized, by a high diversity of species, appears appropriate for a population that seeks to satisfy a wide variety of its daily needs through harvesting its own fields and forests. With little access to markets, the Bora cannot and do not specialize in the production of any particular plant species. Rather, they seek to produce a little of many different things. This pattern, which may have been typical of Amazonian communities in the past, is not a universal strategy among the rural folk of the Department of Loreto.

4 Market-oriented agroforestry in Tamshiyacu

4 Market-oriented agroforestry in Tamshiyacu

The town of Tamshiyacu, located 30 km upstream from Iquitos, is known as one of the most important native fruit-producing areas in the region. The two principal products marketed by Tamshiyaquinos, pineapples and umarí (Poraqueiba sericea), are cultivated within a well-designed, marketoriented agroforestry system. The two products are harvested at different phases of this system. Although Tamshiyacu farmers do grow products destined for household consumption, a market emphasis pervades virtually every phase of this particular agroforestry system. (For more detailed information on Tamshiyacu agriculture see de Jong, 1987; Hiraoka, 1986; Padoch et al., 1985.)

In contrast to the Bora, the people of Tamshiyacu are not officially recognized as Indians, that is, natives of the region, although few if any Tamshiyaquinos are recent arrivals. They belong to a large group of Amazonians known locally as ribereños (Padoch, 1988), who are largely offspring of detribalized natives and immigrants who arrived during or soon after the great Rubber Boom of the turn of the century. This largest group of rural people in the lowland Peruvian Amazon has gone surprisingly unnoticed by scientists.

Tamshiyacu is one of the larger towns in the area, and the majority of its approximately 2,000 inhabitants are farmers who frequently visit Iquitos and its markets. Transportation is provided by colectivos (river buses), several of which leave the village for the city early in the morning, or pass by from upstream and stop at the village dock. Many producers, however, choose not to make the trip themselves but sell their products to middlemen in the village itself.

In Tamshiyacu, which is both more populous and longer-settled than Brillo Nuevo, mature forest is found at a greater distance from the community centre. Tamshiyacu farmers, like the Bora, begin the swidden-fallow agroforestry cycle by clearing patches of forest, both primary and secondary. As in Brillo Nuevo, valuable forest species including palms and fruit trees may be spared. However, in contrast to the practice of Bora farmers and other tribal shifting cultivators, part of the cut and slashed vegetation is not burned in the field but is turned into charcoal for sale in Iquitos. Sites where charcoal is made are then often reserved for the planting of nutrient-demanding crops.

Most of the crop species planted in Brillo Nuevo can also be found in Tamshiyacu fields. In the town, however, the selection of cultigens, and especially the quantity planted of any species, is determined more by recent and expected urban market price trends than by household needs. The first crops to be harvested are manioc and plantains, which are also planted first. Both these crops, meant for market and household consumption, are commonly replanted after the first harvest. The planting of a field is, however, a complex process that continues for a over a year or more. After manioc and plantains, other annuals and several perennials are planted. The timing of many plantings depends on the availability of seeds or seedlings.

However, a field may already show a large diversity of crops at an early stage. After about two years, manioc and plantain production is phased out and pineapples start producing in quantity. At this point, the fields change primarily into cash crop plantations. Pineapple is harvested until the fourth year. By then, tree crops have come into prominence.

Apart from umarí (Poraqueiba sericea), trees commonly planted in agroforestry fields are fast-growing but short-lived species such as cashew (Anacardium accidentale), uvilla (Pourouma cecropiaefolia), and caimito (Pouteria caimito). The production of these fruits starts early and does not last very long; their productive life is from two to six years. Together with pineapple these fruits provide cash income during the transition stage from swidden to mature fruit orchard. Once production of these species has declined, umarí, a much-favoured fruit in the Iquitos market, becomes the pre-eminent product. An umarí stand normally yields for at least thirty years, although examples of sixty-year-old umarí orchards can be found. Yields of up to 80,000 fruits per hectare of umarí orchard are not unusual. A second important cash crop species in older orchards is Brazil nuts (Bertholletia excelsa). This long-lived tree crop has traditionally been planted in small quantities, although farmers now are increasingly including this crop in their fields.

Although total labour input declines when fields become older, Tamshiyacu farmers clean their field thoroughly as long as they are being harvested. In the swidden stage, while manioc is in production, weeding and harvesting are done frequently. When pineapple becomes the principal crop, fields continue to be weeded once or twice a year. However, as perennials take over, weeding changes in intensity. The more careful weeding in early swiddens gives way to a more rapid slash weeding in older fields. The shade provided by the fruit trees allows for more open understorey in older fields and a much faster weeding. Some cleaning, however, remains essential even in old fields, since umarí fruits must be gathered when they fall to the ground. Although several naturally occurring fruits or other useful species may survive the periodic weeding of an umarí orchard, Tamshiyacu agroforesters commonly protect or encourage far fewer forest species than do the Bora.

After thirty years or so, production in many umarí fruit orchards declines. The stand is then cut, with many of the umarí trees made into charcoal. Brazil nut trees, however, are not felled since they con tinue to fruit for a much longer time. After cutting, the umarí field may be fallowed for five to ten years, but it is not unusual to see farmers plant new swiddens immediately, and begin the cycle again.

Although we have presented a generalized picture of agroforestry in Tamshiyacu, individual differences among farmers, their fields, and the management techniques they employ are notable. Some farmers never use forest slash to make charcoal, some never convert annual crop fields to perennial-dominated orchards, some market most of their produce, others market little. Furthermore, it must be pointed out that many residents of Tamshiyacu engage in other economic pursuits. Some have diverse agricultural holdings in the Amazon floodplain (Hiraoka, 1986). Most farmers are also hunters, fishermen, occasional waged labourers, or even market vendors. The majority of Tamshiyacu farmers, however, obtain the bulk of their income from selling cultivated fruits from agroforesty fields (Padoch, 1988). While incomes vary enormously, Tamshiyacu producers as a group are exceptional in the level of incomes they report from sales of agroforestry products. Owners of large umarí orchards have incomes several times the Amazonian average. Yearly household incomes of approximately US$5,000 were enjoyed by several Tamshiyacu umarí producers, a level few other rural residents of the region ever reach.

5 Conclusions and recommendations

5 Conclusions and recommendations

We have presented descriptions of agroforestry practices found in two communities in the Peruvian Amazon. One of our sample villages is inhabited by members of a tribal group, the other is a settlement of non-tribal, but none the less largely indigenous Amazonians known as ribereños The Tamshiyacu and the Bora systems show some obvious similarities; ribereño practices should quite clearly be considered a transformation and adaptation of indigenous tribal patterns. However, the two systems described differ dramatically as well. In Tamshiyacu, production in older fields becomes highly specialized reflecting the relatively easy access the town's farmers have to the Iquitos market. Brillo Nuevo farmers, on the other hand, do not have the same market opportunities. Their system is much more generalized, diverse in species, and successful in satisfying the many needs of farmers in an area far from markets.

The differences in species composition and production between the two Amazonian settlements largely reflect differences in management. Tamshiyacu orchards are managed more intensively over a longer period, while Bora farmers allow, even encourage, more forest vegetation in their ageing fallows.

While the choice to specialize does help Tamshiyacu farmers frequently to earn high incomes, it also has drawbacks. In years of particularly high umarí production throughout the Iquitos region, a considerable part of the harvest cannot be sold and farmers are not adequately compensated for their labour investments. To date, no suitable process for preserving the bland, oily pulp of the umarí fruit has been developed and all fruit must be consumed fresh.

We have noted that the two descriptions given here are not an exhaustive summary of agroforestry types found in the lowland Peruvian Amazon. Our research indicates that agroforestry systems in the Peruvian Amazon are widespread and greatly varied (see Padoch and de Jong, 1987). There is no one Amazonian agroforestry system. Local patterns of swidden-fallow agroforestry are highly flexible, adaptable to various environmental and economic conditions. The examples that we have presented are also highly dynamic. Active experimentation occurs in both communities, as farmers attempt to accommodate their agroforestry practices to changing environmental and economic conditions.

In focusing much of our discussion on the market-oriented system of Tamshiyacu, we would particularly like to emphasize the need to study the traditional systems of non-tribal communities. These forgotten villages of the Peruvian Amazon (and of other Amazonian states) often present useful models to resource use planners. Yet it is the remote, tribal communities that have been given more attention. Most ribereños or mestizos living within a day's travel of Iquitos participate more actively in market trade than do natives on small isolated tributaries. Rather than using such tribal systems as models, and speculating on how these might serve the Amazonian who is heading toward increased market involvement, ribereño patterns may offer a more accessible model.

Our investigations in the Iquitos area have pointed out the great importance of market access and market demand in determining the configuration of a particular agroforestry system. The example taken from Tamshiyacu demonstrates conclusively that swiddeners can and do respond strongly to market opportunities. However, the market of Iquitos, like many tropical regional markets, is limited and often cannot consume even the yields that are produced. Research and extension in agroforestry systems must be accompanied by sound economic knowledge and advice. If the small farmers of Iquitos and similar regions elsewhere are to be encouraged to augment their production of fruits and other agroforestry products, they must also be presented with opportunities for processing and extraregional export of those products.



Acevedo-Rodriguez, P. 1990. "The occurrence of piscicides and stupefactants in the plant kingdom." In G.T. Prance and M.J. Balick (eds.), New directions in the study of plants and people (Advances in Economic Botany 8, New York Botanical Garden, New York), 1-9.

Alcorn, J.B. 1984. "Development policy, forests, and peasant farms." Economic Botany 38 (4): 387-409.

Anderson, A.B. 1988. "Use and management of native forests dominated by acai palm (Euterpe oleracea Mart.) in the Amazon estuary." In M.J. Balick (ed.), The palm - tree of life: Biology, utilization and conservation (Advances in Economic Botany 6, New York Botanical Garden, New York), 144-54.

. (ed.). 1990. Alternatives to deforestation: Steps toward sustainable use of the Amazon rainforest. Columbia University Press, New York.

Balée, W. 1985. "Ka'apor ritual hunting." Human Ecology 13 (4): 485-510.

Bergman, R.W. 1980. Amazon economics: The simplicity of Shipibo Indian wealtk. Dellplain Latin American Studies, No. 6. University Microfilms International.

Bodmer, R.E., T.G. Fang, and L.M. Ibañez. 1988. "Ungulate management and conservation in the Peruvian Amazon." Biological Conservation 45: 303-10.

Denevan, W.M. 1976. "The aboriginal population of Amazonia." In W.M. Denevan, (ed.), The native population of the Americas in 1492 (University of Wisconsin Press, Madison), 205-35.

. 1984. "Ecological heterogeneity and horizontal zonation of agriculture in the Amazon floodplain." In M. Schmink and C.H. Wood, (eds), Frontier expansion in Amazonia (University of Florida Press, Gainesville), 311-36.

Denevan, W.M., and C. Padoch. 1988. Swidden-fallow agroforestry in the Peruvian Amazon. Advances in Economic Botany 5, The New York Botanical Garden, New York.

Emmons, L.H. 1982. "Ecology of Proechimys in southeast Peru." Tropical Ecology 23: 28090.

. 1990. Neotropical forest mammals: A field guide. University of Chicago Press, Chicago.

Galvão, E. 1955. Santos e visagens. Companhia Editôra Nacional, São Paulo.

Goulding, M. 1981. Man and fisheries in an Amazonian frontier. W. Junk, The Hague.

Hecht, S.B., A. B. Anderson, and P. May. 1988. "The subsidy from nature: Shifting cultivation, successional palm forests, and rural development." Human Organization 47 (1): 25-35.

Hiraoka, M. 1985. "Floodplain farming in the Peruvian Amazon." Geographical Review of Japan B58: 1-21.

. 1986. "Zonation of mestizo farming systems in northeast Peru." National Geographical Research 2: 354-71.

. 1990. "The role of swidden fallow in the provision of game: Examples from the Peruvian Amazon." Paper presented at the Annual Meeting of the Association of American Geographers, Toronto, 18-22 April.

Irvine, D. 1987. "Resource management by the Runa indians of the Ecuadorian Amazon." Ph.D. dissertation, Department of Anthropology, Stanford University.

Junk, W.J. 1983. "As agues da região Amazônica." In E. Salati, H.O.R. Schubart, and A.E. Oliveira, (eds.), Amazonia, desenvolvimento, integração, e ecologíc (Editôra BrasilienseCNPq, São Paulo), 45-100.

Junk, W.J., and K. Furch. 1985. "The physical and chemical properties of Amazonian waters and their relationship with the biota." In G.T. Prance and T.E. Lovejoy (eds.), Key environments: Amazonia (Pergamon Press, New York), 3-17.

Linares, O. 1976. "Garden hunting in the American tropics." Human Ecology 4: 331-49.

Lowe-McConnell, R.H. 1977. Ecology of fishes in tropical waters. E. Arnold, London.

Nietschmann, B. 1973. Between land and water: The subsistence ecology of the Miskito Indians in Nicaragua. Seminar Press, New York.

Padoch, C., J. Chota Inuma, W. de Jong, and J. Unruh. 1985. "Amazon agroforestry: A market-oriented system in Peru." Agroforestry Systems 3: 47-58.

Padoch, C., and W. de Jong. 1990. "Santa Rosa: The impact of the minor forest products trade on an Amazonian place and population." In G.T. Prance and M. Balick (eds.), New directions in the study of plants and people (Advances in Economic Botany 8, New York Botanical Garden, New York), 151-58.

Parker, E.P. 1985. "The Amazon caboclo: An introduction and overview." In The Amazon caboclo: Historical and contemporary perspectives (Studies in Third World Societies 32. College of William and Mary, Dept. of Anthropology, Williamsburg), xvii-li.

Pierret, P.V., and M.J. Dourojeanni. 1966. "La caza y la alimentacion humane en las riberas del Rio Pachitea, Peru." Turrialba 16 (3): 271-7.

Posey, D.A. 1983. "Indigenous ecological knowledge and development of the Amazon." In E. Moran (ed.), The dilemma of Amazonian development, Westview Press, Boulder.

Posey, D.A., and W. Balée, (eds.). 1989. Resource management in Amazonia: Indigenous and folk strategies. Advances in Economic Botany 7. New York Botanical Garden, New York.

Redford, K.H., and J.G. Robinson. 1987. "The game of choice: Patterns of Indian and colonist hunting in the Neotropics." American Anthropologist 89 (3): 650-67.

Rodriguez, M.A., and W.M. Lewis. 1990. "Diversity and species composition of fish communities of Orinoco floodplain lakes." National Geographical Research 6 (3): 319-28.

Roosevelt, A.C. 1989. "Lost civilizations of the Lower Amazon." Natural History 66: 74-83.

. (1991). Mound builders of the Amazon: Geophysical archaeology on Marajó Island, Brazil. Academic Press, New York.

Smith, N.J.H. 1974. "Destructive exploitation of the South American turtle." Yearbook of the Association of Pacific Coast Geographers 36: 85-102.

. 1981. Man, fishes, and the Amazon. Columbia University Press, New York.

.1985. "The impact of cultural and ecological changes on Amazonian fisheries." Biological Conservation 32: 355-73.

Smythe, N. 1978. The natural history of the Central American agouti. Smithsonian Contributions to Zoology, no. 257. Smithsonian Institution Press, Washington, DC.

Villarejo, A. 1979. Asi es la selva, 3rd edn. CETA, Iquitos.