Universidade Federal do Acre

Floristics and Economic Botany of Acre, Brazil
Florística e Botânica Econômica do Acre, Brasil

Richard Wallace (Tropical Conservation and Development Program, 319 Grinter Hall, University of Florida, Gainesville, Florida 32611). The traditional marketing system in the three communities in the chico mendes extractive reserve: implications for the extraction and marketing of other non-timber forest products. Adv. Econ. Bot. x: xx-xx. 1999. Although there is a growing body of research on non-timber forest product marketing in the Amazon, we know little about how new extractive products might fit into the current market system for traditional extractive products, Hevea rubber and Brazil nuts. This paper describes and analyzes the current marketing system for these two products for three forest communities in the Chico Mendes Extractive Reserve. It then examines the implications of this system for market-oriented extraction and production of other non-timber forest products. I conclude that the incorporation of new market-oriented non-timber forest product activities into current market channels will face barriers: at least initially, new products will not provide important access to services and goods, credit and basic supplies, which play a critical role in the market system as they will not have the exchange value of traditional products. In addition, many families face transport difficulties, including long travel time to markets, limited access to transport and seasonal constraints due to seasonal rain patterns. Community families working together might best confront these problems. This will reduce the labor and transport costs time constraints of market-oriented activities of individual families.

Key words: extractive reserves, marketing, non-timber forest products

AA borracha não dá ajuda a gente, não@

ARubber does not help the people, no@

Rubber tapper in seringal Floresta, Xapuri

The low price of Hevea rubber (Hevea brasiliensis) has led many rubber tappers in the Chico Mendes Extractive Reserve in Southeast Acre to search for new and identify economic opportunities to maintain and increase family income. The need to diversify their income sources may be critical for survival in the forest over the long term (Anderson 1989).

Many families are seeking wage labor opportunities (Campbell 1996), when available, or are focusing on small farmer activities such as producing more subsistence crops, raising small farm animals, processing farinha, or manioc (Manihot esculenta) flour and agroforestry production. This is not surprising given that the limited extension assistance received by rubber tappers living in the reserve promotes these activities (personal observation).

Another diversification activity getting increasingly more attention is the extraction and marketing of other non-timber forest products. These include nuts, fruits, oils, resins and their processed derivatives. Forest dwellers in Acre are particularly well placed for this possibility as the forest is rich in diversity (Daly, Silveira, this issue), offering a range of potential products from which forest families might choose.

As new products and markets are identified, it is important to analyze how these Anew@ products might fit with current forest marketing systems, whether these systems can be adapted for new products, or whether new marketing systems, i.e. channels and players, will need to be developed. Although there is a growing body of literature on non-timber forest products marketing in the Amazon in general (Browder 1992a 1992b, Coomes 1995, Clay and Clements 1993, Homma 1992, 1993, Nepstad and Schwartzman 1992, Padoch 1987, 1988, 1990, 1992, Plotkin and Folmare 1992, and Richardson 1995) and more specifically on Acre (Clay 1992a and 1993a, ECOTEC 1993, Gama e Silva 1992 and Lafleur and Bryon 1993) we still know little about how products might integrate with current market channels.

In this paper, I describe and analyze the current marketing system for rubber and Brazil in three communities within the Chico Mendes Extractive Reserve and examine the motivations behind rubber tapper trade patterns. Building on these findings, I discuss the implications of the current trading patterns for marketing other non-timber forest products, including how new activities fit with current market-oriented and subsistence farming activities. I focus the analysis on the possibility of extraction and marketing three palm fruits native to the region: açaí (Euterpe precatoria), bacaba (Oenocarpus mapora) and patauá (Oenocarpus bataua). These three palms are native to the region and produce fruits which are already commercialized in the capital city of Rio Branco, the closest regional market.

I conclude that the incorporation of market-oriented non-timber forest product activities into current socio-economic activities may face barriers: current in-forest trading channels may not be suitable for new products and market oriented extraction of new products may compete with current subsistence and extraction activities. To confront these barriers, families will need to pool resources, principally labor and transport, and develop direct ties with processors in urban areas as current channels do not currently trade these products. The success of a few innovating families may encourage others to undertake riskier Anew@ activities.

Study Site

Field research took place in the Chico Mendes Extractive Reserve , located in the south-east region of the state, and the capital of Rio Branco, found near the eastern state border at between . The Chico Mendes Reserve was selected for study implementation because of its relative proximity to the Rio Branco market and the limited attention given to market-oriented studies in the area.

The Chico Mendes Reserve was created in 1990, the result of what was a violent struggle for land rights and social justice pitting rubber tappers against ranchers (see Hecht and Cockburn 1989 and Schwartzman 1992). The reserve is named after Francisco AChico@ Mendes, a rubber tapper, rural union leader and founder of the Xapuri Agro-Extractive Cooperative (CAEX). He was one of many poor rubber tappers murdered in this struggle.

The concept of extractive reserves was first proposed by the rubber tappers at a meeting in the Brazilian capital Brasilia, in 1985. At this meeting, the National Council of Rubber Tappers was established and rubber tappers proposed the creation of extractive reserves, defined by Allegretti (1990) as Apublic lands designated for the specific purpose of sustainable use of forest products...[with] property rights designated according to traditional patterns of land use rather that imported models of occupation,@ In other words, the boundaries of these landholdings, or colocações, are determined by location of resources rather than by conventional geographic shapes: rubber tree trails and/or Brazil nut trees mark boundaries between landholdings. Rubber tappers living in the reserve were given 30 year usufruct rights to the forest resources on their landholdings (see also Schwartzman, 1989).

The Chico Mendes Extractive Reserve is the largest extractive reserve in Brazil with an area of approximately 970,570 hectares stretching across five municipalities: Xapuri, Brasileia, Assis Brasil, Sena Madureira and Rio Branco. There are an estimated 1,834 families living in the reserve spread over 46 seringais, or rubber tree forests, which fall in part or whole in the reserve (CNS, 1992). These forest areas were previously established and owned by seringalistas, or rubber barons, who allowed rubber tappers to live and tap rubber on landholdings on these estates. They now serve as the basis for the organization of many cultural, social and economic activities in the reserve, including community organization, elementary education, religious practice, cooperative post trading and even soccer teams.

Three forest areas, or communities, in the Municipality of Xapuri were selected for study implementation: seringal Floresta, the community of Rio Branco; seringal Boa Vista, the community of São João de Guarani, and; seringal Filipinas, the community of Terra Alta. They were identified and selected in coordination with the Xapuri Association of Rubber Tappers of the Chico Mendes Reserve (AMOREX), a community-based association which oversees socio-economic development activities in the area of the Municipality of Xapuri which falls in the reserve.

The selection of the communities of Rio Branco and São João de Guarani, approximately five and 12 hours walking distance, respectively, to Xapuri, provided an opportunity to consider transport time as a factor in the rural marketing system. Travel from São João de Guarani to Xapuri during the rainy season with an animal loaded with rubber or Brazil nuts can take up to 16 hours direct, making it a 1 2 to two day trip allowing an evening stop on the way. With a day to allow the animal to rest in Xapuri, round-trip travel can take up to five days. The selection of the community of Terra Alta provided the opportunity to examine river transport as a marketing component. Although community families can travel to Xapuri by river or land, river travel by canoe is the principal means of transport. Travel time from Filipinas to Xapuri by canoe ranges from 4-8 hours depending on the season and power of the canoe motor.

Methods

A marketing systems approach was used to analyze the rural production system. La Gra's (1990) commodity systems assessment methodology was particularly useful in identifying key system components for analysis, and was adapted to the study area. Prior studies carried out in the Peruvian Amazon by Padoch (1987, 1988, 1990 and 1992), Padoch and de Jong (1989), were also helpful in conceptualizing the study. Important market system components, including labor constraints, transport, infrastructure, and current trading channels and patterns were analyzed.

In the Chico Mendes Reserve, the study initially included five volunteer families from each of the three forest communities. Guidelines for family participation included: 1) families should reside in the reserve: 2) families should live in the seringal, or forest area, of the community (one exception was made); 3) families could not be acting as marreteiros, or intermediaries, in the selling process, but rather would be traditional extractivist families; 4) families should live within approximately two hours hiking distance from the community center, and; 5) preference would be given to families that indicated experience of household use of at least one of the three palm fruits being studied.

Data collection involved various tools and activities. A total of three visits were made to each of the core group of families in the study. These visits were timed over different seasons to allow for observation of both rainy and dry season periods. An initial study visit of three days was made to each of the five volunteer households of each community to gather preliminary data on the extraction and use of the selected products, and collect data on the current system of commercialization for traditional extractive products - rubber and Brazil nuts. On this initial visit, informal interviews, participant observation and a participatory map-making activity in which the entire family participated were utilized.

On the second visit, which included one day and one evening at 14 of the 15 initial households, a formal questionnaire was implemented. This questionnaire was developed based on data collected during the first visit and focused on obtaining detailed information in key marketing system components. During a third visit to each community, formal interviews were conducted with additional families to increase the sample size and provide richer information trade patterns. A total (including the initial 14 families) of 42 families were interviewed and 39 included in the study.

The history of extraction and marketing of Hevea rubber in the Amazon has been well documented (See Dean 1987, Weinstein 1983, Schwartzman, 1991, Bakx, 1988, Barbosa da Costa 1989). One important element of this history is an understanding of the debt-peonage relationships under the aviamento system which characterized the trading system for rubber from the mid 1800s until the post-World War II period. Under this system, rubber tappers purchased basic supplies and tools at inflated prices from rubber barons at the outset of working the landholding and in turn were committed to selling their rubber at the low price dictated by the barons. This system maintained the rubber tappers in perpetual debt. Plantation produced rubber in Malaysia weakened this system from 1910 to the beginning of World War II, although it was revived when the Axis powers cut-off trade routes.

In the post World War II period rubber prices fell again with the influx of cheap Malaysian rubber onto the world market. Bakx (1988) argues that the effects of this lower price changed considerably socio-economic relations in the seringal. With credit curtailed from supply houses in Manaus, urban-based rubber barons began to substitute the low value subsistence supplies for rubber tappers with higher value items for urban inhabitants. With fewer supplies available for rubber tappers, they permitted rubber tappers to produce subsistence crops which greatly changed the face of the debt-peonage ties: rubber tappers began to have positive balances at years= end, they began to seek out cheaper supplies from urban merchants, and they became more mobile in the forest, seeking out a patrão who paid more for their production. As a result, many rubber barons abandoned or sold their landholdings in the post-World War II period. Yet, while rubber tapper enjoyed greater economic freedom from their patrão, in many cases, marreteiros, or trade intermediaries traveling in the forest, assumed the baron=s role, furnishing basic supplies, clothes and other products at inflated prices and purchasing rubber and Brazil nuts, and sometimes small animals and basic food crops at low prices.

The Xapuri Agro-Extractive Cooperative (CAEX), founded by rubber tappers in 1988, was an effort to provide rubber tappers an outlet for selling and trading their products at better prices than paid by intermediaries in the forest. Membership requires a payment of 100 kilos of rubber (approximately US$64.80). Benefits of membership include selling products at in-forest trading posts or CAEX in Xapuri at prices higher than offered by intermediaries, and at least initially, access to credit to purchase basic supplies. CAEX opened in-forest trading posts in six forest communities, including Rio Branco and Terra Alta, to facilitate trade with rubber tapper families and better compete with intermediaries. The prices paid by the in-forest posts for products is slightly higher than paid at CAEX=s operations in Xapuri. The price of supplies is also higher in-forest.

The traditional market system in the seringal: products traded and trade outlets

In the Chico Mendes Extractive Reserve, families sell and trade principally rubber and Brazil nuts and also artisan crafts, such as brooms and baskets woven from forest fibers and various agricultural products, including basic crops, such as rice and beans and small animals. Figure 1 presents a detailed picture of trading patterns of study families for 1996, identifying principal trade channels and the products traded or sold through in-forest and urban (Xapuri) channels. Non-members can trade with the trading posts as well but receive a lower price for their product

A detailed analysis of trade patterns for rubber and Brazil nut

Tables 2 and 3 presents a more detailed breakdown of rubber and Brazil nut trade by study families. They show the trade outlets for products, the range of prices received by forest families and the quantity of product sold by each community to each channel.

In each of the three study communities, CAEX continues to play an important, although often unreliable, role in rubber and Brazil nut trade. Its impact is greater in the communities of Rio Branco and Terra Alta as CAEX maintains a núcleo, or in-forest trading post, in these two communities, where rubber and Brazil nuts are purchased by note, redeemable for a check in Xapuri, or traded for supplies.

As mentioned above, both cooperative members and non-members may trade products at the trading posts, although the latter receive a lower price and during the research period, were not provided credit to purchase supplies. The trading posts closed in January 1997 and when the Rio Branco post re-opened in the middle of the year (the Terra Alta post remained closed), the official policy was that even members would not be provided credit to purchase supplies. However, the cooperative manager at Rio Branco indicated that he did allow some cooperative members who he knew would pay their debt quickly, to purchase supplies on credit. Among study families, 8 of 13 families interviewed in both the community of Rio Branco and Terra Alta were cooperative members. No families among the 13 Guarani study families were cooperative members. In Guarani, families are five or more hours from the closest cooperative trading post in Rio Branco and had less to gain from membership. Families in Rio Branco and Terra Alta who were not CAEX members gave various reasons for not being members, including an inability to pay the entrance fee, or seeing limited benefits in the near term. (Campbell (1996) identified similar reasons for non-cooperative membership). One rubber tapper was unhappy with the administration of the cooperative and was not a member, although his son was.

Although families in Guarani have very limited trade with CAEX, they do benefit indirectly from its activities, principally from increased competition fostered by the in-forest trading posts, which rubber tappers indicated has forced intermediaries to lower the price they charge for supplies.

CAEX also retains a significant role in trade in the communities where it maintains a trading post. CAEX=s role is particularly strong in Rio Branco where it commands a majority of rubber and Brazil nut sales, with non-members also trading with the in-forest trading post. In Terra Alta CAEX has a lesser role, particularly with regard to rubber sales. This is likely due to the local trading posts regular lack of supplies, leading even cooperative members to trade with intermediaries and merchants. This was true for Brazil nut trade in Terra Alta as well. No non-cooperative members traded with CAEX in Terra Alta. Although not noted on the figures, it is worth noting that in 1996 an intermediary who lives near the Terra Alta community purchased approximately 5,000 latas, or cans (a can equals approximately 12 kilos) of Brazil nuts, approximately 30 tons of nuts with shell, (including landholdings outside the study families) from both CAEX members and non-members for US$1.01 per and sold the nuts to CAEX in Xapuri in Xapuri for US$1.82 per can. In Guarani, only two families traded rubber and Brazil nuts with a CAEX trading post, both small transactions with the trading post in the community of Rio Branco.

Tables 2 and 3 show that the intermediary also has a strong role in trade among study families in the communities of Guarani and Terra Alta, particularly for Guarani. Intermediaries who service Guarani families generally operate from the city, in this case Xapuri, traveling with pack animals on forest paths, some as often as every 15 days during the rubber tapping season. A few operate from landholdings within the reserve. In Guarani, where families have limited access to supplies, intermediaries function as the provider of basic goods to many of them. Intermediaries sell supplies on credit, extremely important to the rubber tappers in Guarani whose majority of income is provided principally by rubber extraction, as this forest area lacks the Brazil nut population of Terra Alta and Rio Branco (based on the author=s personal conversations with rubber tappers in the community). Their income is concentrated during and just after the rubber tapping months, April through November or December (depending on when the rains begins). Purchasing on credit obligates debtors to sell their products, principally rubber but also Brazil nuts and other agricultural products, to intermediaries in repayment. In Guarani and Terra Alta, 10 of 13 families in each community used at least a portion of 1996 rubber or Brazil nut production to pay debt, principally to intermediaries in payment of supplies. In the community of Rio Branco, this ratio fell significantly, with only 3 of 13 study families using rubber and Brazil nuts for debt payment in 1996. Financial difficulties of the CAEX trading post in Terra Alta likely increased the role of the intermediary in this community.

The importance of the trading of supplies to the income of intermediaries is critical to understanding their operations. The following example demonstrates how the two-way transaction benefits him. Using 1997 prices, let=s say a rubber tapper has 10 kilos of rubber to sell. If the intermediary purchases these 10 kilos for cash he would pay approximately R$.56 per kilo in-forest (a total of US$5.60) and sell them for US$.74 per kilo to a wholesaler in Xapuri, with a profit before expenses of US$1.80. If however, he trades supplies for these 10 kilos he makes a return on the sales of the supplies as well. For example, he sells a one kilo sack of sugar for US$.93, purchased from a local merchant for a cost of US$.42 per kilo (the price per kilo would be lower when he buys in bulk). By selling these 6 kilos of sugar (US$5.60 value in trade for the rubber) he earns an extra US$3.06 (R$.51 per kilo for 6 kilos) more than the initial rubber sale. For this transaction, the sale of supplies accounts for approximately 62 percent of total revenue. (The profit from sales of supplies will fluctuate depending on the product). Because of the importance of this dual transaction, intermediaries often require rubber tappers to use a minimum 50 percent of the selling price for the purchase of supplies, with the balance, paid in cash if desired. Intermediaries may be reluctant to simply purchase products from rubber tappers.

Other trade outlets included exchange with other rubber tapper families in the community and the use of rubber and Brazil nuts to pay the Amortgage@ on landholdings. Although the amounts traded are small, no study families in Rio Branco used either of these products for these activities.

What differentiates river transport from land transport is that traveling by canoe entails an actual cash outlay to purchase gas (gasoline or propane gas) and oil. Travel from Terra Alta by canoe costs approximately US$16.67 (US$11.11 for gas and US$5.56 for oil) round-trip. Canoe motors must also be properly maintained and occasionally repaired, placing an additional cash burden on families.

Families in Terra Alta can also hire a canoe to get goods to Xapuri, although this is costly. Canoes for hire charge R$74.07 per round trip for freight from the Terra Alta community to Xapuri. This usually involves two days travel for the canoe operator, one to travel to Terra Alta with a return the next day. There are three principal canoe operators in Xapuri and they all charge this same rate, bringing an appearance of price collusion. I did however find this price to be negotiable and made the trip for as low as US$46.30 for a round trip made on the same day. I payed the owner cash in advance. Other times I paid the full US$74.07. Operators indicated that the limited traffic which they share among them is the reason for the high price, with some operators making less than one trip a week to various communities located along the Acre and Xapuri Rivers and their tributaries.

Second, the above discussion spent considerable time analyzing the role of the intermediary in the traditional marketing system. The potential role for the intermediary in the commercialization of Anew@ products is a complicated one. The cooperative was created to lessen rubber tapper dependency on them, yet this study reveals that they play an important role in the marketing system, providing services which the cooperative and other commercial entities are unable to. Would it then be a good idea to identify a role for the intermediary in the marketing of complementary extractive products? Possibly there might be an opportunity to take advantage of services the intermediary might offer, such as use of large teams of pack animals (the principal intermediary in Xapuri has 35 pack animals), rather than bringing him in as a buyer. The larger question is whether an intermediary would be interested in purchasing perishable products, such as palm fruits or even providing a service, such as transport.

The suitability of pack animals and canoes for transporting fruits is another important question? For river travel, fruit, normally transported in 60 kilo capacity nylon sacks, can easily be loaded in canoes and covered to protect it from the sun. However, for land transport, with sacks roped to the back of pack animals, fruit might be damaged over long distances. Although all three fruits are hard to the touch, rubber tappers indicated that açaí is softer and the thin mesocarp may be sensitive to being jostled on the back of an animal over a long distance. A container with a hard shell, such as plastic milk jugs already used to transport in-forest processed Brazil nuts to Xapuri, might be used to transfer fruit as well.

Rubber and Brazil nuts are particularly important when families have limited off-farm wage labor options, such as the community of Guarani. In the community of Rio Branco, some families are already turning to alternative income generating activities and this need for complementarity is diminished. New activities will also have to complement subsistence farming activities which greatly reduce the rubber tappers= cash needs.

Açaí ripens in the dry season and this will benefit communities which rely on land transport. Families in Guarani could potentially travel round-trip to Xapuri in two days, although three is more realistic and allows time for the animal to rest. Communities which rely on river transport may find transport difficult in this period, principally toward the end of the açaí harvest season when rivers are low.

The harvesting of bacaba would also compete with extraction and farming activities including the collecting and shelling of Brazil nuts, the clearing of rubber tapper trails and the harvesting of corn and rice. Families without Brazil nuts on their landholdings may be in a better position to allocate labor toward this activity as Brazil nut collection and shelling is primarily an adult male task -- shelling Brazil nuts requires the heavy swing of a machete to break through the hard shell. In Guarani, many families do not have Brazil nut trees.

The ripening of bacaba occurs in the rainy season which will add considerable transport time for communities distant from Xapuri, such as Guarani. Closer communities such as Rio Branco will also be affected but less so. Communities traveling by river, such as Terra Alta, will find transport easier during these months due to higher water levels. During this period the cooperative also periodically travels to this area, although, as discussed above, priority is given to hauling Brazil nuts during this period and there is often only limited room for other commodities.

The marketing of traditional products, rubber and Brazil nuts remains, for many families, a dual transaction, with families producing and selling their products in trade for important supplies. Families seek out trading partners who are able to meet their credit and supply and subsistence needs. They trade with a variety of different actors both in the forest, and in some cases the city. Key to rubber tapper participation in market activities is minimizing risks. The price or trade value of rubber may be low, or unstable, in the case of Brazil nuts, but these markets are essentially guaranteed. Rubber tappers know they can sell these products. As they are non-perishable, they can do so when access to transport and family labor permits.

Because capital and cash resources are limited, families are looking for markets which minimize risk, and offer similar Aguarantees.@ They may not exist. New products will not, at least initially, provide access to credit and supplies nor serve as payment for debts or provide the guarantee to which families are accustomed and families may be unwilling to shift labor and transport resources to extract and transport them. Currently there are no buyers of açaí, bacaba or patauá in the Chico Mendes Reserve: neither CAEX, its trading posts or intermediaries or merchants purchase or trade fruit products. Families without debt may be best placed to undertake new and experimental market-oriented activities. Marketing channels will have to be established which minimize potential losses to rubber tappers willing to test new ventures.

New products can differ in terms of how they fit with current market-oriented activities depending on their harvest season. Communities which travel by land will find it easier to harvest and transport goods during the dry season. Even distant communities such as Guarani could participate in markets. Communities which rely on river travel will be benefitted by the rainy season when canoes can navigate essential without problems. Local river traffic increases during this period increasing transport opportunities.

As capital and labor resources among families are scarce, communities, or teams of families, rather than individual families may be better placed to supply these products. Teamwork will reduce the labor burden and time constraints of extraction on each individual family and through team work, transportation resources such as animals and canoes can be pooled and costs can be shared to reduce cash outlays of each family. Individual families can take turns transporting fruit from the community to Xapuri, reducing significantly the time allocation to accompany sales in the city. Families in distant communities such as Guarani might consider a cooperative membership to gain access to the cooperative=s pack animals to facilitate transport.

There are clearly considerable challenges to identifying products and markets, but markets for some products already exist and as Amazon towns grow the demand for natural fruit drinks and pulps will only increase. More research is needed on all fronts, particularly ecological research to better understand how to manage extractive resources. Local markets, with limited demand provide a good opportunity to measure the ecological impact of market-oriented extractive activities.

To link into markets, communities will need to take a pro-active approach to identifying a market niche which matches their resource constraints. They will need to make contact with market channel players to evaluate where they can best enter the market system. Local and regional markets, closer to forest communities and with lower demand are a good place to start. Communities with not dissimilar circumstances are already successfully marketing products there (Rocha and Ferraz 1998). Processing activities to add-value to products and increase the communities share of the final product value might be considered.

Local research and extension organizations can and should play a assisting role by helping communities make contact with buyers, analyzing regional transport options and costs, and even support local in-forest community organization in the formation of extraction teams and training in non-predatory extractive methods. Research and assistance must be participatory and in coordination with association and family objectives. Researchers should be cognizant that these may be different. Families can participate according to their interest and resources. Research laying out how markets function, the players and pricing can help forest communities better identify and analyze potential opportunities.

An initial challenge, and maybe the largest, is getting non-timber forest products on the menu of options. While important research and extension activity has focused on agroforestry, and traditional agricultural activities, non-timber forest products, other than rubber and Brazil nuts, has been largely ignored with a few exceptions. Researchers at the Botanical Park of Federal University of Acre are carrying out ecological research on açaí, bacaba and patauá and, in coordination with researchers of other local organizations, are doing interesting work on management of copaiba (Copaifera spp) for oil extraction. Without an increased focus on non-timber forest products as viable options, forest communities are likely to focus on agroforestry and agriculture and remain largely ignorant, as we all will be, of the economic potential of extractivism.

The author would like to thank Connie Campbell, Reginaldo Castela, Arthur Leite, xxxxxx Rego and Ronei xxxx for comments on a very early draft of this paper while the author was conducting research in Acre. He would also like to thank AMOREX and CAEX in Xapuri, Acre for assistance in field research in the Chico Mendes Extractive Reserve. The author is also grateful to the Rainforest Alliance which provided a grant under which this research was carried out.