Forest, field and factory : changing livelihood strategies in two extractive reserves in the Brazilian Amazon


Material Information

Forest, field and factory : changing livelihood strategies in two extractive reserves in the Brazilian Amazon
Physical Description:
ix, 194 leaves : ill. ; 29 cm.
Campbell, Constance E
Publication Date:


Subjects / Keywords:
Forest products industry -- Brazil   ( lcsh )
Nut industry -- Brazil   ( lcsh )
Land use, Rural -- Brazil -- Acre   ( lcsh )
Forest management -- Brazil -- Acre   ( lcsh )
Natural resources, Communal -- Brazil -- Acre   ( lcsh )
Anthropology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Anthropology -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1996.
Includes bibliographical references (leaves 179-193).
Additional Physical Form:
Also available online.
General Note:
General Note:
Statement of Responsibility:
by Constance E. Campbell.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 36772267
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
        Page iv
        Page v
    Table of Contents
        Page vi
        Page vii
        Page viii
        Page ix
    Chapter 1. The political ecology of extractive reserves in southeastern acre
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
    Chapter 2. Transforming economies and social organizations in the extractive reserves
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
    Chapter 3. Transforming traditional economies
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
    Chapter 4. Decentralized Brazil nut processing and the rubber tapper household economy
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
    Chapter 5. Conclusions
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
    Biographical sketch
        Page 194
        Page 195
        Page 196
Full Text





I dedicate this dissertation to my grandmother, Francis Mullins Campbell. Her first job was as a teacher in a one-room schoolhouse in rural Illinois. She went on to be one of my best teachers. From her I have learned the value of hard work done with Scottish Presbyterian stubbornness combined with the joy of family.


This dissertation was supported by research and institutional support from the Tropical Conservation and Development Program and the Center for Latin American Studies at the University of Florida. I was honored to be the first recipient of a Charles Wagley Endowed Research Fellowship for the final stages of my field research. The Florida-Brazil Institute provided travel assistance as part of this fellowship.

During my coursework, the Department of Anthropology provided travel funds for several conferences. The Center for Latin American Studies provided me with assistantships and research stipends through the "Agroforestry Development Program for Small Producers in the State of Acre, Brazil", funded by the U.S. Agency for International Development and managed by Dr. Marianne Schmink. The Center also provided office space and financial support for conference trips. In addition, my coursework was supported by an assistantship from the Women in Agricultural Development Program which also provided office space while I wrote this dissertation. I also was honored to receive two years of support as a P .E.O. Scholar Awards recipient.

From preliminary conversations revolving around my decision to enter the doctoral program to the final printing of this dissertation, I have been blessed with the unfailing support of my committee chair, Dr. Marianne Schmink. I am very grateful to have had the invaluable experience of learning from her and becoming a close friend and colleague. My iii

doctoral work has been greatly enriched by the dynamic interdisciplinary environment at the University of Florida. My committee members contributed to that experience and I thank Drs. Helen Safa, Anthony Oliver-Smith, Pete Hildebrand and Clyde Kiker for their guidance and insights over the course of my doctoral program. Special thanks to the staff of the Center for Latin American Studies and the Tropical Conservation Development Program including Margarita Gandia, Carmen Myers, and Lydia Gonzalez. I am very grateful to Carla Riley and Kevin Veach for their special efforts in producing this dissertation.

While doing my field research, many friends and colleagues provided invaluable companionship and refreshment in liquid and spiritual form. I thank Karen Kainer, Jon Dain, Richards Wallace, Peter Cronkleton, Marianne Schmink, Samantha Joy Wood, Gary Shaeff and Phyllis Nix for many fond memories of our communal house in Rio Branco. As an anthropologist in an interdisciplinary program, I valued the unique opportunity of conducting field research in the same community as a fellow doctoral candidate from UF's Department of Forestry. I know that this dissertation has benefitted directly from this interdisciplinary experience and I am grateful to Karen Kainer, a faithful friend and colleague, for sharing data insights and valuable gossip with me. My academic and professional development also benefitted greatly from the Grupo PESACRE, which has been a constant source of institutional and personal support.

They say that the friends that you make in graduate school are with you for life. I hope that the ties that I wove with many people over the years at Grinter Hall will remain pliable and strong. Those who made this an enriching and fun experience include all of those individuals named above as well as Gay Biery-Hamilton, Peggy Lovell, Avecita Chicch6n, iv

Richard Piland, Pennie Magee, Diego Hay, Karen Kraft, Jackie Jeffihies, Nathalie Lebon, Martha Cuba and Marcia Good-Maust. A newer generation of students gave me a shot of energy and enthusiasm and I thank them for their friendship and for putting a new light on the academic experience for me. These wonderful people include Amanda Stronza, Katie Lynch, Holly Payne, Kevin Veach, Elena Bastidas and Christine Housel.

Without the constant support which emanated from the home front, this dissertation would never have seen the light of day. Polly and Wilson Harris kept the mailboxes full of letters, cookies and encouraging e-mail messages. Reg and Roni Campbell provided me and Eduardo with much needed support, including respites from graduate school during shared vacations. Catherine and Carol Campbell each came through with sisterly love and, along with the rest of the clan, took my wanderings back and forth to Brazil in stride and always kept me rooted in the strength of family.

Finally, to the four men who energized and sustained me during this time: Dr. Pepper, Ben and Jerry, and my husband, Eduardo. It is to Ed whom I owe the existence of this dissertation. I eagerly look forward to the challenge of being as patient, creative and constant in supporting his doctoral work as he has been in mine. It will be a tough act to follow.




A CK N O W LED G M EN T S .............................................................................. ill

A B S T R A C T .................................................................................................... v iii


RESERVES IN SOUTHEASTERN ACRE ............................... I

Extractive Reserves and the Movement in Xapuri ............. ............. 2
Political Ecology of the Extractive Reserves .................................. 7
Non-Timber Forest Products within the Political
Ecology of the Extractive Reserves ............................................ 14
R esearch Sites and M ethods ........................................................... 18

R E SE R V E S ........................ ....................................................... 2 1

A cre an d X apuri ............................................................................ 22
The R ubber B oom in A cre ............................................................. 27
The Colocacao: A Changing Socioeconomic Space
in th e F o re st ................................................................................ 3 1
The Seringal: From Rubber Estate to Community ......................... 43



Management and Processing of Brazil Nut
(castanha) Bertholletia excels .................................................. 63
Post-H arvest Processing ................................................................ 74
The Decentralized Brazil Nut Project of the CAEX ....................... 77
Productivity and Income Levels of the
D ecentralized P roject .................................................................. 94
C o n clu sio n s ................................................................................... 10 2


The Decline of Rubber Tapping in the Study Area ........................ 107,
Transforming Livelihood Strategies ............................................... 120
Changing Gender Roles in Household Economies ......................... 132
Defining Rights and Responsibilities .............................................. 142
M ig ratio n ............................................ ........................................ 14 7
Perceptions of Household Economic Well-Being ........................... 156
C o n clu sio n s ................................................................................... 16 3

5 C O N C L U SIO N S ........................................................................... 165

Solidarity and Institutionalization of the Reserves .......................... 165
Transforming Household Economies ............................................. 169
C hanging Social Identities ...................... ...................................... 172
Increased Production through NTFP Post-Harvesting ................. 173
Challenges for the Extractive Reserves .......................................... 175

R E FE R E N C E S ....................................................................................... .......... 179

B IO G R A PH IC AL SK E TCH ................................................................................ 194


Abstract of Dissertation Presented to the Graduate School of the University of Florida in
Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FOREST, FIELD AND FACTORY:


Constance Elaine Campbell

December 1996

Chairperson: Marianne Schmink
Major Department: Anthropology

This study examines socioeconomic changes in Acre, the westernmost state in the Brazilian Amazon. In the southeastern municipality of Xapuri, extractive producers have created and sustained a socio-political movement that has resulted in a new model of communal land management called extractive reserves. Extractive reserves are federal tracts of forested land that have been designed to provide an alternative to Amazon development policy by securing the usufruct rights of forest residents to remain in protected areas and to continue practicing traditional, nondegradatory livelihoods.

This study explores the impacts of a project within two extractive reserves that was designed to generate rural employment and add value to Brazil nuts, a traditional product harvested from the forest. Through post-harvest processing (shelling and drying) of the nuts


in the extractive reserves, the project sought to increase the income levels of forest dwellers, thus stabilizing the rural population by improving the general economic well-being of extractive reserve residents.

Using a political ecology framework, I examine this social movement, the extractive reserves and the Brazil nut project in order to determine their ability to secure sustainable livelihood strategies for traditional forest dwellers. I argue that such transformations of local economies and the construction of new forms of community in the reserves combine with Acre's unique history, ecology and socio-political conditions and force us to rethink the abstract inevitability of proletarianization of the peasantry on an expanding capitalist frontier and the relegation of small producers to subsistence-level production.



This study examines socioeconomic changes in Acre, the westernmost state in the Brazilian Amazon. In the southeastern municipality of Xapuri, extractive producers have created and sustained for over two decades a socio-political movement that has resulted in a new model of communal land management called extractive reserves. Extractive reserves are federal tracts of forested land that have been designed to provide an alternative to Amazon development policy by securing the usufruct rights of forest residents to remain in protected areas and to continue practicing traditional, nondegradatory livelihoods.

With the demarcation of the reserves, these small-scale commodity producers have been able to maintain their autonomy and effective control of their land in the face of rapid capitalist expansion. Their experience differs from that of producers in the eastern Amazon, where it has been argued that the expanding frontier and capitalist penetration results in a general transformation from autonomous peasant producers to a proletarianized work force (Foweraker 1981). The culturally and historically unique character of these Amazonian peasants, or caboc/os (Nugent 1993), in large part determines the ongoing transformation of their traditional extractive economies into a new, distinct "house model" (Gudeman and Rivera 1990) within the reserves. Using a political ecology framework, I argue that such


transformations of local economies and the construction of new forms of community in the reserves combine with Acre's unique history, ecology and socio-political conditions and force us to rethink the supposed inevitability of proletarianization of the peasantry on an expanding capitalist frontier.

This study centers on the transformation of econonues and the continuing emergence of communities in the extractive reserves. One alternative production strategy, the postharvest processing of Brazil nuts, is examined in light of these change processes. Research questions focus on the impact of this Brazil nut project in conjunction with other organizational and economic shifts in the reserves and in the social movement of local producers. How does the project affect income levels of the households and individuals involved? How is this income distributed within the household and how are decision-making strategies changing? Are gender roles in the household and in the community changing because of the project? What other shifts in productive activities and gender roles are taking place? What are the implications of these changes for the future viability of the reserves in this study? These questions are addressed within a political ecology framework which contextualizes the Xapuri rubber tappers' movement in the historically specific experience of Acre as the westernmost state in the Brazilian Ai-nazon.

Extractive Reserves and the Movement in Xapuri

The rubber tappers' movement in Xapuri and the eventual establishment of the extractive reserves was a response to the federal government's Amazon development policy of the 1970s and 1980s. Federal and state governments promoted the expansion of capital-


intensive agricultural enterprises in the Amazon, most of which consisted of cattle ranches that enjoyed significant benefits in the form of tax write-offs and reduced-interest financing and which required clearing of large tracts of forest (Browder 1986, 1988; Hecht 1982; Schmink and Wood 1992). Although there were fewer subsidies awarded for these activities in Acre than in the eastern part of the basin, many rubber tapper families were expelled from the forest as ranchers claimed the old rubber estates, or seringais. Some families fled to Bolivia, where they continued to tap rubber, while others took up residence on the periphery of Rio Branco, the capital of Acre.

While the BR-364 highway was being paved through the neighboring state of Rond6rua on its way to Acre, bringing widescale deforestation and social conflict, the rubber tappers in Xapuri organized themselves to defend the forest and secure their livelihoods. The movement was led by Chico Mendes, a rubber tapper and president of the rural workers' union who was assassinated by a local ranching family in 1988. Chico Mendes did not oppose construction of the road, but he organized the rubber tappers to protect their claim to forested areas that lay within the area affected by the BR-3 64.1

The rubber tappers' most effective tool in defense of the forest was the empate. In these demonstrations in the forest, women, men and children would gather at the site of a forest clearing. Unarmed, they approached the ranchers' hired laborers, encouraging them to stop the deforestation by appealing to class interests. The ranchers responded with increasingly violent measures through their alliances with local police. The rubber tappers

'See 111GB (1990); Schmink and Wood (1992, p.1 14-117) for more on this road, affected areas and related projects.


were often met by armed police who had been called to the area (Allegretti 1990, 1994; Schwartzman 1989). In crucial empates women and children formed the front line of defense, placing themselves between the forest and the ranchers' chain saws (Campbell forthcoming).

In their struggle for social justice, the rubber tappers sought to avoid eviction and defend their livelihood.' Together they shared a common goal with national and international environmentalists -- that of saving the Amazon rainforest. Alliances with key individuals and organizations in the environmental arena strengthened the Xapuri movement (Keck 1995, Schmink and Wood 1992). With assistance from various non-governmental organizations and prominent supporters, the Xapuri rubber tappers worked together with extractive producers throughout the Amazon basin to hold a national meeting in 1985 in Brasilia. During that meeting, they created the National Rubber Tappers' Council (CNS) and put forth the proposal to create a new type of land reform, extractive reserves (Allegretti 1990, 1994).

The rubber tappers' movement in Xapuri achieved major victories with the establishment of the first extractive settlements in 1987 and the later creation of extractive reserves in 1990. The role of the state in creating the extractive settlements and reserves was somewhat akin to the policy of "crisis colonization" carried out in the eastern Amazon (Schmink and Wood 1992, p. 172). In the extractive case, some areas were declared settlements or reserves in order to diffuse conflict between rubber tappers and ranchers. In others, the state used areas that were indemnified and slated to become colonization projects,

2 See Juan Martinez-Alier and Thrupp (1992); Joan Martinez-Alier (1991);
Schwartzman (1991, 1992b) on parallels between this and other social movements and ecology.


irrespective of the rubber tappers' movement identification of areas that it considered to be priorities for extractive reserves. This led to accusations on the part of the movement that the government created reserves in areas that were not unionized, knowing that the extractive reserve model most likely would not work well there, with the intention of discrediting the movement and the concept of the extractive reserves. 3

The concept of the extractive reserves is based loosely on that of indigenous reserves, in which traditional peoples are guaranteed usufruct rights on federally protected land. The reserves are an alternative to other development initiatives at the federal level because they grew out of a grass-roots movement, were based on collective use and did not require division of the land into smaller, individual plots. The reserves are federal property under the jurisdiction of EBAMA, the federal agency for natural resource management. The outer boundaries of the reserves are demarcated, but the area within the reserve is common property to be managed collectively by an association comprised of all residents.' Each household signs a 20-year contract with the federal government, agreeing to abide by the management plan of the reserve in which the rules and regulations for resource management are defined. Households within the reserve are responsible for management of their own forest tract, the boundaries of which are not demarcated but which are respected through traditional use claims.

See Allegretti (1994, 1995); Gomes and Felippe (1994) for a discussion of federal extractive reserve policy.

See Silberling (1992) for a discussion of the common property aspects of the extractive reserves.


The defense of the forest via the extractive reserves was only one priority for the movement in Xapuri. The movement also sought to protect and improve the lives of the forest-dwelling rubber tappers by establishing political and economic autonomy. Through the creation of a rural workers union, construction of schools and health posts, and establishment of the region's first agro-extractive cooperative, the movement has been able to realize significant changes for the lives of some of its participants (Allegretti 1990, 1994; Campbell 1990; Schwartzman 1989). Families in the area served by the union and the cooperative are now autonomous in that they can sell their products freely to the cooperative or to other traders. These families are thereby released from the conditions of debt-peonage resulting from the patron-client system which was established during the nineteenth century rubber boom in Amazonia. These victories are even more significant because of the very low levels of literacy, the sheer physical distances between families in the reserves and their isolation from urban markets.

During a seminar held by the CNS in Rio Branco in December, 1992, participants discussed directives for extractive reserve programs geared to address these challenges of marketing and social services. They emphasized that the original concept of the extractive reserve model was to preserve the way of life of agroextractive populations and to protect the forests in which they lived from conversion to pasture (CNS 1992a). In this study I examine the reserve model in light of an alternative technological innovation which has been recently introduced into the reserves in southeastern Acre. The incorporation of this production practice represents the ongoing process of constructing the extractive reserves.

The participants in the 1992 conference highlighted the importance of this process.


The praxis of the reserves is changing rapidly in response to the problems encountered in current forest resource systems. We find ourselves at a critical juncture in the development of various potential extractive systems, based on current practices. Thus the extractive reserve is a relatively new proposal that is defining itself as problems are understood and as strategies surface which could resolve them. . The development of viable extractive reserves is a long-term process and will not simply
result automatically from demarcation of the reserves. (CNS 1992a)

Using a political ecology framework, I examine the impact of one such strategy on the construction of new forms of social organization and the transformation of household economies in two reserves in southeastern Acre. This strategy is the decentralized postharvest processing of Brazil nuts. By linking small, community and household-level processing facilities in the extractive reserves with the central factory located in the city of Xapuri, the cooperative has devised a modified extractive system which increases productivity and generates employment in the reserves without putting, additional pressure on the resource base. The mechanics of the project are presented in Chapter 3 The remainder of this chapter is devoted to contextualizing this production alternative in the broader political ecology of the extractive reserves,

Political Ecology of the Extractive Reserves

Political ecology links an ecologically grounded social science approach within a political economy framework by focusing on the relationships between people, natural resources and the surrounding socio-political structures. In applying an analytical framework of political ecology to the extractive reserves in southeastern Acre, I address several new


directions within the specific context of frontier development.' In the following sections, I examine the refinement of political economy in relation to social change, especially in frontier areas.

Peet and Watts point out that one of the innovative areas of inquiry within political ecology has to do with refinements to the concepts and application of political economy (Peet and Watts 1993). They criticize earlier political ecology authors, such as Blaikie and Brooksfleld (1987), for being too all-encompassing in their treatment of the world economy and for taking an almost voluntaristic stance on social factors and causes. The same criticism has been made of broader applications of political economy, although not necessarily within the political ecology framework. For example, William Roseberry notes that the related political and economic questions that became known as the agrarian question (having to do with the fate and revolutionary potential of the peasantry in the face of expanding capitalism) were, at one level, largely political questions to which largely economic answers were given. A focus on class-based issues led to a very poor understanding of the various political leanings and actions of peasants (Roseberry 1993). Sherry Ortner also argues that political economists are too economiustic and materialist in their approach, assuming that the capitalist system is ever-pervasive and that, therefore, anthropologists must recognize that our fieldwork is shaped by this system (Ortner 1984). This capitalist-centered world view of political economy can prevent its practitioners from truly knowing the full richness and variety of human experiences in relation to nature, especially as people's "ethnoscapes" (Appadurai

See Peet and Watts (1993) for a review of political ecology and new areas of inquiry in the field, including the refinement of political economy.


1991) and "imagined communities" (Benedict Anderson 1991) are broadened far beyond the physical limits of their geographical context. Particularly in the Amazon basin, peasant or caboclo production systems have been characterized as static and undifferentiated, when in fact just the opposite is true (Nugent 1993).

The presumed predominance of the capitalist system is increasingly challenged as researchers explore the variety of economic trends and livelihood strategies in which people are engaged (Chibnik 1994; Deere 1990; Garcia Canclini 1993;- Guha 1989; Lockwood 1993; Nugent 1993; Reed 1995; Schmink and Wood 1992; Sorensen 1988; Wilk 1991). Agricultural smallholders in various parts of the world (Netting 1993) and in various agroecosystems (Altieri 1990) have devised livelihood strategies by which they resist proletarianization and retain control of their land through subsistence and market.

In order to strengthen the link between society and nature in political economy studies, William Roseberry calls for a materialism that explores the culture and consciousness of a certain population (Roseberry 1989). In the same vein, David Cleary's analysis of the Amazon calls for "a reconstructed political economy [which] should begin with the region and work outwards" (Cleary 1993, p. 338). This approach is taken in this study by working from the experiences of the rubber tappers in Xapuri who based their resistance and social movement on the defense of their nature-society relationship (Joan Martinez-Alier 1991; Juan Martinez-Alier and Thrupp 1992).

Classic Marxist political economy predicts that, with the development of frill-scale, capitalist agriculture, the peasantry will become increasingly articulated within the capitalist system and will differentiate into either rural bourgeoisie (commodity producers in


agriculture) or rural proletariats (rural wage workers). Thus, peasants or small commodity producers eventually disappear with the development of capitalism because they constitute a transitory part of a social class (de Janvry 198 1). However, the penetration of capitalism into agriculture varies widely and the transitions from one mode of production to another are not distinct, such that there may be multiple modes of production in existence at the same time, particularly in frontier zones (Wood 1982). Thus, in the transition towards capitalism, it is possible that small producers will be neither true proletariats or true peasants but rather "semi-proletarians" (de Janvry 1981), reliant upon a combination of subsistence and commodity production which may include wage labor.

Does this ine of reasoning apply in frontier situations? The application of the political economy model to the expansion of the frontier in Arnazonia has been instrumental in explaining social change as the capitalist market system penetrates the hinterland (Foweraker 1981). However, Ortner's criticism of the strongly economistic concentration of the political economy model (Ortner 1984) is affirmed by several reviewers of Foweraker's theory who note that it is difficult to apply this abstract framework to a variety of complex social and political relationships (Cleary 1991; Dean 1984;1 Schmink and Wood 1987). Regional variation within the basin calls for closer attention to differences within the peasantry (Nugent 1993), urbanization (Becker 1995), social activism and different frontier dynamics (Bakx 1988; Cleary 1991, 1993; Hay 1988; Martins 1990; Rudel and Horowitz 1993; Schmink and Wood 1992).

1 argue that the dynamics of social change, state policy and economic development on the Amazon frontier call for modification to traditional political economy, in particular its


transhistorical, evolutionary stages of development and social labels of "peasant," "proletariat' and "capitalist." I agree with David Cleary who argues that it is crucial to recognize the difference between capitalization and monetization of the economy in the Amazon frontier. This incomplete penetration of the region by capitalism results in a complex environment in which the outcome of migration, urbanization and proletarianization trends fail to fit the empirical predictions of traditional political economy. Cleary thus calls for a modified political economy which takes into account the importance of the informal economy, various forms of human agency, the changing role of the state and the diversity of livelihood strategies in which frontier populations are engaged (Cleary 1993).

In moving away from traditional models of capitalism which apply evolutionary or linear stages of development such as "pre-capitalist" or "capitalist" in the Amazon basin (Foweraker 1981; Hay 1988), Charles Wood encourages a political economy approach which allows for the recognition of various modes of production in a particular area at the same time (Wood 1982). Cleary concurs in arguing that the complexity of changes and processes in the Amazon make use of political economy's hegemonic, evolutionary model of noncapitalistprecapitalist-capitalist stages obsolete.

Questioning the assumption of pervasive capitalist penetration is essential in order to recognize the variety of econom-ic trends and livelihood strategies being practiced throughout the region, strategies that call into question the use of terms that have become somewhat vague such as "peasant" (Cleary 1993), "small faniner," or "cattle rancher" (Browder 1995a). Stephen Nugent calls for a closer examination of the Amazonian peasant caboclo, a social category whose cultural and economic distinctiveness is often overlooked (Nugent 1993).


1 have chosen to use here the term which people in the two research communities use to describe themselves; rubber tapper or seriiigueiro ,

This refinement of political economy requires that anthropologists carry out household studies that build on Chayanovian econom-ic analysis (Roseberry 1993), Within the extractive reserves, I argue that these producer households do seek to balance the need to balance cash and subsistence requirements. Through new production alternatives, households and communities are investing their labor, working capital and future earnings to build autonomous processing facilities through which they add value to a traditional extracted forest product. Households and individuals intend to use the increased cash income from such enterprises for the purchase of transport animals, cattle, urban schooling for their children, improved housing, and infrastructural improvements on their land. In Chapters 3 and 4, I will show that rubber tapper households are not economically stagnant, as Nugent has argued for caboclo economies in general (Nugent 1993). Reserve residents are actively participating in defining new livelihood strategies as the extractive reserve model is constructed.

These new livelihood strategies bring about significant changes in gender roles, the division of labor and the allocation of resources within rubber tapper households. As will be seen in this study, changing market values for forest products (Leach 1994) has implications for the gendered social spaces and resources to which different household members have access (Rocheleau 1988; Rocheleau et al. 1995). With the introduction of wage labor in the

6 For a gendered discussion on the female equivalent of this term, see Campbell 1996. For discussion of the Amazon caboclo, see Nugent 1993; Parker 1985.


processing units, rubber tapper households are following a pattern noted by Carmen Diana Deere who recognizes that households are often the site of multiple class interests with important and varied rendered roles (Deere 1990; Deere et al. 1995). These shifting roles in the household call for attention to intra-household cooperation and conflict (Sen 1990). As Deere has argued in her analysis of peasants in Peru, the producers in this study have been able to resist fiill proletarianization and escape from poverty by employing a multitude of income-generating activities in a rendered division of labor. These household-level implications will be explored in Chapter 4.

Community-level shifts in social organization and economic relations are also taking place within the extractive reserves. Within these new economies, I argue that rubber tapper households are afforded the opportunity to make investments in post-harvest processing only because their communities are politically and socially mobilized. Thus, they have access to credit opportunities or external funds to finance start-up costs of production alternatives. These opportunities exist through the national and international alliances formed between the rubber tappers' social/ecological movement and environmental allies and donor agencies. These relationships, the opportunities and the roadblocks faced by the rubber tappers in Xapuri are due to the fact that the extractive reserves are a confluence of discourses, interests and actions on several levels. Each reserve will be different depending on the local discourse and practice and its relationships to other discourses and practices. The next section in this chapter brings in one such discourse and debate, which is the promotion of non-timber forest products as a "conservation" and "development" alternative for the extractive reserves.


Non-Timber Forest Products within the Political Ecology of the Extractive Reserves

The debate over the suitability of promoting non-timber forest products draws in several key issues within the political ecology framework. The viability of smallholder production, the danger of increased reliance on unstable markets, overexploitation of native forest stands and diverting attention from the structural issues of land tenure, state responsibility for market development and human rights are all hotly debated in relation to the marketing of non-timber forest products. In this study, I explore this debate in light of an ongoing project in the extractive reserves--the post-harvest processing of Brazil nuts.

This study focuses on a community-level project in the western Amazon which is at the confluence of the conservation-development debate. In defining alternatives to past development practices that resulted in widescale deforestation and severe social costs, local communities, municipal governments and donor agencies seek to meet developmental goals of increased household income and stabilized rural and urban-fringe populations, and conservation goals of decreasing pressure on Amazonian rainforests. One alternative that has been enthusiastically embraced by some researchers and rainforest communities is that of adding value locally to non-timber forest products (NTFPs) that are either gathered from the forest or cultivated in agroforestry plots.

The marketing of NTFPs through the so-called "rainforest harvest" is a development initiative that is intensely debated by its promoters and detractors in both the academic and popular literature. As pointed out by Anthony Anderson, the debate has become polarized to the point of obscuring objective analyses of extractivism as a development option (Anthony B. Anderson 1994). 1 agree with others who caution that the promotion of non-timber forest


products or other forest management practices could serve to draw attention and resources away from structural issues such as land tenure, human rights and the need for local communities to define their own alternatives to development (Corry 1993; Escobar 1995,Nugent 1991). However, it is important to note that the non-timber forest product examined in this study was initiated and carried out by a local cooperative of agro-extractive producers with technical assistance from an non-governmental organization. Also, the communities involved in this project have secure land tenure and an established cooperative marketing organization through which they market their production. Thus, they differ from other rainforest groups who may have become involved in other non-timber forest projects that paid less attention to such matters and focussed instead on marketing novel products.

In order to define cominon ground for this discussion on non-timber forest products, I present here Jefferson Fox's (1995:189) definition:

NTFPs are defined both by what they are and by how they are harvested. They encompass all biological materials that are extracted from natural forests for human use. These uses include foods, medicines, resins, latexes, dyes, fodder and fiber grasses, wildlife (products and live animals), fuelwood, rattan, bamboo, and smallwood. Methods of harvest include the collection of natural products from the forest and smal.1-scale production of products that originally grew naturally. Thus the definition includes wood harvested for local use for housing, construction or export
from the region of origin but not wood harvested by large-scale timber interests.

While this definition is fairly clear, the debate over whether management of these products can benefit local communities and conserve tropical forests over the long term is very murky. Adding to the complexity of the debate are the differences between cultivated and gathered non-timber forest products. Most of the literature to date on NTFPs has examined extracted products and attempted to determine the economic value of the forest,


without adding value to the products through post-harvest processing. Yet there are significant differences between cultivated and gathered products and between those products sold en natural and those to which value is added. The literature generally has examined NTFPs marketed by individuals. There are significant differences which must be examined between cases of cooperative marketing versus those of individual producers marketing NTFPs directly.

Those studies that have focused on socioeconomic issues have tended to concentrate on comparative income levels from NTFPs as opposed to other management strategies. Other key variables that must be examined in evaluating NTFPs as a conservation and development alternative include intra-household access to and control over certain resources, intra-household decision-making strategies, availability of labor for production and processing, social organization for production and marketing and changes in population settlement and urban migration due to the level of centralization in the production, processing and marketing stages. The implications of these variables for different genders and socioeconomic classes within participating households and communities are far reaching and profound. In the end, such implications could determine the performance and equity of projects which promote processing of non-timber forest products.

This study varies on two fronts from the majority of the NTFP literature. First, this study examines the socioeconomic impacts that result from post-harvest processing NTFPs. Second, this study focuses on gender issues at the household and community levels and examines the socioeconomic changes resulting from the introduction of post-harvest processing and cooperative marketing of non-timber forest products. These two points link


the NTFP debate back to issues within the political ecology framework that I raised earlier. In the construction of new economies in the extractive reserves in which autonomous producers are involved in communal management of forest resources, will processing initiatives significantly shift labor relations within the reserve due to an increase in wage labor? Households that invest in autonomous processing enterprises may be better able to resist proletarianization than others. As market values shift for certain products, will some households or individuals benefit unevenly because of a priori differences in resource access and control? As will be seen in Chapters 3 and 4, household production and resource management strategies in the extractive reserve are quite varied. Thus, households and individuals define different relationships with particular resources and post-harvest processing.

I argue along with other researchers that the collection, cultivation and marketing of NTFPs in extractive or agroforestry systems can be key sources of household income for gatherers and cultivators, providing certain conditions are met (Allegretti 1995; Anthony B. Anderson 1992; Anderson and loris 1992; Arguello 1995; Butler 1992; Clay 1992a, b; Clay and Clement 1993;1 Padoch and Denevan 1987;- Padoch et al. 1985;, Panayotou and Ashton 1992; Pinedo-Vasquez; et al. 1992; Plotkin and Famolare 1992; Salafsky et al. 1993;Schwartzman 1992). In the case presented here, the marketing of NTFPs strengthens ethnic or community identity and socio-political cohesion, as others have found elsewhere (Clay 1992a; Nations 1992; Reed 1995).


Research Sites and Methods

This study examines the process of socioeconomic change in two extractive reserves in southeastern Acre related to the introduction of a new technology in post-harvest processing. The reserves in this study are the Cachoeira and Chico Mendes Extractive Reserves. The Chico Mendes reserve covers 976,570 hectares (Estado do Acre 1991) and has a population of 3000-4000 families (CNS et al. 1992). The seringal "Floresta' lies within this reserve and is one of the two areas in this study. The Cachoeira Extractive Reserve covers 24,473 hectares (Estado do Acre 1991). Population estimates range from 68 families (Estado do Acre 1991) to 80 families (CNS et al. 1992).

The NTFP project that is the focus of this study is the decentralized processing of Brazil nuts. In this project, nuts are harvested, shelled and dried in the extractive reserves before being brought to the central factory in the town of Xapuri for final packaging. The project consists of community and household-level processing units in the Floresta and Cachoeira, extractive reserve areas. The project began in 1991 with the installation of three pilot sites, one each in Floresta and Cachoeira and a third in another area.

Within a political ecology framework, I focus on the farming systems of the producers in studying the relationship between the household division of labor, household budgets, decision-making, and access to and control of resources as well as community organization before and during the project activities. Field research began in September, 1991, and ran through December, 1994, with field trips of three to six months conducted each year.

I used four methodological techniques in this research project. I gathered maps and other information about the reserves and the communities from their representative


organizations to complement data I had previously obtained while conducting my master's research in some of the same seringais in Xapuri in 1988 and while working on a University of Florida agroforestry project from 1990 to 199 1. Secondly, I conducted household-level interviews before and after project initiation. Third, I conducted individual and focus group interviews with those employed by the project in both sites. Lastly, I carried out semistructured interviews with project personnel and organization leaders for project-level data and their evaluations of the project.

A stratified, random sample of households was used. Using a map of the Floresta seringal which depicts households, rivers, roads, schools and health posts, I divided the map into distinct geographical sections such that the various divisions encompass differences in physical access (proximity to roads and rivers) and services (schools, health posts, cooperative outposts). I randomly selected 48% of the households, or roughly 19 families, divided among these divisions relative to the population density. Accompanied by a community member, I traveled by foot within the seringal, conducting roughly one interview per day. A similar approach was followed in the Cachoeira reserve. Households in the reserve are anywhere from five minutes to two hours by foot on forest trails to the nearest neighbor. Within my sample, households were from two hours to a two-day walk from the city of Xapuri.

In the following chapter, I present the changing socioeconomic context of the extractive reserves in this study. I argue that the historical specificity of the reserves in Xapuri, the impact of the rubber boom in this area during the last century and the role of the state determine in part the community and household-level changes that are documented in


Chapters 3 and 4. Based on the historical framework of Chapter 2 and the trajectory of the rubber tappers' social movement, Chapter 3 ) focuses on a specific productive activity in the extractive reserves. The decentralized processing of Brazil nuts is contextualized within the extractive reserves and its socioecononiic impacts are examined within the reserve and the social movement in general. Chapter 4 explores the changing economies of households and individuals in the extractive reserves in light of the project intervention of decentralized Brazil nut processing. In this chapter, I argue that the traditional extractive economy of the reserve is being transformed and that households are becoming sites of multiple class relations and shifting gender roles as wage labor and post-harvest processing increase. Chapter 5 presents my conclusions and identifies research and organizational themes that deserve further attention.


This chapter explores the ongoing, implementation of two extractive reserves in southeastern Acre. Individuals, households and communities are transforming their economies fi7orn pure extractivism. to more diversified strategies of production which include investments in "landesque capital" (Blaikie and Brookfield 1987). Economic changes go hand-in-hand with new and modified forms of social organization. Putting the extractive reserves into practice calls for new relationships between individuals, households and other social groups. Changing demands are put on producer organizations as the expectations of what the extractive reserve would bring conflict with the reality of serving a physically and organizationally dispersed population in remote and vast forest areas. In this chapter, I explore these transforming economies and social relations in light of the extractive reserves in this corner of the Brazilian Amazon.

The questions that I examine in this chapter reflect the dynamism of the extractive reserves, The reserves were not designed to be stagnant encapsulations of extractive practices and dyadic socioeconomic relations between producers and traders. What changes are occurring as result of extractive reserve implementation? How does the organization of



production in the extractive reserves differ from earlier forms of organization? What are the implications of these differences for the state and local economy and for household and individual economies? How have social relations changed over time and what are the impacts on resources use and social organization in the reserves? This chapter examines these questions within the larger framework of Acre's historic role in the rubber boom and the Amazon frontier as well as within the context of the rubber tappers' social movement in Xapuri.

Acre and Xapuri

In this first section I present geographic and socioeconomic data on the state of Acre and the municipality of Xapufi, where the extractive reserves of this study are located.' This review places the extractive reserves and the changing economic and social character of the reserves' populations in the dynamic state and regional context of conflictive frontier processes of urbanization, environmental degradation and class formation.

The state of Acre is the westernmost state in Brazil and borders Peru and Bolivia. The state covers 152,589 km' and constitutes 3.07% of the Brazilian Amazon. The state lies between 7'N and 12' S while the capital city, Rio Branco lies at 9' S and at 160m in altitude (Estado do Acre 1991a). Acre's population totalled 417,718 in 1991, with 62% of the population living in urban areas (IBGE 1991). Average annual rainfall (measured in Rio

Acre has since created ten new municipalities, one of which, Epitaciolandia, includes the Cachoeira extractive reserve.


Branco) from 1988 to 1990 was 2230 mm and the average annual temperature was 24.70 C (Estado do Acre 1991a).

Acre has 88.24% of its territory in humid terra fit-me forest, 11.37% in varzea or seasonally inundated, lowland forest, and 0. 3 9% in pasture or plains. The state's forests are rich in rubber (Hevea brasiliensis), castanha or Brazil nut (Bertholettia excelsa) and a great diversity of timber and non-timber forest species. The distribution of rubber is statewide, but Brazil nut is only found in the southeastern region of the state which is drained by the Purus river (IAC 1991). As of 1991, 11.3% of the state was in extractive reserves, more than was dedicated to either indigenous reserves or colonization projects (Estado do Acre 1991a).

The soils of Acre are largely red-yellow podsols with some areas of cambisols and latosols. The rubber tapping communities in this study are in areas with podsols. These soils fall within the general classification of normala, which indicates that they are good to normal for annual or perennial crops but not as well suited for pasture (IMAC 199 1).

Acre's economy traditionally has relied on the extraction of valuable products from its forests. However, an industrial base that would add value to Acre's natural resources of rubber, Brazil nuts and timber has not been formed. A study of Acre's timber industry found that this sector was severely limited by weak business administrative skills, a lack of infrastructure such as all-weather roads and poor financing opportunities. Industry particularly was hindered by the lack of a forest sector policy. This, combined with the effects of regional development policy which favored expansion of agriculture and livestock, added to degradation of native forests and inhibited the practice of sound forest management (Govemno do Estado do Acre 1986). As will be seen in Chapter 3, the same impediments to


industrial development have confronted the agro-extractive cooperative of Xapuri in its attempts to add value to Brazil nuts, one of Acre's valuable forest resources which could yield significantly higher returns to the region.

The lack of investment in Acre's primary resources and the state's emphasis on promoting agriculture and livestock are not reflected, however, in the state's tax base. The bulk of Acre's 1991 tax income (92%) came from commerce and service sectors while industry provided only 4% of the state's tax base. Of the remaining 4%, rubber, timber and Brazil nuts together constituted 2.2% of Acre's 1991 tax income, while agriculture and livestock jointly contributed only 1 .9% (Estado do Acre 1991a),

The state's weak economic base is further reflected in Acre's reliance on federal subsidies and public salaries. A collaborative study by a state agency and a nongovernmental organization examined Acre's state-generated income from 1968 to 1988. Acre's average annual income from the state's own economic sources (exclusive of federal grants and subsidies) over this 20-year period was only 12.4% of the whole state budget (CNS et al. 1992b). This trend has worsened, such that in 1991, only 9.8% of Acre's state finances originated from taxes on commerce and production. The remaining 90.2% of the state government's budget came from sources such as credit programs and transfer payments (Estado do Acre 1991a). Poor management of such transfer payments, largely due to corruption, result in a lack of investment in basic infrastructure such as adequate roads, marketing and communication services. Combined with high commercial taxes and expensive, mandatory social benefits, this hinders the establishment and operations of businesses that seek to add value to Acre's forest resource base (George 1996), such as those


of the cooperative in Xapuri. In order to contextualize this difficult economic environment, I conclude this section with a brief exploration of socioeconomic changes in Xapuri.

The municipality of Xapuri was founded in 1905 and covers 8,137km2, or 5.3% of the state. The city of Xapuri lies at 100'S, at the confluence of the Xapuri and Acre rivers. Via overland routes, Xapuri lies 188 km from Rio Branco along the BR-3 17 road (Estado do Acre 1991a). In 1991, the municipality of Xapuri constituted 3.0% of Acre's population (IiIGE 1991).

In 1980, federal census officials recorded the total population of Xapuri to be 14,692, with 79% (11,570) living in the urban area and 22% (3,122) living in rural areas (Estado do Acre 1991a). Acre's 1991 population data were based on surveys carried out by the National Health Service which has a reputation for tenacity in reaching all rural households. According to this survey, the total Xapuri population in 1991 was 14,803, with 26% of the population in urban areas and 74% in the rural areas (Estado do Acre 1991 a), almost a complete inversion of the rural to urban ratio recorded in 1980 and one which would be counter to the growing trend of urbanization seen elsewhere throughout the basin (Becker 1995). The results from the 1991 federal census help to clarify these conflicting figures somewhat. In 1991, Xapuri's total population was 12,366 with 41% in the urban areas and 59% of the population in rural residences (IBGE 1991). 1 would argue that the rural population has been undercounted by the federal census and that the population trends in Xapuri are increasingly urban.'

8 Based on the number of domestic hook-ups with Eletroacre, the manager of the state's electric company facility in Xapuri suggested that the urban population in 1994 was in the range of 10, 000 people.


During last century's rubber boom, Xapuri was an important trading post. Today, many of the storefronts of the trading houses along the river still bear the names of the Lebanese traders who came to Acre durin- the rubber boom. Rubber and Brazil nuts are still delivered by boat to the city of Xapuri and carried up the steep riverbanks to be sold in these trading houses. Due to the wealth generated in Xapuri during the boom, the city was well designed and retains some of the charm from the turn of the century, with brick-paved streets lined with towering mango trees, large parks and stately townhomes and stores.

However, as was the case throughout the basin, profits from rubber were not reinvested in production but were extracted from the region through the chain of merchant capitalism dominant at the time (Weinstein 1983). Today, Xapuri still relies on rubber and Brazil nuts for its income, although the greatest source of income for the municipality comes from the federal government in the form of public salaries and grants. Xapuri's main streets are now full of deep ruts. The roads leading out to the many neighborhoods of clapboard houses that have risen up around the city are poorly maintained dirt paths that quickly turn muddy and almost impassable during the rainy season. Water and electricity are provided on a somewhat precarious schedule, relying on a diesel-powered plant.

As with the state, the economy of the municipality of Xapuri is also heavily dependent on transfer payments from federal or state sources. In 199 1, such transfers constituted 97% of Xapuri's financial statement, Xapuri's tax base that year relied heavily on livestock and commercial services, which accounted for 29% and 45% of tax income, respectively. Rubber constituted 21% of Xapuri's tax income in 1991 while Brazil nuts and timber contributed 9.4% and 2.7%, respectively. There was no recorded tax income from the sale of agricultural


goods in 1991 (Estado do Acre 1991a). As of 1991, 53% of the state was in extractive reserves (Estado do Acre 1991 a) which could serve as a source for additional income from forest products.

The Rubber Boom in Acre

The fabled richness of Acre's forests made the region a target for the rubber barons during the nineteenth century (Tocantins 1979). During the rubber boom, which lasted from the mid-1I880s to the early 1920s, rubber barons or bosses recruited laborers into the Amazon to work as rubber tappers. Most rubber tappers came from the Brazilian Northeast, an area which experienced a crippling drought in 1877, thus leaving many farmers desperate for work and vulnerable to the tales of riches to be made in the Amazon.' Many of the rubber tappers in this study are descendants of these nordestinos or Northeasterner.

At the outset of their journey to the Amnazon, workers were advanced dry goods and their passage was covered by the boss, an initial debt from which many would never be free. The rubber tappers formed the lowest rung of the debt-peonage system known as aviamento, a vertical chain of merchants and middlemen linking the rubber tapper in the forest to the large shipping houses of Bel~m and Manaus, After an often long and tiring journey, rubber tappers arrived at a small port village and were supplied with enough dry goods for several months. They were then taken further upstreamn to the forest tract or seringal where they would tap rubber. They were shown how to cut trails in the forest to access the 300-500

See Weinstein (1983); Dean (1987), and Santos (1980) for complete histories of the Amazon rubber boom.


rubber trees they would tap. Once a rudimentary shelter was constructed and the tappers had the basic skills needed to provide the boss with a good harvest of rubber at the end of the season, they were left alone in the forest for months at a time. Many died from malaria, snake bites or other maladies.

The tales of easy riches to be gained in the forest for the rubber tappers proved to be false as the accounts were tallied up in the boss' book, always showing a debt owed to the boss. During the rubber boom (1850-1910), the aviamento system ensured that the bosses maintained effective control over the rubber tappers' labor. The rubber tapper was kept constantly in debt to the boss through the latter's manipulation of the accounting books and the exorbitant prices charged for dry goods in exchange for rubber. Rubber tappers were forced to purchase all of their dry goods through their boss and were prohibited from selling rubber to another buyer. Retribution for selling rubber to another boss was swift and violent; at times the offending rubber tapper was killed by tying balls of rubber around his neck and then setting them on fire. Thus, the boss maintained effective control over the rubber tappers through economic and non-economic means."'

Women were scarce in theseringal (Santos 1980), especially in the upper reaches of the Amazon basin (da Cunha 1986). Indigenous women and children were often kept in brutal slave conditions as prostitutes or rubber workers (Simonian 1995). The few nonindigenous women who did live in the seringal were the wives of the political and economic elites (Tocantins 1979).

For an excellent, first-hand account of life as a rubber tapper in Acre during the turn of the century, see Yungjohann (1989).


Those controlling the latex extraction industry preferred to recruit single males from northeastern Brazil as laborers. Very few rubber tappers brought, or were allowed to bring, their wives and families with them (Reis 1953). As one observer put it, "[T]he seringueiro [rubber tapper] was a machine to produce rubber. No women dared accompany him to the desolate forest. Women would be dangerous for social stability, . .objects so rare would cause envy and result in covetous crimes of passion" (Tocantins 1979, p. 166).1 Women's presence would also undermine the bosses' control-, by contributing their labor and services in subsistence agriculture and health care they would decrease the rubber tapper's purchases of the bosses' dry goods and medicines (UNIFEM 1990).

Rubber tappers in more isolated regions such as Acre requested that their bosses supply them with women. In response, the large commercial trading houses in Bel~m and Manaus began sending women to the .seringai. They were delivered in much the same way as dry goods and utensils -- at inflated prices which were charged on the seringueirs account in the boss's store. In one instance, the Governor of the State of Amazonas ordered the Manaus police to round up 150 women from the city's brothels and cabarets. These women were then shipped to and distributed in the Acrean city of Cruzeiro do Sul (Reis 1953). One immigrant who arrived in Amazonas in 1942 recalled how his uncle had made his fortune by taking advantage of the scarcity of women and selling his five sisters to local men (Benchimol 1992).

Ligia Simonian also notes that women were seen as the cause of high crime rates in the seringal (Simonian 1995).


In 1910, the Brazilian monopoly of the rubber market was broken when Malaysian rubber carne on the market. As the price of rubber fell, so did the bosses' means of control over the rubber tappers. The aviamento system weakened to the point that the bosses were unable to get the credit and dry goods they needed to supply and control their labor force. The rubber tappers were thus allowed to hunt, fish and -plant their own crops. In addition, rubber tappers often sent back home for their wives and families, Others married local women and began to raise families in the seringal,

This dramatically changed the rubber tapping production system. With the inclusion of female and child labor, the household was able to diversify its production for subsistence and marketing. Hunting, fishing, Brazil nut gathering and extraction of other forest products became part of the household econorny. The household gained a certain degree of autonomy in decision-making regarding labor allocation and livelihood strategies. However, most tappers were still in the vestiges of the debt-peonage system with no marketing opportunities other than itinerant traders or patrons who still were able to maintain economic control of the rubber tapper household.

During World War II, Japanese forces cut off valuable supply routes for rubber from Malaysia, leaving the Allied forces to seek rubber from the Ainazon for the war effort. With financial backing from the U.S., credit sources Were created to rejuvenate many rubber estates that had been inactive and rubber bosses enjoyed a resurgence of power, subsidized by generous marketing assistance (Dean 1987; Santos 1980). Workers were recruited from the northeast of Brazil and were classified as soldiers or soldados da borracha.


With the decline in prices after the war, a similar process to what occurred after the rubber boom took place. Many of the former rubber bosses abandoned their seringal. Those bosses who did remain were able to keep the rubber tappers in debt by the practice of manipulating the books and prices in their favor. Other rubber tappers were left vulnerable to itinerant traders or marreteiros who served as a patron or boss (Dean 1987; Martinello 1988; Santos 1980). Many of the soldados dci bortracha stayed on in Acre and continue to live and work in the seringal (Martinello 1988). 12

The Coloca9Clo: A Changing Socioeconomic Space in the Forest

During the days of the rubber boom in the last century, the forest tract or estate controlled by a boss or patron was referred to as a seringal, in reference to the local name for the Hevea brasiliensis rubber tree, sering-uehia. The value of a seringal depended on the number of rubber trails it encompassed (Tocantins 1979). Since a seringal designated an area of social control and not necessarily a standard geographic area, there was considerable variation in size from one seringal to another. Today's extractive reserves might encompass anywhere from one seringal, as is the case with the Cachoeira reserve, to 91 seringais in the Chico Mendes reserve (CNS 1992; Estado do Acre 1991b).

Within each seringal are numerous smaller holdings which originally designated the area to be worked by the various rubber tappers or seringueiros under the control of the boss. Each of these smaller holdings is a colocaqdo. This term comes from the verb colocar,

12 Recent legislation awarded these soldados a monthly pension which is collected by several men or their widows in this study.


meaning to place or to put. The coiocaqdco refers to the area where a rubber tapper was placed or colocado in the forest by the boss during the rubber boom. Both terms, co/ocaqlo and seringal are the standard names for rubber-tapping forest tracts today.

The reference for the size of a coiocaqdo is the number of rubber trails that it has. Acre's earliest colocaqbes during the rubber boom of the last century consisted of one to six men whose sole occupation was the extraction of latex."3 The only infrastructure on a colocagdo was a small hut for sleeping and a covered shed for smoking and processing the latex (Yungjohann 1989). Rubber tappers were not permitted to spend time with subsistence agriculture, hunting or fishing because this would detract from the boss' profit margin (Tocantins 1979; Weinstein 1983).

The colocaqdo of the rubber boom was designed to meet the demands of a oneproduct economy of pure extractivism dominated by merchant capital. Today's colocagdo combines a variety of subsistence and market production strategies to meet the needs of autonomous households integrated with the capitalist market. Most households in the extractive reserve still tap rubber and also gather Brazil nuts as their primary sources of cash income. Households collect other products from the forest including various fruits, honey and housing materials. They also engage in hunting and fishing for subsistence needs. Annual crops are grown in small areas of slash-and-burn agriculture and are used mostly for home consumption with some sales of surplus harvests. Small livestock such as chickens

13 One rubber tapper recounted that he and his companions opened a with 14 trails for seven men, six of whom died before the first season was over (Yungjohann 1989).


and pigs are raised near the house where some fruit trees might also be planted. Some households have a few head of cattle and perhaps a horse or mule for transportation.

The main building in a colocaydio is the house. Other infrastructure might include a storage shed, pens for small livestock, small vegetable gardens or a casa de farinha, a covered area for milling and toasting manioc flour. Some households have added value to their colocaqdo with the installation of a mini-factory for drying, shelling and roasting Brazil nuts. Thus, the diversity of livelihood strategies, gender roles and market integration of a typical colocaqtio today is quite distinct and more complex compared to the days when rubber held a monopoly on productive activities and single men lived in isolated huts in the forest.

Although the economic dominance of rubber is waning, it remains the principal factor upon which land distribution is defined. Boundaries and land ownership in the extractive reserve today still are based on the location of the rubber trails in a coloca~do.'" Accordingly, the large majority of the households that I surveyed did not know the size of their in hectares. In 1991, responses from five households indicated an average holding of 450 hectares (range: 200-900). In 1991, four households reported their holding sizes, which averaged 840 hectares (range: 600-1200)."1 The average colocaqdo in Cachoeira. has been

14 The average number of rubber trails per household in my sample was five (range: 0-12, N=27).

15 Changes in the reported size of holdings were due to some internal migration within my sample and the impact of a socioeconomic survey team that had been working in the reserve in 1993. Part of their survey involved estimating the size of the colocaqes, for which they used a calculation of 100 hectares per rubber trail. Several families recalled that calculation when responding to my question about the size of their holding in 1994.


estimated to be 370 hectares (Estado do Acre 1991), while researchers found a range of holdings in the Porongaba seringal (in the Chico Mendes reserve) of between 200-600 hectares (Nepstad et al. 1992).

Land Use Strategies in a Coloca-cdo

While a typical colocaqdo is mostly mature forest, there are usually several cleared areas within the family's holding. Where a family builds their house, storage shed, pens for small animals or a small garden area is one such area. Houses typically have a kitchen with a wood-fired, clay stove, a front room and one or two rooms for sleeping. Houses are built several feet off the ground and usually are constructed from local palm wood of paxizba (Socratea exhorrhiza) for the floor and walls and the roof is thatched with ouricuri (Attalea excelsa). Some households that own a chainsaw or can afford to pay for hired help with a saw might have a house constructed of sawn timber. The preferred timber for construction in the seringal is from the Brazil nut tree, the wood of which can legally be used for housing in the reserve but the sale of which is prohibited. The house is located near a creek to facilitate access to water which must be hauled up to the house for cooking. Bathing and clothes-washing are done at the creek. None of the households in my survey had a well.

Small areas of mature forest are cut and burned each year for planting annual crops of rice, beans, corn and manioc. Few households own chainsaws or have the means to rent the use of one. The arduous labor involved in clearing and burning combines with poor market returns on annual crops to keep the size of cleared areas low. By allowing these agricultural fields to lie fallow and then using them again for annual crops, some long-standing


households have been able to meet their cropping needs without having to clear primary forest for several years. In 1991, the average size of the agricultural plots was 1.01 hectares (range: 0.25 to 2.00 ha; N=27). In 1994, the size of agricultural plots averaged 0.90 hectares (range:

0.25 to 2.00; N=23).

.Fall ow/Secondary
Annual Crops

Figure 2.1 Relative Sizes of Altered Forest Areas on 28 Holdings in 1994 (ha)

Fallow agricultural fields and other areas of secondary growth that are more advanced in the process of forest regeneration constitute a third form of clearing in a coloca~do. Some such areas may have fruit trees which are still harvested. Others may be older fallow areas that have reverted to scrub forest. The average size of these areas in 1991 was 2.99 hectares (range: 0.0-8.0; N=16). In 1994, the average size was 1.39 hectares (range: 0.5-6.00; N=22). A paired samples t-test found a statistically significant difference between the size of these areas in the two sampling years (t-value=2.50; dfr 13;- p=.02'7). This difference could be due to the fallow cycle as households take secondary forest areas out of fallow and re-cut and


bum them for a cycle of annual crops. Quite a few of the households in my survey reported that they had made use of secondary forest areas and had not cut mature forest for many years.

The fourth type of cleared area in a coloca(:do is pasture for cattle, horses or mules. The average size of pasture per household in 1991 was 2.31 hectares (range: 0.0-15.0,; N=23). In 1994, the average pasture was 2.92 per household (range: 0.0- 17.5; N=22). There was no significant difference in the mean pasture areas between these two years according to a paired samples t-test (t-value=-.63;, df=-18;- p=.538). The average difference in pasture between 1991 and 1994 was 0.38 hectares (range: -7.0 to 7.0; N=19), indicating that the average pasture area per colocado is not increasing much.

Combining these three latter forms of clearings (agricultural fields, fallow or secondary forest areas and pasture) resulted in an average total altered forest area per household of 5.18 hectares in 1991 (range: .75-15.75;- N=25) and an average of 4.89 hectares in 1994 (range: 1.00-17.75; N=23).

Z Forest
Cleared Areas


Figure 2.3 Land Use Areas in a Typical 500 ha Forest Holding (1994)

An average of 5 ha of cleared areas and/or secondary forest per household in my sample indicates that the percentage of land within the coloca~do that has been deforested is very small. Taking an average colocaydo size of 400 hectares for my sample (which is compatible with the Porongaba and Cachoeira measurements) indicates that less than 2% of the land is cleared.

Regulation of Land Use in the Extractive Reserve

In this section I examine the relationship between these clearing rates, the householdlevel decision-making processes involved and the rules defined by the residents' association which were designed to maintain existing forest cover in the extractive reserves. Before the advent of the extractive reserve management plans, each household decided independently how large each of these clearings would be, based on available labor, consumption needs and livestock. Unwritten rules about the size and location of forest clearing combined with limited resources to keep such areas small in relation to the primary forest tract controlled by each household. Rubber tappers avoided locating their agricultural fields and pasture in areas where there were rubber or Brazil nut trees. Some households also avoided clearing forest on the edge of creekbeds. Vigilance by neighboring families and community pressure usually was sufficient to keep families from breaking these tacit understandings about land use in the seringal.


There were cases, however, of small scale ranchers or of rubber tapper households who cleared larger tracts of forest for pasture. Over the years that I have visited these seringais, I have seen and heard of numerous instances of families who cleared and burned right up to both sides of the creekbank, even when this meant that their own water supply was severely limited or contaminated as a result. There were also cases of families who regularly felled rubber and/or Brazil nut trees as they cleared forest for their agricultural fields. Grumblings from neighboring families were heard, and the dissatisfaction went so far as to exclude certain households from community activities in some cases. There were no institutionalized means by which rubber tappers could either enforce these commonly understood land use rules or have such rules brought to bear by an outside regulator.

These unwritten rules have now become written in the management plans of each extractive reserve. Previously tacit understandings about the duty of rubber tappers to preserve the forest became legally binding regulations. The autonomy of a household's decision-making as to the size and location of forest clearings has changed. Each extractive reserve must have a management plan, jointly agreed upon by the residents' association, the CNS and IBAMA (Instituto Brasileiro do Melo Ambiente or the Brazilian Environmental Institute), which details the regulations and enforcement procedures to guarantee the selfsustainability of the reserve. Each resident household must sign an agreement with the federal government, agreeing to abide by such land use rules or face a fine and eventual expulsion from the reserve if infringements continue to occur.

For the Chico Mendes extractive reserve, the management plan was in draft form at the end of my research trip (CNS 1994). The plan outlines commonly-agreed upon land use


regulations for the reserve including the amount of land within the colocaqdo that can be dedicated to "complementary activities" such as agriculture and livestock. Rules also apply to limit the maximum size of these areas relative to the whole colocaVcto. Extractive techniques and harvesting schedules, the use of timber, fishing practices, and other management issues are addressed in the plan.

In 1994, 1 asked the families in my survey sample who was responsible for enforcement of these regulations and general oversight of the reserves. Of the 18 responses, I I households (6 1 %) responded with the correct answer: that IB AMA was responsible for enforcing reserve regulations. Of the Femaining seven households, five (28%) did not know who the regulatory agency for the extractive reserves was and two households (I I%) thought that the community was to be the enforcing institution. The designation of a regulation agency does not translate into effective oversight. Antonio, a former union delegate lamented that, "...the problem is that it's not well-defined [who is supposed to enforce the limits and how]. It's not going to work with rubber tappers doing the monitoring. It has to be IBAMA but they're so corrupt."

Each household is permitted to use up to 10% of their colocaqdo for activities complementary to extraction, such as agriculture, livestock or fish ponds. The raising of large animals must be restricted to a maximum of 50% of the area dedicated to such complementary activities. Areas dedicated to agriculture, pasture and secondary growth may not exceed 5% of the total area of the colocaqiio. Although it was not written in the draft use plan, officials in the union and the CNS informed me that the maximum amount of mature forest (mata bruta) that each household may clear per year is one hectare. Up to two


hectares of fallow or secondary growth (capoeira) may be cleared per year. Thus the maximum clearing limit is a combined total of 3ha/yr.

When I asked the surveyed families whether there was a limit on the amount of forest that could be cleared each year in the reserve, 18 of the 20 households (90%) indicated that a limit existed. I then asked what the clearing limits were. Only five of the 18 responses (18%) were correct. The responses of the other 13 households ranged from a clearing limit of 16 ha/year (more than five times the permitted area) to various combinations of limits on the size of primary and secondary forest areas that could be cleared.

To see if the forest clearing rates of the households in my sample fell within the regulated limits of the extractive reserve, I examined the differences in total cleared areas between 1991 and 1994. We saw earlier that the combined areas of agricultural fields, fallow and/or secondary forest and pasture averaged 5.18 hectares in 1991 and 4.89 hectares in 1994. 1 calculated the average difference in clearing over these three ye ars to be -0.58 hectares. Responses ranged from a low of -8.75 hectares (indicating a decrease in cleared or secondary areas) to a high of 8.50 hectares.

Taking the highest annual clearing rate for 1991 to 1994 in my sample of 8.5 hectares as the worst possible case in terms of forest clearance and dividing it by three, yields an average of 2.8 hectares cleared per year by this household. Comparing this clearing rate of 2.8 ha/yr with the extractive reserve limit of 3 ha/yr indicates that overall households in my sample fall below the prescribed forest clearance regulations. Considering that the average clearing rate for all the households in the sample was -0. 5 8 ha over three years (or -. 19 halyr),


the extractive reserve regulations on forest clearing appear to be quite generous in terms of the surveyed households' current clearing practices."

If this clearing limit is enforced, it effectively puts a hold on additional pasture formation in the colocaq5es. "This limit is okay for planting but not for pasture", said Jose, a reserve resident. Many interviews were full of comments regarding the inability of those households that do not already have some pasture in their coloca do to raise additional cattle. Those residents who already had a large area of pasture can now either continue to raise roughly the same amount of cattle they currently have, or rent out their pasture to neighbors, a common practice in the reserves.

When I asked the families in the survey whose job it was to monitor these clearing limits, there were mixed responses. Of the 18 families that responded, nine (50%) gave the correct response, that reserve residents were in charge of monitoring forest clearing. The next most frequent response was given by four households (22% of the sample) who said that no one was responsible for enforcement. Two families (11%) thought that the union was in charge of monitoring forest clearing, while another two families said that they did not know whose job this was. The remaining household (6% of the sample) responded that enforcement was to be done by community agents who had been recently hired to work on

development project funded by the Austrian (Yovernment.

In choosing my sample in order to study the impact of the Brazil nut project, I stratified the population in the seringcd and then randomly selected rubber tapping households. Small ranchers and traders, one of whom has 100 head of cattle, were thus excluded from my sample, such that existing areas of significant clearing in the reserves are not calculated here.


The management plan for the reserve charges each rubber tapper with monitoring his holding and those of other rubber tappers to ensure that the regulations of the plan are observed. This enforcement plan assumes that rubber tappers will be willing and able to monitor and report breaches of the forest clearing limit by nearby households who may be long-standing neighbors, relatives, or, in the case of those involved in the Brazil nut project, supervisors or employees. Such an assumption was regarded favorably by some of the people that I spoke with in the reserves, but many others expressed their concerns about neighboring rubber tappers monitoring forest clearing. An older man told me, "It doesn't work to have one rubber tapper monitoring the other. There's a lot of people here who don't understand the rules. It could create fights; it has to be someone from outside [the reserve to enforce clearing limits]."

There was general agreement that only a few reserve residents were disobeying the rules. The union delegate reported, "[o]nly a minority don't obey the clearing rules ... Some of them have dreams of being ranchers". People's perceptions of infringements were, of course, based on their own interpretation of the clearing limit. One man who thought people were allowed to clear only one hectare/year said, "A lot of rubber tappers don't respect the limit. This year a lot of forest fell," There was also a consensus on the lack of orientation and education for the reserve residents. Altercations between neighbors could easily arise if someone who misunderstood the regulations were to confront another reserve resident about what he or she considered to be an infringement. Several people noted that others disobey the limits, but that this was due more to a lack of awareness than an intentional disrespect of the extractive reserve rules.


The Seringal: From Rubber Estate to Community

During the rubber boom and through the middle of this century, life in the seringal was one of physical isolation because of the distribution of rubber trees in the forest and of social isolation because of the organization of production and marketing. Tocantins notes that there has been a historical social division in the seringal which distinguishes between two social groups, the rubber tappers and the boss with his associates. Those affiliated with the boss included the manager, assistant, accountant, hunters and fishermen, boat operators, field hands for pasture and livestock management. The other group was the rubber tappers who worked in the "siberian isolation of the colocaqdo, without social interaction, just him and the forest, him and the rubber tree". Those who lived and worked at the barracto, or the boss' trading post had daily contact with other employees. In contrast, the rubber tapper's life was one of social isolation (Tocantins 1979).

These social groupings gave rise to distinct social spaces within the seringal. As recorded by Mary Allegretti, these social spaces still exist in some areas. The margem, or margin, is usually on the bank of a river and represents the hub of activity in the seringal. Here, rubber and dry goods exchange hands at the barracao (Allegretti 1979). An especially well-to-do boss might even construct a simple chapel near his house for use by the occasional traveling priest (Tocantins 1979a). In contrast to the margem, the centro, or center, is the area of the rubber tappers. Deeper in the forest, isolated from social interaction as dictated by the location of the rubber trees, the rubber tappers live and work (Allegretti 1979).

In the 1970s, as roads penetrated areas that had previously been accessible only by boat, the predominance of the nhargein over the centro was reversed at the whim of


mapmakers. Along the BR-317 road which runs from Rio Branco to Xapuri, an older man who used to tap rubber now sells soft drinks and beer to travellers who pass by his door. He says with a laugh, "This coloca2ao used to be called 'Fim do Mundo', or 'The End of the Earth' because its a five hour walk to the edge of the river where the barracdo was. Now we call this place '0 Centro do Mundo', or 'The Center of the World'!".

The scope of space and resource use for the rubber tapper during the boom was limited to extraction and exchange. Rubber tappers' space was restricted to the rubber trails and the connections between his colocaiio in the centro and the barracdo of the boss at the margem. Today's rubber tapper households have a much broader range of social and geographic spaces which encompass a much more diversified production and natural resource management strategy and which are integrated into a more complex economic and sociopolitical system.

Today, even though the household-level coloca 5es remain physically isolated, there are multiple forms of organization and social relations within the seringal and surrounding areas. These have emerged hand-in-hand with changes in the organization of physical spaces in the seringal. Although the household-level colocagdo is the principal land area, some households have chosen to work together to create community spaces. There is an increasing prevalence of common-use areas in the seringal such as soccer fields, communal agroforestry plots or fish ponds, schools, health posts and CAEX supply posts. Collectively, these represent a changing social and spatial organization in the extractive reserves.

The base communities of the Catholic church and the union movement, which grew in response to the roads and accompanying land conflicts, were responsible for breaking the


traditional socio-spatial organization of isolated households in the seringal. In some seringais, such as those in this study, the dichotomy of centro/nargem and boss/rubber tapper has been replaced by a more decentralized production system of autonomous households. The dyadic ties between the boss and the rubber tapper have been replaced with a multi-stranded web of social, economic and political ties within the seringal and between rubber tapper households and their representative organizations, the union, the CAEX and the CNS. I argue that, as the extractive reserve plans are further implemented, there will be a continued emergence of community areas within the reserves. While the concept of "community" differs from a more urbanized one of concentrated population, reserve residents refer to certain areas within a seringal (or the whole seringal itself in the case of a smaller tract) as comunidade Sao Francisco or a coninidade Floresta.

Due to the physical separation of households in the seringal, communities as such are numerically small. "Communities" in the reserves may consist of five to 25 households that meet fairly regularly, either through the Sunday morning base community of the church, union meetings with the local delegate or to clear pasture area for the CAEX's transport mules. The community also serves as an address and as a social identity. Many times in union assemblies in the city of Xapuri, rubber tappers would refer to another group by the name of their seringal or the central colocaydo in that area. These central colocaq&es oftentimes are the location of the school, health post and CAEX supply store.

The communities that are being formed within the extractive reserves and the larger "community" or movement of rubber tappers in Xapuri are not without internal conflict, mistrust or unneighborly disputes. As the management plans for the extractive reserves are


implemented, and as the CAEX and union amplify their services and activities, I postulate that there will emerge a variation of such communities within the reserves. This process will have varying results, depending on the history of socio-political mobilization, different levels of participation in the Catholic church's base communities, or perhaps the number and strength of intra- or inter-familial ties in the area. These social variations, combined with structural factors such as the variation in resource access, transportation and physical isolation will most likely result in a mosaic of heterogeneous communities throughout the reserves with some being much more active in the movement's institutions and others being less so. In the final section of this chapter, I discuss shifting social and gender roles and the difficulties faced by these organizations in meeting the changing needs of extractive reserve residents.

Creating New Social Roles and Spaces

Gender roles and the sexual division of labor have changed dramatically from the days of single men living in isolated huts in the forest. This study pays particular attention to gender as a critical variable in detennining access to and control over resources, the dynamics of which will determine in part the viability of extractive reserves. Awareness of genderdifferentiated roles in livelihood strategies is crucial to understanding the complex dynamics of household resource management (Agarwal 1994; Deere 1990, 1995; Poats et al. 1988; Sage 1993). In the extractive reserves, knowing who has access to and control over certain resources or particular physical spaces sliould be considered when drawing up management plans which must be subscribed to by the resident population. These variables of access and control vary significantly between and within households. Such variation in access and


control is a key determinant of resource management within households. Along with this variation in productive aspects, it is also critical that those involved in such projects, from participating households to donors, understand the value and fluidity of traditional social roles in the households and communities of the.svringal.

Lack of recognition of these social roles can be devastating to extractive reserve projects and those involved. One such project involved hiring reserve residents as community agents. As dictated by the donor agency, pr Ject staff hired men and women to work together on teams. The job of these men and women was to travel to different seringais within the reserves and encourage comi-nLinity organization through meetings and household visits. Antonia, a mother of five and an active participant in her church group, was hired by this project. She traveled with the three men on her team to neighboring, areas and, according to project staff, carried out herjob well in spite of the difficulties of being the only woman in a very challenging situation. Upon returning home, her participation in the project exacerbated existing difficulties between Antonia and her husband. She ended up taking several of the younger children and leaving, him. In a later discussion with a leader of the CAEX, he used Antonia's case as an example.

Once she earned her own money, she started making demands and made him jealous because of what she had earned. She wanted to live her life and it didn't work out and
now they're separated.

Hiring men and women to make overnight trips together within the.svringal reflected a lack of sensitivity and awareness of social roles on the part of project staff. As a sole women working with three men, Antonia's trips with her team were inappropriate in the eyes of the


community and her husband. The fall-out from this experience could well be an unwillingness on the part of union and CAEX organizers to involve women as leaders in future projects. Women may also refrain (or be withheld by their husbands) from participating in other community efforts.

As will be seen in Chapter 3, the Brazil nut project that is the focus of this study provides socially acceptable opportunities for women to work in different social arrangements and to gain new skills. Women and men work together in community-level factories, one of which is managed by a young woman. As extractive reserve residents work together to define new production strategies and forms of community organization, shifts in traditional gender roles are bound to occur. However, these shifts should not be externally imposed, albeit via well-intentioned directives from donor auencies.

As I have discussed elsewhere, gender roles in the rubber tappers movement are in a process of re-definition as individuals, households and communities invest themselves in defLning and managing the extractive reserves (Campbell forthcoming). In some families, the male head of household maintains the traditional role of holding sole responsibility for marketing decisions and transactions. In other cases, the oldest son may be the family's political representative in the union and the cooperative. Women who work in the family's agricultural fields and in domestic tasks near the home may be perceived as making less of a contribution to the household, thus negatively affecting perceptions of what women are entitled to and decreasing their bargaining or breakdown position (Jaquette 1993; Sen 1990). The work of those household members who engage in activities which bring in visible cash income, such as rubber and Brazil nuts, may be considered more valuable to the household's


livelihood. Such valuations can affect decisions as to who should attend school, who should be allowed to participate in the community meeting or the church gathering, who should join the union and vote in the upcoming elections or who should work in the Brazil nut project.

In Chapter 4, it will be shown that the traditional role of the male head of household as the earner and controller of the family's cash income is changing dramatically. In a few households involved in the Brazil nut project, the wife now earns more than her husband. Several women in this study who have turned 55 years of age now receive an agricultural worker's retirement package from the state which is worth more than the household's other sources of income. As a project organizer said of the Brazil nut project and other initiatives in the extractive reserve plan,

..the introduction of producer groups and other projects such as the Brazil nut
processing affects the basic economic and social structure of seringueiro households.
It changes the relationship between men and women. With these economic changes, if there isn't discussion in the family and in the community before the project begins, it might happen that [the man would say], 'Now, she earns more than me! Who's in charge here?'. This generates conflict. The extractive reserve concept proposes to liberate and give value to everyone, including women .... The projects might liberate women but you'll have these conflicts. If you mess with the family structure and don't encourage or amplify this discussion, you'll have conflict. The survival of the extractive reserves depends on these discussions. You have to work on the day-to-day
issues. 17
In this study's examination of changing livelihood and decision-making strategies, it will be shown that these issues of access and control are central to the well-being and survival of households in the extractive reserve.

Organizing Effectively to Meet Chan(,zing,, Needs

'~Field interview with the Project Coordinator, Rede Acreana de Mulheres e Homens. October 25, 1994;1 Rio Branco, Acre.


I close this chapter with an examination of the social organizations that the rubber tappers created as part of their social movement. This discussion focuses on the ability and the potential of the rural workers' union and the CAEX to meet the needs of the extractive reserve residents. Whether or not the union and the CAEX can complement this process over the long-term could determine the viability of the reserves. The transformed social and economic relations in the seringal rnean that the reserve residents increasingly rely on these organizations for cooperative marketing, political representation and emergency assistance-services that traditionally were provided by the boss. Here I present the opinions of surveyed households regarding the performance of the union and the CAEX in light of the social and economic transformations within the seringal.

The Union of Rural Workers in Xapun had a total membership of 4,264 by the end of September 1994. Of these, 233, or 5.5% were women. Union officials and reserve residents all concurred in noting that the total number of members was not indicative of active participation. Of the over 4,000 members whose signatures appeared on the registration book, over 1,000 of these individuals either died or are no longer active, dues-paying members. The clerk in the union office estimated that there were only approximately 150 members who regularly pay their monthly dues, which were US $0.85 in September 1994."

This represents a vicious circle for the union and its membership. Without general operating funds, elected union officials (who receive no salary) have no means to make phone

A former union president told rne that this has been a perennial problem. Even in the union's heyday when Chico Mendes was president, there were over 2,000 "members" as reported in the press and promoted by union leaders. However, less than 10% of these members were actively paying monthly dues at the time.


calls or write proposals for project funding." They are then easy targets for complaints by union members who told me that the elected leaders, "spend all their time in town and don't come out to the areas (seringais) anymore". In 1994, when I asked union members if they were satisfied with the union's operations, 55% of the 20 responses were negative. The most telling sign of this dissatisfaction came from comparing responses from 1991 and 1994 to my question about the benefits provided by the union. In 1991, 29% of the respondents indicated that union membership brought thern no benefits. By 1994, this response had doubled to 60%.

Evaluations of the union today and how it might be in the future were mixed. When I asked the families what they saw for the future of the union, their answers were more optimistic. Of 18 responses, 67% indicated that the union would improve and continue to grow. Those who were less optimistic consisted of 22% of the households which expected the union to cease functioning. The remaining 11% responded that they weren't sure what the union would be like in the future.

Opinions were also mixed about the opportunities for women to play a more active and influential role in the union. Of 20 union member households that I surveyed in the reserve, 65% thought that women would be increasingly active and influential in the union. Of these same households, 70% responded that a woman could be elected president of the union within the next five years. However, of the 25% who thought it would be impossible, there were several women who were very active in the women's group of the union. They

While I was working in the union office with the clerk, the lights went out because the union had not paid the electric bill for several months.


indicated that no women had the opportunity to gain the necessary leadership skills and that machista attitudes toward women in the seringal and in the union would prevent such an


Some households credited the union with the existence of the extractive reserves and services such as the schools, health posts and the CAEX. The general tone, however, was that the union was less and less able to i-neet the demands of its membership. Tl- s may well be due to changes in union membership. More young people and women have joined the union in recent years. Table 2.1 shows the changes in union membership of the households that I surveyed in 1991 and 1994. Increasing numbers of women and young men joined the union over the period of my survey according to field interviews and union records.

Table 2.1 Frequency of Union Membership of Selected Groups of Surveyed Reserve

Male Heads of Female Heads of Sons over 18 years

Household Household of age
85% 3 2% 46%
(22 of 26) (9 of 28) (5 of 11)

86% 3 7% 64%
1994 (18 of 21) (8 of 22) (7 of 11)

21 See Campbell (forthcoming) for more on the changing role of women in the rubber tappers' movement.


Union records show that, prior to 199 1, single young women comprised just 26% of the female membership. From 1991 to 1994, this group accounted for over 50% of women members. This is an important shift, considering that 70% of the first women to join the union were widows and were the sole political representative of the household.

Today, with more single and married women and more young men in the union, the tights and responsibilities of political representation are shared with husbands or fathers. The increasing dissonance between members' opinions of union operations and the ability of elected leaders to meet members' expectations is likely due to this shifting membership composition and could be attributed in part to generational differences in attitudes and ideas about social organization.

A very important benefit that can be credited to the union's work is the increasing access to state services of the households that I surveyed. Some households acknowledged the union's assistance in filling out paperwork and educating them about their rights as rural workers. As will be seen in Chapter 4, many households in the reserve now receive either agricultural workers' retirement payments, pensions for service by a soldado da borracha during World War II, or other payments from the federal health fund for those with physical disabilities. The role of the union in educating members about their rights and access to these services is often not recognized but is crucial to many household economies in the reserve.

The other organization in the rubber tappers' movement that will be discussed here is the CAEX. The next chapter examines one of the CAEX's project activities and presents more general information on CAEX's operations. Here, I discuss CAEX in relation to the social and economic transformations within the extractive reserves.


One of the main challenges facing the CAEX in regards to its membership is the ideological shift from a dyadic, patron-client relationship to a collective economy comprised of autonomous households. In the process of constructing communities within this transformed economy, households in the reserve are creating a web of inter-household social and economic relations which is tied to the CAEX Those households which have not joined in this process and are not members of the CAEX, are forced to depend on increasingly unreliable local traders. As will be shown in Chapter 4, households that are not CAEX members had significantly lower income levels and weaker economic outlooks than those who have invested in the process of constructing a community within a transformed, collective economy.

In 1991, 3 1% of the men over 18 years of age in my survey sample were CAEX members. By 1994, this had grown to 44% of the adult men in my survey. As of December 1994, there were 251 CAEX members listed in the registration book. There are approximately 4,000 families in the Chico Mendes reserve alone. The task of providing market services to even a small fraction of this many households is daunting. Many people mn my survey noted how fortunate they were to have the CAEX. They often remarked on the difficulties faced by families who lived more in the centro of the seringal where the CAEX had not yet established supply posts or regular transportation. Some of these households are several day's walk from the city of Xapuri via forest trails. A study which included marketing channels within the seringal found that 60% of the households in the Chico Mendes reserve relied on marreteiros, or travelling middlemen, for their dry goods supplies. This differs


dramatically from the situation in the Cachoeira reserve, which is much smaller, wherein 79% of the residents use the CAEX as their main supplier (CNS et. at, 1992).

When I asked CAEX members about their opinions of CAEX prices, both for buying rubber and Brazil nuts and for selling diy goods, all of the responses indicated that CAEX pays more than local traders for rubber and Brazil nuts. The responses on the relative prices of dry goods were mixed. All in all, members were satisfied with the economic performance of the CAEX On the administrative side, however, there were numerous complaints. Concerns included the large number of employees at the central office who were seen as doing very little work, the constant travelling by the CAEX president and the inattention of other elected leaders to the day-to-day operations of the cooperative. Specifically related to the Brazil nut project which is highlighted in the next chapter, CAEX was late in issuing paychecks due to a lack of operating capital. An engineer from ECOTEC was responsible for project management and received very high praise from those working on the project. Households participating in the Brazil nut project expressed concern about the future of project management after the scheduled departure of this engineer from the CAEX Despite these complaints, members expressed overall satisfaction with the CAEX and expected it to grow in terms of membership and to improve in coming years.

CAEX membership tends to be limited to male heads of household. There were no cases of any households having more than one CAEX member and there were no cases of any female heads of household members in my sample. Even in the case of a female-headed household, she had transferred her membership into her son's name. This seems to be on the


verge of changing, however, as more women gain the economic means to pay the initial membership dues and the social confidence to Join the CAEX (Campbell, forthcoming).

Even though some households have not joined the CAEX, this does not signifyr an unwillingness to do so. Initial dues to join the cooperative are 50kg of rubber, or the equivalent in another product. At the end of 1994, this amounted to roughly US$45.00. As shown in Table 2.2, this payment has kept some female heads of household from joining. Overall, 39% of those interviewed indicated that they intended to join or would like to join if they could come up with the initial dues payment.

Table 2.2 Attitudes of Non-CAEX Members Regarding CAEX Membership Male heads Female heads Sons over 18 ATTITUDES REGARDING
of household of household years of age MEMBERSHIP IN TBlE CAEX
(N=5) (N=8) (N=5)

Intend to join 40% 12.5% 60%

(2) (1) (3)

Not interested/CAEX has no 40% 50% 40%

benefits (2) (4) (2)

Put membership in son's name 20% 12.5%

Would like to join but cannot pay 25%

the dues 1(2) _________


Several households in my survey who are not CAEX members said that they would like to join but that the CAEX did not meet their needs with its present level of service. They wanted the CAEX to fulfill its promise to construct a supply post nearer to their home before they joined. For the time being, they had to depend on an increasingly unreliable local trader. Several households told me that it was almost impossible for them to buy or sell goods in the seringal. Local traders were so decapitalized that they could not get credit with which to purchase dry goods for resale to reserve residents.

These accounts of economic hardships were reinforced by a former rubber boss during a meeting that we had in Xapuri. He told me,

When I came here from Ceardt in 1954, everything was different. Before, in the days when the bosses were in charge, you could go out to the seringal for days at a time and you didn't need to take anything with you. Not a hammock or extra food or anything. They offered you lots of food and were offended if you didn't accept.
When they got sick, it was the responsibility of the boss to take care of them. Now they're out there dying of hunger. There aren't any traders that go out there because they can't make any money. All the rubber tappers want to leave the seringal because they can't make it and because they don't have a boss to take care of them and manage
the seringais.

One of the greatest challenges for the CAEX, the union and the rubber tappers involves the ideological shift from the traditional, dyadic patron-client relationships with the boss or trader to a cooperative, empowered relationship among other rubber tapper households. Adelberto, one of the CAEX leaders told me of the difficulties he experienced in his own seringal when attempting to stimulate community work days for an agroforestry project. His neighbors were reluctant to invest their time in the project because they were unaccustomed to such initiatives. Even though they have secure land tenure, many of his


neighbors still think in terms of being controlled by the boss, by whom their productive and marketing options were limited. He said, "That's how we have to think. We have to feel like we're the boss now"

This is a difficult process for both men and women in the extractive reserves. Elsewhere I have explored the challenges that face women who seek to expand their role in the household, the CAEX and the union (Campbell, forthcoming). Even for men who are accustomed to doing all of the buying and selling for their households, either via middlemen or the CAEX, the concept of being a voting member in the cooperative and having some say in its operations is quite different from their earlier marketing experiences. Joao is a young seringueiro who manages one of the Brazil nut project sites which are described in the next chapter. He told me that it was very difficult for him to get up the nerve to speak up at the CAEX when the employees were not doing their job to his satisfaction. "We have to remember that we, the CAEX members, hired these guys and we are their bosses now. If they're just sitting around or if I need to get going and I need them to get my paycheck ready, I'm not afraid to tell them, 'Look, I'm a member here and you work for us'. I get frustrated with seringueiros who complain about the service at the CAEX but they won't speak up and say anything".

There may be generational differences in the rubber tappers' conceptualization and operationalization of these shifts from the patron-client days to today's autonomous, collective production and marketing systems. Adelberto told mne of a conversation that he had with his father.


I've been working in the moi)ilvenlo [the rubber tappers' social movement] for 14 years. The other day my dad came shaking his finger at me and saying, 'See, if you'd been tapping rubber all this time, you'd be better off' And I felt like telling him,
'Look, you've been tapping, rubber your whole life and look where it's gotten you.'

These are some of the most pressing challenges facing the rubber tappers movement, the union and the CAEX David Bray has noted this process in other producer groups as they move from a shared history of struggle in a socio-political movement to one of cooperative production and marketing, a process that he has termed "from protest to production" (Bray 1991, p. 126). The rubber tappers' social movement has gone through a similar transition from the days of the empales which focused on civil protest in order to secure their rights to their traditional forest tracts. As the movement has become institutionalized through the union, school and health projects, the CNS, the CAEX and various extractive reserve projects in research and extension, the seringueiros and their families are creating communities out of their shared history of struggle combined with their ability to formalize this cooperation through these organizations. My earlier research found that this shared history of struggle was an essential element in this process of forming informal and formal ties between neighbors to sustain community activities. In that study, the experiences of two extractive reserves in relation to a popular education project differed greatly due to the shared history of sacrifice and bloodshed in the empales of one area versus the lack of prior cooperation between neighbors in an area which received numerous extractive reserve benefits without much active solicitation (Campbell 1990).

Many people discussed the current econornic situation of families who live in the more remote parts of the reserves, where there are few, if any, services provided by either the state


or the movement via CAEX or the union. One woman noted, "During our struggle, the rubber tapper became free from the boss. But for many of them, he also is now free of the union and the CAEX Providing sufficient and effective collective marketing services for all reserve residents is an immense task. While households within the extractive reserve are creating communities and transforming social and economic relationships, there is another process of significant proportion taking place at the interface of the reserve and marketing channels, government agencies and donor agencies. This is the process of local institution building (Schmink 1992), which is essential to the viability of the reserves, The ability of the CAEY, the union and the CNS to collaborate with the emerging community structures in the reserves to implement alternative production strategies is a key facet of the extractive reserve model. The next two chapters examine the impact of one such production alternative in two extractive reserve communities.


One of the economic objectives of the extractive reserve model is to add value to forest products through new technologies and cooperative marketing which would eliminate the traditional exploitative relationships such as aviamento, the chain of debt-peonage of the rubber boom era. In this chapter, I look specifically at one ongoing project intervention in two extractive reserves in light of the broader goals of the reserve model and the needs of reserve populations. The project under examination is the decentralized, post-harvest processing of Brazil nuts in the reserves which is linked to the CAEX factory in Xapuri. This initiative was designed to address the question of economic viability of the reserves through four strategies: (1) decentralization of the post-harvest processing of Brazil nuts; (2) increased efficiency in processing by the CAEX through technological and labor changes; (3) promotion of rational use of Brazil nuts in the forest and in agroforestry systems, and, (4) evaluation of current economic policies relating to Brazil nut production and marketing and the definition of alternative policies to encourage the production of Brazil nuts and other potentially sustainable extractive products (ECOTEC and CNS 1990).

Here and in Chapter 4, 1 address primarily the first objective which sought in part to increase economic returns to extractive reserve residents through decentralized processing



of Brazil nuts.' My research questions focus on the relationship between the economic autonomy of rubber tapper households and the sale of wage labor due to the project. While the classic Marxist fine of reasoning assumes that the creation of wage labor work eventually leads to proletarianization, I argue that this project instead offers opportunities for households and individuals to combine piece-work for cash income with traditional subsistence activities. Due to the forms of production within the project, households and individuals can increase their economic autonomy by investing in community or household-level processing units. In this way, they are able to increase their cash income through the sale of their labor, without losing control of the production process.

I also explore the impact of the project on the cultural identity and social organization of rubber tapper households within the extractive reserve. As production alternatives are introduced in the extractive reserves in order to diversify and improve income levels, there is a strong potential for alteration of traditional work relations and the human-nature relationships upon which the rubber tappers' movement was founded. An increasing reliance upon wage labor (from the project and other sources) as one component of extractive reserve household economies could intensify an ongoing process of abandonment of rubber tapping. Changes in the cultural and market values of non-timber forest products affect household and individual decisions regarding the allocation of labor in certain activities as opposed to another. This reflects the "ecological rationality" (Toledo 1990) of production systems in the extractive reserves. I also examine the relative income earned by reserve residents in these

Other project objectives have been addressed by other researchers and
evaluators, See Hecht et al. (1994); Rebohle (1991); Richardson (1995); and Anthony Anderson (1996).


different decentralized processing arrangements to determine the extent to which the project affects socioeconomic differentiation within the extractive reserve.

In order to contextualize the Brazil nut project within the broader ecological and socioeconomic arena of the extractive reserves in southeastern Acre, I first present an overview of the botanical, marketing and technological aspects of Brazil nut processing. This background will help to assess the potential for rational use of the resource, and for increased efficiency in processing to improve potential income.

Management and Processing of Brazil Nut (castanha) Bertholletia excelsa

Although the Brazil nut is an important source of income for the Amazon basin, and almost a of the current Brazil nut crop is harvested from wild trees (Mori and Prance 1990), Virgilio Viana points out that little research has been done to date on the management of natural populations (Viana et al. n.d.). Since deforested areas in Acre cover less than 10% of the state, and conventional agricultural and forestry research do not meet the needs of extractive producers for improved natural forest management practices, he and his colleagues call for increased funding for applied ecological research on Brazil nut and other species (Viana et al. n.d.). Such research could provide the means for increased productivity and practice of natural forest management, including practices related to the native Brazil nut populations in the two extractive reserves in this study.

Stephen Nugent (199 1) assiduously warns that such techniques and practices could be detrimental to the livelihood of small producers unless they are part of a larger "agrostrategy" which encompasses land tenure, market stability and the local variation of


production and marketing strategies. However, the economic importance of Brazil nuts to the producers in this study is increasing due to the post-harvest processing initiatives of CAEX. Applied research into management practices, reproduction biology and productivity of this species can be effective within the context of the extractive reserves, in which producers have guaranteed land tenure and have effectively organized for political lobbying and for cooperative marketing.

Distribution. Regeneration and Productivity of Brazil Nut Trees

The Brazil nut tree is a member of the Lecythidaceae family. Its natural territory is primary, lowland moist forest and it does best on well-drained oxisols and ultisols. The average rainfall for areas where Brazil nut is found is 1400-2800 mm and the average annual temperature is 24-27 degrees Celsius. The Brazil nut is a canopy tree and sometimes grows to emergent size (Clement 1993) with an average height of 40-50 meters and a diameter of 2 meters when mature (Cavalcante 1991). There are reports of two trees that each reached 62 meters in height and had trunks 4.3 meters in diameter.2 Both trees were probably between 800-1200 years old (Cavalcante 199 1).

The Brazil nut tree is a native of Amazonia with its current distribution ranging over almost all of the basin in Bolivia, Peru, Colombia, Venezuela and the Guyanas (Cavalcante 1991; Clement 1993). Charles Peters noted that Brazil nut generally occurs at fairly low densities in the forest (Peters 1992). Clement reports that the density of Brazil nut trees is

2 One of these two trees was measured after it had been cut down in the Jani plantation area,


usually less than 1 tree/ha (Clement 1993), while Richardson reports densities from 0.1 to

2.5 trees/ha (Richardson 1995).

In contrast to the generally low density in most forest areas, Brazil nut trees are often found in fairly extensive groupings known as castanhais (Cavalcante 1991) or manchales in Peru (Clement 1993). There is strong evidence for human dispersal of the Brazil nut by Amerindians who planted seeds and/or seedlings to create today's "natural" castanhais (Posey 1985). Densities of up to 15-20 mature trees/ha may be reached in these areas which range from 5-10 ha or more in size. Some castanhais in the southeastern regions of the states of Pard and Acre are 50-100 ha in size (Clement 1993).

However, the large size of these castanhais in Acre does not necessarily indicate a generally high occurrence of Brazil nut trees or of productive trees throughout the state. Many households that I surveyed in the Floresta seringal in southeastern Acre have very low levels of productivity from the Brazil nut trees on their holdings and there are no Brazil nuts reported in the Alto Jurudi extractive reserve (Almeida and Menezes 1994).

In one of my research sites, which is known for its relatively higher Brazil nut production, researchers found a density of 5.86 plants/ha over 1 meter in height, with the density of adult trees (> 40 cm dbh) being 1.92 trees/ha (Viana et al. n.d.). In the Porongaba seringal within the Chico Mendes reserve, Daniel Nepstad and colleagues found an average of 350 Brazil nut trees/colocaqdo (Nepstad et al. 1992). A survey that encompassed the Chico Mendes extractive reserve found an average of 257 Brazil nut trees/holding, over half of which were productive adult trees (CNS 1992).


In 1991, when I asked 28 families in the Floresta and Cachoeira seringais if they knew how many Brazil nut trees were in their forest holding, only six (21%) gave a positive response. Their answers ranged from 150 to 1000 trees with an average of 325 Brazil nut trees/holding. Their lack of knowledge about the number of trees is likely due to the fact that all of these trees are wild and that they are managed or visited on an infrequent basis. As opposed to rubber trees which are visited up to two times per week during the lengthier tapping season, Brazil nut trees might be visited only several times during the year. Only one of the 28 households that I visited reported having ever planted Brazil nut trees. In this case, they had planted 25 trees several years earlier.

One component of the Brazil nut project sponsored by CAEX sought to carry out research and promote rational use of Brazil nut trees in the forest and in agroforestry systems (ECOTEC and CNS 1990). Although some steps of this component were carried out, including research on Brazil nut density and seedling performance, the establishment of seed beds and the production of some seedlings, the project did not fulfill the stated objectives in this area. High staff turnover and lack of accompaniment through extension efforts resulted in abandonment of the project areas by reserve residents. A few households planted seedlings that were made available to them through the project but there were no systematic results in agroforestry or natural forest management plots. Karen Kainer, a University of Florida doctoral candidate who conducted research on the autoecology of Brazil nut seedlings in Cachoeira in 1993-1994, noted that the few cases of Brazil nut plantings by residents in that area were done with some technical assistance by the project. While people in other areas have transplanted seedlings from the forest to agricultural plots, producing seedlings from


seed is more difficult and most likely would require some technical assistance to achieve a successful level of production.

Regeneration of Native Stands

Brazil nut seeds may take 12-36 months to germinate, and even then only 60% germinate successfully (Clement 1993). In the eastern Amazon, Chris Miller noted that there were no juvenile seedlings in the Jai plantation area, but that seedlings of all age classes were found in the forests of the Marabd and Carajas areas (Miller 1990). Closer to the seringais in this study, residents of the Porongaba area in the Chico Mendes reserve reported a lack of juvenile Brazil nut trees (Nepstad et al. 1992). This could be due to a variety of factors, including the practice of burning leaves and debris under the mature trees to facilitate collection, low regeneration rates, overharvesting, reduced numbers of dispersal agents due to hunting, and seed fungus (Nepstad et al. 1992). Some of these factors could be addressed through research and management.

In contrast to the findings of low numbers ofjuvenile trees in the Porongaba seringal, residents in the Floresta seringal in this study noted that there were plenty of seedlings and adult trees in the forest. However, these did not produce worthwhile quantities of nuts. Unlike Porongaba and Floresta, researchers in the Cachoeira seringal found a higher density of juvenile plants (3.78 plants/ha) than adults (2.02 plants/ha) in their sampled transects (Viana et al. n.d.). Kainer reported that reproduction in native stands did not appear to be a limiting factor in the Cachoeira area, but that research on this issue was urgently needed in this and other seringais. (Kainer, pers. comm.).


Harvest Levels

The fruit or ourigo of the Brazil nut tree weighs between 500-1500 g and measures 10-15 cm in diameter. Each fruit contains between 15-24 seeds (what we call the nut). Each nut is 4-7 cm in length and weighs 4- 10 grams (Clement 1993). In the Cachoeira seringal, Viana and colleagues found that the number of kgs of nuts in the shell ranged from 1.5 to 105 kg/tree/yr and averaged 24.0 kg/tree (Viana et al. n.d.). The nuts are roughly 50% kernel and 50% shell by weight (Richardson 1995). Estimates for the CAEX project calculated that 66% of the weight of the nut is in the shell, all of which would be removed through the decentralized processing, thus significantly reducing transport costs (ECOTEC and CNS 1990).

The Brazil nut flowers during the dry season and fruits 15 months later during the rainy season, roughly December-March. First fruiting of Brazil nut trees usually occurs at 12 years of age but this can be reduced to 6 years through grafting, with cases of first fr-uiting in 3.5 years (Cavalcante 1991). Once the fruit of the Brazil nut ripens and falls to the forest floor from the canopy, harvesters break open the fruit with a machete and remove the nuts.

As with reproduction, what is known about Brazil nut harvests shows fluctuation over time and significant differences in productivity between seringais. My data show that the productivity of Brazil nut trees varies significantly from one seringal to another, as measured mn harvesting levels. A study in the eastern Amazon noted that harvest levels also vary fr-om one year to the next and tend to exhibit fluctuations in two to three year cycles (Miller 1990), a trend that also was reported by households in my survey sample.


In southeastern Acre, different seringais have produced varying harvest levels. Research conducted in the Chico Mendes extractive reserve found that the annual production of in-shell Brazil nuts per holding averaged 1375 kgs (CNS 1992b).3 In the Porongaba seringal which lies within this reserve, the average annual production rate was 9 kg/tree or 3150 kg/holding (Nepstad et al. 1992). In the Cachoeira reserve, annual production estimates ranged from 4,500 kg/holding (Schwartzman 1989) to 5,500 kg/holding (ECOTEC and CNS 1990).4

Table 3.1 indicates that there were significant differences in the harvest levels between the Floresta and Cachoeira seringais in 1991 and 1994. This table does not directly measure production per tree. Rather, because households in both seringais reported that they harvested most of the Brazil nuts in their holdings, this table reflects the relative productivity of the trees in each research site. Floresta residents stated that they had numerous Brazil nut trees (including seedlings) in their coloca96es but that the productivity of these trees was very low. Thus, the distribution and productivity of this resource places important limits on its potential for increasing income for reserve residents.

This report noted the average annual harvest/holding to be 125 latas, which I converted to kgs using a calculation of 11 kgs/lata.

4 For Cachoeira, Schwartzman reported a mean annual harvest per family of "325 latas (or 4,500) kg)" (Schwartzman 1989, p. 155), using a generous calculation of 13.85 kg/lata. ECOTEC reported an estimated annual harvest per holding of 500 latas, which I converted to 5,500 kg/holding using Kainer's estimate of approximately 11 kg/lata.


Table 3.1 Comparative Brazil Nut Harvests (kg of in-shell nuts/holding) 1991 1994
Floresta Cachoeira Floresta Cachoeira

median 495 6490 302.5 6600

(n=19) (n=8) (n=16) (n-7)
range 0-5500 1650-17,600 0-3300 3080-9020
mean difference 5672 5612

t value 4.66 8.17
df 25 21
p .000 .000
Source: Field interviews by the author
Note: Respondents' answers in latas were converted to kgs using the conversion of 11

If households continue to rely on existing forest stands of Brazil nut, those in the Floresta seringal will be at a significant disadvantage to those in Cachoeira. According to the average harvest/holding reported by the CNS for the Chico Mendes reserve in general (of which Floresta is part), residents in this large reserve share this lack of access to productive stands of Brazil nut trees (CNS 1992). As will be shown later in this chapter, harvest levels of Brazil nut are important since they determine whether or not households are eligible for involvement in the different processing arrangements available in the project.

In the extractive reserves, each household controls access to the forest resources in its colocagdo. This prevents harvesters other than reserve residents from gaining access to the Brazil nut trees, thus denying competitive harvesting on the same forest area. The boundary between one household and another within the reserve is respected to the degree


that neighboring households share the harvest from a Brazil nut tree which lies along the boundary, with each household gathering those nuts that had fallen on their side of the boundary (Schwartzman 1989).

Thus, access to and control over a more productive stand of Brazil nut on one's colocagdo could signify a much higher potential for increased income, both from harvest levels and from value-added through post-harvest processing. Forest management practices and post-harvest handling and storage techniques could increase the quality and quantity of nuts that are sold. These practices and techniques are discussed in the following section. Additional income increases due to post-harvest processing (shelling and drying) are discussed later in this chapter along with analysis of the Brazil nut project.

Potential Increases Through Forest Management and Post-Harvest Handling

Clement argues that production could be increased by harvesting trees that currently are left unexploited because of the low price paid to collectors. He maintains that these prices are low due to weak international demand and the high number of middlemen who effectively reduce payment made to the collector. He notes that the extractive reserves and marketing agreements with Cultural Survival should address these issues (Clement 1993). However, my own data from this and previous research projects show that collectors surveyed in two extractive reserves in Acre currently gather almost all available Brazil nuts. In 1991, 85% of the 26 households I interviewed indicated that they gather all the Brazil nuts in their colocaVdo. In 1994, 79% of the 24 surveyed households gave the same response. In both of these survey years and in previous research in the same area, families that did not harvest


all the nuts from the trees to which they had access reported that they "leave some for the animals", indicating that they do not leave large quantities unharvested (Campbell 1990 unpublished data). While there are a few families in the reserve who refrain from harvesting nuts, the potential increase in harvest levels due to a price increase would likely be minimal compared to present extraction levels from the study area.

Improved storage facilities in the extractive reserves to reduce post-harvest spoilage could, however, be implemented in response to a higher market price for Brazil nuts. Higher value for this product could encourage households to harvest nuts more frequently during the season so that nuts do not remain soaked in rainwater on the forest floor for as long a period of time. Harvesters tend to gather the Brazil nut pods together in piles in the forest until they have a sufficient quantity built up. They then spend the better part of a day at these sites, breaking open the pods with a machete and removing the nuts. In the meantime, some pods become water-logged and the nuts inside are prone to rotting.

Once the Brazil nuts are harvested from the forest, they often sit in storage for up to several months before being transported or processed. Since the nuts are often wet from time spent on the forest floor prior to collection or due to poor storage facilities, there are often high losses from fungus. However, if the nuts are kept away from outside moisture, they can remain stored for up to a year without losing flavor.

Post-harvest losses could be reduced by modifying current harvesting, storage and transportation practices. Nuts that are gathered during the harvest (December-March) and


then stored until processing (May-November) have a much higher chance of rotting.' One of the rnini-factory managers in the reserve suggested that nuts should be stored in two separate sheds to avoid losses. Nuts harvested from December through January tend not to be as wet and would be stored in one shed. Nuts harvested in February and March, the peak of the rainy season, are usually waterlogged and would go into a separate shed. They would first process the wetter nuts in the February/March shed when the processing season begins. They could then move to the drier December/January nuts later in the processing season.

With these practices, it is possible that higher market value could incrementally increase the quantity and the quality of nuts that producers sell due to improved handling and storage of nuts gathered from currently productive forest stands. However, these increases would be due more to efficient harvesting and storage practices, rather than from larger harvests from existing natural stands of Brazil nuts. (Other management practices such as those described below, also could increase the productivity of native stands). Thus, it should be possible to increase efficiency and rational use of native stands of Brazil nuts without increasing current levels of environmental impact.

Along with modifications in harvesting and storage practices, natural forest management practices could be implemented to increase the productivity of natural stands of Brazil nut. Viana and colleagues call for research on such practices as enrichment plantings in secondary forests, forest gaps, agroforestry systems and liberation thinning of suppressed, non-reproductive individuals in natural populations (Viana et al. n.d.). Based on research in

The manager of the urban factory reported that they began processing in June, 1994. He hoped that they would be able to start processing as early as February in the following year, thus reducing storage time.


the Chico Mendes extractive reserve, Nepstad et al. note that the lack of juvenile Brazil nut trees is due to a combination of factors. They report that, in addition to other variables, the overharvesting of Brazil nuts, overhunting of certain game animals which act as dispersal agents, the practice of burning debris underneath adult trees to facilitate collection and the dependence of seedlings on forest gaps could each contribute to low reproduction of Brazil nut in the forest (Nepstad et al. 1992). Some of these factors could be ameliorated through modified practices and research called for by Viana et al.

Post-Harvest Processing

While most of the potential improvements in harvesting techniques, storage and natural forest management practices are still under examination, one profitable method of increasing returns from existing Brazil nut harvests is post-harvest processing. Post-harvest processing of Brazil nuts yields several products. Brazil nuts are high in oil, 75% of which is unsaturated fatty acids that make the oil very attractive for cooking. Cooking oil is taken from the first pressing of the nuts while the second extraction can be used for soap-making and lamp fuel (Mori and Prance 1990). The factory in Xapuri currently presses and sells oil locally and plans to market the oil on a wider basis.' At the factory and in the reserves, broken or rancid nuts are used as animal feed. Several women in the reserves make Brazil

6 In November, 1994, the CAEX charged US $0.60/liter of Brazil nut while the price of a liter of soy oil (the most common cooking oil purchased by rubber tapper households) was US $ 1. 10. Brazil nut oil was available only in the CAEX's central store in town and not in the five supply outposts. The exchange rate for that month averaged US $0.83 to Real $1.00.

nut soap from the rancid nuts. Most of this soap is for home use and several families reported that they no longer have to purchase commercially produced soap for laundry and bathing. Sales of Brazil nut soap are increasing in Xapuri where the CAEX store regularly has Brazil nut soap on its shelves. The CAEX has been negotiating with the state Secretary of Education in an attempt to have ground Brazil nuts added to the flour used in the stateorganized school lunch program. The CAEX also has been working with the Institute for Amazon Studies (IEA), a research and advocacy NGO, on a contract for a health food bar with Nutrimental, a large Brazilian food processor (Butler 1992).

The most lucrative market for Brazil nuts lies in whole nuts either shelled or in-shell. Richardson reports that all marketed nuts come from wild sources (Richardson 1995). Minor plantations exist in Kuala Lumpur and Ghana (Mori and Prance 1990), and one plantation has been established in the state of Amazonas but had not yet sold any of its nut production (Mori 1992), Brazil's national agricultural research agency, EMBRAPA has also established a Brazil nut plantation that has not yet been able to produce fruit (Karen Kainer, pers. comm.). This lack of economically successful plantations bodes well for continued investment in research and production improvements of native stands as there is no apparent inuninent danger of substitution of wild Brazil nut production from cultivated sources.

Domestic consumption of Brazil nuts is very low and almost all of Brazil's production is exported to the U.S. and western Europe (Mori and Prance 1990). From 1980 to 1989, close to 95% of Brazil's national production was for export, with the U.S. ranked as the

The residue from pressing Brazil nuts can also be used for soap-making and
animal feed (Mori and Prance 1990). In late 1994, an urban women's group in Xapuri was considering approaching the CAEX factory to propose a Brazil nut soap-making project.


number one importer of in-shell Brazil nuts. The U.K. ranks first in imports of shelled nuts. Of the shelled nuts exported to the U.S., 80% is in mixed nut products. Of the $2.3 billion global nut trade, Brazil nuts captured 1.5% of the market (LaFleur 1992). The FOB value of shelled Brazil nut is usually twice that of in-shell nuts (Clement 1993).

Brazil nut production averaged 38,000 MT/year of in-shell nuts from 1980-1989 and reached a maximum of 50,000 MT in 1983. Since then, production fell steadily to just over 25,000 MT in 1989 (Richardson 1995). Harvest levels in 1991 and 1992 were 35,838 MT and 25,303 NIT, respectively (13GB 1994). During these two years, Acre's production of inshell nuts exceeded that of the state of Pard, the perennial leader in Brazil nut production. Acre produced 14,630 MT in 1991 and 11, 156 MT in 1992 compared to Pard's 9,456 MT and 10,962 MT for 1991 and 1992, respectively (113GB 1994). Exports of shelled nuts from Acre totalled 187 MT in 1989 and 179 MT in 1990 (Bstado do Acre 1991). According to the factory manager in Xapuri, CAEX exported 50 MT of shelled Brazil nuts in 1994.

Due to the heavy emphasis on exports, processing centers have been placed in major port cities, far from collection sites, resulting in high losses of 15-30% during transport from the forest to the processing centers (LaFleur 1992). Production has declined in the Marabd area in the eastern Amazon due to deforestation, rural-urban migration, flooding of castanhais and perhaps problems with pollination because of pasture establishment (Mori 1992). The resulting shift in harvesting from the eastern Amazon to the western states of Acre and Amazonas adds higher transport costs to these more remote western states which could undermine the capacity of Brazil nuts to compete in international markets (Anderson, Anthony B. 1992). However, employees at the CAEX were optimistic about the potential


to increase production for their operations through decentralized processing and through processing and marketing of different by-products as discussed above.

The Decentralized Brazil Nut Pro-ject of the CAEX

CAEX was the first agro-extractive cooperative formed by extractivist producers in the Amazon without external funding. Since its inception in 1988 with 34 founding members, CAEX had grown to over 200 members by December 1994. CAEX has its administrative offices in the town of Xapuri, as well as a dry goods store and the Brazil nut processing factory, through which it buys and sells rubber and Brazil nuts and produces Brazil nut oil. CAEX also has five supply posts in different seringais throughout the municipality, at which it buys rubber, Brazil nuts, and other products and sells dry goods. With financial assistance from various organizations over the years, CAEX has broadened its scope of operations from its original function as a collective buyer and seller of rubber. Grants and/or loans from organizations including Cultural Survival, the Inter-American Foundation, the Ford Foundation and World Wildlife Fund have assisted CAEX in its operations. The Canadian and Austrian governments have also provided assistance to CAEX.

The remainder of this chapter explores one facet of the CAEX's present activities, the decentralized processing of Brazil nuts in the extractive reserves. First I present an overview of the history of post-harvest processing by CAEX and the shift from centralized operations in an urban factory to an effort to decentralize this activity with rural labor. Then I examine specific questions related to the organization of the project. These questions include the ability of the project to efficiently decentralize post-harvest processing, to muster community


and household investment in the pilot phases of the project and to provide equitable employment for reserve residents.

Urban Decentralized Processing at the Chico Mendes Factory

In 1990, CAEX purchased an empty government warehouse in Xapuri and established the Chico Mendes Processing Factory to shell and dry Brazil nuts for national and international markets. For the next year and a half, the factory was the largest employer in town, with 140 employees and CAEX generated 60% of the tax income for Xapuri's municipal coffers (CNS et al. 1993). However, Brazil's employment legislation requires that all employers pay social benefits to full-time employees, benefits that cost employers almost as much as they already pay in salaries. Unable to meet these high labor costs in the face of competition from Bolivian nut processors who were not saddled with similar costs, CAEX switched to decentralized processing with home-based shelling of Brazil nuts in workers' home in the city of Xapuri. Under this arrangement, nut shellers who had previously worked during shifts at the factory were now paid for working at home on a piece-work basis.

As of November 1994, the factory had 92 people who worked at home as shellers, most of whom were women. The CA.LX Vice President reported that there also were 30 flalltime, salaried employees at the Brazil nut factory working in the drying, sorting and packaging stages. The processing operation consisted of pre- and post-shelling steps at the factory, with the shelling taking place in the shelters' homes. At the factory, in-shell nuts were placed in a metal warming bin for 12 hours in order to draw the nut away from the shell. The in-shell nuts were then soaked in water for five hours in order to soften the shell. CAEX then


delivered 3 latas (approximately 33 kgs of in-shell nuts) to the shellers, each of whom had a manual machine at home. Once the nuts were shelled, the home-based worker brought them to the factory the following day for weighing, final drying and sorting. Nuts were dried in ovens fired by fuelwood and Brazil nut shells. The shellers received a monthly paycheck based on their production. Nuts were sorted into three classifications, based on the size and condition of the nut. As of December, 1994, shellers received US$0.30/kg for first class, shelled nuts, $0.25/kg for second class and $0.20/kg for third class.'

Decentralized Processing in the Extractive Reserves

In order to generate employment in the reserves, to increase the output of the factory, and to strengthen CAEX's economic viability by avoiding the high labor costs of full-time, urban employees, CAEX began a project of decentralized Brazil nut processing in 1991. The project has received financial backing from the Ford Foundation, the Inter-American Foundation and World Wildlife Fund. Technical assistance was provided by ECOTEC, a nongovernmental organization based in Recife.

The project envisioned the establishment of mini-factories in the extractive reserve which would each employ 5-10 nut shellers. This original plan of operations has since expanded and there are now three arrangements for decentralized processing in the reserves. First, there are three community-level mini-factories in the reserves where all processing steps are carried out. These mini-factories each have from five to twelve nut shellers, a manager

8 For more on the factory and labor relations with urban shellers, see Anthony Anderson (1996); and Hecht et al. (1994).


and an assistant manager. Second, there are several families who live near one of the minifactories and who take nuts home to be soaked and shelled and later delivered back to be dried in the oven at the mini-factory. These families are paid the same rate as the nut-shellers in the mini-factory, the only difference being that they work at home. Third, there are four household-level units where all of the processing steps are carried out at home. These families have constructed facilities similar to the mini-factories but that are scaled down for household-level production, each with between two and four nut crackers.

My household survey sample includes two of the three mini-factories and two of the four household-level units that were in operation in 1994. 1 was able to accompany these two mini-factories from pre-construction stages beginning in 1991 through several years of operation. Also included in my household study are three of the seven households participating in home-based shelling in conjunction with this mini-factory. This chapter relies on data from my household surveys and extended interviews as well as data from CAEX project records and interviews with CAEX and project staff. Project-wide information on production levels and salaries for all decentralized processing was available from the CAEX office for all those individuals who had worked on the decentralized Brazil nut project during 1993-1994.9 To complement thi s information, I applied a questionnaire regarding employee's attitudes and practices in the project. Using this questionnaire in Floresta and Cachoeira, I interviewed all of the employees in the two mini-factories and three of the four families who

Records for the 1992 season were lost in the CAEX computer when an employee inadvertently cleared the hard drive.


operated a household-level unit. The analyses of income levels and project performance are based on this project-wide CAEX data and these questionnaires.

The steps in the decentralized processing in the reserves are the same as those taken at the central factory, but the technologies are scaled down for community or household-level processing. First, the nuts are laid out onto a drying area, usually a raised wooden platform, Nuts are left to dry in the sun for one-three days. This has the same effect as in the factory where nuts are heated in order to pull the kernel away from the shell. Nuts are then soaked overnight in water to soften the shell. Breaking open the shell is the next step. This is done using a manual nut-cracker that is bolted to a table. The shell is gently cracked by hand with this machine and is then removed by prying it off with one's fingers or a small knife. Care is taken not to bruise the nut by pulling too hard on the nutcracker's weighted lever because a higher price is paid for whole, unbroken nuts.

Once the nuts are shelled, they are placed into a pile on the breakers' worktable. At the end of the day, shelled nuts are classified into piles of first, second and third quality nuts and each processor's production is weighed and recorded. Nuts are then placed on trays with wire bottoms and placed into the wood-fired oven for final drying, during which the humidity of the nuts must be reduced to 4%. Fuelwood is gathered from the nearby forest. Drying takes approximately 12 hours. Nuts are then placed into sealed plastic drums for transport into town. There the nuts will be sorted again according to industrial classifications (tiny, midget, medium, large, chipped, broken) and packaged for final shipping."

'0 For other technical and marketing aspects of the decentralized project, see
Richardson (1995); Anthony Anderson (1996); Rebohle (199 1); and Hecht et al. (1994).


At the end of the 1994 processing season, there were 37 people employed in the extractive reserves with the decentralized processing project. This includes the three pilot mini-factories, home-based shellers and the four household units. My household survey sample in the Floresta and Cachoeira areas encompassed 15 of these 37 persons, or 41% of those employed in the project. These 15 employees constituted 16% of the adult population in the two areas. By late 1994, the CAIEX had already undertaken steps to construct additional mini-factories and there were many solicitations by CAEX members who wanted to build their own household units. The CAEX vice president estimated that, once a rotating capital find was established from the monthly payments of existing units and mini-factories, CAEX could build 12 household units per year within the next two years. In 1994, decentralized processing accounted for 24% of the CAEX factory's total output of shelled nuts. The central factory manager hoped to increase this percentage to 50% in the coming years. Depending on CAEX administration of the decentralized project and the central factory, construction of additional mini-factories and household units, combined with improved processing and storage practices, should significantly increase the project's employment opportunities in the extractive reserves in the near future.

The three pilot mini-factories experienced some technological and personnel difficulties as the project got underway in the extractive reserves. While the positions of manager, assistant manager and most of the shelling retained a stable work force, there was a fairly high turnover rate as some reserve residents worked for a month or two as a sheller and later left the project. From 1993 to 1994, 26% of those who were employed in 1993


decided not to work at the project for the 1994 season." This turnover rate fell to 10% in 1994 as four people decided to leave the project out of the total of 41 who were employed over the course of the season. All but one of the 17 people employed by the project in 1994 whom I interviewed planned to continue working in the 1995 season. Managers and project staff told me that there was constant demand for employment at the mini-factories.

The sex composition of the work force in the project remained fairly constant over the 1993 and 1994 seasons. In 1993, the work force consisted of 49% men and 5 1 % women. The following year, men's participation accounted for 46% of the total work force while women accounted for 54% of the employees. As shown in Table 3.2, men and women shared differentjob categories in the project. Only one woman was employed in management, and only one man worked as a home-based seller. This chapter will later show that there is no statistically significant difference in the income levels between these job categories. Thus, the job categories and the income within the project are fairly evenly distributed among the male and female employees,

In my household sample there were 15 people employed by the project in 1994 (i.e. those whose names were listed on the payroll). Their ages ranged from 15 to 55 with a median of 27 years of age. The 15 year old was a young man who had previously stopped attending school and chose to work on the project instead of tapping rubber on his family's colocagdo. In the household-level units and in the families who worked as home-based sellers, there were cases of eleven and twelve-year old children helping out with the

Reasons for leaving the project had to do with health problems, job dissatisfaction and families moving from the area.


processing. Of the six children under 16 years of age who worked with their families on the project, three of these attended school in the morning and helped with shelling or some other task in the afternoon. The other three were 14 and 15 year old boys who had already passed third grade and chose to work on the project with their family. The project thus does not rely heavily on the labor of children, nor have families taken children out of school in order to work on the Brazil nut processing."

An evaluation team has argued that the project did result in children leaving
schooling in order to shell Brazil nuts and that young children have taken over the task of nut-shelling once the adult women tired of this work. Their conclusions were based on very brief visits to two sites and differ significantly from my observations which were taken over the course of several years. See the concluding section of this chapter.


Table 3.2 Job Category and Sex Composition of the Work Force in the Brazil Nut
Project (1994)

Mini- Manager Asst. Home- HH Unit Total

Factory Manager based Manager

Sheller Sheller3
Men # 11 2 3 1 2 19
% 50% 66% 100% 22% 50% 46%

Women# 11 1 0 8 2 22
% 50% 33% --- 88% 50% 54%

Total # 22 3 3 9 4 41
% 100% 100% 100% 100% 100% 100%
Source: CAEX project records

Mini-Factories: community investment in increased productivity

The three mini-factories in operation at the end of 1994 each employed a site manager, an assistant manager and from five to ten nut shellers. Nut shellers worked a daily shift of 4 to 8 hours in the mini-factories. They were paid on a piece-work basis, similar to the rates paid to urban nut shellers. The decentralized rates were higher than those paid to urban shellers because of the lower operating costs of the decentralized mini-factories. The rural nut shellers were paid US$.399/kg of first class shelled nuts, $.299 for second class and $.20

"3 These data reflect the payroll sheets of CAEX. In many cases, the home-based shelling relies on the labor of several household members for transportation, sun-drying, soaking and shelling the nuts. This is often an activity in which most older children and adults in the household are involved. However, only the name of one person from the household is listed on the payroll.


for third class. Urban nut shellers received US$0.30/kg for first class shelled nuts, $0.25/kg for second class and $0.20/kg for third class.

The mini-factory managers and assistants were responsible for sun-drying, soaking and oven-drying the nuts as well as keeping the shellers' tables supplied and weighing and recording al of the production. Their pay was based on the output of the mini-factory. The manager was paid US$0.044/kg of all nuts produced by the unit. The assistant manager received US $0.03 7/kg.

The site managers and assistants were responsible for processing steps which had to be maintained around the clock. There was a continual need to keep the nuts moving from the sun-drying tables to the soaking bin and then to the sheller's tables. From there, nuts had to be placed on racks in the oven for drying and then into barrels for transport to the CAEX for final packaging. The wood-burning ovens needed almost constant attention. One of the managers stated that gathering firewood was the most time-consuming task at their minifactory. At another, soaking the nuts required the manager and the assistant to carry 220 liters of water every day from the nearest stream which was approximately 300 yards away. They expected to install a water pump provided by CAEX for the following season. The other mini-factory had a water wheel to bring water up to the soaking tank from a nearby pond.

The three mini-factories were constructed with communal labor and locally available wood. Other materials and equipment such as the nut crackers and some of the stove components were provided by the CAEX Each mini-factory is obliged to reimburse the CAFX for the cost of these materials. This debt payment is made monthly by putting all of


the third quality nuts processed at the mini-factory towards the outstanding amount. Upon full payment, the mini-factory will belong to the community.

Extractive reserve residents have made communal investments to improve the infrastructure and productivity of the reserve through the mini-factories. Community members invested their time and labor in constructing the mini-factory, even though they had no written agreement that their families would have guaranteed access (i.e. employment) to the mini-factory. One way in which families could maintain some degree of control in the mini-factories was through continued membership in the CAEX and active participation in community meetings where some management issues of the factory were decided. Since the project was not donated to the community by the CAEX, it avoided many of the pitfalls of a welfare or assistencialisno approach which can make communities dependent beneficiaries as opposed to empowering participants who make their own decisions and investments regarding the project.

Home-based shelling with a mini-factorv

The second arrangement for decentralized processing involves home-based shellers who work in conjunction with one of the mini-factories. Households involved in this arrangement lived from five minutes to an hour from the mini-factory. The choice to work at home was their own and was often made based on the more flexible work schedule with home-based shelling. For reasons described in this section, families realized that there were advantages and disadvantages to this arrangement. Those who lived farther away or who didn't have pack animals probably would have had higher returns to their labor if they had worked in the mini-factory. Others used the home-based shelling experience in their decision


to construct their own household unit. Those who lived closest to the mini-factory still preferred the home-based shelling.

Home-based nut sellers have somewhat higher incomes than those working in the mini-factory. This could be due to longer working hours or the assistance of other family members throughout the day so that the shelling machine rarely was idle. Although homebased shelling can yield higher returns per machine than in the mini-factory, it also requires more household labor in terms of transportation, drying and soaking tasks. Someone from these households has to go the mini-factory to gather Brazil nuts from the central shed, take them home, lay them in the sun for at least one day and then soak them overnight. The soaked nuts then go to the breaking table on the porch of their house or in the front room for shelling. That same day, the shelled nuts have to be delivered back to the center for ovendrying.

The households that shell the nuts at home must make the same contributions to paying off the mini-factory construction costs as those who work at the factory itself, even though they also had to construct drying and shelling tables at home as well as make arrangements for soaking the nuts. A man involved in home-based shelling complained, "We basically have to do all the work except for the [oven] drying. But the manager and his assistant at the mini-factory still get the same cut from us even though we do almost all of their job". Another declared that he was frustrated at having to subsidize the construction payments of the mini-factory because he would not be able to realize a long-term benefit. Since he planned to move -Mthin the coming year, he felt that he had been investing fruitlessly


in the mini-factory. He said, pointing to the roof, "It's not like I can take that plank or that bench or that machine with me when I move, but I know that part of this place is mine".

Another drawback to this arrangement is transportation. One afternoon I met with a young man who had come to the mini-factory to gather nuts from the central shed to take to his coloca do, an houes walk away. As he had no pack animals, he had to rent the horses of a man who lived near the mini-factory. However, all three of the horses were out in the pasture, meaning that he would have to spend time to round up the horses and load the nuts, on top of having to pay the standard daily rate of US$10 for the use of three animals.

Most of those involved in home-based shelling reported that this arrangement works best for households who have their own transportation and/or who live near enough to the mini-factory to be able to take advantage of the centralization by picking up already soaked nuts, thus avoiding the tasks of sun-drying and soaking the nuts at home. The time spent transporting the nuts and carrying out the other processing steps (aside from the shelling) is substantial for some households. Several households were trying out the home-based shelling for a season in order to decide whether they wanted to invest in an independent, householdlevel processing unit, which is the third decentralized arrangement. Household-level units

This decentralized arrangement consists of household- level processing units where a steps of the processing are carried out at home. There were four such units in operation in December 1994 with several others soon to be constructed. As with the mini-factories, the CAEX offers to provide equipment and specialized labor for construction of the stove while


the family builds the drying tables, shelling facilities and covered area for the oven. The costs of these materials and hired labor must all be paid back to the CAEX by the household. In the same fashion as the mini-factories, montl-Ay payments are deducted directly from the unit's production, in the value of 50 kg of first quality nuts (approximately US $32.00) to cover construction and material costs,"

Based on labor availability, their cash on hand and the level of debt that they wish to incur, households may choose to have between two and four machines. The CAEX project manager estimated that average start-up costs for a two-sheller household unit were US $700.00, but that comers could be cut to bring the cost down to approximately $500." Depending on these costs and the market value per kg of Brazil nuts, he calculated that it would take approximately three years (based on a six-month processing season per year) for each household to pay its debt. The coloca do would then have a significant value added to its infrastructure. This would increase the sale or exchange value if and when the household decided to sell or transfer its coloca do. The units allow the households to add value to their primary products and increases labor opportunities for household members who otherwise might not have access to cash income.

The four households had no problems in making these monthly payments, which are deducted from their paychecks. In one instance, the CAEX honored a request by one of the households to refrain from taking the monthly deduction because of a tight cash flow. This flexibility and sensitivity to the realities of the economies of participating households made them more secure in taking on such a large debt with the CAEX.

Costs for some equipment such as a wheelbarrow, buckets for hauling water, soaking tanks, etc. could be avoided if the family already had similar items.


The CAEX paid US$0.64/kg of nuts processed by the household units in 1994. This figure applied to Brazil nuts that the CAEX had already purchased from various gatherers during the harvest and had stored in the central shed near the mini-factory. When households that operate their own unit could harvest and process their own Brazil nuts, the CAEX paid US$1.32/kg.

Of the four household-level units in operation in December 1994, two of them lay outside the extractive reserve boundaries. Although these did not have their land tenure guaranteed within the reserve, they had made a significant investment of their labor and future earnings in their colocagdo. These families were both very active in the Xapuri union and the CAEX and had participated in the empates responsible in part for the establishment of the reserves. Although they did not reside in the reserve, the families had a strong enough sense of security in their land rights as posseiros (squatters), and in the movement's ability to defend that tenure, that they were not inhibited from taking on the risk of investing in a Brazil nut processing unit.

Of the four families who worked as home-based shellers in 1993, three of them went on to construct their own household-level unit in 1994. Another couple that had worked at one of the mni-factories since the project began in 1991 also planned to build their own unit for 1995. One of the households that was working with home-based shelling in 1994 planned to do the same and construct their own unit for the 1995 season. Working with the project through one of the mini-factories gave these households a chance to evaluate their work preferences and the profitability of setting up their own household unit. Also, due to their many years of residence in the same co/ocagdo, they were very well aware of the harvest